Quality of life (QOL) evaluation within a randomized phase III study of gemcitabine plus S-1 (GS) versus S-1 versus gemcitabine (GEM) in unresectable, advanced pancreatic cancer (PC) in Japan and Taiwan: GEST study.

Abstract

9070 Background: The GEST study, a randomized 3-arm phase III study, was designed to assess the non-inferiority of S-1 alone and the superiority of GS to GEM. Primary endpoint was overall survival (OS). The non-inferiority of S-1 was confirmed, GS was superior to GEM but the difference was not statistically significant. QOL was a secondary endpoint. Here we report QOL results from this study. Methods: Chemotherapy-naïve patients with unresectable advanced PC, ECOG PS of 0-1, with adequate organ functions were enrolled. Patients were equally randomized to one of three arms: GEM (1000 mg/m2, iv, d1, 8 and 15, q4w) , S-1 (80/100/120 mg/day based on BSA, po, d1-28, q6w) , or GS (GEM 1000 mg/m2, iv, d1 and 8 plus S-1 60/80/100 mg/day based on BSA po, d1-14, q3w). The patients were assessed with EQ-5D questionnaire at baseline, and at 6,12,24,48 and 72 weeks from first administration. Changes over time of EQ-5D utility score and Quality-adjusted life years (QALY) were calculated. In addition, we explored the relationship between QOL and clinical characteristics. Results: Among 832 patients, full analysis set of this study, 736(88%) had completed both baseline questionnaire and at least one questionnaire after randomization. In the EQ-5D utility score analysis by including deaths as score 0, there was a statistically significant difference in score over time between GS and GEM (p=0.0031, repeated measures ANOVA), and no statistically significant difference between S-1 and GEM. In the analysis of QALY, there was a statistically significant difference between GS and GEM (p=0.0008, generalized Wilcoxon test), but no statistically significant difference between S-1 and GEM. In the population of PS 1, GS is remarkably better than GEM in favor of EQ-5D utility score and QALY. Conclusions: Patients treated with S-1 resulted in clinically equivalent efficacy outcomes without any impact on patient QOL compared with GEM. Patients treated with GS had a statistically significant difference in QOL compared with GEM, in both overall and PS1 populations. Further research is needed to confirm the efficacy of GS in unresectable advanced PC.