Phosphorylated Erk (Perk) As Biomarker In Patients With Advanced Pancreatic Cancer Treated With Erlotinib Within A Randomized Phase Iii Trial (Aio-Pk0104)

189 Background: Validated biomarkers for patients (pts) with advanced pancreatic cancer (APC) treated with anti-EGFR drugs have not yet been defined. pERK represents the downstream target of the RAS-RAF-MEK cascade. Methods: Within AIO-PK0104 281 pts with APC were randomized between gemcitabine/erlotinib followed by capecitabine and capecitabine/erlotinib followed by gemcitabine. Archival formalin fixed paraffin embedded (FFPE) tumor tissue for central pERK analysis by immunohistochemistry (IHC) was available from 153 pts. Within a retrospective subgroup analysis, pERK data (either as dichotomous or continuous variable) were correlated with AIO-PK0104 biomarker results on KRAS mutation status and EGFR expression and with efficacy study endpoints using a Cox regression model. Results: A semi-quantitative immunohistochemistry scoring system considering cytoplasmic and nuclear pERK expression (staining intensity & % of positive cells) was developed (score 0-12). FFPE samples with a score of 6-12 were defined a pERK high . 98/153 pts were classified as pERK high and 55/153 pts as pERK low , respectively. The median pERK score was 7 (range 0-12). No significant correlation of pERK with baseline characteristics like stage of disease, gender, age, KPS or CA 19-9 was detected. Median OS in pERK high pts was estimated with 5.7 months and with 6.2 months in pERK low pts (HR 1.29, 95% CI 0.90-1.83, p=0.16). However, when analysing pERK as continuous variable, a significant association between the pERK score and OS was found (HR 1.06, 95% CI 1.0-1.12, p [log rank]=0.050, p [likelihood ratio]=0.047), indicating an increase in the hazard for death by a factor of 1.06 for each score level of pERK expression. There was no correlation of pERK expression with the objective disease control rate (OR 0.99, p=0.91) or the occurrence of skin rash (OR 0.93, p=0.41). Pts with a KRAS mutation had a similar rate of pERK high expression compared to patients with KRAS wildtype (61% vs 71%, p=0.32). Conclusions: pERK expression may have an impact on OS of APC pts treated with the anti-EGFR agent erlotinib. Prospective validation of these results is recommended. Clinical trial information: NCT00440167.