CHAPTER 237 Thoracic Outlet Syndrome
Vascular TOS is relatively uncommon and accounts for less than 1% of cases of TOS, but it is usually easy to detect on clinical examination or vascular imaging modalities.1 Neither vascular type of TOS (arterial or venous) directly affects the brachial plexus but may cause neurological symptoms related to ischemia.
History of Thoracic Outlet Syndrome
TOS has had a long and tumultuous history. Galen first described the presence of a cervical rib in 150 AD, and Vesalius further characterized this anomaly in the 16th century.2 In 1742, Hunauld introduced the cervical rib and its associated symptomatology, and in 1860, Wilshire noted the relationship between a cervical rib and upper extremity paresthesias.2 One year later, Coote described surgical removal of an exostosis from the C7 transverse process for the treatment of an ischemic hand.3 In 1910, Murphy published his experience with first rib resection and its clinical efficacy,4 and in 1927, Adson advocated scalenotomy without rib resection.5 Until the 1930s, resection of the first rib was the mainstay of treatment. In 1929, Naffziger and Grant described neurovascular compression at the thoracic outlet secondary to scalene muscle anomalies, and in the 1930s, they performed scalenectomies for relief of the symptoms. In 1935, Ochsner and colleagues termed this condition scalenus anticus syndrome.6 It was not until 1956 that the term thoracic outlet syndrome was coined; in this paper, Peet and coauthors reported a series of 55 patients with TOS, 71% of whom improved with conservative treatment.7 In the 1960s there was a resurgence of first rib resection procedures for the treatment of TOS after Clagett’s description of the posterior approach and Roos’s description of the transaxillary approach.8,9 Dr. Kline has advocated the posterior subscapular approach to brachial plexus decompression in selected cases,10,11 a procedure similar to the one described by Clagett. More recently, the choice of approach has usually depended on the surgeon’s preference.
Anatomy
The thoracic outlet refers to the communication of the thoracic cavity with the root of the neck. Although considered to be a misnomer by anatomists, it is designated an outlet because important vessels and nerves emerge to enter the neck and upper extremities.12
Three sites within the thoracic outlet where neurovascular compression may occur have been well described: the interscalene triangle, the costoclavicular space, and the subpectoral tunnel.13 The most important passageway clinically is the interscalene triangle, which is bordered by the anterior scalene muscle anteriorly, the middle scalene muscle posteriorly, and the medial surface of the first rib inferiorly. This triangle contains the trunks of the brachial plexus and the subclavian artery. It is important to note that the subclavian vein runs anterior to the anterior scalene muscle. Immediately distal to the interscalene triangle, the neurovascular bundle enters the costoclavicular triangle, which is bordered anteriorly by the middle third of the clavicle, posteromedially by the first rib, and posterolaterally by the upper border of the scapula (Fig. 237-1). The neurovascular bundle then enters the subcoracoid space, also referred to as the retropectoralis minor space, beneath the coracoid process deep to the pectoralis minor tendon.
Compression or irritation of the brachial plexus, or both, have been described within each of these spaces. Demondion and coworkers used magnetic resonance imaging (MRI) to demonstrate narrowing of these anatomic spaces with certain upper limb movements. Hyperabduction and external rotation of the arm produced compression of the neural elements within the costoclavicular space; arm elevation compressed these elements within the subcoracoid space.14 Using computed tomographic (CT) angiography, Remy-Jardin and associates have also demonstrated vascular compression within these spaces with similar arm movements.15
For practical purposes, the vast majority of neurogenic TOS cases involve compression within the interscalene triangle. Anomalous structures such as cervical ribs, hypertrophied musculature, and fibrous bands may further constrict this space.16,17 Anomalous bands are more common than cervical ribs and may originate from a cervical or thoracic rib, the transverse process of C7, the suprapleural membrane (Sibson’s fascia), or the scalene muscles.16–18
Clinical Findings
In true or classic neurogenic TOS, the typical patient is a young, thin female with a long neck and drooping shoulders,19,20 but this condition is also seen in athletes with overdeveloped scalene musculature.21 Dull aching in the lateral aspect of the neck, shoulder, axilla, parascapular region, and inner portion of the arm is often described. Discomfort may be provoked by repetitive use of the extremity, particularly with overhead activities. Sensory disturbances can include numbness or paresthesias, or both, in a dermatomal pattern over the ulnar aspect of the forearm and hand. Vasomotor disturbances such as changes in skin color and temperature may be seen in advanced cases, presumably related to compression of sympathetic fibers in the lower trunk, C8, or the T1 spinal nerve, versus the circulatory alterations seen in vascular TOS. Muscle wasting and atrophy in the hand can be seen in advanced cases of neurogenic TOS. A classic finding is the so-called Gilliatt-Sumner hand,22 in which a dramatic degree of atrophy occurs in the abductor pollicis brevis and lesser atrophy in the interosseous and hypothenar muscles. It is important to note that the motor findings in patients with true neurogenic TOS include muscles in both the median and ulnar nerve distributions whereas the sensory findings are confined to the ulnar nerve distribution. In most cases the upper and middle trunk elements are not involved, and thus the median nerve sensory distribution is spared. These clinical findings correspond to the typical findings on electrodiagnostic studies.
