Safety, efficacy, and pharmacokinetics (PK) of TS-1 combined with cisplatin in advanced cancer patients with severe hepatic dysfunction.

Abstract

e15001 Background: Advanced cancer patients with severe hepatic dysfunction have limited systemic therapy options due to impaired drug metabolism and increased toxicity risks and are often excluded from clinical trials. Few cancer drugs are systematically studied for safety in this population. While cisplatin and 5-fluorouracil (5FU) are considered safe in hepatic dysfunction, TS-1, an oral 5FU prodrug activated by the liver, lacks formal safety data and is not recommended in severe hepatic dysfunction. We evaluated the safety, efficacy and PK of TS-1 in combination with cisplatin in cancer patients with severe hepatic dysfunction and compared them to those with normal, mild and moderate dysfunction. Methods: In this phase II study, patients with advanced solid tumors with no further standard treatment options were eligible. Patients were stratified into 4 cohorts by liver function (NCI-ODWG criteria): (1) normal (total bilirubin [TB] & ALT/AST≤ULN); (2) mild impairment (TB>1-1.5xULN & ALT/AST≤5xULN); (3) moderate impairment (TB>1.5-3xULN & ALT/AST≤5xULN), and (4) severe impairment (TB>3-10xULN and/or ALT/AST>5-20xULN). Cohort 4 had a larger sample size to account for higher attrition. Patients received cisplatin 60mg/m 2 3-weekly with TS-1 30mg/m 2 BID (days 1-14) until progression or intolerable toxicity. Primary endpoints were PK and safety; secondary objectives were objective response rate (ORR), 3-month clinical benefit rate (3M-CBR) and progression-free survival (PFS). Results: 61 patients were enrolled (Cohorts 1: 2: 3: 4 = 11: 10: 13: 27). Median age was 55 years (23-75); 74% were female. Breast cancer was the most common tumor type (57%). PK analysis showed lower 5FU C max and longer T max in Cohort 4 compared to Cohort 1 (p<0.05), reflecting slower activation of TS-1 to 5FU in severe hepatic dysfunction. However, 5FU AUC was similar across the 4 cohorts. Safety and efficacy signals observed in Cohort 4 were comparable to that seen in Cohorts 1-3 (Table 1). No grade 5 treatment-related adverse events (TRAEs) occurred. Notably, 22.2% of patients in Cohort 4 achieved PFS ≥6 months. PK changes did not correlate with clinical benefit or toxicity, suggesting that severe hepatic dysfunction did not lead to lower efficacy or greater toxicity to TS-1. Conclusions: Patients with severe hepatic dysfunction had similar 5FU AUC and toxicities from cisplatin + TS-1 as those with normal or less severe hepatic dysfunction despite slower activation of TS-1 to 5FU. 22.2% had PFS ≥6 months, supporting this combination as a clinically viable treatment for these patients who often have no further cancer treatment options. Clinical trial information: NCT03519074 . Key TRAEs and efficacy outcomes in all cohorts. Cohort 1 2 3 4 G3/4 TRAEs (%)  Neutropenia 36 20 15 22  Anemia 9 20 15 11  Thrombocytopenia 0 0 15 11 Efficacy  ORR (%) 18.2 20.0 0 22.2  3M-CBR (%) 54.5 60.0 15.4 40.7  Median PFS (months) 3.0 3.0 1.8 2.8  PFS ≥ 6 months (%) 18.2 10.0 7.7 22.2