Accuracy of 2D BLADE Turbo Gradient- and Spin-Echo Diffusion Weighted Imaging for the Diagnosis of Primary Middle Ear Cholesteatoma

Objective To evaluate the diagnostic accuracy of 2D BLADE turbo gradient- and spin-echo diffusion weighted imaging (TGSE BLADE DWI) for primary middle ear cholesteatoma diagnosis, using qualitative and quantitative methods. Study Design Retrospective case series. Setting University hospital. Patients Participants included those with suspected primary middle ear cholesteatoma after assessment by clinical otorhinolaryngologists combined with magnetic resonance imaging (MRI) examination. Finally, of the 85 ears from 65 patients enrolled in the study, 73 had cholesteatoma, and 12 had otitis media. Intervention Two radiologists independently assessed images and measured apparent diffusion coefficient (ADC) values. Sensitivity, specificity and accuracy were evaluated. Kappa (k) statistics, the intraclass correlation coefficient (ICC), the Kolmogorov-Smirnov normality test, the independent t test, and receiver operating characteristic (ROC) analysis were used for statistical analysis. Pair-wise comparison of the area under the ROC curve (AUC) was also performed using the Delong test. Main Outcome Measures Imaging and histopathologic findings. Results The mean ADC value of cholesteatoma group (mean, 0.923 +/- 0.246 x 10(-3) mm(2)/s) was significantly lower than that of noncholesteatoma group (mean, 1.744 +/- 0.205 x 10(-3) mm(2)/s; p < 0.001). In <= 3 mm cholesteatoma group, the AUC of qualitative DWI was 0.846; the sensitivity, specificity, and accuracy for diagnosing cholesteatoma were 69.23%, 100%, and 84%, respectively; while the AUC of quantitative diagnosis was significantly increased to 1.0 (p = 0.0209); and based on the optimal threshold of ADC, <= 1.352 x 10(-3) mm(2)/s, the sensitivity, specificity and accuracy improved to 100%. For >3 mm cholesteatoma group, there were no significant differences in diagnostic performance. Excellent interobserver agreement and ICC for the qualitative and quantitative evaluations (k = 0.90 and ICC = 0.80, respectively) was noted between reviewers. Conclusion TGSE BLADE DWI is useful for the detection of primary middle ear cholesteatomas, especially <= 3 mm lesions.