Effect of Genetic Variations in Drug-Metabolizing Enzymes and Drug Transporters on the Pharmacokinetics of Rifamycins: A Systematic Review.

Background: Rifamycins are a novel class of antibiotics clinically approved for tuberculosis chemotherapy. They are characterized by high inter-individual variation in pharmacokinetics. This systematic review aims to present the contribution of genetic variations in drug-metabolizing enzymes and transporter proteins to the inter-individual variation of rifamycin pharmacokinetics. Method: We followed Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement guidelines. The search for relevant studies was done through PubMed, Embase, Web of Science, and Scopus databases. Studies reporting single nucleotide polymorphism in drug transporters and metabolizing enzymes' influence on rifamycin pharmacokinetics were solely included. Two reviewers independently performed data extraction.
Results: The search identified 117 articles of which 15 fulfilled the eligibility criteria and were included in the final data synthesis. The single nucleotides polymorphism in the drug transporters SLCO1B1 rs4149032, rs2306283, rs11045819, and ABCB1 rs1045642 for rifampicin, drug metabolizing enzyme AADAC rs1803155 for rifapentine and CES2 c.-22263A>G (g.738A>G) for rifampicin partly contributes to the variability of pharmacokinetic parameters in tuberculosis patients.
Conclusion: The pharmacokinetics of rifamycins is influenced by genetic variation of drug-metabolizing enzymes and transporters. Controlled clinical studies are, however, required to establish these relationships.

Introduction

Tuberculosis (TB) is an infectious disease, which remains a major public health problem globally. In the year 2020, the estimated number of people who died from tuberculosis is 1.3 million among HIV-negative people and 214,000 among HIV-positive.1 Current pharmacotherapy of tuberculosis involves a combination of at least four drugs. Rifamycins are key components of pharmacotherapy for both active and latent TB.

Rifamycins are a class of antibiotics isolated from Amycolatopsis in 1957. Four distinct semi-synthetic rifamycin analogs (rifampicin, rifabutin, rifapentine, and rifaximin) are approved for clinical use. Rifampicin, rifabutin, and rifapentine are used for the treatment of TB and chronic staphylococcal infections.2 Rifapentine given once weekly for 12 weeks with isoniazid is effective and well tolerated in the treatment of latent TB.3 Rifaximin is poorly absorbed from the gastrointestinal tract and is indicated for the treatment of traveler’s diarrhea, functional bloating, irritable bowel syndrome, and small bowel bacterial overgrowth.4

Variable exposure to anti-TB drugs may be associated with unfavorable treatment outcomes.5 Factors associated with drug exposure variability of anti-TB drugs, such as age, gender nutritional status, human immune-deficiency virus, diabetes, and genetic polymorphism, were described in various previous studies.6–9 There has been a notable development in recent years on how genetic variations in drug-metabolizing enzymes and transporters contribute to variation in exposure and response to the drugs.10,11 As the local and systemic concentrations of anti-TB drugs are affected by genetic variations in drug-metabolizing enzymes and transporters, pharmacokinetic and pharmacogenetic studies are increasingly performed to optimize TB treatments.12,13

Rifamycins are thought to be metabolized by microsomal hepatic carboxylesterases (CES), and serine esterase arylacetamide deacetylase (AADAC) to 25-deacetylrifamycins.14,15 The uptake, distribution, and excretion of rifampicin are mediated by membrane drug transporters. There are two transporters superfamilies; the solute carrier (SLC) transporters and the adenosine triphosphate (ATP)-binding cassette (ABC) transporters.16 SLC superfamily consists of more than 400 membrane-bound family proteins. Multiple studies revealed that the SLCO1B1 sinusoidal influx transporter influences rifampicin influx,17,18 and the SLCO1B1 *15 haplotype is associated with rifampin-induced liver injury.19 Most ABC transporters in eukaryotic cells mediate the efflux of the substrate from the cells. ABC transporters influence the hepatocellular concentration of rifampicin.20–23 Rifamycins are substrates of P glycoprotein (P-gp), coded for by the polymorphic ABCB1 gene.24 Rifampicin also induces ABCB1 gene expression.25 Although SLCO1B1 and ABCB1 gene products have been reported to influence rifamycins pharmacokinetics, there is no candidate gene identified so far for therapeutic drug monitoring.

