Parotid Warthin's tumor: novel MR imaging score as diagnostic indicator

Objectives: Despite known characteristic radiologic and clinical features, differentiation between Warthin's tumor (WT) and other parotid tumors remains challenging. The purpose of this study was to more precisely assess the MR imaging features of WT and to develop a scoring system combining the most specific characteristics. Methods: A total of 208 patients with parotid gland tumors and presurgical MRI were included. Tumors were divided into 5 histological subtypes, and different MRI features were compared between groups. An MRI scoring test was developed including MR parameters that contributed significantly in distinguishing WT from other tumors. Results: The best MRI features for differentiating between WTs from other tumors included bilaterality (P = 0.002), multifocality (P < 0.001), ADC values <905.1 (P < 0.001), and high signal intensity on T1-W images (P < 0.001). Six or more points on the 14-point scoring MRI scale was associated with an area under the curve of 0.99 (Accuracy of 98%), while a cut-off value of 7 indicated 100% specificity and 100% positive predictive value. Conclusions: Ill-defined margins, low T1-W signal, and location in the upper 2/3 of the parotid gland excluded WTs in 100% of cases. The proposed scoring method allows WTs to be distinguished from other tumors with high accuracy. Key points: • Despite known characteristic radiologic and clinical features, differentiation between Warthin's tumor (WT) and other parotid tumors remains challenging. • The best MRI features for differentiating between WTs from other tumors included bilaterality, multifocality, ADC values <905.1, and high signal intensity on T1-W images. • Six or more points on the 14-point scoring MRI scale was associated with an area under the curve of 0.99, while a cut-off value of 7 indicated 100% specificity and 100% positive predictive value.