Genetic polymorphisms affect the metabolism of antituberculosis drugs

Introduction

According to the UN Sustainable Development Goals, eliminating the tuberculosis epidemic by 2030 is a priority in the field of healthcare [1, 2]. Despite the fact that in 2021 the Russian Federation left the list of countries with a high burden of tuberculosis, the threat of the spread of drug-resistant tuberculosis and co-infection (HIV infection/tuberculosis) remains. Patients with multidrug-resistant tuberculosis (MDR-TB) and HIV infection are the most difficult cohort to treat. The need to use a large number of drugs is accompanied by drug-drug interactions, which increase the risk of adverse events, mainly of toxic genesis, and interruptions in treatment, which entails an expansion of the spectrum of drug resistance of the pathogen, a decrease in the effectiveness of treatment and the preservation of a reservoir of infection in society [3]. Among the adverse events, the leading place was occupied by hepatotoxic reactions (59.3% [4], 76.2% [5], 9.0–27.4% [6]. The frequency of their occurrence depends on several factors, among which the leading one is the combination of anti-tuberculosis drugs (ATDs) and concomitant pathology [7]. Achievements in genetics and genetic research methods have shown that the safety profile of pharmacotherapy and the creation of a bactericidal concentration of the active substance in the patient's blood serum are significantly affected by some genetic polymorphisms, including liver enzymes involved in the metabolism of ATDs [8–16].

In some cases, gene polymorphism becomes an obstacle to taking anti-TB drugs in doses prescribed for the prevention of tuberculosis infection due to isoniazid-induced liver damage [2.6% (95% CI 1.7–3.7%), of which mortality is 0.02%] [17]. Knowledge of this and the ability to influence the implementation of genetic information open up new opportunities for improving the quality of medical care for patients receiving anti-TB drugs for therapeutic and prophylactic purposes. In connection with the above, the purpose of this work was to review studies devoted to studying the effect of polymorphic markers of genes of enzymes involved in the metabolism of anti-TB drugs on the safety profile of such drugs.

Materials and methods

The clinical guidelines "Tuberculosis in adults" (2024) list anti-TB drugs, the use of which may be accompanied by the development of hepatotoxic reactions: isoniazid, rifampicin and its derivatives, pyrazinamide, prothionamide/ethionamide, bedaquiline, thioureidoiminomethylpyridinium, para-aminosalicylic acid, and amoxicillin/clavulanic acid. In addition, there are reports in the literature of such reactions when using fluoroquinolones [18], ethambutol [19], linezolid [20], and delamanid [21]. An online search of articles in the Medline PubMed (https://www.ncbi.nlm.nih.gov/pubmed/), Embase (https://www.elsevier.com) and Google Scholar (https://www.scholar.google.ru) databases was conducted using the names of the above-mentioned TB drugs together with the terms “tuberculosis” and “polymorphism”. Sixty-two publications were analyzed. A separate search of information was conducted in the leading international resource on pharmacogenomics, the Pharmacogenomics Knowledge Base (https://www.pharmgkb.org/) to assess the levels of evidence for certain recommendations on drugs obtained on the basis of the results of pharmacogenetic studies.

To obtain instructions on dose adjustment depending on the patient's genotype, the instructions for the medical use of the drugs included in the review were analyzed in the Russian state register of medicines (https://grls.rosminzdrav.ru/Default.aspx) and similar registries of other countries (FDA, USA, EU, etc.).

A chemotherapy regimen consisting of isoniazid, rifampicin, pyrazinamide, and ethambutol is used to treat drug-sensitive tuberculosis.

Results and discussion

Most studies on the effect of the genetic polymorphism of liver enzymes on the metabolism of anti-TB drugs have focused on rifampicin or isoniazid.

Isonicotinic acid hydrazide (IAH) is used for treating drug-susceptible TB and remains the standard for assessing the efficacy of newly developed drugs in scientific research. Its uniqueness is due to its bactericidal action against M. tuberculosis [22]. Despite this, the metabolism of IAH is still the subject of research, much of which has focused on the pharmacodynamic and toxicological aspects of its metabolites [23]. In the liver, IAH is metabolized by the enzymes N-acetyltransferase-2 (NAT2), cytochrome P450 2E1 (CYP2E1), and glutathione-S-transferase (GST) with two isoforms, GSTT1 and GSTM1.

