Obstruction of the subclavian artery or vein and pressure on the lower trunk of the brachial plexus may occur in the thoracic outlet due to a number of anatomical abnormalities, the best recognized being a cervical rib. The first successful removal of a cervical rib was undertaken by Coote in 1861. The clinical syndrome of thoracic outlet obstruction is a complex of clinical features, which may be either predominantly vascular or predominantly neurological. The syndrome has been given a number of descriptions over the years, each suggesting an aetiology, such as costoclavicular syndrome, scalenus anticus syndrome, and hyperabduction syndrome. However, the varied symptoms and signs in these patients are best grouped under the broad term thoracic outlet obstruction syndrome.

The subclavian vein and artery cross the anterior half of the first rib, separated by the insertion of scalenus anterior into the scalene tubercle of the first rib (Fig. 1) 312. The brachial plexus lies behind the vessels; the lower trunk of the plexus formed by the roots of C8 and T1 is the most inferior structure of the plexus and lies on the first rib. As these structures cross the first rib they lie behind and below the clavicle and the subclavius muscle attaching to the inferior aspect of the clavicle. After crossing the first rib the neurovascular bundle passes into the axilla below the coracoid process of the scapula with its attached pectoralis minor.

The most common cause of obstruction or pressure at the thoracic outlet is a cervical rib which passes beneath the brachial plexus and subclavian artery to attach to the scalene tubercle on the first rib. Not infrequently a cervical rib is incomplete but is replaced by a fibrous band which extends from the tip of the rib again to the scalene tubercle. The lower trunk of the brachial plexus may be stretched over a cervical rib or a fibrous band, as may the subclavian artery. This may also be squeezed in the angle between the scalenus anterior insertion and the cervical rib in certain positions (Fig. 2) 313. Pressure on the brachial plexus and its lower trunk is sometimes due to a fibrous band lying along the medial border of scalenus medius as it passes to its insertion in the first rib. The insertion of scalenus anterior, which is quite fibrous, may sometimes be abnormal, with an extension running back along the first rib beneath the subclavian artery and obstructing it when the arm is in certain positions. The space between the clavicle and the first rib may be diminished in those with an old malaligned fracture of the clavicle or a hypertrophied subclavius muscle. This space also decreases as a result of poor posture: it is not uncommon for women to present with symptoms of thoracic outlet obstruction, either with or without a cervical rib, as they approach middle age.

The axillary vein may be obstructed as it crosses the first rib in front of scalenus anterior to become the subclavian vein if the space between the clavicle and first rib is decreased: this is usually associated with a large subclavius muscle or a fibrous anterior extension of the insertion of scalenus anterior.

Patients present with any of a variety of symptoms and signs which may be predominantly vascular, predominantly neurological, or a combination of both. The majority of patients are women, usually between the ages of 20 and 40. Those with evidence of arterial obstruction may present with a history of Raynaud's phenomenon, nearly always unilateral, or of inability to use the arm on the involved side above the head for any period of time because of weakness in the arm and hand or pain and numbness in the fingers. On examination the subclavian artery may be prominent in the root of the neck when elevated by a cervical rib, which may indeed itself be palpable, and obliteration of the pulse can be readily achieved by a variety of manoeuvres, such as hyperabduction of the arm and bracing of the shoulders, which will be associated with pallor of the same hand. A bruit may be heard at rest or, more commonly, as the artery is compressed with one of the above manoeuvres. Occasionally older patients present with features of a subclavian aneurysm (which is really a post-stenotic dilation of the artery beyond an obstruction), giving rise to distal emboli from thrombus within the lumen of the aneurysm. Emboli from a sub-clavian aneurysm can produce distal signs ranging from splinter haemorrhages to gangrene of the finger tips (Fig. 3) 314.

Patients with predominantly neurological symptoms often represent a difficult diagnostic group. Classically, pressure on the lower trunk of the brachial plexus results in paraesthesia of the ulnar aspect of the hand and forearm associated with certain posture, such as carrying a heavy basket or attempting to use the hand above the head. By the time of presentation there may be obvious weakness and wasting of the small muscles of the hand. There may be a tender point in the root of the neck over the plexus as it crosses a cervical rib. The patient often complains of pain with no appropriate segmental distribution in the shoulder, upper arm, or neck. Finally many patients present with a combination of both vascular and neurological symptoms.

