Abstract 5479: PGx of gemcitabine in pancreatic ADK: Where did we go wrong

Abstract

Abstract There is a rising concern about how to limit toxicities and improve treatment outcome upon gemcitabine (GEM)-based treatment since evidence has been made that cytidine deaminase (CDA) is the major enzyme responsible for its inactivation into dFdU. CDA is subject to a wide inter-individual variability in terms of activity, due to genetic variations, leading to decreased or increased activity. We studied the effect of CDA phenotype and genotype on incidence of early severe hematologic toxicities upon GEM. We also conducted a pharmacokinetic (PK) study of GEM and dFdU. 120 patients with pancreatic ADK were enrolled in a prospective clinical trial FFCD-1004, all receiving GEM in monotherapy (1000mg/m2, 30-min IV infusion). Blood samples were collected prior to treatment to evaluate CDA status by a double approach: a phenotypic and a genotyping tests. PK study was conducted with a LSS: on 1rst course, samples were withdrawn prior administration, at the end of infusion, and 90 and 120 min after the end of infusion. GEM and dFdU levels were measured by UPLC-MSMS on plasma fraction. GEM and dFdU kinetics were analyzed using non-linear mixed effect modeling software Monolix 4.3.2®. 109 patients were analyzable for CDA activity, with a mean age of 66 years. 40 patients displayed early severe toxicities, including 26 hematological toxicities and 20 other toxicities. 39 kinetics could be collected. CDA activity range from 0.86 to 8.56 U/mg. Mean and median CDA activity were respectively 2.6 and 2.2 U/mg, which is not different from our previous study. There was not difference in CDA activity between patients with early severe hematological toxicities versus no toxicities. For rs2072671 SNP, 38 patients were wild-type, 41 heterozygous and 14 were homozygous variant (92 samples). Patients with a C/C genotype were not at higher risk of early severe toxicity. PK of GEM and dFdU was best described by a 3-compartment model, with a first order transfer rate. BSA was covariate of dFdU volume distribution and clearance of first dFdU compartment, whereas age was a covariate of volume of distribution of the second dFdU compartment. We evidenced CDA as a covariate of inter-compartmental clearance of dFdU, and not as a covariate of GEM to dFdU clearance as expected. This multicenter prospective clinical trial doesn't confirm predictive role of CDA status in occurrence of early severe toxicities in patients with pancreatic adenocarcinoma treated with GEM only. Genotypic data could neither explain discrepancies in toxicity. Furthermore, we designed a sparse data pharmacokinetic study allowing to estimate population and individual parameters of GEM from few samples. However, this trial doesn't reflect the wide inter-individual variability observed in routine. Selection in patients must introduce a bias, especially by selecting pancreatic ADK patients. Predictive role of CDA in occurrence of toxicity should be investigated in a larger cohort including various cancers. Citation Format: Cindy SERDJEBI, Florence GATTACCECA, Martina SCHNEIDER, Carole MONTERYMARD, Jaic CARIO, Karine LE MALICOT, Madani RACHID, Bruno LACARELLE, Jean-François SEITZ, Joseph CICCOLINI, Laetitia DAHAN. PGx of gemcitabine in pancreatic ADK: Where did we go wrong. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 5479. doi:10.1158/1538-7445.AM2015-5479