260   Predictive Power of P- Glycoprotein and CYP2D6 Polymorphisms for Plasma Levels of Antiretroviral Agents.

J. Fellay*, C. Marzolini, G. Greub, T. Buclin, C. B. Eap, and A. Telenti.
CHUV. Lausanne, Switzerland.

Background:Important inter-individual variability is observed in drug plasma levels among HIV-infected patients receiving antiretroviral treatment (ART), which may explain differences in toxicity and efficacy. Numerous allelic variations have been described within genes coding for P-glycoprotein (PGP) and enzymes of cytochrome P450 (CYP), which are important in the metabolism of ART. We analyzed the association between various alleles and drug plasma levels.

Methods:We studied 63 Caucasian patients with plasma viremia <400 copies/ml while receiving a PI- or Efavirenz-containing regimen. They were distributed in tertiles according to their drug plasma levels. PGP (promoter and exon 26), CYP3A4 (promoter and exon 7), CYP2C19 (exon 4 and 5) and CYP2D6 (selected alleles and gene duplication) were analysed.

Results:Homozygosity CC at PGP codon 1145 (exon 26) was present in 12 (44%) of 27 patients with high drug levels as compared to 3 (8.3%) of 36 subjects with lower levels (p = 0.002). Homozygosity TT at the same codon was present in 7 (44%) of 16 patients with low drug levels as compared to 6 (12.8%) of 47 subjects with higher levels (p = 0.014). For CYP2D6, 10 (37%) of 27 patients with high levels were homo- or heterozygous for at least one allele defining the poor metabolizer (PM) phenotype, as compared to 3 (8.3%) of 36 subjects with lower levels (p = 0.01). The only patient with a multicopy CYP2D6 gene (ultrarapid metabolizer = UM) had low levels. No other polymorphism in analysed genes was correlated to drug plasma levels. Overall, in the low drug level group, 50% of patients had either a TT allele (PGP) or UM genotype (CYP2D6); in the high level group, 63% had a CC and/or PM genotype. In multivariate logistic regression adjusted for age and sex, the odds ratio of having a CC allele (PGP) and/or PM genotype (CYP2D6) was 2.2 (CI95: 1.5—3.4) for the high drug level group. For the low level group, the odds ratio of having a TT allele (PGP) and/or UM genotype (CYP2D6) was 8.4 (2— 35.4).

Conclusions:The marked inter-patient variability in drug plasma levels could be explained to a significant extent by the allelic pattern of two genes involved in the metabolism or transport of antiretrovirals. This information could be of potential use to better predict drug levels, treatment failure and toxicity.