Predictive Biomarkers For Chemotherapy Through Master Regulator Analysis

The two primary chemotherapy regimens to treat pancreatic cancer are called mFOLFIRINOX and gemcitabine/nab-paclitaxel. Even though both are considered appropriate therapies to give to patients in the first line, there is no universal test to predict which is more effective for a specific patient. This leads to worse outcomes for patients and unnecessary side effects, since mFOLFIRINOX is generally less well tolerated. This project will look at patient molecular data and clinical outcomes from PanCAN's SPARK database to discover predictors that can help doctors pair patients with the most appropriate chemotherapy.Dr. Bakir will do this using a technique called master regulator analysis. A master regulator is a protein that is more important than other proteins in making a cell behave the way it does. There are sophisticated computational techniques that allow scientists to sequence genetic material known as mRNA to make predictions about what the master regulator proteins are in a particular patient's tumor. Training a computer to study large sets of patient mRNA information allows scientists to make predictions for what the master regulators are.This project uses PanCAN's SPARK database to determine whether there are master regulators that make patients either more or less sensitive to mFOLFIRINOX or gemcitabine/nab-paclitaxel. SPARK contains the mRNA levels from the tumors of 350 patients with pancreatic cancer as well as their treatment history. By knowing both the mRNA levels in the tumor and how the patients responded to chemotherapy, Dr. Bakir and team can correlate the patient response to the master regulators that are active in their tumor. If promising candidate master regulators are found, the investigators will then look at biopsies and tumor tissue from surgeries to see if the expected master regulators are expressed on a protein level. This proposal therefore uses computational techniques with the goal of discovering markers that can help doctors predict which chemotherapy is best for a given patient.