Background: It has not yet been defined if KRAS has a prognostic value or is a predictive biomarker for the efficacy of erlotinib in advanced pancreatic cancer (PC).
Methods: AIO-PK0104 was a phase III trial comparing gemcitabine/erlotinib followed by capecitabine with capecitabine/erlotinib followed by gemcitabine in advanced PC. For this post hoc subgroup analysis, biomarker data on the KRAS exon 2 mutation status were correlated with objective response to 1st-line therapy and with overall survival after start of 2nd-line chemotherapy (OSc).
Results: KRAS codon 12 was mutated in 121 of 173 (70 %) patients. The KRAS status showed no association with objective response (p = 0.40), but KRAS wildtype patients had an improved OS (HR 1.68, p = 0.005). A trend for a survival benefit was also observed during (non-erlotinib containing) 2nd-line chemotherapy, with a HR of 1.47 (p = 0.10) for the OSc.
Conclusion: This post hoc analysis of AIO-PK0104 supports the assumption that KRAS is rather a prognostic than a predictive biomarker in PC.