Hepatic and renal toxicity and associated factors among HIV-infected children on antiretroviral therapy: a prospective cohort study.

OBJECTIVES: The aim of the study was to investigate the prevalence of renal function and liver enzyme abnormalities among HIV-infected children, changes in prevalence with time on combination antiretroviral therapy (cART), and the factors associated with these abnormalities. METHODS: A prospective cohort study was conducted among HIV-infected children < 18 years old (n = 705) who were on first-line cART. Liver enzymes, renal function, haematology, immunology and virological response were assessed at enrolment and followed bi-annually for 18 months. Liver fibrosis and cirrhosis were assessed using noninvasive markers including the aspartate aminotransferase (AST) to platelet ratio index (APRI) and fibrosis score (FIB-4). RESULTS: The median age was 12 [interquartile range (IQR) 8-14] years; 53.3% of patients were male. At enrolment, the median cART duration was 3.3 (IQR 1.1-6.1) years; 177 (25.1%) and 83 (11.8%) patients had elevated AST and alanine aminotransferase (ALT), respectively. A tenth of the children had an APRI score > 0.5, suggesting liver fibrosis. Being on a zidovudine (ZDV)- or nevirapine (NVP)-based regimen and having a viral load > 1000 HIV-1 RNA copies/mL were significantly associated with elevated ALT. Twenty-four (3.4%) and 84 (12.1%) patients had elevated creatinine and blood urea nitrogen (BUN), respectively. As cART duration increased by 6 months, median BUN increased by 1.6 [95% confidence interval (CI) 0.4-2.7] mg/dL (P = 0.01); the glomerular filtration rate (GFR) decreased by 35.6 (95% CI 17.7-53.4) mL/min/1.73 m2 (P < 0.0001); and AST and ALT decreased by 1.4 (95% CI 0.4-2.5) IU/L (P = 0.01) and 1.4 (95% CI 0.2-2.6) IU/L (P = 0.01), respectively. CONCLUSIONS: A high prevalence of liver enzyme and renal function abnormalities was observed at enrolment. Decreasing liver enzyme levels during follow-up are possibly reassuring, while the progressive reduction in GFR and the increase in BUN are worrisome and require further study.

CYP2B6*6 genotype and high efavirenz plasma concentration but not nevirapine are associated with low lumefantrine plasma exposure and poor treatment response in HIV-malaria-coinfected patients.

We investigated the influence of efavirenz (EFV)- or nevirapine (NVP)-based antiretroviral therapy (ART) on lumefantrine plasma exposure in HIV-malaria-coinfected patients and implication of pharmacogenetic variations. A total of 269 HIV patients with uncomplicated falciparum malaria on NVP-based ART (NVP-arm), EFV-based ART (EFV-arm) or not receiving ART (control-arm) were enrolled and treated with artemether-lumefantrine. Day-7 lumefantrine, baseline EFV and NVP plasma concentrations, and CYP2B6*6,*18, CYP3A4*1B, CYP3A5*3,*6,*7, ABCB1 c.3435C>T and ABCB1 c.4036A>G genotypes were determined. The median day-7 lumefantrine plasma concentration was significantly lower in the EFV-arm compared with that in NVP- and control-arm. High EFV plasma concentrations and CYP2B6*6/*6 genotype significantly correlated with low lumefantrine plasma concentrations and high rate of recurrent parasitemia. No significant effect of NVP-based ART on lumefantrine exposure was observed. In conclusion, owing to long-term CYP3A induction, EFV-based ART cotreatment significantly reduces lumefantrine plasma exposure leading to poor malaria treatment response, which is more pronounced in CYP2B6 slow metabolizers.

Pharmacokinetic and pharmacogenomic modelling of the CYP3A activity marker 4ß-hydroxycholesterol during efavirenz treatment and efavirenz/rifampicin co-treatment.

