Individual NAT2 genotype identity data should be enriched to prevent Isoniazid-induced liver injury (IDILI) and optimize the dose of Isoniazid (INH). Therefore, this study aims to present the level of IDILI risk for specific genotype alleles. The data collection involves literature indexed by Google Scholar, Scopus, and Pubmed databases. The search uses a combination of the following keyword variants “INH” OR “INH”, “liver injury” OR “hepatotoxicity”, “polymorphism” OR “pharmacogenomic”, and “N-acetyltransferase 2” OR “NAT2”. Furthermore, the screening results of library sources were narrowed to 11 original articles that met the inclusion criteria. The IDILI risk assessment analysis due to NAT2 enzyme polymorphism following the odds ratio has a 95% confidence interval. The results showed that the IDILI risk level of the slow acetylator group was 3.11 times higher than other populations. Meanwhile, the rapid and intermediate acetylator groups were not at risk. Three variants related to *6 allele were classified as high risk; *6A/*6A risk 5.76 times, *6A/*7B (5.54 times), and *6/*7 (4 times). The three allele configurations of the *5 and *7 were also classified as a risk; *5B/*7B (5 times), *7B/*7B (3.23 times), and *5/*7 (2,74 times).

Hepatotoxicity, Isoniazid dose adjustment, pharmacogenomic, polymorphism

Isoniazid (INH) has been used as an anti-tuberculosis (TB) drug since 1952, and reports of INH-induced liver injury (IDILI) are ongoing and common (Erwin et al. 2019). Previously, different studies were conducted to identify IDILI susceptibility in certain populations, demographic characteristics, and its hepatoprotective ability as a follow-up solution (Zhang et al. 2020). These studies also involve the polymorphisms study of enzymes responsible for INH metabolism (Perwitasari et al. 2015; Erwin et al. 2019; Yang et al. 2019; Zhang et al. 2020).

INH metabolic activity related to IDILI are associated with polymorphisms of several genes including N-acetyltransferase II (NAT2), Cytochrome P450 2E1 (CYP2E1), and glutathione S transferases (GST1) (Perwitasari et al. 2015; Erwin et al. 2019; Yang et al. 2019). Studies involving the NAT2 genotype are very important compared to those involving enzymes triggering (Zhang et al. 2020). The single nucleotide polymorphism (SNP) in NAT2 affects the INH metabolism rate. In addition, it affects the variations in treatment efficacy and frequency of adverse reactions (Swaminathan and Ramachandran 2012). Populations with NAT2 enzymes in the slow acetylator group were shown to be susceptible to IDILI exposure (Swaminathan and Ramachandran 2012; Perwitasari et al. 2015; Yang et al. 2019), and this was increased with the standard dose of INH (Shi et al. 2015).

Further studies on the genetic polymorphism of NAT2 are expected to include different ethnic populations (Perwitasari et al. 2015). The literature enrichment guides clinicians in screening patients to predict and prevent the IDILI through optimal pharmacotherapy determination (Swaminathan and Ramachandran 2012; Wang et al. 2012; Perwitasari et al. 2015). Meanwhile, the screening for NAT2 polymorphisms requires the identity of two haplotypes (diplotypes). This is because the identity of the individual NAT2 genotype is very important in adjusting the INH dose (Wichukchinda et al. 2020).

Some closely related slow acetylator NAT2 alleles are identified as IDILI trigger genotypes and are related with *5, -*6, and -*7 (Cai et al. 2012; Lv et al. 2012). Previous studies have shown the increased risk of IDILI in slow NAT2 acetylators due to the CYP2E1*c1/c2 (Santoso et al. 2021) or its association with the GSTM1 null genotype (Cai et al. 2012).

The study priority for the causative factors of IDILI is based on the NAT2 polymorphism, and the group with slow-acetylator was also susceptible to the exposure. Individual NAT2 genotype identity data should be improved by clinicians to optimize INH dose. This study attempts to present the IDILI risk level for specific genotype alleles since there are limited publications that identify the related topic. Therefore, these results should be the basis for adjusting the INH dose in genotype variants susceptible to IDILI.

