¹Schulthess Klinik, Zurich, Schweiz
²Philipps-Universität Marburg, Marburg, Deutschland
³Universität Bern, Bern, Schweiz
⁴Policlinica del Rosario, Ibiza, Spanien
⁵Charité-Universitätsmedizin Berlin, Berlin, Deutschland

Text

Objectives and questions: This study aimed to determine patient-associated and implant design-related risk factors predictive of acromial and scapular spine fracture risk in Reverse Shoulder Arthroplasty (RSA).

Material and methods: A local shoulder arthroplasty registry was reviewed for primary RSA patients documented to have undergone a radiologic assessment and a minimum 2-years follow-up. Preoperative diagnoses included cuff tear arthropathy, primary osteoarthritis, instability arthropathy, post-traumatic, inflammatory arthropathy, congenital dysplasia, and implant revision. Patient-related risk factors (female gender, prior surgery, acromioplasty, trauma, osteoporosis) and prosthetic design were analyzed. Prosthetic designs were classified into three groups: distalized-medialized or Grammont design (DM), lateralized (L) and lateralized-distalized (LD).

Results: Between March 2006 and December 2023, a total of 2,080 cases of primary RSA (1672 patients, 408 bilateral) with a minimum follow-up of 24 months. Of this cohort of patients, 87 had a fracture of the acromion or the scapular spine (incidence 4.2%). The mean time to fracture was 12.6 months (SD 19.3, range 0-80.2). Most patients with fracture were female (72%) and 686 patients (33%) were diagnosed with osteoporosis. The incidence of stress fracture in the DM design cohort was 4.4%, 4.5% in the L and 1.9% in the LD cohort. According to the Levy classification, 14.9% were type 1, 49.4% type 2, and 35.6% type 3. Most Levy type 3 fractures had a DM design (74.2%), whereas type 1 and 2 had mostly a L design (69.2% and 53.5%, respectively). Within the Levy type 3 group, 6 fractures (19.3%) occurred in shoulders that had os acromiale preoperatively, displaced after implantation of the prosthesis, or subsequent to acromial fractures (bifocal fractures). The treatment was conservative in 86.2% of the patients. Only 39% of the fractures consolidated. Among patients treated conservatively, the healing rate was 34.7% while among those treated surgically 66.7%.

Discussion and conclusions: In this cohort, acromial and scapular fractures were significantly more common in female patients and a non- negligible percentage of fractures is associated with the diagnoses of osteoporosis. Implant design is relevant for occurrence of acromial and scapular spine fractures. Medialized prosthetic designs have an increased risk of producing scapular spine fractures, while lateralized design resulted in increased risk of acromial fractures. Patients who underwent osteosynthesis of fractures showed higher healing rates than those treated conservatively. These data allow surgeons to better understand this complication.