Comparison between 3D isotropic and 2D conventional MR arthrography for diagnosing rotator cuff tear and labral lesions: A meta-analysis

Background: Although 3D-isotropic MR arthrography has been characterized as a substitute imaging tool for rotator cuff tear (RCT) and labral lesions, it has not been commonly used in clinical practice because of controversy related to image blurring and indistinctness of structural edges. Purpose: To perform a comparison of the diagnostic performance of 3D-isotropic MR arthrography and 2D-conventional MR arthrography for diagnosis of RCT (solely RCT, full/partial-thickness supraspinatus [SST]-infraspinatus [IST] tear, or subscapularis [SSc] tear) and labral lesions. Study Type: Meta-analysis. Population: Patients with shoulder pain. Field Strength/Sequence: 3D-isotropic and 2D-conventional MR arthrography at 3.0T or 1.5T. Assessment: PubMed and EMBASE were searched following the PRISMA guidelines. Statistical Tests: Bivariate modeling and hierarchical summary receiver operating characteristic modeling were performed to compare the overall diagnostic performance of 3D-isotropic and 2D-conventional MR arthrography. Multiple-subgroup analyses were performed for diagnosing RCT, full/partial-thickness SST-IST tear, SSc tear, and labral lesions. Meta-regression analyses were performed according to subject, study, and MR arthrography characteristics including 3D-isotropic sequences (turbo spine echo [TSE] vs. gradient echo [GRE]). Results: Eleven studies (825 patients) were included. Overall, 3D-isotropic MR arthrography had similar pooled sensitivity (0.90 [95% CI, 0.87–0.93]) (P = 0.95) and specificity (0.92 [95% CI, 0.87–0.95]) (P = 0.99), relative to 2D-conventional MR arthrography (sensitivity, 0.91 [95% CI, 0.86–0.94]); specificity, 0.92 [95% CI, 0.87–0.95]). Multiple-subgroup analyses showed that sensitivities (P = 0.13–0.91) and specificities (P = 0.26–0.99) on 3D-isotropic MR arthrography for diagnosing RCT, full/partial-thickness SST-IST tear, SSC tear, and labral lesions were not significantly different from 2D-conventional MR arthrography. On meta-regression analysis, 3D-TSE sequence demonstrated higher sensitivity (P < 0.01) than 3D-GRE for RCT and labral lesions. Data Conclusion: 3D-isotropic MR arthrography can replace 2D-conventional MR arthrography. 3D-isotropic MR arthrography using TSE sequence is recommended for more accurate diagnosis, as it demonstrates increased sensitivity. Level of Evidence: 2. Technical Efficacy: Stage 3. J. Magn. Reson. Imaging 2018;48:1034–1045.