Management of Food Allergens
Management of Food Allergens Edited by
Jacqueline Coutts and Richard Fielder Gen-Probe Life Sciences PLC, BIOKITS Products, Deeside, Flintshire, UK
A John Wiley & Sons, Ltd., Publication
This edition first published 2009 C 2009 Blackwell Publishing Ltd Blackwell Publishing was acquired by John Wiley & Sons in February 2007. Blackwell’s publishing programme has been merged with Wiley’s global Scientific, Technical, and Medical business to form Wiley-Blackwell. Registered office John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, United Kingdom Editorial offices 9600 Garsington Road, Oxford, OX4 2DQ, United Kingdom 2121 State Avenue, Ames, Iowa 50014-8300, USA For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell. The right of the author to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. Library of Congress Cataloging-in-Publication Data Management of food allergens / edited by Jacqueline Coutts and Richard Fielder. p. cm. Includes bibliographical references and index. ISBN 978-1-4051-6758-1 (hardback : alk. paper) 1. Food allergy. 2. Food industry and trade. Jacqueline. II. Fielder, Richard. RC596.M36 2009 616.97’5–dc22 2009013261 A catalogue record for this book is available from the British Library. R Set in 10/12 pt Times by Aptara Inc., New Delhi, India Printed in Singapore
1
2009
I. Coutts,
Contents
Contributors Preface
xi xiii
PART I RISK ASSESSMENT 1
2
The reality of food allergy: the patients’ perspective David Reading 1.1 Background 1.2 Consumer reaction 1.3 Supporting consumers 1.4 Allergy services 1.5 Teenagers and young adults 1.6 Food labelling 1.7 Allergen thresholds 1.8 Food alerts 1.9 Our work with industry 1.10 The work of the FSA 1.11 Schools 1.12 Eating out 1.13 Daily life with a food allergy 1.14 Hopes for the future References
3 4 6 7 10 10 14 15 16 17 18 20 21 22 24
Clinical incidence of food allergy Zsolt Sz´epfalusi and Thomas Eiwegger
26
2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 3
3
Introduction Case 1 – Severe anaphylaxis to an unknown food product Case 2 – Idiopathic anaphylaxis Case 3 – Cross-reactivity or contamination? Case 4 – To vaccinate or not in egg allergy Case 5 – Adrenalin auto-injector for all egg-allergic patients? Case 6 – Immunotherapy for oral allergy syndrome? Conclusion References
Identification and characterisation of food allergens E.N. Clare Mills, Philip Johnson, Yuri Alexeev, and Heimo Breiteneder 3.1
Introduction
26 29 30 31 33 34 36 37 37 42 42
vi
Contents
3.2 3.3 3.4 3.5
4
Classification of food allergens Plant food allergens Animal food allergens Conclusions References
52 53 57 59 60
Coeliac disease: allergy or intolerance? Norma McGough
70
4.1 Introduction 4.2 About Coeliac disease 4.3 Prevalence and diagnosis 4.4 What is gluten? 4.5 The gluten-free diet 4.6 Gluten-free foods 4.7 Prescriptions 4.8 Allergen labelling 4.9 Food production 4.10 The Codex standard 4.11 Gluten testing 4.12 Gluten-free catering 4.13 Cross-contamination 4.14 Nutritional adequacy 4.15 Lactose intolerance 4.16 Coeliac UK References
70 70 70 71 71 72 72 72 74 74 75 75 76 76 77 77 77
PART II RISK MANAGEMENT 5
Risk management – the principles Ren´e Crevel 5.1 5.2 5.3 5.4 5.5 5.6
6
Introduction Allergen management: the issues Development of allergen management plans: principles and considerations Objectives Application Concluding remarks References
Risk management – operational implications Anton J. Alldrick 6.1 6.2 6.3 6.4
Introduction Identifying the hazard Managing the hazard Conclusion References
83 83 84 85 87 93 98 99 102 102 103 104 113 113
Contents
7 Choices for cleaning and cross-contact Steve Bagshaw 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 7.10 7.11 7.12 7.13 7.14
Allergen management and cleaning The cleaning process Principles of cleaning Open plant cleaning Dry cleaning Manual cleaning Foam and gel cleaning Cross-contamination Floor cleaning Tray and rack washing machines Cleaning-in-place Management of allergen cross-contamination Cleaning management The cleaning programme References
8 Validation of cleaning and cross-contact Helen M. Brown 8.1 8.2 8.3 8.4 8.5
Introduction Validation of a cleaning regime Sampling to validate cleaning What to measure to validate cleaning Summary References
9 Validation, standardisation and harmonisation of allergen activities in Europe and worldwide Bert Popping 9.1 9.2 9.3 9.4
10
Analytical methods Method validation Standardisation of methods Harmonisation References
Standardisation of analytical methodology with special reference to gluten analysis Ingrid Malmheden Yman 10.1 10.2 10.3 10.4
Introduction Methods and standards Gluten analysis Gluten determination References
vii
114 114 115 117 119 119 121 122 123 125 125 127 131 131 132 137 138 138 139 141 145 148 148
150 150 150 151 152 153
154 154 154 157 160 164
viii
11
12
Contents
Analytical choices Marie-Claude Robert
166
11.1 11.2 11.3 11.4 11.5
166 166 168 172 182 182
Development of allergen testing Test formats Commercial test kits Analytical issues specific to immunoassays Conclusions References
Food allergen method development programme at Health Canada: support to standard setting and consumer protection Samuel Benrejeb Godefroy, Michael Abbott, Terry Koerner, Dorcas Weber, and Theresa Paolisini 12.1 12.2 12.3
Rationale to ACT on preventing food allergy incidents in Canada Health Canada’s food allergen methodology programme Conclusion References
184
184 185 191 193
PART III RISK COMMUNICATION 13
Finished product labelling and legislation Sue Hattersley 13.1 13.2 13.3 13.4 13.5 13.6
14
Introduction Legislation on allergen labelling – European Directive 2003/89/EC and subsequent amendments Allergen cross-contamination and advisory labelling (Such as ‘May Contain’ statements) Provision of allergy information for foods that are not pre-packed ‘Free from’ foods Conclusions References
197 197 197 203 208 210 210 211
Guidelines for manufacturing and certification programmes Neil Griffiths
212
14.1 14.2 14.3 14.4 14.5 14.6 14.7 14.8 14.9 14.10 14.11
212 212 213 214 217 218 220 221 222 224 225
Preface Introduction The law Voluntary information Guidelines Certification schemes Training The use of risk assessment Management The environment Labelling and communication
Contents
14.12 14.13 14.14
15
Thresholds Testing Conclusions References
Risk communication – a manufacturer’s perspective Clive Beecham 15.1 15.2 15.3 15.4 15.5 15.6 15.7 15.8
Background The process of going ‘nut free’ The factory changes People disciplines Verification Retailer reaction What is nut free? – the problem of evolving science The need for thresholds
Appendix: Useful web links Index The colour plate section follows page 146
ix
226 228 229 231 233 233 234 237 238 241 243 244 245 248 253
Contributors
Michael Abbott Bureau of Chemical Safety, Food Directorate, Health Products and Food Branch, Health Canada, Ottawa, Ontario, Canada
Samuel Benrejeb Godefroy Bureau of Chemical Safety, Food Directorate, Health Products and Food Branch, Health Canada, Ottawa, Ontario, Canada
Yuri Alexeev Institute of Food Research, Norwich, UK
Neil Griffiths Neil Griffiths Consultants, Portslade, Brighton, UK
Anton J. Alldrick Campden BRI, Chipping Campden, Gloucestershire, UK Steve Bagshaw Holchem Laboratories Ltd, Haslingden, Rossendale, Lancashire, UK Clive Beecham Kinnerton (Confectionery) Company Ltd, London, UK Heimo Breiteneder Department of Pathophysiology, Medical University of Vienna, Alsergrund, Vienna, Austria Helen M. Brown Campden BRI, Chipping Campden, Gloucestershire, UK Ren´e Crevel Safety & Environmental Assurance Centre, Unilever, Colworth Science Park, Sharnbrook, Bedford, UK Thomas Eiwegger Department of Pediatrics, Medical University of Vienna, Alsergrund, Vienna, Austria
Sue Hattersley Food Standards Agency, Aviation House, Kingsway, London, UK Philip Johnson Institute of Food Research, Norwich, UK Terry Koerner Bureau of Chemical Safety, Food Directorate, Health Products and Food Branch, Health Canada, Ottawa, Ontario, Canada Norma McGough Coeliac UK, High Wycombe, Buckinghamshire, UK E. N. Clare Mills Institute of Food Research, Norwich, UK Theresa Paolisini Bureau of Food Regulatory and International Affairs, Food Directorate, Health Products and Food Branch, Health Canada, Ottawa, Ontario, Canada Bert Popping Eurofins, Pocklington, Yorkshire, UK David Reading The Anaphylaxis Campaign, Farnborough, Hampshire, UK
xii
Contributors
Marie-Claude Robert Nestl´e Corporate Allergen Management, Nestl´e Research Centre, Lausanne, Switzerland Zsolt Sz´epfalusi Department of Pediatrics, Medical University of Vienna, Alsergrund, Vienna, Austria
Dorcas Weber Bureau of Chemical Safety, Food Directorate, Health Products and Food Branch, Health Canada, Ottawa, Ontario, Canada Ingrid Malmheden Yman Livsmedelsverket (National Food Administration), Uppsala, Sweden
Preface
This book is about one of the greatest challenges facing the food industry: providing food to an ever-increasing number of allergic consumers through a global supply chain. The subject is complex and in its infancy. Keeping up to date with the rapid advances from the fragmented knowledge base is a challenge for the reader, especially when it is drawn from such diverse sources. Any investigation into food allergens and their management inevitably produces many more questions than answers, and it is this challenge that the book seeks to address with author contributions from recognised specialists active in the field. A risk analysis perspective has been taken in order to discuss information relevant to evolving best-practice approaches for managing food allergens in the process environment. The topic is treated as a food safety issue rather than a product labelling one. It aims towards helping all those with a vested interest in understanding how to protect consumer health through good manufacturing practice and providing clear labelling advice. Discussion excludes the catering sector, where foods consumed are perceived to pose a greater risk to the allergic consumer than packaged foods. The need for this book has been driven by the wider public awareness of food allergies and intolerances and the growth in the number of people who follow a restricted diet, at least for a period in their lives. Regardless of whether those individuals have been clinically diagnosed, the number of family members, friends and colleagues affected is appreciable. In the USA, more than ten million people are affected, with a similar number in Europe. The choice of foods available is important to sustain a restricted diet and a healthy lifestyle. Not surprisingly, the market for ‘free from’ foods has grown dramatically in recent years. The market for such gluten-free and dairy-free foods shows no signs of abating in the foreseeable future. The trend for healthier foods has led to the inclusion of diverse natural ingredients in the diet (such as nuts, seeds and grains), many of which are potentially allergenic. Whilst such ingredients can help to prevent and counteract ill health in some, to others their presence can prove fatal. The global economy means that such ingredients are sourced worldwide and can simply be selected from countries where costs are the lowest. Ingredient specifications, including any possible contaminants, must be declared by suppliers. Avoiding contamination in food production can be as much about control during supply as during manufacture and distribution. Without the means to control every stage, the risks for the food producer and consumer alike can never be reduced to zero – in spite of the ‘free from’ claim. ‘Hidden’ allergens (i.e. where contaminants are not declared, for whatever reason) continue to be the largest single cause of global product recalls. Most recalls could be averted by stricter supplier and factory controls (both prerequisite and HACCP). This is indicative of the early developmental stage of best-practice approaches to validate and verify the effectiveness of manufacturers’ risk management systems. The role of testing is deemed to be an integral element in any approach, and much work is needed to develop effective programmes of routine testing, particularly at a time when
xiv
Preface
there are intense external pressures to do so. Such pressures come from a number of directions: r Increased labelling legislation: e.g. European Directive 2007/68/EC added lupin and molluscs to the list of allergenic foods requiring labelling as an ingredient, whilst some non-allergenic ingredients gained permanent exemption from labelling. In Japan, lobster and crab are proposed to be added by 2010. r Guidance documents: on allergen management have been provided in many countries, e.g. in the UK and in Australia and New Zealand. r Codes of practice: to make food retailers’ advice more explicit for their suppliers. r Manufacturing standards: consumer groups continue to liaise with the industry to develop criteria which should be met when producing ‘allergen-free’ foods. r Thresholds: in reality, how much is too much for the allergic consumer and how does this relate to statutory levels for enforcement and management levels for process control? Different specialists’ perspectives are offered in this book to address the huge challenge that undeclared food allergens pose to the industry. An extremely important part of allergen control practices is the documentation of every aspect of the process, activities and operations of food manufacture. Documentation serves the need for traceability in the event of future problems and product recalls from the market. However, it cannot be relied upon alone without audits, inspection and testing to demonstrate that the control systems are working. The need for validation of previously agreed specifications is another key message of the book, and this has to include manufacturing, cleaning and analytical testing. Until such times that new techniques (e.g. clinical interventions, genetic modification) bring an end to food allergy and intolerance, allergic consumers are reliant on manufacturers’ systems of allergen control. However, minimising sufferers’ exposure will undoubtedly help to enhance their quality of life. Richard Fielder Jacqueline Coutts
Part I
Risk Assessment
1
The reality of food allergy: the patients’ perspective
David Reading
1.1 BACKGROUND Most of us take our food for granted. We know where the next meal is coming from and we can be fairly confident that it is going to be safe to eat. But people who deal every day with a life-threatening food allergy do not have that confidence. They can find mealtimes a source of enormous stress. For what is nutritious to the majority can be an extreme danger to a minority. Anyone who has been rushed to casualty suffering a life-threatening allergic reaction to a food, or has seen that happen to his or her child, knows how life changing that experience is. One mother wrote to us saying: ‘My seven-year-old daughter is allergic to nuts, milk, eggs and fish. Our lives are pretty much ruled around food. I worry terribly about the future.’ That is typical of the messages we receive from families dealing with this problem every day of their lives. The fact that food allergy can be deadly serious has been well known for decades to the medical profession, but it was not until the early 1990s that the public at large, the media and the food industry began to understand that it can kill a small but significant number of people. And for many, food allergy can be lifelong. Before the 1990s, allergy was generally regarded as a minor inconvenience, if it existed at all. It was out there on the fringes of medicine, affecting (so it was thought) people who let their imaginations run riot. At the most, it was generally considered that food allergy amounted to no more than a bit of indigestion or an itchy rash. Then in late 1993, the UK media reported that four people had died suddenly in a short space of time from allergic reactions to nuts. They included my 17-year-old daughter, Sarah, who died after eating a dessert-containing peanut in a town centre restaurant. She had an inkling that she was allergic to peanuts but had no idea how serious it could be. The condition from which she died is called anaphylaxis. Her throat swelled and closed up, she suffered a severe asthma attack and her blood pressure plummeted. The resulting media publicity spread alarm throughout the food industry and placed immense pressure on the government to take action. For if people’s lives were at risk from a common, everyday food that was nutritious for the majority, what were the implications for food production and labelling? The author is shown with his daughter, Sarah, who died from her allergy, in Figure 1.1. As a result of those four tragic deaths, and others that came to light, the Anaphylaxis Campaign was formed in January 1994 with me as its co-founder and chairman. A small core group of a dozen of us (people with food allergy and the parents of children with food allergy) met in a flat near Baker Street, London, in January of that year. We had never met
4
Management of Food Allergens
Fig. 1.1 David Reading with his 17-year-old daughter, Sarah, who died in 1993 from an allergic reaction to peanuts.
before, but made contact as a result of the huge wave of media coverage that occurred. That informal London get together was the inaugural meeting of the Anaphylaxis Campaign, which began its life as a small pressure group run from people’s homes. Our clear objective was to save lives, and in those early months we set out to be a focal point for the spread of information. This proved to be more straightforward than we had dared to hope. After the Campaign’s launch, and the resulting media publicity, I was receiving 60–70 letters a day through my letterbox, many of them telling heart-rending stories of children who had been rushed to casualty suffering from extreme allergic reactions to food. Along with the information provided to us by medical professionals, this influx of case histories from people with allergies provided a comprehensive file from which to develop guidance for patients and the industry alike. We met the Food Minister, Nicholas Soames, in February of 1994, a few of the major retailers and manufacturers in March and the Department of Health in April. We told Soames that more than 1000 people had contacted the Campaign in less than 2 months; 700 had signed up, and more were joining daily; 85% of those were allergic to peanuts; 80% of the total were the parents of allergic children younger than 15 years. Soames said that he was ‘staggered’ by the scale of the food allergy problem. He said that the food labelling issue had to be fought in Europe, but agreed to launch an awareness drive throughout the UK food industry. The Ministry of Agriculture, Fisheries and Food (MAFF) began that year to commission allergy-related research, and this has grown into a formidable programme in the hands of the Food Standards Agency (FSA). See Tables 1.1 and 1.2 for a breakdown of the current membership profile of the Anaphylaxis Campaign.
1.2 CONSUMER REACTION Several facts became apparent from that flood of letters to the Campaign: nut allergy was far more common than had been recognised; it was causing terrible anxiety and disruption to
The reality of food allergy: the patients’ perspective
5
Table 1.1 Anaphylaxis Campaign membership profile by age (March 2008). Number Number Number Number Number
of of of of of
members members members members members
with recorded allergies aged 0–11 aged 12–20 aged 21–30 aged 31 and older
7566 2738 2995 660 1173
people’s lives; and many had been labouring under the mistaken impression that they were among only a handful of families affected. One mother of a nut-allergic 13-year-old boy said: ‘Until we found your organisation, we were forced to cope on our own. We felt we were playing Russian roulette every mealtime. The future looked very bleak.’ We learned that teenagers and young adults seemed to be those most vulnerable to potentially fatal reactions; a range of foods seemed to be implicated (not just nuts); and people were reacting to tiny amounts of the offending allergen. On the face of it, it was a nightmare scenario for the food industry. Since that time, the Campaign has remained in the forefront of efforts to raise the level of debate. With our membership of almost 8000 people (Spring 2008), we have been able to comprehend quite clearly what people with food allergies want. Primarily, they want to be understood. Many believe their needs have been bypassed. Those who are allergic to foods other than peanuts, for example, will often complain that industry (and indeed the Campaign itself) is fixated on peanut allergy and that other foods are played down. It is important to understand that any food that contains protein could potentially cause anaphylaxis for someone, somewhere. To the outsider, the degree of anxiety expressed by families who care for a food-allergic child can appear out of all proportion. The level of apprehension becomes understandable when you realise that there is a frightening unpredictability to severe food allergy and very often inadequate medical guidance. Many of those affected have experienced sudden life-threatening episodes requiring an emergency dash to hospital. One parent of an
Table 1.2 Anaphylaxis Campaign membership profile by allergens (March 2008). Peanuts and tree nuts Egg Sesame Milk Fish Shellfish Wheat/gluten Kiwi Soya Lupin Mollusc Mustard Celery
6680 1554 905 845 546 508 195 427 192 22 21 19 14
6
Management of Food Allergens
allergic child answered one of our surveys with a bullet point list of food allergy’s ‘side effects’:
r r r r r r r r
Holidays are limited to within 10 miles of a large hospital that can speak English. None of the family likes my son being out of our direct care. We can’t go anywhere without medicine of some sort being administered, i.e. antihistamine daily. We pay for private education in a small school that is nut free and allows no packed lunches. Many parents will not have my son for tea let alone sleepovers. Pub food or restaurant visits involve a long and embarrassing plough through lists of food sheets to be told my son can have pasta – without sauce. That’s if they have a food list. Many leisure facilities are not keen to let me leave him in their care. Schoolwork is sporadic at best for my son, and he suffered hair loss from a virus recently. His self-confidence is low.
Although the proportion of the population at risk of anaphylaxis is relatively small, the impact is felt much wider. If it is a child who has the allergy, the burden is carried by parents, siblings, grandparents, aunts and uncles, friends and their parents (particularly if a party is planned) and the local school.
1.3 SUPPORTING CONSUMERS Over the years, it has become apparent that most food allergy problems are manageable – both for the patient and for the industry. But both parties need comprehensive, reliable information, and sometimes that is lacking. Furthermore, there are many allergy myths that have to be dispelled. The most destructive of these myths relates to levels of risk. From the early days, it has been common to hear the comment that ‘even the smell of a peanut will kill my child’ or ‘my child is unlikely to reach adulthood’. This is an understandable fear, because the media will always focus on the worst-case scenario of death. Many of the headlines were sensational from the start: ‘The deadly allergy we should all fear’ was how one magazine put it. Others were equally alarmist – with peanuts described as ‘the hidden killer’ and another announcing that ‘allergies can kill in seconds’. Instead of playing up the risks to gain attention, the Campaign found itself having to keep a cool head and dampen down the hysteria (see Figure 1.2). In order to address this alarm and despondency, the Campaign instigated a programme of support and education to help both those affected by life-threatening allergies and the food industry. Thanks to a National Lottery grant, we opened an office in Farnborough, Hampshire; we developed information tools including fact sheets and training videos; we set up a helpline; we began a series of interactive workshops for allergic teenagers; and we continued to inform and encourage industry. The picture emerging from calls to our helpline showed that many people were receiving woefully inadequate medical guidance, and therefore they were unable to cope with what is fundamentally a manageable condition. Patients need an accurate diagnosis, they need to understand exactly what foods to avoid and where they might encounter those foods, and
The reality of food allergy: the patients’ perspective
Fig. 1.2
7
Newspaper cuttings related to food allergies.
they need to know how to treat themselves when things go wrong. But it was common to hear it reported that a general physician (GP) faced with a young allergic patient had told the child’s parents: ‘If you think he’s allergic to nuts, the solution is simple – don’t give them to him.’ This is easier said than done, and frequent close calls and occasional deaths occur. Crucially, those at risk need to carry their own medication for self-treatment, should they inadvertently encounter an allergen that causes them harm. The front-line treatment is an injection of adrenaline, to be administered as soon as a serious reaction is suspected. People at risk are prescribed their own self-treatment kits (e.g. EpiPen or Anapen), and these must be carried at all times, with no exceptions.
1.4
ALLERGY SERVICES
But first the patient needs a good diagnosis, and as several influential reports have pointed out, allergy services in most parts of the United Kingdom are poor and many patients are unable to obtain the advice they need. In the early days, there were promising developments. During the Campaign’s first few months, the Chief Medical Officer gave the nation’s GPs some clear guidance in one of his regular bulletins (CMO’s Update, May 1994). He told doctors never to advise a peanutallergic patient to test his or her reaction by eating peanuts, stressed the importance of prescribing adrenaline, and asked them to note that repeat injections were often needed. But although we attended several meetings with ministers and civil servants at the Department of Health, allergy services have been painfully slow to improve. A report by the
8
Management of Food Allergens
Royal College of Physicians (2003) declared that 18 million people suffered from allergy at some time in their lives, but there were only six full-time allergy clinics run by an allergy specialist. Other clinics existed, the report said, but most of these were run by specialists in other fields (such as ear, nose and throat) and were a part-time ‘add-on’. People told us frequently that their GP was unable to help. They were floundering with food labelling that confused them, restaurant staff who were reluctant to serve them and (in the case of children) school staff who had received little or no guidance on how to manage what is sometimes a life-threatening condition. The Royal College report said that hospital admissions due to anaphylaxis had increased sevenfold from 1993 to 2003 and doubled from 1999 to 2003. Comments we received during a 2005 survey of our members included the following: The wait to see a specialist is two years. We have known about our son’s allergy since he was nine months and have never been seen. The GP said he had no training in allergies and a slow reintroduction of nut traces may be the way to go. Without the Anaphylaxis Campaign our lives would be much poorer. We would have many more hospital visits. After she suffered a reaction after eating banana, we took her to A&E. We were told it couldn’t have been the banana. We gave it to her again, and she had another reaction. Our son spent months visiting skin and eye specialists, who failed to diagnose his allergy. Figure 1.3 shows a graph of survey results.
Thinking about your experiences, how do you feel the NHS is equipped to manage the needs of people with severe allergy? 500 Series1
450 400 350 300 250 200 150 100 50 0 Very well
Fairly well
Not particularly well
Not at all well
Fig. 1.3 Results of a 2005 Anaphylaxis Campaign Survey looking at people’s perception of allergy services. (Reproduced with permission.)
The reality of food allergy: the patients’ perspective
9
In 2005, the House of Commons Select Committee on Health reported categorically that Britain is in the grip of an allergy epidemic and the present NHS services cannot cope (House of Commons Health Committee, 2004). The committee’s report said that 1 in 50 children in England is now allergic to nuts – almost a quarter of a million children – and there has also been an alarming rise in other allergies, including sesame and latex. The committee documented a huge amount of evidence of unmet need. In 2007, the House of Lords Science and Technology Select Committee again emphasised the scale of the problem after it carried out its own inquiry into allergy (House of Lords Science and Technology Committee, 2007). The committee’s recommendations included the following:
r r r r
At least one allergy centre should be established in each Strategic Heath Authority. This should be run by a full-time allergy specialist. These allergy centres should encourage and coordinate the training of local GPs and other health care workers in allergy. The allergy centre should act as a lead in providing public information and advice. The lead allergist in each allergy centre should be responsible for maintaining a patient database to support clinical research.
We are still waiting for a firm commitment to allergy from the Department of Health. Government reports and statements have proved deeply disappointing. A campaign for change is being led by the National Allergy Strategy Group, a partnership of medical professionals and allergy charities. Early in 2008, a hopeful sign emerged when it was revealed that an All Party Parliamentary Working Group had been formed to ensure allergy maintains a high profile. Deep concern over the lack of activity displayed by the Department of Health is demonstrated by a letter that appeared in The Times on 31 January 2008. Signed by some of the leading allergy experts of the United Kingdom, the letter stated:
We are in the midst of an allergy epidemic with about 20 million children and adult allergy sufferers in the UK. Indeed, we have one of the highest prevalences of allergic diseases worldwide – diseases such as asthma, anaphylaxis, drug and food allergy, eczema, rhinitis and insect-sting allergy. There is an enormous burden on patients and their families, and costs to the NHS are rising.
The letter stated that many problems could be eased or prevented by expert diagnosis and management. The Department of Health had acknowledged the need for improvement, but nothing had been done to implement the key requirements. Instead, responsibility had been passed to local agencies, which could not solve the problem because it required a national solution. The letter ended by urging the Health Minister to adopt the Lords’ recommendations without further delay.
10
Management of Food Allergens
1.5 TEENAGERS AND YOUNG ADULTS Although people of all ages have food allergy, teenagers and young adults are at most risk from fatal reactions. This has been demonstrated by published studies (Pumphrey, 2000; Pumphrey and Gowland, 2007) and our own experience. Many in this age group find it hard to cope. Time and time again, we hear reports of problems faced by teenagers and so this age group became a particular focal point. The experience gained from our 1-day workshops for allergic teenagers is that the participants are often poorly informed and lack confidence in managing their allergies. Many admit they take risks. They know they should carry their adrenaline, but choose to leave it at home, particularly if they do not anticipate that they are going to eat anything. And they are so fed up with defensive labelling (e.g. ‘may contain nuts’) that they just choose to ignore it. The first of our workshops for teenagers took place in London in 1999 and since then we have held almost 100 across the United Kingdom. We found early on that teenagers respond well to the interactive format, in which they share ideas, discuss best practice and act out challenging situations (such as confrontations with difficult restaurant staff). Reporting on one of our early workshops, a young participant wrote in the Campaign’s newsletter: We gained a lot of useful information from each other during the day. We shared experiences and shared ways of dealing with situations and how to tell people about our allergies. We used role play to help devise ways of telling people we have allergies. For example, how do you ask someone you’re going to kiss whether they’ve been eating peanuts? And how do you tell someone who thinks allergies are a joke that it’s a very serious condition. Another wrote: It was a great feeling to know I’m not on my own out there and there are many other people just the same as me. Also I think people should be more aware of other allergies such as dairy products, eggs, fruit and non-food causes of allergy, as most publicity seems to be given to nut allergy.
1.6
FOOD LABELLING
From the early days, it became clear that one of the burning issues for the consumer was food labelling – in particular, the use of ‘may contain’ warning statements. These warning labels are believed to have originated in the confectionery industry in 1994 when it became clear that production methods carried a risk that small quantities of nut could potentially be picked up in the factory by products that were not intended to contain nut. To warn the allergic consumer, food companies began to adopt advisory statements on their packaging. This practice spread like wildfire, and within a few years customers were complaining that ‘just about everything carries a warning these days’. In 2005 and 2006, the Campaign organised two surveys of people with food allergy with the objective of providing evidence to the Department of Health on people’s concerns and needs. One question asked online was: ‘How well or badly does the current system of food labelling work for people with severe allergy?’
The reality of food allergy: the patients’ perspective
11
The results were:
r r r r r
Very well 0.8% Fairly well 32.5% Fairly badly 39.1% Very badly 26.3% Don’t know 1.3%
This online survey, and another survey in which we wrote to every one of our 7695 members, drew many comments about food labelling. Allergic consumers rely on the accuracy of labelling and one respondent expressed the views of many others when she said: ‘A weekly shop takes ages as checking all ingredients is a must.’ The most frequent complaint focused on the increasing use of ‘may contain’ statements: ‘We were looking for a birthday cake in (supermarket X) and all cakes said “may contain traces of nuts”. In a sponge cake with buttercream and jam filling, where would the nuts be? What do I do, take a chance or trail around until I find one that’s OK?’ ‘So many products sold by (supermarket Y) “cannot guarantee nut free” – nearly all the products I once used now say this. What am I supposed to do?’ ‘What’s the difference between “may contain nuts” and “made in a factory that handles nuts”?’ ‘I went through every item in (an in-store) bakery and did not find one which I could eat.’ ‘Recently I saw a bag of peanuts labelled “may contain traces of nuts” – what were they trying to convey?’ ‘At the end of the day the “may contain” label is all down to cleanliness of production lines. (Manufacturer X) have got it sorted, they are fantastic.’ Many respondents had formed the view that ‘may contain’ was simply a defensive device with no substance and some said they disregarded the warnings believing the risk was not genuine. One person wrote: ‘We ignore all “may contain traces of nuts” warnings since we found one of these warnings on a cabbage.’ And another wrote: ‘(My son) eats many things that are made in factories where nuts are handled and a couple of times has had a reaction. But he has to eat!’ Sometimes, labelling proves to be confusing, inaccurate and even dangerous. Respondents ranged from having an understanding attitude, even though life was difficult, to displaying anger. There was a strong feeling that information should be more consistent.
12
Management of Food Allergens
Accuracy of information proved to be a serious concern. Some respondents reported mistakes and discrepancies that they believed to be extremely hazardous. People said they were constantly telephoning manufacturers and retailers to obtain clarity. A few became aware that manufacturers sometimes moved production to different sites, where there were different allergen issues. They had become used to eating a ‘safe’ product only to find that it was now out of bounds. People offered criticism or praise of specific food outlets, depending on their experiences. There were those who felt neglected, such as the respondent who wrote: ‘Some supermarkets are very helpful, but a lot are terrible. The attitude seems to be that “we have enough customers so why bother to improve product safety or choice of food for a minority”.’ Industry has pointed out that ‘may contain’ warnings are usually adopted for good reason – and only after all possible measures have been taken to minimise or eliminated crosscontamination. In our experience, many members of the public are sceptical, especially when they encounter nut warnings in the most unlikely places, such as a bag of mixed salad or even on a packet of nuts. In the two examples given, there may be good reason why the warnings appear, but the average shopper is baffled. The Campaign’s view is that ‘may contain’ warnings are a necessary short-term solution to a difficult problem, but we hope that eventually they will become unnecessary. One possible solution – the Campaign’s certification programme – will be explored later. A study we carried out for the FSA on ‘may contain labelling’ (FSA, 2002) indicated that it takes 39% longer to shop for a nut-allergic person and costs 11% more than for someone without food allergy in the family. Alarmingly, the study also revealed that one in ten ‘may contain’ labels was missed by shoppers reading food packets. Frustration over advisory labelling is just one of many issues that consumers raise with our helpline staff. We are often the first port of call when a reaction occurs. If you can imagine what it can be like for you, or your child, to be rushed to casualty suffering from anaphylaxis, you will probably understand that there is confusion as well as distress. What was it that caused the reaction? People will often jump to conclusions, and these may be the wrong ones. They may blame allergen contamination or – if they are fish allergic – they make think the company has added omega-3 without declaring it. A whole range of scenarios are presented to our helpline. As a first step, if a particular product is believed to be responsible for the reaction, we encourage the consumer to have a full and open dialogue with the manufacturer or retailer, and try to get to the bottom of the problem. We may also contact the company ourselves and explore whether there was indeed allergen contamination or a mistake during production. Or has the consumer become newly sensitised to an allergen? We urge food companies to be honest and open, and consumers not to play the blame game. But if a serious mistake appears to have been made, we will expect local enforcement officers to become involved. Occasionally, a sample of a suspect product is sent off for analysis. Reactions sometimes occur when a food product contains an ingredient that is properly declared but unexpected. We urge people to read the label every time they shop, but people lead busy lives and want short cuts. Consumers have failed to notice egg present in Edam cheese, kiwi extract in an Easter egg moulding kit and casein in a wide variety of pre-packed goods where you might not expect to find milk. Often consumers make mistakes, but on occasions they can be forgiven for being confused. The average person has never been inside a food factory and will have no concept about risk levels. Does ‘made in a factory that handles nuts’ signify a real risk or can the statement be disregarded? Sometimes the packaging tells a contradictory story. The consumer may expect different-sized but similar products to carry the same allergen risk but it may be wrong to
The reality of food allergy: the patients’ perspective
13
make this assumption. In other cases, you find an allergen statement on the outer wrapper but not on the inner wrapper, or vice versa. Confusion may also occur where one brand of pesto sauce contains pine nuts but another is made with cashews. Some Bakewell tarts may be made with almonds, others with apricot kernel. All of this serves to make life difficult for the consumer. Some people’s demands are unrealistic. They expect to see a ‘contains nuts’ warning if coconut is present (when in fact coconut is not strictly speaking a nut); they want frontof-pack allergen information that will save them the trouble of going to the ingredient list (we advise people to read the list of ingredients every time); in extreme cases, they may want an allergen banned from the environment, such as in schools. This last point may sound unreasonable if you are not affected by allergy – and the Campaign itself has serious reservations about allergen bans – but such demands demonstrate the understandable anxiety that people feel. Other consumer demands are more reasonable. If a new allergen has been introduced into a product, they are right to expect to see some indication on pack (such as a ‘new recipe’ flash). A company’s information should be consistent. For example, people should see the same information on pack that appears on the company’s website, which does not always happen. And food companies should be as scrupulous with their ‘free from’ lists as they are with their product labelling. There are occasions when someone makes a mistake for which the manufacturer is partially responsible. This happened with tragic consequences in the case of a 20-year-old mother. She died from an allergic reaction to peanuts following a terrible mistake that involved a ‘may contain’ label. The young woman regularly ate products that carried ‘may contain’ warnings because she had never reacted to them and believed she was fine with them. In October 2003, she prepared a pack of vegetarian sausages for herself and her son. She saw that the pack carried a warning, ‘may contain traces of nuts’, and looked no further. Had she gone next to the ingredient list, she would have seen peanuts listed as an intentional ingredient. As she ate her meal, the young woman immediately realised she was having an allergic reaction. She was treated with adrenaline at her local surgery but by then too much time had elapsed. An inquest recorded a verdict of misadventure. At the time of her death, the young woman had not renewed her EpiPens. It was suggested at the inquest that she might have been frightened of having them. The Campaign’s food adviser, Hazel Gowland, who was an expert witness at the inquest, commented at the time: This was a shocking example of what can happen out there in the real world. We all hope that people with food allergy will always read food labels thoroughly and have adrenaline to hand at all times, but life in the real world is often different. It’s a fact that some people will eat a product carrying a warning label, find they don’t react and believe this means they can eat other similar products. In this young woman’s case, there was an additional factor that led her to making a mistake: the vegetarian sausages carried a ‘may contain’ warning plus the word ‘peanuts’ among the ingredients. She made a wrong assumption that ‘may contain traces of nuts’ was the only reference to nuts. However, it is important to strike a positive note, as well as to list the problems. Industry has made huge strides since 1994. There is a genuine concern for people with food allergies
14
Management of Food Allergens
and even if this is only partially altruistic, it is a fact that you can walk into a shop and see comprehensive allergen labelling that would have been unheard of just over a decade ago. Some companies have acknowledged the needs of the allergic public by developing products ‘free from’ certain allergenic ingredients and although this has sometimes brought its own problems (there have been recalls of products in ‘free from’ ranges), nevertheless this is one more sign that food allergy is well and truly on the industry’s agenda.
1.7
ALLERGEN THRESHOLDS
In the tragic case outlined above, peanut was an intentional ingredient that was present as a significant amount. But a question frequently asked by industry and consumers alike is: When do unintentional traces of an allergen constitute a real risk? Is there a limit below which people won’t react? Scientists around the world are working to develop an understanding of threshold doses for food allergens – the lowest amount that can trigger an allergic reaction. The big questions are: How much is too much? Is there a measurable level below which people wouldn’t react? If so, this would offer much-needed reassurance. Industry could make efforts to reduce cross-contamination to below these levels and consumers might feel safer. However, not everyone is happy with this approach. Some consumers tell our helpline that they want to see every possibility of cross-contamination removed. They know that minuscule traces can trigger reactions and want zero risk. The perception that microscopic traces can kill may be founded largely on scare stories and extreme cases, but it is true that reactions can be triggered by small quantities of allergen. They may not be life threatening but any symptoms requiring treatment are unpleasant and alarming. Furthermore, an allergic person’s own threshold can vary from day to day. How much they react to at any given time may depend on factors such as their general state of health, how well their asthma is controlled, whether they have been exercising strenuously or drinking alcohol, and other factors. Even if industry works to agreed thresholds, would these limits be misleading if a person can react to lower amounts at certain times? People question whether any industry action based on agreed thresholds will protect 100% of the allergic population all of the time. The answer is that there may be a very small minority who are so susceptible that they could react to an amount below the threshold. It is highly unlikely that this would be a life-threatening reaction – but do we know that for sure? Establishing thresholds will not totally remove uncertainty. On the other hand, many consumers agree that establishing thresholds would hold benefits both for them and for the food industry. Total elimination of risk is impossible in any area of life, but risk minimisation is achievable. The most severe allergic reactions are normally caused not by traces, but by significant quantities of allergen, intentionally added to the food. In such cases, there is usually a major error made somewhere along the way, either by the person supplying the food or the person eating it, or both. Food industry action based on agreed thresholds could lead to a reduction in ‘may contain’ labels. We suspect that many of these warnings are applied when a food company is hampered by a lack of knowledge about what constitutes a significant risk. Knowledge of thresholds will make industry’s job easier and have a positive knock-on effect for the allergic consumer. The Anaphylaxis Campaign leans towards the pro-threshold argument. We sympathise with those who want zero risk. They want to see separate factories, or at least segregated
The reality of food allergy: the patients’ perspective
15
production lines, to eliminate cross-contamination altogether, and this is something the Campaign supports wholeheartedly. But this will take some time to achieve and meanwhile risk minimisation is the goal. In our experience, most life-threatening reactions are caused by significant quantities of an allergen. Anxiety is fuelled by the myth that even touching or smelling a peanut is likely to cause death. There is no doubt that unpleasant reactions do occur when an allergen such as peanut or milk gets on to the hands of an allergic person. However, these are more likely to be localised reactions where the contact occurred. The myth that a trace is likely to kill needs to be dispelled. In an attempt to offer practical advice to people, we point out that they can lower their own personal risk by managing their asthma and eczema, taking special care with food when they feel unwell, run down or stressed and ensuring they do not take any risks with food if they have been drinking alcohol, which can increase the severity of allergic reactions.
1.8
FOOD ALERTS
In order to protect its members, the Campaign operates an early alert system that warns people when a mistake has been made by industry and products are on sale that pose a risk. After investigating the circumstances, and becoming sure of the facts, we target our allergic members by first class mail. Our database of members can tell us specifically who we need to target. For example, if necessary, we could pull out all members aged 0–7 years in Bolton with milk allergy. Our first alerts occurred in the late 1990s and the number rose steeply when the European Union’s regulations governing the mandatory labelling of allergens took effect. The Campaign sent out 36 product alerts to allergic members during 2006, compared with 17 in 2004. In 2007, the figure rose to 58. The following incidents are examples of those that have occurred over the years. A company specialising in food for babies and infants announced the recall of four batches of cheese and tomato bake for infants from 4 months. These batches contained milk ingredients but had been incorrectly labelled as milk and lactose free. Table 1.3 shows food alerts during 2007. Due to a packaging error, a batch of children’s ham snacks was recalled because they contained the hot dog variety. The affected packs did not show the correct ingredients or allergy advisory statement and there was a risk of contamination with nuts, milk, soya,
Table 1.3 Anaphylaxis Campaign food alerts by allergen during 2007. Milk Peanuts/nuts Wheat/gluten Soya Sulphites Egg Seafood Celery Mustard Sesame
26 7 7 7 6 5 2 1 1 1
16
Management of Food Allergens
poppy, sesame and sunflower seeds. The error was said to be due to an isolated manufacturing incident at a factory on the Continent. Muesli products marketed by two separate companies were found to contain undeclared nuts – in one case pecan, hazelnut and flaked almond and in the other case whole hazelnuts and Brazils. A retailer’s chicken product containing couscous was labelled ‘gluten free’, but couscous is derived from wheat. Bags of yoghurt-coated raisins were found to contain a number of yoghurt-coated peanuts. The packaging carried a warning statement: ‘This product may contain traces of nuts.’ In a separate incident, the mother of a peanut-allergic child found 13 chocolate-covered peanuts in a bag of chocolate raisins. The product was withdrawn after the company was contacted by environmental health officers. And there was yet another case involving a different company’s yoghurt raisins. These were withdrawn from the shops after a customer found ten containing peanuts in the packet. A major retailer withdrew stocks of fresh shortcrust pastry because, although butter was clearly listed among the ingredients, the words ‘contains milk’ were inadvertently omitted from the allergen box. The product contained a small quantity of milk protein. The message here is that if an allergen advice box is used, it should be comprehensive. In the vast majority of cases, the company with the problem agrees to pay for the mail alert to our members. In a small number of cases, the Campaign has to pay for the alert out of its own funds. The FSA also runs an alert system, based on SMS text messages sent to subscribers. The Campaign welcomes this scheme but decided to remain with its own system after a postal survey among members came out overwhelmingly in favour of alerts by post.
1.9 OUR WORK WITH INDUSTRY Industry faces significant challenges in ensuring that food is safe and properly labelled for people with allergies. What food companies require is information and guidance based on good science. In March 2003, we set up a membership scheme for the food industry in which we offer our subscribers regular, high-quality news bulletins and the opportunity to attend seminars, where problems can be discussed and analysed. Eighty companies had joined this scheme by the end of 2007. Bulletins contain important information on subjects such as peanut allergy research, the need for thresholds, allergy to individual foods such as lupin, poppy seeds and pine nuts, and European legislation. However, we concluded early on in the Campaign’s history that fundamentally what consumers really need is the knowledge that there are consistent, high-quality standards in place for the control of food allergens. Primarily people want to know that they can eat safely and they wish to see a significant reduction in ‘may contain’ warnings. Progress was made in 2007 when the Campaign became the first organisation in the United Kingdom to develop a standard that specifically aims to promote good allergen management and labelling (Anaphylaxis Campaign, 2007). The standard had been developed thanks to a grant from the FSA in 2006. The work was done by recognised experts from within the industry and a rigorous consultation exercise was launched in the autumn of that year. Pilot audits were held and the standard was revised in line with the findings and with the comments that emerged during the consultation. The United Kingdom Accreditation Service (UKAS) completed a technical review of the standard in August 2007 and it became available for sale
The reality of food allergy: the patients’ perspective
17
Fig. 1.4 The logo that food companies will be invited to use on products that have been certified against the Anaphylaxis Campaign Standard. The logo denotes good allergen control during the food production process. (Reproduced with permission.)
in November of that year. In developing the scheme, the Campaign had always intended that companies adopting the standard would have the option of participating in a certification scheme to show compliance with the requirements of the standard. With that in mind, a logo was designed for use on food packets to show the product had been independently certified. The Campaign is working with Doncaster-based Highfield.co.uk Ltd. to seek to ensure that the standard is widely adopted, thus bringing a high quality of allergen control to food production. Highfield was given the task of marketing and selling the standard, and also took on the crucial role of training those people who are required to carry out audits against the standard as well as the staff of food companies that participate in the scheme. Highfield have now launched the Anaphylaxis Campaign allergen training courses for manufacturers, retailers, auditors and trainers (see Figure 1.4). However, it would be completely wrong to infer that all of the good news relating to food allergy has been generated by consumer groups. Specific food companies, trade bodies and the FSA have been in the forefront of numerous positive developments. For example, the British Retail Consortium and Food and Drink Federation have both written realistic and constructive guidance documents for their members on issues such as the handling of nuts and the practical requirements of European allergen legislation. Both organisations have also provided positive input into important FSA initiatives relating to food allergens.
1.10 THE WORK OF THE FSA Since our 1994 meeting with Nicholas Soames, a succession of government ministers and civil servants have shown a strong commitment to food allergy.
18
Management of Food Allergens
In 1997, the Ministry of Agriculture, Fisheries and Foods (MAFF) instigated a poster and leaflet campaign aimed at the catering industry. This was re-branded by the FSA and became the popular ‘Be Allergy Aware’ pack. The FSA produced web guidance for caterers and for newly diagnosed patients, commissioned work on catering, shopping and consumer attitudes, and worked with the Anaphylaxis Campaign to produce a poster aimed at allergic students and leaflets translated into Asian languages. The FSA also published a guidance document for retailers and manufacturers on allergen management, focusing largely on the ‘may contain’ issue (FSA, 2006). At the same time, the FSA developed training for enforcement officers and added food allergy to the curriculum for caterers. In early 2007, the FSA published its allergen guidance for caterers and businesses that sell food loose (FSA, 2008). Furthermore, the FSA’s allergy research programme now funds projects to the tune of more than one million pounds a year.
1.11 SCHOOLS Food allergy has a disconcerting effect on schools and playgroups. Saying goodbye to your severely allergic child at the school gate can be a stressful experience. Parents may be accused of being overprotective, but anyone who has first-hand knowledge of anaphylaxis will understand why a high degree of anxiety exists. Our surveys have found many examples of good practice in schools. As one head teacher stated, it is possible to have excellent policies: All children have to be registered with the school as having an allergy. We have to hold two EpiPens (the expiry dates are noted and parents reminded). The ‘pens’ are held in the office in a secure cupboard with easy access. Each child has the box clearly marked with photos on the box, also consent forms with medication to be given by school from parent, plus a pencil and checklist in each box. The teachers/staff are encouraged to attend annual courses for anaphylaxis training. But there have been many other cases where understanding and awareness were poor. ‘At playgroup, I had to organise the EpiPen training myself, and had to sit in the car outside the school in case of emergency as staff were not confident.’ ‘Our daughter with severe nut allergy was excluded from playgroup for 2 weeks until they could investigate their insurance cover.’ Our surveys showed up inconsistencies with regard to training of school and preschool staff in the procedures necessary to care for allergic children. Where teachers have not been trained, or have forgotten the training, there can be serious issues to deal with: ‘He cannot have school dinners because they feel unable to take the risk . . . this has been hard for a 4-year-old to understand.’ ‘My son was rejected by his nursery school when he was prescribed an EpiPen, which now has a “no EpiPen” policy.’
The reality of food allergy: the patients’ perspective
19
We ask schools to be aware of the dangers of bullying, as everyone who is different is a potential target. The parent of a nut-allergic boy told us: ‘One pupil even tried to make him eat something containing nuts.’ Most schools have at least one child who carries adrenaline, and most will have several. School staff cannot be compelled to administer a child’s injection in the event of a reaction occurring, but in most cases there are enough volunteers to give reassurance to the parents and their child. Educating the school staff is crucial. To meet this need, the Anaphylaxis Campaign has launched a major programme of training. Thanks to the receipt of several grants, including substantial amounts from the American Peanut Council and several food companies, the Campaign devised a comprehensive training package for school nurses. The project was piloted in five areas of the United Kingdom in 2006 and was so successful that a countrywide launch began in 2007 and will continue at least until 2009. The aim is to train all school nurses in the country so that they are able to pass down their knowledge to school staff. Subjects covered in the training packs and seminars include avoidance of allergens, symptom recognition and treatment of reactions, as well as a wide expanse of background knowledge. Even within the first few months, 12 newly trained nurses had led 44 training sessions in schools and reached no fewer than 630 members of school staff. In a separate programme, the Campaign has launched an awareness drive to provide information and guidance for allergic students and for caterers in colleges and universities. This high-risk group forms an important part of the Campaign’s work. The attempt to improve safety in schools and other educational settings may have been hampered by government guidance that actively encourages the sale of nuts and seeds in school tuck shops and vending machines. Under 2007 rules formulated by the School Food Trust, snacks such as chocolate and crisps are banned from sale in schools, and instead they are being told to sell products that are considered healthy. Government guidelines promote nuts and seeds as healthy options. We believe this could lead to an increase in risk for allergic children. Good food labelling will help, but primarily the major risks are likely to be through cross-contamination. Nut proteins tend to become transferred easily from children’s sticky hands to desks, chairs, computer keyboards and other surfaces. The danger is that through ‘casual contact’ with these allergens, susceptible children may suffer reactions. Whilst these reactions may not necessarily be severe, they will certainly be unpleasant and disruptive. The potential risk was demonstrated in a paper published by a German medical team (Lepp et al., 2002). They reported the case of a 32-year-old man with peanut allergy who suffered a serious allergic reaction during a card game. His friends were eating peanuts and peanut protein from their fingers found its way on to the playing cards. As the cards often stuck together, the player with the allergy licked his thumb to separate them. It was this that caused his serious reaction. In an attempt to assess whether our fears are founded, we launched an online survey in which we encouraged families to report whether there was an increase in nuts and seeds in individual schools around the country and whether they perceived the risks to have risen. We are encouraged by the initial findings. Although these are early days, results suggest that there has not been the expected rush by schools to introduce nut products. However, the exercise will have to be repeated to establish whether snack trends in schools will change to the advantage or disadvantage of children with allergies.
20
Management of Food Allergens
1.12 EATING OUT Our more general online survey in 2006 asked respondents about the degree to which they were affected by everyday situations. A high level of concern was given to eating out as a family, with 78% mentioning this as having a great effect on them. In our experience, supported by published studies, people with serious food allergies face significant risks when they eat out. Clearly because this is, in part, due to the fact that the usual food labelling rules do not apply and people are relying on menu statements and verbal assurances by staff. Pumphrey (2000) and Pumphrey and Gowland (2007) have shown that a significant proportion of the reported deaths occurred when food was bought in catering establishments, such as restaurants, hotels and takeaways. Peanuts or tree nuts were frequently implicated. In some cases where the victim had asked for a meal without nuts, the person serving (and in several cases even the caterer) had not been aware that the food contained nuts. In other cases, the request for nut-free food had either been misunderstood or forgotten. This shows that some highly allergic patients know that they need to be extremely vigilant at mealtimes, but mistakes still occur. In our view, most documented cases highlight the importance of prescribing adrenaline for ‘at risk’ allergy patients and educating them about their use. Our surveys showed that eating out ranges from difficult to impossible for people with severe food allergies and their families. People complained about a lack of understanding and said catering staff were frequently unable to offer accurate advice. One person told us: ‘Most waiters just say, “You should be OK.” One waitress said to me: “Please don’t die here.” ’ Respondents noted that some national chains had put in place good systems of allergen control and communication with allergic customers. Implicit in these comments was the idea that if it could be done by one or two companies, why could not others do it? People complained of a lack of consistency among restaurants, hotels and takeaways: ‘Some chefs are brilliant, making special meals, but otherwise you get a lethargic response.’ ‘Restaurants and cafes refuse to serve you.’ ‘We get very embarrassed. People think we are just being fussy if we ask for the ingredients.’ Respondents reported widespread lack of understanding and a need for training and education: ‘Restaurants have no understanding of how life threatening it is to eat a food with nuts.’ ‘(My daughter) had a reaction due to nut-containing foods in a restaurant where two different waiters had assured us there were no nuts.’ ‘There is a widespread belief that only peanuts cause allergies. One restaurant owner guaranteed there were no allergy-causing nuts in a dish, but it arrived with cashew nuts.’
The reality of food allergy: the patients’ perspective
21
‘In a caf´e, I bought chocolate chip cookies labelled “may contain traces of nuts” but found white bits in them which turned out to be peanuts.’ Deaths are rare, but when they occur they generate dramatic media publicity which has a demoralising effect on people affected. My own daughter’s death in 1993 is one of many that have been well publicised. A 13-year-old girl died after eating a small amount of curry sauce made with peanut butter. There was some mention in her doctor’s notes about peanut allergy, but there had been no proper diagnosis and no prescribing of adrenaline. Her family had no idea that she was at risk of a fatal reaction. Another teenage girl collapsed and died during a formal dinner at university after she ate a dessert that – unknown to her – contained nuts. She knew she had nut allergy and had asked her GP for help. Her GP had led her to believe that nothing could be done. There had been no proper allergy diagnosis or advice and no prescribing of adrenaline. Fatalities are thankfully rare but near misses are probably more common than most people realise. A young woman reported a severe reaction requiring hospitalisation following a meal in a cafe. The dessert menu had nut logos on some dishes and she asked about the tiramisu, which did not have a logo. The staff checked and checked again and she was served the tiramisu. She began to have a reaction, which became severe, and was taken to hospital. The restaurant double checked the box and found that the tiramisu she had eaten contained hazelnut crumb. A young London man reported a severe reaction from a meal in his local curry house, where he ate regularly. He said that the staff had told him twice that his menu choice did not contain nuts. He was taken by ambulance to hospital, spent the weekend there and then a week recovering. One must never forget that the onus is on each allergic diner to be clear with catering staff about what he or she cannot eat. This is an important message for people with food allergy. Despite the problems outlined above, many catering businesses now have excellent allergen management systems and effective controls in place. Furthermore, the FSA’s guidance (2008) offers caterers a useful reference point if they wish to take allergy seriously. People with food allergy can be assured that things are improving. But here, as ever, we return to the problem of inadequate allergy advice under the NHS. How can people protect themselves adequately if they do not have an expert diagnosis and high-quality guidance in the first place?
1.13 DAILY LIFE WITH A FOOD ALLERGY People’s lives are affected in many other ways, some of which are demonstrated by these comments that arose from our surveys: ‘No woodland walks, house plants, or fresh flowers, no holding my grandchildren’s cuddly toys.’ ‘I am smell-sensitive to nuts. This makes it impossible for me to go to the cinema.’ ‘I can’t spontaneously embrace my husband in case he has eaten anything I’m allergic to.’ ‘My husband has had to give up his job to look after my daughter during school holidays.’
22
Management of Food Allergens
‘We had problems booking flights with one airline. They wanted a form completed by the doctor before letting our son fly, so we booked with another airline which just asked for a doctor’s letter.’ ‘If airlines provide nut-free meals, this is usually reliable on the outbound journey but frequently forgotten on the inbound.’ ‘I was made to feel like a second-class citizen. Why should I be denied the social pleasures everyone else enjoys because of ignorance?’ ‘When my son leaves home for university, he is likely to eat out more. Because of the years that have passed without an anaphylactic reaction, he may become less inclined to carry his EpiPen and less careful about reading ingredients labels.’ However, there are positive experiences among the negatives: ‘Whilst on holiday with (travel company A) recently, it was fantastic to hear the captain announce the need for all passengers to avoid all nuts as there were two passengers on board with severe nut allergies.’
1.14 HOPES FOR THE FUTURE What does the future hold for people with food allergy? There is some cause for concern. Pumphrey and Gowland (2007) showed that almost a third of the people who die from food allergy are not actively avoiding the particular food allergen. Some may have had minor symptoms in the past, but they were not expecting symptoms to become so severe so quickly. This presents a challenge to GPs, specialists and all those who diagnose allergy. It is also the job of researchers to devise ways of identifying the patients who are most at risk: something that cannot be done with any certainty yet. The quality of life of people with food allergy is also a burning issue. One study (Avery et al., 2003) showed that quality of life was more severely impaired in peanut allergy than insulin-dependent diabetes. The great challenge for the future is to improve the lives of allergic people by developing the clinical services and information systems, but also by better food manufacturing procedures and labelling to allow people to make accurate and appropriate risk assessments. Clearly there is the chance, too, that further allergens will be identified as causing problems. When the Campaign first began its work, just a few food allergens were considered worthy of attention. Then kiwi fruit appeared on the radar. And others, such as lupin flour, began to emerge as potential problems. A young man with peanut allergy suffered a severe allergic reaction in 2002 after eating a chicken and ham pie in his office canteen. He suspected there were peanuts in the meal, but laboratory tests proved negative. Then it was pointed out that the pastry – imported from France – contained lupin flour. He saw an allergist and skin prick tests showed him to be extremely allergic to lupin. Until these emerging allergens are better understood, GPs cannot be expected to know about them, let alone know how to offer advice. There is some cause for unease. However, there is also much cause for optimism. The Campaign’s standard promoting consistent allergen control provides a real opportunity for food companies to get to grips with allergens for the benefit of the allergic public. We hope
The reality of food allergy: the patients’ perspective
23
to see more and more companies opting for certification, but even without it, the standard will improve life considerably for industry and consumers alike. On the schools front, the Campaign’s national training programme will inevitably lower risks for allergic pupils. The present commitment by the FSA to allergy is to be applauded and we hope the good work will continue. The campaign to achieve an improvement in allergy services will continue, despite the lack of attention given to allergy so far by the Department of Health. The National Allergy Strategy Group will continue campaigning until the point is reached where everyone with an allergy gets optimum diagnosis, treatment and guidance. But what people really want for the future is a cure. Can the effects of a severe allergic reaction be reduced by drugs, and can allergy even be switched off altogether? Many parents who face the grim prospect of their allergic child remaining severely allergic for life – with the fear of death never far away – ask us passionately about the likelihood of medical science finding the elusive cure. Only this will end their anxiety. The answer is that there is good work going on worldwide with this objective in mind. Oral immunotherapy offers the possibility that people with food allergy can become desensitised by eating increasing amounts of the culprit food over a long period. Their immune systems would eventually become tolerant to the food. There are several research teams, in the United Kingdom and elsewhere, hoping to demonstrate that this can be done. The US researchers have had success with egg (Buchanan et al., 2007) and are trying to do the same with peanut. Children in their study had reduced symptoms to egg when the team increased the amount they ate over a 2-year period. After the 2 years of desensitisation, all the children tolerated a higher dose of egg than at the outset of the study, and this was more than would typically be encountered in an accidental exposure. Most of the children could tolerate two scrambled eggs with no adverse reaction by the end of the study. In those who did react, the reactions were less severe. Eventually, the study team hopes to induce lasting tolerance to egg. Anti-IgE therapy is intended to block the action of IgE, the antibody responsible for triggering the cascade of symptoms in patients with allergies. The drug could be delivered by injection once or twice a month indefinitely to people at risk of severe allergic reactions. People might still react, but the severity of reactions would be diminished. People would still be advised to avoid eating peanuts, but they might not have to worry about having a life-threatening reaction if they eat a small amount by mistake. Researchers in the United States are experimenting with a vaccine based on Escherichia coli bacteria (Li et al., 2003). These are killed and then used as a carrier for modified peanut proteins. These would be administered to the allergic person by suppository. The treatment seems to activate the immune system to turn off the IgE response. Canadian researchers believe a protective enzyme found in the blood decreases the severity of allergic reactions (Vadas et al., 2008). The enzyme acetylhydrolase breaks down PAF (platelet-activating factor), a chemical produced by the body as part of a severe allergic response. People with low levels of acetylhydrolase appear to have more severe reactions than those with higher levels of the enzyme. If the findings are verified, drugs might be developed that would treat life-threatening allergic reactions when they occur and possibly even protect people from experiencing severe responses in the first place. Furthermore, scientists are also working on the question of what makes people allergic in the first place – and can this trend be reversed? Since 1994, the government has wrestled with the problem of why peanut and other food allergies are on the increase. The famous Isle of
24
Management of Food Allergens
Wight studies showed initially that 1 in 200 children was peanut allergic and a decade later that the figure had risen to 1 in 70 (Grundy et al., 2002). Parents of a peanut-allergic child have their own theory about how their child became sensitised. Many mothers become convinced it was because they ate peanuts while pregnant or breastfeeding, and are consumed by guilt. Although such anecdotal evidence is unreliable, the Department of Health gave credence to this hypothesis in 1998. Drawing on evidence provided by the Committee on Toxicity, the Department of Health issued guidelines to women suggesting that they ‘may wish’ to avoid nuts and peanuts during pregnancy and while breastfeeding if there is allergy in the immediate family (i.e. if they or their partners are allergic, or if they have an allergic child, e.g. with asthma, eczema, hay fever or food allergy). The guidelines also state that peanuts and tree nuts should not be introduced to children until after 3 years of age (Committee on Toxicity, 1998). Recently, serious doubts have been expressed about this advice. A team of London-based researchers believe that they may demonstrate that introducing peanut-containing foods early into children’s diets will actually prevent peanut allergy from developing. The team has been recruiting infants (aged 4–10 months) with severe eczema and/or egg allergy from the London area. Half the children are receiving peanut-containing snacks and the other half are advised on the avoidance of peanuts (as per the Department of Health recommendations). When the children have reached 5 years of age, a determination of the relative rates of peanut allergy will be made. It is a case of ‘watch this space’. There are many experts who believe this study offers hope that the apparent rise in food allergy can be reversed. In this chapter, I have tried to demonstrate that whilst there is much good work taking place to improve life for people with food allergy, the level of anxiety that some experience is unbearable. That is summed up poignantly in this final quotation: Our daughter suffered from an allergy to peanuts. There was no specialist support system in place to help her. In particular, we had no knowledge of the fact that peanut allergy sufferers may also be allergic to lupin flour – an ingredient in high grade pastry. At the age of 15 she had this potentially fatal anaphylactic reaction to lupin flour following the consumption of school food. The paramedic responded well but the local hospital was lamentable following the first attack. The doctor dealing with our daughter had never before seen an EpiPen. He succeeded in squirting it up a wall and suggested we find a pharmacy to replace it since the hospital did not have any. We have no doubt that the cavalier treatment Alice received affected her psychologically. She took her own life on the eve of starting at university and having just completed her Duke of Edinburgh Award Gold.
REFERENCES Anaphylaxis Campaign (2007) The Anaphylaxis Campaign Standard to Increase Trust in Information about Allergens in Food. The Anaphylaxis Campaign, Farnborough, Hampshire. Avery N., King R., Knight S. and Hourihane J. (2003) Assessment of quality of life in children with peanut allergy. Pediatric Allergy and Immunology, 14(5), 378. Buchanan A., Green, T., Jones, S. et al. (2007) Egg oral immunotherapy in non-anaphylactic children with egg allergy. The Journal of Allergy and Clinical Immunology, 119(1), 199–205. Committee on Toxicity (1998) Report on Peanut Allergy. Department of Health (Government report), London. Available at http://cot.food.gov.uk/cotreports/cotwgreports/cotpeanutallergy, accessed 25 January 2008.
The reality of food allergy: the patients’ perspective
25
FSA (Food Standards Agency) (2002) ‘May Contain’ Labelling – The Consumer’s Perspective. Food Standards Agency, London. Available at http://www.food.gov.uk/foodlabelling/researchandreports/ maycontainsummary, accessed 25 January 2008. FSA (Food Standards Agency) (2006) Guidance on Allergen Management and Consumer Information. Food Standards Agency, London. Available at www.food.gov.uk/multimedia/pdfs/maycontainguide.pdf, accessed 25 January 2008. FSA (Food Standards Agency) (2008) The Provision of Allergen Information for Foods that Are Not Prepacked. Food Standards Agency, London. Available at http://www.food.gov.uk/foodindustry/ guidancenotes/labelregsguidance/nonprepacked, accessed 25 January 2008. Grundy J., Matthews S., Bateman B. et al. (2002) Rising prevalence of allergy to peanut in children: data from two sequential cohorts. Journal of Allergy and Clinical Immunology, 110(5), 784–789. House of Commons Health Committee (2004) The Provision of Allergy Services. Sixth Report of Session 2003–2004, Vol. 1. Parliamentary Publication (UK Government), London. Available at http://www. publications.parliament.uk/pa/cm200304/cmselect/cmhealth/696/69602.htm, accessed 25 January 2008. House of Lords Science and Technology Committee (2007) Allergy. Sixth Report of Session 2006–2007. Parliamentary Publication (UK Government), London. Available at http://www.publications.parliament. uk/pa/ld200607/ldselect/ldsctech/ldsctech.htm, accessed 25 January 2008. Lepp U., Zabel P. and Schocker F. (2002) Playing cards as a carrier for peanut allergens. Allergy, 57(9), 864. Li X., Srivastava K., Grishin A. et al. (2003) Persistent protective effect of heat-killed Escherichia coli producing engineered, recombinant peanut proteins in a murine model of peanut allergy. Journal of Allergy and Clinical Immunology, 112(1), 159–167. Pumphrey, R.S.H. (2000) Lessons for the management of anaphylaxis from a study of fatal reactions. Journal of Clinical and Experimental Allergy, 30, 1144–1150. Pumphrey R.S.H. and Gowland M.H. (2007) Further fatal reactions to food in the United Kingdom 1999–2006. The Journal of allergy and clinical immunology, 119, 1018–1019. Royal College of Physicians (2003) Allergy the Unmet Need: A Blueprint for Better Patient Care. Royal College of Physicians, London. Available at www.rcplondon.ac.uk/pubs/books/allergy/allergy.pdf, accessed 25 January 2008. Vadas P., Gold M., Perelman B. et al. (2008) Platelet-activating factor, PAF acetylhydrolase, and severe anaphylaxis. New England Journal of Medicine, 358(1), 28–35.
2
Clinical incidence of food allergy
Zsolt Sz´epfalusi and Thomas Eiwegger
2.1 INTRODUCTION The understanding of food allergy and hypersensitivity has progressed substantially since 1950, when Loveless first suggested using masked food challenges to objectively determine the veracity of patients’ histories of food allergy. May (1976) codified this approach and established the double-blind, placebo-controlled food challenge (DBPCFC) as the gold standard for the investigation and accurate diagnosis of adverse reactions to food. The definition of the term allergy has been replete with misunderstanding when applied to foods. This chapter restricts the use of the term allergy to those adverse food reactions that have been shown to have an immunologic basis of strong immunologic association. A series of crucial clinical case histories detailed below will lead the reader through the areas of diagnosis and management of anaphylaxis due to food allergy:
r r r r
making a diagnosis, deciding who needs a special diet with particular restrictions, deciding who needs a self-injectable adrenaline device, and for how long and considering new allergens causing food allergy.
Anaphylaxis is a common paediatric emergency with less frequent incidences in adulthood (see Figure 5 in Poulos et al., 2007) (Figure 2.1). There is no current consensus about the specific criteria for diagnosis of anaphylaxis, as reflected in very different incidence assessments among countries. A widely accepted definition would be a ‘severe, life-threatening generalised or systemic hypersensitivity reaction’ (Johansson et al., 2004) or a ‘serious allergic reaction that is rapid in onset and may cause death’ (Sampson et al., 2006). In allergic anaphylaxis, there is an IgE-mediated systemic release of mediators from mast cells and basophils. This leads to the development of the cutaneous, respiratory and/or cardiovascular symptoms and/or signs that are seen in anaphylaxis (Sheikh and Walker, 2005). The development of such IgE-mediated symptoms is shown in Figures 2.2a–2.2d. These photographs show before and after food challenge with microgram amounts of peanut in a small child. Figures 2.3a and 2.3b show the development of urticaria and eczema in a baby challenged with food material containing milk. The incidence of anaphylaxis is unclear, but probably ranges from 1 to 20 per 100,000 person-years (Yocum et al., 1999; Macdougall et al., 2002; Braganza et al., 2006). Routine UK hospital admission data suggest an increase of 250% in severe anaphylaxis between 1995 and 1999 (Sheikh and Alves, 2000) (see Figure 2.4). Among the causes of anaphylaxis in infants and children, foods are the most frequent, with peanuts, tree nuts, milk, egg and
Clinical incidence of food allergy
27
Image not available in this electronic edition.
Fig. 2.1 Australian hospital admissions for anaphylaxis caused by food by broad age group, 1994–1995 to 2004–2005. (Reprinted from Poulos et al . (2007), with permission from Elsevier.)
(a)
(b)
(c)
(d)
Fig. 2.2 (a) Child prior to challenge. (b) Child showing cutaneous and respiratory distress shortly after consumption of small amount of peanut material. (c) Further progression of the Ig-E-mediated allergic response-producing urticaria. (d) Urticaria on hand.
28
Management of Food Allergens
(a)
(b)
Fig. 2.3 (a) Baby showing signs of urticaria and discomfort when challenged with small amount of milk protein. (b) Progression of the allergic response causing itching and spread of urticaria.
fish being the most common precipitants (Pumphrey, 2000; Bock et al., 2001; Mehl et al., 2005; Bock et al., 2007; Pumphrey and Gowland, 2007). While the majority of children outgrow their allergy to milk, egg, wheat and soya, allergies to peanut, tree nuts, fish and shellfish are often lifelong. Although any food can cause anaphylaxis, the most commonly implicated foods for severe allergic reactions are peanuts, tree nuts, fish and shellfish (Burks et al., 1999).
Image not available in this electronic edition.
Fig. 2.4 Number of hospital discharges with the primary diagnosis of anaphylaxis per 100,000 episodes of hospital discharge and cause of anaphylaxis (in England). (Reprinted from Sheikh and Alves (2000), with permission from BMJ Publishing Group Ltd.)
Clinical incidence of food allergy
29
2.2 CASE 1 – SEVERE ANAPHYLAXIS TO AN UNKNOWN FOOD PRODUCT A 7-year-old boy complains of an itchy mouth at a birthday party. Ten minutes later, he starts to develop an itchy rash on his face. Within minutes his lips become swollen and he indicates to his friend’s parents that he is experiencing difficulties in breathing. The parents of his friend give him some cetirizine syrup and call the ambulance. He is treated in the emergency department with oxygen, subcutaneous adrenaline, nebulised salbutamol and intravenous prednisolone. He recovers after a couple of hours without developing any other symptoms. The dietary history reveals that he consumes all the common food sources without any problem but he avoids peanuts since a school friend had experienced an anaphylactic attack in the classroom while eating peanut chips. Due to this event, the child attended a consultation at an outpatient allergy clinic in a university hospital. The central question was whether or not the boy had true peanut allergy. In his history, the boy had eczema assigned as ‘atopic’ by his paediatrician. The examination is unremarkable. Skin prick testing (SPT) with peanut, a selection of tree nuts (walnut, cashew, almond) and a panel of common food allergens (milk, egg, wheat, soy, fish) revealed a positive reaction only to peanuts (6-mm wheal). The histamine control gave a 4-mm wheal; the saline control was negative. Specific IgE levels to peanuts were elevated with 12 kU/L (Roberts, 2007). The questions arising from this case can be summarised as follows:
r r r r
Is the child allergic to peanuts? Is the anaphylaxis-inducing food peanut? Should a diet be recommended? Is a self-injectable adrenaline to be prescribed?
If the child has experienced his first adverse reaction to a specific food (i.e. peanut), the positive predictive value for correctly identifying the allergen is 50% (Sampson and Ho, 1997). If the adverse reaction to the same food has already occurred three times, the positive predictive value for that food to be the allergen is around 100% (Eggesbo et al., 2001). When the history is not diagnostic, testing by skin prick or serum-specific IgE levels is required. SPT is less expensive and the result is immediately available. Serum-specific IgE may be used when a greater range of allergens needs to be tested and, in contrast to prick testing, concomitant medication (e.g. antihistamines, steroids) does not affect the result. It has been recognised that the greater the wheal response on SPT or the higher the serum-specific IgE level, the more likely a positive result will indicate clinical allergy. Many different groups have generated cut-off levels for different populations (Sampson and Ho, 1997; Sporik et al., 2000; Boyano Martinez et al., 2001; Garcia-Ara et al., 2001; Sampson, 2001; Rance et al., 2002; Bernard et al., 2003; Perry et al., 2004). The use of these cut-off levels may reduce the need for food challenges (Table 2.1). An SPT result of <3 mm or a serum-specific IgE of <0.35 kU/L is generally regarded as disproving allergic reactivity. However, a negative result can also be seen in individuals with severe food allergy. Thus, if the history strongly suggests an allergic reaction, despite negative skin prick and serum-specific IgE results, a food challenge should be performed.
30
Management of Food Allergens
Table 2.1
Positive predictive values at >95% of specific IgE levels and skin prick testing.
Specific IgE levels (kU/L)a Hens egg Cows milk Peanuts Tree nuts Fish
Age >2 years 7 15 15 15 20
Infants <2 years 2 5
Skin prick test positivityb (wheal diameter in mm) Hens egg Cows milk Peanut
7 8 8
5 6 4
Note: Negative allergy tests (specific IgE levels and/or skin prick tests) may still be associated with clinical reactivity. Allergy tests should never be interpreted in the absence of a thorough allergy history. a Sampson and Ho, 1997; Boyano Martinez et al., 2001; Garcia-Ara et al., 2001; Sampson, 2001; Rance et al., 2002. b Sporik et al., 2000; Clark and Ewan, 2003; Roberts and Lack, 2005.
2.3 CASE 2 – IDIOPATHIC ANAPHYLAXIS A 30-year-old medical trainee is admitted to the emergency department with dyspnoea, low blood pressure, generalised urticaria and angioedema. Stabilisation is achieved with volume supply of steroids and antihistamines. Allergological work-up reveals normal total and specific IgE-levels to common inhalant and nutritive allergens. SPT is negative and the histamine wheal is 6 mm. Serum tryptase taken at admission is normal. The patient has not had any immediate reactions in the past. The family history reveals neither allergies nor bronchial asthma or allergic rhinitis. Four months prior to the admission, the patient was suffering from pneumonia with mycoplasma pneumoniae IgM titer of 1:128 suggestive of mycoplasma pneumonia. Upon discharge, an oligoallergenic and histamine-reduced diet was recommended. The patient tolerated this diet for 4 weeks, then abandoned both diets and was well until 6 months later when a second anaphylactic event occurred resulting in severe dyspnoea with oxygen desaturation, blood pressure drop and collapse. Emergency treatment was initiated. Allergological investigations were repeated and again revealed no allergic (IgE-mediated) sensitisation. Serum tryptase was slightly elevated at that time. The patient was discharged, a recommended diet was given again and a daily antihistamine was prescribed. The following questions shall be answered:
r r r
Shall a diet be recommended? Is a self-injectable adrenaline device to be prescribed? What is the cause of the anaphylactic reaction?
