Impact of Pharmacogenetics Markers of Human NAT2 Gene in Tuberculosis Treatment

Abstract

Tuberculosis (TB), mainly caused by Mycobacterium tuberculosis, accounts for 10 million cases worldwide per year, remaining a major problem for public health. The anti-TB drug isoniazid (INH) is recommended by the WHO. Despite of effective drugs, some individuals do not respond to standard treatment that can result in varying degrees of adverse drug reactions. One of the factors related to the variability in individual response to treatment is the presence of polymorphisms in genes encoding drug-metabolizing proteins, which can alter the protein’s activity. The NAT2 gene encodes Arylamine N-acetyltransferase 2 (NAT2), the main enzyme responsible for INH metabolism. Genetic variants found in NAT2 coding region affect N-acetylation. The rate of substrate metabolism defines the phenotype of individuals as fast, intermediate, slow, or ultra-slow acetylators. The slow phenotype has been associated with high risk of hepatotoxicity during TB treatment, and this risk is shown to be greater when an ultra-slow acetylator is identified. Furthermore, fast phenotype could be associated with extensive TB treatment due to greater drug clearance and therefore lower bioavailability of INH. The identification and use of biomarkers for phenotype prediction could minimize unfavorable therapeutic outcomes and optimize the effectiveness of tuberculosis treatment.

1. Introduction

Tuberculosis (TB) disease is mainly caused by Mycobacterium tuberculosis, and it is estimated that about a quarter of the world’s population is infected by this bacterium. Out of this total, approximately 10% will develop active tuberculosis, while the remaining 90% will have noncommunicable tuberculosis. According to data published by the WHO in 2022 for the previous year, it is estimated that 10.6 million people fell ill with tuberculosis in 2021, resulting in 1.6 million deaths worldwide. These numbers highlight the prevalence of cases in low- and middle-income countries accounting for 98% of reported TB. TB is the 13th leading cause of death globally and ranks first in deaths caused by a single infectious agent with only COVID-19 in 2020 (surpassing HIV/AIDS) having higher numbers. Thus, TB remains a major health problem for the world population. In 2021, according to WHO data, TB cases are distributed as follows: Southeast Asia (45%), Africa (23%), Western Pacific (18%), Eastern Mediterranean (8.1%), Americas (2.9%), and Europe (2.2%). Thirty countries with the highest TB burden accounted for 87% of all estimated incidence cases worldwide, and eight of these countries contribute to two-thirds of the global total: India (28%), China (7.4%), Indonesia (9.2%), the Philippines (7.0%), Pakistan (5.8%), Nigeria (4.4%), Bangladesh (3.6%), and Democratic Republic of the Congo (2.9%) [1]. The estimated TB incidence rates for 2021 can be observed in Figure 1.

Figure 1.

Estimated TB incidence rates, 2021 [1].

Isoniazid (INH) [2], rifampicin (RIF) [3], pyrazinamide (PZA) [4], and ethambutol [5] (EMB) are the four antibiotics indicated as first-line treatment for TB. The standard TB treatment comprises a first phase with isoniazid, rifampicin, pyrazinamide, and ethambutol, for 2 months with fixed dosages, followed by a second phase that includes the use of only INH and RIF for an additional 4 months. The first 2 months of TB treatment are responsible for killing of metabolically active and dormant bacilli, and most of the sputum smear-positive patients turn negative within this period [3, 4]. Traditionally, the prescription of standard fixed-dose (FDC) combination drugs has been recommended to simplify and facilitate treatment and adherence. These combinations of four drugs have some advantages, such as cost-effectiveness and universal access, low relapse rate (estimated at 3–5%), and intermittent administration in most cases.

