The feasibility and results of an arthroscopic removal of humeral locking plates and glenohumeral arthrolysis after proximal humeral fractures.

Purpose: This study describes the use of standard shoulder arthroscopy techniques to remove a proximal humerus locking plate following proximal humerus fracture. The goal of this study was to assess the feasibility and results of this technique.

Methods: This was a retrospective non-comparative study. Inclusion criteria were fracture union when hardware was removed, significant residual glenohumeral stiffness after 6 months of physiotherapy, arthrogenic screw(s) and/or osteonecrosis (partial or complete) of the humeral head resulting in significant pain.

Results: Eleven patients were included in this study. Surgery was successful in all cases, and surgery lasted a mean of 105 ± 10.5 min. Patients' mean age was 54.6 ± 10.6 years and the mean hospital stay 1.6 ± 0.8 days. No patient was lost to follow-up. At the mean last follow-up of 17.7 ± 23.4 months, pain and all functional parameters improved significantly: the Constant score (43.4 ± 8.8 vs 60.5 ± 0.3, p = 0.003), the visual analogue pain score (4.7 ± 1.5 vs 2.8 ± 2.3, p = 0.012), the shoulder abduction (77.7 ± 18.6 vs 104.5 ± 27.3, p = 0.004), the flexion (85.9 ± 30.7 vs 97.7 ± 27.7, p = 0.026) and the external rotation (15 ± 12 vs 31.8 ± 13.6, p = 0.004). Internal rotation improved from L3 to T12. All seven patients who practiced sports before the initial fracture had returned to sports approximately 6 months postoperatively. No infections or wound dehiscence occurred.

Conclusion: Shoulder arthroscopy was found to be feasible for removal of hardware following proximal humeral fracture and can be associated with diagnostic and therapeutic arthroscopy and glenohumeral arthrolysis if required.