Diagnosis of long head of biceps tendinopathy in rotator cuff tear patients: correlation of imaging and arthroscopy data

Morgane Rol, Luc Favard, Julien Berhouet

The goal of this prospective study was to assess the reliability of pre-operative cross-sectional imaging for the diagnosis of long head of biceps (LHB) tendinopathy in patients with a rotator cuff tear. Cross-sectional imaging with MRI or CT arthrography data from 25 patients operated upon because of a rotator cuff tear between 1 October 2015 and 1 April 2016 was analysed by one experienced orthopaedic surgeon, one experienced radiologist and one orthopaedic resident. The analysis consisted of determining whether the LHB was present, the extrinsic tendon abnormalities (dislocation, tendon coverage) and intrinsic abnormalities (fraying, inflammation, degeneration). These findings were then compared to intra-operative arthroscopy findings, which were used as the benchmark. The interobserver correlation between the three different examiners for the cross-sectional imaging analysis as well as the correlation between the imaging and arthroscopy data were determined. The correlation between the imaging and arthroscopy data was the highest (80%) for the determination of LHB dislocation from the bicipital groove. The other diagnostic elements (subluxation, coverage and tendon degeneration) were difficult to discern with preoperative imaging, and correlated poorly with the arthroscopy findings (45% to 65%). The interobserver correlation was moderate to strong for the diagnosis of extrinsic tendon abnormalities. It was low to moderate for intrinsic abnormalities. Except for LHB dislocation, pre-operative imaging is not sufficient to make a reliable diagnosis of LHB tendinopathy. Arthroscopy remains the gold standard for the management of LHB tendinopathy, as diagnosed intra-operatively.