Although the standard 6-month treatment regimen is highly effective for drug-susceptible Mycobacterium tuberculosis, some limitations are related to TB treatment, such as long duration, access to treatment in low- and middle-income country with limited financial capacity, and adverse drug reactions (ADRs). The most common ADRs associated with INH are peripheral neuropathy and hepatotoxicity, while reactions with an unestablished frequency include nausea, vomiting, stomach pain, fever, lymphadenopathy, skin rash, and vasculitis [2]. For RIF, ADRs include nausea, vomiting, and hepatitis, whereas the frequency of hyperbilirubinemia and cholestasis has not been established [3]. One severe ADR, life-threatening liver injury (about 14%), is related to the use of 3 g daily of PZA, while arthralgia, vomiting, anorexia, general malaise, sideroblastic anemia, urticaria, and increased uric acid were not frequency determined [4]. Finally, optic neuritis is the most frequent ADR related to EMB, followed by arthralgia, abdominal discomfort or pain, malaise, headache, vertigo, and mental confusion with undetermined frequency [5].

The antituberculosis drug-induced liver injury (ATDILI) affects from 2 to 28% of patients treated with multidrug therapy for TB and is potentially serious and fatal, leading to treatment interruption and failure [6].

Individuals on the same therapeutic regimen show a variability in their response. ADRs are one of the main causes of morbidity and mortality in developed countries, affecting both children and adults. Additionally, they have a negative socioeconomic impact, leading to hospitalizations and prolonged treatment times. Several factors contribute to this variability, including age, sex, nutritional status, general medical condition, lifestyle, concomitant therapy, presence of comorbidity, and genetic factors. The genetic background of individuals can play a significant role in determining the outcomes of TB treatment [7].

Variations in the human genome sequence, such as single-nucleotide polymorphisms (SNPs), involve a replacement of one nucleotide base with any one of the other three, occurring at approximately once every 1000 bases in the human genome. When SNPs are located on genes encoding metabolic enzymes, transport proteins, or cell surface receptors, they may affect the pharmacokinetics (PK—absorption, distribution, metabolism, excretion) and pharmacodynamics (PD—drug-target interaction and dose-effect relationship) of the drug. Pharmacogenetics can be defined as the branch of pharmacology that studies how variations influence drug response and its effects. Pharmacogenomics is an innovative approach aimed at minimizing ADRs and maximizing treatment effectiveness.

The significant differences in individuals’ drug responses are evident, considering that it is estimated that individuals differ from each other every 300–1000 nucleotides, resulting in approximately 10 million SNPs. Identifying which of these variants or combinations of variants have functional consequences for long-term drug effects will be enable the development individualized therapy based on the patient’s genetic sequence, leading to an adequate response and prevention adverse reactions [8, 9, 10, 11].

Certain genes are correlated with the pharmacokinetics and pharmacodynamics of drugs used for TB treatment, such as NAT2, CYP2E1, GSTT1, GSTM1, and SLCO. When these genes encode proteins with low activity, there is a higher incidence of hepatitis (Table 1).

Drug	Gene	Allele	Nucleotide Change(s)/rs Identifiers	Amino Acid Change(s)	Protein Function	Risk Phenotype	Reference
Isoniazid	NAT2	NAT2*6	c.590G > A (rs1799930)	R197Q	Slow	Slow acetylation	[12, 13]
		NAT2*7	c.857G > A (rs1799931)	G286E	Slow		[12, 14]
		NAT2*14	c.191G > A (rs1801279)	R64Q	Slow		[12, 15, 16]
	CYP2E1	CYP2E1*1A	−1053 C > T	—	Slow	Slow detoxification	[17, 18]
		CYP2E1*5	-1293G > C	—	Slow		[17, 18]
	GST	GSTM1*0	null genotype	—	Absence	Slow detoxification	[18]
		GSTT1*0	null genotype	—	Absence		[19]
Rifampicin	SLCO1B1	SLCO1B1*5	c.521 T > C (rs4149056)	V174A	Slow	Decreased Transport	[20]
		SLCO1B1*37	c.388A > G (rs2306283)	N130D	Slow		[21]
		SLCO1B1*4	c.463C > A (rs11045819)	P155T	Slow		[22]
		SLCO1B1*15	c.50611A > G (rs2306283); c.52422 T > C (rs4149056)	N130D V174A	Slow		[23]
	ABCB1	—	c. 3435C > T (rs1045642)	I1145I ^(synonymous)	Slow	Decreased Transport	[24]
		—	c.2677G > T (rs2032582)	S893A	Slow		[24]
		—	c.1236C > T (rs1128503)	G412G ^(synonymous)	Slow		[24]

Table 1.

