Abstract Diffuse gliomas are the most prevalent primary malignant brain tumors in adults and are characterized by invasive growth and poor prognosis. Tumor cells infiltrate the brain parenchyma and spread to distant sites. The aim of this study is to characterize the spatial heterogeneity, mode, and direction of tumor development to identify key areas for localized treatment. Neuronavigation was employed to collect n=134 image-guided samples from n=16 adult patients with diffuse gliomas, including n=7 IDH wild-type (IDHwt) and n=9 IDH mutant (IDHmut) tumors from regions with and without MRI abnormalities. Collected samples were subjected to targeted and shallow whole-genome sequencing. Somatic variants were used to determine the evolutionary processes governing heterogeneity. Examination of the dN/dS ratio, Tajima’s D score, and MOBSTER model suggested that all tumors are consistent with neutral evolution, as previously reported. Considering that genetic heterogeneity was a result of stochastic mutations, we used phylogeographic relationships between samples to dissect the spatiotemporal development of these tumors. Evolutionary (patristic) and cartesian (Euclidean) distances between sample pairs from the same patient were correlated in n=2/9 IDHmut and n=2/7 IDHwt patients, suggesting that migratory or punctuated evolutionary processes play a major role in tumor diffusion. An examination of spatial diffusion revealed that the latest descendant appeared peripherally in n=5/9 IDHmut patients, suggesting boundary-driven growth. In the remaining cases, three patients exhibited centrally-driven growth, whereas a fourth patient displayed a mixed pattern. In contrast, in n=4/7 IDHwt patients, descendants appeared closer to the tumor center, suggesting centrally-driven growth. In the remaining three IDHwt patients, one patient showed a diffusion that was consistent with boundary-driven growth, while the other two patients showed a mixed pattern. Taken together our data suggest that IDHwt tumors tend to develop centrally, whereas IDHmut tumors preferably develop outward. Our data could aid in tailoring localized treatment based on projected growth patterns.
Read full abstract