Is molecular diagnostics ready for its cancer close‐up?

In January, the 2 leading manufacturers of high-end DNA sequencing machines boasted that their latest models could sequence an entire human genome in a single day. The following month, the Cancer Gene Census maintained by the Wellcome Trust Sanger Institute in Cambridge, United Kingdom, provided another stunning number. The running tally of human genes linked to cancer had reached 474, implicating roughly 2% of our genes. “We hold the seeds of our disaster within us,” said Michael Stratton, PhD, director of the institute, at a symposium on cancer genomics held during the American Association for the Advancement of Science (AAAS)'s annual meeting in Vancouver, British Columbia, February 16 to 20, 2012. Dr. Stratton's talk underscored the diverse ways those seeds can sprout into lethal cancers. The convergence of rapid-fire technology and large-scale research efforts, however, is beginning to map out the routes needed to identify each threat and eliminate or keep it controlled. Eventually, the accruing data might help lead cancer researchers toward the goal of more personalized therapeutic interventions. However, teasing out the relative contributions of mutations in hundreds of genes may first require a fundamental shift in how pathologists and other clinicians test and classify tumors. At the AAAS symposium, Samuel Aparicio, MA, PhD, the Nan and Lorraine Robertson Chair of Breast Cancer Research at the University of British Columbia/BC Cancer Agency in Canada, summarized some of his group's molecular research that suggests breast cancer could be subdivided into 10 major groups. This finding, he said, suggests that what we know as one entity might be fragmented into many diseases. The human epidermal growth factor receptor 2-linked subtype, accounting for approximately 15% of cases, has been successfully targeted by trastuzumab therapy, although Dr. Aparicio said the potential new classification scheme raises questions concerning whether patients diagnosed with other subtypes will have sufficient therapeutic options. Therapy may well lag behind diagnostics, but Jennifer L. Hunt, MD, MEd, chair of pathology and laboratory services at the University of Arkansas for Medical Sciences in Little Rock, said in a separate interview that researchers cannot figure out the best treatments until they can work out the subdivisions. More broadly, researchers are beginning to accept the idea that cancer is no longer a body site-specific occurrence but rather a mutation-specific event. Elaine Mardis, PhD, codirector of the Genome Institute at the Washington University School of Medicine in St. Louis, Missouri, said that, “in the future, instead of a patient saying, ‘I have breast cancer,’ they might say something like, ‘I have PIK3CA-mutated, P10-deleted breast cancer of the luminal B subtype.’ ” A larger burst of research based on whole-genome sequencing will likely accelerate the trend. Last June, researchers with The Cancer Genome Atlas published a comprehensive molecular portrait of serous ovarian adenocarcinoma tumors, which are responsible for 85% of all deaths from ovarian cancer.1 After analyzing the molecular traits of nearly 500 tumors and sequencing every protein-encoding gene in 316 of the specimens, the team found that more than 96% of tumors contained a mutation in TP53, a gene that encodes a tumor suppressor protein. Roughly 22% of the tumors also harbored mutations in the breast cancer-linked BRCA1 and BRCA2 genes. Combined, the data suggested that clinicians could divide ovarian tumors into multiple subtypes, and that survival predictions might be improved significantly by considering the combined activity patterns of large gene sets. Similar profiles for another 20 tumor types are being developed, funded largely through the American Recovery and Reinvestment Act of 2009. As the information accumulates, researchers hope to go back to the underlying biology and understand the disease mechanisms. “In the process of figuring out the ‘how to’ for all this, it not only empowers the pathology exercise, but it also is really informative [for] basic scientists, who want to understand what drives cancer biology,” Dr. Mardis said. Taking a more holistic view can lead to unexpected connections. The drug imatinib has proven adept at targeting the oncogenic BCR-ABL gene fusion in patients with chronic myelogenous leukemia. Established research suggests that the drug is also effective for certain gastrointestinal stromal tumors, thanks to a shared molecular pathway.2 Diagnostic tests are now running the gamut from a narrow focus on mutational hotspots to more unbiased sequencing of a few dozen genes. As sequencing times and costs continue to plummet, sequencing the whole genome of patients' tumor samples could become standard, requiring pathologists to adapt. Kevin Roth, MD, PhD, chair of pathology at the University of Alabama at Birmingham, said the pathology field is already reinventing itself by placing a far greater emphasis on molecular diagnostics and informatics. The ability of cytopathology to isolate cancer cells or even DNA from circulating blood and other fluids (an advantage over tissue-based surgical pathology) has further altered the dynamic. Cytopathologists are used to doing more with less, Dr. Roth said, and those with molecular training will be key not only for diagnostics but also for monitoring therapies and perhaps even helping to rapidly modify treatments that are not working. Pathologists, he said, also may be called on to act as gatekeepers in helping to decide which laboratory-developed test “is scientifically and medically justifiable and what is simply a fishing expedition,” positioning them more centrally in patient care discussions. “It would be naive to think that money is limitless and we can institute any type of testing we want regardless of financial consequences,” Dr. Roth said. Conversely, he said, “if we wait for all of these tests to pay for themselves, patients will suffer.” No one can yet predict the overall impact of cancer genomics and molecular diagnostics on medical costs, but Dr. Hunt said a new calculus is needed. “We don't think about the cost of care when we develop an assay; we think of the cost of the assay,” she said. “These are expensive, but if you think about the cost of care, you could be decreasing the cost dramatically if you tailor therapy based upon the diagnostics, and that's what we don't do very well.” Improved diagnostics might similarly aid clinical trials by excluding those patients who are unlikely to respond because their tumor-associated mutations are not targeted by the experimental drug, thereby saving time and money while preventing needless side effects. “It's a tremendously exciting time for pathologists. I think we have to grab a hold of it,” Dr. Hunt said. Seizing the moment does not mean letting go of other assays, Dr. Hunt and other experts warn. If a melanoma tumor includes a mutation in the BRAF gene, for example, clinicians now know that prescribing the anti-BRAF drug vemurafenib often delivers a temporary reprieve. However, in patients with colorectal cancer who carry the identical BRAF mutation, vemurafenib has no effect at all, according to Dr. Stratton. Clinicians who do not account for the cancer's histological type as well as its molecular profile, he said, would be making a preventable mistake in prediction. The accounting will only get more complicated, with genomic DNA, physical DNA modifications, RNA, and protein levels all providing additional layers of information to morphological and immunohistochemical tests. However, a well-integrated view may offer the best chance to recognize and neutralize a lot of “bad seeds.” CytoSource Reader Poll # The Rise of Molecular Diagnostics Q: Within the next decade, molecular diagnostics based on a tumor's genomic DNA will: A. Dominate the field of cancer pathology. B. Be important, but not essential, to cancer pathology. C. Constitute only 1 part of a balanced cancer pathology toolkit. D. Be supplanted by diagnostics based on other molecular traits Take the poll online at www.cancercytojournal.com. The results will be published in the August 25, 2012 issue. DECEMBER POLL RESULTS Q: How have the new ACGME rules, which went into effect in July and limit the work hours of medical residents and prevent first year residents from taking call, affected pathologists at your academic institution? 33% Not applicable (I am not at an academic institution). 11% We have not been affected at all. 22% We are happy to adopt these new rules and easily made adjustments. 33% It has been a real challenge for our department to meet these new rules. Content in this section does not reflect any official policy or medical opinion of the American Cancer Society or of the publisher unless otherwise noted. © American Cancer Society, 2012.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.