Primary anterior cruciate ligament repair: magnetic resonance imaging characterisation of reparable lesions and correlation with arthroscopy.

Objectives: A recent treatment algorithm suggests that proximal anterior cruciate ligament (ACL) tears with good-to-excellent tissue quality are amenable to primary repair. Our primary objective was to assess the ability of MRI to determine the exact tear location and tissue quality, using arthroscopy as a reference standard.

Methods: In an initial sample of 71 patients with prior ACL surgery (repair or reconstruction), the diagnostic accuracy of MRI was assessed using arthroscopy as a reference standard. Each native ACL tear was graded according to Sherman's arthroscopic classifications during the surgical procedure. MRI scans were retrospectively reviewed for grading, blinded to arthroscopic findings and in consensus by two musculoskeletal radiologists. Tear location and tissue quality were graded using the MRI Sherman tear location (MSTL), MRI Sherman tissue quality (MSTQ) and simplified MRI Sherman tissue quality (S-MSTQ) classifications. Intra- and inter-observer agreement was assessed on a second sample of 77 patients. MRI classification accuracy was compared by McNemar's tests. Intra- and inter-observer agreement was assessed using Cohen's kappa coefficient.

Results: Regarding tear location, diagnostic accuracy was 70% (50/71) based on the MSTL classification. Diagnostic accuracy for tissue quality was 52% (15/29) based on the MSTQ classification and 90% (26/29) for the S-MSTQ classification (p = 0.003). Inter-observer agreement was good for MSTL (κ = 0.78) and moderate-to-good for the MSTQ and S-MSTQ classifications (κ = 0.44 and 0.63 respectively).

Conclusions: MRI seems to be accurate in assessing tear location and tissue quality and may help clinicians to predict the reparability of ACL tears.

Key Points: • MRI seems to be accurate in assessing tear location and tissue quality and may help clinicians to predict the reparability of ACL tears. • High intra-observer agreement was demonstrated when grading the tear location into one of five types. • The diagnostic accuracy of the simplified MRI tissue quality classification, involving deletion of the ligament stump signal criterion, was better than that observed with the MRI Sherman tissue quality classification, but was moderate to good in terms of inter- and intra-observer agreement.