Personalized Cancer Care

From MIT World

In the final of four symposia on pathbreaking cancer research, Tyler Jacks expresses “great optimism that we’re getting close, that we can see over the horizon...and we will be successful in controlling the disease in the not too distant future.” Personalized medicine will pave the way to this future, explains moderator Michael Yaffe. Scientists must gain an understanding “of what makes each tumor unique, and design a custom strategy to try to turn that knowledge into a cure,” he says. Yaffe’s panelists describe the great challenges involved in generating “the right drug, for the right time, in a dose that maximizes the chances of killing the tumor without killing the patient.”



“Tumor genomes are very complex,” says
Michael Hemann, but their makeup can now be modeled outside the patient, modified by removing or adding DNA, and even tested for responses to possible therapeutic compounds. Bioengineering using such tools as retroviruses proves especially helpful in exploring two proteins that appear to be involved in a large percentage of human cancers: ATM and p53. Researchers know that mutations in these correlate with different prognoses in cancer patients, and have developed drugs to help silence these proteins selectively. The problem is drug resistance. But Hemann believes it will be possible to “make a drug-resistant tumor drug sensitive,” by manipulating ATM and p53 in a manner specific to a patient’s own tumor profile. By identifying the molecular signature of many different tumors, through genome scale analysis of alterations in a large collection of genes, scientists can build a “toolset to predict patient outcome and optimize strategies for …overcoming drug resistance.”

Normal cells do a good job of repairing DNA damage, but in some types of cancers, such as those associated with the loss of BRCA1 or 2 genes, cells cannot fix breaks in DNA, explains David Livingston. These tumor suppressor genes fix double strand breaks in DNA, “which is for a cell like a medical emergency,” and if not corrected, leads to cell death. Livingston describes efforts to help address BRCA1 or 2 deficiencies, with targeted drug treatments that combine inhibiting the activities of DNA-damaging proteins inside the cell with chemotherapy to kill micrometastases. Livingston believes “the most urgent need of all is for some sort of analysis method, a robust, tractable biomarker,” that can report on the sensitivities of tumor cells to certain kinds of drugs, and that might help predict clinical outcome as well as personalize a patient’s therapy.

“What in the world do we do with all this information?” asks Daniel Haber, referring to an enormously complex “wiring diagram” of a cancer. Scientists already have in hand novel therapeutics that target different nodes in this picture, with some dramatic successes in certain types of lung cancer and melanomas. But “there has been a tremendous amount of serendipity to how some of these have evolved,” says Haber. He wants to achieve more systematic progress in matching the right drug with the right patient, preclinically; applying information “that’s wonderful in the lab, in the patient;” and then “serially monitoring cells during therapy.”

Haber argues for robotic screening of thousands of cancer cell lines to identify biomarkers for drug sensitivity, and genotyping cancers taken in biopsy “in real time for known and clinically significant mutations.” Finally, Haber is developing a microfluidic device to capture circulating tumor cells (“a huge challenge”) to determine how many cells move from the primary tumor into the blood, and whether they act as metastatic precursors. This would open a window onto cancers as they evolve, permit monitoring of drug resistance, and maybe even suggest opportunities for adjusting treatment as a tumor regrows.

Watch the video here.