Pharmacokinetics of nevirapine in HIV-infected children receiving an adult fixed-dose combination of stavudine, lamivudine and nevirapine

Objective: To evaluate the steady state pharmacokinetics of nevirapine (NVP) in HIV-infected children receiving a fixed-dose combination of stavudine, lamivudine and NVP.

Methods: This cross-sectional study enrolled 34 children (18 girls) who had received GPO-VIR S30 (30 mg stavudine, 150 mg lamivudine and 200 mg NVP) for at least 8 weeks. Tablets were divided into quarter fractions (1/4, 1/2, 3/4 or 1 tablet) to attain the NVP dosages of 120-200 mg/m² every 12 h. Plasma NVP levels were measured at predose, and at 2 and 6 h after drug administration.

Results: The median age was 8.4 years (range, 3-15). Median CD4 lymphocyte count and percentage at study entry was 576 × 10⁶ cells/l and 20.25%, respectively. The median pharmacokinetics parameters were area under the curve at 12 h, 78.4 h × μg/ml; minimum plasma drug concentration, 5.98 μg/ml; plasma half-life, 25.5 h; apparent oral clearance, 0.079 l/kg per h; and volume of distribution, 2.95 l/kg. Only one child had a minimum plasma drug concentration < 3.4 μg/ml (2.57 μg/ml). Of the 13 children who received GPO-VIR as their first-line regimen, 12 had plasma HIV-1 RNA < 400 copies/ml at 6-18 months, with a median CD4 lymphocyte increase of 216 and 433 × 10⁶ cells/l at 6 and 12 months of treatment, respectively.

Conclusions: The administration of GPO-VIR S30 fixed-dose combination tablets in fractions or as a whole tablet to children resulted in appropriate NVP exposure and satisfactory virological and immunological benefit. This finding confirms the effectiveness of using a fixed-dose combination as a 'transitional option' while waiting for a paediatric fixed-dose combination drug formulation.

Introduction

Nevirapine (NVP) is recommended as a component of the first-line highly active antiretroviral therapy (HAART) regimen for HIV-infected infants and children, especially those younger than 3 years ^[1,2]. In developing countries, NVP is the most affordable non-nucleoside reverse transcriptase inhibitor for both children and adults. In 2001, the Thai Governmental Pharmaceutical Organization (GPO) introduced the first fixed-dose combination for adults, called GPO-VIR; this is composed of stavudine (30 or 40 mg), lamivudine (150 mg), and NVP (200 mg). Currently, GPO-VIR is the first-line regimen provided by the Ministry of Public Health in Thailand. Because of its convenience and low cost, and the limited alternatives for children, GPO-VIR has also been used in children. Concerns of whether using this adult fixed-dose combination in children, which often requires breaking the tablet, could result in inappropriate dosing of the different components in various age and weight groups are pertinent.

Maintaining adequate exposure for all antiviral compounds in fixed-dose combination is important for long-term clinical success. It is particularly critical that optimal NVP plasma concentrations are achieved as NVP has a low resistance barrier and a high level of cross resistance to the other non-nucleoside reverse transcriptase inhibitors ^[3]. Long-term virological response has been associated with NVP plasma concentrations > 3.4 μg/ml ^[4]. The association between either stavudine or lamivudine plasma concentration and response is less clear, likely because of the required intracellular activation of the nucleoside analogues. Therefore, paediatric dosing strategies for this fixed-dose combination have focused on achieving standard paediatric NVP dosing. The measurement of NVP plasma concentrations will help to determine if adequate absorption and dosing of this fixed-dose combination are attained. The aim of this study was to evaluate the steady-state pharmacokinetics of NVP in Thai children who had been taking the adult fixed-dose combination, GPO-VIR, either in whole tablets or in broken fractions.

Methods

Study population

HIV-infected children who had been receiving GPO-VIR S30 (stavudine 30 mg, lamivudine 150 mg and NVP 200 mg) for more than 8 weeks at Siriraj Hospital, Bangkok were enrolled after providing written informed consent by parent(s) and assent by older children. For these children, antiretroviral treatment was provided under the national antiretroviral access programme supported by the Ministry of Public Health. The dosage of GPO-VIR was based upon a NVP dosage of 120-200 mg/m² every 12 h and was achieved by dividing the tablet if necessary to 1/4, 1/2, 3/4 or 1 tablet. In order to provide the 'lead in' period for NVP dosing, during the first 2 weeks of treatment GPO-VIR was given in the morning, and only stavudine and lamivudine were given in the evening. Exclusion criteria were non-adherence to the treatment (defined as < 95% adherence rate when assessed by pill counts and questionnaires), concomitant use of drugs that might interfere with cytochrome P450 metabolism (rifampin, clarithromycin, erythromycin, itraconazole, ketoconazole, phenobarbital, phenyltoin), liver dysfunction, gastrointestinal problems that might affect drug absorption, toxicity related to GPO-VIR, and active opportunistic infections or malignancy.

