4.11 The clavicle and sternoclavicular joint

The word clavicle is derived from the Latin clavicula, ‘small key’, which refers to the musical symbol of similar shape. It describes an unusual and interesting bone in its development, function, and the disorders to which it is susceptible. It is the only human long bone that forms by intramembranous ossification. Although the central ossification centre is the first in the body to form, and is responsible for longitudinal growth in the first 5 years of life, the medial and lateral centres, which appear between 12–19 years of age, are among the latest to fuse (22–25 years).

The clavicle is the only bony connection between the upper limb and trunk, forming unusual and incongruous synovial joints at either end, each with a fibrocartilaginous intra-articular meniscus. Medially, there is a saddle-shaped sternal articulation which includes a facet with the first rib in 25% of cases. Both joints are primarily stabilized by strong extra-articular ligaments rather than capsular condensations (Figure 4.11.1). These lie some distance from the joints themselves and act as pivots, enforcing a degree of translational movement. This shearing tends to occur in the two compartments of the sternoclavicular joint with predominantly anteroposterior movements occurring between the sternum and meniscus, and superoinferior movements in the lateral compartment. At rest, the interclavicular ligament helps the superior capsular ligament to produce shoulder ‘poise’, resisting the upward force on the medial clavicle produced by the weight of the arm pulling down from laterally, across the fulcrum of the first rib and rhomboid ligaments.

The medial clavicle and its sternal articulation appear uniquely susceptible to a number of rather unusual disorders, both traumatic and atraumatic, with a very confused terminology.

Whilst subcutaneous and easily amenable to clinical examination (see Chapter 4.1), imaging is difficult. The 40-degree upwards angle ‘serendipity’ plain radiograph is the most useful (Figure 4.11.2), but cross sectional imaging with side-to-side comparison will usually be required.

Fractures and dislocations of the medial clavicle are rare, particularly by comparison with injuries to the shaft and lateral end, comprising less than 3% of injuries. Fractures are caused by high-energy compression and are almost always associated with multisystem trauma. Dislocations, however, may be entirely atraumatic anteriorly (very rarely posteriorly), especially in young adults with ligamentous laxity. In these cases, supportive treatment is usually successful, the results of open surgery poor, and the surgical risks high.

Acute anterior dislocation is caused by a blow to the front of a retracted shoulder with axial compression. Closed reduction under anaesthetic by scapular retraction over a bolster with direct pressure over the medial clavicle is usually (at least temporarily) successful, and may be maintained with a figure-of-eight sling, but redislocation is expected.

Acute posterior dislocation is much less common, caused by axial compression from a blow to the back of a protracted shoulder and may compress the structures of the neck. Closed reduction should be attempted, if necessary assisted by a towel clip, if within 7 days or so of injury, but intervention delayed longer than this is less likely to succeed and more likely to damage the adjacent vessels. Before skeletal maturity (NB 25 years!), the injury is often an occult physeal separation.

In chronic dislocations, the risks and benefits of surgery need to be carefully balanced. Preoperative imaging will include vascular studies. Excision arthroplasty is the preferred procedure with or without one of the stabilization procedures (with a thoracic surgeon on standby). Interosseous metalwork should never be used to maintain reduction in view of the risk of migration causing serious injuries including death.

The investigation and diagnosis of lesions of the medial clavicle is made more difficult by the confusing descriptions and nomenclature of the very large number of eponymous and acronymous conditions beloved of rheumatologists, which are characterized by a spectrum of local and systemic inflammatory and non-inflammatory features. Having excluded tumour and hamartoma, there are four possible aetiopathologies: inflammatory (auto-immune and idiopathic), infective (acute, chronic low grade, often multifocal, and recurrent), ischaemic (Friedrich’s disease and radionecrosis), and degenerative.

Condensing osteitis of the clavicle was first described by Brower in 1974 and typically occurs unilaterally in women of late premenopausal years, sparing the sternoclavicular joint. It shares morphological and radiological features with osteitis of the ilium and pubis, and histologically, despite the title, there is a relative absence of inflammatory reaction. The features are often indistinguishable from aseptic necrosis, which is more usually described in adolescents, has a similarly good prognosis, and also usually spares the sternoclavicular joint. Both disorders run a relapsing and remitting course, eventually clinically resolving. Both may be best described as a variant of osteochondrosis, and the interesting observation has been made that the clavicle, ilium, and pubis all have a fibrocartilage covering, perhaps explaining the predilection of the condition for these sites. Other suggestions have included mechanical compression and various endocrine changes.

Osteoarthritis tends to affect the dominant limb in perimenopausal women.

The prognosis is good for all of the listed conditions after a relapsing natural history with standard supportive measures that may include intra-articular steroid injections.

