What Is the 40-year Survivorship of Periacetabular Osteotomy Free From Conversion to THA?

Periacetabular osteotomy (PAO) is a widely accepted technique for the surgical treatment of symptomatic acetabular dysplasia with reproducibly favorable short- and midterm outcomes. Long-term follow-up studies are essential because PAO is performed in young patients with the aim of preserving the native hip for many decades. However, follow-up of 30 years remains scarce, and no study has reported 40-year outcomes after PAO.

(1) What is the cumulative 40-year survivorship free from conversion to THA? (2) How did hip function change and osteoarthritis progress during follow-up of the surviving hips? (3) Which preoperative factors were associated with conversion to THA? (4) What clinically relevant thresholds can be identified for preoperative factors associated with conversion to THA?

This was a retrospective case series of a consecutive, single-surgeon cohort followed longitudinally and previously reported at 10, 20, and 30 years. Between 1984 and 1987, a total of 75 hips (63 patients) were treated with PAO for acetabular dysplasia at the Department of Orthopaedic Surgery of the Inselspital Bern by the originator of the procedure. Of those, 57% (43 of 75) of hips had reached the endpoint (conversion to THA or arthrodesis) before our previous report on this group (published at a minimum of 27 years). Over the entire follow-up, 5 hips (4 patients) in which the patient died with the hip preserved were treated as competing events. Since the previous report, a further 8 hips were converted and 8 hips (6 patients) were lost to follow-up, leaving 11 hips (7 patients) with complete clinical and radiographic follow-up here, all at a minimum of 36 years. During the study period (1984 to 1987), the only indication for PAO was hip dysplasia with closed triradiate cartilage. The study group here consisted of 83% (52 of 63) women, the median (range) patient age at surgery was 26 years (12 to 55), and the median (range) BMI was 22 kg/m2 (16 to 28). A total of 27 hips had previous surgery to attempt sufficient coverage. Clinical and radiographic follow-up was performed at 10-year intervals, at 10, 20, 30, and 40 years after PAO. Survivorship free from conversion to THA was estimated using a competing-risks analysis with death without conversion treated as a competing event. Twelve patients underwent bilateral PAO and contributed two hips each. Because two hips from the same patient are not independent observations, all confidence intervals (CIs) and p values were calculated using a robust variance estimator with the patient as the clustering unit. Clinical scores, hip ROM, and osteoarthritis grade were described at each follow-up. Preoperative factors associated with conversion to THA were analyzed using Cox proportional hazards regression with a robust variance estimator clustered on patient. Thresholds for predictive factors were identified using maximally selected rank statistics. This approach tests every observed value of a factor as a potential dividing line and selects the value producing the greatest difference in survivorship between the two resulting groups.

At 40 years, the survivorship free from conversion to THA was 27% (95% CI 12% to 39%). Merle d'Aubigné-Postel and Harris hip scores improved during the first decade after surgery and remained above the preoperative level at each follow-up visit. Two preoperative factors were independently associated with a higher risk of conversion to THA: a higher Tönnis osteoarthritis grade (HR 2.3 per 1-grade increase [95% CI 1.4 to 3.9]; p < 0.01) and a positive anterior impingement test result (HR 3.9 versus a negative test result [95% CI 1.9 to 7.7]; p < 0.01). Threshold analysis, which identifies the value of a preoperative factor that best separates hips likely to be preserved from those likely to be converted, identified 28 years for age at surgery. Younger age was therefore associated with a higher likelihood of retaining the native hip.

PAO provides durable joint preservation in acetabular dysplasia and can delay THA for decades, with 27% of hips remaining free from conversion at 40 years. Preoperative osteoarthritis and a positive anterior impingement test result were associated with conversion to THA. Older age was also associated with conversion, with a threshold of 28 years, although age was related to the degree of joint degeneration at presentation. Patient selection should therefore be based primarily on the preoperative condition of the joint, with younger age at surgery as an additional consideration.

Level III, therapeutic study.