We retrospectively analysed the clinical results of 30 patients with injuries of the sternoclavicular joint at a minimum of 12 months’ follow-up. A closed reduction was attempted in 14 cases. It was successful in only five of ten dislocations, and failed in all four epiphyseal disruptions. A total of 25 patients underwent surgical reduction, in 18 cases in conjunction with a stabilisation procedure.
At a mean follow-up of 60 months, four patients were lost to follow-up. The functional results in the remainder were satisfactory, and 18 patients were able to resume their usual sports activity at the same level. There was no statistically significant difference between epiphyseal disruption and sternoclavicular dislocation (p > 0.05), but the functional scores (Simple Shoulder Test, Disability of Arm, Shoulder, Hand, and Constant scores) were better when an associated stabilisation procedure had been performed rather than reduction alone (p = 0.05, p = 0.04 and p = 0.07, respectively).
We recommend meticulous pre-operative clinical assessment with CT scans. In sternoclavicular dislocation managed within the first 48 hours and with no sign of mediastinal complication, a closed reduction can be attempted, although this was unsuccessful in half of our cases. A control CT scan is mandatory. In all other cases, and particularly if epiphyseal disruption is suspected, we recommend open reduction with a stabilisation procedure by costaclavicular cerclage or tenodesis. The use of a Kirschner wire should be avoided.
Posterior sternoclavicular dislocation is a rare injury, usually sustained during contact sports.1–5 It is due either to a direct anteroposterior force to the medial aspect of the clavicle, or to an indirect mechanism with a blow to the postero-lateral aspect of the shoulder, displacing it anteriorly.5,6 The clinical diagnosis is sometimes difficult. Standard radiographs in Heinig’s projection are not easily obtained in an emergency setting in a patient who is in pain and mobilising with difficulty. If the diagnosis is merely suspected, a CT scan must be done,3,5,7,8 as it will confirm the diagnosis and reveal associated complications.2,3,7,9,10
In young subjects, before closure of the growth plate of the medial end of the clavicle at 22 to 25 years of age,11 the injury may present as an epiphyseal disruption of the medial end of the clavicle,3,4,12 with posterior displacement, rather than as posterior dislocation. However, it is often very difficult to differentiate these two situations, as the circumstances and mechanism of injury, the clinical presentation and the potential immediate complications are identical.3–5,13 Standard radiographs do not enable differentiation between them, which may also be difficult on a CT scan.10,12,14 The distinction can then only be made in the event of a surgical procedure. Robinson et al10 described epiphyseal disruption as a pseudodislocation.
These injuries are of considerable importance. The proximity of the trachea, oesophagus, and retrosternal vascular and neural structures account for the frequent internal complications, which may be life-threatening.1,2,5,7,9,15,16 Also, some authors have reported cases of post-traumatic degenerative arthritis of the sternoclavicular joint, causing pain which is sometimes intense and which may appear soon after the injury.17 As the sternoclavicular joint is very mobile10 and is subjected to considerable demands in the course of manual and sporting activities, this can endanger the functional prognosis in young, active workers and sports participants with high expectations.
There are few significant published series3,13 with sufficient follow-up to allow a consensus on management to emerge, and this is reflected in the diversity of surgical techniques described in the literature.3,10,14,18–22 There was no consensus in our department as to the optimal management, or as to which surgical procedure might be carried out. The aim of our study was therefore to analyse the functional impact on the upper limb of posterior dislocation of the sternoclavicular joint or epiphyseal disruption with posterior displacement on recreational and occupational activity in young sports players and workers, with a minimum of 12 months’ follow-up, in order to assess the efficacy of the various surgical procedures and to propose a rational programme of management.
