Efavirenz levels and 24-week efficacy in HIV-infected patients with tuberculosis receiving highly active antiretroviral therapy and rifampicin

Background: Concomitant use of efavirenz and rifampicin is common for treatment of HIV and tuberculosis. Plasma efavirenz levels can be reduced by rifampicin, but the appropriate daily dosage of efavirenz is unclear.

Methods: HIV-infected patients with active tuberculosis, receiving rifampicin > 1 month, were randomized to receive stavudine and lamivudine plus efavirenz 600 or 800 mg daily. Plasma efavirenz levels were measured (at 12 h after dosing and on day 14) by high-performance liquid chromatography. Plasma HIV RNA was assessed at 16 and 24 weeks after antiretroviral therapy.

Results: Baseline characteristics were comparable in the 84 patients (two groups of 42). Median plasma efavirenz levels were 3.02 mg/l (range, 0.07-12.21) in the 600 mg group and 3.39 mg/l (range, 1.03-21.31) in the 800 mg group (P = 0.632). Plasma efavirenz levels were < 1 mg/l in 3 of 38 (7.9%) patients in the 600 mg group and in none of the 800 mg group (P = 0.274). Approximately 40 and 45% of patients had efavirenz levels > 4 mg/l, respectively. There was no significant difference in time to HIV RNA < 50 copies/ml (P = 0.848).

Conclusions: Median plasma efavirenz levels were comparable among both groups. Efavirenz 600 mg/day should be sufficient for most Thai HIV-infected patients receiving rifampicin with body weight approximately 50 kg. These results may not be applicable to other ethic populations who have higher body weights. However, the study of long-term virological and immunological outcomes is needed and under further investigation.

Introduction

HIV infection has been one of the most health-threatening problems in Thailand for two decades. Highly active antiretroviral therapy (HAART) has reduced the risk of HIV progressing to AIDS, the incidence of opportunistic infections and the mortality rate ^[1-3]. There is strong evidence for treating patients when their CD4 cell count is < 200 × 10⁶ cells/l or they have major opportunistic infections ^[4]. Efavirenz, a non-nucleoside reverse transcriptase inhibitor drug ^[5], is currently available ^[4,6,7]. An efavirenz-based HAART regimen has been used because of its effectiveness plus manageable side effects and affordable cost. Most recent guidelines have recommended an efavirenz-based regimen as an initial preferred HAART regimen in a naive HIV-infected patient ^[4].

In Thailand as in other developing countries, a patient often presents late with advanced AIDS and major opportunistic infections ^[8-10]; tuberculosis is the most frequent major opportunistic infection in HIV-infected patients ^[8,11,12]. In patients with advanced AIDS and active tuberculosis, HAART should be administered concurrently with the tuberculosis treatment because - if HAART is deferred - another opportunistic infection may superimpose itself and accelerate HIV disease progression ^[13,14]. Therapy with antiretroviral drugs can reduce the short-term risk of death for an HIV-infected patient diagnosed with tuberculosis ^[15]. The optimal timing of such therapy is still unknown because adherence, adverse events and paradoxical reactions are important problems in concurrent treatment of tuberculosis and HIV. The standard short-course, antituberculous regimen (i.e., rifampicin, isoniazid, pyrazinamide and ethambutol) has been used extensively ^[16]. The potential for drug interactions is one of the major concerns in HIV-infected patients receiving both treatments concurrently, since both efavirenz and rifampicin can induce cytochrome P450 isoenzymes in the liver ^[17,18]. Although, coadministration of efavirenz 800 mg/day with two nucleoside reverse transcriptase inhibitors and rifampicin is currently recommended as an option in treating HIV and tuberculosis coinfection, particularly when rifabutin is not available ^[4], previous observational studies have shown that efavirenz 600 mg/day tends to be sufficient to treat an HIV-infected patient receiving rifampicin ^[19,20]. However, there are few clinical studies on the pharmacokinetic effect of rifampicin on plasma efavirenz levels ^[21]. The present study is a randomized controlled trial with the primary objective to compare plasma efavirenz levels after two dosages, 600 or 800 mg daily, in HIV-infected patients with active tuberculosis receiving a rifampicin-containing antituberculous regimen and HAART. The secondary outcomes were the number of patients with plasma efavirenz levels < 1, 1-4 and > 4 mg/l and time to virological success, defined as plasma HIV RNA < 50 copies/ml after 16 and 24 weeks of antiretroviral therapy.

Methods

The study design was an open-labelled randomized controlled trial involving 84 Thai HIV-infected patients from two centres: 50 patients from the Infectious Diseases Clinic, Ramathibodi Hospital, Mahidol University, Bangkok and 34 patients from the Bamrasnaradura Institute, Ministry of Public Health, Nonthaburi, Thailand. Enrollment was from July 2003 to February 2004. Inclusion criteria were age > 15 years; diagnosis of active tuberculosis by clinical features, positive acid-fast stain and/or positive culture for Mycobacterium tuberculosis; receiving rifampicin-containing, antituberculous regimen > 1 month prior to enrolment; and willing to participate and signing consent form. Exclusion criteria were previous antiretroviral therapy, pregnancy or receiving a medication that may have a drug interaction with efavirenz or rifampicin. The institutional ethics committees of both hospitals approved the study.

