Total hip arthroplasty is the most widely used procedure to treat avascular necrosis (AVN) of the femoral head. Few studies have compared the outcomes of THA in femoral head AVN and primary hip osteoarthritis. Therefore we performed a case-control study to compare THA for femoral head AVN vs. primary hip osteoarthritis in terms of: (1) prosthesis survival, (2) complication rates, (3) functional outcomes and radiographic outcomes, (4) and to determine whether specific risk factors for THA failure exist in femoral head AVN.

THA survival is similar in femoral head AVN and primary hip osteoarthritis.

We compared two prospective cohorts of patients who underwent THA before 65 years of age, one composed of cases with femoral head AVN and the other of controls with primary hip osteoarthritis. In both cohorts, a cementless metal-on-metal prosthesis with a 28-mm cup and an anatomical stem was used. Exclusion criteria were THA with other types of prosthesis, posttraumatic AVN, and secondary osteoarthritis. With α set at 5%, to obtain 80% power, 246 patients were required in all. Prosthesis survival was assessed based on time to major revision (defined as replacement of at least one implant fixed to bone) and time to aseptic loosening. The other evaluation criteria were complications, Postel-Merle d’Aubigné (PMA) score, and the Engh and Agora Radiographic Assessment (ARA) scores for implant osseointegration.

The study included 282 patients, 149 with AVN and 133 with osteoarthritis. Mean age was 47.8±10.2 years (range, 18.5–65) and mean follow-up was 11.4±2.8 years (range, 4.5–18.3 years). The 10-year survival rates were similar in the two groups: for major revision, AVN group, 92.5% (95% confidence interval [95% CI], 90.2–94.8) and osteoarthritis group, 95.3% (95% CI, 92.9–97.7); for aseptic loosening, AVN group, 98.6% (95% CI, 97.6–98.6) and osteoarthritis, 99.2% (95% CI, 98.4–100). The AVN group had higher numbers of revision for any reason (19 vs. 6, P=0.018) and for dislocation (8 vs. 1, P=0.031). Mean PMA scores at last follow-up were comparable in the AVN group (17.65±1.27 [range, 10–18]) and osteoarthritis group (17.59±1.32 [range, 14–18]) (P=0.139). Osseointegration was also similar in the two groups: global Engh score, 26.51±1.81 (range, 14–27) for AVN and 26.84±0.91 (range, 19.5–27) for osteoarthritis (P=0.065); femoral ARA score, 5.83±0.46 (range, 3–6) for AVN and 5.90±0.42 (range, 3–6) for osteoarthritis (P=0.064); and cup ARA score, 5.74±0.67 (range, 3–6) for AVN and 5.78±0.66 (range, 3–6) for osteoarthritis (P=0.344).

Survival in this study was good and consistent with recent data on AVN, with no difference between AVN and osteoarthritis. Revisions for any cause or for dislocation were more common after THA for AVN. Functional outcomes were similar in the AVN and osteoarthritis groups. An anatomical cementless prosthesis combined with metal-on-metal 28-mm bearing provides durable good outcomes.

III, non-randomized comparison of two prospective cohorts.