Subgroup A patients receiving T+C (n=143), had a median PFS of 9.4 months (M), hPFS of 10.8M and TTDQoL of 5.7M vs patients receiving C (n=160), had a median PFS of 7.6M (Hazard ratio (HR): 0.64, 95% confidence interval [CI]=0.47, 0.87; 1-sided p=0.0020), hPFS of 7.6M (HR: 0.53, 95% CI=0.39, 0.73; 1-sided p< 0.0001) and TTDQoL of 3.9M (HR: 0.65, 95% CI=0.46, 0.91; 1-sided p=0.0063). Additionally, the CEA percentage change from baseline was found to impact the treatment effect (T+C or C) on TTDQoL (2-sided p=0.0061 from a Cox regression interaction test). CEA response and ORR in Subgroup A patients with baseline CEA > 5 ng/mL are shown below. Similar results, with a larger difference, were seen in a smaller Subgroup B (patients with KRAS-wild type and ECOG 0 or CEA< 35 ng/mL).
Subgroup analysis identified patients with further improved outcomes, including TTDQoL, for T+C vs C. More patients with baseline CEA >5ng/mL treated with T+C had a CEA reduction consistent with a better ORR with T+C vs C, supporting CEA use as an early marker for tumor response and progression with T+C.