Association of genetic variants with risk of liver damage induced by anti-tuberculosis drugs.

ND: not determined.

Isoniazid is a prodrug that requires activation by the catalase/peroxidase enzyme (KatG) of M. tuberculosis, resulting in the production of reactive oxygen radicals (superoxide, hydrogen peroxide, and peroxynitrate) and organic radicals. These radicals inhibit the formation of mycolic acid in the bacterial cell wall, causing damage to the DNA and subsequent death of the bacillus. The most common mechanism of resistance to isoniazid involves mutations in katG, which decrease its activity, preventing the conversion of the prodrug into its active metabolite. Isoniazid is considered a primary drug and is used in the treatment of all forms of tuberculosis caused by strains of M. tuberculosis that are sensitive to it. Its biotransformation occurs in the liver through acetylation. In humans, there is genetic heterogeneity in the rate of isoniazid acetylation and liver disease may prolong the clearance of isoniazid. It inhibits the synthesis of mycolic acid, an important component of the wall of mycobacteria, not acting against other types of bacteria.

Isoniazid (INH) is metabolized in the liver via N-acetyltransferase (NAT2) generating acetylhydrazine; when oxidized by cytochrome P4502E1 (CYP2E1), it can form hepatotoxic intermediates [25, 26]. Another route of INH metabolism is direct hydrolysis to hydrazine, a potent hepatotoxin. These reactive metabolites can destroy hepatocytes by interfering with cellular homeostasis or triggering immune reactions in which reactive metabolites bound to plasma proteins in hepatocytes can act as haptens [27]. Figure 2 shows the estimated frequency of therapeutic failure and treatment toxicity with INH for tuberculosis in populations worldwide. Standard TB therapy, including INH, results in more than 40% of patients experiencing ADR and more than 30% with treatment failure in populations around the world. Therefore, pharmacogenomics-guided therapy is highly cost-effective and directly impacts INH-induced liver injury and treatment response [28].

Figure 2.

Estimated frequency of therapeutic failure and treatment toxicity with INH for tuberculosis in populations worldwide [12]. Adapted from [28].

2. N-acetyltransferase 2

Human N-acetyltransferase 2 is a phase II biotransformation enzyme predominantly expressed in the liver, small intestine, and colon [29]. NAT2 was first identified in humans in 1960 and is a 33.79 kDa [30] cytosolic enzyme with the role of catalyzing N-acetylation and O-acetylation, through the transfer of the acetyl group from the acetyl-Coenzyme A (acetyl-CoA) cofactor to the nitrogen terminal of hydrazines, arylamines, and heterocyclic amines [31, 32]. NAT2 is involved in the metabolism and inactivation of substances that include carcinogens and drugs used in the treatment of infections and chronic diseases, such as tuberculosis, leprosy, and arterial hypertension, among others [32, 33].

The high-resolution structure of human NAT2 was determined by Wu et al. [31] and defines three protein domains. The first, named as the N-terminal domain, comprises residues 1–83 and consists of five helices and a short β strand present between α2 and α3 helices. The second domain, comprising residues 84–192, is composed of nine β strands and two short helices. These two domains are connected to the third domain, called C-terminal (residues 230–290), through the α-helical inter-domain (residues 193–229; helices α8– α10), which is composed of four antiparallel β strands and one helix [31]. The catalytic triad, located in the N-terminal domain of the protein structure, consists of three highly conserved residues: cysteyne at position 68, Histidine 107, and aspartic acid 122. Their interaction occurs through the donation of the acetyl group of the acetyl-CoA cofactor to a sulfhydryl residue of the catalytic site cysteine, forming acetylcysteinyl as an intermediate structure. Subsequently, the acetyl group is transferred to the amino terminus of the substrate [34, 35, 36].

