RESUMEN
BACKGROUND: High-grade pineal region tumors are rare and heterogeneous types of primary central nervous system neoplasms; radiological differential diagnosis is challenging but it is important because it has a therapeutic relevance. PURPOSE: To discriminate among high-grade pineal region tumors by combining apparent diffusion coefficient (ADC) volumetric values and qualitative features in order to predict their histology. MATERIAL AND METHODS: Twenty-two patients with high-grade pineal region tumors were assessed by qualitative and quantitative analysis. Margins, T2-weighted signal intensity, contrast enhancement, hemorrhage, calcifications, different volumetric ADC fractions (ADCmean, ADCmax, ADCmin) were evaluated and were compared to the histopathologic findings (cell count and proliferation index). RESULTS: Our qualitative imaging data showed that only margins were different among different tumors and each tumor type showed peculiar age onset. ADCmean was found the best quantitative value to discriminate high-grade tumors of the pineal region. ADCmean correlated with proliferation index but not with cell count. ADCmean values were lower in tumors with higher proliferation rate and a significant difference in ADCmean values were found between germinomas and pineoblastomas, between germinomas and papillary tumors and between papillary tumors and pineoblastomas. Moreover, the cut-off value of 0.865 × 10-3 mm2/s for ADCmean (ADC mean threshold value) could differentiate germinoma from pineoblastomas with the best combination of sensitivity and specificity. CONCLUSION: The ADCmean value measured on the whole tumor, reflecting tumor proliferative activity, may be a practical and non-invasive marker for predicting tumor histology in high-grade pineal region lesions and might be useful in preoperative assessment.
