Cancer, an acquired disease of somatic genetic/epigenetic alterations and dysregulated protein expression, is currently classified by morphological criteria and treated empirically on a tissue-of-origin basis with toxic chemotherapy drugs. Major efforts are currently underway to develop a molecular classification of cancer on the mRNA and DNA levels by systematically quantifying (profiling) gene expression and copy number in tumor lysates using high-throughput genomics technologies. But tumors are both morphologically (composed of different cell types) and molecularly (different alterations in subpopulations of tumor cells) heterogeneous and it is widely thought that this heterogeneity ultimately leads to treatment resistance and patient mortality. With advances in microscopy, computer-enhanced image analysis, and tissue microarrays, genomic abnormalities and protein expression can now be profiled in-situ without destroying tissue morphology and the spatial relationships among different constituent cell types. Such spatial in-situ proteomics data will complement existing genomic techniques and aid discovery of:

Knowledge gained through in-situ proteomics may ultimately permit oncologists to more effectively individualize therapy with drugs targeted to the specific dysregulated proteins present in an individual patient's tumor.

My laboratory focuses on gliomas, a morphologically and molecularly heterogeneous group of brain tumors with potentially dismal patient outcomes for which few effective drugs are available. The paradigms of personalized chemotherapy for solid tumors, namely Herceptin/trastuzumab and Her-2/neu (ErbB-2) testing in breast cancer as well as Gleevec/imatinib and c-kit testing in gastrointestinal stromal tumors (GIST), serve as scientific blueprints for our approach to coordinated drug-diagnostic biomarker co-development with the following goals:

1. Development of a molecular classification of human gliomas. Tumors will be profiled using high-throughput assays on tissue microarrays, including quantitative fluorescence immunohistochemistry (Q-IHC, Camp, et al. Nature Med 8(11):1323 Nov 2002) and fluorescence in-situ hybridization (FISH).
2. Molecular characterization of genetically engineered murine (GEM) models of gliomas and comparison to their human counterparts with the above techniques
3. Investigation of the molecular mechanisms of drug sensitivity and resistance in vitro and in GEM murine models to identify promising drug candidates for application to the above molecularly-defined tumor subtypes in human clinical trials.