Postoperative computed tomography‐derived bundle morphology ratios are associated with noncontact graft failure after double‐bundle anterior cruciate ligament reconstruction in adolescents

Purpose To evaluate whether postoperative computed tomography (CT)‐derived bundle‐specific morphology ratios are associated with noncontact graft failure following anatomic double‐bundle anterior cruciate ligament (ACL) reconstruction in adolescents. Methods This single‐centre retrospective cohort study included 133 consecutive patients aged 15–19 years who underwent primary anatomic double‐bundle ACL reconstruction with routine postoperative CT at 1 week postoperatively and a minimum follow‐up of 24 months; all patients had a complete 2‐year outcome assessment. Patients with a lateral femoral condyle ratio (LFCR) ≥ 62.18% were excluded, as this threshold has been associated with substantially elevated graft rerupture risk [2]. Only noncontact graft failures within 2 years were considered events. Preoperative MRI was used to assess femoral morphology, and postoperative CT was used to quantify anteromedial (AMR) and posterolateral (PLR) bundle morphology ratios. Receiver operating characteristic analysis and multivariable logistic regression were performed to evaluate discriminative ability and independent associations with graft failure. Results Both AMR and PLR were significantly associated with graft failure and demonstrated moderate‐to‐good discriminative ability (area under the curve 0.804 and 0.799, respectively). In multivariable analysis adjusting for age, sex, BMI, activity level and meniscal injury status, only AMR remained associated with graft failure in an exploratory multivariable model (OR: 1.15 per 0.01‐unit increase, 95% CI: 1.02–1.31; p = 0.026). Intraobserver reliability was good to excellent across all parameters (intraclass correlation coefficients 0.878–0.949). Conclusion Postoperative CT‐derived bundle morphology ratios are associated with noncontact graft failure following anatomic double‐bundle ACL reconstruction, with AMR bundle morphology demonstrating the strongest association in an exploratory multivariable model. These preliminary findings suggest that quantitative morphology ratios may be useful for future risk stratification, but prospective validation is required before clinical implementation. Level of Evidence Level III, retrospective cohort study.