No title

Orthopaedics and Trauma Elsevier, ISSN: 1877-1327, http://www.sciencedirect.com/science/journal/18771327 Volume 25, Issue 1, Pages 1-78 (February 2011) 1

Editorial Board, Page i

Mini-Symposium: Shoulder 2

(i) Degenerative rotator cuff disease and impingement, Pages 1-10 Robin M. Seagger, Andrew L. Wallace

3

(ii) Frozen shoulder, Pages 11-18 Tim Bunker

4

(iii) Imaging of shoulder pathology, Pages 19-29 James Teh

5

(iv) Disorders of the acromioclavicular joint and distal clavicle, Pages 30-36 A.P. Wright, I.A.R. MacLeod, S.C. Talwalker

Arthroplasty 6

Leg length inequality following total hip replacement, Pages 37-42 Anthony McWilliams, Todd D. Stewart, Andrew J. Grainger, Philip J. O’Connor, Derrick White, Anthony Redmond, Martin H. Stone

Hand 7

Fractures of the metacarpals and phalanges, Pages 43-56 Benjamin J.F. Dean, Christopher Little

Basic Science 8

The basic science and clinical applications of neurophysiological investigations, Pages 57-63 S.R. Vollans, S.S. Hasan

Medico-Legal 9

Medical negligence: legal theory and surgical practice, Pages 64-74 N.V. Todd

CME Section 10

CME questions based on the Mini-Symposium on “The Shoulder”, Pages 75-76

11

Answers to CME questions based on the Mini-Symposium on “Specific Bone Tumours”, Page 77

Book Reviews 12

AAOS (American Academy of Orthopaedic Surgeons) Clinical Practice Guideline. Treatment of carpal tunnel syndrome, Page 78 R. Amirfeyz, I.J. Leslie

13

Complications in orthopaedics: open fractures, Page 78 Nik K. Kanakaris

Orthopaedics and Trauma Orthopaedics and Trauma presents a unique collection of International review articles summarizing the current state of ­knowledge in orthopaedics. Each issue begins with a focus on a specific area of the orthopaedic knowledge syllabus, covering ­several related topics in a mini-symposium; other articles complement this to ensure that the breadth of orthopaedic learning is supplemented in a 4 year cycle. To facilitate those requiring evidence of participation in Continuing Professional Development there is a questionnaire linked to the mini-symposium that can be marked and certified in the Editorial office.

Editor-in-Chief D Limb BSc FRCS Ed (Orth) Leeds General Infirmary, Leeds, UK

Editorial Committee M A Farquharson-Roberts (Gosport, UK), I Leslie (Bristol, UK) M Macnicol (Edinburgh, UK), I McDermott (London, UK), J Rankine (Leeds, UK)

Editorial Advisory Board D C Davidson (Australia) J Harris (Australia) G R Velloso (Brazil) P N Soucacos (Greece) A K Mukherjee (India) A Kusakabe (Japan) M-S Moon (Korea) R Castelein (The Netherlands) R K Marti (The Netherlands) G Hooper (New Zealand)

Emeritus Editor Professor R A Dickson MA ChM FRCS DSc Leeds General Infirmary, Leeds, UK

A Thurston (New Zealand) E G Pasion (Philippines) L de Almeida (Portugal) G P Songcharoen (Thailand) R W Bucholz (USA) R W Gaines (USA) S L Weinstein (USA) M Bumbasirevic (former Yugoslavia)

MINI-SYMPOSIUM: SHOULDER

(i) Degenerative rotator cuff disease and impingement

relatively short lever arm of the shoulder muscles acting on the significantly long lever arm of the upper limb, often with extra load in the hand, leads to very high loads through the tendons and large reaction forces across the joint surfaces. Shoulder pain is common. One in 40 of the United Kingdom population seeks advice from primary care services each year for new and recurrent shoulder complaints.1 The majority of these presentations, and up to 70% of all referrals to shoulder clinics, are related to rotator cuff pathology.2 Primary impingement of the rotator cuff tendons was historically thought to be due to external factors, such as the prominence of the anterior-inferior acromion and the development of bony spurs extending into the coraco-acromial ligament, particularly anterior degenerative spurs. Clinical and basic science research over the years has now led to the widely held consensus that rotator cuff tendinopathy is a multifactorial pathology with intrinsic, extrinsic and environmental factors all contributing to the development of the condition.

Robin M Seagger Andrew L Wallace

Abstract Impingement and tears of the rotator cuff are common. 2.5% of the population of the United Kingdom seek advice for a shoulder complaint each year. 70% of these referrals are due to rotator cuff related pathology. Codman and Neer both initially postulated that impingement and tendinopathy were due to extrinsic factors with the antero-lateral acromion ‘impinging’ on the superior surface of the rotator cuff. This view has subsequently been modified and the general consensus now agrees that it is a multifactorial condition with both extrinsic, intrinsic and environmental factors all playing a role. A clear history and examination findings are crucial to exclude subtle causes of impingement that are atypical, e.g. instability. This review clearly describes the general shoulder examination and specific tests. The benefits and disadvantages of ultrasound and magnetic resonance imaging are discussed in the diagnosis of rotator cuff pathology. The treatment options for proven impingement vary from analgesia and physiotherapy, injection therapy to surgical intervention, both open and arthroscopic. The treatment needs to be tailored to the individual patient and is often dictated by the initial response to injections and radiological findings as well as examination findings. Outcomes of arthroscopic subacromial decompression are favourable in most studies and evidence does suggest that the surgery reduces the incidence of rotator cuff tears when compared to an unoperated cohort. There are however contradictory studies claiming physiotherapy is comparable to subacromial decompression.

Anatomy The rotator cuff is essential for normal function of the glenohumeral joint. The function of the rotator cuff is twofold, to rotate the glenohumeral joint and to centre the humeral head within the glenoid to allow the efficient function of other muscle groups, for example the deltoid. The ‘dynamic’ rotator cuff comprises four muscles, and its function in stabilizing the shoulder is complemented by the static structures, including the labrum, acromion, coracoid and coracoacromial ligament. The supraspinatus (SSP) muscle arises from the supraspinous fossa of the scapula. It is a fusiform muscle, passing anterolaterally under the acromion, where it inserts into the ‘footprint’ of the superior greater tuberosity over an area of approximately 23  16 mm.3 In one-third of people the SSP tendon also partly inserts into the lesser tuberosity.4 It is innervated by the suprascapular nerve (C5 and C6) having arisen from the upper trunk of the brachial plexus and passing below the transverse scapular ligament in the suprascapular notch. This is an area of potential compression of the nerve. The function of the supraspinatus is reported as the initiation of forward elevation and abduction over the first 15 . It is also vital in stabilizing the humeral head in the glenoid. This function is demonstrated in large and massive cuff tears, where the humeral head begins to sublux superiorly under the action of the deltoid. The infraspinatus (ISP) arises from the infraspinous fossa and passes laterally inserting into the postero-superior aspect of the greater tuberosity into a footprint of approx. 29  19 mm3. It is also supplied by the suprascapular nerve (C5 and C6). The main function of the infraspinatus muscle is external rotation of the glenohumeral joint along with teres minor, and again, stabilizing the humeral head. Teres minor (TM) originates from the upper two-thirds of the dorsal, axillary border of the scapula. It passes laterally just inferior to the lower border of infraspinatus. It inserts into the inferior aspect of the greater tuberosity and is the second of only two external rotators of the glenohumeral joint. It is innervated by the axillary nerve (C5 and C6). The subscapularis (SSC) muscle is a large triangular muscle that arises from the subscapular fossa. It passes laterally and

Keywords impingement; injection; rotator cuff; shoulder

Introduction The shoulder has the greatest range of motion of any joint in the human body. Whilst this affords many benefits, such as the ability to accurately place and manipulate the hand in a vast space, it does come at a cost. The shoulder is inherently unstable and requires bony, muscle and ligamentous support to remain congruent. Humans have evolved to undertake many overhead activities in everyday life, in the workplace and in the sporting arena. The

Robin M Seagger MSc FRCS Calvert Shoulder Fellow, The Shoulder Unit, Hospital of St John & St Elizabeth, 60 Grove End Road, London NW8 9NH, UK. Conflicts of interest: none. Andrew L Wallace PhD MFSEM FRCS FRACS Consultant Shoulder & Elbow Surgeon, The Shoulder Unit, Hospital of St John & St Elizabeth, 60 Grove End Road, London NW8 9NH, UK. Conflicts of interest: none.

ORTHOPAEDICS AND TRAUMA 25:1

1

Ó 2010 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

anteriorly to the capsule of the shoulder joint, with which it blends, and inserts into the lesser tuberosity over an area of 40  20 mm3. It is innervated by the upper and lower subscapular nerves (C5 and C6 respectively). The function of the subscapularis is internal rotation of the glenohumeral joint and humeral head stabilization. The rotator cuff muscles and tendons form the boundaries of a number of crucial anatomical landmarks. These include the quadrilateral space that contains the axillary nerve and the posterior circumflex vessels. The boundaries are teres minor superiorly, teres major inferiorly, the shaft of humerus laterally and long head of triceps medially. The triangular space contains the scapular circumflex artery and radial nerve. It comprises the teres minor and major, superiorly and inferiorly respectively, and long head of triceps laterally. The long head of biceps tendon has also been found to play an important role in subacromial impingement. The long head of biceps tendon arises from the supraglenoid tubercle within the glenohumeral joint and the initial 2e3 cm of its course is intraarticular. It exits the shoulder joint as it passes into the bicipital groove of the proximal humerus. The groove is formed by the posterior aspect of the lesser tuberosity and the anterior aspect of the greater tuberosity.

Pathophysiology Figure 1

The specific cause of rotator cuff degeneration remains unclear. Initially many believed that it was purely a mechanical process resulting from extrinsic wear of the bursal surface of the tendon beneath the superior arch of the acromion and many studies support this theory. Fu et al. attributed impingement to extrinsic forces in the majority of non-atheletes.5 Wuelker et al. presented convincing cadaveric biomechanical evidence6 and Neer presented his surgical outcome results based on this theory.7 In 1983 Neer described the three stages of impingement (Table 1). In 1986 Bigliani classified the morphology of the shape of the acromion into three types, Type I is flat, Type II is curved and Type III has an anterior beak or hook8 (Figure 1). However not all surgeons were convinced that extrinsic factors were the only cause of impingement and cuff degeneration, so there were alternative contemporary theories. Rathburn and Macnab postulated that microvascular insufficiency9 leads to areas of prolonged hypoxia and, as a result, degeneration. Uhtoff and Sarkar suggested microdegeneration of the collagen within tendon10 and this has prompted a greater interest in intrinsic biological factors operating in the cuff.

Intrinsic factors Muscle dysfunction: Nirschl11 postulated that impingement may be due to muscle dysfunction. This occurs in athletes and overhead manual workers secondary to overload of the supraspinatus tendon. When the arm is in the overhead position, the supraspinatus experiences eccentric contraction to balance the internal rotation and adduction of the arm. In a repetitive activity, this may lead to overload, due to the difference in lever arm length as previously described.12 It has also been suggested that as the muscle fatigues in overuse, (i) the function of the rotator cuff to centralize the humeral head in the glenoid and (ii) proprioceptive feedback from the tendon mechanoreceptors are both diminished, and the result can be upward subluxation under the coraco-acromial arch, possibly resulting in impingement.13,14 Overuse: impingement due to overuse is a diagnosis of exclusion.12 It is more common in younger patients who partake in repetitive overhead manual work or competitive sports men and women, particularly, throwing athletes. In these cases, the microtrauma suffered by the tendon outweighs the microrepair. This in turn results in chronic inflammation of the rotator cuff. Chronic inflammation of the tendons causes friction and wear as they pass through the limited aperture of the subacromial arch.15e17

Neer’s stages of impingement (1983) Stage I Stage II Stage III

Degenerative tendinopathy: a study of 200 cadaveric shoulders demonstrated that all shoulders with abnormal acromial morphology (see below) demonstrated some damage to the bursal surface of the cuff. However as all tendons with evidence of degeneration did not have a corresponding acromial abnormality, the authors concluded that there must be an underlying degenerative element within the tendon.18

Oedema and haemorrhage of cuff and bursa. Reversible Irreversible. Fibrosis and tendinitis of rotator cuff tendons Partial or full thickness tears of rotator cuff tendons. Irreversible without intervention

Table 1

ORTHOPAEDICS AND TRAUMA 25:1

2

Ó 2010 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

Many studies have implicated degeneration of the rotator cuff tendons as the primary aetiology of shoulder impingement. It is thought that the resulting dysfunction of the tendons may reduce their ability to centre the humeral head in the glenoid and therefore lead to cephalad subluxation and hence impingement.10 Recent histological studies from biopsies at surgery have shown that the cellularity of the rotator cuff decreased as the tear size increased.19 The authors found that there was a small increase in macrophage, mast cell and leucocyte populations as a small tear appeared, but then as the tear propagated these population sizes all diminished below the levels of control tendons with no tears. The same phenomenon was also seen with respect to tendon vascularity and neoangiogenesis. Cell populations and cell type are also affected by degeneration and tearing of the tendon. The numbers of tenocytes within the tendon decreases and they are replaced by increasing numbers of chondrocytes in degenerate tendon. Deposition of Type II cartilage within the collagen fibres of the tendon has been demonstrated. There is almost certainly a genetic element. The incidence of cuff tears in the siblings of patients with the same diagnosis was 59% whereas the incidence in matched controls with no siblings with a similar history was only 25%.20 The role of hypoxia and apoptosis (programmed cell death) in rotator cuff degeneration and subsequent tears has been investigated. It has been that, when compared to controls with no impingement, there was evidence of relative hypoxia in intact but impinging tendon and smaller cuff tears. There was also evidence of increased levels of pro-apoptotic proteins in partial and small tears but this reduced in massive tears.21

subtle instability of the glenohumeral joint. The rotator cuff muscles, as mentioned previously are important dynamic stabilizers of the glenohumeral joint and electromyography shows that they are active throughout elevation of the upper limb.27 The subtle movements of the humeral head within the glenoid can dramatically affect the biomechanics of the shoulder, particularly in overhead activities related to occupation and sport, leading to secondary impingement of the rotator cuff tendons.5,28 This theory is supported in numerous research studies that have shown that young athletes with impingement type symptoms are not improved with surgical acromioplasty alone.28,29 The treatment of these patients must include correction of the underlying cause of the problem. Scapulothoracic dysrhythmia: scapular dysrhythmia can present as a ‘secondary’ clinical sign due to an underlying pathology. Athletes, particularly those engaged in swimming and throwing sports, may develop subtle dysfunction of the scapulothoracic muscles that in turn causes ‘winging’ or abnormal protraction of the scapula on the thoracic wall.5,30e32 Scapular dysrhythmia or scapular muscle weakness will lead to an increase in the forces across the glenohumeral joint causing secondary extrinsic subacromial impingement. Abnormal function of the long thoracic nerve (supplying serratus anterior) or the spinal accessory nerve (supplying trapezius) may also result in scapular winging and secondary impingement. Acromioclavicular joint (ACJ) degeneration: the role of the ACJ in impingement is unclear, but degenerative change in this joint is a common accompanying feature. Large inferior osteophytes arising from the ACJ can occasionally be seen to indent the superior surface of the rotator cuff tendons and subacromial bursa and in these cases it may play a role.7,33e35 The joint, however, should only be resected in the presence of symptoms and degenerative change.12

Extrinsic Acromial shape: the shape and inclination of the acromion have been postulated as the main extrinsic cause of impingement for many years.22,23 Neer reported in 1972 that shape and angle of the anterior acromion (particularly, an anterior spur) were causative for subacromial impingement.7 It was based on this fact that he developed his treatment in the form of an anterior acromioplasty. The spur is thought to arise as a result of repetitive microtrauma due to traction, at the insertion of the coracoacromial ligament.12 As previously mentioned Bigliani classified the shape of the acromion into three simple and common morphologies (Figure 1).8 It was reported that the incidence of full thickness rotator cuff tears was higher in patients with Type III, hooked acromions and indeed 80% of later patients with a visible tendon tear on arthrogram had Type III morphology.12 More recently, however the interobserver reliability of the Bigliani classification using standard shoulder ‘outlet’ views has been bought into doubt.24,25 The difficulty may arise due to the radiographic projection angle rather than the classification being flawed.26

Impingement by the coraco-acromial ligament: impingement against the anterior edge of the acromion is an accepted cause of subacromial impingement. The coraco-acromial ligament is another structure that has been implicated in the development of extrinsic impingement,7,15,33,36,37 and the name ‘snapping shoulder’ attached to this condition.38 Several studies, cadaveric and surgical, have shown enlargement and thickening within the tendon39,40 and it is accepted that in the presence of swollen and inflamed subacromial structures, a prominent and unyielding coraco-acromial ligament can lead to extrinsic impingement of the rotator cuff tendons. Os acromiale: originally described in 1863, an os acromiale is a mobile segment of the acromion due to failure of fusion of the distal acromial epiphysis. The prevalence is between 1 and 15% of the normal population.41,42 Due to the anterior pull of the deltoid, the relatively mobile section of the acromion can lie in a ‘flexed’ position and cause impingement on the cuff surface or the traction of the coraco-acromial ligament can displace the os into a similar positon.41

Glenohumeral instability: underlying ligamentous laxity and resulting instability must be in the forefront of the surgeon’s mind when a young patient (<35 years) presents with the clinical signs and symptoms of impingement. It is uncommon, but not impossible, for this age group to develop true primary subacromial impingement and it is more commonly the result of

ORTHOPAEDICS AND TRAUMA 25:1

3

Ó 2010 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

The classification of os acromiale morphology is shown in Figure 2. Treatment of a symptomatic os acromiale is tailored to the size of the fragment. Small fragments are excised, larger fragments are reduced and fixed using a tension band system or screws.43,44

During repeated loading cycles, attritional wear may occur leading to tendinitis and eventually a possible tear and labral damage. This can occur more quickly as muscle fatigues.12

Presentation Patients are normally in their forties or fifties (average of 51.5 years for tendinopathy4) and complain of pain. The pain has often been present for many months after an insidious onset and has not settled despite analgesia and physiotherapy. Patients may recall a specific incident when they have injured the shoulder that precipitated the start of the pain. Care must be taken in diagnosing younger patients that present with similar symptoms with impingement syndrome, as they may have impingement as a result of underlying instability. Clearly, treating the impingement and not the underlying cause is unlikely to improve their symptoms. The pain is generally felt in the upper arm and may be worse at night and on reaching or elevation and rotational movements. They may complain of a generalized background ache with severe exacerbations on certain movements, which may assist the diagnosis. Codman and Neer both clearly described the classic presentation of impingement and cuff lesions. Common features include patients undertaking repeated over-head activity, a ‘painful arc’ between 70 and 120 of elevation, tenderness over the supraspinatus tendon and crepitus.4 Pain may not be the only complaint. Patients may also be experiencing stiffness and weakness of the shoulder. These two symptoms may be as a result of pain inhibition but if they persist, after subacromial injection, the examiner should consider other pathologies such as (i) arthritis or frozen shoulder if stiffness remains, and (ii) rotator cuff tears or neurological lesions if weakness persists. A detailed history of neurological symptoms must also be taken. In neural compression pain may be experienced around the shoulder. The patient may also experience symptoms of weakness in the shoulder, elbow or hand, and they may also report altered sensation or numbness in the upper limb. It will then be important to delineate any dermatomal pattern to the sensory changes or neurological deficit. Occupation and a history of other activities are crucial, as impingement may be more common in people who undertake a lot of overhead activity or throwing. It is also important to question the patient about previous treatment and any response they may have had to non-steroidal anti-inflammatories, subacromial injections and physiotherapy for example.

Sub-coracoid impingement: sub-coracoid impingement is an uncommon diagnosis but one the shoulder surgeon should consider.45,46 The patient generally describes pain over the anterior aspect of the shoulder that often radiates down into the upper arm. The pain is often made worse by forward elevation and particularly internal rotation.12 The theory is that the pain is caused by the impingement of the subscapularis tendon beneath the coracoid arch. Diagnosis on clinical grounds alone is difficult but MRI scanning has helped in the diagnosis of this condition. Evidence of inflammation within the subscapularis tendon or beneath the bony architecture of the coracoid in conjunction with the above symptoms should make the examiner consider this diagnosis. It has been demonstrated that the interval between the coracoid and lesser tuberosity in patients with sub-coracoid impingement was about half (6 mm vs. 11 mm) that of patients without evidence of the condition.45 Treatment of subacromial impingement was initially recommended to include excision of the coracoid process and transposition of the conjoint tendon.47 However, as arthroscopic techniques have improved, an arthroscopic coracoidplasty can now be undertaken to increase the space for the subscapularis tendon to pass. Internal impingement (postero-superior impingement): internal or postero-superior impingement occurs most commonly in overhead workers and throwing athletes.48,49 The symptoms are felt as pain in the ‘cocking’ phase of the throwing action. In this position the shoulder is in maximal abduction, external rotation and extension and the articular side surface of the rotator cuff tendons impinge on the postero-superior aspect of the glenoid. This may be exacerbated in anterior instability, as clearly this is also the position of maximal anterior translation of the humeral head.50

Examination and specific tests Physical examination should confirm what has already been postulated after a full and thorough history. A systematic ‘Look, Feel, Move’ approach should reliably elicit the clinical signs present. Inspection The patient should be stripped to the waist to fully expose the anterior and posterior aspects of the shoulder, torso and upper limb. Scars, sinuses, muscle wasting, swelling and asymmetry can then clearly be visualized.

Figure 2 Diagram showing areas of failed apophyseal fusion and os acromiale morphology.

ORTHOPAEDICS AND TRAUMA 25:1

4

Ó 2010 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

Palpation Palpation alone has a limited role in shoulder examination. The landmarks that should be palpated and can yield useful diagnostic information include the acromioclavicular joint (ACJ), the long head of biceps tendon (LHB) and the antero-lateral rotator cuff tendons. The ACJ is a common pain generator in all age groups from the third decade onwards. Pain is generally felt directly over the joint and palpation can reproduce the pain. Cross body adduction or ‘the scarf test’ are reliable in diagnosing the condition. The LHB can be palpated in the bicipital groove on the anterior aspect of the shoulder but generally the dynamic tests (see below) of the LHB are more reliable. The LHB is frequently pathological in patients with impingement. The tendon should specifically be tested during examination and inspected at the time of surgery. The rotator cuff tendons, particularly the supraspinatus tendon, can be palpated at the antero-lateral corner of the acromion. In some patients with tendinopathy and cuff tears this may elicit discomfort and if a full thickness cuff tear is present it is described that a defect may be physically palpable.

static and dynamic winging of the scapula can be present in many pathologies but give a general clue that something is amiss. Early scapulothoracic movement of one side may represent stiffness within the glenohumeral joint resulting in compensatory movement of the scapula. Active and passive range of motion can then be assessed. Examination of flexion, abduction, internal and external rotations is usually sufficient to assess function, but extension and adduction can be assessed, and also contribute to more specific tests, as described in Table 2. Active flexion is crucial in impingement. As described by Neer,7 as the arm is elevated in the scapular plane the patient experiences pain, usually between 70 and 120 . A loss of active flexion or weakness of flexion in the presence of full passive flexion suggests a tear of the rotator cuff tendons. Stiffness may represent impingement or a tear or glenohumeral pathology which will become more evident during the rest of the examination. The passive range of external rotation is an excellent indicator of glenohumeral pathology. It is best assessed initially with the patient stood with their elbows tucked into the waist and flexed to right angles. The range of rotation is measured from the neutral position with the forearms pointing directly forwards. Reduced external rotation may occur in rotator cuff pathology due to pain inhibition but it is more likely to represent osteoarthritis or frozen shoulder. Internal rotation is normally assessed by asking the patient to ‘reach up their back’ and the vertebral level reached can be compared to the unaffected side. The strength generated during flexion/elevation and rotational movements is important, particularly in rotator cuff disorders. Profound weakness may well reflect a tear of the cuff, due to pain inhibition, or less commonly a neurological deficit. If a patient less than 35 years old presents with signs and symptoms of impingement, then the examiner must consider subtle instability of the glenohumeral joint as the underlying

Movement The majority of the examination of the shoulder takes place with the patient and examiner standing. Examination of the shoulder should involve a thorough examination of the movements of the cervical spine. In the presence of degenerative changes of the cervical spine, nerve root compression or radiculopathy of the C5 nerve root may present as pain over the shoulder. Flexion, extension, lateral flexion and rotation of the cervical spine can easily be undertaken. With the patient stood facing away from you, scapular movement, symmetry and rhythm can be assessed. Ask the patient to slowly elevate their arms in front of them to approximately 90 . The arms are then lowered again. Observation of

Specific tests of impingement Neer’s sign Jobe’s test

HawkinseKennedy test

Neer’s test Yergasson’s test (LHB)

Speed’s test (LHB)

Pain on elevation of arm in plane of scapula. Pain normally felt between 70 and 120 of elevation. Arm elevated to approx. 60 . Arm placed into full internal rotation and elevation resisted by examiner. The same is then repeated with the arm in full external rotation. A positive test evokes pain on IR but the pain is less when repeated in ER. The arm is placed into 90 of elevation in the scapular plane. The elbow is flexed to 90 also and the examiner passively internally rotates the shoulder. A positive test elicits pain as the greater tuberosity is rotated beneath the acromion. Neer’s test. If there is a positive sign, local anaesthetic is infiltrated into the subacromial bursa. After a few minutes the sign is repeated. A positive result is the abolition of pain or increased range of motion and strength. Speed’s test is undertaken with the shoulder in neutral and the elbow flexed at 90 at the waist. The forearm is actively held in full supination as the examiner tries to pronate the forearm. The test is positive when pain is felt over the anterior shoulder over the LHB. The thumb of the examiner’s other hand may be used to palpate the tendon. The shoulder is elevated in the scapular plane to approximately 30 . The elbow is extended and forearm supinated. Resisted elevation of the shoulder is undertaken. The test is positive when pain is felt over the anterior shoulder over the LHB. The thumb of the examiner’s other hand may be used to palpate the tendon.

Table 2

ORTHOPAEDICS AND TRAUMA 25:1

5

Ó 2010 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

cause. If this is the case a full assessment of the stability of the shoulder should be undertaken.

Radiology Plain radiography Plain radiographs are useful in the diagnosis of impingement, despite the fact that the tendons themselves are not visualized, there are often clues to the underlying disease process. At least two views of the shoulder are required. A standard anteroposterior view, lateral scapula, outlet view or axial views are all commonly utilized. In the presence of impingement alone the antero-posterior radiograph alone may demonstrate subtle signs. The ‘sourcil sign’ (French for ‘eyebrow’ (Figure 3)) is present when there is increased sclerosis of the undersurface of the acromion. Degenerative changes of the ACJ may lead to large inferior osteophytes that may also lead to impingement. In the presence of a rotator cuff tear or cuff insufficiency there may be evidence of subtle or profound cephalad subluxation of the humeral head in the glenoid. The outlet view of the shoulder may show anterior spurs on the inferior acromion and also allow the surgeon to assess the shape of the acromion according to the Bigliani classification (Figure 1).

Figure 4 USS image of calcific deposit in supraspinatus tendon. From scale on left of image, deposit appears approx. 10 mm  6 mm.

Magnetic resonance imaging MRI is a reliable imaging method for the diagnosis of rotator cuff tears, both full and partial thickness. It is as reliable at detecting tendinopathies and tendinitis as USS, due to its ability to demonstrate inflammatory processes and oedema within the tendon. It has been shown in canine subjects that MRI is accurate in the preoperative diagnosis of abnormalities within the rotator cuff tendons and the long head of biceps tendon.52 MRI is a reliable method of diagnosing tendinopathy of the rotator cuff, however the reliability is significantly better in the hands of a senior, suitably trained radiographer53 (Figure 5).

Ultrasonography Ultrasound (USS) assessment of the rotator cuff tendons is now generally accepted to be comparable to magnetic resonance imaging (MRI) in the diagnosis of full thickness tendon tears. In a recent meta-analysis it has also been reported to be comparable in the diagnosis of partial thickness tears.51 Anecdotal evidence, however, suggests that USS is not as reliable in diagnosing partial thickness tears and tendinopathy. USS clearly has the benefit that it is a dynamic imaging modality and the patient can place the arm into the painful range of motion whilst the ultrasonographer looks for evidence of impingement or catching of the bursa or rotator cuff tendons beneath the acromion. USS can also clearly visualize the presence of calcific deposits within the substance of the tendon, a relatively common cause of impingement-type symptoms (Figure 4).

Injection therapy The use of injection therapy is common in the shoulder. Local anaesthetic and steroid injections can be used therapeutically but also as a diagnostic tool. Relief from symptoms after an injection into the subacromial space is reassuring evidence that it is the contents of this space, i.e. the tendons or the subacromial bursa, as opposed to the glenohumeral joint or cervical spine, that is the root of the problem. It is common practice to infiltrate the subacromial space with 40 mg of steroid, e.g. triamcinalone or methylprednisolone, diluted in up to 10 ml in local anaesthetic. Strict sterile techniques should be undertaken. The risk of ‘septic bursitis’ is exceeding low with very few reported cases,54,55 however the morbidity associated with it is significant. Injections of local anaesthetic can make up part of a standard shoulder assessment as in Neer’s test (Table 2) to assess for an immediate reaction of reduced pain or increased strength. If the patient responds well to the local anaesthetic or in the longer term, the steroid, then it confirms that the subacromial space is the source of the symptoms and occasionally patients can be cured with a single injection. Treatment Treatment of impingement of the shoulder clearly depends on the underlying aetiology. If subtle instability is the underlying cause, physiotherapy to enhance dynamic stability and proprioception

Figure 3 AP radiograph of the right shoulder showing sclerosis of the undersurface of the acromion e ‘sourcil sign’. Note also the large anterolateral bony spur on the acromion.

ORTHOPAEDICS AND TRAUMA 25:1

6

Ó 2010 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

Figure 5 Two images from shoulder MRI. Image a, on the left demonstrates increased signal (white arrow) within the supraspinatus tendon. This represents the tendinopathy within the substance of the tendon. Image b on the right clearly shows a full thickness tear of the supraspinatus tendon (green arrow).

would be the likely first line of treatment. This section however, will concentrate on the treatment of a tendinopathic tendon with evidence of an extrinsic bony spur from the acromion. Some patients will gain good long-term relief from a subacromial injection and need no further treatment. Others will gain excellent symptomatic relief for months at a time but the symptoms predictably return. In these cases it may be acceptable to repeat the injections if there is prolonged relief. Up to three injections annually are generally accepted as the maximum ‘safe’ dose as it has been shown that repeated exposure to steroid leads to a reduction in the tensile strength of the collagen fascicles within the tendon and therefore increased risk of rupture.56 There is evidence that physiotherapy can improve the symptoms in impingement. The treatment concentrates on strengthening the rotator cuff muscles, particularly the anterior subscapularis and posterior infraspinatus. The theory is to

maximize the effect of these muscles as a ‘force couple’ to centre the humeral head more inferiorly in the glenoid. This in turn increases the subacromial space and hence reduces the pressure on the superior rotator cuff. For recurrent or resistant impingement in a suitable patient with a clear history, clinical examination findings and ideally radiological evidence, surgery may be considered. Numerous operative procedures have been described over the years with the majority of the early descriptions involving an open procedure. Early reports suggested the excision of the entire acromion or the majority of the anterior process57 but the results of this procedure were poor in the long-term. The operation of acromionectomy resulted in defunctioning of the deltoid and if a rotator cuff tear developed later the outcome was disastrous. Neer described limited resection of the anterior-inferior third of the acromion (acromioplasty) and excision of the coraco-acromial ligament at its

Figure 6

ORTHOPAEDICS AND TRAUMA 25:1

7

Ó 2010 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

insertion.7 The description again was of an open procedure, but this was the surgical management of choice for many years. Arthroscopic subacromial decompression has gained favour since it was first described by Ellman.58 The surgical technique involves excision of the inflamed bursa, division of the coracohumeral ligament and removal of the anterior acromion and any other spurs using a high-speed burr. The completed acromioplasty must meet the same criteria as the open procedure, i.e. cover the full width of the acromion and all spurs must be removed. The 2e5-year results were good with the vast majority reporting satisfactory or excellent outcomes and very few complications.59 There is convincing long term evidence supporting the use of surgery in the management of subacromial impingement. The long-term follow-up of 105 shoulders between 8 and 11 years post-subacromial decompression reveals favourable outcomes with statistically significant improvement in range of motion, strength, pain and functional outcome scores.60 Recent meta-analyses, however, have questioned the efficacy of the surgical management of impingement. There is evidence that conservative treatment, including anti-inflammatories, injection therapy and physiotherapy can offer as good an outcome as arthroscopic subacromial decompression, although more high quality randomized controlled trials are needed to confidently draw this conclusion.61,62 Subacromial decompression has been shown to reduce the prevalence of rotator cuff tears. Patients that underwent subacromial decompression were investigated 15 years postdecompression for partial and full thickness rotator cuff tears. The prevalence in the study group was only 18%, approximately half of the reported prevalence in a similar age group of over 60 years, 40%.63 Subacromial decompression is a successful surgical procedure when undertaken by an experienced surgeon on the ‘right’ patient. The initial management of the patient should include an accurate and detailed history and full clinical assessment (Table 2) to eliminate other underlying pathologies that are causing ‘secondary’ impingement. Various surgical techniques have been described for arthroscopic subacromial decompression. Many surgeons perform the technique with the arthroscope in the standard posterior portal for viewing and the ‘working’ portal laterally. Radiofrequency ablation64 is commonly used in conjunction with a soft tissue resector to clear the bursa and release the coraco-acromial ligament from the undersurface of the acromion. It is crucial to adequately clear the soft tissue to be able to fully visualize and assess the contour of the acromion and hence the surgical resection once undertaken. In longer procedures, such as a rotator cuff repair, this clearance is even more important as over time the residual soft tissues can swell due to saline being pumped through the space resulting in irretrievable loss of view. An alternative surgical technique, and one favoured by the senior author, is the so-called, ‘cutting block’ technique. This requires an initial resection of the anterior spur and front edge of the acromion as described above. The arthroscope is then switched to the lateral portal and the posterior portal is used as the ‘working’ portal. This allows the surgeon to use the posterior aspect of the undersurface of the acromion as a guide to the level of resection of bone required to produce a flat undersurface in two planes (Figure 6).

ORTHOPAEDICS AND TRAUMA 25:1

Summary Rotator cuff tears and tendinopathy are a common presentation to the specialist shoulder surgeon. Impingement was initially considered to be a problem arising from the antero-lateral acromion causing extrinsic pressure on the superior surface of the supraspinatus tendon. More recent research has presented convincing evidence that it is in fact a multifactorial condition that involves both extrinsic factors, such as compression of the tendon, intrinsic factors such as hypoxia and collagen degeneration and environmental factors, such as occupation and sporting activities. The standard orthopaedic practice of ‘look, feel, move’ should be utilized in making the diagnosis. A clear idea of the diagnosis should be apparent from the history and examination alone in the majority of patients using the specific tests seen in Table 2. The use of injection therapy and specific radiology are used as adjuncts in the confirmation of surgeons’ ideas. It is a common diagnosis from the late fourth decade onwards. Making the diagnosis in younger patients should be undertaken with caution and only having ruled out more common conditions such as subtle instability. Once the diagnosis has been confirmed, the treatment is tailored to the patient. Analgesia and physiotherapy to improve muscle tone and proprioception may be all that some patients require, however some may benefit from a single or repeated injections depending upon the length of benefit each affords. Many patients, however, come to surgical decompression. This can be carried out as an open procedure but generally the arthroscopic technique has gained favour. The basic premise, however, remains the same, complete decompression of the tendon over the entire depth and width of the acromion, excision of the bursa and division of the coraco-acromial ligament. A

REFERENCES 1 Linsell L, Dawson J, Zondervan K, et al. Prevalence and incidence of adults consulting for shoulder conditions in UK primary care; patterns of diagnosis and referral. Rheumatology (Oxford) 2006; 45: 215e21. 2 Oh LS, Wolf BR, Hall MP, Levy BA, Marx RG. Indications for rotator cuff repair: a systematic review. Clin Orthop 2007; 455: 52e63. 3 Curtis AS, Burbank KM, Tierney JJ, Curran AR. The insertional footprint of the rotator cuff. An anatomic study. Arthroscopy 2006; 22: 609. 4 Kolts I. A note on the anatomy of the supraspinatus muscle. Arch Orthop Trauma Surg 1992; 111: 247e9. PMID: 1389774. 5 Fu FH, Harner CD, Klein AH. Shoulder impingement syndrome. A critical review. Clin Orthop Relat Res 1991 Aug; 269: 162e73. 6 Wuelker N, Plitz W, Roetman B. Biomechanical data concerning the shoulder impingement syndrome. Clin Orthop Relat Res 1994 Jun; 303: 242e9. 7 Neer C. A preliminary report. Anterior acromioplasty for the chronic impingement syndrome in the shoulder. J Bone Joint Surg Am 1972; 54: 41e50. 8 Bigliani LU, Morrison DS, April EW. The morphology of the acromion and it’s relationship to rotator cuff tears. Orthop Trans 1986; 10: 228. 9 Rathbun JB, Macnab I. The microvascular pattern of the rotator cuff. J Bone Joint Surg Br 1970 Aug; 52: 540e53.

8

Ó 2010 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

10 Uhthoff HK, Sarkar K. Anatomy and pathology of the rotator cuff. Orthopade 1995 Nov; 24: 468e74. 11 Nirschl RP. Rotator cuff tendinitis: basic concepts of pathoetiology. Instr Course Lect 1989; 38: 439e45. 12 Bigliani LU, Levine WN. Subacromial impingement syndrome. J Bone Joint Surg Am 1997 Dec; 79: 1854e68. 13 Carpenter JE, Blasier RB, Pellizzon GG. The effects of muscle fatigue on shoulder joint position sense. Am J Sports Med 1998 MareApr; 26: 262e5. PMID: 9548121. 14 Jerosch J, Castro WH, Sons HU, Moersler M. Etiology of sub-acromial impingement syndromeea biomechanical study. Beitr Orthop Traumatol 1989 Sep; 36: 411e8. 15 Uhthoff HK, Hammond DI, Sarkar K, Hooper GJ, Papoff WJ. The role of the coracoacromial ligament in the impingement syndrome. A clinical, radiological and histological study. Int Orthop 1988; 12: 97e104. 16 Jobe FW, King WD. Orthopedics: the impingement syndrome and the rotator cuff. West J Med 1989 Jun; 150: 681. PMID: 18750587. 17 McCann PD, Bigliani LU. Shoulder pain in tennis players. Sports Med 1994 Jan; 17: 53e64. 18 Ozaki J, Fujimoto S, Nakagawa Y, Masuhara K, Tamai S. Tears of the rotator cuff of the shoulder associated with pathological changes in the acromion. A study in cadavera. J Bone Joint Surg Am 1988 Sep; 70: 1224e30. 19 Matthews TJ, Smith SR, Peach CA, Rees JL, Urban JP, Carr AJ. In vivo measurement of tissue metabolism in tendons of the rotator cuff: implications for surgical management. J Bone Joint Surg Br 2007 May; 89: 633e8. 20 Gwilym SE, Watkins B, Cooper CD, et al. Genetic influences in the progression of tears of the rotator cuff. J Bone Joint Surg Br 2009 Jul; 91: 915e7. 21 Benson RT, McDonnell SM, Knowles HJ, Rees JL, Carr AJ, Hulley PA. Tendinopathy and tears of the rotator cuff are associated with hypoxia and apoptosis. J Bone Joint Surg Br 2010 Mar; 92: 448e53. 22 Goldthwait J. An anatomic and mechanical study of the shoulder joint, explaining many of the cases of painful shoulder, many of the recurrent dislocations and many of the cases of brachial neuralgias or neuritis. J Bone Joint Surg Am 1909; S2e6: 579e606. 23 Zuckerman JD, Kummer FJ, Cuomo F, Greller M. Interobserver reliability of acromial morphology classification: an anatomic study. J Shoulder Elbow Surg 1997 MayeJun; 6: 286e7. 24 Jacobson SR, Speer KP, Moor JT, et al. Reliability of radiographic assessment of acromial morphology. J Shoulder Elbow Surg 1995 NoveDec; 4: 449e53. 25 Duralde XA, Gauntt SJ. Troubleshooting the supraspinatus outlet view. J Shoulder Elbow Surg 1999 JuleAug; 8: 314e9. 26 Matsen FA, Arntz CT. Subacromial impingement. In: Rockwood CA, Matsen FA, eds. The shoulder. Philadelphia, PA: WB Saunders Co, 1990: 623e46. 27 Bandholm T, Rasmussen L, Aagaard P, Jensen BR, Diederichsen L. Force steadiness, muscle activity and maximal muscle strength in subjects with subacromial impingement syndrome. Muscle Nerve 2006; 34: 631e9. 28 Glousman RE. Instability versus impingement syndrome in the throwing athlete. Orthop Clin North Am 1993 Jan; 24: 89e99. 29 Jobe FW, Kvitne RS, Giangarra CE. Shoulder pain in the overhand or throwing athlete. The relationship of anterior instability and rotator cuff impingement. Orthop Rev 1989 Sep; 18: 963e75.

