Lung cancers driven by EGFR mutations and ALK-translocations are exquisitely sensitive to small molecule tyrosine kinase inhibitors, however drug resistance invariably develops leading to disease relapse. We work closely with the MGH Thoracic Oncology Group to identify and characterize mechanisms of acquired resistance in lung cancer patients treated with targeted therapies that target EGFR, ALK and many other oncogenic drivers.

By analyzing tumor biopsies or tumor DNA isolated from blood, we are often able to detect mutations and other genomic alterations that cause drug resistance. We also perform functional studies on cell lines or mouse PDX models generated from resistant patients to identify pathways that contribute to drug resistance. These models provide a platform for testing novel therapies to select the most promising for evaluation in clinical trials.

Although “next-generation” TKIs can overcome some resistance mechanisms, acquired resistance develops to these new agents as well. To halt this perpetual cycle, novel strategies are needed that can alter the evolutionary trajectory of acquired drug resistance, but little is known about how cancers evolve during treatment in the clinic. We persistent drug tolerant clones that survive initial therapy can serve as

a cellular reservoir from which heterogeneous genomic mechanisms of resistance may evolve. We are working to isolate and characterize these persistent drug tolerant cells from patient tumors. By identifying vulnerabilities and mechanisms driving tumor evolution in these cells, we hope to develop therapeutic strategies that will disrupt this perpetual cycle of acquired drug resistance.

Genetic, epigenetic and phenotypic heterogeneity can lead to differences in the behavior of individual cancer cells within the same tumor. It is intuitive that this will lead to variable drug sensitivity between tumor clones, yet in most cases, the precise link between these features and variable clinical response to therapy has not been defined. We are applying single cell genomics and other techniques to interrogate the cell states

of tumor cells in clinical tumor samples patient-derived models during therapy. Our goal is to identify intrinsic tumor cell and microenvironmental factors that determine whether a cell will be sensitive or resistant to therapy, and to use this knowledge to design novel therapeutic strategies that can lead to complete responses.

The efficacy of targeted therapies is due, in part, to the induction of a robust apoptotic response. Signaling pathways activated by driver oncogenes modulate the BCL-2 family to suppress apoptosis. When targeted therapies are insufficient to trigger apoptosis, BH3 mimetics that inhibit pro-survival BCL-2 family proteins such as

MCL-1 and BCL-XL can increase sensitivity to TKI treatment. We are working to understand how the evolution of apoptotic dependencies may be shaped by co-ocurring genetic alterations. Ultimately, we hope to develop rational drug combinations that target the specific apoptotic dependencies in each tumor.

To enable our discovery and therapeutic development efforts, we have established an integrated platform for generation of patient-derived cell lines and mouse patient-derived xenograft (PDX) models from clinical tumor samples obtained through an IRB-approved repeat biopsy program. We have generated over 250 cell lines and 100 PDX models from patients treated

with FDA-approved and experimental targeted therapies, as well as early-stage and metastatic treatment naïve lung cancers. These models are a powerful resource for patient-specific modeling of drug sensitivity and resistance and play an instrumental role in the development of new targeted therapeutic strategies.