Proximal Femoral Replacement in Revision Total Hip Arthroplasty: A 20-Year Experience

Proximal femoral replacement (PFR) is a salvage procedure in revision total hip arthroplasty for extensive femoral bone loss. This study aimed to evaluate implant survivorship, complications, patient mortality, and clinical outcomes of PFR for non-oncologic indications at mid-term follow-up (five years).

We reviewed 61 PFRs for nononcologic indications performed between 2000 and 2022 at a single academic institution. The most common indications were periprosthetic fracture and reimplantation after periprosthetic joint infection (each 33%), followed by aseptic loosening (31%) and dislocation (3%). Femoral bone loss was severe in all patients, with 49% Paprosky type IIIb defects and 51% Paprosky type IV defects. Constrained liners were used in 26 patients (43%), standard head-liner constructs were used in 25 patients (41%), and dual-mobility constructs in 10 patients (16%). The mean patient age was 77 years, with 67% being women. The mean follow-up was five years.

The 5-year cumulative incidence of revision for aseptic loosening was 2%. The 5-year cumulative incidence of any revision of the PFR was 8%. The 5-year cumulative incidence of any revision was 24%. Dislocation was the most common reason for revision (N = 8), followed by periprosthetic joint infection (N = 4), periprosthetic fracture (N = 2), and aseptic loosening (N = 1). Among the 10 patients who dislocated, eight patients (80%) underwent revision to a constrained liner at a mean of two years. The 5-year cumulative incidence of any reoperation was 30%. The mean Harris Hip Score increased from 46 preoperatively to 78 at five years. The 2- and 5-year mortality rates were 15 and 42%, respectively.

In this complex cohort of 61 PFRs for nononcologic revision total hip arthroplasties, aseptic loosening was rare. However, revisions and reoperations for any reason were prominent, with dislocation being the primary failure mode. In addition, the 5-year mortality rate remained high at nearly one in two.

III, retrospective cohort study.