Recently, advances in technology and scientific discoveries in the medical arena have enabled the practitioner to individualize drug therapy. The keen interest to personalize TB treatment has been a point of discussion over the last decade.26–29 The use of pharmacokinetics and pharmacogenetics of anti-tubercular drugs as tools for TB treatment optimization has been discussed previously.13,18 However, there is a scarcity of comprehensive data on the pharmacogenetics of rifamycins. This systematic review was, therefore, designed to evaluate the influence of genetic polymorphism in rifamycins metabolizing enzymes and transporters on their pharmacokinetics.

Methods

This systematic review was carried out following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statements (). The protocol has been registered at PROSPERO with registration number CRD42020206029.

Search Strategy

Relevant studies were identified through a search of PubMed, Web of Science, Embase, and Scopus databases. The following combination of words was used: pharmacokinetics OR concentration OR “drug concentration” AND rifamycins OR rifampin OR rifampicin OR rifabutin OR rifapentine OR rifaximin AND SLCO1B1 OR ABCB1 OR carboxylesterase OR CES OR Arylacetamide deacetylase OR AADAC AND “Genetic polymorphism” OR pharmacogenetics OR pharmacogenomics OR “single nucleotide polymorphisms” OR SNP. Further, a hand-search was done from reference lists of studies included to identify eligible studies. There was no limitation on the dates of publication or publication status. Publications available only in the English language were included. The search was refined to studies of human participants.

Eligibility Criteria

The following were the eligibility criteria for the inclusion of studies: 1. Human participant studies; 2. Studies that reported on pharmacokinetic parameters of rifamycins; 3. Studies in which study participants were genotyped for rifamycins metabolizing enzyme or transporters gene; and 4. Studies that reported on the pharmacokinetic parameters of rifamycins and the effect of genetic variation on pharmacokinetics.

Quality Assessment

Validated tools exist for genetic association studies methodological quality assessment. We used the quality of genetic association studies (Q-Genie)30 tool to assess the quality of included studies. Using the checklist adopted () from Q-Genie TS assessed the quality of selected studies.

Data Extraction

Two (TS and GM) independently extracted data from all included publications using a pre-prepared data extraction format which included items as follows: first author, publication year, study drug, sample size, type of pharmacokinetic parameters assessed, a country in which the study was conducted, participant characteristics, genetic polymorphism investigated, pharmacokinetic parameter results and its association with genetic polymorphism. The disparity between the two reviewers during data extraction was resolved through discussion.

No contact with the authors was done for missing data and the data presented in this review were extracted from the articles.

Results

Included and Excluded Study

A total of 115 articles related to genetic polymorphism of drug-metabolizing enzymes and drug transporters with the pharmacokinetics of rifamycins were retrieved from PubMed, Web of Science, Scopus, and Embase databases. Hand search identified two additional articles which were not obtained during the database search. As shown in the PRISMA flowchart (Figure 1) 51 duplicates were removed. The remaining 66 articles were screened by title and abstract for predefined criteria, and 47 were excluded. The reasons for exclusion of studies from titles and abstracts were (1) review articles (N=3); (2) studies focusing on drugs other than rifamycins (N=26); (3) studies that did not have information on the pharmacokinetics of rifamycins but only genetic information reported (N=8); and (4) studies in which only pharmacokinetics data were reported without genetic information (N=10). Furthermore, four articles were excluded after reading them fully. Of the four articles excluded; one article did not contain rifamycins data, one study was done on healthy participants and the other two articles did not contain pharmacokinetic parameters.

Figure 1

PRISMA flow diagram showing the literature search for studies that investigated the effect of genetic variations in drug metabolizing enzymes and drug transporters on the pharmacokinetics of rifamycins.

Characteristics of Included Studies

Of the 15 articles selected for qualitative data synthesis, most of the studies (N=14) focused on SLCO1B1 gene polymorphism association with the pharmacokinetics of rifamycins (). Specifically, seven studies evaluated the association of SLCO1B1 gene polymorphism and pharmacokinetics,31–37 three studies SLCO1B1 and ABCB1 gene polymorphism with pharmacokinetics,38–40 one study SLCO1B1 and AADAC gene polymorphism with pharmacokinetics,41 one study SLCO1B1, and CES gene polymorphism with pharmacokinetics,42 and two studies SLCO1B1, AADAC, and CES gene polymorphism with pharmacokinetics.43,44 Only one study investigated the association between CES gene polymorphism with pharmacokinetics.45 The most studied rifamycins are rifampicin (thirteen studies) and rifapentine (two studies). No study is available that reported the pharmacokinetic-pharmacogenetic association for rifabutin and rifaximin.