The molecular study of human N-acetyltransferases revealed two homologous genes, encoding NAT1 and NAT2, and a pseudogene, pNAT [37]. These loci are located on chromosome 8 (8p22). N-acetyltransferase 2 is encoded by NAT2, an intronless gene that has 873 bp [38]. SNPs in the NAT2 gene-coding region can alter enzymatic activity, resulting in three different phenotypes: fast acetylators (FAs) with two functional alleles, intermediate acetylators with one functional allele combined with a “slow” allele, and slow acetylators (SAs) with two “slow” alleles [39]. According to the official NAT2 nomenclature website, the arylamine N-acetyltransferases (NATs) database (http://nat.mbg.duth.gr/) [40], more than 100 SNPs and 108 haplotypes have been identified. The reference NAT2*4 allele was first identified in the Japanese population and does not have any SNP in its coding region The protein encoded by NAT2*4 confers FA phenotype [13, 41, 42]. The NAT2 alleles or haplotypes are characterized by the combination of up to six simultaneous SNPs. They are commonly clustered into specific allelic groups based on signature SNP: haplotypes belonging to the NAT2*5 allelic group share the c.341 T > C, SNP signature just as the NAT2*6 allelic group share the c.590G > A, SNP signature. The c.857G > A SNP characterizes NAT2*7 cluster, and c.191C > A is common for NAT2*14 alleles [43]. The frequencies of NAT2 allelic groups among different populations can be seen in Table 2.

Population (N)	NAT2 haplotypes, % (2 N = 100%)							Reference
	*4	*5	*6	*7	*12	*13	*14
America
Brazilian (404)	20	38	27	4	4	3	4	[44]
Ngawbe (105)	72.4	2.4	0	23.3	0	1.9	ND	[45]
Embera (136)	61	9.9	3.7	22.8	0	2.6	ND	[45]
Africa
Tswana (101)	13.4	32.2	20	0	20.8	6.4	8.4	[46]
Mandenka (97)	9.3	36.1	17	6.7	15.5	5.2	10.3	[47]
Sudanese (127)	8.7	47.2	28.7	3.1	8.3	0.8	3.1	[48]
Europe
Spanish (1312)	22.2	45.7	26.7	1.2	2.6	0.3	1.4	[49]
German (844)	22.7	46.5	27.8	1.3	ND	1.5	0.1	[50]
US Caucasian (387)	24.2	45.9	26.6	1.9	0.4	0.9	0.1	[51]
Russian (364)	23.5	45.6	27.2	3.2	0.5	0	0	[52]
Asia
Chinese (212)	59.2	4	20.8	14.9	0.7	0.5	0	[53]
Japanese (200)	69.5	0.5	19.8	8.8	ND	1.3	ND	[54]

Table 2.

Frequencies of NAT2 allelic groups among populations worldwide.

ND: not determined.

Mutations can cause various effects on the activity and function of a protein through four main mechanisms: decreased mRNA levels, altered protein levels, altered protein stability, and direct alteration of protein activity [43]. Although SNPs can influence NAT2 activity, it is essential to consider the haplotype to fully understand the genotype-phenotype relationship. Previous studies of enzymatic activity with the most frequent NAT2 alleles/haplotypes have enabled the characterization of the acetylation profile, making them potential biomarkers for enhancing therapeutic treatment or preventing the increase of the number of cancer cases. However, understanding the prior genotypic profile of the population is required [43].

