Differentiation of brain metastases by percentagewise quantification of intratumoral-susceptibility-signals at 3 Tesla

Introduction: Evaluation of intratumoral-susceptibility-signals (ITSS) in susceptibility-weighted-imaging (SWI) has been reported to improve diagnostic performance for solitary enhancing brain lesions. Due to the distinct morphologic variability of ITSS, standardized evaluation proved to be difficult. We analyzed, if a new postprocessing method using percentagewise quantification (PQ) of ITSS enables differentiation between different entities of cerebral metastases and may thus improve differential diagnosis in cases of unknown primary. Materials and methods: SWI and contrast enhanced T1-weighted MR images were acquired from 84 patients with intracerebral metastases (20 patients with mamma carcinoma (MC), 15 patients with malignant melanoma (MM), 49 patients with bronchial carcinoma (BC)) at 3 Tesla MR. Images were co-registered and enhancing lesions were delineated on T1-weighted images and the outline transferred to the corresponding SWI map. All voxels within the lesion presenting values below a reference value placed in the ventricular system were determined and percentagewise calculated. Results: Diagnostic performance of percentagewise quantification (PQ) of ITSS, as determined with area under the receiver operating characteristic curve (AUC) was excellent (AUC = 0.96; 95% confidence interval (CI) 0.90, 1.00) to discriminate MM from MC, good for the discrimination of MM and BC (AUC = 0.81, 95% CI 0.70, 0.92) and poor for the discrimination of MC and BC (AUC = 0.60; 95% CI 0.47, 0.73). Conclusion: PQ is a new approach for the assessment of SWI that can be used for differential diagnosis of intracerebral metastases. Metastases of MM and MC or BC can be distinguished with high sensitivity and specificity.