Pharmacogenetics of drug metabolizing enzymes and clinical implications in selected Kenyan populations
Much of the inter-individual and interethnic variability in drug response is now attributable to genetic differences in drug metabolism. Drug metabolizing enzymes exhibit genetic polymorphism with certain allelic variants displaying striking variable ethnic distribution. The pharmacogenetics of drug metabolism has been extensively studied in Caucasian and Asian populations, yet much remains to be done in African populations. The main objective of this thesis was to investigate the genetic polymorphisms and clinical implications of genes encoding drug metabolizing enzymes namely, CYP2B6. CYP2C19, CYP2D6. N-acetyl transferase (NAT2) and Glutathione- S- transferases (GSTs) in selected Kenyan populations. The first specific objective was to determine the distribution of pharmacogenetically relevant single nucleotide polymorphisms (SNPs) of CYP2B6. CYP2C19. CYP2D6. NAT2, GSTMl and GSTTl in selected Kenyan populations. A population of 350 Kenyans belonging to three ethno-linguistically distinct populations was recruited. The populations studied were the Western Nilotes (Luo) (100), the Eastern Nilotes (Maasai) (152) and the Bantu (Kikuyu) (102). Genotyping of allelic variants CYP2B6*6(516G>T); CYP2C19*2(681G>A); CYP2C19*3(638G>A); CYP2D6*4(1846G>A); CYP2D6*5(deletion); CYP2D6* 17(1 023C>T); CYP~D6*29(3183G>A); NAT2*5(341T>C);.NAT2*6(590(G>A); NAT2*7(857G>A); .NAT2*14 (191G>A); GSTMl and GSTTl was undertaken by polymerase chain reaction coupled with restriction fragment length polymorphism (PCR-RFLP). Results showed that the distribution of CYP2D6, GSTMl and GSTTl allelic variants was highly variable between the three Kenyan ethnic populations. CYP 2D6*4 was surprisingly higher in the Eastern Nilotes (9%) than in the Western Nilotes (3%) and the Bantu (2%) (P=0.002). CYP2D6*17 was more prevalent in the Bantu (34%) than in the two Nilotic populations (18-23%) (P= 0.003). The frequency of GSTMl*O/*O (deletion) in the Eastern Nilotes (16%) was nearly half that in the Western Nilotes and the Bantu (30%) (P=0.009) whereas that of GSTTl *01*0 (deletion) in the Eastern Nilotes (41 %) was nearly double that found in the Bantu and Western Nilotes (22-26%) (~=0.005). The null allele CYP2C19*3 was undetected in the Bantu but was found at <1 % in the Nilotes. There was homogeneity in the distribution of CYP2C19*2 (10-18%), CJP2B6*6 (34-37%) and the NAT2 slow acetylator alleles (NAT2*5 (33-38%), NAT*6 (23- 27%), NAT2*7(3-6%), NAT2*14 (9-14%) between the three Kenyan ethnic populations. The intermediate metabolizer (lM) was: for CYP2D6 (20-27%), CYP2C19 (20-27%), CYP2B6 516TT (12-16%) whereas NAT2 slow acetylation was 54-64%. The second specific objective was to analyze clinically relevant genetic variants of CYP2D6 and CYP2C19 and evaluate their impact on psychotropic medication patterns in Kenyan psychiatric in-patients. A total of 193 psychiatric in-patients at Mathari Hospital were recruited and genotyped for CYP2C19*2 and CYF2C19*3; CYP2D6*2, CYP2D6*4, CYP2D6*5, CYP2D6*17 and CYP2D6*29 using PCRRFLP methods. CYP2D6 and CYP2C 19 metabolic phenotypes were predicted from observed genotypes according to published methods and related to the pattern and CYP metabolic profiles of psychotropic medications in the patient population. The distribution of SNPs of CYP2C19 and CYP2D6 in psychiatric inpatients was similar to that in the general population except for CYP2C 19*3 which was higher in patients (8%) than in the general population ﾫ1%). Predicted intermediate metabolizer (IM) of CYP2D6 and CYP2C19 was 24% and 26% respectively. Predicted CYP2D6 lM phenotype was strongly associated with the use of high doses (15-20mg) of anticholinergic medication (68%) (P<O.OOI) indicative of serious extrapyramidal side effects; frequent hospital admission of 5-6 times (67%) per year (P<O.OOI) and exposure to antipsychotic polypharmacy of >5 drugs (P = 0.002). Psychotropic dosage regimens in the in-patients were higher than the defined daily doses (DDD) for chlorpromazine (CPZ) (58%), haloperidol 4 (94%) and amitriptyline (50%). Female gender (P=0.006) and non-exposure to substance abuse (P=0.031) were associated with lower doses ofCPZ. There was a significant potential for drug-drug interactions with 96% of the prescriptions containing concurrent CYP2D6 substrates and inhibitors whereas 9% had both CYP2C 19 substrates and an inducer. The main enzyme inducers were carbamazepine and phenobarbitone. The third specific objective was to determine the prevalence ofCYP2B6 516 G>T genotypes and assess their impact on nevirapine plasma levels and therapeutic outcomes in Kenyan HIV patients. A sample of 110 adult HIV positive patients on nevirapine based HAAR T at 6 months was recruited. They were genotyped for the CYP2B6 516G>T variant by real time polymerase chain reaction (RT-PCR). Nevirapine trough plasma levels were determined by a validated HPLC method. The frequency of CYP2B6*6 (516T) variant allele was 44% which was higher than the 32% seen in Kenyan general populations. CYP2B6 516 GT was 50% whereas CYP2B6 516 TT was 19%. Nevirapine plasma levels displayed great inter-individual variability ranging from 640 - 11800 nglrnL and was strongly correlated to CYP2B6 genotypes with a significant gene dose effect. Patients bearing the CYP2B6 516 TT had 55% higher plasma levels than CYP2B6 516 GG (P< 0.001). Sixty six percent ofHIV patients attained nevirapine levels above 4300 nglmL which reportedly offers durable viral suppression. However 17% had levels of3100- 4300 ng/rnL defined as the mutant selection window whereas another 17% had levels below 3100 nglmL associated with viral failure. Nevirapine plasma levels were not associated with toxicity and were not influenced by demographic characteristics except by the use of social drugs (16%) that resulted in 68% lower plasma levels (P = 0.006). Change in CD4 cell counts at 6 months of HAART was 170%, but was not influenced by nevirapine plasma levels or CYP2B6 genotypes. This thesis research has provided novel genetic data on clinically important genes affecting drug therapy in Kenyan populations not studied before. Intra-ethnic, inter-ethnic and inter-population genetic variability was demonstrated by the variable didtribution of CYP2D6*4, CYP2D6*17, GSTM1*0 and GSTTl*O. These observations underscore the risk of generalizing population genetic data even within a country and the need for further studies covering other Kenyan ethnic populations. The clinical implications of CYP2D6*17(34%) is of concern since CYP2D6 metabolizes 25% of all drugs in clinical practice, notably for cancer, cardiovascular and central nervous system. The NAT2 slow acetylation (66%) could increase toxicity to isoniazid and co-trimoxazole, first-line agents for tuberculosis and pneumocystis pneumonia prophylaxis respectively. The deletion ofGSTMI (30%) and GSTTI (40%) has been associated with increased incidence of hepatocellular carcinoma in the presence of hepatitis and aflatoxicosis, conditions that are common in Kenyan populations. The clinical implications of pharmacogenetics have been demonstrated in this thesisis in psychiatric and HN patients. These diseases afflict many Kenyans and are subject to inadequate treatment outcomes with attendant financial implications. Nevirapine phannacokinetic studies incorporating CYP2B6 516 T (44%) could define plasma levels for optimal viral suppression in Kenyan HN patients and stem antiretroviral resistance. Psychotropic dosage. titrations incorporating CYP2D6 llv1 phenotype could provide a better algorithm for dosage optimization, improved therapy outcomes and reduced incidence of extrapyramidal effects instead of the observed rampant use of anti-cholinergic prophylaxis in psychiatric in-patients. Considerable genetic variability between Kenyan and Caucasian populations was evident, yet most drugs undergo clinical evaluation in largely Caucasian cohorts. There is need for local clinical studies incorporating the pharmacogenetics of CYP2D6 1M, CYP2D6*17, CYP2B6*6, CYP219*2, GSTM1*0, GSTTI *0 and NAT2 slow acetylation to optimize dosage regimens for Kenyan populations.