No provocative tests for neurogenic TOS are very reliable, but the 90-degree shoulder abduction and external rotation test and a Tinel sign over the supraclavicular brachial plexus seem to have the best predictive value. Classic provocative maneuvers include the Roos test (elevated arm stress test to induce reproduction of the neurological symptoms), the Adson test (full neck extension and head rotation toward the side being examined, during deep inhalation, to detect a reduction in radial pulse amplitude), and the Wright test (progressive shoulder abduction to reproduce the symptoms).23,24 Overall, these tests have a sensitivity and specificity of 72% and 53%, respectively.23 False-positive results may occur in as many as 45% of patients with the Adson test, 77% with the Roos test, and 61% with the supraclavicular pressure test, particularly in patients with other entrapment neuropathies.25
Unfortunately, the more common finding in patients with neurogenic TOS is a chronic pain syndrome with features suggestive of brachial plexus irritation or intermittent compression. These patients usually fall into the disputed neurogenic TOS category.26 Physical examination is difficult in these patients because of the tendency to guard the extremity, provide an unreliable sensory examination, and demonstrate a give-way type of weakness.
Vascular-type TOS usually occurs in young adults with a history of vigorous arm activity. Patients with arterial TOS often have complaints of a cold, diffusely painful arm that is easily fatigued with activity. Clinical signs can include coolness, pallor, and cyanosis of the affected hand with diminished or absent distal pulses. In later stages of this disease, gangrene of the digits may occur. A supraclavicular mass or bruit may also be present. Thrombosis of the subclavian vein, also known as Paget-von Schrotter syndrome, is manifested as upper extremity edema and cyanosis with distended superficial veins of the shoulder and chest, often without complaints of pain1 (Fig. 237-2).
Diagnostic Evaluation
Neurophysiologic testing, including electromyography (EMG) and nerve conduction studies (NCSs), can be helpful in the evaluation of patients with suspected neurogenic TOS (Table 237-1). In true neurogenic TOS, EMG and NCSs are most useful. EMG of the affected hand muscles reveals a reduced number of motor units under voluntary control. There may also be an increased incidence of large, long-duration polyphasic potentials, but abnormal spontaneous activity is unusual. Needle examination of the cervical paraspinal muscles should produce normal results. Maximal motor conduction velocity may be slowed in the median nerve but normal in the ulnar nerve, and distal motor latencies for both nerves are normal. Compound motor action potentials recorded over the thenar muscles are reduced in situations of marked axonal loss, whereas those recorded over the hypothenar muscles are generally normal. Sensory nerve action potentials recorded at the median nerve in the wrist have normal amplitude and latency, but they are often small or absent when recorded from the ulnar nerve after stimulation of the fifth finger. Ulnar somatosensory evoked potentials may or may not be abnormal. In disputed neurogenic TOS, electrophysiologic studies are usually normal.27 Nerve conduction velocities for the medial antebrachial cutaneous nerve have been reported to be abnormal in patients with neurogenic TOS in the absence of other electrophysiologic findings28,29 (Fig. 237-3).
Spinal | Cervical disk disease or foraminal stenosis |
Cervical spinal cord tumor | |
Cervical syrinx | |
Peripheral nerve | Brachial plexitis |
Median nerve entrapment neuropathy | |
Ulnar nerve entrapment neuropathy | |
Nerve sheath tumor | |
Orthopedic | Shoulder abnormalities (rotator cuff injury) |
Other | Complex regional pain syndrome |
Fibromyalgia | |
Apical lung lesion (Pancoast’s tumor) |
Radiographic studies are an important component in the evaluation of TOS. Imaging studies of the cervical spine, such as plain radiography, CT myelography, and MRI, are recommended to identify spondylotic disease and to rule out spinal tumor or syrinx. Plain radiographs of the cervical spine or chest, or both, should be obtained to rule out a cervical rib or an enlarged C7 transverse process (Fig. 237-4). The finding of an elongated, tapering, downward-pointing (“beaked”) C7 transverse process or a partial cervical rib strongly suggests the presence of a fibrous band extending to the first thoracic rib. Sites of thoracic outlet compression may be visualized on MRI, and studies have correlated decreased size of the costoclavicular space with TOS symptomatology, particularly with provocative maneuvers.14,30,31 MRI of the brachial plexus is useful to identify mass lesions such as nerve sheath tumors or a Pancoast tumor. If a vascular cause is suspected, a vascular study (MR angiography or venography, CT angiography or venography, ultrasonography, or catheter angiography) of the subclavian vessels should be performed. MR neurography, a technique based on enhancing signal differences between nerves and surrounding tissues, has increasingly been used for the diagnosis of neurogenic TOS.32