Reconstruction of the anterior cruciate ligament: meta-analysis of patellar tendon versus hamstring tendon autograft.

PURPOSE: No graft tissue has consistently shown superiority over others for reconstruction of the anterior cruciate ligament (ACL). Bone-patellar tendon-bone (BPTB) and doubled hamstring tendon (semitendinosus and gracilis) (HT) are the most commonly used autologous grafts. We performed a meta-analysis to compare the effectiveness of ACL reconstruction using either BPTB or HT grafts.

TYPE OF STUDY: Systematic review and meta-analysis.

METHODS: We searched the MEDLINE database (1966 to April 2003) for English-language randomized or prospective studies comparing BPTB and 3- or 4-strand HT grafts used for ACL reconstruction. For inclusion, studies were required to follow identical rehabilitation protocols within each study, and provide subjective or objective outcome data after a minimum average 2-year follow-up. Comparison data between BPTB and HT for each identified outcome measure were combined using a random-effects model meta-analysis.

RESULTS: Eleven reports fulfilled the criteria for inclusion. Outcomes favoring BPTB were found in the following outcome measures: normal Lachman examination (relative risk [RR], 0.91; 95% confidence interval [CI], 0.83-0.99; P = .025), normal pivot-shift (RR, 0.94; 95% CI, 0.88-1.0; P = .067), KT-1000 manual-maximum side-to-side difference of < or = 3 mm (RR, 0.75; 95% CI, 0.55-1.01; P = .057), and fewer reconstructions resulting in flexion loss > 5 degrees (RR, 1.41; 95% CI, 1.01-1.96; P = .04). Intermediate level laxity was more common with the HT graft, as shown by higher rates of abnormal Lachman > 0 (RR, 1.22; 95% CI, 0.99-1.5; P = .06), pivot-shift > 0 (RR, 1.3; 95% CI, 0.96-1.75; P = .09), and KT-1000 manual-maximum side-to-side differences > 3 mm (RR, 1.64; 95% CI, 1.13-2.39; P = .01). Outcome measures that favored HT were absence of patellofemoral crepitance (RR, 1.08; 95% CI, 1.01-1.15; P = .03), fewer results with extension loss > 5 degrees (RR, 0.56; 95% CI, 0.3-1.03; P = .06), and kneeling pain. The incidence of instability, as defined by Lachman grade 2, pivot-shift grade 2, or KT-1000 manual-maximum side-to-side difference > 5 mm, was not significantly different between the 2 grafts. All other outcome measures were not significantly different.

CONCLUSIONS: The data presented in this meta-analysis show that the incidence of instability is not significantly different between the BPTB and HT grafts. However, BPTB was more likely to result in reconstructions with normal Lachman, normal pivot-shift, KT-1000 manual-maximum side-to-side difference < 3 mm, and fewer results with significant flexion loss. In contrast, HT grafts had a reduced incidence of patellofemoral crepitance, kneeling pain, and extension loss. The choice of graft by the patient and surgeon must be individualized, and the results of this meta-analysis can aid in the decision by clarifying the risks and benefits of each surgical approach.

LEVEL OF EVIDENCE: Level I.