OUTCOMES FOLLOWING ACL RECONSTRUCTION USING PATELLAR TENDON VERSUS HAMSTRING TENDON AUTOGRAFT IN ADOLESCENT ATHLETES: AN AGE-MATCH, SEX-MATCHED, AND BODY MASS INDEX-MATCHED COHORT STUDY

Background: Anterior cruciate ligament reconstruction (ACLR) remains one of the most commonly performed orthopaedic surgeries in adolescents and young adults. However, selection of the optimal graft for the individual patient remains controversial, considering multiple graft options, subtle technique differences, and wide-ranging bodies of evidence for each graft option in largely adult populations. However, multiple epidemiologic studies clearly demonstrate adolescents as the most greatly affected age-based sub-population, with the highest rates of re-tear and global morbidity, due to psychological and physical effects of long absences from athletic activity. Nevertheless, methodologically rigorous comparative studies specific to the adolescent athlete are largely absent from the literature. The purpose of the current study is to therefore investigate the outcomes of ACLR in adolescents receiving either patellar tendon/bone-tendon-bone (BTB) autograft (ACLR-BTB) or hamstring tendon autograft (ACLR-HS), using retrospective propensity score matching techniques to generate comparable cohorts. Methods: The study was designed as a retrospective chart review, with collection of cross-sectional data from a single pediatric tertiary care center over a thirteen year period (2003-2015). An initial cohort of patients, aged 13-19 years at time of primary ACLR-BTB or primary ACLR-HS by one of 5 surgeons, was identified through a comprehensive electronic medical record database query. Propensity score matching of ACLR-BTB and ACLR-HS subjects through a logistic regression model, with caliper restriction of 0.1 standard deviations, was then applied, based upon specific baseline characteristics that are frequently used by some surgeons to guide graft selection: age, sex, and body mass index (BMI). The resultant cohort of patients obtained a 1:2 (ACLR-BTB: ACLR-HS, 83 patients:186 patients) match. A chart review was performed to identify patient demographics, surgical data, and post-operative outcomes including graft rupture/re-tear and length of clinical follow-up. Patients were additionally contacted to obtain longer term sport participation and re-injury data that may have been investigated or addressed at outside institutions, as well as to obtain functional outcomes using region-specific questionnaires (Pedi-IKDC and HSS-Pedi FABS), which are validated for both this younger population and adult populations. Due to previous literature-based reports on greater anterior knee pain following ACLR-BTB, compared to ACL-HS, scores from the questions specific to knee pain or kneeling pain were extracted from the Pedi-IKDC results and analyzed to assess the potential influence of anterior knee pain on overall scores within this adolescent population. Results: At time of ACLR, the mean age and BMI of the matched ACLR-BTB patients were 17.1 years (sd: +/-1.14 years) and 25.8 (sd: +/-5.04), compared to 16.9 years (sd: +/- 1.38, p: 0.18) and 24.9 (sd: +/- 5.01, p: 0.18) for matched ACLR-HS patients, with similar male preponderance (ACLR-BTB: 77%, ACLR-HS: 72%, p: 0.47) and mean length of follow-up (ACLR-BTB: 42.9 months; ACLR-HS: 43.3 months). While mean Pedi-IKDC scores were slightly higher in the ACLR-BTB group (88.8; sd: +/-10.2, range: 56.5-100) than the ACLR-HS group (85.2; sd: +/- 16.1, range: 23.9-100), this difference was not statistically different (p: 0.50). When scores on questions regarding knee pain and kneeling difficulty were analyzed, no differences were seen between graft cohorts (ACLR-BTB: median knee pain score: 4, ACLR-HS: 4 p: 0.73; ACLR-BTB median kneeling difficulty score: 4, ACLR-HS: 4, p: 0.21). Activity level, based on the HSS-Pedi FABS score, was also similar (ACLR-BTB: 16, range: 3-29 vs. ACLR-HS: 17, range: 0-30; p: 0.96). Rates of ACL graft re-tear were not significantly different between groups, both when calculated based all possible follow-up methods (ACLR-BTB: 11%, ACLR-HS: 11%, p: 0.91), or when those patients responding to long-term follow up questionnaires were isolated (ACLR-BTB: 15%, ACLR-HS: 16%, p: 0.85). Conclusion/Significance: Despite previous studies, including large, multi-country registry-based analyses, demonstrating lower re-tear rates following ACLR-BTB than ACLR-HS, the current study, focused exclusively on adolescent cohorts matched for age, sex, and BMI, showed no difference in graft re-tear rates between the two cohorts at medium-term follow-up. Moreover, activity scores and functional knee outcome scores appear to be equivalent between these younger graft-based populations. In contrast to previous studies associating BTB autograft with greater long term anterior knee pain, adolescents appear to have equal degrees of knee pain or kneeling pain, regardless of graft choice, which may stem from greater healing and biologic remodeling capacity at graft harvest sites than adults or more modern techniques of patellar bone grafting and paratenon closure. Given the consistently high adolescent ACL re-tear rates in this and other studies, continued focus on ACL prevention programs and optimizing the timing and safety of return to play remains critical.