Sacrotuberous Ligament: Relationship to Normal, Torn, and Retracted Hamstring Tendons on MR Images

Abstract

To evaluate continuity of the sacrotuberous ligament (STL) in normal and abnormal hamstring (HS) tendons on magnetic resonance (MR) images and to test the hypothesis that greater degrees of HS retraction are correlated with STL discontinuity. The institutional review board approved this retrospective HIPAA-compliant study and waived informed consent. Control cohort comprised 33 patients (mean age, 54.1 years) without HS abnormalities at hip MR arthrography. Study cohort comprised 100 patients (mean age, 55.3 years) with HS abnormalities at pelvic or hip MR imaging. Two musculoskeletal radiologists independently assessed STL continuity with the ischium and semimembranosus (SM) and conjoined biceps femoris and semitendinosus (BF-ST) tendons and evaluated these tendons for tendinopathy, partial tear, or rupture. A third musculoskeletal radiologist measured retraction of ruptured tendons. Inter- and intraobserver agreement was calculated with weighted κ or intraclass correlation coefficients. HS abnormalities in the cohorts were compared with Mann-Whitney test. In patients with tendon rupture, relationships between qualitative (STL and HS attachments) and quantitative (tendon retraction measurements) data were analyzed with analysis of variance and linear regression with Bonferroni correction. STL was continuous with ischium in all patients. In control patients, STL was always continuous with BF-ST but never continuous with SM. In study patients, BF-ST tendon alone, SM tendon alone, and both BF-ST and SM tendons showed abnormalities in 17, six, and 77 patients, respectively. HS rupture occurred in 24 patients; it involved BF-ST tendon alone in 13 patients and both BF-ST and SM tendons in 11. STL was continuous with BF-ST tendon in 12 patients and discontinuous in 12 patients. Retraction of BF-ST tendon (mean, 33 mm; range, 5-81 mm) was independently correlated with STL continuity with BF-ST (P = .0001) and SM (P = .0004) tendon rupture. Retraction was significantly greater (P ≤ 0.01) when STL was discontinuous and SM tendon was ruptured. Inter- and intraobserver agreement was very good or excellent in categorization of HS abnormalities and measurement of retraction. STL showed continuity with both ischium and BF-ST tendon but not SM tendon. In HS rupture, tendon retraction was significantly less when STL remained attached to BF-ST tendon.