Design: Prospective multicentre registry, inverse probability weighting
Population: 489 patients with primary ACL reconstruction, isolated or with MCL injury
Primary outcome: IKDC subjective score
Effect: Non-hamstring vs hamstring β 7.89 (95% CI 3.45 to 12.33); graft × MCL interaction not significant (P=0.095)

A concern with hamstring autograft in knees with a medial collateral ligament injury is that harvesting semitendinosus and gracilis weakens a dynamic medial stabiliser. A French multicentre registry analysed 489 primary ACL reconstructions from 2020 to 2022, using propensity weighting to compare hamstring and non-hamstring grafts across isolated ACL tears and combined ACL–MCL injuries.

Non-hamstring grafts were associated with IKDC subjective scores about 8 points higher overall (95% CI 3.5 to 12.3), and with better return-to-sport readiness and subjective knee value. Combined injuries managed without MCL surgery had lower scores than isolated ACL tears. The interaction test did not show that MCL status changed the effect of graft type.

A third of eligible patients lacked complete data, and graft choice was surgeon-driven, so residual confounding is likely. The study was also not large enough to detect a modest interaction.

The result does not settle graft choice in the MCL-injured knee, but it gives no specific reason to avoid hamstrings in that setting.

Patients with combined ACL–MCL injuries reported worse outcomes than isolated ACL tears
Non-hamstring grafts were associated with better patient-reported scores overall
There was no clear evidence that an MCL injury makes hamstring grafts worse specifically
Assess and document MCL grade before choosing graft and planning rehabilitation

Why it matters

It tests a widely held assumption about hamstring harvest in medial-side injury.

The statistics, in plain English

The interaction estimate was −7.0 with a confidence interval from −15.3 to 1.2, which crosses zero. That means the study cannot exclude a meaningful interaction; it simply did not show one.