Elimination diets should be undertaken with caution, especially if a number of foods or food groups are to be restricted. Several reports have documented the misuse of elimination diets resulting in inadequate caloric intake and failure to thrive in different patient populations (Bierman et al., 1977; Lloyd-Still, 1979; David et al., 1984; Liu et al., 2001). Accurate identification of the offending allergens is critical not only to get resolution of symptoms
Clinical incidence of food allergy
31
by eliminating the food, but also to avoid unnecessary restriction of foods from the diet. When elimination diets are used in the diagnostic process, it should be for a specified trial period. The symptoms attributed to the food should resolve and then reappear when the food is reintroduced (Bindslev-Jensen and Poulsen, 1998; Sicherer, 1999). In the present case, a suspected food or food group possibly responsible for the anaphylactic reaction could not be defined. However, an oligoallergenic diet, a so-called ‘eat only’ diet, was recommended, comprising rice and rice products, corn and corn products, cooked vegetables (broccoli, carrots, spinach, sweet potatoes) and chicken. Without any feedback reassessment, the patient abandoned the diet after 4 weeks and tolerated a normal diet for another 6 months. It is likely that most patients with uncontrolled diet recommendations would abandon this procedure if no change in health is observed. In general, food restricted from the diet should only include those items for which the patient has either a clear history of clinical reaction or for which symptom resolution during a diagnostic diet recommendation has been achieved. Oligoallergenic diets should never be recommended as a long-term procedure without procuring the assistance of a dietitian, assessing a full nutritional status so that sufficient nutrients are provided to promote appropriate growth and to ensure proper institution of alternate food sources in the diet. The question of whether or not to prescribe a self-injectable adrenalin device is largely restricted to the history the individual has reported. Unfortunately, anaphylactic reactions had occurred twice without elucidation of the causative agent/food. Despite the fact that no elicitor had been defined, the clinical reactions revealed a rapid progression from the first symptoms up to respiratory and cardiovascular distress (collapse/shock). Thus, as summarised in Table 2.2, only one aggravating factor for the increased risk of an anaphylactic reaction in foodallergic children is fulfilled (age over 5 years), but immediate course of disease progression would support recommending an adrenaline auto-injector. In this instance, the cause for the anaphylactic reaction could not be elucidated. For such cases, also called idiopathic anaphylaxis, some authors recommend a sustained application of H1-antihistamines for a certain period of time, as recommended in chronic urticaria patients (Ring et al., 2004; Asero, 2007; Wedi and Kapp, 2007). However, evidence-based data on the effectiveness of this treatment are not available.
2.4 CASE 3 – CROSS-REACTIVITY OR CONTAMINATION? A 3-year-old boy experienced an anaphylactic reaction with respiratory distress (wheezing and oxygen desaturation) on first known peanut ingestion. Hospitalisation and emergency treatment with intramuscular adrenaline, intravenous volume supply and intravenous corticosteroids stabilised the anaphylactic reaction. SPT and IgE measurements revealed a 12-mm wheal reaction to peanut extract and a peanut-specific IgE of 23 kU/L. Strict peanut avoidance was recommended, an adrenaline auto-injector was prescribed and the parents were taught how to use it appropriately. Two years later, at the age of 5 years, the boy is admitted to the emergency department of the University Children’s hospital with wheezing, generalised urticaria and angioedema after consumption of a soya pudding. History revealed that the boy was eating soya pudding and other soya products on a regular basis without any symptoms. At this time, due to low pricing, the parents had bought a large amount of soya pudding from a different brand. At the first consumption of this ‘new’ soya pudding, the boy experienced his anaphylactic reaction.
32
Management of Food Allergens
Table 2.2
Prevalence of food allergy (epidemiology, sensitisation and food challenge).
Steinke et al., 2007
Age group (years)
Prevalence of food allergy
Methodology
2–3
7.2% 1.7–11.7%
Reported (interview) Austria-Finland
No
34.9%
Yes
3.6% 2.5%
Reported (questionnaire + interview) DBPCFC DBPCFC + sensitisation (sIgE)
Zuberbier et al., 2004 0–80
Food challenge
Roehr et al., 2004
0–17
38.4% 4.2%
Reported (questionnaire) DBPCFC
Yes
Osterballe et al., 2005
<3 to adults
7.2–10.3% 2.3–3.2%
Reported (questionnaire) Sensitisation (SPT) Food challenge
Yes
Pereira et al., 2005
11–15
11.6% 5.1% 1.0% 0.1%
Reported (questionnaire) Sensitisation (SPT) Food challenge DBPCFC
Yes
Venter et al., 2006a
1
14.2% 2.2% 1.4–2.8% 0.9–2.5%
Reported Sensitisation (SPT) Food challenge DBPCFC
Yes
Venter et al., 2006a
6
11.8% 3.6% 2.5% 1.6%
Reported (questionnaire) Sensitisation (SPT) Sensitisation + food challenge Sensitisation + DBPCFC
Yes
Venter et al., 2008
1–3
5.3% 6.0% 5.0%
Sensitisation (SPT) Sensitisation + food challenge Sensitisation + DBPCFC
Yes
Dean et al., 2007
1–3
2.8–3.9%
Sensitisation (SPT)
Yes
Ostblom et al., 2008
4
1.6%
Sensitisation (sIgE) + symptoms No
The following questions shall be answered:
r r r
Was the reaction due to soya or another food? Is IgE cross-reactivity between soya and peanut responsible? Shall the parents be advised to avoid soya products as well as peanut?
Soybean allergy has been well described in more than 20 studies (Sicherer et al., 2000; Besler et al., 2001; Kleine-Tebbe et al., 2002). In general, individuals report cutaneous, gastrointestinal and respiratory symptoms with some studies specifically reporting oral allergy syndrome. There is also evidence of exercise-induced anaphylaxis to soya (Taramarcaz et al., 2001). SPT with commercial extracts seems to underestimate clinically reactive soya allergy proven by DBPCFC. In contrast to that, many SPT positive children remain food
Clinical incidence of food allergy
33
challenge negative (Eigenmann and Sampson, 1998). IgE cross-reactivity between soya and peanut is described as both foods belong to the family of Leguminosae. Eigenmann et al. (1996) showed that there is extensive binding of anti-peanut IgE to soya proteins and vice versa. However, the affinity of the antibodies is reduced. After removal of peanut-specific IgE, strong binding was detected to soya proteins at 21 kDa and 46 kDa. Cases of anaphylaxis, four of them fatal, in peanut-allergic individuals after eating soya have been reported (Foucard and Malmheden Yman, 1999). However, it has been argued that the food may have been contaminated by peanuts and that peanut-allergic patients rarely react clinically to soya (Sicherer et al., 2000). The molecular IgE-recognition profile to peanut and soybean allergens has not been investigated in the 5-year-old boy. It is most likely that the reactivity was caused by some peanut contamination, since the clinical reactions with anaphylactic symptoms were similar to those experienced with peanut ingestion 2 years earlier. Even low-grade clinical symptoms to soya pudding consumption were not observed during previous years. Thus, question 2 might be answered inconclusively with a no; cross-reactivity is not likely to be the reason for the clinical reaction. In consequence to the answer of question 2, the parents should not be advised to avoid soya products from now on. In the present case, to have better evidence for this recommendation, a double-blinded soya challenge with a commonly tolerated soya pudding and the incriminated ‘new’ soya pudding brand was performed. The child experienced severe clinical reactions with respiratory distress at very low doses of the ‘new’ brand soya pudding, but tolerated the ‘old’ soya pudding. SDS-Page electrophoresis revealed similar content of total and specific proteins content in both preparations. IgE-immunoblotting and immunoblot inhibition was not performed. In summary, the challenge experiments together with the protein analysis reveal that contamination rather than cross-sensitisation is most likely the reason for this child’s clinical reactivity to one particular soya pudding. Thus, so far the parents were not advised to restrict the intake of soya products by the child, but the child itself refused to eat soya products by the time of the diagnostic challenge.
2.5 CASE 4 – TO VACCINATE OR NOT IN EGG ALLERGY An 11-month-old girl is presented with presumptive egg allergy. On contact with raw egg on her hand at the age of 2 months, she developed localised urticaria which disappeared spontaneously. A couple of weeks later she again developed urticaria on her cheek after being kissed by her mother, who had eaten ham and eggs. Egg was never knowingly introduced in her diet. SPT with egg white and egg yolk revealed a 9-mm wheal. SPT to cow’s milk, wheat and soy were negative. MMR vaccination is shortly due and her mother wonders whether she should allow this since the package insert says MMR vaccination is contraindicated in egg allergy (Kemp, 2007). The following questions shall be answered:
r r r
What is the risk of MMR vaccination? What is the risk of influenza vaccination? Has egg allergy resolved meanwhile?
34
Management of Food Allergens
Transient low levels of IgE antibodies against egg proteins often occur in healthy infants without any symptoms (Hattevig et al., 1984; Bonnelykke et al., 2008; Pfefferle et al., 2008). The prevalence of IgE-mediated reactions to egg in population-based studies of children younger than 2 years is reported to be between 1 and 4.2% (Eggesbo et al., 2001; Roehr et al., 2004; Zuberbier et al., 2004; Osterballe et al., 2005; Pereira et al., 2005; Venter et al., 2006a, 2006bb; Dean et al., 2007; Steinke et al., 2007). Allergic reactions to egg protein are commonly observed within the second 6 months of life. Most frequently, the symptoms are cutaneous with erythema and urticaria occurring in 90% of children with a median age of 10 months at presentation (Boyano Martinez et al., 2001). Symptoms usually occur within 30 min of egg contact. Delayed reactions for 1–2 h may occur in a minority of cases (Boyano Martinez et al., 2001). Contact urticaria on areas in contact with egg protein is a common finding (de Boissieu and Dupont, 2006). The application of raw egg may sometimes be followed by systemic symptoms of erythema/urticaria outside the application area. Gastrointestinal symptoms, abdominal discomfort and vomiting occur in 40–50% of cases following the ingestion of egg. Clinical reactions do occur at the first known exposure with egg (Monti et al., 2002; de Boissieu and Dupont, 2006). Monti et al. described a series of 107 infants with atopic dermatitis with no known egg ingestion, but evidence of IgE-sensitivity to egg and positive reactions to an oral food challenge in 67%. Sensitisation may occur either by transplacental or transamniotic transfer of egg proteins (Holloway et al., 2000; Sz´epfalusi et al., 2000) through breast milk (Palmer et al., 2005) or potentially through the skin in eczematous skin lesions in infancy (Lack, 2008). It is worth mentioning that there is no evidence that egg avoidance in pregnancy decreases sensitisation to egg proteins and egg allergy in infants (Falth-Magnusson and Kjellman, 1992). Egg allergy is regarded as one of the contraindications for an MMR vaccination. This issue still raises many uncertainties since measles vaccines are produced in a culture of chick embryo fibroblasts. Despite very low numbers of minor reactions (n = 4), the majority of 410 egg-allergic children (406 of 410; 99%) tolerated MMR vaccination (Aickin et al., 1994). For influenza vaccination, anaphylactic reactions have been reported in egg-allergic children (Bierman et al., 1977; Retailliau et al., 1980). In severe life-threatening reactions to egg, influenza immunisation is generally contraindicated. In non–life-threatening reactions to egg, after SPT and intradermal testing with the vaccine, a sequential application of the 1:10-diluted vaccine can be given every 20 min, five times (Sugai et al., 2007) or a two-dose injection protocol (1/10, followed 30 min later by 9/10 dose) (James et al., 1998). Thus, in children with non–life-threatening reactions to egg in whom influenza vaccination is indicated, SPT with undiluted influenza vaccine should be performed. In case of no reaction, a two-dose step vaccination should be performed. Table 2.3 shows the prevalence of food allergy, as reported across a number of different studies in a range of countries.
2.6 CASE 5 – ADRENALIN AUTO-INJECTOR FOR ALL EGG-ALLERGIC PATIENTS? A 24-month-old boy is referred because his doctor wants to know whether he should have an adrenaline auto-injector. At 9 months of age, following consumption of a quarter teaspoon of cooked egg, he developed facial erythema, lip swelling, a hoarse cry and possibly some
Clinical incidence of food allergy Table 2.3
35
Aggravating factors in food-allergic children.
Age over 5 years
Most deaths have been observed in school-age children and teenagers
History of bronchial asthma
Associated with more deaths; in particular, if controlling medication is needed
Peanut and tree nut allergy
Particular risk for severe and fatal reactions
Small allergen amounts or traces eliciting clinical reaction
High association with fatalities in older patients (no data on younger children and toddlers)
Strongly positive skin prick test
High association with fatalities
Kemp (2003).
respiratory distress which resolved spontaneously over a couple of hours. Medical help was not sought. Again, at 13 months of age, following consumption of another quarter teaspoon of whole cooked egg, he had a similar reaction with hoarse cry. He has not reacted to other foods and has not been exposed to peanut, other nuts or fish products. Skin prick tests were positive to egg yolk (7-mm wheal) and egg white (8-mm wheal). The following question has been asked:
r
Is an adrenaline auto-injector to be prescribed?
This child also experienced some respiratory symptoms, in addition to cutaneous symptoms. Are respiratory symptoms common or uncommon in egg allergy? Respiratory symptoms occur in a minority of cases, with approximately 12% of egg-allergic children having respiratory symptoms upon exposure to egg (Boyano Martinez et al., 2001). Symptoms of cough, hoarse voice or cry, wheeze and stridor indicate involvement of the respiratory tract, some of which have been reported in the present case, suggestive of laryngeal involvement. Fatality from egg anaphylaxis is very uncommon (Pumphrey, 2000; Allen et al., 2007; Pumphrey and Gowland, 2007). However, involvement of the respiratory tract indicates a potentially more severe reaction, and authorities and consensus guidelines have suggested that airway involvement is an indication for prescribing an adrenaline auto-injector in foodallergic children (Bock et al., 2001; Sicherer and Simons, 2007). The weight of the child was 14 kg which fits into guidelines for provision of self-injectable adrenaline which suggest that children of >10 kg weight can be prescribed an adrenaline auto-injector with a fixed dose of 0.15 mg of adrenaline (Sicherer and Simons, 2007). According to Table 2.1, this child’s SPT to egg white with an 8 mm response would be highly suggestive of a positive clinical reactivity to egg, with a positive predictive value >95% (Sporik et al., 2000; Sampson, 2001). However, quantitative SPT and specific IgE levels are not particularly predictive of the severity of a clinical reaction. In the described case, the doctor asked particularly whether or not to prescribe an adrenaline self-injector, and this question is more related to the severity than to the presence of any reaction to egg consumption or exposure. Some additional factors may help in that decision (Kemp, 2003). Aggravating factors are shown in Table 2.3, these being: age over 5, since most deaths have been observed in school-age children or teenagers; a history of asthma requiring controlling medication (inhaled steroids); peanut or tree nut sensitivity; reactions induced by traces or small amounts of allergen; and a strongly positive SPT. The presence of asthma in general is considered to increase the risk of death (Bock et al., 2001; Macdougall et al., 2002) or
36
Management of Food Allergens
a severe reaction (Sicherer et al., 2001). This conclusion is derived predominantly from older individuals with peanut and tree nut allergy. It remains to be clarified whether this also applies for younger children and infants with egg allergy. In the present case, asthma was not reported, but at that age involvement of the respiratory tract might represent the first manifestation of wheezy bronchitis, which might later be assigned as bronchial asthma. Thus respiratory symptoms as reported with cough would support the prescription of an adrenaline self-injector. Consequently, the answer to the doctors’ question would be yes. Importantly, prescription is not enough to prevent a severe reaction upon unexpected egg exposure. Without parental and individual teaching how to use an auto-injector, the likelihood to be applied correctly is very low. Some surveys on that issue highlight that only around 30% of adrenaline auto-injectors were used appropriately in a subsequent anaphylactic reaction (Gold and Sainsbury, 2000; Mehl et al., 2005). Instructions on auto-injector administration and provision of a clear and simple anaphylaxis action plan are just as important as the provision of the devise itself (Hu et al., 2007).
2.7 CASE 6 – IMMUNOTHERAPY FOR ORAL ALLERGY SYNDROME? A 35-year-old man has been suffering from tree pollen allergy for 3 years and approaches his doctor because of tingling in his mouth and throat while eating his favourite apple variety. His allergies during birch pollen season are well controlled. He regularly takes antihistamines and topical nasal steroids to treat his blocked nose. He has heard from friends and read in newspapers that immunotherapy might also help to relieve his symptoms associated with ingestion/contact with apples. Birch pollen sensitisation has been proven by SPT (wheal of 13 mm) and specific IgE-antibodies of 25 kU/L. Prick-to-prick test with different apple varieties induced wheals in the range of 10 mm. The patient clearly asks for immunotherapy as treatment of his oral allergy syndrome to apple associated with his birch pollen allergy affecting the upper airways. The following questions shall be answered:
r r
Shall an immunotherapy be recommended? Is the sublingual route as effective as the subcutaneous route?
Allergen-specific immunotherapy in birch pollen allergies is recommended as a first-line treatment. Short-term and long-term efficacy, as well as safety data, strongly support such a treatment in birch pollen-allergic individuals (Bousquet et al., 1998). In addition, successful allergen-specific immunotherapy (in particular the subcutaneous route) has been associated with an increase in allergen-specific IgG4-antibodies, reduced proliferative and cytokine responses in allergen-stimulated peripheral blood cells in vitro and a shift from a diseaseeliciting TH2 cells towards a more non-pathogenic TH0/1-like phenotype (Secrist et al., 1993; Jutel et al., 1995; Ebner et al., 1997). The balance between allergen-specific IL-10producing Treg and TH2 cells has also been proposed to be pivotal for a physiological immune response to allergens (Akdis et al., 2004). As an alternative of the subcutaneous route, the sublingual route is regarded as a validated option, particularly in adults (Wilson et al., 2005). Thus, an allergen-specific immunotherapy in either form (subcutaneous or sublingual) is
Clinical incidence of food allergy
37
recommended. However, the patient asked for the allergen-specific immunotherapy in order to control his oral allergy syndrome caused by apples. Recent data support the view that current strategies applying allergen-specific immunotherapy in birch pollen allergies have limited effects on concomitant food allergy to apple (Hansen et al., 2004; Mari et al., 2005; Kinaciyan et al., 2007), although beneficial effects, including partial long-lasting effects, have been reported in a non-randomised study (Asero, 1998, 2003). Another randomised study revealed significant statistically calculated improvements, but the treatment did not result in symptom-free consumption of apples (Bucher et al., 2004). In addition to missing clinical efficacy, immunological changes observed in Bet v1-specific responses in vitro did not affect the Bet v1-homolog in apple Mal d1 (Bohle et al., 2007; Kinaciyan et al., 2007). One study has, however, described the successful benefits of tree pollen immunotherapy on fennel, cucumber and melon allergy (Asero, 2000). Thus, despite improvements in the understanding of how pollen-associated food allergies develop, manifest and affect patients, treatment with a tree pollen targeted immunotherapy is currently not recommended for tree pollen-allergic individuals.
2.8 CONCLUSION The evaluation of food allergy should follow an orderly approach. The history may not allow the physician to accurately make the diagnosis, but it can be used to formulate a plan that will include skin testing, determination of food-specific serum IgE, diagnostic laboratory (or endoscopic studies), the use of limited or oligoallergenic food elimination diets and ultimately the decision to perform food challenges. The rigour of challenge is dictated by the nature of the problem and the setting in which the challenges are to be performed, recognising that families can be greatly aided without having to involve a tertiary medical centre. It has become one of the more straightforward areas of allergy diagnosis and treatment provided that the steps outlined and discussed above are followed.
REFERENCES Aickin R., Hill D. and Kemp A. (1994) Measles immunisation in children with allergy to egg. BMJ, 309, 223–225. Akdis M., Verhagen J., Taylor A. et al. (2004) Immune responses in healthy and allergic individuals are characterized by a fine balance between allergen-specific T regulatory 1 and T helper 2 cells. Journal of Experimental Medicine, 199, 1–10. Allen C.W., Campbell D.E. and Kemp A.S. (2007) Egg allergy: are all childhood food allergies the same? Journal of Paediatrics and Child Health, 43, 214–218. Asero R. (1998) Effects of birch pollen-specific immunotherapy on apple allergy in birch pollenhypersensitive patients. Clinical and Experimental Allergy, 28, 1368–1373. Asero R. (2000) Fennel, cucumber, and melon allergy successfully treated with pollen-specific injection immunotherapy. Annals of Allergy, Asthma, and Immunology, 84, 460–462. Asero R. (2003) How long does the effect of birch pollen injection SIT on apple allergy last? Allergy, 58, 435–438. Asero R. (2007) Chronic unremitting urticaria: is the use of antihistamines above the licensed dose effective? A preliminary study of cetirizine at licensed and above-licensed doses. Clinical and Experimental Dermatology, 32, 34–38. Bernard H., Paty E., Mondoulet L. et al. (2003) Serological characteristics of peanut allergy in children. Allergy, 58, 1285–1292.
38
Management of Food Allergens
Besler M., Steinhart H. and Paschke A. (2001) Stability of food allergens and allergenicity of processed foods. Journal of Chromatography. B, Biomedical Sciences and Applications, 756, 207–228. Bierman C.W., Shapiro G.G., Pierson W.E. et al. (1977) Safety of influenza vaccination in allergic children. Journal of Infectious Diseases, 136(Suppl), S652–S655. Bindslev-Jensen C. and Poulsen L.K. (1998) Accuracy of in vivo and in vitro tests. Allergy, 53, 72–74. Bock S.A., Munoz-Furlong A and Sampson H.A. (2001) Fatalities due to anaphylactic reactions to foods. Journal of Allergy and Clinical Immunology, 107, 191–193. Bock S.A., Munoz-Furlong A. and Sampson H.A. (2007) Further fatalities caused by anaphylactic reactions to food, 2001–2006. Journal of Allergy and Clinical Immunology, 119, 1016–1018. Bohle B., Kinaciyan T., Gerstmayr M. et al. (2007) Sublingual immunotherapy induces IL-10-producing T regulatory cells, allergen-specific T-cell tolerance, and immune deviation. Journal of Allergy and Clinical Immunology, 120, 707–713. Bonnelykke K., Pipper C.B. and Bisgaard H. (2008) Sensitization does not develop in utero. Journal of Allergy and Clinical Immunology, 121, 646–651. Bousquet J., Lockey R. and Malling H.J. (1998) Allergen immunotherapy: therapeutic vaccines for allergic diseases. A WHO position paper. Journal of Allergy and Clinical Immunology, 102, 558–562. Boyano Martinez T., Garcia-Ara C., Diaz-Pena J.M. et al. (2001) Validity of specific IgE antibodies in children with egg allergy. Clinical and Experimental Allergy, 31, 1464–1469. Braganza S.C., Acworth J.P., McKinnon D.R. et al. (2006) Paediatric emergency department anaphylaxis: different patterns from adults. Archives of Disease in Childhood, 91, 159–163. Bucher X., Pichler W.J., Dahinden C.A. and Helbling A. (2004) Effect of tree pollen specific, subcutaneous immunotherapy on the oral allergy syndrome to apple and hazelnut. Allergy, 59, 1272–1276. Burks W., Bannon G.A., Sicherer S. and Sampson H.A. (1999) Peanut-induced anaphylactic reactions. International Archives of Allergy and Immunology, 119, 165–172. Clark A.T. and Ewan P.W. (2003) Interpretation of tests for nut allergy in one thousand patients, in relation to allergy or tolerance. Clinical and Experimental Allergy, 33, 1041–1045. David T.J., Waddington E. and Stanton R.H. (1984) Nutritional hazards of elimination diets in children with atopic eczema. Archives of Disease in Childhood, 59, 323–325. De Boissieu D. and Dupont C. (2006) Natural course of sensitization to hen’s egg in children not previously exposed to egg ingestion. European Annals of Allergy and Clinical Immunology, 38, 113–117. Dean T., Venter C., Pereira B. et al. (2007) Patterns of sensitization to food and aeroallergens in the first 3 years of life. Journal of Allergy and Clinical Immunology, 120, 1166–1171. Ebner C., Siemann U., Bohle B. et al. (1997) Immunological changes during specific immunotherapy of grass pollen allergy: reduced lymphoproliferative responses to allergen and shift from TH2 to TH1 in T-cell clones specific for Phl p 1, a major grass pollen allergen. Clinical and Experimental Allergy, 27, 1007–1015. Eggesbo M., Botten G., Halvorsen R. and Magnus P. (2001) The prevalence of allergy to egg: a populationbased study in young children. Allergy, 56, 403–411. Eigenmann P.A., Burks A.W., Bannon G.A. and Sampson H.A. (1996) Identification of unique peanut and soy allergens in sera adsorbed with cross-reacting antibodies. Journal of Allergy and Clinical Immunology, 98, 969–978. Eigenmann P.A. and Sampson H.A. (1998) Interpreting skin prick tests in the evaluation of food allergy in children. Pediatric Allergy and Immunology, 9, 186–191. Falth-Magnusson K. and Kjellman N.I. (1992) Allergy prevention by maternal elimination diet during late pregnancy – a 5-year follow-up of a randomized study. Journal of Allergy and Clinical Immunology, 89, 709–713. Foucard T. and Malmheden Yman I. (1999) A study on severe food reactions in Sweden – is soy protein an underestimated cause of food anaphylaxis? Allergy, 54, 261–265. Garcia-Ara C., Boyano-Martinez T., Diaz-Pena J.M. et al. (2001) Specific IgE levels in the diagnosis of immediate hypersensitivity to cows’ milk protein in the infant. Journal of Allergy and Clinical Immunology, 107, 185–190. Gold M.S. and Sainsbury R. (2000) First aid anaphylaxis management in children who were prescribed an epinephrine autoinjector device (EpiPen). Journal of Allergy and Clinical Immunology, 106, 171– 176.
Clinical incidence of food allergy
39
Hansen K.S., Khinchi M.S., Skov P.S. et al. (2004) Food allergy to apple and specific immunotherapy with birch pollen. Molecular Nutrition and Food Research, 48, 441–448. Hattevig G., Kjellman B., Johansson S.G. and Bjorksten B. (1984) Clinical symptoms and IgE responses to common food proteins in atopic and healthy children. Clinical Allergy, 14, 551–559. Holloway J.A., Warner J.O., Vance G.H. et al. (2000) Detection of house-dust-mite allergen in amniotic fluid and umbilical-cord blood. Lancet, 356, 1900–1902. Hu W., Grbich C. and Kemp A. (2007) Parental food allergy information needs: a qualitative study. Archives of Disease in Childhood, 92, 771–775. James J.M., Zeiger R.S., Lester M.R. et al. (1998) Safe administration of influenza vaccine to patients with egg allergy. Journal of Pediatrics, 133, 624–628. Johansson S.G., Bieber T., Dahl R. et al. (2004) Revised nomenclature for allergy for global use: Report of the Nomenclature Review Committee of the World Allergy Organization, October 2003. Journal of Allergy and Clinical Immunology, 113, 832–836. Jutel M., Pichler W.J., Skrbic D. et al. (1995) Bee venom immunotherapy results in decrease of IL-4 and IL-5 and increase of IFN-gamma secretion in specific allergen-stimulated T cell cultures. Journal of Immunology, 154, 4187–4194. Kemp A.S. (2003) EpiPen epidemic: suggestions for rational prescribing in childhood food allergy. Journal of Paediatrics and Child Health, 39, 372–375. Kemp A.S. (2007) Egg allergy. Pediatric Allergy and Immunology, 18, 696–702. Kinaciyan T., Jahn-Schmid B., Radakovics A. et al. (2007) Successful sublingual immunotherapy with birch pollen has limited effects on concomitant food allergy to apple and the immune response to the Bet v 1 homolog Mal d 1. Journal of Allergy and Clinical Immunology, 119, 937–943. Kleine-Tebbe J., Vogel L., Crowell D.N. et al. (2002) Severe oral allergy syndrome and anaphylactic reactions caused by a Bet v 1- related PR-10 protein in soybean, SAM22. Journal of Allergy and Clinical Immunology, 110, 797–804. Lack G. (2008) Clinical practice. Food allergy. New England Journal of Medicine, 359, 1252–1260. Liu T., Howard R.M., Mancini A.J. et al. (2001) Kwashiorkor in the United States: fad diets, perceived and true milk allergy, and nutritional ignorance. Archives of Dermatology, 137, 630–636. Lloyd-Still J.D. (1979) Chronic diarrhea of childhood and the misuse of elimination diets. Journal of Pediatrics, 95, 10–13. Loveless M. (1950) Milk allergy: a survey of its incidence. Journal of Allergy, 21, 489–499. Macdougall C.F., Cant A.J. and Colver A.F. (2002) How dangerous is food allergy in childhood? The incidence of severe and fatal allergic reactions across the UK and Ireland. Archives of Disease in Childhood, 86, 236–239. Mari A., Ballmer-Weber B.K. and Vieths S. (2005) The oral allergy syndrome: improved diagnostic and treatment methods. Current Opinion in Allergy and Clinical Immunology, 5, 267–273. May C.D. (1976) Objective clinical and laboratory studies of immediate hypersensitivity reactions to food in asthmatic children. Journal of Allergy and Clinical Immunology, 58, 500–515. Mehl A., Wahn U. and Niggemann B. (2005) Anaphylactic reactions in children – a questionnaire-based survey in Germany. Allergy, 60, 1440–1445. Monti G., Muratore M.C., Peltran A. et al. (2002) High incidence of adverse reactions to egg challenge on first known exposure in young atopic dermatitis children: predictive value of skin prick test and radioallergosorbent test to egg proteins. Clinical and Experimental Allergy, 32, 1515– 1519. Ostblom E., Lilja G., Ahlstedt S. et al. (2008) Patterns of quantitative food-specific IgE-antibodies and reported food hypersensitivity in 4-year-old children. Allergy, 63, 418–424. Osterballe M., Hansen T.K., Mortz C.G. et al. (2005) The prevalence of food hypersensitivity in an unselected population of children and adults. Pediatric Allergy and Immunology, 16, 567–573. Palmer D.J., Gold M.S. and Makrides M. (2005) Effect of cooked and raw egg consumption on ovalbumin content of human milk: a randomized, double-blind, cross-over trial. Clinical and Experimental Allergy, 35, 173–178. Pereira B., Venter C., Grundy J. et al. (2005) Prevalence of sensitization to food allergens, reported adverse reaction to foods, food avoidance, and food hypersensitivity among teenagers. Journal of Allergy and Clinical Immunology, 116, 884–892.
40
Management of Food Allergens
Perry T.T., Matsui E.C., Kay Conover-Walker M. and Wood R.A. (2004) The relationship of allergen-specific IgE levels and oral food challenge outcome. Journal of Allergy and Clinical Immunology, 114, 144–149. Pfefferle P.I., Sel S., Ege M.J. et al. (2008) Cord blood allergen-specific IgE is associated with reduced IFN-gamma production by cord blood cells: the Protection against Allergy-Study in Rural Environments (PASTURE) Study. Journal of Allergy and Clinical Immunology, 122, 711–716. Poulos L.M., Waters A.M., Correll P.K. et al. (2007) Trends in hospitalizations for anaphylaxis, angioedema, and urticaria in Australia, 1993–1994 to 2004–2005. Journal of Allergy and Clinical Immunology, 120, 878–884. Pumphrey R. (2000) Lessons for management of anaphylaxis from a study of fatal reactions. Clinical and Experimental Allergy, 30, 1144–1150. Pumphrey R. and Gowland M. (2007) Further fatal allergic reactions to food in the United Kingdom, 1999–2006. Journal of Allergy and Clinical Immunology, 119, 1018–1019. Rance F., Abbal M. and Lauwers-Cances V. (2002) Improved screening for peanut allergy by the combined use of skin prick tests and specific IgE assays. Journal of Allergy and Clinical Immunology, 109, 1027–1033. Retailliau H.F., Curtis A.C., Storr G. et al. (1980) Illness after influenza vaccination reported through a nationwide surveillance system, 1976–1977. American Journal of Epidemiology, 111, 270–278. Ring J., Brockow K. and Behrendt H. (2004) History and classification of anaphylaxis. Novartis Foundation Symposium, 257, 6–16. Roberts G. (2007) Anaphylaxis to foods. Pediatric Allergy and Immunology, 18, 543–548. Roberts G. and Lack G. (2005) Diagnosing peanut allergy with skin prick and specific IgE testing. Journal of Allergy and Clinical Immunology, 115, 1291–1296. Roehr C.C., Edenharter G., Reimann S. et al. (2004) Food allergy and non-allergic food hypersensitivity in children and adolescents. Clinical and Experimental Allergy, 34, 1534–1541. Sampson H.A. (2001) Utility of food-specific IgE concentrations in predicting symptomatic food allergy. Journal of Allergy and Clinical Immunology, 107, 891–896. Sampson H.A. and Ho D.G. (1997) Relationship between food-specific IgE concentrations and the risk of positive food challenges in children and adolescents. Journal of Allergy and Clinical Immunology, 100, 444–451. Sampson H.A., Munoz-Furlong A., Campbell R.L. et al. (2006) Second symposium on the definition and management of anaphylaxis: summary report – second National Institute of Allergy and Infectious Disease/Food Allergy and Anaphylaxis Network symposium. Annals of Emergency Medicine, 47, 373– 380. Secrist H., Chelen C.J., Wen Y. et al. (1993) Allergen immunotherapy decreases interleukin 4 production in CD4+ T cells from allergic individuals. Journal of Experimental Medicine, 178, 2123–2130. Sheikh A. and Alves B. (2000) Hospital admissions for acute anaphylaxis: time trend study. BMJ, 320, 1441. Sheikh A. and Walker S. (2005) Anaphylaxis. BMJ, 331, 330. Sicherer S.H. (1999) Food allergy: when and how to perform oral food challenges. Pediatric Allergy and Immunology, 10, 226–234. Sicherer S.H., Furlong T.J., Munoz-Furlong A. et al. (2001) A voluntary registry for peanut and tree nut allergy: characteristics of the first 5149 registrants. Journal of Allergy and Clinical Immunology, 108, 128–132. Sicherer S.H., Sampson H.A. and Burks A.W. (2000) Peanut and soy allergy: a clinical and therapeutic dilemma. Allergy, 55, 515–521. Sicherer S.H. and Simons F.E. (2007) Self-injectable epinephrine for first-aid management of anaphylaxis. Pediatrics, 119, 638–646. Sporik R., Hill D.J. and Hosking C.S. (2000) Specificity of allergen skin testing in predicting positive open food challenges to milk, egg and peanut in children. Clinical and Experimental Allergy, 30, 1540–1546. Steinke M., Fiocchi A., Kirchlechner V. et al. (2007) Perceived food allergy in children in 10 European nations. A randomised telephone survey. International Archives of Allergy and Immunology, 143, 290–295. Sugai K., Shiga A., Okada K. et al. (2007) Dermal testing of vaccines for children at high risk of allergies. Vaccine, 25, 3454–3463. Sz´epfalusi Z., Loibichler C., Pichler J. et al. (2000) Direct evidence for transplacental allergen transfer. Pediatric Research, 48, 404–407. Taramarcaz P., Hauser C. and Eigenmann P.A. (2001) Soy anaphylaxis. Allergy, 56, 792.
Clinical incidence of food allergy
41
Venter C., Pereira B., Grundy J. et al. (2006a) Prevalence of sensitization reported and objectively assessed food hypersensitivity amongst six-year-old children: a population-based study. Pediatric Allergy and Immunology, 17, 356–363. Venter C., Pereira B., Grundy J. et al. (2006b) Incidence of parentally reported and clinically diagnosed food hypersensitivity in the first year of life. Journal of Allergy and Clinical Immunology, 117, 1118–1124. Venter C., Pereira B., Voigt K. et al. (2008) Prevalence and cumulative incidence of food hypersensitivity in the first 3 years of life. Allergy, 63, 354–359. Wedi B. and Kapp A. (2007) Evidence-based therapy of chronic urticaria. Journal der Deutschen Dermatologischen Gesellschaft, 5, 146–157. Wilson D.R., Lima M.T. and Durham S.R. (2005) Sublingual immunotherapy for allergic rhinitis: systematic review and meta-analysis. Allergy, 60, 4–12. Yocum M.W., Butterfield J.H., Klein J.S. et al. (1999) Epidemiology of anaphylaxis in Olmsted County: a population-based study. Journal of Allergy and Clinical Immunology, 104, 452–456. Zuberbier T., Edenharter G., Worm M. et al. (2004) Prevalence of adverse reactions to food in Germany – a population study. Allergy, 59, 338–345.
3
Identification and characterisation of food allergens
E. N. Clare Mills, Philip Johnson, Yuri Alexeev, and Heimo Breiteneder
3.1 INTRODUCTION As part of its normal humoral immune responses, in a process generally known as sensitisation, the body produces various different types of molecules called immunoglobulins (Igs or antibodies) which are designated as IgA, IgG, IgM and IgE. These molecules comprise a binding domain which is capable of recognising, usually with high affinity and specificity, ‘non-self’ molecules which include those found in microbes, parasites, environmental agents such as pollen and dust, as well as dietary proteins. In one type of allergic disease known as type I hypersensitivity, the repertoire of antibodies is altered, the body producing larger quantities of IgE, the antibody class normally produced in response to parasitic infections. These IgE responses are directed towards a variety of environmental agents, the target molecules recognised by IgE being termed allergens. It is these molecules that are responsible for triggering an allergic reaction, a process also known as elicitation. IgE works by becoming associated with cells, such as mast cells, which are packed full of inflammatory mediators like histamine. On binding multivalent allergen, the surface bound IgE becomes ‘cross-linked’, triggering release of the mediators, such as histamine, which actually cause the symptoms associated with allergic reactions. Allergens are usually proteinaceous in nature, the sites recognised by the IgE being known as epitopes. These epitopes can comprise either linear sequences of amino acids (also termed continuous) or disparate portions of a protein’s amino acid sequence which are brought together by the folding of the polypeptide chain to form a discontinuous, or conformational, epitope. It is thought that the majority of epitopes are conformational in nature (Van Regenmortel, 1992) and consequently can be difficult to define in relation to food allergens where processing may either disrupt conformational epitopes found in the native, unprocessed protein, or even introduce new epitopes. In the absence of any treatment, the only option for food allergic individuals is to avoid the food they are allergic to, and, if appropriate, rescue medication is given in case of an accidental consumption of a problem food. Such food avoidance strategies can be difficult to implement. As a consequence, legislation has been brought in around the world which makes it mandatory to label certain foods and derived ingredients, irrespective of the level to which they are added to a foodstuff (Mills et al., 2004c). The list of ingredients which must be labelled does vary across the world and in the European Union it has already been updated to include molluscs and lupin. Table 3.1 shows the major foods that have to be labelled under this legislation, together with the allergens that have been isolated and characterised from them. In addition to these foods, there are many fresh fruits and vegetables that trigger food allergies and are associated with allergies to other substance such as pollen.
Peanut (Arachis hypogaea)
P34
Vacuolar thiol protease
Protein family
2S albumin Conglutin, 2S albumin n Arachin, 7S seed storage globulin Actin-binding proteins
Ara h 2
Ara h 6
Ara h 3,4
Ara h 5
Gly m Bd 28 k
None Conarachin, vicillin, 7S seed storage globulin
Gly m Bd 30 k
None
Ara h 1
11S seed storage globulin, Glycine Glycinin, individual gene products known as G1, G2, G3, G4 and G5.
None
Profilin
Cupin
Prolamin; 2S albumin
Prolamin; 2S albumin
Cupin
C-protease (unconfirmed)
C-protease
Cupin
Profilin Bet v 1
SAM22, PR-10 protein
Gly m 3
Gly m 4
Not known
Gly m 1
Soybean (Glycine max )
Other names
Gly m 2
IUIS allergen name
AF059616
(Continued )
Kleber-Janke et al . (1999)
Restani et al . (2005)
Koppelman et al . (2005) AAL37561, 1W2Q A, AAD56337 AAC63045, AAD47382, AAM46958, AAM93157, ABI17154
Clarke et al . (1998)
Koppelman et al . (2004)
Ogawa et al . (1991), Hiemori et al . (2004)
Bando et al . (1996)
Nielsen et al . (1989)
Crowell et al . (1992)
Rihs et al . (1999)
Codina et al . (2002)
Gonzalez et al . (1995)
Reference
AAM78596, AAN77576, CAC41202
P43237, AAT00596, AAT00595, P43238
BAB21619, P22895, AAB09252
P22895, AAB09252, BAA25899
CAA26723, CAA 332154895 (G1), CAA33216 (G2), CAA33217 (G3), CAA37044, CAA26478, CAA60533 (G4), AA33964, AAA33965, CAA55977 (G5)
CAA42646
O65809, CAA11755
A57106 (partial sequence only)
AAB34755, ABA7, ABA54898
Sequence accessions
The major foods that have to be labelled under EU labelling legislation, together with the allergens that have been isolated and characterised from them.
Food
Table 3.1
Identification and characterisation of food allergens 43
Cashew nut (Anacardium occidentale)
Almond (Prunus dulcis)
Tree nuts
Legumin-like, 11S seed storage globulin 2S albumin
Ana o 3
60S acidic ribosomal protein
Pru du 5
Ana o 2
Major almond protein, Amandin, 11S seed storage globulin
None
Vicilin-like, 7S seed storage globulin
Prunus Seed allergenic protein 2, Conglutin-γ
None
Ana o 1
Prolamin; 2S albumin
2S, Prunus Seed allergenic protein 1
None
Prolamin
Cupin
Cupin
Ribosomal 60S superfamily
Cupin
Prolamin; 2S albumin
Profilin
Cupin
Pru du 4
Congluten-β
Oleosin-like
Lup an 1
Bet v 1 Oleosin
Protein family
None
Other names
Ara h 8
IUIS allergen name
(Continued )
Lupin (Lupinus luterus)
Food
Table 3.1
AAL91665
AAN76862
AAM73729, AAM73730
DQ836316
S51942, CAA55009
Robotham et al. (2005)
Wang et al. (2003)
Wang et al. (2002)
Abolhassani and Roux (2007)
Garcia-Mas et al. (1995), Sathe et al. (2002)
Poltronieri et al. (2002)
Poltronieri et al. (2002)
P82944 1, P82944 2 (partial sequences only) P82952 (partial sequences only)
Scheurer et al. (2001)
Peeters et al. (2007)
Pons et al. (2002)
Mittag et al . (2004b)
Reference
AAL91662, AAD29411, CAD37201
EU352876
Q6J1J8
AAQ91847
Sequence accessions
44 Management of Food Allergens
Soybean (Glycine max )
Hazelnut (Corylus avellana)
C-protease (unconfirmed)
Cysteine protease
Gly m Bd 30k, P34 Gly m Bd 28 k
Cupin
11S seed storage globulin, Glycine Glycinin, individual gene products known as G1, G2, G3, G4 and G5.
None
None
Profilin BetV1 family
SAM22, PR-10 protein
Not known
Gly m 4
Defensin
Gly m 2
Vacuolar thiol protease
Cupin
Gly m 3
P34, Hydrophobic protein
Vicilin-like, 7S seed storage globulin
Cor a 11
Gly m 1
Legumin-like, 11S seed storage globulin
Cor a 9
Cupin
Profilin Prolamin; LTP family
Non-specific lipid transfer protein (nsLTP)
Cor a 2
Cor a 8
Bet v 1
Cor a 1
BAB21619, P22895, AAB09252
P22895, AAB09252, BAA25899
CAA26723, CAA33215 (G1); CAA33216 (G2); CAA33217 (G3); CAA37044, CAA26478, CAA60533 (G4); AAA33964, AAA33965, CAA55977 (G5).
CAA42646
O65809, CAA11755
A57106 (partial sequence only)
AAB34755, ABA54897, ABA54898
AAL86739
AAL73404
AAK28533
Q9AXH5
CAA50325, CAA50326, CAA50328, Q08407, CAA96548, CAA96549, AAD48405, AAG40329, AAG40330, AAG40331
(Continued )
Ogawa et al. (1991), Hiemori et al. (2004)
Bando et al. (1996)
Nielsen et al. (1989)
Crowell et al. (1992)
Rihs et al., 1999
Codina et al. (2002)
Gonzalez et al. (1995)
Lauer et al. (2004)
Beyer et al. (2002a)
Schocker et al. (2004)
Flinterman et al. (2008a)
L¨uttkopf et al. (2002)
Identification and characterisation of food allergens 45
Almond (Prunus dulcis)
Tree nuts
Lupin (Lupinus luterus)
Prolamin
2S albumin, Prunus Seed allergenic protein 1 2S albumin, Prunus Seed allergenic protein 2, Conglutin gamma Amandin, Pru du amandin, 11S globulin 60S acidic ribosomal prot. P2
None
None
None
Pru du 5
Ribosomal 60S superfamily
Cupin
Prolamin
Profilin
Cupin
Oleosin-like
Bet v 1 family
Profilin
Cupin
Prolamin
Prolamin
Cupin
Protein family
Pru du 4
Conglutin beta
Oleosin
None
Lup an 1
Actin-binding protein
Arachin, 7S seed storage globulin
Ara h 3,4
PR-10 protein
Conglutin, 2S albumin
Ara h 6
Ara h 5
2S albumin
Ara h 2
Ara h 8
Conarachin, vicillin, 7S seed storage globulin
Ara h 1
Peanut (Arachis hypogaea)
Other names
IUIS allergen name
(Continued )
Food
Table 3.1
DQ836316
S51942, CAA55009
Abolhassani and Roux (2007)
Garcia-Mas et al. (1995), Sathe et al. (2002)
Poltronieri et al. (2002)
Poltronieri et al. (2002) P82944 1, P82944 2 (partial sequences only) P82952 (partial sequences only)
Scheurer et al. (2001), Tawde et al. (2006)
Peeters et al. (2007)
Pons et al. (2002)
Mittag et al. (2004aa, 2004bb)
Kleber-Janke et al. (1999)
Restani et al. (2005)
Koppelman et al. (2005)
Clarke et al. (1998)
Koppelman et al. (2004)
Reference
AAL91662, AAD29411, CAD37201
EU352876
Q6J1J8
AAQ91847
AF059616
AAC63045, AAD47382, AAM46958, AAM93157, ABI17154
AAL37561, 1W2Q A, AAD56337
AAM78596, AAN77576, CAC41202
P43237, AAT00596, AAT00595, P43238
Sequence accessions
46 Management of Food Allergens
2S albumin
Ana o 3
Oleosin (14–16 kDa)
Cor a 13
2S albumin
Oleosin (17 kDa)
Cor a 12
Jug r 1
Vicilin-like, 7S seed storage globulin
Cor a 11
Walnut (Juglans regia)
Legumin-like, 11S seed storage globulin
Cor a 9
No allergens characterised, clinically relevant type-1 hypersensitivity observed
Prolamin
Non-specific lipid transfer protein (nsLTP)
Cor a 8
Prolamin
Oleosin-like
Oleosin-like
Cupin
Cupin
Profilin
Bet v 1 family
Prolamin
Cupin
Cupin
Cor a 2
PR-10 protein
Legumin-like, 11S seed storage globulin
Ana o 2
Cor a 1
Vicilin-like, 7S seed storage globulin
Ana o 1
Macadamia nut (Macadamia integrifolia, Macadamia tetraphylla)
Hazelnut (Corylus avellana)
Cashew nut (Anacardium occidentale)
AAB41308
AY224599
AY224679
AAL86739
AAL73404
AAK28533
Q9AXH5
CAA50325, CAA50326, CAA50328, Q08407, CAA96548, CAA96549, AAD48405, AAG40329, AAG40330, AAG40331
AAL91665
AAN76862
AAM73729, AAM73730
(Continued )
Teuber et al. (1998)
Pallares (2000)
Akkerdaas et al. (2006)
Akkerdaas et al. (2006)
Lauer et al. (2004)
Beyer et al. (2002b)
Schocker et al. (2004)
Flinterman et al. (2008b)
L¨uttkopf et al. (2002)
Robotham et al. (2005)
Wang et al. (2003)
Wang et al. (2002)
Identification and characterisation of food allergens 47
Wheat (Triticum aestivum)
Celery (Apium graveolens)
Sesame (Sesamum indicum)
Black Mustard (Brassica nigra)
Unknown
Api g 5
Prolamin
Profilin
Api g 4
Alpha gliadins
Bet v 1 family
Api g 1
FAD binding oxidase homologue
Beta-globulin Cupin
7S globulin
Prolamin
Glucose and ribitol dehydrogenases
Prolamin
Ses i 2
2S albumin
Seed maturation protein
2S albumin
Cupin
Prolamin
Cupin
Cupin
Protein family
Ses i 3
Ses i 1
Bra j 1
11S-type or legumin-like globulin
Sin a 2
Legumin-like, 11S seed storage globulin
Jug r 4
2S albumin
Vicilin-like, 7S seed storage globulin
Jug r 2
Sin a 1
Other names
IUIS allergen name
(Continued )
White mustard (Sinapis alba)
Mustard
Food
Table 3.1
Q41545, Q41509, P04721, P04723, P04725, P18573, Q9M4M1, Q9M4L7, Q9M4M6
P81943
Q9XF37
P49372, P92918
Q9AUD0
Q9XHP1
AF240005, AF091841
AF449242
P80207
AAX77383, AAX77384
CAA62909, CAA62910, CAA62911, CAA62912, CAA62908, P15322
AAW29810
AAF18269
Sequence accessions
Sandiford et al. (1997)
Ganglberger et al. (2000)
Scheurer et al. (2000), Scheurer et al. (2001)
Breiteneder et al. (1995)
Beyer et al. (2002b)
Tai et al. (2001)
Tai et al. (1999), Pastorello et al. (2001b)
Beyer et al. (2002a)
Palomares et al. (2005)
Palomares et al. (2007)
Monsalve et al. (1993)
Wallowitz et al. (2006)
Teuber et al. (1999)
Reference
48 Management of Food Allergens
Cows milk (Bos Taurus)
Prolamin Serine protease inhibitor family
Omega-5-gliadin Serine protease inhibitor
Tri a 19
Beta-Lactoglobulin Immunoglobulin
Bovine Serum Albumin Lactoferrin
Bos d 5
Bos d 7
Bos d 6
None
Beta-Casein Alpha-Lactalbumin
Bos d 4
Kappa-Casein
Bos d 8
Casein
Alpha S2-Casein
Bos d 8
Bos d 8
Transferrin
Serum albumin family
Immunoglobulin family
lipocalin
lysozyme/alpha-lactalbumin
Casein
Casein
Casein
Alpha S1-Casein
Bos d 8
Thoredoxin-fold family
Thioredoxin glutenin
Tri a 25
Tri a 26
Prolamin
Lipid Transfer Protein (LTP1)
Tri a 14
Profilin
Profilin-1 Isolectin A, WGA1
Tri a 12
Trypsin/alpha-amylase inhibitor
CM3
Tri a 18
Trypsin/alpha-amylase inhibitor
Alpha-amylase inhibitor 0.53
P24627
P02769
(Continued )
Pierce et al. (1991)
Restani et al. (2004)
Kontopidis et al. (2004) Bernhisel-Broadbent et al. (1991)
P02754
Hurley and Schuler (1987), Neyestani et al. (2003).
Jimenez-Flores et al. (1987)
Chatchatee et al. (2001)
Busse et al. (2002)
Mercier et al. (1971), Bernard et al. (1998)
Battais et al. (2003)
Weichel et al. (2006)
Constantin et al. (2008)
DuPont et al. (2000), Lehto et al. (20030
Asero et al. (2000)
Sutton et al. (1984)
Rihs et al. (1994)
Salcedo et al. (2004)
Carbonero and Garcia-Olmedo (1999)
XP 593266, AAB37381.2, AAB37380.1, XP 587538.1, AAC18409.1
P00711
P02666
P02668
P02663
P02662 (minor variants)
X12928
AJ404845
EU051824
Q40215
P24296
P49232
P17314
P01084
Identification and characterisation of food allergens 49
Lysozyme α-livetin
Gal d 4
Gal d 5
Calcium-binding EF-hand
Calcium-binding EF-hand
Calcium-binding EF-hand Calcium-binding EF-hand Collagen
Cyp c 1, β-parvalbumin
Sco j 1, β-parvalbumin
β-parvalbumin Thu o 1, β-parvalbumin Gelatine (denatured collagen aggregate)
Carp (Cyprinus carpi)
Mackerel (Scomber japonicus)
Salmon (Salmo salar )
Tuna (Thunnus tonngol )
Sal s 1
Calcium-binding EF-hand
β-parvalbumin
Cod (Gadus morhua)
Calcium-binding EF-hand
Transferrin
Serpin
The c 1, β-parvalbumin
Gad c 1
Serum albumin family
Ovotransferrin
Gal d 3
Fish Alaska pollack (Theragra chalcogramma)
C-type lysozyme
Ovalbumin
Gal d 2
Kazal-type srine protease inhibitor
Ovomucoid
Gal d 1
Protein family
Hen’s egg
Other names
IUIS allergen name
(Continued )
Food
Table 3.1
U23822 (red seabream), O93484 (rainbow trout)
n.d.
X97824
P59747
CAC83658
P02622
Q90YK7
P19121
P00698
P02789
P01012
P01005
Sequence accessions
Hansen et al. (2004)
Bugajska-Schretter et al. (1998)
Lindstrøm et al. (1996)
Hamada et al. (2003)
Bugajska-Schretter et al. (2000)
Bugajska-Schretter et al. (1998)
Van Do et al. (2005)
de Blay et al. (1994)
Mohan et al. (2003)
Awade et al. (1994)
Honma et al. (1996)
Besler et al. (1997)
Reference
50 Management of Food Allergens
Cha f 1
Crab
He as 1
Hal m 1
Cra g 1
Snail (Helix aspersa)
Abalone (Haliotis midae)
Oyster (Crassostrea gigas)
n.d., not determined.
Tod p 1
Squid (Todarodes Pacificus)
Molluscs
Pan s 1
Spiny lobster (Panulirus stimpsoni)
Tropomyosin
Unknown
Tropomyosin
Tropomyosin
Tropomyosin
Tropomyosin
Tropomyosin
Arginine kinase
Pen m 2
Hom a 1
Tropomyosin
Pen m 1, Pen a 1, Met e1
American lobster (Homarus americanus)
Shrimp
Crustacea
Calcium-binding EF-hand
Unknown
Calcium-binding EF-hand
Calcium-binding EF-hand
Calcium-binding EF-hand
Calcium-binding EF-hand
Calcium-binding EF-hand
Phosphagen kinase
Calcium-binding EF-hand
Q95WYO
n.d.
Y14855
n.d.
Q9N2R3
Q61379
O44119
Q819P7
AAZ76743, Q25456
Ishikawa et al. (1998)
Lopata et al. (1997)
Asturias et al. (2002)
Miyazawa et al. (1996)
Leung et al. (1998b)
Leung et al. (1998a)
Leung et al. (1998a)
France et al. (1997), Binder et al. (2001)
Daul et al. (1994)
Identification and characterisation of food allergens 51
52
Management of Food Allergens
The World Health Organization and the International Union of Immunological Societies (IUIS) produce an official list of allergens, which is designated by the Allergen Nomenclature Sub-Committee (Hoffman et al., 1994). Allergens included in this listing must induce IgEmediated (atopic) allergy in humans with a prevalence of IgE reactivity above 5%. An allergen is termed major if it is recognised by IgE from at least 50% of a cohort of allergic individuals but does not carry any connotation of allergenic potency; allergens are otherwise termed ‘minor’. The allergen designation is then based on the Latin name of the species from which it originates and is composed of the first three letters of the genus, followed by the first letter of the species finishing with an Arabic number, for example Ara h 1 relates to an allergen from Arachis hypogea (peanuts). The numbers are determined by the order in which allergens are identified and are common to all homologous allergens (also known as isoallergens) in a given species. Isoallergens are defined on the basis of having a similar molecular mass, an identical biological function, if known, for example enzymatic action and >67% identity of amino acid sequences. For those species where the first three letters of a genus and the first letter of a species are identical, the second letter of the species is also used. The IUIS designations for those allergens in foods for which labelling is mandatory are also given in Table 3.1.
3.2 CLASSIFICATION OF FOOD ALLERGENS It is generally held that the vast majority of food allergies are caused by a limited number of foods (Bush and Hefle, 1996), but a large number of foods have been documented as causing food allergies, reflecting the diversity of species that humans consume. Over the past 20 years, there has been an explosion in the number of allergens that have been identified and characterised and subsequently there have been efforts to classify them in order to identify common properties and motifs which may be predictive of allergenicity. Observations have been made that food allergens are restricted to certain protein families (Mills et al., 2004a), and subsequently a classification of plant food allergens based on the membership of allergens to certain protein families and superfamilies has been proposed (Breiteneder and Radauer, 2004). Using post-genomic bioinformatic tools such as Pfam (Bateman et al., 2004), an analysis of plant food allergen families showed that they belonged to only 27 of the then identified 8183 Pfam families, indicating that conserved structures and biological activities play a role in determining or promoting allergenic properties of proteins (Jenkins et al., 2005). Three plant food allergen protein families/superfamilies were found to predominate: the prolamin superfamily, the cupin superfamily and the Bet v 1 family, which together with the profilins accounted for more than 65% of all plant food allergens. A similar situation was found for pollen allergens which were classified into 29 Pfam families representing a 0.35% section of today’s classified protein universe (Radauer and Breiteneder, 2006). A similar distribution was found for food allergens of animal origin (Jenkins et al., 2007) with three protein families: the tropomyosins, parvalbumins and caseins dominating. Thus, the repertoire of allergenic proteins identified is small compared to the vast array of different proteins found in biology. The explanations for this are lacking but may in part result from conservation of surface structures in certain families, such as the Bet v 1 and parvalbumin superfamilies, which promotes IgE cross-reactivity (Jenkins et al., 2005, 2007).
Identification and characterisation of food allergens
53
3.3 PLANT FOOD ALLERGENS 3.3.1
Fresh fruits and vegetables
Allergy to fresh fruits and vegetables is frequently associated with inhalant allergies to agents such as birch and grass pollen and latex. It is thought that individuals initially become sensitised to the inhalant allergens in pollen and latex and subsequently go on to develop allergies to foods because of IgE cross-reactivity to closely related homologues of the pollen and latex allergens. Symptoms are often mild and confined to the oral cavity, which has given rise to the term oral allergy syndrome (OAS) and frequently (although not always) processing removes allergenicity. Thus, many individuals with the fruit/vegetable-pollen or fruit/vegetable-latex allergies can safely consume cooked fruits and vegetables but not fresh produce. The pollen-related fruit and vegetable allergies tend to have a geographic distribution related to pollen distribution. Allergens involved include those homologous to the major birch pollen allergen, Bet v 1, whose role in plants has yet to be defined, although it does belongs to family 10 of the pathogenesis-related proteins (Breiteneder et al., 1989; Hoffmann-Sommergruber, 2002). As it can bind plant steroids in a central tunnel, one suggestion is that it functions as a steroid carrier in plants (Markovic-Housley et al., 2003) (see Plate 3.1). Homologues involved in pollen-fruit cross-reactive allergies have been identified in a very large number of fruits and vegetables, some of the most important include the Rosaceae fruits such as apple (Mal d 1; Vanek-Krebitz et al., 1995), cherry (Pru av 1; Neudecker et al., 2001) and peach (Pru p 1; Gaier et al., 2008). Homologues have also been identified in fruits such as kiwi fruit which are emerging as important allergenic foods in Europe (Act d 8; Oberhuber et al., 2008) and exotic fruits not generally consumed in Europe which may pose a risk such as Sharon fruit (Bolhaar et al., 2005) and jackfruit (Bolhaar et al., 2004). In addition, allergenic Bet v 1 homologues have also been identified in vegetables notably celery (Api g 1; Breiteneder et al., 1995) and carrot (Dau c 1; Hoffmann-Sommergruber et al., 1999). In general, the IgE-binding sites on Bet v 1 are conformational in nature (Gajhede et al., 1996; Neudecker et al., 2001) and consequently IgE reactivity is lost following processing that causes unfolding of the protein. They are also labile to gastrointestinal digestion. A second group of widespread IgE cross-reactive allergens involved in pollen-fruit cross-reactive allergies are the profilins, which were originally identified as the birch pollen allergen Bet v 2 (Valenta et al., 1991). Profilins have a key role in biological systems promoting the polymerisation of actin, but whilst being widespread in nature only those found in plants have been described as allergens. Whilst some studies have cast doubt about the clinical relevance of IgE to profilins (Wensing et al., 2002), others have demonstrated that profilins play an important role in eliciting symptoms in certain patients (Radauer et al., 2006). There is second type of fruit and vegetable allergy which is generally found in the Mediterranean area which is not associated with pollen allergy and tends to be expressed with much more severe, even life-threatening allergic reactions. It involves a distinctly different group of allergens, the non-specific lipid transfer proteins (LTPs) (Fernandez-Rivas et al., 2006). These proteins have been characterised as allergens in fruits such as apple (Mal d 3; Sanchez-Monge et al., 1999b), peach (Pru p 3; Pastorello et al., 1999) and grape (Vit v 1; Pastorello et al., 2003), and vegetables such as asparagus (Diaz-Perales et al., 2002) and cabbage (Bra o 3; Palac´ın et al., 2006). Whilst originally identified in plants through their ability to transfer lipids in an in vitro system, their function in vivo is probably quite different and they appear to have some role in plant protection as they belong to PR group 14 (Breiteneder and Mills, 2005). The observation that those LTPs involved in food
54
Management of Food Allergens
allergies are located in epidermal tissues (Douliez et al., 2000) along with their lipid-binding characteristics has led to suggestions that they play a role in transporting cutin and suberin monomers to the outer layers where they are polymerised to form the outer waxy layers. Like many other members of the prolamin superfamily, they are highly resistant to gastric and duodenal digestion (Asero et al., 2000), with major IgE-binding sites remaining intact as indicated by the fact that simulated gastrointestinal digestion does not alter their ability to elicit skin reactions in vivo as observed for grape LTP (Vassilopoulou et al., 2006). A third group of relevant fruit allergens are those involved in the latex-fruit cross-reactive allergy syndrome which include the class-I chitinases. A group of carbohydrases with a role in protecting plants from pathogens, these proteins are found widely in plants they have been termed panallergens (Salcedo et al., 2001). A number of allergens have been described including ones from avocado (Pers a 1, Sowka et al., 1998), banana (Mus p 1.2 SanchezMonge et al., 1999a) and chestnut (Cas s 1, Diaz-Perales et al., 1998). Other allergens involved in IgE cross-reactive allergies between foods and latex include patatin; a storage protein from potato has also been shown to be cross-reactive with the latex allergen Hev b 7 along with other proteins from avocado and banana (Sowka et al., 1999). Other minor fruit allergens include the highly disulphide-bonded proteins known as thaumatin-like proteins (TLPs), C-proteases (Pastorello et al., 1998a) and a variety of lectins and Kunitz inhibitors identified in potato (Seppala et al., 2001). Many have anti-fungal and/or anti-bacterial activity and therefore may have a role in plant protection. One example of an important emerging allergenic fruit which contains several of these is kiwi fruit in which is found both a TLP (Act d 2; Gavrovic-Jankulovic et al., 2002) and a thiol-protease actinidin (Act c 1; Pastorello et al., 1998a). The eight disulphide bonds in TLP are probably reponsible for their stability to proteolysis (Smole et al., 2008) and the fact that grape TLP retains its allergenicity even after fermentation during wine production (Flamini and De Rosso, 2006). Other less widely found allergens include the flavin adenine dinucleotide (FAD)-containing oxidase allergen of celery, Api g 5, an M r 53-57-kDa protein which is extensively glycosylated and posesses cross-reactive glycans (Bublin et al., 2003) and a germin-like protein which has been identified in bell pepper (Leitner et al., 1998) and orange (Cit s 1, Crespo et al., 2006) for which the N-linked glycans have been found to be important for IgE binding (P¨oltl et al., 2007).
3.3.2
Nuts and seeds
In addition to the pollen-fruit cross-reactive allergy syndromes, it is emerging that Bet v 1 homologues in various nuts and seeds can cause similar allergies. These have been especially well documented for hazelnut where an isoform, Cor a 1.04, has been identified in the nut which resembles Bet v 1 more closely than the allergenic Bet v 1 homologue from hazelnut pollen Cor a 1.01 (L¨uttkopf et al., 2002). There are also reports of LTPs found in nuts and seeds triggering allergies similar to those observed in fruits such as peach, including LTP allergens from walnut (Jug r 3; Pastorello et al., 2004) and hazelnut (Cor a 8; Pastorello et al., 2002) the latter having been shown to be an allergen in a population from Northern Europe (Flinterman et al., 2008a). However, the major allergens in these foods include other members of the prolamin superfamily, the 2S albumins and the cupin seed globulins, both of which often function as a protein store in the seed (Jenkins et al., 2005) (see Plate 3.2). The 2S albumins are usually synthesised in the seed as single chains of M r 10,000–15,000 which maybe post-translationally processed to give small and large subunits which usually
Identification and characterisation of food allergens
55
remain joined by disulfide bonds. The type of this processing depends on the plant species with those in sunflower being single chain albumins, whilst those in Brazil nut are two-chain albumins (Shewry and Pandya, 1999). They have been identified as important allergens in nuts including walnut allergen Jug r 1 (Teuber et al., 1998), almond (Poltronieri et al., 2002) and Ber e 1 from Brazil nut (Pastorello et al., 1998b), and in seeds such as oriental and yellow mustard allergens Bra j 1 and Sin a 1 (Monsalve et al., 1993; Menendez-Arias et al., 1988), Ses i 1 and 2 from sesame (Pastorello et al., 2001b; Beyer et al., 2002b; Wolf et al., 2004) and the 2S albumin from sunflower seeds SFA-8 (Kelly et al., 2000). In addition to the 2S albumins, a major group of allergens found in nuts and seeds are the 11S and 7S seed storage globulins which belong to the cupin superfamily. The 11S globulins, sometimes termed legumins because they are particularly found in legume seeds, are hexameric proteins of M r ∼300,000–450,000. Each subunit is synthesised in the seed as a single chain of M r about 60,000, which is post-translationally processed to give rise to acidic (M r about 40,000) and basic (M r about 20,000) chains, linked by a single disulfide bond and are rarely, if ever, glycosylated (Mills et al., 2004b). The 7/8S globulins, also termed vicilins, are somewhat simpler, comprising three subunits of M r ∼40,000–80,000, but typically about 50,000. Seed storage protein allergens have been described in a variety of nuts and seeds with both 11S and 7S proteins having been reported as allergens in hazelnut (Cor a 11 (7S globulin) and Cor a 9 (11S globulin); Pastorello et al., 2002; Beyer et al., 2002a, b), cashew nut (Ana c 1 and Ana c 2; Wang et al., 2002, 2003), whilst only the 7S globulins of walnut (Jug r 2; Teuber et al., 1999), sesame seed (Ses i; Beyer et al., 2002b) and mustard seed (Palomares et al., 2005, 2007) having been identified. The 11S globulins have also been shown to be allergens in almond, also known as almond major protein (AMP) (Roux et al., 2003). Another group of potentially important allergens that has been identified in the last few years is the oleosins, a group of proteins associated with oil bodies where they play an important role in packaging and stabilising the oil droplet surface, having a portion of the protein structure buried in the oil phase with a second domain on the aqueous facing surface. These have been identified as allergens in sesame (Leduc et al., 2006) and hazelnut (Akkerdaas et al., 2006). They may be important if they find their way into crudely refined oil which, unlike highly refined oils, has sufficient protein to trigger allergic reactions (Crevel et al., 2000).
3.3.3
Legumes
Many of the types of allergens found in other plant foods have also been identified in allergenic legumes. Thus, allergens involved in the cross-reactive pollen syndromes have been identified in several legumes including the Bet v 1 homologues in soybean (KleineTebbe et al., 2002; Mittag et al., 2004a), peanut (Ara h 8; Mittag et al., 2004b) and mung bean (Mittag et al., 2005) and peanut profilin (Kleber-Janke et al., 1999). In addition, the 7S (β-congrlcinin) and 11S (glycinin) seed storage globulins have been described as allergens in soybean (Ogawa et al., 1995; Burks et al., 1988; Beardslee et al., 2000) (see Plate 3.3). There is also some evidence that the 2S albumins of soy (Shibasaki et al., 1980) and chickpea (Vioque et al., 1999) are allergenic. In addition, the 7S (Ara h 1; Burks et al., 1991) and 11S (Ara h 3; Rabjohn et al., 1999; Beardslee et al., 2000) seed storage globulins as well as the prolamin superfamily albumin, Ara h 2, 6 and 7 (Burks et al., 1992; Kleber-Janke et al., 1999), are important allergens in peanut. In contrast to the 2S albumins which are relatively resistant to simulated gastrointestinal proteolysis (Suhr et al., 2004), the 7S globulins are
56
Management of Food Allergens
highly susceptible to pepsinolysis. A number of lower molecular weight polypeptides appear to persist following digestion of the peanut Ara h 1, although they retain their IgE-binding capacity following proteolysis (Shin et al., 1998) and simulated gastrointestinal digestion (Eiwegger et al., 2006). Seed storage globulin allergens have also been identified as allergens in lentil (Len c 1; Lopez-Torrejon et al., 2003) and pea (Pis s 1; Sanchez-Monge et al., 2004) which can be cross-reactive with peanut (Wensing et al., 2003). Such cross-reactivity is particularly problematic with lupin (Moneret-Vautrin et al., 1999) with proteins such as conglutin-β having been identified as a major allergen, Lup an 1 (Goggin et al., 2008). Lup an 1 is a 7S seed storage globulin with significant homology to the peanut allergen Ara h 1 and hence may be responsible for the clinical cross-reactivity observed between these two legumes. Other allergens identified in peanut include an oleosin (Pons et al., 2002) and a lectin, peanut agglutinin (Burks et al., 1994), whilst in soybean a Kunitz trypsin inhibitor (Moroz and Yang, 1980; Burks et al., 1994) and a member of the cysteine protease family, the 34 kDa so-called oil body-associated protein, known as Gly m 1, and Glym Bd 30 k (Ogawa et al., 1993), have been identified as allergens in soybean. Another soybean allergen which is of relevance in countries such as Japan is the M r 23 kDa protein known as Gly m 28 k which is glycosylated and contains important IgE-reactive glycans also found in a derived 23 kDa peptide (Hiemori et al., 2004). Found in the protein storage vacuoles, the protein has an as yet unknown function in the plant.
3.3.4
Cereals
Cereals have been found to trigger two types of IgE-mediated allergic disease: the occupational allergy known as Baker’s asthma, which results from inhalation of flour particles in dusty working environments such as bakeries, and food allergies resulting from ingestion of cereal containing foods. There also appear to be some individuals who react to wheat proteins as a result of prior sensitisation to grass pollen, who are serologically distinct from those who are exposed to flour in a work environment (Sander et al., 1997). However, IgE-mediated allergy to wheat products does not appear to be as widespread as allergies to foods such as egg and peanut, despite a public perception that wheat allergy is prominent. Diagnosis of wheat allergy is further complicated by the low solubility of cereal seed storage prolamins in the dilute salt solutions routinely used in clinical diagnosis, which may mean that cereal allergy may remain undiagnosed. The seed storage prolamins of cereals, also known more commonly as gluten, are usually associated with causing the food intolerance syndrome, coeliac disease, but can also trigger allergies to cereals, both by ingestion and through inhalation (Sandiford et al., 1997) including conditions such as atopic dermatitis (Varjonen et al., 1995, 1997) and exercise-induced anaphylaxis (EIA) (Varjonen et al., 2000). The latter is a severe allergic reaction that certain patients experience only if they exercise after eating a problem food; two allergens have been described as triggering such reactions, including γ -, α- and ω-5 gliadins (Palosuo et al., 1999, 2001; Matsuo et al., 2004). Other prolamin storage proteins have been identified as major cereal allergens, including both the polymeric HMW and LMW subunits of glutenin as well as the monomeric γ and an α-gliadins (Watanabe et al., 1995; Tanabe et al., 1996; Maruyama et al., 1998; Simonato et al., 2001a). Cooking appears to affect allergenicity, and one study suggested baking may be essential for allergenicity of cereal prolamins (Simonato et al., 2001b).
Identification and characterisation of food allergens
57
In addition to the gluten protein fraction, other wheat proteins have been implicated as food allergens including a single M r ∼15,000 subunit corresponding to an trypsin/αamylase inhibitor identified as an allergen (James et al., 1997), whilst another trypsin/αamylase inhibitor, termed CM3, has been identified as an allergen triggering atopic dermatitis (Kusaba-Nakayama et al., 2000). α-Amylase inhibitors have also been implicated in allergies to other cereal-based foods, including an Mr 16,000 beer allergen which originates from barley (Curioni et al., 1999) and an Mr 16,000 protein which is a major allergen in maize (Pastorello et al., 2000). A number of α-amylase inhibitors with M r of about 14,000–16,000 including one M r ∼16,000 subunit, termed RA 17, have been described as allergens in rice (Nakase et al., 1994). Another group of proteins which have been described as cereal allergens is the type 1 non-specific lipid transfer proteins (LTPs), including ones from maize (Pastorello et al., 2000), spelt (Pastorello et al., 2001a) and wheat (Tri a 14; Pastorello et al., 2007). Species differences in response to processing are emerging, since cooking wheat did not modify the IgE-binding capacity of the α-amylase inhibitors but some patients lost their IgE-binding capacity towards the LTP, in contrast to maize (Pastorello et al., 2007). In addition, barley LTP has been found to trigger allergic reactions in beer (Curioni et al., 1999; Asero et al., 2001; Garcia-Casado et al., 2001).
3.4 ANIMAL FOOD ALLERGENS 3.4.1
Cow’s milk
The major allergens in cow’s milk are the caseins, a group of structurally mobile proteins which bind calcium through clusters of phosphoserine and/or phosphothreonine residues. Caseins are a heterogeneous mixture of proteins, the product of expression of a polymorphic multigene family which undergoes post-translational proteolysis and phosphorylation. They show a heterogeneity of IgE-binding properties. Thus, IgE cross-reactivity studies in a group of cow’s milk allergic infants showed that whilst all but 10% had serum IgE against α S2 casein, only around half recognised α S1 -casein and only a small proportion (15%) had IgE against β-casein (Natale et al., 2004). The high level of homology between caseins form different mammalian species explains their IgE cross-reactivity. Extensive cross-reactivity has been observed between the milks of cow, sheep and goat (Spuergin et al., 1997) and between the milks of cow, ewe, goat and buffalo, but not of camel (Restani et al., 1999). Thus individuals with cows’ milk allergy generally reacting when undergoing oral challenge with goat’s milk (Bellioni-Businco et al., 1999) whose caseins have sequence identities of over 90% with bovine caseins. Lower sequence identities of 22–66% may be associated with reduced IgE cross-reactivity, which may explain why some individuals with cow’s milk allergy can tolerate mare’s milk (Businco et al., 2000) and do not show IgE cross-reactivity to milk proteins from species such as camel (Restani et al., 1999). In addition, it has also been suggested that mare’s milk and donkey’s milk might be used in selected cases of cow’s milk allergy after appropriate modification to make them suitable for human infants (Businco et al., 2000; Muraro et al., 2002). More recently, allergies to goats’ or sheep’s milk have been emerging, although the IgE reactivity appears to be confined to the casein fraction (Ah-Leung et al., 2006). The other important allergens in cows’ milk are the whey proteins β-lactoglobulin, the only lipocalin which acts as a food allergen (Virtanen, 2001), and α-lactalbumin, which like the egg allergen lysozyme belongs to the glycoside hydrolase family 22 clan of the
58
Management of Food Allergens
O-glycosyl hydrolase superfamily (Wal, 2002). Lastly, one minor allergen identified in milk is the iron-binding protein, lactoferrin (Wal, 2002).
