RESUMO
OBJECTIVE: The aim of our study was to evaluate the importance of a non-injected T1-weighted spin-echo sequence (T1WSE) combined with a non-echo-planar diffusion-weighted (non-EPDW) sequence for the pre-operative detection of cholesteatoma by the radiologist on MRI, compared to surgery. MATERIALS AND METHODS: In this retrospective case review, 113 patients with chronic otitis underwent surgery (gold standard) for a clinical suspicion of cholesteatoma. Our primary outcome was to compare non-EPDW images + a contrast-free T1WSE sequence for cholesteatoma detection. Our secondary outcome was to quantify the signal intensity value of the suspected lesions, relative to the signal intensity of the cerebellum (Sic) to calculate Signal Intensity Ratios (SIR = SI/Sic). The SIR values of cholesteatomatous and non-cholesteatomatous tissue were compared to surgical findings. Receiver-operating characteristic curve analysis determined an optimum SIR cut-off value for the prediction of cholesteatoma. RESULTS: The sensitivity (96.9%) of non-EPDW for the diagnosis of cholesteatoma was high, with good specificity (74.2%), and increased to 85.5% when combined to a T1WSE sequence. Additionally, the mean SIR values (on T1WSE) of cholesteatoma were significantly lower than non-cholesteatomatous tissue (p < 0.05). When nonEPDW and T1WSE were combined, a cut-off SIR value < 1.04, diagnosed cholesteatoma pre-operatively with very high specificity and sensibility (92.7% and 90.3% respectively). CONCLUSION: Our study showed that combining a nonEPDW sequence with the quantitative analysis of contrast-free T1W SE sequence in pre-operative patients enables the correct diagnosis of cholesteatoma with good sensitivity and specificity (> 90%) and reduces risks of false-positive cases for surgeons.