Abstract
I read with great interest the recent article by Rizvanovic et al 3 comparing patient-reported outcomes and revision rates after primary anterior cruciate ligament (ACL) reconstruction with quadriceps tendon (QT), hamstring tendon (HT), and patellar tendon (PT) autografts in the Swedish Knee Ligament Registry. The authors should be commended for addressing an increasingly relevant clinical question through a large national cohort and, importantly, for presenting sex-stratified analyses with clinically interpretable KOOS4 (mean score of the Knee injury and Osteoarthritis Outcome Score [KOOS] Pain, Symptoms, Sports/Recreation, and Quality of Life subscales) thresholds. Their finding that women receiving QT autografts had lower odds of achieving a minimal important change and Patient Acceptable Symptom State, together with higher odds of treatment failure, compared with women receiving HT autografts is both clinically relevant and hypothesis generating.
At the same time, several issues deserve further consideration before these findings are translated into firm graft selection recommendations. First, the interpretation of the 2-year patient-reported outcome analyses should be tempered by the possibility of differential follow-up bias. Only 44% of the registry cohort contributed 2-year KOOS data, and response proportions differed across graft groups, with the lowest response observed in the QT group. In addition, responders and nonresponders differed with respect to age, sex, injury mechanism, and associated intra-articular pathology. Because the principal signal in the study was sex specific and depended on responder-derived thresholds such as minimal important change, Patient Acceptable Symptom State, and treatment failure, even modest selection effects may have influenced the apparent disadvantage of QT grafts in women. Sensitivity analyses based on inverse-probability weighting or multiple imputation could have helped quantify the robustness of this conclusion.
Second, residual confounding by indication likely remains an important concern. Although the regression models were carefully adjusted for several measured covariates, the registry did not capture the rationale for graft selection, details of rehabilitation exposure, return-to-sport progression, or serial objective strength recovery. These omissions matter because recent evidence suggests that QT harvesting may be followed by greater quadriceps strength deficits than HT or bone–patellar tendon–bone grafting, 2 and sex-specific work has shown delayed quadriceps strength and hop recovery in women undergoing ACL reconstruction with QT autografts compared with HT autografts. 1 In that context, the lower patient-reported outcomes observed in women with QT grafts may reflect, at least in part, a modifiable postoperative recovery pathway rather than an inherent disadvantage of the graft itself.
Third, the absence of between-graft differences in 2-year revision rates should be interpreted cautiously. The revision analysis is valuable, but the number of QT revisions was small, and a 2-year horizon may be insufficient to fully characterize graft durability in subgroups defined by sex and graft type. A combined framework incorporating patient-reported outcomes, objective quadriceps performance, return-to-sport status, and longer-term graft survival would likely provide a more complete basis for individualized graft counseling.
These comments are offered in a constructive spirit. Rizvanovic et al 3 have made an important contribution by highlighting that sex may modify the relationship between graft choice and recovery after ACL reconstruction. The next step, in my opinion, is to determine whether the observed female-specific signal persists after accounting more explicitly for missing outcome data, rehabilitation variables, and postoperative muscle function. Such work could meaningfully refine patient counseling and help identify women in whom QT autografts may require more tailored rehabilitation or closer functional surveillance.