ORTHOPAEDICS AND TRAUMA 25:1

30 McMaster WC, Long SC, Caiozzo VJ. Isokinetic torque imbalances in the rotator cuff of the elite water polo player. Am J Sports Med 1991 JaneFeb; 19: 72e5. 31 Warner JJ, Micheli LJ, Arslanian LE, Kennedy J, Kennedy R. Scapulothoracic motion in normal shoulders and shoulders with glenohumeral instability and impingement syndrome. A study using Moire´ topographic analysis. Clin Orthop Relat Res 1992 Dec; 285: 191e9. 32 Kamkar A, Irrgang JJ, Whitney SL. Nonoperative management of secondary shoulder impingement syndrome. J Orthop Sports Phys Ther 1993 May; 17: 212e24. 33 Neer 2nd CS. Impingement lesions. Clin Orthop Relat Res 1983 Mar; 173: 70e7. 34 Petersson CJ, Gentz CF. Ruptures of the supraspinatus tendon. The significance of distally pointing acromioclavicular osteophytes. Clin Orthop Relat Res 1983 Apr; 174: 143e8. 35 Kessel L, Watson M. The painful arc syndrome. Clinical classification as a guide to management. J Bone Joint Surg Br 1977; 59: 166e72. 36 Pujadas GM. Coraco-acromial ligament syndrome. Proceedings of the American of Surgeons. J Bone Joint Surg 1970; 52A: 1261e2. 37 Ogata S, Uhthoff HK. Acromial enthesopathy and rotator cuff tear. A radiologic and histologic postmortem investigation of the coracoacromial arch. Clin Orthop Relat Res 1990 May; 254: 39e48. 38 McLaughlin HL. Lesions of the musculotendinous cuff of the shoulder: I. The exposure and treatment of tears with retraction. J Bone Joint Surg Am 1944; 26: 31e51. 39 Flatow EL, Soslowsky LJ, Ticker JB. Excursion of the rotator cuff under the acromion: patterns of subacromial contact. Am J Sports Med 1994; 22: 779e88. 40 Soslowsky LJ, Carpenter JE, DeBano CM. Development and use of an animal model for investigations on rotator cuff disease. J Shoulder Elbow Surg 1996 SepeOct; 5: 383e92. 41 Mudge MK, Wood VE, Frykman GK. Rotator cuff tears associated with os acromiale. J Bone Joint Surg Am 1984; 66: 427e9. 42 Edelson JG, Zuckerman J, Hershkovitz I. Os acromiale: anatomy and surgical implications. J Bone Joint Surg Br 1993; 75: 551e5. 43 Warner JJ, Beim GM, Higgins L. The treatment of symptomatic os acromiale. J Bone Joint Surg Am 1998 Sep; 80: 1320e6. 44 Kurtz CA, Humble BJ, Rodosky MW, Sekiya JK. Symptomatic os acromiale. J Am Acad Orthop Surg 2006 Jan; 14: 12e9. 45 Friedman RJ, Bonutti PM, Genez B. Cine magnetic resonance imaging of the subcoracoid region. Orthopedics 1998 May; 21: 545e8. 46 Gerber C, Terrier F, Ganz R. The role of the coracoid process in the chronic impingement syndrome. J Bone Joint Surg Br 1985 Nov; 67: 703e88. 47 Dines DM, Warren RF, Inglis AE, Pavlov H. The coracoid impingement syndrome. J Bone Joint Surg Br 1990 Mar; 72: 314e6. ¨l E. Postero-superior glenoid 48 Walch G, Liotard JP, Boileau P, Noe impingement. Another impingement of the shoulder. J Radiol 1993 Jan; 74: 47e50. 49 Davidson PA, Elattrache NS, Jobe CM, Jobe FW. Rotator cuff and posterior-superior glenoid labrum injury associated with increased glenohumeral motion: a new site of impingement. J Shoulder Elbow Surg 1995 SepeOct; 4: 384e90. No abstract available. PMID: 8548442. 50 Jobe CM. Posterior superior glenoid impingement: expanded spectrum. Arthroscopy 1995 Oct; 11: 530e6. PMID: 8534293. 51 de Jesus JO, Parker L, Frangos AJ, Nazarian LN. Accuracy of MRI, MR arthrography, and ultrasound in the diagnosis of rotator cuff tears: a meta-analysis. Am J Roentgenol 2009 Jun; 192: 1701e7.

9

Ó 2010 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

52 Murphy SE, Ballegeer EA. Magnetic resonance imaging findings in dogs with confirmed shoulder pathology. Vet Surg 2008 Oct; 37: 631e8. 53 Sein ML, Walton J, Linklater J, et al. Reliability of MRI assessment of supraspinatus tendinopathy. Br J Sports Med 2007 Aug; 41: e9. Epub 2007 Feb 8. 54 Drezner JA, Sennett BJ. Subacromial/subdeltoid septic bursitis associated with isotretinoin therapy and corticosteroid injection. J Am Board Fam Pract 2004; 17: 299e302. 55 Karthikeyan S, Kwong HT, Upadhyay PK, Parsons N, Drew SJ, Griffin D. A double-blind randomised controlled study comparing subacromial injection of tenoxicam or methylprednisolone in patients with subacromial impingement. J Bone Joint Surg Br 2010 Jan; 92: 77e82. 56 Haraldsson BT, Langberg H, Aagaard P, et al. Corticosteroids reduce the tensile strength of isolated collagen fascicles. Am J Sports Med 2006 Dec; 34: 1992e7. Epub 2006 Aug 10. 57 Armstrong JR. Excision of the acromion in treatment of the supraspinatus syndrome; report of 95 excisions. J Bone Joint Surg Am 1949 Aug; 31B: 436e42. No abstract available. PMID: 18148779. 58 Ellman H. Arthroscopic subacromial decompression: analysis of oneto three-year results. Arthroscopy 1987; 3: 173e81.

ORTHOPAEDICS AND TRAUMA 25:1

59 Ellman H, Kay SP. Arthroscopic subacromial decompression for chronic impingement. Two- to five-year results. J Bone Joint Surg Br 1991 May; 73: 395e8. 60 Klintberg IH, Svantesson U, Karlsson J. Long-term patient satisfaction and functional outcome 8e11 years after subacromial decompression. Knee Surg Sports Traumatol Arthrosc 2010 Mar; 18: 394e403. Epub 2009 Oct 23. 61 Kromer TO, Tautenhahn UG, de Bie RA, Staal JB, Bastiaenen CH. Effects of physiotherapy in patients with shoulder impingement syndrome: a systematic review of the literature. J Rehabil Med 2009 Nov; 41: 870e80. 62 Kelly SM, Wrightson PA, Meads CA. Clinical outcomes of exercise in the management of subacromial impingement syndrome: a systematic review. Clin Rehabil 2010 Feb; 24: 99e109. ¨rnsson H, Norlin R, Knutsson A, Adolfsson L. Fewer rotator cuff 63 Bjo tears fifteen years after arthroscopic subacromial decompression. J Shoulder Elbow Surg 2010 Jan; 19: 111e5. Epub. 64 Taverna E, Battistella F, Sansone V, Perfetti C, Tasto JP. Radiofrequency-based plasma microtenotomy compared with arthroscopic subacromial decompression yields equivalent outcomes for rotator cuff tendinosis. Arthroscopy 2007 Oct; 23: 1042e51.

10

Ó 2010 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

(ii) Frozen shoulder

original research into frozen shoulder. This means that many reviews and chapters are written by “experts” who have never done any original research into frozen shoulder and they fail to question its prevalence, demographics, and natural history but merely regurgitate what others have written before as gospel. This article will question the gospel.

Tim Bunker

Abstract Pathology unexplained? Today, thanks to 2 decades of research, we understand the symptoms and signs of frozen shoulder, the macroscopic and microscopic pathology, and we possess effective treatment protocols. We are even making inroads into understanding the basic science, molecular biology, and control of this disease.

The term ‘frozen shoulder’ is overused and misused. In the past 2 decades we have made significant progress into understanding the pathological processes involved and we are starting to make inroads into it’s molecular biology, which may unlock future therapeutic interventions. This article outlines the clinical management of patients with frozen shoulder and appraises current basic science research. Whilst arthroscopic release is currently the mainstay of surgical managment, it is possible that in the future treatments manipulating cellular control could help us control the course of the disease in susceptible individuals.

The condition remains subject to misinformation, therapeutic nihilism, quackery and neglect? The search engine Google will give you 3,360,000 websites that promise a surgery-free, instant cure for frozen shoulder. These treatments range from herbal pastes applied to the feet to 12 weeks of osteopathic stretching, many with celebrity affidavits and newspaper cuttings attached! This reminds me of the illustration in the back of Codman’s book. In this picture Codman sits looking out of a window at a man walking down the street with shoulder pain. The man is heading down a street lined with advertisements for fringe practitioners and fraudsters who know nothing about shoulder conditions but are willing to fleece the man and make him no better, whilst Codman, who has devoted his life to scientific study and surgery of the shoulder, and is able to correctly diagnose and effectively treat him, looks on impotently for he is a doctor and cannot advertise his expertise! So why has it taken over 70 years from Codman’s original description of frozen shoulder for us to resolve this enigma? In order to comprehend this protracted gestation period, we have to travel back to 1934 and Codman’s original description.5 Codman stated, ‘This is a class of cases which I find difficult to define, difficult to treat, and difficult to explain from the point of view of pathology. Yet these cases form a fairly distinct clinical entity’.

Keywords contracture; frozen shoulder; myofibroblast; shoulder arthroscopy

Introduction It is incredible that, to this day, frozen shoulder is so commonly overdiagnosed and misdiagnosed, its nature ill-understood, its pathology unexplained, and patients with the condition remain subjected to misinformation, therapeutic nihilism, quackery and neglect. Overdiagnosed? So what is the prevalence? The condition is less common than the oft-quoted figure of 2% of the population. This figure was arrived at 40 years ago, when shoulder disease was ill-understood and frozen shoulder was used as a dustbin diagnosis for any stiff and painful shoulder. When Hazelman performed arthrograms of 36 patients diagnosed as frozen shoulder 40 years ago, 11 had complete tears of the rotator cuff. So of that study 30% were proven misdiagnosed, and probably 50% were actually overdiagnosed. Studies in the UK 30 years ago showed that only 50% of patients referred as having frozen shoulder actually had visual/ tactile evidence of the disease (50% overdiagnosed), and studies from Canada 20 years ago showed that only 37 of 150 patients referred with the diagnosis of frozen shoulder actually qualified for the proper diagnosis (76% overdiagnosed). I doubt GPs do any better today! The author’s repeated studies on frozen shoulder (all verified by arthroscopy)1e4 show that it only accounts for 5% of shoulder disease, and since shoulder disease affects, at most, 15% of the population then it would be reasonable to suggest that the real incidence of capsular contracture is about 0.75% of the population. So frozen shoulder is an overused as well as misused term.

Symptoms and signs Codman was an extremely astute clinician, and a keen observer, so he was able to define this condition very precisely. He stated that these patients have 12 features in common: ‘The condition comes on slowly; pain is felt near the insertion of deltoid; inability to sleep on the affected side; painful and incomplete elevation and external rotation; restriction of both spasmodic and adherent type; atrophy of the spinati; little local tenderness; X-rays negative except for bony atrophy; the pain was very trying to every one of them; but they were all able to continue their daily habits and routines.’ (Table 1). Unfortunately, he did not go on to analyze each of these 12 symptoms and signs in detail, as he did when discussing the 18 features of rotator cuff tear, for, had he done so, he could well have solved the enigma forthwith. One of the frustrations of frozen shoulder is that it shares many features with those far more common shoulder disorders; impingement, partial thickness and full-thickness rotator cuff tears. In particular its onset, the site of pain, awakening at night, restricted elevation and muscle spasm are all found in rotator

Nature ill-understood? There are many myths handed down from review to review, book to book. The problem here is that there is a paucity of

Tim Bunker MD MCh FRCS Consultant Orthopaedic Surgeon, Nuffield Health Exeter Hospital, Wonford Road, Exeter, Devon EX2 4UG, UK. Conflicts of interest: none.

ORTHOPAEDICS AND TRAUMA 25:1

11

Ó 2011 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

found in arthritis and locked dislocation; and secondly, contracture of the ligaments that bind the humerus to the glenoid. Certainly, if you are going to be pedantic there are some rare muscular conditions, such as deltoid contracture (a handful of cases in the world literature), which also cause restricted passive movement, but in pragmatic terms, if the radiograph is normal and the joint shows passive restriction, then this can only be caused by contracture of the ligaments. The ligaments contract in the late stages of massive rotator cuff tears and in frozen shoulder. Therefore frozen shoulder is caused by contracture of the ligaments of the shoulder capsule QED. Indeed, as our story unfolds, you will see that this is, in fact, the case. So Codman actually had the solution within his grasp, and if only he had realized this and called the condition ‘Contracted shoulder’, instead of frozen shoulder, three generations of orthopaedic surgeons would have been spared from puzzling over this elusive condition.

Codman’s 12 criteria for frozen shoulder 1. The condition comes on slowly 2. Pain is felt near the insertion of deltoid 3. Inability to sleep on the affected side 4. Painful and incomplete elevation 5. e and external rotation 6. Restriction of both spasmodic 7. e and adherent type 8. Atrophy of the spinati 9. Little local tenderness I0. X-rays negative except for bony atrophy 11. The pain was very trying to every one of them 12. e but they were all able to continue their daily habits and routines. Table 1

Investigations cuff disease. However, I would disagree with Codman about wasting, for this is rarely seen in frozen shoulder, whilst it is commonly seen in cuff disease. This has led to frozen shoulder being a diagnosis of exclusion, a dustbin diagnosis. So, let us look critically at three aspects of Codman’s definition, for they hold the key to the condition.

Radiology Radiographs: radiographs, as we have said, are normal in frozen shoulder. However the arthrogram is pathognomonic. Neviaser performed arthrograms on the shoulders of patients with frozen shoulder and showed that the capsule of the shoulder is contracted. The joint has a diminished volume, there is absent filling of the infraglenoid recess and the subscapular recess and bicipital tunnel are obliterated in frozen shoulder. Nobody does arthrograms these days, unless part of an MRA (MR arthrogram).

The condition comes on slowly This does not get us far, for there are many disorders of the shoulder, such as impingement, which are far more common and also come on slowly. Codman had noticed that ‘they usually give a story of slight trauma or overuse’. This association with minor trauma is well known to all orthopaedic surgeons who, for instance, always caution patients with a Colles’ fracture to keep the shoulder moving lest they develop a frozen shoulder. Surgery may be another initiating factor, for instance breast surgery. It had been thought that it was the immobilization that led to the development of the frozen shoulder, but you will see, as our story unravels, it is more likely the molecular response to the injury or surgery that is responsible.

Magnetic resonance imaging (MRI): it has not shed much further light on the condition. Emig et al6 described thickening of the joint capsule and synovium in frozen shoulder. Tamai et al7 have described a technique of gadolinium-enhanced dynamic MRI, which has shown an increased blood flow to the synovium of the shoulder in frozen shoulder. Ultrasound can show thickening of the coracohumeral ligament and increased blood flow on Doppler ultrasound. However the worst pathological changes occur under the coracoid process, and ultrasound cannot see through bone.

Painful and incomplete external rotation We now come to the first distinguishing feature of frozen shoulder, which is limitation of external rotation. There are only four shoulder conditions that restrict external rotation; arthritis, locked posterior dislocation, the late stage of a massive cuff tear and frozen shoulder. All of these have specific radiographic changes. Arthritis shows diminution of joint space, inferior osteophytes, sclerosis and occasional cysts; locked posterior dislocation shows a ‘light bulb sign’ on the anteroposterior film and posterior dislocation on the axillary view; massive cuff tear shows upward subluxation of the head with a break in Shenton’s line of the shoulder and irregularity of the greater tuberosity; whilst frozen shoulder shows an entirely normal radiographic appearance of the shoulder.

Arthroscopy There have now been numerous studies that have detailed the arthroscopic findings in frozen shoulder. In the early stages the major finding is angiogenesis, or new blood vessel formation. This can be quite spectacular, with fan shaped areas of blood vessel formation, vascular fronts, petechial haemorrhages and even glomeruli (Figures 1e3). Within the infraglenoid recess the vessels line up in a radial fashion that we term the ‘lava flow’. Granulation tissue that is red, highly vascular, with a villous or fronded appearance of the synovium occurs in the rotator interval area. Angiogenesis occurs on the adjacent glenoid labrum, as well as around the base of the long head of the biceps tendon. The granulation tissue may extend on to the top surface of the subscapularis tendon, and on to the anterior edge of the supraspinatus tendon. It is interesting that angiogenesis is a feature of diabetes, for frozen shoulder is common in diabetics. In the late stages the angiogenesis diminishes. The joint is less red, but thick bands of scar tissue can be found obliterating the

Limitation of the spasmodic and mildly adherent type Terminology has changed and we would now state ‘limitation of active and passive movement’. The key to the puzzle is the limitation of passive movement that, in the shoulder, can only be caused by two things: firstly, irregularity of the joint surface, as is

ORTHOPAEDICS AND TRAUMA 25:1

12

Ó 2011 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

Figure 3 Macroscopically, the tissue was an inextensible nodular fleshy band. Histological examination of the tissue showed a background matrix of dense collagen, arranged in nodules and laminae.

The joint volume is reduced, making insertion of the arthroscope and navigation around the contracted joint difficult. As was found arthrographically, the infraglenoid recess is contracted, and the synovium is moderately inflamed in this area as well. So, to summarize at this point, the symptoms and signs of frozen shoulder suggest that there is a contracture of the shoulder joint capsule. Arthrography demonstrates a contracted joint. MRI and ultrasound show thickening of the joint capsule. Arthroscopy confirms a contracted joint space with granulation tissue in the rotator interval area. A pattern is beginning to emerge, and it is a contracture.

Figure 1 The major finding is a consistent abnormality arising out of the subscapularis bursa, at the origin of the long head of biceps. This abnormality consists of ‘granulation tissue’ which is red, highly vascular, with a villous or fronded appearance of the synovium.

normal structure of the capsule. The superior glenohumeral ligament becomes thickened, obliterating the rotator interval. Scar can cover the top edge of subscapularis, and we term this the ‘involucrum’. Occasionally the long head of biceps can be scarred to the cuff. The middle glenohumeral ligament is thickened, as indeed is the whole of the capsule. The joint surface is usually normal. In our series of 35 consecutive patients with frozen shoulder whom we arthroscoped,1 there were three patients with minor fibrillation of the cartilage of the humeral head, two of whom had minor changes on the glenoid but this is normal for the population studied, who had an average age of 54 years. There are no intra-articular adhesions in this condition, a fact that has been documented in eight recent arthroscopic studies on frozen shoulder.

Blood tests What of the blood tests? The full blood count, white cell count and ESR are all normal in this condition. CRP may be elevated in the early stages. Calcium, phosphate, serum globulins and bone alkaline phosphatase are all normal. Cholesterol and triglycerides may be elevated in frozen shoulder8 but clearly this is not a specific test. However, what is interesting is that cholesterol and triglycerides may also be elevated in that other contractile disease, Dupuytren’s contracture, one of many associations shared by these two contractile diseases.

Associations Meulengracht and Schwarz9 found evidence of Dupuytren’s disease in 18% of their patients with frozen shoulder. Schaer10 found that 25% of their patients with frozen shoulder had Dupuytren’s contracture. We found 58% of our patients with frozen shoulder had evidence of palmar contracture, usually in the form of a pit or nodule, and more rarely a band of Dupuytren’s disease. This association becomes even more intriguing when reviewing the literature on the pathology of frozen shoulder, brief though it is, for two authors11,12 state that the histological appearance of shoulder capsule in patients with frozen shoulder is identical with Dupuytren’s disease of the palm and, as will be seen, we have confirmed this appearance.13 The other strong association with frozen shoulder is diabetes. Diabetics have a 20% risk of developing frozen shoulder, which rises to 36% in insulin-dependent diabetics. Of patients with

Figure 2 Arthroscopy in frozen shoulder typically reveals ‘granulation tissue’ – highly vascular, fronded tissue, arising out of the subscapularis bursa at the origin of the biceps tendon.

ORTHOPAEDICS AND TRAUMA 25:1

13

Ó 2011 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

bilateral frozen shoulders, 42% are diabetic. Frozen shoulder is often severe in patients with diabetes, and more difficult to treat.14 Likewise, Dupuytren’s contracture is common in diabetics, and is difficult to treat. Patients with diabetes may have raised serum lipid levels.15 These associations may be interesting, but they are just pieces of a puzzle which may, or may not, fit together to form a complete picture. Orthopaedic surgeons, however, are practical people; they like to hold things in their hands, and inspect them with their eyes, in order to understand them. So what about the surgical findings in frozen shoulder?

ligament may become shortened, and may have to be released at surgical reconstruction to restore external rotation’.19 Ozaki surgically explored 17 patients with recalcitrant chronic frozen shoulder20 and stated, ‘At operation, the major cause of the restricted glenohumeral movement was found to be contracture of the coracohumeral ligament and rotator interval. Release of the contracted structures relieved pain and restored motion of the shoulder in all patients’. In a consecutive series of 25 patients with severe frozen shoulder, which had failed to resolve with manipulation, we found a consistent abnormality of the coracohumeral ligament and rotator interval area.21 This area of the capsule was abnormal, thickened and vascular. There seems to be new vessel formation (much like Codman’s bloodshot eye). The thickening prevented easy determination of the edges of the rotator interval (Simmonds, no obvious demarcation). When the shoulder was forced into external rotation the contracted coracohumeral ligament became taut and stood out as a palpable thickening (DePalma’s checkrein), often as thick as the surgeon’s little finger. When the tissue was incised it bled, often forcefully, and was found to be very adherent (Neviaser’s adherence) to the underlying long head of biceps, and the incision was accompanied by release of the passive restraint to glenohumeral external rotation (as stated by Neer and Ozaki). Surgical release gave us the ability to inspect the tissue histologically.

Surgical findings Codman was probably the first to surgically explore the frozen shoulder. He stated, ‘I used to give ether and open the (subacromial) bursa. The appearance of the floor of the bursa was always the same e a congestion over the supraspinatus tendon on the base of the bursa like that of a bloodshot eye. Adhesions were often found. The congestion was in the synovia, not in the tendon below it.’ He does not state whether he ever opened the shoulder joint itself. Neviaser16 reported on 10 patients with frozen shoulder upon whom he had operated. He approached the shoulder using a deltopectoral approach, incising subscapularis vertically to give a good view of the capsule. The capsule he found to be thickened, adherent to the humeral head and, when it was released, it was found to be under such tension that it sprung apart and could not be re-approximated. Although he never mentioned the word contracture, his description speaks for itself. Simmonds17 described the rotator cuff as ‘looking like a vascular, leathery hood with no obvious demarcation between the tendons.’ We now term the demarcation between the tendons the ‘rotator interval’. The rotator interval is the triangle formed between supraspinatus and subscapularis, the base being the coracoid process. Normally, the capsule here is quite thin, being strengthened by the superior glenohumeral ligament and also by the coracohumeral ligament, which runs from the base of the coracoid process to the biceps sulcus, the area between the lesser and greater tuberosities of the humerus. This, remember, is the area that appears so abnormal at arthroscopy, being obliterated with granulation tissue. DePalma stated that ‘The coracohumeral ligament is converted into a tough inelastic band of fibrous tissue spanning the interval between the coracoid process and the tuberosities of the humerus. It acts as a powerful checkrein. division of the coracohumeral ligament allows early restoration of scapulohumeral movement’.18 This is truly the most elegant description of a contracture of the coracohumeral ligament. DePalma was clearly a hairsbreadth away from resolving the enigma of frozen shoulder when he got distracted by the long head of biceps, and concluded that tethering of biceps was the cause of frozen shoulder. Lundberg exposed the shoulder in 20 patients with frozen shoulder.12 He noted peri-articular inflammatory changes, especially at the insertion of the cuff on to the greater tuberosity. In six patients, he performed an arthrotomy and found thickening of the capsule and no intra-articular adhesions. Neer stated, ‘The senior author has found that in pathological conditions such as frozen shoulder, old fractures, or arthritis, the coracohumeral

ORTHOPAEDICS AND TRAUMA 25:1

Pathology of frozen shoulder We examined the tissue excised from the rotator interval area of 12 consecutive patients with severe frozen shoulder.22 Macroscopically, the tissue was an inextensible nodular fleshy band. Histological examination of the tissue showed a background matrix of dense collagen, arranged in nodules and laminae. The cell population was moderate to high (Figure 4). The samples were prepared for immunocytochemistry to demonstrate precisely what type of cell was present. This staining showed that the cells were fibroblasts, with some transformation to the contractile fibroblast that has been termed the ‘myofibroblast’. The samples showed that the tissue was particularly vascular. As for inflammatory cells, none were present within the thick collagen of the contracture itself, but some were present both in the synovium and around blood vessels. Such an appearance, of a dense collagen matrix populated by fibroblasts

Figure 4 The cell population was moderate to high.

14

Ó 2011 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

and myofibroblasts, can be found in healing scar tissue and in contracture. This appearance is very similar to the palmar contracture of Dupuytren’s disease. Historically Neviaser examined his tissue biopsies microscopically16 and found considerable or extensive fibrosis in 6 of 10 cases. Simmonds reported dense collagen fibres, increased vascularity and the presence of fibroblasts. DePalma found fibrosis, increased vascularity, thickening of the synovial membrane and cellular infiltration. Despite the evidence of fibrosis, all three authors concluded that the changes represented low-grade inflammation. Lundberg12 reported a compact, dense capsule with an increase of cells that were fibroblasts; he was the first to mention the similarity of the appearance of this contracture with palmar contracture, although he attributed the idea to his pathologist Norden. Kay and Slater11 also noted the resemblance between the histology of shoulder joint capsule in frozen shoulder and the palmar contracture of Dupuytren. Ozaki20 noted fibrosis in their tissue, and Hannifin et al found diffuse capsular fibroplasia, thickening and contracture. In the modern era, further to our immunocytochemical studies, Killian’s group have confirmed the presence of fibroblasts laying down collagen within the capsule. They performed electron microscopical studies showing that the collagen structure was grossly abnormal, with thickened fibrils in the frozen shoulder group. Carr’s group from Oxford confirmed the presence of fibroblasts laying down collagen. Like us they found some inflammatory cells, but they have also shown the presence of mast cells and postulate that these cells may be modulating some inflammatory process that triggers the fibroblastic response. To recapitulate, the symptoms and signs of frozen shoulder suggest that there is a contracture of the shoulder joint capsule. Arthrograms demonstrate a contracted joint. MRI shows a thickening of the joint capsule. Arthroscopy confirms a contracted joint space with granulation tissue in the rotator interval area. Frozen shoulder shares associations with the palmar contracture of Dupuytren. Surgical exploration reveals a contracture of the shoulder capsule, mainly around the coracohumeral ligament, which acts as a checkrein to passive movement. Histological examination shows this contracture to be formed of a dense collagen matrix, populated with fibroblasts actively forming collagen, and with some transformation to the contractile myofibroblast, the active cell of contracture. A pattern of contraction continues to develop. This must lead us on to consider the basic science of the fibroblast and the myofibroblast in healing and contracture.

Fibroblasts are controlled by certain cytokines. Cytokines are peptide molecules that act as cell messengers. They control many aspects of cell migration and growth, acting in minute concentrations by binding to receptors on the target cell. Cytokines regulate fibroblast chemotaxis, fibroblast proliferation and collagen synthesis. Because the fibroblast appears to be the key cell in frozen shoulder, we elected to measure cytokine and growth factor levels in 17 consecutive patients with severe frozen shoulder.23 The reverse transcription polymerase chain reaction (PCR) method was used to measure the levels of cytokines and growth factors in tissue biopsied from these patients with frozen shoulder. The tissue showed an over-expression of a number of cell-signalling molecules. The intensity of signal for the fibrogenic growth factors such as transforming growth factor beta (TGFb), platelet-derived growth factor alpha (PDGFa), and fibroblast growth factor (FGF) was elevated and was higher than the pro-inflammatory cytokines such as interleukin 1 and tumour necrosis factor (TNF), although there was a high level of interleukin 6. This work has been elegantly confirmed by Colville’s group, who took joint fluid from patients with capsular contracture and found that this tissue caused a 5000% increase in in-vitro fibroblast proliferation compared with control groups. These elevated cytokine levels in frozen shoulder have also been demonstrated by Rodeo, Hannafin and Warren24 using monoclonal antibody techniques. They found TGFb and PDGF to be elevated, and suggested that these cytokines may act as a persistent stimulus causing capsular fibrosis and the development of frozen shoulder. Hamada, in Japan, has found increased levels of vascular endothelial growth factor (VEGF) in stiff shoulders that may account for the angiogenesis that is seen at arthroscopy.25 Ryu et al demonstrated strong expression of VEGF and angiogenesis in diabetic frozen shoulders. Killian et al showed an increase in alpha-1(I) mRNA transcription in both frozen shoulder and Dupuytren’s contracture. Since Insulin like growth factor is known to stimulate fibrosis in connective tissue they measured the serum IGF-1 and IGF-1 receptor levels but found them to be similar in contracted shoulder and control capsule. Patients with palmar contracture also have elevated levels of cytokines and growth factors which, although not identical to those changes found in frozen shoulder, show a higher intensity for the fibrogenic growth factors than the inflammatory ones.26 Of course, when we look at the development of contractures, we have to examine not only the factors that may act as a persistent stimulus to scar formation, but we must also look at the opposite side of the equation, the failure of remodelling. The remodeling of the extracellular collagen matrix is undertaken by a family of enzymes that used to be called collagenases, but now go by the glorious name of the matrix metalloproteinases (MMPs). MMPs share five basic attributes: they degrade the extracellular matrix, they contain zinc, they are secreted in a latent pro-form, they are inhibited by tissue inhibitors (TIMPs) and they share common amino-acid sequences. We decided to examine the levels of MMPs and TIMPs in frozen shoulder, once again using the reverse transcription PCR technique.23 We found a strong expression of MMPs in frozenshoulder tissue, particularly MMP2. However, we found an even greater expression of their natural inhibitor TIMP.

The basic science of healing and contracture The healing process is divided into three phases: an early inflammatory phase; a repair phase with the formation of granulation tissue; and finally scar formation, maturation and contracture. Contracture may be physiological, when the healing wound contracts to pull the wound edges into apposition, or it may be pathological, where there is an imbalance between scar formation and remodelling, resulting in abnormal scarring and contracture. It is the late stages of healing, the formation of granulation tissue, scar formation, contracture and remodelling that hold the key to understanding the molecular biology of frozen shoulder.

ORTHOPAEDICS AND TRAUMA 25:1

15

Ó 2011 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

This leads us to speculate as to whether there may not be a failure to remodel in frozen shoulder due to persistent high levels of T1MP. A broad-spectrum TIMP has been synthesized (MarimastatÒ, British Biotech Ltd, Oxford, UK) and a remarkable study has been carried out with it. Twelve patients with inoperable gastric carcinoma were enrolled into a study to see whether TIMP could slow down the progression of their disease by inhibiting MMPs (which are found to be elevated in gastric carcinoma), thereby preventing the dissemination of the tumour through the extracellular matrix and naturally encasing the tumour in scar tissue. Of the 12 patients treated, six developed bilateral frozen shoulder within 4 months of starting treatment and three also developed palmar contractures.27 This remarkable in-vivo experiment would appear to confirm our thoughts on the role of TIMP in the formation of frozen shoulder. But why should patients with frozen shoulder have an abnormal healing and remodeling? Could there be some genetic predisposition to this?

arthroscopic release), or reverse the abnormality of molecular biology that is causing the contracture (manipulate the cytokines or MMPs). Treatment should also be tailored to the course of the disease, which has three phases: the painful phase (inflammatory phase), the stiff phase (contracture), and the remodelling phase. There has been a great deal of debate about the natural history of frozen shoulder. The literature shows great variation. This allows the optimistic surgeon to ‘spin’ this information so that it appears the condition always gets better in 2 years, and the pessimistic surgeon to ‘spin’ in the opposite direction and say that, if you look critically, no one gets better. As usual, the answer lies half way between these extremes, but it is well worth looking at some of the evidence. Codman stated, ‘even the most protracted cases get better, with or without treatment, in about 2 years’. It is this statement that has got into the textbooks, has become dogma, and has become established, without question, through three generations of orthopaedic surgeons. However, if you read the original text, you will find that Codman was not dogmatic, for he qualified his statement in two ways. Firstly he stated that, ‘recovery is by degrees, it is pretty hard even for the patients to say when they are well’, and secondly what he meant by entire recovery was that the joint was ‘not left deformed or otherwise permanently damaged’. Contrast this with the statement of Simmonds17 ‘complete recovery. is not my experience’ and of DePalma18 ‘It is erroneous to believe that in all instances restoration of function is attained’. Most papers show that 10% of patients have an ongoing disability. Reeves’ very detailed study30, that had a 10-year follow-up, showed that of 41 patients, 25 had a permanent loss of movement, although there was only a functional loss in three. The best paper on the natural history of frozen shoulder31 evaluated 62 patients both objectively and subjectively for an average follow-up of 7 years. They found that 50% still had either mild pain or stiffness of the shoulder or both, whilst 60% still demonstrated some restriction of motion compared with study-generated control values. Marked restriction of movement, when present, was most common in external rotation. The authors concluded that complete resolution was not universal and brought them to question whether this is a self-limiting condition, for half their patients remained symptomatic, and more than half had a measurable loss of motion, although this caused little functional disability. Griggs et al32 confirmed these findings and stated that ‘even amongst the patients who were satisfied, a substantial number were not pain free’; 10% had mild pain at rest, and 27% had mild or moderate pain with activity. 40% of the satisfied patients had abnormal shoulder function. Our own studies at 2e5 years showed that although 86% had an improvement in their level of pain, this did not mean that they had no pain. Only 53% had no pain, 33% had an occasional pain and 14% had marked residual pain. These findings have been confirmed by the largest ever study from Oxford on 273 patients followed for up to 20 years. Using the Oxford Shoulder Score they demonstrated that 41% of their patients had mild to moderate persistent symptoms at 7 years and 6% had severe ongoing symptoms with pain and functional loss.33 In all patients, there are two aspects that need to be treated: the pain and the contracture. Pain is an early manifestation of the

The cytogenetics of frozen shoulder Clonal chromosomal abnormalities have been discovered in a variety of contractile diseases. In particular trisomy 7 and 8 have been found in Dupuytren’s disease,28 and multiple clonal chromosomal abnormalities in Peyronie’s disease.29 We therefore elected to see if there were any clonal chromosomal abnormalities in frozen shoulder. We took capsular tissue from 10 consecutive patients with frozen shoulder, cultured the cells in tissue culture and then performed metaphase arrest, followed by in-situ G banding to look for abnormal karyotypes. To our surprise, we found clonal chromosomal abnormalities in frozen shoulder. These abnormalities were trisomy of chromosomes 7 and 8. A twin study examining 865 pairs of heterozygous and 963 pairs of homozygous twins estimated a heritability of 42% for frozen shoulder and stated that genetic factors are implicated in the aetiology of frozen shoulder. So where does all this fancy science take us? We can now say that the symptoms and signs of frozen shoulder make us postulate that this is a contractile disease. This is confirmed by arthrography, by MRI, by arthroscopy, by its associations with palmar contracture, by surgical exploration, by histology, and by immunocytochemistry. In frozen shoulder, fibrogenic growth factors are dominant, remodelling is prevented by high levels of TIMP, and treating cancer sufferers with TIMP causes both frozen shoulder and palmar contracture. Finally, both frozen shoulder and palmar contracture demonstrate clonal chromosomal abnormalities with duplication of the same chromosomes. A pattern appears to be developing, and that is that frozen shoulder is a contracture of the capsule of the shoulder.

