The relationship between pemetrexed exposure (AUC) and clinical outcome in Japanese NSCLC patients

Abstract

18041 Background: The efficacy of pemetrexed (PEM) 500 mg/m2 administered with vitamin B12 and folic acid (FS) has been established in previously treated NSCLC. The 500 mg/m2 dose is based on the MTD determined for PEM without FS. Recently completed phase I studies have shown the recommended dose of PEM single agent with FS to be 900–1,000 mg/m2. A randomized phase II study conducted in Japan (ASCO 2007) showed PEM 500 mg/m2 and 1,000 mg/m2 with FS to have similar efficacy for patients with 2nd and 3rd line NSCLC. PEM is renally excreted and a PK model characterizing the relationship between CrCL and PEM AUC has been previously established. The exposure response relationships (ERRs) between AUC and progression free survival (PFS), individual patient’ tumor response, and toxicity were explored in these analyses. Methods: Eligible patients had a histologic or cytologic diagnosis of NSCLC, had been previously treated by 1 or 2 chemotherapy regimens, and were randomized to 500 mg/m2 (N = 108 pts) or 1,000 mg/m2 (N = 108 pts) of PEM once every 3 weeks. A previously established PK model was used to estimate AUC based on patient CrCL and dose administered. AUCs were linked to clinical outcome to identify ERRs. Efficacy models assumed a distribution of events with hazard dependent on prognostic factors and mean AUC at cycle over the treatment period. The AE model assumed probability of toxicity as a function of AUC at each cycle. Results: Median CrCL for patients in 500 mg/m2 and 1,000 mg/m2 groups were 85 ml/min (42–167) and 83 ml/min (38–146), respectively; median estimated PEM AUCs were 158 mg.h/L (107–221) and 318 mg.h/L (215–468). The relationship between model-estimated AUC and PFS was not significant. This is consistent with the clinical findings: PFS for the 500 mg/m2 and 1,000 mg/m2 groups of 3.0 months and 2.4 months, respectively; overall response rate (RR) of 18.5% and 14.8%, respectively. ERR analyses for individual patient tumor response, neutropenia and thrombocytopenia are ongoing at this time. Conclusions: Estimated PEM AUCs, based on CrCL, does not correlate with PFS for pretreated NSCLC at 500 mg/m2 and 1,000 mg/m2. The ERRs for individual tumor response and toxicity are ongoing and will be presented at the meeting. No significant financial relationships to disclose.