I am a Physician Scientist in the Departments of Radiation Oncology, Neurological Surgery, and Pathology at the University of California San Francisco (UCSF). Eighty percent of my time is dedicated to research, and my lab is focused on understanding genomic, biochemical, and cellular drivers of brain tumor heterogeneity and evolution. To do so, we integrate human samples with evolutionarily-diverse preclinical models to develop new biomarkers and new therapies to improve treatments for patients. I am the Robert and Ruth Halperin Endowed Chair in Meningioma Research, and the Director of the Preclinical Therapeutics Core in the UCSF Brain Tumor Center. As a practicing physician, I use radiotherapy and radiosurgery to treat patients with brain tumors.

Neurofibromatosis cancer predisposition syndromes are associated with meningiomas, Schwann cell tumors, ependymomas, gliomas, and other tumors that are derived from the neural crest, a multipotent embryonic cell population with remarkable molecular diversity. Using DNA methylation profiling integrated with genetic, transcriptomic, biochemical, proteomic, single-cell, mechanistic, and functional approaches, we showed meningiomas are comprised of 3 molecular groups with distinct clinical outcomes, biological drivers, and therapeutic vulnerabilities. Building on these foundational data, we analyzed 1856 meningiomas from 12 medical centers across 3 continents to develop a targeted gene expression biomarker that improves prediction of meningioma outcomes compared to all other classification systems and identifies which meningiomas benefit from postoperative radiotherapy. Clinical trials based on our biomarkers and therapeutic discoveries are under development, and we are now using single-cell technologies, spatial techniques, functional genomics, and preclinical models to study therapeutic vulnerabilities underlying tumor heterogeneity and evolution. We showed that radiotherapy is sufficient for interconversion of brain tumor molecular groups through epigenetic and metabolic reprogramming, that subclonal populations of brain tumor cells underly tumor heterogeneity, and that organoid models can be used to test drugs that target intratumor heterogeneity.