Ron Keizer (1), John Fors (1), Terry Blaschke (2,3), Steven Kern (3), Rada Savic (1)
(1) Department of Bioengineering and Therapeutic Sciences, UCSF, San Francisco, CA; (2) Division of Clinical Pharmacology, Department of Medicine, Stanford University, Stanford, CA; (3) Bill & Melinda Gates Foundation, Seattle, WA
Objectives: To increase access to antiretroviral drugs in sub-Saharan Africa it has been proposed to use efavirenz (EFV) at a reduced dose (400 mg vs 600 mg). However, antiretroviral drugs require high patient adherence to achieve therapeutic effect and avoid viral resistance, especially if co-treated with tuberculosis (TB) drugs. Using simulations we investigated whether reduced-dose efavirenz can obtain similar treatment effectiveness as standard dose.
Methods: The PK model included effects of genotypes of CYP2B6 and N-Acetyl-transferase type 2 (NAT2), and co-medication with TB drugs.[1] To correct for the auto-inductive effects of efavirenz, we speculated that at 400 mg compared to 600 mg, CLss is 25% lower, but that the inductive effect of CYP2B6 could be increased up to 100% by TB treatment. The PK model was linked to a model describing viral dynamics.[2-4] Adherence was implemented either as a Markov model or by sampling actual MEMS-caps data from HIV-treated patients. Simulations were performed in R, with ODEs solved numerically in C++/Boost. Levels of adherence that were investigated were “100%”, “80%”, “50%” and “MEMS”. Variables recorded in the simulations were the plasma concentrations over time (PK), the % of time plasma concentration
Results: Most patients with CYP2B6 519TT genotype achieved treatment success, irrespective of dose or TB co-treatment. However, even at 100% adherence, in patients at highest risk of sub-therapeutic dose (519GG patients with fast NAT2, co-treated with TB), treatment success was achieved for only 45% of patients at 400 mg, vs 62% at 600 mg. With “MEMS” adherence, treatment success dropped to 33% vs 49%, respectively, which was slightly worse than at 80% adherence. Overall, patients with 519GT genotypes achieved slightly higher success rates than 519GG patients.
Conclusions: Our results indicate that reduced-dose EFV is safe only for a subgroup with CYP2B6 519TT genotype and not for the 519GG and 519GT subgroups, especially when co-treated for TB. As patients are often co-treated for TB, and HIV therapy programs in sub-Saharan Africa rarely achieve high adherence levels, there is considerable risk and possibly increased costs associated with reduced-dose EFV.
References:
[1]. Bertrand et al. JID 2014:209
[2]. Perelson and Nelson. SIAM Rev1999
[3]. Piana et al. ESPACOMP 2011
[4]. Wu et al. JAIDS 2005-39(3)
Reference: PAGE 23 () Abstr 3246 [www.page-meeting.org/?abstract=3246]
Poster: Drug/Disease modeling - Infection