Arthroscopically assisted shoulder arthrodesis: Is it an effective technique?

Purpose: The purpose of this study was to compare clinical and radiographic outcomes of open and mini-open arthroscopic arthrodesis. Methods: Twelve patients underwent arthroscopically assisted glenohumeral arthrodesis over a 5-year period; none were lost to follow-up. Surgery was performed with the patient in the lateral decubitus position, with the arm positioned in 30° each of flexion, internal rotation, and abduction. The articular cartilage was arthroscopically removed from the humerus and glenoid, creating flat opposing surfaces. One to 2 Kirschner wires were inserted percutaneously through the deltoid and across the glenohumeral joint in the center of the articulation; screws were then inserted arthroscopically. The glenohumeral joint was accessed through a mini-open posterior approach beneath the deltoid and was then reassessed before cannulated screws were tightened completely to compress the joint. Two dynamic compression plates were applied to the posterior glenohumeral joint to neutralize rotatory forces. Each patient was immobilized for 6 weeks. Follow-up radiographic imaging was performed at 2, 6, 12, and 24 months after surgery. These patients were then compared with a similar group who underwent a classic open approach. Success of arthrodesis was determined by bone growth across the glenohumeral joint as visualized on axillary radiographs. Results: At 2-year follow-up, complete fusion was achieved in 12 (100%) arthroscopically treated patients, with 2 patients (17%) having early bone grafting (within 6 weeks) through a percutaneous approach. Four patients in the classic open approach group ("open group") required additional grafting. Two patients in the group undergoing arthroscopic surgery ("arthroscopic group") had solid fusion but persisting infection from previously failed operations. Comparison of the 2 groups showed no difference in patient satisfaction or infection rates (2 in each group, all of whom had active infection at the time of the index surgery). One patient in the open group sustained a humeral shaft fracture 5 years after arthrodesis, which required additional surgery. No patient who underwent an arthroscopic procedure required additional surgery other than the 2 early bone grafts. Conclusions: Arthroscopically assisted mini-open glenohumeral arthrodesis provides results that are at least equal to those of open arthrodesis, with a much less invasive approach. Success and complication rates are the same as for the completely open procedure. Level of Evidence: Level IV, therapeutic case series.