Treatment of frozen shoulder We now have a great advantage over our orthopaedic forebears: we know that frozen shoulder is a contracture of the capsule of the shoulder. This allows us to treat it logically. So, depending on how severe the contracture is (for it varies in its intensity from patient to patient), we can either stretch it out (home exercises or physiotherapy), break the contracture (manipulation under anaesthetic), excise the contracture (open surgical release or

ORTHOPAEDICS AND TRAUMA 25:1

16

Ó 2011 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

disease. It may be caused by the inflammatory component; that red area of granulation tissue that is seen arthroscopically, which is rich in cytokines. It may actually be caused by bleeding into the joint, for the new blood vessels are fragile and leaky. This could account for the jerk pain that patients get. Steroids have been shown to have no benefit over home exercises.34 However all studies can be severely criticized as they enrolled painful stiff shoulders, in other words primary and secondary frozen shoulder, so many of the patients would have had other shoulder disease. One of these papers did arthrograms of the study group and 11 of 36 had cuff tears, yet were kept in the study! This is a recurring criticism of so many papers on capsular contracture, the diagnosis is wrong. A recent randomized double blind study of a 3-week course of oral Prednisolone showed no significant difference between the active and placebo arms of the study at 6 weeks or 3 months. The best paper on physiotherapy is that of Diercks et al,35 which showed that intensive physiotherapy prolonged the natural history of the disease from 15 months to 24 months and achieved a lower Constant Score of 76 compared to 87 in the control group who did home exercises. Once again we must stress that what the patients want is not for their disease to be prolonged from 15 to 24 months, but for it to end TODAY. Manipulation under anaesthetic has historically been used by many orthopaedic surgeons, and is still used by some today. However, it has had a chequered career. Professor Sir John Charnley36, before he became famous for his hip replacement, was intrigued by frozen shoulder. He published a paper on his personal results of manipulation of frozen shoulders in 1959. Before performing the study, he sent a questionnaire to his colleagues in the British Orthopaedic Association, asking for their views on manipulation of frozen shoulder. Seventy percent said they would never perform a manipulation, as all would eventually get better, and some could be harmed. Lesser men would have been put off by such a reply, but not Charnley. In a consecutive series of 35 patients he found that he did no harm, pain was eased by manipulation, and however long the duration of the disease, most were free of symptoms by 10 weeks. Andersen, Sjobjerg and Sneppen37 have shown that 79% of patients with frozen shoulder are relieved of their pain, and 75% regain a near normal range of movement after manipulation. We have arthroscoped patients before and after manipulation to discover exactly what is happening.1 Essentially, what we found was that elevation, or abduction, tears the capsule from the neck of the humerus, releasing the inferior capsule, and this occurs with relative ease. It is much harder to free rotation, but forced external rotation tears the coracohumeral ligament. This is an extra-articular ligament, so what is seen arthroscopically is haemorrhage in the rotator interval. Often, the coracohumeral ligament is so contracted that it will not tear and the patient is left with limitation of external rotation. Loew has shown that manipulation is not without complications. Arthroscopic release, in the hands of the expert shoulder surgeon, has transformed the management of capsular contracture. Many of the studies can be criticized for purporting to show the results of treating capsular contracture when the index group was actually made up of any stiff shoulder including fractures, cuff disease and post-surgical stiff shoulders and then pooling the results. For instance one paper started with 1720 stiff shoulders of

ORTHOPAEDICS AND TRAUMA 25:1

which only 11 had an arthroscopic release for primary adhesive capsulitis. Four articles are worthy of study. Ogilvie Harris38 compared the results of manipulation versus arthroscopic release in their hands. Although both groups gained the same substantial improvement in range of motion the arthroscopic group had significantly better pain relief and function, to the extent that twice as many were graded excellent. The following year J.P.Warner39 showed a 49 increase in elevation, 42 increase in external rotation and improvement in Constant Scores from 13 to 77/100. Harryman and Matsen published a year later and demonstrated fantastic results. The range of motion went from 41% of the opposite side to 78% on the first postoperative day and 93% at the end of the study. Before surgery 6% could sleep and after 73%. They were the first to show the dramatic speed of recovery following treatment, which is the very thing that patients want. Berghs and Bunker21 confirmed this with a dramatic improvement on day one post surgery in 36% and 88% improvement within 2 weeks. Pain improved from 3.6/15 to 12.6/ 15 and the partial Constant Scores from 20/75 to 62/75. There were no complications in three of these studies, but one transient axillary neurapraxia in the Harryman study. Arthroscopic release appears to show great promise for it delivers what the patient wants; relief of pain, undisturbed nights and improved function TODAY, or if not today THIS WEEK, in the majority of people, with minimally invasive, keyhole, day-case surgery. However it is not a panacea, for it appears that 10% of patients fail to improve whatever is thrown at them. This group can be predicted to a certain extent. Those who fall into this worse group are the men, diabetics, those with marked Dupuytren’s disease, bilateral disease, severe contractures and those with failed previous treatment. Arthroscopic release gives a good improvement in forward elevation, and external rotation, but internal rotation can be disappointing. Several authors have seen whether this can be improved by posterior release but Snow and Funk40 showed no significant difference in range of motion with the addition of a posterior release. Against this Pouliart and Gagey have described variations in the superior capsuloligamentous complex41, and explain the ramifications of this. It may limit internal rotation in contracture, and the release needs to be extended postero-superiorly in many patients.

Summary Frozen shoulder is a contracture of the shoulder joint capsule. Although the disease causes a global contracture of the shoulder joint, it appears maximal in the rotator interval area, and particularly around the coracohumeral ligament. The contracture can be visualized by arthrography and arthroscopy, and the capsule is seen to be thickened and vascular on MRI. Surgical exploration confirms the capsular contracture, and the fact that it is maximal around the coracohumeral ligament. Histology shows that this contracture is made of a dense collagen matrix, which shows a high degree of cellularity, and the cells are fibroblasts and myofibroblasts. Fibroblasts are under the control of fibrogenic cytokines and growth factors, which are found to be elevated in patients with frozen shoulder. Remodelling may be slow in frozen shoulder due to high levels of TIMP. In a bizarre study, frozen shoulder has been shown to be produced by administering TIMP to humans. Arthroscopic release now gives us

17

Ó 2011 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

21 Berghs BM, Sole-Molins X, Bunker TD. Arthroscopic release of adhesive capsulitis. J Shoulder Elbow Surg 2004 MareApr; 13: 180e5. 22 Omari A, Bunker TD. Open surgical release for frozen shoulder: surgical findings and results of the release. J Shoulder Elbow Surg 2001; 10: 353e7. 23 Bunker TD, Reilly J, Baird K, Hamblen D. Cytokines and growth factors in frozen shoulder. Presented at the Combined Meeting of the Orthopaedic Associations of the English Speaking World, Auckland, New Zealand, 1998 24 Rodeo S, Hannafin J, Tom J, Warren R, Wickiewicz T. Immunolocalisation of cytokines in adhesive capsulitis of the shoulder. J Shoulder Elbow Surg 1997; 6: 174. 25 Handa A, Gotoh M, Hamada K, et al. Vascular endothelial growth factor 121 and 165 in the subacromial bursa are involved in shoulder joint contracture in type II diabetics with rotator cuff disease. J Orthop Res 2003; 21: 1138e44. 26 Baird KS, Crosan JF, Ralston SH. Abnormal growth factor and cytokine expression in Dupuytren’s contracture. J Clin Pathology 1993; 46: 425e8. 27 Davies TRC. Frozen shoulder after treatment with Marimastat. Personal communication 1998 28 Bonnici AV, Birjandi F, Spencer JI, Fox SP, Berry AC. Chromosomal abnormalities in Dupuytren’s contracture and the carpal tunnel syndrome. J Hand Surg 1992; 17B: 349e55. 29 Guerneri S, Stioui S, Mantovani F. Multiple clonal chromosomal abnormalities in Peyronie’s disease. Cancer Genet Cytogenet 1991; 52: 181e5. 30 Reeves B. The natural history of frozen shoulder. Scand J Rheumatol 1975; 4: 193e6. 31 Shaffer B, Tibone JE, Kerlan RK. Frozen shoulder: a long term follow up. J Bone Joint Surg 1992; 74A: 738e46. 32 Griggs SM, Ahn A, Green A. Idiopathic adhesive capsulitis. A prospective functional outcome study of nonoperative treatment. J Bone Joint Surg [Amer] 2000 Oct; 82-A: 1398e407. 33 Hand C, Clipsham K, Rees JL, Carr AJ. Long-term outcome of frozen shoulder. J Shoulder Elbow Surg 2008 MareApr; 17: 231e6. 34 Arroll B, Goodyear-Smith F. Corticosteroid injections for painful shoulder: a meta-analysis. Br J Gen Pract 2005 Mar; 55: 224e8. 35 Diercks RL, Stevens M. Gentle thawing of the frozen shoulder: a prospective study of supervised neglect versus intensive physical therapy in seventy-seven patients with frozen shoulder syndrome followed up for two years. J Shoulder Elbow Surg 2004 SepeOct; 13: 499e502. 36 Charnley JC. Periarthritis of the shoulder. Postgrad Med J; 1959: 384e8. 37 Andersen NH, Sojbejerg JO, Johannsen HV, Sneppen O. Frozen shoulder: arthroscopy and manipulation. J Shoulder Elbow Surg 1998; 7: 218e22. 38 Ogilvie-Harris DJ, Myerthall S. The diabetic frozen shoulder: arthroscopic release. Arthroscopy 1997; 13: 1e8. 39 Warner JJP, Allen A, Marks PH, Wong P. Arthroscopic release for chronic refractory adhesive capsulitis of the shoulder. J Bone Joint Surg [Amer] 1996; 78: 1808e16. 40 Snow M, Boutros I, Funk L. Posterior arthroscopic capsular release in frozen shoulder. Arthroscopy 2009 Jan; 25: 19e23. 41 Pouliart N, Somers K, Eid S, Gagey O. Variations in the superior capsuloligamentous complex and description of a new ligament. J Shoulder Elbow Surg 2007; 16: 821e36.

an effective, rapidly working, day case minimally invasive treatment for this condition. Understanding the nature of frozen shoulder allows us to apply effective treatments to the condition, and opens the doors to the possibility of manipulating the course of the disease, so that patients who develop this common, disabling, painful and protracted condition may, in the future, enjoy effective early resolution of their disease. A

REFERENCES 1 Bunker TD, Lagae K, DeFerm A. Arthroscopy and manipulation in frozen shoulder. J Bone Joint Surg 1994; 76(B)(supp 1): 53. 2 Bunker TD. Frozen shoulder: unravelling the enigma. Ann R Coll Surg 1997; 79: 210e4. 3 Bunker TD. Frozen shoulder. In: Norris T, ed. Orthopaedic knowledge update: shoulder and elbow. Rosemont IL: American Academy of Orthopaedic Surgery, 1997. 4 Bunker TD. Frozen shoulder. In: Bunker TD, Schranz PJ, eds. Clinical challenges in orthopaedics: the shoulder. Oxford: Isis, 1998. 5 Tendinitis of the short rotators. In: Codman EA, ed. Ruptures of the supraspinatus tendon and other lesions in or about the subacromial bursa. Boston: Thomas Todd, 1934. 6 Emig EW, Schweizer ME, Karasick D, Lubowitz J. Adhesive capsulitis of the shoulder. MRI diagnosis. Am J Radiol 1995; 164: 1457e9. 7 Tamai K, Yamato M. Abnormal synovium in frozen shoulder. A preliminary report with dynamic magnetic resonance imaging. J Shoulder Elbow Surg 1997; 6: 534e43. 8 Bunker TD, Esler CAN. Frozen shoulder and lipids. J Bone Joint Surg 1995; 77B: 684e6. 9 Meulengracht E, Schwartz M. Course and prognosis of periarthritis scapulohumerale. Acta Med Scand 1952; 143: 350e60. 10 Schaer H. Die Atiologie der Periarhritis humeroscapularis. Ergebn Chir Orthop 1936; 29: 11. 11 Kay NRM, Slater DN. Fibromatosis and Dupuytren’s disease. Lancet 1981; 2: 303. 12 Lundberg BJ. The frozen shoulder. Acta Orthop Scand 1969; 119: 1e59. 13 Bunker TD, Anthony PP. The pathology of frozen shoulder. J Bone Joint Surg 1995; 77B: 677e83. 14 Janda DH, Hawkins RJ. Shoulder manipulation in patients with adhesive capsulitis and diabetes mellitus. J Shoulder Elbow Surg 1993; 2: 36e8. 15 Havel RJ. Disorders of lipid metabolism. In: Beeson PB, McDermott W, Wyngaarden JB, eds. Cecil textbook of medicine. 15th edn. Philadelphia, PA: WB Saunders, 1979: 202e2011. 16 Neviaser JS. Adhesive capsulitis of the shoulder. J Bone Joint Surg 1945; 27: 211e21. 17 Simonds FA. Shoulder pain: with particular reference to the frozen shoulder. J Bone Joint Surg 1949; 31(B): 426e32. 18 DePalma AF. Loss of scapulohumeral motions (frozen shoulder). Ann Surg 1952; 135: 194e204. 19 Neer CS, Saterlee CC, Dalsey RM, Flatow EL. The anatomy and potential effects of contracture of the coracohumeral ligament. Clin Orthop 1992; 280: 182e5. 20 Ozaki J, Nakagawa Y, Sakurai G, Tamai S. Recalcitrant chronic adhesive capsulitis of the shoulder. J Bone Joint Surg 1989; 71A: 1511e5.

ORTHOPAEDICS AND TRAUMA 25:1

18

Ó 2011 Elsevier Ltd. All rights reserved.

IMAGING

(iii) Imaging of shoulder pathology

instability, magnetic resonance arthrography (MRA) is the preferred technique. This article will outline the imaging of common causes of the painful shoulder.

James Teh Impingement Impingement can be subdivided into extrinsic and intrinsic causes. Classic primary extrinsic impingement is very common and refers to mechanical rotator cuff impingement by the structures that combine to form the coraco-acromial arch. There is a vicious cycle of tendon microtrauma, resulting in tendon swelling and bursitis, and in turn further impingement, which may eventually result in a tendon tear. Impingement syndrome is a clinical diagnosis characterized by pain and limitation of motion. There may be difficulty reaching up behind the back, pain with overhead use of the arm and weakness. Neer described three stages of extrinsic impingement1: Stage 1 is generally encountered in younger patients, with reversible oedema and haemorrhage in the rotator cuff. Stage 2 is characterized by tendinopathy. Stage 3 is generally encountered in older patients, with rotator cuff tears. Most cases of impingement syndrome settle with conservative management, including physiotherapy and subacromial steroid injections. Intrinsic or internal impingement results from the impingement of the rotator cuff and joint capsule on the glenoid, or between the glenoid and the humerus2 (Figure 1). Often it occurs due to repetitive overhand activity such as throwing. It is believed that varying degrees of glenohumeral instability, posterior capsular contracture, and scapular dyskinesis may play a role in the development of symptomatic internal impingement. The role of imaging in impingement syndrome varies according to local practice. Imaging is often reserved for cases that do not respond to conservative therapy, if a tear is suspected or if the diagnosis is uncertain. Normal clinical practice is to offer arthroscopic subacromial decompression for cases that do not settle with conservative measures. This involves shaving the undersurface of the acromion and dividing the coraco-acromial ligament. The treatment of rotator cuff tears is more controversial. Tears may be encountered in the asymptomatic population, so clinical correlation is essential. The symptoms from rotator cuff tears may respond to injections or subacromial decompression alone, and there is little good evidence that routine rotator cuff repair improves outcomes.3 In particular the repair of chronic or large cuff tears is often unrewarding. The choice of imaging modality for the assessment of impingement syndrome will depend on local practice. Plain radiographs are often performed as the initial imaging modality as this may show rotator cuff calcific tendinopathy, subacromial spurs, a downward sloping acromion, loss of the acromiohumeral distance (less than 7 mm suggests a large rotator cuff tear) or glenohumeral arthropathy. Direct visualization of the rotator cuff requires MRI or ultrasound, however. MR arthrography is considered the most sensitive and specific technique for diagnosing both full and partial thickness rotator cuff tears, but requires an invasive procedure, and so is not routinely used for evaluating rotator cuff pathology. Ultrasound and conventional MRI are accurate techniques for diagnosing cuff tears and are comparable in both sensitivity and specificity.4 Ultrasound has the advantages of being a quick and

Abstract The shoulder is a complex joint with numerous structures contributing to mobility and stability. Shoulder pain is a common clinical complaint that may be due to a wide spectrum of disorders, including rotator cuff disease, instability and arthropathy. Conditions extrinsic to the joint such as cervical spine disease, neurovascular disease and neoplasm may also mimic shoulder pathology. Imaging therefore plays an important role in the assessment and management of suspected shoulder problems. This article will outline the imaging of common causes of the painful shoulder.

Keywords Impingement; instability; MRI; pathology; rotator cuff tear; shoulder; ultrasound

Introduction The shoulder is a complex joint with numerous structures contributing to mobility and stability. Shoulder pain is a common clinical complaint that may be due to a wide spectrum of disorders, including rotator cuff disease, instability and arthropathy. Conditions extrinsic to the joint such as cervical spine disease, neurovascular disease and neoplasm may also mimic shoulder pathology. Imaging therefore plays an important role in the assessment and management of suspected shoulder problems. Plain radiographs remain the first line modality for the evaluation of the painful shoulder. They are particularly valuable in the context of demonstrating fractures and dislocations, detecting calcium deposits and evaluating bony lesions and osteoarthritis. However, plain radiographs are poor at detecting the soft tissue damage following instability and have a limited role in assessing rotator cuff pathology. Computed tomography (CT) is the gold standard for detailed assessment of bony injury, and has a key role in pre-operative planning for trauma. Ultrasound is an operator-dependent technique that allows detailed, dynamic assessment of the rotator cuff complex, biceps tendon and soft tissue masses. It may also be used for assessing joint effusions and for performing injections under imaging guidance. Ultrasound cannot assess the glenoid labrum and gives only limited bony detail. Its use is inevitably dictated by the availability of local expertise. Magnetic resonance imaging (MRI) has revolutionized the imaging of shoulder pathology. Conventional (nonarthrographic) MRI allows direct visualization of most of the important structures of the shoulder but its ability to delineate the glenoid labrum and glenohumeral ligaments is limited when there is no effusion. Therefore, for evaluating shoulder

James Teh BSc MBBS MRCP FRCR Consultant Musculoskeletal Radiologist, Radiology Department, Nuffield Orthopaedic Centre, Windmill Road, Headington, Oxford OX3 7LD, United Kingdom.

ORTHOPAEDICS AND TRAUMA 25:1

19

Ó 2010 Elsevier Ltd. All rights reserved.

IMAGING

Figure 1 Internal impingement syndrome. Axial proton density fat-saturated sequence demonstrating cystic change in the postero-superior humeral head (arrow).

dynamic technique for evaluating the rotator cuff. It also allows guided injection of the subacromial space. It is therefore particularly useful in the setting of one-stop clinics. MRI has the advantage of allowing evaluation of the bony anatomy and glenohumeral joint. However MRI is expensive, time-consuming and is often disliked by patients. Furthermore, in many institutions it is not readily available. If the clinical question is simply whether or not there is a cuff tear, the preferred imaging modality at the author’s institution is ultrasound.

Figure 2 Small full thickness supraspinatus tear. Coronal ultrasound demonstrating loss of the fibrillar echotexture and loss of normal convexity (arrowheads) of the anterior leading edge of supraspinatus.

MRI Although the gold standard for the imaging diagnosis of a full thickness rotator cuff tear is to perform an MR arthrogram, noncontrast MRI is usually sufficient to make an accurate diagnosis.4 The shoulder is typically scanned with the patient in the supine position with the arm at the side in the neutral or slightly externally rotated position. It is recommended that images should be obtained in axial, coronal oblique and sagittal oblique planes. On MRI the most reliable sign of a full thickness tear is fluid signal on a T2 fat-saturated or STIR sequence, extending from the superior to inferior surface of the tendon8,9 (Figure 7). Less intense high T2 signal may be seen with tendon degeneration. Massive cuff tears manifest as a large gap in the cuff with the retracted tendon end seen beneath the acromion (Figure 8). Partial thickness tears manifest as focal areas of high T2 (fluid) signal extending into a defect on either the articular or bursal

Rotator cuff tears Ultrasound Rotator cuff tears are characterized on the basis of size and location. Small tears measure less than 1 cm, medium tears measure 1e3 cm, large tears measure 3e5 cm and massive tears measure more than 5 cm. On ultrasound a full thickness rotator cuff tear may appear as discontinuity of the fibres with a fluid filled gap.5 There may be no significant fluid, and loss of the convexity of the superior surface may be the key feature of a tear (Figure 2). Full thickness tears most often occur at the anterior leading edge of supraspinatus, but may occur in the midsubstance (Figure 3). The tear may extend to involve the infrapinatus and subscapularis tendons. With massive rotator cuff tears there is retraction of the tendon, with a bare humeral head lying against the overlying deltoid (Figure 4). Tears involving the subscapularis tendon may be associated with damage to the transverse humeral and/or coracohumeral ligament, which holds the long head of biceps in situ within the humeral groove. Injury to this biceps pulley complex may result in medial subluxation of the long head of biceps (Figure 5). Partial thickness tears are more difficult to diagnose than full thickness tears on ultrasound.6 There is loss of continuity of fibres with focal low echogenicity within the substance of the tendon, which does not extend the full thickness of the tendon (Figure 6). Articular surface tears are more common than bursal surface tears. Usually these are treated non-operatively, but there may be a role for surgery in selected patients particularly in the younger, athletic population. Surgery may involve a combination of debridement of the partial tear, acromioplasty and/or repair of the cuff.7

ORTHOPAEDICS AND TRAUMA 25:1

Figure 3 Small full thickness supraspinatus tear. Sagittal ultrasound image demonstrating a low signal fluid gap at the anterior leading edge of supraspinatus (SST). The supraspinatus tendon should lie adjacent to the long head of biceps (LHB).

20

Ó 2010 Elsevier Ltd. All rights reserved.

IMAGING

Figure 4 Massive cuff tear. Coronal ultrasound image demonstrating a bare humeral head, covered by the deltoid muscle. The region of the torn cuff contains minor fluid (arrowheads). The retracted supraspinatus is shown (arrow).

Figure 6 Articular surface partial tear of supraspinatus. Sagittal ultrasound image demonstrating a low echogenicity defect (arrowheads) in the anterior leading edge of supraspinatus, adjacent to the long head of biceps (LHB).

surfaces. There may be increased joint or bursal fluid respectively. The ‘‘rim rent’’ tear describes a small articular surface partial tear at the insertion of supraspinatus (Figure 9). The accuracy of MRI for diagnosing partial thickness tears is not as good as for diagnosing full thickness tears.10 MRI has the advantage over US of allowing assessment of fatty atrophy of the muscles following a rotator cuff tear, the presence of which has been correlated with poor outcome following cuff repair11 (Figure 10).

accurate than US for the diagnosis of cuff tendinosis. On a STIR or T2 fat-saturated sequence there is focal or diffuse increased signal (but not fluid signal, which would indicate a tear) within the tendon substance (Figure 11). The tendon may also appear swollen. There may be associated subacromial bursitis or fluid. On US there is focal or diffuse decreased reflectivity in the tendon with loss of the normal fibrillar structure, with tendon swelling. Subacromial bursitis may be evident with bunching of the bursa on dynamic scanning.

Rotator cuff tendinopathy Rotator cuff tendon degeneration or tendinosis is a common occurrence, usually associated with impingement syndrome. It may also be considered part of the normal ageing process. At a microscopic level there are microtears which undergo a process of healing by fibrosis and regeneration of fibres, with an associated angiofibroblastic response. MRI is considered more

Intrinsic impingement Certain imaging features are characteristic of intrinsic impingement. These include injury to the posteroesuperior glenoid

Figure 5 Medial dislocation of long head of biceps. Axial ultrasound image demonstrating medial dislocation of the long head of biceps (arrowheads) from the humeral groove (arrow). There was disruption of the subscapularis tendon.

ORTHOPAEDICS AND TRAUMA 25:1

Figure 7 Small full thickness supraspinatus tear. Coronal oblique T2 fat-saturated image demonstrating a fluid filled gap in the supraspinatus tendon.

21

Ó 2010 Elsevier Ltd. All rights reserved.

IMAGING

Figure 10 Fatty atrophy of the supraspinatus muscle belly. Coronal oblique T1-weighted image demonstrating fatty atrophy of the supraspinatus muscle belly associated with a cuff tear.

Figure 8 Massive supraspinatus tear. Coronal oblique STIR image demonstrating a bare humeral head with retraction of the supraspinatus tendon. There is fluid communicating between the glenohumeral joint and subacromial space.

with severe pain and localized redness and swelling, which may be mistaken for infection. On plain radiographs there is discrete or cloudy calcific deposition in the rotator cuff (Figure 12). An axial view is often required to demonstrate involvement of the infraspinatus and subscapularis tendons. On US calcific deposits appear as echogenic foci with posterior acoustic shadowing12 (Figure 13). The tendon may be swollen, resulting in impingement, and subacromial bursitis is commonly associated. Occasionally calcific material extrudes into the

labrum with chondral defects and subchondral cysts at the posteroesuperior aspect of the humeral head.2 The rotator cuff articular surface fibres degenerate and may tear at the junction of supraspinatus with infraspinatus. On MRI, contact of the posteroesuperior glenoid labrum with the rotator cuff may be seen on the abducted, externally rotated position (ABER view).

Calcific tendinopathy Calcific tendinopathy of the rotator cuff is a painful disorder of unknown aetiology, involving the intra-tendinous deposition of hydroxyapatite crystals. It most commonly affects the supraspinatus tendon, followed by the infraspinatus tendon. The subscapularis tendon is rarely involved. Patients may present

Figure 11 Supraspinatus tendinosis. Coronal oblique proton density fatsaturated sequence demonstrating a thickened tendon with increased signal (arrowheads). Note that there is osteoarthritis of the acromioclavicular joint with inferior marginal osteophyte formation that may be causing impingement (arrow).

Figure 9 Rim rent (partial articular surface) tear. Coronal oblique STIR sequence demonstrates a small focus of fluid signal at the articular surface of the supraspinatus insertion (arrow).

ORTHOPAEDICS AND TRAUMA 25:1

22

Ó 2010 Elsevier Ltd. All rights reserved.

IMAGING

primarily clinical with patients presenting with pain and progressive limitation of motion. In most idiopathic cases the condition is self-limiting, but may have a prolonged course over 2e3 years.15 Imaging is not routinely required, but is often requested to exclude underlying arthropathy or rotator cuff pathology, which may present in a similar fashion. On MRI there is often low signal thickening of the joint capsule particularly in the region of the anterior interval and axillary recess, with thickening of the coracohumeral ligament16 (Figure 14). The fat in the subcoracoid space may also be obliterated. If arthrography is attempted, the joint is often found to be of diminished volume. On ultrasound, thickening of the coracohumeral ligament and hypoechoic texture in the anterior interval may be present, in addition to increased vascularity on Doppler17 (Figure 15). The management of adhesive capsulitis remains controversial. Suggested treatments include physiotherapy, oral steroids subacromial injections, joint injections, joint distension, manipulation under anaesthetic and surgery.15,18

Figure 12 Calcific tendinosis of the rotator cuff. Plain radiograph demonstrating rotator cuff calcification (arrow).

subacromial space. Ultrasound guided barbotage involves injection of the subacromial space with local anaesthetic and steroid, followed by needling of the calcific deposits.13 If the calcium is immature, percutaneous irrigation and lavage of the calcific deposits can also be performed. This procedure has been shown to be as effective as arthroscopic treatment.14 On MRI calcific deposits usually appear as focal areas of signal void, and may therefore be difficult to appreciate on a background of a low signal tendon. If suspected on an MRI scan, plain radiographs or ultrasound should be performed.

Instability Glenohumeral instability refers to subluxation or dislocation of the humeral head in relation to the glenoid fossa. There are several classifications for describing instability of the shoulder. These take into account the underlying cause and direction. Essentially two broad categories exist. Traumatic, unidirectional instability with a Bankhart lesion requiring surgery (TUBS) and atraumatic multidirectional instability with bilateral laxity helped by rehabilitation or capsulorrhaphy of the inferior capsule (AMBRI).19

Frozen shoulder

Unidirectional instability The most common shoulder dislocation is anterior, occurring in over 95% of cases, usually after a fall on the outstretched externally rotated, abducted arm. A spectrum of abnormalities may be seen including injury to the antero-inferior labro-ligamentous structures, an enlarged anterior pouch, a stretched subscapularis and a HilleSachs deformity.20 Imaging has an important role in

Frozen shoulder or adhesive capsulitis is a poorly understood condition, which may occur spontaneously or as a result of trauma, infection or calcific tendinopathy. The diagnosis is

Figure 13 Calcific tendinosis. Coronal ultrasound image demonstrating a focus of high echogenicity (arrow) in the supraspinatus tendon with posterior acoustic shadowing (arrowheads).

ORTHOPAEDICS AND TRAUMA 25:1

Figure 14 Adhesive capsulitis. Sagittal T1-weighted MRI scan demonstrating thickening of the coracohumeral ligament (arrowheads) and obliteration of the subcoracoid fat.

23

Ó 2010 Elsevier Ltd. All rights reserved.

IMAGING

Figure 15 Adhesive capsulitis. Sagittal ultrasound image demonstrating increased vascularity and thickening of the coracohumeral ligament (arrowheads), in the anterior interval low signal thickening around the long head of biceps (LHB). There is a tear and tendinosis of the anterior portion of supraspinatus (arrow).

Figure 16 Bankart lesion. Axial T1 fat-saturated MR arthrogram demonstrating a detached Bankart lesion (arrow) and a HilleSachs deformity (arrowheads).

to the glenoid articular cartilage. The cartilaginous injury may range from a flap tear to an osteochondral impaction fracture. This injury often occurs as a result of impaction of the humeral head against the glenoid rather than a dislocation.

guiding stabilization surgery and labral repair. The changes that occur following dislocation are best demonstrated with the joint distended on MR arthrography.21 Non-contrast MRI can show the abnormalities particularly if performed in the acute setting (within 4 weeks of injury), when the joint is distended by an effusion or haemarthrosis.22 After the acute phase when the joint is not distended, the accuracy of non-contrast MRI decreases. CT arthrography does not offer the same soft tissue contrast but can be used if MRI cannot be performed.

Capsular and glenohumeral ligament injury Anterior glenohumeral dislocation often results in tears of the glenohumeral ligaments. The inferior glenohumeral ligament is composed of anterior and posterior bands interposed by an axillary pouch. The anterior band is an important stabilizer of the glenohumeral joint, and injury is often associated with clinically evident instability. Humeral avulsion of the inferior glenohumeral ligament is known as the HAGL lesion. This results in inhomogeneity or frank disruption of the anterior capsule with a redundant, J-shaped axillary pouch24 (Figure 20). In 20% of cases there is bony avulsion, known as a BHAGL lesion.

Anteroeinferior labral tears The vast majority of anterior labral injuries involve the anteroe inferior portion.23 Occasionally, following anterior dislocation there may be a tear of the entire anterior labrum, extending from the base of the biceps tendon down to the insertion of the inferior glenohumeral ligament. Isolated tears of the anteroesuperior portion of the labrum are very rare and the possibility of normal variation should be considered if fluid signal is seen within the labrum in this region. The soft tissue or fibrous Bankart lesion is avulsion of the labro-ligamentous complex from the anteroeinferior labrum with disruption of the scapular periosteum (Figure 16). A glenoid rim fracture is termed an osseous Bankart lesion (Figure 17). The HilleSachs deformity describes an impaction injury of the posteroesuperior humeral head, resulting in a hatchet shaped defect or flattening of the humeral head. Several Bankart variants have been described, when the periosteum is not fully detached:  The Perthes lesion is an avulsed anterior antero-inferior labrum with medial stripping of the scapular periosteum, which remains attached at its base (Figure 18).  The ALPSA lesion is an anterior labro-ligamentous periosteal sleeve avulsion injury characterized by medial displacement and retraction of the torn anterior labrum, which is like a rolled up sleeve (Figure 19).  The GLAD lesion is a glenoid labral articular disruption, characterized by a torn anteroeinferior labrum that is held in situ by an intact scapular periosteum, with an associated injury

ORTHOPAEDICS AND TRAUMA 25:1

SLAP lesions The superior labral anteroeposterior (SLAP) lesion is a superior labral tear that extends anterior and posterior to the biceps

Figure 17 Bankart lesion. Coronal oblique CT reformat demonstrating an osseous Bankart lesion (arrow).

24

Ó 2010 Elsevier Ltd. All rights reserved.

IMAGING

Figure 18 Perthes lesion. Axial T1 fat-saturated MR arthrogram demonstrating a detached anteroeinferior labrum with periosteal stripping that remains attached at the base.

Figure 20 Humeral avulsion of the glenohumeral ligament (HAGL). Coronal oblique T1 fat-saturated MR arthrogram demonstrates an irregular and redundant inferior pouch.

anchor.25 SLAP lesions usually occur due to repetitive overhead throwing activity, as seen in baseball, basketball and American football. MR arthrography is the gold standard for the diagnosis of SLAP lesions. Linear high signal is seen within the superior labrum extending laterally away from the glenoid rim (Figure 21). Concurrent shoulder injuries are often present including rotator cuff tears, marrow oedema in the humeral head, capsular laxity, HilleSachs or Bankart lesion.26 Many variants of SLAP lesion have been described, according to the involvement of the biceps anchor and labrum. The main types are: Type 1 superior labral degeneration and fraying. Type 2 avulsion of the superior labrum with the biceps anchor. Type 3 inferiorly displaced bucket handle tear of the superior labrum.

Type 4 extension of the bucket handle tear to involve the biceps anchor. Rather than concentrating on categorization, a descriptive analysis of the injury is considered more important. Paralabral cysts A paralabral cyst is a loculated fluid collection that arises from extrusion of joint fluid through a labral tear. The cyst may extend into the spinoglenoid or suprascapular notch resulting in entrapment neuropathy of the suprascapular nerve, with corresponding denervation and muscle wasting of the infraspinatus. The muscle wasting appears as fatty infiltration on T1-weighted sequences, with high signal seen on STIR or T2 fat-saturated sequences indicating oedema27,28 (Figure 22).

Figure 19 Anterior labro-ligamentous periosteal sleeve avulsion (ALPSA). Axial gradient echo sequence in acute setting following anterior shoulder dislocation demonstrating irregularity of the anteroeinferior labrum with bunching up of a periosteal sleeve avulsion (arrow). A large effusion is present.

ORTHOPAEDICS AND TRAUMA 25:1

Figure 21 SLAP tear. Coronal oblique T1 fat-saturated MR arthrogram demonstrating contrast in the superior labrum extending away from the glenoid (arrow).

25

Ó 2010 Elsevier Ltd. All rights reserved.

IMAGING

Figure 22 Spinoglenoid notch ganglion. Axial T2-weighted image demonstrating a fluid signal mass in the spinoglenoid notch (arrowhead) with fatty atrophy of the infraspinatus tendon (arrow).

Normal variants There are several anatomical variants seen on MR arthrography that may be mistaken for pathology.29  A sublabral foramen is located at the 2 O’clock position, anterior to the biceps anchor and represents localized detachment of the labrum from the glenoid rim. This may be difficult to distinguish from a SLAP lesion. Increased distance between the labrum and the glenoid, an irregular appearance of the labral margin, or lateral extension of the separation may suggest an SLAP lesion rather than a normal anatomic variant.  The sublabral cleft is a normal synovial reflection between the cartilage of the glenoid cavity and the superior labrum. The cleft is smooth and tapering and points towards the

Figure 24 Buford complex (normal variant). Axial T1 fat-saturated MR arthrogram showing a small anterior labrum (arrowhead) with a large cord-like middle glenohumeral ligament (arrow), which should not be mistaken for a detached labrum.

glenoid, as opposed to a SLAP lesion that usually points away from the glenoid (Figure 23).  The Buford complex refers to an absent or small anteroe superior labrum with a cord-like thickened middle glenohumeral ligament. This appearance may simulate an avulsed anterior labrum (Figure 24).  Three types of anterior capsular insertion have been described. Type 1 inserts onto the anterior margin of the glenoid. Type 2 inserts onto the neck of the glenoid and type 3 inserts onto the

Figure 23 Sublabral recess. Coronal obliqueT1 fat-saturated MR arthrogram demonstrating a sublabral recess (arrow). Note how the recess points towards the glenoid rather than away, unlike a SLAP tear. The biceps anchor is indicated by the arrowhead.

ORTHOPAEDICS AND TRAUMA 25:1

Figure 25 Anterior capsular stripping. Axial T1 fat-saturated MR arthrogram demonstrating stripping of the anterior capsule (arrowheads) which is attached to the body of the scapula (arrow) with a large redundant anterior joint recess.

26

Ó 2010 Elsevier Ltd. All rights reserved.

IMAGING

Figure 26 ParsonageeTurner syndrome (acute brachial neuritis). Axial T2 fat-saturated sequence demonstrating marked oedema in the infraspinatus (arrowheads). No mass lesion was demonstrated to indicate compression of the suprascapular nerve.

Figure 28 Haemophilic arthropathy. Coronal STIR sequence demonstrating moderate glenohumeral arthropathy with joint space loss (arrowheads). There is low signal synovial deposition in the axillary pouch in keeping with haemosiderin (arrow).

scapula more than 1 cm from the labral base. Studies have shown that the type of insertion does not correlate well with the presence of instability,30 however it is recognized that type 3 insertions may be due to previous capsular stripping following anterior dislocation31 (Figure 25).

atrophy of the musculature primarily innervated by the suprascapular nerve32 (Figure 26). Glenohumeral joint synovitis Shoulder pain resulting from synovitis is a common complaint in patients with inflammatory arthropathy and patients on haemodialysis. Joint effusions and synovitis are readily diagnosed on MRI.33 The presence of fluid signal distending the glenohumeral joint may represent an effusion or synovial thickening. Unless intravenous contrast is administered it may be difficult to differentiate between these, and often both are present. Common

Acute brachial neuritis Acute brachial neuritis or ParsonageeTurner syndrome is an acute painful neuromuscular condition characterized by involvement of the brachial plexus and shoulder girdle. The aetiology remains obscure but viral infection has been suggested. Men are more commonly affected than women. The typical presentation is of severe sudden onset of pain followed by weakness. Most cases resolve spontaneously although there may be residual weakness. On MRI there is muscle oedema and

Figure 29 Avascular necrosis of the humeral head. Coronal proton density fat-saturated image demonstrating the double line sign (arrows) with early subchondral collapse. There is surrounding bone oedema indicating an unstable lesion.

Figure 27 Glenohumeral joint synovitis. Coronal STIR sequence demonstrating effusion and synovitis of the glenohumeral joint (arrows) with erosions of the greater tuberosity (arrowhead).

ORTHOPAEDICS AND TRAUMA 25:1

27

Ó 2010 Elsevier Ltd. All rights reserved.

