Efavirenz (EFV) is a nonnucleoside reverse transcriptase inhibitor and in combination with two nucleoside reverse transcriptase inhibitors (NRTI) is the preferred first-line antiretroviral therapy for HIV-infected children >3 years and weighting >10 kg.1
A concentration–response relationship for EFV has been reported. In adults, EFV plasma concentrations (collected 8–20 hours postdose) <1.0 mg/L were associated with virologic treatment failure and concentrations >4.0 mg/L were associated with central nervous system side effects.2 Among children initiating a first-line EFV-based regimen, a higher percentage of children with a minimum EFV concentration >1.1 mg/L had an HIV-1 RNA viral load decreases greater than 2 log10 copies/mL after 3 months of treatment compared with those below this threshold.3
Several studies have suggested that the US Food and Drug Administration (FDA)-approved weight-band dosing may not provide optimal drug exposure in all children. Two studies in Africa reported up to 40% of children with an EFV trough concentration (C24) < 1.0 mg/L.4,5 A study in Thailand found 15% of children with C24 < 1.0 mg/L.6 Indeed, WHO has proposed higher dosing (14 to <20 kg, 25 to <30 kg and 35 to <40 kg). The proportion of children below this target is also influenced by host genetic polymorphisms, such as the cytochrome P450 (CYP) 2B6 516G>T gene polymorphism, which is strongly associated with higher plasma concentrations.7
Using data from HIV-infected children in Thailand, we developed a population pharmacokinetic model to describe the EFV concentration-time course and studied the association between EFV concentrations and probability of HIV RNA viral loads >400 copies/mL. In addition, using simulations, we estimated the percentage of children with an EFV mid-dose concentration <1.0 mg/L while receiving US FDA weight-band dosing, and their probability of viral replication >400 copies/mL.
MATERIALS AND METHODS
EFV concentrations data were available from 188 children prescribed EFV initially following US FDA approved weight-band dosing: 148 with random sparse sampling within a cohort study (NCT00433030) and 40 with intensive sampling from a pharmacokinetic study.6 Fifty-three children were antiretroviral-experienced before their current EFV-based regimen but none had detected nonnucleoside reverse transcriptase inhibitor resistance mutations before EFV use. Seventeen children had previously received nevirapine before EFV and their treatment was switched for clinical reasons. Written informed consent was obtained from the parent/legal guardian. Children receiving medications known to interact with EFV, such as rifampicin, were excluded. Adherence was assessed by pill count. Ethics Committee approvals were provided at local and national levels. All children were prescribed 200–600 mg EFV, as capsules or tablets once a day to be taken in the evening without regard for food.
Measurement of EFV Plasma Concentrations
EFV plasma concentrations were determined using a validated high-performance liquid chromatography assay.7
Population Pharmacokinetic Analysis
Data were analyzed using the nonlinear mixed effect modeling program NONMEM (version VI) driven by Wings for NONMEM (http://wfn.sourceforge.net). Multiplicative and proportional error models were investigated to describe residual variability. An exponential model was used for interindividual variability. Covariates were tested via an upward–backward model building procedure. For evaluation of the goodness-of-fit, diagnostic graphics were obtained using RfN (http://wfn.sourceforge.net/) with R software. As individual CYP2B6 genotypes data were not available, a mixture model was used to identify subgroups of patients with difference clearance, that is, “fast” and “slow” metabolizers. Validation of the final population model was performed using a visual predictive check and a bootstrap resampling method.
Association Between EFV Concentrations and Viral Replication
Individual EFV C12 and C24 was estimated for each child. Fisher’s exact test was used to compare the proportion of children with a C12 above/below 1.0 mg/L and with/without viral replication. Using data from 145 children in the cohort study, a mixed effects logistic regression model including EFV concentration [C12 or C24] as a fixed effect and children as random effects was used to describe the risk of viral replication >400 copies/mL according to the child’s individual EFV concentration determined during the same period the viral load was performed.
We simulated a population with a weight distribution following that of 873 HIV-infected children participating in the PHPT pediatric HIV cohort. For each sex and 1 year age category between 3 and 18 years, 30 weights were generated from the normal distribution observed in this cohort. Pharmacokinetic parameters were estimated for children in the simulated population aged 3–18 years receiving US FDA-approved weight-band doses (100 Monte Carlo simulations per child). Individual EFV C12 and C24 concentrations were then derived from the pharmacokinetic parameters.
