|
|
|
|
|
Session 22
Oral Abstracts Clinical Pharmacology: New Agents, Interactions, and Predictors of Virologic Response Thursday, 10 am - 12:30 pm Presentation Time: 11:00 am Ballroom A |
Background: Efavirenz (EFV) and nelfinavir (NFV) are metabolized primarily by cytochrome P4502B6 (CYP2B6) and CYP2C19, respectively. In a recent 24-week analysis involving 154 subjects (NWCS214), a single nucleotide polymorphism (SNP) more frequent in blacks than in whites (CYP2B6 516G>T) was associated with sustained plasma EFV elevations and week-1 central nervous system side effects. To better define the relevance of this and other SNP for longer-term response to antiretroviral therapy, and for EFV and NFV pharmacokinetics, NWCS213 used data from a different large, randomized trial with up to 3-year follow-up (ACTG 384).
Methods: We studied
antiretroviral-naïve subjects who received EFV and/or NFV plus 2 nucleoside
analogues (zidovudine [AZT]/lamivudine
[3TC] or stavudine [d4T]/didanosine
[ddI]) in ACTG 384, and contributed specimens to the
AACTG Human DNA Repository under Protocol A5128. Population pharmacokinetic parameters were estimated from a
non-linear mixed effects model assuming one compartment and first-order
absorption. A total of 9 SNP in CYP2B6, 3A4, 3A5, 2C19 and
MDR1 were characterized by oligonucleotide ligation. Univariate comparisons were carried out using
non-parametric tests.
Results: The 504 study subjects included 340 EFV and 348 NFV recipients
(184 received both EFV and NFV), 246 (49%) whites, 155 (31%) blacks, and 93
(19%) Hispanics. CYP2B6 516G>T was
associated with higher plasma EFV AUC24h values in all subjects and
in white, black, and Hispanic subpopulations (p ≤ 0.001 for all 4 comparisons). Among all subjects, median
EFV AUC24h according to G/G, G/T, and T/T genotype was 49 (n = 187),
58 (n = 148), and 101 (n = 32) μg*h/L, respectively. Similarly, CYP2C19
681G>A was associated with higher plasma NFV levels in all subjects (p < 0.001) and in each subpopulation
(all p < 0.01). Median NFV AUC12h
according to G/G, G/A, and A/A genotype was 25 (n = 252), 31 (n = 85), and 34
(n = 10) μg·h/L,
respectively. CYP3A4, 3A5, and MDR1 SNP were not
significantly associated with EFV or NFV pharmacokinetics. By univariate analysis, no single SNP was consistently
associated with time to viral load, toxicity-related, or all-cause initial
treatment failure over 3 years of follow-up, or CD4 increases at 48 weeks.
Conclusions: Despite strong associations between genetics variants and plasma pharmacokinetics, we did not find significant associations between CYP2B6 or CYP2C19 variants and long-term responses to EFV- or NFV-containing regimens in this dataset. Further analyses of these SNP to consider gene-gene interactions and other outcome variables are warranted.
Keywords: pharmacogenetics; cytochrome P450; MDR1
![]() |