Reconstruction of displaced acromio-clavicular joint dislocations using a triple suture-cerclage: description of a safe and efficient surgical technique

For this retrospective data analysis we evaluated the records of all patients operatively treated due to AC joint dislocation Rockwood[1] type III to V between 2007 and 2009 in a level 1 trauma centre. In all cases an anatomic reconstruction of the CC and the AC ligaments using cerclages with resorbable sutures was performed. For inclusion in the study, which was approved by an ethic committee, all patients had to declare their written consent to participate. Furthermore the patients had to be skeletally mature and at least 2 years out of surgery. Patients with chronic AC-joint dislocation (delay from injury to surgery of more than 21 days), previous AC- joint surgery, and/or associated fractures of the clavicle, acromion or the coracoid process were excluded from the study. Of the 39 patients that met the inclusion criteria, 33 (85%) were available for clinical evaluation. From these 33 patients three refused the x-ray at follow up. The average patient age was 39 years (range, 18–71 years) at time of surgery. According to the Rockwood classification[1] there were nine type-III, nine type-IV and fifteen type-V injuries in 30 male and 3 female patients (see Table1). In 19 patients the dominant arm was affected. Twenty-five patients had suffered an isolated injury of the AC-joint. Three patients had additional ipsilateral rib fractures, one patient had a fracture of the contralateral clavicle, and another patient had a concomitant type B fracture of the pelvis.

The predominant injury mechanisms were bicycle/motorbike accidents in thirteen patients and falls during skiing or snowboarding in eight patients. Thirty patients were employed and twenty patients were previously engaged in recreational sporting activities.

The mean interval between injury and operation was 5 days (range 0–9 days). General anaesthesia was used in all cases and patients were placed in “beach-chair” position. An anterior approach using a vertical skin incision medial to the AC-joint (“saber cut”) was performed in all cases. After exposure of the coracoid process, a subcoracoid passage of two 1.5 mm PDS-cords was performed using a Deschamps. Next, two drill holes are made into the clavicle, considering the anatomic insertions of the CC ligaments[20]. The reduction was performed under visual and fluoroscopic control and the sutures were tightened after transosseous passage of the clavicle. An anatomic reduction of the joint was performed, though published data have shown a an elongation of the PDS material[2, 21‐23]. In addition, the AC ligament complex was reconstructed by use of a transosseous 1.0 mm PDS® cerclage (see Figure1). Finally, the trapezoid/deltoid fascia was reconstructed and torn disci in the AC- joint were removed.

For postoperative management, the shoulder was immobilized in a sling (Medisling®, Medi, Bayreuth, Germany) for six weeks. Passive and active (gravity-assisted) exercises out of sling were started under physiotherapists supervision two days after surgery limiting the range of motion to 90° of shoulder flexion and abduction. Six weeks after surgery, the sling was removed and full range of motion allowed. Heavy weight lifting and overhead sporting activities were restricted for 12 weeks postoperatively.

At time of follow-up all patients were invited for clinical evaluation. Personal interviews and physical examinations were carried out by an independent investigator (L.M), not involved in the patients initial treatment. The Constant- Murley score[24], the DASH Score, the ASES score, the Simple Shoulder Test and the VAS were recorded.

The radiological evaluation included weighted shoulder radiographs (10 kg each side) in anterior-posterior (A-P) direction (“panorama view”, see Figure2) pre-operatively, post-operatively and at time of follow-up. Additionally, we performed axial radiographs of the operated shoulder to describe the vertical and the A-P translation of the AC-joint. According to Hessmann et al.[25] the radiographs were analyzed by measuring the CC distance (CCD = distance between the inferior rim of the clavicle and the superior rim of the coracoid) on both shoulders. Moreover, the vertical AC distance (vACD = distance between the superior rim of the clavicle and the acromion) was evaluated. On the axial view, the horizontal acromio-clavicular distance, meaning the distance between the anterior rim of the clavicle and the anterior border of the acromion, was measured. The congruity of the AC joint, the presence of degenerative changes of the AC joint, and periarticular ossifications were investigated by three authors independently (GHS, FM, SS). CC ligament calcifications were classified as absent, minor (i.e. spots or small ossicles) and major (i.e. almost complete bridging between the coracoid process and the clavicle)[18].

On self-evaluation, 31 patients (94%) graded their functional results as excellent or good, and two graded their results as fair and poor. The results of the self-evaluation could be confirmed by the results in the different evaluated shoulder scores (see Table1).