There was variation among studies in sample size, the type of study participants, and the pharmacokinetics parameter compared with gene polymorphism. The smallest sample size was 34,39 while the largest was 256.34 The study participants were TB patients from 13 different countries and races. The majority of the studies were done on adults, but one study data were obtained from children.42 In some studies, participants were TB-HIV co-infected patients. The pharmacokinetics parameters commonly compared with gene polymorphism were maximum concentration (Cmax), AUC (area under the curve), and clearance. However, methods for blood sample collection and pharmacokinetic parameter determination varied among studies.

Association Between Drug Transporter and Rifamycins Pharmacokinetics

Association Between Polymorphism of SLCO1B1 and Rifamycins Pharmacokinetics

SLCO1B1 gene encodes for an Organic Anion Transport Proteins 1B1 (OATP1B1). It is located on chromosome 12. OATP1B1 is a transmembrane protein involved in the uptake of various drugs including rifamycins from the blood into the hepatocyte.46 Currently, 191 clinical variants have been reported. SLCO1B1c.521T>C (rs4149056), where the valine amino acid changed to alanine at position 174, was reported to affect drug response.47 Eight studies assessed the effect of rs4149056 SNPs on rifamycin pharmacokinetic parameters. Among these studies, only Huerta-García et al reported increased AUC among heterozygous CT for SLCO1B1 521T>C than the other genotypes. However, the observed increase in AUC was not statistically significant.39 A summary of specific transporters influence on pharmacokinetics is presented in Table 1.

Table 1

Summary of the Studies Reported the Drug Transporter (SLCO11 and ABC1B) Gene Polymorphisms Association with Rifamycins Pharmacokinetics Variation