Study design

To evaluate the adequacy of plasma NVP concentrations, blood samples at steady state were taken at predose and at 2 and 6 h after drug intake. CD4 lymphocytes were measured routinely at the time of GPO-VIR treatment initiation as baseline, and every 6 months thereafter. In children who received GPO-VIR as their first antiretroviral treatment regimen, plasma HIV-1 RNA was measured using Roche Amplicor HIV-1 Monitor version 1.5 (Cobas; (Roche Molecular Systems, Branchburg, New Jersey, USA), with a limit of detection of 400 copies/ml.

Sample preparation and nevirapine assay

NVP plasma drug concentrations were measured at the Faculty of Associated Medical Sciences, Chiang Mai University by a validated high-performance liquid chromatography assay as described previously ^[5]. The lower limit of assay quantification was 50 ng/ml. The laboratory participates in two international external quality control programmes, the US AIDS Clinical Trial Group Pharmacology Quality Control (Precision Testing) program and the French ASQUALAB Quality Control programme, and has successfully passed for all drugs ^[6].

Pharmacokinetic analysis

The concentration-time data were fitted to a one-compartmental model with first-order absorption and elimination using maximum a posteriori probability (MAP)-Bayesian estimation. The prior parameter values and variances used in the Bayesian estimator were derived from a population pharmacokinetic analysis of 495 infants and children participation in the paediatric ACTG studies ^[7].

Individual NVP pharmacokinetic parameters (apparent oral clearance and volume of distribution) were estimated using the post hoc subroutine in the computer program NONMEM ^[8]. Area under the concentration-time curve (AUC) from 0 to 12 h was calculated as bioavailability × dose/clearance. Half-life was calculated as 0.693 × volume of distribution/apparent oral clearance. The minimum NVP concentration was defined as the observed predose concentration. Because of the limited blood samplings, the maximum plasma drug concentration was not determined.

Results

Thirty-five children (18 girls and 16 boys) were enrolled; one patient was excluded because of non-adherence. The median age (range) was 8.4 years (range, 3-15) and the median CD4 cell count and percentage at the time of study entry was 576 × 10⁶ cells/l (range, 35-1443) and 20.2% (range, 1.73-30.16), respectively. The clinical stages by Centers for Disease Control and Prevention classification were class A in eight, class B in 14, class C in nine and none in two. The NVP dosage averaged (±SD) 164 ± 27 mg/m² every 12 h. The average lamivudine and stavudine daily dosages were 4.9 ± 0.7 mg/kg and 1.0 ± 0.1 mg/kg every 12 h, respectively (Table 1). Twenty-four subjects (71%) were receiving broken tablets as part of their standard dose. Thirty-two children swallowed the tablets without crushing. Four children were also receiving indinavir boosted with ritonavir combined with GPO-VIR for salvage treatment.

Table 1

The NVP pharmacokinetics parameters are shown in Table 2. There was no bias in the composite curve of predicted versus observed concentrations. One child receiving 190 mg/m² every 12 h of NVP had a minimum NVP concentration less than the target concentration of 3.4 μg/ml (2.57 μg/ml); however, at 2 and 6 h after NVP drug intake the drug levels were 4.93 and 4.64 μg/ml, respectively, and a good virological, immunological and clinical response was observed. NVP plasma levels were similar in the four children who also received IDV boosted with RTV, except for one child who had a very high minimum NVP concentration of 24.37 μg/ml, and an AUC of 257.6 h × μg/ml while receiving 167 mg/m² NVP, 4.2 mg/kg lamivudine, 0.83 mg/kg stavudine, 330 mg/m² indinavir and 83 mg/m² ritonavir every 12 h. This patient's indinavir trough and 2 h postdose concentration were satisfactory: 0.554 μg/ml (target > 0.1) and 8.620 μg/ml (target < 10), respectively.

