Presentation Abstract

Abstract Number: LB-174
Presentation Title: Use of massively parallel, next-generation sequencing to identify gene mutations beyond KRAS that predict response to panitumumab in a randomized, phase 3, monotherapy study of metastatic colorectal cancer (mCRC)
Presentation Time: Monday, Apr 19, 2010, 4:15 PM - 4:25 PM
Location: Room 146, Washington Convention Center
Author Block: Marc Peeters, Kelly S. Oliner, Alex Parker, Jing Huang, Salvatore Siena, Yves Humblet, Jean-Luc Van Laethem, Thierry André, Jeffrey Wiezorek, David Reese, Scott D. Patterson, Eric Van Cutsem. University Hospital Ghent, Ghent, Belgium, Amgen Inc., Thousand Oaks, CA, Ospedale Niguarda Ca’ Granda, Milan, Italy, Centre du Cancer de l'Universite Catholique de Louvain, Brussels, Belgium, Erasme University Hospital, Brussels, Belgium, Hôpital Pitié-Salpétrière, Paris, France, University Hospital Gasthuisberg, Leuven, Belgium
Abstract Body: Introduction: KRAS gene mutation is a well-established biomarker for a lack of response to anti-EGFR antibodies in CRC. We used next-generation sequencing technology to investigate whether mutation of other genes known to be altered in CRC is predictive of response to panitumumab.
Methods: Banked patient tumor samples from a randomized, phase 3, monotherapy clinical trial that compared panitumumab plus best supportive care (BSC) to BSC alone (the 20020408 study, primary endpoint is progression-free survival [PFS]) were sequenced using 454 pyrosequencing technology (Roche). Archival patient tumor samples from 320 patients that had previously been analyzed for KRAS exon 2 mutations using allele-specific PCR (DxS, Manchester, UK) were analyzed for mutations in 9 genes: AKT1, BRAF, CTNNB1, EGFR, KRAS (exon 3), NRAS, PIK3CA, PTEN, and TP53.
Results: 288 patient tumor samples, balanced between the two treatment arms, yielded results for a mean of 7.85 genes per patient; data completeness for each gene ranged from 84% to 99%. 5.7 million DNA sequence reads comprising a total of 1.26 billion base pairs were obtained using 51 primer pairs to amplify 43 exons. Mutations were detected in: 1/250 (0.4%) AKT1, 18/243 (7.4%) BRAF, 5/256 (2.0%) CTNNB1, 3/280 (1.1%) EGFR, 7/284 (2.5%) KRAS (exon 3), 14/282 (5%) NRAS, 24/255 (9.4%) PIK3CA, 15/272 (5.5%) PTEN, and 167/277 (60.3%) TP53 patient tumor samples. Fifty tumors had more than one mutant gene, and 20 had more than one mutation in a single gene. A higher than expected rate of simultaneous mutation at KRAS and either BRAF or NRAS was observed. Panitumumab significantly improved PFS in the patients with KRAS wild type (WT) (exons 2+3) tumors (N=153; hazard ratio [HR]=0.39; 95% CI=0.28,0.56) and had no effect on PFS in patients with KRAS mutant (MT) (exons 2+3) tumors (N=124; HR=1.03; 95% CI=0.71,1.50). In patients with KRAS WT tumors, the effect of panitumumab compared to BSC on PFS for additional tumor genotypes were as follows: NRAS WT (N=138; HR=0.39, 95% CI=0.27,0.56), NRAS MT (N=11; HR=1.94; 95% CI=0.44,8.44); BRAF WT (N=115; HR=0.37; 95% CI=0.24,0.55), and BRAF MT (N=15; HR=0.34; 95% CI=0.09,1.24).
Conclusions: Next-generation sequencing technology can be employed to assess mutation status of multiple gene loci in archival patient tumor samples from a randomized phase 3 study. Observed mutation rates in this study were consistent with previous reports in CRC; however, the superior sensitivity of 454 sequencing revealed unexpected genotypic complexity in many patient tumor samples. Implications of tumor genotype at these nine loci for the effect of panitumumab on progression free survival and objective tumor response will be presented.