Pediatric Neuro-Oncology: Time to Go Molecular

May 15, 2016

One recurring theme from genomic studies of pediatric CNS tumors (and almost all cancers, for that matter) is that tumors that historically appeared to be a single entity based on examination under the microscope and routine immunohistochemical staining actually harbor molecularly distinct subgroups when analyzed by genomic sequencing techniques.

The treatment of childhood cancers is regarded as one of the great successes in oncology, mainly due to the dramatic improvements in survival for children with leukemia observed over the past 50 years. However, the treatment of children with central nervous system (CNS) tumors has witnessed more modest improvements in survival, with some CNS tumors, such as diffuse intrinsic pontine glioma (DIPG), showing no discernable improvement in survival in the same time period. Nonetheless, the last few years have been very exciting for the pediatric neuro-oncology community, on account of novel insights gleaned from the applications of genomic sequencing technologies to pediatric CNS tumors. In this issue of ONCOLOGY, Dr. Parsons and colleagues review how genomic sequencing technologies have advanced our understanding of pediatric CNS tumors, and discuss the promise these new technologies hold for advancing the care of children with CNS tumors in the not so distant future.[1]

One recurring theme from genomic studies of pediatric CNS tumors (and almost all cancers, for that matter) is that tumors that historically appeared to be a single entity based on examination under the microscope and routine immunohistochemical staining actually harbor molecularly distinct subgroups when analyzed by genomic sequencing techniques. Furthermore, there are clear regional differences in genomic events that drive CNS oncogenesis. Notable examples include high-grade gliomas, in which mutations in H3 are region-specific: gliomas that arise in the midline of the CNS commonly harbor K27M H3 mutations, and those that arise in the cerebral cortex can harbor G34R/V H3 mutations but almost never K27M H3 mutations.[2-4] Regional differences also exist between supratentorial and posterior fossa ependymomas, as reviewed by Parsons et al.[1] Importantly, these genomic studies are guiding the World Health Organization in its efforts to improve the classification of CNS tumors-most notably with the designation, in 2016, of midline gliomas with K27M H3 mutations as a separate category, as well as the recognition of three molecular groups of medulloblastomas.[5] Classification of another type of CNS tumor, primitive neuroectodermal tumor (PNET), is undergoing some major revamping in 2016. First, it was realized that a subset of these tumors, when subjected to molecular analysis, actually comprises different tumors-eg, if they harbor G34R/V H3 mutations, they are actually gliomas with immunohistochemical features of PNET.[6] More recently, detailed genomic analysis has revealed that this tumor entity comprises several new molecular subgroups.[7]

While the concept of applying novel genomic sequencing technologies to the diagnosis and treatment of children’s CNS tumors is exciting, several caveats regarding their routine application are worth discussing. Perhaps the most obvious challenge is tumor heterogeneity, which is particularly relevant to high-grade gliomas,[8] although less so to other CNS tumors, such as medulloblastoma. If a biopsy samples only a small portion of a tumor, the genomic alterations present in the biopsy specimen may not include all of the important genetic events in the tumor. One way to get around this challenge is to sample several regions of the tumor, but this may not be a trivial issue for certain CNS tumors (such as DIPG, with its delicate location in the brainstem). In addition, clinical trials for children with CNS tumors that define tumors based on genomic analysis may result in smaller cohorts of patients with a particular genetically defined neoplastic entity-and thus may require more time to complete the specified accrual for determining whether a particular new therapy meets its predefined efficacy endpoints.

Perhaps the most complex question is which genomic sequencing applications are best to use in guiding clinical trials. This is a much-debated topic, which will likely take some time to sort out. Some candidate applications are DNA methylation platforms, which have been very successful in classifying gliomas and PNETs.[3,7] Another is RNA sequencing analysis, which has been instrumental in identifying novel fusion transcripts in CNS tumors such as supratentorial ependymomas.[9] Other applications, such as predefined gene panels of common oncogenes and tumor suppressors, are less experimental.[10] These latter types of analysis can be performed both in academic centers and by commercial entities such as FoundationOne or TGen. However, for these new diagnostic tools to be widely adapted, the economic aspects of their use also need to be addressed. Because higher costs are associated with genomic sequencing technologies (even though their price is rapidly dropping), insurance companies need to step up and reimburse patients for these tests. Finally, while the incorporation of genomic sequencing technologies into efforts to identify relevant molecular targets in childhood CNS tumors is critical in evaluating new therapeutics, it is also important to assess whether the therapeutic agents reach and inhibit their molecular targets. This can help with the interpretation of clinical trial results, particularly if they are negative.

In conclusion, the genomic era has brought renewed optimism for advancing the treatment of children with CNS tumors. There is already an ongoing clinical trial for children with DIPG that is using genomic sequencing technologies to identify relevant molecular targets (ClinicalTrials.gov identifier: NCT02274987), while another clinical trial is using molecular tools to subgroup children with newly diagnosed medulloblastoma (ClinicalTrials.gov identifier: NCT01878617). It is likely that progress in the treatment of these tumors will be accelerated by the use of genomic sequencing technologies to identify molecular targets and match them with therapies that target them. Perhaps more importantly, in addition to utilizing genomic analysis for molecular target identification, we should also incorporate a pharmacodynamics component to ensure that the novel therapies we are prescribing for our patients successfully inhibit their target(s).

Financial Disclosure:The author has no significant financial interest in or other relationship with the manufacturer of any product or provider of any service mentioned in this article.

References:

1. Bavle AA, Lin FY, Parsons DW. Applications of genomic sequencing in pediatric CNS tumors. Oncology (Williston Park). 2016;30:411-8, 422-3, 425.

2. Schwartzentruber J, Korshunov A, Liu XY, et al. Driver mutations in histone H3.3 and chromatin remodelling genes in paediatric glioblastoma. Nature. 2012;482:226-31.

3. Sturm D, Witt H, Hovestadt V, et al. Hotspot mutations in H3F3A and IDH1 define distinct epigenetic and biological subgroups of glioblastoma. Cancer Cell. 2012;22:425-37.

4. Wu G, Broniscer A, McEachron TA, et al. Somatic histone H3 alterations in pediatric diffuse intrinsic pontine gliomas and non-brainstem glioblastomas. Nat Genet. 2012;44:251-3.

5. Taylor MD, Northcott PA, Korshunov A, et al. Molecular subgroups of medulloblastoma: the current consensus. Acta Neuropathol. 2012;123:465-72.

6. Gessi M, Gielen GH, Hammes J, et al. H3.3 G34R mutations in pediatric primitive neuroectodermal tumors of central nervous system (CNS-PNET) and pediatric glioblastomas: possible diagnostic and therapeutic implications? J Neurooncol. 2013;112:67-72.

7. Sturm D, Orr BA, Toprak UH, et al. New brain tumor entities emerge from molecular classification of CNS-PNETs. Cell. 2016;164:1060-72.

8. Paugh BS, Broniscer A, Qu C, et al. Genome-wide analyses identify recurrent amplifications of receptor tyrosine kinases and cell-cycle regulatory genes in diffuse intrinsic pontine glioma. J Clin Oncol. 2011;29:3999-4006.

9. Parker M, Mohankumar KM, Punchihewa C, et al. C11orf95-RELA fusions drive oncogenic NF-kappaB signalling in ependymoma. Nature. 2014;506:451-5.

10. Nikiforova MN, Wald AI, Melan MA, et al. Targeted next-generation sequencing panel (GlioSeq) provides comprehensive genetic profiling of central nervous system tumors. Neuro Oncol. 2016;18:379-87.