IMAGING

REFERENCES 1 Neer 2nd CS. Impingement lesions. Clin Orthop Relat Res; 1983: 70e7. 2 Giaroli EL, Major NM, Higgins LD. MRI of internal impingement of the shoulder. AJR Am J Roentgenol 2005; 185: 925e9. 3 Oh LS, Wolf BR, Hall MP, Levy BA, Marx RG. Indications for rotator cuff repair: a systematic review. Clin Orthop Relat Res 2007; 455: 52e63. 4 de Jesus JO, Parker L, Frangos AJ, Nazarian LN. Accuracy of MRI, MR arthrography, and ultrasound in the diagnosis of rotator cuff tears: a meta-analysis. AJR Am J Roentgenol 2009; 192: 1701e7. 5 Papatheodorou A, Ellinas P, Takis F, Tsanis A, Maris I, Batakis N. US of the shoulder: rotator cuff and non-rotator cuff disorders. Radiographics 2006; 26: e23. 6 Kolla S, Motamedi K. Ultrasound evaluation of the shoulder. Semin Musculoskelet Radiol 2007; 11: 117e25. 7 Fukuda H. The management of partial-thickness tears of the rotator cuff. J Bone Joint Surg Br 2003; 85: 3e11. 8 Zlatkin MB, Iannotti JP, Roberts MC, et al. Rotator cuff tears: diagnostic performance of MR imaging. Radiology 1989; 172: 223e9. 9 Farley TE, Neumann CH, Steinbach LS, Jahnke AJ, Petersen SS. Fullthickness tears of the rotator cuff of the shoulder: diagnosis with MR imaging. AJR Am J Roentgenol 1992; 158: 347e51. 10 Traughber PD, Goodwin TE. Shoulder MRI: arthroscopic correlation with emphasis on partial tears. J Comput Assist Tomogr 1992; 16: 129e33. 11 Gladstone JN, Bishop JY, Lo IK, Flatow EL. Fatty infiltration and atrophy of the rotator cuff do not improve after rotator cuff repair and correlate with poor functional outcome. Am J Sports Med 2007; 35: 719e28. 12 Martinoli C, Bianchi S, Prato N, et al. US of the shoulder: non-rotator cuff disorders. Radiographics 2003; 23: 381e401. quiz 534. 13 del Cura JL, Torre I, Zabala R, Legorburu A. Sonographically guided percutaneous needle lavage in calcific tendinitis of the shoulder: shortand long-term results. AJR Am J Roentgenol 2007; 189: W128e34. 14 Maugars Y, Varin S, Gouin F, et al. Treatment of shoulder calcifications of the cuff: a controlled study. Joint Bone Spine 2009; 76: 369e77. 15 Buchbinder R, Green S, Youd JM, Johnston RV, Cumpston M. Arthrographic distension for adhesive capsulitis (frozen shoulder). Cochrane Database Syst Rev; 2008. CD007005. 16 Mengiardi B, Pfirrmann CW, Gerber C, Hodler J, Zanetti M. Frozen shoulder: MR arthrographic findings. Radiology 2004; 233: 486e92. 17 Lee JC, Sykes C, Saifuddin A, Connell D. Adhesive capsulitis: sonographic changes in the rotator cuff interval with arthroscopic correlation. Skeletal Radiol 2005; 34: 522e7. 18 Levine WN, Kashyap CP, Bak SF, Ahmad CS, Blaine TA, Bigliani LU. Nonoperative management of idiopathic adhesive capsulitis. J Shoulder Elbow Surg 2007; 16: 569e73. 19 Thomas SC, Matsen 3rd FA. An approach to the repair of avulsion of the glenohumeral ligaments in the management of traumatic anterior glenohumeral instability. J Bone Joint Surg Am 1989; 71: 506e13. 20 Zlatkin MB, Bjorkengren AG, Gylys-Morin V, Resnick D, Sartoris DJ. Cross-sectional imaging of the capsular mechanism of the glenohumeral joint. AJR Am J Roentgenol 1988; 150: 151e8. 21 Chandnani VP, Yeager TD, DeBerardino T, et al. Glenoid labral tears: prospective evaluation with MRI imaging, MR arthrography, and CT arthrography. AJR Am J Roentgenol 1993; 161: 1229e35. 22 Wintzell G, Haglund-Akerlind Y, Tengvar M, Johansson L, Eriksson E. MRI examination of the glenohumeral joint after traumatic primary anterior dislocation. A descriptive evaluation of the acute lesion and at 6-month follow-up. Knee Surg Sports Traumatol Arthrosc 1996; 4: 232e6.

associations include bursitis, tenosynovitis and tendinosis. As there is a normal communication between the joint and the long head of biceps tendon sheath, the latter may be distended when there is a joint effusion. Joint erosions manifest as well-defined high subarticular lesions on STIR or T2 fat-saturated sequences, often with surrounding bone oedema. The erosions are of corresponding low signal on T1-weighted images (Figure 27). In many cases the finding of synovitis is non-specific, but if there is low signal on STIR or T2-weighted sequences within the synovium, the presence of haemosiderin deposition should be considered, raising the possibility of pigmented villonodular synovitis or other causes of recurrent haemorrhage such as haemophilia (Figure 28). In patients with renal failure amyloid deposition disease may also result in a similar appearance. The presence of multiple loose bodies within the joint raises the possibility of synovial osteochondromatosis. Ultrasound with Doppler interrogation has been shown to reliably diagnose shoulder synovitis and humeral head erosions34 but MRI offers a more complete evaluation. Ultrasound maintains an important role in guiding aspiration of the joint in cases of suspected infection or crystal arthropathy. Avascular necrosis Avascular necrosis (AVN) is not a specific disease entity but the final common pathway of a number of conditions leading to bone death. There are many predisposing factors, including trauma, corticosteroids, alcoholism and connective tissue disorders but it is often idiopathic. The humeral head is the second most affected site following the femoral head. MRI has an important role in the early detection of AVN, which may not be visible on plain radiographs. In recent years the role of bone scintigraphy has diminished significantly with the increased availability and greater accuracy of MRI. The typical appearance of early AVN is of a well-defined geographical area of subchondral abnormality, with associated bone oedema. As the lesion progresses the ‘‘double line sign’’ may be seen on T2-weighted images, which represents a high signal line of granulation tissue surrounded by a low signal line of fibrosis and bone sclerosis. Eventually there is subchondral collapse and secondary osteoarthritis. The presence of bone oedema with a double line sign (as opposed to no bone oedema) has been shown in one study to have a worse prognosis as judged by progression to total hip arthroplasty35 (Figure 29). A

Suggested imaging protocol for shoulder pain.

ORTHOPAEDICS AND TRAUMA 25:1

28

Ó 2010 Elsevier Ltd. All rights reserved.

IMAGING

23 Waldt S, Burkart A, Imhoff AB, Bruegel M, Rummeny EJ, Woertler K. Anterior shoulder instability: accuracy of MR arthrography in the classification of anteroinferior labroligamentous injuries. Radiology 2005; 237: 578e83. 24 Tirman PF, Steinbach LS, Feller JF, Stauffer AE. Humeral avulsion of the anterior shoulder stabilizing structures after anterior shoulder dislocation: demonstration by MRI and MR arthrography. Skeletal Radiol 1996; 25: 743e8. 25 Waldt S, Burkart A, Lange P, Imhoff AB, Rummeny EJ, Woertler K. Diagnostic performance of MR arthrography in the assessment of superior labral anteroposterior lesions of the shoulder. AJR Am J Roentgenol 2004; 182: 1271e8. 26 Chang D, Mohana-Borges A, Borso M, Chung CB. SLAP lesions: anatomy, clinical presentation, MR imaging diagnosis and characterization. Eur J Radiol 2008; 68: 72e87. 27 Tirman PF, Feller JF, Janzen DL, Peterfy CG, Bergman AG. Association of glenoid labral cysts with labral tears and glenohumeral instability: radiologic findings and clinical significance. Radiology 1994; 190: 653e8. 28 Tung GA, Entzian D, Stern JB, Green A. MR imaging and MR arthrography of paraglenoid labral cysts. AJR Am J Roentgenol 2000; 174: 1707e15.

ORTHOPAEDICS AND TRAUMA 25:1

29 De Maeseneer M, Van Roy F, Lenchik L, et al. CT and MR arthrography of the normal and pathologic anterosuperior labrum and labralebicipital complex. Radiographics 2000; 20(Spec No): S67e81. 30 Palmer WE, Caslowitz PL. Anterior shoulder instability: diagnostic criteria determined from prospective analysis of 121 MR arthrograms. Radiology 1995; 197: 819e25. 31 Shankman S, Bencardino J, Beltran J. Glenohumeral instability: evaluation using MR arthrography of the shoulder. Skeletal Radiol 1999; 28: 365e82. 32 Gaskin CM, Helms CA. ParsonageeTurner syndrome: MR imaging findings and clinical information of 27 patients. Radiology 2006; 240: 501e7. 33 Hermann KG, Backhaus M, Schneider U, et al. Rheumatoid arthritis of the shoulder joint: comparison of conventional radiography, ultrasound, and dynamic contrast-enhanced magnetic resonance imaging. Arthritis Rheum 2003; 48: 3338e49. 34 Bruyn GA, Naredo E, Moller I, et al. Reliability of ultrasonography in detecting shoulder disease in patients with rheumatoid arthritis. Ann Rheum Dis 2009; 68: 357e61. 35 Mitchell MD, Kundel HL, Steinberg ME, Kressel HY, Alavi A, Axel L. Avascular necrosis of the hip: comparison of MR, CT, and scintigraphy. AJR Am J Roentgenol 1986; 147: 67e71.

29

Ó 2010 Elsevier Ltd. All rights reserved.

MINI-SYMPOSIUM: SHOULDER

(iv) Disorders of the acromioclavicular joint and distal clavicle

Surgical anatomy The acromioclavicular joint is a diarthrodial joint containing a fibrocartilaginous disc, which may be incomplete. Stability is achieved by static and dynamic structures. Static stability in the horizontal plane is provided by the superior acromioclavicular ligament (which blends with the fibres of the deltoid and trapezius muscles). The inferior, anterior and posterior acromioclavicular ligaments are all condensations of the joint capsule (Figure 1). The coracoclavicular ligament confers considerable stability to the ACJ, predominantly in the vertical plane. This ligament has fibres passing from the outer inferior surface of the clavicle to the base of the coracoid process. The coracoclavicular ligament consists of two portions; the conoid and the trapezoid. The conoid is cone shaped, with its apex attaching on the posteromedial side of the coracoid base. The base of the conoid attaches to the conoid tubercle on the under surface of the clavicle. The trapezoid arises anterior and lateral to the attachment of the conoid on the coracoid. The insertion on the clavicle is similarly anterior and lateral to the conoid insertion. The trapezius and deltoid provide an element of dynamic stability. Considerable variation in the osseous configuration of the joint has been observed.1 Cadaveric, CT and radiological analysis have demonstrated that the joint may be classified as flat, oblique or curved in both the transverse and coronal planes.

A P Wright I A R MacLeod S C Talwalker

Abstract The acromioclavicular joint (ACJ) is a common source of pathology in patients of all age groups and lifestyles. Young patients, particularly those engaged in contact sports, often suffer injuries that result in instability of the ACJ. Advances in arthroscopic surgery and implant technology have opened new avenues for the treatment of these injuries. Despite this, in many cases, there remains considerable controversy as to whether surgical intervention is warranted in the acute setting. Fractures of the distal clavicle may occur in any age group, and have a reputation for having a high incidence of non-union. The development of anatomically contoured locking plates offers the potential for improved outcomes in this difficult group of fractures. The ACJ is a common site for degenerative change and advances in arthroscopic techniques have changed the way this condition is managed. The aim of this article is to give the general reader an overview of the current understanding of these conditions and provide an up-to-date account of the treatment options available.

Evaluation of ACJ conditions Although subluxation and dislocation are by far the most common conditions affecting the ACJ, it is also prone to all conditions involved within a surgical sieve. Degenerative or post-traumatic arthropathy, inflammatory arthropathy, crystal arthropathy, osteolysis and neoplastic conditions must all be considered as part of a differential diagnosis requiring a history, examination and appropriate investigation to confirm the diagnosis.

Keywords acromioclavicular joint; osteoarthritis; osteolysis; shoulder arthroscopy; sports injuries

Acromioclavicular joint separation Introduction

Injuries to the ACJ represent 12% of all shoulder girdle injuries. They commonly result from high-energy impact to the point of the shoulder, as frequently occurs in contact sports. This forces the acromion downwards and medially. There is a strong male preponderance and injuries are most common in the first three decades of life. The acromioclavicular ligaments, coracoclavicular ligaments and deltopectoral fascia may fail either alone or in combination. With a complete injury to all soft tissue constraints the distal end of the clavicle may “button hole” superiorly, posteriorly, or inferiorly. These injuries are generally clinically obvious, with pain and tenderness localized to the ACJ. The apparent deformity may vary with respect to the degree and direction of displacement but it is predominantly a downward displacement of the scapula. Plain radiographs are sufficient to confirm the diagnosis, although weight-bearing views are sometimes helpful to determine the degree of instability. Several classification systems exist for ACJ dislocations. The most widely accepted is that developed by Rockwood (Figure 2). Treatment of acromioclavicular injuries depends on whether they are incomplete or complete. The vast majority of Type I and Type II injuries require no surgical intervention, however, there

Disorders of the acromioclavicular joint (ACJ) are common, and may affect individuals at any age. Advances in implant technology have created new solutions for conditions that have historically seldom been treated surgically. This article aims to give an overview of conditions affecting the ACJ and distal clavicle, an approach to patient assessment and an appraisal of the available evidence for treatment methods.

A P Wright FRCS(Tr & Orth) Upper Limb Fellow, Upper Limb Unit, Wrightington Hospital, Hall Lane, Appley Bridge, Wigan, Lancashire WN6 9EP, UK. Conflicts of interest: none. I A R MacLeod FRCS(Tr & Orth) Upper Limb Fellow, Upper Limb Unit, Wrightington Hospital, Hall Lane, Appley Bridge, Wigan, Lancashire WN6 9EP, UK. Conflicts of interest: none. S C Talwalker FRCS(Tr & Orth) Consultant Upper Limb Surgeon, Upper Limb Unit, Wrightington Hospital, Hall Lane, Appley Bridge, Wigan, Lancashire WN6 9EP, UK. Conflicts of interest: none.

ORTHOPAEDICS AND TRAUMA 25:1

30

Ó 2010 Published by Elsevier Ltd.

MINI-SYMPOSIUM: SHOULDER

Clavicle lar vicu iocla ts m o Acr gamen li

ts

oacro

mial li

game

id ezo p a r T

nt

id no

Corac

amen Coracoclavicular lig

Co

ion om Acr

coid a r Co

Figure 1 Osseo-ligamentous anatomy of the acromioclavicular joint.

remains huge controversy regarding which complete injuries require surgical intervention. There are few if any prospective trials comparing non-operative to operative treatment in two well matched groups.

stabilization of the ACJ and repair of both the acromioclavicular and coracoclavicular ligaments. Outcome was assessed at 3 months and 13 months by the use of a scoring system that stratified results into poor, fair, good and excellent. At 3 months, patients treated conservatively scored better than those treated surgically, although this difference was not seen at 13 months. Bannister et al. randomized 60 patients to either non-operative treatment or open reduction and fixation with a coracoclavicular screw.3 Fifty-eight patients were followed for 4 years. They found that patients treated non-surgically returned to work and sport significantly earlier, and at both 1-year and 4-year follow-up outcomes as determined using the same non-validated scoring system employed by Imatani were better in the conservatively treated group. In the conservatively treated group the initial displacement of the ACJ improved by 5 mm over 12 months. In contrast 35% of patients treated surgically reduction was lost following removal of the coracoclavicular screw. When 12 patients with displacement more than 2 cm were analysed separately, there were a greater number of good or excellent outcomes associated with surgical treatment than conservative treatment.

Type I A type I injury represents a simple sprain of the acromioclavicular ligaments. Clinically, this is manifested as pain and tenderness at the ACJ with normal radiographs. Treatment is symptomatic, with early mobilization as comfort allows. Type II Type II injuries involve a complete tear of the acromioclavicular ligaments, with intact coracoclavicular ligaments. In addition to localized tenderness, there may be slight displacement of the acromion inferiorly. Treatment is the same as for type I injuries. The long-term prognosis is good. Type III This injury involves injury to both the acromioclavicular ligaments and the coracoclavicular ligaments. In addition to localized tenderness there is a clinically obvious deformity manifested as a prominent distal end of clavicle. Radiographs demonstrate overriding of the distal clavicle over the acromion by at least 100% (Figure 3). It is important to realize that the deformity is an inferior displacement of the scapula, rather than superior displacement of the clavicle. The treatment of acute type III injuries is the most controversial aspect of this condition. Most surgeons advocate symptomatic treatment as for type I and II injuries, although some suggest that anatomic reduction and ligament reconstruction is beneficial in the younger, more active patient. Only three randomized or quasirandomized studies exist. Imatani et al. studied 23 patients who were allocated on an alternating basis to non-operative treatment or fixation with open reduction and fixation with either a coracoclavicular screw or Steinman pin fixation across the ACJ.2 Outcome was assessed using a 100-point scale developed for the study. Outcomes were categorized into excellent, good, fair and poor. There were more excellent and good outcomes in the conservatively treated group than in the surgically treated group, although a statistical analysis was not performed. Larsen et al. followed 84 patients who had been randomized to either conservative treatment or open reduction with K-wire

ORTHOPAEDICS AND TRAUMA 25:1

Type IV The acromioclavicular and coracoclavicular ligaments are torn and the lateral end of the clavicle is driven posteriorly through the trapezius muscle and lies in the subcutaneous tissue. An AP radiograph demonstrates subtle widening of the acromioclavicular distance and without careful clinical examination or an oblique X-ray the severity of this injury may be underestimated. Treatment is with open reduction and internal fixation with ligament repair or reconstruction. Type V The acromioclavicular and coracoclavicular ligaments are torn and the lateral end of the clavicle is buttonholed superiorly through the deltotrapezial fascia such that it tents the skin (Figure 4). X-rays demonstrate a dramatically increased coracoclavicular distance. Treatment is the same as for type IV injuries. Type VI This is by far the rarest injury subtype. The lateral end of the clavicle is displaced inferiorly beneath the coracoid and conjoined tendon. These are high-energy injuries and there is an increased risk of co-existing injuries to the brachial plexus,

31

Ó 2010 Published by Elsevier Ltd.

MINI-SYMPOSIUM: SHOULDER

n

AC L L

ACL = acromioclavicular ligament CAL = coracoacromial ligament

oid

CA

Clavicle Tra pez oid

n Co

io om Acr

id

o Co rac

Grade I

Grade II

ius ez p Tra

Grade IV (seen from above)

Grade V

Grade VI

Conjoint tendon

Grade III

Figure 2 Rockwood classification of ACJ dislocation.

vasculature or pleura. Treatment is the same as for type IV injuries.

Treatment of acromioclavicular joint dislocation Non-operative A broad arm sling is used to resist the tendency of the shoulder girdle to sag. This is discontinued as pain settles, which typically takes 7e10 days. It is unusual to require supervised rehabilitation as stiffness is seldom a problem. Surgical treatment of acute dislocations An exhaustive knowledge of the numerous methods for the fixation of the acute ACJ dislocation is not required for the majority of surgeons; however, an understanding of the principles is useful. The patient is positioned in a beach chair position.

Figure 3 Radiograph of type III ACJ dislocation.

ORTHOPAEDICS AND TRAUMA 25:1

32

Ó 2010 Published by Elsevier Ltd.

MINI-SYMPOSIUM: SHOULDER

Surgical treatment of chronic dislocations In the case of a chronic dislocation, there is no potential for healing of the ligamentous complex and thus it must be replaced, either by ligament transfer, allograft or autograft reconstruction. The WeavereDunn reconstruction, described in 1972,6 is perhaps the most well known method for the treatment of long-standing instability of the ACJ. The joint is approached in the same manner as for acute injuries. The lateral 2 cm of the clavicle is resected obliquely, such that the inferior part of the osteotomy overlies the coracoid. The coracoacromial ligament is identified and detached at the acromial end with a thin wafer also taken from the acromion. A heavy non-absorbable suture is placed in the ligament and two drill holes are made in the superior cortex of the lateral end of the clavicle. The suture is passed through the drill holes; the arm is supported to elevate the scapula whilst downward pressure on the clavicle is applied. The suture is tensioned to pull the ligament within the medullary canal of the clavicle and the suture is tied. Excess ligament is trimmed. Although not in the original description, it has become common practice for the ligament transfer to be protected with a coracoclavicular screw or a PDS loop/nylon tape around the coracoid and clavicle. Post-operatively, the arm is supported in a broad arm for 3 weeks during which time only pendular exercises are permitted. Elevation of the arm beyond 90 must be avoided until the coracoclavicular screw (if used) is removed. Dynamic stabilization of the ACJ has been described.7,8 The coracoid process is osteotomized and transferred to the under surface of the clavicle with the attached conjoined tendon. In doing so the conjoined tendon is converted to a depressor of the clavicle. This does not directly address the pathomechanics of ACJ separation in which, rather than the clavicle being displaced superiorly, it is the scapula that descends. Despite this, the technique has been used in both acute and chronic ACJ dislocations with satisfactory results.7,8 Excision of the distal portion of the clavicle has been proposed as a solution to the painful type III ACJ dislocation. In the absence of repair or substitution this fails to resolve symptoms, merely shifting the problem proximally along the clavicle.

Figure 4 View of superior aspect of right shoulder demonstrating type V ACJ dislocation.

The joint is approached via an incision starting just medial to the ACJ and extending inferiorly to the coracoid process. Full thickness skin flaps are raised and the deltopectoral fascia is incised and the deltoid is elevated, exposing the ACJ. Often the coracoclavicular ligaments have failed by peeling off their clavicular attachment complete with a periosteal sleeve, which may be repaired. Once reduced, the ACJ may be held directly with K-wires or a hook plate or indirectly with a coracoclavicular (Bosworth) screw, by passing suture or tape beneath the coracoid and through the clavicle, by securing a suture anchor in the coracoids and tying it around the clavicle or a synthetic ligament secured with an endobutton through the clavicle and coracoid at either end (Figure 5). The shoulder is rested in a broad arm sling for 3 weeks, after which time active mobilization begins. When choosing a method for fixation of the ACJ, it should be considered that any rigid implant will require removal before full mobilization of the shoulder is possible. Arthroscopic repair of acute ACJ dislocations has been described.4,5 The surgical detail is beyond the scope of this article. The general reader should be aware that this technique is technically challenging and, as yet, has no proven benefit over open techniques.

Complications of treatment Complications of treatment of ACJ injuries occur with both nonoperative and operative treatments. Primary concerns with nonoperative treatment include residual deformity and late onset degenerative changes within the joint. There may also be problems related to soft tissues interposed within the joint or joint stiffness due to prolonged immobilization. Problems associated with operative treatment include infection, periprosthetic fracture, and problems related to the implants used, such as migration of the metalwork and the need for subsequent removal prior to mobilization.

Osteolysis of the distal clavicle First described in 1950,9 this is a phenomenon that most commonly results from trauma to the shoulder or as a stress reaction to intolerable levels of exercise, usually in weightlifters. It is characterized by dull ache, weakness and pain in flexion and abduction. In some individuals the ACJ is the weakest point in the shoulder girdle and microfractures or a subchondral fracture can be sustained.

Figure 5 Intra-operative fluoroscopic image demonstrating coracoclavicular reconstruction using the endobutton.

ORTHOPAEDICS AND TRAUMA 25:1

33

Ó 2010 Published by Elsevier Ltd.

MINI-SYMPOSIUM: SHOULDER

intervention. In type II fractures there is disruption of the coracoclavicular ligaments and complete displacement at the fracture site is the norm. The non-union rate has been estimated at 21e25% in this injury group.12,13 In Robinson’s prospective study of 101 patients with a displaced fracture of the lateral end of the clavicle who had a minimum of 6 months non-operative treatment, only 14% went on to require surgery.12 Further analysis suggested no difference in the Constant score between patients with a non-union and those who had united. Several options exist for fixation of a lateral clavicle fracture. If the fracture is to be treated acutely, reduction of the fracture can be secured indirectly by the use of a coracoclavicular screw, PDS loop/ nylon tape around the clavicle and coracoid or a tightrope fixation device. As with ACJ stabilization, there is the potential to perform these techniques arthroscopically. Plate and screw devices may be used as an alternative to or in addition to these methods. The clavicular hook plate (Figure 7) is an option for both distal clavicle fractures and ACJ disruption. The hook is placed posterior to the ACJ, beneath the acromion, whilst the proximal limb of the plate is secured to the diaphysis of the clavicle. Screw holes in the distal portion of the plate allow fixation if bone stock permits. The hook plate provides fixation across the ACJ, even in the most distal or comminuted lateral clavicle fractures. As a result, the plate limits movement across the ACJ and removal is mandatory before elevation or abduction of the shoulder beyond 90 is permitted. With excessive movement in these planes there is a risk that the hook may erode into the acromion. The hook portion poses additional risks by the nature of its proximity to the subacromial space and rotator cuff. A study of 15 cadaver shoulders showed that the subacromial bursa was penetrated in 13 out of 15 shoulders and impingement on the rotator cuff occurred in nine out of 15.14 A retrospective clinical study of 18 patients who had undergone hook plate fixation demonstrated a non-union rate of 22%, and a 28% incidence of acromial osteolysis.15 The mean Constant score was 88.5. Recent developments in locking plate technology have allowed the manufacture of anatomically precontoured locking plates. Small calibre locking screws allow fixation of the distal clavicle fragment without crossing the ACJ. It has been suggested that these implants provide less rigid fixation than the hook plate, and some surgeons

Figure 6 Radiograph demonstrating osteolysis of the distal clavicle following ACJ dislocation.

Radiography may demonstrate localized osteoporosis, tapering, loss of subchondral bone detail, osteophyte formation and cystic changes in varying degrees in the distal clavicle with no bone changes in the acromion (Figure 6). A bone scan demonstrates increased uptake in the region of the distal clavicle. Magnetic resonance imaging may demonstrate oedema within the distal clavicle, fluid in the acromioclavicular joint and cysts or erosions within the distal clavicle.10 This gives rise to bone resorption and localized pain. If unilateral osteolysis is observed a differential diagnosis should include gout, neoplasia and Gorham’s massive osteolysis. When bilateral osteolysis is noted systemic disorders must be considered such as rheumatoid arthritis, hyperparathyroidism and scleroderma. If activity modification fails to bring about relief, surgical treatment takes the form of excision of the distal clavicle.

Fractures of the distal end of the clavicle Distal fractures comprise up to 30% of all clavicular fractures11 and can be classified into types I and II depending upon the integrity of the coracoclavicular ligaments and type III if the fracture extends into the ACJ. In type I fractures the coracoclavicular ligaments are intact and there is sufficient soft tissue stability to ensure that these fractures do not displace and heal uneventfully without surgical

Figure 8 Post-operative radiograph of locking plate fixation of a lateral clavicle fracture supplemented with an endobutton coracoclavicular device.

Figure 7 Post-operative radiograph following hook plate fixation of an ACJ dislocation.

ORTHOPAEDICS AND TRAUMA 25:1

34

Ó 2010 Published by Elsevier Ltd.

MINI-SYMPOSIUM: SHOULDER

direction produces pain, has the greatest positive predictive value for ACJ pathology. Plain radiographs are sufficient to confirm the diagnosis in most cases, although where there is evidence of co-existing glenohumeral joint pathology, ultrasonography or magnetic resonance imaging may be helpful. As with other degenerative musculoskeletal conditions, nonoperative management is appropriate for most patients in the first instance. Simple analgesics and activity modification may be all that is required. Injection of corticosteroid into the ACJ is often employed as a precursor to surgical intervention, either as a diagnostic or therapeutic manoeuvre. Jacob and Sallay investigated the longevity of pain relief following ACJ injection, determining that the mean duration of relief was 20 days, with 81% of patients failing to benefit from long-term pain relief.19 Peri-articular injection, in which the needle misses its target, may be responsible for a proportion of patients who fail to respond to injection. It has been suggested that image-guided injection may greatly improve the accuracy of injection.20 The mainstay of surgical treatment for ACJ osteoarthritis is joint excision. The approach for open surgery is the same as previously described for open fixation of the dislocated ACJ. The deltotrapezial fascia is opened in line with the skin incision and the joint capsule is elevated from the ACJ. The lateral 15 mm of the clavicle is excised using an oscillating saw or osteotome. It is essential to identify the orientation of the articular surface of the lateral end of the clavicle as the plane of the osteotomy must be parallel to this to avoid incomplete excision. Failure to remove residual osteophytes may also give rise to ongoing symptoms. Following soft tissue and skin closure, a sling is applied. Active range of movement exercises are commenced immediately and the sling is discarded as soon as pain permits. Arthroscopic distal clavicle excision is an alternative to open surgery. This is an especially useful tool in cases where there is additional bursal or glenohumeral pathology. The clavicle may be excised via a direct superior portal or an indirect subacromial portal. As with open surgery, it is essential to determine the orientation of the joint and ensure complete excision of all osteophytes. Freedman et al. undertook a small prospective, randomized controlled trial comparing open excision with arthroscopic excision.21 There was a non-significant trend towards better pain relief and function in patients treated with arthroscopic surgery. Intra-articular pathology was seen and treated in 50% of patients in the arthroscopic group.

Figure 9 Osteoarthritis of the ACJ.

advocate augmenting the fixation with a coracoclavicular device, such as the endobutton (Figure 8). These constructs do not have to be routinely removed and unrestricted shoulder movement can be permitted early in the post-operative period.

Osteoarthritis of the acromioclavicular joint Degeneration of the acromioclavicular joint (Figure 9) is common, with deterioration of the meniscus commencing as early as the second decade of life. This leads to progressive degeneration within the joint, with up to 57% of elderly patients exhibiting radiographic evidence of arthritic change.16 Typically, a patient will present with an aching pain on the superior aspect of the shoulder, made worse with heavy lifting or abduction at the glenohumeral joint. Clinically, it can be difficult to distinguish ACJ pain from other sources of pain around the shoulder girdle. In severe cases there may be swelling in the region of the ACJ. Localized tenderness is often present. Typically, abduction beyond 100 at the glenohumeral joint will exacerbate the pain. The cross-arm adduction test involves the affected arm being placed on the opposite shoulder and the examiner applying pressure across the shoulders. Pain in the region of the ACJ indicates a positive test, although this test may also be positive in other shoulder conditions such as posterior capsular tightness, impingement or rotator cuff tears. The O’Brien test was developed to detect labral pathology but has also been used to assess the ACJ.17 The patient holds the arm elevated at 90 and adducted 15 medial to the sagittal plane with the elbow in extension. The arm is internally rotated so that the thumb is pointing towards the floor whilst the examiner applies a downward force on the arm. This manoeuvre is repeated with the arm supinated. The test is considered positive if there is discomfort in the region of the ACJ that is more severe with the arm pronated than supinated. O’Brien found that 52% of patients with a positive test had radiographic evidence of ACJ degeneration. A study by Walton et al.18 determined that the most sensitive test for ACJ pathology is localized tenderness over the joint, although this has a low specificity. In contrast, the O’Brien test was found to have a sensitivity of 16% but a specificity of 90%. The authors suggest that a positive radioisotope scan and a positive Paxinos test, in which the application of direct pressure between the thumb and index finger across the ACJ in an anteroposterior

ORTHOPAEDICS AND TRAUMA 25:1

Other conditions affecting the ACJ As previously mentioned, other conditions must be considered when a patient presents with a painful ACJ, such as rheumatoid arthritis and septic arthritis. Petersson found that the incidence of rheumatoid arthritis affecting the ACJ was related to the severity of the underlying condition, and treatment was successful with ACJ excision and subacromial decompression.22 Septic arthritis is extremely rare but cases have been reported involving Staphylococcus aureus, Streptococcus viridans and tuberculosis. Treatment was successful with a combination of appropriate antibiotics, with excision in some cases.

Conclusion There are a wide variety of conditions that may affect the acromioclavicular joint in any age group. Failure to consider the ACJ

35

Ó 2010 Published by Elsevier Ltd.

MINI-SYMPOSIUM: SHOULDER

10 Kassarjian A, Llopis E, Palmer WE. Distal clavicular osteolysis: MR evidence for subchondral fracture. Skeletal Radiol 2007; 36: 17e22. 11 Robinson CM, Court-Brown CM, McQueen MM, Wakefield AE. Estimating the risk of nonunion following nonoperative treatment of a clavicular fracture. J Bone Joint Surg Am 2004; 86: 1359e65. 12 Robinson CM, Cairns DA. Primary nonoperative treatment of displaced lateral fractures of the clavicle. J Bone Joint Surg Am 2004; 86: 778e82. 13 Nordqvist A, Petersson C, Redlund-Johnell I. The natural course of lateral clavicle fracture. 15 (11e21) year follow-up of 110 cases. Acta Orthop Scand 1993; 64: 87e91. 14 ElMaraghy AW, Devereaux MW, Ravichandiran K, Agur AM. Subacromial morphometric assessment of the clavicle hook plate. Injury 2010; 41: 613e9. 15 Tambe AD, Motkur P, Qamar A, Drew S, Turner SM. Fractures of the distal third of the clavicle treated by hook plating. Int Orthop 2006; 30: 7e10. 16 Horvath F, Kery L. Degenerative deformations of the acromioclavicular joint in the elderly. Arch Gerontol Geriatr 1984; 33: 259e65. 17 O’Brien SJ, Pagnani MJ, Fealy S, McGlynn SR, Wilson JB. The active compression test: a new and effective test for diagnosing labral tears and acromioclavicular joint abnormality. Am J Sports Med 1998; 265: 610e3. 18 Walton J, Mahajan S, Paxinos A, et al. Diagnostic values of tests for acromioclavicular joint pain. J Bone Joint Surg Am 2004; 86: 807e12. 19 Jacob AK, Sallay PI. Therapeutic efficacy of corticosteroid injections in the acromioclavicular joint. Biomed Sci Instrum 1997; 34: 380e5. 20 Pichler W, Weinberg AM, Grechenig S, Tesch NP, Heidari N, Grechenig W. Intra-articular injection of the acromioclavicular joint. J Bone Joint Surg Br 2009; 91: 1638e40. 21 Freedman BA, Javernick MA, O’Brien FP, Ross AE, Doukas WC. Arthroscopic versus open distal clavicle excision: comparative results at six months and one year from a randomized, prospective clinical trial. J Shoulder Elbow Surg 2007; 16: 413e8. 22 Petersson CJ. The acromioclavicular joint in rheumatoid arthritis. Clin Orthop Relat Res 1987; 223: 86e93.

as a potential source of pathology in a patient with an injury to the shoulder girdle or with non-traumatic pain affecting the shoulder girdle will expose the patient and surgeon to the risk of disappointing treatment outcomes. Advances in operative techniques and implant technology are adding to what has previously been a very limited surgical armamentarium. A

REFERENCES 1 Colegate-Stone T, Allom R, Singh R, Elias DA, Standring S, Sinha J. Classification of the morphology of the acromioclavicular joint using cadaveric and radiological analysis. J Bone Joint Surg Br 2010; 92: 743e6. 2 Imatani RJ, Hanlon JJ, Cady GW. Acute, complete acromioclavicular separation. J Bone Joint Surg Am 1975; 57: 328e32. 3 Bannister GC, Wallace WA, Stableforth PG, Hutson MA. The management of acute acromioclavicular dislocation. A randomised prospective controlled trial. J Bone Joint Surg Br 1989; 71: 848e50. 4 Wolf EM, Pennington WT. Arthroscopic reconstruction for acromioclavicular joint dislocation. Arthroscopy 2001; 17: 558e63. 5 Chernchujit B, Tischer T, Imhoff AB. Arthroscopic reconstruction of the acromioclavicular joint disruption: surgical technique and preliminary results. Arch Orthop Trauma Surg 2006; 126: 575e81. 6 Weaver JK, Dunn HK. Treatment of acromioclavicular injuries, especially complete acromioclavicular separation. J Bone Joint Surg Am 1972; 54: 1187e94. 7 Bailey RW, Metten CF, O’Connor GA, Titus PD, Baril JD, Moosman DA. A dynamic method of repair for acute and chronic acromioclavicular disruption. Am J Sports Med 1976; 4: 58e71. 8 Berson BL, Gilbert MS, Green S. Acromioclavicular dislocations: treatment by transfer of the conjoined tendon and distal end of the coracoid process to the clavicle. Clin Orthop Relat Res 1978; 135: 157e64. 9 Werder H. Posttraumatic osteolysis of the extremity of the clavicle. Schweiz Med Wochenschr 1950; 80: 912e3.

ORTHOPAEDICS AND TRAUMA 25:1

36

Ó 2010 Published by Elsevier Ltd.

ARTHROPLASTY

Leg length inequality following total hip replacement

to tolerate what would be otherwise considered an acceptable deformity. This paper focuses on the important elements in the history, examination and investigation with relevance to LLI and total hip replacement, particularly considering those at greater risk of symptoms. The paper also discusses the background, definitions as well as the management strategies for an unacceptable LLI.

Anthony McWilliams

Keywords arthroplasty; hip; leg length inequality; replacement

Todd D Stewart Andrew J Grainger Philip J O’Connor

Introduction

Derrick White After a hip replacement many patients have a small difference in their leg lengths. Most remain unaware of this. As the difference in the leg lengths increases the proportion of patients who are both aware of the difference and are unhappy with it increases. Leg length inequality (LLI) is an independent risk factor in the outcome of total hip replacement (THR) and can result in a disappointing result despite an otherwise satisfactory operation.1 It is important to remember though that not everyone with an inequality will be symptomatic, though the proportion increases with increasing inequality.2,3 Although the consequences of leg lengthening were recognized by Charnley when the technique of arthroplasty was popularized and subsequent papers were published during the 1970s and 1980s, it was not until the 1990s that LLI following total hip replacement came to prominence in the literature. This may be due the indications for arthroplasty broadening to include a more demanding and younger population who require more function from an arthroplasty as opposed to simple pain relief. This article focuses on LLI following THR particularly in the identification of those likely to result in a symptomatic LLI, mechanisms to minimize the deformity and treatment options in those identified with a post-operative LLI. It also highlights important parts of the history, examination and investigation that would typically take place in an arthroplasty clinic but are of particular relevance to LLI. While both shortening and lengthening are issues with THR, patients tend to tolerate shortening more than lengthening.2 As a result the bulk of this paper primarily concerned with lengthening of the post-operative limb. As well as the associated morbidity, LLI following total hip replacement is having increasing medico-legal consequences.

Anthony Redmond Martin H Stone

Abstract Total hip replacement is a technical and demanding operation and subsequent leg length inequality (LLI) is a recognized complication. Although noted when the operation was popularized in the 1960s it was in the 1990s that it increased in prominence. LLI following total hip replacement is an independent risk factor in the outcome of total hip replacement. While not everyone with a post-total hip replacement LLI will be symptomatic, those who are can complain of mechanical problems, pain and neurological deficit. LLI also has increasing medico-legal consequences. Although any patient undergoing total hip replacement is at risk of a symptomatic LLI there are identifiable populations who are less likely

Anthony McWilliams BSc MB BS MRCS Specialty Registrar, Yorkshire and Humber Deanery, UK. Conflict of interest: none. Todd D Stewart PhD Senior Lecturer in Medical Engineering, University of Leeds, Leeds LS2 9JT, UK. Conflict of interest: none. Andrew J Grainger MRCP FRCR Consultant Muscloskeletal Radiologist, Dept of Radiology, Chapel Allerton Hospital, Leeds, UK. Conflict of interest: none. Philip J O’Connor MRCP FRCR FFSEM Consultant Muscloskeletal Radiologist, Dept of Radiology, Chapel Allerton Hospital, Leeds, UK. Conflict of interest: none.

Definitions Derrick White PhD BHSc (Hons) BSc (Hons) PGCE NIHR Leeds Musculoskeletal Biomedical Research Unit, Leeds Teaching Hospitals NHS Trust, Leeds, UK. Conflict of interest: none.

LLI after THR, as with any other cause of LLI, may be divided into true and apparent inequality. Apparent, sometimes called functional LLI, occurs when the lower limbs appear to be of different lengths when measured from a fixed midline reference point but are the same length when measured individually. An example of an apparent/functional LLI is when tight anterior structures result in a flexion deformity. True or structural LLI occurs when the limb length is either shortened or lengthened due to the implant itself.3,4 True or structural LLI can be further subdivided. Type I is lengthening directly due to the component, i.e. if the femoral stem is proud or the cup has been placed too low (Figure 1). Type

Anthony Redmond MSc PhD Senior Lecturer in the Section of Musculoskeletal Disease, Leeds Institute for Molecular Medicine, NIHR Fellow at the Leeds Musculoskeletal Biomedical Research Unit, Chapel Allerton Hospital, Leeds, UK. Conflict of interest: none. Martin H Stone MB ChB MPhil FRCS (Ed) FRCS (Eng) Consultant Orthopaedic Surgeon, Chapel Allerton Hospital, Leeds LS7 4SA, UK. Conflict of interest: none.

ORTHOPAEDICS AND TRAUMA 25:1

37

Ó 2010 Elsevier Ltd. All rights reserved.

ARTHROPLASTY

than 30 mm is, under normal circumstances, definitely not. However, there remains a grey area in-between that is contentious.

