Hamstring vs. Patellar Tendon: Which is Best?

Abstract & Commentary

Synopsis: This study was performed to determine differences in outcome by graft type for anterior cruciate ligament reconstruction. There were no significant differences between the two groups measured by Lysholm knee score or IKDC scores.

Source: Corry IS, et al. Arthroscopic reconstruction of the anterior cruciate ligament. Am J Sports Med 1999;27(3): 444-454.

Although both techniques for anterior cruciate ligament (ACL) reconstruction using patellar tendon and hamstring autographs have been well established as effective, considerable debate persists as to which is better. Previous comparative studies used different fixation methods for each graft type, such as interference screws inside the tunnel for patellar tendon grafts and suspensory fixation outside the tunnel for hamstring grafts. This difference may affect graft stiffness and confound the results. In this study, Corry and colleagues have compared patellar tendon and hamstring ACL reconstructions with identical fixation methods for the first time.

Prior to 1994, all of their patients had patellar tendon grafts; after this point, hamstring grafts were used. All grafts were placed with a single-incision arthroscopic technique and fixed with "blunt-threaded" metal interference screws inside the tunnels. Only isolated ACL tears were included in the study, excluding patients with chondral lesions, other ligament injuries, prior surgery, pre-existing arthritis, or excisions of greater than one-third of the meniscus. Demographics were similar between the two groups. Ninety consecutive patients were available in each group, with follow-up at a minimum of two years available for more than 90% of patients. Postoperative rehabilitation was identical between groups without the use of braces.

Repeat traumatic rupture occurred in two patellar tendon grafts and three hamstring grafts. One additional patient in each group had atraumatic graft failure. There were no significant differences between the two groups measured by Lysholm knee score or IKDC scores. The patellar tendon group had greater thigh atrophy at one year, but this was resolved by two years. Kneeling pain (55% vs 6% at 1 year) and patella crepitus were significantly greater for the patellar tendon group, indicative of donor site morbidity.

Perhaps the most interesting finding involved the KT-1000 arthrometry results for laxity. The mean side-to-side difference for female patients was 2.5 mm for hamstring grafts, which was significantly greater than the 1.0 mm for patellar grafts. Male patients had equal values of 0.9 mm for both groups. This finding was independent of the other variables studied.

Comment by David R. Diduch, MS, MD

This study by Corry et al represents the most balanced, comparative study to date to determine differences in outcome by graft type for ACL reconstruction. Either hamstring or patellar tendon grafts are acceptable choices based upon these results. There were no significant differences by Lysholm or IKDC scores, range of motion, general symptoms, or return to sports. However, differences between groups were demonstrated that may help surgeons and patients with prospective choices.

Although the patellar tendon group was more likely to return to running and jumping-type sports (level I), they had more thigh atrophy and donor site pain. These differences were less pronounced by the two-year follow-up. The most alarming finding was the increased laxity by KT-1000 and Lachman tests for female patients with hamstring grafts. This may be due to differences in bone quality that affect fixation strength of soft tissue with interference screws. Any differences would be more evident for soft tissue grafts that take longer to heal within the bone tunnel than does a bone plug.

The stiffness of a graft is determined not only by its load to failure but also by its fixation method. The load to failure for four-strand hamstring grafts and patellar tendon grafts has been shown to exceed that of the intact ACL. Fixation of the graft closer to the joint line with interference screws increases the stiffness over suspensory fixation at distant sites. Although both experimental groups had interference fit fixation within the tunnels, it would appear that the purchase of the hamstring grafts by the screws within less dense bone in female patients may be sufficiently less to allow slight slippage of the graft. However, despite the difference in measured laxity, major differences were not evident between the groups based on activity levels or knee scores. Hence, this difference in laxity may not matter clinically. Given that both grafts produce acceptable results, it is up to the surgeon and patient to decide which option is best for them. The procedure yielding the most predictable results for a surgeon may be the best option.