Baseline carcinoembryonic antigen as a predictive factor of ramucirumab efficacy in RAISE, a second-line metastatic colorectal carcinoma phase III trial

• Published in: European journal of cancer (1990) 2017-06, Vol.78, p.61-69
• Main Authors: Yoshino, Takayuki, Obermannová, Radka, Bodoky, György, Garcia-Carbonero, Rocio, Ciuleanu, Tudor, Portnoy, David C, Kim, Tae Won, Hsu, Yanzhi, Ferry, David, Nasroulah, Federico, Tabernero, Josep
• Format: Article
• Language: English
• Subjects: 5-Fluorouracil; Adult; Aged; Antibodies, Monoclonal - therapeutic use; Antibodies, Monoclonal, Humanized; Antigens; Antineoplastic Agents - therapeutic use; Bevacizumab; Carcinoembryonic antigen; Carcinoembryonic Antigen - metabolism; CEA; Chemotherapy; Clinical trials; Colonic Neoplasms - blood; Colonic Neoplasms - drug therapy; Colonic Neoplasms - mortality; Colorectal cancer; Colorectal carcinoma; Confidence intervals; Double-Blind Method; Female; Folinic acid; Hazards; Health hazards; Hematology, Oncology and Palliative Medicine; Humans; Immunotherapy; Irinotecan; Kaplan-Meier Estimate; Male; Metastases; Metastasis; Metastatic; Middle Aged; Monoclonal antibodies; Neoplasm Metastasis; Oxaliplatin; Patients; Phase III clinical trial; Predictive; Ramucirumab; Rank tests; Rectal Neoplasms - drug therapy; Rectal Neoplasms - mortality; Statistical models; Subgroups; Survival; Targeted cancer therapy; Treatment Outcome
• ISSN: 0959-8049; 1879-0852
• DOI: 10.1016/j.ejca.2017.03.007

Abstract Background The RAISE phase III clinical trial demonstrated that ramucirumab + (folinic acid plus 5-fluorouracil plus irinotecan) FOLFIRI significantly improved overall survival (OS) versus placebo + FOLFIRI for second-line metastatic colorectal carcinoma (mCRC) patients failing bevacizumab- and oxaliplatin-based chemotherapy (hazard ratio [HR] = 0.84, 95% CI = 0.73–0.98, P = 0.022). Post hoc analyses of RAISE patient data examined the association of carcinoembryonic antigen (CEA) subgroups with efficacy parameters. Methods CEA subgroups (≤10 versus >10 ng/ml) were based on 2X upper limit of normal (ULN) (5 ng/ml). The Kaplan–Meier method estimated the median OS and the progression-free survival (PFS). Log-rank test compared the survival distributions within the subgroups. Hazard ratio (HR) (95% confidence interval [CI]) and treatment-by-subgroup interaction p-values were calculated by Cox proportional hazards model. Results Ramucirumab treatment prolonged survival for the CEA ≤10 subgroup (HR = 0.68; 95% CI = 0.50–0.92; P = 0.013) and CEA >10 subgroup (HR = 0.90; 95% CI = 0.76–1.07; P = 0.233). However, the ramucirumab OS benefit over placebo was greater for the CEA ≤10 subgroup than for the CEA >10 subgroup (median OS: 3.6 versus 0.8 months greater, respectively). The interaction P-value between CEA level and treatment effect on OS was 0.088. This trend was observed across randomisation strata and to a lesser extent for PFS (P = 0.594). Conclusions Although patients in both high- and low-CEA subgroups derive OS and PFS benefits from ramucirumab treatment, the low baseline CEA level may identify a subgroup of patients with mCRC who obtain greater benefit from ramucirumab.