SNPs discovery studies of the NAT2 gene have been performed in diverse populations. Among the seven SNPs commonly found in NAT2 (Table 3), four (c.191G > A/ rs1801279, c.341 T > C/rs1801280, c.590G > A/ rs1799930, and c.857G > A/ rs1041983) are non-synonymous mutations that result in a significant decrease in acetylation capacity, and the other three (c.282C > T/ rs1799929, c.481C > T/ rs1799931, and c.803A > G/rs1208) are synonymous SNPs or do not change the acetylation phenotype [11, 13, 39, 55]. There are 13 NAT2 haplotypes; eight of them are widely distributed, with three associated with rapid acetylation phenotype (NAT2*4, NAT2*12A, and NAT2*13A), and the other five are related to slow acetylation phenotypes (NAT2*5B, NAT2*6A, NAT2*7B, NAT2*5C, and NAT2*5A) [56].

SNP	rs Identifiers	Amino Acid Change(s)	Protein Function	Reference
c.191G > A	rs1801279	R64Q	Slow	[13, 39]
c.282C > T	rs1799929	Y94Y (synonymous)	Rapid	[13, 39]
c.341 T > C	rs1801280	I114T	Slow	[13, 39]
c.481C > T	rs1799931	L161L (synonymous)	Rapid	[13, 39]
c.590G > A	rs1799930	R197Q	Slow	[13, 39]
c.803A > G	rs1208	K268R	Rapid	[13, 39]
c.857G > A	rs1041983	G286E	Slow	[13, 39]

Table 3.

The seven SNPs commonly found in NAT2.

The identification of the mutation frequency allows a virtual pre-definition of the phenotypic profile in a closed ethnic population. African populations exhibit the highest level of diversity within a population, characterized by a low frequency of the NAT2*4 haplotype but a high frequency of the other two fast haplotypes, NAT2*12A and NAT2*13A. In fact, NAT2*12A is called the hallmark of this population, as it is rarely found outside Africa. However, this population also has a high frequency of slow acetylator alleles. In European populations, the NAT2*5B and NAT2*6A haplotypes, associated with the slow acetylation phenotype, are predominant over the fast NAT2*4 haplotype. Descendants of European and sub-Saharan Africans with slow acetylator genotypes represent about 50% of the population [56].

In relation to Asia, characterized by one of the highest frequencies of NAT2*4 in the world (Table 2) and a low diversity of haplotypes, only three other alleles occur with frequency > 0.01: NAT2*6A, NAT2*7B, and NAT2*14, which confer a low acetylation protein [11, 56]. On the other hand, the American continent has a high level of population diversity, like the African population, indicating remarkable heterogeneity for NAT2 variation in this region [56]. The difference found can be easily explained by the presence of several small populations isolated as a result of the displacement of people over the years from larger populations, including those with European and African ancestry. This explains the presence of the NAT2*6A, NAT2*7B, and NAT2*14 haplotypes in this region [26]. In Brazilians, an admixture population, some studies have shown the high allelic diversity of NAT2, including the description of new or rare SNPs and alleles (NAT2*5O, *6 M and *12E) with unknown functional effect [11, 55]. The high frequency of slow acetylators found in the Brazilian population may be related to the high rate of miscegenation between the European (colonizer) and African (slave) populations [55].

Several studies have shown that low activity of NAT2 enzymes can significantly impact drug treatment of individuals and pose risks. Recently, two independent studies have further refined the low enzymatic activity in slow acetylators and ultraslow acetylators, and their heterogeneity was observed when comparing genotypes of the same classification. The difference in heterogeneity is measured through the Odds Ratio (OR), which represents the chance of an event occurring in a group or between groups. Specific NAT2*5 genotypes (c.341 T > C), NAT2*6 (c.590G > A), NAT2*7 (c.857G > A), and NAT2*14 (c.191G > A) show amino acid substitutions in the protein (I114T, R197Q , G286E, and R64Q , respectively) [57, 58]. This sub-characterization of the markers allows to the identification of risks associated with each cluster/haplotype, enabling better therapeutic targeting and reanalysis of the literature, which might have overlooked data with a lower risk of occurrence of the class.