3.4.2 Egg A number of allergens have been described in egg, in particular the Kazal inhibitor known as ovomucoid, the dominant hen’s egg white allergen Gal d 3 (Bernhisel-Broadbent et al., 1994), which is extensively glycosylated and may act to stabilise the protein against proteolysis (Cooke and Sampson, 1997). Another egg allergen is also a protease inhibitor, the serpin serine protease inhibitor namely ovalbumin Gal d 1 (Bernhisel-Broadbent et al., 1994). A third type of hydrolase, this time a glycosidase belonging to the glycoside hydrolase family 22 clan of the O-glycosyl hydrolase superfamily, namely lysozyme has been described as a minor hen’s egg allergen, known as Gal d 4 (Nitta and Sugai, 1989). It probably acts as a muramidase, hydrolysing peptidoglycans found in bacterial cell wall. Lastly, the sulphur-rich iron-binding glycoprotein ovotransferrin has also been identified as minor allergens in hen egg white (Holen and Elsayed, 1990; Aabin et al., 1996).
3.4.3
Fish
The major allergen identified so far in fish is the white muscle protein known as parvalbumin, a protein which contains a calcium-binding domain known as an EF-hand which is widely found throughout living systems (Lewit-Bentley and Rety, 2000). Loss of calcium results in a major change in conformation which also results in a loss of IgE binding capacity (Bugajska-Schretter et al., 1998, 2000). The parvalbumins family comprises two distinct subtypes, known as α- and the β-parvalbumins, and whilst their three-dimensional structures are very similar, with only one exception all food allergens belong to the β-parvalbumin family (Jenkins et al., 2007) (see Plate 3.4). The only α-parvalbumin reported to be an allergen is that of frog (Hilger et al., 2002). The codfish allergen, Gad c 1, was the first allergenic fish parvalbumin to be described (Aas and Jebsen, 1967; Elsayed and Bennich, 1975) but a number have now been identified in many different fish species and can therefore be considered to be the pan-allergens in fish (Bernhisel-Broadbent, 1992). Clinical crossreactivity to multiple fish in individuals with fish allergy based on the major fish allergen parvalbumin is a common observation (Sicherer, 2001).
3.4.4
Shellfish and Crustacea
A family of closely related proteins present in muscle and non-muscle cells, tropomyosins are major allergens in the two invertebrate groups, Crustacea and Mollusca, which are generally known as shellfish and are generally assumed to be responsible for seafood allergies. The first to be characterised were those identified in shrimp, and are now acknowledged to be invertebrate pan-allergens by several laboratories (Shanti et al., 1993; Daul et al., 1994; Leung et al., 1994; Reese et al., 1999). The first two residues of the IgE binding region (epitope) in the C-terminal portion of the protein appear to be crucial for IgE binding and is not found in vertebrate tropomyosin. As a consequence of the lack of homology in the IgE epitopes, there is no reported cross-reactivity between IgE from shellfish allergic individuals, and animal muscle tropomyosins. As a result of their extensive homology, tropomyosins exhibit IgE cross-reactivity between various crustacean and mollusc species
Identification and characterisation of food allergens
59
(Motoyama et al., 2006), but whilst clinical reactions to multiple crustacean species seem to be fairly common, this is less clear regarding mollusc reactivity which may be restricted to cross-sensitisation (Sicherer, 2001). The proteins appear to be generally heat stable, their allergenicity being unaltered by boiling (Naqpal et al., 1989), with tropomyosins being detected in the cooking water (Lehrer et al., 1990). A minor group of allergens identified in shrimp is the arginine kinases (Yu et al., 2003) which have also been identified as crossreactive allergens in the Indian meal moth, king prawn, lobster and mussel (Binder et al., 2001).
3.5 CONCLUSIONS The last 20 years have seen a great expansion in our knowledge of food allergens, particularly with regard to their identification and characterisation with most of the allergens identified in the major allergenic foods. This knowledge is facilitating development of novel diagnostic approaches, in particular component resolved diagnosis using individual purified and highly characterised allergens to relate specific symptoms with the profile of allergens recognised. Such approaches, should they prove effective, have many benefits, as currently the most effective way to diagnose a food allergy is to undertake a food challenge with all its attendant risks to the patient their with time-consuming nature. Novel diagnostic platforms, such as protein arrays, have the potential to deliver improved diagnostic capability whilst utilising only small quantities of serum (Asero et al., 2007). This knowledge of allergens is also leading to the development of new approaches regarding the development of therapeutics, in particular allergens engineered to lose their IgE epitopes, yet retain sufficient immunological activity to allow them to be used to desensitise allergic individuals. Such approaches appear promising, based on studies in animal models using fish parvalbumins (Swoboda et al., 2007). This large body of knowledge is also allowing us to begin to study what makes one protein become an allergen, but not another. There are indications that the food processing and the matrix itself may modulate any intrinsic allergenic potential of an allergen per se. For example, the potency of peanut allergens in a chocolate matrix to elicit an allergic reaction was shown some years ago to be modulated by addition of vegetable fat, the higher fat content masking early oral reactions, resulting in greater consumption of the food and subsequently more severe reactions (Grimshaw et al., 2003). Our extensive knowledge of the allergens responsible for food allergies will make it possible to begin tackling such issues, despite the lack of adequate animal models for food allergy. This will be important for managing the risks associated with food allergies, whether it be predicting the allergenic potential of a novel protein or food, or managing allergens in foods within existing factory environments. Certainly, new methods for detecting allergens on foods based on the revolution in protein mass spectroscopy that has taken place in the past 10 years will need to draw on this body of knowledge in the future.
Acknowledgements This work was supported by the competitive strategic grant from BBSRC to IFR.
60
Management of Food Allergens
REFERENCES Aabin B., Poulsen L.K., Ebbehoj K. et al. (1996) Identification of IgE-binding egg white proteins: comparison of results obtained by different methods. International Archives of Allergy and Applied Immunology, 109, 50–57. Aas K. and Jebsen J.W. (1967) Studies of hypersensitivity to fish. Partial purification and crystallization of a major allergenic component of cod. International Archives of Allergy and Applied Immunology, 32, 1–20. Abolhassani M. and Roux K.H. (2007) cDNA cloning, expression and characterization of an allergenic 60s ribosomal protein of almond (Prunus dulcis). Journal of Allergy and Clinical Immunology, Unpublished (direct submission to NCBI accession number ABH03379.1). Ah-Leung S., Bernard H., Bidat E. et al. (2006) Allergy to goat and sheep milk without allergy to cow’s milk. Allergy, 61, 1358–1365. Akkerdaas J.H., Schocker F., Vieths S. et al. (2006) Cloning of oleosin, a putative new hazelnut allergen, using a hazelnut cDNA library. Molecular Nutrition and Food Research, 50, 18–23. Asero R., Ballmer-Weber B.K., Beyer K. et al. (2007) IgE-mediated food allergy diagnosis: current status and new perspectives. Molecular Nutrition and Food Research, 51(1), 135–147. Asero R., Mistrello G., Roncarolo D. et al. (2001) A case of allergy to beer showing cross-reactivity between lipid transfer proteins. Annals of Allergy, Asthma, and Immunology, 87, 65–67. Asero R., Mistrello G., Roncarolo D. et al. (2000) Lipid transfer protein: a pan-allergen in plant-derived foods that is highly resistant to pepsin digestion. International Archives of Allergy and Applied Immunology, 122, 220–232. Asturias J.A., Eraso E., Arilla M.C. et al. (2002) Cloning, isolation, and IgE-binding properties of Helix aspersa (brown garden snail) tropomyosin. International Archives of Allergy and Immunology, 128(2), 90–96. Awade A.C., Moreau S., Molle D. et al. (1994) Two-step chromatographic procedure for the purification of hen egg white ovomucin, lysozyme, ovotransferrin and ovalbumin and characterization of purified proteins. Journal of Chromatography A, 677(2), 279–288. Bando N., Tsuji H., Yamanishi R. et al. (1996) Identification of the glycosylation site of a major soybean allergen, Gly m Bd 30K. Bioscience, Biotechnology, and Biochemistry, 60(2), 347–348. Bateman A., Coin L., Durbin R. et al. (2004) The Pfam protein families database. Nucleic Acids Research, 32(Database issue), D138–D141. Battais F., Pineau F., Popineau Y. et al. (2003) Food allergy to wheat: identification of immunogloglin E and immunoglobulin G-binding proteins with sequential extracts and purified proteins from wheat flour. Clinical and Experimental Allergy, 33(7), 962–970. Beardslee T.A., Zeece M.G., Sarath G. and Markwell J.P. (2000) Soybean glycinin G1 acidic chain shares IgE epitopes with peanut allergen Ara h 3. International Archives of Allergy and Applied Immunology, 123, 299–307. Bellioni-Businco B., Paganelli R., Lucenti P. et al. (1999) Allergenicity of goat’s milk in children with cow’s milk allergy. Journal of Allergy and Clinical Immunology, 103, 1191–1194. Bernard H., Creminon C., Yvon M. and Wal J.M. (1998) Specificity of the human IgE response to the different purified caseins in allergy to cow’s milk proteins. International Archives of Allergy and Immunology, 115(3), 235–244. Bernhisel-Broadbent J., Dintzis H.M., Dintzis R.Z. and Sampson H.A. (1994) Allergenicity and antigenicity of chicken egg ovomucoid (Gal d III) compared with ovalbumin (Gal d I) in children with egg allergy and in mice. Journal of Allergy and Clinical Immunology, 93, 1047–1059. Bernhisel-Broadbent J., Scanlon S.M. and Sampson H.A. (1992) Fish hypersensitivity. I. In vitro and oral challenge results in fish-allergic patients. Journal of Allergy and Clinical Immunology, 89, 730– 737. Bernhisel-Broadbent J., Yolken R.H. and Sampson H.A. (1991) Allergenicity of orally administered immunoglobulin preparations in food-allergic children. Pediatrics, 87(2), 208–214. Besler M., Steinhart H. and Paschke A. (1997) Allergenicity of hen’s egg-white proteins: IgE binding of native and deglycosylated ovomucoid. Food and Agricultural Immunology, 9(4), 277–288. Beyer K., Bardina L., Grishina G. and Sampson H.A. (2002a) Identification of sesame seed allergens by 2dimensional proteomics and Edman sequencing: seed storage proteins as common food allergens. Journal of Allergy and Clinical Immunology, 110, 154–159.
Identification and characterisation of food allergens
61
Beyer K., Grishina G., Bardina L. et al. (2002b) Identification of an 11S globulin as a major hazelnut food allergen in hazelnut-induced systemic reactions. Journal of Allergy and Clinical Immunology, 110(3), 517–523. Binder M., Mahler V., Hayek B. et al. (2001) Molecular and immunological characterization of arginine kinase from the Indianmeal moth, Plodia interpunctella, a novel cross-reactive invertebrate pan-allergen. Journal of Immunology, 167, 5470–5477. Bolhaar S.T., Ree R., Bruijnzeel-Koomen C.A. et al. (2004) Allergy to jackfruit: a novel example of Bet v 1-related food allergy. Allergy, 59, 1187–1192. Bolhaar S.T., van Ree R., Ma Y. et al. (2005) Severe allergy to sharon fruit caused by birch pollen. International Archives of Allergy and Applied Immunology, 136, 45–52. Breiteneder H., Hoffmann-Sommergruber K., O’Riordain G. et al. (1995) Molecular characterization of Api g 1, the major allergen of celery (Apium graveolens), and its immunological and structural relationships to a group of 17-kDa tree pollen allergens. European Journal of Biochemistry, 233, 484–489. Breiteneder H. and Mills E.N.C. (2005) Molecular properties of food allergens. Journal of Allergy and Clinical Immunology, 115, 14–23. Breiteneder H., Pettenburger K., Bito A. et al. (1989) The gene coding for the major birch pollen allergen Betv1, is highly homologous to a pea disease resistance response gene. EMBO Journal, 8, 1935–1938. Breiteneder H. and Radauer C. (2004) A classification of plant food allergens. Journal of Allergy and Clinical Immunology, 113, 821–830. Bublin M., Radauer C., Wilson I.B. et al. (2003) Cross-reactive N-glycans of Api g 5, a high molecular weight glycoprotein allergen from celery, are required for immunoglobulin E binding and activation of effector cells from allergic patients. FASEB Journal, 17, 1697–1699. Bugajska-Schretter A., Elfman L., Fuchs T. et al. (1998) Parvalbumin, a cross-reactive fish allergen, contains IgE-binding epitopes sensitive to periodate treatment and Ca2+ depletion. Journal of Allergy and Clinical Immunology, 101(1), 67–74. Bugajska-Schretter A., Grote M., Vangelista L. et al. (2000) Purification, biochemical, and immunological characterisation of a major food allergen: different immunoglobulin E recognition of the apo- and calciumbound forms of carp parvalbumin. Gut, 46, 661–669. Burks, A.W., Brooks, J.R. and Sampson, H.A. (1988) Allergenicity of major component proteins of soybean determined by enzyme-linked immunosorbent assay (ELISA) and immunoblotting in children with atopic dermatitis and positive soy challenge. Journal of Allergy and Clinical Immunology, 81, 1135–1142. Burks A.W., Cockrell G., Connaughton C. et al. (1994) Identification of peanut agglutinin and soybean trypsin inhibitor as minor legume allergens. International Archives of Allergy and Immunology, 105, 143–149. Burks A.W., Williams L.W., Connaughton C. et al. (1992) Identification and characterization of a second major peanut allergen, Ara h II, with use of the sera of patients with atopic dermatitis and positive peanut challenge. Journal of Allergy and Clinical Immunology, 90, 962–969. Burks A.W., Williams L.W., Helm R.M. et al. (1991) Identification of a major peanut allergen, Ara h I, in patients with atopic dermatitis and positive peanut challenges. Journal of Allergy and Clinical Immunology, 88, 172–179. Bush R.K. and Hefle S.L. (1996) Food allergens. CRC Critical Reviews in Food Science and Nutrition, 36(Suppl), S119–S163. Businco L., Giampietro P.G., Lucenti P. et al. (2000) Allergenicity of mare’s milk in children with cow’s milk allergy. Journal of Allergy and Clinical Immunology, 105, 1031–1034. Busse P.J., J¨arvinen K.M., Vila L. et al. (2002) Identification of sequential IgE-binding epitopes on bovine alpha(s2)-casein in cow’s milk allergic patients. International Archives of Allergy and Immunology, 129(1), 93–96. Carbonero P. and Garcia-Olmedo F. (1999) A multigene family of trypsin/alpha-amylase inhibitors from cereals. In: Seed Proteins (eds P.R. Shewry and R. Casey). Kluwer, Amsterdam, pp. 617–634. Chatchatee P., Jarvinen K.M., Bardina L. et al. (2001) Identification of IgE and IgG binding epitopes on betaand kappa-casein in cow’s milk allergic patients. Clinical and Experimental Allergy, 31(8), 1256–1262. Clarke M.C., Kilburn S.A., Hourihane J.O. et al. (1998) Serological characteristics of peanut allergy. Clinical and Experimental Allergy, 28(10), 1251–1257. Codina R., Ardusso L., Lockey R.F. et al. (2002) Identification of the soybean hull allergens involved in sensitization to soybean dust in a rural population from Argentina and N-terminal sequence of a major 50 KD allergen. Clinical and Experimental Allergy, 32(7), 1059–1063.
62
Management of Food Allergens
Constantin C., Quirce S., Grote M. et al. (2008) Molecular and immunological characterization of a wheat serine proteinase inhibitor as a novel allergen in baker’s asthma. Journal of Immunology, 180(11), 7451–7460. Cooke S.K. and Sampson, H.A. (1997) Allergenic properties of ovomucoid in man. Journal of Immunology, 159, 2026. Crespo J.F., Retzek M., Foetisch K. et al. (2006) Germin-like protein Cit s 1 and profilin Cit s 2 are major allergens in orange (Citrus sinensis) fruits. Molecular Nutrition and Food Research, 50, 282–290. Crevel R.W., Kerkhoff M.A. and Koning M.M. (2000) Allergenicity of refined vegetable oils. Food and Chemical Toxicology, 38, 385–393. Crowell D.N., John M.E., Russell D. and Amasino R.M. (1992) Characterization of a stress-induced, developmentally regulated gene family from soybean. Plant Molecular Biology, 18(3), 459–466. Curioni A., Santucci B., Cristado A. et al. (1999) Urticaria from beer: an immediate hypersensitivity reaction due to a 10-kDa protein derived from barley. Clinical and Experimental Allergy, 29, 407. Daul C.B., Slattery M., Reese G. and Lehrer S.B. (1994) Identification of the major brown shrimp (Penaeus aztecus) allergen as the muscle protein tropomyosin. International Archives of Allergy and Applied Immunology, 105, 49–55. De Blay F., Hoyet C., Candolfi E. et al. (1994) Identification of alpha livetin as a cross reacting allergen in a bird-egg syndrome. Allergy Proceedings, 15(2), 77–78. Diaz-Perales A., Collada C., Blanco C. et al. (1998) Class I chitinases with hevein-like domain, but not class II enzymes, are relevant chestnut and avocado allergens. Journal of Allergy and Clinical Immunology, 102, 127–133. Diaz-Perales A., Tabar A.I., Sanchez-Monge R. et al. (2002) Characterization of asparagus allergens: a relevant role of lipid transfer proteins. Journal of Allergy and Clinical Immunology, 110, 790–796. Douliez J.-P., Michon T., Elmorjani K. and Marion D. (2000) Structure, biological and technological function of lipid transfer proteins and indolines, the major lipid binding proteins from cereal kernels. Journal of Cereal Science, 32, 1–20. Dupont F.M., Vensel W.H., Chan R. and Kasarda D.D. (2000) Characterization of the 1B-Type omegaGliadins from Triticum aestivum Cultivar Butte. Cereal Chemistry, 77, 607–614. Eiwegger T., Rigby N., Mondoulet L. et al. (2006) Gastro-duodenal digestion products of the major peanut allergen Ara h 1 retain an allergenic potential. Clinical and Experimental Allergy, 36, 1281–1288. Elsayed S. and Bennich H. (1975) The primary structure of allergen M from cod. Scandinavian Journal of Immunology, 4, 203–208. Fernandez-Rivas M., Bolhaar S., Gonzalez-Mancebo E. et al. (2006) Apple allergy across Europe: how allergen sensitization profiles determine clinical expression of plant food allergies. Journal of Allergy and Clinical Immunology, 18, 481–488. Flamini R. and De Rosso M. (2006) Mass spectrometry in the analysis of grape and wine proteins. Expert Reviews in Proteomics, 3, 321–331. Flinterman A.E., Akkerdaas J.H., den Hartog Jager C.F. et al. (2008a) Lipid transfer protein-linked hazelnut allergy in children from a non-Mediterranean birch-endemic area. The Journal of Allergy and Clinical Immunology, 121(2), 423–428. Flinterman A.E., Akkerdaas J.H., Knulst A.C. et al. (2008b) Hazelnut allergy: from pollen-associated mild allergy to severe anaphylactic reactions. Current Opinion in Allergy and Clinical Immunology, 8(3), 261–265. France R.M., Sellers D.S. and Grossman S.H. (1997) Purification, characterization, and hydrodynamic properties of arginine kinase from gulf shrimp (Penaeus aztecus). Archives of Biochemistry and Biophysics, 345(1), 73–78. Gaier S., Marsh J., Oberhuber C. et al. (2008) Purification and structural stability of the peach allergens Pru p 1 and Pru p 3. Molecular Nutrition and Food Research, 52(Suppl 2), S220–S229. Gajhede M., Osmark P., Poulsen F.M. et al. (1996) X-ray and NMR structure of Bet v 1, the origin of birch pollen allergy. Nature Structural Biology, 3, 1040–1045. Ganglberger E., Radauer C., Grimm R. et al. (2000) N-terminal sequences of high molecular weight allergens from celery tuber. Clinical and Experimental Allergy, 30, 566–570. Garcia-Casado G., Crespo J.F., Rodriguez J. and Salcedo G. (2001) Isolation and characterization of barley lipid transfer protein and protein Z as beer allergens. Journal of Allergy and Clinical Immunology, 108, 647–649.
Identification and characterisation of food allergens
63
Garcia-Mas J., Messeguer R., Arus P. and Puigdomenech P. (1995) Molecular characterization of cDNAs corresponding to genes expressed during almond (Prunus amygdalus Batsch) seed development. Plant Molecular Biology, 27(1), 205–210. Gavrovic-Jankulovic M., Irkovic T., Vuckovic O. et al. (2002) Isolation and biochemical characterization of a thaumatin-like kiwi allergen. Journal of Allergy and Clinical Immunology, 110, 805–810. Goggin D.E., Mir G., Smith W.B. et al. (2008) Proteomic analysis of lupin seed proteins to identify conglutin beta as an allergen, Lup an 1. Journal of Agricultural and Food Chemistry, 56(15), 6370–6377. Gonzalez R., Varela J., Carreira J. and Polo F. (1995) Soybean hydrophobic protein and soybean hull allergy [letter]. Lancet, 346, 48–49. Grimshaw K.E., King R.M., Nordlee J.A. et al. (2003) Presentation of allergen in different food preparations affects the nature of the allergic reaction – a case series. Clinical and Experimental Allergy, 33, 1581– 1585. Hamada Y., Tanaka H., Ishizaki S. et al. (2003) Purification, reactivity with IgE and cDNA cloning of parvalbumin as the major allergen of mackerels. Food and Chemical Toxicology, 41(8), 1149–1156. Hansen T.K., Poulsen L.K., Stahl Skov P. et al. (2004) A randomized, double-blinded, placebo-controlled oral challenge study to evaluate the allergenicity of commercial, food-grade fish gelatin. Food and Chemical Toxicology, 42(12), 2037–2044. Hiemori M., Ito H., Kimoto M. et al. (2004) Identification of the 23-kDa peptide derived from the precursor of Gly m Bd 28K, a major soybean allergen, as a new allergen. Biochimica et Biophysica Acta, 1675, 174–183. Hilger C., Grigioni F., Thill L. et al. (2002) Severe IgE-mediated anaphylaxis following consumption of fried frog legs: definition of alpha-parvalbumin as the allergen in cause. Allergy, 57, 1053–1058. Hoffman D., Lowenstein H., Marsh D.G. et al. (1994) Allergen nomenclature. Bulletin of the World Health Organization, 72, 796–806. Hoffmann-Sommergruber K. (2002) Pathogenesis-related (PR)-proteins identified as allergens. Biochemical Society Transactions, 30(Pt 6), 930–935. Hoffmann-Sommergruber K., O’Riordain G., Ahorn H. et al. (1999) Molecular characterization of Dau c 1, the Bet v 1 homologous protein from carrot and its cross-reactivity with Bet v 1 and Api g 1. Clinical and Experimental Allergy, 29, 840–847. Holen E. and Elsayed S. (1990) Characterization of four major allergens of hen egg-white by IEF/SDSPAGE combined with electrophoretic transfer and IgE-immunoautoradiography. International Archives of Allergy and Applied Immunology, 91, 136–141. Honma K., Kohno Y., Saito K. et al. (1996) Allergenic epitopes of ovalbumin (OVA) in patients with hen’s egg allergy: inhibition of basophil histamine release by haptenic ovalbumin peptide. Clinical and Experimental Immunology, 103(3), 446–453. Hurley W.L. and Schuler L.A. (1987) Molecular cloning and nucleotide sequence of a bovine alphalactalbumin cDNA. Gene, 61, 119–122. Ishikawa M., Ishida M., Shimakura K. et al. (1998) Tropomyosin, the major oyster Crassostrea gigas allergen and its IgE-binding epitopes. Journal of Food Science, 63(1), 44–47. James J.M., Sixbey J.P., Helm R.M. et al. (1997) Wheat α-amylase inhibitor: a second route of allergic sensitisation. The Journal of Allergy and Clinical Immunology, 99, 239–244. Jenkins J.A., Breiteneder H. and Mills E.N.C. (2007) Evolutionary distance from human homologs reflects allergenicity of animal food proteins. The Journal of Allergy and Clinical Immunology, 120(6), 1399– 1405. Jenkins J.A., Griffiths-Jones S., Shewry P.R. et al. (2005) Structural relatedness of plant food allergens with specific reference to cross-reactive allergens: an in silico analysis. Journal of Allergy and Clinical Immunology, 115, 163–170. Jimenez-Flores R., Kang Y.C. and Richardson T. (1987) Cloning and sequence analysis of bovine beta-casein cDNA. Biochemical and Biophysical Research Communications, 142, 617–621. Kelly, J.D., Hlywka, J.J. and Hefle, S.L. (2000) Identification of sunflower seed IgE-binding proteins. International Archives of Allergy and Applied Immunology, 121, 19–24. Kleber-Janke T., Crameri R., Appenzeller U. et al. (1999) Selective cloning of peanut allergens, including profilin and 2S albumins, by phage display technology. International Archives of Allergy and Applied Immunology, 119, 265–274.
64
Management of Food Allergens
Kleine-Tebbe J., Vogel L., Crowell D.N. et al. (2002) Severe oral allergy syndrome and anaphylactic reactions caused by a Bet v 1- related PR-10 protein in soybean, SAM22. Journal of Allergy and Clinical Immunology, 110, 797–804. Kontopidis G., Holt C. and Sawyer L. (2004) Invited review: beta-lactoglobulin: binding properties, structure, and function. Journal of Dairy Science, 87(4), 785–796. Koppelman S.J., de Jong G.A., Laaper-Ertmann M. et al. (2005) Purification and immunoglobulin E-binding properties of peanut allergen Ara h 6: evidence for cross-reactivity with Ara h 2. Clinical and Experimental Allergy, 35(4), 490–497. Koppelman S.J., Wensing M., Ertmann M. et al. (2004) Relevance of Ara h1, Ara h2 and Ara h3 in peanutallergic patients, as determined by immunoglobulin E Western blotting, basophil-histamine release and intracutaneous testing: Ara h2 is the most important peanut allergen. Clinical and Experimental Allergy, 34(4), 583–590. Kusaba-Nakayama M., Ki M., Iwamoto M. et al. (2000) CM3, one of the wheat alpha-amylase inhibitor subunits, and binding of IgE in sera from Japanese with atopic dermatitis related to wheat. Food and Chemical Toxicology, 38, 179–185. Lauer I., Foetisch K., Kolarich D. et al. (2004) Hazelnut (Corylus avellana) vicilin Cor a 11: molecular characterization of a glycoprotein and its allergenic activity. Biochemical Journal, 382(2), 327–334. Leduc V., Moneret-Vautrin D.A., Tzen J.T. et al. (2006) Identification of oleosins as major allergens in sesame seed allergic patients. Allergy, 61, 349–356. Lehrer S.B., Ibanez M.D., McCants M.L. et al. (1990) Characterization of water-soluble shrimp allergens released during boiling. Journal of Allergy and Clinical Immunology, 85, 1005–1013. Lehto M., Palosuo K., Varjonen E. et al. (2003) Humoral and cellular responses to gliadin in wheat-dependent, exercise-induced anaphylaxis. Clinical and Experimental Allergy, 33(1), 90–95. Leitner A., Jensen-Jarolim E., Grimm R. et al. (1998) Allergens in pepper and paprika. Immunologic investigation of the celery-birch-mugwort-spice syndrome. Allergy, 53, 36–41. Leung P.S., Chen Y.C., Gershwin M.E. et al. (1998b) Identification and molecular characterization of Charybdis feriatus tropomyosin, the major crab allergen. Journal of Allergy and Clinical Immunology, 102(5), 847–852. Leung P.S., Chen Y.C., Mykles D.L. et al. (1998a) Molecular identification of the lobster muscle protein tropomyosin as a seafood allergen. Molecular Marine Biology and Biotechnology, 7(1), 12–20. Leung P.S., Chu K.H., Chow W.K. et al. (1994) Cloning, expression, and primary structure of Metapenaeus ensis tropomyosin, the major heat-stable shrimp allergen. Journal of Allergy and Clinical Immunology, 94, 882–890. Lewit-Bentley A. and Rety S. (2000) EF-hand calcium-binding proteins. Current Opinions in Structural Biology, 10, 637–643. Lindstrøm C.D., van Dˆo T., Hordvik I. et al. (1996) Cloning of two distinct cDNAs encoding parvalbumin, the major allergen of Atlantic salmon (Salmo salar). Scandinavian Journal of Immunology, 44(4), 335–344. Lopata A.L., Zinn C. and Potter P.C. (1997) Characteristics of hypersensitivity reactions and identification of a unique 49 kd IgE-binding protein (Hal-m-1) in abalone (Haliotis midae). Journal of Allergy and Clinical Immunology, 100(5), 642–648. Lopez-Torrejon G., Salcedo G., Martin-Esteban M. et al. (2003) Len c 1, a major allergen and vicilin from lentil seeds: protein isolation and cDNA cloning. Journal of Allergy and Clinical Immunology, 112, 1208–1215. L¨uttkopf D., M¨uller U., Skov P.S. et al. (2002) Comparison of four variants of a major allergen in hazelnut (Corylus avellana) Cor a 1.04 with the major hazel pollen allergen Cor a 1.01. Molecular Immunology, 38(7), 515–525. Markovic-Housley Z., Degano M., Lamba D. et al. (2003) Crystal structure of a hypoallergenic isoform of the major birch pollen allergen Bet v 1 and its likely biological function as a plant steroid carrier. Journal of Molecular Biology, 325, 123–133. Maruyama N., Ichise K., Katsube T. et al. (1998) Identification of major wheat allergens by means of the Escherichia coli expression system. European Journal of Biochemistry, 255, 739–745. Matsuo H., Morita E., Tatham A.S. et al. (2004) Identification of the IgE-binding epitope in omega-5 gliadin, a major allergen in wheat-dependent exercise-induced anaphylaxis. Journal of Biological Chemistry, 279, 12135–12140.
Identification and characterisation of food allergens
65
Menendez-Arias L., Moneo I., Dominguez J. and Rodriguez R. (1988) Primary structure of the major allergen of yellow mustard (Sinapis alba L.) seed, Sin a I. European Journal of Biochemistry, 177, 159–166. Mercier J.-C., Grosclaude F. and Ribadeau-Dumas B. (1971) Primary structure of bovine alpha-s1 casein. Complete sequence. European Journal of Biochemistry, 23, 41–51. Mills E.N.C., Jenkins J.A., Alcocer M.J. and Shewry P.R. (2004a) Structural, biological, and evolutionary relationships of plant food allergens sensitizing via the gastrointestinal tract. Critical Reviews in Food Science and Nutrition, 44, 379–407. Mills E.N.C., Jenkins J.A. and Bannon G.A. (2004b) Plant seed globulin allergens. In: Plant Food Allergens (eds E.N.C. Mills, P.R. Shewry). Blackwell Publishing, Oxford, pp. 141–157. Mills E.N.C., Valovirta E., Madsen C. et al. (2004c) Information provision for allergic consumers: where are we going with food allergen labelling? Allergy, 59, 1262–1268. Mittag D., Akkerdaas J., Ballmer-Weber B.K. et al. (2004b) Ara h 8, a Bet v 1-homologous allergen from peanut, is a major allergen in patients with combined birch pollen and peanut allergy. Journal of Allergy and Clinical Immunology, 114, 1410–1417. Mittag D., Vieths S., Vogel L. et al. (2004a) Soybean allergy in patients allergic to birch pollen: clinical investigation and molecular characterization of allergens. Journal of Allergy and Clinical Immunology, 113, 148–154. Mittag D., Vieths S., Vogel L. et al. (2005) Birch pollen-related food allergy to legumes: identification and characterization of the Bet v 1 homologue in mungbean (Vigna radiata), Vig r 1. Clinical and Experimental Allergy, 35, 1049–1055. Miyazawa H., Fukamachi H., Inagaki Y. et al. (1996) Identification of the first major allergen of a squid (Todarodes pacificus). Journal of Allergy and Clinical Immunology, 98(1), 948–953. Mohan S., Sinha N. and Smith-Gill S.J. (2003) Modeling the binding sites of anti-hen egg white lysozyme antibodies HyHEL-8 and HyHEL-26: an insight into the molecular basis of antibody cross-reactivity and specificity. Biophysical Journal, 85(5), 3221–3236. Moneret-Vautrin D.A., Guerin L., Kanny G. et al. (1999) Cross-allergenicity of peanut and lupin: the risk of lupin allergy in patients allergic to peanuts. Journal of Allergy and Clinical Immunology, 104, 883– 888. Monsalve R.I., Gonzalez de la Pena M.A., Menendez-Arias L. et al. (1993) Characterization of a new orientalmustard (Brassica juncea) allergen, Bra j IE: detection of an allergenic epitope. Biochemical Journal, 293, 625–632. Moroz L.A. and Yang W.H. (1980) Kunitz soybean trypsin inhibitor: a specific allergen in food anaphylaxis. New England Journal of Medicine, 302, 1126–1128. Motoyama K., Ishizaki S., Nagashima Y. and Shiomi K. (2006) Cephalopod tropomyosins: identification as major allergens and molecular cloning. Food and Chemical Toxicology, 44, 1997–2002. Muraro M.A., Giampietro P.G. and Galli E. (2002) Soy formulas and nonbovine milk. Annals of Allergy, Asthma, and Immunology, 89(6 Suppl 1), 97–101. Nakase M., Adachi T., Urisu A. et al. (1994) Rice (Oryza sativa L.) alpha amylase inhibitors of 14–16 kDa are potential allergens and products of a multi gene family. Journal of Agricultural and Food Chemistry, 44, 2624–2628. Naqpal S., Rajappa L., Metcalfe D.D. and Rao P.V. (1989) Isolation and characterization of heat-stable allergens from shrimp (Penaeus indicus). Journal of Allergy and Clinical Immunology, 83, 26–36. Natale M., Bisson C., Monti G. et al. (2004) Cow’s milk allergens identification by two-dimensional immunoblotting and mass spectrometry. Molecular Nutrition and Food Research, 48, 363–369. Neudecker P., Schweimer K., Nerkamp J. et al. (2001) Allergic cross-reactivity made visible: solution structure of the major cherry allergen Pru av 1. Journal of Biological Chemistry, 276, 22756– 22763. Neyestani T.R., Djalali M. and Pezeshki M. (2003) Isolation of alpha-lactalbumin, beta-lactoglobulin, and bovine serum albumin from cow’s milk using gel filtration and anion-exchange chromatography including evaluation of their antigenicity. Protein Expression and Purification, 29(2), 202–208. Nielsen N.C., Dickinson C.D., Cho T.J. et al. (1989) Characterization of the glycinin gene family in soybean. Plant Cell, 1(3), 313–328. Nitta K. and Sugai S. (1989) The evolution of lysozyme and α-lactalbumin. European Journal of Biochemistry, 182, 111–118.
66
Management of Food Allergens
Oberhuber O., Bulley S.M., Ballmer-Weber B.K. et al. (2008) Characterization of Bet v 1 related allergens from kiwifruit relevant for patients with combined kiwifruit and birch pollen allergy. Molecular Nutrition and Food Research, 52(Suppl 2), S230–S240. Ogawa T., Bando N., Tsuji H. et al. (1991) Investigation of the IgE-binding proteins in soybeans by immunoblotting with the sera of the soybean-sensitive patients with atopic dermatitis. Journal of Nutritional Science and Vitaminology (Tokyo), 37(6), 555–565. Ogawa T., Bando N., Tsuji H. et al. (1995) Alpha-subunit of β-conglycinin, an allergenic protein recognized by IgE antibodies of soybean-sensitive patients with atopic dermatitis. Bioscience, Biotechnology and Biochemistry, 59, 831–833. Ogawa T., Tsuji H., Bando N. et al. (1993) Identification of the soyabean allergenic proteins, Gly m Bd 30K, with the soybean seed 34kDa oil-body-associated protein. Bioscience, Biotechnology and Biochemistry, 57, 1030–1033. Palac´ın A., Cumplido J., Figueroa J. et al. (2006) Cabbage lipid transfer protein Bra o 3 is a major allergen responsible for cross-reactivity between plant foods and pollens. Journal of Allergy and Clinical Immunology, 117, 1423–1429. Pallares D.E. (2000) Allergy to macademia nut. Annals of Allergy, Asthma, and Immunology, 85(5), 385–386. Palomares O., Cuesta-Herranz J., Vereda A. et al. (2005) Isolation and identification of an 11S globulin as a new major allergen in mustard seeds. Annals of Allergy, Asthma, and Immunology, 94(5), 586–592. Palomares O., Vereda A., Cuesta-Herranz J. et al. (2007) Cloning, sequencing, and recombinant production of Sin a 2, an allergenic 11S globulin from yellow mustard seeds. Journal of Allergy and Clinical Immunology, 119(5), 1189–1196. Palosuo K., Alenius H., Varjonen E. et al. (1999) A novel wheat gliadin as a cause of exercise-induced anaphylaxis. Journal of Allergy and Clinical Immunology, 103, 912–917. Palosuo K., Varjonen E., Kekki O.M. et al. (2001) Wheat omega-5 gliadin is a major allergen in children with immediate allergy to ingested wheat. Journal of Allergy and Clinical Immunology, 108, 634–638. Pastorello E.A., Conti A., Pravettoni V. et al. (1998a) Identification of actinidin as the major allergen of kiwi fruit. Journal of Allergy and Clinical Immunology, 101, 531–537. Pastorello E.A., Farioli L., Conti A. et al. (2007) Wheat IgE-mediated food allergy in European patients: alpha-amylase inhibitors, lipid transfer proteins and low-molecular-weight glutenins. Allergenic molecules recognized by double-blind, placebo-controlled food challenge. International Archives of Allergy and Applied Immunology, 144, 10–22. Pastorello E.A., Farioli L., Pravettoni V. et al. (1999). The major allergen of peach (Prunus persica) is a lipid transfer protein. Journal of Allergy and Clinical Immunology, 103, 520–526. Pastorello E.A., Farioli L., Pravettoni V. et al. (2004) Lipid transfer protein and vicilin are important walnut allergens in patients not allergic to pollen. Journal of Allergy and Clinical Immunology, 114, 908–914. Pastorello E.A., Farioli L., Pravettoni V. et al. (1998b) Sensitization to the major allergen of Brazil nut is correlated with the clinical expression of allergy. Journal of Allergy and Clinical Immunology, 102, 1021–1027. Pastorello E.A., Farioli L., Pravettoni V. et al. (2000) The maize major allergen, which is responsible for food-induced allergic reactions, is a lipid transfer protein. Journal of Allergy and Clinical Immunology, 106, 744–751. Pastorello E.A., Farioli L., Pravettoni V. et al. (2003) Identification of grape and wine allergens as an endochitinase 4, a lipid-transfer protein, and a thaumatin. Journal of Allergy and Clinical Immunology, 111, 350–359. Pastorello E.A., Farioli L., Robino A.M. et al. (2001a) A lipid transfer protein involved in occupational sensitization to spelt. Journal of Allergy and Clinical Immunology, 108, 145–146. Pastorello E.A., Varin E., Farioli L. et al. (2001b) The major allergen of sesame seeds (Sesamum indicum) is a 2S albumin. Journal of Chromatography. B, Biomedical Science Applications, 756, 85–93. Pastorello E.A., Vieths S., Pravettoni V. et al. (2002) Identification of hazelnut major allergens in sensitive patients with positive double-blind, placebo-controlled food challenge results. Journal of Allergy and Clinical Immunology, 109, 563–570. Peeters K.A., Nordlee J.A., Penninks A.H. et al. (2007) Lupine allergy: not simply cross-reactivity with peanut or soy. Journal of Allergy and Clinical Immunology, 120, 647–653. Pierce A., Colavizza D., Benaissa M. et al. (1991) Molecular cloning and sequence analysis of bovine lactotransferrin. European Journal of Biochemistry, 196, 177–184.
Identification and characterisation of food allergens
67
P¨oltl G., Ahrazem O., Paschinger K. et al. (2007) Molecular and immunological characterization of the glycosylated orange allergen Cit s 1. Glycobiology, 17, 220–230. Poltronieri P., Cappello M.S., Dohmae N. et al. (2002) Identification and characterisation of the IgE-binding proteins 2S albumin and conglutin gamma in almond (Prunus dulcis) seeds. International Archives of Allergy and Applied Immunology, 128, 97–104. Pons L., Chery C., Romano A. et al. (2002) The 18 kDa peanut oleosin is a candidate allergen for IgE-mediated reactions to peanuts. Allergy, 57(Suppl 72), 88–93. Rabjohn P., Helm E.M., Stanley J.S. et al. (1999) Molecular cloning and epitope analysis of the peanut allergen Ara h 3. Journal of Clinical Investigation, 103, 535–542. Radauer C. and Breiteneder H. (2006) Pollen allergens are restricted to few protein families and show distinct patterns of species distribution. Journal of Allergy and Clinical Immunology, 117, 141–147. Radauer C., Willerroider M., Fuchs H. et al. (2006) Cross-reactive and species-specific immunoglobulin E epitopes of plant profilins: an experimental and structure-based analysis. Clinical and Experimental Allergy, 36, 920–929. Reese G., Ayuso R. and Lehrer S.B. (1999) Tropomyosin: an invertebrate pan-allergen. International Archives of Allergy and Applied Immunology, 119, 247–258. Restani P., Ballabio C., Cattaneo A. et al. (2004) Characterization of bovine serum albumin epitopes and their role in allergic reactions. Allergy, 59(Suppl 78), 21–24. Restani P., Ballabio C., Corsini E. et al. (2005) Identification of the basic subunit of Ara h 3 as the major allergen in a group of children allergic to peanuts. Annals of Allergy, Asthma, and Immunology, 94(2), 262–266. Restani P., Gaiaschi A., Plebani A. et al. (1999) Cross-reactivity between milk proteins from different animal species. Clinical and Experimental Allergy, 29, 997–1004. Rihs H.P., Chen Z., Rueff F. et al. (1999) IgE binding of the recombinant allergen soybean profilin (rGly m 3) is mediated by conformational epitopes. Journal of Allergy and Clinical Immunology, 104(6), 1293– 1301. Rihs H.P., Rozynek P., May-Taube K. et al. (1994) Polymerase chain reaction based cDNA cloning of wheat profilin: a potential plant allergen. International Archives of Allergy and Immunology, 105(2), 190–194. Robotham J.M., Wang F., Seamon V. et al. (2005) Ana o 3, an important cashew nut (Anacardium occidentale L.) allergen of the 2S albumin family. Journal of Allergy and Clinical Immunology, 115, 1284–1290. Roux K.H., Teuber S.S. and Sathe S.K. (2003) Tree nut allergens. International Archives of Allergy and Applied Immunology, 131, 234. Salcedo G., Diaz-Perales A. and Sanchez-Monge R. (2001) The role of plant panallergens in sensitization to natural rubber latex. Current Opinions in Allergy and Clinical Immunology, 1, 177–183. Salcedo G., Sanchez-Monge R., Garcia-Casado G. et al. (2004) The cereal alpha-amylase/trypsin inhibitor family associated with baker’s asthma and food allergy. In: Plant Food Allergy (eds E.N.C. Mills and P.R. Shewry). Blackwell Science, Oxford, pp. 70–86. Sanchez-Monge R., Blanco C., Diaz-Perales A. et al. (1999a) Isolation and characterization of major banana allergens: identification as fruit class I chitinases. Clinical and Experimental Allergy, 29, 673–680. Sanchez-Monge R., Lombardero M., Garcia-Selles F.J. et al. (1999b) Lipid-transfer proteins are relevant allergens in fruit allergy. Journal of Allergy and Clinical Immunology, 103, 514–519. Sanchez-Monge R., Lopez-Torrej´on G., Pascual C.Y. et al. (2004) Vicilin and convicilin are potential major allergens from pea. Clinical and Experimental Allergy, 34, 1747–1753. Sander I., Raulf-Heimsoth M., Duser M. et al. (1997) Differentiation between cosensitization and crossreactivity in wheat flour and grass pollen-sensitized subjects. International Archives of Allergy and Applied Immunology, 112, 378–385. Sandiford C.P., Tatham A.S., Fido R. et al. (1997) Identification of the major water/salt insoluble wheat proteins involved in cereal hypersensitivity. Journal of Clinical and Experimental Allergy, 27, 1120–1129. Sathe S.K., Wolf W.J., Roux K.H. et al. (2002) Biochemical characterization of amandin, the major storage protein in almond (Prunus dulcis L.). Journal of Agricultural and Food Chemistry, 50(15), 4333–4341. Scheurer S., Wangorsch A., Haustein D. and Vieths S. (2000) Cloning of the minor allergen Api g 4 profilin from celery (Apium graveolens) and its cross-reactivity with birch pollen profilin Bet v 2. Clinical and Experimental Allergy, 30, 962–971. Scheurer S., Wangorsch A., Nerkamp J. et al. (2001) Cross-reactivity within the profilin panallergen family investigated by comparison of recombinant profilins from pear (Pyr c 4), cherry (Pru av 4) and celery
68
Management of Food Allergens
(Api g 4) with birch pollen profilin Bet v 2. Journal of Chromatography. B, Biomedical Sciences and Applications, 756(1–2), 315–325. Schocker F., Luttkopf D., Scheurer S. et al. (2004) Recombinant lipid transfer protein Cor a 8 from hazelnut: a new tool for in vitro diagnosis of potentially severe hazelnut allergy. Journal of Allergy and Clinical Immunology, 113(1), 141–147. Seppala U., Majamaa H., Turjanmaa K. et al. (2001) Identification of four novel potato (Solanum tuberosum) allergens belonging to the family of soybean trypsin inhibitors. Allergy, 56, 619–626. Shanti K.N., Martin B.M., Nagpal S. et al. (1993) Identification of tropomyosin as the major shrimp allergen and characterization of its IgE-binding epitopes. Journal of Immunology, 151, 5354–5363. Shewry P.R. and Pandya M.J. (1999) The 2S albumin storage proteins. In: Seed Proteins (eds P.R. Shewry and R. Casey). Kluwer Academic Publishers, Dordrecht, pp. 563–586. Shibasaki M., Suzuki S., Tajima S. et al. (1980) Allergenicity of major component proteins of soybean. I. International Archives of Allergy and Applied Immunology, 61, 441–448. Shin D.S., Compadre C.M., Maleki S.J. et al. (1998) Biochemical and structural analysis of the IgE binding sites on Ara h1, an abundant and highly allergenic peanut protein. Journal of Biological Chemistry, 273, 13753–13759. Sicherer S.H. (2001) Clinical implications of cross-reactive food allergens. Journal of Allergy and Clinical Immunology, 108, 881–890. Simonato B., De Lazzari F., Pasini G. et al. (2001a) IgE binding to soluble and insoluble wheat flour proteins in atopic and non-atopic patients suffering from gastrointestinal symptoms after wheat ingestion. Clinical and Experimental Allergy, 31, 1771–1778. Simonato B., Pasini G., Giannattasio M. et al. (2001b) Food allergy to wheat products: the effect of bread baking and in vitro digestion on wheat allergenic proteins. A study with bread dough, crumb, and crust. Journal of Agricultural and Food Chemistry, 49, 5668–5673. Smole U., Bublin M., Radauer C. et al. (2008) Mal d 2, the thaumatin-like allergen from apple, is highly resistant to gastrointestinal digestion and thermal processing. International Archives of Allergy and Immunology, 147, 289–298. Sowka S., Hafner C., Radauer C. et al. (1999) Molecular and immunologic characterization of new isoforms of the Hevea brasiliensis latex allergen Hev b 7: evidence of no cross-reactivity between Hev b 7 isoforms and potato patatin and proteins from avocado and banana. Journal of Allergy and Clinical Immunology, 104, 1302–1310. Sowka S., Hsieh L.S., Krebitz M. et al. (1998) Identification and cloning of Prs a 1, a 32-kDa endochitinase and major allergen of avocado, and its expression in the yeast Pichia pastoris. Journal of Biological Chemistry, 273, 28091–28097. Spuergin P., Walter M., Schiltz E. et al. (1997) Allergenicity of alpha-caseins from cow, sheep, and goat. Allergy, 52(3), 293–298. Suhr M., Wicklein D., Lepp U. and Becker W.M. (2004) Isolation and characterization of natural Ara h 6: evidence for a further peanut allergen with putative clinical relevance based on resistance to pepsin digestion and heat. Molecular Nutrition and Food Research, 48, 390–399. Sutton R., Skerritt J.H., Baldo B.A. and Wrigley C.W. (1984) The diversity of allergens involved in bakers asthma. Clinical Allergy, 14, 93–107. Swoboda I., Bugajska-Schretter A., Linhart B. et al. (2007) A recombinant hypoallergenic parvalbumin mutant for immunotherapy of IgE-mediated fish allergy. Journal of Immunology, 178, 6290–6296. Tai S.S.K., Lee T.T.T., Tsai C.C.Y. et al. (2001) Expression pattern and deposition of three storage proteins, 11S globulin, 2S albumin and 7S globulin in maturing sesame seeds. Plant Physiology and Biochemistry, 39, 981–992. Tai S.S.K., Wu L.S.H., Chen E.C.F. and Tzen J.T.C. (1999) Molecular cloning of 11S globulin and 2S albumin, the two major seed storage proteins in sesame. Journal of Agricultural and Food Chemistry, 47, 4932–4938. Tanabe S., Arai S., Yanagihara Y. et al. (1996) A major wheat allergen has a Gln-Gln-Gln-Pro-Pro motif identified as an IgE-binding epitope. Biochemical and Biophysical Research Communications, 219, 290–293. Tawde P., Venkatesh Y.P., Wang F. et al. (2006) Cloning and characterization of profilin (Pru du 4), a crossreactive almond (Prunus dulcis) allergen. Journal of Allergy and Clinical Immunology, 118, 915–922.
Identification and characterisation of food allergens
69
Teuber S.S., Dandekar A.M., Peterson W.R. and Sellers C.L. (1998) Cloning and sequencing of a gene encoding a 2S albumin seed storage protein precursor from English walnut (Juglans regia), a major food allergen. Journal of Allergy and Clinical Immunology, 101, 807–814. Teuber S.S., Jarvis K.C., Dandekar A.M. et al. (1999) Identification and cloning of a complementary DNA encoding a vicilin-like proprotein, Jug r 2, from English walnut kernel (Juglans regia), a major food allergen. Journal of Allergy and Clinical Immunology, 104, 1111–1120. Valenta R., Duchene M., Pettenburger K. et al. (1991) Identification of profilin as a novel pollen allergen; IgE autoreactivity in sensitized individuals. Science, 253, 557–560. Van Do T., Elsayed S., Florvaag E. et al. (2005) Allergy to fish parvalbumins: studies on the cross-reactivity of allergens from 9 commonly consumed fish. Journal of Allergy and Clinical Immunology, 116(6), 1314–1320. Van Regenmortel M.H.V. (1992) Molecular dissection of protein antigens. In: Structure of Antigens, Vol 1 (ed. M.H.V. Van Regenmortel). CRC Press Inc., Boca Raton, pp. 1–28. Vanek-Krebitz M., Hoffmann-Sommergruber K., Laimer da Camara Machado M. et al. (1995) Cloning and sequencing of Mal d 1, the major allergen from apple (Malus domestica), and its immunological relationship to Bet v 1, the major birch pollen allergen. Biochemical and Biophysical Research Communications, 214, 538–551. Varjonen E., Vainio E. and Kalimo K. (1997) Life-threatening, recurrent anaphylaxis caused by allergy to gliadin and exercise. Clinical and Experimental Allergy, 27, 162–166. Varjonen E., Vainio E. and Kalimo K. (2000) Antigliadin IgE – indicator of wheat allergy in atopic dermatitis. Allergy, 55, 386–391. Varjonen E., Vainio E., Kalimo K. et al. (1995) Skin-prick test and RAST responses to cereals in children with atopic dermatitis. Characterization of IgE-binding components in wheat and oats by an immunoblotting method. Clinical and Experimental Allergy, 25, 1100–1107. Vassilopoulou E., Rigby N., Moreno F.J. et al. (2006) Effect of in vitro gastric and duodenal digestion on the allergenicity of grape lipid transfer protein. Journal of Allergy and Clinical Immunology, 118, 473–480. Vioque J., Sanchez-Vioque R., Clemente A. et al. (1999) Purification and partial characterization of chickpea 2S albumin. Journal of Agricultural and Food Chemistry, 47, 1405–1409. Virtanen T. (2001) Lipocalin allergens. Allergy, 56(Suppl 67), 48–51. Wal J.-M. (2002) Cow’s milk proteins/allergens. Annals of Allergy, Asthma, and Immunology, 89(6 Suppl 1), 3–10. Wallowitz M., Peterson W.R., Uratsu S. et al. (2006) Jug r 4, a legumin group food allergen from walnut (Juglans regia Cv. Chandler). Journal of Agricultural and Food Chemistry, 54, 8369–8375. Wang F., Robotham J.M., Teuber S.S. et al. (2002) Ana o 1, a cashew (Anacardium occidental) allergen of the vicilin seed storage protein family. Journal of Allergy and Clinical Immunology, 110, 160–166. Wang F., Robotham J.M., Teuber S.S. et al. (2003) Ana o 2, a major cashew (Anacardium occidentale L.) nut allergen of the legumin family. International Archives of Allergy and Applied Immunology, 132, 27–39. Watanabe M., Tanabe S., Suzuki T. et al. (1995) Primary structure of an allergenic peptide occurring in the chymotryptic hydrolysate of gluten. Bioscience, Biotechnology, and Biochemistry, 59, 1596. Weichel M., Glaser A., Ballmer-Weber B. et al. (2006) Wheat and maize thioredoxins: a novel cross-reactive cereal allergen family related to baker’s asthma. Journal of Allergy and Clinical Immunology, 117(3), 676–681. Wensing M., Akkerdaas J.H., van Leeuwen W.A. et al. (2002) IgE to Bet v 1 and profilin: cross-reactivity patterns and clinical relevance. Journal of Allergy and Clinical Immunology, 110, 435–442. Wensing M., Knulst A.C., Piersma S. et al. (2003) Patients with anaphylaxis to pea can have peanut allergy caused by cross-reactive IgE to vicilin (Ara h 1). Journal of Allergy and Clinical Immunology, 111, 420–424. Wolff N., Yannai S., Karin N. et al. (2004) Identification and characterization of linear B-cell epitopes of beta-globulin, a major allergen of sesame seeds. Journal of Allergy and Clinical Immunology, 114, 1151–1158. Yu C.J., Lin Y.F., Chiang B.L. and Chow L.P. (2003) Proteomics and immunological analysis of a novel shrimp allergen, Pen m 2. Journal of Immunology, 170, 445–453.
4
Coeliac disease: allergy or intolerance?
Norma McGough
4.1 INTRODUCTION Coeliac disease (CD) is not an allergy or simple food intolerance. It is a life-long disease which affects the immune system, i.e. it is an autoimmune disease, and although it mainly affects the gut, it also affects other parts of the body. It is caused by an intolerance to gluten and is therefore triggered by eating gluten, a sequence of amino acids (protein) in the cereals wheat, rye and barley. Some people with CD are also sensitive to oats, one study reported this sensitivity in 1 in 20 people with CD (Lundin et al., 2003). Gluten causes inflammation of the gut in people with CD and the finger-like projections become flattened. This is known as villous atrophy (see Plates 4.1 and 4.2). This results in a variety of symptoms including stomach pain, bloating, sickness, diarrhoea and constipation. This damage affects the absorption of nutrients and can result in a range of nutritional deficiencies and other clinical manifestations. There is no cure for CD, but the gluten-free diet provides a complete treatment.
4.2 ABOUT COELIAC DISEASE Screening studies indicate that CD affects 1% of the population (West et al., 2003; Bingley et al., 2004), making CD one of the most common chronic autoimmune disorders and the most common cause of malabsorption in the United Kingdom. Complications associated with CD include gut cancer (Collin et al., 1994), osteoporosis (Kemppainen et al., 1999) and infertility (Sanders, 2003). CD is more common in those with other autoimmune diseases such as type 1 diabetes (Holmes, 2001) and autoimmune thyroid disease (Cuoco et al., 1999).
4.3 PREVALENCE AND DIAGNOSIS CD can present and be diagnosed at any age, from weaning onto gluten-containing cereals through to late old age. However, it is more commonly diagnosed later in life and is therefore more common in adults than children. Most people are diagnosed between 40 and 50 years of age and more women are diagnosed than men. Underdiagnosis of CD is significant; evidence suggests that only 1 in 8 cases are currently diagnosed (van Heel and West, 2006). CD is diagnosed by antibody blood tests and a small intestinal biopsy. It is important that gluten is not taken out of the diet until all the tests for CD are complete. Following
Coeliac disease: allergy or intolerance?
71
a gluten-free diet prior to testing can result in false negative results (BSG, 2002; CREST, 2006).
4.4
WHAT IS GLUTEN?
‘Gluten’ is a general term used to cover the alcohol-soluble proteins; gliadins in wheat, hordeins in barley and secalins in rye; these proteins are toxic to people with CD. The glutenfree diet involves avoidance of wheat, rye and barley and ingredients derived from these cereals, e.g. wheat starch and barley malt. Some individuals with CD are also sensitive to oats (Lundin et al., 2003; Silano et al., 2007). A significant problem with most oat products is contamination from wheat, rye or barley (Thompson, 2005). The allergen labelling directive currently lists oats as a gluten-containing cereal (EC, 2007). The Draft Revised Codex Standard for Foods for Special Dietary Use for Persons Intolerant to Gluten makes a special reference to oats (CAC, 2007).
4.5
THE GLUTEN-FREE DIET
Dietary management provides a complete treatment for the disease, resulting in gut healing which in turn improves absorption of nutrients. Since the elimination of gluten-containing cereals from the diet results in elimination of the staples in the diet, substitute staple foods including gluten-free bread, gluten-free pasta, gluten-free flour, gluten-free crackers, glutenfree biscuits and gluten-free breakfast cereals are produced by a range of manufacturers including specialist gluten-free manufacturers and retailers. The gluten-free diet is made up of a combination of gluten-free substitute foods to replace the gluten-containing staples of a normal diet; naturally, gluten-free foods such as rice, corn, meat, fish, cheese, eggs, milk, fruit, pulses and vegetables, and mainstream processed foods that also happen to be gluten free. Although the gluten-free diet may sound ‘black and white’, this is not the case and the tolerance to gluten varies amongst people with CD. Over the years, different ranges of substitute products have been developed, which contain acceptably low levels of gluten, known as Codex wheat starch, to replace staple foods on a gluten-free diet. The standard for labelling of these products has been governed by a WHO/FAO body under the auspices of the Codex Committee on Nutrition for Special Dietary Uses (FAO/WHO, 2008).
4.5.1 Oats Historically, oats have been considered unsafe for people with CD and it is only recently that it has become accepted practice for people to include pure uncontaminated oats as part of a balanced gluten-free diet. Although most people with CD may be able to tolerate uncontaminated oats without a problem, research has suggested that some people who have CD may still react (Lundin et al., 2003; CAC, 2007). However, two systematic literature review have shown that pure uncontaminated oats may be incorporated safely into the gluten-free diet (Thompson, 2003; Haboubi et al., 2006), suitability has to be decided on an individual basis Oats are a good source of soluble fibre, which can help to lower cholesterol as well as adding variety to the diet. However, as oats are included in the Directive 2003/89/EC of
72
Management of Food Allergens
allergens, even uncontaminated oats cannot be used in a product or recipe advertised as gluten-free (see Section 4.8; Haboubi et al., 2006). In addition, acceptance of oats differs on the gluten-free diet varies from one country to another with countries like Canada, the USA, Australia and Italy avoiding the use of oats completely.
4.6
GLUTEN-FREE FOODS
The gluten-free checklist provides a guide to what products generally contain gluten (see Table 4.1). There are different ranges of specialist gluten-free foods, e.g. bread and pasta, to replace standard wheat-containing varieties. Some of these foods are available on prescription, from health-food shops, mail order, pharmacies, the internet and supermarkets, with some retailers producing their own ranges of gluten-free substitute products as part of a ‘free from’ range. Bulk catering supplies are also available from suppliers, distributors and the company directly. For more information refer to Appendix 1 of the Coeliac UK Food and Drink Directory.
4.7 PRESCRIPTIONS Gluten-free foods available on prescription tend to be staples such as bread, flour, bread mixes, pizzas, crackers and pasta. It is advantageous for people with CD to be able to obtain samples of products from manufacturers in order to help choose products that they want to obtain on prescription. Coeliac UK provides information via their helpline (Tel: 0870 4448804), website (www.coeliac.org.uk), publications and events. Registered dietitians working in the NHS also provide advice on prescriptions for gluten-free products. The gluten-free foods available on prescription are listed in the National Health Service Drug Tariff for England and Wales (NHS, 2007), following successful submission to the Advisory Committee on Borderline Substances (ACBS). People with CD in England and Scotland are currently still required to pay for prescriptions, unless they meet the criteria for exemption, i.e. children, those in full-time education, those older than 60 years or those individuals on income support or other benefits. Prescriptions in Wales are free of charge. There are voluntary guidelines available on reasonable quantities of gluten-free foods for individuals, based on nutritional requirements, which have been incorporated into PCSG guidance for CD management (PCSG, 2006).
4.8
ALLERGEN LABELLING
Clear use of allergen labelling including management of allergens is essential. The advent of allergen labelling guidance in November 2005, which included gluten as one of the 12 named allergens (currently 14 allergens), makes it possible to select suitable gluten-free food by checking the label. The guidance which makes it necessary for all food pre-packaged and sold in countries belonging to the European Union (EU) to list all deliberate ingredients and clearly identify any food allergen included in a food as a deliberate ingredient (EC, 2007). The presence of gluten must appear on the ingredients list as the grain itself, e.g. wheat, barley, rye, oats, triticale, kamut and spelt. The use of allergen advice boxes is recommended, but not
Coeliac disease: allergy or intolerance? Table 4.1
73
Gluten-free ingredients checklist.
Gluten free
May contain gluten
Cereals and flour Corn, corn flour, rice, rice flour, Flavoured savoury rice arrowroot, amaranth, buckwheat, products, cereal bars, corn and millet, teff, quinoa, sorghum, soya rice-based breakfast cereals flour, potato starch, modified starch, potato flour, gram flour,
Gluten containing Wheat, bulgar wheat, durum wheat, wheat bran, wheat germ, wheat starch, semolina, couscous, barley, malt, malted barley, rye, triticale, kamut, spelt Bread, crackers, pasta, oats, cereals, muesli
Meat, poultry, fish, cheese, eggs All fresh meats, poultry, fish, Meat and fish pastes, pates, shellfish, smoked meats and fish, sausages, burgers, fish in cured pure meats, smoked, fish in sauce, oil/brine, cheese, eggs
Meat, poultry, fish cooked in batter or breadcrumbs, faggots, rissoles, haggis, breaded ham
Milk and milk products Fresh, UHT, dried, condensed, evaporated, goat’s, sheep’s milk, fresh and soured cream, buttermilk, cr`eme frˆ aiche
Milk with added fibre, artificial cream, yogurt and fromage frais containing muesli or cereals
Coffee and tea whiteners, oat milk, flavoured yoghurt and fromage frais
Fats and oils Butter, margarine, lard, cooking oils, ghee, low-fat spread Fruits and vegetables All fresh, frozen, canned and dried pure fruits and vegetables Savoury snacks Plain potato crisps, homemade popcorn Soups, sauces and pickles Tomato and garlic puree, individual herbs and spices, vinegars, mixed herbs and spices, ground pepper,
Suet
Oven, microwave and frozen chips, instant mash, fruit pie fillings, waffles
Vegetables and potatoes in batter, breadcrumbs or flour, potato croquettes
Flavoured crisps
Snacks made from wheat, rye, barley and oats, pretzels
Gravy, stock cubes, soups, Shoyu (Chinese soy sauce), sauces, mixes, tamari, mustard, stuffing mix mayonnaise, salad cream, dressings, pickles, chutney, blended seasoning, curry powder
Preserves and spreads Jam, conserves, honey, golden Mincemeat, lemon curd syrup, treacle, marmalade, peanut and other nut butters Drinks Tea, coffee, fruit juice, squash, clear fizzy drinks, cocoa, wine, spirits, cider, sherry, port Miscellaneous Gelatine, bicarbonate of soda, cream of tartar, yeast, artificial sweeteners,
Drinking chocolate, cloudy drinks
Malted milk drinks, barley waters/squash, beer, lager, ales, stouts
Tofu, cake decorations, marzipan, baking powder, ready to use icings
Ice cream cones and wafers
74
Management of Food Allergens
compulsory, and is used to highlight the presence of gluten, i.e. contains ‘wheat gluten’. The guidance and legislation regarding identification of allergens in food products is explained in more detail in Chapter 13. The following provides a basic list of those ingredients that are gluten free, and those that are not. The source of gluten-free ingredients in food production is critical. Flours are high-risk ingredients regarding contaminants.
4.8.1
r r r r r r r r
Maize starch Modified starch Modified maize starch Maltodextrin Glucose syrup Dextrose Monosodium glutamate Fructose syrup
4.8.2
r r r r r r r r r r r r
Gluten-free ingredients
Not gluten-free ingredients
Wheat starch Modified wheat starch Wheat flour Wheat rusk/wheat bran Barley flour Barley malt Rye flour Oat bran Bulgar wheat Couscous Spelt Semolina
4.9 FOOD PRODUCTION In all food production, it is necessary to apply a system of quality control with Hazard Analysis and Critical Control Point (HACCP). The FSA has produced guidance on allergen management and awareness of gluten-containing ingredients. This incorporates a set of principles to pre-empt and apply control measures to potential hazards throughout production, processing, manufacture and distribution (FSA, 2006).
4.10 THE CODEX STANDARD The first development regarding a standard for gluten-free food was in 1981 when the WHO, FAO, Codex Alimentarius set a standard for foods labelled gluten free of 0.05 g nitrogen
Coeliac disease: allergy or intolerance?
75
per 100 g of dry food (Codex Alimentarius Commission, 1981). This was interpreted as 200 parts per million (ppm) for ‘gluten-free’ labelling purposes. This standard has enabled manufacturers to produce gluten-free substitute products that contain a specially washed ingredient derived from wheat starch (Codex wheat starch) which improves the texture and palatability of the products. However, there has never been universal acceptance of this standard and in some countries, such as the USA, Canada, Australia and some countries in Southern Europe, products containing Codex wheat starch were not adopted. In these countries, only naturally gluten-free ingredients have been permitted in gluten-free substitute products and a standard of less than 20 ppm has been applied. Recent debate has produced a new draft dual standard known as the Codex standard for Foods for Special Dietary Use for Persons Intolerant to Gluten (WHO/FAO) which lowers the standard for gluten-free labelling purposes to less than 20 ppm. This development has occurred in the light of new research showing gut damage at very low levels of gluten ingestion and the knowledge that there is a greater risk of exceeding the safe level of intake the higher the standard, due to an additive effect of consuming gluten-free substitute products (Catassi et al., 2007). People with CD eat different amounts of substitute products, like gluten-free bread and gluten-free flour, so the suitability of products containing Codex wheat starch or those that are naturally gluten free is managed on an individual basis (Gubert et al., 2006). According to the proposed standard, substitute staple products containing Codex wheat starch may be considered suitable for people with CD if they have a level of gluten between 20 and 100 ppm. However, these products cannot be labelled gluten free and precise terminology for labelling purposes is still to be decided at this point in time at European Union level and implemented at a national level.
4.11
GLUTEN TESTING
Another important development associated with the standard for gluten-free labelling purposes has been in the science behind the methods of detection of gluten in foods below the standard of 200 ppm. The Codex Committee on Methods of Analysis and Sampling (CCMAS) approved the enzyme-linked immunoassay (ELISA) R5 Mendez method as a type I method in 2006. This method detects gluten reliably at a level as low as 5 ppm. It is considered to be the most accurate method currently available. As a type 1 method, it is the recommended method to be used in sampling and detection of gluten in foods (CAC, 2006).
4.12
GLUTEN-FREE CATERING
Careful checking of ingredients is essential. It is also important to make sure that all prepackaged ingredient information is retained. Gluten-free options or choices can be incorporated into menus with some additional care in menu planning and preparation (see Table 4.1). Soy sauce, mustard, sauces, spice mixes, curry powder, mayonnaise and stock cubes may all contain gluten. Table 4.2 provides a useful guide to gluten-free catering alternatives.
76
Management of Food Allergens
Table 4.2
Guide to gluten-free catering alternatives.
Food
Alternative
Soup containing barley, pasta or wheat Sauces containing wheat flour Chips fried in oil with wheat battered products Stock cubes containing wheat
Use potato or other root vegetables to thicken Use cornflour Use a separate fryer just for chips Make stock from fresh meat, poultry or vegetable products or select those which happen to be gluten free
4.13 CROSS-CONTAMINATION Complete avoidance of gluten can be a challenge for individuals with CD trying to ‘eat out’. Dry gluten-containing ingredients like flour and breadcrumbs are high-risk ingredients for contamination and cross-contamination when you are producing gluten-free meals. Therefore, careful sourcing of gluten-free flours and substitute products is necessary if you intend to use them. Since flour can easily travel from one surface to another, it is essential to consider best practice guidance to ensure that gluten is not passed between surfaces or utensils. Steps to avoid contamination include:
r r r r r r
Cleaning surfaces immediately before their use (because flour can take hours to settle and can subsequently contaminate surfaces and utensils). Use clean frying oil for chips and gluten-free foods – DO NOT reuse oil that has cooked breaded or battered products. Keep all pans, utensils and colanders separate during preparation and cooking and wash in-between use. Using a clean grill, separate toaster or toaster bags to make gluten-free toast. Use separate butter or spreads to prevent contamination between gluten-free and glutencontaining foods. Use squeezing bottles to avoid contamination through the dipping of spoons or knives.
4.14 NUTRITIONAL ADEQUACY CD is a disease of malabsorption and therefore optimal nutrition is imperative. There may even be a case for increasing recommendations for specific nutrients in appropriate foods. There are reports of deficiencies in iron, fibre, folate, thiamine, riboflavin and niacin compared to standard diets. In addition, there are particular circumstances whereby people with CD may be more at risk of nutritional deficiencies, as symptoms of intestinal damage or noncompliance are often unrecognised, resulting in associated anaemias, including iron, folate and B12 vitamin deficiency (Thompson, 2005; Silano et al., 2007). As diagnosis commonly occurs later in life and is associated with the concomitant diagnosis of osteoporosis, an increased calcium intake is recommended (BSG, 2007). Staple gluten-free substitute products are classed as ‘foods for particular nutritional uses’ and they are also included in allergen labelling (FSA, 2006). However, in legislative terms, there is no equal guidance on fortification for gluten-free flours compared to wheat flour. This means that gluten-free substitute products do not have an equivalent nutritional composition to standard wheat-containing products. In the absence of legislation, gluten-free
Coeliac disease: allergy or intolerance?
77
product ranges may vary in nutritional composition with only a few products having added calcium, iron and B vitamins. Therefore, consideration of current fortification policies on the nutritional status of gluten-free substitute products is required.
4.15 LACTOSE INTOLERANCE Lactose is a sugar found in mammalian milk (human, cows, sheep and goats), not in soya or rice milk. The enzyme lactase, which breaks down lactose, is found in the lining of the villi. When people are first diagnosed with CD, the lining of the gut is damaged, which can mean that not enough lactase is produced or that the enzyme does not work effectively. Once established on a gluten-free diet, the gut is able to heal and lactose digestion returns to normal. Lactose intolerance is therefore usually temporary (Murphy et al., 2002; Ojetti et al., 2005). However, although in reality levels of tolerance to lactose vary, catering for a lactose-free diet means avoiding all mammalian milks and their products (e.g. cow’s, goat’s and sheep’s milk, cream, ice cream, whey and evaporated milk).
4.16 COELIAC UK Coeliac UK is the leading charity in the UK working to improve the lives of people with CD and dermatitis herpetiformis (DH). Our mission is to improve the lives of people living with the condition through support, campaigning and research. Our vision is that the needs of people with CD and DH are universally recognised and met. We work to achieve this vision by:
r r r r
providing expert and independent information to help people manage their health, campaigning on their behalf to improve access to fast diagnosis, good subsequent, improving health care and safe foods – in the home and out and researching new treatments and the possibilities of a cure.
Each year, we raise awareness and campaign for change. This year, we had an awareness week in May to ensure eating out, whether through need or pleasure, is freed from restrictions. Our Awareness Week campaign, ‘Food Without Fear’ aims to raise awareness and understanding of a gluten-free diet amongst chefs and caterers, health care professionals and hospital ward staff, parliamentarians and the general public to ensure that eating out, through pleasure or need, is freed from restrictions. We provide a wide range of services including a dedicated helpline for members, health care professionals, manufacturers, suppliers and caterers, Crossed Grain magazine, annual Food and Drink Directory, which is updated monthly online at www.coeliac.org.uk. For more information, please contact the Coeliac UK helpline on 0870 4448804 or customer services on 01494 437278.
REFERENCES Bingley P.J., Williams A.J. and Norcross A.J. et al. (2004) Longitudinal studies of parents and children study team. British Medical Journal, 328, 322–323.
78
Management of Food Allergens
BSG (British Society of Gastroenterology) (2002) Guidelines for the Management of Patients with Coeliac Disease. BSG, London. Available online at www.bsg.org.uk, accessed 20 January 2008. BSG (British Society of Gastroenterology) (2007) Guidelines for Osteoporosis in Inflammatory Bowel Disease and Coeliac Disease. BSG, London. Catassi C., Fabiani E., Iacono G. et al. (2007) A prospective, double-blind, placebo-controlled trial to establish a safe gluten threshold for patients with coeliac disease. American Journal of Clinical Nutrition, 85, 160– 166. CREST (Clinical Resource Efficiency Support Team) (2006) Guidelines for the Diagnosis and Management of Coeliac Disease in Adults. CREST, Belfast. Available online at www.crestni.org.uk, accessed 20 January 2008. Codex Alimentarius Commission (1981) Standard for Gluten-Free Foods. Codex Standard 118 – 1981. Rome, CAC. Available online at http://www.codexalimentarius.net/web/standard list.jsp, accessed 12 January 2008. Codex Alimentarius Commission (2006) Report of the Twenty Seventh Session of the Codex Committee on Methods of Analysis and Sampling. CAC, Budapest. Codex Alimentarius Commission (CAC) (2007) Report of the 29th Session of the Codex Committee on Nutrition and Food for Special Dietary Uses (ALINORM 08/31/26). CAC, Rome. Collin P., Reunala T., Pukkala E. et al. (1994) Coeliac disease – associated disorders and survival. Gut, 35(9), 1215–1218. Cuoco L., Certo M., Jorizzo R.A. et al. (1999) Prevalence and early diagnosis of celiac disease in autoimmune thyroid disorders. Italian Journal of Gastroenterology and Hepatology, 31(4), 283–287. European Commission (EC) (2007) COMMISSION DIRECTIVE 2003/89/EC of the European Parliament and of the Council 10 November 2003 amending Directive 2000/13/EC. Official Journal of the European Union, L308, 15–18. European Commission (EC) (2007) COMMISSION DIRECTIVE 2005/26/EC of 21 March 2005 establishing a list of food ingredients or substances provisionally excluded from Annex IIIa of Directive 2000/13/EC of the European Parliament of the Council. Official Journal of the European Union, L75, 33–34. European Commission (EC) (2007) COMMISSION DIRECTIVE 2007/68/EC of 27 November 2007 amending Annex IIIa to Directive 2000/13/EC of the European Parliament and of the Council as regards certain food ingredients. Official Journal of the European Union, L310, 11–14. FAO/WHO (Food and Agricultural Organisation/World Health Organisation) (2008) Codex Alimentarius. FAO/WHO, Italy. Available online at http://www.codexalimentarius.net/, accessed 20 March 2008. FSA (Food Standards Agency) (2006) Guidance on Allergen Management and Consumer Information. Best Practice Guidance on Managing Food Allergens with Particular Reference to Avoiding CrossContamination and Using Appropriate Advisory Labelling (e.g. ‘May Contain’ Labelling). FSA, London. Gubert A., Espadlaler M., Angel Canela M. et al. (2006) Consumption of gluten-free products: should the threshold value for trace amounts of gluten be at 20, 100 or 200 ppm. European Journal of Gastroenterology and Hepatology, 18(11), 1187–1195. Haboubi N.Y., Taylor S. and Jones S. (2006) Coeliac disease and oats: a systematic review. Postgraduate Medical Journal, 82(972), 672–678. Holmes G. (2001) Coeliac disease and Type 1 DM – the case for screening. Diabetic Medicine, 18, 169– 177. Kemppainen T., Kroger H., Janatuinen E. et al. (1999) Osteoporosis in adult patients with celiac disease. Bone, 24, 249–255. Lundin K.E.A., Nilsen E.M., Scott H.G. et al. (2003) Oats induced villous atrophy in coeliac disease. Gut, 52, 1649–1652. Murphy M.S., Sood M. and Johnson T. (2002) Use of the lactose H2 breath test to monitor mucosal healing in coeliac disease. Acta Paediatrica, 91(2), 141–144. NHS (National Health Service) (2007) The Electronic Drug Tariff. NHS, London. Available online at http://www.ppa.org.uk/ppa/edt intro.htm">http://www.ppa.org.uk/ppa/edt intro.htm, accessed 4 March 2008. Ojetti V., Nucera G., Migneco A., Mauricio G., Lauritano C., Silvio D., Zocco M.A., Nista E.C., Cammarota G. and De Lorenzo A. (2005) High prevalence of celiac disease in patients with lactose intolerance. Digestion, 71(2), 106–110.
