NEW YORK (GenomeWeb) – A new study suggests that single biopsies are unlikely to catch all of the key mutations that exist in clones from any given medulloblastoma tumor — and may lead to an outsized focus on targetable mutations found in just one or a few clones.
"The lack of clonal actionable driver mutations that are ubiquitously present across all regions of a given brain tumor suggests that monotherapies that target a single gene from a single biopsy are unlikely to have dramatic effects in terms of improving the lives of patients with brain tumors," co-corresponding authors Michael Taylor, with the Hospital for Sick Children and the University of Toronto, and the BC Cancer Agency's Marco Marra, and their colleagues wrote.
As they reported in a paper appearing online today in Nature Genetics, Taylor and Marra led an international team that did genomic and transcriptomic profiling on samples taken from multiple regions apiece in almost three-dozen individuals with medulloblastoma, high-grade glioma, or renal cell carcinoma.
Based on the extensive spatial heterogeneity they encountered when assessing somatic mutations in the tumors, the study's authors argued that future targeted therapy trials for medulloblastoma "should first ensure the spatially ubiquitous nature of the target mutation."
For the analysis, the team collected biopsy samples from between four and 11 regions apiece from nine primary medulloblastomas, 16 high-grade gliomas, and 10 renal cell carcinomas. Data for 10 of the high-grade gliomas was limited to array-based expression profiling. The remaining samples were subjected to expression, copy number, and somatic mutation profiling.
The researchers found that the medulloblastoma tumors did appear to have some intra-tumor transcriptomic consistency, making it possible to define expression-based medulloblastoma subtypes from lone tumor samples. But that was at odds with the pronounced heterogeneity they saw across multiple biopsy samples when considering somatic mutations, small insertions and deletions, and copy number changes.
While some potentially targetable mutations were present in single clones, the team also saw examples of tumors with targetable mutations shared across biopsies, and everything in between.
Based on their findings, the authors noted that several biopsies may be needed to see most actionable alterations in a given medulloblastoma tumor. Still, they recognized that it is often difficult to obtain many biopsies and proposed starting with two biopsies as a way of getting a sense of the general genetic heterogeneity of the tumor at hand.
"The extent of the spatial heterogeneity of somatic mutations observed in our cohort suggests that clinical trials of molecularly targeted therapy should first assess the ubiquitous distribution of the target," the authors wrote.