Abstract
BACKGROUND AND PURPOSE: Echo-planar diffusion-weighted imaging (DWI) and delayed postcontrast T1-weighted MR imaging (DPI) have been proposed in previous studies to detect residual middle ear cholesteatomas, with varying results. We assessed and compared these 2 techniques in patients with canal wall-up tympanoplasty. MATERIALS AND METHODS: This was a prospective cohort study. Patients who underwent surgery for middle ear cholesteatoma had CT scanning 9 months after the surgery. If opacity was observed (64%) on CT scans, DWI and DPI were performed before second-look surgery. CT, MR imaging, and surgical data were available for 31 patients. Charts were reviewed independently by 3 blinded examiners. Interobserver agreement for MR imaging was calculated (Cohen kappa). Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated for these techniques: 1) alone or in association, and 2) according to the residual cholesteatoma size measured during surgery. RESULTS: Interobserver agreement was better for DWI (kappa = 0.81) than for DPI (kappa = 0.51). Sensitivity, specificity, PPV, and NPV values were 60%, 72.73%, 80%, and 50%, respectively, with DWI; and 90%, 54.55%, 78.26%, and 75%, respectively, with DPI. With cholesteatomas >5 mm, the sensitivity and specificity of DWI reached 100% and 88%, respectively, with values for DPI reaching 100% and 80%, respectively. The association of both techniques only allowed improvements in the specificity for lesions >5 mm. CONCLUSIONS: Both techniques gave acceptable results for residual cholesteatoma detection. DWI is more specific but less sensitive than DPI. Their concurrent use may benefit patients by avoiding undue surgery.