Venous obstruction at the thoracic outlet is an uncommon presentation, but this diagnosis should always be suspected in young people who present with an axillary vein thrombosis or intermittent swelling of the arm. They may participate in sports that can lead to hypertrophy of the subclavius muscle, such as surf board riding or butterfly swimming.

A radiograph of the neck and upper chest will demonstrate cervical ribs (Fig. 4) 315, but the absence of a rib does not exclude the diagnosis of thoracic outlet obstruction due to a fibrous remnant of an incomplete rib. A cervical rib, either complete or incomplete, will be visible on the radiograph, as will a prominent transverse process of the seventh cervical vertebra in the presence of a fibrous band. The presence of a cervical rib does not necessarily mean that this is the cause of the presenting symptoms. Cervical ribs are not uncommon and in most people do not give rise to problems.

Magnetic resonance imaging (MRI) may be a useful addition to our diagnostic procedures, as it has the capacity to identify the brachial plexus and angulation of the lower trunk of the plexus over a fibrous band (Fig. 5) 316. Our experience of MRI in the investigation of thoracic outlet obstruction in Oxford suggests that it is useful.

An arteriogram will demonstrate obstruction of the artery at the outlet with the arm abducted for example (Fig. 6) 317, but this can be established clinically and does not need to be confirmed by angiography. The main indication for an arteriogram is to exclude the presence of thrombus within a possible subclavian aneurysm. However, the presence of both an aneurysm and intraluminal thrombus can be established by ultrasound, which further reduces the need for angiography.

A venogram will always be performed in a young patient presenting with a venous thrombosis and in the patient with intermittent swelling of the arm. A characteristic picture may be seen in the presence of obstruction, and sometimes a clinical diagnosis of axillary vein thrombosis will not be confirmed on venography, the clinical features being due to obstruction of the vein only at the thoracic outlet (Fig. 7) 318.

Neurophysiological studies are essential in establishing a neurological component to thoracic outlet obstruction and can provide a high degree of diagnostic accuracy in establishing the site of the problem at the brachial plexus level rather than at the level of the cervical spine or more distally.

In established ‘classical’ thoracic outlet syndrome there is a characteristic pattern of electrophysiological abnormality. The compound muscle action potential recorded from the thenar eminence is low, but median motor conduction velocity in the forearm and distal motor latency are normal. Preserved sensory nerve action potentials from the thumb, and index and third fingers, in the distribution of the C6 and C7 roots of the plexus, with normal conduction velocities across the wrist, serve to differentiate the condition from carpal tunnel syndrome. In contrast, the sensory nerve action potential from the fifth finger, in the C8 dermatome, is relatively reduced in amplitude or is absent. The hypothenar compound muscle action potential may also be reduced in amplitude, but less so than the response from abductor pollicis brevis. Normal motor and sensory conduction between the mid-arm and wrist exclude a peripheral ulnar nerve lesion.

Electromyography of the intrinsic hand muscles usually shows signs of chronic partial denervation consisting of scanty fibrillation potentials, prolonged polyphasic unstable motor unit action potentials which may have high amplitudes, and reduced density of the interference pattern. These findings are more prominent in the abductor pollicis brevis than in the first dorsal interosseous muscle, and in a few patients similar changes are seen in the forearm flexor compartment.

Other neurophysiological procedures which may be of help in diagnosis include measurement of F response (a late spinal response) latency, reflex latency measurement, recording of sensory nerve action potentials from the medial cutaneous nerve of the forearm, and cortical somatosensory responses following dermatomal stimulation. However, these tests are unlikely to be of value if the more characteristic changes are not observed. Measurement of proximal ulnar motor conduction velocity across the brachial plexus is now generally disregarded as a method of diagnosing the thoracic outlet syndrome. Typical findings in a patient with thoracic outlet compression due to a cervical rib are shown in Table 1 189.

The diagnosis of a vascular thoracic outlet obstruction is usually relatively straightforward, but may present problems in patients with bilateral Raynaud's phenomenon. However, symptoms due to thoracic outlet obstruction usually have an earlier onset in one hand. If the presenting feature is unilateral Raynaud's phenomenon, it is important to exclude an occupational cause such as the use of hand-held drills. It should also be remembered that arterial obstruction may occur proximal to the outlet, at the origin of the left subclavian or innominate arteries, and where this possibility exists an arch aortogram will be required.