The percentage of children with C12 and C24 < 1.0 mg/L, 1.0–4.0 and >4.0 mg/L were calculated. Using the simulated C12 and C24 and the parameters of the logistic regression model derived to predict viral replication, the average probability of viral replication >400 copies/mL was determined for each weight band.
A total of 621 plasma concentrations from 188 HIV-infected children (47% male) were available for pharmacokinetic evaluation. At the time of first EFV concentration measurement, median (interquartile range) age was 10 years (7.5–12.7), weight 23.5 kg (18–34.5), height 125 cm (114–142) and EFV dose 13 mg/kg (11.4–14.3). Subjects with intensive and sparse sampling were not significantly different according to their characteristics at the time of pharmacokinetic assessment. Median duration of EFV treatment before pharmacokinetic assessment was 26.2 months (6.6–53.4). At time of EFV initiation, median CD4 cell count was 8% (2–16), HIV RNA load was 4.8 log10 copies/mL (3.8–5.4). Duration of EFV-based treatment was 5.1 years (3.8–6.3).
EFV Population Pharmacokinetics
A one-compartment model adequately described the data. The absorption phase was modeled using two transit compartments and residual variability using a proportional error model. A mixture model for EFV clearance, that is, two groups of patients separated into fast and slow metabolizers provided the best fit. Within the study population, 178 children (95%) were classified as fast metabolizers and 10 children were classified as slow metabolizers. Only body weight influenced apparent oral clearance (CL/F), and volume of distribution (Vd/F) and allometric scaling best described this relationship.
The final EFV population parameters (median, interindividual variability) were transit time 1.35 h−1 (0.59), CL/Ffast metabolizers 16.10 L/h/70 kg (0.43), CL/Fslow metabolizers 2.88 L/h/70 kg (0.43) and Vd/F 490 L/70 kg (0.79). Proportional residual variability was estimated to be 28%. The relative standard error was 6% for CL/Ffast metabolizers, 17% for CL/Fslow metabolizers and 19% for Vd/F. All other structural parameters had a relative standard error below 35% and variability parameters below 46%. The visual predictive check showed acceptable overlap of the observed data and modeled concentration profiles, including at C12 and C24; and the bootstrap (1000 runs) provided acceptable population parameter estimates (See Fig., Supplemental Digital Content 1, http://links.lww.com/INF/C215).
For the entire study population (independently of metabolizer phenotype), the median (interquartile range) EFV AUC0–24 was 44.4 (31.0–56.4) mg·h/L, C12 3.1 (2.0–4.1) mg/L and C24 1.2 (0.7–1.6) mg/L. Overall, 15 of 145 children (10%) had a predicted C12 < 1.0 mg/L. Eight of the 15 children (53%) with a C12 < 1.0 mg/L had viral replication compared with 27 of 130 (21%) children with a C12 ≥ 1.0 mg/L (P = 0.01).
Association Between Drug Concentrations and Viral Replication
The risk of viral replication >400 copies/mL increased with lower mid-dose and trough concentrations. The odds of viral replication increased by 1.8 for each 0.5 mg/L decrease in EFV C12 (odds ratio: 1.8, 95% confidence interval: 1.5–2.2; P = 0.03), while it increased by 2.1 for each 0.5 mg/L decrease in EFV C24 (odds ratio: 2.1, 95% confidence interval: 1.4–3.0; P = 0.03).
Proportion of Children with Subtherapeutic Concentrations Using US FDA Weight-band Dosing and Probability of Viral Replication
Overall, based on the simulated population of children weighing 10–75 kg, it was predicted that on average 15% had a C12 < 1.0 mg/L and among these children the average probability (%) of viral replication >400 copies/mL was 23%. Figure 1A, B shows the percentage of children with a C12 <1.0, 1.0–4.0 and >4.0 mg/L for fast and slow metabolizers. The probability of viral replication for children with a C12 within each concentration category is also presented. The model estimated that on average 15% of fast and <1% of slow metabolizers had a C12 < 1.0 mg/L across the weight bands. Fast metabolizers had an average probability of viral replication >400 copies/mL of 9% over the weight range while slow metabolizers had a probability of <1% (Fig. 1C, D).
