Researchers from The Cancer Genome Atlas (TCGA) Research Network have completed the largest, most diverse tumour genetic analysis ever conducted, revealing a new approach to classifying cancers. The study not only revamps traditional ideas of how cancers are diagnosed and treated, but could also have a profound impact on the future landscape of drug development.
Since 2006, much of the research has identified cancer as not a single disease, but many types and subtypes and has defined these disease types based on the tissue in which it originated. In this scenario, treatments were tailored to which tissue was affected, but questions have always existed because some treatments work, and fail for others, even when a single tissue type is tested. The extent to which genomic signatures are shared across tissues is still unclear.
In their work published in the journal Cell, the TCGA researchers analysed 3,527 specimens from 12 different cancer types to see how they compared to one another, the largest data set of tumour genomics ever assembled. They found that cancers are more likely to be genetically similar based on the type of cell in which the cancer originated, compared to the type of tissue in which it originated.
The TCGA researchers performed an integrative analysis using five genome-wide platforms and one proteomic platform. The analysis revealed a unified classification into 11 major tumour subtypes. Five subtypes were nearly identical to their tissue-of-origin counterparts, but several distinct cancer types were found to converge into common subtypes.
Lung squamous, head and neck, and a subset of bladder cancers coalesced into one subtype typified by p53 alterations, p63 amplifications, and high expression of immune and proliferation pathway genes.
However, bladder cancers split into three pan-cancer subtypes, one virtually indistinguishable from lung adenocarcinomas, and another most similar to squamous-cell cancers of the head and neck and of the lungs. The findings may help explain why patients with bladder cancer often respond very differently when treated with the same systemic therapy for their seemingly identical cancer type.
Another striking example of the genetic differences within a single tissue type is breast cancer. The breast, a highly complex organ with multiple types of cells, gives rise to multiple types of breast cancer. In this analysis, the basal-like breast cancers at molecular level have more in common with ovarian cancer and cancers of a squamous-cell type origin, a type of cell that composes the lower-layer of a tissue, rather than other cancers that arise in the breast. Commonly referred to as "triple-negative," basal-like cancers are particularly aggressive and are more prevalent among African-American women and younger women.
The multiplatform classification, while correlated with tissue-of-origin, provides independent information for predicting clinical outcomes. Although follow-up studies are needed to validate and refine this newly proposed cancer classification system, it will ultimately provide the biologic foundation for an era of personalised cancer treatment that patients and clinicians eagerly await.
All data sets are available for data-mining from a unified resource to support further biological discoveries and insights into novel therapeutic strategies. The data sets and results have been made available to other researchers through the Synapse website. The bioinformatics company Sage Bionetworks created Synapse as a data repository for the Pan-Cancer Initiative.
The TCGA was launched in 2006 with the goal of compiling genomic atlases of more than 20 types of cancer. As the project proceeded, however, commonalities across cancer types began to emerge, which led to the launch of the TCGA Pan-Cancer Initiative in October 2012.
The TCGA is jointly funded and managed by the USA National Cancer Institute and the National Human Genome Research Institute, both part of the National Institutes of Health. TCGA-generated data are freely available in advance of publication at the TCGA Data Portal.
Hoadley KA, Yau C, Wolf DM, et al. Multiplatform Analysis of 12 Cancer Types Reveals Molecular Classification within and across Tissues of Origin. Cell 2014 Aug 6. pii: S0092-8674(14)00876-9. doi: 10.1016/j.cell.2014.06.049. [Epub ahead of print]