Trial Comparing Two Two Sequences of Therapy in Colorectal Metastatic Patients

Target population: Patients with histologically confirmed metastatic colorectal cancer progressed after a first line treatment containing FOLFIRI and BEV Inclusion criteria: * Age \>18 \< 75 years of age * Diagnosis of histologically proven adenocarcinoma of the colon or rectum, stage IV * K-ras wild-type * ECOG performance status 0-1 at study entry Endpoints: \- Response Rate, Disease control rate, The duration of overall response, Overall survival, PFS, Time to treatment failure, Quality of Life, Incidence of AEs, Frequency and nature of serious adverse reactions (SADRs), Premature withdrawals Statistical methods: Assuming a randomization ratio of 1:1, 282 deaths are required in order to achieve a power of 80% of detecting a hazard ratio of 0.72 in favour of one of the two sequences, translating in an increase of median survival time from 10 to 14 months, with a type I error of 5%, two-sided, using the Mantel-Cox version of the log-rank test. With a uniform accrual period of 3 years and a follow-up of 18 months, about 350 patients will be needed to reach the target number of events. All statistical analyses will be based on an intention-to-treat approach. CONSORT rules will be applied to describe study flow and protocol deviations. All OS and PFS curves will be drawn with the Kaplan-Meier method. Results will be presented as Hazard Ratio (HRs) and their 95% Confidence Interval (CIs). On annual basis, starting from the second year, an interim analysis will be conducted. In principle, no formal stopping rule will be applied, unless otherwise suggested by the DSMC. Safety reports will be drawn on annual basis.