Abstract #LB-270: Prospective use of optimized ras mutation diagnostic in Phase 2 trial of mutation specific Ras-targeted immunotherapy (GI-4000) in adjuvant pancreas cancer

AACR Annual Meeting-- Apr 18-22, 2009; Denver, CO Background: Ras mutations are important predictive markers for EGFR-targeted treatment of non-small cell lung cancer (NSCLC) and colorectal cancer (CRC) and are found in 71% of pancreas ductal adenocarcinomas (COSMIC database). GI-4000 is a targeted immunotherapy designed to eliminate tumor cells expressing specific Ras mutations in exons 2 (G12V, C, D or R) or exons 3 (Q61L, R or H). Ras genotyping is a prerequisite for enrollment into the randomized double blind placebo controlled Phase 2 trial in newly resected pancreas cancer (GI-4000 plus gemcitabine vs. gemcitabine alone). Eligible patients are only treated with the GI-4000 product that matches the mutation identified in their tumor. The gold standard diagnostic method, namely PCR amplification and double strand sequencing limited to Ras exon 2 is susceptible to a risk of false negatives due to the presence of wild-type ras DNA from heterozygous tumor cells and low tumor cell content in the paraffin embedded formalin fixed specimens. Eliminating false negatives should improve clinical decision making for all mutant ras driven cancers. Here we describe how use of an optimized ras genotyping method improved patient screening. Experimental procedures: Tissue was obtained by surgical resection of tumors from 221 subjects with Stage I and II pancreas cancer to date from sites in US, India and Bulgaria. Tumors were genotyped for K-ras exons 2 and 3 using an optimized bi-directional sequencing of DNA from nested PCR amplification with peptide-nucleic acid (PNA) oligomer clamping. Results: Compared to the gold standard Ras genotyping method, the inclusion of PNA oligomers suppressed amplification of wild type Ras sequences (as previously reported) and substantially reduced the false negative rate. This optimized method detected ras mutations down to 1 ng of genomic DNA, and to 2% heterozygous mutant tumor cells in the sample. Using the optimized method, tumor genotyping revealed that 86% of pancreas cancers screened for the Phase 2 study had K-ras mutations, predominantly G12V or G12D. In addition, tumors harbored G12R (13%) or Q61H (4%) mutations at greater frequencies than previously reported. Non-canonical mutations were also detected, highlighting a potential source of false negative results for assays with restricted mutation coverage. Conclusions: Prospective Ras testing in an ongoing Phase 2 study of newly resected pancreas cancer demonstrates the importance of optimizing the diagnostics used to screen tumors for the purpose of personalized cancer therapies. We observed substantial reduction in the rate of false negatives and increased detection of non-canonical mutations in ras. Improved sensitivity and specificity of Ras diagnostic methods are also predicted to have important implications in treatment decisions and outcomes in the context of EGFR-targeted approaches for NSCLC and CRC. Citation Information: In: Proc Am Assoc Cancer Res; 2009 Apr 18-22; Denver, CO. Philadelphia (PA): AACR; 2009. Abstract nr LB-270.