Literature data have demonstrated the great importance of functional knowledge of N-acetylation for patients treated with anti-TB drugs such as INH. The slow acetylation state contributes to the occurrence of hepatotoxicity (adverse hepatic effects) during TB treatment, characterized by liver damage caused by excessive ingestion of chemicals or risk of antituberculosis drug-induced liver injury (ATDILI), defined by an increase of more than twice in the upper limit of normal value (ULN) in alanine transaminase (ALT) or aspartate aminotransferase (AST) and long-term bilirubin [55, 59, 60].

Individuals carrying the genotype NAT2*4/*5 and NAT2*4/*6 exhibit similar acetylation values. However, when the wild-type allele is absent, the variant alleles NAT2*5 and NAT2*6 confer different acetylation capacities [56]. A gene dose effect can be observed for these variant alleles within slow acetylation phenotype, as there is a statistically significant trend toward slower than expected acetylation capability when genotypes are combined: NAT2*5/*5 > NAT2*5/*6 > NAT2*6/*6. NAT2*6/*6 shows a reduction of almost 30% in their enzymatic activity compared to the NAT2*5/*5 homozygote [61].

Comparing individuals with a high-activity enzyme to slow acetylators, Teixeira et al. in 2011 observed a 2.8-fold increased risk of ATDILI occurrence (95% CI: 2.20–3.57; P = 5.73E − 18) for the slow acetylation genotype [11]. Meanwhile, while Suvichapanich et al. in 2018 observed that the ultraslow subgroups (based on combined genotypes *6A/*6A, *6A/*7B, and *7B/*7B) were able to achieve a 3.6-fold increased risk (95% CI: 2, 30–5.63; P = 1.76E −08) of ATDILI. Furthermore, N-acetylation enzymatic kinetics assays with recombinant human NAT2 demonstrated that the activity of NAT2*4 was 7.6–22-fold greater than that of NAT*5 and *6, respectively. Additionally, the Vmax/Km value of NAT2*5B was 3.2 and 4.7 times higher than those of *6A and *7B, respectively, indicating better catalytic efficiency when compared to other low-activity alloenzymes [57]. Therefore, it is evident that ultraslow acetylators may contribute to a higher risk of ATDILI than other classes [61].

It is believed that in INH metabolism, the decrease in NAT activity can alter the elimination of hydrazine and N-acetylhydrazine, leading individuals with these genotypes to be exposed to hepatotoxins for a longer time, increasing susceptibility to ATDILI. Additionally, ATDILI occurrence may be due to the redirection of INH pharmacological metabolism to hydrolysis, a minor pathway that results in a greater amount of hydrazine, given the decreased NAT activity [62].

The pharmacokinetic variability caused by the diverse distribution of polymorphisms in NAT2 has posed a challenge for implementing a standard dosage of INH. While the relationship between the slow acetylation profile and the occurrence of adverse reactions is clear, the relationship between the fast acetylation profile, inadequate serum levels of INH, and the occurrence of unfavorable outcomes is not well understood, with conflicting results in the main studies. Fast acetylators are characterized by the absence of SNPs in the NAT2 coding region or by the presence of variants that do not affect N-acetylation, such as c.282C > T, c.481 > T, c.803A > G, and c.845A > C or a combination of some of them (NAT2*11A, NAT*12A, NAT2*12B, NAT2*12C, NAT2*13A, and NAT2*18 alleles) [15, 40, 50, 63, 64, 65].