Background The morbidity associated with a post-THR LLI can vary from the patient only becoming aware of the problem when noticed by others (often the physiotherapist or spouse), to other patients who experience unpleasant symptoms on recovering from the anaesthetic. Patients can complain of mechanical problems or pain or may suffer nerve damage.7,8 The mechanical symptoms of LLI range from instability and dislocation to a limp. Limping is itself a manifestation of disability and the resulting gait abnormality can increase the physiological demands of mobility. A breakpoint around 40 mm in healthy individuals has been reported at which this occurs; however, in those with reduced physiological reserve this may be as low as 20 mm of LLI or less. This is important as although the limit for a healthy patient would be considered a rare and unacceptably high inequality, 20 mm is well within the ‘grey area’ and is within the range of some of the post-operative LLI published.9 While patients tend to notice lengthening more than shortening, it is important to remember that an excessively shortened limb may result in loss of soft tissue support, instability and dislocation as well as many other of the mechanical complaints. Lower back pain is more common in patients with an LLI and is thought to be largely due to the resulting functional scoliosis which, along with altered muscle action, can make it difficult to stand erect even with an inequality of 10 mm.10 Scoliosis can be detected with an LLI of 6 mm.4 Similarly, both ipsi and contralateral hip, knee and ankle joint pain can arise due to compensatory mechanisms. Golightly et al. demonstrated that an LLI of greater than 20 mm (of all causes, not just arthroplasty) is associated with progressive pain and osteoarthritis in the knee and hip, generally in the short limb.11,12 Pain and arthritic changes may be present following as little as 9 mm LLI.4 Nerve damage due to surgical limb lengthening tends to present earlier than mechanical symptoms. These patients complain of immediate paraesthesiae or pain on recovery from the anaesthesia. The sciatic nerve is particularly vulnerable, resulting in potential motor and sensory deficit. As well as the problems for the patient detailed above, there are mechanical consequences of LLI for the articulation of the hip joint itself. While there is an agreement of the concept that LLI will result in altered wear properties of the arthroplasty, there is disagreement about the precise mechanisms. Behave et al. noted that with an LLI (of any cause) the longer limb bears greater load for longer in the gait cycle than the shorter limb.8 However, Barnett et al. found that lengthening resulted in a varus deformity causing reduced range of motion at the hip, leading to predictably lower wear rates.13 In addition, even if LLI does result in an increased wear rate per million steps, if a patient has a significantly symptomatic LLI they are likely to be less mobile, with a corresponding reduction in wear associated with the reduction in activity.

Figure 1 An example of a Type I LLI. In this case the stem is incompletely inserted, which is directly leading to the inequality.

II lengthening occurs when component mal-positioning (such as excess ante- or retroversion) results in instability. In an effort to stop the hip dislocation the surgeon may increase length and offset to increase soft tissue tension (Figure 2).5,6 Due to the technical and complex nature of the biomechanics of arthroplasty it is also possible to have a mixed picture of true and apparent LLI. It is vital however that the major cause of a symptomatic post-operative LLI is identified as misdiagnosis may result in unnecessary or inappropriate revision surgery. The issue of what defines an unacceptable and perhaps even negligent LLI is more difficult and is not only clouded by the lack of agreement of significance but also by the fact that for any given LLI only a proportion will be symptomatic. Within the literature the consensus is that less than 10 mm is acceptable and that greater

The patient in the clinical setting Figure 2 This radiograph demonstrates a Type 2 structural LLI. In this case the cup has been inserted in an open position, which causes instability of the joint replacement. In an effort to increase soft tissue tension, a stem with increased neck length and offset has been inserted, resulting in an LLI.

ORTHOPAEDICS AND TRAUMA 25:1

Identification of the patient at particular risk Heightened awareness by the surgeon of an ‘at risk’ patient for LLI will encourage increased care with leg lengths at the time of

38

Ó 2010 Elsevier Ltd. All rights reserved.

ARTHROPLASTY

surgery. There are specific sub-groups of patients who are more likely to suffer a symptomatic LLI and if these can be identified before hip replacement this can be discussed with them at that stage. The populations can be divided into those susceptible and those sensitive. Patients who are susceptible to an LLI are those with atypical bony anatomy such as a narrow or bowed femur or poor acetabular bone stock. These anatomical problems make insertion and alignment of the components difficult. Due to the increased technical demands and the difficulty in performing intra-operative checks, patients with very high BMI may also be at greater risk of an LLI. The patients that are more sensitive to LLI are those who have a decreased ability to cope with or compensate for any given inequality. The assessment prior to any decision to operate must identify this group as they may not even tolerate what would otherwise be considered an acceptable LLI. This group includes patients with a pre-existing scoliosis, contra-lateral leg shortening for any reason, females and patients with a short stature.14 Contractures around either hip, such as fixed flexion, abduction or adduction, not only affect the patient’s gait pre-operatively, but also additionally can cause an apparent LLI. Hence, contractures can potentially reduce the patient’s ability to compensate for any change in leg length. A short varus femoral neck may not only predispose a patient to be susceptible, but also, with stretching of previously short and tight abductors, can result in an abduction contracture. The resulting pelvic tilt exaggerates any lengthening. Any history of a limp or LLI should be cause for further enquiry. Questions such as: ‘When did it start?, How was it first noticed?, Any obvious cause (e.g. trauma)?, Which leg feels longer?, What are the present symptoms? and Are there other symptoms, such as lower back pain, that can be caused by an unidentified LLI?’ should also be discussed with the patient. Pre-existing LLI of either leg must be documented. If the patient already is short on the non-operative side (whether this is perceived or not), this may result in a relatively modest increase in leg length becoming symptomatic post-operatively. In addition, any patient who has bilateral hip disease resulting in ‘equal’ shortening on both sides should be made aware that the process of arthroplasty, which aims to restore the centre of rotation, is likely to cause an inequality until such time as the other hip is replaced.

Figure 3 The typical stance adopted by a patient with an LLI. The knee of the longer leg is flexed to assist in the maintenance of a level pelvis.

type of deformity as a structural scoliosis will render the patient potentially sensitive to any change in leg length, whereas a nonstructural scoliosis may be caused by LLI and could be exacerbated by injudicious surgery. Examination of the resting position of the leg and comparison of the range of movement of the hips, as well as employing techniques such as the Thomas test, should provide information regarding the presence of soft tissue contracture, though the picture can be blurred by both pain inhibition and reduced range of movement due to arthropathy, which are typical in the patients presenting to an arthroplasty clinic. Adduction and abduction contractures should be tested for fixed deformity and its cause must be identified as it can be a source of pre-operative LLI and associated symptoms as well as a cause of post-operative functional LLI and resulting morbidity. Assessment of pre-existing LLI should be carried out as a matter of routine. In the general population LLI has been found to be as great as 20 mm in asymptomatic individuals.4,15 This can be performed while the patient remains standing using the ‘indirect method’. This is accomplished by placing blocks of known height underneath the shorter leg until the pelvis is square. The leg lengths may also be measured directly: prior to employing this method the pelvis must be levelled and the legs placed in a symmetrical position. Apparent LLI should be measured using a fixed midline reference to the medial malleoli and true LLI measured from the greater trochanter to the medial mallioli. The latter can be further examined using the Galeazzi test to identify whether the inequality is above or below the knee. If above the knee, examining Bryant’s triangle will identify LLI at the level of the neck of femur.16 Clinical measurement is important in identifying LLI, however, it can be

Pre-operative physical examination The patient’s stance and then gait must be examined and documented. A patient with an LLI of any cause is likely to flex the knee of the longer limb in an attempt to correct pelvic tilt (Figure 3). Particularly of interest is the long leg gait, indicating a pre-existing inequality, which can be the result of a true LLI in any part of either leg or be the result of a functional nature. A Trendelenburg gait may indicate abnormal abductor function and thus a patient with a Trendelenburg gait may be susceptible to LLI, which must be considered when planning surgery. Gait abnormalities due to previous arthropathy on either side must also be recognized. The footwear should be examined for raises, asymmetric wear or the presence of orthotics. While the patient is still standing the spine should be assessed for any scoliosis, and if present it should be identified as being structural or non-structural using techniques such as the Adam’s forward bending test. It is particularly important to identify the

ORTHOPAEDICS AND TRAUMA 25:1

39

Ó 2010 Elsevier Ltd. All rights reserved.

ARTHROPLASTY

inaccurate and therefore must not be relied upon to be the single quantitative measure. A radiographic LLI of 10 mm or more can appear equal on clinical examination. Investigations The primary investigation in the arthroplasty clinic is the plain AP radiograph, centered at the symphysis, to include the pelvis and both hips. Ideally this should be performed in a standardized way. Typically this can done with the pelvis square (using blocks if required) and the hips in either neutral or internally rotated to allow adequate views of the hip/pelvic anatomy.17 Two methods of measurement of LLI on plain AP radiographs are common in the literature. Williamson et al. describes using an inferior interischial line as a reference, then drawing a line parallel to this at the level of the most prominent part of the lesser trochanter and measuring the perpendicular distance between the two lines. This is repeated for the other hip and the two measurements are compared (Figure 4).18 The second method (Woolson et al.) is similar but uses a line at the level of the inferior part of the acetabular teardrop as a reference and a parallel line at the level of the middle of the lesser trochanter (Figure 5).19 To separate femoral mal-positioning from acetabular cup malpositioning, the authors have presented work comparing these methods to other methods and advocate the use of the ‘Leeds method’ of measuring LLI on plane AP pelvis radiographs. The Leeds method uses an initial reference at the level of the centre of femoral rotation (CFR), two further parallel lines are drawn, one at the level of the inferior acetabular teardrop and the other at the midpoint of the lesser trochanter, and repeated for the other hip. This provides three measurements per hip, the CFR to teardrop, which corresponds to any change in leg length associated with the cup, the teardrop to trochanter for LLI due to the stem and an overall measurement of leg length; these can be of particular use in the audit of practice and in the bilateral or revision setting. The authors have found the Leeds method to be similarly accurate for the measurement of LLI but with better repeatability than the two

Figure 5 The Woolson et al. technique for measuring LLI on an AP pelvis radiograph. The reference line is at the level of the inferior part of the acetabular teardrop. The two further parallel lines are at the level of the midpoint of the lesser trochanter. Measurement WoA for the arthroplasty side is greater than the WoN, indicating that the arthroplasty side is longer. Note the contrast with the Williamson et al. method.

prominent methods in the literature (Figure 6).20 The authors recommend that a lateral hip radiograph should also be obtained to check for any atypical bony anatomy in the hip or femoral shaft. More complex investigations such as CT scans or use of a gait lab tend to be either for the investigation of a large or unexplained LLI or in the revision arthroplasty setting.

Figure 6 The Leeds method for measuring LLI on a radiograph. The reference line bisects the centres of femoral rotation. Two further parallel lines per hip are then drawn, the first at the level of the acetabular teardrop and the second at the level of the midpoint of the lesser trochanter. This produces three measurements per hip, C equating to any LLI due to the cup position, S any LLI due to the stem and overall measurement O. In this case the C measurements are similar, S is greater on the arthroplasty side and the overall measurement, O, is greater. Therefore the lengthening in this arthroplasty is predominantly due to the stem.

Figure 4 The Williamson et al. method of measuring LLI on a radiograph. The reference line is at the inferior interischial line and two further parallel lines are drawn through the most prominent part of the lesser trochanters. The perpendicular distance between is measured and compared. In this case the distance WiA for the arthroplasty is smaller than WiN for the native hip, indicating that the operated side has been lengthened.

ORTHOPAEDICS AND TRAUMA 25:1

40

Ó 2010 Elsevier Ltd. All rights reserved.

ARTHROPLASTY

Techniques to minimize LLI Current literature separates techniques employed to minimize LLI at the time of surgery into pre-operative and intra-operative techniques. The main pre-operative technique is to use templates as described by Muller21; originally using overlay templates but more recently with computerized templating for digital films. Using this technique the acetabulum is analyzed for version, size and abnormal bony anatomy (such as roof deficiency, osteophytes or subchondral cysts), allowing the surgeon to determine the planned size of acetabular component. The femur is then similarly assessed for abnormal anatomy and planning of the neck resection and size and offset of stem. Intra-operative techniques to assess and minimize LLI can be subdivided into indirect and direct methods. Indirect techniques are either those to assess soft tissue tension or direct comparison with the contra-lateral lower limb. The indirect techniques are an assessment of soft tissue balancing more than specifically of LLI; as such they are subjective and dependant on other factors such as the patient positioning, the type of anaesthetic and the amount of muscle relaxation achieved. However, in the hands of an experienced arthroplasty surgeon intra-operative techniques can be a useful ‘surrogate’ evaluation of LLI. A number of direct measurement techniques have been published with varying clinical success. These direct measurement techniques are broadly based around using a fixed reference on the pelvis and measuring to a marker usually on the greater trochanter (typically either a suture or electrocautery mark). The devices vary from either a Steinman pin or K-wire bent to act as a marker or caliper, to specific proprietary devices. Other published methods include the use of an ‘L’ shaped caliper to allow measurement of the long axis of the femur,17 a carpenter’s or spirit level for horizontal alignment,22 dual pin calipers to minimize loosening23 or direct measurement with a ruler.24 The authors found a mean LLI for all published techniques of 0.3e9.0 mm and the total range of all results for LLI was
22 to þ35 mm.24e27 A major difficulty with all of these techniques results from the significant change in leg length that results from changes in adduction or abduction of the leg between measurements. Up to 17 mm error in the measurement of leg length can be caused by 10 of adduction or abduction malpositioning.28 Many surgeons measure the calcar cut from the lesser trochanter using the surgeon’s index finger as the measuring ruler. The distance from the shoulder of the stem or stem rasp from the tip of the greater trochanter is also used as a good reference distance. Clinical success in the reporting of intra-operative techniques are varied; Hofmann et al. reported that in 50% of cases the preoperative plan was altered intra-operatively. Others report good series with 97% of cases with a post-surgery LLI within 1 cm.19 Radiographic errors can also cause a problem pre-operatively as the image magnification may be up to 129% in error, resulting in sizing issues.29 Advocates of templating state that it formalizes a process of thought in which the surgeon plans the operation, components and likely obstacles.

an X-ray should be taken and advice from a second surgeon should be sought before the patient is woken from the anaesthetic. If the LLI is greater than 3 cm, consideration should be given to an immediate re-operation to correct this problem. The action taken will depend on the facilities and the competence and experience of the staff available at the time. For an LLI of less than 3 cm, a wait and see approach may be considered more appropriate. Each case must be decided on its merits. Neurological deficit will tend to present sooner than pain or mechanical symptoms. Nerve palsy in severe LLI can present in the immediate post-operative period and must be investigated with a plain X-ray. Other causes of nerve palsy should be excluded, such as direct trauma intra-operatively, traction, ischaemia, thermal injury or compression due to haematoma or cement. Extent of the palsy is a useful indicator as mild to moderate deficits may resolve, however, severe palsy in the presence of significant LLI is unlikely to improve without surgical correction. Retention of the prostheses or early return of motor function has a better prognosis for recovery in these patients.30,31 Patients presenting after hip replacement with pain or mechanical symptoms should be assessed to ensure that the LLI is the likely cause of the symptoms. These are much more likely to be functional or structural as opposed to nerve palsies which, especially in earlier/ severe presentation, would tend to be purely structural. The mainstay of non-operative treatment of a structural LLI is the shoe raise. Over 90% of patients can have gait correction and significant improvement with this option alone.13,31 However, functional LLI should not be aggressively treated with orthoses as correction of the soft tissue deformity could be limited by them; physiotherapy is more appropriate for these patients. Operative management in LLI should be considered a last resort for most patients and is usually carried out in specialist centres. Many patients’ perception of the LLI will improve and often resolve in the first year after the surgery. Surgery in cases of functional LLI predominantly involves soft tissue release or soft tissue advancement. The management of structural LLI would tend to be revision arthroplasty. It is therefore important that prior to any surgery the classification of the structural LLI is known. Without this the patient may undergo extensive and unnecessary surgery. If the problem is due to a Type I structural LLI, where the lengthening is directly due the prosthesis, then the individual component or, with modular systems just part of the component (e.g. changing to a reduced offset head in a proud stem), may be revised. In a Type II structural LLI the surgeon has lengthened the leg to provide compensatory tightening of the soft tissues at the time of the initial hip replacement for a mal-aligned component, usually the cup. This malpositioned component must be revised initially and the other component reviewed on the table. Simple correction of the alignment may not take into account the extra lengthening factored into the other component to maintain stability, causing a greater instability.

Summary LLI following total hip replacement, though not always symptomatic, can be a cause of major dissatisfaction despite an otherwise successful operation. It is likely that with the broadening indications for hip replacement surgery and an increasingly demanding population the prominence of LLI is going to increase.

Management of the patient with a post-operative LLI The management of the patient with a symptomatic LLI will depend on the timing, type of symptoms, severity, and cause. If the surgeon notices a significant LLI at the end of the operation

ORTHOPAEDICS AND TRAUMA 25:1

41

Ó 2010 Elsevier Ltd. All rights reserved.

ARTHROPLASTY

15 Edeen J, Sharkey PF, Alexander AH. Clinical significance of leg-length inequality after total hip arthroplasty. Am J Orthop (Belle Mead NJ) 1995; 24: 347e51. 16 Harris N, Stanley D. Advanced examination techniques in orthopaeadics. Cambridge; 2005. 17 Shiramizu K, Naito M, Shitama T, Nakamura Y, Shitama H. L-shaped caliper for limb length measurement during total hip arthroplasty. J Bone Joint Surg Br 2004; 86: 966e9. 18 Williamson JA, Reckling FW. Limb length discrepancy and related problems following total hip joint replacement. Clin Orthop Relat Res 1978; 134: 135e8. 19 Woolson ST, Hartford JM, Sawyer A. Results of a method of leg-length equalization for patients undergoing primary total hip replacement. J Arthroplasty 1999; 14: 159e64. 20 McWilliams A, Grainger A, O’Connor P, et al. Assessing reproducibility for radiographic measurement of leg length inequality for total hip replacement. Sheffield: British Hip Society Annual Scientific Meeting, 2010. 21 Muller ME. Surgery of the musculoskeletal system, vol. III. New York: Churchill Livingstone, 1983. 22 Bose WJ. Accurate limb-length equalization during total hip arthroplasty. Orthopedics 2000; 23: 433e6. 23 Takigami I, Itokazu M, Itoh Y, Matsumoto K, Yamamoto T, Shimizu K. Limb-length measurement in total hip arthroplasty using a calipers dual pin retractor. Bull NYU Hosp Jt Dis 2008; 66: 107e10. 24 Hofmann AA, Bolognesi M, Lahav A, Kurtin S. Minimizing leg-length inequality in total hip arthroplasty: use of preoperative templating and an intraoperative x-ray. Am J Orthop (Belle Mead NJ) 2008; 37: 18e23. 25 White TO, Dougall TW. Arthroplasty of the hip. Leg length is not important. J Bone Joint Surg Br 2002; 84: 335e8. 26 Konyves A, Bannister GC. The importance of leg length discrepancy after total hip arthroplasty. J Bone Joint Surg Br 2005; 87: 155e7. 27 Jasty M, Webster W, Harris W. Management of limb length inequality during total hip replacement. Clin Orthop Relat Res 1996; 333: 165e71. 28 Sarin VK, Pratt WR, Bradley GW. Accurate femur repositioning is critical during intraoperative total hip arthroplasty length and offset assessment. J Arthroplasty 2005; 20: 887e91. 29 Gonzalez Della Valle A, Slullitel G, Piccaluga F, Salvati EA. The precision and usefulness of preoperative planning for cemented and hybrid primary total hip arthroplasty. J Arthroplasty 2005; 20: 51e8. 30 Schmalzried TP, Amstutz HC, Dorey FJ. Nerve palsy associated with total hip replacement. Risk factors and prognosis. J Bone Joint Surg Am 1991; 73: 1074e80. 31 Pritchett JW. Nerve injury and limb lengthening after hip replacement: treatment by shortening. Clin Orthop Relat Res 2004; 418: 168e71.

The morbidity associated with LLI ranges from mechanical symptoms to pain and nerve palsy. While LLI is a recognized risk following total hip replacement, there are populations of patients that are more likely to be symptomatic following total hip replacement. It is vital that these are identified pre-operatively and counselled appropriately. Post-operative LLI following total hip replacement is becoming more prominent in the literature and increasingly recognized as having medico-legal consequences. This paper aims to highlight those steps of a normal consultation that are of particular importance in the consideration of LLI and provide some background in terms of definition and management. A

REFERENCES 1 Mancuso CA, Jout J, Salvati EA, Sculco TP. Fulfillment of patients’ expectations for total hip arthroplasty. J Bone Joint Surg Am 2009; 91: 2073e8. 2 Ranawat CS, Rao RR, Rodriguez JA, Bhende HS. Correction of limblength inequality during total hip arthroplasty. J Arthroplasty 2001; 16: 715e20. 3 Maloney WJ, Keeney JA. Leg length discrepancy after total hip arthroplasty. J Arthroplasty 2004; 19-4(Suppl 1): 108e10. 4 Gurney B. Leg length discrepancy. Gait Posture 2002; 15: 195e206. 5 Parvizi J, Sharkey PF, Bissett GA, Rothman RH, Hozack WJ. Surgical treatment of limb-length discrepancy following total hip arthroplasty. J Bone Joint Surg Am 2003; 85-A: 2310e7. 6 Theruvil B, Kapoor V. Surgical treatment of limb-length discrepancy following total hip arthroplasty. J Bone Joint Surg Am 2004; 86-A: 1829. author reply. 7 Friberg O. Biomechanical significance of the correct length of lower limb prostheses: a clinical and radiological study. Prosthet Orthot Int 1984; 8: 124e9. 8 Bhave A, Paley D, Herzenberg JE. Improvement in gait parameters after lengthening for the treatment of limb-length discrepancy. J Bone Joint Surg Am 1999; 81: 529e34. 9 Gurney B, Mermier C, Robergs R, Gibson A, Rivero D. Effects of limblength discrepancy on gait economy and lower-extremity muscle activity in older adults. J Bone Joint Surg Am 2001; 83-A: 907e15. 10 Friberg O. Clinical symptoms and biomechanics of lumbar spine and hip joint in leg length inequality. Spine (Phila Pa 1976) 1983; 8: 643e51. 11 Golightly YM, Allen KD, Helmick CG, Schwartz TA, Renner JB, Jordan JM. Hazard of incident and progressive knee and hip radiographic osteoarthritis and chronic joint symptoms in individuals with and without limb length inequality. J Rheumatol 2010; 37(10): 2133e40. 12 Golightly YM, Allen KD, Helmick CG, Renner JB, Jordan JM. Symptoms of the knee and hip in individuals with and without limb length inequality. Osteoarthritis Cartilage 2009; 17: 596e600. 13 Barnett J, Redmond A, White D, et al. Biomechanical analysis of leg length discrepancy following total hip replacement. University Of Leeds, United Kingdom, Leeds Teaching Hospitals NHS Trust. 14 Ranawat CS, Rodriguez JA. Functional leg-length inequality following total hip arthroplasty. J Arthroplasty 1997; 12: 359e64.

ORTHOPAEDICS AND TRAUMA 25:1

Acknowledgements This work was supported by the NIHR (National Institute for Health Research) through funding of the LMBRU (Leeds Musculoskeletal Biomedical Research Unit). In addition this work was partially funded through WELMEC, a Centre of Excellence in Medical Engineering funded by the Welcome Trust and EPSRC, under grant number WT 088908/Z/09/Z.

42

Ó 2010 Elsevier Ltd. All rights reserved.

HAND

Fractures of the metacarpals and phalanges

A clear history as to the precise mechanism of injury should be obtained to help reveal both the type of injury sustained and any potentially associated injuries. Axial loading frequently results in shearing articular or metaphyseal compression fractures, while one should maintain a high index of suspicion for injuries proximal to the hand. A bending element is often required for joint dislocations or diaphyseal fractures, while torsional forces may result in spiral fractures (with a high likelihood of mal-rotation) or more complex dislocation patterns. Higher energy injuries with crushing forces typically result in much greater soft tissue trauma, which increases the likelihood of an associated open injury and neurovascular compromise (including compartment syndrome) and can significantly influence the available treatment options. The common symptoms associated with hand fractures and dislocations include pain, swelling, weakness, stiffness and deformity. Patients should be questioned about the initial deformity and reduction manoeuvres that may have been attempted. Numbness and subjectively altered sensation indicate nerve involvement, which may be as a direct result of the injury or a secondary effect of swelling. With an open injury it is important to ascertain the nature of the environment in which the injury was sustained, the degree of wound contamination and the likelihood of foreign bodies must be considered. The time from injury to presentation is also a vital piece of information.

Benjamin JF Dean Christopher Little

Abstract Hand fractures and dislocations are amongst the most frequently encountered orthopaedic injuries. A thorough assessment including history, clinical examination and radiological investigations is essential before deciding upon an appropriate management plan. The vast majority of injuries can be managed non-operatively, but knowing when surgery is likely to improve the functional outcome is the key in selecting when to proceed with a specific surgical intervention. This review article aims to summarize the key concepts in the management of these injuries so that the reader can better understand the rationale behind the surgical decision making process.

Keywords dislocations; fractures; hand; metacarpal; phalangeal

Introduction/epidemiology

Clinical examination and signs

Hand fractures and dislocations are amongst the most common orthopaedic injuries presenting to Accident and Emergency (A&E) departments, comprising approximately a quarter of all patients attending with a fracture or dislocation.1 Males are typically affected more commonly than females; sport related injuries are particularly common in the 3rd decade and workplace related injuries in the 5th decade. A large prospective study from Norway found that fractures were distributed around the hand as follows: 46% phalangeal, 36% metacarpal, with nearly 10% of the overall total affecting the little finger metacarpal neck.2 The borders of the hand is most commonly affected, with injuries to the thumb and the little finger ray predominating. Hand fractures and dislocations represent a considerable burden upon society in terms of medical costs and reduced workplace productivity, so minimizing the functional loss that may occur following injury is important not only in terms of the patient’s quality of life but also to minimize the overall cost to society.

Tenderness, swelling, bruising, crepitus, deformity, restricted motion and instability are common signs of injury. The skeletal examination can often be carried out more thoroughly following the use of local anaesthetic, which can be used either locally or as a block to specific nerves; obviously the documentation of the neurological status of the digit/s is vital prior to the administration of any local anaesthetic agents. Such blockade also allows tendon and ligament integrity to be assessed without pain, whilst facilitating the irrigation of wounds. Remembering to examine adequately for the presence of rotational deformity is essential. Mal-rotation of one bone segment that may occur as a result of a fracture is best appreciated from the alignment of the next most distal segment (best observed when the intervening joint is flexed to 90 degrees) and from the assessment of the relative alignment of the nail plates of the injured digit in relation to those of the adjacent uninjured digits. Apparent digit mal-rotation may also be a manifestation of unequal swelling in the palm; if doubt exists (having considered the injury mechanism and fracture pattern, twisting injuries are more likely to be associated with mal-rotation than axial compression injuries), re-assessing the hand after treatments to reduce swelling (elevation and ice) may be appropriate as long as not too long a time elapses.

History and mechanism of injury A well-taken history is important in order to tailor the management to the specific needs of the patient. Age, occupation, hand dominance, recreational activities and medical co-morbidity are all key parts of the general information that must be factored into the decision making process.

Imaging Plain radiographs should be undertaken in at least two planes centred at the level of the suspected injury, typically consisting of orthogonal postero-anterior (PA) and lateral views, with additional oblique views and cross-sectional imaging being obtained where diagnostic doubt exists, or to aid treatment planning. Ultrasound scanning performed by an experienced person may be useful to assess ligament and tendon integrity

Benjamin JF Dean MRCS(Ed) Orthopaedic Registrar Nuffield Orthopaedic Centre, Windmill Road, Headington, Oxford OX3 7LD, UK. Christopher Little FRCS (Orth) Consultant Hand Surgeon Nuffield Orthopaedic Centre, Windmill Road, Headington, Oxford OX3 7LD, UK.

ORTHOPAEDICS AND TRAUMA 25:1

43

Ó 2010 Elsevier Ltd. All rights reserved.

HAND

and to detect radiolucent foreign bodies. MR imaging (MRI) is helpful to detect occult fractures, particularly in the carpus3 (Figure 1). Computed tomography (CT) scans may be used to plan the approach to and fixation of complex intra-articular injuries, or to assess the skeleton where plain films are hard to interpret due to overlapping bony shadows (such as assessing the carpo-metacarpal joints).

Associated injuries Open injuries require prompt surgical debridement and antibiotic prophylaxis should be standard for these injuries. Significantly displaced distal phalangeal fractures are associated with the disruption of the nail matrix. Soft tissue coverage can be achieved by a variety of means ranging from skin grafts to local or free flaps. Tendon disruption may occur with specific injuries, a common example being the terminal extensor tendon with distal interphalangeal joint (DIPJ) injuries, the bony mallet injury (Figure 2). Neurovascular injuries are rare following simple closed fractures and dislocations of the hand, but common with open hand trauma, particularly in the case of injuries as a result of sharp objects. The term ‘combined injuries’ refers to a hand fracture associated with one or more of the aforementioned soft tissue injuries (skin, tendon or neurovascular injury). Mutilating hand trauma is a complex and highly challenging area that is covered elsewhere.4

Figure 2 A lateral radiograph demonstrating the bony mallet injury, note the extensor lag of the distal interphalangeal joint.

and those where non-union will not impair function (many small palmar plate avulsion injuries; note that those associated with joint instability are different injuries requiring a different approach). A stable fracture can be considered as one that will maintain an acceptable position at rest and when early mobilization protocols are initiated, with fractures that will not retain an acceptable position under these circumstances being defined as unstable. When instability is present the best treatment option is the least invasive that can adequately restore skeletal stability. In this way early motion is permitted and the potential adverse effects of surgery and of prolonged immobilization upon the tissues are minimized. Fractures that are likely to displace with early mobilization are considered ‘potentially unstable’ and can be managed with a period of immobilization (undesirable in the fingers as stiffness may ensue), surgical intervention to stabilize the fracture, or early mobilization with a protective resting splint with frequent clinical and radiographic monitoring to allow prompt surgical intervention in the event that unacceptable displacement occurs. The most critical elements in deciding between operative and non-operative treatment are the assessments of rotational malalignment and stability. Non-operative treatment avoids the complications of surgical treatment and is cheaper, but may not restore and maintain skeletal position and stability. Where reduction and early motion are deemed necessary, a closed reduction with percutaneous K-wire stabilization (CRPP) may be considered. Open surgery increases soft tissue trauma, but can potentially achieve the most stable and anatomic reduction and so allow aggressive early mobilization. Generally a good rule is that the more aggressive the surgeon’s intervention has been, the more aggressive the post-operative rehabilitation must be.

Principles of management Most digit fractures will unite without intervention, so the overriding aim of treatment of the skeletal elements of hand injuries is to restore function to the hand by maintaining joint mobility and stability. Early mobilization of the digits is the best way to maintain/regain mobility. Union of a fracture in its anatomical position is often not necessary for this goal to be achieved, and it is important that both the surgeon and the patient have a clear idea of the treatment objectives from the outset. Surgeons treating patients with hand fractures should have a good understanding of the injuries that require anatomic reduction (mal-rotated finger fractures and many intra-articular injuries), those that can be allowed to malunite without long-term function being compromised (many border digit metacarpal fractures),

Distal phalangeal fractures Anatomy and classification The human distal phalanx is characterized by a broad, spade-like tuberosity (the tuft) with a proximal projecting spine and a wide diaphysis containing a palmar concavity to create the ungual fossa.5 The soft tissue of the palmar surface can be divided into

Figure 1 Axial MRI of the wrist demonstrating a trapezial ridge fracture that was not visible on plain radiographs.

ORTHOPAEDICS AND TRAUMA 25:1

44

Ó 2010 Elsevier Ltd. All rights reserved.

HAND

Dorsal base fractures/the mallet fracture A mallet fracture is an injury to the extensor mechanism at the dorsum of the DIPJ associated with an intra-articular fracture at that level. As long as the joint remains reduced, mallet splintage (with check films in the splint over the first 2 weeks) is all that is needed (Figure 3). There is controversy as regards the management of mallet injuries that involve large displaced articular fragments. One large series has shown that the non-operative treatment of all mallet fractures, even those with DIPJ subluxation and involving significant portions of the joint, resulted in excellent function and pain free motion in a majority of patients.8 Bony mallet injuries with larger avulsed fragments are prone to joint instability (the palmar element subluxing); this may require closed reduction and trans-articular wiring of the joint, with the extensor fragment sometimes requiring separate control using a dorsal blocking wire (a technique described by Ishiguro).9 The wires pass through the joint, so care and attention to the pin sites with a low threshold for early wire removal is important to minimize the risk of a septic arthritis. Internal fixation has an unacceptably high complication rate.

the proximal and distal pulp. Nail bed injuries may accompany distal phalangeal fractures; the presence of a subungual haematoma should alert one to this possibility. The dorsal portion of the distal phalanx and the nail plate create the outer layers of a sandwich, which contains the nail matrix as the filling. When a fracture breaches the seal between the nail plate and the hyponychium, it constitutes an ‘open’ injury and should be treated as such. The relationship between the bone and soft tissue structures is of prime importance when considering the management of these injuries. Distal phalanx fractures may be of the tuft, the shaft or of the base. A key consideration is the attachment of the terminal extensor and flexor tendons to the dorsum of the epiphysis and palmar metaphysis of the distal phalanx respectively; fractures distal to the tendon insertions will be separated from any internal deforming forces as a result, whereas those that pass between the insertions (such as the Seymour fracture6 seen in skeletally immature people) will displace with the basal fragment extending and the distal fragment flexing. Seymour injuries are often open, with the nail plate pulling out from the eponychial fold; they require wash-out, reduction, nail bed repair and, often, CRPP. Dorsal base articular rim fractures are inherently unstable as a result of these deforming forces; these injuries represent extensor insertion avulsions and are referred to as bony mallet fractures (Figure 2).

Volar base fractures These may be either a small basal rim fracture representing a palmar plate avulsion or a closed avulsion injury of the flexor digitorum profundus (FDP) tendon with a bony fragment attached. FDP avulsion injuries can be type 1, 2 or 3 according to the Leddy and Packer classification system,10 which reflects the degree of retraction of the tendon stump (in practice, one into the palm, two to the level of the PIP joint and three at the level of the DIP joint). If the bony fragment is large enough it may be fixed with a compression screw, otherwise the injury is treated in accordance with the modern principles of flexor tendon reinsertion (Figure 4).

Tuft fractures Rim fractures of the tuft are often caused by crush injuries and are usually fairly stable, being splinted by the nail matrix dorsally and the pulp on the volar surface; while the rim elements not uncommonly fail to unite, the functional outcome is generally good and initial expectant management is appropriate. A splint may be use for comfort (as a thimble) but early motion of the digit and desensitization is important. Any significant displacement in the distal portion of the distal phalanx that supports the nail matrix will often be associated with an open injury; treatment should focus on formal exploration, wash-out and repair of the nail bed with the fracture being reduced and pinned to provide support to the surgical repair of the nail matrix unless the reinserted nail plate gives sufficient stability. Surgical exploration of the nail matrix is not obligatory if the fracture is closed, regardless of the presence or size of a subungual haematoma. The haematoma may need to be drained for pain relief (trephining the nail plate at its base). Generally active range of motion exercise can be begun at around a week following matrix repair provided the DIPJ has not been pinned; if the DIPJ has been pinned then pin removal is typically at the 3e4-week mark.7 The patient must be aware of the potential for nail deformity and the time required for stable nail regrowth, which is typically 5 months.

Proximal and middle phalanx fractures Anatomy and classification Fractures of the middle and proximal phalanges can be grouped by the anatomic regions of head, neck, shaft and base. Axial

Shaft fractures The majority of shaft fractures can be managed with splintage and this involves immobilization of the DIPJ. Active range of motion exercises involving the DIPJ can begin once the fracture has had time to consolidate at approximately the 3-week mark. Significantly displaced shaft fractures are prone to non-union, CRPP is usually sufficient for these fractures unless there is interposed tissue blocking the reduction which would then necessitate open reduction.

ORTHOPAEDICS AND TRAUMA 25:1

Figure 3 Lateral radiograph of the bony mallet injury of Figure 2 treated with a mallet splint, this radiograph confirms a satisfactory reduction with a fully extended joint.

45

Ó 2010 Elsevier Ltd. All rights reserved.

HAND

a and b. This variation of an FDP avulsion had a multi-fragmentary articular element and a fracture across the metaphyseo-diaphyseal junction. The tendon was reattached with a suture anchor, which could be inserted into the medulla by hyper-extending at the fracture site; the joint element was secured with a screw. Figure 4

loading mechanisms of injury may produce unicondylar or bicondylar fractures of the head or intra-articular fractures of the base. Base fractures may be subdivided into dorsal base, volar base or the more comminuted complete articular fractures that are often known as ‘pilon’ fractures. The periosteal surfaces of the proximal and middle phalanges are intimately related to the flexor and extensor tendons. The tendons of flexor digitorum superficialis (FDS) and FDP join the fibres of the fibro-osseous tunnel to cover the entire palmar surface of each bone. Consequently displaced shaft fractures result in tendons being exposed to sharp bone edges, which can result in the formation of troublesome adhesions. As a result it is important to achieve tendon gliding across the fracture site during healing. Proximal phalangeal fractures will normally angulate with an apex volar deformity. The interossei flex the proximal fragment while the distal fragment is hyperextended by the central slip’s pull on the base of the middle phalanx. The deformity associated with middle phalangeal fractures is less predictable. The more proximal middle phalangeal fractures tend to angulate with an apex dorsal deformity, FDS flexes the distal fragment and the central slip extends the proximal fragment. The more distal fractures tend to angulate with an apex volar deformity with FDS flexing the proximal fragment. The extensor mechanism is reliant upon the dorsal length of the proximal phalanx; bony malunion and consequent phalangeal shortening can result in a loss of interphalangeal extension (fixed flexion deformity). Middle phalangeal shortening tends to

ORTHOPAEDICS AND TRAUMA 25:1

produce PIP joint hyperextension due to the loss of distal joint extension and the characteristic swan neck deformity results. Management The majority of fractures can be managed non-operatively. Transverse shaft fractures will tend to be much more stable than oblique, spiral or comminuted fractures. The degree of displacement is related to the inherent stability of the fracture. Stable fractures can be managed with splintage and early protected motion. Excellent results can be achieved by immobilization of the affected digit in the ‘safe’ or ‘intrinsic plus position’ (metacarpophalangeal (MCP) joint flexion to 90 degrees with full extension of the interphalangeal joints) for 2e3 weeks, followed by buddy strapping until clinical union has been achieved. Extremely stable undisplaced fractures can be managed by buddy strapping from day 1 without a period of immobilization. It is important to check for the maintenance of the reduction, consequently all patients should be reviewed at 1 week for repeat radiographs. A commonly used treatment for fractures that are displaced and unstable after reduction is CRPP.11 Great care must be taken when using this method in the phalanges as a result of the broad extensor mechanism; in the proximal two thirds of the proximal phalanx it is almost impossible to avoid tethering the extensor mechanism. Kirschner wires (K-wires) can be inserted in a crossed or longitudinal manner via an antegrade or retrograde approach. Typically an extra-articular approach through the base is preferred for the middle phalanx, while proximal

46

Ó 2010 Elsevier Ltd. All rights reserved.

