Background: Despite the large number of anterior cruciate ligament reconstructions performed each year, there remains a significant controversy regarding the effect of the graft source on the functional outcome of patients.
Hypothesis: There is no difference in outcomes of autograft versus allograft anterior cruciate ligament reconstructions.
Study design: Systematic review.
Methods: The authors systematically identified prospective studies (Oxford level of evidence I or II only) that included autograft patients, allograft patients, or both. Objective outcomes that were reported were meta-analyzed; this included pivot-shift results, KT-1000 arthrometer results, International Knee Documentation Committee (IKDC) scores, Lysholm Scores, graft failures, and postoperative complications. Two statistical analyses were performed. First a primary statistical analysis was performed comparing pooled autograft data (bone-patellar-tendon bone and hamstrings combined) and pooled allograft data (bone-patellar-tendon bone and hamstrings combined). To have a more comprehensive understanding of the differences between each specific graft source, a secondary analysis was performed without pooling the data; this directly compared the 4 types of graft sources that were studied.
Results: Over 400 scientific manuscripts were initially reviewed; 31 manuscripts fulfilled all of the search criteria. There were very few statistically significant differences between autograft and allograft tissue. The KT-1000 arthrometer laxity testing revealed a mean of 1.4 +/- 0.2 mm (weighted mean +/- standard error of the mean) for the allograft group compared with 1.8 +/- 0.1 mm for the autograft group (t = 2.40; P <.02). However, this difference was only for the mean score; there was no statistical significance when considering KT-1000 arthrometer measurements of greater than 3 or 5 mm. The percentage of patients receiving a final IKDC score of "A" (normal knee) was statistically significant for allograft tissue (43.9% +/- 5.5%) versus autograft tissue (28.2% +/- 1.0%) reconstructions. There was no statistically significant difference between the percentages of IKDC scores of A or B for patients receiving pooled allograft (82.9% +/- 4.2%) versus pooled autograft (87.2% +/- 0.9%) anterior cruciate ligament reconstruction (t = 1.01; P > .1). The graft failure rate was 4.7 +/- 0.5 per 100 for autograft reconstructions and 8.2 +/- 2.1 per 100 allograft reconstructions; although this may represent a trend, it is not statistically significant (t = 1.49; P > .1). The complication rate was slightly higher for autograft reconstructions at 3.5 +/- 0.4 complications per 100 autograft reconstructions compared with 2.4 +/- 1.1 complications per 100 allograft reconstructions, but not significant (t = 1.41; P > .1).
Conclusion: After a comprehensive examination and statistical analysis of the modern literature, the authors could not identify an individual graft source that was clearly superior to the other graft sources. This led them to believe that, with currently available data, the graft source has a minimal effect on the outcome of patients undergoing anterior cruciate ligament reconstruction.