759 
Pharmacogenetics of Plasma Drug Exposure and Treatment Outcomes with Efavirenz-containing Regimens: An ACTG Study
David Haas*1, H Ribaudo2, A Motsinger3, B Schackman4, R Gulick4, E Acosta5, M Schwab6, E Schaeffeler6, G Morse7, G Robbins8, and The AIDS Clinical Trials Group
1Vanderbilt Univ, Nashville, TN, US; 2Statistical and Data Analysis Ctr, Harvard Sch of Publ Hlth, Boston, MA, US; 3North Carolina State Univ, Raleigh, US; 4Weill Med Coll of Cornell Univ, New York, NY, US; 5Univ of Alabama at Birmingham, US; 6Margarete Fischer-Bosch Inst for Clinical Pharma, Stuttgart, Germany; 7State Univ of New York at Buffalo, US; and 8Harvard Univ, Boston, MA, US
Background: Efavirenz (EFV) is metabolized by cytochrome P450
(CYP) 2B6. Single nucleotide polymorphisms (SNP) in CYP2B6 delay EFV
clearance, whereas ABCB1 SNP have been associated with treatment
response. We characterized relationships between SNP, pharmacokinetics, and
response to EFV-containing regimens.
Methods: Associations between EFV pharmacokinetics and 22
SNP in CYP2B6, ABCB1 and CYP3A5 were investigated among
individuals with pharmacokinetic data from the AIDS Clinical Trials Group (ACTG)
protocol A5097s (pharmacokinetic substudy of A5095) and ACTG 384. Among
subjects randomized to EFV in A5095 we assayed CYP2B6 516G>T,
983T>C, and 5 SNP in ABCB1. We defined CYP2B6 516/983
metabolizer genotype as rapid (516GG+983TT); intermediate (516GT or 983TC);
slow (516TT, 983CC, or [516GT+983TC]). Stepwise selection models investigated
genotype associations with time to: EFV discontinuation; first central nervous
system (CNS) adverse event; and virologic failure. Analyses adjusted for body
mass index, and self-reported adherence at week 12 (virologic failure only).
Genotyping was by MALDI-TOF (A5097s, ACTG 384) and Taqman (A5095).
Results: Of 22 SNP assayed in A5097s (n = 123) and
ACTG 384 (n = 174), pharmacokinetic associations were
dominated by CYP2B6 516G>T and 983T>C. Other alleles were
infrequent (<2%) or had little independent association with pharmacokinetics.
Of 765 subjects randomized to EFV-containing regimens in A5095, 287 whites, 225
blacks, and 131 Hispanics were genotyped. Among whites, slow metabolizer CYP2B6
genotype (compared to rapid) was associated with an increased hazard of CNS adverse
events (HR = 2.40; 95%CI 1.07 to 5.37; p = 0.034) and EFV discontinuation
(2.18; 1.03 to 4.63, p = 0.04), but not virologic failure (1.16; 0.27 to
4.94, p = 0.85). Among blacks with slow metabolizer genotype, there was
some evidence of a lower hazard of virologic failure (0.34; 0.12 to 1.00, p =
0.050), but not CNS adverse events (1.18; 0.51 to 2.71; p = 0.69) or EFV
discontinuation (0.69; 0.38 to 1.26; p = 0.23). Among Hispanics,
significant associations with slow metabolizer genotype were not seen for virologic
failure (1.0; p = 0.99), CNS adverse events (0.98; 0.37 to 2.62; p =
0.97), or EFV discontinuation (0.90; 0.19 to 4.32; p = 0.89). A
previously reported association between ABCB1 3435C>T and decreased
hazard of VF was not seen.
Conclusions: Associations with EFV pharmacokinetics were
dominated by CYP2B6 516G>T and 983T>C. Relationships between SNP
in CYP2B6 and ABCB1, CNS AEs, EFV discontinuation, and virologic
failure were not consistently identified among populations. A unifying model
that relates pharmacogenetics to EFV treatment response remains elusive, and may
be affected by additional genetic or non-genetic factors.
|