This retrospective comparative study was conducted following receipt of approval fromour institutional review board. The inclusion criteria were as follows: 1. age >18 years; 2. unilateral injury; 3. acute injury(<4 weeks); and 4. high-grade ACJ dislocation (Rockwood type III–VI). Patients with the following conditions were excluded: additional fractures (clavicle, scapulae, or proximal humerus) in the same shoulder, ACJ arthritis, or rotator cuff injury. Patients with a previous injury to the same shoulder and those who were followed-up for less than one year were also excluded from the study. From May 2010 to December 2018, 267 patients with high-grade ACJ dislocations (Rockwood classification type III–V) underwent surgical interventions. The existence of controversy from surgeon to surgeon, the modalities could be divided into CC reconstruction with horizontal K wires fixation and hook plate fixation. Forty-nine of the 267 patients were treated using hook plate fixation (DePuy Synthes 3.5 mm LCP® Clavicle Hook Plate or Aplus® Distal Clavicle HOOK Locking Plate System) and another 218 patients treated with intra-articular Kirschner wires fixation, loop suspension fixation or biologic ligament reconstruction. There were totally six surgeons applied the hook plate fixation in the beginning and the additional CC reconstruction with non-absorbable, braided, sterile polyester surgical suture (Mersilene® Polyester Fiber Suture, Ethicon, Cincinnati, OH, USA) were developed later for the enforcement of vertical soft tissue stability instead of hook plate only (Figure 1).
Finally, 38 patients were recruited in this study. Nineteen patients underwent hook plate fixation alone (HP group), and the other 19 patients underwent hook plate fixation with CC reconstruction using mersilene suture (HM group). Two patients were injured due to falling from a standing height, and the others were involved in motorcycle accidents. All hook plates were removed 3 to 6 months after the index surgery with retained mersilene suture. For the symptomatic subacromial osteolysis, the hook plates will be removed three months after the index surgery. Otherwise, the implants retained until six months.
Pre-and postoperative assessment
Demographic and clinical data were recorded, including age, gender, mechanism of injury, Rockwood classification, interval between injury and surgery, and timing of implant removal. Shoulder functional assessment was conducted using the University of California at Los Angeles (UCLA) Shoulder Score, the Constant Murley Score (CMS) and ACJ-specific Taft score, which includes subscales to assess pain (0–10), night pain (0–5), strength (0–25), activities of daily living (0–20) and range of motion (0–40) and Taft score is the sum of subjective evaluation (pain and stiffness), objective evaluation (ROM and strength compare with opposite shoulder) and radiographic examination (AC joint subluxation, dislocation or post-traumatic osteoarthritis). The subjective pain score was measured using a visual analog scale (VAS). All clinical evaluations were carried out at 1, 3, 6 and 12 months postoperatively
Patients were placed in a beach-chair position under general anesthesia. The approach began from the AC joint at the anterior one-third of the distal clavicle with a 5-6-cm transverse incision, then the ruptured meniscus and hematoma in the ACJ were debrided. The ACJ was reduced and provisionally fixed usingk-wire. In the HP group, an appropriate clavicular hook plate was inserted directly posterior to the ACJ, with the hook portion under the acromion, and the clavicle was part-fixed with screws. In the HM group, CC reconstruction was performed at the beginning by passing two mersilene sutures just underneath the coracoid process with right-angle dissectors; then, two clavicle tunnels of a 2.7 mm width were created 3–4-cm medial to the distal clavicle end between the trapezoid and conoid ligament. The passed mersilene sutures were tied through the clavicle bone tunnels under slight over-reduction of the ACJ. Then, the hook plate applied with the sparing of the clavicle tunnel from screw insertion (Figure 2). Finally, the ACJ capsule and deltotrapezial fascia were repaired using absorbable sutures. All patients tolerated the procedure well, there were no major complications such as neurologic or vascular injury.
A series of plain films, including AP and outlet views, was obtained prior to surgery, on postoperative day 1, and 1, 3, 6 and 12 months postoperatively. In the radiographic assessment, three lines were drawn horizontal to the ground: the coracoidal parallel line was drawn through the superior cortex of the coracoid; the acromial parallel line was drawn through the inferior acromial cortex; and the clavicular parallel line was drawn through the inferior clavicular cortex. The absolute coracoclavicular distance (aCCD) refers to the distance between the clavicular parallel line and the coracoid parallel line, while the absolute acromiocoracoid distance (ACD) was defined as the distance between the acromial parallel line and the coracoidal parallel line. The relative coracoclavicular distance (rCCD) was defined as the ratio of the aCCD to the ACD (aCCD/ACD*100%) (Figure 3). Subacromial osteolysis refers to radiolucent signs around the hook and subacromial space. Distal clavicle osteolysis assessed after remove hook plate. All radiographic examination measurement conducted by a single orthopedic surgeon who did not participate in the surgeries.
The shoulders operated upon were protected by the use of ashoulder sling for six weeks. Passive exercise was initiated immediately after surgery via low-grade forward flexion and pendulum exercises. Active and rotational motion was carried out four weeks postoperatively, and muscle strengthening was initiated after 6 weeks under tolerable pain. Full ROM was permitted before removal of the hook plate. All patients received the similar protocol in our institute, but 1-2 weeks variations existed due to individual difference of recovery.
Continuous variables are expressed as the mean with one standard deviation unless otherwise specified. Categorical variables were evaluated using the Fisher exact test for nonparametric statistics due to the small sample size. The two-tailed Mann-Whitney U test was used for all continuous variables. The significance level was set at 0.05 (p < 0.05). Data were analyzed using SPSS 22.0 for Windows (SPSS, Inc., Chicago, IL, USA).