Coeliac disease: allergy or intolerance?
79
PCSG (Primary Care Society for Gastroenterology) (2006) The Management of Adults with Coeliac Disease in Primary Care. PCSG, Rickmansworth. Sanders D.S. (2003) Coeliac disease and subfertility: association is often neglected. British Medical Journal, 327(7425), 226–227. Silano M., Dessi M., De Vincenzi M. et al. (2007) In vitro tests indicate that certain varieties of oats may be harmful to patients with celiac disease. Journal of Gastroenterology and Hepatology, 22(4), 528–531. Thompson T. (2003) Oats and the gluten-free diet. Journal of the American Dietetic Association, 103(3), 376–379. Thompson T. (2005) Contaminated oats and other gluten-free foods in the United States. Journal of American Dietetic Association, 105, 348. van Heel D. and West J., Logan R.F., Hill P.G. (2006) Recent advances in coeliac disease. Gut, 55, 1037–1046. West J., Logan R.F., Hill P.G. et al. (2003) Seroprevalence, correlates, and characteristics of undetected coeliac disease in England. Gut, 52, 960–965.
Part II
Risk Management
5
Risk management – the principles
Ren´e Crevel
5.1 INTRODUCTION Food allergy has been long recognised as a clinical phenomenon, with numerous reports in the twentieth-century medical literature (Prausnitz and K¨ustner, 1921; Tuft and Blumstein, 1942; Loveless, 1950). Epidemiological studies published in the last 10–15 years indicate that significant proportions of the population of most countries surveyed suffer from food allergy. Early studies by Young et al. (1994) in the United Kingdom estimated a population prevalence of 1.4–1.9%, while more recent studies suggest figures of the order of 3.5–4% in the industrialised world (Young et al., 1994; Bruijnzeel-Koomen et al., 1995; Kanny et al., 2001; Sicherer et al., 2004; Pereira et al., 2005; Venter et al., 2006; Rona et al., 2007). The prevalence among children is generally accepted to be higher, of the order of up to 8% (Sampson, 2005). Such figures equate to upwards of 10 million people with the condition in regions such as the EU or USA, and therefore a significant burden of ill health and reduced quality of life. Avoidance of the offending allergen is currently the only treatment for food allergy and an immediate consequence is that having a member with food allergy affects the whole family and, to a lesser extent, the social circle of the allergic individual. Over the last two decades, food allergy has thus evolved from a problem for the food-allergic individual to one of significant public health importance. This recognition has led to initiatives by public authorities and subsequently legislators. Stages in this development include the proposal by the Nordic countries in the early 1990s to amend the Codex Alimentarius’s 25% rule, leading to the FAO–WHO consultation in 1995 which identified eight major foods or food groups as important causes of food allergy (FAO/WHO, 1997). These early stages were followed in the early years of the twenty-first century by legislation in many countries and regions, including Australia, New Zealand, Japan, the European Union (Directive 2003/89/EC), the USA (FALCPA, 2004) and Canada, while proposals have been formulated in South Africa. In parallel, many food-manufacturing companies recognised the importance of food allergy and their responsibilities to their foodallergic consumers. Experience in the food industry, now going back a decade or longer, has demonstrated that allergen management shares many of the principles underlying management of other risks, such as toxicological or microbiological ones. In particular, safety cannot be ensured through end product batch testing, but must be built upon the careful analysis of the hazards at each stage of the product cycle. A key principle is therefore the development of an integrated approach that considers each stage of manufacturing, from the raw materials to the product that is delivered to the consumer, and involves all levels within a food company from the shop floor to the boardroom, as reflected in recently published guidelines (DG
84
Management of Food Allergens
Table 5.1
Sources of allergen management issues (from FDA–CFSAN 2004).
Issue
Details
Raw materials Manufacturing processes and procedures Labelling or packaging Equipment design Training
Adventitious presence of allergen, rework Lack of cleaning, physical separation, carryover Lack of labelling review process Old equipment, difficult to clean and inspect Insufficient and not including appropriate people
Sanco HACCP guidance 2005; ISO 22000:2005). However, responsibility for ensuring the food safety of allergic consumers does not (and cannot) rest solely with the food industry, but is a responsibility shared by other stakeholders, including health professionals, regulators and the allergic patients or their carers themselves. This chapter considers the main issues associated with, the principles underlying allergen management, its objectives, and how these have been applied in a large food-manufacturing company.
5.2
ALLERGEN MANAGEMENT: THE ISSUES
The increasing number of recalls triggered by regulatory inspections (Vierk et al., 2002), by consumer complaints (Hefle and Lambrecht, 2004) or by system failures, which result in the presence of undeclared allergen at hazardous levels, indicate that unresolved issues exist with allergen management systems. A report commissioned by the US FDA as part of a review of good manufacturing practices (GMPs) in the food industry summarised issues arising from the operation of allergen management systems (FDA–CFSAN, 2004). These issues spanned the whole manufacturing supply chain (see Table 5.1).
5.2.1 Raw materials Adventitious presence of allergen was a problem, with incomplete knowledge of the formulation of ingredients used as raw materials, and the use of reconditioned ingredients and raw materials. Suggestions for overcoming these issues included developing a close working relationship with suppliers and performing regular audits. Training for the supplier could also be considered to improve understanding of the issues. Rework, which could be classed as a raw material, should be handled according to a defined and documented plan.
5.2.2 Manufacturing processes and procedures The presence of undeclared allergen in a product was found to result from a multiplicity of sources during manufacture, including inadequate cleaning, lack of physical separation at crossover points in production lines, lack of separation between production runs, carryover of allergen from shared storage equipment or from maintenance tools. The report recommended that allergens should be included in an overall hazard analysis, using a system such as the hazard analysis critical control point (HACCP) approach or equivalent. Specific recommendations included cleaning to an appropriate degree of thoroughness, recognising that this may exceed what is required to assure microbiological safety. If appropriate, cleaning
Risk management – the principles
85
should include disassembly of the equipment. Cleaning protocols may need to be validated analytically. Other measures included production scheduling, building in physical barriers to prevent allergen ingress at crossover points and providing dedicated storage facilities and maintenance equipment for specific allergenic materials.
5.2.3
Labelling or packaging
The report found that many companies failed to review their labelling, as a result of which ingredient declaration could be inaccurate. Recommendations included regular label reviews, together with policy to ensure old labels cannot be reused.
5.2.4
Equipment design
The report identified old equipment as an important issue, either because it is difficult to clean, for instance because it cannot be dismantled, or because it lends itself to allergen accumulation which cannot be readily detected by inspection. Hygienic design should overcome much of this problem as old plant is renewed.
5.2.5 Effective training of employees The report identified training as one of the areas where effectiveness was perceived to be lowest. Issues included training of the wrong people, not training enough people and not providing enough training. A key point made was that training should be delivered by people familiar with the plant rather than by consultants.
5.3
DEVELOPMENT OF ALLERGEN MANAGEMENT PLANS: PRINCIPLES AND CONSIDERATIONS
Allergen management systems form part of a family of food safety management systems, which include general toxicological and microbiological food safety. The requirements of allergen management may sometimes be at variance with those needed to control other food safety hazards. For instance, wet cleaning of dry mix lines would be a very effective way of removing allergenic residues, but at the risk of subsequent microbiological contamination, if water could not be adequately removed. Allergen management therefore needs to be integrated into the overall food safety management system, a fact now explicitly recognised in new international guidance documents such as ISO 22000:2005. Allergen management systems must meet both minimum statutory requirements and those that the company itself may have defined, after considering its objectives in this area of safety. These company requirements will arise from its own understanding of its operations and of the needs of its allergic consumers. Development of an allergen management system could proceed through the following series of steps:
r
Definition of objectives. Defining objectives forms a key stage and merits careful consideration. Without clear objectives, an allergen management plan will be difficult both to communicate to those who must apply it operationally and to other stakeholders. It will
86
r
r
Management of Food Allergens
also make it difficult to evaluate success and compliance. Absence of explicit objectives does not mean that no objectives exist, only that they are implicit. Such a situation poses a risk that objectives could be externally imposed without consideration for their achievability. Clear objectives also provide a company with a defensible position in the event of challenge. Setting out a policy. Once objectives have been defined, these can be set out in a policy which will be a statement of what the company aims are and an outline of how they are to be achieved. It can also set out broad responsibilities for ensuring that these aims are achieved. The policy enunciates the commitment of the company through its senior management. It also provides the basis for defining subsidiary objectives and mechanisms which will ensure compliance. Drawing up guidelines. As indicated above, the policy document is not a day-to-day working document. Guidelines fulfil this purpose by detailing mechanisms and procedures to implement the broad objectives at a practical level. Guidelines will therefore cover detailed operations at all levels of the organisation. They will be adapted to cover the diversity of the company’s operations, including raw material types and sourcing, types of operation, etc.
An allergen management system must also balance a number of other requirements, the main one being to ensure a high degree of food safety with respect to allergens without impairing other aspects of food safety, but also without putting at risk the economic viability of the manufacturer. Such systems will therefore need to take into account a number of factors, as detailed below.
5.3.1 Nature of food manufacturing operations Food manufacturing is typically a complex operation, which, for purposes of allergen management, may be divided into several stages. These include the following:
r
r
r
Raw material selection and specification: depending on the operation, raw materials range from primary, minimally processed ingredients, e.g. wheat flour, to complex formulated ingredients. The specification of these raw materials will limit what can be achieved in terms of residual allergen content. For instance, Codex Standards for wheat allow up to 3% other grains, while the Codex Standard for maize (corn) allows up to 2% (Codex Alimentarius Commission, 1995a,b). Manufacturing operations: these vary considerably in complexity, depending on the type of products that are made at the facility. However, in almost all operations, many different products with different formulations are made on the same lines, with a constant risk of inadvertent transfer of residues from the previous formulations onto the following ones. An added complexity is that some equipment can be extremely difficult to clean down between production runs because of its design or the characteristics of the product made. Thus allergen management in dry mix plants, where water or aqueous solvents cannot usually be used, poses challenges which differ from those in wet process lines. Similarly, thermal treatments of the product often alter the ease with which allergenic residues can be removed by sanitation procedures. Delivery to the consumer: most food manufacturers will deliver through retail or other outlets. In terms of allergen management, the manufacturer will need to consider how
Risk management – the principles
87
relevant information will be conveyed to the final consumer in such a way that it retains its integrity. In addition, food safety needs to be integrated into product development and innovation such that new products do not degrade inadvertently the safety of existing operations when they are introduced. This can occur, for instance, if a new product brings in an allergen which is new to a facility without consideration of additional control measures.
5.3.2
Biological and clinical characteristics of food allergens and food allergy
As with other toxicological hazards, food allergens may arise at any point in the food chain. However, they differ from most other chemical hazards as they pose a risk only to a limited and reasonably well-defined proportion of the population and are harmless to the vast majority at almost any level of intake. Known food allergens are exclusively proteins or polypeptides in foods, which provoke an immune response mediated by IgE antibodies in susceptible individuals. Allergenic proteins from foods belong to a relatively small number of protein families (Breiteneder and Mills, 2005) and tend to share certain molecular characteristics (e.g. thermal stability, resistance to pepsinolysis, possession of intramolecular disulphide bonds), although there is no single set of features which can discriminate between an allergenic and a non-allergenic protein. Food allergy, because of its nature, affects the wider household and community, modifying their food-buying habits and generally decreasing their quality of life (Gudgeon et al., 2005; Hefle et al., 2007). Allergic reactions to foods can be life threatening and can occur in some individuals in response to exposure to milligram quantities of the relevant food (Taylor et al., 2002). Avoidance of the offending allergenic food, together with the use of rescue medication, is the only therapy. Typically, the range of minimum doses which elicit reactions in foodallergic individuals varies over many orders of magnitude. For peanut, for instance, this may be at least four to five orders of magnitude among patients which have been challenged. Furthermore, the reactivity of patients varies over time, depending on extraneous factors such as environment, concurrent physical activity, etc. Although reliable information is beginning to emerge on the distribution of minimum eliciting doses within selected population groups, little information is available on how symptoms in any one individual vary with the amount of allergen to which they are exposed. Thus it is difficult to define the safety margin between the dose provoking a slight reaction, which does not pose a threat to health and a dose which provokes a severe reaction. Protection of the food-allergic consumer can thus pose difficult problems for food businesses. ILSI Europe has recently published an excellent overview of the scientific and clinical aspects of food allergy and the issues involved in its management in the form of a concise monograph (Jackson, 2003).
5.4
OBJECTIVES
Risk management of food allergens, as of other hazards, requires a clear definition of objectives. These need to recognise and balance sometimes the divergent interests of different stakeholders, a process which is often difficult to make explicit in the absence of formal
88
Management of Food Allergens
mechanisms. In this context, the concept of food safety objectives, analogous to those proposed by the ICMSF for microbiological hazards and as elaborated by the Codex Alimentarius Commission (2003), could form a useful basis for eliciting a consensus. The primary objective behind allergen risk management must be to prevent adverse reactions in food-allergic individuals while not limiting their food choices unnecessarily. However, the apparent simplicity of this statement hides a number of complex issues. Firstly, in a matter of public health such as allergen management, the interests and needs of different stakeholders, which may diverge, must be balanced. These stakeholders include not only the allergic population but also the general population, as well as the food industry and health professionals. Most common food allergens play an important nutritional role in the diet of the population as a whole, and it is not practicable, let alone desirable to eliminate them from products. Furthermore, the majority of the population is not exposed to any risk from food allergens, irrespective of the amounts to which they may be exposed. This situation differentiates food allergens from other food hazards, such as chemical or microbiological contaminants, from which no individual derives any benefit, and where the whole population is potentially at risk. This has two implications: firstly, the general population derives no direct benefit from allergen control measures, and indeed must indirectly bear their cost; secondly, because most product lines are shared, the presence of adventitious specific residual allergens will be inevitable in many products. As discussed above, the threshold of reactivity to allergenic foods ranges over many orders of magnitude, while allergic reactions to foods span a considerable range of severity, from the barely noticeable to life-threatening anaphylaxis. There is thus a small and poorly defined fraction of the allergic population who react severely to very small amounts of allergenic food. Procedures to reduce unintentional presence of allergens may not be sufficient to protect such people, and other allergen management measures will be required to protect them. These observations underline the importance of setting clear objectives for allergen risk management, and ensuring that they are clearly communicated to stakeholders. These objectives should define who the risk management activities aim to protect and against what. They can then serve to evaluate the effectiveness of measures taken to achieve them.
5.4.1
Who are we trying to protect and against what? – the concept of tolerable risk
Because of the characteristics of food allergens and food allergy discussed earlier, particularly the difficulties in defining the proportion of the population which reacts severely to small amounts of allergen, the concept of protecting the whole allergic population against all reactions, however mild, is unrealistic. Design of food allergen management plans must therefore consider what level of protection they purport to provide, or in other words what can be considered a tolerable risk. Defining the tolerable risk level involves considering the probability of an adverse event, as well as its consequences in terms of severity. The two aspects are clearly related, inasmuch as the tolerable frequency for severe adverse events will be much lower than the tolerable frequency of all adverse events, among which events of low severity will predominate. Establishing the tolerable level is a societal issue, but little explicit guidance exists in the area of food allergy itself. One approach would be to consider the disease burden attributable to the less common food allergens, which are not subject to statutory regulation. This burden of disease could be considered to be that tolerated
Risk management – the principles
89
implicitly by society. In practice, however, information is still lacking on the disease burden attributable to common, regulated allergens, let alone the less common ones, making this approach impracticable. Another approach might be to look to other situations analogous to food safety. Assuring the safety of drinking water shares characteristics with the protection of food-allergic people. In particular, any risk posed by drinking water is imposed on those who consume it and, for practical purposes, is usually impossible to avoid. The World Health Organization has debated drinking water standards and proposed draft guidelines for drinking water quality, which translate to a probability of disease for any one individual of 10−3 per year or approximately one in ten over a lifetime (WHO, 2003). This figure, of course, refers to all disease, irrespective of severity, so cannot be applied without qualification to allergic reactions to food. However, it could provide a basis for discussion among stakeholders.
5.4.2
Hazard versus risk-based approaches
The simplest way to protect allergic consumers would be to ensure that the allergenic constituents to which they are reactive are either declared on the product label in all circumstances or totally excluded. This hazard-based approach is broadly appropriate to deliberately added ingredients, and indeed has been adopted in several pieces of legislation regulating food allergens (e.g. Directive 2003/89/EC). However, food-allergic individuals are also at risk from small amounts of residual allergen inadvertently carried over from other products because of the sharing of manufacturing lines, or because of admixture into raw materials such as grains during storage and transport. The hazard-based approach focuses on the presence of the allergen and does not recognise the existence of any amount of allergenic protein below which there is virtually no risk. As a result, it provides no mechanism to decide on the magnitude of the danger that the presence of residual allergen constitutes. It is arguably therefore not the most effective way to protect the vast majority of food-allergic people while balancing the interests of different stakeholders. At the limit, it leads to declaration of the presence of allergen in amounts which present little or no risk to health. Application of such an approach thus decreases the food choices of the allergic population and increases the likelihood that they will suffer nutritional deficiencies as well as reducing their quality of life (Avery et al., 2003). When applied to cross-contact allergens, it is reflected in excessive use of precautionary (‘may contain’) labelling. While such a label on an individual product may well add to the protection of allergic individuals who might otherwise eat that product, the uncontrolled proliferation of such labels is likely to decrease it as trust in the label reduces and it is ignored (Hefle et al., 2007). The possibility also exists that some allergic people will misinterpret such a label, if they consume the product and fail to react, wrongly concluding that they are no longer allergic. The theoretical diagram (see Figure 5.1) attempts to illustrate these observations. Maximising public health outcomes thus ineluctably leads to a risk-based approach to allergen management. Risk can be defined as the probability that a hazard will become manifest, and is often expressed as a function of the intrinsic hazard and the exposure to that hazard. This definition is sometimes expanded to include severity of the resulting adverse effect. Thus, risk is defined in ISO/IEC Guide 51 as the combination of the probability of occurrence of harm and the severity of that harm (ISO 22000:2005). The key concept is therefore that of probability. In line with that concept, risk management does not seek to eliminate the risk, which is generally regarded as impossible unless there is no exposure, but to reduce the probability of harm to a level which is considered tolerable.
90
Management of Food Allergens
Probability of adverse event Frequency of labelling associated with minimum risk 0 Frequency of precautionary labelling Fig. 5.1 Relationship between the extent of precautionary labelling and risk to allergic consumers. (Reprinted from Crevel, R. (2006) Allergen management in the food industry. In: Managing Allergens in Food (eds C. Mills, H. Wichers and K. Hoffmann-Sommergruber). With permission from Woodhead Publishing Ltd, UK.)
5.4.3
Risk assessment principles for allergens in food manufacturing: do we have the tools needed, can the available data be used with those tools?
Kroes et al. (2000) acknowledge in their paper on the application of the ‘threshold of toxicological concern’ that ‘a particular challenge is the evaluation of food allergens and components causing other forms of intolerances, and how to determine the levels present and actual intakes vs. the limited knowledge of amounts needed for induction or elicitation of a response’. The authors in fact decided to exclude consideration of this issue from their paper. However, food allergens make a significant impact on public health, and allergen management is mandated under general food law. Management of the risk they pose is thus imperative. Before a risk can be managed, it must first be assessed. Both risk assessment and risk management form part of the process of risk analysis which the Codex Alimentarius Commission (2003) defines as being composed of three components:
r r r
Risk assessment: a scientifically based process consisting of hazard identification, hazard characterisation, exposure assessment and risk characterisation. Risk management: the process, distinct from risk assessment, of putting in place measures which minimise the risk, taking into account the interests of all those affected by it. Risk communication: the interactive exchange of information and opinions throughout the risk analysis process concerning hazards and risks, risk-related factors and risk perceptions, among risk assessors, risk managers, consumers, industry, the academic community and other interested parties, including the explanation of risk assessment findings and the basis of risk management decisions.
Risk assessment as a process thus begins with hazard identification, followed by hazard characterisation, exposure assessment and risk characterisation to scope the overall risk to the population (FAO/WHO, 1995, 1997). Translating this process to food allergens, the hazard itself is already defined as their intrinsic allergenicity, in this context their ability to provoke an allergic reaction. Beyond this, risk assessors require information about the characteristics of the hazard, which in this context can mean the response characteristics of the population
Risk management – the principles
91
at risk (distribution of minimum eliciting doses – thresholds, dose–response), the size of the population at risk and the extent of exposure. In ideal circumstances, risk assessors could calculate the number of reactions that would occur for any given level of residual allergen in a food product if people allergic to that food consumed it. In practice, information on all the elements required for a risk assessment remains extremely limited. Indeed, important reports from both the EFSA (2004) and the US FDA (Threshold Working Group, 2008) have questioned whether adequate data currently exist to establish thresholds, one of the key measures characterising the risk from food allergens. Studies to describe the reactivity of the allergic population in terms of minimum eliciting doses (thresholds) are still largely limited to selected populations, such as clinic patients, and further work is needed to map their reactivity onto the general allergic population. It is also limited to certain allergenic foods. Most available information is still based on studies undertaken for other purposes, further limiting their value for establishing minimum eliciting doses because of the type of data generated. Only anecdotal information is available on the relationship between the dose of allergen to which an individual is exposed and the severity of the reaction experienced, as obtaining that information systematically presents considerable ethical difficulties. Further complexities arise from differences in the extent to which allergenic proteins are released from different food matrices, and become available to be recognised by the immune system. Uncertainty also exists over the size of the population at risk, given that epidemiological data are scarce for most regions. In most countries, there is also a surprising lack of detailed information about the incidence of reactions, even severe and fatal ones. Minimum eliciting doses (thresholds) clearly form a key piece of information for risk assessors. A recent report by FDA–CFSAN (Threshold Working Group, 2008) identified four ways in which they might be determined: analytically based, statutorily derived, using a safety assessment approach and using a risk assessment approach. They indicated a preference for the risk assessment approach on the grounds that it was the most scientifically robust, as well as transparent, an important consideration in communicating with other stakeholders. Of those four approaches, only the safety and risk assessment ones use biological data on the reactivity of allergic patients, generated under controlled conditions. The safety assessment approach applies an uncertainty factor either to the highest dose of allergenic protein observed not to provoke a reaction or to the lowest one which does provoke a reaction (depending on what is available). In recent years, well-designed clinical studies have begun to define the distribution of the minimum doses of allergen that elicit responses in members of the allergic population. These new data have stimulated the development of new approaches (Bindslev-Jensen et al., 2002; Crevel et al., 2007) to overcome this problem. These approaches use statistical modelling of the population distribution of minimum eliciting doses (thresholds) to characterise the allergenic hazard. Coupled with estimates of exposure to the relevant allergen and knowledge of prevalence, these dose-distribution models can generate quantitative estimates of risk, helping to prioritise allergen management measures (Spanjersberg et al., 2007; Kruizinga et al., 2008). Recently, we described in detail how such an approach might be used as well as its limitations (Crevel et al., 2007). The approach is illustrated conceptually in Figure 5.2 and takes the following steps: 1. Good quality data from well-controlled challenge studies constitute the foundation of the approach. Such studies should be conducted using the agreed protocol proposed by Taylor et al. (2004) or the EuroPrevall partnership (Crevel et al., 2008). 2. A tolerable limit for the proportion (p) of allergic individuals who might react is established for the allergens of interest. The level p is defined in consultation with other stakeholders
92
Management of Food Allergens
(regulators, patients, industry, clinicians). This part of the process requires other factors to be accounted for such as the prevalence of allergy to the allergen of interest, the actual rate of reported reactions and their severity. 3. Data from challenge studies are analysed and a statistical distribution is fitted to the data. The model is used to predict the EDp and the lower confidence interval on the EDp for the p defined in step 2. Precedents from other areas of food safety support this approach, although there are distinct differences, related to the type of data that are available or can be generated in each case. Thus Buchanan et al. (1997) estimated a conservative dose–response relationship for listeriosis using surveillance data on the number of cases, together with food sampling data on Listeria contamination of a food responsible for most such cases. In our approach, we use statistical distributions to describe data from clinical challenge studies performed under carefully controlled conditions. In microbiological terms, such data are more akin to controlled feeding studies with defined numbers of microorganisms in human volunteers. Inevitably, the populations tested in these studies contain an element of bias. Firstly, they will tend to exclude individuals who have suffered life-threatening reactions, as these may be less willing to participate and the clinicians more reluctant to accept them in a study on account of the risks. Perhaps more significantly, the process of selecting for challenge studies will result in a bias towards the more severely affected of the allergic population, other than the category previously mentioned. This arises because volunteers are chosen from the population which attends tertiary referral clinics, and who are therefore sufficiently motivated by the impact of their condition to visit a clinical expert. Publication bias will also occur in the case of analyses based on the published literature, which will tend to focus on the more interesting challenge responses. Limitations also exist when using such data as a basis for risk management, because they do not take into account some of the factors which modulate the allergic response, such as alcohol intake, exercise, etc. However, it can be argued that compared to other areas of toxicology, these data present the major advantage of being generated in the species of interest. Currently, available data do not permit a direct Proportion of reactions (in clinical study)
Extrapolation
Experimental range
100%
50%
10% 5%
ED5 ED10
Fig. 5.2
ED50
Dose (mg protein)
Illustration of the characterisation of food allergen risks using the population dose distribution.
Risk management – the principles
93
application of the methods proposed in Buchanan et al. (1997). Specifically, data on the incidence of even severe allergic reactions to foods are not reliable, and usable published data on the distribution of undeclared allergen in food are very limited and do not anyway purport to provide a statistically representative picture of the extent of residual allergens in foods (Hefle and Lambrecht, 2004; Pele et al., 2007). Nevertheless, the next step in the development of this approach is to validate the predictions by comparing the number of predicted reactions with their actual incidence. Risk assessment could provide a sound scientific basis to deal with unintentional allergenic ‘cross-contact’ through the definition of an appropriate upper limit for non-ingredient allergenic food components. For example, Switzerland requires the declaration of specified allergenic constituents whenever present in concentrations greater than 0.1% (1000 mg/kg), whether as ingredients or otherwise (Conseil f´ed´eral suisse, 2002). However, a threshold of 0.1% would almost certainly not be considered sufficiently protective of public health, since less than 1 mg of peanut protein has been shown to elicit adverse reactions in allergic subjects (reviewed in Taylor et al., 2002; Threshold Working Group, 2008). Obviously, an upper limit for non-ingredient allergenic food components also needs to consider the No Observed Adverse Effect Level (NOAEL) reported for each of the important allergenic foods. Thus, while allergic consumers are protected against undeclared allergenic ingredients, they remain at risk from non-ingredient allergenic components, while the food industry lacks clear guidance from regulatory authorities. This absence of a regulatory threshold also has other consequences, which may reduce the protection afforded the allergic consumer. Allergenic ingredients present in insignificant quantities must be declared. For example, an ingredient containing refined peanut oil, with almost undetectable protein, could be a component of another ingredient used in very small amounts, such as a flavour. Yet products containing this ingredient must be labelled as containing peanut and allergic consumers who eat them could erroneously conclude that their allergy had resolved when in fact the amounts were too small to trigger a reaction.
5.5
APPLICATION
5.5.1 From policy to guidelines: the need for an integrated approach Allergen management policies establish principally intent, but guidelines are required to set out the detailed mechanisms whereby this intent is made operational. As already mentioned, allergen management requires consideration of all the elements of the life cycle and production of a product from its design, through selection of raw materials to production processes and finally delivery to the consumer. This marks out allergen management as an activity which requires commitment throughout a company from senior management all the way to those working on production lines. Procedures that are put in place for allergen management may also impinge on other aspects of food safety. Effective delivery of such commitment therefore requires that allergen management be integrated into overall food safety management and encompass all stages of the life cycle of the product from the raw materials to the end product which the consumer buys. This has recently been recognised in external standards for food safety management systems such as ISO 22000:2005 and the EU’s DG Sanco HACCP guidance 2005. As already discussed, this integrated approach starts with the development of policies, which provide the framework for the operation of the allergen
94
Management of Food Allergens
management plans. Typically, such policies will specify the aims of the plan, indicate what will be done to achieve those aims, and define major responsibilities for their fulfilment. Policies thus enshrine the corporate commitment to the plan. However, implementation requires more detailed guidance documents, which can be tailored to specific operational requirements and provide practical advice to individual manufacturing units. Guidelines will therefore differentiate between different types of food processing, such as dry mix lines and wet lines, but will also cover elements such as the cleaning of specific pieces of equipment within a factory. This approach ensures that a high minimum standard exists for the handling of allergens throughout the company. Existing food safety management systems can be helpful in applying these principles operationally.
5.5.2
Food safety management systems and food allergens
A number of formal systems have been developed to deal with food safety hazards, of which the best known are GMPs and HACCP systems. More recently, the International Standards Organization (ISO) has published ISO22000:2005, which describes a fully integrated approach to food safety management.
5.5.3 Good manufacturing practice GMP regulations (CFR 21 Pt 110, 1999) are probably the earliest food safety management systems. These evolved originally to counter chemical contamination, adulteration and misrepresentation of products to the consumer. The current content of these regulations reflect these origins, as well as the later focus on microbiological safety. Many aspects of the current GMPs relating to human food production contain elements with application to allergen management. They thus cover provisions for personnel, including health, education and training and supervision arrangements, provisions for buildings and facilities, including construction, layout, maintenance, sanitary operations and facilities, provisions for equipment design, construction and maintenance, in particular with regard to sanitation. The GMPs also cover production and process controls, including raw materials as well as manufacturing operations. Main requirements are as follows:
r
r
r
Personnel: Personnel working in food handling or processing facilities are required to be free of disease, to maintain adequate cleanliness to avoid contamination of the food, to take other necessary precautions to avoid contamination. If they are responsible for identifying food contamination, they must have an adequate level of education and training. Furthermore, responsibility for compliance must be assigned to suitably competent supervisory personnel. Buildings and facilities: Plant buildings and structures must be of sufficient size, and adequate construction, and design to facilitate maintenance and sanitary operations for food-manufacturing purposes. The guidelines lay particular emphasis on sufficient size to ensure adequate separation of operations to avoid contamination. Facilities must also have adequate arrangements for cleaning of equipment as well as for employees to clean their hands. Equipment: Equipment must be designed to be cleaned adequately and must be properly maintained.
Risk management – the principles
r
95
Production and process controls: All operations from the raw material selection and handling to the final product must be conducted in accordance with adequate sanitation principles. Again, stress is laid on the requirement for adequate supervision. The regulations require that ‘All reasonable precautions shall be taken to ensure that production procedures do not contribute contamination from any source’.
Even a cursory consideration of the requirements of current GMPs shows that although they are heavily focused on microbiological control, their principles should constitute a good starting point for allergen management. The original US regulations date from 1986, when food allergy was not deemed a significant public health problem. A report commissioned by the FDA into modernisation of the GMPs recommended revision to incorporate, among other elements, guidance on allergen control (FDA–CFSAN, 2004). This recommendation was endorsed by many of those who offered comments during the consultation phase on the report.
5.5.4 Hazard analysis and critical control points HACCP, unlike GMP, is not a system of regulations which specify what must be done. Instead, it is a set of principles which can be applied as part of an approach to food safety management. It was originally proposed by the Codex Alimentarius Commission and is now considered by various regulatory authorities (e.g. EFSA) to be an appropriate tool to control hazards in food businesses. In contrast to GMP, HACCP is process-specific, and indeed can be considered complementary to GMP. The structure of HACCP means that it can be applied to almost any hazard. HACCP consists of the following seven stages: 1. hazard analysis: identify any hazards that must be prevented, eliminated or reduced to tolerable levels, 2. identify the critical control points, 3. establish critical limits at critical control points, 4. establish and implement effective monitoring procedures at critical control points, 5. establish corrective actions when monitoring indicates that a critical control point is not under control, 6. establish procedures to verify that the measures outlined in paragraphs 1 to 5 are working effectively, 7. establish documents and records commensurate with the nature and size of the food business to demonstrate the effective application of the measures outlined in steps 1 to 6. Some regulatory authorities, rather than defining GMPs, have mandated the use of HACCP principles as part of food safety management. Thus, EU Regulation 852/2004 on the hygiene of foodstuffs requires the application of HACCP principles. However, the guidance document (EU, 2006) also stresses the flexible application of the concept, commensurate with the complexity of the business.
5.5.5 ISO 22000:2005: food safety management systems ISO 22000:2005 starts from the premise that the most effective food safety systems are established, operated and updated within the framework of a structured management system and
96
Management of Food Allergens Table 5.2 Key elements of food safety management systems according to ISO22000:2005. Element Role of management Communication Resources Documentation Procedures to deal with non-conformity Evaluation
incorporated into the overall management activities of the organisation (ISO 22000:2005). The standard aims to provide the structure for a comprehensive food safety management system, integrated into other management systems (e.g. quality management systems, such as ISO 9001:2000). Key elements (Table 5.2), which it highlights, include:
r
r r r r r
Role of management: the standard recognises the essential role of senior management commitment, demonstrated by definition of the food safety policy, its documentation and its communication within and beyond the company. It also details a requirement for operation of the policy to be reviewed periodically by senior management, and for attribution of key responsibilities. Communication: the standard lays considerable stress on communication, both within the company to promote its effective application, but also beyond to suppliers, retailers and consumers. Resources: compliance with the standard can only be achieved if adequate resources both in equipment and trained personnel are available. Documentation: the safety management system must be documented, both in terms of its different elements (e.g. HACCP plans) and evidence that it is being adhered to (e.g. audit results). Non-conformity: plans to deal with breaches of the policy (e.g. product non-conformity) need to be documented, including a system for managing recalls. Evaluation: the food safety management system must be evaluated periodically to ascertain whether it continues to cover the company’s requirements and takes into account the most recent information on the food safety hazards subject to control.
The principles of HACCP, as enunciated by the Codex Alimentarius Commission and outlined in the previous section, form an integral part of ISO 22000:2005. The standard differs from the GMP regulations inasmuch as it provides a framework for companies who do not have a food safety management system to develop and implement one. It also permits those who have already implemented such a system to review it against the standard and, if appropriate, update or add elements to it. It does not provide detailed guidance, for instance, on how to design and operate a factory to minimise the risk of allergen cross-contact. It links to GMPs through the requirement that prerequisite programmes, covering for instance hygiene, are in place prior to implementation of the system.
Risk management – the principles Table 5.3
97
Key factors covered by Unilever’s allergen guidelines.
Factor Innovation and product development Supply chain Manufacturing protocols Packaging, promotion and advertising Training Retailers Food professionals
5.5.6 Food allergen management in Unilever Unilever, in parallel with other large food companies, identified food allergy as public health issue long before it was addressed in international and regulatory documents dealing with food safety. Although its approach thus evolved separately, it contains many of the elements which have now been identified as key to food safety management. Unilever thus has a policy for dealing with allergens, which states that it shall declare the presence in its products of any allergen which is a common cause of allergic reactions. At a minimum, any allergen required by local regulations will be declared. However, beyond that, the allergenic risk from foods not considered commonly allergenic may be assessed if clinical or epidemiological data indicate the need. If classed as a common cause of allergic reactions in accordance with the company’s criteria, this food component would then be declared on labels and included in allergen management plans. Unilever also undertakes to inform any consumer on request about the presence of uncommon allergens in specific products. The custodian of this policy, as of all Unilever policies, is the Executive Committee, thereby demonstrating senior management commitment. This approach enables the company to react more rapidly to emerging allergens than if it relied solely on reacting to external regulation. The policy, however, is made operational through specific guidelines, which exist for dry mix plants, wet savoury plants, ice cream factories and other operations. Allergen management guidelines (Table 5.3) need to ensure that allergens are correctly and intelligibly declared in products, but also to make sure that allergen is not present inadvertently at levels likely to cause adverse health effects. Such guidelines need to address all stages in the product life cycle, from its design, through the sourcing of ingredients to manufacture, labelling and distribution. Specifically, they need to deal with the following:
r
r
Innovation: the product developer needs to consider whether the allergenic ingredient plays a functional role in the product or whether an equivalent non-allergenic ingredient could serve as well. Similarly, could an ingredient already present in the product replace an allergenic one; for instance, could wheat flour be used as a carrier for a flavour in a product already containing wheat? A further question that needs consideration is whether the use of the allergenic ingredient in a product would alter the risks arising from existing operations, for instance by bringing a new allergenic ingredient into a plant. Supply chain: control of allergens in the supply chain requires a close relationship with suppliers, so that they understand the needs of the manufacturer and can meet its requirements. Typically, within Unilever, the starting point of the supplier assessment will be a questionnaire about allergens handled and precautions in place to avoid cross-contact, including the existence of a HACCP plan. This is backed up by periodic audits of the
98
r
r
r r
r
Management of Food Allergens
suppliers’ facilities. Taking together, the resulting information will permit an assessment of the risk that an allergen is present inadvertently through the raw materials, and will form the basis for risk management measures, which might include a change of supplier. Additionally, where applicable, suppliers are required to seek agreement to any change in the formulation of the ingredient they supply. Manufacturing protocols: the main considerations are the inclusion of common allergens in HACCP plans, production scheduling to minimise cross-contact, validated cleaning procedures and clear labelling and separation of specific allergenic ingredients within the factory. Procedures need to cover rework, where sound product is not packaged but ‘recycled’. Finally, the same degree of attention is needed whether the company’s own manufacturing facility is concerned or that of co-packers. Packaging, promotion and advertising: packaging carries the label and therefore the allergen information. Care is required to ensure that information remains with the product until it reaches the consumer; the allergen management plan needs to assess the probability that packs containing multiple individually packed items will be split, and the outer packaging separated from the product. Other considerations include the use of warnings if the formulation has changed to include an allergenic ingredient previously not present. Training: staff at all levels need to understand the importance of allergen control procedures and their own role in ensuring compliance. Training is thus vital and improves support for what could otherwise be perceived as irksome restrictions. Retailers: generally, the manufacturer’s allergen information will be sufficient. However, situations such as in-store promotions require care to ensure that the consumer is fully informed. Sound product, which fails to meet all standards for general sale, may be repackaged and sold on in specialised outlets or even in a different market. The manufacturer needs to ensure that appropriate allergen information is retained and available to the ultimate consumer. Food professionals: most allergic reactions to foods occur outside the home, in conditions where the product is often not labelled and even when asked, food professionals fail to provide correct information. Where pre-prepared food is provided to that sector, the manufacturer has a responsibility to ensure that accurate allergen information is provided and conveyed to the consumer.
This brief summary of the elements of allergen management in Unilever illustrates the close parallels between the evolution of thinking within the company and the emerging consensus in the food safety management community. Although implementation in Unilever has characteristics specific to the company, with the key elements of a system integrated with other safety management systems, and the ultimate responsibility lying with the senior leadership of the company, it is more striking by its similarities to that consensus than by its differences from it.
5.6 CONCLUDING REMARKS Food allergy has emerged over the last two decades as a fully fledged public health issue, which it became imperative for food manufacturers to address. Food allergens possess some very distinct characteristics compared to other food safety risks, not least that they affect a small but significant proportion of the population, that many are used in
Risk management – the principles
99
a multitude of applications and that small amounts can sometimes provoke severe reactions. However, experience and research have shown that the risks posed by food allergens could be handled using methodologies developed in the context of other hazards. Considerable progress has been made in assessing the risks from food allergens, but confidence in the reliability of such risk assessments among a significant proportion of stakeholders continues to be limited by the scarcity of data, as well as by difficulties in evaluating its quality and interpreting it. Notwithstanding these limitations, the food industry has faced the need to protect its allergic consumers, which has led to the development of allergen management systems. A key observation arising from this experience is that optimal outcomes require consideration of the whole life cycle of a product, which can best be achieved through full integration of allergen risk management with management of other food safety risks. Newly developed standards, such as ISO 22000:2005, build on this crucial observation. Development of allergen management policies and procedures by food manufacturers contributes significantly to protection of allergic consumers by ensuring that the inadvertent presence of allergenic ingredients in products is minimised.
REFERENCES Avery N.J., King R.M., Knight S. and Hourihane J.O. (2003) Assessment of quality of life in children with peanut allergy. Pediatric Allergy and Immunology, 14, 378–382. Bindslev-Jensen C., Briggs D. and Osterballe M. (2002) Can we determine a threshold level for allergenic foods by statistical analysis of published data in the literature? Allergy, 57, 741–746. Breiteneder H. and Mills E.N. (2005) Molecular properties of food allergens. Journal of Allergy and Clinical Immunology, 115(1), 14–23. Bruijnzeel-Koomen C., Ortolani C., Aas K. et al. (1995) Adverse reactions to food. European academy of allergology and clinical immunology subcommittee. Allergy, 50, 623–635. Buchanan R.L., Damert W.G., Whiting R.C. and Van Schothorst M. (1997) Use of epidemiologic and food survey data to estimate a purposefully conservative dose–response relationship for Listeria monocytogenes levels and incidence of listeriosis. Journal of Food Protection, 60(8), 918–922. Codex Alimentarius Commission (1995a) Codex Standard For Maize (Corn). Codex Stan 153 – 1985 (Rev. 1 - 1995). Codex Alimentarius Commission (1995b) Codex Standard For Wheat And Durum Wheat. Codex Stan 199 – 1995. Codex Alimentarius Commission (2003) Draft Principles for the Risk Analysis of Foods Derived from Modern Biotechnology, at Step 8 of the Elaboration Procedure. ALINORM 03/34, Appendix II. Codex Alimentarius Commission, Joint FAO/WHO Food Standards Programme, Food and Agriculture Organisation, Rome. Available at ftp://ftp.fao.org/codex/alinorm01/al0134ae.pdf, accessed 21 April 2009. Conseil f´ed´eral suisse (2002) Ordonnance sur les denr´ees alimentaires. RO, 573. Crevel R.W., Ballmer-Weber B.K., Holzhauser T. et al. (2008) Thresholds for food allergens and their value to different stakeholders. Allergy, 63(5), 597–609. Crevel R.W.R., Briggs D., Hefle S.L. et al. (2007) Hazard characterisation in food allergen risk assessment: the application of statistical approaches and the use of clinical data. Food and Chemical Toxicology, 45(5), 691–701. Directive 2003/89/EC of the European Parliament and Council (2003) Official Journal of European Union L308/15, 25.11.2003. Available at http://eur-lex.europa.eu/LexUriServ/LexUriServ .do?uri=OJ:L:2003:308:0015:0018:EN:PDF, accessed 21 April 2009. EFSA (European Food Safety Authority) (2004) Opinion of the scientific panel on dietitic products, nutrition and allergies on a request from the commission relating to the evaluation of allergenic foods for labelling purposes. EFSA Journal, 32, 1–197.
100
Management of Food Allergens
EU (European Union) (2006) Guidance Document. Key questions related to import requirements and the new rules on food hygiene and official food controls. Available at http://ec.europa.eu/food/ international/trade/interpretation imports.pdf, accessed 21 April 2009. FAO/WHO (1995) Application of Risk Analysis to Food Standards Issues, Report of the Joint FAO/WHO Expert Consultation, Geneva, Switzerland, 13–17 March 1995, WHO/FNU/FOS/95.3. World Health Organization, Geneva. Available at http://www.who.int/fsf/Documents/applira.pdf, accessed 21 April 2009. FAO/WHO (1997) Risk Management and Food Safety, Report of a Joint FAO/WHO Consultation, Rome, Italy, 27–31 January 1997, FAO Food and Nutrition Paper 65. Food and Agriculture Organisation of the United Nations, Rome. Available at http://www.fao.org/docrep/W4982E/W4982E00.htm, accessed 21 April 2009. FDA–CFSAN (2004) Good Manufacturing Practices (GMPs) for the 21st Century – Food Processing. Available at http://www.cfsan.fda.gov/˜dms/gmp-toc.html, accessed 21 April 2009. FALCPA (Food Allergen Labeling and Consumer Protection Act) (2004) (Title II of Public Law 108–282) s.741–15-21. Available at http://www.cfsan.fda.gov/˜acrobat/alrgact.pdf, accessed 21 April 2009. Gudgeon L.A., Trewin J.B., Grimshaw K.E.C. and Hourihane J.O.B. (2005) Patients find low dose threshold challenges useful in the management of their peanut allergy. American Academy of Allergy, Asthma and Immunology, San Antonio. Hefle S.L., Furlong T.J., Niemann L. et al. (2007) Consumer attitudes and risks associated with packaged foods having advisory labeling regarding the presence of peanuts. Journal of Allergy and Clinical Immunology, 120, 171–176. Hefle S.L. and Lambrecht D.M. (2004) Validated sandwich enzyme-linked immunosorbent assay for casein and its application to retail and milk-allergic complaint foods. Journal of Food Protection, 67(9), 1933–1938. International Standard 22000 (2005) Food safety management systems – requirements for any organization in the food chain. First Edition 1 September 2005. International Standards Organization, Geneva, Switzerland. Jackson W.F. (2003) Food Allergy. ILSI Europe Concise Monograph Series. International Life Sciences Institute (ILSI), Brussels, Belgium. ISBN 1–57881-160–0. Kanny G., Moneret-Vautrin D.A., Flabbee J. et al. (2001) Population study of food allergy in France. Journal of Allergy and Clinical Immunology, 108(1), 133–140. Kroes R., Galli C., Munro I. et al. (2000) Threshold of toxicological concern for chemical substances present in the diet: a practical tool for assessing the need for toxicity testing. Food and Chemical Toxicology, 38, 255–312. Kruizinga A.G., Briggs D., Crevel R.W.R. et al. (2008) Probabilistic risk assessment model for allergens in food: sensitivity analysis of the minimum eliciting dose and food consumption. Food and Chemical Toxicology, 46(5), 1437–1443. Loveless M.H. (1950) Milk allergy: a survey of its incidence; experience with masked ingestion test. Journal of Allergy, 21, 489. Pele M., Broh´ee M., Anklam E. and Van Hengel A.J. (2007) Peanut and hazelnut traces in cookies and chocolates: relationship between analytical results and declaration of food allergens on product labels. Food Additives & Contaminants, 24, 1334–1344. Pereira P., Venter C., Grundy J. et al. (2005) Prevalence of sensitization to food allergens, reported adverse reaction to foods, food avoidance, and food hypersensitivity among teenagers. Journal of Allergy and Clinical Immunology, 116, 884–892. Prausnitz C. and K¨ustner H. (1921) Studien u¨ ber die Ueberempfindlichkeit. Zentralbl Bakteriol Abt Orig, 86, 160–169. Rona R.J., Keil T., Summers C. et al. (2007) The prevalence of food allergy: a meta-analysis. Journal of Allergy and Clinical Immunology, 120(3), 638–646. Sampson H.A. (2005) Food allergy – accurately identifying clinical reactivity. Allergy, 60(Suppl 79), 19–24. Sicherer S.H., Munoz-Furlong A. and Sampson H.A. (2004) Prevalence of seafood allergy in the United States determined by a random telephone survey. Journal of Allergy and Clinical Immunology, 114(1), 127–130. Spanjersberg M.Q., Kruizinga A.G., Rennen M.A. and Houben G.F. (2007) Risk assessment and food allergy: the probabilistic model applied to allergens. Food and Chemical Toxicology, 45(1), 49–54.
Risk management – the principles
101
Taylor S.L., Hefle S.L., Bindslev-Jensen C. et al. (2004) A consensus protocol for the determination of the threshold doses for allergenic foods: how much is too much? Clinical and Experimental Allergy, 34(5), 689–695. Taylor S.L., Hefle S.L., Bindslev-Jensen C. et al. (2002) Factors affecting the determination of threshold doses for allergenic foods: how much is too much? Journal of Allergy and Clinical Immunology, 109, 24–30. Threshold Working Group (2008) Approaches to establish thresholds for major food allergens and for gluten in food. Journal of Food Protection, 71(46), 1043–1088. Tuft L. and Blumstein G.I. (1942) Studies in food allergy. Sensitization to fresh fruits: clinical and experimental observations. Journal of Allergy, 13, 574–581. Venter C., Pereira B., Grundy J. et al. (2006) Prevalence of sensitization reported and objectively assessed food hypersensitivity amongst six-year-old children: a population-based study. Pediatric Allergy and Immunology, 17, 356–363. Vierk K., Falci K., Wolyniak C. and Klontz K.C. (2002) Recalls of foods containing undeclared allergens reported to the US Food and Drug Administration, fiscal year 1999. Journal of Allergy and Clinical Immunology, 109(6), 920–922. WHO (2003) Guidelines for Drinking-Water Quality, 3rd edn. World Health Organization, Geneva. Young E., Stoneham M.D., Petruckevitch A. et al. (1994) A population study of food intolerance. Lancet, 343(8906), 1127–1130.
6
Risk management – operational implications
Anton J. Alldrick
6.1 INTRODUCTION Modern approaches to food safety management are often based on some form of risk evaluation and the institution of appropriate management systems to ensure that the risk (probability) of an adverse event occurring is reduced to an acceptable level. Such a process involves a number of steps. These are:
r r r
Hazard identification Estimation of risk Management of risk
Most experienced food-safety practitioners have the intellectual ability to deal with these concepts; however, experience has shown that in the case of food allergens the subject can be challenging. To get some idea of the scope of the problem in a 12-month period beginning March 2007, 78 allergen product recalls/withdrawals were notified to the UK Food Standard Agency; this compares with 26 alerts for foreign-body contamination (Food Standards Agency, www.food.gov.uk/safereating/allergyintol/alerts/, www.food.gov.uk/enforcement/alerts). Arguably, one of the reasons for this is that as a class of contaminants, food-allergens often provoke challenges to existing patterns of thinking in terms of food-safety management. Fundamentally, these arise out of the following dichotomy. On the one hand (generally in the case of most food-safety hazards):
r r
most hazards represent some form of threat to the entire population; any adverse reaction can, in many cases, be related to a minimum eliciting dose.
On the other hand (as discussed in a recent paper published by the European Food Safety Authority (2004), in terms of food allergens):
r r r r
the eliciting agent is usually an integral component of the food or ingredient; generally speaking food allergy only affects a small proportion of the consumer base; sometimes inevitable levels of cross-contamination within a production system, while low enough not to compromise product quality, are sufficiently large enough to induce adverse reactions in susceptible individuals; there is no ‘rule of thumb’ as to what amount of a particular allergen in a particular food matrix will not induce an adverse reaction in a particular individual (in other words, there is no preset threshold value).
Risk management – operational implications
103
As discussed in this chapter, while managing food allergens may require a rigorous intellectual approach, the operational techniques used are no different from those required normally to achieve ‘good manufacturing practice (GMP)’.
6.2 IDENTIFYING THE HAZARD It can be argued that the underlying principles of modern food-safety management philosophy are twofold:
r r
The correct definition of the hazard faced by the consumer. The subsequent identification and implementation of those steps necessary to remove the hazard completely or to reduce the risk of it occurring within a food business to an acceptable level.
Thus the first step in any sensible food-safety management system is to identify the hazard which has to be managed. Given the subject matter covered in this chapter and for convenience, in the case of food-allergy within a food-processing business, the hazard might be defined as ‘The inadvertent consumption of a food allergen by a sensitive individual’. Fundamentally, therefore the risk of the hazard occurring revolves around two factors:
r r
The use (or lack thereof) of appropriate communication mechanisms whereby the foodallergic consumer can be informed of the presence of allergenic materials within a particular food. Minimising the inadvertent inclusion of food allergens in products where they would not be expected to be present.
Allergenic materials can be present in the food by virtue of the fact that either they are an ingredient or they are present as the result of cross-contamination consequential to other activities within the food business providing the food in question. In the latter case, it is reemphasised that a level of cross-contamination insufficient to provoke a consumer complaint in terms of overall product acceptability (e.g. taste or appearance) will often be high enough to provoke a response in an allergic individual. Potentially most foods can act as an allergen to at least one individual. In 1996, Hefle et al. published a list of over 150 foods demonstrated to have caused an allergic response on at least one occasion. Thus, in terms of identifying which food allergens to focus on, those responsible for food-safety management are faced with performing a triage exercise to identify those allergens which present the most significant threat to the food business’s consumer base. In the European Union, this exercise has to a large degree been achieved through labelling legislation. At the time of writing (2008), food labelling requirements (including how food ingredients may be declared) are governed by Directive 2000/13/EC as amended. A key amendment to this legislation relates to the communication of foodallergen-related information; this has progressed through various forms and is currently encapsulated in Commission Directive 2007/68/EC. This document not only details how food-allergen-related information should be transmitted but also supplies a list of allergenic foods (Annex IIIa) covered by the regulation. The British Retail Consortium (2008) in version 5 of its ‘global food standard’ identified the same list. Within the European Union, therefore,
104
Management of Food Allergens
food-safety practitioners have a basic list of food allergens which forms the basis of their hazard risk assessments. It is important to note that this list acts as a starting point. Depending on the product and consumer base other potentially allergenic foods might have to be considered. It should be remembered that a number of food businesses were considering food allergens such as lupins and molluscs within their food-safety management systems, before their subsequent inclusion within Annex IIIa. Furthermore, in order to reflect local susceptibilities, national food-safety authorities often include other foods in addition to those listed on the Annex IIIa list, when giving advice to consumers. For example, on its ‘eat well be well’ website, the UK Food Standards Agency (www.eatwell.gov.uk) lists a further nine foods or food types, which include rice and kiwi fruit. Thus in arriving at a list of allergenic foods to be managed attention needs to be paid not only to those allergens considered significant by legislation but also those that the consumer segment at which the food is targeted might be considered vulnerable to. This principle forms not only part of good food-safety management but is, in a number of jurisdictions, also underpinned by law. For example, within the European Union, Article 14 of Regulation 178/2002 (the basic food law) requires that food to be sold shall be safe (i.e. not injurious to health, nor unfit for human consumption). The article goes on to require that decisions concerning the safety of food take into account (amongst other things) the processes used to produce the food, information provided to the consumer and ‘the particular health sensitivities of a specific category of consumers where that food is intended for that category of consumers’ (Article 14.4c). A reasonable interpretation of this part of the legislation would be that unless a food was specifically labelled as being unsuitable for a person allergic to a food listed in Annex IIIa, it should be considered suitable unless appropriate precautionary labelling (e.g. ‘may contain’) was applied.
6.3 MANAGING THE HAZARD 6.3.1
Management principles
Having identified the hazard and the likely causes, the next step is to reconcile the food production process with the information that the sensitive consumer will be provided with at the point of sale. In general terms, the consumer relies on three principal aspects of the label: the product description, the ingredients declaration, and any allergen advice (this may highlight the presence of a particular allergen and/or contain precautionary (e.g. ‘may contain’) labelling discussing the circumstances under which the product was made. The first point to be made is that given their nature, food allergens can rarely, if ever, be processed out of the product. Application of modern GMP philosophy, as proposed by Codex Alimentarius (2003) would therefore indicate that successful management of the food-allergen hazard is through appropriate validated prerequisite programmes which take into account the appropriate food allergens and are verified on an appropriate basis. Another challenge to the food safety manager of any particular manufacturing facility is that while he/she can exercise some degree of management control on that facility that level of control cannot be exercised either on the suppliers of raw materials or on the distribution of the finished product. Furthermore, given the normal dynamics of business – activities from other functions within the organisation have the potential of aggravating the risk of a food-allergen-related hazard occurring. Such functions would include those responsible for raw materials procurement, new product development, sales and marketing.
Risk management – operational implications
105
Potential sources of risk aggravation Procurement, new product development, external contractors, sales and marketing
Dispatch
Packaging
Production
Receipt and store
Raw materials
Rework
Supplier quality assurance, sanitation, maintenance, production control Supporting prerequisite programmes Fig. 6.1 Schematic showing production flow through an idealised manufacturing facility (shaded area) and influence of associated management functions.
A simple schema showing the dynamics of food production and the associated management systems relating for food production is shown in Figure 6.1. This shows not only the idealised production flow within a manufacturing facility, but also the various other major activities with potential to influence the risk of the hazard occurring. Alldrick (2006) coined the acronym ‘PIPE’ (people ingredients process enforcement) to describe the underpinning approach to the successful management of the hazard presented by food allergens. In essence, PIPE requires that:
r r r r
People within a food business should understand how their own activities can impact on the risk of food-allergen-linked incidents occurring and how they can minimise that risk. This responsibility extends from the chief executive officer to the temporary menial worker and everyone else within the food business. Ingredients (raw materials) should be sourced from suppliers who can demonstrate competence in providing materials of defined food-allergen risk. Once delivered on site, systems have to be in place that ensure the integrity of the packing of high-risk materials and that these are handled and stored appropriately. This also applies to rework. Processes and supporting systems used must ensure that the risk of inadvertent foodallergen consumption is minimised. This is achieved through a number of routes including segregation of production lines, scheduling of production where segregation is infeasible and application of appropriate sanitation regimes. Enforcement mechanisms are in place. These should be designed to not only ensure compliance but also verify on a continuing basis that the food-allergen management systems in place remain fit for purpose.
To one degree or another, these principles apply throughout the process flow. The challenge faced by the food-safety practitioner is to ensure that the information presented to the
106
Management of Food Allergens
consumer on the wrapper matches the food within the wrapper. How this can be achieved is discussed in the following sections using the schematic in Figure 6.1 as a template.
6.3.2 Raw materials In some respects, the supply and sourcing of raw materials represents that part of the process over which the food-safety practitioner has least control. This is certainly the case until raw materials have been delivered to the manufacturing facility. In order to have confidence in any raw material’s integrity, there must be appropriate prerequisite programmes in place. These revolve around the concept of supplier quality assurance (SQA). In essence, SQA requires that suppliers are evaluated on a regular basis as to their competence to supply raw materials that consistently comply with specifications. The initial steps with regard to SQA often rely on a self-assessment questionnaire, depending on the overall risk presented by the raw material. This is usually followed up either by an audit of the raw material production facility by the purchaser (second party audit) or by a requirement that the facility concerned is accredited to a food-manufacturing standard (e.g. BRC Global Standard for Food Manufacture – referred to above), the so-called third party audit option. In terms of the self-assessment questionnaire, it is now usual for it to include a section concerning whether or not the raw material supplied contains a particular allergen. Two major problems with such a questionnaire lie in how the question is posed and the format it is delivered in. Frequently, the question is (apparently) simply put as: ‘Is the material free from?’ followed by a list of allergenic materials against which are placed two check boxes (either yes or no). Such an approach suffers from problems of linguistics and psychology. Dealing with the linguistics question first. In terms of the consumer to state that a product is ‘free from’, something would imply that the supplier had taken all reasonable steps to both ensure and confirm the absence of the something. However, in terms of allergen control and when used in terms of specification, a more commonly held interpretation is an application of the ‘free from’ definition used in the British Retail Consortium ‘guidelines for the handling of nuts’ British Retail Consortium (2004), viz. Free from nuts means no nuts . . . must be used in a product as any of the following: (i) An ingredient; (ii) A compound ingredient; (iii) A ‘processing aid’. It should be noted that strict application of the definition means that a raw material cross-contaminated in some way by an allergen could still be declared as being ‘free from’. Furthermore, use of this definition does not require a very high level of verification (e.g. positive release after analysis). The situation is further complicated if the English civil law ‘man on the Clapham omnibus’ principle (Greer, 1933) is applied. Thus, while a precedent has been set at a technical level (specifications), it would be unreasonable for such a definition to be inferred from a ‘free from’ endorsement placed on packaging of any raw material or food. In these cases, the inference would be that active steps had been taken to ensure the absence of the allergen in any raw material or food whose wrapper/label bore such an endorsement. A second problem relates to the psychology of completing such questionnaires. Simple reliance on tick boxes
Risk management – operational implications
107
usually means that once a person has begun entering down one column they will continue to do so even if the answer is incorrect, with the probability of error increasing in proportion to the length of the questionnaire. There are, however, routes to circumvent these challenges. The first is to refine the nature of the questions asked. Thus in addition to asking whether a product is ‘free from’ a particular allergen, questions along the lines of:
r r
Is this ingredient suitable for use in foods intended for persons with an allergy to a particular food? Is there a risk of this ingredient being contaminated with a particular allergen as a consequence of other activities undertaken by the supplier?
In order to overcome some of the psychological problems associated with simply checking boxes, questionnaires can be designed to require the respondent to answer either ‘yes’ or ‘no’ and to justify those answers. In any event, best practice in food-safety management requires that information obtained from suppliers be verified on a periodic basis. Verification can take a number of forms. In the case of food allergens, the two principal ones are chemical analysis and audit. It should be emphasised that any verification process reflects a sampling process and gives only a snapshot of what is probably happening at a defined point in time. The frequency of verification will depend on the food concerned, the extent of any precautionary labelling and the target consumer. Essentially, the frequency of verification progressively increases, as products are sold which do not have recourse to precautionary labelling and further increases for those products targeted at persons with particular food allergies. The need for ongoing verification at this stage of the food manufacturing cannot be overemphasised. Within any food business, it can be possible for other elements within the business to act unilaterally and without reference to those responsible for food-safety management. One consequence could be for other elements within the business to identify and source raw materials without first determining the competence of the supplier to meet food-allergen requirements. This is not as unlikely as it might sound, since even in the most heavily managed businesses there are often contingency arrangements to purchase materials from ‘non-approved’ suppliers albeit on ‘concession.’ Thus, depending on the nature of the product and the information given to the consumer, management systems have to be in place to reduce the probability of such events happening.
6.3.3
Goods receipt and storage
Once raw materials have been received, they must be checked to ensure that they (at least visually) meet specification and stored appropriately prior to dispense into the production facility. Numerous texts have been written concerning what steps should be taken to minimise cross-contamination of different potentially allergenic ingredients (e.g. Food Standards Agency (2006). It is not the purpose of this chapter to rehearse them in detail; however, the basic principles will be discussed here. These relate to how such materials should be stored and what steps need to be taken in the event of a spillage. Most of the precautions necessary therefore relate to the physical characteristics of the material rather than their biological properties.
108
Management of Food Allergens
In all cases, the emphasis is on segregation to both minimise and contain the effects of any spill. Storage in segregated areas (when possible) minimises the risk of crosscontamination in the event of a spill. Storage of high-risk material at low level and – particularly in the case of liquid materials storage – in bounded areas will assist in containing the spread of such material, if a spill occurs. Any facility must also have in place systems to deal with spills which occur and which minimise the risk of allergen crosscontamination. The three elements needed in such a system are documented procedures, suitable equipment and appropriately trained personnel. The lower the complexity of such systems – the better; a case in point relates to particulate allergenic materials. Food businesses operating to modern GMP standards (e.g. BRC Global Food Standard) already have systems in place to deal with glass and hard plastic breakages. Such systems include the provision for:
r r r r r
Speedy notification of management and implementation of containment and control measures (see below). Areas where spills have been detected to be quarantined pending satisfactory clean-up procedures. Dedicated cleaning equipment, and protective clothing (all of which are used only once and then safely discarded) for use by trained staff. Appropriate decontamination procedures. Audit facilities to verify the efficacy of cleaning to determine whether or not the quarantine should be lifted.
Depending on the nature of the food business, such systems can be developed to include the possibility of allergen spills. Another area within the raw materials handling activities of a business and key to successful food-allergen management concerns rework. Within many food businesses, rework is a fact of life either because of its technological contribution to the recipe or the cost of raw materials. In food businesses which manufacture a range of products, it is essential that those businesses ensure that rework containing a particular allergenic material is not included in a food where that allergen is supposed to be absent. In order to achieve this, it is important that rework is considered to be an ingredient in its own right and managed in the same way as any other ingredient. In other words, stock control systems should be suitably rigorous to differentiate between different types of rework and ensure that the rework is correctly used (usually through the application of a rework matrix). In certain industries, allergen cross-contamination due to rework may probably exhibit an ‘iceberg’ effect (i.e. a large number of cases are currently going undetected). A case in point relates to the use of chocolate as an ingredient. A recent paper by Pele et al. (2007) performed a survey of chocolate products sold in a number of member states within the European Union together with (then) candidate countries. They found that over 50% of chocolate products with no ingredients declaration or precautionary labelling relating to tree nuts contained detectable amounts of hazelnut protein as measured by ELISA. In the case of peanuts, the figure was 25%. These figures may, in part, reflect poor rework management systems but also cross-contamination due to either poor in-line sanitation or other activities within the factory.
Risk management – operational implications Raw material 1%
109
‘Allergen free’ claim 5%
Manufacture 46%
Label details 48% Fig. 6.2 Broad analysis of 78 UK Food Standards Agency food-allergen-related alerts notified in the period March 2007 to March 2008 inclusive.
6.3.4 Production An analysis of allergen-food-related alerts issued by the UK Food Standards Agency for the period March 2007 to March 2008 inclusive (Food Standards Agency, www.food.gov.uk/safereating/allergyintol/alerts/) indicates that a little under half of them reflected failures within the manufacturing environment (see Figure 6.2). Out of 78 alerts, 36 related to events taking place within the food-manufacturing environment. Of these, 14 related to cross-contamination events and the rest to human error. Further analysis of ‘human-error’ related alerts indicated that 16 represented products being placed into incorrect packaging. These statistics highlight the need to ensure that prerequisite programmes are in place, operatives are suitably trained and that appropriate mechanisms are in place to enforce and ensure compliance. In terms of prerequisite programmes, the obvious point to consider is the risk of crosscontamination. The risk of cross-contamination can be reduced by effective segregation of production as identified in a number of codes of practice and standards (e.g. Food Standards Agency, 2006; British Retail Consortium, 2008). As already discussed above, segregation begins at the ingredients dispense stage where the handing of allergenic ingredients should be kept separate from other ingredients and moves through to the use of dedicated production facilities. Complete segregation of production through the use of dedicated production facilities is frequently not possible for valid commercial reasons. In such cases, for raw materials, consideration should be given to the use of dedicated utensils, the use of appropriate personal protective equipment and the appropriate management systems in place to not only train all relevant personnel in their use but also the enforcement of their application. In terms of the production line itself, the use of dedicated lines is desirable; however, not always practical. Where dedicated lines are used, care must be taken in ensuring that activities on one line do not compromise those on another (non-allergenic) line. Of particular concern is the transfer of allergenic materials either through the air by means of fine particulates or through aerosols or as a consequence of plant design; for example, where one conveyor system carries an allergen-containing product over another which carries product free from that allergen. In this case, care must be taken that residues accumulating on belts and bearings of one conveyor are not ‘snowing’ on the other. Where dedicated production lines are infeasible, care must be taken to minimise allergen transfer to non-allergen-containing products. This can be achieved by scheduling of production and appropriate sanitation programmes between production runs. Given that the hazard
110
Management of Food Allergens
to be managed is ‘the inadvertent consumption of a food allergen by a sensitive individual’ (see above), producing allergen and non-allergen-containing foods on the same line arguably provides the food-safety practitioner with the greatest challenge. It is the efficacy of the prerequisite programmes (in particular sanitation), which will determine the need or otherwise for precautionary labelling (‘may contain’ allergen X). It must also be remembered that the commercial desirability of precautionary labelling for any particular brand or product will be determined to a large degree by other elements within the food business (in particular marketing and sales). The content of a food label is not something usually within the control of those immediately responsible for local food-safety management. Care must be taken therefore, not only in scheduling manufacture of different products along a particular line but also in adopting the appropriate sanitation practices which can effectively decontaminate the line to minimise the risk of cross-contamination. As discussed earlier, the primary source of manufacturing-related alerts reflected human error. The single largest cause was using incorrect packaging (16 alerts out of a total of 76), the remainder involved factors such as incorrect ingredient addition and in one case the finding of peanuts in a manufacturing zone designated as ‘nut (and peanut)-free)’. All the points highlight the need for prerequisite programmes to be extended to account the significance of food allergy and also for constant vigilance. Higher levels of vigilance are required for products which carry a claim that they are ‘free from’ a particular allergen, of the 78 alerts reported 4 (5.1%), were attributable to some form of contamination with other ingredients containing a particular allergen (either milk or gluten); the presence of which was found only on analysis after the product had been released for sale. Products containing ‘allergen-free’ claims merit particular care in the food safety management systems employed. By placing such a claim on the product, the manufacturer is actively selecting for consumers sensitive to that particular allergen and therefore has a higher duty of care. This not only means having appropriate prerequisite programmes in place but also far more rigorous verification systems to ensure that they are functioning effectively. These would probably involve some sort of positive-release system on the basis of laboratory analysis of an appropriate sample demonstrating levels of the allergen concerned being below a particular limit (usually set by the sensitivity of the assay).
6.3.5 Packaging As already discussed, packaging and the information carried on it play a significant role in managing the risk of inadvertent allergen consumption by a sensitive individual. In dealing with packaging issues, it has already been shown that one route by which transmission of the correct information can be corrupted is through the use of inappropriate practices within the process environment. At their most extreme level, these can result in product declared as being free of an actual allergen, actually containing it (4 out of 78 alerts). A more frequent process error associated with packaging was use of incorrect packaging (16 alerts). However, a more fundamental error relates to the information provided on the label either in the form of an ingredients’ declaration or allergen advice panel. Examination of Figure 6.2 reveals that 37 (49%) of all food-allergen-related alerts issued by the UK Food Standard Agency were associated with label design problems. Of these, 29 were related to ingredients declarations and 8 to allergen information panels. The three main routes by which this occurred were: a failure to make a correct ingredients’ declaration (in a number of cases this involved more than one significant food-allergen);
Risk management – operational implications
111
defective allergen advice panels; and using the wrong packaging for the product concerned. The first two cases might be described as the ‘correct product – faulty label’ syndrome, while the third might be described as the ‘wrong product – correct label’. The ‘wrong product – correct label’ syndrome has been discussed in the Production section above. The ‘correct product – faulty label’ syndrome alone was the largest reason for the recalls (approximately 50%). Although the supporting information published was limited, it is possible to categorise the causes of failures in label design under one of three broad categories:
r r r
Human error: In some cases, the recipe had simply not been completely copied over to the wrapper in accordance with legislation. Compound ingredients: This might reflect human error but also ‘hidden allergen’ syndrome. In this case, it is subsequently discovered that a compound ingredient contains a food allergen hitherto not declared. Given some of the products concerned, this might also be a reflection of rework practices (discussed further below). Incorrect description: In a number of cases, recalls took place because although an appropriate ingredient description was made the terminology used could not be readily understood by the consumer. As in the case of errors relating to compound ingredients, a large number of these referred to milk-derived ingredients (in particular, the use of the words ‘whey’ and ‘casein’). Other alerts related to derivatives of other Annex IIIa listed allergens that were not subject to an exemption under the provisions of Commission Directive 2007/68/EC. One such example involved the declaration of lecithin prepared from soya simply as ‘lecithin’, rather than including its origin.
6.3.6
Prerequisite programmes and the roles of other functions within the food business
No food-processing activity operates in isolation. Figure 6.1 describes some of the other functions that impact on the activity; reference has already been to elements of prerequisite programmes such as supplier quality assurance and sanitation. Within many foodmanufacturing processes, there is rarely going to be a particular process step within it, ‘at which control can be applied and is essential to prevent or eliminate a food safety hazard or reduce it to an acceptable level’, in other words, a ‘critical control point’ (as defined by Codex Alimentarius, 2003). Prerequisite programmes are thus the first line of defence in protecting the food allergic individual. Reference has already been made to supplier quality assurance and sanitation. However, going back to the principles of PIPE, arguably the key aspect of any prerequisite programme that includes food allergy within its mandate is training. It is axiomatic, but nevertheless worth re-emphasising that the efficiency of any food-safety management system is a reflection of the quality of training given to all elements of the business. It is important to note that training must be inclusive of all functions, including those who either have intermittent access to the production or whose activities are peripheral to the production process. In terms of functions ancillary to food manufacture itself, two key functions which can, if improperly managed, compromise the allergen status of a production line are sanitation and maintenance. A failure to have validated and reliable cleaning systems can rapidly lead to the risk of cross-contamination. A similar point applies to maintenance activities; appropriate clean down has to be undertaken, for both the area under maintenance and the equipment used. In the case of production areas dedicated to make products for which an allergen
112
Management of Food Allergens
absence (‘free from’) claim is made, similar practices to those adopted for microbiological high care units should be implemented including the use of dedicated tools and protective clothing. Other functions that can impact and potentially compromise allergen management practices are invariably physically dislocated from the production line and can, in extreme circumstances, be located in an entirely separate country. Marketing, new product development (NPD) and purchasing functions activities fall within this category. Examples of events (potentially) compromising the food-allergen safety management of food-manufacturing facilities dealt with by the author as over a number of years include the following:
r r r
Proposals to manufacture a product containing an allergen not previously used in the production plant and without consideration of the implications for cross-contamination and relevant precautionary labelling for the other products manufactured there. On a similar theme, proposals by new product development (NPD) functions to run production trials for a new product containing a food allergen not previously used and without consideration for the consequences of potential cross-contamination for existing products. Changes in the suppliers of raw materials without sufficient application of SQA procedures, leading to the sourcing of allergen compromised materials.
Thus, irrespective of their function within the food businesses, everyone needs to have appropriate training in food-allergen awareness – a point recently reinforced in the Anaphylaxis Campaign’s food manufacturing standard (Anaphylaxis Campaign, 2008).
