Abstract
Background: Core decompression (CD) combined with bone marrow aspirate concentrate (BMAC) is used as a joint-preserving treatment for osteonecrosis of the femoral head (ONFH), but factors associated with treatment failure remain unclear. This study evaluated factors associated with total hip arthroplasty (THA) conversion and structural progression after CD with BMAC. Methods: This small retrospective cohort study included 56 patients (56 hips) treated with CD and BMAC for ONFH between August 2020 and December 2025. Corticosteroid exposure, preoperative Association Research Circulation Osseous (ARCO) stage, and clinical outcomes were evaluated. The primary outcome was THA conversion; secondary outcomes included radiographic progression and changes in clinical scores. Multivariable logistic regression was used to evaluate associations with THA conversion and progression to ARCO stage IV. Kaplan–Meier analysis was used to estimate THA-free survival. Median observation time from CD+BMAC to THA or the final native-hip assessment on 1 January 2026 was 28.5 months (IQR, 9.5–34.25). Results: Seventeen hips (30.4%) underwent THA. THA conversion increased from 15.4% in hips without corticosteroid exposure to 83.3% in those exposed for >6 months (p = 0.007). Increasing corticosteroid exposure duration was associated with THA conversion in the two-variable model (aOR, 2.48; 95% CI, 1.31–4.67; p = 0.005). The association between ARCO stage III and THA conversion was not statistically significant; this does not exclude an association. Progression to ARCO stage IV occurred in 92.9% of stage III versus 28.6% of stage II hips (p < 0.001). Both ARCO stage III (aOR, 38.03; 95% CI, 3.90–371.17; p = 0.002) and increasing corticosteroid exposure (aOR, 2.58; 95% CI, 1.20–5.57; p = 0.016) were associated with stage IV progression in the two-variable model. Despite significant clinical improvement, 71.4% of hips showed radiographic progression. Kaplan–Meier THA-free survival was 79.1% at 12 months, 68.3% at 24 months, and 64.3% at 36 months; median THA-free survival was not reached. At the final native-hip assessment, patients who subsequently underwent THA showed significantly smaller improvements in VAS, HHS, and WOMAC scores than those who remained free from arthroplasty (all p < 0.001). Conclusions: Corticosteroid exposure duration and preoperative ARCO stage showed associations with clinical and radiographic outcomes after CD+BMAC. Longer corticosteroid exposure was associated with THA conversion and radiographic progression, whereas advanced ARCO stage was associated with radiographic progression. These associations require validation in prospective studies.