Project Summary ERBB2 is amplified in ~20% of Gastric and esophageal adenocarcinomas (GEAs) and metastatic ERBB2+ GEAs are treated with a combination of chemotherapy and the antibody Trastuzumab. However, Trastuzumab is only modestly effective in GEA, and all other targeted agents in ERBB2+ breast cancer have failed in GEA clinical trials. We propose to directly address the two primary factors that we hypothesize to mediate failure of ERBB2 therapy in GEA: adaptive resistance and genetic complexity. To perform these studies, our team of investigators with complementary skill sets will both perform detailed assessment of optimal approaches to stably inhibit ERBB2 using an array of patient-derived model systems. Furthermore, we will perform a prospective clinical collection spanning multiple large academic medical centers in which we evaluate the genetic evolution of ERBB2+ GEAs during therapy, define genetic alterations that accompany resistance and then functionally validate mechanisms of resistance and optimal combination therapy. We will also explore the role of cell-free (cf)DNA genomic profiling to guide therapy in the face of genomic evolution of the disease during therapy. The overall goal will be to validate candidate resistance mechanisms and seek to define optimal combination therapies that can overcome them. We therefore propose the following Specific Aims: Aim 1: To define mechanisms of adaptive resistance to ERBB2 therapy in GEA patient samples and to develop optimal targeted combinations to stably inhibit ERBB2 activity in GEA model systems. Aim 2: To evaluate genetic etiologies of resistance by determining how frequently resistance results from ERBB2-negative subclones or from secondary genomic alterations in ERBB2+ tumor cells. Aim 3: To validate mechanistically the capacity of secondary genomic alterations to promote Trastuzumab resistance and to test combination therapies to overcome resistance. In summary, we will define genetic and non- genetic mechanisms of resistance to ERBB2 therapy. Ideally, our studies will lead to the development of active/optimal candidate therapies that work well in first-line therapy as well as in in tumors marked by acquired resistance to current therapy.