In our cohort, a higher proportion of children with a C12 < 1.0 mg/L had viral replication compared to those with a C12 > 1.0 mg/L, and a lower C12 and C24 were associated with a higher odds of viral replication. Our pharmacokinetic model predicted that approximately 15% of children would have a C12 < 1.0 mg/L when receiving US FDA-approved weight-band dosing (<1% for children with a slow metabolizer phenotype) and these children had a 23% risk of viral replication. EFV C12 was <1.0 mg/L in 14% of 87 HIV-infected children on EFV-based regimen (median dose 11.4 mg/kg) in Rwanda; however, no association between EFV concentrations and virologic failure was observed.8
EFV pharmacokinetics in our population were consistent with previous reports.3,4 The recent results of the ENCORE 1 trial demonstrated noninferiority in the time-to-loss of virological response of 400 mg versus 600 mg EFV in antiretroviral naive adults but an advantage in terms of tolerance.9,10 As expected, a significantly higher proportions of patients had a C12 < 1.0 mg/L receiving 400 mg EFV compared with 600 mg (25% vs. 6%)10 indicating that a high proportion of patients with a C12 < 1.0 mg/L maintain viral suppression. Indeed, the original article by Marzolini et al2 identifying this target threshold reported that 50% of patients with EFV concentrations <1.0 mg/L (8–20 hours postdose) had virologic failure. Thus, this target may need refinement.
Our data are reassuring regarding the efficacy of US FDA dosing and there are already extensive safety data, although it is important to note that only 3% of our population weighed 10–15 kg and/or were <5 years of age, and our predictions may be less reliable in this subpopulation. The latest WHO guideline recommends higher EFV dosing in three weight bands (14 to <20 kg, 25 to <30 kg and 35 to <40 kg). It is unclear whether this change will help significantly decrease the risk of virologic failure but it may increase the risk of toxicities in children, especially for slow metabolizers. Thus, there may be a need to assess the efficacy and safety of the higher EFV dosing guidelines in settings where access to pharmacogenetic testing is not widespread.
We thank all members of the PHPT, Siriraj hospital and Chiang Mai University hospital teams and the patients who were enrolled. Site staffs: Rawiwan Hansudewechakul (Pediatrician), Chiangrai Prachanukroh Hospital; Suparat Kanjanavanit (Pediatrician), Nakornping Hospital; Chaiwat Ngampiyaskul (Pediatrician), Prapokklao Hospital; Suchat Hongsiriwon (Pediatrician), Chonburi Hospital; Prapaisri Layangool (Pediatrician), Bhumibol Adulyadej Hospital; Pornchai Techakunakorn (Pediatrician), Phayao Provincial Hospital; Pimpraphai Thanasiri (Pediatrician), Samutsakhon Hospital; Sakulrat Srirojana (Pediatrician), Kalasin Hospital; Pornpun Wannarit (Pediatrician), Lamphun Hospital; Noppadon Akarathum (Pediatrician), Sanpatong Hospital; Boonyarat Warachit (Pediatrician), Hat Yai Hospital; Achara Puangsombat (Pediatrician), Samutprakarn Hospital; Warit Karnchanamayul (Pediatrician), Rayong Hospital; Sookchai Theansavettrakul (Internist), Phan Hospital; Sudanee Buranabanjasatean (Internist), Mae Chan Hospital; Preecha Sirichithaporn (Hospital Director), Doi Saket Hospital; Maneeratn Nantarukchaikul (Pediatrician), Somdej Prapinklao Hospital; Temsiri Hinjiranandana (Pediatrician), Queen Sirikit Hospital; Narong Lertpienthum (Pediatrician), Buddhachinaraj Hospital; Pornsawan Attavinijtrakarn (Pediatrician), Phaholpolpayuhasaena Hospital; Airada Saipanya (Internist), Chiang Dao