6.3.7
Ensuring compliance (enforcement)
The acronym PIPE discussed earlier includes the word, ‘enforcement’. In this context, the term refers to not only making sure that management systems put in place to control the hazards presented by food allergens are functioning but also that their effectiveness is assessed and improvements made as required. Enforcement therefore involves elements of supervision, verification and management review. Some aspects of verification can be regarded as effectively forms of supervision, for example, personnel being asked to demonstrate their knowledge and practice of the allergen management practices put in place either by line managers or during audit. Verification itself can take the form of determining compliance on a historical basis through examination of appropriate records including those concerned with operation and sanitation of the facilities. It can also measure the current efficacy of any system operating through periodic analysis of both material (ingredients and finished product) and the environment (e.g. analysis of swabs and/or rinse waters). In commissioning such laboratory analyses, it is important that there is a full understanding of the limitations of such analyses. These include the nature of the analyte (protein versus DNA), in the case of ELISA-based assays the antibody used and, for both types of analyses, limits of detection/quantification (discussed in further detail elsewhere in this book). Information gathered from these exercises should be collated and assessed with other quality-management information and the continuing developments taking place in our understanding of food allergy to determine the ongoing efficacy of the food-allergen management system.
Risk management – operational implications
113
6.4 CONCLUSION At first sight, the broad attributes of food allergy (wide range of allergenic foods, low eliciting dose, diverse range of symptoms and the (relatively) small size of the affected populations) seemingly present food processors with an insurmountable challenge in terms of food-safety management. A more considered analysis shows that this is not the case. Application of the basic principles of food-safety management based on an understanding of food allergy and the processes within the food business can lead to systems which can support the manufacture of foods both with and without particular allergens under conditions where the risk to the food-allergic consumer is minimal. Achieving this state of affairs requires high levels of commitment from each and every person within the food business from the chief executive down. This can be achieved only by appropriate staff education and management.
REFERENCES Alldrick A.J. (2006) Managing allergy issues. The World of Food Ingredients (October/November), 64–65. Anaphylaxis Campaign (2008) The Anaphylaxis Campaign Standard to Increase Trust in Information about Allergens in Food. The Anaphylaxis Campaign, Farnborough. British Retail Consortium (2004) BRC guidelines for the handling of nuts. Available at www.brc.org.uk/showDoc04.asp?id=2402&moid=2553, accessed 21 April 2009. British Retail Consortium (2008) Global Standard for Food Safety. The Stationery Office, London. Codex Alimentarius Commission (2003) Recommended International Code of Practice General Principles of Food Hygiene, CAC/RCP 1-1969, Rev. 4–2003, pp. 31. Available at www.codexalimentarius.net/ download/standards/23/cxp 001e.pdf, accessed 21 April 2009. Commission Directive 2007/68/EC of 27 November 2007 amending Annex IIIa to Directive 2000/13/EC of the European Parliament and of the Council as regards certain food ingredients. Official Journal of the European Union, L310, 1–4. Directive 2000/13/EC of the European Parliament and of the Council of 20 March 2000 on the approximation of the laws of the member states relating to the labelling, presentation and advertising of foodstuffs. Official Journal of the European Communities, L109, 29–42. European Food Safety Authority (2004) Opinion of the Scientific panel on dietetic products, nutrition and allergies on a request from the commission relating to the evaluation of allergenic foods for labelling purposes. EFSA Journal, 32, 1–197. Food Standards Agency (2006) Guidance on Allergen Management and Consumer Information. Available at www.food.gov.uk/multimedia/pdfs/maycontainguide.pdf, accessed 21 April 2009. Greer L.J. (1933) in Hall v. Brooklands Auto Racing Club (1933) 1 KB 205. Hefle S.L., Nordlee J.A. and Taylor S. (1996) Allergenic foods. Critical Reviews in Food Science and Nutrition, 36(S), S69–S89. Pele M., Broh´ee M., Anklam E. and van Hengel A.J. (2007) Peanut and hazelnut traces in cookies and chocolates: relationship between analytical results and declaration of food allergens on product labels Food Additives and Contaminants, 24, 1334–1344.
7
Choices for cleaning and cross-contact
Steve Bagshaw
7.1
ALLERGEN MANAGEMENT AND CLEANING
With the ever increasing awareness and importance of producing foods that clearly label if a product contains known allergens, either as a deliberate ingredient or as a possible contaminant, allergen risk assessments and management must be introduced to the food process. Allergens should be managed to avoid their unintentional presence in products. This management involves evaluation of the likelihood of allergen cross-contamination associated with every step of the food production process, from sourcing raw materials through to marketing of a finished product. Existing good manufacturing practice (GMP) controls will assist with allergen management, for example avoiding cross-contamination by segregation, cleaning, using separate utensils, etc. The introduction of allergen management into a food business can be seen as an extension of existing food safety management rather than a completely new system. Cleaning for allergen control is required only where that allergen is not an intentional ingredient of the food being produced. Allergenic material is generally a protein. Very small amounts of some allergens, such as peanut, can cause adverse reactions; the severity will vary but for certain individuals this may be a fatal anaphylactic shock. Existing legislation relating to allergens exists The Food Labelling (No 2) (Amendment) Regulations 2005; however, this covers only the labelling aspects of the issue. The Anaphylaxis Campaign standard has set standards relating to the level of cleanliness required. All equipment, surfaces, utensils identified by risk assessment as subject to contamination by allergens shall be cleaned to a demonstrably visually and physically clean standard, or equivalent validated standard to remove any potential cross-contamination residues. Where design of equipment prevents achievement of the above standard of cleanliness the use of “may contain’ labelling shall be considered. The company must be able to demonstrate why cleaning to the above standard is not possible and this can only happen where the decision is the result of the allergen management review process. (Anaphylaxis Campaign Standard, 2007) Cleaning practices that are satisfactory for hygiene purposes may not be sufficient for the removal of allergens from surfaces and equipment. Any cleaning process developed for allergen removal must be validated to ensure allergens are removed from the target surface and that no risk of cross-contamination to other food contact surfaces occurs. It
Choices for cleaning and cross-contact
115
must be remembered that, unlike microbial contamination, allergenic material is generally unaffected by heat (Taylor and Lehrer, 1996) or chemicals. Cleaning is used to remove contamination but it can in itself be a source of contamination. Potential sources of cross-contamination from cleaning include overspray from wash-down guns, aerosols, cleaning tools, personnel and recovered cleaning solutions. The time of cleaning, methodology of clean and cross-contamination controls must be carefully considered to avoid cleaning causing allergen cross-contamination. The cleaning regime can be tested and compared with other regimes by checking for levels of contamination after the clean. The assessment could look at residual soil level which can be tested by looking at ATP or protein, though specific allergen tests are more widely recognised as offering a better means for control after the cleaning and rinsing stages.
7.2
THE CLEANING PROCESS
With all food production processes, both equipment and surfaces become contaminated with food residues, foreign bodies and microbial contamination. The removal of these contaminants, soil, is the process of cleaning. Cleaning should always be considered an essential and integral part of the production process for many reasons. Namely:
r r r r r r r r
Legislation – The Food Safety Act 1990, UK and EU regulations require that effective hygiene standards are implemented for anyone who handles, manufactures or serves food For microbial control – reducing bacterial numbers to an acceptable level for the product being produced Removing physical or chemical contamination Foreign body control Performance of the plant (for example removal of build-up on heat exchangers) Safety of the plant and operatives Discourage pests A pleasant and clean work environment to create the right impression for all
Cleaning needs to be carried out in such a manner that is both effective and efficient. This means without causing damage to personnel, equipment or surfaces; and without causing cross or re-contamination. The timing and frequency of clean needs to be set by risk assessment of the product and process. During the production process, product contact surfaces will become contaminated with soil and microorganisms. This contamination may lead to deterioration in product quality and eventually an unacceptable product. By assessing product quality against production running time, a frequency of clean can be determined that ensures that product quality always remains acceptable. Potential cross-contamination risks from the cleaning process must also be identified and risk assessed. Production planning should ensure that free from allergen products are produced on a line at the start of the day following a hygiene clean. When equipment is used to process both foods containing allergens and those free from allergens, then by maximising production runs the number of changes where thorough cleaning is required can be reduced.
116
Management of Food Allergens
Fig. 7.1
Effect of periodic cleaning.
7.2.1 Inter-product cleans An inter-product clean is used when a production line is changing product. This removes physical contamination such that the subsequent product does not contain components of the previous production run. For example a vegetarian-labelled sandwich should not contain chicken from the previous production. Inter-product cleans, typically seen in food manufacturing, control physical and microbial contamination to an acceptable level where allergen cross-contamination is not an issue. With most fixed open plant food production equipment, an inter-product clean does not involve a full strip down and therefore there exists a potential for product residue to be left on the equipment. With equipment that can be taken offline or where the equipment item is naturally remote from a production area thorough cleaning can be achieved. For example on semi automated conveyor lines depositors can be taken away after a given period and replaced by a clean unit. The dirty depositor is then cleaned in a dedicated washroom where a full strip down and hygiene clean can be carried out.
7.2.2
Break cleans
Break cleans are carried out at natural breaks in production and are used for tidying/clean as you go.
7.2.3
Timed clean
Cleaning after a defined time period is appropriate for certain equipment. With continuous production, a potential for increasing soil and microbial loading on food contact surfaces exists; this can be controlled by suitable periodic cleaning (Figure 7.1).
7.2.4
Hygiene clean
A hygiene clean, or an end of production clean, refers to a thorough strip down of equipment to allow full access to all surfaces and routine entrapment areas. This is usually carried out
Choices for cleaning and cross-contact
117
when no other production is taking place nearby and when all work in progress or finished product has been cleared away from the vicinity.
7.2.5 Deep clean Certain areas or parts of equipment such as electrical control cabinets may not be accessed during the hygiene clean. These areas of the equipment are not deemed to pose a day-to-day contamination risk and are cleaned for instance every 3 months; these are often referred to as deep cleans.
7.2.6
Good cleaning practices
Cleaning a food production environment requires considerable coordination by the cleaning crew to ensure that it is effectively and efficiently cleaned without introducing crosscontamination risks. Some of the considerations are
7.2.6.1 Preparation of the area and equipment
r r r
Remove all product from area before starting Cover electrical components Ensure all cleaning equipment is clean and disinfected – brushes, cloths, pads, etc.
7.2.6.2 Cleaning
r r r r r r r r r r
Adhere to the agreed method; paying attention to detail Follow correct procedure and sequence of clean Carry out checks on cleaning process including level of strip of equipment, chemical strengths, temperatures and contact times with detergents and disinfectants Run hosepipes underneath machinery and equipment Don’t place dirty items onto clean or vice versa Work in a top to bottom manner Clean in a safe manner – don’t take risks, wear personal protection equipment Don’t clean machinery parts or equipment on the floor Keep clean and dirty items separate Wash hands regularly, particularly after handling dirty items
7.2.6.3 Reassembly
r r
Inspect all surfaces before reassembly and re-clean, if necessary Disinfect surfaces before and during reassembly using only hygiene trained personnel
7.3
PRINCIPLES OF CLEANING
Cleaning involves detaching soil from a surface and the removal of it to waste. In wet cleaning, this means the detergent has to detach the soil and then suspend it in the cleaning solution
118
Management of Food Allergens
ic
al
en
er gy
Fig. 7.2
Thermal energy
Ph ys
al
y rg e n
e
ic
em
Ch
Cleaning energies.
until it reaches a water treatment plant. With dry cleaning, it general involves removing to a waste bin via a brush and dustpan or via a vacuum cleaner.
7.3.1
Cleaning energies
With wet cleaning, thermal, chemical and physical energy are used in combination to remove the soil from a surface (Figure 7.2). The balance of energies depends on the physical process and the detergent type; some involve low chemical energy (such as neutral detergents) combined with high physical energy (scrubbing) and others high chemical energy (acidic descaler) with little physical energy. The greater the attachment of the soil, to a surface, the greater the combined energy required for cleaning. Contact time with a detergent is necessary to maximise its performance; generally, the longer the contact time with a detergent, the less physical energy will be needed to remove the soil from the surface. Temperature is important since it increases chemical reaction speeds between detergent and soil and is also essential for the effective emulsification of fats where present.
7.3.2
Choice of detergent
The role of the detergent is to assist in the removal of soil from a surface. The choice of detergent has to be made when a number of factors have been considered. These include soil type, method of application, materials of construction of the surfaces being cleaned, water hardness, health and safety and many others. Most detergents combine emulsification properties with some type of chemical reaction. There is no evidence to suggest that any chemical reaction occurring between the detergent (or disinfectant) and allergenic protein has any effect on its allergenicity. Therefore, the effective removal of soil by the cleaning process is essential to achieve the removal of allergenic material.
Choices for cleaning and cross-contact
119
Although there are thousands of detergent products available in the professional and domestic market, they break down, broadly speaking, into the following:
r r r
r r
General purpose neutral or mildly alkaline detergents are used for hand cleaning by spraying or for soak cleaning (sinks). They rely largely on emulsification and suspension of soiling and are particularly effective on fats and oils. Sanitisers are neutral, mildly acidic or mildly alkaline detergent disinfectants that combine the cleaning properties of a neutral detergent with a degree of disinfection. Highly alkaline detergents are used where heavier soiling is encountered. They rely partly on chemical reaction to hydrolyse proteins or saponify fats. They can be applied in a number of formats including gels or foams for long contact times with open surfaces or as low foam products for recirculation cleaning such as dishwashing, tray washing and cleaning-in-place (CIP). They are effective on highly carbonised or polymerised soils. Highly alkaline chlorinated detergents, either as foams or low foam products for recirculation, are used because of their excellent removal of the fats and also proteins. Acidic detergents can be used for mineral scale removal and protein removal.
7.4
OPEN PLANT CLEANING
Open plant cleaning can be carried out either as a wet or dry process depending on the nature of the soiling present, the product, the process and type of production equipment. Dry cleaning is used mainly for processes where dry or particulate products are handled. Wet cleaning is employed wherever possible due to the higher efficacy of the wet cleaning process (better soil removal from surfaces).
7.5
DRY CLEANING
Dry cleaning (Figure 7.3) refers correctly to cleaning where no liquid detergents or disinfectants are used; however, it is also commonly used to refer to cleaning where disposable impregnated wet wipes or damp disposable cloths are used. In the food industry, it is usually found in processes where the presence of water could affect the quality and consistency of the product such as bread, pastry, biscuits, cereals and so on or create conditions that enhance microbial growth. Dry cleaning is a purely mechanical process that relies on the soil being physically removed (brush, vacuum). The nature of brushes and vacuums mean that 100% soil removal of soils will not be achieved (Holah et al., 2004). Consideration must be given to following brushing or vacuuming with wiping with detergent/disinfectant dampened cloths to increase overall soil removal.
7.5.1 Cross-contamination by dry cleaning Tools used for cleaning can become a major route of cross-contamination. Cloths and scourers should be disposed of after use. Brushes, scrapers and other tools must be cleaned, disinfected
120
Management of Food Allergens
Fig. 7.3
Dry cleaning.
and stored hygienically for later use. Tools should be clearly defined for area of use, for example floor use only or food contact use only. In addition, specific tools need to be kept for cleaning of surfaces that have allergenic material on them. A colour code system can be used (Figure 7.4) for this. It is vitally important that the system is clearly defined and managed.
Fig. 7.4
Colour coding. (Reproduced with permission from Vikan Professional Cleaning Solutions.)
Choices for cleaning and cross-contact
Fig. 7.5
121
Atomisation.
Vacuum cleaners tools such as nozzles and hoses will become contaminated and will need thorough cleaning. In addition, thought should be given to specific hose and nozzle sets for times when allergenic material is being handled. The vacuum cleaner, if mobile, rather than a fixed central installation will exhaust air locally. Regular cleaning and replacement of filters to ensure no particulate materials are blown out of the unit is critical. Air lines are sometimes used to dislodge soil from difficult to access areas; essentially moving it to an area that can be accessed by a dustpan and brush or vacuum. Unfortunately, air lines will impart energy to fine particulate matter making it airborne and allowing it to spread over large areas (Figure 7.5).
7.6 MANUAL CLEANING Manual cleaning refers to the cleaning process where the detergent is applied via a cleaning tool such as cleaning cloth, scourer or brush (Figure 7.6). It also refers to cleaning of parts that are put to soak in detergent solution before physical action. Manual cleaning of machinery, equipment and surfaces is the most common method employed throughout the food-manufacturing industry. Manual cleaning provides a flexible method of cleaning for a variety of equipment and surfaces and has little risk of crosscontamination caused by aerosols or overspray; however, the control and cleaning of cleaning tools is vital to ensure no cross-contamination. With manual cleaning, there is a substantial contact between the operative and the chemical being used to clean; therefore, the type of detergent used must be carefully considered. It will usually involve a 1–2% v/v solution, at typically 40–45◦ C, of either a neutral detergent, a quaternary ammonium compound (QAC)-based detergent or a light/medium duty alkaline detergent. These light duty detergents will not perform as well as the higher alkalinity products on fat or protein soils.
122
Management of Food Allergens
Fig. 7.6
Manual cleaning.
7.7 FOAM AND GEL CLEANING Foam cleaning refers to the cleaning process where the main detergent is applied as foam, and gel cleaning where the main detergent is applied as a gel (Figure 7.7). Gels can also be aerated during application; this is called a mousse.
Fig. 7.7
Foam cleaning.
Choices for cleaning and cross-contact
123
With increasing commercial and technical pressure placed on the food manufacturing industry, the time window and manpower required for cleaning have been squeezed and decreased. Foam cleaning has proven to be a very effective, efficient and popular method for cleaning of rooms and equipment. The improvement in foam technology, such as long cling foams, and the introduction of different types of foam detergents have made it a process that can be used, with benefit, in many situations. Foam is created by mixing water, detergent and air together and applying it via a hose with a special nozzle or lance onto the surfaces and equipment. The foam detergent will typically be applied at 3–5% v/v, depending on the soil to be removed and water hardness. The main advantages of foam and gel cleaning in comparison to manual cleaning are as follows:
r r r r r r
The detergent solution can be applied to large and difficult to reach areas in a short period of time An extended detergent contact time between the soil and the detergent A reduction in the time of clean Less manpower required Control of detergent use Safer application of hazardous detergents
A common misconception of foam and gel cleaning is that it negates the need for any type of physical action (such as scrubbing with a brush or scourer). Physical energy must be applied after suitable detergent contact time. The physical energy can be applied by either scrubbing or by energy from a water jet typically either high or medium pressure.
7.8 CROSS-CONTAMINATION 7.8.1
By wet cleaning
A wash-down system provides the hygiene operative with an efficient tool for rinsing away soil and detergent (Figure 7.8). The water jet provides a degree of physical energy to a surface that assists in the removal of the soil – the higher the pressure of the system, the higher the impact energies available. Medium pressure systems operate at typically 20 bar with a flow rate at the nozzle of 30 L/min. This gives very similar cleaning energies to a high pressure systems operating at 70 bar and with a flow rate at the nozzle of 15 L/min. Low pressure systems operate at typically 5 bar with a nozzle flow rate of 40–50 L/min. Both medium- and high-pressure systems provide sufficient cleaning energy to remove most soiling if the correct foam or gel detergent has been used. With low pressure rinsing, however, this is not the case and it is essential that surfaces are scrubbed prior to rinsing. All wash-down systems whether low, medium or high pressure will cause overspray which can lead to cross-contamination if no controls are put in place; however, high pressure systems also create aerosols which add another vector of cross-contamination. As discussed in the Section 7.5, tools are a major source of contamination. Cleaning tools and their management generally receive low priority; this was aptly demonstrated in a food industry wide survey by Campden BRI which showed that of all cleaning tools tested 35% were Listeria positive. The correct choice of tools, the management of cleaning of tools and the designation of specific tools for different purposes (usually identified by a colour coding system) is essential.
124
Management of Food Allergens
Fig. 7.8
Wet cleaning.
7.8.2
By personnel
Cross-contamination by personnel can cause major issues with the end result of the clean. The vehicles of contamination include the following:
r r r
Hands/gloves Protective wear – mainly overalls, aprons or wet suits Footwear
Transfer of personnel from low risk to high care and an allergen area to a non-allergen area should be avoided, if possible. If not, then hand washing and changing procedures should be adopted where all protective wear including overalls, hairnet, hat, gloves and footwear are changed. The use of specific overalls for allergen areas should be considered. The laundering of overalls used in allergen areas needs to be managed by the site and the laundry. Checks on laundered items should be carried out to establish the removal of allergens by laundering. Footwear should remain, where possible, captive to an allergen area. If this is not possible, consideration must be given to footwear washing to avoid cross-contamination on floors. Production personnel involved in cleaning (typically on an inter-product clean) must adopt very strict apron and glove changing and hand washing after cleaning and prior to production. Hand washing is important for both microbial control and allergen control. A proper hand-wash procedure including the use of a soft nail brush should be trained and monitored. Hand-washing compliance can be poor but increases when the hand-wash facilities are good, the water temperature is comfortable, the hand soap mild and most importantly personnel understand the reason for hand washing.
Choices for cleaning and cross-contact
Fig. 7.9
7.9
125
Floor cleaning.
FLOOR CLEANING
Floor cleaning can be carried out using manual methods such as a mop/brush and bucket, by utilising the wash-down system or by a dedicated floor-cleaning machine (Figure 7.9). The most appropriate method will depend on the access to the floor area, time of cleaning and the size of the floor area. With manual and wash-down methods, the detergents used are generally the main detergent used for the equipment and wall cleaning, since the type of soil encountered will be the same. With floor cleaning machines, the detergent is usually a low foam alkaline detergent. All methods create overspray which can travel vertically onto food contact surfaces. Work by Campden BRI showed that the potential for cross-contamination of food contact surfaces from floors was real and measurable. The distance of vertical and horizontal travel by floor cleaning solutions varies depending on the method of clean (Holah et al., 1993). The greatest vertical and horizontal travel of cleaning solutions occurs when using a high pressure wash-down gun, followed by medium pressure wash-down gun, low pressure wash-down gun, floor cleaning machine with mopping and brushing creating the least travel.
7.10 TRAY AND RACK WASHING MACHINES Washing machines are used to automatically or semi-automatically clean trays, racks, utensils and in the case of dishwashers, crockery and cutlery (Figure 7.10). Because of the high volume of trays, baskets, tubs, containers, etc., required by many modern food businesses, it has become unrealistic and uneconomic to clean manually. A large bakery, for example, may require 2500 baskets an hour to keep a continuous production flow. Washing machines
126
Management of Food Allergens
Fig. 7.10
Tray washing.
come in many shapes and sizes and are generally built for the cleaning of a specific size and type of item. Washing machines should only be used for the original purpose they were intended. The short contact time with detergent and the relatively low impact energies of the wash nozzles means that to clean effectively high chemical energies are required. In most situations, a high alkaline low foam detergent at 0.25–1% v/v is used at 55–75◦ C. The washing machines must be managed correctly with regular cleaning of the machine and filters, regular changing of wash solutions, inspection of wash and rinse nozzles and control/monitoring of detergent temperatures and strengths. Washing machines are either tunnel-type machines or single tank machines. With tunnel machines (Figure 7.11), items are placed on a conveyor which transports them through a number of stages of the washing process. Each stage occurs in a different section of the washer: pre-rinse, wash with detergent, rinse, disinfect chemically or by temperature and
Fig. 7.11
Tunnel traywash.
Choices for cleaning and cross-contact
127
then dry. With single tank machines, the item is placed in a cabinet and the wash process takes place in the cabinet by sequencing the cleaning stages (pre-rinse, wash, rinse, disinfect). The efficacy of a clean is dependent on correct original design of the machine for the items being cleaned and correct operation and maintenance.
7.10.1
Cross-contamination by washing machines
With all modern washing machines, certain cleaning solutions are reused, for instance the detergent solution. The detergent solution will become contaminated with soil from the washed items and therefore provide a transfer mechanism for allergens. The potential for this cross-contamination can be controlled by having specific dedicated washing machines for items that have allergen contamination, by washing such items separately (for instance, in a three sink system) or by validating the operation of the washing machine and then putting management controls in place to ensure its correct operation.
7.11
CLEANING-IN-PLACE
CIP is the cleaning of pipework or vessels (tanks) by passing cleaning fluids through the pipework or spraying inside the vessel (Figure 7.12). The flow rate and volume of cleaning fluids required can be very high and it is most common practice to recirculate and sometimes recover these fluids to make most use of them. The principles of cleaning are always the same, whether it is manual cleaning of a utensil in a sink or in a dishwasher or the CIP of a vessel. Energy in the form of chemical energy, thermal energy and physical energy are all required to remove the soil from the surface.
Fig. 7.12
Cleaning-in-place.
128
Management of Food Allergens
It is important to note that for equipment to be effectively cleaned by CIP, it must be designed to do so. For instance, an automatic depositor head designed for CIP will incorporate a mode where all of the depositor head seals become exposed to cleaning solutions. With a normal depositor, the head will require full dismantling for manual cleaning and it is not possible to effectively clean this type of depositor by recirculation or CIP.
7.11.1
Chemical choice
A wide range of CIP detergents and disinfectants are available. One common feature to all is that the products are low foaming and in some circumstances act as defoamers. This is particularly important where caustic detergents are used to remove fatty deposits. The reaction between the caustic and fats produces soap which will naturally foam in recirculation situations. The choice of detergent and disinfectant are based, as in open plant cleaning, on a number of factors including soil type, materials of construction of the surfaces being cleaned, water hardness and many others. In many CIP situations, a caustic-based, low foam detergent is used at typically 0.25–1% w/v (NaOH) at 65–75◦ C. With CIP, the physical energy comes from the movement or impact of the fluid.
r r
With vessel cleaning, the physical energy comes from the impact or flow of the cleaning fluid over the internal vessel surface. In the case of pipework, the turbulence of the fluid flowing creates a physical scouring action on the internal wall of the pipe. Turbulent flow is achieved at flow velocities of greater than 1.5 m/s.
A CIP set delivers cleaning fluids (water, detergent and disinfectant) to vessels or pipework that need to be cleaned. The cleaning solutions are routed to the relevant vessel or pipework route by either a flow plate (manual connection panel) or via an automatic valve bank.
7.11.2
Vessel cleaning
Production vessels come in vast range of sizes. A large brewery or dairy vessel could hold 100,000 L (for instance a bright beer tank), while a small cooking vessel in a food factory may hold 500 L. Both can be cleaned by CIP as long as all the vessel internals can be exposed to the cleaning fluids. With any vessel, it is wise to seek advice from a spray-head manufacturer to ensure the correct spray-head(s) is used to give complete coverage (computer-aided design will be used to ensure theoretical coverage). When vessels have internal structures (e.g. scraped surface paddles in a cooker), it is vital that the spray-head design and position also ensures coverage of these internals as well as the vessel surface.
r r r
Spray devices should be regularly removed for inspection and cleaning. Internal pipework, vessels, welded joints, unions and valves should all be of a hygienic cleanable standard (Curiel et al., 1993). Ensure no pooling of liquids in the bottom of the vessel occurs; if feed rate is higher than scavenge rate, then consider burst rinsing.
Choices for cleaning and cross-contact
Fig. 7.13
Total loss CIP set.
7.11.3
Pipework routes
129
CIP feed pump and circuit design should ensure a minimum flow rate of 1.5 m/s; this ensures that the fluid flow in the pipework is turbulent.
r r r r
Pipework routes should be of consistent diameter. A change in diameter will lead to lower cleaning velocities in the larger diameter pipework and therefore potentially poor cleaning. CIP flow rates should not be too high otherwise pipe hammer can occur. Pipe hammer is caused by rapid changes in liquid velocity and can lead to seal and pipework damage. Internal pipework welded joints, unions and valves should all be of a hygienic cleanable standard. Dead-legs should be avoided.
7.11.4
Types of CIP set
7.11.4.1 Total loss CIP sets A total loss CIP set recirculates the cleaning fluids through the vessel or pipework to be cleaned. After the end of the detergent recirculation stage, the fluid is discharged to drain and not recovered to a storage tank (Figure 7.13). The buffer tank is generally small (200 L) and is used to maintain a feed to the CIP feed pump. A typical cleaning cycle will be (times will vary depending on cleaning circuit) as shown in Table 7.1.
130
Management of Food Allergens
Table 7.1 circuit).
A typical cleaning cycle for a total loss CIP set (times will vary depending on cleaning
Cleaning stage
Description of stage
1. 2.
Pre-rinse Detergent
Feed clean water, to circuit, and return to drain. Feed for 5 min. Feed clean water, until circuit made, and return to buffer tank; while recirculating run caustic dosing pump. A plate heat exchanger is used to heat the solution to the required temperature. Allow to recirculate for 20 min.
3. 4.
Rinse Disinfect
Feed clean water, to circuit, and return to drain. Feed for 5 min. Feed clean water, to circuit, and return to buffer tank; while feeding water, run disinfectant dosing pump. Allow to recirculate for 3 min.
5.
Drain
Drain system and buffer tank
7.11.4.2 Recovery CIP sets A recovery CIP set recirculates the cleaning fluids through the vessel or pipework to be cleaned (Figure 7.14). During and after the detergent recirculation stage, the fluid is recovered to a storage tank and is then available for use on a subsequent clean. There are different degrees of recovery with some sets recovering water in a rinse recovery tank, detergent in the detergent tank and disinfectant in a disinfectant tank. The set illustrated (Figure 7.14) has two large recovery tanks; one is used to hold the detergent and the other recovered final
Fig. 7.14
Recovery CIP set.
Choices for cleaning and cross-contact Table 7.2
131
A typical cleaning cycle of a recovery CIP set (times will vary depending on cleaning circuit).
Cleaning stage
Description of stage
1.
Pre-rinse
Feed water, to circuit, from rinse recovery tank and return to drain. Feed for 5 min.
2.
Detergent
Feed detergent, to circuit, from detergent tank and return to detergent tank. A plate heat exchanger is used to heat the solution to the required temperature. Allow to recirculate for 20 min.
3. 4.
Rinse Disinfect
Feed clean water, to circuit, and return to rinse recovery. Feed for 5 min. Feed clean water, to circuit, and return to rinse recovery tank; while feeding water, run disinfectant dosing pump. Disinfectant dosing for typically 3 min.
5.
Drain
Drain system returning all solutions to the rinse recovery tank
rinse water. A typical cleaning cycle will be (times will vary depending on cleaning circuit) as shown in Table 7.2
7.11.5 Cross-contamination by CIP As with an open plant surface, the best method of reducing cross-contamination by allergens in closed equipment such as vessels and pipework is to have separate process plant for allergenic and non-allergenic food products. If this is not feasible then the vessels or pipework must be capable of being cleaned free from allergens, the method validated and then verified on an on-going basis. An additional cross-contamination route exists with cleaning in place because of the commonality of the CIP set which is often used for cleaning many vessels and pipework routes. Any deposits in the CIP set, including filters, may be transferred between cleans. With recovery CIP sets, the recovered solutions are a potential for cross-contamination; but with total loss CIP sets, the potential for cross-contamination is reduced because no solutions are recovered or reused on subsequent cleans.
7.12
MANAGEMENT OF ALLERGEN CROSS-CONTAMINATION
See Table 7.3.
7.13 CLEANING MANAGEMENT All food manufacturers should have a company hygiene policy and included within this should be provisions for an effective approach to the hygiene management system. The policy should make the following provisions:
r r
A statement of commitment, at the highest level, that cleaning is essential for product integrity State the responsibilities of directors, management and operatives
132
Management of Food Allergens
Table 7.3
Management of allergen cross-contamination.
Cross-contamination
Controls
Dry airborne particles
• • • •
Aerosols from wash-down guns Overspray from wash-down guns
• •
Cleaning tools
• • • •
Cleaning personnel
• •
Use of disposable gloves and aprons Tray and rack washers
CIP sets
r r r r r
• •
Limit the use of air lines Ensure vacuum cleaners are regularly maintained and filters changed Use local extraction from processes where allergens may be released Use positive pressure in rooms where non-allergen foods are being handled Use low or medium pressure wash-down systems; in which case aerosols are not created Maximise physical separation of equipment used for allergen and non-allergen production Use barrier screens/curtains to control overspray Consider sequence of cleaning Consider timing of cleaning Have specific tools for specific allergen containing foods. Use colour coding and training to assist with compliance Cleaning programme for the cleaning of tools Correct hand-washing procedures. Ensure compliance through training and education Effective cleaning/laundering of overalls Effective footwear washing and storage
• Wash allergen containing utensils separately from the washer where possible • If using washer, then ensure final rinse is clean and not recycled water and validate clean achieved • Carry out checks for allergen presence after the washer • Use dedicated vessels, pipework and CIP sets where possible • Use total loss CIP sets where possible • Manage cleaning of CIP sets including filter cleaning • With recovery CIP sets, manage periodic cleaning of recovery vessels • With all sets, validate CIP clean for no cross-contamination and put in place controls to ensure CIP clean follows validated procedure • Carry out routine end of clean checks for allergen presence
State the standards required by the manufacturer, customers, third party auditors and legislation Identify the resources that will be made available such as labour, equipment, chemicals, water and so on Systems for monitoring, controlling, improving the hygiene system, including documentation Training of management and operatives A system of review
7.14 THE CLEANING PROGRAMME The cleaning methodology and management controls must be appropriate to the process and the risks to the food. The flow chart shows the process required to set up and review the cleaning programme (Figure 7.15).
Choices for cleaning and cross-contact
133
Setting the standard Agree standards required
Developing cleaning methodology Methodology developed to meet standard required and then validated
Risk assessment Assessment based on risks from both process and cleaning
Create cleaning schedule documentation Scheduling is a record of the validated methodology
Training Training is carried out against schedule and detailed instructions
Control and audit of cleaning Controls put in place to ensure methodology followed
Review Periodic review and review on change
Fig. 7.15
The flow chart showing the process required to set up and review the cleaning programme.
7.14.1 Setting the standard The first step for implementing a successful hygiene management system is to ensure that the standards required by the manufacturer, customers, third party auditors and legislation are clearly identified, recorded and communicated to all employees.
7.14.1.1 Example only Standard required prior to start of production in low-risk salad preparation area:
r r r
Visually clean Free from allergenic material (if relevant) Food contact surfaces – total viable count (TVC) < 102 CFU/cm2
134
Management of Food Allergens
Table 7.4
Example only: cleaning program for sandwich assembly area.
Frequency
Time
Type
Team
Daily Prior to product change Production breaks
14:00 to 18:00 10 min between products 22:00 to 22:30 02:00 to 02:30 06:00 to 06:30 10:00 to 10:30 18:00 to 14:00
Hygiene clean Inter-product clean Inter-product clean
Hygiene team Production Hygiene team
Clean and tidy of floor and waste
Hygiene team
On-going
Standard required during production in low-risk salad preparation area:
r r r
Visually tidy Free from allergenic material (if relevant) Food contact surfaces – TVC < 104 CFU/cm2
7.14.2
Developing cleaning methodology
The cleaning methods need to be developed and trialled to see if they meet the standards required. Competent and trained hygiene management together with third party consultants, such as reputable chemical suppliers, using the risk assessment approach should define the frequency and devise the methodology of cleaning for each area of the process. The cleaning process then needs to be risk assessed to see if any direct or cross-contamination issues occur by utilising the methodology. It is useful at this stage to summarise the cleaning programme (see Table 7.4). The cleaning programme should be validated to ensure that it meets the standards set. Once validated, the methodology can then be recorded as the cleaning schedule.
7.14.3
Design of verification programme
The verification process is part of the design and review of cleaning methodology. The process is used to validate a proposed cleaning methodology and then as part of management review to monitor performance against standards. This then enables development of the process in the drive for continuous improvement. Microbiological testing, ATP monitoring, protein testing or specific allergen tests can be used for a number of purposes; however, we are specifically considering their use for the verification of cleaning. Samples can be taken (sampled) from environmental surfaces; this includes food contact surfaces such as the cutting blade on a slicer, indirect food contact surfaces such as the control panel for a slicer and non-food contact surfaces such as the framework of a slicer. Any sample taken will as name suggests only be a sample and cannot reflect all surfaces that the food will come into contact with in production. As such the sample sites need to be chosen to best reflect the effectiveness of the clean. Process samples can be taken from the equipment as another measure of the cleanliness of the equipment and therefore the effectiveness of the clean. These can be:
r
Product samples taken as first off product from a line. As such this first off product will have contacted all food contact surfaces.
Choices for cleaning and cross-contact
r
r
135
Rinse water samples. These are usually taken in closed plant cleaning systems and involve collecting the last rinse water from a CIP system. Great care needs to be exercised in the ‘clean’ sampling of this rinse water. In addition, any fault in the CIP such as a failure of a section of pipe to be routed during the clean will lead to an unsuccessful clean but may also give a clean rinse water sample (because the last rinse would also not have flowed through the unrouted section). A synthetic process sample can be used as a medium for sampling closed plant. For instance, a buffered saline solution can be passed through a process, such as cooking, cooling and filling. This can then be sampled and used as a measure of the cleanliness of the plant.
7.14.4
Create cleaning schedule documentation
It is the responsibility of the food manufacturer to document the cleaning procedures; however, the chemical supplier may assist with the process. The purpose of the documentation is:
r r r
For training of hygiene/production operatives As a reference for hygiene/production operatives To enable auditing of the process
The documentation can be very extensive since it needs to cover a number of different types of clean, the detail of cleaning and any controls required as defined by the risk assessment process. A typical structure of cleaning programme is illustrated in Figure 7.16.
7.14.5 Training Training of management and operatives is important to ensure that staff carry out their duties correctly and fulfil their potential by understanding:
r r
Their responsibilities within the team The standards required
With good training, confidence is promoted, job satisfaction increased, team spirit developed, performance improved and the amount of supervision required is reduced. Those companies that invest in the time and resources for training tend to reap the rewards of increased profitability. Training should be tested and recorded.
7.14.6
Control of cleaning
It is important to monitor the energies that are used for cleaning to ensure that they are in line with those established when the methodology was validated. A record of the factors should be kept and if any is out of specification corrective action must be followed to restore to that defined in the methodology.
136
Management of Food Allergens
Cleaning plan Plan itemising all cleaning in a process area, including break, inter-product, hygiene and deep cleaning
Sequence of clean Summary of the stages of cleaning within an area
Walk-away standard The condition that the process area should be left in by production prior to the clean
Cleaning method card Detailed instructions on cleaning method for surfaces and equipment within the area
Cleaning record A signed document as proof of cleaning against method
Inspection record Signed documents of inspection
Fig. 7.16
Cleaning standard The condition that the process area should be left in after cleaning
A typical structure of a cleaning programme.
For open plant cleaning this would involve:
r r r r
Checking that the physical strip down of equipment meets that defined on the schedule. Checking the chemical strengths of detergent and disinfectant as applied to surfaces. This will generally involve a simple test kit. Checking the strength of disinfectants on surfaces after application to ensure that no dilution or inactivation is occurring. This can be carried out using disinfectant specific test strips. Checking the water temperature and pressure of any wash-down system. Water temperature and pressure are often critical to achieving a successful clean. For CIP this would involve:
r
Checking the physical set-up of the cleaning route and the correct functioning of sprayheads.
Choices for cleaning and cross-contact
r r
137
Checking the chemical strengths of detergent and disinfectant. Modern cleaning systems will generally monitor and record return flow conductivity (relates directly to chemical strength), return flow temperature and return flow rate. Unless the CIP system automatically ensures the cleaning process parameters are met, the cleaning record should be analysed to verify that all parameters meet or exceed those achieved during validation.
7.14.7
Audit of cleaning
Visual inspections after the cleaning process are important to ensure that the required visual standards are achieved and maintained. The inspection should be thorough and all areas should be covered including those outside the daily cleaning programme. Inspections after the cleaning process should make provisions for pass, caution or re-cleans. The results of any inspection (positive or negative) should be firmly communicated to relevant members of staff with a corrective action loop in place. Where specific testing, such as microbiological or allergen, testing of product and the environment is required. It is important to have a sampling plan in place so that clear trends can be built up and any problems identified and actioned.
7.14.8
Review
As with any system, it is important to build in a review procedure. The hygiene system should be reviewed on a regular, continuous basis but no less than annually. In addition, a review must be carried out if there is a change in the process or products.
REFERENCES Anaphylaxis Campaign Standard (2007) The Anaphylaxis Campaign Standard to Increase Trust in Information about Food Allergens in Food. The Anaphylaxis Campaign, Farnborough, Hampshire. Curiel G.J., Hauser G., Peschel P. et al. (1993) Hygienic Design of Closed Equipment for the Processing of Liquid Food, EHEDG Report. Campden and Chorleywood Food Research Association, Gloucestershire. Holah J.T., Middleton K.E., Smith D.L. et al. (2004) Cleaning Issues in Dry Production Environment, R&D Report 192. Campden and Chorleywood Food Research Association, Gloucestershire. Holah J.T., Taylor J.H., Holder J.S. et al. (1993) The Spread of Listeria by Cleaning Systems, Part II. Technical Memorandum 673. Campden and Chorleywood Food Research Association, Gloucestershire. Taylor S. and Lehrer S.B. (1996) Principles and characteristics of food allergens. Critical Reviews in Food Science and Nutrition, 36(S), S91–S118.
8
Validation of cleaning and cross-contact
Helen M. Brown
8.1 INTRODUCTION Most food-manufacturing sites produce more than one product using common manufacturing facilities (storage, personnel, manufacturing line, packaging) so there is the potential for an ingredient or product to come into contact with another product to which it is not intentionally added. If cross-contact results in adventitious cross-contamination of a nonallergen-containing food with an allergen, or with a new undeclared allergen, there may be significant implications for both the allergic consumer and the manufacturer. The risk of cross-contact must therefore be controlled carefully, or the consumer be informed through product labelling. In the USA, it is estimated that 8% of product recall actions that took place between 1999 and 2003 were attributed to ineffective use of sanitation principles (Vierk et al., 2002; FDA, 2004). To avoid allergen cross-contamination, the whole operation of a manufacturing site needs to be reviewed to determine how and where allergens may be introduced (including workers lunch boxes as well as the more obvious routes), what route the allergens will take through the site and how they will be handled and by whom. The process of generating an allergencontrol plan is covered in more detail in other chapters of this book. An allergen-control plan will identify potential points of cross-contact and assess the risk of cross-contamination (i.e. that the residue will be transferred). It will also identify prerequisite programmes in the manufacturing process that may provide relevant control measures. Prerequisite programmes are basic operations within food manufacturing that are necessary to ensure good manufacturing practice. Examples of prerequisite programmes relevant to food safety are staff training, supplier quality assurance, pest control and sanitation. Cleaning is a prerequisite that is frequently identified as a process that may reduce or prevent allergen cross-contact. Where this is the case, evidence is needed to show that the cleaning process is effective in reducing or removing the allergen to an acceptable level, i.e. it has to be validated, and it has to be verified on an on-going basis. Validation is required for established cleaning regimes if the allergen-control plan identifies them as critical, but most validation studies will be for new products and processes in which case validation of cleaning regimes should take place prior to commercial manufacture of the product. No one validation study fits all cleaning processes and changes made to the manufacturing process need to be assessed for their possible impact on the efficacy of cleaning. This includes changes in product reformulation, re-configuration of equipment, or changes to the cleaning regime such as cleaning fluids and the time lapse between processes. If such changes are considered likely to affect the efficacy of cleaning, a new validation study of the changed process is required. Even if no changes are made, the cleaning process requires re-assessment
Validation of cleaning and cross-contact
139
at defined intervals, at least annually. Manual cleaning methods are likely to be more variable than clean-in-place (CIP) systems, so should be re-assessed at more frequent intervals.
8.2 VALIDATION OF A CLEANING REGIME Validation of a cleaning regime entails generating data to show that the cleaning process is effective in removing allergens to a pre-defined acceptable level. These data may also inform risk assessments, and will certainly be required for audits of the allergen-control plan where cleaning has been identified as a critical process in preventing cross-contamination. In the pharmaceutical industry, where similar issues exist concerning the safety of products manufactured using shared equipment, there is a requirement to scientifically demonstrate that equipment cleaning processes can remove a given contaminant to below a pre-determined acceptance level (FDA, 1993, 1998). An effective validation study requires careful planning, protocols and documentation. It is best done in the context of an overall allergen-control plan, and after all possible measures to reduce and prevent cross-contact have been implemented. This pre-validation work also helps to support claims about the consistency of the cleaning process. Where limited resources are available, the pre-validation study is crucial to the effectiveness of the validation programme. Clear simple protocols for cleaning validation studies include the aim of the study, the scope of the study, what to measure (the analytical procedure to use), what to sample (the sample type), where and when to sample, how many samples to take and the criteria that define cleaning as effective. Traditionally, in the pharmaceutical industry, a minimum of three validation runs have been performed for cleaning validation, despite there being no statistical justification for this. Ultimately, it is for the manufacturer to decide on the required number of runs of the cleaning process to demonstrate validity. The scope of the cleaning validation study outlines the ingredients and products, equipment (production line, utensils, protective clothing), cleaning regimes and processes that are covered by the study. Validation may not be necessary for every product or piece of equipment and it may be possible to use groupings of similar products and select the worst case. For example, in a validation of the cleaning of equipment used to prepare slurries as an intermediate stage in the preparation of food flavourings, recipes with the highest concentration of celery and peanut were used (Stephan et al., 2004). Grouping on the basis of product and process may enable validation to focus on products that are known to be the most difficult to clean, for example, the most sticky product or the heat process that dries on residues. A standard operating procedure for the cleaning process will cover the details of the equipment to be cleaned, the cleaning procedure including cleaning materials to be used, and any temperature and time holds. It should also make clear the training requirements for staff performing the procedure. Documentation of validation studies depends on the complexity of the system and cleaning procedures. A checklist for the validation protocol (see Table 8.1) is also suitable as a checklist for documentation required. A final validation report outlining the results should be produced and approved by management. In the absence of thresholds or legally permitted levels for allergens in foods (except sulphur dioxide), the manufacturer or retailer is responsible for defining the criteria to establish whether cleaning is effective. The rationale for selecting the allowable residue
140
Management of Food Allergens
Table 8.1
Checklist for a cleaning validation protocol.
The aim of the validation study Responsibilities For performing the validation study For approving the validation study Scope of the validation study Description of equipment to be used. Other equipment that it may apply to with rationale. The product for which the cleaning regime is to be used. Other products that it may apply to with rationale. The cleaning process Standard operating procedure for the cleaning process The interval between the end of production and the beginning of cleaning The amount of product that could be produced before the clean is carried out. General rules of the validation Choose the worst case, e.g. the process temperature likely to cause residues to stick to the surface; the clean to be after products containing the highest concentration of allergen Sampling and analytical testing The type of samples to be taken Sampling locations The analytical method(s) to be used Rationale for why these are to be used Validation requirements for sampling and analytical methods The number of samples to be taken for testing How many cleaning runs to be done to validate the cleaning Acceptance criteria State acceptance criteria with rationale The number of cleans needed to give consistent results to demonstrate efficacy Verification procedures What they will be and the frequency required Re-validation Interval and frequency
limits should be logical, based on the allergen or allergens involved and use current up-todate information. The limits need to be practical, achievable and verifiable, so need to be defined with knowledge of sampling and analytical methods. Very low levels of food allergens cause allergic reactions in sensitized individuals. On this basis, one criterion often applied to the cleaning process is that it removes allergen residues to below the limit of detection of the appropriate, analytical method. Regular reviews of such criteria are required to take account of new information as it becomes available. Historically, the limits of detection of analytical methods decrease as methods improve. This would place increasing demands on the cleaning process. Whether this is necessary to protect the majority of food allergy sufferers awaits further information from studies on allergen thresholds. In its guidance document, the UK Food Standards Agency states ‘Food producers for the time being can adopt “a visually and physically clean” standard for assessing the risk of possible allergen cross-contamination. This requires a thorough visual inspection of the production line (following cleaning) and the final product’. (FSA, 2006). When using such a criterion ‘no quantity of residue to be visible on the equipment after cleaning procedures are performed’, it would be advisable to back this up by performing some testing to determine the concentration of the allergen present when carry-over residues are just visible.
Validation of cleaning and cross-contact
141
8.3 SAMPLING TO VALIDATE CLEANING Sampling must ensure recovery of allergen residues from the equipment surface or from final product, and the level of recovery and repeatability has to be sufficient to meet the criteria set to define cleaning as effective. Where samples are taken, when they are taken and how they are taken can influence the outcome of a validation study – a negative test result may be the consequence of poor sampling technique or cross-contaminants not being homogeneously dispersed throughout the sample, or due to samples being taken from a point where allergen residue is unlikely. The stage at which samples are taken during the cleaning process is significant. If the aim is to demonstrate a comparative decline in allergen residues removed during cleaning, samples need to be taken at suitable time intervals from the onset of cleaning. Samples taken at intervals throughout a clean can be used to demonstrate that allergen residues are not held up within a system and subsequently released at a later stage of the clean. Besides taking samples to show the efficacy of cleaning, samples taken before cleaning can be used to confirm that the sampling and testing methods are capable of recovering and detecting allergen residues, i.e. positive controls. Sampling points will be dependent on the equipment and cleaning process that is being validated. Sampling points are selected to challenge the efficacy of cleaning, targeting difficult to clean areas and areas that present opportunity for significant cross-contact. Selection of sampling points is best done in consultation with engineers and operatives who are familiar with the process. They will have knowledge of the movement of product through the system and points that are difficult to clean or may retain or hold up product. They will also be aware of access points, how to use them and the type of sample that it is possible to take.
8.3.1
Rinse samples – rinse waters, wash waters or purging/flushing materials
Samples of rinse water are taken when there is no other way to access the equipment, for example CIP systems or connecting pipe work. Wash water, particularly if is recycled or part of a common washing point for utensils or trays, could be a vehicle for cross-contamination and should be included in the sampling programme. In isolation, a single sample of final rinse water in which the allergen is below the predefined acceptable level is not sufficient to validate the efficacy of cleaning. It does not provide any information about whether allergen residues have been removed or the level of allergen residues remaining on the equipment. However, if a pre-wash rinse can be sampled, or rinse waters can be sampled from the onset of cleaning, then at intervals through to the final rinse water and these samples show a declining level of allergen, by implication the level remaining on the equipment has been reduced. Sampling of rinse water for subsequent verification of cleaning may then be justified. In a cleaning validation study on CIP equipment used to prepare peanut slurry for flavouring intermediates (Stephan et al., 2004), detection of the allergen in the pre-rinse water confirmed that the sampling and measurement approach were working (equivalent to a positive control). This gave confidence in the sampling and analytical method particularly when, as expected, the allergen was not detected in rinse waters following subsequent alkaline and acid washes and not in the final product (see Table 8.2).
142
Management of Food Allergens
Table 8.2 peanut.
Analysis of rinse water samples and first final product after a clean for the presence of
Stage in cleaning process
ELISA result for peanut (µg/mL)
General protein Bradford Coomassie Plus, Pierce (µg/mL)
Result
Rinse water Pre-wash After alkaline cleaning After acidic cleaning
98.5–1379a ND ND
85.9–917.6a ND ND
Positive Negative Negative
Product Final product after clean
ND – 1.1c
NDb
Negative
Data from Stephan et al . (2004). ND, not detected. a the range from 4 cleans, 3 samples per clean. b the final product after clean was an artificial mix of water and maltodextrin, so did not contain protein. c mg/kg, peanut was detected after only one run and was traced back to an application error during the manufacturing process.
Collecting rinse water samples is not always easy or reproducible. It may involve sampling from a fast-moving stream of sometimes hot water. Once the sample has been collected, it needs to be chilled or frozen and analysed as soon as possible as the allergen may be unstable in water due to sedimentation of microbial growth and lost before it can be tested. Rinse or wash water samples known to contain cleaning fluids may affect the method of analysis, limiting the validity of such samples. In a study of a tray washer as a possible source of cross-contact, samples of wash water containing 1% alkaline foam detergent were taken for testing (Arrowsmith and Brown, 2006). Tests using a positive control sample in the presence of this concentration of detergent gave a false-negative result (see Table 8.3). Alternative sampling points were required to confirm that tray washing was not a route of cross-contamination. In dry manufacturing environments, cleaning may be by flushing or purging a system with a dry, inert material such as flour, diatomaceous earth, salt or starch. Removal of allergen residues can be followed by sampling the purging materials after known volumes have passed through the system. Data that show a decline in the level of the allergen until it reaches the pre-defined acceptable level or below demonstrate the efficacy of cleaning. Table 8.4 shows declining levels of soya when whey was used to purge the system of a soya product (Hefle, 2005). Table 8.3 An example of interference of a cleaning fluid in a wash water sample on the result of on an ELISA allergen test for prawns. Test material
ELISA result for tropomyosin (µg/mL)
Prawn extract diluted to 5 µg prawn/mL Prawn extract diluted to 5 µg prawn/mL in the presence of 1% alkaline foam cleaning fluid
0.08
Prawn extract diluted to 10 µg prawn/mL Prawn extract diluted to 10 µg prawn/mL in the presence of 1% alkaline foam cleaning fluid
0.21
Data from Arrowsmith and Brown (2006). LLOQ, lower limit of quantification.
Validation of cleaning and cross-contact Table 8.4 purging.
143
Demonstration of the decline of soya residues in whey powder used for cleaning by
Purge of soya product using whey powder Purge (volume 1) Purge (2 × volume 1) Purge (3 × volume 1)
ELISA result soya (mg/kg) >5000 527 <0.5
Data from Hefle (2005).
Each of the sampling approaches provides an indication that cleaning is reducing the levels of allergen residue remaining. They rely on comparison to subsequent samples to show a decrease in the level allergen residues. For the pharmaceutical industry, the USFDA (FDA, 1993, 1998) recommends a direct method to sample the cleanliness of the equipment, arguing that it is the residues remaining on the cleaned equipment, not what is in the rinse water that will affect the product that is produced after cleaning.
8.3.2 Direct surface samples Cleaning of areas that are hardest to clean and are reasonably accessible can be validated by taking samples directly from the surface. Equipment and points to target are those likely to trap and build up soil such as dead-legs, coarse or rough welds or joints, O-rings. Sampling may require some dismantling of equipment for access. Other surfaces that are potential sources of cross-contact are the hands and protective clothing of operatives. Swabbing is the most commonly used method to sample from surfaces. An adsorptive material (a swab) is wet using a pre-wetting agent that facilitates the removal and transfer of residues from the surface to an extraction solution. The pre-wet swab is wiped over a defined surface area using a specified action to remove and capture residues. The captured residues are then recovered into an extraction solution ready for analysis. The swab and the pre-wetting agent/extraction solution must be compatible with the analytical method used for detection. Testing unused swabs that have been pre-wetted and extracted confirms that the swabs do not give a false-positive result. Pre-wetting solution spiked with a known amount of allergen can be used to confirm that there are no inhibiting substances in the swab or extraction solution that interfere with the chosen analytical method. As there may be a delay between sampling and testing, especially if samples are to be sent away to a laboratory for testing, the stability of allergen residues in the extraction buffer prior to analysis needs to be known and arrangements for sampling and testing organised accordingly. Swabs specifically for allergen sampling accompanied by a pre-wetting/extraction solution are commercially available. Some of these swabs have been validated for use with specified analytical methods and validation data are available [for example Tepnel (www.tepnel.com, accessed 14 September 2008) and HAVen (http://www.hallmarkav.com/, accessed 14 September 2008)]. For these systems, once the surface is sampled and the swab returned to the pre-wetting/extraction solution, the swabbed samples can be stored at room temperature or below for at least 4 days, or stored frozen below −20◦ C if longer storage time is needed. Ideally, the swabbing method gives a consistent recovery for a specified allergen present at a specified concentration on a surface. For cleaning validation in the pharmaceutical industry,
144
Management of Food Allergens
the FDA recommends that this overall recovery is known. Work at Campden BRI has shown recovery of food allergens from surfaces using swabs can be very variable, depending on the operator, the surface area covered by swabbing and how residues are extracted from the swab. If the recovery of the allergen in the swabbing technique is unknown, the results obtained from sampling and testing can only be indicative of the extent of the clean. Schlegel et al. (2007) measured the recovery of peanut proteins allowed to dry on stainless steel coupons R flat zapped-head foam swabs. Contaminant levels on the stainless steel using Critical Swab plates ranged from 1 to 1.2 µg/cm2 and recovery of the protein was measured using a general protein assay. Using a single swab, recovery varied from 45 to 75%, suggesting that a single swab could not uniformly sample the specified area. By swabbing the area for a second time using a new swab, recovery was increased to 93%. This level of recovery was considered to be comparable with recoveries obtained in pharmaceutical settings. It illustrates that a larger swab or more swabbing steps may be required to recover higher levels of contaminants or to sample larger surface areas. Other direct surface sampling methods include the physical removal of residues that are insoluble or dried, coupons placed on surfaces that can subsequently be removed and residues rinsed off using a known volume of extraction solution or pieces of protective clothing cut from heavily soiled areas and subjected to the extraction steps in the analytical method. Allergens in the form of dry powders, for example nut dusts, flours, milk and whey powder, egg powder, can be readily moved about in air flows or aerosols generated when cleaning. The potential for sprays and compressed air jets to transfer material from one surface to another is widely recognised. Provided sufficient time for settlement of particles has been allowed, direct sampling from surfaces is one way to confirm this movement. Alternatively, settle plates located at ‘worst case’ points can be used to collect and weigh material settling over a defined period. The potential cross-contamination risk ‘due to cleaning’ can then be estimated.
8.3.3 Final product samples The first product taken from the manufacturing line after a clean is an essential sample for most cleaning validation studies. Residues of the previous product remaining after cleaning are likely to be concentrated in the first product to exit the line following cleaning. At least the first three units that are produced following the clean should be sampled for testing. Samples taken from the middle and at the end of the run will show if there is further release of residues. The final product represents the recovery of residual allergen and its dilution in the new product. If different portion sizes of product are produced, then each portion size should be sampled. If after cleaning, the presence of a cross-contact allergen in the final product is consistently detected at a level greater than the pre-defined limit, then the cleaning regime is not effective. The cleaning process needs to be amended and validation of the new regime started again. Alternatively, other allergen management routes such as scheduling, segregation and use of dedicated lines need to be revisited or considered. It may also be advisable to check that raw materials are not inadvertently the source of the apparent cross-contamination. If the allergen is not detected in any sample after cleaning, the level of allergen residue in the sample is not greater than the detection level of the analytical test. It does not mean that no allergen residues remain after cleaning.
Validation of cleaning and cross-contact
8.4 8.4.1
145
WHAT TO MEASURE TO VALIDATE CLEANING Visually clean
Visual inspection is one fairly quick and simple way to get an indication of whether foodcontact surfaces, floors, walls, overhanging pieces of equipment and operatives (hands and coats) are a source of cross-contact. It can also be used to identify areas to target when sampling. Dust from flours and pre-mixes that settle on equipment surfaces can easily be seen, as can food particles such as prawns, sesame seeds and nuts, and films that dry out on surfaces after manual cleaning can indicate areas that were not scrubbed or wiped. Whilst there is no evidence to correlate the presence of visible residues with the presence of allergen residues, the presence of visible residues tends to suggest a failure to clean. Visual inspection can be used to target sampling of accessible ‘notoriously’ difficult to clean areas such as tight corners in trays and bins, coarse or rough welds or joints and O-rings for the build-up of soil. Visual assessment does have limitations and it is unlikely that it can be relied upon as the only way to detect the presence of allergen residues that will lead to cross-contamination in the final product. Not all surfaces are accessible, so its use is limited to areas that can be visually inspected. The ability to see food soils, for example egg, can be very much dependent on the lighting and how light is reflected or the colour of the surface on which the film is viewed – milk soils form a white layer readily visible on stainless steel surfaces but much harder to discern on a white plastic conveyor. The fastidiousness and visual ability of inspectors will differ, so consistency may be an issue. Whilst the value of visual assessment should not be underestimated, a negative finding cannot be relied upon, and in a validation study it should be confirmed by other analytical methods.
8.4.2 Analytical tests Ideally, analytical methods used to measure or detect allergen residues after cleaning should be specific for the particular allergen. This is not always possible as specific methods for all allergens are not readily available. The alternative is to use analytical methods that measure the presence of another analyte, such as DNA, total protein or ATP, as a marker for the presence of an allergen. In some circumstances where allergens may be denatured leading to a false-negative result, these methods may be preferred or used in combination with the allergen-specific tests. Using a combination of tests can confirm the validity of the methods selected for validation and for ongoing verification. For example, in a study to monitor cleaning of equipment used for preparation of a celery slurry, washing water samples were tested using a general protein test and a celery-specific DNA-based real-time polymerase chain reaction (PCR) test. Despite protein being measured in all pre-wash washing water samples, PCR products were detected in only 3 of the 12 samples. It was unclear whether the apparent failure of the DNA-based test was due to the conditions in the celery processing increasing DNA degradation or because of the presence of PCR inhibitors in the template DNA (Stephan et al., 2004). By using a combination of methods, the limitations of the DNA-based test were realised. Most allergens with exception of sulphur dioxide are proteins and are therefore susceptible to conformational changes and denaturation. This can affect solubility and therefore recovery during extraction. It can also affect whether the allergen is detected in the analytical method. This can be a considerable limitation when using specific allergen testing methods to validate cleaning.
146
Management of Food Allergens
Analytical methods that are currently readily available with the required sensitivity and specificity to detect specific allergen residues after cleaning are enzyme linked immunosorbent assays (ELISAs). Other methods such as mass spectrometry, surface plasmon resonance (SPR) and electrophoresis are either not readily available or amenable or do not have the required sensitivity or specificity. ELISA test kits, each for a specific allergen, are commercially available from different manufacturers. Availability currently covers many, but not all, of the allergens in the EU allergen-labelling directive (Directive 2003/89/EC). Allergen-testing methods are covered in more detail elsewhere in this book. The specificity of ELISA and SPR tests is achieved by the use of antibodies that recognise a particular region of the protein (an epitope) that is unique to the allergen. In the test, antibodies are used to capture the allergen by binding at the epitope. The presence of trapped allergen is then measured in the test. Changes to specific epitopes on allergen proteins mean that antibodies no longer recognise and capture the protein, so its presence is not detected. Cleaning regimes that use very hot water, acidic or alkaline washes, powerful oxidising agents or quaternary ammonium compounds may result in changes to protein epitopes such that allergen residues are no longer recognised. If this is the case for a particular cleaning regime, then an alternative analytical method to validate cleaning is required. If used to test the first product produced following a clean (a final product sample), the selected method must be capable of detecting the carry-over allergen residue at low levels in the food matrix of the final product. This can be confirmed by spiking a portion of the first final product off the line with the allergen. There may be a number of factors that influence whether the carry-over allergen will be detected in the final product. For example, in the presence of soya milk, gluten is not detected when extracted using 60% (v/v) aqueous ethanol and tested in the R5 Mendez ELISA (Malmheden Yman, 2006). If validating a cleaning process for the removal of gluten, and the product produced after the clean contains soya milk, an alternative extraction (the cocktail solution) is required to avoid false-negative results. The method must be shown not to give false positives due to interference from the final product. For example, sunflower seeds are known to give false positives in one commercially available ELISA test for hazelnut. Obviously, this method would be unsuitable for final product testing when validating cleaning to remove hazelnut residues if the following product contained sunflower seeds. ELISA tests can be affected by the presence of detergent residues present in samples taken for testing. Detergent residues which remain on surfaces after cleaning and are collected during allergen sampling such as terminal sanitizers or cleaning fluids present in rinse or wash waters may interfere directly in the ELISA (see Table 8.3). Again an alternative analytical method may have to be used to validate cleaning. Tests to validate cleaning for the purpose of hygiene, based on detection of ATP, are familiar to the food industry. They are faster than allergen tests and less expensive, but whilst there is a considerable historical data for use of ATP to validate and verify cleaning processes for hygiene, its suitability for validating cleaning for allergens cannot be extrapolated. To use ATP as a marker for allergens, data showing a correlation between ATP and allergen residue level are required. ATP tests detect both microbial ATP and ATP associated with the food residue. ATP is present in living cells as a source of energy, so the levels of ATP associated with a food residue vary depending on the food type, for example a muscle-based food will contain higher levels of ATP than do a non-muscle-based food such as milk powder. The variation of ATP levels in different food residues suggests that using the presence of ATP as a marker for the presence of allergen residues following cleaning may be suitable in some situations and not others. When comparing methods to assess cleaning of surfaces following
Validation of cleaning and cross-contact Table 8.5
147
Comparison of the sensitivity of different analytical methods when used to validate cleaning.
Analytical method applied to sample collected by swabbing
Positive control No clean
Interim clean
Full clean
ELISA result Tropomyosin (µg/mL) General protein (Coomassie Plus Bradford, Pierce) (µg/mL)
>ULOQ Positive >ULOQ Positive
>ULOQ Positive
Hygiene protein (Pro-tect TM , Biotrace) (RCU)
540 Positive
250 Negative
225 Negative
Hygiene ATP (Clean-TraceTM , Biotrace) (RLU)
82,000 Positive
2800 Positive
27 Negative
ULOQ, upper limit of quantification; LLOQ, lower limit of quantification.
production of products containing prawns (see Table 8.5), swabs were taken from surfaces before cleaning (positive control), after a wet interim clean and after a full clean. The allergen (tropomyosin) was detected in the positive control and after the wet interim clean. It was no longer detected after the full clean. The ATP test was positive for the positive control and after the interim clean (at a reduced level), and was negative after the full clean. The magnitude of change in the ATP signal measured between the positive control, the interim clean and the final clean suggests that in this case, ATP may be useful as a marker of allergen residues remaining after cleaning. However, when used to detect milk residues allowed to dry on stainless steel plates, ATP tests did not detect residues until 125 mg milk protein/L was applied, whereas the allergen test detected when 0.025 mg milk protein/L was applied (Arrowsmith et al., 2006). In this situation, ATP testing was not suitable as a marker when compared to the specific allergen test. A similar conclusion that conventional ATP tests lacked the sensitivity of ELISA tests was reached in other studies using milk in solution dried onto a stainless steel surface (Salter et al., 2005). Work using an ATP test with greater sensitivity than conventional ATP tests has indicated that in terms of comparable sensitivity to ELISA methods, it may be useful for validating wet cleaning procedures (Jackson et al., 2008). In practice, generating data to demonstrate the correlation between ATP and allergen tests on a case-by-case basis may be more onerous than using a targeted allergen test. Acceptance criteria for ATP are likely to vary depending on the product. Multiple acceptance criteria for ATP within a manufacturing site requires additional staff training and may be an unhelpful source of confusion leading to mistakes. Caution is advised if adopting this approach. The presence of total protein has been used as a marker of the presence of allergens in cleaning validation studies and for ongoing verification of cleaning. Total protein methods are non-specific so measure microbial protein as well as residual food proteins. Therefore, these methods are applicable only to rinse water samples and direct surface samples. They are unsuitable for final product testing unless the final product does not contain any ingredients that include traces of protein. As the test does not discriminate between allergenic and non-allergenic protein residues, it may lead to more cleaning than necessary to remove allergen residues, for example for ‘high protein but not allergen’ foods. However, the lack of specificity can also be an advantage, in that total protein tests are universally applicable for allergen cleaning validation studies, removing the necessity to identify the allergen to test for and an analytical method suitable for that allergen. This can greatly simplify testing method
148
Management of Food Allergens
requirements in protocols for validation of allergen cleaning. Total protein tests are generally cheaper and faster to perform than specific allergen tests. Pre-prepared test solutions and standard solutions, usually of bovine serum albumin are commercially available, for example Coomassie Plus BradfordTM assay kits (Pierce), bicinchonic microassay (MicroBCA) (Pierce). Simple swab-based tests for hygiene monitoring designed for use at line are availR (Biotrace), but for cleaning validation or verification of cleaning able, for example Pro-tect the sensitivity of a general protein test needs to be comparable to that of specific allergen tests. Simple swab-based tests designed for hygiene monitoring for use at line are available but the quoted range for detection 500–5000 µg/mL BSA equivalent is not comparable to the sensitivity of laboratory-based protein or allergen tests. The sensitivity of laboratory-based general protein tests can be as low as 0.1 µg/mL for the bicinchonic microassay (MicroBCA) (Schlegel et al., 2007) and 1 µg/mL for the Coomassie Plus BradfordTM assay kits (Pierce). The sensitivity of specific allergen-based tests is of the order of 1–3 µg/mL or ten times less than this when applied directly to rinse waters and direct surface samples where no sample extraction step is required. The sensitivity of a general protein test (Coomassie Plus BradfordTM assay) was claimed to be similar that achieved with a specific ELISA test for peanut when used in a cleaning validation study of CIP equipment used to prepare a peanut slurry (Stephan et al., 2004). The magnitude of results for the two tests followed the same trend in pre-wash rinse water samples where the level of residues was high. However, residues were not detected using either method in rinse waters after cleaning, so no comparison at low allergen levels could be made (see Table 8.2). In a cleaning validation study of a sandwich conveyor, the sensitivity of an ELISA test for tropomyosin and the Coomassie Plus BradfordTM total protein assay was compared by testing before cleaning, after an interim clean and after a full clean (see Table 8.5). Residues were detected by both methods in the positive control but after the interim clean they were not detected in the total protein test, whereas they were detected in the ELISA (Arrowsmith and Brown, 2006), indicating that the sensitivity of the ELISA was greater than the total protein test. The ELISA appeared to be more suited to validating cleaning in this case. Again, this shows the need to select methods for cleaning validation on a case-by-case basis and the value of using a combination of methods, especially to inform the methods to use for ongoing verification of cleaning.
8.5
SUMMARY
To demonstrate that cleaning processes are effective in preventing or reducing allergen cross-contact, appropriate validated sampling and analytical methods are needed, including information about the percentage recovery of allergen residues when sampling and analysis is combined. Protocols for generating suitable cleaning validation data have to be developed on a case-by-case basis in the context of an allergen-control plan. Data are likely to be gathered using a combination of different sampling approaches and methods to detect allergen residues or markers of allergen residues. The validity of the data is enhanced by performing cleaning validation studies in the context of an allergen-control plan.
REFERENCES Arrowsmith H.E. and Brown H.M. (2006) Effect of cleaning fluids on detection of allergens. Campden and Chorleywood Research Summary Sheet, 2006–67.
Validation of cleaning and cross-contact
149
Arrowsmith H.E., Smith D., Holah J. and Brown H.M. (2006) Detection of milk allergen on a stainless steel surface: interpretation of swabbing results. Campden and Chorleywood Research Summary Sheet, 2006–70. Directive 2003/89/EC of the European Parliament and Council (2003) Official Journal of European Union L308/15, 25.11.2003. Available at http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2003: 308:0015:0018:EN:PDF, accessed 21 April 2009. FDA (1993) Guide to Inspections of Validation of Cleaning Processes. Reference material for FDA Investigations and Personnel. United States Food and Drug Administration, Washington, DC, pp. 1–6. FDA (1998) Guidance for Industry: Manufacture, Processing of Active Pharmaceutical Ingredients. United States Food and Drug Administration, Washington, DC, pp. 1–53. FDA (2004) Food GMP Modernization Working Group: report summarizing food recalls, 1999–2003. United States Food and Drug Administration. Available at http://www.cfsan.fda.gov/˜dms/cgmps2.html, accessed 14 September 2008. FSA (Food Standards Agency) (2006) FSA Guidance on Allergen Management and Consumer Information. Best Practice Guidance on Managing Food Allergens with Particular Reference to Avoiding Crosscontamination and Using Appropriate Advisory Labeling (e.g. ‘May Contain’ Labeling). Appendix IV: Allergen testing Methods. Available at http://www.food.gov.uk/multimedia/pdfs/maycontainguide.pdf, accessed 14 September 2008. Hefle S. (2005) Allergen Detection Methods. Food Allergy Research and Resource Program. Woodhead Publishing, Cambridge. Jackson L.S., Al-Taher F.M., Moorman M. et al. (2008) Cleaning and other control and validation strategies to prevent allergen cross-contact in food-processing operations. Journal of Food Protection, 71(2), 445–458. Malmheden Yman I. (2006) Analysing gluten in food. In: Hidden Food Allergens Testing the Limit. Tepnel Biosystems, Central Hall Westminster, London. Salter R.S., Hefle S., Jackson L.S. and Swanson K.M.J. (2005) Use of a sensitive adenosine triphosphate method to quickly verify wet cleaning effectiveness at removing food soil and allergens from food contact surfaces. Abstract of Annual Meeting of International Association for Food Protection 2005. International Association for Food Protection, Des Moines, Iowa, P2–06. Schlegel V., Yong A. and Foo S.Y. (2007) Development of a direct sampling method for verifying the cleanliness of equipment shared with peanut products. Food Control, 18, 1494–1500. Stephan O., Weisz N., Veiths S. et al. (2004) Protein quantification, sandwich ELISA, and real-time PCR used to monitor industrial cleaning procedures for contamination with peanut and celery allergens. Journal of AOAC International, 87(6), 1448–1457. Vierk K., Falci C., Wolyniak C. and Klontz K.C. (2002) Recalls of foods containing undeclared allergens reported to the U.S. Food and Drug Administration, fiscal year 1999. Journal of Allergy and Clinical Immunology, 109, 1022–1026.
9
Validation, standardisation and harmonisation of allergen activities in Europe and worldwide
Bert Popping
Since allergen regulations have become the focus of public opinion as legislators have implemented numerous directives for labelling of allergens, allergen testing has increased exponentially. Typically, ELISA techniques are being used but some older (e.g. Rocket Electrophoresis) and newer techniques (e.g. LC-MS/MS) are also being employed.
9.1
ANALYTICAL METHODS
At present, the overall number of allergen tests using ELISA exceeds the number of tests performed using other methods. However, none of the techniques mentioned above are perfect and all have their drawbacks. Especially with ELISA, specificity issues are prominent. This becomes particularly apparent with ELISA kits for gluten where the antibodies used in commercial kits have different targets and give in processed products often significantly different responses, varying from 50 to 5000 ppm in the same sample. Not only does this have limited use for the food manufacturer wanting to assess the effectiveness of cleaning procedures and reassure consumers of the absence of allergens in their products, but it also makes legal enforcement impossible.
9.2
METHOD VALIDATION
In order to provide useful information, there is an urgent need to have reliable results which will stand up in court and also be useful for assessing the effectiveness of cleaning procedures. Typically, this is done first in single laboratory analysis (IUPAC Technical Report, 2002) and later on in validation studies in accordance with AOAC/IUPAC harmonised guidelines (IUPAC, 1995). However, the first approaches to validating allergen detection assays were performed by the European Commission’s Joint Research Laboratory only as recently as 2004 (Poms et al., 2005), validating several peanut kits. Generally, all five ELISA test kits performed well in the concentration range of 5–10 mg/kg, rather than at lower concentrations (2.0 or 2.5 mg/kg). The variation in the found recoveries of peanut between the different test kits had a range of 44–191% across all concentrations. The quantification characteristics between test kits differed significantly at the very low mg/kg level. Two test kits performed well even at concentrations below 5 mg/kg with reproducibilities of 27–36% for biscuits and 45–57% for chocolate.
Validation, standardisation and harmonisation of allergen activities in Europe
151
This validation happened more than a year after the first allergen-labelling regulation was published (2003/89/EC) and only validated ELISA kits for the detection and quantification of peanuts in certain food products. On the list of Annex IIIa of the food allergen-labelling regulation, 12 groups could be found (of which not all fall into the category of IgE-mediating allergens). None of the kits for any of the tree nuts (of which nine nuts are named in the annex) had been validated at that time in inter-laboratory validation study. Also, none of the kits for detecting cereals containing gluten, crustacean, eggs, fish, soya bean, milk, celery, mustard or sesame had undergone any inter-laboratory study at that time. One would have thought that with the implementation of a European Directive the very tools needed to control carry-over of allergens would have been made available as fully validated ELISA kits to be used by food manufacturers and enforcement officers alike. With the European Community having more member states than groups of allergens in the Annex IIIa list, it should have been easily possible to spread the cost and effort for full validation studies between the member states, thereby allowing enforcement of the directives which the members were duty bound to implement.
