Therapeutic Combinations Initiative

Comprehensive mapping of drug resistance mechanisms and processes

Understanding the drivers of therapeutic resistance will help unlock the door to achieving durable therapeutic responses and provide the information critical for fulfilling the promise of genome-guided therapies. Thus, we are undertaking an ambitious approach to mapping the drivers of therapeutic resistance by:

  • Characterizing relapsed patient tumors using standard and liquid biopsies
  • Functionally testing resistance drivers using functional genomic techniques
  • Creating cellular models of drug resistance across a diverse portfolio of diseases and therapeutics (see more about GPP, GP, Blood biopsy team, CCLF)

Predicting effective drug combinations from maps of tumor dependencies

The Broad is cataloging all of the genes and proteins needed by tumor cells for viability. The goal of this effort is to identify all possible ways to target cancer cells and associate them with the genetic features that predict cancer cell killing. This project will exploit those data to identify co-dependencies that nominate drug targets to be used in combinations, effectively closing all possible escape routes and preventing the outgrowth of resistance.

Testing thousands of drugs in combination with current cancer therapies to identify those with improved tumor-killing potential

Many current cancer therapies are very effective at initially reducing tumors in patients. This project will focus on identifying other drugs that, when combined with current treatments, will produce lasting responses (see PRISM).

Identify effective therapies with non-overlapping resistance mechanisms

If a combination of drugs could be found that stopped all possible resistance mechanisms, resistance would not develop. Thus, we are using a new technology developed at the Broad to empirically and prospectively identify therapies that have non-overlapping resistance mechanisms (see GPP). Subsequently, therapies with dissimilar patterns of resistance will be combined in pre-clinical models to determine their effectiveness at engendering long-term responses.