NEW YORK (GenomeWeb) – New research suggests that diverse glioblastoma (GBM) tumors falling into several distinct methylation-based subgroups tend to share early driver mutations, which appear to have arisen long before individuals' initial GBM diagnoses and influence genetic features found in the tumors present at disease recurrence.
"In contrast to this common early path," the authors reported in Cancer Cell today, "relapsed tumors acquired no stereotypical pattern of mutations and typically regrew from oligoclonal origins, suggesting sparse selective pressure by therapeutic measures."
The team, led by investigators at the German Cancer Research Center (DKFZ), did whole-genome sequencing, transcriptome sequencing, targeted gene panel sequencing, and/or array-based methylation profiling on matched tumor-normal and primary tumor-relapse tumor sets from dozens of individuals with IDH-wild type GBM, a form of the disease known for its rapid growth and recurrence.
By integrating these molecular data in phylogenetic and tumor growth, mutation, and evolution modeling, the researchers tracked down apparent initiating mutations involving chromosome 7 gains or chromosome 9 and 10 losses that appeared an estimated two to seven years before the patients' GBM diagnoses, along with mutations affecting the TERT promoter that appeared to mark tumor transitions to a rapid growth phase.
The long, pre-diagnosis mutation buildup "is hardly conceivable in view of the extremely rapid growth of glioblastomas," first author Verena Körber, a graduate student in co-senior author Thomas Höfer's theoretical systems biology lab at DKFZ and Heidelberg University, said in a statement. Such patterns appear to reflect a model in which "many cancer cells die at the beginning," Körber said. "[O]nly at a certain moment does the rapid growth in size start."
For their analyses, the researchers did whole-genome sequencing on tumor-matched blood sample controls from 21 GBM patients, along with RNA sequencing on primary and recurrent tumor samples. They also considered pairs of primary and recurrent tumors from 43 patients with IDH-wild type GBM that were profiled with targeted sequencing on 50 glioma-related genes, and used Illumina BeadChip arrays to assess DNA methylation levels across tumor samples from both groups.
With these data, the team clustered tumors into known methylation-based GBM subgroups, got a look at infiltrating immune cells and other non-tumor cells present in the tumor samples, and retraced the events leading up to mutational signatures found in the primary and recurrent tumors.
Although the 21 tumors from the initial GBM cohort contained several mutational signatures and dozens of driver genes prone to mutation in the GBM tumors, for example, the primary tumors tended to show characteristic chromosome 9 or 10 losses or chromosome 7 gains. Likewise, the investigators identified recurrent TERT promoter mutations that appear to occur somewhat later, based on their presence in some but not all tumor clones, apparently providing a selective growth advantage to tumors.
The researchers' analyses of the recurrent tumors, meanwhile, revealed many of the same mutations detected in primary tumors. From those and other findings, they concluded that treatment-related selection may not lead to significant increases in the genetic diversity of recurrent tumors. Rather, the authors suggested that IDH-wild type GBM tumors "attained their full aggressiveness already before initial diagnosis."
"Our findings imply that standard therapy exerted little selective pressure on most recurrent tumors" they suggested, since "the vast majority of driver mutations were acquired prior to initial diagnosis and only few drivers were acquired after initial treatment."