9.3 STANDARDISATION OF METHODS In 2004, Germany became the first country to start and successfully validate two hazelnut ELISA kits. Through the German Official Methods Group §64 LFGB (formerly known as §35 LFGB) a ring-trial was conducted in accordance with AOAC/IUPAC harmonised guidelines which demonstrated that the two validated kits were fit for purpose to detect hazelnut in dark chocolate. It was later found that the kits exhibit different specificities to non-allergenic ingredients and that an assessment of cross-reactivities to a wide range of related ingredients was necessary to minimise the risk of false-positive results.
9.3.1
CEN
The validation data were submitted to CEN (European Standardisation Committee) for consideration as European standards. The CEN Committee, consisting of three national experts per EU member country, for detection methods of allergens in foodstuffs (CEN/TC 275/WG 12) was established in 2003 and spent several years developing base protocols for the standardisation of methods. In this committee, four subgroups were formed: (a) General Considerations and Validation of Methods; (b) Molecular Biological Methods; (c) Chromatographical Methods; and (d) Protein-based methods for immunological detection of allergenic food and food ingredients (immunology). The General Considerations and Validation of Methods ad hoc group is dealing with terms, definition, preparation of samples, extraction aspects, modifications and changes to a submitted assay, interpretation and expression of results and reference materials to name but a few. The Molecular Biological Methods ad hoc group is concerned with specific aspects for detecting the DNA of food allergens. Here, particular aspects of laboratory layout, conditions for polymerase chain reaction (PCR) are considered. This group also deals with any molecular biological method submitted for assessment as a standard. The Chromatographical Methods group is currently looking at developments in the massspectrometry field for detecting allergens. Here, the technology has been successfully applied
152
Management of Food Allergens
in some laboratories in the USA and Canada but is not yet ready for routine application or to become a standard. It is unlikely to see any standards in this area in the short term. The Immunological ad hoc group is concerned with particular aspects for immunological methods, like variability of ELISA assays, method specifics and actual methods submitted for consideration as a standard, for example the two hazelnut ELISA assays which were validated by the German §64 LFGB group.
9.3.2
Allergen community
Similar to the working groups in Europe for standardisation of allergen detection methods, a group of AOAC International (formerly known as the Presidential Taskforce on Food Allergens and now renamed as the Allergen Community) has developed a draft validation protocol which will shortly be finalised. The Allergen Community has brought together government agencies from Europe, Canada and the USA, the food industry, kit manufacturers and testing laboratories. The purpose of this validation protocol is to provide faster procedures for validation, streamlining the process and making it more cost effective. The effort of developing this validation protocol specifically for allergen detection methods also took several years and still requires agreement by all parties involved. However, it will be a milestone on the path towards enforcement, by providing fully validated detection methods for food allergens. Initially, it is likely that the focus, once agreement is achieved, will be on egg and milk as the target allergens.
9.4 HARMONISATION Such a protocol would also feed into a newly funded sixth framework project of the European Commission called MoniQA (www.moniqa.org, accessed 31 March 2008). This project has the goal to harmonise all aspects of certain topical subjects, of which an important one is allergens. This working group is lead by Eurofins and co-lead by the Institute of Food Research. In the allergen working group, all aspects of allergens are addressed. These include consumer views on the meaningful labelling of allergens, the safety of foods which are not labelled or those only labelled with ‘may contain’ statements. It includes industry views on the costs of changing labels and implementing appropriate HACCP concepts to control allergens, information to the consumer and the numerous changes in legislation which often requires re-labelling products. It also deals with the regulatory aspects such as enforcement of labelling regulations in the absence of fully validated methods, and consumer advice and information. From the kit manufacturers and testing laboratories perspective, it covers issues such as appropriate validation, fitness for purpose, measurement of uncertainty and confirmatory analyses. In the first phase of the project, the status quo will be assessed and gaps in the different areas identified. In the second phase of the project, ways forward to bridge these gaps will be discussed. These are likely to include recommendations to legislators, industry, kit manufacturers, laboratories and consumers on how to avoid situations where confusion has arisen in the past. The MoniQA project is transnational, transorganisational and transinstitutional and as such as the potential to achieve a harmonisation level beyond method-focused groups such
Validation, standardisation and harmonisation of allergen activities in Europe
153
as CEN, or regulation-focused groups such as national or EC-legislators, or industry- and consumer-focused groups working on their own. The current phase of the project has already demonstrated that there are numerous gaps to close. Activities in CEN, AOAC, §64 LFGB, the food industry, with consumer groups, kit manufacturers and testing laboratories are all helping to make progress in this area. This should help provide meaningful information to industry and consumers and make enforcement of allergen directives possible.
REFERENCES IUPAC (1995) Protocol for tech design, conduct and interpretation of method performance studies. Pure and Applied Chemistry, 67(2), 331–343. IUPAC Technical Report (2002) Harmonized guidelines for single-laboratory validation of methods of analysis. Pure and Applied Chemistry, 74(5), 835–855. Poms R.E., Agazzi M.E., Bau A. et al. (2005) Inter-laboratory validation study of five commercial ELISA test kits for the determination of peanut proteins in biscuits and dark chocolate. Food Additives and Contaminants, 22(2), 104–112.
10
Standardisation of analytical methodology with special reference to gluten analysis
Ingrid Malmheden Yman
10.1 INTRODUCTION Food safety is the priority for consumers with food allergies and/or food intolerances. Safety of food products is also an issue for the producer, i.e. the industry. Food authorities’ principal mandate is to ensure safe food for consumers. Therefore, it must be assured that the measurement results of analytical methods are reliable and relevant for the investigated food. Various methods used for the same sample need to generate reliable, reproducible and comparable results. The detection and particularly the quantitative determination of proteins such as milk, egg, nuts, peanuts and gluten in food products can be influenced by several factors. Interactions with compounds from the food matrix, e.g. polyphenols (tannins) could be one such factor. Many of the conventionally used food-processing technologies have also an important impact on the allergenicity and antigenicity of allergenic foods (Poms and Anklam, 2004). For antibody-based methods, differences in antibody specificity and affinity play an important role for the final result as well as cross-reactivity to related and even non-related food components.
10.2 METHODS AND STANDARDS 10.2.1
Needs for reference materials
Reference materials (RMs) offer the possibility to compare results obtained in different laboratories, by different operators, using different or the same methods. An RM must be stable and useful for calibration and standardisation of methods and is necessary to achieve a relevant degree of safety. An RM is a material or substance whose properties are sufficiently homogeneous and well established to be used for the assessment of a measurement method. A certified reference material (CRM) is accompanied by a certificate which establishes its traceability to the unit in which the property values are expressed. Each certified value is accompanied by an uncertainty at a stated level of confidence. CRM and non-CRM are used for method development as well as for validation and calibration. Such material can also be used within a laboratory for statistical quality control, e.g. for establishing control charts. All these applications allow an increasing comparability of measurements between laboratories and improving the confidence in analytical data.
Standardisation of analytical methodology with special reference to gluten analysis
10.2.2
155
Reference materials for food allergens
So far no material for food allergens fulfils the quality criteria described above for CRMs has become available. It is important to have real food matrix materials, not only food matrices spiked with extracts of the allergenic food available. The development and production of such RMs for the analysis of food allergens faces several challenges and the availability of the perfect RMs for allergens in food rarely exists. However, even if the perfect RM does not exist, other material can provide a common reference point for comparison purposes. Several food matrix CRMs are available from producers of RMs, e.g. National Institute of Standards and Technology (NIST) in the USA and Institute for Reference Materials and Measurements (IRMM), European Commission in Belgium. Such CRMs may also contain allergens such as milk powder, egg powder and peanut butter. Even though these materials are certified for components other than allergens, they can be used as allergen RMs.
10.2.3
CEN
In 2003, the European Standardisation Organisation (CEN) established a new working group, WG 12, to elaborate standards for ‘allergens’. The scope was adopted by the CEN Technical Committee 275 in June 2003 and reads ‘Standardisation of methods for the detection and determination of potentially allergenic substances in foodstuffs or marker for their presence, including but not limited to immunological and molecular-biological methods’.
r
r r
CEN is a non-profit technical organisation set up under Belgian law and is composed of the National Standards Bodies of 28 European countries. In addition to full members, CEN has also associated and affiliated members. The European Commission and the EFTA secretariat act as CEN’s Counsellors for policy issues (www.cenorm.be). Although most standards are initiated by industry, a significant number has been developed to support European legislation. CEN has an agreement with the International Organization for Standardization (ISO) through which common European and International Standards can be developed in parallel. Once the draft of a European Standard reaches a mature stage, it is released for public comment, a process known in CEN as the CEN enquiry. After adoption and publication of a European standard, it becomes a standard in the 28 member countries of CEN.
To be adopted as a standard, the method has to be validated, preferably through a collaborative study including a certain number of laboratories analysing a certain number of samples (Collaborative study guidelines, 1995). However, methods which have been ‘in house’ validated by a single laboratory can also be considered. Another requirement for being adopted as a standard is full transparency. In the case of DNA methods, information on target sequence, primers, probes, buffers, etc., must be given. In the case of immunological methods, full details of all buffers must be given as well as information about the specificity of the antibodies (monoclonal or polyclonal). The order of priority for which standards should be progressed, as selected by CEN working group 12, was peanut, hazelnut, milk, egg, cereals and soya. The foods are based on the list appearing in the EU labelling Directive 2003/89/EC.
156
Management of Food Allergens
10.2.4
Codex Committee on Methods of Analysis and Sampling
Codex Alimentarius Commission is the Joint FAO/WHO food standards programme. One of the general subject committees, Codex Committee on Methods of Analysis (CCMAS), defines criteria appropriate for methods and serves as coordinating body with other international groups working on methods of analysis and sampling as well as quality assurance for laboratories. The committee further specifies reference methods of analysis and sampling appropriate to Codex standards and considers, amend and endorse, as appropriate, methods of analysis and sampling proposed by Codex Committees.
10.2.5 Methods for allergens Methods for detecting various food allergens have been available for a number of years. Many of these methods involve the use of ELISA-based techniques to detect specific proteins in food matrices. Validation data from the method developer must be available to the user to show that the methods perform as claimed, i.e. that the methods ensure reliability, accuracy and precision. To guide method developers of ELISA kit for allergens, a document for a harmonized validation protocol has been elaborated by a working group under the auspices of the AOAC Community on Food Allergens (Abbott et al., in press).
10.2.6
Accreditation
Laboratories performing official food control must be accredited according to ISO 17025. However, in many countries, like the Nordic countries, the methods used must also be accredited. The laboratory must present in-house validation data to the accreditation body and calculate inter- and intra-laboratory variation.
10.2.7
Certification of test kits
In 1991, the AOAC Research Institute (AOAC RI) was established as a subsidiary of AOAC International. The AOAC RI administers a performance-testing programme for test kits including a review of the test kit performance claims. Test kits found to be in conformance with their claims are granted ‘performance-tested methods’ status, which assures the test kit user that an independent assessment had been conducted demonstrating that the test kit performs as claimed. The programme provides also a rapid entry point into the AOAC validation process and provides a quicker approval process. Several test kits for allergens have been approved by AOAC RI.
10.2.8
Proficiency testing
It is a requirement of ISO 17025 accreditation that laboratories take part in proficiencytesting schemes, if suitable schemes exist. For laboratories entrusted with official control of food and feeds, Article 12 of EU Regulation (EC) 882/2004 requires such laboratories to be assessed and accredited in accordance with ISO 17025, i.e. proficiency testing is a legal requirement for these laboratories. Together with the use of validated methods, proficiency testing is an essential element of laboratory quality assurance and an independent proof of
Standardisation of analytical methodology with special reference to gluten analysis
157
competence. For several years the Food Assessment Proficiency Analysis Scheme (FAPAS) has been distributing samples for allergen analysis in ‘FAPAS Proficiency Test 27 Allergens’. Also, a company in Germany (DLA, Dienstleistung Lebensmittel Analytik GbR) recently announced that they provide proficiency samples for allergen testing.
10.3 GLUTEN ANALYSIS 10.3.1
Classification of cereals
Cereals are member of the grass family Poaceae (earlier Gramineae) and are taxonomically classified from family to subfamilies down to tribe and species. Wheat, rye and barley all belong to the tribe Triticeae, while oat belongs to another tribe, Aveneae, rice to Oryzae, millet to Paniceae and teff, an African grass to the tribe Eragrostideae (see Figure 10.1). Early studies by Osborne (1926) categorised wheat cereal proteins into four fractions on the basis of their solubility in a series of solvents. Albumins are water soluble, globulins are soluble in dilute salt solutions and gliadins are soluble in aqueous alcohol, while the glutenins (glutelins) are soluble in dilute alkali or acids.
10.3.2
Gluten proteins
Gluten is the grain storage protein, originally in wheat, but the word gluten is also used for the grain storage proteins in rye, barley and oats. Gluten is a cohesive elastic mass, which remains if starch granules are washed out from a water–flour dough. Two main protein fractions constitute the gluten: the alcohol-soluble gliadins and the insoluble glutenins. The gliadins are sometimes referred to as prolamins from wheat, while those from rye are named secalin, those from barley hordeins and those from oat avenins. The ethanol-soluble gliadins comprise over 40 structurally similar proteins, which can be further classified into α-, γ- and ω-gliadins on the basis of their electrophoretic mobility (Shewry and Halford, 2002). The glutenins can be differentiated by their molecular weights into low and high molecular weight subunits (see Figure 10.2).
Poaceae (Graminae) The grass family Subfamily
Bambusoideae Chloridoideae
Panicoideae
Pooideae
Tribe
Oryzae
Rice
Eragrostideae
Teff
Paniceae
Millet
Triticeae
Aveneae
Wheat Rye Barley
Fig. 10.1
Relationship between some important cereals from the grass family.
Oat
158
Management of Food Allergens
Water–flour dough Solid material: ~15% protein ~75% starch granules Knead in excess water to wash out starch granules
Gluten Cohesive elastic mass, consisting mainly of storage proteins Soluble
Gliadins
Insoluble
Extract with alcohol–water solution
Glutenins
α-type
LMW-subunits
γ-type
HMW-subunits
ω-type Prolamins of wheat – gliadins Prolamins of rye – secalins Prolamins of barley – hordeins Prolamins of oats – avenins Fig. 10.2
Gluten protein fractions. LMW, low molecular weight; HMW, high molecular weight.
The name prolamin reflects the fact that the grain storage proteins are notably rich in the amino acids proline and glutamine, which are present in highly repetitive sequence motifs. Also unique sequences exist. The gliadins and glutenins are responsible for the physical dough structure such as viscoelasticity, water binding and gel building by interaction between the two. This interaction allows air bubbles to be trapped inside, providing strength for the bread dough to rise.
10.3.3
Coeliac disease
Gluten intolerance or coeliac disease (CD) is a complex autoimmune disease, which affects genetically disposed individuals. The disease is linked to certain histocompatibility genes, which appear in around 35% among humans. In normal individuals, as well as in coeliacs, digestion in the gastrointestinal tract breaks down gluten into smaller peptides that are absorbed by the small intestine. However, one peptide of 33 amino acids is resistant to acids and enzymes. One enzyme, tissue transglutaminase (tTG), has the ability to modify the peptide in such a way that glutamine acid residues are deamidated (see Figure 10.3). The peptides pass through the epithelial cell wall and are then attached to cells, carrying the histocompatibility genes HLA-DQ2 or HLA-DQ8 (Shan et al., 2002.). These cells present the gluten fragment to T-cells, which stimulate the immune cells to attack the intestine. A defence mechanism is triggered, leading to inflammation and damage of the villi. At the same time, specific antibodies of IgA and IgG type are present locally in the mucosa as well as in the blood.
Standardisation of analytical methodology with special reference to gluten analysis
159
Gliadin is degraded by enzymes to smaller peptides A peptide of 33 residues resists proteolysis
Food proteins are digested in the stomach and the intestine
tTG
The peptide is a potent T cell stimulator after deamidation by tissue transglutaminase (tTG) Fig. 10.3
Antigen presentation to T cells of a 33 amino acid peptide.
Nearly all people suffering from CD carry one or both of the immune molecule types DQ2 or DQ8, in fact 90% carry the DQ2 molecules and most of the rest carry the DQ8 molecules. These are genetically inherited and genetics determines whether a person carry either one or both these disease-associated HLA antigens. Other contributing factors are almost certainly involved in the onset of the disease because for unknown reasons there are a few people with the disease not having either DQ2 or DQ8 HLA types. There are also people carrying DQ2 and DQ8 HLA types who do not develop the disease (Sollid, 2000).
10.3.4
Diagnosis and prevalence of coeliac disease
CD is diagnosed after biopsy of the small intestinal mucosa, which shows typical lesions with villous atrophy, crypt hyperplasia and an increased number of inflammatory cells. Analysis of serum also shows typical antibodies of IgG and IgA type against gliadin, tissue transglutaminase and endomysium. The prevalence of CD in Western countries is calculated between 1 of 200 and 1 of 100 (M¨aki et al., 2003). Some of those diagnosed have only antibodies against gliadin, tissue transglutaminase or endomysium but no symptoms of damage in the intestinal mucosa – they are silent coeliacs (Ivarsson et al., 1999). CD is a life-long disease and the only remedy for those with the diagnosis is to base their food either on naturally gluten-free food like rice, maize, buckwheat, millet, etc., or on products containing wheat starch that has been rendered gluten free. Both types of products are at the moment sold under the label ‘gluten free’ (see proposed change regarding ‘gluten free’ by Codex).
160
Management of Food Allergens
10.3.5 Wheat allergy Wheat proteins are also involved in IgE-mediated allergy. Sensitisation to inhaled wheat flour leads to baker’s asthma, whereas sensitisation to ingested wheat via the gastrointestinal tract may cause food allergy and exercise-induced anaphylaxis. Several proteins responsible for eliciting allergy have been identified in the albumin/globulin fraction, the gliadin fraction, especially the ω-5 gliadin, as well as several low molecular weight glutenin subunits (Palosuo, 2003; Pastorello et al., 2007).
10.4 GLUTEN DETERMINATION 10.4.1
Polyclonal antibodies
About 20 years ago, when antibody-based methods was introduced for food analysis, polyclonal antibodies against gliadin were raised in rabbits and immunological methods were tested as a possible choice for the specific analysis of gluten in gluten-free food. By using specific antibodies to gliadin, other proteins such as those from milk, egg and soya were not detected and the immunological methods could thus be used for analysis of both baking mixtures as well as final food products. One application of the polyclonal antibodies was in dot blot techniques (Janssen et al., 1987). The lack of quantitative results was one disadvantage of the dot blot method. The low reactivity of the antibody with gliadin in highly processed food products also limited its use.
10.4.2
Monoclonal antibody to ω-gliadin
Monoclonal techniques allowed Australian researchers to produce antibodies to different wheat prolamin fractions. One of the antibodies reactive with ω-gliadin was used for the establishment of an enzyme immunoassay (Skerritt and Hill, 1990). ω-Gliadin has the advantage of being heat stable which enables the antibody to be equally reactive to heated as well as to non-heated products. The disadvantage with the Skerritt monoclonal was the low reactivity to barley compared to wheat and rye. The monoclonal did not react to oats. The ω-gliadin immunoassay was collaboratively studied and became an AOAC-approved method (Skerritt and Hill, 1991). Gluten could be measured from 0.02 to 10% by weight. The level 0.02% corresponds to 200 ppm gluten. However, lower levels became important to be able to trace according to the Codex standard (see below). Therefore providers of test kits, based on the Skerritt monoclonal, lowered the quantitation level to 20 ppm, and later to 10 ppm. Today, the lower limit of quantitation is 3 ppm gluten for one of these kits.
10.4.3
Monoclonal antibody to toxic peptide in secalin (R5)
Efforts of establishing a method that could detect barley equally well as the other toxic cereals wheat and rye resulted in a method based on a monoclonal antibody (R5) to the toxic peptide QQPFP from secalin, the rye equivalent to gliadin (Vald´es et al., 2003). No reactivity was shown to oats. The method was collaboratively studied by 20 laboratories analysing 12 different samples with 2 types of test kits (M´endez et al., 2006). The lowest level of gliadin in the samples tested was 13 ppm. The limit of quantitation of the test kit provided is now at 2.5 ppm gliadin, equivalent to 5 ppm gluten.
Standardisation of analytical methodology with special reference to gluten analysis
161
10.4.4 Extraction of gluten The traditional extraction procedure for gluten from food samples is by using ethanol in a concentration between 40 and 60%. Some food matrices like cocoa, chocolate and beer (hops) contain tannins, which are polyphenols with the capacity of binding proteins tightly to the matrix. By addition of milk proteins or fish gelatin, these binding effects can be overcome (Loomis, 1974). Another attempt to fully extract gluten (gliadins) from heated and highly processed samples such as cornflakes was made by changing the extraction solvent to a mixture (cocktail) consisting of 250 mM 2-mercaptoethnol, 2 M guanidin in phosphate buffer. The latter extraction has been introduced as the routine procedure in the R5 method (Garc´ıa et al., 2005). However, for all food types, this harsh treatment is not necessary. For malt analysis in cornflakes, it was shown to be a prerequisite. Denaturing and reducing agents like guanidin and mercaptoethanol interfere with the antigen–antibody interaction (Do˜na et al., 2008). Therefore, samples treated with such solvent have to be diluted, at least 500 times, before analysis.
10.4.5 Reference material for gliadin The measured gluten content of a sample determined by ELISA methods can vary significantly depending on the origin and type of protein(s) used for the calibration of the enzyme immunoassay, i.e. for the establishment of the standard curve. It varies also with the antibody used since the specificity of the Skerritt monoclonal is not the same as that of the R5 antibody (see Figure 10.4; Skerritt et al., 1989; van Eckert et al., 2008). To be able to compare results, there was an urgent need for a gliadin RM. Therefore, the Working Group on Prolamin Analysis and Toxicity (WGPAT/PWG) decided to organise a large-scale preparation of a reference gliadin material. The kernels of 28 common European wheat cultivars were selected for the production of the RM. The isolation of the gliadin from the defatted flour mix was performed according to Wieser et al. (1998). The characteristics of the PWG gliadin were described (van Eckert et al., 2006). During a short time period, the material was provided by IRMM as an RM of gliadin. It was later withdrawn due to impurities in the material, mainly from glutenins. However, at the moment, this calibrant is used by the Skerritt antibody-based test kits as well as the R5 antibody test kits.
10.4.6 Codex gluten standard In one of the Codex Alimentarius Commission general subject committees, the Codex Committee on Nutrition and Foods for Special Dietary Uses (CCNFSDU), a standard for gluten-free foods was elaborated in 1981 (Gluten standard 118-81). The analysis of residual gluten in wheat starch was performed by Kjeldahl nitrogen analysis. The analysis method was intended solely for purity testing of wheat starch and not applicable to foods containing other proteins like milk, egg, soya, etc. Revision of the standard began shortly after, with the intention to include methods for analysis of ready-to-eat products, i.e. final food products. The suggestion was an immunological method for the specific detection of gliadins. The creation of a standard is a stepwise procedure, ending at step 9. However, the standard was postponed for years due to the lack of appropriate methods of analysis. The first draft of the revised standard appeared in an ALINORM in 1996. The draft was advanced from step 3 to step 5. Distinct claims were
162
Management of Food Allergens 3
(a)
IRMM reference material Transia standard
2.5
OD 450 nm
2
1.5
1
0.5
0 0.01
0.1 Gliadin concentration µg/ml
1
1.6
(b)
IRMM reference material Ridascreen standard Transia standard
1.4
OD 450 nm
1.2 1 0.8 0.6 0.4 0.2 0 1
10 Gliadin concentration ppb
100
Fig. 10.4 (a) Comparison between the Transia kit gliadin standard (Timgalen wheat variety) and the IRMM reference material (28 European wheat varieties) using Transia Plate Gluten assay. The IRMM reference material and the standard were reconstituted in 40% ethanol. (b) Comparison between the Ridascreen kit standard, the IRMM reference material (28 European wheat varieties) and the Transia kit standard (Timgalen wheat variety) using Ridascreen Gliadin assay. The Ridascreen standard is provided with the kit, ready to use. The IRMM reference material and the Transia standard were reconstituted in 60 and 40% ethanol, respectively.
proposed for the terms ‘gluten-free’ and ‘low or reduced in gluten’ which would more accurately reflect the current definition with two levels. The levels 20 ppm (naturally gluten free) and 200 ppm gluten (products on wheat starch rendered gluten free), respectively, were proposed as limits. Approval was dependent on challenge studies and intake data being available. At the 28th session of the CCNFSDU in 2006, the threshold value of 200 ppm
Standardisation of analytical methodology with special reference to gluten analysis
163
gluten for products rendered gluten free was lowered to 100 ppm gluten. The finalization of the standard was considered to be possible due to the availability of recent data on acceptable daily intake levels (Collin et al., 2004; Gibert et al., 2006; Catassi et al., 2007) and methodology. The CCMAS at the same time adopted the enzyme-linked immunoassay R5 method as the preferred method (type 1) for the detection and quantification of gluten in food. This adoption was included in the method part of the gluten standard. The gluten standard is at step 8 of the Codex procedure and will be adopted by Codex Alimentarius Commission in 2008.
10.4.7 Analysis of gluten in fermented products Recently, a commercial competitive assay (Ferre et al., 2004) based on the R5 monoclonal antibody was launched. By using the competitive approach, hydrolysed products such as beer, soya sauces and glucose syrups could be detected. The result of the assay is expressed in peptide fragments in µg/ml. These values are then converted to ppm gliadin by dividing the value with 250. However, the degree of hydrolysis can be variable and the conversion factor 250 is therefore only an approximation. Therefore, no direct relation of the test results to the Codex standard limits for gluten-free food can be made, as stated by the provider of the test kit (R-Biopharm1 ).
10.4.8
DNA methods for the detection of cereals
As a complement to protein methods for gluten detection, DNA-based methods have been introduced. Initially, such methods were based on the end product detection in agarose gels, stained with ethidium bromide. The development of instruments for amplification and detection in real time allowed quantitative methods to be established (Dahinden et al., 2000). By choosing specific primers and probes for wheat, barley, rye and oat, respectively, identification of contamination in gluten-free products by species was possible (Sandberg et al., 2003). Quantitative or semi-quantitative assays were developed. However, the conversion factor between DNA and protein is still under debate. Most real-time PCR methods are at the moment less sensitive than protein assays. Inhibitors of the PCR reaction as well as problems and difficulties in extracting DNA of good quality still need to be resolved.
10.4.9 Future methodology CD is caused by inflammatory cell response to gluten peptides bound to HLA-DQ2 or DQ8 molecules. To screen food products for the presence of T-cell stimulatory epitopes involved in CD, a monoclonal test was developed that specifically recognize those peptides. Monoclonals were developed for the T-cell epitopes from α-gliadin, γ-gliadin and low and high molecular weight glutenins (Spaenij-Dekking et al., 2004). The test which detects both intact protein and small protein fragments is now undergoing validation. Recently, monoclonal antibodies to the main immunogenic wheat peptide (33 amino acids) from α-2 gliadin were obtained (Mor´on et al., 2008). They were reported to detect gliadin at low levels as well as toxic fractions present in wheat, rye and barley. However, the test also detects oats but to a low level. 1
Package insert from Ridascreen Gliadin Competitive immunoassay, Art No R7011, R-Biopharm.
164
Management of Food Allergens
REFERENCES Abbott M., Godefroy S.B., Yeung J.M. and Popping B. Guidelines for a harmonized validation protocol for ELISA methods for the determination of food allergens. To be published. Catassi C., Fabiani E., Iacono G. et al. (2007) A prospective, double-blind, placebo-controlled trial to establish a safe threshold for patients with celiac disease. American Journal of Clinical Nutrition, 85, 160–166. Collaborative study guidelines (1995) AOAC International guidelines for collaborative study procedures to validate characteristics of a method of analysis. Journal of AOAC International, 78(5), 143–159 Collin P., Thorell L., Kaukinen K. et al. (2004) The safe threshold for gluten contamination in gluten-free products. Can trace amounts be accepted in the treatment of coeliac disease? Alimentary Pharmacology and Therapeutics, 19, 1277–1283. Dahinden I., Stadler M. and L¨uthy J. (2000) Evaluation of real-time PCR to detect coeliac-toxic components and comparison to the ELISA method analysing 35 baby food samples. Mitteilungen aus Lebensmitteluntersuchung und Hygiene, 91, 723–732. Directive 2003/89 of the European Parliament and of the Council. Published in the Official Journal of the European Union, 25.11.2003, L308, 15–18. Do˜na V.V., Fossati A.C. and Chirod F.G. (2008) Interference of denaturing and reducing agents on the antigen/antibody interaction. Impact on the performance of quantitative immunoassays in gliadin analysis. European Food Research and Technology, 226 (3), 591–602. Ferre S., Garc´ıa E. and M´endez E. (2004) Measurement of hydrolysed gliadins by a competitive ELISA based on monoclonal antibody R5: analysis of syrups and beers. In: Proceedings of the 18th Meeting of the Working Group on Prolamin Analysis and Toxicity, October 2–5, 2003, (ed. M. Stern). Verlag Wissenschaftliche Scripten, Zwickau, Stockholm, Sweden, pp. 65–69. Garc´ıa E., Llorente M., Hernando A. et al. (2005) Development of a general procedure for complete extraction of gliadins for heat processed and unheated foods. European Journal of Gastroenterology and Hepatology, 17, 529–539. Gibert A., Espadaler M., Canela M.A. et al. (2006) Consumption of gluten-free products: should the threshold value for trace amounts of gluten be at 20, 100 or 200 ppm? European Journal of Gastroenterology and Hepatology, 18, 1187–1195. ˚ Juto P. et al. (1999) High prevalence of undiagnosed coeliac disease in adults: a Ivarsson A., Persson L.A., Swedish population based study. Journal of Internal Medicine, 245, 63–68. Janssen F.W., Voortman G. and de Baaij J.A. (1987) Rapid detection of soya protein, casein, whey protein, ovalbumin and wheat gluten in heat-treated meat products by means of dot blotting. Fleishwirtschaft, 65(5), 577–580. Loomis W.D. (1974) Overcoming problems of phenolic and quinones in the isolation of plant enzymes and organelles. Methods in enzymology, 31, 528–544. M¨aki M., Mustalahti K., Kokkonen J. et al. (2003) Prevalence of celiac disease among children in Finland. The New England Journal of Medicine, 348, 2517–2524. M´endez E., Vela C., Immer U. et al. (2006) Report of a collaborative trial to investigate the performance of the R5 enzyme linked immunoassay to determine gliadin in gluten-free food. European Journal of Gastroenterology and Hepatology, 17, 1053–1063. ´ Manyani H. et al. (2008) Sensitive detection of cereal fractions that are toxic to celiac Mor´on B., Cebolla A., disease patients by using monoclonal antibodies to a main immunogenic wheat peptide. American Journal of Clinical Nutrition, 87, 405–414. Osborne T.B. (1926) The Vegetable Proteins, 2nd edn. Longmans, Green and Co, London. Palosuo K. (2003) Update on wheat hypersensitivity. Current Opinion in Allergy and Clinical Immunology, 3, 205–209. Pastorello E.A., Farioli L., Conti A. et al. (2007) Wheat IgE-mediated food allergy in European patients: α-amylase inhibitors, lipid transfer proteins and low-molecular-weight glutenins. International Archives of Allergy and Immunology, 144, 10–22. Poms R.E. and Anklam E. (2004) Effects of chemical, physical, and technological processes on the nature of food allergens. Journal of AOAC International, 87(6), 1466–1474. Sandberg M., Lundberg L., Ferm M. et al. (2003) Real time PCR for the detection and discrimination of cereal contamination in gluten free foods. European Food Research and Technology, 217, 344–349.
Standardisation of analytical methodology with special reference to gluten analysis
165
Shan L., Molberg Ø., Parrot I. et al. (2002) Structural basis for gluten intolerance in celiac sprue. Science, 297, 2275–2279. Shewry P.R. and Halford N.G. (2002) Cereal seed storage proteins: structures, properties and role in grain utilization. Journal of Experimental Botany, 53, 947–958. Skerritt J.H. and Hill A.S. (1990) Monoclonal antibody sandwich enzyme immunoassays for determination of gluten in foods. Journal of Agricultural and Food Chemistry, 38(8), 1771–1778. Skerritt J.H. and Hill A.S. (1991) Enzyme immunoassay for determination of gluten in foods: collaborative study. Journal – Association of Official Analytical Chemists, 74(2), 257–264. Skerritt J.H., Jenkins K.L. and Hill A.S. (1989) Monoclonal antibody based two-site enzyme immunoassays for wheat gluten proteins. 2. Specificity analysis. Food and Agricultural Immunology, 1, 161–171. Sollid L.M. (2000) Molecular basis of celiac disease. Annual Review of Immunology, 18, 53–81. Spaenij-Dekking L., Kooy-Winkelaar Y., Nieuwenhuizen W.F. et al. (2004) A novel and sensitive method for the detection of T cell stimulatory epitopes of alpha/beta- and gamma-gliadin. Gut, 53(9), 1267–1273. Vald´es I., Garc´ıa E., Llorente M. et al. (2003) Innovative approach to low-level gluten determination in foods using a novel sandwich enzyme-linked immunsorbent assay protocol. European Journal of Gastroenterology and Hepatology, 15, 465–474. Van Eckert R., Berghofer E., Ciclitira P.J. et al. (2006) Towards a new gliadin reference material – isolation and characterisation. Journal of Cereal Science, 43, 331–341. Van Eckert R., Bond J., Rawson P. et al. (2008) Reaction of selected antibodies with gliadin from European wheat. Proceedings of the 22nd meeting of WGPAT, Barcelona, Spain, p. 23. Wieser H., Antes S. and Seilmeier W. (1998) Quantitative determination of gluten protein types in wheat flour by reversed-phase high-performance liquid chromatography. Cereal chemistry, 75, 644–650.
11
Analytical choices
Marie-Claude Robert
11.1 DEVELOPMENT OF ALLERGEN TESTING Food allergy testing began in the early 1960s with skin testing (Josephson and Glaser, 1963). Provocative sublingual testing and radioallergosorbent test appeared in the early 1970s (Breneman et al., 1974; Gillespie et al., 1976) followed by the first published double-blind study in the late 1970s for the confirmation of food allergy (Rapp, 1978). In parallel to food allergy testing for clinical purposes, methods dedicated to the detection of allergens in food were developed in the late 1960s. The first reports described gliadin detection by means of double diffusion against rabbit anti-gliadin antiserum on agar gel (Keyser and Mahler, 1973). It was then only two decades later, in 1985, that a method for the specific detection of an allergen was published, with the description of a radioimmunoassay for the quantification of wheat gliadin in bread (Ciclitira et al., 1985). Even though the publication referred exclusively to the gluten hypersensitivity coeliac disease, and not wheat allergy, this was the first description of an immunoassay for the detection of an allergen in food. In 1994, the first enzyme-linked immunosorbent assay (ELISA) using both monoclonal and polyclonal antibodies was developed (Hefle et al., 1994), and the use of ELISA kits became the method of choice for allergenic protein detection and quantification. ELISA development was rapidly followed by the creation of a new easy-to-use dipstick test format (Mills et al., 1997). In 2000, Holzhauser et al. developed and reported the first polymerase chain reaction (PCR) based methods. In the same year, the use of biosensors for allergen detection was adopted (Yeung et al., 2000). Finally, liquid chromatography mass spectrometry (LC-MS/MS) and LC-Q-TOF method developments were reported in 2004 and 2006, respectively (Shefcheck and Musser, 2004; Weber et al., 2006), further extending the choice of methodologies.
11.2 TEST FORMATS 11.2.1
Enzyme-linked immunosorbent assays
ELISA is a simple biochemical technique used to detect a specific molecule, such as a food allergen, in a sample. It is routinely used as a diagnostic tool and quality control check in various industries and as the analytical method of choice by official food control agencies for food allergen detection.
Analytical choices
167
11.2.1.1 Sandwich ELISA The commonly used sandwich solid-phase assay requires two antibodies that bind to different epitopes of the antigen. In the sandwich ELISA, antibodies detecting specifically the target analyte (‘capture’ antibodies) are immobilized on an inert surface such as polystyrene. The solution being assayed for the antigen is applied to the antibody-coated surface under conditions which are optimised for the antibody to bind the antigen. Unbound products are then removed with a washing step. A second antibody (the ‘detection’ antibody), linked covalently to an easily assayed enzyme, binds to the captured antigen, thus completing the ‘sandwich’. After washing away any unbound antibody-linked enzyme, the enzyme in the immobilized antibody–antigen–antibody–enzyme complex is assayed using a colorimetric substrate. The converted substrate colour indicates the amount of antigen present.
11.2.1.2 Competitive ELISA This ELISA uses a competitive approach to quantify an antigen. In an example of one such design, the antigen is first immobilised on the microwell plate. The free antigen in the test sample and a limited amount of antibody are incubated to form an antigen–antibody complex. The antibody competes to bind either to the antigen found in solution or to the coated antigen. The unbound products are then removed with a wash and a second enzyme-linked antibody is added to bind to the antigen–antibody complex attached to the microwell plate. The substrate is added, developing a colouration inversely proportional to the quantity of antigen in the test sample. The higher the original antigen concentration, the weaker the eventual signal.
11.2.2 Immunochromatographic lateral flow or strip tests A lateral flow test is a simple device used to detect the presence or absence of a target analyte in a sample matrix. The immunoassay is performed on dipsticks that are coated with antibodies specific for the target protein analyte. The extracted test samples solution is applied on the sample pad at one end of the strip test. The signal reagent is made of coloured particles labelled with antibodies raised to the target antigen. It is solubilised and binds to the antigen in the sample and flows through the membrane by capillary action. When specific antigens are present, the signal reagent binds to it, a second antibody in the nitrocellulose captures the complex and a coloured line develops. Antigens immobilized on the lateral flow device will catch the second antibodies and work as a positive control line.
11.2.3
Polymerase chain reaction
The PCR relies on the ability of Taq polymerase, an enzyme that catalyses DNA-copying, to remain stable at high temperatures. To copy DNA, polymerase requires three other components: template DNA, a supply of the four nucleotide bases and two primers. The first primer attaches to the 5 end of the gene and makes a new strand in this direction. The primer bound to the opposite strand progresses in a counter direction. The original DNA template is melted, the primers anneal, and the polymerase makes two new strands. The result is a dramatic amplification of the DNA that exists between the primers. These cycles are repeated 20–40 times, each cycle providing two new templates for the next cycle. The amount of amplification is therefore expressed as 2 raised to the power of ‘n’; where ‘n’ represents the number of cycles that are performed.
168
Management of Food Allergens
11.2.4
Biosensors
Biosensors can be small portable devices that may be used for on-site analyses. The biosensor technology makes it possible to measure specific molecular interactions in real time. By the immobilisation on a sensor onto a chip surface, whether it is an antibody recognising the allergenic protein or a specific DNA fragment encoding for an allergen recognising a complimentary strand, the binding interaction between the two can be measured quantitatively. Quantification occurs by measuring the changes in refractive index of a laser striking the chip surface.
11.2.5 Liquid-chromatography linked to mass spectrometry Protein identification and characterisation using LC-MS/MS has become a standard, because complex mixtures of proteins can be analysed and high levels of sensitivity attained. To identify proteins present in a sample, a trypsin digest of the protein extract must be prepared and the resulting peptides analysed by LC-MS/MS. The masses of the parent ions and their fragments are determined and used to search databases of known protein sequences in order to match to in silico digestion patterns of the proteins. Matches are made using search algorithms to test for goodness of fit and scored.
11.3 COMMERCIAL TEST KITS It was in the mid-90s that the first commercial, ready to use ELISA kits for food allergens were made available for the laboratories. These kits allow the quantification of residual allergenic proteins in a sample, by comparison with a standard protein or standard protein extract provided by the kit manufacturer. Following the successful commercialisation of the kits, and the growing demand, the assortment of ELISA test kits has been continuously increasing and is now almost covering the complete range of major food allergens. In parallel, other commercially available methods, such as immunochromatographic strip tests (lateral flow devices) and PCR-based kits have been developed to complement ELISAbased allergen detection kits. These methods generally offer a qualitative, ‘yes/no’ detection. Lateral flow devices for allergen testing are simple to use, provide an analytical result rapidly and without any investment in equipment or a laboratory. The variety of lateral flow test kits commercially available is basically expected to become as widespread as for ELISA kits. On the contrary to immunoassay-based methods, the PCR-based kits allow the indirect detection of allergenic protein in a sample through specific marker DNA amplification. The amplified DNA is then supposed to be correlating with the presence of the allergenic protein. As for other kit formats, the PCR-based methods develop rapidly and the range of allergens covered is more and more important.
11.3.1
Commercial immunological test kits
11.3.1.1 Commercial ELISA kits Commercial ELISA kits are sold as a unit consisting of all essential components necessary for detecting qualitatively or quantitatively allergenic proteins. The kit components include an antibody-coated microwell plate, wash and diluent buffers, three to five ready-to-use
Analytical choices
169
allergenic protein standard solutions or one protein standard to be sequentially diluted by the analyst, an antibody–enzyme conjugate, a chromogenic substrate, and an acidic reagent to be used as stop solution. An extraction solution concentrate is sometimes also included in the kit package. The preparation of the reagents, the procedure of analysis as well as the description of materials required for the analysis but not provided, are described extensively in a leaflet included as a kit insert. Commercial ELISA kits available at the time of writing this chapter are described in Table 11.1.
11.3.1.2 Commercial lateral flow devices The package contains usually 10–25 lateral flow devices, plus sample test tubes or reagent wells, sample diluent, extraction buffer and/or additive. The preparation of the reagents and the test procedure are described in a test kit insert provided by the kit manufacturer. Commercial lateral flow devices available at the time of writing this chapter are described in Table 11.1.
11.3.2
Commercial PCR-based test kits
The reagents needed to perform DNA sample extraction and the Taq polymerase are usually not included in PCR-based kits. All commercial PCR-based kits aimed at detecting food allergens and available at the time of writing are described in Table 11.2.
11.3.2.1 Real-time PCR kits One kit contains generally an allergen reaction mix, an inhibition control (spike), a positive control, Taq polymerase and a fluorescence detection enhancer, which are sufficient for approximately 100 target DNA specific reactions and 100 inhibition controls. The inhibition reactions help to minimise possible false-negative test results caused by inhibition of the PCR process. The qualitative analysis of sample DNAs can be performed with a detection limit of ∼10 DNA copies. The reagents needed to perform DNA sample extraction are usually not provided with the kits.
11.3.2.2 PCR kits for agarose gel analysis After amplification of allergen-specific DNA, PCR products are visualised by agarose gel electrophoresis. Kits include PCR mastermixes, requiring only the addition of template and Taq polymerase. The mastermixes are generally already coloured with a gel-loading dye so that a later additional procedural step is eliminated. Spike and control DNAs are also included to ensure the reliability of the tests.
11.3.2.3 PCR–ELISA kits PCR–ELISA test kits allow the detection of specific DNA targets by means of the PCR. Once the target DNA is amplified, it is subsequently detected via a labelled hybridisation probe in an ELISA-like technique. The kit contains a PCR binding buffer for both the rehydration of streptavidin and binding of the PCR products; a buffer for the denaturation of the PCR products, an initial wash buffer for both the removal of the denaturation buffer and unbound
Almond BLG Casein Egg Gluten Hazelnut Peanut Sesame Shellfish Walnut
Almond Casein Egg Gluten Hazelnut Peanut Shellfish
ELISA
Lateral flow device
Gluten Hazelnut Peanut
Almond BLG Casein Egg Gluten (2×) Gluten peptides Hazelnut Peanut (2×)
R-Biopharm
Peanut Milk
Almond Egg Gluten Hazelnut Milk Peanut Soya
Neogen Almond BLG Buckwheat Casein Crustacea Egg Gluten Hazelnut Mustard Peanut Sesame Soya
Elisa Systems BLG Casein Egg Gluten Peanut
Morinaga Egg Peanut
3M Tecra Lupin Gluten
Hallmark
Almond Casein Crustacea Hazelnut Peanut
Abkem Iberia
BLG Casein Egg
Generon
NOTE: Transia/BioControl and SafePath are mentioned in the text and in Table 11.3. Tepnel: www.tepnel.com; R-Biopharm: www.r-biopharm.com; Neogen: www.neogen.com; Elisa Systems: www.elisas.com.au; Morinaga: www.miobs.com; 3M Tecra: www.tecra.net; Hallmark: www.hallmarkav.com; Abkem Iberia: www.abkemiberia.com; Generon: www.generon.it.
Tepnel
Kit manufacturer and food allergen detected
Commercial immunological test kits for the detection of food allergens (at the time of writing, December 2008).
Kit format
Table 11.1
170 Management of Food Allergens
Peanut Soya Milk
Tepnel
Peanut Hazelnut Almond Soya
Almond Celery Crustacea Fish Gluten Hazelnut Lupin
Molluscs Mustard Peanut Sesame Soya Walnut
R-Biopharm
Almond Apricot Celery Corn Crustacea Gluten Hazelnut
Lupin Molluscs Peach Peanut Pistachios Sesame Soya
Incura
Almond Barley Brazil nut Cashew Celery Hazelnut Mustard Oats
Pecan nut Peanut Pine kernel Pistachios Sesame Soya Walnut Wheat
Generon
Kit manufacturers and food allergen detected
Commercial PCR-based test kits for the detection of food allergens (at the time of writing, December 2008).
Mustard Peanut Pistachio nut Sesame Shrimp Soya Walnut Wheat
The Food Safety Laboratories Limited
Almond Brazil nut Cashew nut Celery Crustacea Gluten Hazelnut Macadamias
Tepnel: www.tepnel.com; R-Biopharm: www.r-biopharm.com; Generon: www.generon.it; The Food Safety Laboratories Limited: www.foodsafetylabs.com.
PCR–ELISA
PCR-agarose gel
Real-time PCR
Kit format
Table 11.2
Analytical choices 171
172
Management of Food Allergens
PCR products; a buffer for the hybridisation of the DNA probes; a stringency buffer for the removal of unbound DNA probes and for stringency testing of the hybridisation; a conjugate buffer for the dilution of the concentrated antibody conjugate; a second wash buffer for the removal of unbound antibody conjugate; a substrate solution for the colorimetric reaction and finally a stop solution to arrest the reaction. The sample analysis is qualitative with a detection limit of 10–20 DNA copies. The reagents needed to perform DNA sample extraction and the Taq polymerase are usually not included in such kits.
11.4 ANALYTICAL ISSUES SPECIFIC TO IMMUNOASSAYS ELISA is probably the method that is used most commonly today by the food industry and official food control agencies; therefore, analytical issues related closely to this technology are addressed in this chapter. For completeness, Table 11.3 summarises all major characteristics of commercial ELISA kits available today so that the reader may compare the different kits easily. The commercialisation of ELISA-based allergen detection kits has brought a real benefit to laboratories. All materials and solutions needed for an analysis are provided together within one box and optimised for the application. But how extensive is this kit validation?
11.4.1
Validation data
The range of matrices for which the kit is suitable is often ill defined, and the laboratory analysts will have to determine on their own if the kit is able to detect the target allergen in the sample matrix of interest. Of course, the extent of the scope of validation is infinite and the kit manufacturers cannot be asked to have validation data available for all matrices. There should nevertheless be a minimal set of data at the disposal of the customer showing the kit performance in selected matrices. Equally, there are few manufacturers who provide timely support and offer to collaborate together on validation studies when the field of application of the kit has not yet been validated as wished by the customer. This service is extremely valuable and appreciated. If, despite all, no validation data is available, the analysts will have to determine for themselves the accuracy, precision, specificity, limit of detection, limit of quantification, linearity, range of measurement, ruggedness and robustness of the method for their sample matrices. Even when some validation data is available for the matrix of interest, analysts will still have to verify the method characteristics associated with precision, such as the estimation of recovery, repeatability and intermediate reproducibility.
11.4.2
Target of the immunoassay analysis
As a kit user, one generally expects the target of the kit to be well defined, either by the kit’s name or then at least in the kit insert. It still may happen that the analyst is using a commercial kit convinced that from reading the insert the kit will detect a specific targeted food allergen protein. But if the kit has been designed to detect another, unrelated, allergenic protein in the target food, the kit user may expect to find a false-negative result. This is misleading to the kit user and may potentially have dramatic consequences for allergic food consumers. Such
Quantitative competitive ELISA Quantitative competitive ELISA Quantitative sandwich ELISA Quantitative sandwich ELISA Qualitative sandwich ELISA
Beta-lactoglobulin
Beta-lactoglobulin
Beta-lactoglobulin
Beta-lactoglobulin
Qualitative sandwich ELISA
Almond proteins
Beta-lactoglobulin
Quantitative competitive ELISA
Almond proteins
Quantitative competitive ELISA
Quantitative sandwich ELISA
Almond proteins
Beta-lactoglobulin
Quantitative sandwich ELISA
Almond proteins
Beta-lactoglobulin (BLG)
Quantitative sandwich ELISA
Almond proteins
Almond
Format
Target
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Antibody
N/A
–
0.05
0.2
0.05
0.3
0.1
2
5.0
2.5
<2 0.2
1
6.25
2.5
2.5
1.0
LOQ
0.5
–
0.6
1.7
0.1
LOD
BLG
Whey proteins
BLG
BLG
BLG
BLG
Almond proteins
Almond proteins
Total almonds
Total almonds
Almond
Results reported as
SafePath
Morinaga
ELISA Systems
Generon
R-Biopharm
Tepnel
Elisa Systems
Abkem Iberia
Neogen
R-Biopharm
Tepnel
Supplier
(continued)
1 std + neg. std 0.05 ppm
7 stds + neg. std 0.3–20 ppm
4 stds + neg. std 0.1–1.0 ppm
5 stds + neg. std 0.2–50 ppm
5 stds + neg. std 5.0–405 ppm
5 stds + neg. std 2.5–40 ppm
3 stds + neg. std 1–5 ppm
7 stds + neg. std 6.25–400 ppm
4 stds + neg. std 2.5–25 ppm
4 stds + neg. std 2.5–20 ppm
5 stds + neg. std 1.0–20 ppm
Standards, range of measurement and inter-laboratory study
Commercially available ELISA test kits for the quantitative and qualitative detection of food allergens (at time of writing December 2008).
Allergenic food
Table 11.3
Analytical choices 173
Quantitative sandwich ELISA Quantitative competitive ELISA Quantitative competitive ELISA Quantitative sandwich ELISA Quantitative sandwich ELISA Quantitative competitive ELISA
Caseins (alpha, beta, kappa caseins)
Casein
Casein
Casein
Casein
Casein
Crustacea/Shellfish
Quantitative competitive ELISA
Casein
Casein
Quantitative sandwich ELISA Quantitative sandwich ELISA
Tropomyosin
Tropomyosin
Quantitative sandwich ELISA
Buckwheat proteins
Buckwheat
Format
Target
(Continued )
Allergenic food
Table 11.3
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Monoclonal
Polyclonal
Polyclonal
Antibody
0.4
0.6
1.0
2
1.0
0.1
1
0.025 0.05
–
–
0.5
1
0.1
0.5
1.6
<1 0.12
2.5
LOQ
1.25
LOD
Supplier
Shellfish proteins
Shrimp tropomyosin
Casein
Milk protein
Skim milk powder
Milk proteins
Casein
Casein
Casein
Tepnel
Elisa Systems
Abkem Iberia
Morinaga
Elisa Systems
Incura
Generon
R-Biopharm
Tepnel
Buckwheat flour Elisa Systems proteins
Results reported as
5 stds + neg. std 1.0–20 ppm
4 stds + neg. std 0.05–0.5 ppm
7 stds + neg. std 0.4–300 ppm
7 stds + neg. std 0.3–20 ppm
4 stds + neg. std 1–10 ppm
5 stds + neg. std 2–20 ppm
5 stds + neg. std 0.1–100 ppm
4 stds + neg. std 0.5–13.5 ppm
5 stds + neg. std 1.6–25 ppm
4 stds + neg. std 2.5–25 ppm
Standards, range of measurement and inter-laboratory study
174 Management of Food Allergens
Egg
Quantitative sandwich ELISA
Quantitative sandwich ELISA Quantitative competitive ELISA Quantitative competitive ELISA
Egg white proteins (ovalbumin, ovomucoid, ovotranferrin, lysozym)
Egg trypsin inhibitor protein
Ovalbumin
Egg white proteins
Egg white proteins Quantitative (ovomucoid and sandwich ELISA ovalbumin)
Quantitative sandwich ELISA
Quantitative competitive ELISA
Tropomyosin
Egg white protein (ovomucoid Gal d1)
Quantitative competitive ELISA
Tropomyosin (pen a1)
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
1.0
1.0
0.1
N/A
0.6
0.1
–
1
2.5
2.0
1.0
1.25
1.0
0.5
0.05
2
Whole dried egg
Egg white proteins
Ovalbumin
Egg proteins
Egg white proteins
Egg white proteins
Shrimp proteins
Crustacean proteins
Neogen
Incura
Generon
SafePath
R-Biopharm
Tepnel
Abkem Iberia
Incura
(continued)
4 stds + neg. std 2.5–25 ppm
4 stds + neg. std 2.0–25 ppm
6 stds + neg. std 0.1–100 ppm
1 std + neg. std 1.25 ppm
4 stds + neg. std 1.0–27 ppm
5 stds + neg. std 0.5–10 ppm
7 stds + neg. std 0.05–200 ppm
4 stds + neg. std 1.0–12.5 ppm
Analytical choices 175
Gluten R5monoclonal R5monoclonal
R5-likemonoclonal R5monoclonal
Quantitative sandwich ELISA
Gliadin Gliadin (prolamins of wheat, barley and rye)
Gliadin (prolamins Quantitative of wheat, barley sandwich ELISA and rye)
Gliadin (prolamins Quantitative of wheat, barley sandwich ELISA and rye
Polyclonal
Gliadin (prolamins Quantitative of wheat, barley sandwich ELISA and rye)
Quantitative sandwich ELISA
Egg proteins
Polyclonal
Skerrittmonoclonal
Quantitative sandwich ELISA
Ovalbumin
Polyclonal
Antibody
Quantitative sandwich ELISA
Quantitative sandwich ELISA
Egg white proteins (ovomucoid and ovalbumin)
Gliadins and glutenins
Format
Target
(Continued )
Allergenic food
Table 11.3
–
0.6
2
1.5
1.0
1.5
5
5
2.5
3
Gliadin
Gliadin
Gliadin
Gliadin
Gluten
Whole egg proteins
< 0.5 0.6
Egg white proteins Egg protein
1.0
LOQ
0.3
–
0.5
LOD
Results reported as
TransiaBioControl
Neogen
R-Biopharm FAST
R-Biopharm
Tepnel
3M Tecra
Morinaga
Elisa Systems
Supplier
5 stds + neg. std 1.5–25 ppm
4 stds + neg. std 5–50 ppm
4 stds + neg. std 5–40 ppm
5 stds + neg. std 2.5–40 ppm AOAC RI approved
5 stds + neg. std 3–50 ppm AOAC OM 991.10
5 stds + neg. std 0.6–15 ppm
7 stds + neg. std 0.3–20 ppm
3 stds + neg. std 1.0–5.0 ppm
Standards, range of measurement and inter-laboratory study
176 Management of Food Allergens
Quantitative sandwich ELISA Quantitative competitive ELISA
Hazelnut proteins
Hazelnut proteins Quantitative sandwich ELISA
Quantitative sandwich ELISA
Hazelnut proteins
Lupin proteins
Quantitative sandwich ELISA
Hazelnut proteins (14,18 and 42 kDa)
Lupin
Quantitative sandwich ELISA
Hazelnut proteins
Hazelnut
Quantitative sandwich ELISA
Omega-gliadin Quantitative competitive ELISA
Quantitative sandwich ELISA
Gliadin
Gliadin peptide (QQPFP)
Quantitative sandwich ELISA
Gliadin
Gluten peptides
Quantitative sandwich ELISA
Gliadin
polyclonal
Polyclonal
Polyclonal
Polyclonal
Monoclonal
Polyclonal
R5monoclonal
Skerrittmonoclonal
Polyclonal
Skerrittmonoclonal
Skerrittmonoclonal
0.78
0.5
2.5
2.5
1.6
<0.25 1.0
–
0.25
1.0
1.5
0.1
Gliadin
Wheat proteins
Gliadin
Gliadin
Hallmark
Morinaga
Elisa Systems
Transia BioControl
Lupin proteins
Hazelnut protein
Hazelnut protein
Total hazelnut
Total hazelnut
Hazelnut
Hallmark
Abkem Iberia
Elisa Systems
Neogen
R-Biopharm
Tepnel
1250 Gliadin peptide R-Biopharm
2.5
<1 922
0.3
2.5
5.0
–
1.25
–
(continued)
5 stds + neg. std 1.0–16 ppm
7 stds + neg. std 0.78–50 ppm
4 stds + neg. std 0.5–5.0 ppm
4 stds + neg. std 2.5–25 ppm
4 stds + neg. std 2.5–20 ppm
5 stds + neg. std 1.6–32 ppm
4 stds + neg. std 1250–10 000 ppm
5 stds + neg. std 2.5–50 ppm
7 stds + neg. std 0.3–20 ppm
5 stds + neg. std 2.5–50 ppm
6 stds + neg. std. 5.0–100 ppm
Analytical choices 177
Quantitative sandwich ELISA
Quantitative sandwich ELISA Quantitative sandwich ELISA Quantitative sandwich ELISA Quantitative sandwich ELISA Quantitative competitive ELISA Quantitative sandwich ELISA Quantitative sandwich ELISA
Mustard seed proteins
Peanut proteins (Ara h2, Ara h1 and other proteins)
Peanut proteins
Peanut protein (Ara h1)
Peanut proteins (incl. Ara h2, Ara h1)
Peanut proteins (incl. Ara h1)
Peanut proteins
Peanut proteins
Peanut proteins
Mustard
Peanut
Quantitative sandwich ELISA
Quantitative sandwich ELISA
Caseins and whey proteins
Milk
Format
Target
(Continued )
Allergenic food
Table 11.3
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Antibody
3.3
<2.5
0.5
–
1
2.5
0.3
2
2.5
1.0
<0.1
1.5
2.5
1.0
1.0
2.5
LOQ
1.0
0.5
0.5
1.0
LOD
Total peanut
Peanut protein
Peanut proteins
Peanut
Peanut
Peanut
Total peanut
Peanut proteins
Mustard seed protein
Skimmed milk powder
Results reported as
4 stds + neg. std 3.3–90 ppm
5 stds + neg. std 1–20 ppm AOAC RI approved
4 stds + neg. std 2.5–25 ppm AOAC RI approved
5 stds + neg. std 1–15 ppm
4 stds + neg. std 1–10 ppm
4 stds + neg. std 2.5–25 ppm
3M Tecra
Morinaga
Incura
4 stds + neg. std 2.5–20 ppm
7 stds + neg. std 0.3–20 ppm
6 stds + neg. std 2.0–25 ppm
R-Biopharm FAST 4 stds + neg. std 2.5–20 ppm AOAC RI approved
R-Biopharm
Tepnel
Neogen
Elisa Systems
Elisa Systems
Neogen
Supplier
Standards, range of measurement and inter-laboratory study
178 Management of Food Allergens
Walnut
Soya
Sesame
Quantitative sandwich ELISA
Qualitative sandwich ELISA
Soya trypsin inhibitor
Walnut proteins
Qualitative sandwich ELISA
Soya trypsin inhibitor
Polyclonal
Polyclonal
Polyclonal
Polyclonal
Quantitative sandwich ELISA
Polyclonal
Soya bean proteins (soya trypsin inhibitor + soya proteins)
Qualitative sandwich ELISA
Sesame seed proteins (2S-albumin protein)
Polyclonal
Polyclonal
Quantitative sandwich ELISA
Sesame seed proteins
Polyclonal
Polyclonal
Polyclonal
Soya flour proteins Quantitative sandwich ELISA
Quantitative sandwich ELISA
Quantitative sandwich ELISA
Peanut proteins (Ara h2)
Sesame seed proteins
Quantitative competitive ELISA
Peanut proteins (incl. Ara h1, Ara h2 and agglutinin)
0.25
N/A
0.5
1.25
1.5
0.5
2.4
50
1.0
2.5
2.5
1.0
1.0
6.0
<1 0.1
1.0
1.56
–
–
Tepnel
Transia
Abkem Iberia
Walnut
Soya proteins
Soya proteins
Soya proteins
Soya flour
Sesame seed protein
Tepnel
SafePath
Elisa Systems
Elisa Systems
Neogen
Elisa Systems
Sesame proteins Tepnel, NEW
Sesame
Peanut proteins
Peanut protein
5 stds + neg. std 2.4–120 ppm
1 std + neg. std 50 ppm
3 stds + neg. std 1.0–5.0 ppm
4 stds + neg. std 2.5–25 ppm
4 stds + neg. std 2.5–25 ppm
3 stds + neg. std 1–5 ppm
5 stds + neg. std 1–20 ppm
5 stds + neg. std 6–100 ppm
3 stds + neg. std 1–5 ppm
7 stds + neg. std 1.6–100 ppm
Analytical choices 179
180
Management of Food Allergens
situations must be avoided by kit manufacturers describing extensively the scope of use of their kits when the specific target of the detection antibodies is not disclosed.
11.4.3
Specificity of the method
Immunoassays are bioanalytical methods in which quantification of the target allergenic protein depends on the reaction of an antigen, the allergenic protein, with an antibody. Demonstrating the specificity of the immunoassay for the allergen of interest is critical because most immunoassays are not preceded by the specific extraction of the allergen from the matrix of interest. Studies aimed at defining the specificity of the antibodies used in the immunoassay must therefore be available to the kit user. Antibodies should be challenged with a minimal set of protein extracts from food commodities closely related to the targeted food allergen. Additionally, if cross-reactivity with other foods has been detected, this should be communicated to all kit users.
11.4.4
Effect of processing
Most of the foods sold and eaten today have undergone some form of preparation for reasons of safety, palatability or flavour. Food preparation often includes processes such as heating (e.g. cooking, pasteurisation), freezing, chopping, drying, fermenting or any other process that may lead to protein denaturation. Immunodominant epitopes of the allergenic proteins may be affected and their allergenic properties modified. Conformational epitopes are typically expected to be more sensitive to process-induced denaturation than linear epitopes. These latter ones can nevertheless be affected by chemical or enzymatic hydrolysis. Linear epitopes may also be chemically altered by deamination, deglycation or even through molecular engineering. All modifications that the allergenic protein undergoes during processing may affect substantially the performance of an immunoassay used for detection. This is why the influence of processing on the ability of the test kits to recognise processed and unprocessed allergenic proteins should be assessed, when relevant to do so.
11.4.5 Accuracy of the calibration line/curve Correct and precise results depend not only upon the laboratory analyst’s ability to conduct the allergen analysis and the effectiveness of the protein extraction buffer, but also from the type of calibration applied in the analytical test. All ELISA methods are based on a relative calibration where standards of known concentration and allergenic protein content are analysed by a spectrophotometer. The responses of samples analysed under the same conditions will be used to calculate corresponding concentrations by interpolation of the response read from the standard calibration curve. In general, linear calibration curves are more desirable because they result in the best accuracy and precision. The calibration curve does not only give an analytical result but it also gives the analyst an idea about the quality of the result. Provided one is operating within the range of linear response the plot should be a straight line, deviations from this straight line giving a good indication about the precision of the result. That is to say, if the data points are spread out, there is less certainty in the result.
Analytical choices
181
Some testing methods, however, do have non-linear calibration curves. When this is the case, more calibration standards are needed to achieve desirable precision. As goodness of fit is more difficult to assess for polynomial curves, non-linear standard curves may also require more reference standards to capture the region of curvature.
11.4.6
Reference material
Appropriate reference materials are missing for most of the allergens listed in the Annex IIIa of the Commission Directive 2007/68/EC. The first reference material made available is IRMM-481 which contains five peanut varieties that have undergone five different types of heat treatments. For a long time, IRMM has been trying to develop a new certified reference material for gluten. In 1999/2000, a reference gliadin was produced from 28 of the most frequent wheat varieties in Europe on behalf of the Working Group on Prolamin Analysis and Toxicity (WGPAT). This material was then taken under charge by the JRC-IRMM official reference material institute and was coded as IRMM-480. Unfortunately, as IRMM is not willing to certify this material further, it means that it is no longer available. As alternatives, NIST (USA-based National Institute of Standards and Technology) Standard Reference Materials (SRM) materials have sometimes been used as reference materials. However, these were not necessarily designed for protein detection purposes, but rather for the determination of compositional and trace constituent element analysis, as well as for selected fatty acids, calories and vitamins in food/agricultural commodities. Recently, however, in a joint effort, representatives from the Food Allergy Research and Resource Program (FARRP), Food Products Association (FPA), Health Canada, Institute for Reference Materials and Measurements (IRMM) and US Food and Drug Administration (FDA) Center for Food Safety and Applied Nutrition (CFSAN) have been evaluating NIST RM 8445, a spray-dried whole egg powder, and have accepted its use as a reference material for allergen detection. This reference material development will hopefully soon be followed by other initiatives in order to cover all major allergens.
11.4.7 Reporting units As long as there continues to be no consensus on the expression of reporting units, allergen detection kit results will remain extremely difficult to compare. This difficulty is enhanced by the absence of reference material that would allow the comparative evaluation and finally the normalization of different allergen detection kits commercially available.
11.4.8 Inter-laboratory studies – batch-to-batch variability When inter-laboratories studies are organized, 8–10 laboratories are generally enrolled. All laboratories assess the same method parameters, using two or more kits in parallel. AOAC International’s key parameters that need to be assessed are the calibration curve (with a minimum of 5 standards, excluding the blank), the behaviour of method with a variety of well-described matrices, the accuracy of the method (including recovery efficiency, precision (repeatability and intermediate reproducibility), a comparison to existing methods, the crossreactivity of the antibodies used for the detection of the allergen, the robustness of the method
182
Management of Food Allergens
(influence of temperature, incubation time, etc.), detection limit, limit of quantification and ruggedness. As inter-laboratory studies are extremely costly and time consuming, the majority of the newly available commercial kits have not undergone such validation. Nevertheless, as long as a proper validation report is provided by the kit manufacturer, all characteristics except the real reproducibility of the data should be available. Although one has also to keep in mind that inter-laboratory validation studies are possible only if it is also possible to keep the variability of the kit as low as possible. This explains why very few studies have been published regarding batch-to-batch or lot-to-lot variability of a single kit type, and why only few kit manufacturers have collected and can provide such data.
11.5 CONCLUSIONS ELISA-based methodologies remain to date the methods of choice for allergen detection. Not only is the knowledge regarding allergen detection by ELISA historically longer, but the technology is still the methodology most commonly used by food industry and official food control agencies. ELISA methods are based on the direct detection of the allergenic protein, and this represents a major advantage when compared to DNA-based detection techniques. ELISA methods are furthermore highly specific with only a few exceptional cross-reactions, and the results are expressed in units of food proteins or of the food residue, avoiding the difficult interpretation of a result expressed as DNA copies. While ELISA-based technologies are well implanted in the laboratories, a more rapid format and application of the technology, called the lateral flow devices, are now proposed to analysts. If the protein extraction procedure to apply in these tests remains the same as for ELISAs, then the time of analysis is much shorter and the format allows its use on the factory floor. These rapid tests are expected to become popular where they are used as a simple screening tool, for example, for verification of production line cleaning in a factory. Despite the widespread availability of analytical techniques for the detection of allergens in food commodities, unambiguous confirmatory methods such as LC-MS methods are also needed. Such method developments have already been reported (Shefcheck and Musser, 2004; Shefcheck et al., 2006; Weber et al., 2006) and further progress is eagerly awaited. Precise quantification at low levels of allergen contamination is the current challenge. LCMS-based methods, once set, will also open new possibilities and insight into allergens from processed foods, current quantification methods being affected by processing-induced protein denaturation or hydrolysis. It has nevertheless to be kept in mind that any method development will become useful for analysts only once it has been validated. This is why the creation of reference materials, to be used for method validation, must be supported at an international level.
REFERENCES Breneman J.C., Cook W.C., Deamer W. et al. (1974) Final report of the food allergy committee of the American college of allergists on the clinical evaluation of sublingual provocative testing method for diagnosis of food allergy. Annals of Allergy, 33(3), 164–166.
Analytical choices
183
Ciclitira P.J., Ellis H.J., Evans D.J. and Lennox E.S. (1985) A radioimmunoassay for wheat gliadin to assess the suitability of gluten free foods for patients with coeliac disease. Clinical and Experimental Immunology, 59(3), 703–708. Gillespie D.N., Nakajima S. and Gleich G.J. (1976) Detection of allergy to nuts by the radioallergosorbent test. Journal of Allergy and Clinical Immunology, 57(4), 302–309. Hefle S.L., Bush R.K., Yunginger J.W. and Chu F.S. (1994) A Sandwich enzyme-linked immunosorbent assay (ELISA) for the quantitation of selected peanut proteins in foods. Journal of Food Protection, 57(5), 419–423. Holzhauser T., Wangorsch A. and Vieths S. (2000) Polymerase chain reaction (PCR) for detection of potentially allergenic hazelnut residues in complex food matrixes. European Food Research and Technology, 211(5), 360–365. Josephson B.M. and Glaser J. (1963) A comparison of skin-testing with natural foods and commercial extracts. Annals of Allergy, 21, 33–40. Keyser J.W. and Mahler R.F. (1973) Detection of gluten in flour. Lancet, 2(7830), 678. Mills C.E.N., Potts A., Plumb G.W. et al. (1997) Development of a rapid dipstick immunoassay for the detection of peanut contamination of food. Food and Agricultural Immunology, 9(1), 37–50. Rapp D.J. (1978) Double-blind confirmation and treatment of milk sensitivity. Medical Journal of Australia, 1(10), 571–572. Shefcheck K.J., Callahan J.H. and Musser S.M. (2006) Confirmation of peanut protein using peptide markers in dark chocolate using liquid chromatography-tandem mass spectrometry (LC-MS/MS). Journal of Agricultural and Food Chemistry, 54(21), 7953–7959. Shefcheck K.J. and Musser S.M. (2004) Confirmation of the allergenic peanut protein, Ara h 1, in a model food matrix using liquid chromatography/tandem mass spectrometry (LC/MS/MS). Journal of Agricultural and Food Chemistry, 52(10), 2785–2790. Weber D., Raymond P., Ben Rejeb S. and Lau B. (2006) Development of a liquid chromatographytandem mass spectrometry method using capillary liquid chromatography and nanoelectrospray ionizationquadrupole time-of-flight hybrid mass spectrometer for the detection of milk allergens. Journal of Agricultural and Food Chemistry, 54(5), 1604–1610. Yeung J.M., Lai E.P. and Mullett W. (2000) Is biosensor a viable method for food allergen detection? (Abstract). Journal of American Chemical Society, 218, U80.
12
Food allergen method development programme at Health Canada: support to standard setting and consumer protection
Samuel Benrejeb Godefroy, Michael Abbott, Terry Koerner, Dorcas Weber, and Theresa Paolisini
12.1 RATIONALE TO ACT ON PREVENTING FOOD ALLERGY INCIDENTS IN CANADA In Canada, the foods most frequently associated with allergic adverse reactions are peanuts, tree nuts (almonds, Brazil nuts, cashews, hazelnuts, macadamia nuts, pecans, pine nuts, pistachios and walnuts), milk, eggs, fish, crustaceans, shellfish, sesame seeds, soya beans, wheat, kamut, spelt and triticale. These foods will be referred to as priority allergens. Current estimates suggest that food allergies affect as many as 6% of children and 3–4% of adults. The percentage of adults affected by food allergies is lower than the percentage of children affected since some children can outgrow their food allergy. Unfortunately, many allergies do persist into adulthood and can be quite serious in nature. An analysis of data from the Canadian Community Health Survey (CCHS) of 2002/03 and the National Population Health Survey (NPHS) of 1994/05 found that food allergies had increased across all age groups in Canada in the 8 years between the surveys (Canadian Council of Social Development, 2006). This observed trend is being supported by a more recent and comprehensive research project recently conducted in Montreal. In this two-part study, the prevalence of peanut allergy in pupils in kindergarten, up to Grade 3, was determined to be 1.3% in 2002. In the second phase of the study, conducted in 2006, preliminary results indicate that the prevalence has increased to 1.8%. It should be noted that this study includes only confirmed cases of peanut allergy (Kagan et al., 2003; Clark et al., 2008). An individual with a food allergy who comes in contact with an allergen, included in the protein portion of a food, may have a reaction that develops quickly and may rapidly progress from mild to severe, including anaphylactic shock and, in extreme cases, death. It is estimated that 1–2% of Canadians live with the risk of anaphylactic reaction [Anaphylaxis Canada (http://www.anaphylaxis.org/#)]. In the USA, food allergy remains a leading cause of anaphylaxis treated in emergency rooms and it is estimated that there are 150 deaths from food-related anaphylaxis per year (Sampson, 2004). Comprehensive data on deaths attributed to food-induced anaphylaxis is not available in Canada. However, a single study examined deaths related to anaphylaxis in the Province of Ontario (Canada) over the period of 1986–2000. In this study, 63 confirmed deaths due to anaphylaxis were identified with 32 were related to adverse reactions to foods (2.3 deaths per year) (Salter et al., 2000).
Food allergen method development programme at Health Canada
185
Coeliac disease (CD) is an autoimmune disorder resulting from exposure of susceptible individuals to dietary gluten found in cereals, such as wheat, rye and barley. Symptoms result from the inflammation of the upper bowel, leading to malabsorption of nutrients, fatigue, anaemia, iron deficiency and osteoporosis as well as the more classical symptoms of diarrhoea and weight loss. It is estimated that up to 1 in 133 persons in Canada is affected by coeliac disease [Canadian Celiac Association (http://www.celiac.ca)]. Canadian food regulatory authorities have clearly identified food allergies and coeliac disease as important public health issues, linked to food consumption. Although some novel food allergy therapies are being explored, for those individuals known to be food allergic or having CD, avoidance of the culprit food allergenic ingredient or of the various sources of gluten is currently the only viable risk management option. Improvement of food-manufacturing practices, prevention of cross-contamination incidents during food processing and handling and clear identification of priority allergens and sources of gluten on food labels are among the initiatives that are being pursued by food industry, regulators and consumer groups, to mitigate food allergy incidents. The Bureau of Chemical Safety in Health Canada’s Food Directorate is supported by other federal agencies such as the Canadian Food Inspection Agency and Agriculture and Agri-Food Canada in developing and implementing programmes aimed at enhancing the protection of Canadian consumers with food allergies and CD. Risk management decisions and regulatory amendments were developed to enhance the labelling of pre-packaged foods by requiring the mandatory declaration of the priority food allergens, gluten sources and added sulphites when present in food. These labelling requirements would enable consumers with food allergies and CD to avoid foods that contain those substances, which may trigger an adverse reaction. Actions in this area have been framed within Health Canada’s Food Allergy Incident Prevention Strategy, which aims to minimise risks associated with inadvertent consumption of undeclared allergens, gluten sources and sulphites in food and to maximise choices of safe and nutritious foods for those with dietary restrictions related to food allergies and CD.
