CEA response as a predictive marker for response according to RECIST in patients with KRAS wild-type metastatic colorectal cancer (KRAS WT mCRC) (pts) treated with panitumumab monotherapy.

Abstract

574 Background: Panitumumab is a fully human monoclonal antibody targeting the epidermal growth factor receptor (EGFR) and is approved as treatment in KRAS WT mCRC in combination with chemotherapy and in chemotherapy-refractory pts as monotherapy. In mCRC, CEA decrease is often used as an early response predictor, but correlative studies with RECIST and formal cutoff criteria are scarcely available. Here we report on an exploratory analysis in KRAS WT mCRC pts treated with monotherapy panitumumab. Methods: A series of 29 KRAS WT mCRC pts who have been treated with panitumumab monotherapy (6 mg/kg once every 2 weeks) as second-line treatment were included (58% male; mean age 65,6 years; performance status was 0-1 in 90%). Every 12 weeks, response evaluation by CT-scan was done routinely. Responses were determined using RECIST v1.1. CEA levels were determined every 6 weeks. A CEA response was defined as two consecutive CEA decreases of ≥10% compared to the previous value. Progression was defined as two consecutive CEA increases of ≥10% compared to the previous value. Pts not meeting these criteria were defined as stable. CEA change at week 6 and CEA response at week 12 week were correlated with RECIST response at week 12 using Spearman’s correlation coefficient. Results: A 10% CEA decrease at week 6 was predictive for response according to RECIST at week 12 (r=0.71, p<0.001). Likewise, at week 12, responses determined by RECIST and CEA strongly correlated (r=0.88, p<0.001). Conclusions: Data from this small study suggest that CEA changes may have predictive potential as early time point response biomarker in pts treated with panitumumab monotherapy. More research is warranted to confirm these retrospective findings. [Table: see text]