P08.26 * CEREBRAL DEEP WHITE MATTER - PROTECTED FROM BRAIN METASTASES?

Abstract

INTRODUCTION: Brain metastases occur in up to 25% of patients with cancer. If there is a non-random intracerebral distribution is under debate. From pathoanatomic studies well vascularised areas at the border of white and grey matter are thought to be locations of preference of cerebral metastases. Vascular or nutritional constraints might restrict the origin of metastases in deep cerebral white matter. In our study we aimed to analyse if and how often brain metastases arise in deep white matter. METHODS: Pre-therapeutic MRI scans of patients with brain metastases from two different centres were analysed. MRI scans were screened by two experienced independent investigators (CS: Neurooncologist, CG: Neuroradiologist) for metastases that appeared to originate from regions of deep white matter (corpus callosum, semioval center). Metastases extending from grey substance, superficial gyrus or ventricles/ventricular lining into white matter were not classified as deep white matter metastases. RESULTS: MRI scans from 198 consecutive patients examined between 2004-2014 in both centres were analysed. 85 patients suffered from non-small-cell lung cancer (NSCLC), 32 patients from small-cell lung cancer (SCLC), 20 patients from breast cancer (BC), 19 from malignant melanoma (MM), 17 from renal cell carcinoma (RC), 10 from colorectal cancer (CRC) and 15 from other tumors. 70% of scans were performed in 3 Tesla scanners with a 3D dataset and slice thickness of 1-1,5 mm. Overall, 29 of 198 patients (14,6%) showed white matter metastases (WMM). 76 patients showed 1-2 metastases, 79 patients showed 3-9 metastases (oligometastatic) and 43 patients showed > 9 metastases (polymetastatic). In the group with 1-2 metastases only 3/76 patients (4%) had WMM, in the oligometastatic group 14/79 (18%) and in the polymetastatic group 12/43 (28%), p = 0,001, chi square). According to primary tumor the rates of patients with WMM were: RC (6/17, 35%), MM (3/19, 16%), SCLC (5/32, 16%), NSCLC (11/85, 13%), BC (1/20, 5%), CRC (0/10, 0%). In direct comparisons (chi-square) the rate in patients with RC was significantly higher than that of NSCLC (p = 0,024), BC (p = 0,019) and CRC (p = 0,033). Between the other entities there were no significant differences. Discussion: Metastasis in deep white matter appears relatively rare in mono-oligo intracerebral metastasis but gets more propable with increasing number of cerebral metastases. Even in polymetastatic cerebral disease only about one quarter of patients show deep white matter metastasis. This might be due to vascular constraints in deep white matter. Interestingly, patients with renal cell carcinoma had more often white matter metastasis, whereas in breast- or colorectal cancer WMM did very rarely occur. If differences in involvement of deep cerebral white matter reflect different vasculogenic mechanisms can only be speculated from our study.