12.2 HEALTH CANADA’S FOOD ALLERGEN METHODOLOGY PROGRAMME The availability of analytical methods to detect and determine levels of markers of priority allergens in foods are of the utmost importance to support standard setting initiatives, the development of compliance and enforcement activities, as well as to provide guidance to industry on implementation of quality control practices, ensuring the effectiveness of allergen-related sanitation techniques. In the early 1990s, Health Canada’s Food Chemical Safety Program launched an ambitious allergen method development programme. Methods were developed for the detection of peanut (Yeung and Collins, 1996; Newsome and Abbott, 1999), soya (Yeung and Collins, 1997), egg (Yeung and Abbott, 2000), hazelnut (Ben Rejeb et al., 2003) and crustacean tropomyosin (Ben Rejeb et al., 2002). When the Canadian Food Inspection Agency (CFIA) was created in 1997, these methods were transferred to the agency laboratories and made the backbone of the laboratory capacity in support of the Canadian compliance programme with regards to undeclared allergens in processed foods. Since that time, a number of commercially available allergen test kits have emerged, enabling the determination of markers of priority allergens in a variety of foods. These kits are based
186
Management of Food Allergens
on either the immunochemical or PCR approaches, and have become more widely available for an increasing number of different allergens. While maintaining its capacity for method development using reagents generated in-house, Health Canada’s food allergen methodology programme has been realigned to adapt to a quickly evolving environment, with a variety of techniques becoming available to determine and measure allergen markers in food. The programme has recently focused an increasing amount of attention on the evaluation and validation of commercially available ELISA-based (enzyme-linked immunosorbent assay) test kits, which are documented in Health Canada’s compendium of food allergen methodologies (Health Canada, 2008). The programme also aims to further collaborate with other international agencies in attempts to ensure a common approach in using analytical results as part of health risk assessments, corresponding to the investigation of food allergy incidents. Finally, the programme is also focusing efforts on the development of alternatives to current bioanalytical methods (ELISA and PCR), in particular, using a combination of protein separation techniques coupled with mass spectrometry for the identification and confirmation of key markers for proteins and/or peptides.
12.2.1
Health Canada’s food allergen method validation activities
Health Canada’s food allergen method validation programme has targeted the development of an evaluation process that meets the requirements of a variety of methodology users. This process would not set specific parameters of performance for the method to be evaluated, but would rather focus on reporting on the performance of the method within its intended scope and the claims of its developer. Nonetheless, this process has aimed at identifying key parameters to be included for such an evaluation. One or more reference materials will be required for each priority allergen. An allergen reference material was defined as a ‘material that is representative of the allergenic food commodity, that is well characterised and that can be produced or supplied with reproducible characteristics’. This material is to be utilised to estimate the method recovery for the food matrix of interest, as part of fortification experiments (naturally incurred or artificial fortification). The method is to be documented using information provided by the test kit manufacturer or the method developer. The characteristics of key reagents such as the capture antibody, the conjugated antibody or test calibrators are to be explicated. In particular, test calibrators or standard solutions will be qualified for their protein content. This step was identified as essential to ensure the possibility of comparing results generated with different test kits for risk assessment purposes. A similar approach was adopted to qualify the designated reference material in protein content and to allow possible correlation between quantification levels expressed through the use of protein calibrators provided with the assay and the reference material chosen for recovery experiments. Despite the variations that may be encountered with protein determination methods, Health Canada’s allergen method development programme has consistently used the bicinchoninic acid assay (BCA test) to estimate protein concentrations. Although subject to discussion, and despite the variation in nature and behaviour of soluble allergen proteins in comparison to bovine serum albumin (BSA), it was agreed to use BSA as a common reference calibrator for the BCA test for reporting purposes. These reference parameters are reported along with the assay results, if used as part of an investigation. The need to ensure that assay results are further expressed in protein concentrations is dictated by Health Canada’s risk assessment procedures, which require that analytical results
Food allergen method development programme at Health Canada
187
for the occurrence of an undeclared allergen are expressed in micrograms of soluble protein per gram of food commodity being investigated. The Health Canada allergen method validation protocol has also set specific parameters to be followed for recovery experiments, conducted as part of inter-laboratory evaluations. For each food matrix included in the evaluation, three levels of fortification are to be prepared and analysed: blank samples, samples fortified at twice the limit of quantification (2 × LOQ) and samples fortified at five times the limit of quantification (5 × LOQ). For each food matrix and each fortification level, the validation protocol has also set the number of replicates to be analysed at 10. A minimum of three independent laboratories is to be involved in generating recovery data. To date, this validation protocol has been applied to two Health Canada in-house developed methods – hazelnut (Ben Rejeb et al., 2004) and Brazil nut ELISAs – and to five commercially available test kits. The results of the latter evaluations were made available on Health Canada’s web-enabled compendium of allergen methodologies (Health Canada, 2008).
12.2.2 Application of Health Canada’s allergen method validation protocol to a milk assay In one example of its use, the Health Canada allergen method validation protocol was applied to generate performance data on one of the commercially available assays targeting a protein marker of milk, beta-lactoglobulin (βLG), one of the major allergens in cow’s milk. This test kit was produced by ELISA Systems, Australia. A representative whole milk powder was chosen to be the designated reference material, according to the previously adopted definition. This material was obtained from the National Institute of Standards and Technology (NIST), USA, and was developed jointly by Agriculture and Agri-Food Canada (AAFC) and NIST. The material is known as the NIST material # 8435. The inter-laboratory evaluation involved the analysis of selected food matrices which had been artificially fortified (spiked samples) with the whole milk powder. A total of five laboratories (four Canadian and one European) were part of the validation exercise. The ELISA Systems Beta-Lactoglobulin kit uses a different standard than do the NIST whole milk powder for its calibration curve. For this reason, it was expected that there would be a difference in the level of response from the kit to the NIST whole milk powder used for spiking purposes and the skim milk powder calibrators supplied with the ELISA Systems test kit. A preliminary investigation showed that 32 and 80 ppm of NIST whole milk powder were required to obtain a response equivalent to 5.0 and 12.5 ppm, respectively. These response levels were targeted because they represent two and five times the LOQ of the kit, which has been set at 2.5 ppm. On the basis of these results, samples for the study were spiked with 0, 32 ppm and 80 ppm of the NIST 8435 standard. Spiking was performed using a suspension of the NIST 8435 whole milk powder in phosphate buffered saline (PBS) solution. Samples were left blank or spiked at one of two levels and given a blind identification code. Samples were spiked using a 440 µg/ml working solution of NIST 8435 whole milk powder. The 440 µg/ml working solution of NIST 8435 was prepared by serial dilution. A 1 g aliquot of the whole milk powder was dissolved in 10 ml of PBS (100 mg/ml) and then diluted (1:10) with additional PBS to give a 10 mg/ml intermediate. Next, 11 ml of 10 mg/ml was diluted to 250 ml to give the final working solution at 440 µg/ml. In order to keep the spiking volumes approximately equal, the blank samples were spiked with 1.0 ml of PBS, the 32 ppm samples with 0.364 ml of 440 µg/ml of working solution and 0.5 ml PBS, while the 80 ppm samples were prepared using 0.909 ml of the 440 µg/ml of working solution.
188
Management of Food Allergens
Three different matrices were included in the evaluation: baby cereal, cookies and dark chocolate. These commodities were included as representatives of some of the matrices which were identified as most likely to contain undeclared milk proteins. The potential for cross-contamination with milk products at low levels exists when commodities that do not contain milk or components of milk as an ingredient are manufactured on the same equipment as commodities that do contain milk or milk components. To supply 5 independent laboratories, 450 samples were required based on 10 replicates at each of 3 spiking levels in each of the 3 commodities. Five gram samples were weighed out into 250 ml screw cap bottles (150 samples for each commodity). The samples were separated into groups of 50 and each group was spiked at one of the 3 spiking levels. Each sample was given a blind code number. The samples were subsequently grouped together for each of the 5 participating laboratories and shipped by courier. Each participating laboratory was provided with the required number of BetaLactoglobulin test kits, as well as the blind-coded samples (30 of each matrix – 90 samples in total per laboratory). Each laboratory extracted the samples following the extraction procedure outlined in the kit instructions. The extraction was performed directly in the sample bottles provided. The sample extracts were then analysed following the instructions provided in the kit insert. As reported in Table 12.1, the results for the 450 samples analysed in the study showed good inter- and intra-laboratory consistency. No false positives were reported. Six false negatives were reported out of the 150 lower level spiked samples. All six of these were reported in the dark chocolate samples. It should be noted that, although below the 2.5 ppm LOQ listed in the kit insert instructions, all these samples gave results close to 2.5 ppm and had optical densities considerably higher than the blank sample. All the higher-level spiked samples gave positive results. For each group of 10 samples at one of the three spiking levels in a particular matrix analysed at the same laboratory (for example the 32 ppm dark chocolate samples from laboratory #2), mean and standard deviation were calculated. Any results more than two
Table 12.1 Summary of results of the inter-laboratory evaluation of the beta-lactoglobulin ELISA (average results for the five participating laboratories). Spiking levela
Expected resultb
Average
Standard deviation
−0.8–0.3 3.8–6.8 6.7–16.1
−0.2 4.9 10.4
0.3 0.7 1.1
−0.7–0.7 2.8–5.6 7.8–10.8
−0.1 4.2 9.0
0.4 0.3 0.4
−0.2 3.0 7.0
0.2 0.1 0.3
Range of results
Baby cereal results for all five laboratories Blanks 32 ppm 80 ppm
5 12.5
Cookie results for all five laboratories Blanks 32 ppm 5 80 ppm 12.5
Dark chocolate results for all five laboratories Blanks −0.9–0.2 32 ppm 5 2.0–4.3 80 ppm 12.5 5.7–8.3 a b
based on designated reference material. based on the kit calibrators.
Food allergen method development programme at Health Canada
189
standard deviations from the mean were considered an outlier and were excluded. Out of 450 samples only 8 samples were excluded for this reason. The z-score for an item indicates how far, and in what direction, that item deviates from its distribution’s mean, expressed in units of its distribution’s standard deviation. A z-score of 2.0 indicates a result that was two standard deviations above the mean, while a score of −2.0 would indicate a result two standard deviations below the mean. z-Scores were calculated for results generated by each laboratory, for each commodity and spiking level. Only three results were above 2.0 and the majority were below 1.0, which shows a good agreement between all the participating laboratories. The highest z-scores were obtained for the blank samples, a result which is expected since small variations in optical densities are proportionally larger for the blanks than for the other samples. The conclusion for this study was that the ELISA Systems Beta-Lactoglobulin kit, evaluated according to Health Canada’s method validation protocol, delivered satisfactory results for the matrices and at the levels tested. Dark chocolate had the lowest recoveries and cereal the highest recoveries. Matrix effect is generally expected for dark chocolate, given the high levels of tannins and other phenolic compounds which can interfere with the extraction of proteins.
12.2.3 Towards a harmonised validation protocol for allergen ELISA-based test kits Availability of validated methods is critical for both method developers and end users. For method developers, validation of an analytical procedure is used to demonstrate that it is suitable for its intended purpose. For end users, validated methods ensure reliability, repeatability, accuracy and precision of the results generated using a particular method. In the absence of a universally recognised reference standard for food allergen ELISAs, many organisations and end users have relied on a variety of method validation protocols and different analytical standards. This kind of inconsistency and duplication will inevitably have a negative economic impact on the allergen analytical community. Health Canada’s food allergen programme has supported various actions leading towards the development of a consensus protocol that could be followed by method developers and end users, to validate the growing number of ELISA-based allergen detection methodologies. These efforts were supported by the activities of the AOAC presidential taskforce on food allergens and the Network of Excellence (NoE) MONIQA (Monitoring and Quality Assurance in the Food Supply Chain), a European-funded initiative dedicated to supporting harmonisation of methods for food analysis. Specific efforts are underway to build on current knowledge and practices of method validation and to design a companion document to the ‘AOAC Official Methods of Analysis Appendix D: Guidelines for Collaborative Study Procedures to Validate Characteristics of a Method of Analysis’ (AOAC, 2002), providing specific guidance for the validation of quantitative ELISA-based methods for food allergens. The protocol is designed to meet or exceed the minimum requirements set forth in Appendix D and is currently being developed with input from a wide range of experts in the area of food allergens. Upon completion, this document will provide a global method validation study protocol to validate the performance characteristics of quantitative food allergen ELISA methods. This practical protocol is also intended to help method developers in designing a validation study to generate appropriate data for possible submission to bodies such as the AOAC International or other regulatory
190
Management of Food Allergens
bodies for recognition. Both the study design and data would be subject to scrutiny prior to acceptance. Much of the guidance being developed will apply to any priority food allergen, as defined by the Codex Alimentarius Committee of food labelling in 1998 (FAO/WHO, 1998). However, due to the nature of food allergens, certain aspects such as choice of reference materials and spiking methods would need to be addressed on a case-by-case basis. Two priority allergens (egg and milk) have been selected for inclusion in the initial protocol. Further guidance for other priority allergens will follow. It is expected that once completed, Health Canada will endorse this protocol in replacement of its current process to design and implement interlaboratory studies for food allergen method validation. It is also expected that Health Canada will use the consensus protocol as the guideline to be followed by method developers, to submit data to Health Canada, in support of test kit evaluation and subsequent endorsement. Health Canada will amend its current process of food allergen method evaluation accordingly and will continue to strive at disseminating data related to performance of food allergen ELISA-based analytical methods on its website (www.healthcanada.gc.ca/foodallergies)
12.2.4
Mass-spectrometry-based analytical methods
In addition to the more traditional methods for allergen detection (ELISA), Health Canada’s allergen method development programme is focusing efforts towards the development of confirmatory procedures using protein separation techniques such as liquid chromatography (LC) and immuno-affinity (IA) coupled with mass spectrometry (MS). MS has the ability to unambiguously identify proteins or peptides in solution and would compliment the quantitative results obtained from established ELISA methods. In addition, MS has the ability to identify proteins in a sample where no other method of analysis exists and theoretically can detect multiple food allergens in a single experiment. The most established method of protein analysis using MS is termed the ‘bottom-up’ approach where the proteins are digested, typically with trypsin, and the fragments are then analysed by MS. Protein identification can be accomplished by searching a database with a list of masses from the tryptic fragments (peptide mass fingerprinting) or each tryptic fragment can be fragmented further in the collision cell of the mass spectrometer to obtain a list of fragments for each peptide. This list of fragments can then be used to search a database to obtain the amino acid sequence for each peptide fragment (MS/MS analysis). The latter method is becoming the more popular because of the rich amount of data that can be obtained in one experiment; moreover, confident protein identification can typically be accomplished with just one or two peptides. Health Canada is currently using the ‘bottom-up’ approach for protein analysis to provide conformational support to established ELISA methodologies. Preliminary studies were performed using defatted roasted peanut extracts separated using 1D SDS-PAGE electrophoresis (Ben Rejeb et al., 2004). Protein bands were excised, digested using trypsin and then analysed by capillary LC-MS/MS. The system was configured with a short C18 pre-column to trap and desalt protein digests followed by an analytical column (PepMap C18). All MS data were acquired using Micromass Masslynx v3.5 under the ‘survey’ mode and analysed either via the Proteinlynx Global Server v1.1 with databases including SWISSPROT, TrEMBL, NR (compact version of NCBI) or via submitting a peak list to Matrix Sciences’ Mascot MS/MS search engine using the NCBI database. The major peanut allergen proteins along with peanut purified allergenic proteins were successfully identified using in-gel digestion of the selected bands. The high resolution and mass accuracy of the MS instrument created data
Food allergen method development programme at Health Canada
191
sets that produced high scores in the search results giving confidence in the match and protein identification. In fact, it sometimes required only a single peptide to confidently identify a protein. These results were very promising; however, it was identified that using electrophoresis for the separation of a complex food matrix would be difficult and time consuming. Therefore, in solution digestion of the proteins followed by LC-MS/MS analysis was applied (Weber et al., 2006). In this study, the emphasis was on casein, which is abundant and thought to be the most allergenic of the milk proteins. In general, the analytical methodology was similar to the previous study with the exception of the preliminary sample purification protocol, which involved ion exchange chromatography followed by a desalting step using molecular weight cut-off filters. In a typical experiment, cookies spiked with milk powder (0–1250 ppm) were extracted, purified, by the above method, and then digested in solution using trypsin. The peptide fragment mixture was then analysed using capillary LC-MS/MS and the resulting data set used to search conventional FASTA-formatted protein databanks (SwissProt or NCBI). Given in Figure 12.1 is a typical result that would be obtained from the database search. First, the output provides a list of proteins that were found in the sample, which comprises proteins that would be expected in a cookie that was spiked with milk powder. The scoring for these proteins is also provided in a bar graph with the most significant hits having the highest score. Finally, the expanded region of the second protein on the list (casein alpha-S1) shows the peptides that were found in the data set that matched for this particular protein. In this case, four peptides were found, which would constitute a significant result. Table 12.2 shows the results extracted for the alpha-casein protein. As expected, the best results come from the highest spike level (1250 ppm) with a Mascot score of 721 with 43% protein coverage and as the spike level is reduced the scores become progressively lower. However, a confident protein confirmation (Mascot score 100) can be obtained for even the lowest spike level (1.25 ppm). These types of analysis demonstrate the power of using LC-MS/MS as a confirmatory method for allergen detection in food matrices. These types of analyses have been applied to cereal grains with promising results and have enabled us to differentiate the sources of gluten in samples of beer, which is something conventional methodologies like ELISA cannot do at this time. Future developments in this area are underway. Methodologies which will exploit the full quantitative potential of the mass spectrometer are being explored. These methods will allow to obtain a second option of allergen marker quantification in food matrices, supporting risk assessment and regulatory initiatives. These techniques will also help respond to areas of concern where there may not be a suitable ELISA methods available for undeclared allergen proteins.
12.3 CONCLUSION The Canadian regulatory environment surrounding the declaration of allergens in food is in constant evolution. Risk assessment methods to investigate food allergy incidents are being updated using newer findings on threshold information as well as the application of chemical health risk assessment techniques used more generally for chemicals in food (deterministic and probabilistic approaches). The reliance on validated allergen analytical methods is the key for stakeholders such as the food industry and regulatory agencies to support their risk assessment and risk management decisions. Enhanced international collaboration is needed to allow the food allergen analytical community to move towards a more harmonised environment in generating data supporting the validation of these methods, hence avoiding
192
Management of Food Allergens
AT R I X { { MSCIENCE
Mascot Search Results
user Email Search title MS data file database Timestamp Warning Significant hits
: : : : : : : : :
dorcas veber
[email protected] C:/Data PKL 21\2004092812.pkl HCBInr 20040923 (2043161 sequences; 685168504 residues) 29 Sep 2004 at 15:43:29 GMT Too many peptide masses in your data file. Mascot has a limit of 1000000 but this system has been configured to have a limit of 308 globulin Beg1 precursor – barley gi I 421987 casein alphaS1 gi I 225632 gi I 34495244 globulin-like protein [Oryza sativa (japonica cultivar–group)] grain softness related protein [Triticum manocococum] gi I 6689417 high molecular weight glutenin subunit 1Ax1 [Triticum aestivum] gi I 21743 granule-bound starch synthase WX-TsD protein [Aegilops speltoides] gi I 6510540 kappa-casein precursor gi I 162811 01cosin Im-II (01cosin 18 kDa) (Lipid body-associated protein L2) gi I 129081 gi I 46098661 hypothetical protein UM03103.1 [Ustilago maydis 521] gi I 21779918 seed globulin [Aegilops tauschii] gi I 20563344 granule-bound starch synthase [Docyniopsis tschonoskii] gi I 32100760 unknown [Triticum aestivum] CHx [Triticum aestivum] gi I 640015
Probability Based Mowse Score
Number of Hits
Ions score is -10*Log(P), where P is the probability that the observed match is a random event Individual ions scores > 52 indicate identity or estensive homology (p<0.05) Protein scores are derived from ions scores as a non-probabilstic basis for ranking protein hits.
30 25 20 15 10 5 0 300 200 Probability Based Mowse Score
100
2.
gi I 225632 casein alphaS1
Mass: 24477
Score : 164
Queries matched: 4
Check to include this hit in error tolerant search Qurey 79 103 164 236
Observed 634.35 692.87 587.32 1158.58
Mr (expt) 1266.69 1383.72 1758.93 2315.14
Mr (calo) 1266.70 1383.72 1758.94 2315.13
Proteins matching the same set of peptides: gi I 30794348 Mass: 24570 casein alpha-S1 [Bos taurus]
Delta Miss Score Expect Rank Peptide −0.01 0 73 0.0006 1 YLGYLEQLLR −0.00 0 35 3.2 1 FFVAPFPEVFGK −0.00 0 36 2.2 1 HQGLPQEVLNENLLR 0.01 0 19 84 1 EPMIGVNQELAYFYPELFR
Score : 164
Queries matched: 4
Fig. 12.1 Typical Mascot (Matrix Science) search report of milk-spiked cookie extract (12.5 ppm). The top portion shows a list of significant hits and the middle portion shows the probability-based Mowse Score. The lower portion shows the information on an individual protein (casein alpha S1) including the mass, the score and the sequence of the matched peptide.
Food allergen method development programme at Health Canada
193
Table 12.2 Comparison of Database search results from cookie extracts spiked with milk (Alpha-S1-Casein Protein). Milk spiked (ppm) 1250 125 12.5 1.25 0
Mascot score
Number of peptides matched
Percent protein coverage
721 391 185 100 –
10 7 4 2 –
43 38 17 10 –
duplication and allowing the dissemination of such methods for use in food industry and regulatory settings. These actions will contribute to increased efforts aiming at preventing unintentional exposure of susceptible individuals to allergenic food commodities and the overall improvement of consumer protection.
Acknowledgements The authors acknowledge the contribution of the Laboratory of Hormonology, Centre ´ d’Economie Rurale, Province of Luxembourg, Belgium, for its support and contribution to the development of Health Canada’s food allergen method development programme.
REFERENCES Ben Rejeb S., Davies D., Cl´eroux C., Langlois D. and Delahaut P. (2002) Presentation at the 116th AOAC International Annual meeting and Exposition, 22–26 September 2002, Los Angeles, CA, Abstract # C-124, p. 102. Ben Rejeb S., Abbott M., Davies D. et al. (2003) Immunochemical-based method for detection of hazelnut proteins in processed foods. Journal of AOAC International, 86(3), 557–563. Ben Rejeb S., Lauer B., Salminen J. et al. (2004) Regulatory and compliance activities to protect food-allergic consumers in Canada: research in support of standard setting and consumer protection. Journal of AOAC International, 87(6), 1408–1416. Canadian Council of Social Development (2006) The Progress of Canada’s Children and Youth. Available at http://www.ccsd.ca/pccy/2006/pdf/pccy health.pdf, accessed 21 April 2009. Clark A.T. and Ewan P.W. (2008) Good prognosis, clinical features, and circumstances of peanut and tree nut reactions in children treated by a specialist allergy center. The Journal of Allergy and Clinical Immunology, 122(2), 286–289. FAO/WHO (1998). Report of the Twenty-Sixth Session of the Codex Committee on Food Labelling. Codex Alimentarius Commission, Ottawa, Canada. ALINORM 26–29 May 1998, 99/22. Health Canada (2008) Web-based Compendium of food allergen methodologies. Available at http://www.hcsc.gc.ca/fn-an/res-rech/analy-meth/allergen/index e.html, accessed 21 April 2009. Kagan R.S., Joesph L. Dufresne C. et al. (2003) Prevalence of peanut allergy in primary school children in Montreal, Canada. Journal of Allergy and Clinical Immunology, 112, 1223–1228. Newsome W.H. and Abbott M.A. (1999) An immunoaffinity column for the determination of peanut protein in chocolate. Journal of AOAC International, 82(3), 666–668. Salter J., Mehra S. Cairns J. et al. (2000) A Study of 32 Food-Induced Anaphylaxis Deaths in Ontario 1986–2000. Anaphylaxis Canada. Available at http://www.anaphylaxis.org/content/programs/ programs research deaths.asp, accessed 21 April 2009.
194
Management of Food Allergens
Sampson H. (2004) Update on food allergy. The Journal of Allergy and Clinical Immunology, 113(5), 805–819. Weber D., Raymond P. Ben-Rejeb S. and Lau B. (2006) Development of a liquid chromatographytandem mass spectrometry method using capillary liquid chromatography and nanoelectrospray ionisationquadrupole time-of-flight hybrid mass spectrometry for the detection of milk allergens. Journal of Agricultural and Food Chemistry, 54, 1604–1610. Yeung J.M. and Collins P.G. (1996) Enzyme immunoassay for determination of peanut proteins in food products. Journal of AOAC International, 79(6), 1411–1416. Yeung J.M. and Collins P.G. (1997) Determination of soy proteins in food products by enzyme immunoassay. Food technology and biotechnology, 35(3), 209–214. Yeung J.M. and Abbott M.A. (2000) Determination of egg proteins in food products by enzyme immunoassays. Journal of AOAC International, 83(1), 139–143.
Part III
Risk Communication
13
Finished product labelling and legislation
Sue Hattersley
13.1 INTRODUCTION There are currently no cures for food allergy or food intolerance. The only way an individual can prevent allergic symptoms occurring is to avoid exposure to the food to which they are sensitive. It is therefore essential for people with food sensitivities that there is accurate and detailed information available about the foods that they wish to purchase to ensure their safety and to allow them to make informed choices. There have been significant improvements in the legal requirements for clear declaration of allergenic ingredients used in foods sold pre-packed in recent years. These were needed to address the exemptions in general ingredients’ listing requirements that meant that people with food allergies or intolerances did not always have full information about a product. The provisions of Directive 2003/89/EC and subsequent amendments, including the development of a list of derived ingredients exempt from the allergen labelling provisions is described in this Chapter. In addition, a number of initiatives that have been developed by the UK Food Standards Agency covering voluntary guidance on areas outside the statutory provisions are also described.
13.2
13.2.1
LEGISLATION ON ALLERGEN LABELLING – EUROPEAN DIRECTIVE 2003/89/EC AND SUBSEQUENT AMENDMENTS Requirements of the legislation
European Union Directive 2003/89/EC, which was published on 25 November 2003, amended Directive 2000/13/EC (which sets out general rules on the labelling of foodstuffs) and introduced a list of allergenic ingredients that must be clearly declared when they are used in food products. European Union member states were required to implement these provisions into national legislation within 12 months (i.e. by 25 November 2004) and foods not complying with these provisions were prohibited from sale as from 25 November 2005, although products placed on the market or labelled before this date could continue to be sold, whilst stocks lasted. This directive introduced a requirement to make a clear reference to a list of 12 ingredients known to cause food allergies or intolerances, whenever they are used in pre-packed foods, regardless of their level of use. All added ingredients and components of added ingredients are covered by these requirements, including carry-over additives, additives used as processing
198
Management of Food Allergens
aids, solvents and media for additives or flavourings and any other substances used as processing aids that are normally exempt from ingredients listing requirements.
13.2.2
Allergens that are covered
The Annex IIIa list in Directive 2003/89/EC containing 12 allergenic ingredients that had to be declared was determined following advice from the European Food Safety Authority (EFSA) on the foods of greatest public health concern in European Union countries (see Table 13.1). The terms used when listing the allergenic ingredients should closely resemble those used in the Annex IIIa list. Thus for cereals containing gluten, it is the cereal (wheat, barley, rye, etc.) that should be declared, rather than gluten per se, although gluten can also be declared in a voluntary allergy advice statement or box, if one is used. The legislation does not name any individual species of crustaceans or fish and therefore all species are covered. Use of common species names, such as crab, prawns, lobster, cod, mackerel or salmon are regarded as sufficient to indicate the use of crustacean or fish ingredients, but it is important that the term ‘crustacean’ or ‘fish’ is used for any unusual species with which consumers might not be familiar. The legislation does not specify the species of eggs or milk, but these should be taken to include eggs from laying hens and other birds, such as duck, turkey, quail, goose, gull and guinea fowls, and milk from cows, sheep, goats, and buffaloes, etc. Many dairy products, such as cheese, butter, fermented milk and cream are considered to be refer clearly to the milk base of these products and for such products further reference to milk is not necessary. However, for unfamiliar dairy products used as ingredients, and for components derived from milk, such as casein and whey, there should be a clear reference to milk when these ingredients are listed. Although peanuts may commonly be referred to as groundnuts or monkey nuts, the term ‘peanuts’ should be used for allergen labelling purposes. The legislation specifies the Table 13.1 Annex IIIa list of allergenic ingredients that have to be clearly declared – taken from Directive 2003/89/EC. Annex IIIa ingredients Cereals containing gluten (namely, wheat, rye, barley, oats, spelt, kamut or their hybridised strains) and products thereof Crustaceans and products thereof Fish and products thereof Eggs and products thereof Peanuts and products thereof Soy beans and products thereof Milk and products thereof Nuts (namely, almond (Amygdalus communis L.), hazelnut (Corylus avellana), walnut (Juglans regia), cashew (Anacardium occidentale), pecan nut (Carya illinoiesis (Wangenh.) K. Koch), Brazil nut (Bertholletia excelsa), pistachio nut (Pistacia vera), macadamia nut and Queensland nut (Macadamia ternifolia)) and products thereof Celery and products thereof Mustard and products thereof Sesame seeds and products thereof Sulphur dioxide and sulphites at levels above 10 mg/kg or 10 mg/L expressed as SO2
Finished product labelling and legislation
199
particular tree nuts that are included in the allergen labelling provisions. There a number of other foods that are also called ‘nuts’, such as coconuts, chestnuts and pine nuts that were not included in the legislation as they were not in the list of nuts of public health concern identified by the EFSA. The term ‘celery’ includes both stick celery and root celery (celeriac) as well as celery seeds and the term ‘mustard’ includes all part of the plant, including sprouted seeds and leaves, as well as mustard powders and oils. The term ‘sesame seed’ also includes products made from sesame seeds, such as sesame oil and tahini. Some foods, such as onions, garlic and some dried fruits, naturally contain sulphur dioxide or sulphites. The legislation, however, covers only added sulphites used as preservatives at levels above 10 mg/kg or 10 mg/L, in the finished product as consumed; that is when prepared according to the manufacturer’s instructions, for example, for concentrated fruit squashes intended to be diluted before consumption. The legislation allows for this list to be revised on the basis of updated scientific evidence and two further allergenic ingredients, lupin and molluscs (gastropods, bivalves and cephalopods), were subsequently added to the Annex IIIa list in 2007 under the provisions of Directive 2006/142/EC. Lupin was added to the Annex IIIa list because of allergy to lupin itself and because of evidence of a significant degree of cross-reactivity with approximately 40% of peanut-allergic people also reacting to lupin. Molluscs were added to the list because there is evidence of allergy to molluscs and also of some cross-reactivity between molluscs and crustaceans. The species of molluscs are not specified in the legislation but all species are included, including squid, mussels and snails. The revised Annex IIIa list now covers 14 major allergens.
13.2.3 Exemptions for certain products derived from allergenic ingredients As Directive 2003/89/EC was being developed, it was recognised that some highly processed ingredients derived from these specified allergenic ingredients would not in practice be allergenic and should therefore be exempt from the allergen labelling provisions. Not exempting such ingredients would not be scientifically justified (as they pose no allergenic risk) and would further reduce the choice of food products that could be eaten by food allergic consumers. Exemptions from the provisions of Directive 2003/89/EC were granted for a number of ingredients derived from the allergenic foods listed in Annex IIIa, on the basis of opinions from the EFSA on dossiers submitted by the food industry. The EFSA opinions on the dossiers that were submitted can be found on the EFSA website at http://www.efsa.europa.eu/ EFSA/efsa locale-1178620753812 ScientificOpinionPublicationReport.htm. It was, however, not possible for industry to compile all the evidence needed to support the exemption of the derived ingredients within the timescale before the provisions of Directive 2003/89/EC came into effect in November 2005. Therefore, the European Commission decided that temporary exemptions would be granted for those products where the existing supporting evidence was considered by EFSA to be strong. The food industry then had a 2-year period to provide further dossiers to support permanent exemption. Directive 2005/26/EC was published on 22 March 2005, and set out a list of derived products that
200
Management of Food Allergens
Table 13.2 List of food ingredients and substances provisionally excluded from the Annex IIIa list of allergenic foods – taken from Directive 2005/26/EC. Ingredients
Products thereof provisionally excluded
Cereals • Wheat-based glucose syrups included dextrose containing gluten • Wheat-based maltodextrins • Glucose syrups based on barley • Cereals used in distillates for spirits Eggs
• Lysozyme (produced from egg) used in wine • Albumin (produced from egg) used as fining agent in wine and cider
Fish
• Fish gelatine used as carrier for vitamin or carotenoid preparations and flavoursa • Fish gelatine or isinglass used as fining agent in beer, cider and wine
Soya bean
• Fully refined soya bean oil and fatb • Natural mixed tocopherols (E306), natural d-alpha tocopherol, natural d-alpha tocopherol acetate, natural d-alpha tocopherol succinate for soya bean sources • Vegetable oil derived phytosterols and phytosterol esters from soya bean sources • Plant stanol ester produced from vegetable oil sterols from soya bean sources
Milk
• Whey used in distillates for spirits • Lactitol • Milk (casein) products used as fining agents in cider and wine
Nuts
• Nuts used in distillates for spirits • Nuts (almonds, walnuts) used (as flavour) in spirits
Celery
• Celery leaf and seed oil • Celery seed oleoresin
Mustard
• Mustard oil • Mustard seed oil • Mustard seed oleoresin
a
Directive 2005/63/EC made a minor correction to this exemption in Directive 2005/26/EC. And products thereof, in so far as the process that they have undergone is not likely to increase the level of allergenicity assessed by the EFSA for the relevant product from which they originated. b
were temporarily exempt from the allergen labelling provisions until 25 November 2007 (see Table 13.2). Member states had to implement the provisions of this directive into national legislation by 25 November 2005. Subsequently, EFSA evaluated the further dossiers submitted by industry to support permanent exemptions. In some cases, industry did not submit any further dossiers to support the continued exemption of some of the derived products and therefore such exemptions were not made permanent. These EFSA opinions can be found on their website at http://www.efsa.europa.eu/EFSA/efsa locale-1178620753812 ScientificOpinion PublicationReport.htm. Directive 2007/68/EC, which was published on 28 November 2007, sets out an amended Annex IIIa list of the allergenic ingredients that have to be clearly declared and the permanent exemptions from these requirements (see Table 13.3). Member states have to implement these provisions into national legislation by 31 May 2008, although foodstuffs placed on the market or labelled before 31 May 2009 that comply with the provisions of Directive 2005/26/EC can continue to be marketed until stocks are exhausted. There may be other ingredients derived from the Annex IIIa list of allergenic foods that are also not an allergenic risk because of the degree of processing that they have undergone.
Finished product labelling and legislation
201
Table 13.3 Amended Annex IIIa list including permanent exemptions – taken from Directive 2007/68/EC. Ingredient
Exemptions
Cereals containing gluten (namely, wheat, rye, barley, oats, spelt, kamut or their hybridised strains) and products thereof
(a) Wheat-based glucose syrup including dextrosea (b) Wheat-based maltodextrinsa (c) Glucose syrups based on barley (d) Cereals used for making distillates or ethyl alcohol of agricultural origin for spirit drinks and other alcoholic beverages
Crustaceans and products thereof Eggs and products thereof
None None
Fish and products thereof
(a) Fish gelatine used as carrier for vitamin or carotenoid preparations (b) Fish gelatine or isinglass used as fining agent in beer and wine
Peanuts and products thereof Soya beans and products thereof
None (a) Fully refined soya bean oil and fata (b) Natural mixed tocopherols (E306), natural d-alpha tocopherol, natural d-alpha tocopherol acetate, natural d-alpha tocopherol succinate for soya bean sources (c) Vegetable oil derived phytosterols and phytosterol esters from soy bean sources (d) Plant stanol ester produced from vegetable oil sterols from soy bean sources
Milk and products thereof (including lactose) except
(a) Whey used for making distillates or ethyl alcohol of agricultural origin for spirit drinks and other alcoholic beverages (b) Lactitol
Nuts (namely, almond (Amygdalus communis L.), hazelnut (Corylus avellana), walnut (Juglans regia), cashew (Anacardium occidentale), pecan nut (Carya illinoiesis (Wangenh.) K. Koch), Brazil nut (Bertholletia excelsa), pistachio nut (Pistacia vera), macadamia nut and Queensland nut (Macadamia ternifolia)) and products thereof
Nuts used for making distillates or ethyl alcohol of agricultural origin for spirit drinks and other alcoholic beverages
Celery and products thereof Mustard and products thereof Sesame seeds and products thereof Sulphur dioxide and sulphites at levels above 10 mg/kg or 10 mg/L expressed as SO2
None None None None
Lupin and products thereof Molluscs and products thereof
None None
a
And products thereof, in so far as the process that they have undergone is not likely to increase the level of allergenicity assessed by the EFSA for the relevant product from which they originated.
However, it is the responsibility of the food industry to submit dossiers to support exemption of particular derived products. If the food industry wishes to obtain further exemptions, supporting dossiers should be submitted to the European Commission for evaluation by EFSA.
202
Management of Food Allergens
13.2.4
UK national legislation and accompanying guidance to help businesses with allergen labelling
The Food Standards Agency has responsibility for food policy in the United Kingdom and has implemented the various pieces of European allergen labelling legislation into national law. As a result of devolution of certain powers to Scotland, Wales and Northern Ireland, food legislation is the United Kingdom is implemented separately in England and in the devolved countries. Table 13.4 sets out the relevant allergen labelling directives and the national implementing regulations across the United Kingdom, together with the main provisions of each piece of legislation. The UK Food Standards Agency also produces guidance notes to accompany legislation, which aim to provide informal, non-statutory guidance on the regulations and include advice and best practice. Whilst these guidance notes are not an authoritative interpretation of the law, they are intended to help the food industry to adopt consistent, transparent labelling practices and to help enforcement authorities to identify and advise on misleading labelling. In this way, consumers should find it easier to interpret and use the labelling information provided on food products. Guidance was produced to accompany the 2004 and 2005 Regulations when they were published. However, when the 2007 Regulations were produced, it was decided to merge Table 13.4 provisions.
EU Directives, corresponding national legislation in the United Kingdom and their main
European directive
National legislation
Main provisions
2003/89/EC
The Food Labelling (Amendment) (England) • Clear declaration of specified (No 2) Regulations 2004 (SI 2004/2824) and allergenic ingredients by corresponding regulations in Scotland, Wales • Removes the ‘25% rule’ on and Northern Ireland. exemption from the labelling provisions for components of compound foods • Allergen declaration on alcoholic drinks (above 1.2% alcohol by volume) using ‘contains allergen X’ statement
2005/26/EC (as amended by 2005/63/EC)
The Food Labelling (Amendment) (England) • Temporary exemptions until (No 2) Regulations 2005 (SI 2005/2057) and November 2007 for some by corresponding regulations in Scotland, Wales products derived from the and Northern Ireland, as amended by The Food allergenic ingredients in the Labelling (Amendment) (England) (No 2) Annex IIIa list (Amendment) Regulations 2005 (SI 2005/2969 and corresponding regulations in Scotland, Wales and Northern Ireland.
2006/142/EC
The Food Labelling (Declaration of Allergens) (England) Regulations 2007 (SI 2007/3256) and by corresponding regulations in Scotland, Wales and Northern Ireland
• Adds lupin and molluscs to the Annex IIIa list
2007/68/EC
The Food Labelling (Declaration of Allergens) (England) Regulations 2008 and corresponding regulations in Scotland, Wales and Northern Ireland
• Provides for a number of permanent exemptions from the Annex IIIa list
Finished product labelling and legislation
203
the two existing sets of guidance notes into a single document and also to include the new provisions within this document. In this way, the guidance on all allergen labelling legislation in the United Kingdom is available within a single document. As the 2008 Regulations are implemented, this document will be further amended to include these new provisions. The guidance notes are available from the Food Standards Agency website at http://www.food. gov.uk/multimedia/pdfs/allergenlabelguide2007.pdf.
13.2.5 Allergy advice boxes or statements The provisions in the legislation require the clear declaration of the use of allergenic ingredients, and this is normally achieved via the ingredients list or the name of the product. However, some food manufacturers and retailers also choose to include additional allergy advice statements or boxes to highlight allergy information for consumers. The use of such statements or boxes is voluntary and therefore they should not be relied on by food allergic consumers who are advised to always read the ingredients list. It is important that if such statements or boxes are used, they list all the relevant allergens from the Annex IIIa list that are used in the product. The wording used in such statements or boxes need not reflect exactly the terminology used in the regulations. For example, the term ‘shellfish’ could be used if crustaceans and/or mollusc ingredients are used in the food, and in addition, the term ‘gluten’ could be used as well as the specific gluten containing cereal that is identified in the ingredients list.
13.2.6
Food withdrawals and recalls
If the labelling of allergenic ingredients used in food products does not meet the statutory requirements or is misleading (for example, if voluntary information about allergens is incorrect), then that food represents a potential risk to health for the allergic consumer. In such cases, the food product should be withdrawn from sale and in some cases may need to be recalled. In the United Kingdom, the most common reason why allergen labelling information is incorrect is because the packaging/labelling has been used for the wrong product, particularly when more than one variety of a product is manufactured (see Section 11.3.2).
13.3 ALLERGEN CROSS-CONTAMINATION AND ADVISORY LABELLING (SUCH AS ‘MAY CONTAIN’ STATEMENTS) 13.3.1 Background Allergens that may be present in foods accidentally as a result of cross-contamination at some point in the food supply chain are not covered by statutory labelling provisions in Europe. Manufacturers may choose to tell the consumer about the risk of allergen crosscontamination via advisory labelling, using phrases such as ‘may contain traces of nuts’. The food industry, although it had produced some industry guidance in this area, asked the UK Food Standards Agency to draw together all the existing best practice advice and produce a single, authoritative guidance document. In addition, consumers had expressed concerns
204
Management of Food Allergens
about the increase in use of these advisory statements in recent years, which devalues their impact and which can severely restrict the choice of certain types of food products for people with food allergies. The Agency therefore decided to produce best practice guidance in this area to help both the industry and consumers. The Agency worked together with the food manufacturers and retailers, as well as allergy support organisations and enforcement authorities, to develop guidance on allergen management and advisory labelling, using the already existing industry guidance as a foundation. This new guidance was published in July 2006 and was accompanied by a short booklet aimed at small and micro-businesses. The guidance can be accessed at http://www.food.gov.uk/ multimedia/pdfs/maycontainguide.pdf. The purpose of this guidance was to help food businesses identify where allergen cross-contamination might occur and how this could be reduced or eliminated, and also to address the issue of the wording in the advisory statements that were used if an unavoidable risk of cross-contamination was identified. It was clear that the diversity of advisory phrases used by different food businesses was confusing for consumers who often try to equate different levels of risk to the different phrases used. For example, many consumers believe that a phrase such as ‘made in a factory where allergen X is handled’ identifies a lower risk than a phrase such as ‘made on equipment that also handles allergen X’. It was also questioned whether the phrase ‘may contain traces of nut’ was appropriate where there was a risk, albeit a remote risk, of cross-contamination with pieces of nut. The guidance was also the subject of a public consultation. Although the initial expectation was some level of quantitative information about accidental allergen levels that might justify the use of advisory labelling could be included in the guidance, it was clear from the comments received to the public consultation that this would not be possible. This was because there is a lack of scientific knowledge regarding the clinical threshold doses of food allergens below which reactions are not triggered in allergic subjects, and the extent of the variability between subjects was also not clear. In addition, there was an absence of recognised and internationally accepted risk assessment procedures to be used to turn clinical thresholds (when they are determined) into management threshold levels to be used by industry. The guidance that was published therefore sets out a qualitative approach to allergen management and the document acknowledges that there will need to be a revision of the guidance as scientific knowledge develops. The focus of the guidance relates to the production of food that will be pre-packed and which is therefore subject to the provisions of legislation governing the labelling of allergenic ingredients. However, the principles set out in the guidance can also be applied to the production of food that will not be pre-packed, that is foods sold loose (such as in bakeries or at delicatessen counters) or in catering situations.
13.3.2
Management of allergen cross-contamination
The guidance sets out an approach to the evaluation of the likelihood of unintentional allergen cross-contamination. It also gives advice on how such contamination could be minimised or eliminated, and if the contamination cannot be avoided, it gives advice on how the risk should be communicated to consumers. Figure 13.1 illustrates the different aspects of the food manufacturing chain where allergen cross-contamination can occur, covering people, raw materials, equipment and premises, cleaning practices and packaging. It should also be noted that allergen management issues should be considered as part of any new product development or product reformulation, as the introduction of a new allergenic ingredient in
Finished product labelling and legislation
205
People
Raw materials and supply chain
Packaging
Allergen management New product development and reformulation
Manufacturing premises, equipment and processes
Cleaning
Fig. 13.1 Aspects of the food manufacturing chain where allergen cross-contamination can occur. (Taken from Allergen Management and Consumer Information, Food Standards Agency, 2006).
a product may lead to complications in allergen management in those premises that could be avoided. All staff involved in food production should be aware of food allergen issues and should be given training in how to avoid allergen cross-contamination, covering issues such as:
r r r r r r
Recognising the ingredients that are of allergenic concern; Identifying situations where cross-contamination is possible; Clothing requirements and hand washing; Cleaning procedures; Management of waste materials and rework and Movement of people and equipment around a site.
Food businesses should check the allergen status of the ingredients that they buy in with their supplier and keep these under review and should consider asking their suppliers to notify them of any changes. Businesses also need to consider how to identify and, if possible, segregate allergenic materials and establish procedures that minimise the risk of allergen cross-contamination during handling within their premises. The implications of any changes in ingredient supplier should also be assessed. Whilst the use of dedicated production facilities may be the ideal approach to allergen management, this is often not possible, especially where production facilities are small or where the product range is large. However, it is possible in many cases to separate the production of product lines containing an allergen from the production of lines not containing that allergenic ingredient. This can be either physical separation by using barriers or different parts of a site or separation by time by appropriate scheduling of production runs and cleaning of equipment between production runs. The business needs to determine how to ensure that
206
Management of Food Allergens
rework or part-prepared material that is held over is correctly identified to prevent allergen cross-contamination. Cleaning practices need to be assessed to try to ensure that allergens are effectively removed from equipment, protective clothing and people’s hands, and it should be noted that cleaning that is sufficient for hygiene purposes may not be adequate to remove certain allergens. These cleaning regimes should be validated, particularly if they are part of the critical control for separation of allergen-containing and non-allergen-containing production runs, and should be monitored. As mentioned in Section 13.2.6, the most common cause of allergen-related food withdrawals or recalls in the United Kingdom is incorrect allergen labelling information, particularly where several varieties of a product are made. In many cases, the product is packaged in the wrong packaging so that the ingredients and allergen information does not relate to the product actually contained within the packaging. Procedures for ensuring that the correct labelling and packaging is used for a particular product are therefore very important and businesses should also ensure that old packaging materials are destroyed following recipe changes. Other reasons for allergy-related food withdrawals and recalls includes inconsistencies within the labelling, such as ingredients’ lists not being consistent with allergy advice statements or boxes or inconsistencies between inner and outer packaging of multi-packs. The guidance produced by the UK Food Standards Agency in 2006 sets out a decision tree approach to the assessment of the risk of potential allergen cross-contamination (see Figure 13.2). This risk assessment should be conducted on a case-by-case basis for any of the allergens for which there is a statutory requirement for labelling when used as an ingredient in a pre-packed food. Whilst the numbers of people affected by some food allergies is higher than others, the severity of the reactions triggered in a sensitive individual may be comparable and therefore it is not possible to assign a ranking of importance of different food allergens for allergen management practices. In the absence of internationally agreed management threshold levels to be used by industry, when reaching a decision on the need, or not, for allergen advisory labelling, a food business needs to consider a number of factors:
r r r r
any information on the amount of the allergen needed to trigger a reaction in a sensitive person; the prevalence of allergy to that particular food in the country in which the food is to be marketed, noting that prevalence does vary between countries, even between different EU member states; the relative allergenicity of the particular ingredient concerned, noting that highly processed ingredients are likely to be less allergenic, and some derived ingredients are in fact exempt from allergen labelling requirements because, on the basis of opinions from the EFSA, they are considered not to be likely to trigger severe allergic reactions and the physical nature of the particular ingredient being used and the geography of the manufacturing environment, as, for example, a powder that may become airborne may represent a greater risk of cross-contamination than an ingredient that is in liquid form.
13.3.3
Advisory labelling
Where a risk of allergen cross-contamination has been identified that cannot be reduced or eliminated, a decision has to be made by the food business on whether or not advisory labelling is appropriate. Whilst consumers are advised to always read the ingredients’ list,
Finished product labelling and legislation
207
Step 1 – Assess risk from intentional presence Is the food manufactured from any of the allergenic foods specified in the current UK legislation or their derivatives (see Appendix I)? YES (Label as necessary. Go to step 7)
NO (Go to step 2)
Step 2 – Assess risk from unintentional presence What is the likelihood, under normal operating conditions, of cross-contamination of the food by specified allergens (Appendix I)? PROBABLE (Go to step 2a)
REMOTE (No action – Go to Step 7)
Step 2a – Check against ingredient labelling Is the potential cross-contaminating allergen already declared on the label as an ingredient? YES (No action – Go to Step 7)
NO (Go to step 3)
Step 3 – Check against exemptions list Is the potential cross-contaminating material exempt from mandatory labelling (see Appendix I)? YES (No action – Go to Step 7)
NO (Go to step 4)
Step 4 – Hazard characterisation Identify the physical form and the characteristics of the potential cross-contaminating allergenic material
Step 5 – Risk management of unintentional presence Can the identified risk of cross-contamination be managed? YES (Go to Step 7)
NO (Go to Step 6)
Step 6 Risk Communication – Include warning on label (Go to Step 7)
Step 7 – Check other relevant allergens Have all relevant allergens been considered? YES (No Action)
NO (Go back to Step 1)
Fig. 13.2 Decision tree approach to the assessment of the risk of potential allergen cross-contamination. (Taken from Allergen Management and Consumer Information, Food Standards Agency, 2006).
208
Management of Food Allergens
it is clear that many consumers use allergen advice statements or boxes as a first point of information about a particular product. Therefore, such statements or boxes should always be placed in close proximity to the ingredients’ list and should clearly distinguish between those allergens that are present as deliberate ingredients (‘contains’) and those that are possible cross-contaminants (using phrases such as ‘may contain’ or ‘not suitable for someone with X allergy’). It is important to note that an assessment by a food business that the risk of possible allergen cross-contamination is so remote that allergen advisory labelling is not deemed appropriate is not the same as determining that a product can use a ‘free from’ or a ‘made in an allergen X free factory’ claim.
13.3.4
Determination of allergen management thresholds
As noted above, the current scientific evidence does not yet allow the determination of internationally accepted allergen management thresholds for industry to underpin labelling decisions. The UK Food Standards Agency together with the EU-funded EuroPrevall project organised a workshop in 2007 to consider whether risk assessment models used in food chemical and contaminant risk assessments could be used in assessing the risks posed by food allergens. There was agreement that approaches such as no observable adverse effect levels (NOAELs) and lowest observable adverse effect levels (LOAELs), benchmark doses and margins of exposure, or probabilistic modelling could be applicable. However, a number of gaps in the existing scientific data were identified that would need to be addressed before such approaches could be used in practice. Much work is already ongoing to address these identified gaps in the scientific data and it is hoped that it will be possible in the near future to start to determine management thresholds for use by the food industry. An allied issue that also needs to be addressed is the standardisation and validation of detection methods for the food allergens of concern. Whilst many commercial manufacturers of allergen detection kits are developing an ever-wider range of kits that can be used by food manufacturers as part of their quality control procedures, there is still a need to establish validated methods for use in enforcement activities. One particular concern is the current lack of the standardised reference materials for the allergens of concern to calibrate existing analytical methodologies.
13.4 PROVISION OF ALLERGY INFORMATION FOR FOODS THAT ARE NOT PRE-PACKED 13.4.1
Background
Member states of the European Union can generally choose which of the mandatory information for pre-packed foods (as set out in Directive 2000/13/EC, as amended) should apply to foods sold non-pre-packed. The United Kingdom, in common with other member states, does not require mandatory labelling of allergenic ingredients in such foods, as there has been a general assumption that the person making the food is available to the consumer and so can answer questions about the ingredients in that product. The term ‘non-pre-packed’ is used to describe food that is sold without packaging (such as food sold loose on market stalls and in restaurants, cafes, coffee shops, etc.) as well as food that is placed in packaging at the point of sale (such as sandwiches or food sold in a bakery or at a delicatessen counter).
Finished product labelling and legislation
209
However, information from clinicians shows that in people who know which foods they are allergic to and who are actively trying to avoid those foods reactions are more likely to occur after eating food that is not pre-packed than food that is subject to the current labelling legislation. The UK Food Standards Agency, which for a number of years had been considering ways of improving the allergen information available to consumers when they are eating out, issued best practice guidance on this in January 2008. The guidance and accompanying leaflet can be accessed at http://www.food.gov.uk/foodindustry/guidancenotes/labelregsguidance/ nonprepacked. Following discussions with stakeholders and a public consultation, the Agency decided that a best practice guidance approach would be more helpful in this area and that any mechanisms for providing such information needed to be practical and workable, given the size and nature of many of the food businesses concerned. Although there are some large catering chains, many of the businesses in this sector of the food industry are small or micro-businesses and the composition of the dishes provided can vary on a daily basis. It was recognised that whilst many consumers do not want detailed information on all the ingredients in a meal when they are eating out, for consumers with food allergies or intolerances, this is essential information. Whilst businesses should provide reasonable descriptions of the dishes being offered, it is not practical to provide full ingredient information in all cases, but this information should be available to be provided on request. The EU announced a review of food labelling in 2005 and the proposal for a new regulation that was issued in January 2008 includes mandatory labelling of allergenic ingredients for non-pre-packed foods. This proposal is currently being negotiated by EU member states and there will need to be careful consideration to address the advantages and disadvantages of the proposed provisions.
13.4.2 Approach to providing allergen information The Agency’s guidance is voluntary and its aim is to help food businesses deal with enquiries from food allergic customers and provide accurate and reliable information about the use of allergenic ingredients and about the risks of possible allergen cross-contamination during preparation and serving. The guidance was developed in partnership with caterers, retailers, consumer organisations and enforcement bodies and includes a number of key messages, together with example scenarios of different types of businesses within the non-pre-packed foods sector.
13.4.3
Key messages
The key objective of the guidance is to improve the dialogue between the person providing the food and the person wishing to purchase the food. There is a responsibility on customers with food allergies and intolerances to ask about the use of the foods to which they are sensitive in the dishes on offer and there needs to be good communication between the customer and the staff, both those serving customers and those preparing the food. It is also important that there is good communication between the different staff so that those serving customers are aware of any changes in recipes, as this may alter the allergen content. Staff should never guess about the allergenic ingredients used in a particular dish and should always check. There may be situations where the standard dishes being produced may not be suitable for someone with a particular allergy but it would be possible for the business
210
Management of Food Allergens
to produce an alternative that did not contain the allergen. For example, a salad could be prepared separately without the nut oil dressing or meat cooked without a sauce containing milk or cream. In order for staff to be able to deal competently with enquiries from food allergic customers, it is important that they receive proper training. There should be an agreed practice within the business for dealing with allergy information requests and all staff should be told about this, including who they should refer such requests to if they cannot answer them themselves. Businesses also need to have systems for accessing information about the ingredients used in the foods they sell. This could be recipes for foods made from fresh ingredients or retention of information about allergenic ingredients in items that are bought in, either as complete dishes or as part-prepared foods. This information needs to be kept up-to-date and to be accessible to staff.
13.5 ‘FREE FROM’ FOODS There is at present no legislation in the European Union governing the use of labelling for foods described as ‘free from’ a particular allergen. There is a growing market for substitute foods made without certain common allergenic foods such as milk, eggs or cereals containing gluten. If businesses wish to provide foods for this market, it is essential that the claims made be based on specific rigorous controls to ensure their validity. Whilst consumers often assume that a ‘free from’ claim is absolute and means that there is no allergen present, in practice, such claims can only be made on the basis that the particular allergen cannot be detected or the level detectable cannot be quantified. The first area where legislation is likely to provide specific limits to substantiate claims that a food is free from a particular allergen is for gluten-free foods. The Codex Alimentarius (a body that sets standards to facilitate international trade in food) first produced a standard for gluten-free foods in 1981, which was revised in 1983. A further revision of this standard was finalised in 2007 by the Codex Committee on Nutrition and Foods for Special Dietary Uses and is expected to be endorsed by the Codex Commission in 2008. This new standard will set a maximum limit of 20 ppm gluten for foods that can be labelled as ‘gluten free’, covering both foods containing naturally gluten-free ingredients and also foods containing ingredients derived from cereals such as wheat, that have been treated to reduce their gluten content. The Codex standard also includes a category of foods containing between 21 and 100 ppm of gluten. The acceptability of this category of foods, and the labelling to be applied to them, can be determined by individual countries but they cannot be described as gluten free.
13.6 CONCLUSIONS In recent years, significant progress has been made in providing allergic consumers with the information they need to make safe and healthy food choices. This progress is particularly evident in the European Union in relation to the labelling of pre-packed foods, where there is now a list of 14 allergenic foods that must be clearly declared on the label when they are used as deliberate ingredients. In addition to these statutory provisions, the UK Food Standards Agency has developed best practice voluntary guidance in two key areas not covered by the current legislation. These are qualitative advice on the management of allergen
Finished product labelling and legislation
211
cross-contamination, including advice on the advisory labelling to warn consumers of the risk of such cross-contamination (such as the phrase ‘may contain’) and guidance on the provision of allergen information for foods sold non-pre-packed. However, much work still remains to be done, particularly in relation to setting quantitative management thresholds for allergen cross-contamination for food businesses to use to inform decisions about the need for advisory labelling as well as for setting thresholds to substantiate claims that a food is free from a particular allergen.
REFERENCES European Commission (EC) (2007) COMMISSION DIRECTIVE 2000/13/EC of the European Parliament and of the Council 20 March 2000. Official Journal of the European Union, L109, 29–41. European Commission (EC) (2007) COMMISSION DIRECTIVE 2003/89/EC of the European Parliament and of the Council 10 November 2003 amending Directive 2000/13/EC. Official Journal of the European Union, L308, 15–18. European Commission (EC) (2007) COMMISSION DIRECTIVE 2005/26/EC of 21 March 2005 establishing a list of food ingredients or substances provisionally excluded from Annex IIIa of Directive 2000/13/EC of the European Parliament of the Council. Official Journal of the European Union, L75, 33–34. European Commission (EC) (2007) COMMISSION DIRECTIVE 2006/142/EC of the European Parliament and of the Council 22 December 2006 amending Annex IIIa of Directive 2000/13/EC. Official Journal of the European Union, L368, 110–111. European Commission (EC) (2007) COMMISSION DIRECTIVE 2007/68/EC of 27 November 2007 amending Annex IIIa to Directive 2000/13/EC of the European Parliament and of the Council as regards certain food ingredients. Official Journal of the European Union, L310, 11–14. Food Standards Agency (FSA) (2006) Guidance on Allergen Management and Consumer Information. Best Practice Guidance on Managing Food Allergens with Particular Reference to Avoiding CrossContamination and Using Appropriate Advisory Labelling (e.g. ‘May Contain’ Labelling). FSA, London.
14
Guidelines for manufacturing and certification programmes
Neil Griffiths
14.1
PREFACE
As the number of people with food allergies or intolerances to food continues to grow, it is not surprising that legislators, who react to consumer concern in this area, should take action. This has resulted in new law requiring food businesses to provide information to consumers about the presence or absence of allergenic ingredients in their food products. In order to provide this information correctly and accurately, it is necessary for the food industry to adopt best practices and act diligently in the provision of this information through labelling or other forms of communication. This chapter not only explains the substantive guidance that has resulted to help the industry achieve best practice, but also attempts to explain why the generation and use of independent certification schemes has also occurred in an attempt to restore trust amongst consumers in the quality of the information supplied. This chapter also examines the use of risk through such procedures as HACCP (hazard analysis and critical control point procedures), management practices, environmental procedures and communication techniques which form the basis of most guidance and certification standards. The importance of training is discussed and the difficulties of testing in this area are also examined. The problems of establishing thresholds are considered.
14.2 INTRODUCTION Those consumers who suffer from food allergens or food intolerance rely heavily on the food industry to provide them with accurate information regarding the absence or presence of food allergens (Figure 14.1). There is growing evidence that this is not currently being achieved with a growing number of high-profile product recalls producing a lack of trust in the label by these allergic or food-intolerant consumers. These same consumers also rely on those who enforce the law to ensure legal compliance by the food industry. Again, there is concern that the correct level of enforcement may not be forthcoming due to overstretched resources and the complexity of the problem. This level of change within a relatively short time scale not only brings significant challenges to food business’s large and small to implement the necessary systems and procedures to reduce the risk to a minimum of supplying wrong allergen information, but also places fresh challenges on the enforcement authorities in their assessment of legal compliance. Both face a steep learning curve in which training has been identified as an essential element. Although the food industry occasionally expresses the opinion that many consumers do not read the label (especially when facing another piece of
Guidelines for manufacturing and certification programmes
213
Confusing food labelling For someone with food allergy, a single mistake could be disastrous. ‘Food labelling is at best confusing and at worst life threatening. Blanket warnings mean that people do not take them seriously.’ (Mother of allergic child) Fig. 14.1
Confusing food labelling.
new labelling legislation), this cannot be said for those who suffer from food allergens or food intolerance. They are forced to read the label avidly. As a result, the accuracy of the information supplied becomes paramount and it is not surprising that the food industry has faced growing pressure to implement best practice. In doing so, there is recognition that it is far better to adopt a quality assurance policy than that of quality control, i.e. it is much better to spend the majority of your resources making sure something is right than testing and rejecting that which is wrong. This principle is well established in food safety systems across Europe and its application to the allergen information field is just as relevant. In establishing best practice in this way, the necessary prerequisite programmes (PRPs) are then established and it is possible for the food industry to embody the principles of HACCP into its overall systems and procedures to reduce the risk substantially of supplying the wrong allergen information.
14.3
THE LAW
The necessity for food businesses to provide good quality information to those who suffer from food allergens or food intolerances has grown substantially over the past decade. Until recently, this information has been of a voluntary nature, but the recent introduction of new European law requiring a list of food allergies and their derivatives to be positively labelled on pre-packed products has produced new requirements on food businesses. There is also considerable pressure by the same consumers for information of a similar nature to be available in non-pre-packed foods and in catering establishments. As a result, the European Regulators have stated their intention to bring about this extension to the current legal requirements (FSA, 2008).
14.3.1
The cost
This law also provides penalties for all food businesses should the provision of such information (whether it is legally required or given voluntarily) be incorrect. The provision of information regarding the absence or presence of allergenic ingredients whether on a label, on a notice, on a noticeboard, in a menu, in an advertisement or promotion or by any other form of communication can lead to legal proceedings if this information is wrong. Such proceedings can cause substantive damage to the reputation of food businesses, apart from the suffering that incorrect information can cause to those with food allergies. It is clear, therefore, that the law is a significant driver for the food industry to adopt best practice and as a result demonstrate the need for guidance and standards embodied in certification schemes. However, the possible substantive damage to the brand name from a major incident such as the
214
Management of Food Allergens
death of a child from anaphylaxis resulting from incorrect allergen labelling may far outweigh the fear of legal action. This fear is probably the biggest driver for the food industry to adopt best practice in this area. It is now well understood that the adoption of best practice goes hand in hand with both the protection of brand and the assurance of legal compliance. What is sometimes more difficult to do is to persuade those that run food businesses that investment in the adoption of best practice can have a positive commercial benefit to their businesses. Apart from the law which requires the presence of the main allergenic ingredients to be labelled on pre-packed food products, the law can be more general such as that contained within the General Food Law Regulation (2002, http://eurlex.europa.eu/pri/en/oj/dat/2002/ l 031/l 03120020201en00010024.pdf, accessed 21 April 2009) which establishes the general principle that labelling shall not mislead the consumer. The law concerned in this area is both criminal (which can result in fines and imprisonment) and civil (which can result in damages).
14.3.2
All reasonable precautions and all due diligence
Whether the law is general or specific, it is important that food businesses that are involved in providing information regarding the presence or absence of allergenic ingredients in their food products should adopt best practices and operate diligently to reduce the risk of providing incorrect information. In the United Kingdom, we have been used to the concept of having available the defence in law of demonstrating ‘all reasonable precautions and all due diligence’. All reasonable precautions means there must be a system. All due diligence means the system must be made to work (Food Safety Act, 1990, http://www.opsi.gov.uk/ACTS/acts1990/Ukpga 19900016 en 1.htm, accessed 21 April 2009). Demonstration of adherence to these principles means the business could successfully defend itself against a criminal charge. This defence recognises that the pursuit of perfection is a credible and worthwhile objective. To prosecute and castigate those who have striven for perfection but failed would be a breach of natural justice. As perfection is an impossible dream, an incentive to keep trying would seem a wise policy. Retailer involvement in utilising the prerequisites of due diligence and establishing the reasonableness of relying on the systems and procedures put in place by their suppliers has put further pressure on suppliers to demonstrate best practice (see Figure 14.2).
14.4 VOLUNTARY INFORMATION In establishing the systems and procedures necessary to establish best practice, it is important to recognise the two distinct areas involved in the provision of both the legal information and the voluntary information. The legal information requires food businesses to identify the presence or absence of deliberately added allergenic ingredients or derivatives thereof to their food products. These ingredients may have been added individually or by presence in a compound ingredient such as a seasoning.
14.4.1
‘May contain’ (or similar) labelling
List of ingredients should include only ingredients deliberately added to the product. The practice of including possible contaminants in the ingredients list (so-called ‘last ingredient
Guidelines for manufacturing and certification programmes
215
Due Diligence ‘It shall be a defence for the person charged to prove that he took all reasonable precautions and exercised all due diligence to avoid the commission of the offence’. Section 21 – The Food Safety Act 1990
‘There is a system’ ‘The system must be made to work’ ‘The pursuit of perfection is a credible and worthwhile objective. To prosecute and castigate those who have striven for perfection but failed would be a breach of natural justice. As perfection is an impossible dream, an incentive to keep trying would seem a wise policy.’ Although ALL due diligence and ALL reasonable precautions must be demonstrated, perfection is not required. Fig. 14.2
Due diligence.
listing’) is illegal in the United Kingdom under Section 15(1)(a) of The Food Safety Act (1990, http://www.opsi.gov.uk/ACTS/acts1990/Ukpga 19900016 en 1.htm) (Article 14(1)(a) of The Food Safety (Northern Ireland) Order 1991) and as a result possible allergen cross-contaminants should be declared separately (see Figure 14.3). There is no legal requirement for food labels to carry ‘may contain’ or similar warnings but many manufacturers and retailers provide this information voluntarily in order to indicate the possible presence of unintentional ingredients that people may be allergic to in pre-packed foods. It is surprising to some the current high level of ‘may contain’ labelling that has occurred, and some suspect that this is due to the cost involved in establishing best practice to avoid or reduce substantially the risk of unintentional addition. It is highly unlikely that the food industry would ever consider labelling their products with ‘may contain pathogens’ (such as Salmonella, Listeria or Escherichia coli 0157) because of the negative connotation this would have on their products. However, the food industry would appear to be more comfortable with ‘may contain’ labelling of allergens mostly because a far fewer number of people currently would be likely to avoid buying their products as a result. It is possible in adopting best practice to reduce the risks substantially of both the unintentional presence of food pathogens and of food allergens. This has led to questions by many as to why ‘may contain’ labelling is needed at all. Many, however, accept that this may be a necessary interim measure as the industry struggles to introduce the best practices necessary to substantially reduce the risk of this unintentional presence of food allergens. However, it
The ‘may contain’ debate ‘I had always thought that food labelling was there to help and protect the consumer. But now I wonder if it really exists to protect the food industry.’ (Mother of allergic child) Fig. 14.3
The ‘may contain’ debate.
216
Management of Food Allergens
is believed by many that ‘may contain’ labelling should be based on proper risk assessment rather than a purely commercial approach. As the number of people who suffer from food allergies and food intolerances grow, then it is clear that greater commercial pressure will be placed on food businesses to implement better systems and procedures to reduce the risk of the unintentional presence of food allergens. The liability involved in not placing a ‘may contain allergen’ label on a product is no different than not placing a ‘may contain pathogen’ label once best practice has been established. If anything the liability is lower for allergens as the risk to the food business of an allergic consumer being exposed to an unintentional allergen presence is lower than the risk of exposing a consumer to an unintentional pathogen presence. This is because there are much fewer consumers who will suffer harm from exposure to an allergen than to a pathogen. As such the need for guidance and standards establishing best practice will grow and the take-up of these will increase. This is already becoming evident as many retailers work with their suppliers to reduce the level of ‘may contain’ labelling on their products. Consumers who suffer from food allergens or intolerances do not find ‘may contain’ labels useful. Does it or doesn’t it contain an allergen? Should they avoid these products or not? Is this labelling just used to protect the manufacturers and avoid any liability? More importantly is its overuse likely to lead to complacency by those who suffer from food allergies. These concerns are recognised as the recent Food Standards Agency Guidance states, ‘There is general agreement between the food industry, consumer support groups and enforcement bodies that excessive use of advisory warning labels about the possible presence of allergens, not only unnecessarily restricts consumer choice, but also devalues the impact of the warning labels.’ Will the wide use of ‘may contain’ labelling cause some allergic consumers to ignore these warnings as superfluous? It is very evident that best practice needs to be established and used by food businesses in the area of ‘may contain’ labelling.
14.4.2
‘Free from’ (or similar) labelling
For those people who suffer from food allergies, the food industry has recognised the marketing opportunities of supplying a range of ‘free from’ or ‘suitable for’ food products. These types of products involve new challenges for the food industry, as a claim of this sort introduces a higher level of legal liability on the business. An ‘allergen free’ claim is an absolute claim, which may be interpreted by consumers to mean a complete absence, whereas the best that can be scientifically demonstrated at present is that samples of the food were shown to be below the analytical limit of detection of a testing method on one or more occasions. At some point in the future, it is possible that thresholds may be set, but currently there is no agreement or recognition of such levels. As a result, there are those that feel that claims of this sort may suggest a level of guarantee that may not be achievable. This was very evident from recent discussions with the United Kingdom Accreditation Service (UKAS) who expressed the view that the ‘free from’ claim could not be a certifiable claim. At the very least, these claims demand a level of best practice that may also demand dedicated sites and/or substantive testing. Currently, there is no legislation to support these claims. There is, however, a Codex Alimentarius standard for gluten free when applied to products with ingredients derived from gluten-containing cereals. There is also a proposed standard for gluten free for products which are naturally gluten free. ‘Free from’ may start to be incorporated into legislation but at present this seems to be unlikely until and if agreement on thresholds can be achieved.
Guidelines for manufacturing and certification programmes
217
14.5 GUIDELINES We have seen the development of substantive guidance documents in recent years. Guidelines have and can be produced by industry itself, enforcement bodies, consumer bodies and regulatory bodies (FSA Guidance on Allergen Management and Consumer Information, http://www.food.gov.uk/multimedia/pdfs/maycontainguide.pdf, accessed 21 April 2009; FSA Clear Labelling, http://www.foodstandards.gov.uk/multimedia/pdfs/clearlabelling.pdf, accessed 21 April 2009; Anaphylaxis Campaign, 2007; BRC Guidance – Allergen Q&A). Of particular note has been the preponderance of guidance generated by regulatory bodies (in consultation with interested parties) to help the implementation and adherence to new legal requirements in an attempt to establish best practice. Guidance of this sort produced by regulatory bodies carries legal significance. Although not statutory in nature, this type of guidance carries greater significance as it has been demonstrated that courts take these guidance notes very seriously when trying to establish whether the defendant has established the UK defence of ‘all due diligence and all reasonable precautions’. It is perhaps interesting to note that although guidance emanating from regulatory bodies carries this greater regulatory significance, it is the extensive consultation this guidance has undergone which carries the greater significance rather than the guidance has been generated by a regulatory body. The fact that industry, enforcement, consumer bodies and other relevant interested parties have had their opportunity to participate in the formation and agreement of guidance increases the kudos of such guidance and removes, in the main, accusations of bias which may result should, for example, guidance only be generated by the food industry. A good example of such guidance is that produced by the Food Standards Agency (FSA) on Allergen Management and Consumer Information – Best Practice Guidance on Managing Food Allergens with Particular Reference to Avoiding Cross-Contamination and Using Appropriate Advisory Labelling (e.g. ‘May Contain’ Labelling). This guidance went through extensive consultation and was supported by the Food and Drink Federation (FDF), the Anaphylaxis Campaign, the British Retail Consortium and LACORS (see Figure 14.4). Unfortunately, although this guidance contains within it many of the good practices necessary to reduce the risk of communicating wrong allergen information, it does not follow that Exercising due diligence – ‘The system must be made to work’ This best practice guidance is going to look really impressive! What a shame none of our suppliers are using it!
Fig. 14.4
Exercising due diligence.
Look at the amount they claim they do. I wonder if they check it has been implemented and is working?
218
Management of Food Allergens
industry automatically embodies these practices within its day-to-day operations. Commercial realities and competitive pressures means food industry bodies themselves have favoured the development and introduction of certification schemes (some embodied in commercial contracts of supply) rather than rely or the willingness of food businesses to introduce best practice embodied in guidance or on enforcement bodies to ensure a level playing field. The British Retail Consortium Global Food Standard is a good example of such an initiative.
14.6 CERTIFICATION SCHEMES The development of certification schemes to demonstrate and give confidence in a food business’ ability to meet good business practice requirements embodied in standards has evolved over recent years. The standards and certification process have either been embodied in law (such as organic or beef labelling) or has evolved through industry initiatives (such as the BRC Global Standard – Food Safety, http://www.food-technologists.co.uk/ PageL2.asp?PageName=NewPage&PageNameL2=BRCEXLEVEL1%252E1, accessed 21 April 2009). The growth of schemes like this has demonstrated the need for such initiatives to supplement internal auditing and self-regulatory practices at a time of high competition and price pressures for the food industry. The fact that these schemes have evolved, both legally and within and between the food industry, supports this contention and gives weight to their need and benefits.
14.6.1
The benefits
Certification schemes based on good food business practices do, and have, reduced the risks of getting it wrong. Many have reduced duplication of effort and the substantive costs that can result on food businesses in product recall or withdrawal. They have aided or avoided costly court action and helped protect company name and reputation. It is important, therefore, that the benefits of implementing certification schemes referred to above should be properly taken into account when food businesses are assessing the costs of properly embodying the requirements of any standard into their business, and the costs of the certification process (see Figure 14.5).
Certification schemes BRC Food/Packaging/Consumer Products IFS ISO EN 22000 Beef Labelling Organic Standards Agricultural Standards (e.g. Eurepgap) Feed Standards (e.g. FEMAS) Many operated under EN 45011 (Product Certification or EN 45012 (Management System Certification) Fig. 14.5
Certification schemes.
