Florence and Herbert Irving Clinical Research Career Award: Finding New Uses for Existing Therapies

February 11, 2020

Novel therapies are often considered the most innovative tools to improve care for cancers with poor prognoses. But a team of Columbia researchers is using a different tactic: they are exploring ways to repurpose drugs approved by the FDA for other conditions and existing experimental therapies in these difficult-to-treat cancers. The researchers are using mathematical modeling to identify how each drug would affect the most aberrantly active proteins, the master regulators, of individual patient’s tumors. Their hope is to deliver effective, individualized treatment plans for cancers with poor outcomes significantly faster than through novel drug development. 

Jovana Pavisic, MD, a prior fellow and now assistant professor in Pediatric Hematology, Oncology, and Stem Cell Transplantation, is a member of the team pursuing this approach. Dr. Pavisic has received a 2019 Florence and Herbert Irving Clinical Research Career Award, a highly competitive grant that enables recipients to dedicate their time more exclusively to research. 

Grounded in the research and computational algorithms created by her mentor Dr. Andrea Califano, Dr. Pavisic’s work uses an algorithm (VIPER) to examine the molecular makeup of individual tumors to identify the master regulators, which are not routinely identified on standard sequencing platforms. She then applies the OncoTreat algorithm, which uses data from large drug perturbation studies in cancer cell lines, to identify existing drugs that will shut down that cancer’s master regulator proteins. The goal is to identify standard chemotherapy and targeted drugs that would not typically be considered in a specific type of cancer, but that are likely to inhibit tumor progression in an individual patient’s tumor. 

Dr. Califano’s lab has pioneered this research in adult cancers, while Dr. Pavisic is bringing this research to pediatrics, a field that is uniquely challenging due to its smaller patient populations and limited data for many drugs in pediatric patients. In several difficult-to-treat pediatric cancers, there are fewer than several hundred cases per year spread across the United States, so Columbia is collaborating with institutions around the country to accrue enough patients and data.

Dr. Pavisic is currently studying osteosarcoma and diffuse intrinsic pontine glioma (DIPG), a type of brain tumor. Osteosarcoma, the most common malignant bone tumor in children, has a notoriously poor survival rate of 60 percent, which drops to less than 30 percent in patients with metastatic or relapsed disease. There have been no significant advances in treatment in over 30 years, and no second line drugs for patients who relapse. Current genomics approaches have had limited success in osteosarcoma, which is marked by catastrophic genetic events affecting the whole genome rather than specific gene mutations; this means that no two patients look alike at the DNA and RNA level. In this case, VIPER is still able to identify common targetable cancer drivers at the protein activity level. OncoTreat can predict drugs that affect a whole module of master regulators, and that can shut down the cancer-driving program in the tumor at a patient-specific level. In DIPG tumors no current therapies, apart from radiation, are effective. The computational pipeline in development by Drs. Pavisic and Califano may similarly identify much-needed new drugs to treat this universally fatal disease.