The formation of metabolites is a result of hydrolysis, cytochrome P450 (CYP)-dependent oxidation, and N-acetyltransferase (NAT) activity. The hepatotoxicity of HINA is associated mainly with metabolites that are capable of generating free radicals and the ammonia they release [22]. The rate of NAT2 acetylation is determined genetically as an autosomal recessive trait [24]. Depending on the NAT2 activity, three main phenotypes can be distinguished: “slow” acetylator (carriers of homozygous genotypes for alleles associated with reduced activity of the isoenzyme), “fast” acetylator (carriers of two alleles associated with equal or higher activity of the isoenzyme compared to the “wild” type allele), and “intermediate” acetylator (carriers of heterozygous genotypes) [25–27]. Due to the reduced rate of metabolism, "slow" NAT2 acetylator are exposed to the increased effects of both isoniazid itself and its toxic metabolites, which explains the increased risk of hepatotoxic reactions. Carriers of the wild-type alleles of NAT*4, polymorphic alleles of NAT2*12 (rs1208, 803A>G) and NAT2*13 (rs1041983, 282C>T) form the phenotype of "fast" acetylator [27], which, on the contrary, can determine the insufficient effectiveness of isoniazid therapy due to the rapid metabolism of the drug to inactive metabolites, but does not exclude the development of hepatotoxicity. Studies devoted to the study of the pharmacokinetics of anti-TB drugs have demonstrated the need for dose adjustment within the limits of its therapeutic effect depending on the type of acetylation and the age of the person [28].

The most significant markers in terms of the safety profile of isoniazid are NAT2*5 (rs1801280, 341T>C), NAT2*6 (rs1799930, 590G>A), NAT2*7 (rs1799931, 857G>A) and NAT2*14 (rs1801279, 191G>A) - their carriers form the phenotype of "slow" acetylator [25]. Similar results were demonstrated in the meta-analysis of Khan S et al. [29]: an increased risk of drug-induced liver damage is observed in carriers of mutant alleles in position NAT2*5 (rs1799929), NAT2*6 (rs1799930) and in position NAT2*7 (rs1799931). The association between the NAT2 polymorphism (NAT2*13A and NAT2*6B) and the development of hepatotoxic reactions in people living with HIV/AIDS [30] and children - NAT2*13 (rs1041983, 282C>T) [31] has been proven.

Another important metabolizer of anti-TB drugs is the cytochrome P450 system. Of particular interest are the genetic markers of the CYP2E1 enzyme, as well as the study of the relationship between the genetic polymorphisms of N-acetyltransferase 2 (NAT2) and glutathione-S-transferase (GST). For patients with the slow acetylation genotype of NAT2, the CYP2E1 RsaI/PstI c1/c1 genotype, and the null GSTM1 genotype, the risk of drug-induced liver injury is significantly increased, which requires a thorough examination before prescribing anti-TB drugs [32, 33]. In the work of Bose PD et al. [34], an association was determined between the carriage of markers NAT2*5, NAT2*6, NAT2*7 and the RsaI variant of the CYP2E1 gene (*5B, rs 2031920, C-1054T) with hepatotoxicity in patients receiving therapy for drug-sensitive tuberculosis.

The study [34] convincingly demonstrated the development of hepatotoxic reactions only in “slow” acetylator with NAT2*6, NAT2*7, and the RsaI allelic variant of the CYP2E1 gene (CYP2E1*5B). In the cohort study by Singla N et al. [35], it was indicated that heterozygous carriage of the CYP2E1*5B allelic variant may contribute to an increased risk of hepatotoxicity against the background of PTP. Similar data were obtained in another study, which assessed the risk of hepatotoxicity depending on the carriage of NAT2 and CYP2E1 allele variants (1024T>C, 1053C>T, 1293G>C). When assessing the separate role of CYP2E1 markers without NAT2, no significant association with hepatotoxicity was found; however, the combination of “slow” acetylation by NAT2 and carriage of the TT homozygote for the 1053C>T gene CYP2E1 significantly increased the risk compared with “fast” acetylator carriers of the same CYP2E1 genotypes [36].