OBJECTIVES: To assess the effect of the major efavirenz metabolizing enzyme (CYP2B6) genotype and the effects of rifampicin co-treatment on induction of CYP3A by efavirenz. PATIENTS AND METHODS: Two study arms (arm 1, n = 41 and arm 2, n = 21) were recruited into this study. In arm 1, cholesterol and 4ß-hydroxycholesterol were measured in HIV treatment-naive patients at baseline and then at 4 and 16 weeks after initiation of efavirenz-based antiretroviral therapy. In arm 2, cholesterol and 4ß-hydroxycholesterol were measured among patients taking efavirenz during rifampicin-based tuberculosis (TB) treatment (efavirenz/rifampicin) just before completion of TB treatment and then serially following completion of TB treatment (efavirenz alone). Non-linear mixed-effect modelling was performed. RESULTS: A one-compartment, enzyme turnover model described 4ß-hydroxycholesterol kinetics adequately. Efavirenz treatment in arm 1 resulted in 1.74 (relative standard error = 15%), 3.3 (relative standard error = 33.1%) and 4.0 (relative standard error = 37.1%) average fold induction of CYP3A for extensive (CYP2B6*1/*1), intermediate (CYP2B6*1/*6) and slow (CYP2B6*6/*6) efavirenz metabolizers, respectively. The rate constant of 4ß-hydroxycholesterol formation [mean (95% CI)] just before completion of TB treatment [efavirenz/rifampicin co-treatment, 7.40 × 10(-7) h(-1) (5.5 × 10(-7)-1.0 × 10(-6))] was significantly higher than that calculated 8 weeks after completion [efavirenz alone, 4.50 × 10(-7) h(-1) (4.40 × 10(-7)-4.52 × 10(-7))]. The CYP3A induction dropped to 62% of its maximum by week 8 of completion. CONCLUSIONS: Our results indicate that efavirenz induction of CYP3A is influenced by CYP2B6 genetic polymorphisms and that efavirenz/rifampicin co-treatment results in higher induction than efavirenz alone.

Pharmacogenetic and pharmacokinetic aspects of CYP3A induction by efavirenz in HIV patients.

We investigated the effects of pharmacogenetic variations and efavirenz pharmacokinetics on inter-individual differences in the extent of CYP3A induction by efavirenz using 4ß-hydroxycholesterol/cholesterol (4ß-OHC/Chol) as a marker for CYP3A induction. Plasma 4ß-hydroxycholesterol and cholesterol concentrations were determined at baseline, and at the 4th, 16th and 48th week of efavirenz-based highly active antiretroviral therapy in antiretroviral therapy-naive HIV patients (n=77). Efavirenz plasma concentrations were quantified at weeks 4 and 16. CYP2B6, CYP3A5, ABCB1, UGT2B7 genotyping were done. Compared with baseline, the median plasma 4ß-OHC/Chol ratio increased at the 4th (257%), 16th (291%) and 48th (165%) week (P<0.0001). CYP2B6*6 genotype significantly influenced 4ß-OHC/Chol ratio at weeks 16 (P=0.02) and 48 (P=0.04) being highest in CYP2B6*6/*6>*1/*6>*1/*1. There were positive correlations between plasma efavirenz and 4ß-OHC/Chol ratios (week 4: P=0.02, week 16: P=0.001). CYP3A enzyme induction by efavirenz is pronounced in CYP2B6 slow metabolizers who have high efavirenz plasma exposure.

High plasma efavirenz level and CYP2B6*6 are associated with efavirenz-based HAART-induced liver injury in the treatment of naïve HIV patients from Ethiopia: a prospective cohort study.

