부산대학교 도서관

Background: In the last years, there has been increasing use of shoulder arthroplasty. As a result, an increasing incidence of periprosthetic humerus fractures (PPHF) is expected. Therefore, this retrospective, multicenter analysis aimed to collect demographic data from patients with PPHF, their treatment strategies, and associated complications.Methods: Demographics of patients with PPHF were collected retrospectively from the database of six trauma hospitals between January 2000 and December 2020. All fractures were categorized according to the Wright and Cofield, and Worland classifications. In addition, the type of treatment for PPHF, as well as subsequent complications, were evaluated.Results: A total of 72 patients with a PPHF were identified. The mean age of the 55 female and 17 male patients was 77 years. PPHF occurred a mean of 50 months after primary arthroplasty implantation, almost exclusively as a result of a fall from standing height (n = 67). In-situ implant was a RTSA in 40 patients and a hemiarthroplasty in 32 patients. More than half of the fractures (n = 38) were type B fractures, according to Wright and Cofield. According to Worland, the majority of patients also had a type B fracture (type B1 (n = 22), type B2 (n = 28), type B3 (n = 10)). Associated comorbidities were found in 21 patients. A total of 31 patients were treated non-operatively, 27 with Open Reduction Internal Fixation (ORIF), and 14 with a revision arthroplasty. Nine primary radial nerve palsies were found (pre-operatively), and seven secondarily (post-operatively). Nine patients suffered from non-union (ORIF 4, non operative 4, stem replacement 1).Conclusion: The treatment strategy for PPHF is primarily tailored to the individual patient. Decisive factors such as patient’s age, fracture type, and primary palsies have to be considered for optimal treatment. In type B1 fractures, there is a recommendation for ORIF because of high non-union rates under non-operative treatment. In contrast, in type B2 fractures, a non-operative approach should be preferred, provided there are no radial nerve lesions.Level of evidence: Level IV, epidemiologic study.