Reference	Gene	SNPs	Characteristics of Study Participant	Rifamycins PK Change Observed
[31]	SLCO1B1	rs2306283rs4149032rs4149056rs4149015	Tuberculosis recurrent black South African of which 127 (73.8) are HIV positive	No significant association between rifampicin pharmacokinetic and all variants of SLCO1B1 gene SNPs studied was observed
[43]	SLCO1B1	rs11045819rs4149032	174 Malawian adults with pulmonary TB of which 98 are HIV-infected patients	No association was reported for both variants of SLCO1B1 gene SNPs studied and the pharmacokinetics of rifampicin
[32]	SLCO1B1	rs4149032	57 newly diagnosed TB-HIV co-infected South African patients	Lower median concentration of rifampicin at 2.5hr; 3.7 μg/mL in heterozygous and 3.4μg/mL in homozygous variants
[38]	SLCO1B1	rs4149056rs2306283	Adult tuberculosis patients 57 study group of 30% are diabetics and 27 validation group of 27% are diabetics	No variation of rifampicin volume of distribution or clearance was observed for both SLCO1B1 gene A388T (rs2306283) and T521C (rs4149056).
	ABCB 1	rs1045642		No effect of rs1045642 SNP on rifampicin pharmacokinetics was observed
[33]	SLCO1B1	rs4149032	100 tuberculosis patients where 50 are HIV positive	No effect of SLCO1B1 rs4149032 genotype on rifampin Median Cmax and Median AUC0–24 was observed
[34]	SLCO1B1	rs4149032rs4149033rs11045819	256 adult tuberculosis patients from India	No significant difference in 2 hr rifampicin plasma concentration for all SNPs studied was observed
[39]	SLCO1B1	rs4149056	34 tuberculosis patients of which 41.2% are diabetics and some are taking other drugs	AG genotype of SLCO1B1 388A>G had lower rifampicin AUC0–24 h compared to AA genotype (83.42 mcg.h/mL versus 108.31 mcg.h/mL) respectively
	ABCB 1	rs1045642 (3435C>T)		Patients with CC or CT genotypes showed lower values in Cmax, and AUC0–24 h compared to those with a TT genotype (Cmax = 9.1 6 mcg/mL versus 15.8 6 mcg/mL; AUC0–24 h = 72.83mcg.h/mL versus 130.35 6 29.5 mcg.h/mL respectively)
[42]	SLCO1B1	rs2306283rs11045819rs4149056rs4149032	113 children aged 3 months to 14 years and 59 (52.2%) were HIV co-infected	In post hoc analysis, the rare SLCO1B1 c.388AA genotype was associated with lower rifampicin Cmax (1.81µg/mL versus 7.11 µg/mL) and AUC0–8h (9.33 µg.h/mL versus 29.50 µg.h/mL) and higher CL/F and V/F compared to those with c.388GG
[40]	SLCO1B1	rs4149032rs4149056rs11045819	60 adult tuberculosis patients aged from 18 to 55 years and 16% were HIV infected.	Patients heterozygous and mutant homozygous for rs4149032 had 18% and 28% reductions in the bioavailability of rifampicin respectively.
	ABCB 1	rs1045642rs2032582rs1128503rs3842		The ABCB1 G2677T (rs2032582) showed no statistically significant increase (19%) in the CL/F and a 19% increase in the mean transit time
[41]	SLCO1B1	rs2306283rs4149032	162 pulmonary tuberculosis from two clinical studies receiving rifapentine in South Africa	No effect on oral clearance, apparent volume of distribution, and F was detected
[37]	SLCO1B1	rs2306283rs11045819rs4149056rs4149032	105 adult patients were newly diagnosed with active pulmonary TB, and Twenty (19%) patients had diabetes mellitus	rs4149032 wild type (TT) had lower oral clearance and higher AUC but no statistically significant differences were detected
[35]	SLCO1B1	rs4149032rs2306283	A cohort of 50 HIV negative patients 25 with rifampicin sensitive pulmonary TB and 25 patients with rifampicin-resistant	When adjusted for all covariates no significant effect of the two SLCO1B1 genotypes on rifampicin pharmacokinetics parameters was identified
[36]	SLCO1B1	rs11045819rs4149056rs59502379rs2306283rs4149015	72 TB patients (37 from Africa and 35 from the United States and Spain) and 16 healthy controls from USA	Patients with the SLCO1B1 c.463C>A (rs11045819) polymorphism had 42% lower rifampicin AUC0–24, 34% lower Cmax, and 63% CL/F
[44]	SLCO1B1	rs2239751 rs2306283rs11045819 rs4149014rs4149032 rs4149056	173 adults of different races and countries of origin of which 12 are HIV positive	None of the SLCO1B gene polymorphism investigated were associated with rifapentine exposure (AUC 24hour)

Abbreviations: AUC, area under curve; PK, pharmacokinetic; SNP, single nucleotide polymorphism; Cl, clearance; F, bioavailability; Cmax, maximum concentration; CL/F, apparent oral clearance; V/F, apparent predicted volume of distribution.

SLCO1B1 g.38664C>T (rs4149032) was reported in twelve studies. rs4149032 is an intronic SNP most common in the African population.48,49 Gengiah et al reported high frequency in the SLCO1B1 (rs4149032) gene polymorphism and its association with low median rifampicin C2.5hr in the heterozygous and homozygous variant carriers.32 Similarly, Chigutsa et al reported high allelic frequency of the SLCO1B1 rs4149032 polymorphism and 28% reductions in the bioavailability of rifampin for homozygous variants.40 No statistically significant increase in the rifampicin exposure for the homozygous TT of g.38664 C > T (rs4149032) was observed in the study of Kim et al.37 However, the large number of studies reviewed here did not report any observed significant effect of SLCO1B1 rs4149032 SNP polymorphism with rifamycin pharmacokinetic variation.

