Evaluation of risk factors for treatment failure in pertrochanteric nonunion repair

To characterize the rate of failure and risk factors for treatment failure when operatively treating pertrochanteric and intertrochanteric femur nonunions.

Retrospective cohort study.

One Level 1 North American trauma center.

Patients from 2008-2022 presenting with intertrochanteric or pertrochanteric (AO/OTA 31A1.2-31A3.3) nonunions that underwent nonunion repair.

The primary outcome measure was treatment failure, which was defined as conversion to arthroplasty for reasons other than progression of pre-existing arthritis or persistent nonunion one year after nonunion repair. Patient demographics, caput-collum-diaphyseal (CCD) angle, the presence of medial bone loss, index implant selection, complications, use of bone graft, and presence of atypical femur fractures were examined for association with treatment failure.

Sixty-three patients met inclusion criteria and had adequate follow-up to union or treatment failure (average 23.3 months, range 3-129). Mean age was 59 (range 24-93, SD 15.6), and 47.6% of patients were male. 42 patients (65.1%) were treated with an angled blade plate (ABP); 19 (30.2%) with a cephalomedullary nail (CMN). The rate of failure was 26.5% (n=17). In 23.8% of cases autograft was utilized (n=15), 11.1% a synthetic biologic (n=7), 17.5% allograft (n=11), and 11.1% with allograft and autograft (n=7). [A1] A biologic augment was more frequently utilized in patients treated with an ABP versus CMN (71.4% versus 42.1%, p=0.02). Of the 42 patients managed with ABP there were 14 treatment failures (33.3%), compared with 3 of 19 patients treated with CMN (15.8%, p=0.15). Active tobacco use was associated with treatment failure; 7 of 14 patients went on to treatment failure compared to 10 of 49 without active tobacco use (50% vs 20.4%, p=0.03). In 14 cases, (22.2%) there was medial bone loss noted at the time of revision. 50% of these cases (7/14) failed to unite compared to 20.4% (10/49) of patients without medial bone loss (p=0.02).

Despite variations in implant choice, use of biologic augments, and patient age a failure rate of 26.5% was observed in patients undergoing pertrochanteric and intertrochanteric nonunion repair. Medial bone loss and tobacco use were each associated with a 50% rate of treatment failure, highlighting consideration of these factors preoperatively when considering undertaking nonunion repair.

Retrospective Cohort Study - Level III. See instructions for Authors for a complete description of levels of evidence.