BackgroundOsteonecrosis of the femoral head (ONFH) is a disabling disease that can ultimately progress to collapse of the femoral head, often resulting in THA. Core decompression of the femoral head combined with cell therapies have shown beneficial effects in previous clinical studies in patients with early-stage (Association Research Circulation Osseous [ARCO] Stage I and II) ONFH. However, high-quality evidence confirming the efficacy of this treatment modality is still lacking.Questions/purposes(1) Is core decompression combined with autologous osteoblastic cell transplantation superior to core decompression with placebo implantation in relieving disease-associated pain and preventing radiologic ONFH progression in patients with nontraumatic early-stage ONFH? (2) What adverse events occurred in the treatment and control groups?MethodsThis study was a Phase III, multicenter, randomized, double-blind, controlled study conducted from 2011 to 2019 (ClinicalTrails.gov registry number: NCT01529008). Adult patients with ARCO Stage I and II ONFH were randomized (1:1) to receive either core decompression with osteoblastic cell transplantation (5 mL with 20 x 10⁶cells/mL in the study group) or core decompression with placebo (5 mL of solution without cells in the control group) implantation. Thirty percent (68 of 230) of the screened patients were eligible for inclusion in the study; of these, 94% (64 of 68) underwent a bone marrow harvest or sham procedure (extended safety set) and 79% (54 of 68) were treated (study group: 25 patients; control group: 29). Forty-nine patients were included in the efficacy analyses. Similar proportions of patients in each group completed the study at 24 months of follow-up (study group: 44% [11 of 25]; control: 41% [12 of 29]). The study and control groups were comparable in important ways; for example, in the study and control groups, most patients were men (79% [27 of 34] and 87% [26 of 30], respectively) and had ARCO Stage II ONFH (76% [19 of 25] and 83% [24 of 29], respectively); the mean age was 46 and 45 years in the study and control groups, respectively. The follow-up period was 24 months post-treatment. The primary efficacy endpoint was the composite treatment response at 24 months, comprising the clinical response (clinically important improvement in pain from baseline using the WOMAC VA3.1 pain subscale, defined as 10 mm on a 100-mm scale) and radiologic response (the absence of progression to fracture stage [≥ ARCO Stage III], as assessed by conventional radiography and MRI of the hips). Secondary efficacy endpoints included the percentages of patients achieving a composite treatment response, clinical response, and radiologic response at 12 months, and the percentage of patients undergoing THA at 24 months. We maintained a continuous reporting system for adverse events and serious adverse events related to the study treatment, bone marrow aspiration and sham procedure, or other study procedures throughout the study. A planned, unblinded interim analysis of efficacy and adverse events was completed at 12 months. The study was discontinued because our data safety monitoring board recommended terminating the study for futility based on preselected futility stopping rules: conditional power below 0.20 and p = 0.01 to detect an effect size of 10 mm on the 100-mm WOMAC VA3.1 pain subscale (improvement in pain) and the absence of progression to fracture (≥ ARCO Stage III) observed on radiologic assessment, reflecting the unlikelihood that statistically beneficial results would be reached at 24 months after the treatment.ResultsThere was no difference between the study and control groups in the proportion of patients who achieved a composite treatment response at 24 months (61% [14 of 23] versus 69% [18 of 26]; p = 0.54). There was no difference in the proportion of patients with a treatment response at 12 months between the study and control groups (14 of 21 versus 15 of 23; p = 0.92), clinical response (17 of 21 versus 16 of 23; p = 0.38), and radiologic response (16 of 21 versus 18 of 23; p = 0.87). With the numbers available, at 24 months, there was no difference in the proportion of patients who underwent THA between the study and control groups (24% [six of 25] versus 14% [four of 29]). There were no serious adverse events related to the study treatment, and only one serious adverse event (procedural pain in the study group) was related to bone marrow aspiration. Nonserious adverse events related to the treatment were rare in the study and control groups (4% [one of 25] versus 14% [four of 29]). Nonserious adverse events related to bone marrow or sham aspiration were reported by 15% (five of 34) of patients in the study group and 7% (two of 30) of patients in the control group.ConclusionOur study did not show any advantage of autologous osteoblastic cells to improve the results of core decompression in early-stage (precollapse) ONFH. Adverse events related to treatment were rare and generally mild in both groups, although there might have been a potential risk associated with cell expansion. Based on our findings, we do not recommend the combination of osteoblastic cells and core decompression in patients with early-stage ONFH. Further, well-designed studies should be conducted to explore whether other treatment modalities involving a biological approach could improve the overall results of core decompression.Level of EvidenceLevel II, therapeutic study.