Molecular profiling of tumors is cancer treatment of the future

In May 2011, Marvin B. Brooks, M.D., HS ’68, a urologist in Palm Springs, Calif., developed severe back pain while performing surgery. X-rays revealed lung cancer that had spread to his liver and vertebrae. He had never smoked.

His oncologist told him that 10 percent of nonsmokers with lung cancer have a mutation in their epidermal growth factor receptor gene (EGFR). If Brooks fell within this category, he might be a candidate for an experimental therapy. The top expert in this field, added the oncologist, is Roy S. Herbst, M.D., Ph.D., professor of medicine (medical oncology) and of pharmacology, chief of medical oncology, and associate director for translational research at Yale. A week later Brooks was in New Haven, and in the same hospital in which he had previously operated and in which his son was born, he had a liver biopsy that went to Yale’s new Molecular Tumor Profiling Laboratory.

Confirming the oncologist’s supposition, the lab detected one of the EGFR mutations that seems to be a driver of many lung cancers. Instead of standard chemotherapy, Herbst recommended one daily pill of the experimental drug erlotinib (Tarceva), a reversible tyrosine-kinase inhibitor that targeted the EGFR mutation in Brooks’ tumor. In October 2011, in January 2012, and again in April 2012, Brooks tested free of cancer. Furthermore, Tarceva eliminated all detectable cancer without inflicting the debilitating side effects associated with chemotherapy.

“I went to Yale thinking I would be dead within a year,” said Brooks.

That’s the sort of story Yale Cancer Center Director Thomas J. Lynch, M.D. ’86, envisioned when he pushed for the profiling lab, which opened in the fall of 2010. “I think cancer treatment in the future will be based entirely on detailed molecular profiling,” said Lynch. “This is what everyone will be doing.”

Analysis from the Profiling Lab often identifies patients as candidates for new drugs that foil known cancer-causing mutations, thereby changing the course of therapy. “In oncology, these drugs have been shockers,” said lab director Jeffrey L. Sklar, M.D. ’77, Ph.D. ’77, professor of pathology and of laboratory medicine, “because they have had such a dramatic effect on tumors that have heretofore been highly resistant to therapy.”

“Right now, in about 20 percent of my patients with lung cancer,” said Herbst, “I find some mutation that will let me give a patient a specific oral drug that is much more effective and less toxic than the standard of care. We are working hard to figure out something new for the other 80 percent.”

The Profiling Lab also identifies mutations associated with other cancers. About half of all melanoma patients have a mutation in a gene called BRAF that responds to a new drug. Other mutations detected by the lab are associated with colon and breast cancer. The lab can currently screen for 66 mutations, each of which requires its own test, all run in a parallel and high-throughput fashion. However, the profiling lab is about to shift to the so-called “next generation” of genomic sequencing, which will detect all 66 mutations with one test. “Genomic sequencing will let us look at many genes all at once and expand indefinitely the number of mutations for which we can routinely screen,” said Sklar. “Very soon, we’re going to be profiling the full set of clinically relevant mutations within the tumor of every cancer patient who walks through the door of Smilow Cancer Hospital.”

The field is accelerating on all fronts. New clinical research and faster sequencing require new developments in software and hardware to process, analyze, and store more and more information. Efforts to develop and utilize these tools are under way at the Yale Center for Genome Analysis, located at West Campus. “Over the next 10 years,” said Richard P. Lifton, M.D., Ph.D., chair and Sterling Professor of Genetics, and a Howard Hughes Medical Institute investigator, “we’ll understand the genetic landscape of all the major human cancers, and we will increasingly be using molecular profiling to select treatment for individual patients. Five years ago we really did not imagine we would be capable of doing the things we’re doing routinely today.”