Comparison of Intramedullary and Extramedullary Osteosynthesis for Two-Part Proximal Humerus Fractures

Abstract

The aim of the study. To evaluate the association between fracture characteristics and the chosen fixation method in two-part proximal humerus fractures, and to compare intramedullary and plate osteosynthesis in terms of functional outcomes, complication rates, and reoperation rates. Methods. A retrospective cohort study included 207 patients with two-part proximal humerus fractures according to Neer classification who underwent surgical treatment between 2013 and 2018 at the Pirogov Russian National Research Medical University. Intramedullary nailing with third-generation implants was performed in 115 patients (55,6%), and locking plate osteosynthesis in 92 patients (44,4%). At 12 months, pain intensity (visual analog scale), Constant Shoulder Score (CSS), American Shoulder and Elbow Surgeons (ASES) score, Disabilities of the Arm, Shoulder and Hand (DASH) score, and shoulder range of motion were assessed. Complications were analyzed within 2 years, and reoperations were recorded at 2 and 5 years after osteosynthesis. Results. The study demonstrated that after plate osteosynthesis for two-part proximal humerus fractures, normal postoperative neck-shaft angle values were achieved in 89.1% of patients versus 80.0% after intramedullary fixation (p = 0.074). At 12 months, plate fixation resulted in better and less variable functional outcomes according to CSS, ASES, and DASH scores. At 12 months, 78% of patients reported no pain; in the remaining patients, residual pain was predominantly mild. A modest advantage was observed in the plate fixation group. However, plate osteosynthesis was associated with longer operative time: median duration was 135 [99; 155] minutes versus 72 [50; 103] minutes in the intramedullary group (p 0.001), as well as a higher mid-term complication rate. Within 2 years, complications occurred in 11 (12.0%) of 92 patients in the plate group versus 5 (4.3%) of 115 patients after intramedullary fixation. Reoperations within the first 2 years were performed in 23 (11.1%) of 207 patients, including 17 implant removals, 6 revision procedures, and 3 delayed arthroplasties. Between 2 and 5 years, among 142 followed patients, only 4 additional reoperations were required, all for nail removal. Conclusion. In two-part proximal humerus fractures, plate osteosynthesis provided significantly better functional outcomes at 12 months based on CSS, ASES, and DASH scores (p 0.001). Patients achieved good range of motion after osteosynthesis, with lower variability in the plate fixation group. Plate osteosynthesis was associated with longer operative time and a higher complication rate. Intramedullary osteosynthesis was associated with shorter operative time and a lower complication rate (4.3% vs 12.0% at 2 years), but more frequently resulted in residual pain and slightly inferior functional outcomes.