Comparison of Graft Failure Rate Between Autografts Placed via an Anatomic Anterior Cruciate Ligament Reconstruction Technique: A Systematic Review, Meta-analysis, and Meta-regression

Background: Recent data from the Danish anterior cruciate ligament (ACL) registry demonstrated increased reoperation rates for hamstring tendon autografts when an anatomic ACL reconstruction is performed. This is consistent with reports of greater time needed for hamstring tendon autografts to mature compared with other autografts. Purpose: To review the literature comparing graft failure rate between patellar and hamstring tendon autografts placed anatomically and to determine if there are differences in return to preinjury activity levels between autografts. Study Design: Systematic review with meta-analysis and meta-regression. Methods: The PubMed, MEDLINE, SPORTDiscus, and CINAHL databases were used to identify studies published from January 1, 2000, through March 7, 2014. To compare postoperative outcomes between patellar tendon and hamstring tendon autografts, summary event rates for graft failure and return to preinjury activity level were calculated. A meta-analysis was performed to calculate a summary odds ratio (OR) for graft failure between autografts using the studies that directly compared the 2 autografts. Meta-regression analyses were performed to assess the influence of postoperative follow-up time on graft failure rate. Results: A total of 28 studies reported graft failures for patellar tendon (6 studies) and hamstring tendon (26 studies) autografts used with anatomic ACL reconstruction; 4 of the 28 were comparison studies. Graft failure rate was not significantly different between patellar tendon (7.0% [95% CI, 4.6%-10.5%]) and hamstring tendon autografts (3.9% [95% CI, 2.7%-5.6%]). The odds of graft failure were slightly higher for hamstring tendon autografts (OR, 1.21 [95% CI, 0.63-2.33]), but this difference was not significant (P = .57). The rate of patients returning to preinjury activity levels was not significantly different between patellar (n = 1 study; 58.1% [95% CI, 40.4%-73.9%]) and hamstring tendon autografts (n = 5 studies; 75.6% [95% CI, 43.7%-92.5%]). Overall graft failure rate was positively associated with postoperative follow-up time, but this effect was only significant with hamstring tendon autografts (P < .05). Conclusion: Differences in graft failure rate between patellar tendon and hamstring tendon autografts were not significant. Although follow-up time was only found to have a significant influence on hamstring tendon graft failure rates, this was likely due to the smaller sample of studies assessing patellar tendon graft failures. Differences in return to preinjury activity levels could not be determined due to the lack of studies assessing that outcome. Both patellar and hamstring tendon autografts demonstrate a low risk of failure and moderately high return to activity level after anatomic ACL reconstruction.