Whether there is a difference in the efficacy of maintenance treatment for metastatic colorectal cancer (mCRC) between patients who achieve complete response (CR)/partial response (PR) and those with stable disease (SD) after induction treatment is controversial. PubMed, Cochrane Systematic Reviews, the Cochrane Collaboration Central Register of Controlled Clinical Trials, ClinicalTrials.gov, and databases of conferences were queried to identify randomized controlled trials evaluating the efficacy of maintenance treatment for mCRC patients. The search included articles dated from the inception of these resources until June 20, 2017. We estimated hazard ratios (HRs) for progression-free survival (PFS) and overall survival (OS). Network meta-analysis was performed to compare the efficacy of four regimens as maintenance treatment. Three randomized controlled trials comprising 1,301 patients were included in this network meta-analysis. Patients who achieved CR/PR after induction therapy benefited more from maintenance treatment than patients who achieved SD (PFS: HR [CR/PR] 1.50, 95% CI 1.09-2.08, vs. HR [SD] 1.35, 95% CI 1.04-1.74; OS: HR [CR/PR] 1.04, 95% CI 0.94-1.15, vs. HR [SD] 1.03, 95% CI 0.99-1.07). The results of network meta-analysis suggested that chemotherapy alone and observation were inferior to chemotherapy plus bevacizumab as maintenance treatment. Patients with mCRC who achieve CR/PR after induction therapy might benefit more from maintenance treatment than those with SD. Chemotherapy plus bevacizumab was the most appropriate regimen for maintenance treatment.
Keywords: complete or partial response; maintenance treatment; metastatic colorectal cancer; network meta-analysis; stable disease.