Impact of early palliative care on overall survival of patients with metastatic upper gastrointestinal cancers treated with first-line chemotherapy: Phase III EPIC trial

Abstract

Background: Palliative care (PC) has usually been offered at end‐life stage, although the World Health Organization recommends providing PC as early as possible in the course of malignancies. Temel et al. (N Engl J Med 2010) have shown that early PC (EPC) provides a favorable effect on quality of life, as well as a surprising benefit on overall survival (OS) (secondary endpoint of this trial) over standard treatment to patients (pts) with metastatic lung cancer. Median OS of pts with metastatic upper gastrointestinal (upGI) cancers does not exceed 10‐11 months, which is as poor as that reported with metastatic lung cancers. Whether or not OS benefit with EPC also applies to pts with metastatic upGI cancers is unknown. Demonstration of such benefit in these pts would lead to an earlier integration of PC in oncologic care. Trial design: EPIC is a randomized phase III trial. It is aimed to estimate the OS benefit of EPC combined with standard oncologic care over standard oncology care only, in pts with metastatic upGI cancers who start 1st‐line chemotherapy. Eligibility criteria also include ECOG PS≤2, and life expectancy > 4 weeks. Main exclusion criteria include: locally advanced tumors, esogastric cancers with unknown or positive HER2 status, dysphagia, and jaundice. Treatments will be randomized in a 1:1 ratio; a minimization procedure will be used to balance pts according to center, PS (0‐1 vs 2) and tumor location (esogastric/gastric, pancreas, and biliary tract). Pts will be recruited nationwide, in 17 university hospitals or cancer centers. OS will be used as a primary endpoint. The content of PC visits will be studied through a specific checklist. Patientreported outcomes (quality of life, depression and anxiety) will be also investigated. Assuming an exponential distribution of survival time, 381 deaths are required to ensure an 80%‐power for an absolute difference of 10% in one‐year OS (40% vs 50.3%, HR=0.75; logrank test two‐sided alpha=5%), leading to a planned sample size of 480 pts enrolled over 3 years. The main analysis will be performed on the intention‐to‐treat dataset. Enrollment began on September 2016.