P-44 Impact of regorafenib dose optimization on clinical outcomes compared to best supportive care and TAS-102 in the treatment of relapsed/refractory metastatic colorectal cancer

Abstract

In a randomized controlled trial (RCT) ReDOS of relapsed/refractory metastatic colorectal cancer (mCRC) patients, regorafenib dose optimization (rego 80+), starting at 80 mg/day with 40 mg weekly increments to the standard dose of 160 mg/day, was found to improve tolerability and clinical outcomes compared to regorafenib at standard dose (rego 160), 160 mg/day for 21 days of a 28-day cycle. This network meta-analysis (NMA) aims to further investigate the impact of rego 80+ on efficacy and safety compared to best supportive care (BSC) and trifluridine/tipiracil (TAS-102). RCTs included in the NMA were identified via a systematic literature review conducted in April 2021. A feasibility assessment was performed to ensure that the included trials did not differ significantly with respect to treatment effect modifiers. Bayesian fixed effect NMAs were performed to simultaneously synthesize hazard ratios (HRs) for overall survival (OS) and progression-free survival (PSF) and odds ratios for safety endpoints of interest, and their respective 95% credible intervals (Crl). The HRs from the NMA were applied to the rego 80+ OS and PFS Kaplan Meier Curves from the ReDOS trial to estimate predicted median OS and PFS for each treatment. Three global or US-only RCTs (ReDOS, CORRECT, RECOURSE) and three Asian-only RCTs (TERRA, Yoshino 2012, and CONCUR) were included in the NMA. When all trials were analyzed, rego 80+ was associated with statistically significant improvements in OS and PFS vs. BSC (HR [95% CrI]: 0.49 [0.33, 0.73] and 0.35 [0.24, 0.53], respectively) and numerically favorable improvements in OS and PFS vs. rego 160 (0.72 [0.49, 1.05] and 0.84 [0.56,1.25]), and TAS-102 (0.71 [0.47, 1.08] and 0.78 [0.52, 1.18]). Median OS (95% confidence interval [CI]) was estimated to be 9.8 (7.5, 11.9) months for rego 80+, 7.5 (4.6, 11.9) months for rego 160, 7.5 (4.4, 11.9) months for TAS-102, and 6.4 (2.5, 10.3) months for BSC. Median PFS (95% CI) was 2.80 (2.0, 2.5) months for rego 80+, 2.07 (1.89, 5.85) months for rego 160, 1.98 (1.8, 5.76) months for TAS-102, and 1.89 (1.17, 2.16) months for BSC. For grade 3+ adverse events (AEs), BSC was the most favorable, with the exception of increased aspartate transaminase (AST). Compared to rego 160, rego 80+ was numerically favorable in all AEs analyzed including hand-foot skin reactions (HSFR), diarrhea, fatigue, hypertension, and increased AST. Compared to TAS-102, rego 80+ had higher incidence of HSFR, comparable fatigue and hypertension, and numerically lower incidence of diarrhea, increased AST, neutropenia, febrile neutropenia, leukopenia, and anemia. All NMA findings were consistent when Asian-only trials were excluded from the analyses. Findings from this NMA indicate that rego 80+ provides additional survival benefits (over standard dose rego 160) when compared to BSC and TAS-102; with gains of median OS of 3.4 months (∼50% risk reduction), and 2.3 months (∼30% risk reduction), respectively. Results also suggest that rego 80+ delays disease progression in comparison. With its more favorable safety profile, rego 80+ should be considered a preferred option for optimizing clinical outcomes in patients with relapsed/refractory mCRC.