The objective of this study was to assess the incidence, timing and identify pharmacogenetic, efavirenz (EFV) pharmacokinetic and biochemical predictors of EFV-based antiretroviral therapy (ART) drug-induced liver injury (DILI). ART-naïve HIV patients (n = 285) were prospectively enrolled. Pretreatment laboratory evaluations included hepatitis B surface antigen and C antibody, CD4 count and viral load. Liver tests were done at baseline, 1st, 2nd, 4th, 8th, 12th, 24th and 48th weeks during ART. Plasma EFV and 8-hydroxyefvairenz concentration was determined at week 4 using liquid chromatography-mass spectrometry. CYP2B6, CYP3A5, ABCB1 3435C/T and UGT2B7*2 genotyping was done using Taqman genotyping assay. Data were analyzed using survival analysis and Cox proportional hazards model. The incidence of DILI was 15.7% or 27.9 per 100 person-years and that of severe injury was 3.4% or 6.13 per 100 person-years. The median time for the development of DILI and severe injury was 2 and 4 weeks after initiation of ART, respectively. There was significant association of DILI with lower baseline platelet, albumin, log plasma viral load and CD4 count (P = 0.031, 0.037, 0.06 and 0.019, respectively). Elevated baseline alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, plasma EFV level and CYP2B6*6 were good predictors for the development of DILI (P = 0.03, 0.01, 0.016, 0.017 and 0.04, respectively). We report for the first time CYP2B6*6 as a putative genetic marker and high plasma EFV concentration as intermediate biomarker for vulnerability to EFV-induced liver injury in HIV patients. CYP2B6 genotyping and/or regular monitoring of EFV and lever enzymes level during early therapy is advised for early diagnosis and management of DILI.

Effect of rifampicin and CYP2B6 genotype on long-term efavirenz autoinduction and plasma exposure in HIV patients with or without tuberculosis.

We performed a prospective comparative study to examine, from a pharmacogenetics perspective, the effect of rifampicin (RIF) on long-term efavirenz (EFV) autoinduction and kinetics. In a study population of patients with HIV receiving EFV with RIF (arm 2, n = 54) or without RIF (arm 1, n = 128 controls), intraindividual and interindividual plasma EFV and 8-hydroxyefavirenz levels were compared at weeks 4 and 16 of EFV therapy. In arm 2, RIF was initiated 4 weeks before starting EFV. In controls (arm 1), the plasma EFV was significantly lower whereas 8-hydroxyefavirenz was higher at week 16 as compared to week 4. By contrast, there were no significant differences in plasma EFV and 8-hydroxyefavirenz concentrations over time in arm 2. At week 4, the plasma EFV concentration was significantly lower in arm 2 as compared to arm 1, but no significant differences were observed by week 16. When stratified by CYP2B6 genotype, significant differences were observed only with respect to CYP2B6*1/*1 genotypes. Ours is the first report of the CYP2B6 genotype-dependent effect of RIF on long-term EFV autoinduction.

Sex and CYP3A5 genotype influence total CYP3A activity: high CYP3A activity and a unique distribution of CYP3A5 variant alleles in Ethiopians.

The objectives of the this study were to assess the influence of CYP3A5 genotype and sex on the variability in total CYP3A activity and to compare 4ß-hydroxycholesterol and omeprazole sulfoxidation as phenotypic markers for CYP3A activity in Ethiopians. Healthy subjects (n=150) were genotyped for CYP3A5*3, *6 and *7 using allele-specific PCR and Taqman genotyping assays. Plasma levels of 4ß-hydroxycholesterol, 3 h post-dose omeprazole and omeprazole sulfone, were determined by gas chromatography-mass spectrometry and high performance liquid chromatography, respectively. The frequency of CYP3A5*1, *3, *6 and *7 was 20.5, 67.3, 12.2 and 0%, respectively. The mean plasma 4ß-hydroxycholesterol level was 35.4 ng ml⁻¹. The mean 4ß-hydroxycholesterol level (P=0.0001) and the 4ß-hydroxycholesterol/cholesterol ratio (P=0.004) were higher in women than in men. CYP3A5 genotype significantly correlated with the plasma 4ß-hydroxycholesterol concentration (P=0.003) and 4ß-hydroxycholesterol/cholesterol ratio (P=0.0002). The omeprazole/omeprazole sulfone ratio was significantly correlated with 4ß-hydroxycholesterol and 4ß-hydroxycholesterol/cholesterol ratio in CYP3A5*0/*0 genotypes but not in individuals carrying the CYP3A5*1 allele. No correlation of omeprazole/omeprazole sulfone ratio with sex or CYP3A5 genotype was observed. A clear gene-dose effect implies plasma 4ß-hydroxycholesterol level as a useful endogenous biomarker for total CYP3A activity (CYP3A5 plus CYP3A4) whereas the omeprazole/omeprazole sulfone ratio reflects mainly CYP3A4 activity. Sex and CYP3A5 genotype influence total CYP3A activity. Ethiopians display high total CYP3A activity and a unique distribution of CYP3A5 variant alleles not described hitherto.

