Evaluation of Osseous Incorporation After Osteochondral Allograft Transplantation: Correlation of Computed Tomography Parameters With Patient-Reported Outcomes

Background: Studies have reported favorable clinical outcomes after osteochondral allograft (OCA) transplantation to treat osteochondral defects and have demonstrated that healing of the osseous component may be critical to outcomes. However, there is currently no consensus on the optimal modality to evaluate osseous healing.

Purpose: To define parameters for OCA healing using computed tomography (CT) and to investigate whether osseous healing identified using CT is correlated with improved pain and function on patient-reported outcomes (PROs) collected closest in time to the postoperative CT scan and at final follow-up.

Study design: Case series; Level of evidence, 4.

Methods: Of 118 patients who underwent OCA transplantation for articular cartilage defects of the knee over the 10-year study period, 60 were included in final analysis based on completion of CT scans at 5.8 ± 1.9 months postoperatively and PROs collected preoperatively and postoperatively. CT parameters, including osseous incorporation, bone density, subchondral bone congruency, and cystic changes, were summarized for each patient relative to the cohort. Parameters were assessed for inter- and intrarater reliability as well as for covariation with patient characteristics and surgical variables. Structural equation modeling was used to assess correlation of CT parameters with change in PROs from preoperatively to those collected closest in time to CT acquisition and at the final follow-up.

Results: Bone incorporation was the most reliable CT parameter. The summarized scores for CT scans were normally distributed across the study population. Variance in CT parameters was independent of age, sex, body mass index, prior surgery, number of grafts, lesion size, and location. No significant correlation (P > .12 across all comparisons) was identified for any combination of CT parameter and change in PROs from baseline for outcomes collected either closest to CT acquisition or at the final follow-up (mean, 38.2 ± 19.9 months; range, 11.6-84.9 months). There was a uniformly positive association between change in PROs and host bone density but not graft bone density, independent of patient characteristics and surgical factors.

Conclusion: CT parameters were independent of clinical or patient variables within the study population, and osseous incorporation was the most reliable CT parameter. Metrics collected from a single postoperative CT scan was not correlated with clinical outcomes at ≥6-month longitudinal follow-up.