Clinical and demographic data improve diagnostic accuracy of dynamic contrast-enhanced and diffusion-weighted MRI in differential diagnostics of parotid gland tumors.

2020 
Abstract Purpose To verify if differential diagnostic algorithms based on selected parameters of dynamic and diffusion-weighted MRI, clinical and demographic variables are accurate enough to differentiate between parotid malignancies, pleomorphic adenomas and Warthin tumors. Materials and methods The study included 84 patients, among them 19 (23%) with postoperative diagnosis of a parotid malignancy, 37 (44%) with Warthin tumors and 28 (33%) with pleomorphic adenomas. Accuracy of dynamic and diffusion-weighted MRI parameters (Tpeak, Tmax, WR and ADC), clinical parameters of the tumor (location in the superficial or deep parotid lobe, single vs. multiple lesion, concomitant lymphadenopathy) and demographic characteristics of the patient (age, sex) was determined on ROC analysis. Results Compared to Warthin tumors, parotid malignancies showed higher Tpeak, Tmax and ADC, lower WR, were more often located in the deep lobe of the parotid and presented as single lesions with concomitant lymphadenopathy. Consideration of all these variables provided 100% sensitivity and 89.2% specificity of the differential diagnostic algorithm. Compared to pleomorphic adenomas, parotid malignancies presented with lower ADC and Tpeak, higher Tmax, were more often located in the deep lobe of the parotid, associated with concomitant lymphadenopathy, diagnosed in men and individuals older than 66 years. Consideration of all these variables provided 100% sensitivity and 96.4% specificity of the differential diagnostic algorithm. Conclusion Combined analysis of clinical data and results of dynamic and diffusion-weighted MRI may provide nearly ideal accuracy in differential diagnostics of parotid malignancies and the two most common histological types of benign tumors.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    25
    References
    2
    Citations
    NaN
    KQI
    []