20078 Background: Pemetrexed is a novel anticancer agent clinically active in several solid tumors, including malignant pleural mesothelioma (MPM), non-small cell lung cancer (NSCLC), and breast cancer. Pemetrexed’s anticancer activity is believed to be related to its inhibition of 3 enzymes involved with folate metabolism and DNA synthesis. Some side effects associated with its clinical activity are more pronounced if a patient (pt) has a folate deficiency and high HC levels (Niyikiza et al, Ann Oncol 1998). Given this connection, we hypothesized that high HC levels might predispose a pt to pemetrexed anticancer activity. In a related question, we examined if baseline HC levels varied among different cancers. Methods: To investigate these hypotheses, plasma HC levels were examined from pts in the pemetrexed-containing arm of a phase III MPM trial (N = 226; Vogelzang, JCO 2003); a phase III NSCLC trial (N = 265; Hanna, JCO 2004); and a phase II breast cancer trial (data pooled from 600 and 900 mg/m2 groups) (N = 92; Llombart, ASCO 2006). For each study, HC (high vs low) was used as a prognostic indicator variable for overall survival (OS) and progression-free survival (PFS). Different cut points distinguished pts with high vs low HC, and final analysis results were based on the optimal cut point that yielded the largest Wald chi-square value (smallest p-value). Results: Average baseline HC levels were similar for each trial: 11.0, 10.6, and 10.6 μmol/L for the MPM, NSCLC, and breast trial, respectively. The MPM data suggested that high HC pts had longer OS (all hazard ratios [HRs] <1). The NSCLC data showed comparable OS times regardless of HC level. MPM and NSCLC data for PFS was similar to OS results. PFS from the breast cancer study was also similar for both HC groups (HR ∼1). Conclusions: Average baseline HC level did not suggest differences in nutritional status of the 3 pt populations. Data examination from the 3 studies yielded contradictory results, suggesting that high HC level is not a predictor of a better efficacy response to pemetrexed. [Table: see text] [Table: see text]