Molecular Biomarker Study in a Randomised Phase III Trial of Irinotecan Plus S-1 versus S-1 for Advanced Gastric Cancer (GC0301/TOP-002)

Aims: Gastric cancer is a common and heterogeneous disease; however, global standard and biomarkers for selecting chemotherapy regimens have not been established. This study was designed retrospectively to identify molecular biomarkers for irinotecan plus S-1 (IRI-S) and S-1 therapy from subset analyses in GC0301/TOP-002, a randomised phase III trial for advanced gastric cancer. Materials and methods: Paraffin-embedded primary tumour specimens were collected from 126 of 326 randomised patients in GC0301/TOP-002. The mRNA was measured for thymidylate synthase, dihydropyrimidine dehydrogenase, topoisomerase I, excision repair cross-complementing gene 1 (ERCC1) and thymidine phosphorylase; categorised into low and high to analyse their association with efficacy end points. Results: There was no significant difference in each mRNA between S-1 and IRI-S groups, whereas there were differences among some clinical characteristics. Multivariate analyses for overall survival showed that mRNA levels were not correlated with prognosis. By comparison, between IRI-S and S-1 arms, low thymidylate synthase, low ERCC1 and high thymidine phosphorylase were associated with better prognosis for IRI-S versus S-1 (hazard ratio = 0.653, 0.702 and 0.709, respectively; P < 0.15 for each interaction). Conclusion: Low thymidylate synthase, low ERCC1 and high thymidine phosphorylase are candidates for predictive biomarkers for first-line treatment in advanced gastric cancer by IRI-S. Further study is warranted to confirm these results in other clinical trials and cohort studies. (C) 2016 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.