The mean absolute Constant-Murley score (CS)[24] was 94.3 ± 7.1 (range 73–100) with an age and gender correlated Constant score (agCS) of 104.2% ± 6.9 (88- 123%). According to Boehm et al.[26] these results were subdivided in the following categories: twenty-six excellent results (100–91 points), six good results (89–80 points) and one satisfactory result (79–70 points). There was no significant correlation between the agCS and the grade of injury according to the Rockwood classification (p = 0.39). The agCS was 103.2% for Rockwood type III injuries, 101.7% for Rockwood type IV injuries and 103,9.3% for Rockwood type V injuries.

The mean DASH score was 3.46 ± 6.6 points, and 18 patients reached the maximum score of 0 points. The additional working module (30 of 33 were at work) was 1.5 ± 3.8 points and twenty-three patients (82%) reached the maximum of 0 points.

Twenty patients regularly participated in sports and reached mean 7.2 ± 11.1 points in the DASH sports module. Of these patients, 10 (50%) reached their pre-injury sports activity level at a mean of 17 weeks with the maximum amount of 0 points.

The mean ASES score was 94.6 ± 9.7, and 17 patients (52%) reached the maximum of 100 points. Evaluating the Simple Shoulder Test (SST), we found that six patients were not able to carry a weight of 9 kg on the injured site. Additionally, four patients could not sleep on the affected side and four patients were not able to throw overhead.

On visual analogue scale twenty-five patients had no pain at all, five patients had very slight pain (i.e. VAS=1) and two patients had intermittently a medium pain revealing a value of 5, a regular pain medication was denied. The mean VAS value was 0.5 ± 0.6.

Clinical examination showed no limitation of the range of motion in abduction, flexion, external and internal rotation compared to the contralateral shoulder. The BMI was on average 25.3 kg/m² and twenty patients had normal body weight while thirteen patients were overweight with a BMI >25 kg/m². There was no significant correlation between the BMI and the functional outcome measured by the age and gender correlated Constant score (p=0.79).

For radiological evaluation, the pre-operative, the post-operative and the x-rays obtained at time of follow-up were reviewed (see Table2).

A significant difference was found in the CCD between pre-operative and post-operative radiographs on day two after surgery (19.1 mm± 4.18 vs. 9.0 mm± 3.84, p<0.001). At time of follow-up, the CCD had significantly increased to 12.2 mm± 3.12 (p<0.001), compared to the post-operative value (9.0 mm) and to the CCD of the contralateral uninjured shoulder (9.7 mm ± 1.84, p<0.001). For detailed analysis the results were subdivided into three groups with respect to the CCD difference to the contralateral side: <5 mm, 5-10 mm and >10 mm. There were 28 patients (85%) with a distance <5 mm, 4 patients with a distance of 5-10 mm (12%), and 1 patient with a CC distance of more than 10 mm. There was no statistical correlation between the clinical outcome (measured by the Constant score) and the radiological outcome (CC- difference to the unaffected side) with a Spearman- correlation coefficient of 0.07, p=0.83.

For the vertical AC distance, a mean of 11.9 mm± 4.11 was found for pre-operative radiographs. Two days after surgery this distance was 0.7 mm± 3.9 (p<0.001) and increased up to 3.5 mm ± 4.1 (p= 0.006) at follow-up. We found a slight over-reduction in twelve patients (12/33, 36%) at day two, but only in one patient at time of follow up.

The axial x-ray was used for evaluation of the acromio- clavicular alter displacement. When measuring the distance between the anterior edge of the clavicle and the acromion on the axial the mean distance was 13.2 mm± 10.2 pre-operatively and 0.91 ± 6.5 at follow up (p< 0.001). In 30 patients (92%) the axial view showed the anterior border of the clavicle within 1 cm ventral or dorsal of the anterior rim of the acromion.

In 10 patients (30%) we found signs of calcifications of the coraco-clavicular ligaments, but all calcifications could be graded as minor according to Dimakopoulos[18]. In 6 patients (18%) radiological signs of osteoarthritis of the AC joint occured at time of follow-up, but only one patients showed clinical signs with soreness, tenderness to palpation and positive cross body test.

There was no significant correlation between the CC calcification and the functional outcome measured by the Constant score (p=0.39), nor between osteoarthritis of the AC joint and the Constant score (p=0.38).

The average surgery time was 79 minutes (range 55–111 min). Regarding complications, an early failure of the cerclage reconstruction with complete re-dislocation was found in three patients (8%). In two cases revision surgery was performed, which revealed a knot breakage of the PDS sutures. At final follow up both patients showed excellent functional results (agCM score 95%, respectively 107%).

The third patient presenting with a CC distance of >10 mm at time of follow-up refused revision surgery due to subjective satisfaction (agConstant score of 102%).

Wound infections or nerval irritations did not occur in any of the cases.