The results confirmed the IDILI risk level of the slow acetylator group to be 3.11 times higher than in other populations. These data complement previous publications, which suggested that the risk ranged from 3.18 to 4.7 times (Cho et al. 2007; Cai et al. 2012; Wang et al. 2012). Moreover, the 6 variants of the NAT2 polymorphism that significantly trigger IDILI were highlighted. Three variants were related to *6 allele and were classified as high risk; *6A/*6A risk 5.76 times, *6A/*7B (5.54 times), and *6/*7 (4 times). Furthermore, three alleles configurations of the *5 and *7 were classified as a significant risk; *5B/*7B (5 times), *7B/*7B (3.23 times), and *5/*7 (2.74 times) (Fig. 2).

Figure 3. 

The role of slow acetylator of N-acetyltransferase 2 (NAT-2) polymorphism induce drug liver injury.

The six variants were the combinations of alleles *5, *5B, *6, *6A, *7, and *7B encoding a slow metabolic phenotype (Boukouvala et al. 2016; Igumnova et al. 2016; Mthiyane et al. 2020). The numbers of NAT2*6A and *7B were spread across many Asian and Caucasian populations (Mthiyane et al. 2020). Meanwhile, the highest IDILI risk involved were found in the population having the *6 alleles, NAT2*6A/*6A risk 5.76 times, -*6A/*7B (5.54 times), NAT2*6/*7 (4 times). This confirmed the predictions of other reports that the NAT2*6 group is highly susceptible to IDILI. The combinations that form the *6/*6 genotype is an ultra-slow metabolic phenotype (Igumnova et al. 2016).

Previous reports suggested the IDILI risk for both Asian and non-Asian is evenly distributed across alleles and ethnic variants of the population (Wang et al. 2012). Moreover, other reports suggested that Asian races are more susceptible to IDILI cases, unlike Caucasians (Cho et al. 2007; Cai et al. 2012; Swaminathan and Ramachandran 2012). However, the development of pharmacogenomic research related to tuberculosis is more involved in the countries of Asia compared to others. Similarly, the cases predominance of this study involved reports from the Asian region (Leiro-Fernandez et al. 2011; An et al. 2012; Ben Mahmoud et al. 2012; Gupta et al. 2013; Santos et al. 2013; Rana et al. 2014; Mushiroda et al. 2016; Wattanapokayakit et al. 2016; Ben Fredj et al. 2017; Suhuyanly et al. 2017; Yuliwulandari et al. 2019), even though Tunisian population caught attention. Zero cases were recorded when various populations showed high cases of NAT2*5/*7, -*6/*7, and -*5B/*7B variants. In the ultra-slow NAT2*6A/*6A variant, the number of cases and controls were balanced. Further studies should be conducted to show that the incidence of IDILI in the slow acetylator variant is not greater than the control (Ben Mahmoud et al. 2012; Ben Fredj et al. 2017).

The identity variants of the NAT2 genotype have consistently coded for the INH metabolic phenotype (Zabost et al. 2013; Seng et al. 2015; Huerta-García et al. 2020). This is because genotype identity is not a single trigger for IDILI. The risk also depends on the configuration of cooperation with other enzymes related to the metabolic activity of INH. Furthermore, it encodes the late phenotype, such as CYP2E1*c1/c2, and the null genotype GSTM1 (Cai et al. 2012; Santoso et al. 2021) to consider the variance of the population weight factor (Huerta-García et al. 2020). For every 10 kg increase in body weight, the concentration of INH is reduced by 1 mg/L [12]–[15], [17](Jung et al. 2015). Therefore, these variants should be considered in the new era of INH dose personalization.

