Introduction: The Köhne risk classification was established in mCRC patients (pts) treated with 5-fluorouracil-based chemotherapy. (Ann Oncol 2002;12:308-317). Afterwards the GERCOR group developed a prognostic model for pts treated with an oxaliplatin-based or irinotecan-based 1st-line chemotherapy. (The Oncologist 2011;16:1228-1238). Bevacizumab (Bev) in combination with FOLFOX/FOLFIRI has shown a survival benefit in the 1st-line therapy and is considered a standard of treatment in mCRC; however, validated prognostic models are not available in these pts treated with antiangiogenics.Methods: Retrospective review of FOLFOX/FOLFIRI-Bev treated pts (2006-2013). Progression-free survival (PFS) and OS were assessed. Univariate and multivariate analysis of prognostic factors for OS was performed. To study the correlation between the Köhne and GERCOR prognostic models, each model's predicted survival was plotted and a Pearson's correlation coefficient was used to analyze the similarity between these two prognostic indexes.Results: 124 pts found. Baseline characteristics in Table 1. With a median follow-up of 65 months (mts) (range 22-112), median (m) PFS and OS was 10.9 (CI95% 9,1-12,8) and 26.8 mts (CI95% 19,4-34,4), respectively. The 1-, 2-, and 5-year OS rate was of 84%, 59% and 14%. Univariate analysis: Performance status (PS) = 2, anemia, hypoalbuminemia, high CA 19.9 baseline level, elevated serum lactate dehydrogenase (LDH) at baseline, > 1 metastatic site, were all significant poor prognostic factors. In the multivariate study, only PS, number of metastatic sites, and serum LDH retained their significant prognostic value. By applying the Köhne model, mOS was 35.2 mts (CI95% 28-42.3) in the low-risk group, 24.1 mts (CI95% 18.2-30.1) in the intermediate-risk group and 3.6 mts (CI95% 0.1-7.9) in the high-risk group. (p-value < 0.0001). According to the GERCOR model, mOS was 38.8 mts (CI95% 25.1-52.4) in the low-risk group, 26.1 mts (CI95% 15.7-36.5) in the intermediate-risk and 17.1 mts (CI95% 14.7-19.5) in the high-risk group, respectively (p-value< 0.0001). A moderate positive relationship was confirmed between them, with a Pearson coefficient of 0,413 that was statistically significant (p-value < 0,0001).Conclusion: PS, serum LDH baseline value, number of metastatic sites were the main prognostic factors in our pts treated with 1st-line FOLFOX/FOLFIRI-Bev. The GERCOR prognostic appears to have advantages over the Köhne model for risk-stratification of patients outside clinical trials, especially in the selection of higher-risk patients.Table: P-281 Introduction: The Köhne risk classification was established in mCRC patients (pts) treated with 5-fluorouracil-based chemotherapy. (Ann Oncol 2002;12:308-317). Afterwards the GERCOR group developed a prognostic model for pts treated with an oxaliplatin-based or irinotecan-based 1st-line chemotherapy. (The Oncologist 2011;16:1228-1238). Bevacizumab (Bev) in combination with FOLFOX/FOLFIRI has shown a survival benefit in the 1st-line therapy and is considered a standard of treatment in mCRC; however, validated prognostic models are not available in these pts treated with antiangiogenics. Methods: Retrospective review of FOLFOX/FOLFIRI-Bev treated pts (2006-2013). Progression-free survival (PFS) and OS were assessed. Univariate and multivariate analysis of prognostic factors for OS was performed. To study the correlation between the Köhne and GERCOR prognostic models, each model's predicted survival was plotted and a Pearson's correlation coefficient was used to analyze the similarity between these two prognostic indexes. Results: 124 pts found. Baseline characteristics in Table 1. With a median follow-up of 65 months (mts) (range 22-112), median (m) PFS and OS was 10.9 (CI95% 9,1-12,8) and 26.8 mts (CI95% 19,4-34,4), respectively. The 1-, 2-, and 5-year OS rate was of 84%, 59% and 14%. Univariate analysis: Performance status (PS) = 2, anemia, hypoalbuminemia, high CA 19.9 baseline level, elevated serum lactate dehydrogenase (LDH) at baseline, > 1 metastatic site, were all significant poor prognostic factors. In the multivariate study, only PS, number of metastatic sites, and serum LDH retained their significant prognostic value. By applying the Köhne model, mOS was 35.2 mts (CI95% 28-42.3) in the low-risk group, 24.1 mts (CI95% 18.2-30.1) in the intermediate-risk group and 3.6 mts (CI95% 0.1-7.9) in the high-risk group. (p-value < 0.0001). According to the GERCOR model, mOS was 38.8 mts (CI95% 25.1-52.4) in the low-risk group, 26.1 mts (CI95% 15.7-36.5) in the intermediate-risk and 17.1 mts (CI95% 14.7-19.5) in the high-risk group, respectively (p-value< 0.0001). A moderate positive relationship was confirmed between them, with a Pearson coefficient of 0,413 that was statistically significant (p-value < 0,0001). Conclusion: PS, serum LDH baseline value, number of metastatic sites were the main prognostic factors in our pts treated with 1st-line FOLFOX/FOLFIRI-Bev. The GERCOR prognostic appears to have advantages over the Köhne model for risk-stratification of patients outside clinical trials, especially in the selection of higher-risk patients. Table: P-281