Other data were obtained in the study by Xiang Y et al., which showed a significant association with the development of drug-induced liver injury for carriers of the NAT2*5 (rs1799929) marker, whereas there was no such association for NAT2*6 (rs1799930) or NAT2*7 (rs1799931). A significant effect of the CYP2E1, GSTM1 or GSTT1 markers was also not proven [37]. The role of CYP2E1 gene polymorphisms and null mutations in the GSTM1 and GSTT1 genes in the manifestation of drug-induced liver injury during tuberculosis treatment was separately studied in the work of Santos EA et al. It was shown that the carriage of at least one mutant allele for CYP2E1 or a null mutation of GSTT1 increases the risk of liver injury by 4.5 times [38]. An observational study by Yu YY et al. [39] demonstrated a correlation between the presence of the allelic variants rs2070676 and rs2515641 CYP2E1 and a higher frequency of adverse events in individuals receiving isoniazid and rifampicin. A number of studies have shown that the GSTM1 and GSTT1 genotype in “fast” NAT2 acetylator with the presence of the CYP2E1 c1/c1 variant increases the risk of developing adverse events associated with taking anti-TB drugs [40], and the slow genotype NAT2, CYP2E1*1A and null GSTM1 have a moderate effect on genetic susceptibility to the development of hepatotoxic reactions [41].

Despite the conflicting data from individual studies on CYP2E1 in the context of drug-induced liver injury during anti-tuberculosis therapy, there are data from a meta-analysis by Liu X et al., where the authors analyzed the results of 29 studies on the association of carriage of CYP2E1 allelic variants with liver injury during combination therapy with isoniazid, rifampicin, pyrazinamide and ethambutol. A significant association of the risk of developing hepatotoxic reactions with the carriage of the CYP2E1 gene polymorphism during combination therapy with first-line drugs was proven only for the RsaI/PstI variants [23]. At the same time, information on a significant association of hepatotoxic reactions with polymorphisms of other genes of cytochrome P450 enzymes remained contradictory. Thus, in a case–control study by Wang Y et al. The role of CYP2B6*6 (rs3745274, c.516G>T) and CYP2B6*4 (rs2279343, c.785A>G) polymorphisms in the development of hepatotoxic reactions was assessed in a sample of 343 patients with tuberculosis who received treatment for drug-susceptible tuberculosis for 2 months. Interest in CYP2B6 was caused by the fact that this enzyme is involved in the metabolism of a wide range of drugs, and in anti-tuberculosis therapy, the fact that it can be significantly induced by rifampicin plays an important role. The data obtained indicated that the homozygous CYP2B6*6 genotype was significantly associated with a reduced risk of developing hepatotoxic reactions among men, whereas no associations were found for CYP2B6*4 [31]. Supporting data on the role of CYP2B6*6 as a safety marker of anti-tuberculosis therapy have also been provided in several studies [42, 43].

