Introduction
The results of our recent multicenter randomized clinical trial investigating anterior cruciate ligament reconstruction (ACLR) with or without lateral extra-articular tenodesis (LET) in patients at high risk of graft failure (Stability Study) suggest that LET reduces the rate of ACL failure at two years post-operative.
Purpose
The purpose of this study was to investigate what morphological and physical factors predict failure within the trial.
Methods
624 patients were randomized with a mean age of 18.9 (range: 14-25), 293 male. A regression analysis was performed to determine what factors would be most predictive of graft failure. Within the model, the primary outcome (combined graft failure + persistent rotatory laxity as measured by an asymmetric pivot shift) was the dependent variable and the following independent variables were investigated: 1) sex, 2) group, 3) age at surgery in years, 4) medial meniscus (no pathology, degenerative, excision, repair), 5) lateral meniscus (no pathology, degenerative, excision, repair), 6) Beighton score, 7) presence of knee hyper-extension, and 8) pivot shift under anesthesia
Results
At two years post-operative, 104/252 (41%) of ACLR alone patients suffered the primary outcome compared to 61/252 (25%) of the ACLR+LET patients. 39 patients had suffered graft rupture, 28/252 (11%) in the ACLR group compared to 11/242 (4.5%) in the ACL+LET group. The most significant predictor of failure was the group allocation i.e ACLR alone or ACLR + LET with ACLR alone having an odds of failure about two times greater than those who got the LET (OR=2.1 95%CI 1.4 to 3.0, p<0.001). After controlling for group, for every year of?age, the odds of failure was?reduced by just over 5%?(OR=0.94, 95%CI 0.93 to 0.96, p<0.001).?Compared to having no medial meniscal pathology, the odds?of failure if there is a partial excision is more than two times greater (OR=2.2, 95%CI 1.2 to 4.3, p=0.01). Sex, pivot shift under anesthesia,?lateral meniscal status, Beighton score, and presence or absence of knee hyper-extension were not significant predictors of failure. When graft failure was used as the dependent variable, group and age remain significant predictors of outcome.
Conclusion
Not performing a LET at the time of ACLR and younger age at the time of surgery are significant predictors of poor outcome when performing a hamstring tendon autograft, single bindle ACLR in patients aged 15-25 years old.