Comparative Study of Multiple High-Signal-Intensity Spots on 3D and 2D Magnetic Resonance Sialography for Patients with Sjögren’s Syndrome

We investigated the sensitivities of 2-dimensional (2D) magnetic resonance sialography (MR-S) and unilateral sagittal and axial 3-dimensional (3D) MR-S using a surface coil and their combination in diagnosing patients with Sjögren’s syndrome (SS). We retrospectively analyzed the 3D and 2D MR-S results of 78 patients with SS. We evaluated the sensitivities of multiple high-signal-intensity spots and staging on MR sialograms and analyzed the efficient imaging methods and cross section for diagnosing patients with SS. The sensitivities of MR-S for detecting abnormal findings (i.e., MR-S stage 1 or higher) were as follows: 65 cases (83.3% [95% confidence interval (CI): 73.2–90.8]) for unilateral sagittal 3D MR-S; 62 cases (79.4% [95% CI: 68.8–87.8]) for axial 3D MR-S; 66 cases (84.6% [95% CI: 74.7–91.8]) for combined unilateral sagittal and axial 3D MR-S; and 32 cases (41.0% [95% CI: 30.0–52.7]) for bilateral sagittal 2D MR-S. The ratio of the abnormal finding of MR-S was tested using the two-tailed Fisher’s exact test. Unilateral sagittal, axial, and combined unilateral sagittal and axial 3D MR-S showed significantly higher sensitivity than bilateral sagittal 2D MR-S, respectively ( P < 0.001). Most cases upstaged by 3D MR-S were those positive (stage 1 or higher) among the stage 0 cases detected by 2D MR-S. Axial 3D MR-S, compared with 2D MR-S, understaged four cases, which was due to the imaging range of the axial 3D MR-S. We concluded that a single unilateral sagittal 3D MR-S was sufficient and axial 3D MR-S was unnecessary for SS staging. T1- and T2-weighted images are essential for investigating the salivary glands in patients with SS. Therefore, we also concluded that bilateral sagittal 3D MR-S of the parotid glands in addition to T1- and T2-weighted imaging is necessary, sufficient, and most efficient for precise MR imaging examination of the salivary glands, including diagnosing SS.