HAND

phalangeal fractures can be managed antegrade through a flexed MCP joint or retrograde through a flexed PIPJ. Post-operative management is tailored to the individual patient but the K-wires are generally protected with some form of external splintage, until the point at which the wires are removed and gentle mobilization can begin. Routinely K-wires are removed 4 weeks following their insertion. If the fracture cannot be adequately reduced by closed means then open reduction and internal fixation (ORIF) is the treatment of choice. Mal-rotation, significant shortening, angulation of more than 10 degrees and less than 50% bone apposition are unacceptable features. Severe open injuries and injuries with

bone loss invariably require ORIF. As a result of ORIF being associated with the most severe injuries and consequently poor prognostic factors, results are often felt to be sub-optimal.12 Options for ORIF include intra-osseus wiring, screw and plate fixation (Figure 5). While the surgical approach chosen should reflect the injury pattern and any open wounds, proximal third fractures of the proximal phalanx are often best approached with a mid line dorsal tendon splitting incision, mid and distal third fractures through one or two mid lateral incisions.13 ORIF must result in rigid fixation to allow early mobilization, a failure to achieve sufficiently secure fixation achieves the worst of both worlds and often a poor outcome.

a, b and c. Unstable proximal phalangeal fracture treated with ORIF via two mid lateral incisions. Figure 5

ORTHOPAEDICS AND TRAUMA 25:1

47

Ó 2010 Elsevier Ltd. All rights reserved.

HAND

Intra-articular fractures Splinting is only appropriate for minimally displaced fractures that are inherently stable. CRPP is an acceptable form of treatment as long as anatomic reduction of the articular surface can be achieved. More rigid fixation can be achieved with screws sometimes in combination with minicondylar plates, but inadequate bone stock may contraindicate ORIF.

collateral ligament (UCL) is a common injury, which is rarely accompanied by a full MP joint dislocation. Rupture of the UCL typically presents with pain and tenderness around the ulnar border of the MP joint. Stress testing in full extension and 30 degrees of flexion should be undertaken to reveal any instability. A Stener lesion is where the distally ruptured UCL is prevented from healing by the interposed adductor aponeurosis. Non-operative management is the mainstay of treatment for thumb MP joint injuries; typically this involves a minimum period of 4 weeks of thumb MP joint immobilization with either a cast or a splint. Only the complete UCL rupture with a Stener lesion requires operative intervention. In a recent prospective study of 24 patients a palpable tender mass on the ulnar side of the MP joint was used as the sole diagnostic criterion for the Stener lesion.17 In this study all patients with a Stener lesion were treated operatively with ligament repair, while all those without were managed non-operatively; outcomes in both groups were equivalent at 1 year. The presence of greater than 30 degrees valgus instability with stress testing in full extension is highly predictive of a Stener lesion.18

Volar base fractures of the middle phalanx Fractures at the base of the middle phalanx can be particularly unstable in direct relation to the percentage of the articular surface involved. When the volar fragment makes up greater than 40% of the articular surface it carries the majority of the collateral ligament and volar plate insertions, and as a result the dorsal fragment and remainder of the phalanx will sublux proximally and dorsally due to the pull of the central slip. Dynamic extension block splinting is an effective management strategy for these injuries provided the volar fragment involves less than 40% of the joint surface (Figure 6). Extension block pinning can be used for fractures involving greater than 40% of the articular surface provided the reduction occurs spontaneously with PIP joint flexion. Other techniques include a single compression screw (provided the fragment is large enough) and volar plate arthroplasty (felt by many to give unsatisfactory results); presently use of a hemihamate bone-graft is gaining popularity.14

Metacarpal fractures Anatomy and classification The metacarpals make up three arches of the hand: these are the two transverse arches that exist at the level of the CMC and MCP joints and the longditudinal arch, which is made up from their long convex dorsal surfaces. The metacarpals are bound together firmly by interosseus ligaments at their bases and by the deep transverse intermetacarpal ligaments distally. These strong ligamentous attachments help to maintain the transverse arches, but flattening may occur with multiple fractures or crush injuries. The distal carpal row forms the fixed transverse arch of the hand, which is the stable foundation upon which the remaining hand structure is constructed. The second and third CMC joints allow for the rigid attachment of the index and middle finger metacarpals, thus forming the fixed unit of the hand; the CMC joints of the ring and little fingers allow for more flexion and extension, important for grip power. Metacarpal fractures may be broken down into those of the head, neck, shaft and base. Shaft and neck fractures typically demonstrate apex dorsal angulation as a result of the deforming pull of the interossei. The normal neck to shaft angle of 15 degrees should be remembered when assessing these injuries radiographically. The term ‘pseudoclawing’ refers to the compensatory MCP hyperextension and PIPJ flexion that may occur in response to the apex dorsal angulation. The relative inflexibility of the index and middle finger CMC joints (where motion is negligible) compared to the ring and little finger CMC joints (motion is 15 and 25 degrees respectively) has important implications for deciding when to intervene surgically. The thumb CMC joint is the most mobile with around 50 degrees of flexion and extension, and 40 degrees of abduction and adduction.

Pilon fractures of the PIPJ The combination of complete articular surface involvement and metaphyseal compaction with bone loss is the most functionally devastating of the pilon fractures involving the PIPJ. Immobilization in flexion, ORIF, extension block splinting, dynamic force couple, skeletal traction and a silicone spacer have all been described in the management of these injuries. Although anatomic restoration of the joint surface should be one’s aim, this is seldom possible due to the amount of articular and metaphyseal comminution. The clinical outcomes following splint immobilization are generally unsatisfactory, while the outcomes following treatment with skeletal traction and ORIF are similar.15 The significant surgical complications associated with ORIF mean that skeletal traction is often the preferred treatment. While a distracting dynamic fixator can be fashioned using bent K-wires with or without rubber bands, several commercially-available modern devices are marketed to apply skeletal traction across the PIPJ while active ranges of motion exercises are begun. Avulsion fractures from the base of the proximal phalanges Marginal fractures of the base of the proximal phalanx involving the MP joint usually represent avulsion fractures of a collateral ligament. There is frequently a substantial palmar bony avulsion fragment. As the risk of non-union with non-operative treatment is high some authors advise early internal fixation with a lag screw using a dorsal approach.16 The advantages of this approach include early mobilization and a reduced risk of non-union.

Management The vast majority of metacarpal neck and shaft fractures may be treated non-operatively; intra-articular fractures of the head and base may also be managed non-operatively as long as the fracture is stable and minimally displaced. Metacarpal fractures with significant rotational deformity or shortening cannot be satisfactorily

Thumb MP joint ligament injuries The prime stabilizers of motion at the thumb MPJ are the radial and ulnar collateral ligaments. Complete rupture of the ulnar

ORTHOPAEDICS AND TRAUMA 25:1

48

Ó 2010 Elsevier Ltd. All rights reserved.

HAND

a and b. This PIP joint fracture-subluxation was unstable after reduction when extending beyond 40 degrees. An extension-blocking splint was fashioned preventing extension beyond 40 degrees, but allowing free flexion. The joint was monitored and, as it remained stable, the amount of extension prevented by the splint was reduced by 10 degrees each week from 3 1/2 weeks after injury, with close monitoring maintained. Figure 6

managed non-operatively. Splinting is directed towards neutralizing deforming forces and pain control while remembering the position of safe immobilization; the MCP joints are held in full flexion in combination with wrist dorsiflexion.19 The splint can be

ORTHOPAEDICS AND TRAUMA 25:1

removed as soon as the patient is comfortably able to move the hand and this should not be later than 3 weeks following the injury. In fact the immediate mobilization of fifth metacarpal neck fractures yields excellent results in spite of high degrees of angulation.20

49

Ó 2010 Elsevier Ltd. All rights reserved.

HAND

Metacarpal head fractures Articular fractures of the metacarpal head are often as a result of direct trauma. Non-operative management is indicated for those which are minimally displaced and stable to active stress testing. Non-operative treatment consists of splint immobilization for 2e3 weeks followed by early protected motion. ORIF is indicated when there is significant loss of articular congruency or instability with active stress.21 Collateral ligament avulsion injuries are difficult to reduce and hold with closed methods, meaning that ORIF is often required. Fixation may be with K-wires, screws or interosseus wires. The metacarpal head is approached dorsally with a tendon splitting approach. Early rigid fixation is important in order to allow early mobilization and prevent extensor mechanism tethering. Screw fixation is possible if the fragment diameter is greater than three times the diameter of the screw hole; movement within a few days is important with this technique. Ideally the fixation devices should be clear of the joint surface (Figure 7). If the fracture configuration means that fixation implants would cross the articular cartilage, either well counter-sunk headed screws or

small headless compression bone screws should be used to ensure no metal is exposed within the joint. Fixation by K-wires results in a stable but non-rigid form of fixation, thus movement must wait until around the 2- week mark. Metacarpal neck fractures Rotational deformity must be corrected and the degree of acceptable angulation depends on the digit involved. Less than 15 degrees of angulation may be tolerated in the index and middle finger metacarpals due to the rigid CMCJs. The more mobile ring and little finger metacarpals can tolerate more angulation with 35 and 45 degrees accepted respectively.22 Treatment is symptomatic and mobilization is begun as early as possible; if needed, splint immobilization involves holding the MCP joints in 90 degrees of flexion with around 30 degrees of wrist extension, while the PIPJs are left free to allow immediate motion. Indications for surgical intervention include the inability to achieve or maintain a satisfactory reduction where one is required. Rotational malalignment is unacceptable,

a, b and c. Pre and post-operative radiographs of a distal metacarpal fracture treated with ORIF. Figure 7

ORTHOPAEDICS AND TRAUMA 25:1

50

Ó 2010 Elsevier Ltd. All rights reserved.

HAND

a and b. The use of an IM K wire to stabilize an angulated fifth metacarpal shaft fracture. Figure 8

while excessive angulation and marked palmar comminution are relative indications for surgery. Intra-medullary (IM) K-wire stabilization is the surgical treatment of choice for most single metacarpal fractures; the use of multiple diverging contoured wires will give rotational control23 e the so-called bouquet technique, although the authors prefer to use direct reductionabsolute stability techniques (ORIF) where the injury is rotationally unstable. Other techniques include the use of crossed K-wires and transfixion to adjacent intact metacarpals. Anatomic reduction is not essential for injuries where there is no mal-rotation and a relative stability fixation technique has been

ORTHOPAEDICS AND TRAUMA 25:1

used. The indications for ORIF include multiple fractures, open injuries, irreducible fractures and injuries that are rotationally unstable. The techniques available for ORIF include K-wires in combination with tension bands and specifically designed hand plates with screws. Metacarpal shaft fractures Less angular deformity can be tolerated in metacarpal shaft fractures in comparison to metacarpal neck fractures due to the longer lever arm that will deliver the metacarpal head into the palm. Transverse, short oblique, oblique and spiral fractures fracture

51

Ó 2010 Elsevier Ltd. All rights reserved.

HAND

a, b and c. An open metacarpal fracture with significant bone loss treated with bridge plating without bone-grafting, radiographs at the 7 month follow up demonstrate bony union. Figure 9

patterns may be seen. Of note oblique and spiral types are particularly prone to acquiring rotational deformity as they shorten. Acceptable angulation is less than 10 degrees for the index and middle finger metacarpals, and less than 20e30 degrees in the ring and little finger metacarpals respectively; no rotational deformity is acceptable, while up to 5 mm of shortening can be tolerated. A moulded splint not including the MCP joints can be employed for shaft fractures proximal to the neck; the splint is discontinued after

ORTHOPAEDICS AND TRAUMA 25:1

around 4 weeks. Indications for surgical intervention include the inability to achieve or maintain a satisfactory reduction, open fractures, multiple fractures and fractures with bone loss. Again a multitude of techniques have been described in the literature, with IM K-wires being increasingly popular (Figure 8). The biomechanical strength of K-wire fixation has been shown to be significantly inferior to plates and screws. The interfragmentary lag screw is extremely effective in resisting loading forces, but two or more

52

Ó 2010 Elsevier Ltd. All rights reserved.

HAND

should be used unless a neutralization plate is applied as well. Plates function most effectively as tension bands (applied on the dorsal surface); with the availability of locking plates, fractures with marked bone loss or segmental fragmentation can be successfully treated with bridge plating techniques without the need for bonegraft in many cases (Figure 9).

CRPP can involve direct pinning at the fracture site or the splintage of the fracture site by securing the metacarpal to the adjacent metacarpal both proximally and distally. It may be difficult to control rotation without ORIF. Interfragmentary lag screw fixation is the treatment of choice for spiral fractures, screw fixation alone is acceptable as long as the

a and b. A Bennett’s fracture treated with ORIF. Figure 10

ORTHOPAEDICS AND TRAUMA 25:1

53

Ó 2010 Elsevier Ltd. All rights reserved.

HAND

fracture length is at least three times the diameter of the metacarpal and at least two screws are used.24 Short oblique fractures can be treated with a lag screw in combination with a dorsal neutralization plate. The complications of ORIF are not infrequent but its great benefit is to allow early mobilization. Thumb metacarpal shaft fractures Thumb metacarpal fractures are common and account for around 25% of all metacarpal fractures. Most fractures are proximal in which the distal fragment is adducted and supinated; the injury may be partial articular, complete articular or epi-basal in nature. Reduction is therefore performed using longitudinal traction with pronation and abduction to correct the angulation, a well-moulded cast is then applied for 3e4 weeks. The thumb CMCJ’s mobility means that angulation is well tolerated and surgical fixation is often not required for nonarticular injuries. Intra-articular fractures of the thumb base The most common of this type is Bennett’s fracture-dislocation. Bennett’s fracture-dislocation is a partial articular injury, involving a volar ulnar portion of the articular surface that is held in place by the volar oblique ligament and the remaining metacarpal fragment, which is displaced dorsally and into supination by the pull of abductor pollicis longus (APL) (Figure 10a). A true lateral and hyperpronated AP view, also known as ‘Robert’s’ view, may be used to best visualize this injury. Due to the articular involvement and unstable nature of the injury, closed reduction and CRPP are indicated on most occasions. Reduction is achieved using longditudinal traction with a combination of extension, abduction and pronation with direct pressure onto the base of the metacarpal, following by percutaneous pinning of the shaft fragment to the trapezium or index finger metacarpal. CRPP is followed by immobilization with a short arm thumb spica cast for around 6 weeks. If attempted closed reduction results in greater than 2 mm of joint surface displacement then open reduction is indicated. Lag screws or pinning may be used depending on the surgeon’s preference (Figure 10b). Rolando’s fracture-dislocation is a complete articular injury and in addition to the palmar fragment found in Bennett’s fracture, it is characterized by a larger dorsal fragment with varying degrees of comminution. Treatment options include closed reduction and casting, soft dressing with early motion, skeletal traction and early motion, and ORIF. Some authors advocate the use of skeletal traction with early motion as the treatment of choice.25 Traction is applied from an obliquely applied metacarpal wire to an outrigger device, while active motion is begun immediately and maintained for 6 weeks. Other authors prefer pinning the thumb metacarpal to the index metacarpal in order to distract the comminuted metaphyseal zone as it is allowed to heal,6 or using an external fixator. ORIF has the advantage of allowing early motion, but runs the risk of devascularizing fracture fragments with an already precarious blood supply and may be technically difficult in the presence of significant comminution. As a result ORIF is best reserved for the irreducible fracture or those with minimal comminution16 (Figure 11).

ORTHOPAEDICS AND TRAUMA 25:1

a and b. A Rolando’s fracture treated with ORIF. Figure 11

Carpo-metacarpal fracture-dislocations The little finger CMC joint is the most commonly affected. The concern is that with no treatment the persistent subluxation and incongruity leads to pain and reduced grip strength. Undisplaced injuries can be managed with cast immobilization and regular checks for the maintenance of reduction. Displacement necessitates operative treatment. To determine if the joint is dislocated, either a CT or use of the intra-medullary axis convergence technique (described by Shewring26) can be used (Figure 12a). CRPP is the treatment of choice providing closed reduction can be achieved. Manual reduction is followed by the placement of K-wires through the metacarpal shaft and into the carpal bones and the adjacent stable metacarpals (Figure 12b). Open reduction is indicated if closed reduction is not possible, and also in cases of excessive swelling, cases of delayed presentation and open

54

Ó 2010 Elsevier Ltd. All rights reserved.

HAND

injuries27; this may also be preferred if there is a single, large fragment off the dorsal lip of the carpus, reduction and stabilization of which would restore stability and allow early motion. Index finger CMC joint dislocations more commonly require open

reduction because of extensor carpi radialis brevis interposition. Open reduction is generally accompanied by K-wire fixation. Whether open or closed reduction is employed, cast immobilization follows and the wires are removed at 4e6 weeks.

a and b. a The intra-medullary axis covergance technique, note that the ring and little metacarpal lines do not meet the index and middle lines at a point in the distal radius suggesting joint dislocations. b demonstrates this injury treated with CRPP. Figure 12

ORTHOPAEDICS AND TRAUMA 25:1

55

Ó 2010 Elsevier Ltd. All rights reserved.

HAND

Conclusions

11 Kozin SH, Thoder JJ, Lieberman G. Operative treatment of metacarpal and phalangeal shaft fractures. J Am Acad Orthop Surg 2000; 8: 111e21. 12 Pun WK, Chow SP, So YC, et al. Unstable phalangeal fractures. Treatment by AO screw and plate fixation. J Hand Surg 1991; 16A: 113e7. 13 Agee JM. Unstable fracture dislocations of the proximal interphalangeal joint: treatment with the force couple splint. Clin Orthop Relat Res 1987; 214: 101e12. 14 Williams RM, Kiefhaber TR, Sommerkamp TG, Stern PJ. Treatment of unstable dorsal proximal interphalangeal fracture/dislocations using a hemi-hamate autograft. J Hand Surg Am 2003 Sep; 28: 856e65. 15 Stern PJ, Roman RJ, Kiefhaber TR, et al. Pilon fractures of the proximal interphalangeal joint. J Hand Surg 1991; 16A: 844e50. 16 Scewring DJ, Thomas RH. Avulsion fractures from the base of the proximal phalanges of the fingers. J Hand Surg Eur Vol February 2003; 28: 10e4. 17 Abrahamsson S, et al. Diagnosis of displaced ulnar collateral ligament of the metacarpophalangeal joint of the thumb. J Hand Surg Am May 1990; 15: 457e60. 18 Heyman P, Gelberman RH, Duncan K, Hipp JA. Injuries of the ulnar collateral ligament of the thumb metacarpophalangeal joint. Biomechanical and prospective clinical studies on the usefulness of valgus stress testing. Clin Orthop Relat Res; 1993 Jul: 165e71. 19 Burkhalter WE. Closed treatment of hand fractures. J Hand Surg 1989; 14A: 390e3. 20 Bansal R, Craigen MA. Fifth metacarpal neck fractures: is follow-up required? J Hand Surg Eur Vol 2007 Feb; 32: 69e73. Epub 2006 Nov 27. 21 McElfresh EC, Dobyns JH. Intraarticular metacarpal head fractures. J Hand Surg 1983; 8: 383e93. 22 Ashkenaze DM, Ruby LK. Metacarpal fractures and dislocations. Orthop Clin North Am 1992; 23: 19e33. 23 Foucher G, Chemorin C, Sibilly A. A new technic of osteosynthesis in fractures of the distal 3d of the 5th metacarpus. Nouv Presse Med 1976 Apr 24; 5: 1139e40. 24 Lee SG, Jupiter JB. Phalangeal and metacarpal fractures of the hand. Hand Clin 2000; 16: 323e32. 25 Breen TF, Gelberman RH, Jupiter JB. Intraarticular fractures of the basilar joint of the thumb. Hand Clin 1988; 4: 491e501. 26 Hodgson PD, Shewring DJ. The ‘Metacarpal Cascade Lines’; use in the diagnosis of dislocation of the Carpometacarpal joints. J Hand Surg Eur Vol June 2007; 32: 277e81. 27 Rawles Jr JG. Dislocations and fracture dislocations at the carpometacarpal joints of the fingers. Hand Clin 1988; 4: 103e12.

Fractures and dislocations of the hand are a complicated and diverse group of injuries. Despite this complexity and diversity they share a common theme in management. The hand is a delicate and sensitive organ that requires both flexibility and stability in order for optimal function. Therefore the individual surgeon must tailor treatment in order to achieve an optimal functional outcome for the patient. This may be with splinting for a stable injury, or with CRPP or ORIF for an unstable injury. A good general rule to remember is that the more aggressive the surgical intervention has been, the more aggressive the rehabilitation regime must be. A successful outcome also involves trying to avoid the potential complications such as nerve hypersensitivity and infection. Finally the management of the patient’s expectations are also important: it is crucial that the patient is aware that fractures and dislocations of the hand produce swelling, stiffness and pain that frequently take more than a year to overcome. A

REFERENCES 1 Packer GJ, Shaheen MA. Patterns of hand fractures and dislocations in a district general hospital. J Hand Surg 1993; 18: 511e4. 2 Hove LM. Fractures of the hand. Distribution and relative incidence. Scand J Plast Reconstr Surg 1993 Dec; 27: 317e9. 3 Brydie A, Raby N. Early MRI in the management of clinical scaphoid fracture. Br J Radiol 2003; 76: 296e300. 4 Henry Mark. Fractures and dislocations of the hand. In: Bucholz RW, Heckman JD, eds. Rockwood and Green’s fractures in adults. 5th edn. Philadelphia, PA: Lippincott Williams and Wilkins, 2002. 5 Shrewsbury MM, Johnson RK. Form, function and evolution of the distal phalanx. J Hand Surg 1983; 8: 475e9. 6 Seymour N. Juxta-epiphyseal fracture of the terminal phalanx of the finger. J Bone Joint Surg 1966; 48: B347e9. 7 Cannon NM. Rehabilitation approaches for distal and middle phalanx fractures of the hand. J Hand Ther ApreJun 2003; 16: 2. 8 Wehbe MA, Schneider LH. Mallet fractures. J Bone Joint Surg Am 1984; 66: 658e69. 9 Ishiguro T, et al. Extension block with Kirschner wire for fracture dislocation of the distal interphalangeal joint. Orthoped Traumatol 1999; 7: 105e11. 10 Leddy JP, Packer JW. Avulsion of the profundus tendon insertion in athletes. J Hand Surg Am 1977 Jan; 2: 66e9.

ORTHOPAEDICS AND TRAUMA 25:1

56

Ó 2010 Elsevier Ltd. All rights reserved.

BASIC SCIENCE

The basic science and clinical applications of neurophysiological investigations

lower motor neurones (anterior horn cells) in the spinal cord is conducted through the nerve fibres to muscles or glands. The PNS can also be sub-classified into somatic (external sensation and skeletal muscles) and autonomic (blood pressure, heart rate, glands, bladder, bowel) systems. Nerve Conduction Studies (NCS), Electromyography (EMG) and Somato-sensory Evoked Potentials (SSEPs) are the three main methods of investigation for the efferent and afferent somatic system. Depending on the diagnostic question, one or more of these techniques may be employed. The investigations help to identify and localize the pathology and determine the duration and severity of the pathology in order to aid management and prognosis. The results are also a useful reference for any future change in patients’ condition, including changes following surgical intervention. Other abnormalities in the PNS and muscles can also be picked up, which may or may not necessarily relate to presenting symptoms and signs, yet may be relevant in the wider management plan. For the surgeon, an understanding of the basic physiological principles behind the investigations enables formulation of a diagnostic question to be addressed in any request for investigations, and helps with interpretation of the results in order to establish a treatment plan.

SR Vollans SS Hasan

Abstract Neurophysiological investigations are primarily requested by orthopaedic surgeons in order to investigate sensory or motor dysfunction in the limbs. They often wish to substantiate their clinical diagnosis before proceeding with surgery, or to confirm a nerve injury. A diagnostic query adequately supported by relevant history and examination is essential to determine the type of neurophysiological investigation required to support the diagnosis. A basic understanding of the available investigations, their indications and limitations, is useful to the surgeon for interpretation of the results in light of the clinical picture. Underpinning all neurophysiological investigations is the knowledge of the basic anatomy and physiology of the nervous system. We discuss the generation and propagation of action potentials, which can be generated by the electrical stimulation of a peripheral nerve and responses may be recorded along the course of the nerve or from the muscle (nerve conduction studies). Electrical activity from a muscle can also be recorded with a needle electrode (electromyography), whilst a stimulus in a peripheral sensory nerve can be recorded in the spinal cord or brain (somatosensory evoked potentials). We discuss the aforementioned tests, looking at the indications for their use. Normal patterns of results as well as those that would be considered to be pathological will also be discussed.

Basic physiology Neurones exhibit a lipoprotein cell membrane with a negative membrane potential (around
70 mV), due to a high concentration of intracellular potassium [Kþ] with respect to sodium [Naþ] and chloride [Cl
] ions. The concentration gradient is maintained by the Naþ/Kþ exchange pump. Voltage-gated ion channels also exist and are paramount in the initiation and propagation of an action potential. Action potentials (Figure 1) are key to nerve signalling, occurring as a result of rapid changes in membrane potential. If a nerve fibre is disturbed physiologically or otherwise, there is alteration in the resting membrane potential such that sodium influx into the cell exceeds potassium efflux. A rise in membrane potential occurs and, if this reaches approximately
65 mV, is said to have reached the threshold for stimulation. At this threshold voltage-gated sodium channels open, allowing more rapid ion flux and therefore depolarization, which promulgates a positive feedback cycle until all the sodium channels are open. The depolarized segments set up local circuits with adjacent resting membranes, opening their ion channels, leading to the propagation of a nerve impulse. Sodium channels start to close within a few milliseconds of opening whilst potassium channels open in turn, resulting in membrane repolarization to the steady state and local cessation of action potential. It is important to note that the action potential elicited at a single point in a nerve will travel in all directions and along any branches of the nerve until the whole nerve is depolarized. This is the all-ornone principle in that, with adequate stimulation, an action potential will propagate over the entire length of a nerve fibre both distal and proximal to the stimulus. If the stimulus is sub-threshold, the membrane does not generate sufficient voltage to stimulate the adjacent segment and the propagation either does not occur or is limited. The structural integrity and health of the cell body (neurone), its fibres (axons), receptors and connections are essential for the

Keywords electrodiagnosis; electromyography; evoked potentials; nerve conduction; neurophysiology

Introduction The nervous system consists of central and peripheral elements. Whilst neurophysiological studies are employed to investigate both of the above, the majority of orthopaedic requests are concerned with the peripheral nervous system (PNS). Conventionally the PNS is considered to comprise afferent and efferent pathways. Afferent (sensory) input from the peripheral receptors is propagated in the peripheral nerve to the sensory neurone in the dorsal root ganglion and efferent (motor) output from the

SR Vollans MRCS(Ed) MFSEM(UK) MBChB BSc Hons Specialty Registrar Trauma & Orthopaedics; York Teaching Hospital NHS Trust, Wigginton Road, York YO31 8HE, UK. Conflicts of interest: none. SS Hasan MBBS MD FRCP Consultant in Electrodiagnostic Neurology/ Clinical Neurophysiologist York Teaching Hospital NHS Trust, Wigginton Road, York YO31 8HE, UK. Conflicts of interest: none.

ORTHOPAEDICS AND TRAUMA 25:1

57

Ó 2010 Elsevier Ltd. All rights reserved.

BASIC SCIENCE

degeneration. When demyelination coexists with axonal damage, or is so prolonged and severe that secondary axonal degeneration occurs with time, a poor prognosis can be expected. Demyelination in a nerve can also result from a direct insult, such as mechanical trauma/pressure or ischaemia, or due to focal or widespread segmental inflammation. Demyelination leads to disruption of the effective transmission of action potentials in a nerve, more specifically slowing conduction through the area of demyelination. High numbers of voltage-gated sodium channels are located around the nodes of Ranvier in nerves with a myelin sheath. The influx of sodium ions at these points is vital for the progression of the action potential. A demyelinated segment in close proximity to the nodes results in ‘conduction block’, leading to distal weakness or paralysis (motor dysfunction) and altered or lost sensory perception, despite the fact that conduction remains normal in the unaffected healthy proximal and distal parts of the nerve. This occurs acutely in a ‘Neuropraxia’, which although of immense functional deficit can be temporary. Provided there has been no axonal damage, it will resolve with restoration of myelin integrity in a relatively predictable timescale and associated with a good outcome. Common clinical examples of this are ‘Saturday night’ radial nerve palsy, affecting the radial nerve in the spiral groove, and patients with an acute inflammatory demyelinating polyneuropathy; GuillaineBarre Syndrome (GBS).

Neurophysiological studies Both CNS and PNS function can be monitored by recording activity with the help of appropriately placed electrodes, which can be placed on the skin or invasively within the tissues. Neuronal signals can be recorded with or without electrical stimulation, and in other ways. In rare cases magnetic stimulation is employed to excite the motor cortex or major nerves, in order to record motor potentials from the target muscles in order to obtain what is called ‘Central Motor Conduction Time’. There is a broad international consensus about the normality or abnormality of the obtained technical data but normal values vary widely. Above all, an expert has to make sense of the results in the light of the patient’s clinical condition and it has to be emphasized that an abnormality found may be incidental or entirely irrelevant. While most patients in day-to-day practice can be studied without much difficulty, there may be limitations in some. Electrical noise from other vital monitoring equipment in acutely ill patients on intensive care may result in sufficient significant interference to corrupt the acquired neuronal signal. Surface oedema, breached skin, dressings and plaster casts, inherent or acquired limb deformity can all curtail assessment. Difficulties are also encountered in patients unable or unwilling to comply because of the existing pain or those who are intolerant of the usually unpleasant electrical current stimuli and needles.

Figure 1 Diagram of an action potential with associated ion channel activity.

effective transmission of neuronal signals. Nerve fibres vary in size and length, depending upon their function. Peripheral nerve fibres are covered by a myelin sheath or are unmyelinated. The speed of impulse transmission is determined by the thickness of the myelin sheath. The thicker myelinated large fibres (A-alpha/ beta) are fast conducting and the unmyelinated (C) and thinly myelinated (A-gamma/delta) small fibres are slow conducting.

Pathophysiology Axonal degeneration This occurs either due to neuronal death or axonal injury such that the distal segment becomes disconnected from the cell body. In case of axonotmesis there is axonal disruption with preservation of the endoneurium. The distal axon with its myelin degenerates by Wallerian degeneration; the so-called “dying back” phenomenon.1 Regeneration takes place slowly within the remaining endoneurium at 1e2 mm/day, with scope for relatively complete functional recovery. Neurotmesis is when an injury divides the entire nerve, completely disrupting the axons and their surrounding tissue elements. Here, regeneration is disorderly and distal function may never return.

Nerve conduction studies

Demyelination This pathology affects the myelin sheath or covering, which may degenerate but may also become restored, either partially or completely, in due course. In isolation, when the axons remain intact, this condition has a better prognosis than axonal

ORTHOPAEDICS AND TRAUMA 25:1

The recording and stimulating electrodes are paired, each with an active and remote terminal An earth electrode is also applied to minimize interference from electrical noise and movements. Stick-on, felt-padded or ring electrodes (for fingers and toes) are used, along with needle electrodes when necessary.

58

Ó 2010 Elsevier Ltd. All rights reserved.

BASIC SCIENCE

The all-or-none principle is applied, so when stimulating a nerve the intensity of the electrical stimulus is kept above the threshold (supramaximal) to ensure adequate action potential generation and its propagation in the nerve. Action potentials can be generated and measured by the above-mentioned non-invasive and invasive means. The recorded potential (response) is a waveform; the time from stimulation to the onset of this wave (or sometimes taken at the peak) is called the latency and the magnitude of this waveform the amplitude. By measuring the length of the conducting nerve segment and dividing this distance by time, one can calculate the conduction velocity. A given nerve can be stimulated at more than one site with the response recorded at a fixed point to aid localization of the site of nerve damage. As mentioned above, demyelination in a nerve will result in slower or blocked conduction with prolonged latency and subnormal conduction velocity. The amplitude of the response depends upon the number of the available functioning axons, so a lower amplitude response is obtained when many axons are either damaged or rendered non-functional by severe demyelination in the nerve.2 A summary of pathological findings in NCS can be seen in Table 1.

Figure 2 Diagram demonstrating the setup and measured variables to calculate nerve conduction velocity. (Re; remote electrode, Ae; active electrode, CD; conduction distance, DML; distal motor latency, PML; proximal motor latency).

Sensory nerve action potentials (SNAPs) SNAPs are potentials obtained by stimulating a pure sensory nerve or a sensory branch of a mixed nerve and recording at a given distance. For example, the superficial radial nerve in the distal forearm or sural nerve in the distal leg are both pure sensory nerves. In contrast, when the sensory branches of the median nerve are stimulated in a finger, the response recorded over the nerve at the wrist is a mixed one, as it contains both sensory and motor axons at this level. When sensory conduction is studied in a mixed nerve segment, for example median or ulnar nerve in the forearm, the response obtained is conventionally known as mixed nerve action potential or NAP. Sensory responses in a nerve can be measured by distal excitation that is ‘normal’ afferent direction of propagation in the nerve (orthodromic) or proximal excitation (antidromic) with a distally placed recording electrode.

Compound motor action potentials (CMAPs) CMAPs (Figure 2) are the potentials recorded at a muscle following stimulation of motor nerve fibres. The recording electrode is placed over the end plate of the nerve entering the muscle (Ae) and over a distal electrically inactive point such as the tendon of the muscle (Re). Most peripheral nerves are stimulated at both distal (near to the muscle) and proximal points, hence the conduction time respectively termed as distal and proximal motor latency (DML/PML). Motor nerve conduction velocity (MNCV) is estimated for the nerve segment between the distal and proximal stimulating points; dividing the conducting distance (CD) by the time obtained by subtracting the DML from PML.

Late responses F-waves Following supramaximal stimulation of a motor nerve, the orthodromically transmitted impulse results in depolarization of the muscle end plate and this produces the CMAP recorded from the muscle. There is also antidromic transmission of the generated impulse, which on reaching the axon’s origin in the lower motor neurone in the spinal cord may allow re-excitation of these motor neurones, which in turn results in a weaker depolarization of the same muscle producing a variable smaller potential later in time (Figure 3). This delayed response, first described by Magladery and McDougal in the 1950s3 following recordings in the foot, was termed the “F-wave” and in this context the initial CMAP is named as the “M-wave”. Because of the way in which F-waves are generated, their amplitude and morphology vary with each excitation, as does the time (latency) to produce this late response from the muscle. In one study,4 200 independent stimulations of motor nerve fibres produced no F-waves in 55% of all the motor units tested. The amplitude of the F-waves is usually less than 10% of the M-wave5 and their latency in an otherwise healthy nerve will still depend upon the age of the individual, the length of the limb and the patient’s height. F-waves can be abnormally delayed or absent when the studied nerve has diffuse pathology or focal abnormalities are present anywhere in the pathway that is responsible for F-wave generation. This depends upon the extent and severity of the

MNCV ðm=sÞ ¼ CD ðmmÞ=PML
DML ðmsecÞ Note that MNCV is not a measure of the loss of numbers of fibres, so can still be near normal in the presence of axonal injury/ degeneration.

Summary of the nerve conduction study findings following axonal damage and demyelination

SNAPS

Axonal loss Absent or low amplitude

Demyelination Normal distal to the block Reduced/absent from proximal to block

CMAPs

Low amplitude

Normal distal to block Reduced from proximal to block

Conduction velocity

Normal or subnormal

Reduced across the block

Table 1

ORTHOPAEDICS AND TRAUMA 25:1

59

Ó 2010 Elsevier Ltd. All rights reserved.

BASIC SCIENCE

extraocular muscles have between 13 and 20 muscle fibres per motor unit, as compared to biceps brachii with approximately 600e1700 fibres per unit. The smaller the motor unit, the lesser the force required for activation but the finer is the control that becomes possible, and vice versa. EMG, as the term implies, is the graphic recording of action potentials within the muscles at resting or contracting states. Surface electrodes can record EMG, but more reliable information is obtained by a needle electrode inserted in the muscle of interest. Most commonly used are bipolar ‘concentric needle electrodes’ consisting of a central platinum wire within a steel cannula up to 0.65 mm diameter. Since muscle fibres have diameters of between 25 and 100 mm, approximately 20 fibres should be in close proximity to the needle tip “pick-up” point. Much finer needles are used to record individual muscle fibre potentials and hence named as ‘single fibre needle electrodes’. The EMG of a normal muscle On piercing a relaxed, resting muscle with the needle, there is transient and brief activity due to mechanical stimulation/ disruption producing what is called the insertional activity. ‘Endplate noise’ can also be audible and seen if the needle tip is very close to the motor end plate; this can be eliminated by a small shift of the electrode tip. Asking the patient to voluntarily contract the muscle results in activation of the motor units and one can record the motor unit action potential (MUAP). Individual MUAPs can be defined at weaker voluntary efforts, with activation of smaller motor units, whilst larger MUAPs appear as the effort is increased. As shown in Figure 4, normal MUAP morphology is usually triphasic. The initial phase is due to the action potential travelling from the motor end-plate towards the electrode, the main spike (upward deflection) is caused by activity within the muscle fibres closest to the electrode and the late phase by those fibres that are more distant. MUAP amplitude and duration can be assessed from the recorded trace. The duration, and to an extent size, of the MUAP is determined by the number of functioning muscle fibres within the motor unit and also the muscle fibre type. When the patient is asked to exert maximal and sustained effort, all small and larger motor units are recruited; it becomes difficult to clearly identify

Figure 3 Diagrammatic representation of conduction paths of the F-wave and H-reflex.

pathology and, in cases where large number of motor axons are still functionally intact, F-wave latency can still be normal. Their importance in clinical practice is most apparent when absent or delayed F-waves can be the only neurophysiological abnormality present in the first few days of an acute demyelinating polyneuropathy (GBS). H-Reflex (Figure 3) The H-reflex is a late response that has been much talked about in the past, but is not so commonly used nowadays, since the information obtained often adds little to what can already be judged by a combination of other studies and clinical examination. The H-reflex in adults can be reliably obtained mainly from the soleus muscle in the calf by stimulating the tibial nerve at the popliteal fossa; hence it is akin to the monosynaptic ankle jerk reflex. Submaximal threshold stimuli are needed for appearance of an H-reflex; with increasing intensity stimuli, the amplitude of the M-wave increases at the expense of the H-reflex which tends to diminish and disappear.

Electromyography (EMG) Basic science A motor neurone in the spinal cord, its axon and all the muscle fibres innervated by that particular axon constitute what is called the motor unit. Nerve endings do not directly enter the muscle fibres but terminate at the muscle end plate. Signal transmission to depolarize the muscle occurs chemically at the neuromuscular junction (NMJ). Neurones that innervate striated muscle are named a-motor neurones; they have large cell bodies and thickly myelinated axons and are therefore fast-conducting fibres. Each axon innervates multiple non-adjacent fibres; for example the

ORTHOPAEDICS AND TRAUMA 25:1

Figure 4 An overview of the motor unit action potential (MUAP).

60

Ó 2010 Elsevier Ltd. All rights reserved.

BASIC SCIENCE

individual MUAPs with the baseline obscured in the resulting trace, producing a picture of full interference pattern (Figure 5).

visible as spontaneous muscle twitches to the naked eye. Other spontaneous discharges, such as complex repetitive discharges, are sometimes seen in a denervated muscle. With a number of muscle fibres atrophied following denervation, the surviving fibres move into closer proximity, so the action potential generated now results in larger MUAP amplitudes. Terminal sprouts lead to reinnervation of the affected fibres, in turn leading to increased numbers of functioning fibres in a motor unit. Because the process continues in a variable fashion, the electrical activity in a motor unit is less synchronized, their action potentials differing in a given time, resulting in dispersed summated responses to produce complex shaped, or multiphasic, MUAPs with instability of the potentials as they appear. A motor unit with five or more phases crossing the baseline (turns7) is normally termed as polyphasic in morphology. With time, the recruited motor units in a previously denervated muscle will appear increasingly large, of longer duration and polyphasic shape. Until such time as the reinnervation is complete, the existing motor units in affected muscles have to work harder to produce a given muscle activity so this increased work rate is reflected in high firing recruitment, as the same unit appears repeatedly many times during EMG. Conversely, low firing rates are seen in a genuinely weak non-denervated muscle due to upper motor neurone pathology and also when the patient’s effort to contract the muscle is voluntarily poor. The depleted functional motor unit numbers also result in a broken or reduced interference pattern (neurogenic or neuropathic recruitment). It is important to realize that in many cases after denervation, all affected muscle fibres will not become adequately, or not at all, reinnervated. The EMG will therefore still show signs of previous denervation with remodelled, large amplitude, long duration polyphasic motor units years after the original insult.

