|Home | About | Journals | Submit | Contact Us | Français|
Seventy-one human immunodeficiency virus-infected patients with tuberculosis who were receiving a rifampin (rifampicin)-containing regimen were initiated on treatment with efavirenz at 600 mg/day plus stavudine-lamivudine. Fasting efavirenz concentrations at 12 h after dosing (C12) were monitored. The mean ± standard deviation efavirenz C12 at weeks 6 and 12 and after rifampin discontinuation were 4.5 ± 4.3, 3.8 ± 3.5, and 3.5 ± 2.7 mg/liter, respectively. High body weight was associated with a low efavirenz C12 at weeks 6 and 12 (P = 0.003, r = −0.255). The efavirenz C12 regression prediction line at 1 mg/liter intercepted a mean body weight of 57.5 kg.
N2R was a randomized clinical trial comparing the clinical efficacies of two nonnucleoside reverse transcriptase inhibitor-based highly active antiretroviral therapy (HAART) regimens for patients coinfected with human immunodeficiency virus (HIV) and tuberculosis who received rifampin (rifampicin) (9). A secondary objective was to assess body weight as a cutoff for selecting the daily dosage of efavirenz. Patients were enrolled and followed through 60 weeks after HAART. Inclusion criteria were described previously (9). All patients were started on twice-daily stavudine-lamivudine and efavirenz at 600 mg/day at bedtime. The dosages of rifampin were 450 mg/day for body weights of ≤50 kg and 600 mg/day for body weights of >50 kg. Fasting efavirenz concentrations at 12 h after dosing (C12) were measured with a validated high-performance liquid chromatography assay at weeks 6 and 12 and at 6 to 12 weeks after rifampin discontinuation. This assay was developed at the Department of Clinical Pharmacology of the University Medical Center Nijmegen, Nijmegen, The Netherlands (5).
All statistical analyses were performed with SPSS version 11.5 (SPSS Inc., Chicago, IL). Pearson's correlations were used to study the relationships between parameters. Factors possibly predictive of efavirenz C12 and an elevated serum alanine aminotransferase (ALT) activity at week 12 were evaluated with a linear-regression model. Inter- or intrapatient variability was expressed as a coefficient of variation (CV). The institutional ethics committees of the Bamrasnaradura Infectious Diseases Institute and the Thai Ministry of Public Health approved this study.
Seventy-one patients were enrolled; 65% were male, and the mean ± standard deviation (SD) baseline body weight was 53 ± 10 kg. The mean ± SD baseline CD4 cell count was 75 ± 68/μl, the median (interquartile range) log plasma HIV type 1 RNA copy number was 5.8 (range, 5.6 to 5.8), and the mean ± SD serum ALT activity was 29 ± 18 U/liter. Six (8.5%) patients were positive for hepatitis B surface antigen, and 18 (25.4%) were positive for hepatitis C antibody. The mean ± SD efavirenz C12 at weeks 6 and 12 and after rifampin discontinuation are shown in Fig. Fig.1.1. After stratification into four groups by body weight (35 to 45, 46 to 55, 56 to 65, and >66 kg) at week 6, the mean C12 were 7.6 mg/liter, 4.1 mg/liter, 2.4 mg/liter, and 1.8 mg/liter, respectively. Three (4.3%) of 70, 2 (3.1%) of 64, and 2 (3.3%) of 60 patients, respectively, had efavirenz C12 of <1 mg/liter at weeks 6 and 12 and after rifampin discontinuation (P > 0.05). Twenty-four (34.3%), 16 (25.0%), and 14 (23.3%) patients had efavirenz C12 of >4 mg/liter at the corresponding periods (P > 0.05). High interpatient variability at weeks 6 (CV, 105%) and 12 (CV, 107%) and after rifampin discontinuation (CV, 77%), as well as high intrapatient variability between weeks 6 and 12 (CV, 136%) were observed in the efavirenz C12. Figure Figure22 shows the relationship between efavirenz C12 at the different time points. In the separate multivariate analysis model, at the different time points, low C12 at weeks 6 and 12 were associated with higher body weights (P = 0.028), as shown in Table Table1.1. This correlation was not significant (P = 0.244) after discontinuation of rifampin.