SLCO1B1 c.388A>G (rs2306283) is another SNP in the SLCO1B1 gene. This SNP causes a change of asparagine amino acid to aspartic at 130, but the effect of this change on the transporter function is not clear yet. Huerta-García et al reported the AG genotype derived from SNP SLCO1B1 c.388A>G was associated with lower rifampicin AUC0–24 h values compared to those with AA genotype.39 In post hoc analysis, Dompreh et al observed that the SLCO1B1 c.388AA genotype was associated with low rifampin concentrations compared to those with c.388GG.42 The five remaining studies did not report any association between rs2306283 SNP and rifamycin pharmacokinetics. The SNP SLCO1B1 c.463 C>A (rs11045819) is another variant allele of the SLCO1B1 gene reported to affect rifamycin pharmacokinetics. According to Weiner et al, patients with SLCO1B1c.463C>A variant allele had 42% lower rifampin exposure, 34% lower peak concentration levels, and 63% greater apparent oral clearance compared with SLCO1B1 c.463CC.36 However, the remaining five studies did not report any association between rs11045819 SNPs and rifamycin pharmacokinetics.

Association Between Polymorphism of ABCB1 and Pharmacokinetics

ABCB1 (ATP-binding cassette sub-family B member 1) genes encode for P-gp also known as multidrug resistance protein 1 (MDR1). P-gp is a transmembrane protein, which acts as an energy-dependent drug efflux pump. It decreases intracellular drug accumulation, thereby decreasing the effectiveness of many drugs.50 The ABCB1c.3435 C>T (rs1045642), ABCB1c.G2677 T/A (rs2032582) and ABCB1c.1236C>T (rs1128503) SNPs are the most common nonsynonymous and synonymous SNPs studied.51 Rifamycins are a substrate and inducer of the ABCB1 gene.52 The decrease in rifampicin exposure with the time of treatment is partly explained by the induction of the ABCB1 gene. Three studies assessed the effect of four ABCB1, rs1045642 rs2032582, rs1128503, and rs3842 (ABCB1c.4036A>G) SNPs. Huerta-García et al demonstrated that the rs1045642 TT genotype is a predictor that explains 34.8% of the variability in rifampicin Cmax and 48.5% of the variability in AUC0–24 h.39 However, the other two studies did not replicate this observed result of Huerta-García et al.38,40

Association Between Drug-Metabolizing Enzyme and Pharmacokinetics

Rifamycins are metabolized by esterase enzymes. The esterase enzymes implicated in the metabolism of rifamycins are hepatic carboxylesterases (CES), and serine esterase arylacetamide deacetylase (AADAC). Two carboxylesterases, CES1 and CES2, are recognized to play major roles in drug metabolism. These enzymes metabolize rifamycins to their respective deacetylrifamycins.14,15,53 Polymorphism of the CES1 and CES2 genes have been shown to influence the metabolism of several drugs.54 However, few studies investigated the effect of CES1 and CES2 gene variants on rifamycin metabolism (Table 2).

Table 2

Summary of the Studies Reported the Drug-Metabolizing Enzyme (AADAC and CES) Gene Polymorphisms Association with Rifamycins Pharmacokinetics Variation

Reference	Gene	SNPs	Characteristics of Study Participant	Rifamycins Pharmacokinetics
[43]	CES 1	rs12149368	174 Malawian adults with pulmonary tuberculosis of which 98 are HIV-infected patients	No associations between rifampicin AUC, Cmax, (CL/F), or V/F and AADAC or CES-1 SNPs polymorphism were identified
	AADAC	rs1803155rs61733692
[42]	CES2	rs3759994	113 children aged 3 months to 14 years and 59 (52.2%) were HIV co-infected	No significant effect of studied CES2 SNPs on rifampicin Cmax, AUC, and CL/F was observed
[41]	AADAC	rs1803155	162 pulmonary tuberculosis patients from two clinical studies receiving rifapentine in South Africa	Patients carrying the AA variant of AADAC rs1803155 were found to have a 10.4% lower rifapentine clearance
[45]	CES2	c.-2548C>Tc.-2263A>Gc.269–965A>Gc.474–152T>Cc.615+120G>Ac.1612+136G>Ac.1613–87G>Ac.1872*69A>Gc.1872*302_304delGAAc.1872*445C>T	35 patients with tuberculosis receiving a first-line antituberculosis treatment and 100 healthy individuals for analysis of the frequency of genetic variations in CES2 in the general population	The plasma rifampicin concentration increased with the number of risk alleles at c.2263A>G, c.269–965A>G and c.1612+136G>A, for example for c.2263A>G 8.9 mg/L versus 13.9mg/L for GG and AA respectively, while the plasma concentration decreased along with an increase in the number of risk alleles at c.1872*302_304delGAA
[44]	AADAC	rs1803155	173 adults of different races and countries of origin of which 12 are HIV positive	Rifapentine exposure (AUC 24) decreased by 10.2% in black participants for AADAC rs1803155 G versus A allele
	CES2	rs8045523rs8192925rs4783745		17.2% increase in rifapentine AUC0-24 for rs8192925 G versus A was observed

Abbreviations: AUC, area under the curve; CES, carboxylesterases; AADAC, arylacetamide deacetylase; Cmax, maximum concentration; CL/F, apparent oral clearance; V/F, apparent predicted volume of distribution.