Sternoclavicular septic arthritis may occur with the same spectrum of organisms as other joints, but is characterized by an insidious onset and a tendency to abscess formation. The joint may also be involved in any of the inflammatory arthritides. Approximately 17% of patients with ankylosing spondylitis have radiological sternoclavicular changes.

In the presence of inflammatory changes extending into the medial clavicle, the terminology is a problem. There is a condition of low-grade inflammation of the medial clavicle that may be recurrent and multifocal (commonly symmetrical). It typically runs a relapsing and remitting course with a good overall prognosis. It was first described in children by Garré in 1893 without the help of any other investigations (just before the birth of radiography) as a chronic, relapsing, and multifocal condition, and is now considered related to a large number of other named non-suppurative inflammatory disorders (Table 4.11.1).

The sternocostoclavicular hyperostosis syndrome (SCCH) was described in middle-aged Japanese men by Sonozaki et al. as a separate entity, involving ossification of the periarticular ligaments of the sternoclavicular joint, and is probably related to diffuse idiopathic skeletal hyperostosis and ankylosing spondylitis. It has been associated with palmoplantar pustulosis in 60% of cases and also psoriasis, suggesting the connection with the SAPHO (synovitis, acne, pustulosis, hyperostosis, and osteitis) syndrome. It may be most appropriate to consider that the SAPHO syndrome is composed of several disorders that share some clinical, radiological, and pathological characteristics, even if the skin disorder is transient or not noticed. Even Tietze’s syndrome has been associated with similar skin lesions. Such are the diagnostic criteria that it may be fair to include many inflammatory conditions under the SAPHO heading since many have a high prevalence of associated skin disorders and therapeutic immunosuppression is reported to be of benefit. On the other hand, the prevalence of sternoclavicular pathology in patients with systemic arthritis and palmoplantar pustulosis is nearly 50% (Figure 4.11.3).

Radiographically, mixed osteolysis with intense sclerosis is seen, often with periostitis, and the enlargement of the subcutaneous bone may become massive, suggesting a diagnosis of fibrous dysplasia, Paget’s disease, or sarcoma.

Histologically, the findings are of chronic inflammation, with mixed new bone and lysis, often with plasmacytosis. These findings are variable, and may sometimes be suggestive of bacterial osteomyelitis. The question of which conditions are caused by an infective agent either in the lesion or elsewhere, and which are best treated by antibiotics and which by immunosuppression is important and has not been answered. It is a concern that Propionibacterium acnes, a common skin saprophyte found in the skin lesions of severe acne, has also been isolated from the bone and joint lesions of patients diagnosed as suffering from both SAPHO and Garré’s disease, as have raised antistaphylolysin titres.

At one end of the spectrum of such disorders is frank osteomyelitis and at the other end, sternocostoclavicular hyperostosis. In between lies a less well-defined variety of musculoskeletal manifestations associated with plamoplantar pustulosis and seronegative arthropathy, perhaps best described as pustulotic arthro-osteitis.

A clinical diagnosis can often be made and biopsy is not always necessary (and in any case usually not helpful). Similarly, an isotope bone scan excludes synchronous lesions. Computed tomography is useful in delineating the extent of any hyperostosis. Magnetic resonance imaging is the investigation of choice for the sternoclavicular joint itself and the periarticular soft tissues. Laboratory tests will exclude metabolic bone disease and help with assessment of the systemic inflammatory response.

In addition to supportive measures, in the presence of inflammation, it would seem reasonable to combine rheumatological treatment of any systemic condition with a therapeutic trial of antibiotics. Although radiation therapy has been suggested for the hyperostotic syndromes it may also be a cause of the clinically similar radiation osteitis! Bisphosphonates have been used but there are few data at present on their effectiveness.

In disease localized to the medial clavicle, whether traumatic, degenerative, or inflammatory, surgical excision may be expected to be curative, but symptoms are usually not sufficiently disabling for patients to undergo the procedure. Good results have been described in small series in many of the discussed conditions following resection of the medial clavicle, but emphasize the importance of either retaining or reconstructing the costoclavicular ligaments.

The medial clavicle is affected by a number of interesting and slightly unusual conditions, many of which are relatively benign and require only supportive measures. There is collection of sclerotic disorders of the clavicle that may be categorized as either inflammatory or non-inflammatory on the basis of clinical, haematological, and radiological signs. In the presence of signs of inflammation, systemic upset, or remote signs, it may be difficult or impossible to exclude infection even on biopsy. In these patients, particularly in children, a trial of antibiotics is recommended. Fortunately, surgical excision of the medial clavicle is rarely necessary, but usually successful when needed.