Imaging Analysis of Hill-Sachs Lesions: A Systematic Review

Abstract

To perform a systematic review of the current literature on the accuracy and reliability of different imaging modalities used to assess the presence and various parameters for Hill-Sachs Lesions in patients with anterior shoulder instability. A systematic review was performed according to the PRISMA guidelines. Inclusion criteria were clinical trials or cadaver studies that assessed the accuracy of humeral head bone loss imaging or reliability and clinical relevance in anterior shoulder instability, and were in the English language. Exclusion criteria were animal studies, imaging studies without measures of accuracy, reliability, or clinical predictive power, shoulder injuries without humeral head bone loss, editorial articles, abstracts, presentations, reviews case reports, and surveys. The systematic review identified 40 studies (2,560 shoulders) after inclusion and exclusion criteria application. For diagnosing the presence of a Hill-Sachs lesion, CTA was found to have the highest accuracy (median = 91%, range = 66-100%), followed by MRA (median = 90.6%, range = 81-93%) and MRI (median = 88%, range = 79-94%). For the same assessment, CTA was also found to have the greatest sensitivity (median = 94%, range = 50-100%), followed by MRA (median = 91%, range = 69-100%) and ultrasound (median = 91%, range = 74-95.6%), and lastly MRI (median = 78%, range = 60-97%). For the quantification of Hill-Sachs lesion parameters, intraobserver reliabilities were reported for 3D CT (ICC range = 0.916-0.999), 2D CT (ICC range = 0.858-0.861), MRI (ICC range = 0.28-0.97), MRA (Kappa = 0.86; ICC = 0.87), and radiographs (Kappa range = 0.464-0.528). For the same quantification parameters, interobserver reliabilities were reported for 3D CT (Kappa range = 0.823-0.975; ICC range = 0.772-0.996), 2D CT (ICC range = 0.721-0.879), MRI (Kappa range = 0.444-0.700), MRA (ICC range = 0.830-0.904), and radiographs (Kappa range = 0.278-0.336). Intraobserver reliabilities for determining if a lesion was “on-” or “off-track” were only reported for 3D CT (Kappa range = 0.730-1.00; ICC range = 0.803-0.901) and MRI (ICC range = 0.770-0.790). This study demonstrates that there is verity of different imaging modalities that provides clinically acceptable accuracy in diagnosing Hill-Sachs Lesions, with CTA being the most reliable. Furthermore, this systematic review also solidifies that 3D CT has the greatest potential in quantifying and determining if a Hill-Sachs lesion will cause persistent anterior shoulder instability.