Several investigations consistently affirm the association between NAT2 slow acetylator status and an elevated susceptibility to (Leiro-Fernandez et al. 2011; Ben Mahmoud et al. 2012; Gupta et al. 2013; Mushiroda et al. 2016; Wattanapokayakit et al. 2016; Ben Fredj et al. 2017; Yuliwulandari et al. 2019). Consequently, when prescribing anti-tuberculosis medications, this genetic factor should be taken into consideration. Research also suggests that NAT2 genotyping has the potential to personalize drug dosing, particularly for isoniazid, based on genetic information, which could optimize treatment outcomes and reduce the risk of hepatotoxicity (Ben Mahmoud et al. 2012; Santos et al. 2013; Suhuyanly et al. 2017). Furthermore, combining NAT2 and CYP2E1 genotype analysis can provide valuable insights into the risk assessment of IILI. It has been observed that slow NAT2 acetylator status, especially when coupled with specific CYP2E1 genotypes, may heighten susceptibility to IILI (An et al. 2012; Ben Fredj et al. 2017). Therefore, monitoring isoniazid serum concentrations is vital, as elevated levels are associated with a greater risk of IDILI (An et al. 2012). Therapeutic drug monitoring (TDM) for isoniazid can help identify individuals at risk and facilitate necessary treatment adjustments.

Moreover, there is a call to develop predictive models that integrate both genetic and clinical risk factors to evaluate the risk of IDILI (Mushiroda et al. 2016). Prospective clinical trials are essential to validate the clinical utility of these models and guide INH dosage adjustments. Large-scale screening for NAT2 and CYP2E1 genotypes, as proposed by studies (Santos et al. 2013; Rana et al. 2014), can effectively predict adverse effects and enhance clinical management. Routine genotyping tests may be warranted for populations at risk of IILI. Recognizing population-specific genetic variations in NAT2 is crucial (Santos et al. 2013; Wattanapokayakit et al. 2016). Clinicians should consider the genetic diversity within their patient populations when making treatment decisions. Finally, we should highlight the feasibility and cost-effectiveness of genetic testing for preventing severe adverse drug reactions within healthcare systems (Wattanapokayakit et al. 2016).

Recently, some reports have followed up on the NAT2 variant phenotype impact by offering a pharmacokinetic model, to establish a new era of INH dose personalization (Huerta-García et al. 2020; Mthiyane et al. 2020). The studies included diverse TB and healthy populations (Chen et al. 2009; Rodriguez et al. 2019; Huerta-García et al. 2020; Jing et al. 2020) co-infected with the human immunodeficiency virus (Bhatt et al. 2014) (Kubota et al. 2011; Seng et al. 2015). However, the development of a pharmacokinetic study involving the identity of the specific genotype is still limited.

The results showed that the genotypes NAT2*5/*7, -*7B/*7B, -*6/*7, -*5B/*7B, -*6A/*7B, and -*6A/*6A, have a risk range of IDILI 2.74 to 5.76 times. Meanwhile, the case finding is dominant in the Asian population, due to the pharmacogenomic study development related to tuberculosis involving more populations from Asian countries. Furthermore, personalization of INH dosage is a solution to reduce the IDILI rate. However, the pharmacokinetic development model should consider the synergistic configuration factor of slow acetylator NAT2 with other enzymes related to the INH metabolic activity. These enzymes also encode the slow and null phenotype, such as CYP2E1*c1/c2 and GSTM1 respectively. The six variants are expected to become priorities for future studies, given the limited database on the IDILI susceptible types.

Three variants related to the NAT2*6 allele were classified as high risk; -*6A/*6A risk 5.76 times, -*6A/*7B (5.54 times), and -*6/*7 (4 times). The three allele configurations of the -*5 and -*7 were also classified as a significant risk; -*5B/*7B (5 times), -*7B/*7B (3.23 times), and -*5/*7 (2.74 times). Based on these findings, incorporating NAT2 genotyping into anti-tuberculosis drug prescriptions, in conjunction with CYP2E1 genotype analysis and predictive models, offers promising prospects for personalized dosing, risk assessment, and treatment monitoring. This approach has the potential to enhance treatment efficacy and reduce the risk of liver toxicity in tuberculosis patients. Therefore, further pharmacokinetic studies to establish INH dosage adjustments for these six variants are warranted.

The authors confirm contribution to the paper as follows: study conception and design: SBS, PP; data collection: SBS, PP; analysis and interpretation of results: SBS, PP, LMI; draft manuscript preparation: SBS, PP, LMI. All authors reviewed the results and approved the final version of the manuscript.

The results are the output of the Institutional Vision Revitalization Research of the Universitas Muhammadiyah Magelang (Unimma). The authors are grateful to the Unimma Research and Development Institute for facilitating the completion of research activities.