Long-term efavirenz autoinduction and its effect on plasma exposure in HIV patients.

We investigated the influence of gender and pharmacogenetic variations on long-term efavirenz autoinduction and disposition among patients with HIV in Tanzania (N = 129). Plasma concentrations (at 16 h) of efavirenz and 8-hydroxyefavirenz were quantified at weeks 4 and 16 of therapy. Genotyping was performed to identify cytochrome P450 (CYP) 2B6*6, CYP3A5*3, *6, and *7, and ABCB1-3435 C/T genotypes. There were reductions in the median efavirenz concentration (Wilcoxon matched-pair test P < 0.001) and efavirenz/8-hydroxyefavirenz ratio (P < 0.001) by 19 and 32%, respectively, at week 16 as compared with week 4. The proportion of patients with efavirenz concentration <1 µg/ml at week 16 was higher by 67, 25, and 5% in CYP2B6*1/*1, *1/*6, and *6/*6 genotypes, respectively. The defined therapeutic range based on observed plasma concentrations is affected by the time point of sampling and the CYP2B6 genotype. The effect of efavirenz autoinduction on reducing plasma exposure continues up to week 16 and predominantly affects CYP2B6 extensive metabolizers. Among CYP2B6 slow metabolizers, the presence of a CYP3A5 genotype allele is associated with greater effects of efavirenz autoinduction on plasma concentrations of the drug. The cumulative induction may influence the long-term antiretroviral therapy outcome, particularly in CYP2B6*1 carriers.

CYP3A5 genotype has an impact on the metabolism of the HIV protease inhibitor saquinavir.

CYP3A is the main enzyme subfamily involved in the metabolism of the HIV protease-inhibitor saquinavir. We hypothesized that individuals homozygous for CYP3A5*1 might have a higher oral clearance of saquinavir, compared with subjects lacking functional CYP3A5 alleles. A single-dose pharmacokinetic trial of saquinavir soft gel capsules, 1,200 mg, was performed in 16 black Tanzanian healthy volunteers with two functional CYP3A5 alleles (*1/*1) and in 18 volunteers without functional CYP3A5 alleles (both alleles being either *3, *6, or *7). The median area under the plasma concentration-time curve (AUC)0-24 reached among subjects with two functional alleles was 1,410 ng h/ml (interquartile range (IQR) 826-1,929), whereas it was 2,138 ng h/ml (IQR 1,380-3,331) in subjects without (P=0.0533, Mann-Whitney U-test). The median ratio of saquinavir over its M2 plus M3 hydroxy metabolites in urine was 64 (IQR 52-73) in subjects with two functional alleles, whereas it was 145 (IQR 89-181) in those without (P=0.000078, Mann-Whitney U-test). In conclusion, saquinavir is metabolized by CYP3A5. The median AUC0-24 for saquinavir among individuals with two functional CYP3A5 alleles was 34% lower than among those with no functional alleles. To clarify the clinical importance of the CYP3A5 polymorphism, further studies should be conducted on saquinavir, dosed to steady state, in the presence of ritonavir boosting.

Genetic polymorphism of cytochrome P450 2C9 in a Caucasian and a black African population.

