Cancer genomes are now rapidly yielding to analysis, showing the heterogeneity of the mutational events and processes contributing to disease. This diversity is enormous and occurs at many levels. Consequently, the first clues from exome sequencing will increasingly need to be complemented by integrated analyses and extended to implicate mutations relevant to treatment outcomes.
It is impossible to appreciate achievements in the field of cancer genomics without being awed by the scale of the problems we have yet to solve. Each new vista brings the opportunity to divide, define and intercept these diseases, subtype by subtype, provided we can identify the altered pathways driving cancer cell proliferation and develop or match to each a targeted intervention. More disease subtypes are coming to our attention, with even the subtypes within an individual diverging into the main neoplasm and subclones. Cancer presents a moving target, a proliferating, varying system that, given enough time and opportunity, can evolve resistance to each intervention.
In this issue, we publish and highlight some of the achievements in this field (on pages 229, 233, 242, 253, 279, 285 and 290; see also Nat. Med., published online 3 February 2013; doi:10.1038/nm.3078). Each study reports some of the notable traits of cancer genomes and a significant part of what we need to know to understand the diversity, mechanisms, natural history and possible vulnerabilities of each type of cancer cell. For example, in some cancer cell types, we can see the relationship of recurrent, dominant oncogenic driver mutations to disease subtypes. In others, we can discern different cells of origin, recurrent mutation among components of common pathways, expression consequences from mutation, differential effects on disease outcome and indicators that might signal a change in therapeutic modality.
An optimist will point out that current exome-based cancer genomics is accretive and a work in progress in which—even at low numbers—recurrent mutations are the first finds from a systematic genomic overview of each cancer type in its turn. Gain-of-function mutations in oncogenes are particularly valuable finds, as they present cancer-specific candidate targets for therapy. Collateral findings may provide clues to underlying mutational processes and environmental initiators and promoters. In counterpoint, the genomic pessimist will point out that cancer-specific alterations occur at a number of regulatory levels that can influence gene function and expression. Sequence alterations in dominant drivers represent a subset of the mutational processes that contribute to these regulatory changes, so it is not surprising that recurrent driver mutations found by whole-exome or even whole-genome sequencing sometimes constitute a small contribution to the genetic component of oncogenesis. In such cases, rather than repeat exome sequencing of ever greater numbers of tumor-normal pairs, it may be more productive to take the evidence at face value that the point mutations found will not add up to a recurrently mutated pathway and to use other methods to identify the aspects of cancer genomes that can be applied to disease classification and outcome prediction and to the identification of mechanisms that can be therapeutically intercepted. Where recurrent drivers are identified, their usefulness as drug targets should actually be demonstrated, so that the pharmaceutical companies can be persuaded to use some of their existing new targeted agents in clinical trials.
Although the progress of cancer genome projects is propulsive, there are things that we can do as a journal to reassure authors that there is time and a will to do the research right for the motives for which it was originally conceived. We want complete and influential studies to be published whole wherever possible. To defray fears of competitive publications undermining the findings of publications in this field, it seems prudent to increase the publishable unit from a single novel recurrently mutated gene in a single tumor type to a significant and actionable conceptual advance supported by several lines of experimental evidence. Coordinated publication is another advantage of Nature and the Nature research journals. Our normal practice, of course, is complete confidentiality and independence, adhering to Nature's standards of excellence and fierce competition for the best papers. As many of the groups working in this field are already coordinating their efforts in large consortia, we offer to coordinate publications across the Nature family of journals and, where possible, even to work with competing journals to expedite and coordinate the simultaneous publication of related papers without slowing the pace of peer review and publication. It is important that we learn to intercept cancer. Fear of being scooped should never be an excuse to think small.