The EMG in an abnormal muscle In the context of the PNS, muscle function can be affected due to pathology at the NMJ or as result of denervation due to damaged motor axons or of the neurone itself. Muscles can also be the site of the primary disease, as in various myopathies or inherited dystrophic pathology. The affliction can be acute, subacute or chronic; the process can be static, stable or due to a progressive pathology. The causes are many, but use of EMG in many cases is an excellent tool to determine the nature of the pathology responsible for muscle dysfunction, either in isolation or as part of a more extensive process; it can also provide information with regard to timing of the insult, severity of the pathology and prognosis in a given patient. The EMG following muscle denervation Denervation in a muscle can result from an insult to nerve axons in the limb, at the plexus, the root or damage to the neurone itself in the spinal cord. Usually muscles are partially denervated, as injury resulting in complete section of the whole nerve to a given muscle is rare except in trauma. Some motor units in a partly denervated muscle are going to be functionless or significantly distressed, leading to atrophy of many muscle fibres. Along with neuronal growth and sprouting (the reinnervation process) the affected motor units start to remodel in shape and size, which changes the morphology of the MUAPs with time. On inserting the needle electrode in a relaxed, denervated muscle, the normally brief insertional activity can be prolonged or increased and other discharging potentials appear spontaneously, depending upon the timing of the neuronal insult. Following a recent nerve injury, or during an active denervation process, the affected muscle is no longer ‘silent’ at rest but exhibits ‘spontaneous activity’ in the form of fibrillations, positive sharp waves or fasciculation potentials. Fibrillation and positive sharp waves reflect spontaneous activity in a single muscle fibre that can only be seen with needle EMG. These usually appear after a few days to 3 weeks following the injury, depending upon the length of the denervated distal nerve segment.6 They either gradually subside with muscle atrophy, or cease upon reinnervation. Fasciculation potentials are generated by depolarization of a group of muscle fibres and may also be

The EMG in primary muscle disease In various types of acquired or inherited myopathy and muscular dystrophy, there is random loss of muscle fibres, thus a motor unit comprises reduced numbers of fibres, many of which can also be functionally poor, affected by the pathology. MUAPs appear short in duration, probably due to elimination of the initial and late phases contributed by the activity in the end plate and remote fibres (Figure 5). While some MUAPs appear normal in shape, variable summation of the electrical potentials in many fibres also results in multiphasic MUAPs, which are still of brief duration, as opposed to what is seen in a neurogenic unit (Table 2). The complex pathophysiology in diseased muscle produces late satellite potentials, as well as smaller amplitude MUAPs in the vast majority of cases. When a muscle is contracted, all available normal and diseased muscle fibres in a myopathic muscle have to come into action, even with weaker force. Hence the resulting interference pattern appears full, with small amplitude, brief duration motor units; myogenic or myopathic recruitment.

Somato-sensory evoked potentials (SSEPs) The presence of the so-called sensory evoked response was first described by Richard Caton (Liverpool, UK) in the 19th century.8 Low threshold electrical stimuli are used repeatedly to stimulate peripheral sensory nerve afferents in the limb, which enables generation of a small amplitude potential that can be recorded

Figure 5 The full interference pattern demonstrating the absence of distinct MUAPs.

ORTHOPAEDICS AND TRAUMA 25:1

61

Ó 2010 Elsevier Ltd. All rights reserved.

BASIC SCIENCE

Summary of the EMG findings within normal, neurogenic and myopathic muscle Normal

Neurogenic

Myopathic

Insertional activity

Normal

LMN [

UMN Normal

Myopathy Normal

Myositis [

Spontaneous activity

Absent

Present

Absent

Absent

Present

MUAP

Normal

[ Amplitude [ Duration

Normal

Y Amplitude Y Duration

Y Amplitude Y Duration

Interference pattern

Full

Reduced High firing units

Reduced Low firing units

Full Y Amplitude

Full Y Amplitude

Table 2

over the CNS structures such as the spinal cord (dorsal column) and the scalp surface (sensory cortex). Usually, major nerves such as the median at the wrist or tibial at the ankle are stimulated. In the case of median nerve stimulation an evoked potential is usually also recorded at the brachial plexus (Erb’s point). SSEPs are time and labour intensive and their value in clinical practice is limited, as a diagnostic query can be reliably answered by other investigations in many cases. The use of SSEPs has thus declined in recent years, even in conditions such as multiple sclerosis where it can be used to address the question of sensory loss in the extremities reported by patients evidently due to demyelinating pathology in the spinal cord or brain. Perhaps the most important role of SSEPs within orthopaedics is in intra-operative monitoring (IOM), reported for the first time in the 1980s; SSEPs still form the most commonly used IOM assessment. Nuwer et al with the European Scoliosis Group performed multi-centre trials in the early 1990s showing the efficacy of SSEPs in reducing the risk of iatrogenic cord injury during scoliosis corrective surgery.9 Commonly, the tibial nerve at the ankle is stimulated and recording electrodes are placed at the level of the second cervical vertebra and on the scalp, overlying the sensory cortex or directly from the epidural space. Subcortical measurements are usually more sensitive during the procedure, as cortical activity decreases with general anaesthesia. The recording electrodes are sensitive to electrical interference from other monitoring equipment the patient is attached to, and to electrical noise such as that generated by the use of diathermy. Other limitations include the delay in acquisition of results, as the time consuming averaging of 300 stimuli must take place in order to obtain a reliable response; thus any change in the recorded response reflects what may have happened minutes earlier. In addition, only the dorsal columns are being monitored by SSEPs so damage elsewhere in the cord may still go undetected and presence of normal SSEPs during IOM may misleadingly provide false reassurance.

reporting chronic locomotor problems associated with local nerve damage or a primary neurological condition leading to secondary musculoskeletal limb disorders or deformities. Examples of the latter scenario would include conditions such as old poliomyelitis or Hereditary Motor and Sensory Neuropathy (HMSN).9 HMSNs are classified on the basis of neurophysiological results as well as genetic testing. A common presentation is in the first or second decade with the development of pes cavus and wasting of the intrinsic foot muscles and peronei, hence the term “peroneal muscular atrophy” used in the original descriptive paper.10 Whilst gross sensory disturbance is noted, disability is predominantly secondary to motor impairment, which is often how they present. As sensation is severely diminished or even absent, assessment of the motor nerves is employed. Type I HMSN, known as Demyelinating CharcoteMarieeTooth (CMT), leads to a marked reduction in MNCV. Type II, known as Neuronal or axonal-type CMT demonstrates more normally preserved MNCV on testing. Patients may present with either sensory and motor symptoms combined, or one of these in isolation. Some also have symptoms of significant autonomic dysfunction. A careful history, with sound knowledge of anatomy and physiology, at least narrows the diagnostic conundrum in many cases. It must be noted that many sensory symptoms, along with pain and weakness, are also reported by patients with no pathology in their PNS, but centrally within the brain or spinal cord. In addition, some musculoskeletal disorders present with subjective symptoms and signs related to the PNS which may in fact be disproved on testing; weakness or muscle wasting seen in osteoarthritis or localized numbness associated with severe tendonitis, for example. Suspected diagnoses can be further refined after examination, including relevant neurological examination. For example, a good ankle jerk reflex in someone with a unilateral drop foot would lead one to consider either a peroneal nerve lesion or L5 radiculopathy rather than tibial or sciatic nerve involvement. Similarly a patient presenting with a globally wasted and weak hand without pain or other sensory symptoms in the limb is unlikely to have neurological damage anywhere from the roots distally, but more likely has pathology in the anterior horn cell, alerting a search for possible signs of more widespread motor neurone disease.

Clinical presentations and neurophysiological studies In the orthopaedic setting, patients presenting with symptoms and signs of PNS dysfunction largely fall into two groups. The first is that referred to the surgeon with suspected or clinically definite compressive neuropathy or possible radiculopathy. The second group of patients is that with traumatic nerve damage as a result of accidents or surgical interventions. In addition, there are patients

ORTHOPAEDICS AND TRAUMA 25:1

Investigatory overview Even in patients where a clinical diagnosis is not really in doubt, neurophysiological studies are useful and on occasions vital in

62

Ó 2010 Elsevier Ltd. All rights reserved.

BASIC SCIENCE

REFERENCES 1 Cavanagh JB. The significance of the “dying back” process in experimental and human neurological disease. Int Rev Exp Pathol 1964; 3: 219e67. 2 Gilliatt RW. Sensory conduction studies in the early recognition of nerve disorders. Muscle Nerve 1978; 1: 352e9. 3 Magladery JW, McDougal Jr DB. Electrophysiological studies of nerve and reflex activity in normal man. I. Identification of certain reflexes in the electromyogram and the conduction velocity of peripheral nerve fibers. Bull Johns Hopkins Hosp 1950; 86: 265e90. 4 Schiller HH, Stalberg E. F-responses studied with single fibre EMG in normal subjects and spastic patients. J Neurol Neurosurg Psychiatry 1978; 41: 45e53. 5 Peioglou-Harmoussi S, Fawcett PR, Howel D, Barwick DD. Fresponses: a study of frequency, shape and amplitude characteristics in healthy control subjects. J Neurol Neurosurg Psychiatry 1985; 48: 1159e64. 6 Buchthal F. Spontaneous electrical activity: an overview. Muscle Nerve 1982; 5: S52e9. 7 Willison RG. Analysis of electrical activity in healthy and dystrophic muscle in man. J Neurol Neurosurg Psychiatry 1964; 27: 386e94. 8 Caton R. The electric currents of the brain. Br Med J 1875; 2: 278. 9 Nuwer MR, Dawson EG, Carlson LG, Kanim LE, Sherman JE. Somatosensory evoked potential spinal cord monitoring reduces neurologic deficits after scoliosis surgery: results of a large multicenter survey. Electroencephalogr Clin Neurophysiol 1995; 96: 6e11. 10 Thomas PK, Calne DB, Stewart G. Hereditary motor and sensory polyneuropathy (peroneal muscular atrophy). Ann Hum Genet 1974; 38: 111e53.

defining and localizing the pathology. They will also provide information on the severity and duration of the pathology, indicate any additional investigations that may be needed and help to direct the choice of treatment interventions and eventual prognosis. Electrodiagnostic assessment begins with a relevant history and targeted neurological examination to define which nerves and muscles require studying, and to what extent, in order to arrive at a diagnosis compatible with the presenting clinical signs. SNAPs and CMAPs are recorded and, if considered necessary, needle EMG of the relevant muscles is carried out. In some cases, depending upon the history or physical examination, or as the nerves and muscles are studied and their responses recorded, an assessment needs to be extended to obtain a logical conclusion. The above exercise enables the determination of whether the pathology is in the peripheral nerve, plexus, root, or at the motor neurone level. In a peripheral nerve the pathology may be in the nerve compartment due to compression or pressure damage, or it may lie outside the compartment. In the latter, an inflammatory process such as acute or chronic demyelination with ‘conduction block’ or ischaemic/vasculitic damage to nerves causing significant distal ‘axonal loss’ will become evident. Even patients presenting with compartment nerve lesions may not be ideal for surgical intervention by means of decompression as they may have a chronically damaged nerve at this site, with irreversible axonal loss. In rare cases of hereditary neuropathy due to susceptibility to pressure palsies (HNPP), a subject may present acutely for the first time with altered sensation and weakness in one nerve territory (for example radial nerve compression palsy in the spiral groove or ulnar nerve at the cubitus) but a neurophysiological assessment will often pick up clinically silent abnormalities in other nerves to aid the wider future management of the patient including familial implication of such a diagnosis.

FURTHER READING American Association of Neuromuscular & Electrodiagnostic Medicine, www.aanem.org/. British Society for Clinical Neurophysiology, www.bscn.org.uk/. European Federation of Neurological Societies, www.efns.org/.

Conclusion Both the CNS and PNS are vital to the appropriate functioning of the musculoskeletal system. Any slight aberration in the delicately balanced system may manifest itself as a functional deficit. By understanding the principles of electrical signalling within the nervous system and applying them to neurophysiological studies, one can draw conclusions regarding the underlying pathophysiology of a specific set of symptoms and signs. In addition, by recognizing the limitations of the various investigations available, it is hoped that an appropriate clinical question can be raised which may be confirmed or refuted by focused investigations. A

ORTHOPAEDICS AND TRAUMA 25:1

Acknowledgements Thanks to Janet Rainford and Claire Hopwood (Neurophysiology Department, Leeds General Infirmary, UK) for the help and information they provided on intraoperative SSEP monitoring during scoliosis corrections.

63

Ó 2010 Elsevier Ltd. All rights reserved.

MEDICO-LEGAL

Medical negligence: legal theory and surgical practice

Duty of care Introduction It is an interesting observation that in jurisprudence the concept that there may be a general duty of care between people arose relatively late (the early 20th century),2 whereas laws relating to contract had been present for hundreds of years. The existence of a doctorepatient relationship will, in almost all cases, impose a duty (or duties) upon the doctor in respect of the care provided to the patient. If there is doubt as to whether the defendant (D) owes a duty of care to the claimant (C) the following tests are applied3: there must be (i) foreseeability of damage (ii) proximity of relationship and (iii) it must be fair, just and reasonable to impose the duty (to which some authorities add) (iv) public policy. The standard of care is the objective and legal test by which the duty of care is measured. Ultimately what actually happened is determined by the Court on the basis of the evidence. What actually happened is then compared to what should have happened. Breach of duty of care may be established by the Court if there is a difference between what actually happened and what should have happened. Breaches of duty can arise by commission (doing something which should not have been done) or omission (not doing something that should have been done). C must demonstrate that a breach of duty of care occurred. In medicine, bad outcomes can occur even if the highest standards of care are employed. C may attempt to say that because they suffered severe injuries as a consequence of surgery this implies negligent care, i.e. “the thing speaks for itself” (res ipsa loquitur). However “res ipsa” will seldom, if ever, succeed if C cannot prove a negligent act or omission.

NV Todd

Abstract Our medical decisions and actions must be made in accordance with accepted standards of care. Surgeons should have a working knowledge of the law so that they can practice in a reasonable, responsible and rational manner. The key areas of law for surgeons are (i) duty of care, (ii) causation and (iii) consent. The standard of care is the objective, legal test by which duty of care is assessed. The tests of breach of duty of care are objective. The primary test for doctors is the Bolam test which states that if your actions (or omissions) would be the usual practice of a reasonable or responsible group of doctors then you satisfy the test. If breach of duty of care is established, the claimant must demonstrate that the breach caused some harm. The test is but-for causation, i.e. if it is established that but-for the breach the bad outcome would not have occurred the claimant will be able to link the breach to the damage done. Consent for surgery is a specific duty of care for doctors. Consent is based upon the principle of autonomy. Consent must be to all material risks of surgery. For surgical procedures, consent must be in writing.

Keywords breach of duty; causation; consent; law; surgery

The Bolam test The test of breach of duty of care is objective and is based upon a test of reasonableness. For doctors the primary test of breach of duty of care was set out in Bolam v Friern Hospital Management Committee [1957].4 Mr Bolam underwent electroconvulsive therapy (ECT) for a depressive illness. ECT was given without anaesthetic. During the convulsive phase of the induced seizure he fractured his pelvis. Mr Bolam sued, claiming that he should have been anaesthetized. It was agreed that if Mr Bolam had been anaesthetized the fracture would not have occurred. In the 1950s there were two schools of thought, the first that ECT induced a chemical change in the brain which would occur whether the patient was anaesthetized or not; the second held that the patient needed to be aware of ECT for the effect to occur and therefore anaesthesia was contraindicated. McNair J held that for doctors:

Introduction Understanding medical negligence is an important part of surgical practice. Our patients have the right to expect that our medical decisions and actions will be made in a professional, knowledgeable and logical way that is commensurate with the medical practice of our peers. Surgeons should have a working knowledge of the law of medical negligence, and its practical applications, in order to practise in a reasonable, responsible and rational manner. This article will focus upon three areas of law that are of particular relevance to surgeons: duty of care, causation and consent. Negligence begins if a legal duty to act, or not to act, (a duty of care) is breached. The claimant must also identify some damage that has occurred, and that that damage is causally related to the breach of duty. The claimant seeks damages for the harm that has been caused (not necessarily every loss that has occurred).1 Thus liability in negligence arises when (i) there is a breach of duty of care which is owed by the defendant (D) to the claimant (C), (ii) that duty is breached and (iii) the breach causes consequential harm. This usually leads to (iv) financial compensation for the harm (damages).

“The test is the standard of the ordinary skilled man exercising and professing to have that special skill. A man need not possess the highest expert skill at the risk of being found negligent. There may be one or more perfectly proper standards; and if a medical man conforms with one of those proper standards then he is not negligent.” That is to say C has to prove that there was no reasonable or responsible group of doctors who would have acted in the manner claimed to be negligent. If the doctor can (as in Bolam) identify a group of doctors that would have acted in the same

NV Todd MD FRCP FRCS Consultant Neurosurgeon and Spinal Surgeon, Regional Neurosciences Centre, Newcastle General Hospital, Westgate Road, Newcastle upon Tyne NE4 6BE, UK. Conflict of interest: none.

ORTHOPAEDICS AND TRAUMA 25:1

64

Ó 2010 Elsevier Ltd. All rights reserved.

MEDICO-LEGAL

way and if that practice is found by the Court to be reasonable and responsible D will succeed even if the group of doctors D relies upon is small. The Bolam (“reasonable and responsible” doctor) test applies to treatment and clinical judgement.5 Dr Jordan delivered Mrs Whitehouse’s baby with forceps. The birth was difficult and prolonged, eventually a Caesarean section was performed. Baby Whitehouse had severe brain injuries. At the initial trial (the trial of first instance) Dr Jordan was held to be liable. On appeal to the House of Lords he was found not to be negligent because the forceps delivery had been carried out reasonably, applying the Bolam “reasonable and responsible doctor” test to clinical judgement. The Bolam test also applies to diagnosis6 and consent.7 In other words all aspects of a surgeon’s practice will be judged by the same, objective test, that of the reasonable and responsible surgeon.

latter view and found in favour of C. On appeal the first instance Judgement was overturned because the Judge had not found one opinion to have been unreasonable, he had accepted that both opinions were reasonable, i.e. there was a reasonable and responsible group of doctors who would have acted as D did. Judges cannot prefer one reasonable opinion against another. If there are two (or more) legitimate standards of care then a doctor who adopts one (or the other) standard of care will not be in breach of duty.9 Practice points In practice the test is considered by the Court based upon the evidence of medical experts plus standards established previously in case and/or statutory law. In our adversarial system of justice, experts will be instructed by both C and D (single joint experts are uncommon in medical negligence cases) and their views may be tested in Court. Doctors must be judged by their peers. It would obviously be inappropriate for the standard of knowledge and care of a Consultant Neurosurgeon to be applied to a General Practitioner. If an Orthopaedic Surgeon taking “general trauma” on-call accepted and managed a patient with an evolving spinal abscess his acts or omissions would be judged by a general orthopaedic expert. If the same patient had been admitted under the care of a specialist spinal orthopaedic surgeon a different, possibly higher, standard of care might apply. In many cases there will be agreement between the expert(s) for each side. If not, each side leads evidence in Court to support its views and the judge decides. On a practical note this is the moment for evidence from basic, simple, undergraduate or postgraduate textbooks, not specialist journals. If, for example, in a case of delayed diagnosis of cauda equina syndrome, every undergraduate textbook states that emergency MRI is required in patients with bilateral radicular leg pain, poor bladder control and/or perineal sensory loss, then it will be difficult for the defendant surgeon to justify a failure to perform MRI under those circumstances. By contrast, if either expert relies upon one case report in the Journal of Exceptionally Rare Orthopaedic Disorders the Judge might rightly conclude that that expert is proposing an exceptionally high standard of care, not the standard of a responsible group of ordinary orthopaedic surgeons. The reasonable surgeon must keep up to date with changing knowledge in the speciality. Once again the test is that of the reasonable and responsible surgeon, not the most up to date Professor of Surgery. Shipbuilders have a duty of care to protect the hearing of their workforce. The availability of effective earprotectors were published in the Lancet in 1951 (not a journal read by many shipbuilders!) but it was found that there was no breach of duty until Government advice was sent out in 1963.10 It is likely to be reasonable for a surgeon not to be aware of a rare risk published in a single specialist journal but he will need to be aware of general guidelines, such as those published by NICE. In Shakoor and Situ11 D prescribed a traditional Chinese herbal medicine for a skin disorder. C suffered acute liver failure and died. Although there had been publications, including letters in the Lancet, describing this complication there had been no such report in the Chinese herbal medicine literature. D succeeded in that the Chinese herbal medicine practitioner exercising ordinary skill need not be aware of letters in the Lancet.

The Bolitho test A further test (of duty of care) is that the acts or omissions of the doctor must have a logical basis.8 Baby Bolitho was a 3-year-old child with croup. He was admitted to hospital with breathing difficulties. An anaesthetic senior registrar attended and said that intubation was not required. Bolitho’s breathing improved. There was then a second episode of respiratory embarrassment. The anaesthetist returned and again said that intubation was not needed. Bolitho’s breathing again improved. There was a third episode of respiratory embarrassment, the anaesthetist was contacted but did not attend, on the basis that she would make the same decision as earlier. Baby Bolitho died. The experts (many) disagreed as to whether the anaesthetist should have attended on the third occasion. The failure of the anaesthetic registrar to attend on the third occasion was found to fall below the appropriate standard (she should have attended). However if she had attended, the Court found that it would have been reasonable for her not to have intubated. Her evidence, based on the two previous attendances was that she would not have intubated. Because not intubating was within the range of reasonable practice and had a logical basis the breach (failure to attend) was not causally linked to the harm (failure to intubate); the claim failed. In Maynard,6 Lord Scarman referred to a “respectable” body of professional opinion. Following Bolitho,8 to the more traditional adjectives, responsible, reasonable and respectable must be added to the term rational (having a logical basis). Preferring medical evidence The Judge is not entitled to favour one body of medical opinion over another provided the views of each body are legitimate, i.e. reasonable and responsible. In Maynard v West Midlands Regional HA,6 Maynard had mediastinal lymphadenopathy without lung lesions. Tuberculosis (TB) was the most likely diagnosis; sarcoidosis or malignancy were possible diagnoses. D performed mediastinoscopy to obtain a biopsy. This caused a permanent recurrent laryngeal nerve palsy. The experts were divided: one group said that mediastinoscopy was reasonable. The other said that there should have been an initial trial of antiTB chemotherapy; mediastinoscopy would never have been required. In the trial at first instance, the Judge preferred the

ORTHOPAEDICS AND TRAUMA 25:1

65

Ó 2010 Elsevier Ltd. All rights reserved.

MEDICO-LEGAL

State of knowledge at the time of the breach The standard of care is based upon the state of knowledge at the time of the breach. In Roe12 the patient, in 1947, was given a spinal anaesthetic. The anaesthetic agent was contained in a glass vial which had been stored in phenol. The phenol had contaminated the anaesthetic agent through “invisible cracks”. Roe was permanently paralyzed when the phenol contaminated anaesthetic was injected into the CSF spaces. The risk of invisible cracks was not widely known until 1951. The case was heard in 1954. The anaesthetist was not held to be liable on the basis of the knowledge available in 1947. As Denning LJ said: “We must not look at the 1947 accident with 1954 spectacles”.

may overburden resources and, if an individual is forced by circumstances to do too many things at once, the fact that he does one of them incorrectly should not lightly be taken as negligence”. Also in Wilsher17 a majority of the Court of Appeal held that inexperienced doctors were not allowed to perform at a lower standard than their more experienced colleagues and the standard of care is not lower for inexperienced doctors. As Glidewell LJ said: “In my view, the law requires the trainee or learner to be judged by the same standard as his more experienced colleagues. If it did not, inexperience would frequently be urged as a defence to an action for professional negligence. If this test appears unduly harsh in relation to the inexperience, I should add that, in my view, the inexperienced doctor called on to exercise a specialist skill will, as part of that skill, seek the advice and help of his superiors when he does or may need it. If he does seek such help, he will often have satisfied the test, even though he may himself have made a mistake.”

The legal standard of care is a matter for the Court It must be understood that a determination of negligence is a matter for the Court. Expert surgeons advise the Court but the Court is at liberty to reject the views of experts even where the expert views are unanimously agreed. In Hucks v Cole13 a patient suffered a Streptococcal infection which was treated with erythromycin. Sensitivities showed that the Streptococcus was not sensitive to erythromycin, it was sensitive to penicillin. The antibiotics were not changed, serious infection ensued. All of the experts stated that management was of an appropriate standard. The judge found that the expert views were not reasonable on the basis that: (i) D was aware that the infecting organism was not sensitive to erythromycin; (ii) changing to penicillin was cheap, and easy to do and (iii) the relative risks and benefits of the decision cannot properly have been assessed. Hucks v Cole13 was supported in Bolitho,9 that, rarely, all expert opinion may lack a logical basis and therefore must be rejected.

Review your own practice All surgeons should consider the following advice regarding their clinical practice: The practice must accord with a “reasonable, responsible, respectable and rational” group of surgical peers. This group may be small. If your acts/omissions are, broadly, in accordance with your peers you will, probably, be fine. If you are, in one or more areas of your practice, “out on a limb” and/or if you regularly find yourself in disagreement with your colleagues, check that your practice would survive the “reasonable, responsible, respectable and rational” test. How easy would it be to get a colleague to support you? Your actions should accord with your surgical peers, i.e. specialist bone tumour surgeons or spinal surgeons would be required to act in accordance with reasonable bone tumour or spinal surgical practice, respectively. A test of logic will be applied: make sure that you can logically explain why you took the actions you did. You must keep, reasonably, up to date: knowledge, and the application of medical knowledge, changes. Where a change in practice becomes broadly adopted, your practice must change too. Be capable of defending your actions with reference to broad, basic, widely available texts or national guidelines (e.g. NICE guidelines) rather than rare, esoteric or personal references.

Special situations The standard of care might vary depending upon special circumstances. Scarce resources can provide a partial defence to negligence. In Knight v Home Office14 there were insufficient resources to provide continuous observation of a mentally ill patient in a prison hospital wing. The patient successfully hanged himself. Because the doctors were acting within the resources available the prison doctors were found not to be negligent. However, the Court will determine what the minimum standard of care should be. In Brooks v Home Office15 a remand prisoner had a high-risk pregnancy. Her referral for obstetric assessment was delayed which led to death of one of the twins. It was found that the standard of care should be the same for all pregnant women. A further example of a minimum standard of care is found in Bull v Devon AHA16: one of Mrs Bull’s twins was severely injured because, during a difficult labour, it took over an hour for an obstetric registrar to arrive. This was found to be an unacceptable standard of care. It may be the case that in an emergency situation a lower standard of care is acceptable. For example, it would not be reasonable to expect a doctor, at the roadside, who is attempting to deal with injuries caused by a major motorway accident, to achieve the same standard as doctors in a properly equipped emergency department. In Wilshire17 Mustill LJ said:

Gross negligence The acts, or omissions, of a doctor can cause the death of a patient. Usually where a negligent act or omission is identified there will be no criminal charge in relation to the patient’s death. However, if the act or omission is found to be “grossly negligent” the doctor can be charged with gross negligence manslaughter. The test is that the negligence must be so gross as to justify a criminal conviction. Dr Adomako18 was an anaesthetic registrar who took over the anaesthetic management in the middle of a case. The ventilation circuit accidentally became disconnected. A number of alarms (O2, CO2 and finally BP alarms) began sounding; Adomako’s response

“I accept that full allowance must be made for the fact that certain aspects of treatment may have to be carried out in what one witness. called “battle conditions”. An emergency

ORTHOPAEDICS AND TRAUMA 25:1

66

Ó 2010 Elsevier Ltd. All rights reserved.

MEDICO-LEGAL

tea that was subsequently found to be contaminated with arsenic. They attended A&E. Dr Banarjee (who was phoned by a nurse) did not attend and told the nurse to send them home, to call their GPs. The men died. Breach of duty of care was established (for the failure to admit to hospital). However, causation failed because the likely presumptive diagnosis would have been food poisoning (no-one would have initially considered arsenic poisoning) and the standard management of food poisoning would not have prevented the deaths. Because it was highly probable (virtually certain) that the deaths would have occurred even had the negligence not occurred, the claimant’s case failed. In many medical cases the unfortunate outcome may have had many potential causes of which only one was caused by the negligence. Unless the contribution of the negligent cause is over 50% the claimant will usually fail to prove their case. Martin Wilsher17 was a premature baby who became blind. He was negligently given excess oxygen, which is a cause of blindness in premature babies. However, Martin had four other potential causes of blindness (apnoea, hypercarbia, intraventricular haemorrhage and a patent ductus arteriosus). The House of Lords found that it was not possible to say that the negligent treatment with excess oxygen was the probable cause of blindness (just that it was a possible cause) and the claimant’s case failed.

was to turn off the alarms (rather than investigate the cause of the problem). The patient died of an hypoxic cardiac arrest. A number of anaesthetic experts advised that Adomako’s response was so poor that it represented gross negligence. Adomako was found guilty of gross negligence manslaughter and he was jailed. Although such a finding is rare, many doctors are not aware that their negligent management of a patient (even if managed in good faith) could lead to a jail sentence. Summary The Court determines whether a doctor’s duty of care to the patient has been breached by applying an objective legal test, the standard of care. The appropriate standard of care is that of the reasonable, responsible and rational doctor skilled in that particular branch of medicine. The standard is based upon the knowledge and practice at the time that the alleged breach occurred.

Causation Introduction A successful claim in negligence leads to the payment of damages. Damages compensate the claimant for the harm that has been suffered. The claimant must prove the harm suffered and the harm must be causally linked to the breach of duty that occurred. Only the harm that is causally linked to the breach leads to recovery of damages.1 There may be many potential causes of the patient’s final disability, not least that they had a primary medical condition requiring treatment. The test is “all or nothing”: if the claimant establishes causation for the harm they recover damages in full. If causation is not established the claimant recovers nothing. The claimant carries the burden of proving that the breach of duty that is alleged was causally related to the harm alleged.17 The test for causation is “on a balance of probabilities”: in other words the claimant must establish there was a more than 50% likelihood that, absent the breach, the harm would not have occurred. If the adverse event leads to a patient’s injury the patient recovers 100% of the quantum of damages. Conversely, if the likelihood of the adverse event occurring is less than 50% then the assumption is it would not have occurred, despite the negligence, and the claimant fails entirely, receiving no damages (not a percentage of the quantified damages). Loveday and Renton19 was a class action seeking compensation for brain injury following the administration of pertussis vaccine. Although there were reports and some evidence linking the use of this vaccine with brain injury the Court found that the proposed causal link was not more than 50% likely, i.e. the claimants failed the balance of probabilities test for causation.

Chester v Afshar modified causation A patient may be able to demonstrate causation on the basis of public policy and patient autonomy. Mrs Chester21 had longstanding low back pain and leg pain; MR imaging had demonstrated lumbar canal stenosis at several levels. Mr Afshar, a neurosurgeon, recommended lumbar laminectomies at several levels. Mrs Chester was concerned about undergoing surgery and brought a friend to the consultation, who took notes. It was agreed that the risk of surgery causing permanent cauda equina syndrome (CES) was never mentioned. Mrs Chester had her operation (privately) a few days later; there was intraoperative damage to the nerve roots of the cauda equina. Five law Lords considered the case. It was decided that (i) the risk of CES was 0.9e2% (in fact probably 0.1e0.5%) (ii) the risk should have been explained and (iii) if told of this risk Mrs Chester would have deferred her operation. A 3:2 majority found for Mrs Chester, largely on the basis that patient autonomy is a basic right which on the grounds of public policy should be upheld. Mr Afshar, in failing to explain the risk of CES, deprived Mrs Chester of the opportunity to consider that risk and, as accepted by the Court in this case, if the risk of CES had been explained Mrs Chester would not have had surgery at that time (and therefore the CES would not have occurred at that time). It was accepted that this was a departure from classical causation. The two dissenting Law Lords accepted that the risk of CES should have been explained to Mrs Chester. They argued that the case failed on causation e that (i) if Mrs Chester would have had the operation on some occasion and (ii) if the risks of CES were the same then (iii) she was just as likely to suffer CES subsequently (i.e. the earlier operation led to CES occurring on that occasion but it would have occurred in any event). In this author’s, respectful, view this may not be correct. In most cases we do not know why a patient has a postoperative CES. In those cases where the cause is known, views can be taken as to whether the CES would have occurred if surgery had not taken

The “but-for” test The “but-for” test is used to determine whether there is a causal link between the breach of duty alleged and the harm suffered. If (i) absent the breach, the harm would not have occurred and (ii) harm followed the breach then (iii) the breach of duty caused the damage. If the same damage would have occurred even if the breach of duty had not occurred then “but-for” causation is not present and the breach did not cause the damage. In Barnett v Chelsea and Kensington Hospital Management Committee20 three nightwatchmen began vomiting after drinking

ORTHOPAEDICS AND TRAUMA 25:1

67

Ó 2010 Elsevier Ltd. All rights reserved.

MEDICO-LEGAL

place on that occasion. Where the cause of intraoperative CES is not known it may be that there are multifactorial causes. For example, it may be that it needed the coincident (i) excessive retraction of the dura plus (ii) a very incomplete discectomy/ decompression plus (iii) a transient drop in the blood pressure etc. to cause the CES. If an operation had been carried out on a different occasion CES would not have occurred unless all of the factors had occurred on that occasion and to the same degree. Although the overall risk of CES is the same with an operation carried out on any occasion, unless there is a particular susceptibility to the risk of CES eventuating in a particular patient, then it is improbable that all of the necessary, and inter-relating factors, would have occurred on a different occasion. Where the causative factor (or factors) are not known it seems, to this author, to be difficult to say that if the operation had been carried out on a different occasion the result would have been the same. Put another way, if the risk of CES, on any occasion is 0.1%, then on any occasion there is a 99.9% probability that CES will not occur. This approach would not apply if the cause of the harm would probably have occurred on any occasion: for example massive unexpected haemorrhage in a patient who was not known to have a congenital bleeding disorder. In such a case, assuming the bleeding disorder was not known on the subsequent occasion, a similar poor outcome would probably have occurred.

House of Lords upheld the standard balance of probabilities test (which Mr Gregg manifestly failed) and rejected the “loss of a chance” claim. However, Lord Nicholls (in a dissenting judgement) was concerned at the potential injustice of this decision in patients whose prognosis was poor in any event. Lord Nicholls noted that if a patient with cancer has a 40% chance of survival with proper treatment which is reduced to 0% because of a negligent delay in treatment there is a real loss to the patient and his family. Thus a patient with a 60% chance of recovery which is reduced to 20% because of the negligent omission, succeeds in full (100%), whereas the same figures applied to a patient with only a 40% chance of recovery with standard treatment leads to loss of the entire claim. This is, as Lord Nicholls said “.rough justice indeed”. Application of the balance of probabilities test seems to mean that, where the outcome is likely to be the same in any event (better than 50% likelihood) doctors can do as they please, making as many breaches of duty as they care to, because the claimant will never be able to demonstrate, on a balance of probabilities, that the actions or omissions of the doctor(s) caused them harm. If this is the legal position, it is, in this author’s view, a curious position. When considering probabilities it is worth noting that the statistical probability of a harmful event occurring in previously reported patients is not the same as the probability of cause and effect in a specific case.24

Loss of a chance is no loss A negligent act, or omission, may reduce the chance of benefit, but unless that chance is better than 50% the claimant will fail. If breach is established but the damage done was only the loss of a chance of a better outcome there is no loss. Negligence leads to compensation for damage that has occurred not loss of a chance of a better outcome. Hotson22 was a 13-year-old boy who fell, fracturing a neck of femur. He was taken to hospital where the diagnosis was not made (breach of duty of care was found). He returned to hospital 5 days later, the diagnosis was made and the fracture treated. Hotson developed avascular necrosis of the femoral head with long-term disability. Hotson was not compensated for the negligence because, on the basis of expert evidence, even if he had been treated on the day of injury there was a 75% chance of avascular necrosis occurring, i.e. on a balance of probabilities (better than 50%) avascular necrosis would have occurred in any event. Interestingly, in Hotson, the Courts of first and second instance awarded damages on the basis of 25% of the full damages, which represented the “lost chance”. The House of Lords rejected this approach and reaffirmed that the test for causation was “on a balance of probabilities” (better than 50%). Where a claimant’s claim for causation fails this test, the claim fails entirely. In a similar fashion if the claim in causation succeeds at the 60% level the final damages are paid in full (not reduced by 40%), i.e. causation is “all or nothing”. This principle was reinforced in Gregg v Scott.23 Mr Gregg had a non-Hodgkins lymphoma. His GP negligently failed to refer him to a specialist. The consequence of the negligence was that Mr Gregg’s chance of survival was reduced from 42% to 25%. The claimant submitted to the House of Lords that Fairchild1 (see below) should apply (even though in Fairchild,1 Wilsher17 had been approved for medical cases). By a majority decision the

Breach of duty by omission Breach of duty of care can be by commission (doing something that should not have been done); or by omission (failing to do something that should have been done). Bolitho8 has been discussed earlier in relation to breach of duty. The Court established that the anaesthetic registrar should have attended on the third occasion, i.e. the breach was the failure to attend. The claimant had to demonstrate that if the anaesthetic registrar had attended, on a balance of probabilities, she would have intubated (which would have prevented the harm, Baby Bolitho’s death from respiratory arrest). The reasonableness of the decision not to intubate was tested, i.e. a doctor cannot avoid negligence if the duty of care was to act (in this case to intubate) but can avoid negligence if it was reasonable not to act. In Bolitho, the claimant failed to establish that intubation would have occurred if the anaesthetist had attended. The claimant could try to establish that the decision not to intubate lacked a logical basis. However if the defendant has expert medical opinion supporting a medical decision (in this case a decision not to intubate) it may be very difficult, if not impossible, to demonstrate that the decision lacked a logical basis.25 The hypothetical decision might be a decision made by someone other than the doctor who omitted to act. In Zarb v Odetoyinbo25 the GP did not refer a patient with sciatica for urgent neurosurgical assessment. By the time Mrs Zarb was referred she had a cauda equina syndrome; there was long-term disability. The claimant needed to establish not only that the GP should have referred, but also that urgent decompressive surgery would have followed. Expert neurosurgical evidence led the Judge to conclude that reasonable neurosurgical practice did not mandate emergency decompressive surgery and the claim failed (on the same principles as set out in Bolitho).

ORTHOPAEDICS AND TRAUMA 25:1

68

Ó 2010 Elsevier Ltd. All rights reserved.

