ACL Reconstruction: Patellar Tendon vs. Hamstring Graft
Abstract & Commentary
Synopsis: Patients who had patellar tendon graft ACL reconstructions had statistically significant better results than patients who had 2-strand hamstring graft ACL reconstructions with or without extra-articular procedures. The extra-articular procedure provided no additional benefit.
Source: Anderson AF, et al. ACL reconstruction. A prospective randomized study of 3 methods. Am J Sports Med. 2001;29(3):272-279.
The debate regarding the optimum graft choice for ACL reconstruction continues to rage. Several recent papers have suggested that hamstring grafts may not be as efficacious as patellar tendon grafts, at least in certain circumstances. The current paper lends further support to patellar tendon advocates, but these results must be carefully scrutinized.
A total of 105 patients with isolated ACL tears were randomized into 3 groups: 1) ACL bone-patellar tendon-bone (BPTB) graft; 2) ACL semitendonosus and gracilis tendon graft with extra-articular iliotibial band tenodesis; 3) ACL semitendinosus and gracilis tendon graft alone. All patients were thoroughly studied pre- and postoperatively with complete International Knee Documentation Committee (IKDC) and Hospital for Special Surgery (HSS) evaluation to include KT-1000, Cybex, and radiographic examination. Although precise details of the procedures are not given in the article, isometry was used to select femoral tunnel location, the patellar tendon grafts were secured distally with 2 staples, and the hamstring grafts were left attached distally and were only 2-strand grafts. Anderson and colleagues obtained follow-up for 102 of the 105 patients at a minimum of 24 months and an average of 35.4 months postoperatively. There was no significant difference in the incidence of patellofemoral crepitus between the groups, although group 3 (hamstring alone) did have the lowest incidence (10% vs 35-40% in the other 2 groups). The mean max-manual KT-1000 side-to-side difference of 2.1 ± 2.0 mm in group 1 was statistically better than the difference in group 3 (3.1 ± 2.3 mm), but not group 2 (2.6 ± 2.2 mm). "Normal" or "nearly normal" results on IKDC evaluation were obtained in 34/35 patients in group 1, 23/34 patients in group 2, and 26/33 patients in group 3. Subjective results and all other objective measures were similar in all groups.
Comment by Mark D. Miller, MD
Although it is tempting to apply these results to our current practice and drastically reduce the number of hamstring ACL reconstructions we do, this study does not definitively solve this problem. The fact that a 2-incision doubled hamstring construct was used in this study (and is no longer used in clinical practice) completely challenges the results. It is clear that an extra-articular procedure adds little to the results, and this has been shown in other studies as well. Note that the patellofemoral crepitus was markedly higher in the BPTB group. Although Anderson et al suggest that this difference was not statistically significant, it must have closely approached significance, and additional subjects may have made it so. Anterior knee pain was not specifically sought out in this study, and this information would have been helpful. These 2 issues are critical, and yet they are not well addressed in this paper. In fact, on face value alone, this paper challenges 2 commonly held axioms: 1) BPTB and hamstring tendon ACL reconstructions yield similar clinical results; and 2) The incidence of anterior knee pain is higher in patients with BPTB ACL reconstructions. Anderson et al conclude that: 1) BPTB ACL reconstructions result in significantly better objective clinical results; and 2) the incidence of "crepitus" (and by inference anterior knee pain) is similar in both groups. I do not believe that these results can be applied to current clinical practice, however, because: 1) current techniques are markedly different (4 strand grafts with improved fixation); and 2) the incidence of anterior knee pain was not thoroughly investigated.