3602 Background: The potential role of AR tissue levels in the prediction of benefit from anti-EGFRs in mCRC pts was suggested by retrospective series. Preclinical and preliminary clinical experiences showed a modulation of plasma EGFR ligands during the treatment with cetuximab. Previous data by our group evidenced that a significant increase of plasma AR occurred one hour after the administration of cetuximab and higher increases were associated with worse clinical outcome in KRAS and BRAFwt irinotecan-refractory mCRC pts receiving cetuximab and irinotecan. Methods: We designed a prospective confirmatory study in the same setting of mCRC pts. To detect a HR for PFS of 2.3 for pts with high AR levels one hour after the administration of cetuximab (1hr-AR) compared to those with low levels, with two-sided a=0.05 and b=0.2, 45 events were required. The median value was adopted as cut-off. Plasma AR levels were assessed by means of validated ELISA kits. Results: Forty-nine KRAS and BRAFwt pts were included. A significant early increase of AR levels was observed (median increase +24.7%; median levels of baseline AR and 1hr-AR: 18.06 and 24.06 pg/mL, respectively; Wilcoxon signed rank test, p<0.0001). At a median follow-up of 20.4 mos, median PFS and OS were 4.6 and 12.1 mos, respectively. No differences in PFS or OS were observed according to 1hr-AR levels (median PFS 5.5 vs 4.6 mos, HR: 0.76 [95%CI:0.40-1.32], p=0.322; median OS: 15.6 vs 13.4 mos, HR:0.77 [95%CI:0.36-1.62], p=0.485). Conclusions: This prospective experience confirms that AR early increases one hour after the administration of cetuximab. Underlying biological mechanisms should be investigated. Nevertheless, this modulation of AR does not predict clinical outcome. Our work underlines the need to prospectively validate retrospective findings in independent series, to assess their reliability. Clinical trial information: 2008-003160-19.