Block randomization, using STATA version 8.0 (Stata Corp., College Station, Texas, USA), was applied to allocate patients to two groups, one receiving efavirenz 600 mg at bedtime and the other receiving efavirenz 800 mg at bedtime. Both groups also took stavudine 30/40 mg and lamivudine 150 mg twice a day and rifampicin 450/600 mg daily. The dosage of stavudine was adjusted by body weight (i.e., stavudine 30 and 40 mg for body weight < 60 and > 60 kg, respectively). The dosage of rifampicin was 450 mg for body weight < 50 kg and 600 mg for body weight > 50 kg. The general characteristics [e.g., gender, age, body weight, body mass index (BMI), previous opportunistic infections, and site of tuberculosis infection] were recorded. Blood samples were obtained at enrollment for measurement of CD4 cell count by flow cytometry and HIV-1 RNA by the polymerase chain reaction using Amplicor Monitor version 1.5 (Roche Diagnostics, Branchburg, New Jersey, USA). Plasma efavirenz levels were analysed at the HIV Netherlands-Australia-Thailand (HIV-NAT) Research Pharmacokinetic laboratory located at Chulalongkorn Medical Research Center by high-performance liquid chromatrography (HPLC). This assay was developed in the Department of Clinical Pharmacology, University Medical Centre Nijmegen, the Netherlands. The HPLC system consisted of a P4000 solvent delivery pump, a AS3000 autosampler, a UV2000 programmable UV detector wavelength 251 nm and the computing integrator for HPLC SN4000 system (Thermo Finnigan, San Jose, California, USA). The analytic column was an Omnisher 5 C18 column (150 mm × 4.6 mm internal diameter, particle size 5 μm) protected by a Chromguard RP column (Varian, Middelburg, the Netherlands). Analytic column runs were processed by Millennium32 software (Water, Etten-Luer, the Netherlands). The efavirenz retention time was 4.7 min. The efavirenz calibration curve was linear over the range 0.20-20.0 mg/l. The lower limit of quantification for efavirenz was 0.20 mg/l. Recovery after extraction from plasma was 106%. Accuracy ranged from 89.0 to 105.0%, and within-day and precisions ranged from 1.44 to 2.38% and from 0 to 2.19%, respectively.

Power and Sample Size version 1.01 was used to calculate sample size by test for equivalence of plasma efavirenz level ^[22]. The equivalence of plasma efavirenz level was a difference of < 1 mg/l between the plasma efavirenz levels in the two treatment groups. Type I and type II errors were 0.05 and 0.20, respectively. The estimation for loss to follow-up was 15%. Sample size needed was 42 patients in each group. Mean (SD) and frequency (%) were used to describe the patients' characteristics for continuous and categorical data, respectively. A Mann-Whitney test was used to compare median plasma efevirenz levels between groups. A chi-square test was used to compare the number of patients with plasma efavirenz levels < 1, 1-4, and > 4 mg/l. The Kaplan-Meier test was used to estimate the probability of virological success. A P value < 0.05 was considered as statistical significant. All analyses were performed using STATA version 8.0.

Results

The patients' general characteristics and CD4 cell and plasma HIV RNA values are summarized in Table 1. Both study groups had very low median CD4 cell counts: 32 × 10⁶ cells/l (range, 1-279) in the group taking 600 mg efavirenz and 37.5 × 10⁶ cells/l (range, 0-384) in the 800 mg efavirenz group. Three patients in the efavirenz 600 mg group and two patients in the 800 mg group were lost to follow-up during week 2. One patient in the efavirenz 600 mg group had to discontinue efavirenz treatment during week 2 as a result of severe maculopapular rash. Hence, the number of patients for plasma efavirenz analyses was 38 in the efavirenz 600 mg group and 40 in the 800 mg group. Distributions of the median plasma efavirenz levels between the two treatment groups are shown in Fig. 1. There was no significant difference between the two treatment groups in the duration of time from dose intake to blood sampling (12.17 ± 1.5 h for efavirenz 600 mg and 12.26 ± 1.9 h for efavirenz 800 mg; P = 0.497). Median plasma efavirenz level was 3.02 mg/l (range, 0.07-12.21) in the efavirenz 600 mg group and 3.39 mg/l (range, 1.03-21.31) in the efavirenz 800 mg group (P = 0.632). The proportions in both groups who had plasma efavirenz levels < 1, 1-4, and > 4 mg/l are shown in Fig. 2. Probabilities of time to virological success between the two groups are shown in Fig. 3. There were no significant differences in the number of patients who had plasma efavirenz levels < 1, 1-4, and > 4 mg/l (P = 0.274) and the time to virological success (P = 0.848). There were three patients in the efavirenz 600 mg group who had plasma efavirenz levels < 1 mg/l. After 24 weeks of antiretroviral therapy, these three patients had plasma HIV RNA levels < 50 copies/ml. Plasma efavirenz levels > 4 mg/l occurred in 15 patients in the efavirenz 600 mg group and 18 patients in the 800 mg group, but there was only one patient received efavirenz 600 mg/day who needed to discontinue antiretroviral therapy owing to severe headache.

