LAUSR

Abstract Background The ML18174 study, which showed benefits of bevacizumab (BEV) continuation beyond progression (BBP) for metastatic colorectal cancer (mCRC), excluded patients (pts) with 1st line PFS shorter than 3 months. Methods The subjects of this multi-institutional retrospective study were mCRC pts who showed disease progression ≤100 days in 1st line BEV containing chemotherapy (Cx) between Apr 2010 and Dec 2016. Their 2nd line-Cx regimens were classified according to BBP and non-BBP, and their efficacy was compared by univariate and multivariate analysis using Cox proportional hazards model to adjust for PS, WBC, ALP, number of metastatic sites, RAS status, sidedness in terms of PFS and OS. Results 61 pts were identified as the subject of this study excluding one pt who received ramucirumab in 2nd line-Cx. Baseline characteristics were numerically different between BBP (n = 36) and non-BBP (n = 25) groups, such as PS (0-1/2: 89/11 and 56/44%), RAS status (wild/mutant/not tested: 28/56/16 and 76/16/8%). 2nd line regimens (BBP/anti-EGFR-containing/cytotoxic alone) were 100/0/0 and 0/64/36% in the BBP and non-BBP groups. Response rates were 5.9 and 8.7%, and median PFS was 3.7 and 2.8 months (HR [95%CI] 0.83 [0.49-1.41], p = 0.486, adjusted HR 0.97 [0.48-1.96], p = 0.932), respectively. The proportions of pts who received 3rd or later line-Cx were 58 and 32% (p = 0.068), and median OS was 7.6 and 5.4 months (HR 0.70 [0.41-1.19], p = 0.193, adjusted HR 0.65 [0.34-1.25], p = 0.195). Conclusions In clinical practice, selection of the 2nd line-Cx, BBP or non-BBP, seemed to depend on pt’s condition after early progression in BEV containing 1st line-Cx, and 2nd line-Cx for these patients showed poor clinical outcomes regardless continuation of BEV or switching to the others.