Guidelines for manufacturing and certification programmes
219
It is important to realise that the use of certification schemes can only reduce the risk of communicating the wrong allergen information but cannot eliminate the risk completely. It is important that this concept is communicated to consumers as it may not be appropriate to give cast iron guarantees of freedom from a specific allergen based on adherence to certification schemes.
14.6.2
Independence
All these schemes independently assess the particular food business’ compliance to the standard and correct any deficiencies identified by the requirements of the standard. They attempt to resolve the ‘marking your own exam paper’ and introduce ‘the fresh pair of eyes’ philosophy to only internally audited systems through the use of independent and UKAS assessed and approved certification bodies (see Figure 14.6). The use of independent certification bodies also addresses the concern (real or imaginary) regarding the ‘proactive’ and ‘reactive’ requirements of such schemes. The concerns expressed by end-users of such schemes that the requirements of such standards and guidance may not be properly or fully put into place due to commercial pressures or otherwise have led to the growth of such schemes and the embodiment of them into commercial contracts of supply. In the main, experience has demonstrated the need for such schemes and has led to reducing the risk of food businesses getting it wrong. These schemes have reduced substantially the duplication of effort and cost involved by enabling reliance on one appraisal rather than many. Further the move away from examination or audit to certification has further improved the reliance on such schemes by moving away from the audit on the day (rather like an MOT) to reliance on an ongoing relationship between the food business and the certification body. However, there are those that suggest that existing schemes, such as the BRC Global Food Standard, which address the generality of food safety within a certification scheme could encompass those systems and procedures necessary to ensure accurate allergen information, thus avoiding the necessity for a separate certification scheme. This may be a longer term objective but to many there is a recognition that the systems and procedures necessary to implement good allergen practice and communication, together with the learning curve required by industry and enforcement, will require a level of concentration that embodiment within a general food safety standard may not achieve. This is not an unusual occurrence, as demonstrated by retailers and enforcers generating guidance and codes of practice to supplement general food safety advice in such areas as foreign bodies, pesticides, HACCP and the like. The independence of such schemes does not just revolve around the certification body. It is important that the standard owners should also recognise the needs of the end-user.
Advantages of certification Provides independent assessment. Generates consumer confidence. Reduces risk of recall, withdrawal, court action, loss of reputation. Reduces duplication of effort and cost. Moves from an audit on the day (like an MoT) to ongoing relationship between business and certification body. Fig. 14.6
Advantages of certification.
220
Management of Food Allergens
Aim of certification The reduction of risk through management of food allergens within the whole supply chain. Increased confidence in the supply chain to ensure accurate communication of risk to consumers. Companies achieving certification will present a lower risk and be better able to meet a ‘due diligence’ challenge. Fig. 14.7
Aim of certification.
The standard should represent good practices and embody all reasonable precautions, whilst the certification process should demonstrate diligence in that the processes embodied in the standard have been made to work. Further, the independence of the standard owner itself may be brought into question. If the standard owner is himself the end-user of the scheme, then the benefit of such schemes can be maximised. However, if the end-user is, for example, the consumer, and the standard owner is a food business or food body, the perceived benefits of such a scheme could be reduced if this results in reduced confidence by the end-user or the courts.
14.6.3
A certification scheme example
A good example of a certification scheme only recently launched is that produced and run by the Anaphylaxis Campaign (2007) of which I was co-author. This was produced with FSA support and was subject to extensive consultation and piloting. The resultant standard and guidance provides best practice in a form which is certifiable through and operated under EN 45011 (Product Certification) established with the specific aim to ‘Increase Trust in Information about Allergens in Food’. This scheme also benefits from the end-user being the standard owner and scheme owner as the Anaphylaxis Campaign was established to represent allergenic consumers (see Figure 14.7). The following sections attempt to summarise what is currently believed to be best practice in the control and communication of allergen information in foods. This summary has placed heavy reliance on the content of the above Anaphylaxis Campaign standard and guidance together with the FSA Guidance on Allergen Management and Consumer Information (http://www.food.gov.uk/multimedia/pdfs/maycontainguide.pdf). Readers of these sections should consult the above documents together with the other references at the end of this chapter if they are attempting to introduce allergen best practice into their food businesses or gain certification for the same. These documents contain far more detail on allergen best practice and certification requirements than has been possible to include in the following summary.
14.7 TRAINING It is obvious that with the speed of change in this area (which would appear to affect most sectors of the food industry) the introduction of new guidelines and standards establishing what is believed to be best practice will not be miraculously introduced into the food industry overnight. Substantive training will be required. This training is not just required only for the food industry but also for those enforcement officers charged with ensuring industry
Guidelines for manufacturing and certification programmes
221
compliance. As well as gaining a thorough understanding of allergen management and an understanding of how cross-contamination can occur, it is also essential that training is given so that an understanding of the effects of allergens on people who have allergic reactions is achieved. The number of trained personnel required will be dependent on the complexity of the business and the procedures put in place for the subsequent training of other personnel in their relevant allergen management responsibilities. All personnel (including temporary staff and contractors) involved in handling ingredients, equipment, utensils, packaging and products should be aware of food allergens and the consequences of their consumption by allergic or food-intolerant consumers. They should be trained in avoiding cross-contamination of foods by the major food allergens. The training should also recognise the difficulties involved with temporary, seasonal staff, contractors and visitors, and specific training should be given to those responsible for communicating allergen information directly to customers and consumers. Documentation and regular review of training will also be needed.
14.8 THE USE OF RISK ASSESSMENT There would appear to be general agreement that the use of risk assessment to assess the capabilities of the food business to produce and communicate allergy information correctly is a good idea. In particular, in order to avoid the unintentional presence of allergenic foods in products, risk assessment can be used to evaluate the likelihood of unintentional allergen cross-contamination across the supply chain, from raw materials through to the finished product. Following completion of such a risk assessment, manufacturers can then determine whether or not allergen advisory labelling is appropriate on the finished product as sold to consumers.
14.8.1
The use of HACCP principles
It would seem eminently preferable that in the assessment, management, communication and review of risk, food businesses should follow HACCP principles consistent with requirements of food hygiene law and to control food safety risk. This is because food businesses are already familiar with the principles, there is a clear structure and HACCP is compatible with quality management systems. Also, the concept of PRPs for allergens is compatible with PRPs for food hygiene. There are, however, people in the food industry who are very vocal about the use of HACCP for the control of allergens and adamant that it is not an appropriate tool. This is because unlike the microbiological arena a critical control point (CCP) in an allergen context cannot be controlled. Further, it is difficult to establish critical limits (a criterion which separates acceptability from unacceptability). It is true that allergens differ from microorganisms. They cannot be controlled by temperature. They cannot be cleaned with detergents or biocides and cleaning processes, but rely on physical removal of all residues. However, the view that HACCP principles cannot be applied is somewhat purist and would appear to attempt to deny the food industry the use of a tried and tested tool for the want of a slightly more flexible approach.
222
Management of Food Allergens
14.8.2
The use of prerequisite programmes
For a microbiological hazard, the control options are prevention, elimination or reduction of the hazard to an acceptable level. In the context of allergens, the options are really to eliminate the CCP via PRPs or changes within the supplier base or premises with warning labelling as the final option. This does not remove the requirement for checking, validating, testing. Removal of CCPs using PRPs does not take away the concept of critical limits except that these limits are likely to be procedural and independent of actual production. Obviously, in identifying any allergen hazards that need to be prevented, eliminated or reduced to acceptable levels, any risk assessment needs to consider published levels for the amount of an allergenic protein which has provoked a reaction in a sensitive patient, the prevalence of reactions to that allergenic protein in the target population, whether there are particular subgroups of the target population that might be at greater risk, the relative allergenicity of the allergenic ingredient being used, the physical nature of the allergenic ingredient and the nature of the production environment.
14.8.3
Applying the principles
If, following risk assessment, too many minor hazards have been identified, this should indicate a review of the PRPs is needed to hopefully eliminate many of these hazards. Otherwise the use of the principles of HACCP can follow the normal practices, as already used for food safety and are demonstrated below. This should include the requisite information gathering to include the target user group, customer complaints, product rejections, supplier audit reports, raw material specifications, manufacturing specifications, finished product specifications and finished product labels. You can then construct a flow diagram. This should include non-operational stages, e.g. transfers, as there may be a contamination risk. Construct a second diagram with details of factory layout, equipment design, cleaning procedures, etc., to help identify contamination risks. Ensure that the verification of the flow chart is carried out across all production shifts (weekends and nights). Ensure that the validation also takes place on the diagram documenting possible contamination issues. Monitoring, corrective actions and verification will operate as for ‘normal’ food safety, although sometimes different options may be needed for different situations. Hazards, CCPs and critical limits have been discussed previously. Target values are still needed for checking, validating and testing. Unfortunately, these target values are not as available as with microbiological standards, but there is a growing volume of literature available. An HACCP review needs to respond to emerging information about allergens as well as to the usual triggers.
14.9 MANAGEMENT 14.9.1
Allergen policy and review
Having undertaken the allergen risk assessment, the business can now define what it is capable and not capable of in terms of allergen communication. This is best achieved through an allergen policy. It is essential that this policy is communicated throughout the business and regularly reviewed. This review should include allergen-related items such as product recalls, product withdrawals, customer complaints, non-conforming products, labelling incidents,
Guidelines for manufacturing and certification programmes
223
supplier-related incidents, cross packaging, failures of change control procedures, adequacy of training, enforcement activity and new developments in relation to allergen information and management. The review should obviously be well documented.
14.9.2 Procedures and documentation Procedures shall be generated for any aspect of allergen management, and any systems and procedures which influence allergen management. These procedures should be regularly audited by the business. Again, record keeping and document control is essential. Also, organisation structures, responsibilities and authorities for all those involved in allergen management needs to be established. Procedures for investigating, correcting, verifying and documenting corrective actions should also put in place together with a system for management of complaints.
14.9.3 Cross-contamination Particular attention must be given to the routes of possible cross-contamination throughout all stages of the supply chain (which are under the control of the business) as identified by risk assessment and the effects of processing. These routes are likely to include people, transport, raw material handling, processing aids, packaging, rework, supply chain, waste from manufacturing and handling activities on the raw material suppliers sites, as well as earlier in the food chain during harvesting and transport.
14.9.4 Traceability and supplier control and approval Traceability of all allergen critical raw materials needs to be established. This may be through audits or from asking suppliers to provide the required information. Food businesses should ensure that materials are ordered against a clear specification and that they ask appropriate questions of their suppliers. A business may wish to ask its suppliers whether an ingredient contains any food allergens either as a major component, a minor component or due to food allergen cross-contamination. As a result, the business or the supplier must have positively assessed the material for the presence of allergens either by deliberate addition or by crosscontamination. A supplier approval procedure with relevant supplier monitoring should therefore be put in place. The FSA guidelines provide a checklist for this area which may prove useful.
14.9.5 Crisis management How to manage an allergen-related crisis and product recall also needs to be considered. It is important that whatever procedures are put in place should recognise that these should be trialled and the effectiveness monitored. Lists of people and organisations to contact in the event of a crisis should be kept and consideration given to the crisis occurring out of normal office hours (which is nearly always the case!).
224
Management of Food Allergens
14.10 THE ENVIRONMENT The best approach to avoid cross-contamination with allergens is to dedicate production facilities to specific allergenic products. This is not always an option particularly in small businesses. In these cases, there are a number of ways of separating the production of allergencontaining products from those that do not contain the allergen. These can include separation by the use of dedicated equipment, using physical barriers between production lines, minimising unnecessary movement of materials, cleaning of equipment between production runs and appropriate scheduling.
14.10.1
Personnel
The importance of training has been discussed earlier. Appropriate procedures on the management of allergens should also be available and/or posted wherever they need to be observed in pursuit of the company’s management policy. Personnel procedures should include clothing requirements including laundering, personal hygiene (including minimising jewellery and the presence of personal possessions in production areas), hand washing, food and drink consumption by personnel and personnel flow (personnel changing production line or site or moving equipment around the site).
14.10.2
Cleaning
Risk assessment should identify all equipment, surfaces and utensils that are subject to contamination by allergens. These should be cleaned to a demonstrably visually and physical clean standard, or equivalent validated standard to remove any potential cross-contamination residues. Within best practice guidance and standards, the use of effective, validated cleaning methods is emphasised. To investigate potential sources of contamination on production equipment and machinery and to check the effectiveness of cleaning regimes, swabbing techniques can be used to good effect. Commercial swabbing kits are available and these can be used to test swab samples for the presence of allergens. It is possible that the current design of some equipment will not lend itself to this level of cleaning and may therefore lead to the need for ‘may contain’ labelling. A responsible person will need to sign off cleaning to a visually and physically clean standard prior to start of production.
14.10.3
Raw material intake
Raw material entering a production environment needs to be carefully controlled and inspected to ensure freedom from cross-contamination due to damage, spillage or other incidents. Particular care is needed during the transfer to storage of the raw material to avoid cross-contamination in particular from transfer equipment such as hoses, chutes or conveyors. Further storage facilities should be managed and designed to avoid cross-contamination with particular attention to tanks, silos, pumps and hoses and associated pipe work. The storage of raw materials should be in accordance with a plan which should have resulted from the risk assessment of the allergen risk. The plan should be based in particular on knowledge of the allergens handled, how they are packed and the nature of the storage facility.
Guidelines for manufacturing and certification programmes
14.10.4
225
Production
The layout and design of the production area should enable effective allergen control. Consideration is required to process flows, equipment, production, scheduling and cleaning. The potential of equipment to produce dust and aerosols may affect its positioning. Rework and work in progress also needs to be managed effectively to avoid cross-contamination. Likewise, sampling for the purpose of testing needs to be undertaken in a way which avoids cross-contamination.
14.11 LABELLING AND COMMUNICATION It is very evident that allergic or food-intolerant consumers will want clear and consistent statements about what they can and cannot eat, with hopefully the same phrases used by all food businesses. Where the food products are pre-packed those ingredients that have been deliberately added (at whatever level) must be listed in the ingredients list. Any advisory labelling should be in close proximity to the ingredients list. This should be a clear statement of potential cross-contamination (i.e. ‘may contain x’ and ‘not suitable for x allergy sufferers’, where x is a specific allergen). Additionally, the presence of deliberately added allergenic ingredients, and/or advice on the possible presence of allergen cross-contamination, may be indicated by means of an allergy information/advice panel/box. These panels/boxes are not a legal requirement but, where such information is given, it is best practice to associate it clearly with the ingredients list. If such devices are employed, all allergenic foods or ingredients as defined by law and used in the food should be listed in such a box, panel or statement. If using a box headed, for example, ‘Allergy Advice’, make sure that there is a clear distinction between allergens that are deliberate ingredients and those that are possible cross-contaminants. The FSA Clear Labelling Guidelines give further advice on the requirements for legibility and intelligibility of such information and should be consulted to ensure best practice is being followed. It is also essential that procedures should be in place to verify the accuracy of such labelling and this should include making sure that consistency occurs between these panels/boxes and other labelling. Businesses should also have procedures in place to control and verify all other allergen communication to include leaflets, advertising, point of sale information, internet, verbal communication, lists of suitable products, care lines, menus, information boards and letters and emails. These procedures should be reviewed regularly. Best practice suggests that detailed explanations of the mechanisms by which contamination occurs (‘made on a line that also handles allergen X’ or ‘made in a factory that also handles allergen X’) may be confusing to consumers who do not have experience of food processing and should be avoided. For those food businesses that decide to gain certification for their efforts in allergen control and communication, it may be possible for them to communicate this achievement to allergic of food-intolerant consumers by use of a proposed logo. The use of such a logo in association with certification schemes has legal precedent in such areas as organic labelling but also in quality or environmentally friendly schemes such as the red tractor or dolphin friendly (see Figure 14.8). However, many in the food industry point out that these schemes are not in the area of food safety and therefore have concerns regarding the use of a logo in the allergen arena. They fear
226
Management of Food Allergens
Liability ‘It is important to realise that the use of a standard of this sort can only reduce the risk of communicating the wrong allergen information but cannot eliminate the risk entirely.’ ‘It is important that this concept is communicated to customers; it is not appropriate to give “cast iron guarantees” of freedom from a specific allergen based on this certification.’ Fig. 14.8
Liability.
that consumers may regard the logo as a ‘100% guarantee’ of correct allergen communication rather than an indication that best practice has been applied and demonstrated by the certified food business. As such they are concerned by the extra liability this logo will engender. This concern has been well rehearsed in the ‘free from’ or ‘suitable for’ debate together with adoption of ‘free from’ lists where the liability associated around the ‘100% guarantee’ would appear to have been accepted by many. However, providing the use of a logo of this sort is well controlled and the communication of its meaning is made very clear to allergic consumers, the potential for misunderstanding of this sort can be substantially reduced. It is obvious that many food businesses that can demonstrate their commitment to best practice may wish to inform its customers in this way in an effort to gain some commercial advantage. It is also evident that many allergic and food-intolerant consumers would find information very useful on how food businesses have demonstrated best practice. As it is intended that the use of a logo of this sort will be voluntary, food businesses can decide whether the use of the logo is consistent with its policies based on assessment of risk and reward.
14.12 THRESHOLDS A threshold could be defined as the amount of an allergen needed to elicit a reaction in a sensitive individual within a given population. The establishment of thresholds would obviously make it easier for the industry to support claims such as allergen free as they provide a limit on which to test against to support the claim (assuming there is a sufficiently accurate testing method available with the necessary detection limit). There are currently no agreed thresholds for allergens (although agreement may be closes for gluten for coeliac disease). There are currently no thresholds for allergens established in law. There is currently no government guidance.
14.12.1
The difficulties in establishing thresholds
Most of the work that is being carried out trying to establish thresholds for allergens is by collaboration between various centres around the world (Taylor et al., 2002, 2004; EFSA, 2004; Flinterman et al., 2006; US FDA, 2006; Crevel et al., 2007). Unfortunately, the prevalence of different allergies and their severity varies around the world. As an example, celery and celeriac is mainly a problem for Switzerland and Germany, whilst mustard is a big problem in France. Within any given population of people with a specific food allergy, there will be some who are more sensitive than others. There is variability between individuals and variation within an individual. Any given individual is likely to respond differently on different occasions to
Guidelines for manufacturing and certification programmes
227
a food allergen simply due to their general state of health and how challenged their immune system is at the point that they encounter the food allergen. Although work has been being carried out in trying to determine threshold levels for a number of years, the different centres were all using different methods and protocols which meant combining results was very difficult. A standard protocol was agreed following a series of round table meetings and published in 2004 (Taylor et al. 2004). In March 2006, the Center for Food Safety and Applied Nutrition, FDA and the US Department of Health and Human Services published a report entitled ‘Approaches to Establish Thresholds for Major Food Allergens and for Gluten in Food’. This 108-page report summarises the current state of scientific knowledge regarding food allergy and coeliac disease, including information on dose–response relationships for major food allergens and for gluten, respectively. The report presents the biological concepts and data needed to evaluate various approaches in order to establish thresholds that would be scientifically sound and efficacious in relation to protection of public health. Each approach has strengths and weaknesses, and the application of each is limited by the availability of appropriate data. It is likely that there will be significant scientific advances in the near future that will address a number of the limitations identified in this report.
14.12.2
Approaches to establish thresholds
The report identifies four approaches that could be used to establish thresholds: Analytical methods-based thresholds are determined by the sensitivity of the analytical method(s) used to verify compliance. Safety assessment based are where a ‘safe’ level is calculated using the no observed adverse effect level (NOAEL) from human challenge studies and an appropriate uncertainty factor (UF) applied to account for knowledge gaps. Risk assessment based examine known or potential adverse heath effects resulting from human exposure to a hazard; quantifies the levels of risk associated with specific exposures and the degree of uncertainty inherent in the risk estimate. Statutorily derived use an exemption articulated in an applicable law and extrapolates from that to other potentially similar situations.
14.12.3
The way forward in trying to establish thresholds
Of the four approaches described above, the report concluded that the quantitative risk assessment-based approach provides the strongest, most transparent scientific analysis to establish thresholds for the major food allergens. However, this approach has only recently been applied to food allergens, and the currently available data were not sufficient to meet requirements. The report recommended that further research should be initiated to develop applicable risk assessment tools and to acquire and evaluate the clinical and epidemiological data needed to support the quantitative risk assessment-based approach. Further, any thresholds established using this approach should be re-evaluated periodically as new data and tools become available. Even though further research is proceeding in this area, it is also possible that because of the complexity involved in establishing thresholds alluded to earlier, the food industry will have to continue without them available for the foreseeable future. This may mean that the
228
Management of Food Allergens
growth of ‘free from’ claims and the like may have to be curtailed unless the industry is prepared to accept the reliance on best practice to avoid the possibility of cross-contamination or at least reduce the risk to a level at which the food business feels it can make the claim. The following section attempts to summarise what is believed to be the current status of allergen testing. This summary has placed heavy reliance on the content of Annex 3 and 4 of the Anaphylaxis Campaign guidance document. Readers of these sections should consult these annexes if they are attempting to introduce allergen testing into their food businesses or use allergen testing for enforcement purposes. These annexes contain far more detail on allergen testing than has been possible to include in the following summary.
14.13 TESTING Following any risk assessment a food business should have determined whether or not testing is required to confirm allergen status. If a need for testing is identified, this should have been documented and scheduled. This schedule may include testing of raw materials, rework, work in progress, finished product, cleaning effectiveness, process verification and waste management. The type of test required should also have been determined by risk assessment.
14.13.1
Testing methods
Testing for food allergens can currently be carried out using two different technologies:
14.13.1.1 Immunoassay-based tests These are antibody-based methods designed to detect proteins present in the allergenic food. These tests, which are almost exclusively supplied in commercially-produced ‘test kit’ formats may non-specifically detect most or all of the proteins present in the allergenic food, or may be targeted towards a specific allergenic protein or proteins from the food in question. In either case it is possible to semi-quantitatively estimate either the amount of total allergenic protein, or the total amount of the allergenic food, in the analytical sample.
14.13.1.2 PCR-based tests These test are designed to detect species-specific DNA derived from the allergenic food. This technique employs the polymerase chain reaction (PCR) technique familiar from its use in criminal forensic testing. Such tests may be supplied in commercially produced test kit formats but are also available as in-house developed methods offered by some contract analytical laboratories. At present, it is not possible to estimate the amount of the allergenic food in the analytical sample, i.e. the methods are qualitative in nature only.
14.13.2
Performance criteria
In order for the food industry or those that enforce allergen labelling law to decide on the most appropriate method to utilise, they should ideally be able to compare methods easily and on a like-for-like basis. There is currently a lack of any coordinated validation programme such as exists in the area of microbiological testing, and to date only a few immunoassay approaches
Guidelines for manufacturing and certification programmes
229
have been formally evaluated, so for most of the allergens there is little third-party-derived information available. As a result, those who are attempting to decide on which method to use must rely on product literature containing the necessary performance criteria available for the requisite kit before purchasing and/or establish performance criteria derived from testing performed in the manufacturer’s own, a third party laboratory or enforcement laboratory in order to the method use and the validity of the results. These performance criteria must include the specificity of the antibody used, the calibration method, the repeatability (within assay variability), the reproducibility (between assay/operator/laboratory variability), the limit of detection, the limits of quantitation, the accuracy, the cross-reactivity, the robustness, any interference and the quality control measures used. It will be hopefully evident from these performance criteria, the advantages and limitations of the method used. If these are correctly communicated to those who interpret the results, this should avoid a lot of potential misunderstandings regarding current allergen-testing limitations. There have been occasions when the results claimed from allergen testing have not been consistent with the performance criteria of the method used. It is also evident that if thresholds were to be established for a range of allergens, many allergen-testing methods would face some considerable need for improvements in the area of formal evaluation, limits of detection, quantification and between laboratory repeatability (with coordinated validation programmes made available).
14.14 CONCLUSIONS With consumers who suffer with food allergies or food intolerances on the increase, the next decade will present some interesting challenges for the food industry. It is unlikely that new drugs, treatments or processes will arrest this growth or provide successful solutions for these conditions in the next decade or two. This is because the time involved in gaining approval for such new drugs or processes are lengthy, even if some of the current promising research can be completed quickly. As a result, the food industry is likely to be faced with a growing sector of its customers demanding accurate information about specific ingredients (or derivatives thereof) their foods contain. It would also appear obvious that the current level of ‘may contain’ (or similar) labelling will need to be reduced. The adoption of best practice and proper risk assessment will provide the food industry with the tools to achieve this reduction whilst recognising the need to reduce the risk of cross-contamination to a minimum. It is likely that the commercial reality ‘may contain’ labelling brings in restricting the sales to a growing sector of allergic or food-intolerant consumers will provide the necessary drivers for the adoption of best practices. However, should this not occur, it is likely that regulators, on the back of growing consumer concern, will impose their own solutions through legislation. One of the main objections to the implementation of best practice through certification is the cost that this may impose upon the food industry. There is also concern that proliferation of certification schemes may occur, adding further cost at a time that this can be ill afforded by many food businesses. There are some in the food industry who believe that best practice could be achieved by allowing the industry to self regulate using the substantive guidance that has been established in the area of allergen control and communication. Enforcement authorities would then oversee whether legal compliance is achieved or not.
230
Management of Food Allergens
However, this reasoning has a number of concerns:
r
r
r
Historically, when left to their own devices, food businesses have felt the need to selfaudit, both within the business and between businesses to ensure that best practice has been adopted properly by its personnel and its suppliers. This substantive activity does not lend itself to demonstrating the food industry’s own confidence that individual food businesses will adopt best practice. As a result, retailers have audited its suppliers and its suppliers have audited their suppliers. In particular, the between-business auditing was the subject of tremendous duplication and cost which lead to the introduction of independent third party certification schemes in the first place. These schemes have demonstrated substantive savings to the industry in between-business auditing (as supported by those who can remember the substantive cost and duplication that occurred before them). The fear of proliferation of certification schemes may seem real. It is accepted that from time to time the principles established for best practice in the general food safety arena (which are currently the subject of comprehensive certification schemes) may require substantive addition and modification to address new and emerging issues. New and emerging issues of this sort, by their very nature, can add substantive cost to the industry as they struggle to establish the systems and procedures to control them. It is right to believe that the use of certification schemes to address these new and emerging issues require a concentration on them which is currently missing in general food safety schemes. However, if some concentration is needed, is this the most cost-effective way of delivering the solution? Is attempting to embody good allergen control and communication into existing general food safety certification schemes currently a good idea? Or will the level of concentration currently needed in this area bias and potentially adversely affect these current schemes? The requirement from the food industry for certification schemes to be cost effective together with the competitive nature of these schemes will ultimately result in resolution of this issue. As new guidance and standards embodying best practice become part and parcel of general food safety the need for special attention to be given to allergen control and communication will reduce. Are the food industry’s current efforts to adopt best practice in the area of allergen control and communication working? There is growing evidence from those who suffer from food allergies or intolerances, those organisations that represent them and regulators that they are becoming increasingly unhappy with the food industries efforts in this area. With the growing number of high-profile product recalls and the high level of ‘may contain’ labelling, there would appear to be a growing level of consumers who are showing a lack of trust in information being communicated to them by the food industry. This growing lack of trust is unlikely to be addressed by industry proclamations that allergenic consumers can trust them to adopt the necessary best practices to address this situation. A degree of independent assessment embodied within the certification process may be the solution, particularly if the owners of the schemes are the end-users. It is beneficial for communication to the allergic and food-intolerant consumer that businesses that have been through a certification process bear a logo on their products for ease of recognition. It is, however, essential that this communication should be about these businesses having demonstrated best practice and not about the total elimination of risk.
Some food businesses would appear to feel comfortable to label their products as ‘Free from’ specific allergens. This is currently in the absence of any agreement on thresholds. It is interesting to note that some in the food industry would appear to be prepared to take on
Guidelines for manufacturing and certification programmes
231
the extra liability that these types of claims embody, without the added precaution of being able to test to an agreed and recognised threshold. Even with dedicated sites, the fear of cross-contamination from raw materials, personnel and the environment means the ability to substantiate these claims, which some maintain indicate a 100% guarantee, would appear problematic. It is also possible that the European Commission may attempt to restrict such claims through regulation. Testing for allergens is currently in a state of evolution. The availability of methods is in some cases limited and some have expressed concerns, particularly about the reproducibility of some of the existing methods between laboratories. For those who seek them out, the performance criteria of these methods provide information on the methods strengths and weaknesses and the applicability of the method to deliver to the food businesses requirements. Current methods may not always deliver the results expected or required by the business, and it is important that food businesses gain an understanding of these limitations particularly when using such allergen testing methods to validate or verify their procedures and systems. There is much debate about the liability issues surrounding allergen communication. When viewed against the existing liabilities already involved in the general food safety arena, these concerns would appear to be over emphasised. To those who challenge the food industry position on this issue, the industry would appear to accept the liabilities involved in ‘free from’ or ‘suitable for’ claims, but also appear to be so concerned about liability there is a persistence in the use of ‘may contain’ labelling. They suspect that the industry is using this issue as an excuse to avoid the introduction of best practice. With the now considerable guidance on best practice in this area together with the availability of good quality certification schemes, the adoption of these is no longer limited by availability. It is for the food industry to decide whether their implementation becomes a commercial necessity.
Acknowledgements The author would like to thank David Reading of the Anaphylaxis Campaign, Dorothy Cullinane of Cullinane Associates Ltd. for their support and help and Phil Goodwin of Haven Limited for his help and support in the allergen testing area.
REFERENCES Anaphylaxis Campaign (2007) Standard and Guidelines to Increase Trust in Information about Allergens in Food (Version 1, September 2007). BRC (British Retail Consortium) Guidance – Allergen Q&A – Directive 2005/26/EC – Handling of ‘Nuts’. Available at http://www.brc.org.uk/policycontent04.asp?iCat=46&iSubCat=330&sPolicy= Food&sSubPolicy=Allergens, accessed 21 April 2009. Crevel R.W.R., Briggs D., Hefle S. et al. (2007) Hazard characterisation in food allergen risk assessment: the application of statistical approaches and the use of clinical data. Food and Chemical Toxicology, 45, 69–701. Flinterman A.E., Pasmans S.G., Hoekstra M.O. et al. (2006) Determination of no-observed-adverse-effect levels and eliciting doses in a representative group of peanut-sensitized children. The Journal of Allergy and Clinical Immunology, 117(2), 448–454. FSA (2008) Proposal for a new regulation on the provision of food information to consumers. Available at http://www.food.gov.uk/consultations/ukwideconsults/2008/infoprovision, accessed 21 April 2009.
232
Management of Food Allergens
EFSA (European Food Safety Authority) (2004) Opinion of the Scientific Panel on Dietetic Products, Nutrition and Allergies on a request from the Commission relating to the evaluation of allergenic food for labelling purposes. The EFSA Journal, 32, 1–197. Available at http://www.efsa.europa.eu/en/science/ nda/nda opinions/food allergy/341.html, accessed 21 April 2009. Taylor S.L., Hefle S.L., Bindslev-Jensen C. et al. (2002) Factors affecting the determinations of threshold doses for allergenic foods: how much is too much? The Journal of Allergy and Clinical Immunology, 109(1), 24–30. Taylor S.L., Hefle S.L., Bindslev-Jensen C. et al. (2004) A consensus protocol for the determination of the threshold doses for allergenic foods: how much is too much? Clinical and Experimental Allergy, 34, 689–695. US FDA (2006) Approaches to Establishing Thresholds for Major Food Allergens and for Gluten in Food. Available at http://www.cfsan.fda.gov/∼dms/alrgn2.html, accessed 21 April 2009.
15
Risk communication – a manufacturer’s perspective
Clive Beecham
15.1 BACKGROUND The Kinnerton story is an interesting one, because if nothing else, it demonstrates that a member of the public can indeed achieve significant measures of change in the way manufacturers or retailers conduct their affairs. I imagine that many of the people reading this book have had to experience the nightmare of going shopping for a loved one who has just been diagnosed as ‘allergic’ (in our case relating to nuts) and is subsequently faced with a bewildering array of negative, confusing or absolutely impossible to find ‘warnings’ on the back of food packs, that only weeks beforehand, would never have caused one to raise an eyebrow. Such circumstances inevitably result in my perceived profile of a typical ‘nut mum’ (because it usually is a mum) being a strident individual who will definitely not take ‘no’ for an answer! So it was that in 1998, a certain Mrs Tracy Cashman picked up the phone to Kinnerton Confectionery and gave me one of the more bruising verbal tongue lashings I have received in my 30 years in the confectionery industry – and that includes my supermarket buyers! Mrs Cashman had a very broad Scottish accent, so broad in fact that I was struggling to grasp her every word; notwithstanding this, the point was made to great effect. Her toddler of a daughter, who was mad on the Teletubbies and recently diagnosed with a nut allergy, had picked up one of our ‘Tinky Winky’ chocolate bars from the shop shelves and proceeded to eat the product before her mother was able to give it her seal of approval. On the back of the Kinnerton product, in bold letters alongside the ingredients, we proclaimed: ‘This Product may contain traces of Nuts’. An incandescent Mrs Cashman was appalled at this. Firstly, why would straightforward milk chocolate have anything to do with nuts if it was not in the ingredients? Wasn’t it just a case of a manufacturer ‘covering his backside’ with a blanket statement designed to allow manufacturing to be ‘responsibility free’? Secondly, why put a statement which, after all, is telling certain consumers that: ‘This product could kill you’, on the back of the pack and thirdly, ‘What am I going to do about it?’ Well, this was not the first time I had had to deal with these arguments and accordingly went into my well-rehearsed routine of:
r
We are confectioners and the tools of our trade happen to include nuts. Nuts for pralines, nuts for marzipan, nuts for walnut whips and many more. As such, our machinery was contaminated by nuts, be they in the atmosphere, on people’s hands or on certain of our production lines that handled both nut and non-nut products. Yes, we did clean down; yes, we did schedule nut production for the end of the week and (after a weekend of
234
r
r
Management of Food Allergens
cleaning), ‘no nut’ products for the start of the week; and yes, we did try to promote good manufacturing practices (GMP) in our factory. But no, we could not make a statement that there was no danger of nut cross-contamination . . . because there definitely was! If we were to put the ‘may contain’ notice on the front of the pack, why stop at nuts? What about the other ingredients which on allergen or even religious grounds carry equal weight? Add in the legislation that needs to go on back of pack, multiply that by a number of languages, and in short order, the back of the pack becomes the front of the pack and the graphics have no-where to go. So, if the back becomes the front and the front becomes the back, how does that leave the consumer any better off? So, what am I going to do about it? Well, without spending shed loads of money, there was nothing further Kinnerton could do under the circumstances.
And then Mrs Cashman tried a different tack. In so many words she said: ‘How can you justify the manufacture of products, specifically aimed at children, in an environment that is potentially lethal?’ There was no off-pat answer for that! Caught out, we agreed to disagree and I was left with a sleepless night grappling with the moral question Mrs Cashman had posed to me. Come morning, I had reached a decision – she was right and we were wrong, simple as that. Kinnerton would endeavour to separate its production facilities into what we then called ‘nut’ and ‘nut free’ and so began a journey in 1999 that cost a million pounds to effect structurally, hundreds of thousands of pounds each year to maintain, considerable angst in verification and allowing oneself to be ‘put up on a pedestal’, where, as hostages to fortune, there is only one way to go.
15.2
THE PROCESS OF GOING ‘NUT FREE’
It’s easy to talk a good game, but achieving the separation of our facilities was a huge task. Being a private and relatively small company, we potentially lacked the technical resource to undertake something that had (to the best of our knowledge) never before been achieved anywhere in the food industry, namely keeping nuts on site but achieving effective separation. However, being private also had its advantages, because in pure commercial terms, the million pounds we spent made absolutely no sense at all and constituted approximately half of our annual profits. This was to be a moral investment because Kinnerton Confectionery, as a manufacturer of children’s chocolate products, believed that it was the right thing to do and there was no need to seek justification from public shareholders . . . as the Nike statement goes – Just Do It! In the planning stage, we took advice from several of our major retail customers, calling on their technical departments to help risk assess our proposed factory layout. The Anaphylaxis Campaign and David Reading (its founder) were particularly helpful and encouraging, though when it came down to the nitty-gritty of the detail, we were made painfully aware that there was no Bible, no rule book for what we were about to undertake – good old common sense and mentally ‘walking’ the various risks through the factory was the only way to go. We were on our own and would have to take the consequences of getting it wrong. It began with the decision to abandon conventional GMP as in my opinion, however good they were, there would always be the risk that systems are only as good as the people who operate and live by them. There were some companies in the United Kingdom that had superb GMP, really first class, but still felt it necessary to make use of the ‘may contain’
Risk communication – a manufacturer’s perspective
Fig. 15.1
235
Communicating between the dividing wall in the factory. (Reprinted with permission.)
nuts statement. What, we thought, is the good of having all that wonderful background work, if you are still advising the people whom you are allegedly trying to help, not to eat the product? So, we decided that we would build a wall down the centre of our chocolate factory in Fakenham, Norfolk, and based on the simple expedient that not even the most mistaken individual could bring nuts through a solid wall with no openings, we had the basis of what we would call a ‘nut-free zone’. Kinnerton, at the time had seven production lines, some of which never needed to use nuts (Advent Calendars, Easter Eggs), some always used them (Chocolate Covered Marzipans, Pralines and Petit Fours) and of course the problem areas were the lines that sometimes did use nuts – these of course would be the most complicated lines in the factory to move (Figure 15.1). But move them we did. Every single production line in the factory bar one, amounting to several millions of pounds worth of equipment, was relocated on its respective ‘side’ – ‘nut’ or ‘nut free’ and a wall was built right the way through the length of the factory, encompassing the warehouse as well. On top of this, a modern chocolate factory has huge amounts of services in the form of heating, water and electricity cabling and feeds as well as chocolate pipe work running off vast 30 tonne chocolate tanks. All these needed to be relocated and re-wired/plumbed during a 6-week period after Easter, when traditionally chocolate sales dip somewhat (Figure 15.2). There were various authorities that we also engaged with to present our plans and to seek their blessing, not least, Environmental Health and Trading Standards, but to my mind, the most important individual body we had to get on board was our insurance company. If we did not get them to buy in to what we were doing and continue to give us public liability cover, then despite everyone’s good intentions, we could not progress with the project. After all, if something went dreadfully wrong, aside from the tragic consequences of a mistake, I may be in the dock, but the insurers would have to bear the financial cost of attempting to put it right, which could be considerable and as such give them every reason to either dramatically raise their premiums or simply say ‘no’. To this end, we thought long and hard about how we would label our product, eventually coming up with the legend: ‘Manufactured in a Nut Free Zone’ (Figure 15.3). If you think about this form of wording, it tries to limit our potential liability somewhat, as it endeavoured to take on board the fact that there were nuts on site (on the ‘nut side’) and therefore rather than simply using the words ‘nut free’, we limited our statement to the area we believed we could control, namely our factory floor. To be honest, I was completely overwhelmed by the stance of our insurers, who couldn’t have been more
236
Management of Food Allergens
Fig. 15.2
All the machinery, pipework and cabling had to be moved. (Reprinted with permission.)
helpful. They accepted the wording and conducted a risk analysis, after which Ron Piper of the Special Risks Department wrote saying: ‘. . . they had better get it right or they are really on the hook. It is good to see someone with the courage of their convictions pioneering what should be standard practice in my opinion.’ Incidentally, there was no increase in our premium rating either. Some tough sales decisions now had to be taken. For instance we have a small bakery on site which produced essentially not only nut-free biscuits, but also almond-based ratafias, which were used in trifles. We had £500,000 worth of retail sales of this product in 1998, but in order to keep them, we would have had to divide the bakery and duplicate our depositors and ovens at considerable cost and with no proper payback. Ratafias were sadly dropped from our portfolio and we lost that business. As I say, sometimes it is better to be a private company!
Fig. 15.3
Old nut free logo. (Reprinted with permission.)
Risk communication – a manufacturer’s perspective
237
Fig. 15.4 Clearly defined ‘step over’ into Nut Zone changing area with stringent changing procedure to avoid cross-contamination. (Reprinted with permission.)
15.3
THE FACTORY CHANGES
Having outlined our plans to effect the physical separation of our plant, we now had to make the whole process work in terms of people and things. The business had to operate with what in effect became two factories operating on one site. Detailed below is a list of some of the more significant changes we had to make to ensure we could stand by our ‘nut-free zone’ statement.
15.3.1
r
r r r r r
Duplication of facilities
Changing rooms. New changing rooms were built to accommodate a ‘nut’ and ‘nut free’ changing regime (Figure 15.4). This involved captive footwear for the nut side as well as a new colour coded clothing regime – essentially yellow and red for nuts and white or blue for nut free (Plate 15.5). Of course, people needed time to change out of their clothes and shoes to go from one side to another, approximately 10 min for each movement. Whereas in the past you would simply walk from one machine to another, now you had to accommodate all the rules relating to a zone change. With up to 700 people now working in our factory, this costs a fortune in lost labour time and also provides a disincentive to management to move freely from one side to another. First aid room Development kitchen Engineering shop Toilet facilities Meeting room facilities for the factory floor
238
Management of Food Allergens
15.3.2 Duplication of machinery and capital items
r r r
r r
r
Mould and tray wash machine. Having only recently invested a considerable amount of money on a new state of the art mould wash, we had to re-invest again. Tempering machines, re-work and re-melt kettles, mixing bowls etc. These are items that we use in the process of getting chocolate and centres to the production lines and had to be allocated to one side or the other, and therefore duplicated. Of course, strict control of the re-work and re-melt generated on either side had to go hand in hand with this. Ten thousand plastic baskets and trays, greaseproof paper, chocolate moulds, carriers. If we sent ‘clean’ nut-free chocolate into the ‘dirty’ nut-contaminated side, in a tray or a vessel of any description, the container could never come back to the ‘nut-free’ side, even if it had never approached any nuts. Quite simply, the fact that a box or a tray entered at all into the ‘nut-contaminated side’ condemned it to never making a return journey, as it would now be ‘contaminated’. Whereas in the past we could move such items freely around our factory, now we could not, adding considerable cost to the way a factory is run. Engineering tools that are colour coded and allocated to each side. Flow-wrapping machinery, box-erecting machinery, case sealers, sellotape holders, pallet trucks and countless other pieces of equipment all the way down to a humble ballpoint pen. If it has been used on the ‘nut side’, then it cannot (without an intensive deep clean followed by positive release after testing) go back. Without really realising it, we were undertaking an inordinate amount of hidden costs that we clearly never anticipated. New air-conditioning systems. Our challenge was to create a system that passed clean air from the ‘nut-free’ side to the ‘dirty air’ nut-contaminated side. This was achieved by regulating the pressure between the two sides so that the air flowed accordingly. In a 50,000 square feet unit that was never designed for this purpose, this was no mean feat.
So there you are, a million pounds spent quite easily, and accomplished in 6 weeks by a Kinnerton team that was almost missionary in its zeal to get it right. To be successful, however, necessitated the recognition that there were items you could not simply throw money at and install.
15.4
PEOPLE DISCIPLINES
This was/is always going to be the toughest issue the factory was going to contend with. Changing the culture of the way Kinnerton staff think and act, moving away from natural instincts, towards considered movement and thought patterns in our daily working lives. You can buy new machinery, build a wall, create the systems but how can you account for human frailties and the fact that people can and will make mistakes? More on this later. The areas detailed below have evolved over the 9 years since embarking on this project. There has been a constant development and learning process. The principles of the systems were based on high/low risk and this is borne out by the changing regimes in place when entering our nut facilities. We made the decision in 1999 that our one of our biggest risks was from within and we must keep nuts ‘in’ the nut side and ‘out’ of nut free. This meant adopting a high/low-risk style of changing regime when entering our nut side.
Risk communication – a manufacturer’s perspective
15.4.1
239
Training
People really are the key when changing systems, protocols and ways of working. You can have the greatest procedures in the world, but unless you train out these procedures they are worthless. Nut segregation protocols are an integrated part of running our business and there are numerous specific work instructions including people movement, product movement, emergency spillage procedures that get trained out to new staff and re-briefed to existing staff on a continuous review basis. In addition, detailed nut allergy training is carried out to bring employees up to speed on the latest thinking in this area. It also serves as a useful reminder of exactly why we operate a nut-segregated facility and the implications if it goes wrong. It is also important not to lose sight of the other allergens handled on site and the level of awareness. In addition to our nut-segregated products, we offer a dairy, gluten, egg and nut-free range of products. Once again, people are key in this process and when the products are being manufactured our staff receive a daily briefing.
15.4.2
Induction
Training really begins from the moment a new employee joins the business. Details of our nut protocol, reasons for it and the practicalities of operating to it are briefed out to staff during the company induction. Anyone joining the business is required to attend and participate in induction training. Long-term contractors to the business are also taken through induction training. This is extraordinarily important for those whose work involves activity in both nut and nut-free factories. Training is essential not only to those working for us but also for those involved in contract services.
15.4.3 Nut Monday It is important for us to maintain awareness of our nut status within the factory and across all our staff. We classify all non-nut factory areas of the business as ‘nut-free zones’, this includes our main canteen and office areas. To assist in maintaining this awareness, we introduced ‘Nut Monday’ monthly, which is essentially a day on which we carry out specific audits, locker checks, staff interviews and other audits to confirm whether the nut protocols are being followed and that there is an understanding of why it is so important.
15.4.4
Visible notices and barriers
Visibility of our segregated facilities is an essential part of informing and reminding staff of their responsibilities. As soon as you enter the Kinnerton reception area, the nut status of our factory is clear with a video constantly explaining what Kinnerton ‘nut-free zone’ represents. As you walk through the business, signs indicate nut and nut-free areas. In today’s multicultural society, consideration also has to be given to signage in foreign languages and the use of pictorial representation to really get the message across.
15.4.5 Changing regimes As detailed at the start of this section, our focus in 1999 right through to date has been to prevent the transfer of nut and nut traces from our ‘nut product manufacturing unit’ to our
240
Management of Food Allergens
‘nut-free’ manufacturing unit. It is generally a given nowadays that factories operating to a good standard of hygiene and housekeeping remove their factory overalls prior to using canteen facilities. This was not the case in 1999 within the confectionery and other perceived low-risk industries. Ironically, it was from a ‘nut control’ perspective that this change was instigated within Kinnerton. Clearly, how could you have a person who has been sorting walnuts for 2 h eating their lunch, in factory overalls, next to a person from the ‘nut-free zone’? Eliminating this risk was just the start. Our refurbished changing rooms included a step over similar to a high-risk factory along with hand-washing requirements when moving out of the nut area. Dedicated yellow and red overalls, mop caps and factory footwear were employed in the nut side, clearly different from those that are used in the ‘nut-free zone’. Our thinking on this has changed over the years and has evolved to a double step over now in place. This includes removal of shoes, step over into a central hand-washing area and further step over into the main nut changing area where dedicated clothing and footwear are donned. In reverse, this includes a further hand-washing step in an attempt to remove any further nut traces from an individual’s hands.
15.4.6
Canteen procedures and bringing food to work
I have already discussed the changing regimes put in place to reduce the nut crosscontamination risk, but what controls are required on personnel bringing nut products to work? The rules employed in this field have again evolved over time. Originally, we clearly saw the risk of cross-contamination between nut and non-nut workers and, as such, nuts are forbidden in the canteen, offices and meeting rooms.
15.4.7
Visitor and contractor controls
Inevitably, the controls implemented for personnel working on site had to be supplemented by controls on those visitors and contractors entering the site. Everyone visiting Kinnerton is subjected to questions over their nut consumption/handling prior to visiting site. If you take something as simple as our pest control contractor, pre-1999 he would have had free access to all parts of the factory. Now, in order to maintain full segregation, separate kits are kept on each side of the factory to prevent any cross-contamination. Any contractor entering the site is questioned over their nut status and that of their equipment. The boundaries of this topic are never ending. Almost every week a new scenario presents itself: the challenge of bringing new items on to site; where do you clean equipment? can you bring certain items into the factory? what happens when a children’s TV program want to film in the factory? These are all changes that our business faces on a regular basis.
15.4.8
Planning and creation of new product
Our business was and is built on flexibility, so the challenges faced when developing new products increase every season. Product testing of nut ingredients and the development of nut products is carried out in the ‘nut side’ in specially constructed ‘nut test’ kitchens. In 1999, it was impossible to predict how the confectionery market would develop in the coming years. The act of segregating the business placed an inevitable cost and complexity within
Risk communication – a manufacturer’s perspective
241
the business which determines many of our manufacturing path decisions from a people and product perspective, so planning is yet another key function within the business that has to have an in-depth understanding of the nut allergy issue and what this means for Kinnerton.
15.5 VERIFICATION 15.5.1 Supplier management A problem we have faced since the beginning of this process is the level of awareness within our supply base. One of the key areas of focus for us in 1999 was a review of our suppliers to confirm whether they were ‘nut free’. At that time, awareness of nut allergy was limited and focused around peanuts. Suppliers would either naively class themselves as ‘nut free’ and fall short of the actual standards employed by Kinnerton or take the risk-averse route and claim ‘may contain traces of nut’ for every product supplied. Our challenges once again were far reaching. How far back do you investigate the supply chain? Do you audit every supplier and the supplier’s supplier? Which suppliers can we actually use for our ‘nut-free’ business? Is the Kinnerton understanding of ‘nut free’ the same as the rest of the industry? There were of course casualties along the way and harsh decisions to be made with our supply base. Unlike other businesses, our supplier assurance program is highly influenced by the nut status and perceived risk of that supplier. Other elements of supplier assurance follow this. Clearly, if a supplier cannot guarantee the nut status of their materials, they cannot form part of the Kinnerton ‘nut free’ supply base regardless of any other accreditation or certification. As with most areas, this is an endless task and we still face the ‘may contain’ challenge with our suppliers. Within the current climate of product withdrawal and recall, this area is getting harder to manage. The different approaches employed by manufacturers is interesting and of course challenging. As an example, there are some businesses that operate on a risk assessment and probability basis, even in some instances a ‘dilution effect’ depending on the type of product and process. It has been difficult to reconcile this approach and its acceptability to our business. There are a number of examples where we have worked directly with suppliers, recommending, implementing and reviewing practices and procedures in order to get to a satisfactory supply position. We have learnt over the years that investment in our supply base is the only way we can deliver innovative products whilst maintaining our nut status. Overall, it is any area we have to work harder at than comparable businesses.
15.5.2
Testing
At the start of our journey, peanut protein tests were commercially available but the interest from the rest of the food industry was pretty limited. Kinnerton of course didn’t use peanuts. It was really from this point onwards that awareness and interest in allergen testing, and specifically nut testing, grew to where we are today. At the time, peanut was seen as the major nut allergen with less focus on other tree nuts. The industry was also unclear as to the relevance of quantification. Presence/absence was the starting point, although semi quantitative testing quickly followed. Our controls and verification were built on those areas described and not around the very limited availability of testing. Science had advanced to the point that in 2006 a brand new, more formalised testing program was introduced into our business. Our retail customers had started to include this aspect within their codes of
242
Management of Food Allergens
practice and the availability of appropriate tests was increasing. Test methods for hazelnut and almonds started to become available in this year and as these nuts were much more prevalent within our factory than (absent) peanuts, testing became much more relevant to us. It is, however, important to remember that there was still no protein test available for walnuts, by far the most widely used nut in our business. DNA testing for allergens including nuts was an area that had started to develop and was utilised on a very occasional basis by Kinnerton. Following a review of nut segregation procedures within the factory and a review of analytical methods available to the business, a new testing regime was introduced in 2006 that took into account new and existing technology; this included daily testing for walnut DNA in product and the environment. In early 2007, on-site test kits for hazelnut and almond were introduced which increased our ability to survey our products, raw materials and environment for these nut proteins. We now test all incoming raw materials using these test kits to supplement our supplier allergy information along with finished products and the environment. The most important change for us in testing has been the development of a walnut protein test as opposed to the previously used DNA test, giving Kinnerton the ability to really verify the controls in place within the factory, focusing on the highest risk nut allergen within the business. Kinnerton began using this protein test as soon as it became available towards the end of 2007, and although not widely available commercially, it has been a key factor in our risk assessments and subsequent changes to our nut statements. Our current position on testing is based around quality assurance principles. It is impossible to ‘verify or guarantee’ the nut status of our products through testing both statistically and from a best practice perspective. Our policy is to maintain the controls within our business as described and to verify that they are working through a surveillance testing program. Risk assessment really drives the type and frequency of testing which is a combination of raw materials, environment, machinery and product testing and utilises all relevant, available analytical test methods. The controls in place within Kinnerton confectionery are better now than they have ever been. The level of understanding and awareness within both the business and our customers and consumers has grown in the last 9 years along with the analytical world. Ironically, the more one tests, the more one learns and the more stringent those tests become, the more one finds. So, whereas years ago one might have passed with a test that had a limit of detection of, say, 5 ppm, you may now pass only at 1 ppm and fail at 2 ppm. You are cleaner than you ever were, but in law, you are possibly operating at a level of failure.
15.5.3
Internal audit
Segregation protocols cannot be effective without an audit process. The impact of segregating a factory was felt in this area. Not only did the site effectively become two factories, a whole new set of procedures was introduced that required audit. These procedures have evolved over the years and still form part of our internal audit process – a combination of formal audits against the nut protocol and informal challenges. Everyone at Kinnerton has a responsibility and an opportunity to challenge and audit systems. Understanding the implications of bringing a nut product on to site helps to maintain this. It is also very easy to have a restricted view of systems and processes. Many people at Kinnerton have seen the nut segregation process develop since 1999 and a vital part of ensuring continued best practice is from an external perspective. Significant changes and improvements have been made within
Risk communication – a manufacturer’s perspective
243
our business from external auditors with experience of other industries and issues. Learning from the audit process both internal and external has heavily influenced our practices today.
15.6
RETAILER REACTION
We had stepped up to the plate, so how did the retailers react to our efforts? Without question, the move on Kinnerton’s part was welcomed, even lauded by the majority of retailers, but then how could it have been received in any other way? The acid test for us was how they would treat it when it came to buying product. Kinnerton specialises in the production of character merchandised chocolates, largely for seasons such as Easter and Christmas, as well as producing chocolate bars and lollipops on an all-year-round basis, using world famous characters such as Thomas the Tank Engine, Barbie or The Simpsons. When submitting our ‘nut-free’ range for a possible listing, the order in which nearly all the retailers would rank acceptability of product (assuming product quality as a given) would be: 1. Price 2. The popularity of the character you are using and then, and only then 3. ‘Nut free’ would tip business over to us As anticipated, we did not think that ‘nut free’ would necessarily be a driver for our business, but from a defensive point of view, we felt that it was the correct thing to do, and should unfortunately anything ever go wrong with a child and chocolate (not necessarily ours) then at least Kinnerton had done the right thing. However, when it came to making private label items for retailers, because factories need to be signed off by their own technical teams (who, due to their training and instincts, positively welcome a ‘nut free’ approach) then the story could be different, and often was, and still is. The commercial function that is only normally interested in numbers can be swayed at this juncture, but it continues to amaze me how some of Britain’s largest retailers would sometimes award complete contracts for say all their hundreds of thousands of advent calendars on the basis of price alone. What we could never understand is that if approximately 5% of your customer base is nut allergic, then why can’t at the very least 5% of your product (if it is available) cater for those customers? When speaking to ‘nut mums’ (which is often), it is clear that all they ask for is that at the very least some, not all, products be ‘nut free’ – give them a choice and do not exclude them, and certainly not for what can amount to less than a penny an advent calendar or Easter egg on the buying price. The retailers’ approach to labelling varied, but broadly speaking, they all removed the nut warning from Kinnerton private label manufactured products, with many adopting the positive: ‘made in a nut-free zone’ legend, or words that carried equal weight, in a similar fashion to Kinnerton’s branded character products. The problems of maintaining the ‘nut-free zone’ were many. The biggest issues always related to making sure that the supplies of the literally hundreds of ingredients that Kinnerton uses were indeed ‘nut free’ and also the control of people movements – both those that worked in the environs of our factory and people who ‘visited’ and ensuring that they too were ‘clean’.
244
15.7
Management of Food Allergens
WHAT IS NUT FREE? – THE PROBLEM OF EVOLVING SCIENCE
Definitions of the words ‘nut free’ possibly mean different things to different people. For the authorities, ‘free from’ is defined by the limits of the level of detection that testing results give at a point in time. As the science gets more efficient at detecting offending proteins, so the ‘acceptable’ parts per million (ppm) level of protein is reduced. So, if science can take us from 5 to 2 ppm then that is a good thing all round, one would have thought? A threshold approach is another way of defining the issue. For instance, in law, ‘gluten free’ complies with a threshold level of detection of 20 ppm that is deemed safe – so it is not ‘gluten free’ in the absolute sense of the word, but legally and presumably scientifically correct for the purposes of safety. Having a threshold though is most helpful because the issue becomes more defined and less subject to: ‘nothing other than 100% perfection will suffice’, which is where we are on most allergens. The threshold is still authenticated by a regime of testing. In the case of manufacturers who make no specific claim, but who would characterise their plants as ‘nut free’ because they had either no nut ingredients on site or deemed that their plant was sufficiently segregated, in many cases almost certainly do not conduct meaningful levels of testing to uphold their assertion, probably for two reasons: (a) It is inordinately expensive to do the job properly. (b) They are probably ‘scared’ about what they will find. Either way they might believe that they were ‘nut free’ but Kinnerton has on several occasions had to not buy from some very large suppliers whose code of practices in this instance fell short of what we would declare as ‘nut free’. When we started in 1999, there was precious little in the way of testing for nuts, just a protein test for peanuts, when ironically peanut was the one nut we did not use on site! However, we did have several other nuts and as such we had to rely on common sense and educated judgement to enforce our regime. Over the years, the science has improved, so there is now a degree of certainty about whether one can characterise your products as ‘nut free’. When I speak to members of the public and ask what they mean by the words ‘nut free’, the words ‘guaranteed no nuts’ follow on very quickly indeed. If you receive a ‘guarantee’ in almost every other walk of industry, it never really means that; it means we will replace what has been supplied, if it is bad or unfit for purpose. A child that has had an anaphylactic shock or worse is not receptive to that standard industrial definition of the word guarantee, but similarly it is impossible for Kinnerton or anyone to ‘guarantee’ product in this sense. When we started, ‘nut free’ really meant doing everything humanly possible to eliminate nuts from the zone but we could not define any potential contamination in terms of ppm present, because the science didn’t really exist. As such, we felt very comfortable with our ‘manufactured in a nut free zone’ statement, because to the best of our knowledge we believed in our robustness and had spent £1,000,000 achieving it. Today, it is now possible to test for nine nuts (walnuts, almonds, hazelnuts, peanuts, cashews, pistachios, Brazils, Macadamias and pecans) to forensic levels utilising the DNA process. There are also ELISA tests for a few nuts that will detect nut protein to levels of 1 ppm, and before long, that figure will again be improved on, so that 1 part per two million or three million are detectable. And where does that leave us, as manufacturers, making some kind of ‘nut free’ statement? Terrified is the short answer. It is difficult for me to try to get you to visualise 1 ppm, but
Risk communication – a manufacturer’s perspective
245
look at this way: Think of Wembley Stadium filled with over 90,000 people. Now extend that to 11 Wembley stadiums and try to find one person in that crowd of a million people – that is what we are testing to here. For the overwhelming majority of people with nut allergy, 1 ppm is reputed to represent no anaphylactic problem at all; however, if a product were found to have 1 ppm in it and the packaging carried a ‘nut free’ statement, then it is currently likely that the manufacturer could be facing the possibility of a complete product withdrawal, which in the case of Kinnerton at say Easter time would bring the company to its knees. And why? Because the words ‘nut free’ in law mean just that, nut free, or to the limits of detection, currently 1 ppm. So 2 ppm, though almost infinitesimal, is not ‘nut free’. Quite simply, the increasing vigilance of science is asking manufacturers to produce in ‘beyond Operating Theatre’ realms of cleanliness and it is going to become untenable. Speaking to one mum recently, she volunteered: ‘Goodness me, if I opened up the window to my son’s bedroom, one part per million could easily fly in, but do I worry about that? I could go to the supermarket with my son and sit on a bus with someone eating peanuts in the seat in front of me, or the bakery counter of the supermarket will have a nut warning, because they are cooking in-house with nut ingredients; not to mention that foods I want to buy are positioned on the shelves right next to genuine nut products and could have been put there by a supermarket operative who had just had a peanut butter sandwich for lunch, all of which represents a lot more risk to me than the unlikely possibility of Kinnerton having 1, 2 or 3 ppm in one of its products!’ Yet Kinnerton could be off-shelf and out of business for 1 ppm. Without a threshold being issued, only zero ppm satisfies the law as it stands and as Professor Steve Taylor, one of the world’s leading experts on nut allergies, says, ‘Zero is a very small number indeed!’ Many retailers are challenging manufacturers on their allergen controls. It is no longer acceptable to consider the allergens handled within the factory and place a ‘may contain’ warning on pack for these without a structured risk assessment. There is, in some cases, a verbal desire to respond to consumer needs and offer solutions to allergy sufferers, to provide consumers with accurate information they can trust in an attempt to reduce risk taking. As a result, retail codes of practice are becoming more and more prescriptive. We appear now to be stuck in a cycle of product recalls, allergy alerts and warnings which are in danger of fixing the industry firmly in the world of ‘may contain’. This, coupled with the developments in testing and the ability to detect minute levels of an allergen in product, increases the risks taken by businesses. Whilst there is no doubt that the safety of the consumer is paramount, there is a danger that the fear factor within the industry and the commercial realities of product recall move the issue away from consumer food safety and towards that of brand protection. Ironically, the ‘may contain’ culture can have a detrimental effect on food safety. The allergic consumer starts to take risks with products when they have limited faith in the labelling, and in many cases frustration at the lack of suitable products, and will base their purchasing habits on products that they have bought safely in the past.
15.8
THE NEED FOR THRESHOLDS
Can we somehow live with a ‘nut free’ regime of 1 ppm or less? You can, if you conduct extensive in-house testing for every batch that the business produces – every product, every
246
Management of Food Allergens
day, every shift, combined with a comprehensive system of positive release. Much of this currently would have to be tested utilising the DNA process which is inordinately expensive, takes approximately 2 days to get a result and is not totally reliable in that they can give out mixed messages. Realistically, it is only when industry gets what are known as onsite protein tests (which we currently have for peanuts, almonds and hazelnuts), which can be done ‘in line’ and are relatively cheap, that one can begin to manufacture in a manageable manner. At the moment, those on-site protein tests for many of the nuts we handle simply do not exist, so Kinnerton have an extensive testing regime overlaying the on-site tests that continues to use DNA testing where necessary, and it is this total package that verifies for us on an audit basis that we are producing in a safe manner. Nonetheless, 1 ppm is still a tolerance that is all but impossible to guarantee. If you visit our factory, we ask you to declare whether you ate any nuts in the previous 24 h, but how do we (or you for that matter) know whether there is a tiny spec of nut protein left in residue on your hands, your eyebrows or elsewhere from the office party 2 days before? The same applies to people working in a factory who have to be free to conduct normal lives outside of work. How can we sensibly guarantee what our workforce was doing for every living minute of their daily lives without transferring the whole food-manufacturing process into producing inside a sterilised air bubble? This challenge is a challenge not just for Kinnerton’s alone, rather it is for the whole ‘free from’ industry. We must be able to test easily and economically and we must have thresholds. Without either of those working in tandem, manufacturers who are ostensibly ‘nut free’ or ‘free from’ are on a hiding to nothing. If a company that has a ‘free from’ claim is prosecuted, successfully or otherwise, for a presence of 1 ppm, then in my opinion there is no future at all in ‘free from’ manufacturing, because the risk to the manufacturer is simply not worth bearing. Ironically, the majority of people who have an informed opinion on the subject would surmise (based on existing research) that approximately 10 ppm represents minimal risk to all but the most hyperallergic. Notwithstanding this, manufacturers still have to abide by a limit of detection which is currently 1 ppm and about to get yet more stringent. As a result, Kinnerton no longer says that its products are ‘manufactured in a nut free zone’. We now say that our products are ‘manufactured with the Kinnerton Nut Safety Promise’ (Figure 15.5) and that promise says that ‘We go to extraordinary lengths to try and make our chocolates safe for people with Nut Allergy.’ Those lengths are detailed above (and on our
Fig. 15.5
New Nut Promise. (Reprinted with permission.)
Risk communication – a manufacturer’s perspective
247
website www.kinnerton.com) and we believe that today we are more robust in our production of chocolates that can be eaten with peace of mind than at any time in the decade since we began the journey. In that time, we have produced over 250 million ‘nut safe’ products that are of course targeted at the nut-allergic population and have not had a verified incident – but in law, we cannot really allow ourselves to say that we are ‘nut free’ any longer, so the Kinnerton ‘Nut Safety Promise’ suffices. Our journey of 9 years has proved that within limitations it is possible for manufacturers to separate their facilities from an allergen point of view. If the legislative bodies want this to become more of a norm and to make it workable for the food industry as a whole, then they need to address the issue of thresholds very quickly indeed. In the meantime, it is worthy of note that 9 years ago a single, very committed mother got on the phone to a chocolate manufacturer and instigated a change in the way food manufacturers go about dealing with food allergies. It can, therefore, be done.
Appendix: Useful web links
GENERAL Internet Symposium Food Allergy Portal
A1.
www.food-allergens.de www.foodallergens.info
ALLERGEN MANAGEMENT GUIDELINES
Australian Food & Grocery Council http://www.afgc.org.au/cmsDocuments/Allergen-Management.pdf Canadian Food Inspection Agency http://www.inspection.gc.ca/english/fssa/labeti/allerg/allerge.shtml UK Food Standards Agency http://www.food.gov.uk/multimedia/pdfs/maycontainguide.pdf US Food & Drug Administration http://www.cfsan.fda.gov/˜dms/wh-alrgy.html Voluntary Incidental Trace Allergen Labelling (VITAL) www.allergenbureau.net
A2.
CONSUMER GROUPS
Anaphylaxis Campaign UK Asthma & Allergy Foundation of America Association of European Coeliac Societies Coeliac UK The Food Allergy & Anaphylaxis Network Food Allergy and Anaphylaxis Alliance Gluten Intolerance Group of North America
A3.
EXPERT GROUPS
AOAC Food Allergen Community www.aoac.org CEN www.cen.eu
www.anaphylaxis.org.uk www.aafa.org www.aoecs.org www.coeliac.org.uk www.foodallergy.org www.foodallergyalliance.org www.gluten.net
Appendix: Useful web links
Gluten Unit http://www.cnb.uam.es/∼gluten/english/principal eng.htm ILSI Food Allergy Task Force http://europe.ilsi.org/activities/taskforces/diet/FoodAllergy.htm WGPAT http://wgpat.com.ar
A.4 LABORATORY ANALYSIS ALcontrol DLA Proficiency scheme Eurofins FAPAS Proficiency scheme IRMM Peanut Test material – 481 NIST material SGS WGPAT Gliadin material
www.alcontrol.com www.dla-lvu.de www.eurofins.com www.fapas.com www.irmm.jrc.be www.nist.gov www.sgs.com http://wgpat.com.ar/PWGgliadin.htm
A.5 ORGANISATIONS Campden BRI Codex Alimentarius Europa European Food Safety Authority (EFSA) FARRP – University of Nebraska Health Canada Institute of Food Research (IFR) Japanese Ministry of Health, Labour & Welfare US Food & Drug Administration (FDA) US Centre for Food Safety & Applied Nutrition
A.6
www.campden.co.uk www.codexalimentarius.net http://europa.eu www.efsa.europe.eu www.farrp.org www.hc-sc.gc.ca www.ifr.ac.uk www.mhwl.co.jp www.fda.gov www.cfsan.fda.gov
PRODUCT RECALL NOTICES
Canadian Food Inspection Agency http://www.inspection.gc.ca/english/corpaffr/recarapp/recaltoce.shtml EUROPA Rapid Alert system for Food & Feed http://ec.europa.eu/food/food/rapidalert/index en.htm Food Standards Australia & New Zealand http://www.foodstandards.gov.au/foodmatters/foodrecalls/currentconsumerlevelrecalls/ index.cfm New Zealand Food Safety Authority http://www.nzfsa.govt.nz/recalls/statistics/index.htm
249
250
Management of Food Allergens
UK Food Alerts http://www.food.gov.uk/enforcement/alerts/ UK Food Allergen Alerts http://www.alert4allergy.org/ US Consumer Product Safety Commission http://www.cpsc.gov/cpscpub/prerel/prerel.html USDA Food Safety & Inspection Service http://www.fsis.usda.gov/Fsis Recalls/index.asp
A.7 Regulations Australia & New Zealand (Standard 1.2.3.) http://www.foodstandards.gov.au/thecode/foodstandardscode.cfm Europe: Directive 2007/68/EC http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2007:310:0011:0014: EN:PDF Japan: Ordinance No.23 of 2001, Labelling of Foods containing Allergens www.mhlw.go.jp/english/topics/qa/allergies/index.html USA: Food Allergen Labelling & Consumer Protection Act (FALCPA) http://www.cfsan.fda.gov/∼acrobat/alrgact.pdf
A.8
RESEARCH PROJECTS
Allergome EuroPrevall FAST GA2 LEN GLOFAL InformAll LEAP MoniQA
www.allergome.org www.europrevall.org www.allergome.org:8080/fast/index.jsp www.ga2len.net www.glofal.org www.informall.eu.com www.leapstudy.co.uk www.moniqa.org
A.9 STANDARDS Codex Standard 118-1979 http://www.codexalimentarius.net/download/standards/291/cxs 118e.pdf The Anaphylaxis Campaign Standard http://www.highfield.co.uk/products/sector/allergens British Retail Consortium Global Food Standard www.brc.org.uk
Appendix: Useful web links
A.10 TEST KIT MANUFACTURERS Abkem Iberia BioControl Biogema ELISA Systems ELISA Technologies Euro-Diagnostica.com Generon Hai Kang Life Hallmark IFP InCura Ingenasa Operon Morinaga Neogen Nippon R-Biopharm Sedium 3M Tecra Tepnel Zeu-Immunotec
www.abkemiberia.com www.biocontrolsys.com www.biogema.sk www.elisas.com.au www.elisa-tek.com www.eurodiagnostica.com www.generon.it www.hkdnachips.com www.hallmarkav.com www.produktqualitaet.de www.incura.it www.ingenasa.eu www.glutentest.com www.miobs.com www.neogen.com www.rdc-nipponham.co.jp www.r-biopharm.com www.sedium-rd.cz www.tecra.net www.tepnel.com www.zeu-immunotec.com
251
Index
A allergen alerts, 15, 109 allergen guidance, 16, 84, 85, 89, 95, 140, 151, 155, 181, 217 allergen testing, 75, 140, 170, 171, 173 allergy services, 7 almond, 44, 46 anaphylaxis, 3, 26 antibodies, 42 antibody therapy, 36 AOAC International, 150, 156, 189 apple, 36, 53 ATP monitoring, 134, 146
E ELISA, 75, 112, 146, 150, 156, 166, 186, 244 EpiPen, 7,18,
B barley, 71 beta-lactoglobulin, 57 British Retail Consortium (BRC), 106, 217
G gliadin, 56, 157, 166 gluten, 56, 71 gluten free, 71, 72, 159, 162, 210 glutenins, 56, 157 good manufacturing practice (GMP), 84, 94, 103, 234 government organisations Agriculture and Agri-Food Canada (AAFC), 187 Canadian Food Inspection Agency (CFIA), 185 European Food Safety Authority (EFSA), 102, 198 FDA/CFSAN, 84, 91, 181 Food Standards Agency (FSA), 17, 102, 202, 220 National Institute of Standards and Technology (NIST), 155, 181, 187
C cashew, 44, 47 celery, 48, 54 cleaning, 115 cleaning in place, 141 cleaning validation, 85, 138 clinical incidence, 26 Codex, 74, 83, 90, 159, 190, 216 coeliac disease, 56, 70, 86, 185 consumer organisations Anaphylaxis Campaign, 3, 5, 114, 217, 220 Canadian Celiac Association, 185 Coeliac UK, 77 consumer survey, 8 cow’s milk, 28, 49, 57 cross-contamination, 76, 109, 123, 132 crustacea, 51, 58 D DNA, 112, 145, 242
F fish, 50, 58 Food Allergy Research and Resource Program (FARRP), 181 food challenges, 26, 91 food labelling advisory, 10, 110, 213, 225, 233, 245 mandatory, 42, 72, 83, 89, 93, 103 free from foods, 106, 246
H HACCP, 74, 95, 98, 105, 212 hazelnut, 45, 47 hen’s egg, 33, 34, 50, 58 L lactose intolerance, 77 lateral flow, 134
254
Index
lipid transfer proteins, 53, 57 lupin, 44, 46 M macadamia, 47 mollusc, 51 MoniQA, 152, 189 mustard, 48 N new product development (NPD), 97 O oats, 71 P packaging, 85, 98, 110 peach, 53 peanut, 26, 31, 43, 46 polymerase chain reaction (PCR), 186 prerequisite programmes, 105, 111, 138 prevalence, 32, 159, 184 product recalls, 102 protein families, 52
R R5 Mendez method, 75 retailers, 72, 98 rework, 108 S sesame, 55 shellfish, 58 skin prick testing, 29 soya, 32, 43, 45 spillages, 107 supplier quality assurance, 106 swabbing, 143 T thresholds, 14, 87, 90, 93, 216 training, 85, 98 W walnut, 47 wheat, 48, 49, 56 Working Group on Prolamin Analysis and Toxicity (WGPAT/PWG), 161, 181
Plate 3.1 Structures of the major birch pollen allergen Bet v 1 (PDB number 1BV1) and birch profilin (PDB number 1CQA). α-Helices are shown as cylinders and shown in cyan. Single β-pleated sheet and loops are shown in magenta and yellow, respectively. The pictures are generated using the Pymol.
Plate 3.2 Structures of prolamins: non-specific lipid transfer protein peach Pru p3 (PDB code 2ALG) and 2S albumin peanut Ara h6 (PDB code 1W2Q). First 12 residues (part of the loop) in 1W2Q are truncated to fit the picture. α-Helices are shown in cyan and loops are shown in yellow. Disulphide bridges common to many of members the prolamin superfamily are shown in green ball-and-stick form. The pictures are generated using the Pymol.
Plate 3.3 Three-dimensional structure of native soybean β-conglycinin trimer (PDB code 1IPK). The structure consists of three chains, A, B and D which are shown in cyan, magenta and yellow, respectively. The picture is generated using the Pymol.
Plate 3.4 Structures of calcium-liganded carp parvalbumin (PDB code 4CPV) and carboxyl terminus of striated alpha-tropomyosin (PDB code 1MV4). Parvalbumin has two calcium binding sites which have the same structural motif formed by an α-helix linked to a second α-helix by a 12-residue loop around the calcium cation. Calcium cations are shown as green spheres. In both structures, α-helices are shown in cyan and loops are shown in yellow. The tropomyosin structure from rat is presented. The pictures are generated using the Pymol.
Plate 4.1
Food passing across flat, damaged villi in the gut lining, without absorption.
Plate 4.2
Healthy gut villi.
Plate 15.1
Kinnerton’s clothing regime. (Reprinted with permission.)