Improving response prediction to neoadjuvant therapy in pancreatic cancer

点击翻译按钮获取中文摘要

PROJECT SUMMARY Localized pancreatic cancers (PC) are rising in incidence and this trend is expected to continue due to increased use of imaging modalities and focused programs on early cancer detection. Although potentially curable, long terms outcomes for localized PC remain poor. This is due to early micrometastatic spread and risk of systemic recurrence which has been reduced with use of systemic chemotherapy. In addition, use of chemotherapy prior to surgery lowers rates of positive margins and these collective observations have led to the adoption of neoadjuvant chemotherapy (NAC), i.e. prior to surgical resection, for majority of patients. NAC extends for multiple months (4-6 months) making real time assessment of response critical. However, this remains a big clinical challenge as pancreatic tumors don’t change much in size with NAC and use of tumor markers is marred by lack of specificity. Pathologic response to NAC on surgical resection specimens is highly prognostic, yet its molecular predictors remain poorly understood. Transcriptional subtypes of PC include classical and basal subtypes; the latter resembles triple negative breast cancer, displays chemotherapy resistance, and is enriched in metastatic disease. Single cell analysis of pancreatic tumors reveals extensive heterogeneity between classical and basal subtypes as well cells with co-expression of both features. In Aim 1, we use our institutional cohort of 174 patients who underwent pancreatic tumor resection after receiving NAC to identify predictors of pathologic treatment response using pre- treatment tumor specimens. We will investigate the classico-basal cell state heterogeneity using a validated multiplex immunofluorescence panel on initial diagnostic biopsies to understand differences in cell state between tumors which exhibit major response vs. no pathologic response. To identify novel predictive transcriptional states beyond classical-basal subtyping we will perform digital spatial profiling using the Nanostring GeoMx platform. In Aim 2, we propose investigating the utility of highly sensitive ctDNA assays to track changes in tumor burden and cell state during NAC. We expect differential ctDNA dynamics during NAC to predict outcomes. In addition using novel ctDNA based epigenetic assays we will assess whether tumoral cell state evolves during emergence of therapeutic resistance to more basal type. This work will lead to identification of novel predictive biomarkers which can help physicians identify high risk cases which are unlikely to respond to traditional NAC or identify early treatment resistance in real time and facilitate earlier switch to alternative chemotherapy or consideration of clinical trials leading to an improvement in outcomes for patients with localized PC. The proposed research will be performed at Dana-Farber Cancer Institute, under the guidance of Drs. Brian Wolpin and Andrew Aguirre. Both are recognized international leaders in PC biology. Importantly, this research strategy is one part of a comprehensive training program to support my transition to an independent research career in translational cancer biology. My scientific advisors (Drs. Myles Brown, Ryan Corcoran, and Eliezer Van Allen) are distinguished experts with experience highly relevant to the proposed research. My long-term goal is to lead an independent academic research group, focused on deciphering the novel biomarkers and mechanisms underlying therapy resistance in PC. The K08 award will provide pivotal protected time for continued mentored research and enable a successful transition to independence.