Patients and Methods
The orthopaedic departments of two university hospitals in the same city (CHU Rangueil as department A and CHU Purpan as department B) were asked to review retrospectively all patients who presented between 1993 and 2006 with an injury to the sternoclavicular joint with posterior displacement after a sports accident. There were 30 patients (26 men, four women) with a mean age of 21 years (15 to 41) at the time of injury included in the study. The injuries were usually sustained in contact sports. The dominant side was affected in 18 cases. At the initial examination all patients had pain, with four showing signs of mediastinal compression. Of these, three complained of dysphagia, two of which were associated with moderate dyspnoea. One patient had oedema of the arm due to venous congestion. In another patient, who was seen 120 days after injury (patient 1), the radial pulse was abolished with the arm in 90° of abduction but with no sensory or motor deficit. Data concerning the accident, including the sport, mechanism of injury, complications, the type of injury and the centre where the patient was managed, are shown in Table I⇓. All patients underwent a non-contrast thoracic CT scan to confirm the diagnosis and to identify potential complications. In five cases the diagnosis was delayed and was established at between six and 120 days by CT scan. A pre-operative CT scan showed epiphyseal disruption of the medial end of the clavicle with posterior displacement in eight patients. The other five cases were diagnosed at operation.
Rationale of treatment.
A total of 11 surgeons managed between one and eight patients each. If epiphyseal disruption was diagnosed on the CT scan, open reduction was the primary treatment. As regards posterior dislocation of the sternoclavicular joint, the two orthopaedic departments taking part in the study differed in their management of the injury. In the first (department A), closed reduction was not attempted after 24 hours, except for unknown reasons in patient 13. If this failed, or if more than 24 hours had passed, all surgeons stabilised the shoulder by tenodesis after open reduction, except for one who carried out only suture of the costaclavicular ligament. In the second department (department B), closed reduction was attempted up to 48 hours after injury. If this failed, open reduction was undertaken with repair of the costoclavicular ligament. If the repair did not appear satisfactory, stabilisation was achieved by costoclavicular cerclage with PDS sutures. After closed reduction, a CT scan was always obtained. The treatment and operative details are summarised in Table II⇓. After operation the shoulder was immobilised for three weeks. Rehabilitation was commenced between three and six weeks, with abduction strictly limited to 90°. Kirschner (K)-wires were removed at six weeks.
At the follow-up we assessed pain, crepitation, instability and stiffness of the sternoclavicular joint during occupational or sports activities. The intensity and frequency of each symptom were scored on a five-point Likert scale as non, mild, moderate, severe or intense. The functional results were evaluated with the Oxford Shoulder Score23 for pain and function, the Constant score,24 the Simple Shoulder Test25 to assess functional disability, the Disability of Arm, Shoulder, and hand (DASH) score,26 with the Single Assessment Numeric Evaluation27 and a global satisfaction index on numerical scale ranging from 0 (very dissatisfied) to 10 (very satisfied). All patients were examined by the same observer (AE).
This was done using Statview software (SAS Institute Inc., Cary, North Carolina) All data are presented as means sd, except for the time from injury to treatment, which is presented as a median. If the distribution of the values was not normal, a non-parametric test of the Wilcoxon or Mann-Whitney type was used to compare data. A p-value < 0.05 was considered significant.
Closed reduction and operative findings.
The median time between injury and treatment was two days (0 to 127). In 14 patients, ten with a dislocation and four with epiphyseal disruption, closed reduction was attempted under general anaesthesia within 48 hours of injury, and was successful in seven. These seven patients all then had a non-contrast thoracic CT scan to confirm the success of the closed reduction. One patient (case 11) had an early recurrence eight days later, and another (case 12) developed disabling anterior subluxation. Both these patients required surgical stabilisation. The closed reduction was only successful in dislocations. All the failed closed reductions for epiphyseal disruption occurred when the diagnosis was not made on the initial CT scan. A closed reduction was attempted 45 days after injury in one patient (case 13), who had an initial diagnosis of dislocation but had an epiphyseal disruption. This attempt failed. Immediate open reduction was undertaken in 16 patients, eight with a dislocation and eight with a disruption, making a total of 25 patients treated surgically. Of the 12 posterior dislocations undergoing operation, six had major meniscal lesions that required meniscectomy. Reduction was associated with K-wire fixation in seven patients. The costoclavicular ligament was sutured in 11 patients, associated in five with costoclavicular cerclage resorbable PDS suture. In eight patients costoclavicular tenodesis was carried out using the subclavius muscle,18 and in one sternoclavicular tenodesis using the sternal head of the sternocleidomastoid muscle.21 Among the 13 epiphyseal disruptions there were two meniscal lesions that were sutured, two ruptures or severe stretching of the costoclavicular ligaments, and one entrapment of the periosteal sheath which led to failure of the closed reduction. After reduction, the periosteum was sutured and the costoclavicular ligaments were repaired if damaged. In two patients costoclavicular tenodesis with the subclavius muscle was performed, and in three costoclavicular cerclage with PDS sutures. Five patients had temporary stabilisation with K-wires. The reductions were confirmed by CT scans.