Sloan et al investigated CES1 rs12149368 SNP effect on rifampicin pharmacokinetics in Malawian tuberculosis patients. The rs12149368 variant does not affect the plasma rifampicin concentration43 (Table 2). Song et al identified 10 variations in CES2 in Korean TB patients. Among the ten variants three closely linked SNPs, c.-2263A>G (rs3759994, g.738A>G), c.269–965A>G (rs4783745, g.4629A>G), and c.1612+136G>A (g.10748G > A), may alter the metabolism of rifampicin by affecting the efficiency of transcription of the gene. In particular, the CES2 c.-2263A>G variant, which is found in the promoter region is associated with increased plasma concentrations of rifampicin.45

Shimazu et al reported that microsomes from a liver sample genotyped as AADAC*3/AADAC*3 showed decreased enzyme activities, compared with others. However, the allelic frequency is low, 1.3% European American, and 2.0% African American. The AADAC*2 (rs1803155) allele, which has a higher frequency has also shown reduced enzyme activity. The recent report of Francis et al and Weiner et al revealed that rs1803155 SNP has a significant effect on rifapentine exposure in tuberculosis patients. The mean AUC-24 of rifapentine decreased by 10.2% in black tuberculosis patient carriers of AADAC rs1803155 G versus A allele.44 The odds increase for GG allele carriers. A similar result was reported by Francis et al. Patients carrying the AA variant of AADAC rs1803155 were found to have a 10.4% lower clearance of rifapentine.41 However, another study from Malawi showed that AADAC rs1803155 SNP did not affect rifampicin pharmacokinetics.43

Discussion

This systematic review provides current updates on the impact of genetic polymorphisms of drug transporters and drug-metabolizing enzymes on the pharmacokinetics of rifamycins. The overall finding suggests that the polymorphism in the drug transporter SLCO1B1 rs4149032, rs2306283, rs11045819, and ABCB1 rs1045642 and metabolizing enzyme AADACrs1803155 and CES2 c.-22263A>G (g.738A>G) of rifamycins partly contributes to the variability of pharmacokinetic parameters in tuberculosis patients.

The SLCO1B1 gene is located on chromosome 12. Fifteen exons and many variants have been identified in the SLCO1B1 gene. The missense mutation of rs4149056 (c.521T>C) where the wild type T is substituted with variant C causes a change in amino acid of OATP1B1 protein from valine to alanine at 174 positions. This change has been implicated in reduced OATP1B1 protein function and is associated with an increased risk for statin-induced muscle toxicity.55 However, an increase in the exposure to rifamycins was not reported in seven studies, and the one study, which reported an increase in AUC for the heterogeneous variant is also statistically non-significant. Lower frequency of rs4149056 CC variant in African populations56 where the majority of studies were done and small sample size may contribute to no difference in the pharmacokinetics. rs2306283 (388A>G) SNP causes a change of asparagine amino acid to aspartic at 130 positions. The consequence of this change on the transporter function is not well elucidated. The patients who were homozygous wild type (AA)42 and heterozygous (AG)39 were reported to have lower rifampicin exposure. Similarly, no myopathy was observed with rs2306283 polymorphism which was observed in other SLCO1B1 genes in patients taking statins suggesting no effect or increased activity of the mutant variant.57

rs11045819, which is located on exon 4, is another missense variant known in SLCO1B1gene. Of the four studies that assessed the impact of rs11045819 SNPs on rifampicin pharmacokinetics, only Weiner et al reported lower rifampicin exposure, lower peak concentration levels and greater apparent oral clearance with the SLCO1B1 rs11045819 variant allele (CA) compared to the wild-type allele (CC).36 This is consistent with a previous report that rs11045819 polymorphism increases OATP1B1 transporter activity and decreases systemic exposure of the OATP1B1 substrate.58,59