It has been noted that liver glutathione S-transferases (GST) catalyze the sulfhydryl conjugation of hepatotoxic intermediates of xenobiotic metabolism, which reduces their toxic effects by eliminating them from the body. The absence of GST activity due to a null mutation in two of its loci, GSTM1 and GSTT, cann lead to the accumulation of toxic intermediates and the resulting liver damage, which makes a person susceptible to the development of drug-induced liver injury [44]. In the work of Perwitasari DA et al. [45] The effect of carriage of allelic variants of the genes NAT2 (rs1799929, rs1799930, rs1799931, rs1801280 and rs1041983), CYP2E1 (rs2031920, rs8192775 and rs2515641), HLA (rs1041981, rs1063355 and rs6906021), null mutation in the genes of glutathione-S-transferases GSTM1 and GSTT1 on the concentration of isoniazid in plasma and their association with the risk of developing drug-induced liver injury was studied in patients with newly diagnosed tuberculosis. The results of this study demonstrated that the greatest number of patients with liver injury had the genotype of "slow" acetylator for NAT2 - 48.9%. No significant association with the risk of hepatotoxicity was found for the polymorphic markers CYP2E1, GSTT1, and GSTM1. Interesting data were obtained for the markers of the HLA gene: carriers of the G allele rs1063355 and the C allele rs1041981 had a reduced risk of developing hepatotoxic reactions compared with non-carriers [45]. Similar data were obtained in the work of other authors, where only the phenotypic profile of the acetylator for Nhad had a significant effect on the risk of developing liver damage during anti-tuberculosis therapy, and the effect of the markers CYP2E1, GSTM1, or GSTT1 on the type of acetylation was not proven [46]. According to the PharmGKB resource, the only marker for which a high (1B) and moderate (2A) level of evidence of the relationship between carriage and toxicity was revealed is the NAT2 gene. Such an association is mainly determined for isoniazid and its combinations with other first-line drugs [47]. With regard to other genes and their significance for the safety profile of anti-tuberculosis drugs, the levels of evidence for such associations do not exceed level 3. Such an association may be based on one study (or preliminary information) or several studies with conflicting information [22]. At the same time, a number of authors have shown that the carriage of a null mutation of only GSTM1 or a combination of null mutations at the GSTM1 and GSTT1 loci was significantly associated with the development of hepatotoxicity (p < 0.02 and p < 0.007, respectively) [48]. These findings correlate with the results of the work of Singla N et al. [35]. For rifampicin, the main metabolic enzyme is cytochrome P450. In a meta-analysis in the study [23], the odds of detecting the fact of CYP2E1 (RsaI/PstI) polymorphism in patients with hepatotoxic reactions (odds ratio when using regimens for the treatment of drug-susceptible tuberculosis with the inclusion of rifampicin when compared with GINC monotherapy were insignificant (OR = 1.18 [95% CI: 0.82–1.71]). Ethambutol is metabolized by the cytochrome P450 (CYP) system, glucuronosyltransferase (UGT), aldehyde dehydrogenase (ALDH), and glutathione S-transferase (GST) [19]. The frequency of hepatotoxic reactions of ethambutol is lower than that of other anti-TB drugs, and the study [19] convincingly demonstrated that the null genotype of GSTM1 rs4025935 is associated with an increased risk caused by anti-TB drugs used to treat drug-sensitive tuberculosis. In addition, the rs7852860 variants in the ALDH1A1 gene are associated with an increased risk of developing hepatotoxic reactions. At the same time, the UGT2B7 rs7662029 variant with the AG genotype may be a protective factor against the development of toxic reactions. Another study [15] convincingly demonstrated the effect of the AG/A mutation in CYP1A2 2159 G>A on a 50% decrease in the relative bioavailability. Modeling showed that doses of 30 mg/kg body weight and 50 mg/kg for G/G and G/A carriers, respectively, will lead to clinically adequate exposure, which requires a 2-fold increase in the dose instead of the recommended ones (15–20 mg/kg to 30 mg/kg). However, the 50 mg/kg dose required to achieve the therapeutic effect in G/A carriers may be impractical due to the dose-dependent toxicity of ethambutol.

For pyrazinamide, the main enzyme of its inactivation is deamidase and xanthine oxidase of the liver. No studies on the genetic polymorphism of this enzyme were found in the available literature. All studies concerned the study of cytochrome P450 when using a combination of PTPs for treating drug-sensitive tuberculosis.

No studies were found on the effect of polymorphism of enzymes involved in the metabolism of protionamide on the development of hepatotoxic reactions. In the study [13], the effect of PTPs on eight CYP-specific reactions in human liver microsomes was studied. Taking into account the treatment regimens for tuberculosis, the most significant drug interactions with a combination of clofazimine and protionamide were polymorphisms in the CYP3A4 and CYP2B6 genes, and for isoniazid and rifapentine - CYP3A4 [49]. The metabolism of levofloxacin is influenced by the polymorphism of cytochrome P450 2C19 (CYP2C19), which mainly determines its bactericidal concentration in the human body [50]. The pharmacokinetic parameters of fluoroquinolones, such as the area under the curve (AUC), creatinine clearance (CCr), maximum plasma concentration (Cmax), half-life (T1/2) and peak time (Tmax), are significantly influenced by the polymorphisms of the genes of glucuronosyltransferase (UGT: UGT1A1, UGT1A9), ATP-binding cassette of P-glycoprotein (encoded by the ABCB1 gene), and solute carrier family of organic anion-transporting polypeptide 1B1 (encoded by SLCO1B1) [18].