No intra- or post-operative complications related to the use of K-wires were observed. One patient (case 17), who had undergone open reduction and isolated suture of the costoclavicular ligament, presented with early redislocation without further injury after 20 days. A further open reduction was carried out with costoclavicular tenodesis. There was no further recurrence.
At a mean follow-up of 59.9 months (14 to 128) 26 patients had been reviewed. Of the remainder, one had died, one was serving a prison sentence and two were lost to follow-up. Pain was experienced in five patients, three minimal and two moderate, during their normal sports activities; eight noted crepitation, six mild, one moderate and one severe, and four complained of minimal stiffness. One patient (case 12), who had been treated surgically for recurrent anterior subluxation, experienced instability during sports activities. The functional results were excellent (Table III⇓).
A total of 18 patients were able to resume their usual sports activity at the same level, three (cases 2, 21 and 25) gave up rugby for another sport with less contact, three (cases 7, 8 and 14) had changed to racket sports (badminton and pelota) using their dominant limb which had been injured, and one had given up sports owing to lack of free time. No redislocation was observed in any of these patients. The functional results, occupation and sports activities are shown in Table III⇑. Only one patient (case 17) considered that his occupational or sports activities were restricted because of his injury.
There was no significant difference between the two centres regarding the functional outcome (p > 0.05 for all parameters studied).
There was no significant difference between patients who had a posterior sternoclavicular dislocation and those with epiphyseal disruption (Table IV⇓). Of the 21 patients operated on and followed up clinically, 14 underwent stabilisation by costoclavicular cerclage, costoclavicular or sternoclavicular tenodesis or K-wiring, and seven had no associated procedure. The mean functional scores (simple shoulder test and DASH) were significantly better and the Constant score was higher and almost reached significance (p = 0.07) (Table V⇓) when an associated stabilisation procedure had been performed. One patient (case 12), treated surgically for subsequent anterior instability, was excluded from this analysis.
This study, including only young workers and sports players, is so far the largest published series of sternoclavicular injuries with posterior displacement.3,13 The follow-up was of moderate duration, but authors who have reported post-traumatic arthritis as the cause of pain have shown that this can appear soon after injury17 because of the demands made on this joint in sports and professional activities. The number of surgeons involved did not allow us to assess the results of any particular technique, but we were able to compare two main types of procedure, namely reduction either alone or followed by surgical stabilisation.
We included patients with posterior sternoclavicular dislocation and epiphyseal disruption because the management of the two conditions has been comparable in many publications,4,5,13 and the circumstances and mechanisms of injury, and the clinical and radiological presentation, are identical.3–5,13 In half of our cases, as in other earlier studies, a CT scan did not diagnose epiphyseal disruption before surgery.10,12,14 Although the clinical outcomes of these two injuries were similar we are not able to recommend that both be managed in a similar way. Because of the systematic failure of attempted closed reduction of epiphyseal disruption due to entrapment of the periosteum, we recommend surgical management. This allows direct control of reduction, suture of the periosteum and repair of the costoclavicular ligaments and meniscus, which are damaged in a significant number of cases. We also recommend stabilisation by costoclavicular cerclage or tenodesis. The use of K-wires is not advised because of the associated risks.