Table 2

Of the 14 treatment- naive children who initiated treatment with GPO-VIR, 13 had received longer than 6 months of treatment. At 6 months, plasma HIV-1 RNA analyses were available in 11 patients and 10 of these patients had a viral load < 400 copies/ml. Two children only had a viral load result available at 12 and 18 months; both were < 400 copies/ml. The 12 children who achieved viral suppression had a satisfactory immunological response with a median CD4 cell count increase of 216 and 433 × 10⁶ cells/l (7.18% and 13.94%) at 6 and 12 months, respectively. One child had poor virological suppression at 6 months but had a transient small increase in CD4 cell count at 6 months. This child had M184V, K103N and M230L resistance mutations.

Discussion

One of the main obstacles to treating HIV infection in children has been the lack of affordable and appropriate antiretroviral formulations. At present, there is no prequalified fixed-dose combination available for paediatric use. NVP-containing fixed-dose combinations for adults are the most available treatment in developing countries. Therefore, it is a common practice in Thailand and many developing countries to use adult fixed-dose combinations in children. This is the first study describing the pharmacokinetics of NVP in children taking an adult fixed-dose combination containing NVP.

Generic drug components in fixed-dose combination have been questioned for their bioavailability and bioequivalence to the original individual drugs. Of these, NVP is the most critical component. Havlir et al. ^[9] found that trough levels of NVP correlated with the virological response. In addition, the risk of virological failure has been reported to be five times higher in patients with plasma concentration < 3 μg/ml when measured from a random sample ^[10]. In a separate study, NVP trough levels < 3.5 μg/ml were associated with a higher rate of treatment failure ^[4]. Based on these data, the guidelines from Department of Health and Human Services in the United States have suggested a minimum NVP target trough level of 3.4 μg/ml ^[11].

The pharmacokinetics parameters in our study were in the ranges reported in studies performed in adults ^[12,13]. The therapeutic adequacy of NVP could be considered by using a minimum concentration of > 3.4 μg/ml or by comparison of AUC values; however, there is no clear defined AUC threshold that correlates with virological response for NVP. In our study, all the children apart from one had trough concentrations > 3.4 μg/ml despite a higher clearance than adults ^[14]. Pill cutting may alter drug absorption and cause inaccuracy of drug dosing, but in this study we demonstrated that this practice gave satisfactory plasma concentrations of NVP.

The efficacy of GPO-VIR has been demonstrated in other studies. A cohort study in Bangkok of 101 antiretroviral therapy-naive HIV-infected adults found that 58% (by intention-to-treat) and 78% (by on-treatment) had viral loads < 50 copies/ml ^[15]. A recent study of GPO-VIR in 55 antiretroviral drug-naive children at 72 weeks of treatment found that 64% (by intention-to-treat) and 71% (by on-treatment) had viral load < 50 copies/ml. ^[16]. Our study found that, among treatment-naive children, 12 of 13 (92%) children had good response, with a viral load < 400 copies/ml within 6 to 18 months.

One of the four children who were also receiving indinavir-boosted with ritonavir had an unexplainable high NVP level (minimum concentration 24.37 μg/ml). Indinavir and ritonavir generally do not significantly alter NVP concentrations ^[17]. While ritonavir is a potent cytochrome P450 CYP3A inhibitor, NVP is metabolized by at least two isoforms, 3A and 2B6, which may explain the limited impact of ritonavir on NVP concentrations. However, if this subject had low CYP2B6 activity, the NVP concentrations might display greater sensitive to CYP3A inhibition by ritonavir.

A limitation of this study was that the NVP concentrations were only obtained at three time points in each subject; therefore, some of the pharmacokinetics parameter estimates, particularly volume of distribution and half-life, have limited accuracy. However, the low NVP fluctuation during the dose interval should have resulted in reasonable precision in parameter estimates for clearance and AUC.

In conclusion, this study demonstrated that NVP exposure in children taking GPO-VIR was comparable or greater than that seen with the original NVP in other populations. It is reassuring that dividing the adult fixed-dose combination tablet GPO-VIR does not adversely affect NVP bioavailability and can be used to generate paediatric doses between 120 and 200 mg/m² to achieve therapeutic levels of exposure. This study encourages the use of adult fixed-dose combination as a transitional measure to support scaling-up antiretroviral therapy in children where the paediatric formulations are not available.

Acknowledgements

Médicines Sans Frontières provided personnel for children and families activities during the study. US National Institutes of Health AIDS Research and Reference Program, Division of AIDS provided NVP which was used in HPLC assay. Abbot Laboratories provided the experimental material A86093, used as internal standard in the HPLC assay.

Sponsorship: This work was supported by Médicines Sans Frontières.