Prognostic factors associated with outcomes after knee cartilage repair across different procedure types: a systematic review.

PURPOSE: To identify and describe the prognostic factors associated with outcomes after knee cartilage repair within distinct procedure and disease contexts.

METHODS

PubMed, Embase, the Cochrane Library, and Google Scholar were searched from inception to 29 August 2026; backward and forward citation searching was also performed. Eligible reports quantitatively evaluated baseline factor-postoperative outcome associations after knee cartilage repair. Two reviewers independently selected reports, extracted association-level data, and applied the quality in prognosis studies tool. Because effect measures could not be harmonized, findings were synthesized without meta-analysis by disease context, procedure, factor, outcome, and follow-up.

RESULTS

Thirty-one reports were included, with report-level sample sizes totaling 13,022; cohort overlap prevented determination of the number of unique participants. Of 37 study-analysis units, 29 were at high risk-of-bias and 8 at moderate risk; none were at low risk. In isolated marrow stimulation, increasing age was associated with less favorable short- and long-term outcomes in two multivariable analyses at high overall risk-of-bias, whereas lesion-size findings varied across outcomes and follow-up. Autologous chondrocyte implantation (ACI)/matrix-induced autologous chondrocyte implantation (MACI) studies reported outcome-specific associations involving age, sex, body mass index, baseline status, defect characteristics, previous surgery, and manufacturing variables. Adjusted osteochondral-allograft failure analyses identified age, body mass index, diagnosis, graft area, and bony realignment as candidate factors; clinical-threshold and machine-learning findings remained context dependent. In osteoarthritis-associated cell-based cohorts, older age and bipolar lesions were associated with long-term failure in one adjusted analysis, while other findings were imprecise or derived from overlapping cohorts. No factor was consistently established across procedures or outcomes. A formal certainty-of-evidence framework was not applied because of marked heterogeneity, sparse context-specific evidence, and generally high risk-of-bias; definitive between-procedure comparisons were not possible.

CONCLUSIONS

Available evidence identifies candidate prognostic associations within specific disease, procedure, outcome, and follow-up contexts. However, substantial clinical heterogeneity, potential confounding, cohort overlap, and sparse context-specific evidence limit the certainty and generalizability of the findings. The evidence does not support universal clinical cutoffs or conclusions that prognostic effects differ between procedures.

LEVEL OF EVIDENCE: level IV, systematic review of level I-IV studies.