Rarely do we see the long-term results of “robbing Peter to pay Paul”1 in orthopaedics, and the paper by Snow et al. is an important addition to the literature in describing the substantial loss of muscle volume and fatty atrophy in the harvested semitendinosus and gracilis muscles at nine to eleven years after harvest for anterior cruciate ligament (ACL) reconstruction. In this study, the authors enrolled ten patients for bilateral magnetic resonance imaging (MRI) scans from a list of 220 patients who had undergone isolated ACL reconstruction with doubled semitendinosus-gracilis grafts by the senior authors nine to eleven years prior to the study. These patients had no complaints of pain or instability, and had returned to their preinjury (recreational) level of sports. Images were acquired with a 1.5-T unit and were digitally analyzed by board-eligible orthopaedic surgeons who had been instructed by a musculoskeletal radiologist. The mean volume of the semitendinosus muscle on the operatively treated side was 58.5% of that on the nonoperatively treated side, and the gracilis demonstrated a similar degree of atrophy, with a mean volume of 54.2% of that on the nonoperatively treated side. These muscles also demonstrated substantial areas of fatty infiltration, and the muscle belly had retracted proximally. Tendon regeneration was quite variable, with six of ten having regeneration proximal to the joint line and only two with regeneration distal. Two patients had no discernible tendon regeneration. The long head of the biceps femoris demonstrated an apparently compensatory mean increase of 8.3%; however, the increase in volume did not equal the loss in the harvested medial hamstrings.
These results by Snow et al. confirm earlier reports at a shorter-term follow-up by other authors, including Williams et al.2 and Burks et al.3, demonstrating significant atrophy in the harvested muscles at six to twelve months. Williams et al., in a study of eight patients, found substantial loss of volume, cross-sectional area, and length in the harvested semitendinosus and gracilis, with extremely variable tendon regeneration. There was also a suggestion of compensatory biceps and semimembranosus hypertrophy. Burks (one of the senior authors of the current paper) and colleagues previously reported that substantial progressive atrophy and retraction of the semitendinosus (the “window shade” effect) occurred in nine patients studied at three and twelve months. This was also correlated with persistent strength deficits in isokinetic testing.
The clinical importance of this persistent muscle volume loss, fatty atrophy, and variable tendon regeneration of the medial hamstrings remains unclear, and it would have been a valuable addition to the paper to have quantitatively described the effects of these anatomical changes on strength and function. Variable degrees of knee flexor weakness have been identified after hamstring tendon harvest, with testing at knee flexion of >70° and internal rotation torque generation demonstrating the greatest deficits. The semitendinosus and gracilis muscles play a greater role at greater degrees of flexion, and the anteromedial insertion of these tendons on the tibia at the pes anserinus provides an internal rotation moment. The remaining knee flexors (biceps femoris and semimembranosus) or proximally regenerated tendons are unable to adequately compensate4 when these specific parameters are tested. However, if the gracilis is preserved, these weaknesses are somewhat mitigated5. This could be explained by compensatory gracilis hypertrophy or by regeneration of the semitendinosus along the gracilis to its distal insertion.
A relationship between tendon regeneration and flexor strength and functional tests has been recently demonstrated by Choi et al.6 in a study of forty-five patients at a minimum follow-up of two years. Individual tendon regeneration was associated with better strength and improved functional performance on the carioca test. However, the Lysholm, Tegner, and International Knee Documentation Committee scores demonstrated no significant differences among any of the three groups.
So what is the ultimate message to be derived from the results described by Snow et al.? The authors do not provide any functional or clinical outcomes, other than the fact that the subjects had returned to regular sports activity and had no instability or pain complaints. The study is based on data from only ten subjects, although they had remarkably consistent deficits at this longer-term follow-up. We have to extrapolate from other short-term data the finding that the muscle volume loss, atrophy, and variable tendon regeneration at longer-term follow-up will likely result in continued deficits in deep knee flexion strength and internal rotation torque. However, the limited data we have also suggest that these somewhat alarming anatomical abnormalities do not adversely affect clinical results. The mild biceps femoris hypertrophy and incomplete medial tendon regeneration may be sufficient to provide enough knee flexion torque to effectively compensate for the persistent deficits in the harvested muscle. It remains to be seen if different degrees of muscle atrophy and tendon regeneration will have any clinically relevant impact on patients at longer-term follow-up.