With a sternoclavicular dislocation seen within 48 hours of injury with no signs of mediastinal compression, closed reduction can be attempted under general anaesthesia,4 assisted if required by a sterile clamp inserted percutaneously.28,29 The failure rate remains high,4,5,12,30 and a CT scan is required to monitor the reduction.3,4,7,28 If reduction is achieved, it is usually stable,6,12,13 but injuries to the articular meniscus31,32 may lead to faulty reduction3,31 or instability with recurrent dislocation.32 Open reduction is then necessary to prevent possible secondary complications.15,16,33 Anterior instability has been described after closed reduction,29 and we had a similar case. In the event of failure of closed reduction for a dislocation in people up to 25 years of age, there is a strong possibility that an epiphyseal disruption has not been diagnosed on the CT scan. Strong traction must be avoided, as the nerves may be stretched and damaged.6
When surgical reduction is necessary the procedure must be carried out under general anaesthesia in a centre where a cardiothoracic surgeon can be rapidly available. After reduction, even if it appears stable, we recommend a stabilisation procedure because of the risk of recurrence or anterior instability. Numerous techniques have been proposed.5,10,18–22,34 In order to limit the functional consequences, treatment must spare the joint surfaces, the stabilising structures of the joint capsule and ligaments, and the physis in the young patient. The use of invasive hardware should be avoided.17 The Balser plate14,34 is not appropriate. The hook inserted into the man-ubrium sterni allows some mobility but leads to damage to the cartilage, articular meniscus and physis. The costoclavicular ligament, which is one of the main stabilisers of the sternoclavicular joint, must be repaired.3,22,32 Suture of this ligament is simple, but there is a risk of failure if this procedure is isolated, particularly if the ligament is stretched or lacerated. If a damaged articular meniscus is resected, this widens the joint gap and increases the risk that the suture will fail. Costoclavicular cerclage with non-resorbable suture material associated with repair of the costoclavicular ligament makes the reduction more stable. Complications such as a pneumothorax have been described after cerclage fixation, and in order to avoid this the cerclage must be transcostal and not subcostal. We prefer stabilisation by costoclavicular tenodesis using the subclavius muscle,18 or by sternoclavicular tenodesis using the sternocleidomastoid muscle.19,21 These two methods make the reduction more stable and take advantage of the excellent mechanical quality of the transplanted muscle. We use these techniques in epiphyseal disruption, but they must be associated with periosteal suture.3 Repair can be undertaken with other structures, such as the palmaris longus, tensor fascia lata, gracilis and semitendinosus. Spencer and Kuhn22 have demonstrated that reconstruction using a semitendinosus autograft placed in a figure-of-eight fashion has given the best peak load to failure and best stiffness compared with the use of the tendon of subclavius. Use of the hamstring tendons requires a second surgical site and increases the morbidity of the procedure.35 However, Spencer and Kuhn22 did not study reconstruction using the sternocleidomastoid muscle, which in our short experience is a simpler and more reliable procedure21 than with the subclavius muscle. The latter is sometimes too short for satisfactory reconstruction22 and may be difficult to harvest. In the early stages of our experience we used temporary sternoclavicular wiring with ligament repair. We have now abandoned this technique, as there have been reports of breakage or migration of pins or wires, leading to potentially life-threatening complications.36,37
In spite of the immediate gravity of the injury caused by posterior displacement of the clavicle, if satisfactory and stable reduction is obtained the functional outcome is excellent. Occupational and sports activities can then be resumed at the previous level. We only recommend an attempt at closed reduction with a posterior sternoclavicular dislocation with no signs of mediastinal compression in the first 48 hours after injury. In all other cases we advocate surgical reduction. The capsular ligamentous structures, and in particular the costoclavicular ligament, must be carefully repaired and a stabilisation procedure undertaken.
No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.
- Received December 12, 2008.
- Accepted July 21, 2009.
- © 2010 British Editorial Society of Bone and Joint Surgery