Progress in Sequencing Human Cancer Genomes: Advancements in Childhood Medulloblastoma

Abstract

Improvements in technology and sequencing of the human genome have provided a template for polymerase chain reaction primers to amplify and sequence coding exons of large protein-coding genes. This has lead to genome-wide analyses of human cancers and sequencing of all protein-encoding genes in more than 80 human cancers. Medulloblastomas (MBs) are the most common malignant brain tumor in children and are diagnosed in approximately 1 in 200 000 children less than 15 years old each year.1 Prior research has identified 2 subgroups of MBs associated with either alterations in genes of the Hedgehog pathway and those mutated in the Wnt pathway.2,3 Alterations in TP53 (8), MYC, and OTX2 have also been uncovered.4,5 Although these alterations have helped define genetic subtypes of MDs, most patients do not have these mutations. In a recent study by Parsons et al, copy alterations were assessed using microarrays and all known protein-coding genes microRNA genes were sequenced.6 The authors found that each tumor had 11 gene alterations, 5 to 10 times fewer than that of other solid tumors sequenced to date. Mutations in known pathways of medulloblastoma were confirmed including Wnt and Hedgehog pathways. Additionally, a number of new mutations were uncovered including histone-lysine N-methyltransferase genes MLL2 or MLL3 in 16% of MB patients. These involved genes were found to be tumor suppressor genes, being responsible for chromatin remodeling and transcriptional regulation. In contrast to germline mutations present in the fertilized egg from which individuals develop, somatic mutations occur in normal cells as tissues progress during postnatal life. Somatic mutations in tumors either have no effect on tumor growth (passenger mutations) or confer an advantage allowing it to expand, invade local tissue, and/or metastasize (driver mutation). Passenger mutations provide a record of the number of divisions a cell has undergone during both normal development and tumor progression. Although most adult cancers such as breast, ovary, colorectal, and pancreas, and glioma7,8 have 1000 to 10 000 somatic mutations, some cancers particularly prone to environmental carcinogens such as melanoma and lung cancer may have more than 100 000.7,9-11 The low number of mutations in medulloblastomas, along with key mutations in developmental pathways, may be key differences differentiating adult from childhood cancers. An improved genetic classification for MB could be beneficial both in terms of defining prognosis as well as developing specific therapies. Currently, large cell/anaplastic MBs associated with MYC amplification have a worse prognosis than desmoplastic MBs that are often associated with PTCH1 or other Hedgehog pathway alterations.3,12,13 The classic MB subtype does not have a defining molecular alterations, and the MLL2/MLL3 mutations do not fit into a definable subtype.6 Furthermore, some cases show no mutations in cancer genes or only one alteration of any gene. Clearly, there are far more alterations to identify. As our understanding of malignant brain tumor biology has progressed, so too has the goal to develop more effective, less toxic, rationally conceived therapeutics. The standard of care for adjuvant therapies are relatively uniform, consisting of a combination of radiation and cytotoxic chemotherapy. As molecular profiling continues to evolve, however, the opportunities for more targeted, individualized intervention are rapidly increasing. Moving forward, under the International Cancer Genome Consortium, large-scale sequencing of each major cancer subtype is underway.14 With improvement in technology and declines in cost, it is likely that whole genome sequencing will be the methodology of the future rather than limiting exploration to sequencing the DNA of exons of protein-coding genes or analysis of transcriptomes. Complete sequencing will have a role on detection, diagnosis, treatment, and prevention, while further defining pathogenesis, progression, and mechanisms of drug resistance. Although each cancer type is likely defined by overlapping sets of mutated genes, a single genomic screening test may become an integral element of treatment planning of newly diagnosed patients and the design of clinical trials. To this end, initial stratification of brain tumors patients into molecularly determined treatment groups has already begun. The 2 best examples are promoter methylation and transcriptional silencing of O-6-methlylguanine-DNA methyltransferase (MGMT), which leads to improved GMB chemosensitivity, and eGFR vIII mutation, which may allow response to the eGFR inhibitor erlotinib. These findings represent a paradigm shift—one that stratifies patients by molecular determinants into treatment groups, thereby guiding personalized, biologically grounded therapies.Figure 1: Axial MRI with and without contrast depicting a classic medulloblastoma.Ricardo J. Komotar Robert M. Starke E. Sander Connolly Michael B. Sisti