In general, tumors are heterogeneous and exhibit different phenotypic characteristics that result from multiple genetic alterations and/or cellular origin. This heterogeneity contributes to both the variation in course and severity of disease and to the variation in response to current treatment regimes. To this end, we recently identified different transcriptome subtypes of pancreatic, colorectal, breast and multiple other cancers (including a combined tissue-independent gene expression subtypes from multiple epithelial cancers) using primary tumor samples and specific gene signatures. These subtypes exhibit differential prognosis, cellular origin/phenotype and responses to current treatment regimens. These suggest that clinically available and perhaps additional drugs targeting specific molecular, biochemical and metabolic pathways unique to each subtype may be effective for cancer patients. Personalized and precise therapy exploiting such distinctions could ultimately be applied in a clinical setting with goals to improve disease management and avoid over and under-diagnosis and treatment.
The team’s long-term research interests are focused on facilitating personalized and precise cancer diagnosis and therapy using these molecular subtypes. Specifically, the goals are to
(i) explore the cellular origin/phenotype (especially as it relates to stemness) of these subtypes (ii) examine the mechanisms by which genome-wide transcriptomic/genomic/metabolic abnormalities contribute to the tumor pathophysiologies/phenotypes of the subtypes
(iii) assess the ways these abnormalities and cellular origin/phenotype influence patient survival and responses to specific therapies
(iv) to pin-point more effective personalized therapies specific to each subtype with lower toxicity.
Together, such strategies will help to refine and/or develop biomarkers that will allow for the stratification of patients into classes. Ultimately, our hope is that this information will be translated into the clinic where such genetic/metabolic markers will be used to subtype cancer patients in order to identify the most appropriate therapy or combination of therapies to confer the greatest benefit to the patient.
The team’s general approach involves hypothesis-driven (with a goal of identifying precise therapies for tumor subtypes) integrated analysis of high-throughput genomics, transcriptomics (measured using microarrays and/or massively parallel sequencing platforms) and metabolomics data (measured using targeted and untargeted mass spectrometry-based profiling platforms) generated from patient tumors, transplanted and genetically engineered animal tumors and cancer cell lines. Results from patient samples are then correlated with clinical information. Moreover, predictions generated with patient samples are also tested and validated at the bench using cell lines and/or mouse model experiments (including preclinical trials). Overall, this strategy follows a systems approach involving both computational biology (analysis of high-throughput data, in silico) and wet-lab (in vitro and in vivo) experiments, which will go hand-in-hand to achieve the overall experimental goals.