Hospital; Sansanee Hanpinitsak (Pediatrician), Regional Health Promotion Centre 6, Khon Kaen; Sirisak Nanta (Pediatrician), Mae Sai Hospital; Sathaporn Na-Rajsima (Pediatrician), Mahasarakam Hospital; Narongdej Pipattanawong (Internist), Sankhampang Hospital; Ratchanee Kwanchaipanich (Pediatrician), Bhuddasothorn Hospital; Paiboon Lucksanapisitkul (Head of Pediatrician), Pranangklao Hospital; Weerasak Lawtongkum (Pediatrician), Vachira Phuket Hospital; Worawut Cowatcharagul (Hospital Director), San Sai Hospital. PHPT: P. Sukrakanchana, S. Chalermpantmetagul, C. Kanabkaew, R. Peongjakta, A. Piyalap, B. Insuyah, B. Ratchanee, J. Chalasin, J. Thonglo, N. Kruenual, N. Krueduangkam, N. Krapunpongsakul, R. Kaewsai, R. Wongchai, S. Thammajitsagul, S. Jinasa, T. Thimakam, W. Khamjakkaew; Laboratory: N. Ngo-Giang-Huong, P. Punyathi, W. Sripaoraya, Y. Tawon; PHPT Data Center: L. Decker, B. Caritey, S. Tanasri, S. Chailoet, R. Jitharidkul, R. Suaysod, W. Wongwai; Administrative support: P. Pirom; Tracking & Supplies: K. Than-in-at, A. Rakpor; Drug distribution center: D. Chinwong. Siriaj Hositpal:K. Chokephaibulkit, O. Wittawatmongkol, S. Sricharoenchai, N. Kongstan, S. Thongnok,S. Malawong. Research Institute for Health Sciences (RIHES): V. Sirisanthana (site PI), T. Sudjaritruk (Pediatrician), N. Wongnum (Research Assistant), C. Khamrong (Study Coordinator), J. Chanthong (Study Nurse), R. Thummalangka (Study Nurse).
1. World Health Organization. Antiretroviral therapy of HIV infection in infants and children
: Toward University Access, Recommendations for a public health approach: 2010 revision 2010 Available at: http://www.who.int/hiv/pub/paediatric/infants2010/en/index.html
. Accessed February 12, 2013.
2. Marzolini C, Telenti A, Decosterd LA, et al. Efavirenz plasma levels can predict treatment failure and central nervous system side effects in HIV-1-infected patients. AIDS. 2001;15:71–75
3. Hirt D, Urien S, Olivier M, et al. Is the recommended dose of efavirenz optimal in young West African human immunodeficiency virus-infected children
? Antimicrob Agents Chemother. 2009;53:4407–4413
4. Fillekes Q, Natukunda E, Balungi J, et al.ARROW Trial Team. Pediatric underdosing of efavirenz: a pharmacokinetic study in Uganda. J Acquir Immune Defic Syndr. 2011;58:392–398
5. Ren Y, Nuttall JJ, Egbers C, et al. High prevalence of subtherapeutic plasma concentrations of efavirenz in children
. J Acquir Immune Defic Syndr. 2007;45:133–136
6. Cressey TR, Aurpibul L, Narkbunnam T, et al. Pharmacological assessment of efavirenz weight-band dosing recommendations in HIV-infected Thai children
. J Acquir Immune Defic Syndr. 2013;62:e27–e29
7. Puthanakit T, Tanpaiboon P, Aurpibul L, et al. Plasma efavirenz concentrations and the association with CYP2B6-516G >T polymorphism in HIV-infected Thai children
. Antivir Ther. 2009;14:315–320
8. Mutwa PR, Fillekes Q, Malgaz M, et al. Mid-dosing interval efavirenz plasma concentrations in HIV-1-infected children
in Rwanda: treatment efficacy, tolerability, adherence, and the influence of CYP2B6 polymorphisms. J Acquir Immune Defic Syndr. 2012;60:400–404
9. Puls R, Amin J, Losso M, et al. Efficacy of 400 mg efavirenz versus standard 600 mg dose in HIV-infected, antiretroviral-naive adults (ENCORE1): a randomised, double-blind, placebo-controlled, non-inferiority trial. Lancet. 2014;383:1474–1482
10. Dickison L, Puls R Efavirenz (EFV) 400 versus 600 mg daily: Results of the ENCORE1 Intensive PK Sub-Study. 2014 USA 21st Conference on Retroviruses and Opportunistic Infections (CROI 2014). Boston, MA