The two major neurological diagnoses to be excluded are cervical spondylosis and carpal tunnel syndrome. MRI is the investigation of choice in the former, and neurophysiological studies the most useful investigation in the latter, but neither replaces a careful history and examination.

Exercises designed to strengthen the muscles of the neck and shoulder girdle and the adoption of a more upright posture can relieve symptoms in many patients, especially women approaching middle age who often begin to droop across the shoulders, resulting in narrowing of the thoracic outlet. This should be the first approach in most patients unless they have florid symptoms of obstruction, definite neurological signs, or arterial changes, in which case it is better to proceed to surgery.

The surgical approach to this condition depends on the cause. Cervical ribs, either complete or fibrous, should be excised. Additional bands in the scalenus medius are divided. If no obvious cause of compression is found, division of the scalenus anterior at its insertion (scalenotomy) has been widely practised in the past. However, definite evidence of vascular or neurological thoracic outlet obstruction in the absence of a cervical rib or fibrous band is now treated by excision of the first rib. If there is evidence of arterial damage such as a significant post-stenotic dilatation or an aneurysm of the subclavian artery, that section of the artery is excised and replaced with a short length of vein or a prosthetic graft such as polytetrafluoroethylene.

The techniques by which the above surgical procedures are performed are several and each has its protagonists. To some extent the technique should be dictated by the cause of the thoracic outlet obstruction.

With the patient in the prone position and a small sandbag behind the shoulders, a transverse incision is made above and parallel to the clavicle at the base of the anterior triangle. This is carried down through the platysma and the supraclavicular pad of fat to expose the scalenus anterior. The phrenic nerve crosses the anterior surface of the muscle from lateral to medial and should be carefully separated from the muscle and protected with a sling. Attention should be paid to the possible presence of an accessory phrenic nerve. The scalenus anterior muscle is then divided near its insertion, to expose the subclavian artery.

If a cervical rib or a fibrous remnant is present or any other fibrous band it will be obvious on palpation, and can be exposed by further dissection. The major advantage of the supraclavicular approach is that it allows the anatomy of the thoracic outlet to be fully visualized, together with any abnormality. A cervical rib is excised, being careful to take the rib back behind the plexus. This last part of the excision is usually best performed from the lateral aspect of the brachial plexus. A fibrous remnant of a cervical rib is divided and, depending on its length, an incomplete cervical rib is removed to ensure that the lower trunk of the brachial plexus is free of any potential pressure. Any other bands that are apparent, such as along the medial border of scalenus medius are also divided, especially in the absence of a cervical rib or fibrous remnant.

The supraclavicular approach is ideal for removal of a cervical rib or a fibrous remnant, and also allows inspection of the artery. If a segment of the artery has to be replaced the axillary artery can be exposed below the clavicle for distal control and placing of the distal anastomosis.

The subclavian vein is not seen through the supraclavicular approach and so it does not allow a possible venous obstruction to be defined; this is best achieved by the infraclavicular approach.

The first rib can be removed from above, but this procedure is less easy as exposure of the rib is often less than satisfactory. Removal of the first rib is best accomplished through the transaxillary approach or the infraclavicular approach.

This is a popular method of resection of the first rib to relieve thoracic outlet compression. Some protagonists of this technique would argue that removal of the first rib and hence detachment of everything that attaches to it, such as the scalenus anterior, cervical rib, and fibrous bands, ensures that the thoracic outlet obstruction is relieved whatever the cause.

With the patient in a full lateral position, the upper arm is abducted to a right angle and forcibly retracted by an assistant. A transverse incision is then made on the medial aspect of the axilla, between the borders of pectoralis major anteriorly and the latissimus dorsi posteriorly, and deepened to expose the first rib. Care must be taken not to damage the nerve to the serratus anterior. The first rib is then excised after separating it from the attaching structures. The secret of good exposure is vigorous retraction of the arm, which usually requires assistants relieving one another at frequent intervals! This approach has the disadvantage that adequate access to the artery and vein is not possible.