By intent-to-treat and on-treatment analyses at 60 weeks, 70.4% (50 of 71) and 81.9% (50 of 61) of the patients achieved HIV type 1 RNA levels of <50 copies/ml, respectively. Two (66.7%) of 3 patients with efavirenz C12 of <1 mg/liter at week 6 and 18 (16.9%) of 67 patients with efavirenz C12 of ≥1 mg/liter at week 6 developed treatment failure. The mean ± SD serum ALT activity increased from the baseline at week 12 (38 ± 32 U/liter; P = 0.046, paired t test). At 12 weeks, two patients developed grade III hepatotoxicity. Multivariate analysis of ALT activity-predictive factors at week 12 showed that positivity for hepatitis B surface antigen was a predictor of high serum ALT activity during concurrent efavirenz and rifampin administration (P = 0.005). Three patients developed grade II/III cutaneous reactions.
The range of acceptable efavirenz C12 is currently proposed to be 1 to 4 mg/liter (7). To date, clinical trials evaluating weight-based cutoffs for efavirenz dosing during coadministration of rifampin are limited (1, 2, 6-8, 10, 12). In the present study, almost one-third of the patients had body weights of ≥60 kg during monitoring. Nevertheless, most achieved concentrations above the minimum recommended level. The efavirenz C12 decreased 16% relative to that at week 6, as patients gained an average of 3 kg of body weight, but this decrease did not change the percentage of patients with a C12 of <1 mg/liter. Interestingly, an efavirenz concentration of 1 mg/liter intercepted a mean body weight of 60 kg at the upper limit of the 95% confidence interval (CI) of the beta value of the correlation coefficient. Therefore, increasing efavirenz dosing, practically to 800 mg/day, in patients weighting >60 kg to compensate for the effect of rifampin coadministration may be considered.
Efavirenz concentrations persistently decreased from 25% to 46% for every 10-kg increase in body weight. In addition, a high body weight was found to be an important predictive factor of a low drug concentration. These models did not include gender because this parameter had a highly positively correlated with body weight (P < 0.001). A cross-sectional study showed that body weight was an independent predictive factor for efavirenz concentration (16). Furthermore, marked interpatient and intrapatient variabilities were observed in this study during concurrent the use of both drugs, and this trend was found to be consistent over time. The previous studies showed low intrapatient variability in patient receiving efavirenz alone (11). This may be explained by the extensive difference of genetic polymorphism that influences CYP2B6 enzyme expression (17).
A number of limitations should be acknowledged. First, this study collected C12 instead of the minimal drug concentration to assess drug exposure. A number of studies attempted to demonstrate the ability of C12 to predict clinical efficacy and adverse reactions (3, 11). Second, there are no studies that clearly show that therapeutic drug monitoring improved patients' clinical outcomes. Third, a persons with the CYP2B6 polymorphism 516G>T had higher efavirenz concentrations (4, 13, 15). The frequency of this allele ranges from 15% in Asians to 50% in Africans (4, 15). Ultimately, previous studies demonstrated that many factors and issues contribute to pharmacokinetic variability, including biological factors, environmental factors, and genetic issues (14).
A standard-dose efavirenz-based regimen is appropriate for HIV patients with tuberculosis who are receiving rifampin and have body weights of <60 kg. In clinical practice, a weight-based cutoff for efavirenz dosing is a practical therapeutic approach. However, further study is needed to explore the appropriate escalating daily efavirenz dosage in patients with higher body weights. In addition, the issue of ethnic pharmacogenetic differences should be taken into account.
This study was funded by grants from the Thailand Research Fund (TRF); the Ministry of Public Health, Thailand; and the Bamrasnaradura Infectious Diseases Institute, Thailand.
We thank all of the patients in this study, as well as Achara Chaovavanich, Palakorn Srinithiwat, Somsit Tansuphaswadikul, Samruay Nilkamhang, Phatchara Tunteerapat, Putthiporn Limpanadusadee, and all of the other staff of the Bamrasnaradura Infectious Diseases Institute, for their support and assistance in conducting this study. We also acknowledge David M. Butler for his comments and help in reviewing the manuscript.
We have no conflicts of interest to declare.
Published ahead of print on 10 August 2009.