AIMS: CYP2C9 is a major enzyme in human drug metabolism and the polymorphism observed in the corresponding gene may affect the therapeutic outcome during treatment with several drugs. The distribution of variant CYP2C9 alleles was therefore investigated in an Italian and an Ethiopian population. METHODS: Allele-specific PCR analysis was carried out in order to determine the frequencies of the two most common variant alleles, CYP2C9*2 and CYP2C9*3 in genomic DNA isolated from 157 Italians and 150 Ethiopians. RESULTS: The frequencies of CYP2C9*1 (80%), CYP2C9*2 (11%) and CYP2C9*3 (9%) found in the Italian population were similar to other Caucasian groups. However in the Ethiopian population CYP2C9*1, CYP2C9*2 and CYP2C9*3 were present at a frequency of 94, 4 and 2% respectively. The 95% confidence intervals in CYP2C9*1, CYP2C9*2 and CYP2C9*3 between Italians and Ethiopians were 0.098, 0.176, 0.040, 0.098 and 0.040, 0.098, respectively. CONCLUSIONS: Our results indicate that the Ethiopian population has a unique relative distribution of the CYP2C9 alleles, which is not similar to any other ethnic group hitherto described.

S-mephenytoin hydroxylation phenotype and CYP2C19 genotype among Ethiopians.

The polymorphic metabolism of S-mephenytoin and the distribution of two known deleterious mutant CYP2C19 alleles was determined among 114 healthy unrelated black Ethiopians. Six subjects (5.2%) were poor metabolizers (PMs) of S-mephenytoin. The frequencies of the defective CYP2C19*2 (CYP2C19m1) and CYP2C19*3 (CYP2C19m2) alleles were 0.14 and 0.02, respectively. Three of the PMs were homozygous for the CYP2C19*2 allele and the remaining three PMs were heterozygous for both the CYP2C19*2 and CYP2C19*3 mutant alleles. It is concluded that the frequency of PMs for S-mephenytoin is similar in Ethiopians, Zimbabweans and Caucasians and that the CYP2C19*3 allele, for the first time identified in a black population, together with the CYP2C19*2 allele account for all of the defective CYP2C19 alleles among the Ethiopian PMs.

Frequent distribution of ultrarapid metabolizers of debrisoquine in an ethiopian population carrying duplicated and multiduplicated functional CYP2D6 alleles.

The debrisoquine hydroxylase (CYP2D6) catalyzes the oxidative metabolism of more than 40 different clinically important drugs. The CYP2D6 gene is highly polymorphic. Defect alleles, causing the poor metabolizer phenotype, and also alleles with duplicated or multiduplicated active genes, causing ultrarapid metabolism, have been described. In the current investigation, we have evaluated the CYP2D6 phenotype (n = 115) and genotype (n = 122) among healthy Ethiopians. Only two subjects (1.8%) exhibited metabolic reaction (MR) for debrisoquine > 12.6 and were classified as poor metabolizers. A mutation in exon 1 causing a 34Pro --> Ser amino acid exchange, typical of the Chinese CYP2D6*10B (Ch1) allele and yielding an unstable enzyme, was present among 16% of the population and the carriers exhibited a high MR (0.9-5.0). Increased MR was also found among 18% of the subjects with a 107Thr --> Ile mutation associated to the CYP2D6*17(Z) allele causing diminished activity of CYP2D6 in vivo. Interestingly, 29% of the population investigated carried alleles with duplicated or multiduplicated CYP2D6 genes, indicative of ultrarapid metabolism. Xbal and EcoRI RFLP analyses identified individuals having new alleles with four or five CYP2D6*2(L) genes. Subjects with duplicated or multiduplicated CYP2D6*2 genes exhibited the lowest MR. These results suggest that the Ethiopian population, in comparison to Caucasian, Oriental and other Black populations, is genetically different with respect to the constitution of the CYP2D locus. The results also show that subjects carrying duplicated or multiduplicated active CYP2D6 genes are very common in certain ethnic groups, implicating this issue of potential global importance.