Recent clinical successes in cancer genome-inspired personalized medicine have led to exciting patient responses, but only a subset of these clinical responses are long-lasting. Tumors are adept at becoming resistant to therapies, and research elucidating the molecular mechanisms of cancer drug resistance is critical to pinpointing strategies to prevent these escape routes.
Since drug combination therapies first proved effective in treating cancer in the 1960s, such therapies have been used successfully against testicular cancer, Hodgkin’s lymphoma, and other advanced cancers. But, despite their effectiveness, new combination therapies have been slow to emerge. It has proven challenging to choose from thousands of potential drugs and to test them via trial and error in a safe and efficient way. However, new computational models and technologies are making it increasingly possible to conduct large-scale searches for drug combinations in a rational, systematic fashion.
Researchers at the Broad are currently using these new approaches to change the way cancer drug combinations are identified, tested, and delivered to patients. They are creating computational predictions of effective therapeutic combinations by integrating knowledge of catalogued resistance mechanisms in patients and cell models, gene-expression profiles produced by drug exposure, and patterns of drug sensitivity generated from the Cancer Dependency Map. Once promising combinations have been rapidly identified, we will use laboratory models to predict and measure the efficacy of these new therapeutic strategies to determine which are the most promising for cancer patients.
Genomic characterization of drug resistance in clinical samples