Integrated genomics elucidates relative spatial homogeneity of embryonal brain tumors

Marc Remke, Florence M.g. Cavalli, A. Sorana Morrissy, Vijay Ramaswamy, David Jones, Roger J. Packer, Éric Bouffet, Gary D. Bader, Arndt Borkhardt, Stefan M. Pfister, Nada Jabado, Marco A. Marra, Michael D. TaylorMolecular 2015 · DOI

Introduction

In the era of precision medicine, clinicians often rely on a single needle biopsy to sequence a tumor’s DNA and select targeted therapies. This approach assumes that the tumor is genetically uniform throughout. However, many aggressive cancers—such as renal cell carcinoma (RCC) and high-grade glioma (HGG)—are known to exhibit high intratumor heterogeneity (ITH). In these cases, a mutation found in one region of a tumor might be absent in another, meaning a single biopsy could miss the very drivers of the cancer's growth or drug resistance.

This study investigates whether embryonal brain tumors (EBTs), which are common in pediatric patients, suffer from this same "mosaic" problem or if they are spatially homogeneous, making them better candidates for biopsy-driven targeted treatment.

Methodology

The researchers employed a multiregion sampling strategy to quantify spatial variation. Instead of a single sample per patient, they analyzed a median of six spatially distant biopsies from each tumor (ranging from 4 to 11 samples per patient). The study utilized an integrated genomics approach to compare these regions:

  • Whole Exome Sequencing (WES): To identify somatic mutations in protein-coding regions.
  • High-resolution DNA Copy Number Analysis (Cytoscan HD): To detect large-scale chromosomal gains or losses.
  • Transcriptional Profiling: To analyze gene expression patterns (the transcriptome) across the tumor mass.

Datasets

The cohort consisted of 39 distinct pediatric and adult tumors, including:

  • Embryonal Brain Tumors: Medulloblastoma (n=9), AT/RT (n=2), and Medulloepithelioma (n=1).
  • Comparison Groups: High-grade glioma (HGG, n=17) and Renal cell carcinoma (RCC, n=10), which served as benchmarks for high heterogeneity.

Key Insights and Results

The study's quantitative findings revealed a stark contrast between EBTs and other aggressive malignancies:

  • Stability of Molecular Subgroups: In medulloblastoma, patients are categorized into subgroups (e.g., WNT, SHH) that dictate prognosis. These subgroup affiliations were found to be stable across all biopsies from the same patient, with no regional variation in molecular classification.
  • Ubiquity of Mutations: Somatic mutations in EBTs were significantly more likely to be "ubiquitous"—present in every biopsy taken from the tumor. This stands in contrast to HGG and RCC, where many mutations are "private" to specific regions.
  • Transcriptomic Homogeneity: Gene expression levels were highly consistent across different spatial areas in EBTs, whereas HGG and RCC showed highly disparate molecular profiles.

These results suggest that the foundational genetic drivers of EBTs occur early in the tumor's development and dominate the entire cell population, leading to a more uniform mass compared to the chaotic evolution seen in adult cancers.

Clinical Implications

The genomic landscape of EBTs appears remarkably stable. Consequently, the authors conclude that limited biopsies are indeed representative of the whole tumor. This has important implications for biological classification and the development of targeted therapies; clinicians can be more confident that a drug targeting a mutation found in a single biopsy will reach that same target throughout the entire tumor.

Limitations and Future Work

As a meeting abstract, this study presents high-level conclusions. While the findings are robust across the 39 tumors analyzed, some specific subtypes (like medulloepithelioma) were represented by only a single patient. Furthermore, the study focuses on spatial homogeneity at a single time point and does not address temporal heterogeneity—how a tumor might evolve following exposure to chemotherapy or radiation. Future research will likely explore how this spatial uniformity can be leveraged to streamline the development of more effective, targeted treatment protocols.

Summary updated

Abstract

Genome-wide profiling and next-generation based sequencing studies have dramatically improved our understanding of embryonal brain tumor (EBT) biology in the recent years. However, the vast majority of these studies are based on the assumption that single biopsies are representative for the entire primary tumor. Intratumor heterogeneity constitutes a common phenomenon previously described in renal cell carcinoma (RCC) and high-grade glioma (HGG). Highly disparate molecular profiles of spatially separated tumor areas within the same tumor may preclude development of molecularly targeted therapies based on single tumor biopsies.
BibTeX
@inproceedings{remke2015integrated,
  author = {Marc Remke and Florence M.G. Cavalli and A. Sorana Morrissy and Vijay Ramaswamy and David Jones and Roger J. Packer and Éric Bouffet and Gary D. Bader and Arndt Borkhardt and Stefan M. Pfister and Nada Jabado and Marco A. Marra and Michael D. Taylor},
  title = {{Integrated genomics elucidates relative spatial homogeneity of embryonal brain tumors}},
  booktitle = {Molecular},
  year = {2015},
  doi = {10.1186/2194-7791-2-s1-a10},
}
Figure