Background: Antiangiogenic drug (AAD)-triggered oxygen and nutrient depletion through suppression of angiogenesis switches the glucose-dependent metabolism to lipid-dependent metabolism. Blocking fatty acid oxidation can enhance AAD-mediated anti-tumor effects in colorectal cancer. Previous reports suggested that polymorphisms of the lipid metabolism-related genes are associated with the increased risk of CRC and poor clinical outcome in CRC. Therefore, we hypothesized that genetic variants in the lipid metabolism pathway may predict first-line treatment outcome in mCRC pts. Methods: Genomic DNA from blood samples of pts enrolled in two independent randomized trials, FIRE-3 and MAVERICC, was genotyped through the OncoArray, a customized array manufactured by Illumina including approximately 530K SNP markers. The impact on outcome of 25 selected SNPs in 10 genes involved in the lipid metabolism pathway (CD36, FABP4, LPCAT1, LPCAT2, PPARG, CPT1A, ACSS2, SREBF1, FASN, ACACA) was analyzed. Those treated with FOLFIRI/ bevacizumab (bev) in FIRE-3 (n = 107) and MAVERICC (n = 163) served as discovery and validation cohorts respectively, while FIRE-3 FOLFIRI/ cetuximab (cet) (n = 129) arm was used as the control. Interaction between each SNP and treatment was evaluated in FIRE-3 (FOLFIRI/bev arm vs. FOLFIRI/cet arm). Results: In the discovery (FIRE-3 bev) cohort, pts with FASN rs4485435 any C allele (N = 21) showed significantly shorter progression-free survival (PFS) (8.69 vs 13.48 months) compared to carriers of G/G (N = 62) in both univariate (hazard ratio [HR] = 2.88; 95% confidence interval [CI]: 1.57-5.29; p = 0.00037) and multivariate (HR = 2.87; 95%CI 1.4-5.9; p = 0.00675) analyses. These data were validated in the MAVERICC bev cohort in multivariate analysis (11.17 vs 14.06 months; HR = 2.07; 95%CI: 1.15-3.74; p = 0.02). Pts carrying any T allele in PPARG rs3856806 (N = 36) showed significantly longer overall survival (OS) (Not reached vs 42 months) than carriers of C/C (n = 93) in the FIRE-3 cet cohort in both univariate (HR = 0.4; 95%CI 0.17-0.92; p = 0.03) and multivariate (HR = 0.37; 95%CI 0.15-0.93; p = 0.02) analyses, but the association was not observed in the bev cohort of MAVERICC and FIRE-3. In the comparison of bev arm vs cet arm in FIRE-3, interactions were shown with FASN rs4485435 (p = 0.017) on PFS and PPARG rs3856806 (p = 0.059) on OS. Conclusions: Our study demonstrates for the first time that FASN polymorphism could predict outcomes of bev-based treatment in mCRC patients; Meanwhile PPARG polymorphism could predict outcomes of cet-based treatment in mCRC patients. These findings support a possible role of the lipid metabolism pathway in contributing to resistance to anti-VEGF/EGFR treatment.