The Personal Cancer Genome and Precision Cancer Trials
By SAMEEK ROYCHOWDHURY, MD, PhD,
assistant professor of Internal Medicine and of Pharmacology
Clinical trials traditionally have been designed to treat patients with a specific disease and to treat all patients on the trial the same way. For example, we would evaluate a new drug for breast cancer by giving the agent to a group of breast-cancer patients, and if enough of them benefited, we would judge the drug a good one. If few patients benefited, the drug would likely be abandoned.
During such disease-based trials, a small number of participants, maybe 1-10 percent, can benefit amazingly from the drug, but further development of the agent is unlikely because clinical trials as usually done are impractical for small numbers of patients.
Genomics is now helping us understand why some patients respond differently to therapy; although all the women in our hypothetical trial have cancer of the breast, not all breast cancers are the same.
Traditionally, we diagnose breast cancer, leukemia or prostate cancer by examining the way cells appear under a microscope. This tissue-of-origin approach has provided a histology-based or “microscopic classification” of cancer that has been used for decades.
Histologically, malignancies such as breast and prostate cancer appear to be homogenous. But cancer genomic studies, such as The Cancer Genome Atlas, demonstrate that these cancers are heterogeneous, with sets of mutations and other genetic changes that allow their grouping into subtypes (breast cancer may have more than 25). Each subtype may be a different disease that requires different therapy based on the tumor’s genomics. These molecular subtypes are driving the development of targeted drugs and influencing how we do clinical trials.
For example, a genetic pathway called PI3 kinase (PI3K) is genomically altered or mutated in up to 25 percent of breast cancers, and we have disease- and mutation-based trials such as OSU-12127 evaluating a PI3K inhibitor for breast-cancer patients with a mutation in a PI3K gene. In this trial, a smart drug specifically inhibits the PI3K kinase gene pathway. In contrast, in other cancers, genomic targets are less common, amounting to just 1-10 percent of cases, which effectively turns most cancers into multiple “orphan diseases.”
But how do we complete clinical trials for a gene mutated in only 1 percent of breast cancer or across many different cancers? If a trial needs 100 patients, 10,000 patients must be screened. That’s very challenging; traditional clinical trials are not feasible. Such orphan diseases will require a personalized therapeutic approach, and, consequently, clinical-trial designs are being developed to accommodate this new reality.
Some cancer centers, including the OSUCCC – James, are developing a new type of trial design that “baskets” together different diseases that share a molecular target in one trial (see figure).
In some instances, if we know enough about the molecular subsets of a disease, we may be able to enroll patients into trials that have a drug that matches the molecular makeup of their cancer. To facilitate these trials, cancer centers must be capable of providing a personalized molecular view of an individual’s cancer.
Gene sequencing technology called next-generation sequencing will enable oncologists to determine which of 200-plus significant genes are altered in a patient’s cancer and to use this information to guide therapy. Ohio State is among the leaders in the country in promoting and championing this precision oncology strategy.
In summary, to move forward into the age of precision oncology we need to accomplish four objectives:
1. Complete the molecular characterization of cancer;
2. Develop drugs that target the molecular defects that drive cancer;
3. Implement next-generation sequencing to molecularly characterize the tumors of patients in the clinic;
4. Launch innovative clinical trials that utilize these molecular portraits.