The well-studied SLCO1B1 gene SNPs believed to affect rifamycin pharmacokinetics is rs4149032. The rs4149032 is an intron-located SNP and is reported to have a high allelic frequency. The effect of SLCO1B1 rs4149032 on gene expression and OATP1B1 protein transporter function is not clear yet. Nevertheless, SLCO1B1 rs4149032 polymorphism was found to be associated with lower rifampicin exposures. Emmanuel et al and Gengiah et al reported that patients who are homozygous mutant and heterozygous for rs4149032 polymorphism have lower bioavailability and Cmax respectively of rifampicin.32,40 In addition, Kim et al observed lower oral clearance and higher rifampicin exposure for rs4149032 homozygous wild type (TT).37

Rifampicin significantly increases gene expression, protein levels, and efflux activity of ABCB1.25,60 It is also a substrate for P-glycoprotein.61 Huerta-García et al demonstrated that the rs1045642 SNPs, which is a silent mutation, is associated with rifampicin pharmacokinetics. Patients with CC or CT genotypes showed lower values of Cmax and AUC 24 compared to those with a TT genotype.39 Although the rs1045642 SNPs is a silent mutation, previous studies have shown that rs1045642 affects the P-gp protein either by being in linkage disequilibrium with other functional SNPs or by allele-specific differences in the codon usage affecting the protein folding and function.62,63 The observed change in the rifampicin pharmacokinetics with rs1045642 SNPs may be attributed to the above explanation.

Rifamycins are metabolized by the esterase enzyme family; microsomal hepatic carboxylesterases (CES), and serine esterase arylacetamide deacetylase (AADAC) to 25-deacetylrifamycins.14 Three esterase enzymes AADAC, CES1, and CES2 have been reported as enzymes responsible for rifamycin deacetylation. Several genetic polymorphisms of the CES1 and CES2 genes have been shown to affect drug metabolism. For example, variations of the CES1 gene have been reported to affect the metabolism of dabigatran oseltamivir, imidapril, and clopidogrel. Similarly, CES2 gene polymorphisms have been found to affect aspirin and irinotecan.54 Few studies are available that report the association of CES1 and CES2 variants and rifamycin pharmacokinetics. Song et al evaluated 10 SNPs of CES2 and found increased plasma rifampicin concentrations with the CES2 c.-22263A>G (g.738A>G) variants.45 Although Dompreh et al did not report similar results,42 the higher frequency of this variant allele warrants further investigation.

AADAC is primarily expressed in the liver and metabolizes clinically important drugs including rifamycins. Three, namely, AADAC*1 (wild-type), AADAC*2, and AADAC*3, where the latter two have decreased enzymatic activity, were reported so far.14,15 Recently, Francis et al and Weiner et al reported AADAC rs1803155 SNPs to have a significant effect on rifapentine metabolism. Shortly, a mutant variant of rs1803155 (AA) has decreased activity and decreased clearance of rifapentine. On the other hand, patients who have the wild type (GG) have shown decreased rifapentine exposure.41,44 Furthermore, Gabriele et al discovered the presence and inter-individual variation of AADAC in the human lung.64 These findings suggest the important role of AADAC pharmacogenetics in tuberculosis drug therapy.

Exposure to rifamycins in particular rifampicin is a crucial variable for successful tuberculosis treatment outcomes. The high inter-individual variability in rifamycins pharmacokinetics have been associated with various factors such as diabetes mellitus65 and partly HIV co-infection.66,67 The majority of studies included in this review included patients with co-morbid conditions. The sample size is also inadequate for some studies.

In conclusion, the genetic polymorphism of drug transporters and drug-metabolizing enzymes has an impact on rifamycin pharmacokinetics. However, based on the available data, it is difficult to identify candidate SNPs in the drug transporters SLCO1B1 and ABCB1 for therapeutic drug monitoring. On the other hand, the effect of drug-metabolizing enzyme SNPs on the rifamycin pharmacokinetics is promising but needs more studies. In general, further controlled clinical studies with adequate sample size are required to characterize the genetic variation influence on the pharmacokinetics of rifamycins for tuberculosis chemotherapy optimization.