Population pharmacokinetic and pharmacodynamic study of efavirenz in HIV–1–infected children treated with first line antiretroviral therapy in South Africa
Highly active antiretroviral therapy (HAART) has improved the life expectancy of HIV–1–infected patients dramatically since it was launched in 1996, but there are still many challenges in the provision of HAART, especially to children in resource limited countries. Efavirenz (EFV), a non–nucleoside reverse transcriptase inhibitor (NNRTI) forms part of the recommended national first line antiretroviral treatment regimen for children older than 3 years and weighing more than 10 kg in South Africa. Limited pharmacokinetic information on EFV plasma concentrations in sub–Saharan HIV–1– infected children is available. EFV is primarily metabolised by hepatic CYP2B6 isoenzymes. The CYP2B6 gene is characterised by extensive inter–individual variability in hepatic expression and activity. The single nucleotide change, 516G>T, on the CYP2B6 gene has consistently been associated with elevated EFV plasma levels in different ethnic populations and these are more frequently observed in populations of African descent. The recommended therapeutic range of EFV plasma levels is 1–4 ug/ml and the Cmin should be above 1 ug/ml. In this prospective study (PK/PD.EFV.07) cohort, 60 black children, both genders, with no prior exposure to antiretroviral therapy and eligible for antiretroviral therapy (ART) were enrolled and followed up at 1, 3, 6, 12, 18 and 24 months post HAART initiation. They all attended the outpatient clinic at Harriet Shezi Children’s Clinic, Chris Hani Baragwanath Hospital, Soweto, South Africa. The required ethics approval was obtained to conduct this study. The objectives of this investigation were to: develop and validate a suitable LCMS/ MS method to accurately determine plasma EFV levels from this study population, determine the prevalence and effect of CYP2B6 516G>T polymorphism on EFV plasma levels, determine the population pharmacokinetic clearance (CL/F) value of EFV, identify covariates that influence the clearance of EFV in HIV–1– infected children, and investigate specific pharmacodynamic effects and therapeutic outcomes of this EFV–based regimen within this paediatric population over the 24 months post–HAART initiation. The main findings of the measured mid–dose EFV plasma concentrations showed that sub–therapeutic concentrations (<1 ug/ml) accounted for 18% (116/649), within therapeutic range (1–4 ug/ml) represented 52.5% (341/649), and concentrations above the therapeutic range (>4 ug/ml) represented 29.5% (192/649). A significant number of the samples (47.5%) were outside the accepted therapeutic range during this 24 month follow–up period. Possible reasons contributing to this include genetic variation in drug metabolism and non–adherence. Genotype results on all 60 study participants were: 23% 516 T/T homozygotes, 42% 516 G/G homozygotes and 35% 516 G/T heterozygotes. The 516 T–allelic variant frequency was relatively high at 41%. This also supports and explains why such a large number (29.5%) of the mid–dose interval plasma samples were above (>4 ug/ml) the accepted therapeutic range. Repeated measures ANOVA confirmed that CYP2B6 516 G/G, G/T and T/T genotypes were consistently predictive of the log EFV concentrations at all times (P = 0.0001). The total median (IQR) EFV plasma concentrations over the 24 months post–HAART when pooled, were 6.36 (3.47 - 7.28) for T/T, 2.55 (1.62 - 3.59) for G/T, and 1.41 (1.02 - 1.74) ug/ml for G/G groups respectively (P<0.00001). Multiple comparisons by groups revealed that the EFV plasma concentrations between the T/T and G/G (P=0.000002) and between G/T and G/G (P=0.009) were statistically significant. However, the differences between the EFV plasma concentrations of the T/T and G/T groups were not significantly different (P=0.074). This supports previous results that the presence of the 516 T–allelic variant is responsible for the higher EFV plasma concentrations within individuals presenting with this single nucleotide mutation on the CYP2B6 gene. This EFV–based treatment was well tolerated even at plasma concentrations above the therapeutic range (>4 ug/ml) and most side effects subsided spontaneously. 89% of the participants were virally suppressed at 24 months post–HAART. The efficacy of this EFV–based treatment did not affect the three genotype groups differently and they showed similar improvement in their immunological (CD4–cell count and CD4%) markers and reduction in viral load over the 24 months post– HAART initiation. We found no association of the CYP2B6 516G>T polymorphism and side effects reported after 1 month of treatment within this study population. The final population pharmacokinetic (PK) estimates for EFV clearance (CL/F) were, 2.46, 4.60, and 7.33 l/h for the T/T, G/T, and G/G respective genotype groups. The volume of distribution (V/F) estimate was 89.52 l. The importance of interoccasion variability (IOV) in a PK model for a longitudinal study was again highlighted by this investigation. To our knowledge, this is the first study in black South African HIV–1–infected children with measured sequential EFV plasma concentrations which also investigated the influence of the CYP2B6 516G>T polymorphism on EFV plasma concentrations and the population clearance (CL/F) value of EFV in a longitudinal study over a period of 24 months post–HAART initiation.
- ETD@PUK