Table 1	Fig. 1	Fig. 2
Fig. 3

Discussion

Mid-dosing plasma efavirenz levels were assessed in HIV-infected patients with active tuberculosis receiving rifampicin. Generally, the usual parameters of drug exposure that are investigated for relationships with virological response are area under the curve (AUC) and trough level. Both AUC and efavirenz trough sampling is not convenient because the drug is normally given at bedtime. Marzolini et al. ^[23] demonstrated that the mid-dosing plasma efavirenz level could predict treatment failure. In our study, there were more males in the efavirenz 800 mg group (83.5%) than in the 600 mg group (64.3%) (P = 0.047). However, Ribaudo et al. ^[24] have recently shown that efavirenz clearance was not associated with gender but was strongly associated with body weight and race; therefore, this gender imbalance between the two treatment groups may not have affected the outcome of the plasma efavirenz analysis. Both the mean body weight and the mean BMI of our patients were relatively low. The patients in our study presented late-stage HIV disease with very low CD4 cell counts and high levels of HIV RNA, which commonly occurs in other developing countries ^[25].

As efavirenz was administered at bedtime to improve the tolerability and reduce adverse events ^[26], blood samples for efavirenz study were obtained for the patients' convenience in the morning at 12 h after the dose intake. The median plasma level was 3.02 mg/l (range, 0.07-12.21) in the efavirenz 600 mg group and 3.39 mg/l (range, 1.03-21.31) in the 800 mg group (P = 0.623). Marzolini et al. ^[23] reported median plasma efavirenz levels in 130 Swiss HIV-infected patients without tuberculosis of 2.18 mg/l. It is of interest that, in spite of the concomitant use of efavirenz and rifampicin in our patients, the median plasma efavirenz level in both our treatment groups was higher than that reported by Marzolini et al. In the same study ^[23], plasma efavirenz > 4 mg/l was shown to be associated with the risk of central nervous system adverse reactions. Interestingly, we found approximately 40% and 45% of the efavirenz 600 mg and 800 mg groups, respectively, had plasma levels > 4 mg/l. This may be explained by a lower BMI in our patients and the racial difference. There was only one patient received efavirenz 600 mg/day who discontinued HAART as a result of severe headache.

Our findings show that the plasma efavirenz level was approximately 11% lower in the efavirenz 600 mg group (3.02 mg/l) than in the efavirenz 800 mg group (3.39 mg/l). Benedek et al. ^[27] studied the pharmacokinetic effect of rifampicin on plasma efavirenz concentrations in healthy volunteers and reported that rifampicin reduced efavirenz's AUC by 26% and its peak concentration by 20%. Lopez-Cortes et al. ^[21] showed that the median peak decreased by 24% and trough concentration by 18%. A 1-4 mg/l range at mid-dosing interval represented an appropriate therapeutic level for dose individualization and virological failure in 50% of patients with plasma efavirenz levels < 1 mg/l ^[23]. High-interpatient and low-intrapatient variabilities suggest that therapeutic drug monitoring may be advisable for individualizing efavirenz treatment ^[23]. However, in a resource-limited setting such as ours, where therapeutic drug monitoring is in limited use and the cost of HAART treatment is high, our data may support the physician's use of efavirenz 600 mg/day for the concomitant treatment of HIV-infected patients with tuberculosis receiving rifampicin.

A limitation of our study is that no correlations were made between the plasma efavirenz levels in the two treatment groups and their long-term immunological and virological outcomes. However, long-term treatment outcome analyses have been planned and are underway. A previous study has shown that plasma efavirenz levels can determine treatment outcomes ^[23]. Another consideration is that the mean body weight of our patients was relatively low and, therefore, the plasma efavirenz levels that we measured with these dosages may not be applicable for patients with higher body weight. After treatment of tuberculosis, almost all our patients gained weight. Consequently, plasma efavirenz levels might reduce over time while patients are taking efavirenz and rifampicin. Our further study should provide details of plasma efevirenz in over multiple timepoints.

In summary, plasma efavirenz levels, the proportions of patients who had plasma efavirenz < 1 mg/l and the time to virological success did not differ between the two groups of HIV-infected patients who had active tuberculosis receiving efavirenz 600 or 800 mg/day plus rifampicin. A HAART regimen containing 600 mg efavirenz daily should be sufficient for patients with HIV and active tuberculosis. However, a study of long-term virological and immunological outcomes is needed and under further investigation.

Acknowledgements

We would like to thank all the attending staff in the Division of Infectious Diseases, the Virology Laboratory, the Immunology Laboratory and the HIV-NAT Research Collaboration, Thai Red Cross AIDS Research Center for their support. We thank Mark Boyd for his comment.

Sponsorship: The study was supported by a research grant from Faculty of Medicine, Ramathibodi Hospital, Mahidol University.