In recent years, the incidence of acromioclavicular joint dislocation has been increasing. The main causes are falling from high places, car accidents and sports injuries. Rockwood et al. classified AC joint dislocation into 6 types, which has certain guiding significance for the choice of treatment methods. The conservative treatment of Rockwood type I and II dislocation is satisfactory, while type III and other dislocations require surgical treatment to repair the completely broken coracoclavicular ligament. A large number of surgical methods for acromioclavicular joint dislocation have been reported in literatures, but the surgical choice is controversial.With the development of biomechanics research and material science, more and more surgeons tend to reconstruct coracoclavicular ligament to treat AC joint dislocation. Most surgeons regard the coracoclavicular ligament as a single bundle without reconstructing the trapezoidal ligament and the conical ligament separately. The double-bundle reconstruction reported in a few literatures is not an exact anatomical reconstruction. In this paper, we introduce a method of complete anatomical reconstruction of coracoclavicular ligament with double-bundle Endobutton loop-plate and Ethibond-2 suture according to the anatomy origin. The method was applied to 22 patients with type III-V acromioclavicular joint dislocation, including fresh or old injuries. They were followed up for at least 12 months to evaluate the early clinical outcome.

1. Patient information: Twenty patients with acromioclavicular joint dislocation diagnosed by Jiangsu People's Hospital from August 2013 to June 2015 were included. Another 2 patients from September to December 2015 in Nanjing Qixia District Hospital were also included in this study. There were 15 males and 7 females, with an average age of 44.5 (17-71) years. The dislocation of acromioclavicular joint was unilateral, including 9 cases on the left side, 13 cases on the right side and 13 cases on the dominant side. The causes of injuries included 12 falls, 7 car accidents and 3 sports injuries. X-ray and CT was performed on bilateral shoulder before operation to determine the Rockwood classification, including 7 cases of type III, 1 case of type IV and 14 cases of type V. The including criteria was Rockwood III or above of acromioclavicular joint dislocation and no other history of fracture, dislocation or trauma of the affected limb. The average operation time from injury to reconstruction of coracoclavicular ligament was 24.5 (2-182) years. Old injuries defined as that the treatment was delayed for more than 3 weeks after injury. There were 6 cases of old dislocation and 16 cases of fresh dislocation. This study was approved by the Medical Ethics Committee of the First Affiliated Hospital of Nanjing Medical University. All patients signed the informed consent and agreed to complete anatomical reconstruction of coracoclavicular ligament for acromioclavicular dislocation. 2. Operative methods: Under general anesthesia, we took the beach chair position. A 10 cm long arc incision was made from the distal clavicle to the tip of coracoid process. The skin, subcutaneous, superficial fascia and deep fascia were incised to expose the upper surface of distal clavicle. During this process, the insertion of deltoid and trapezius muscle were dissected minimally, and the coracoid process was exposed through the groove between deltoid and pectoralis major muscles, so as to avoid injuring the cephalic vein.According to the anatomical position of the trapezoidal ligament and the conical ligament, the coracoid process and clavicle were drilled with a 2.0 mm K-wire. The insertion point on clavicle is the projection of trapezoidal and conical ligament on supraclavicle surface, and the outlet point is the attachment of the trapezoidal and conical ligament on the subclavicle surface. The insertion point on the coracoid process is the attachment of conical and trapezoid ligament on the upper surface. The outlet point on coracoid process is on basal surface. The drill diameter should be 3.0 mm to avoid penetrating the cortex of the coracoid process and keep the adjacent bone channels at a certain distance.A double-folded No.2 Ethibond suture A was passed through the clavicle and coracoid tunnel in turn under the guide of suture passer. The acromioclavicular joint is repositioned by manipulation. The acromioclavicular joint is temporarily fixed with a 2.0 mm K-wire. The full reduction of acromioclavicular joint is confirmed by fluoroscopy during operation. The length of the loop is measured by comparing two ethibond sutures of the same length. A suitable size of endobutton (A) is selected according to the measured length. Then pass another No.2 ethibond suture B through the first hole and the fourth hole of the looped plate A, and then take a No.2 ethibond suture C through the loop . Under the guidance of suture A, suture B and C were passed through the tunnel of clavicle and coracoid process. Suture B is gently pulled to place the plate A at the base of coracoid process. Suture C is gently pulled to make the loop pass through the tunnel of coracoid process and clavicle. A plate B without loop is placed in the loop passing through the first and fourth hole of plate B, respectively. After removing K-wire, fluoroscopy showed satisfactory reduction and reliable fixation . Then we explore the acromioclavicular joint, using absorbable suture to strengthen the acromioclavicular joint capsule. After a large amount of saline irrigation, close the wound layer by layer. 3. Postoperative rehabilitation: The patients were encouraged to increase the active ROM of shoulder after 6 weeks. Gradual resistance training was postponed to 8 weeks after operation. Normal physical work and physical exercise could be resumed 6 months after operation. 4. Functional Evaluation: VAS score and Constant score were performed at 3, 6 and 12 months after operation. AP and axillary view of bilateral shoulder joints were re-examined. Coracoclavicular space of affected side and healthy side was measured before and after operation. Coracoclavicular space refers to the vertical distance between the lowest point of the subclavicle surface and the highest point of the upper surface of the coracoid process. All scores and measurements were independently completed by three physicians. 5. Statistical Analysis: Each data is averaged by three times of evaluation results, and all data are expressed by mean±standard deviation. Normality test was performed on each group of data. Paired t test and Wilcoxon test were used before and after operation. P<0.05 showed significant difference. Data were processed by SPSS l8.0 software.

The patients were followed up for 12 to 24 months (the average follow-up time was 17.7 + 4.0 months) . The visual analogue score of pain decreased from 5.0 to 0.2 after 12 months (P<0.001) , and the Constant score increased from 44.3 to 93.7 after 12 months (P<0.001) . The average coracoclavicular space of affected side recovered from 21.0 mm to 8.5 mm 12 months after operation (P<0.001) . There was no significant difference between the coracoclavicular space of affected side and healthy side after operation (P>0.91,) . Four patients with AC joint arthritis complained of mild pain in the shoulder joint. Nonsteroidal anti-inflammatory and analgesic drugs were given orally with good results. Constant score did not decrease significantly. During the follow-up period, no serious complications such as re-dislocation and clavicular or coracoid process fracture were found.

Complete anatomical reconstruction of coracoclavicular ligament with double-bundle Endobutton is a safe, reliable and novel surgical method. It has been applied to the treatment of Rockwood III-V fresh or old acromioclavicular dislocation with good clinical results.