The positioning of the patient is all important. The patient lies with a sandbag between the shoulder blades and a transverse incision is made beneath the clavicle from the coracoid process laterally towards the midline for some 5 cm. This incision is carried down to the pectoralis major, which is then divided with diathermy to expose the acromioclavicular fascia, incision of which allows the axillary vein and artery to be identified with the plexus behind the artery. The shoulder is then lifted forward, opening up the costoclavicular space and allowing the vein and artery to be retracted from the first rib, the anterior half of which is readily visualized. The rib is then excised as far forward as required and then back beyond the plexus. This approach is ideal for patients with venous or arterial obstruction, but has one disadvantage in that as excision of the rib proceeds posteriorly to take it back beyond the plexus, the view of the rib is sometimes poor.

This is practised infrequently, for although it provides a good view of the neck and posterior part of a cervical rib or a first rib, the artery is not accessible, and it is a much more destructive approach than the others described. It is sometimes useful if a cervical sympathectomy needs to be performed in a patient who has had already undergone several neck explorations.

When an operation is performed for arterial obstruction, with a clearly established diagnosis the results are excellent (Table 2) 190. If neurological compression at the thoracic outlet has been diagnosed without doubt, surgical relief of the obstruction will also produce a good outcome. If there is muscle wasting of the small muscles of the hand recovery will be slow; if this is long-standing it is unlikely to improve, although sensory symptoms and pain are relieved immediately. However, the diagnosis of neurological thoracic outlet compression is often far from clear, and in some instances surgery is performed almost as a diagnostic procedure, when both the patient and neurologist are desperate. Although some vascular surgeons feel that to be certain of relieving symptoms the first rib should always be removed even if a cervical rib is present, the Oxford experience does not support this concept (Table 2) 190. Recurrence of symptoms usually occurs in the first 12 months but may occur later in a small number of patients due to fibrous scarring resulting in bands which produce pressure on the lower trunk of the plexus once again, or due to regrowth of a resected first rib, which may occur if the rib is removed subperiosteally. For this reason ribs should be removed with the periosteum.

Adson AW, Coffey JR. Cervical rib: a method of anterior approach for relief of symptoms by division of the scalenus anticus. Ann Surg 1927; 85: 839–57.

Coote H. Exostosis of the left transverse process of the seventh cervical [022] vertebra surrounded by blood vessels and nerves: successful removal. Lancet 1861; i: 360.

Davies AH, Walton J, Stuart E, Morris PJ. Surgical management of the thoracic outlet syndrome. Br J Surg 1991; 78: 1193–5.

Eastcott HHG. Reconstruction of the subclavian artery for complications of cervical rib and thoracic outlet syndrome. Lancet 1962; ii: 1243–6.

Kieffer E. Arterial complications of the thoracic outlet syndrome. In: JJ Bergan, JST Yao, eds. Evaluation and Treatment of Upper and Lower Extremity Disorders. Orlando: Grune & Stratton, 1984: 249–75.

Lewis T, Pickering GW. Observations on maladies in which the blood supply to digits ceases intermittently or permanently, and upon bilateral gangrene of digits; observations relevant to so-called Raynaud's disease. Clin Sci 1934; 1: 327–66.

Murphy JB. Case of cervical rib with symptoms resembling subclavian aneurysm. Ann Surg 1905; 41: 399–406.

Murphy T. Brachial neuritis caused by pressure of 1st rib. Aust Med J 1910; 15: 582–5.

Peet RM, et al. Thoracic outlet syndrome: evaluation of a therapeutic exercise programme. Mayo Clin Proc 1956; 31: 281–7.

Roos DB. Transaxillary approach for 1st rib resection to relieve thoracic outlet syndrome. Ann Surg 1966; 163: 354–8.

Roos DB. New concepts of thoracic outlet syndrome that explain etiology, symptoms, diagnosis and treatment. Vasc Surg 1979; 13: 313–21.

Strange-Vognsen HH, et al. Resection of the first rib, following deep arm vein thrombolysis in patients with thoracic outlet syndrome. J Cardiovasc Surg 1989; 30: 430–3.

Thompson JF, Webster JH. First rib resection for vascular complications of the thoracic outlet syndrome. Br J Surg 1990; 77: 555–7.

Tilney P, Griffiths HTG, Edwards EA. Natural history of major venous thrombosis of the upper extremity. Arch Surg 1976; 101: 792–6.

Vollmar J, Heyden B. Thoracic outlet syndrome with vascular complications. J Cardiovasc Surg 1979; 20: 531–6.