American Journal of Neuroradiology 2009;30:591.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
American Journal of Neuroradiology
DOI 10.3174/ajnr.A1412
HEAD & NECK
Diffusion-Weighted Echo-Planar MR Imaging of Primary Parotid Gland Tumors: Is a Prediction of Different Histologic Subtypes Possible?
From the Department of Diagnostic and Interventional Radiology (C.R.H., C.A., L.D., G.A.), Diagnostic Center; Department of Psychiatry and Psychotherapy (K.U.P.), Center of Psychosocial Medicine; and Department of Oto-, Rhino-, Laryngology (F.R., M.J.), Center for Head Care and Dermatology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany; Siemens Healthcare (J.G.), Hamburg, Germany; and Center of Oto-, Rhino-, Laryngology Regensburg (J.O.U.), Regensburg, Germany.
Please address correspondence to Christian R. Habermann, MD, University Medical Center, Diagnostic Center, Department of Diagnostic and Interventional Radiology, Martinistr. 52, 20246 Hamburg, Germany; e-mail: c.habermann{at}uke.uni-hamburg.de
BACKGROUND AND PURPOSE: Our aim was to determine the value of echo-planar diffusion-weighted MR imaging (epiDWI) in differentiating various types of primary parotid gland tumors.
MATERIALS AND METHODS: One hundred forty-nine consecutive patients with suspected tumors of the parotid gland were examined with an epiDWI sequence by using a 1.5T unit. Image analysis was performed by 2 radiologists independently, and the intraclass correlation coefficient was computed. Histologic diagnosis was obtained in every patient. For comparison of apparent diffusion coefficients (ADCs), a paired 2-tailed Student t test with a Bonferroni correction was used.
RESULTS: In 136 patients, a primary parotid gland tumor was confirmed by histology. Among the observers, a high correlation was calculated (0.98). ADC values of pleomorphic adenomas were significantly higher than those of all other entities, except for myoepithelial adenomas (P = .054). ADC values of Warthin tumors were different from those of myoepithelial adenomas, lipomas, and salivary duct carcinomas (P < .001, 0.013, and .037, respectively). Mucoepidermoid carcinomas, acinic cell carcinomas, and basal cell adenocarcinomas were not differentiable from Warthin tumors (P = .094, .396, and .604, respectively).
CONCLUSION: epiDWI has the potential to differentiate pleomorphic adenoma and myoepithelial adenomas from all other examined entities. Due to an overlap not only within the group of benign and malignant lesions but also between groups, diagnoses should not be addressed on the basis of ADC values solely. Therefore, further studies combining DWI, morphologic criteria, and probably other MR imaging techniques seem warranted.