Summary
The estimated risk of knee donor-site morbidily following autologous osteochondral transplantation for osteochondral lesions of the talus was less than 10% and larger studies tended to report lower risk than smaller studies.
Abstract
Background
Despite the promising clinical results of autologous osteochondral transplantation (AOT) for treating osteochondral lesions of the talus (OLT), the risk of knee donor-site morbidity (DSM) remains a cause of concern. The risk of DSM is not well established in part because of variation in outcomes from a paucity of published reports. Therefore, the purpose of this systematic review and meta-analysis was to estimate the risk of knee DSM in patients receiving AOT in the treatment of OLT.
Methods
A systematic search of the MEDLINE, EMBASE and Cochrane Library databases was performed in February 2018 by two independent reviewers. Included studies were evaluated with respect to quality of evidence, sample size, mean patient age, study design, mean follow-up time, and observed risk of knee DSM. Surgical procedure characteristics, including lesion size, donor-site for the grafts, diameter and number of the grafts required, were collected for descriptive purposes. Random-effects models were used to estimate the average risk of DSM, either assuming no patient lost to follow-up experienced DSM or all patients lost to follow-up experienced DSM. Multivariable meta-regression was used to estimate the association between study characteristics and the observed risk of DSM. If there was evidence of an association between a study characteristic and risk, a subgroup analysis was performed.
Results
Twenty-six studies with 915 ankles (904 patients) were included in the systematic review and meta-analysis. The most common donor-site was the lateral femoral condyle. The mean number of grafts was 1.7, and the mean diameter of grafts was 9.9 mm. The mean estimated risk of DSM was found to be 6.7% (95% CI: 2.8, 11.8; I2=82.1%) assuming no patient lost to follow-up experienced DSM. If all patients lost to follow-up experienced DSM, the means estimated risk of DSM rose to 10.8% (95% CI: 4.8, 18.3; I2=88.7%). There was an association between study sample size and risk of DSM (P < 0.001), and larger studies reported lower DSM risk than smaller studies. In larger studies (N = 30), the mean estimated risk of DSM was 2.8% (95% CI: 1.2, 5.0; I2=47.6%) assuming no patient lost to follow-up experienced DSM, and 5.0% (95% CI: 2.1, 9.0; I2=74.5%) assuming all patients lost to follow-up experienced DSM. High between-study heterogeneity could not be completely explained by variability in study sample size, mean patient age, design, or mean follow-up time.
Conclusions
Based on the current literature, the mean estimated risk of DSM following AOT for OLT was 6.7% assuming no patient lost to follow-up experienced DSM in this analysis. Notably, larger studies tended to report lower DSM risk than smaller studies. Because of the heterogeneity of studies included in this analysis with small sample size and low level and quality of evidence, meaningful analysis is challenging. This data should be considered carefully when considering AOT for the treatment of OLT.