According to Peloquin et al. [66], fast acetylators have a clearance rate of isoniazid 1.4 and 3.6 times higher than intermediate and slow acetylators, respectively. This results in faster drug elimination and less absorption, leading to lower exposure to INH than what is considered effective. Fast acetylators are more likely to present a positive sputum culture after 2 months of treatment [66, 67]. A prolonged therapeutic response may have consequences such as extended treatment time, low patient adherence, and higher financial cost, due to possible hospitalizations and greater demand for medication. Additionally, the drug may not reach the minimum concentration necessary to be effective against the bacilli, potentially leading to the selection of resistant strains [67, 68].

According to Jing et al. [69], even with the administration of a high dose of INH (900 mg/day), only 66.1% of the patients with fast acetylation reached the maximum concentration. However, when using the standard dosage (300 mg/day), more than 98% of fast acetylators, 89.1% of intermediate acetylators, and only 26.9% of slow acetylators achieved Cmax values lower than the minimum recommended dose (3 μg/mL) [69]. These findings show that the standard used INH dosages used for intermediate and fast acetylators are insufficient, and adjusting the INH dose based on NAT2 metabolism status may be necessary to achieve an improved balance between risk and benefit during treatment.

Pharmacokinetic assays have been increasingly incorporated into clinical routine, especially in the developed countries, with the aim of proposing appropriate dosages for new drugs. Studies conducted in Asia and the United States have demonstrated significant differences in INH clearance based on different acetylation phenotypes [62, 66]. Moreover, the variations in acetylation profiles among different populations underscore the variable influence of NAT2 polymorphisms, emphasizing the importance of genotypic identification before administering isoniazid [69]. Pharmacogenetic research involves candidate genes that can function as biomarkers associated with drug metabolism and transport, enabling the prediction of drug toxicity and efficacy.

3. Conclusions

According to the model proposed by Russmann et al. [70], several factors may be responsible for the variability of the therapeutic response, such as nongenetic factors (age, sex, nutritional status, general medical condition, lifestyle, concomitant therapy, or presence of comorbidity, etc.) and genetic factors. Among them are the SNPs that when located in genes encoding metabolic enzymes, transporters, or receptors, may be able to change pharmacokinetics and pharmacodynamics of the drug [70].

An important study tool is pharmacogenetics that could be used to identify variants in the NAT2 gene. This characterization allows for the prediction of possible therapeutic outcomes unwanted in TB treatment. According to the data obtained, it was possible to visualize that slow acetylators are more likely to develop adverse effects than fast acetylators. Besides, it is important to refine the phenotype prediction in subcategories, to identify the “ultraslow” acetylators and analyze the real risk for this subcategory.

The NAT2 slow acetylator and the risk of antituberculosis drug-induced liver injury (ATDILI) is confirmed. Among them, the recently proposed subgroup of ultraslow acetylators stands out, with patients with NAT2 genotypes of *6A/*6A, *6A/*7B, and *7B/*7B, contributing to an increased risk of ATDILI (OR: 3.60; 95% CI: 2.30–5.63; P = 1.76E − 08) than all NAT2 slow acetylators (OR: 2.80; 95% CI: 2.20–3.57; P = 5.73E − 18), as well as fast acetylators [61].

It is known that the acquisition of drug resistance is related not only to the pathogen but also to the host. It was observed that in certain populations with high rates of therapeutic failure and acquisition of bacterial resistance, patients under anti-TB treatment had lower-than-expected serum drug levels, mostly fast acetylators [66, 71].

The study set of studies in relation to pharmacogenomics diagnosis, the effect on clinical outcomes, and patient development for everyday clinical use should be developed. Accumulated data on NAT2 clinical significance support the replacement as non-weighted standard therapy practices with pharmacogenomics-guided therapies. Thus, researchers look for factors that justify the development of new approaches to direct the treatment of tuberculosis, such as, the use of pharmacogenomics for NAT2. This practice will benefit treatment by reducing adverse effects and increasing efficiency [13, 41].

Acknowledgments

To the support of the research team involved in this work, and to the support of Oswaldo Cruz Foundation/IOC and funding from CAPES.