3506 Background: This multicenter, randomized study assessed efficacy & safety for 3 irinotecan/fluoropyrimidines combinations in previously untreated mCRC. In a 3 × 2 factorial design, we also assessed whether celecoxib added to chemotherapy (CT) improved CT efficacy and/or reduced toxicity. Methods: Pts were randomized to: FOLFIRI - irinotecan (I) 180 mg/m2, leucovorin (LV) 400 mg/m2, 5-FU bolus 400 mg/m2, & infusional 5-FU 2400 mg/m2 over 46 hours q 2 wks; modified IFL (m-IFL) - I 125 mg/m2, LV 20 mg/m2, & bolus 5-FU 500 mg/m2 wkly × 2, q 3 wks; or CapeIri - I 250 mg/m2 day 1 & capecitabine 1000 mg/m2 po BID × 14 days, q 3 wks. Pts were also randomized to concurrent celecoxib (400 mg po BID) or placebo in a double-blind fashion. Time to progression (TTP) was the primary endpoint. Results: 430 pts were enrolled from 2/03 to 4/04, prior to an amendment that added bevacizumab to CT arms. Baseline characteristics were balanced. TTP for FOLFIRI (median = 8.2 mos) was significantly better than for either m-IFL (6.0 mos; p = 0.01) or CapeIri (5.7 mos; p = 0.01). Overall survival (OS) also favored FOLFIRI (median = 23.1 mos) compared to either m-IFL (17.6 mos; p=0.10) or CapeIri (18.8 mos; p = 0.19). Common grade ≥ 3 toxicities are listed below. CapeIri had the highest rates of nausea, vomiting, diarrhea, dehydration & hand-foot syndrome, whereas FOLFIRI had lower rates. Among all 430 pts, median TTP did not differ for pts randomized to celecoxib compared to placebo (6.9 vs 6.9 mos; p=0.71). Median OS was also similar for celecoxib vs placebo (19.5 vs 18.8 mos; p=0.63). CT toxicities did not differ for celecoxib vs placebo. Rates for MI/stroke were 1.5% for celecoxib and 1.9% for placebo. Conclusions: First-line FOLFIRI offers a superior TTP when compared to m-IFL or CapeIri; OS & toxicity analyses also favored FOLFIRI. Celecoxib neither improved CT efficacy nor reduced CT toxicity. Updated survival data & data on pts enrolled after the addition of bevacuzimab will be presented. [Table: see text] [Table: see text]