NEW YORK (GenomeWeb) – A new meta-analysis of pediatric gliomas suggests the childhood brain tumors fall into several distinct subgroups with varying prognoses and prevalence in younger or older individuals.
"We found that tumours that have historically been lumped together under one diagnosis are in fact comprised of many, remarkably different, diseases," senior author Chris Jones, a molecular pathology and cancer therapeutics researcher at the Institute of Cancer Research, said in a statement.
Jones led a team from the UK, the US, and Switzerland that analyzed data from more than 1,000 pediatric high-grade gliomas or diffuse intrinsic pontine gliomas (DIPG), using hotspot mutation data, methylation profiles, copy number data, genome sequences and/or exome sequences for published and unpublished cases.
The researchers' integrated molecular analyses of these data were described in Cancer Cell today, pointing to glioma clusters comprised of tumors with or without H3.1 or H3.3 histone mutations as well as subgroups characterized by IDH1 mutations, altered methylation profiles, low-grade glioma-like features, or mutations in genes such as EGFR, NF1, or MYCN. Certain subgroups had better outcomes than others, they explained, and molecular features in the gliomas tended to cluster with age of onset.
"Genomic aberrations increase with age, highlighting the infant population as biologically and clinically distinct," Jones and his coauthors wrote, noting that "[u]ncommon pathway dysregulation is seen in small subsets of tumors, further defining the molecular diversity of the disease, opening up avenues for biological study and providing a basis for functionally defined future treatment stratification."
For their analysis, the researchers considered data from 1,067 grade III or IV glioma or grade II to IV DIPG cases, representing children or young adults under 30 years old. The collection spanned 157 previously unpublished cases and 910 cases assessed for 20 studies reported in the past.
The team had access to new or existing hotspot mutation data for more than 900 of the tumors, along with copy number profiles for 834 tumors and array-based methylation profiles for 441 cases. The set also included 118 tumors assessed by whole-genome sequencing, 247 exome-sequenced tumors, and seven tumors tested by both exome and genome sequencing.
Along with analyses focused on tumor location, the researchers looked at molecular clusters coinciding with patient and outcomes. For example, they noted glioma subgroups marked by H3.1 histone mutations or low-grade glioma-like mutations were somewhat more common in young children with the brain cancer. Youngsters with tumors in the latter subgroup often had better outcomes, as did other subgroups lacking histone or IDH1 mutations.
The team's analysis suggested tumors with H3.3 histone mutations found in combination with mutations in TP53 or ATRX typically occur in the brain's cerebral hemispheres, while tumors with mutations in the H3.1 histone frequently formed in the pons portion of the brainstem.
"We need to start thinking about these as completely different cancers and diagnosing and treating them based on their genetic faults," Jones said. "It's exciting that several types look like they could be clearly treatable using either existing drugs on the market or other treatments under development."