Proximal Implant Breakage after Cephalomedullary Nailing of Extra-capsular Proximal Femur Fractures: A Multi-Center Retrospective Comparative Analysis.

Abstract

To compare rates of nail breakage of three common cephalomedullary nails (CMNs) for the treatment of AO/OTA 31A1-3 femur fractures. Design: multi-center retrospective study. 13 Level I trauma centers. Adult patients with AO/OTA 31A1-3 femur fractures treated between 2014 and 2021with the Trochanteric Fixation Nail-Advanced (TFNA), Gamma3, or Trigen InterTAN were included. The primary outcome was implant (nail or head element) breakage. The secondary outcomes included nonunion, cut-out/cut-through and overall reoperation rate. Univariate and multivariable analyses were performed to compare breakage rates between implants while controlling for age, sex, AO/OTA 31A1-2 vs 31A3 fracture patterns, low vs high-energy mechanisms and post-operative neck shaft angle (NSA). 2,130 patients were included: 770 (36.2%) TFNA, 1,073 (50.4%) Gamma3 and 287 (13.5%) InterTAN. The InterTAN group had younger patients (median age: InterTan: 74, TFNA: 77, Gamma3: 80, p<0.001), more high-energy mechanisms (InterTan: 23%, TFNA: 14%, Gamma3: 10%, p=0.001), and more varus malreductions (NSA<128.5: InterTan: 46%, Gamma3: 39%, TFNA: 34%, p=0.001). The TFNA group was more likely to have an AO/OTA 31A3 fracture pattern than the Gamma3 (TFNA: 17.3%, InterTAN: 14.3%, Gamma3: 12.1%, p=0.002). The overall rate of implant breakage was 1.4% in the InterTAN group, 1.2% in the TFNA group, and 0% in the Gamma3 group. All breakages occurred in the presence of a nonunion or delayed union. Only two of the 31A3 fracture patterns resulted in breakage, both in the TFNA group, while the remainder occurred in 31A1-2 fracture patterns (p = 1.0). After controlling for confounding factors listed above, the TFNA and InterTAN groups were associated with a marginally increased odds of implant breakage compared to the Gamma3 (TFNA vs Gamma3: OR 0.04, 95% CI <0.01-0.5, p=0.034; InterTAN vs Gamma3: OR 0.04, 95% CI <0.01-0.5, p=0.037), while there was no difference between the TFNA and InterTAN (p=0.991). Post-operative neck shaft angle was not independently predictive of breakage (p = 0.55). This study suggests that the TFNA may be slightly more prone to breakage than the Gamma3 when used for extracapsular proximal femur fractures. However, breakage is a rare event after cephalomedullary nailing with all implants evaluated, is always associated with nonunion or delayed union, and the magnitude of this difference may not be clinically relevant. III.