The metabolism of bedaquiline included in MDR-TB chemotherapy regimens was influenced by a polymorphism in the AGBL4 gene, which determined the clearance of the drug, the risk of developing hepatotoxic reactions, and its therapeutic concentration [14]. A study conducted among South Africans [51] demonstrated an association of the rs776746 CYP3A5*3 polymorphism with a slower elimination of bedaquiline. Delamanid is metabolized through the CYP system, and at a therapeutic dose it did not induce CYP1A2, CYP2C9, and CYP3A4 activity in human hepatocytes, and no increase in CYP1A2, CYP2B6, CYP2C9, and CYP3A4 mRNA levels was observed. This was confirmed by Kour G et al. [10]. In the work [20], the participation of CYP3A4, CYP1A1, CYP2C8, CYP2C18, CYP2C19, CYP2J2, and CYP1B1 in the metabolism of bedaquiline and linezolid was proven, which determined the risk of developing adverse events and the concentration of drugs in the serum. Based on the above, the genetic polymorphism of the genes NAT2, CYP, GST, UGT, and ALDH influenced the metabolism of some anti-TB drugs, determining their safety and efficacy in the treatment of TB, including by changing the concentration in the serum. The maintenance of the therapeutic (bactericidal) concentration is regulated by the genes of the organic substance transporters: ATP-binding cassette of P-glycoprotein (encoded by the ABCB1, AGBL4 gene), the family of solute carriers of organic anion-transporting polypeptide 1B1 (encoded by SLCO1B1), transcriptional regulators of pregnane X receptor (PXR), constitutive androstane receptor (CAR), and transcription gene (Forkhead box protein O1, FOXO1) [8, 52–56].

The study [9] examined polymorphisms of individual nucleotides in the genes ABCB1, PXR, pleiotropic transcription factor (VDR), CYP24A1, and CYP27B1 and determined the effect of ABCB1 3435 CT/TT (p = 0.023) and CYP24A1 8620 AG/GG (p = 0.030), as well as Cdx2 AG/GG (p = 0.004) on the concentration of ethambutol in the serum of patients, mainly decreasing it, which did not affect the development of hepatotoxic reactions. At the same time, a study by Russian scientists established a weak correlation between the polymorphism of genes (NAT2 (590G>A (rs1799930), ABCB1 (3435T>C (rs1045642), as well as the allele of the ABCB1 gene (T) with the risk of developing hepatotoxic reactions in patients with pulmonary tuberculosis [57]. In the work of Zhang J et al. [58], the association of single nucleotide polymorphisms rs2755237 and rs4435111 in the regulatory region of the Forkhead box protein O1 (FOXO1) transcription gene, activating the expression of the metabolic enzyme aminolevulinate synthase 1, capable of leading to drug-induced liver damage during combined therapy with isoniazid and rifampicin, was studied. It was shown that the carriage of the C allele for the rs2755237 polymorphism and the T alleles for the rs4435111 polymorphism were associated with a reduced risk of developing hepatotoxic reactions.

Thus, the efficacy and safety of anti-TB drugs depend not only on the genetic polymorphism of the genes that determine liver function but also on the genes that transport organic substances.

An analysis of the instructions for use of the reviewed drugs in the registries of different countries showed that some references to the influence of pharmacogenetic markers on toxicity were found only for isoniazid in the FDA registry. The labeling of isoniazid drugs approved by the FDA varies slightly among manufacturers. Remedyrepack Inc. does not directly mention the NAT2 gene, but notes that “slow” acetylation may lead to higher levels of drug exposure and, consequently, to an increase in toxic reactions. Mikart Inc. mentions that the rate of acetylation is genetically determined in different ethnic groups, showing differences in the rate of drug inactivation, and that “slow” acetylation may lead to increased blood levels of the drug and, therefore, to an increased risk of toxic reactions. The labels of Rifater (a combination of rifampicin, isoniazid, pyrazinamide) contain similar information. All labels contain a warning about the risk of hepatitis associated with the use of isoniazid, but none of the companies mention the role of NAT2 or the possibility of genetic testing to minimize side effects.

Conclusion

The review collected and summarized information on the results of studies demonstrating the effect of pharmacogenetic markers on the risks of drug-induced liver damage during the use of anti-tuberculosis drugs. It was found that the most complete data with a sufficiently high level of evidence are available for allelic variants of NAT2 encoding the phenotypes of "slow" acetylator. Associations for NAT2 are relevant mainly for isoniazid and its combinations with other drugs for treating patients with tuberculosis. The reliability of data on the effect of the carriage of gene variants of CYP2E1 and glutathione-S-transferases (GSTM1 and GSTT1), UGT, and ALDH on the safety of anti-tuberculosis pharmacotherapy remains controversial due to the disparity of research results and, in some cases, their insufficiency. The obtained information on the influence of the polymorphism of genes providing for the transfer of organic substances on therapeutic concentrations of anti-TB drugs causes in some cases the ineffectiveness of TB chemotherapy and requires dose adjustment taking into account pharmacogenetic parameters. The articles included in this review were not limited in time due to the limited number of relevant studies. This review is subject to a small potential bias, including the influence of the authors’ personal views and gaps in the literature search and selection methods, which may lead to the omission of relevant studies.