RETURN TO ABSTRACT LISTING


Objective:
Hamstring tendons autografts are commonly used for anterior cruciate ligament reconstruction (ACLR) with low morbidity reported. Post-harvest regrowth of semitendinosus (ST) and gracilis (G) tendons is common, but with variable degrees of “neotendon” formation described, mostly in short- and mid-term reports. This long-term MRI study describes the end-result of hamstring neotendon formation in patients 10+ years after autograft hamstring ACLR.

Materials and Methods:
51 patients treated with unilateral hamstring autograft ACLR in the Multicenter Orthopedic Outcome Network (MOON) nested cohort underwent 3T knee and thigh MRI of both limbs at a follow-up period exceeding 10 years. Qualitative review of operated knees categorized patients by the degree of regeneration and insertion site location. Manual measurements of all knees included ST and G insertion location relative to the joint line (JL), tendon cross-sectional area (CSA) 4 cm above the joint and myotendinous junction (MTJ). Mean insertion-JL distance and tendon CSA for operated and non-operated limbs were compared by paired Student t-tests.

Results:
Three grades of tendon regeneration of hamstring tendons in MRI scans were defined by insertion site: full (at pes anserinus), partial, or none (no tendon visible in thigh or knee). 61% of ST tendons and 48% of G tendons showed full regeneration, while 20% of ST and 13% of G tendons had no regeneration. Partially regenerated ST and G tendons were attached to the popliteal fascia at varying distances from the joint line. Most of the partially regenerated tendons (80% of ST and 58% of G) had a characteristic thin low-signal tissue which extended from the neotendon insertion to the pes anserinus suggestive of scar tissue from remnant tendon sheath. The mean CSA of completely regenerated ST tendons in operated limb was significantly higher at 4 cm above the joint line (0.15±0.01.cm²; p < 0.05) and the MTJ (0.17±0.01 cm²; p < 0.01) compared to the non-operated limb (0.12±0.03 cm² at both levels). However, there were no significant differences found in the CSA of the G tendon either at 4 cm above the joint (p = 0.92) or the MTJ (p = 0.87). Additionally, the mean insertion point of the regenerated tendons was significantly more proximal than non-operated tendons with the distance from the joint line in operative limbs (3.9±0.6 cm) vs. non-operated limbs (4.3±0.6 cm; p < 0.05).

Conclusion:
Most previously harvested ST and G tendons exhibit full or partial regeneration 10 years after ACLR. However, for a number of patients, roughly 1 out of 5 ST and 1 out of 8 G tendon harvests, no neotendon was seen. These post-harvest tendon healing results, i.e. reattachment site, scar tissue, and CSA, highlight the variability of tendon regeneration that likely influence the hamstring muscles’ functional capacity in the long term. Recognition of the tendon harvest healing response patterns on knee MRI may alert clinicians to postoperative hamstring muscle status. Further studies on the cross-sectional muscle area of the harvested hamstring tendons will determine muscle function.