Our current clinical trial proposal includes a short course of pre-operative, single agent erlotinib followed by post-operative erlotinib-gemcitabine in a neo-adjuvant/adjuvant approach to the treatment of patients with resectable pancreatic adenocarcinoma. The short course pre-operative erlotinib treatment serves two objectives: 1) erlotinib, through its cytostatic effects, may hinder the ability of tumor cells to metastasize at the time of surgical resection with minimal toxic effects and no delay in surgical treatment; 2) the true effects of erlotinib on potential determinants of response can be best assayed on the pancreatic cancer itself since no other model can better resemble the patients authentic tumor microenvironment. The post-operative treatment component of the protocol attempts to improve the demonstrated effects of gemcitabine by adding erlotinib in the adjuvant setting, a drug combination that, as mentioned, has already been proven to be advantageous for patients with advanced pancreatic cancer.
Traditionally, chemotherapy for cancer has been conducted in an empiric fashion by first selecting a regimen based on clinical trial evidence and clinical parameters, and then by assessing the objective response to that regimen employing clinical and radiographic imaging means. This approach suffers from many disadvantages, most conspicuously the inability to select the better patient candidates prior to the initiation of therapy, thus sparing the risk and expense of ineffective treatment for patients who are unlikely to respond. Since pancreatic cancer expression of EGFR protein by itself is not predictive of therapeutic response, alternative methods of patient selection seem to be essential for the success of EGFR-targeted treatment. The understanding of EGFR molecular signaling has allowed the drug development process to shift from an empiric random screening approach to a more rational and mechanistic, target-directed approach. Among multiple attempts to identify molecular determinants of tumor cell sensitivity to EGFR inhibitors, there are two main paradigms that stand out: 1) activation of downstream pharmacodynamic effectors associated with response to the drug (i.e. phosphorylation of Akt or extracellular signal-regulated kinase (ERK); expression of c-fos), and 2) prediction of sensitivity based on a "static" analyte (detection of epidermal growth factor receptor (EGFR) sensitizing mutations; epithelial-mesenchymal transition profile). However, given the complexity of factors governing pancreatic cancer response to erlotinib, the biologic heterogeneity of malignant phenotype, and overall relatively low response rates, we believe that it is unlikely that analysis of a single biomarker will be useful for patient selection. A comprehensive analysis of a dynamic panel of biomarkers relevant to EGFR signaling and erlotinib mechanism of action seems more useful in that sense. Furthermore, the limited ability of pancreatic cancer tissue sampling often precludes biomarker correlation to assays worked out in xenograft models or in-vitro conditions.
The translational rationale of this proposal is therefore to develop predictive chemotherapy sensitivity-resistance assays (CSRA) for pancreas carcinoma patients treated with erlotinib and gemcitabine. This will represent a major advance, because the CSRA would enable prediction of clinical response prior to initiation of therapy.