MEDICO-LEGAL

In Joyce v Merton, Sutton and Wandsworth HA26 cardiac catheterization was performed via the right brachial artery. The brachial artery was closed with sutures; the anterior and posterior walls were inadvertently sutured together occluding the brachial artery (something that was not found to be negligent). Breach of duty was found for a failure to warn the patient to return if there were vascular problems. Brachial artery occlusion was diagnosed in outpatients 12 days later. Vascular reconstruction was attempted. A subsequent aortogram dislodged thrombus resulting in brainstem infarction and a locked-in syndrome. Expert evidence was (i) if the arm had been reexplored within 48 h the angiogram would not have been needed and the stroke would have been avoided, but (ii) even if the patient had re-attended within 48 h there would have been no immediate operation. The negligent omission (failure to warn) was not therefore causally linked to the subsequent brainstem stroke.

accident that brought Mr Hogan to the care of the doctors, i.e. it was the original accident that put him in the position where medical negligence could lead to the second injury. It is probable that in most cases a very high degree of negligence (probably gross negligence or recklessness) on the part of treating doctors will be required to break the chain of causation between an original injury and the final outcome. Material contribution There are cases where the damage was probably caused by more than one mechanism. If only one of the mechanisms was a consequence of breach of duty it may be impossible to satisfy the “but-for” test (that the damage was more than 50% likely to have been caused by the mechanism for which breach is established). It may be possible to demonstrate causation using a test that is less strict. The test has been mainly applied to industrial injury, not to medical negligence cases.1,31,32 Where damage is caused by more than one cause and the contribution of each cause is unknown, causation can be established if the cause (for which breach is established) made a material contribution to the damage. Mr Wardlaw32 was exposed to silica dust from two sources: (i) swing grinders (a negligent cause) and (ii) pneumatic hammers (a non-negligent cause). Mr Wardlaw developed pneumoconiosis from the silica dust. Although he could not establish that the silica dust from the swing grinders was more likely to have caused pneumoconiosis than silica dust from the pneumatic hammers causation was established because, on a balance of probabilities, silica dust acquired negligently (the swing grinders) materially contributed to the development of pneumoconiosis. Mr McGhee31 was exposed to brick dust at work (not negligent). There were no showers at work (found to be in breach). Mr McGhee had to cycle home before he could wash off the brick dust which caused the brick dust to be on his skin for a longer period of time. He developed dermatitis which was caused by the brick dust. Medical evidence was unable to decide to what extent the increased exposure to brick dust increased the risk of dermatitis. Causation was established because the risk of developing dermatitis would have been materially reduced if showers had been provided at work. Until recently the Courts had been reluctant to apply the principle of material contribution for medical negligence cases. In Wilsher17 (see above) a negligent cause of blindness (oxygen toxicity) was found by the House of Lords not to be causative because oxygen toxicity was not more than 50% likely to have been the cause of blindness (there were four other potential causes of blindness). However oxygen toxicity could potentially have been a contributory factor to the final blindness. Miss Bailey33 was admitted to the MOD-run Hasler Hospital for removal of common bile duct stones via ERCP. There was postoperative haemorrhage which was negligently managed leaving Miss Bailey in a poor state. She also developed pancreatitis (a non-negligence consequence of ERCP) which contributed to her poor state. Miss Bailey was transferred to intensive care at another hospital. She was transferred to a ward, from ITU, where she drank lemonade, vomited and aspirated, suffering a cardiac arrest and hypoxic brain damage. The Court of Appeal found that Miss Bailey aspirated because of her “weakened state” which had

Breaking the chain of causation (novus actus interveniens) The chain of causation may be broken by a new event (a novus actus interveniens). The initial negligent act or omission may not be causative of the final injury. Where, for example, a psychiatric patient has a recognized risk of suicide, and because of a negligent failure to observe the patient appropriately, the patient commits suicide then the chain of causation is not broken and the claim succeeded.27 Where the treating doctors are not aware of a risk of suicide, suicide may break the chain of causation.28 Mr Hyde was being treated in hospital for physical problems and he became depressed, which was not known to his treating doctors. Mr Hyde jumped out of a window suffering long-term disability because of his injuries. Because his doctors were not aware of his depression (and the risk of attempted suicide) the act of jumping from the window was regarded as a “novus actus” which breached the chain of causation. The hospital treating his initial physical injuries (which were in fact the cause of his depression) were not held liable for the injuries that were consequent upon Mr Hyde jumping from the window. Breaking the chain of causation in cases of clinical negligence is uncommon. It is unusual for the acts or omissions of doctors to break the chain of causation unless there is a very high degree of negligence on the part of the doctors. In Hogan v Bentinck West Hartley Collieries (Owners) Ltd29 Mr Hogan injured his thumb at work. This injury ought to have been capable of treatment with no long-term functional disability but, unfortunately, his medical treatment was of a poor standard and this led to amputation of the thumb. The question was whether the medical negligence broke the chain of causation, i.e. was the original injury the ultimate cause of amputation of the thumb, or was it the medical negligence? What was found was that in that case medical negligence did not break the chain of causation, i.e. the original injury that occurred at work was the cause of amputation of the thumb. In Webb,30 Mr Webb tripped at work injuring his leg. The second D, the doctor negligently amputated the leg above the knee. This negligence was found not to break the chain of causation. Where further injury (in this case above knee amputation) is a foreseeable consequence of the primary injury, the primary wrongdoer is liable for the subsequent injury unless the second defendant (the doctor in this case) is grossly negligent. One way of looking at this is to consider that it was the original

ORTHOPAEDICS AND TRAUMA 25:1

69

Ó 2010 Elsevier Ltd. All rights reserved.

MEDICO-LEGAL

two main causes (i) the negligent resuscitation and (ii) the nonnegligent pancreatitis. Although the claimant could not demonstrate that the negligent cause was more likely than not to have caused the weakened state, it was enough that the negligent cause materially contributed to the weakened state. A material contribution is a contribution to the harm that is more than negligible. The Court of Appeal stated that:

without valid consent a battery has been committed by the surgeon upon the patient. This can lead to legal action in the civil court for damages. If the battery was committed recklessly or intentionally, criminal charges might be brought. Patient autonomy Respect for the autonomy of others is an ethical principle based upon the moral authority of respecting the choices made by others. This principle was applied to consent for surgery as far back as 191438:

“In a case where medical science cannot establish the probability that ‘but for’ an act of negligence the injury would not have happened but can establish that the contribution of the negligent cause was more than negligible, the ‘but for’ test is modified, and the claimant will succeed.” This approach has been applied to psychiatric injury.34 Mrs Dickins had a psychiatric disorder for which there were more than one potential contributing factors, one of which was a negligent failure to address stress at work. The Court of Appeal found that stress at work was a material contribution to the harm; Mrs Dickins succeeded in causally linking stress at work to the final psychiatric disorder.

“.every human being of adult years and sound mind has a right to determine what shall be done with his own body; and the surgeon who performs an operation without his patient’s consent commits an assault, for which he is liable in damages” Informed consent can be considered from the point of view of (i) the patient or (ii) the surgeon. For the patient, consent is the agreement for the proposed surgery, freely given by a competent adult, following an appropriate explanation of the risks, benefits and/or alternatives to the proposed operation. For the surgeon consent is the legal justification for the proposed operation (which in the absence of consent would be a, potentially actionable, battery). For adults there are three bases for consent: (i) legal capacity, (ii) consent must be voluntary and (iii) consent must be informed.

Multiple defendants If several defendants have contributed to damage the claimant can recover damages. The quantum of damages is recovered from each defendant in proportion to the risk of harm for which each defendant was responsible.35 Sometimes a claimant will be unable to establish which defendant was more, or less, likely to have caused the harm. In a non-medical case1 the claimant was employed by two different employers. He was exposed to asbestos dust in both employments. Inhalation of asbestos dust caused mesothelioma, from which he died. Mr Fairchild could not demonstrate “butfor” (greater than 50%) causation for either employer, nor that any breach materially contributed to damage. However Mr Fairchild established causation, probably, largely upon the grounds of natural justice.

Consent is a shield Consent is a shield, used by the surgeon to defend himself. It cannot be used by a patient as a sword, to enlarge the duties owed to a patient or to force a doctor to go beyond what the doctor believes is clinically necessary.39 Volenti non fit injuria Consent, in part, derives from the principle “volenti non fit injuria”, which means that no harm can be carried out to a willing person. As an example, if in a cage-fight one man is seriously injured, the other fighter (who if the injuries occurred in a street fight would face serious charges) is absolved because both fighters entered the ring voluntarily, aware of the potential consequences. Similarly, if a patient is properly aware of a risk of surgery, and voluntarily accepts that risk in order to achieve the potential benefits of surgery, then if the risk eventuates, informed consent may be a complete defence (a complete shield) to any claim.

Informed consent Introduction All surgical procedures require the surgeon to obtain informed consent. This is a specific duty of care. The law is complex and has, relatively recently, been modified. It is crucially important for surgeons to understand the principles and practice of law in this area. Consent for surgical procedures must always be in writing. Consent is, essentially, a meeting of minds. The surgeon cannot assume that the patient has any knowledge or appreciation of the nature of the proposed surgery, its magnitude, its risks and/or benefits and/or the alternatives to surgery. These must be explained. For complex surgery consent might well be a process rather than something that happens on a single occasion. If any force is applied to another without lawful justification a battery has been committed. If any words or actions cause another to believe there is an immediate threat of battery, an assault has been committed. There is an absolute prohibition against touching.36 Exceptions include day-to-day physical contact, for example being jostled at a railway station,37 or statutory powers of arrest. There is no exception made for any form of medical treatment. If a surgical procedure is performed

ORTHOPAEDICS AND TRAUMA 25:1

Form of consent There is no common law requirement that consent must be in any particular form. Consent can be (i) by conduct (e.g. putting out an arm for a blood test or vaccination40) (ii) verbal (e.g. explaining the need for internal examination) or (iii) in writing. Written consent, of course, is the standard practice for surgical procedures. Capacity Adults are presumed to have capacity unless it can be shown that they lack capacity.41 An adult may lack capacity if they do not have one or more of the following: (i) an understanding of the relevant information, (ii) an ability to retain the information for long enough to make an informed decision, (iii) the ability to use

70

Ó 2010 Elsevier Ltd. All rights reserved.

MEDICO-LEGAL

the information to make the decision or (iv) the ability to communicate the decision.42 If a patient has capacity (mental competence) there is an absolute right to accept or refuse medical treatment even if the patient’s decision is felt by others to be bizarre, irrational or nonexistent, and if the decision results in the death of patient, or in pregnancy, the child. A 41-year-old lady was quadriplegic and ventilator-dependent following haemorrhage from an intramedullary spinal cord cavernous heamangioma.43 B requested disconnection of the ventilator. It was found that she was competent to make the decision to stop the ventilation that was keeping her alive. Similar decisions have been made in a case of ventilator-dependent quadriplegia following GuillaineBarre syndrome44 and nasogastric feeding in a patient with cerebral palsy.45 There are a number of “mother-v-baby” cases. In one case46 a mother, who was not a Jehovah’s Witness, refused blood transfusion having been influenced by her mother, who was a Jehovah’s Witness. An emergency Caesarean section was required. The mother was unconscious and required blood transfusion. It was found that T lacked capacity and the transfusion was given. S47 was a 31-year-old lady, 36 weeks pregnant, with severe pre-eclampsia. Caesarean section was recommended and refused by the patient. S was detained under the Mental Health Act 1983 and following the decision of a Court at first instance Caesarean section was performed. S appealed the decision and succeeded: S was competent, the Caesarean section was unlawful and there was a battery. As Lady Butler-Schloss said:

necessary to save the man’s life). There are limits to what may be done in the emergency situation. Surgery is limited to what is required at that moment. If there are options that can wait and reasonably be discussed with the patient (upon reversal of the anaesthetic) the surgeon should not proceed to the further measures at that time.51,52 Lord Goff51 held that: “Where, for example, a surgeon performs an operation without consent on a patient temporarily rendered unconscious in an accident, he should do no more than is reasonably required, in the best interests of the patient, before he recovers consciousness. I can see no practical difficulty arising from this requirement, which derives from the fact that the patient is expected, before long, to regain consciousness and can then be consulted about longer-term measures”. What this means is that, if during an emergency operation, you come to a point where there are options for treatment, for example in orthopaedic surgery a risk/benefit analysis for amputation of a limb; if there was no harm to the patient in allowing the patient to wake up and discuss the options with the patient, then that is what should be done. This is a process that may be difficult for doctors who, by and large, are keen to carry out the definitive treatment at the time of the first operation, not to split up the surgical treatment into two, or perhaps more, operations. However, if there is a real issue in relation to what the patient would probably desire, then if emergency surgery is not required for that part of the operation it should be deferred and the patient’s consent obtained.

“A decision to refuse medical treatment by a patient capable of making the decision does not have to be sensible, rational or well considered. If a doctor overrides the decision of a competent adult and operates they commit a trespass”

Permanent incapacity Where the mental capacity is permanent views should be taken in the “best interests” of the patient. F51 was a 36-year-old woman with a mental age of four. She had been institutionalized since the age of 12. At the time she was in a sexual relationship with another patient. It was accepted that pregnancy would be “disastrous”. It was found that (i) incapacity was permanent, waiting was pointless, (ii) there was an obvious need for care and (iii) the “best interests” of F were to undergo sterilization.51 Tony Bland53 suffered a hypoxic brain injury having been crushed at the Hillsborough football stadium. He was left alive but in persistent coma (a persistent vegetative state). It was found that (i) there was permanent incapacity (ii) with no prospect of improvement (iii) his doctors were under no absolute duty to prolong his life (iv) medical treatment includes tube feeding and antibiotics (v) medical treatment can be withdrawn (vi) the test is the “best interests” of the patient. It was determined that:

Pre-existing mental illness A patient may have a profound mental illness and still retain capacity to consent to medical treatment. In re C48 a man with an exceptionally low IQ and paranoid schizophrenia (in Broadmoor hospital) was found to have the capacity to refuse amputation of a gangrenous toe (from which he died). However, in re C49 a woman with gangrene of both feet who refused amputation, refused to recognize the gangrene and refused to accept the condition might cause her death was found to lack capacity.49 A 25-year-old lady with learning difficulties and behavioural problems refused dialysis for renal failure.50 It was found that she had capacity to make a competent decision.

“The condition of the patient who is totally unconscious and in whose condition there is no prospect of any improvement is such that life prolonging treatment is properly regarded as being, in medical terms, useless.”53

Temporary incapacity Incapacity may be temporary or permanent. The obvious examples of temporary incapacity are the unconscious patient who requires emergency surgery or where there is an unexpected finding, intraoperatively, in an anaesthetized patient. The doctor is entitled to act in accordance with what the doctor feels is in the “best interest” of the patient. Technically the doctor still commits an assault, but the defence is necessity; (for example if a bystander pulled a man out of the path of a runaway lorry and, in so doing, dislocated the man’s arm the bystander has committed an assault, but defends his actions as having been

ORTHOPAEDICS AND TRAUMA 25:1

Tube feeding was withdrawn and Tony Bland died. However in another case54 where a patient was in persistent coma as a consequence of medical negligence it was determined that provision for her care at home should be made, in her best interests, even though it was accepted that she was totally unaware of her surroundings. We do not need to understand the ratio of the decision in this case to appreciate that in patients with

71

Ó 2010 Elsevier Ltd. All rights reserved.

MEDICO-LEGAL

permanent incapacity treatment decisions will be made in the “best interests” of the patient based upon the facts, as determined by the Court. It must, however, be understood that what a doctor, or doctors, believe is the patient’s best interests is not necessarily so. Ultimately the Court is the arbiter of a patient’s best interests. In another case of persistent coma and tube feeding55 Thomas Bingham MR found that:

judgement Lord Scarman felt that a patient had a right to know of a “real” or “material” risk of surgery; and that: “Ideally the Court should ask itself whether in the particular circumstances the risk was such that this particular patient would think it significant if he was told it existed. what would a reasonably prudent patient think significant.” Lord Scarman here substituted a prudent-patient principle for the doctor-centred principle. The Sidaway principle (the reasonable practice of responsible medical men) was upheld in Pearce59 (where it was found to be reasonable that Mrs Pearce was not told that prolonging an overdue pregnancy was associated with a 0.1e0.2% risk of stillbirth; stillbirth eventuated). The Sidaway principle was rejected in Wyatt60 where it was found not to be reasonable to fail to explain a small risk of congenital abnormalities following chickenpox infection during pregnancy. In another jurisdiction Bolam and Sidaway were felt to be inappropriate approaches in relation to consent.61 In the High Court in Australia it was set out that there was a duty to warn of material risks of treatment if the reasonable patient would attach significance to the information or if the doctor should be aware that the patient would attach significance to the information.61 In Chester v Afshar21 (for details of the case please see above) all five law lords agreed that the appropriate test for consent was that of material risk. All material risks should be explained to the patient. These include common risks of surgery, such as superficial infection, even if this is of no longstanding significance, and rare risks that might lead to long-term harm.

“.there should not be a belief that what the doctor says is the patient’s best interest is the patient’s best interest. For my part I would certainly reserve to the Court the ultimate power and duty to review the doctor’s decision in the light of all the facts”55 This does not mean that the Courts wish to, or could, review all end of life decisions. If an elderly patient with a major, disabling, hemisphere stroke was to deteriorate with pneumonia, a consensus decision of doctors and others to limit treatment would, almost always, be reasonable. However, where there was a decision to limit treatment which would lead to the death of a child or young adult, or permanent sterilization, or where there was disagreement between the medical or paramedical staff, or patient, judicial review of treatment-limiting decisions will usually be required. Nature of the information Patients need to understand the nature of the proposed treatment, the risks and benefits and any alternatives. This includes (i) an explanation of the proposed treatment, for example in a lumbar microdiscectomy that the operation is performed under general anaesthetic, the operation is from a posterior approach to remove the disc prolapse, etc and (ii) the risks and benefits of the treatment. Fraud always negates consent. Garrett56 was a dentist who performed unnecessary dental procedures for profit. Deliberate misinformation meant that there was no voluntary consent for the procedures. Garrett committed a battery. The Court distinguishes intentional misinformation (as above) from unintentional negligent misinformation. Reibl57 was offered a left carotid endarterectomy. The 10% risk of stroke and the 4% risk of death were not discussed with him. Reibl had the procedure and suffered a permanent stroke. He claimed he would not have had the procedure if he had known of the risk of stroke. It was found that Reibl should have been told of the risks but there was no deliberate intent to mislead, there was a negligent mistake. Reibl succeeded in negligence.

Alternatives to treatment Alternatives to the proposed treatment should be discussed. Mrs Birch62 underwent cerebral catheter angiography. She consented to a 1% risk of stroke. Stroke eventuated. Mrs Birch was not told that MR angiography (MRA) was an option that carried no risk of stroke (although diagnosis was less certain with MRA). Because Mrs Birch did not know that MRA was an option: “She was denied an informed choice. the failure to discuss these matters could not be described in law as reasonable, responsible or logical” Which doctor and when? GMC guidelines are that consent should be taken by (i) the surgeon who is to undertake the operation or (ii) someone who is capable of carrying out the operation. The point is that only a surgeon who is capable of carrying out the procedure can properly understand the risks and benefits of surgery and therefore explain all material matters to the patient. If consent is taken by a junior doctor in training who is not capable of carrying out the operation the presumption may be that informed consent was not obtained. When informed consent should be obtained is less clear. For complex procedures, including some surgical procedures, it may be that consent is a process rather than something that occurs on a single occasion. Material consequences of surgery explained after the operation (irreversibility of sterilization) do not constitute informed consent.63 Mr Moses64 underwent a lumbar microdiscectomy, postoperatively he had a cauda equina syndrome (CES). Mr Moses

Level of information What level of information should be given to patients? Initially the courts took a doctor-centred approach. In Bolam4 it was found to be reasonable not to give Mr Bolam the option of an anaesthetic for ECT (which would have avoided the consequent fracture) because some doctors were of the view that anaesthesia was not needed. Mrs Sidaway58 had a cervical myelopathy requiring decompressive cervical laminectomies. The risk of quadriplegia was found to be 1e2%. Mrs Sidaway was not told of this risk; postoperatively she was quadriplegic. The majority judgement was doctor-centred on the basis that the majority of neurosurgeons in the 1980s did not warn of this risk and therefore it was reasonable for Mrs Sidaway not to be informed. In a partially dissenting

ORTHOPAEDICS AND TRAUMA 25:1

72

Ó 2010 Elsevier Ltd. All rights reserved.

MEDICO-LEGAL

B74 was a 10-day-old girl with Down’s syndrome who suffered an acute bowel obstruction requiring surgery. The parents refused consent. Treatment was approved on the basis that Down’s syndrome does not condemn the patient to a life of appalling quality and therefore it was in B’s best interests to be treated. Courts have approved, against the wishes of the parents, blood transfusion75 and bone marrow transplantation.76 Sterilization to prevent pregnancy has both been approved77 and not approved.78

claimed he had never been consented to a risk of CES. The surgeon claimed that CES had been explained on the day of surgery. The Judge found that even if CES had been discussed on the day of surgery this would not represent informed consent because the patient would not have had sufficient time to digest and reflect upon this new and material risk. All material risks of surgery must be explained in advance of the procedure. How far ahead is not known. It may be that explaining new material risks following admission to hospital, on the evening before major, planned, surgery would not represent informed consent for the same reasons as set out in Moses.64 Current reasonable practice is, probably, to see all patents who require planned surgery in a preadmission clinic 1e2 weeks preoperatively; to go over the material risks of surgery and obtain written consent (copied to the patient). This gives the patient time to consider/reconsider the options and cancel the operation if, upon reflection, the patient decides to defer, or postpone indefinitely, the operation.

Euthanasia Euthanasia was, and remains, unlawful even if the doctor’s intention is to prevent suffering. There has, however, only been one case where a doctor was convicted of attempted murder in an attempt to relieve suffering.79 Dr Cox, a rheumatologist, had a patient, Mrs Boyes, who had multiple medical problems and was in severe uncontrolled pain. Mrs Boyes asked Dr Cox (and others) to end her life, which to her was intolerable. Dr Cox injected Mrs Boyes with potassium chloride with the intended purpose of causing death. He was convicted of attempted murder.

Consent in children Children achieve majority at the age of 18. Children are presumed to achieve capacity to consent to medical treatment at the age of 16.65 Below the age of 16 children can demonstrate that they understand the nature of the treatment in sufficient detail to give consent (Gillick competence).66 Until a child becomes Gillick competent the authority to consent lies with the parents (or those with parental responsibility). Gillick65 was a mother of five girls, all under the age of sixteen. Mrs Gillick sought an injunction that it would be unlawful to prescribe contraceptive medication to girls aged less than 16 without parental consent. Gillick failed provided the child under the age of 16 had the understanding and intelligence to make an informed decision. P67 was a 15-year-old girl who was living is a hostel bringing up her first child. She became pregnant again and requested an abortion. Abortion was opposed by her parents, on religious grounds. It was found that P had the understanding to decide whether or not to have an abortion, i.e. she had the capacity to consent to abortion. Where a child under 16 refuses treatment they may be assumed not to be Gillick competent. W68 was a 16-year-old girl suffering from anorexia nervosa. She required nasogastric (NG) feeding (a medical treatment). There was an initial application to permit treatment. W at that time consented to nasogastric feeding and was found to be Gillick competent to consent to NG feeding. She subsequently refused NG feeding. The Court of Appeal permitted NG feeding against her will on the grounds that (i) she was probably not Gillick competent to refuse the treatment and (ii) NG feeding was in her best interests. A girl69 and a boy70 both aged 15 were not Gillick competent to refuse blood transfusion required to treat thalassaemia and leukaemia respectively. A 15-year-old girl was not Gillick competent to refuse cardiac transplantation.71

Summary All doctors risk a claim, against them, in medical negligence. All doctors need to understand the principles by which, legally, their practice will be judged. I hope this article assists you in understanding what the legal tests are and how they can ensure that your decisions fall within what is currently understood as the appropriate standard of care for doctors. A

REFERENCES 1 Fairchild v Glenhaven Funeral Services Ltd [2002] UKHL 22. 2 Donoghue v. Stevenson [1932] All ER Rep 1; [1932] AC 562; House of Lords. 3 Caparo Industries plc v Dickman 2 AC 605. 4 Bolam v Friern Hospital Management Committee [1957] 1 WLR 583. 5 Whitehouse v Jordan [1981] 1 WLR 246 HL at 256. 6 Maynard v. West Midlands Regional Health Authority [1985] 1 All ER 635. 7 Sidaway v. Bethlem Royal Hospital Governors [1985] AC 871. 8 Bolitho v. City and Hackney Health Authority 4 All ER 771. 9 Hunter v Hanley 1955 SC 200. 10 Thompson and others v Smiths Shiprepairers (North Shields) Ltd [1984] QB 405; [1984] 1 All ER 881, QBD. 11 Shakoor v Situ [2000] 4 All ER 181. 12 Roe v Minister of Health [1954] 2 All ER 131. 13 Hucks v Cole [1993] 4 Med LR 393. 14 Knight v Home Office [1993] All ER 237. 15 Brooks v Home Office [1999] 2FLR 233 QBD. 16 Bull v Devon AHA [1993] 4 MED LR 117 (CA). 17 Wilsher v Essex Area Health Authority 1988 AC 1074 (HL). 18 R v. Adomako [1994] 3 All ER 79. 19 Loveday v Renton [1990] 1 Med ER 117, CA 133. 20 Barnett v. Chelsea and Kensington Hospital Management Committee, [1969] 1 QB 428.

Differing opinions Where treatment of an incompetent child is being considered there may be differences of opinion as to what should, or should not, be done. These cases are particularly pointed where one or other party feels that the child should be allowed to die. In such cases review by the Court is required.72 Such review can be made urgently. The Court will determine the best interest of the child.73

ORTHOPAEDICS AND TRAUMA 25:1

73

Ó 2010 Elsevier Ltd. All rights reserved.

MEDICO-LEGAL

21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50

51 52 53 54

Chester v. Afshar UKHL 41. Hotson v East Berkshire Area Health Authority [1987] AC 750 3. Gregg v Scott [2005] UKHL 2. Meadow v GMC [2006] EWHC 146 (Admin). Zarb v Odetoyinbo [2006] EWHC 2880 141. Joyce v Merton, Sutton and Wandsworth HA [1996] 7 Med LR1, CA 140-1. Kirkham v. Chief Constable of the Greater Manchester Police [1990] 2 QB. 283 [2001]. Hyde v Tameside AHA, The Times, 15 April 1981. Hogan v Bentinck West Hartley Collieries (Owners) Ltd [1949] 1 All ER 588. Webb v Barclays Bank Plc and Portsmouth Hospitals NHS Trust. McGhee v National Coal Board [1972] 3 All ER 1008, HL 144. Bonnington Castings v Wardlaw [1956] 1 All ER 615, HL 143. Bailey v Ministry of Defence [2008] EWCA Civ 883. Dickins v O2 plc [2008] EWCA Civ 1144. Barker v Corus [2006] UKHL 20. Mallette v Schulman [1990] 67 D.L.R. (4th). Wilson v Pringle [1987] QB 237. Schloendorf v Society of New York Hospital 103 NE 92 (1914). Burke v GMC [2005] EWCA Civ 1003, CA 162-3. O’Brien v Cunard SS Company [1891] 28 NE 266. S1 (2) Mental Capacity Act 2005. SS 1,2,3,4 Mental Capacity Act 2005. Re B (adult: refusal of medical treatment) [2002] EWHC 429 (fam); [2002] 2 All ER 449, CA. Nancy B v Hotel de Dien de Quebec [1992] 86 DCR (4th) 385. Bouvia v Superior Court [1986] 225 Cal Rptr 297 (Cal CA). Re T (adult: refusal of medical treatment) [1992] 4 All ER 649, CA. Fam 95. St George’s Healthcare NHS Trust v S [1998] 3 All ER 670, CA. Re C (Adult: Refusal of Treatment) [1994] 1 WLR 290. State of Tennessee v Northern [1978] 563 SW 2d 197 [Tenn Ct App]. Re JT (adult: refusal of medical treatment) [1998] 1 FLR 48.

ORTHOPAEDICS AND TRAUMA 25:1

55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79

74

Re F. West Berkshire Health Authority [1990] 2 A.C. 1. Mental Capacity Act 2005 Section 4(3). Airedale NHS Trust v Bland [1993] 1 ALL ER 821. Ahsan v University Hospitals Leicester NHS Trust [2006] EWHC 2624 (QB). Frenchay Healthcare NHS Trust v S [1994] 2 All ER 403, 1 WLR 601. Appleton v Garrett [1997] 8 Med LR 75. Reibl v Hughes [1980] 114 DLR [3d] 1, Can Sup Ct. Sidaway v. Bethlem Royal Hospital Governors AC 871. Pearce v United Bristol Healthcare Trust (1998) 48 BMLR 118. Wyatt v Curtis & Central Nottinghamshire Health Authority [2003] EWCA Civ 1779. Rogers v. Whitaker - [1992] 109 ALR 625. Birch v UCL Hospitals Trust [2008] EWHC 2237 (QB). Lybert v Warrington HA [1996] 7 Med LR 91. Moses v Vafadis (unreported). s8 Family Law Reform Act 1969. Gillick v West Norfolk and Wisbeck Area Health Authority [1985] 3 All ER 402 HL. Re P [1986] 1 FLR 272 Fam. Re W [a minor] [medical treatment] 4 All ER 627 CA. Re S [a minor] [consent to medical treatment] [1994] 2 FLR 1065. Re E [a minor] [1993] 1 FLR 386. Re M [1999] 2 FLR 1097. R v Portsmouth Hospital NHS Trust ex parte Glass [1999] EWCA Civ 1914. Re MB [1997] 8 Med LR 217. Re B [a minor] [wardship: medical treatment] [1981] [1990] 3 All ER 927 CA. Re S [a minor] [medical treatment] [1993] 1 FLR 376 Fam. NHS Trust v A (a child) [2007] EWHC 1696 Fam. Re B [a minor] [wardship: sterilisation] [1987] All ER 206, HL. Re D [a minor] [wardship: sterilisation] [1976] 1 All ER 326 QBD. R v Cox (1992) 12 BMLR 38.

Ó 2010 Elsevier Ltd. All rights reserved.

CME SECTION

CME questions based on the Mini-Symposium on “The Shoulder” D Posterior acromioclavicular ligament E Superior acromioclavicular ligament

The following series of questions are based on the MiniSymposium on ‘‘The Shoulder”. Please read the articles in the Mini-Symposium carefully and then complete the selfassessment questionnaire by filling in the square corresponding to your response to each multiple-choice question. After completing the questionnaire, either post or fax the answer page to the Orthopaedics and Trauma Editorial Office at the address at the bottom of the RESPONSE sheet. Please photocopy this page if you wish to keep your copy of Orthopaedics and Trauma. Replies received before the next issue of the journal is published will be marked and those reaching an adequate standard will qualify for three external CME points. You will be notified of your marks and a CME certificate will be despatched, via email, for your records.

5 In a cadaveric study in which hook plates were applied to 15 shoulders, in how many did the hook penetrate the subacromial bursa? A None B Three C Four D Eight E Thirteen

6 Which of the following conditions is least likely to be responsible for unilateral osteolysis of the distal clavicle? A Gorham’s massive osteolysis B Gout C Infection D Scleroderma E Tumour

Questions 1 From which cervical root level(s) do the fibres of the axillary nerve supplying Teres minor arise? A C5 B C5 and C6 C C6 D C6 and C7 E C7

7 What does loss of the superior convexity of the rotator cuff on ultrasound scan indicate? A Bursal side partial thickness tear B Calcific tendinitis C Full thickness cuff tear D Joint side partial thickness tear E Normal variation

2 Which of the following passes through the triangular space, medial to the long head of triceps? A Axillary nerve B Circumflex scapular artery C Musculocutaneous nerve D Posterior circumflex humeral artery E Posterior cutaneous nerve of the arm

8 In which of the following lesions is an intact anteroinferior labrum held in situ by an intact scapular periosteum with an associated injury to the glenoid articular cartilage? A ALPSA lesion B Bankart lesion C GLAD lesion D PASTA lesion E Perthes lesion

3 A patient presents with a full thickness rotator cuff tear. Statistically, what is the approximate chance that any sibling they have also has a cuff tear? A 5% B 15% C 30% D 45% E 60%

9 Which of the following muscles is most likely to appear oedematous on an MRI scan of a patient with Parsonage Turner syndrome? A Infraspinatus B Latissimus dorsi C Serratus anterior D Subscapularis E Trapezius

4 Which ligament principally provides stability in the horizontal plane to the Acromioclavicular joint? A Anterior acromioclavicular ligament B Coracoclavicular ligament C Inferior acromioclavicular ligament

ORTHOPAEDICS AND TRAUMA 25:1

75

Ó 2011 Published by Elsevier Ltd.

CME SECTION

Responses

10 Which of the following blood tests can return elevated results in the early stages of frozen shoulder? A Alkaline phosphatase B CRP C ESR D Serum globulins E Serum phosphate

Please shade in the square for the correct answer. B C D E 1A 2A B C D E 3A B C D E 4A B C D E 5A B C D E 6A B C D E 7A B C D E 8A B C D E 9A B C D E 10 A B C D E 11 A B C D E 12 A B C D E

11 Approximately what proportion of insulin dependent diabetics will develop a frozen shoulder? A One in a hundred B One in fifty C One in ten D One in three E More than half

Your details (Print clearly) 12 Which metallic element is found in the matrix metalloproteinases? A Calcium B Cobalt C Iron D Magnesium E Zinc

NAME ............................................................................ ADDRESS ............................................. ........ ............................................. EMAIL ................................................. ........ RETURN THE COMPLETED RESPONSE FORM by fax to þ44-113-392-3290, or by post to CME, Orthopaedics and Trauma, Academic Department of Orthopaedic Surgery, “A” Floor Clarendon Wing, Leeds General Infirmary, Great George Street, Leeds LS1 3EX, UK.

Please fill in your answers to the CME questionnaire above in the response section provided to the right. The return address and fax number are given below the response section.

ORTHOPAEDICS AND TRAUMA 25:1

76

Ó 2011 Published by Elsevier Ltd.

CME SECTION

Answers to CME questions based on the Mini-Symposium on “Specific Bone Tumours” Please find below the answers to the Orthopaedics and Trauma CME questions from Vol. 24, issue 5 which were based on the Mini-Symposium on “Specific Bone Tumours”

Answers 1a

b

c

d

e

2a

b

c

d

e

3a

b

c

d

e

4a

b

c

d

e

5a

b

c

d

e

6a

b

c

d

e

7a

b

c

d

e

8a

b

c

d

e

9a

b

c

d

e

10 a

b

c

d

e

11 a

b

c

d

e

12 a

b

c

d

e

ORTHOPAEDICS AND TRAUMA 25:1

77

Ó 2010 Published by Elsevier Ltd.

BOOK REVIEWS

for the surgeon/statistician to ponder and ascertain the validity of the recommendations. This book is a valuable add on to the library of every practice or hospital department treating carpal tunnel syndrome. The relevant literature from April 2007 to-date is not part of this analysis and should be sought separately. A

AAOS (American Academy of Orthopaedic Surgeons) Clinical Practice Guideline. Treatment of carpal tunnel syndrome AAOS 2009, 167 pages, $100, ISBN 10: 0-89203-637-0, ISBN 13: 978-089203-637-0, Published 2009

Throughout the developed world, the cost of healthcare has escalated beyond the control of governments who now wish to ration treatment which is regarded as non-essential or not affordable. Carpal tunnel release is one such treatment which has been under scrutiny and health funding bodies look for evidence that treatment other than surgery should be used. Evidence will be produced to support the case. It must, however, be the best evidence available and not evidence that just suits the purpose. Surgeons in all parts of the globe will be faced with the threat to withhold surgery for carpal tunnel syndrome in favour of cheaper non-operative methods. It is thus essential that clinicians have, readily available, references to the literature so that they can join in the debate about what is best treatment for their patients, whether that be surgery or not. The Clinical Practice Guideline on the treatment of carpal tunnel syndrome published by the American Academy of Orthopaedic Surgeons is a valuable source of information for the clinician and also for those who administer the funding of healthcare. This is a thorough and extensive review of the literature as defined by Cochrane Collaboration. The work group panel consisted of a group of seven American/Canadian hand surgeons and a physician of Physical medicine and Rehabilitation. The final draft was reviewed by an outside advisory panel and internally by the Council on Research Quality Assessment and Technology. The content of the book has been approved by the AAOS Evidence Based Practice Committee and the Board of Directors of the AAOS. The Work Group Panel reviewed MEDLINE, EMBASE and Cochrane Database up to 6 April 2007. They also included CINAHL from 1982 to 12 June 2007. The recommendations were graded based on the level and consistency of the available evidence from Grade A: Good evidence (Level I with consistent findings), Grade B: Fair evidence (Level IIeIII Studies with consistent findings), Grade C: Poor quality evidence (Level IV or V) and Grade I: Insufficient or conflicting evidence not allowing a recommendation. The language used for each treatment was based on this grading system with level A was “recommended”, B “suggested” and C defined as an “option”. 332 articles were initially reviewed with 94 passing the strict inclusion criteria. Based on this, nine recommendations were made to cover all aspects of treating the carpal tunnel syndrome including operative, other procedures during operation, many types of non-operative treatments, immobilization after surgery and outcome assessment instruments. The only successful treatment with Grade A Level I evidence is carpal tunnel release. Steroid injections are described as providing clinical improvement for up to 6 months but the evidence to support this is Grade B Level II and III. The book summarizes these recommendations briefly at the beginning, obviating the need to go through the detail if one is only interested in the outcome. However the excellent methodology, discussion and conclusions are detailed in chapters 1e3 where 354 references appear. Chapter 4 is the Appendix and is the largest section covering pages 29e167. It provides all the detail of the thorough methodology used. This is

ORTHOPAEDICS AND TRAUMA 25:1

R Amirfeyz

FRCS

& IJ Leslie

FRCS

Hand Surgery Unit, Department of Trauma & Orthopaedics, Bristol Royal Infirmary, Bristol, United Kingdom.

Complications in orthopaedics: open fractures L Scott Levin, ed, American Academy of Orthopaedic Surgeons; 2010; ISBN: 9780892036394, Pages: 100; Price: $45

Distinct monograph structured in the form of case examples. A variety of representative cases of open fracture complications and their management options discussed in 90 pages including more than 35 illustrative photos and X-rays, as well as three comprehensive algorithms. The contributing authors, as well as the editor of this monograph of the American Academy of Orthopaedic Surgeons, are all well known clinicians in the field of open fracture and soft tissue management, offering a small distil of their large experience. The reader of this monograph, is anticipated to be familiar with the principles of open fracture management, and should expect from this AAOS supplement an all inclusive presentation of associated complications and their successful management strategy. Its reading offers a compact set of examples of difficult cases that require expertise and careful planning for their treatment. However, besides this accumulation of well-illustrated examples what appears as particularly attractive is the discussion that follows each of the seven chapters, where the strategy and key points of the surgeon’s decision-making process are generously presented. Principles of managing bone loss, associated vascular injuries, limb salvage decision making, soft tissue coverage strategies, as well as management of associated complications of open fractures like infection, neuropathic pain, malunions and nonunions complete the panel of separate chapters of this condensed monograph. The inexperienced or semi-experienced in these types of clinical problems, surgeon or trainee, can use this monograph as a reference of methods of treating such difficult cases, and as a guide for further reading on the wide subject of open fractures. The specialized surgeons on trauma reconstruction, dealing with similar problems at their daily practice, are mostly to benefit from this collection of challenging cases and from the rare accumulation of experts’ opinion on the subject of open fractures and their spectrum of associated pathology and complications.A

Nik K Kanakaris

MD PhD

Consultant Orthopaedic Surgeon, Academic Unit of Orthopaedic Surgery, Leeds General Infirmary, Leeds LS2 9NS, UK.

78

Ó 2010 Published by Elsevier Ltd.