Abstract
N‐acetyltransferase 2 (NAT2) genetic polymorphisms might alter isoniazid metabolism leading to toxicity. We reviewed the impact of NAT2 genotype status on the pharmacokinetics, efficacy, and safety of isoniazid, a treatment for tuberculosis (TB). A systematic search for research articles published in Scopus, PubMed, and Embase until August 31, 2023, was conducted without filters or limits on the following search terms and Boolean operators: “isoniazid” AND “NAT2.” Studies were selected if NAT2 phenotypes with pharmacokinetics or efficacy or safety of isoniazid in patients with TB were reported. Patient characteristics, NAT2 status, isoniazid pharmacokinetic parameters, early treatment failure, and the prevalence of drug‐induced liver injury were extracted. If the data were given as a median, these values were standardized to the mean. Forty‐one pharmacokinetics and 53 safety studies were included, but only one efficacy study was identified. The average maximum concentrations of isoniazid were expressed as supratherapeutic concentrations in adults (7.16 ± 4.85 μg/mL) and children (6.43 ± 3.87 μg/mL) in slow acetylators. The mean prevalence of drug‐induced liver injury was 36.23 ± 19.84 in slow acetylators, which was significantly different from the intermediate (19.49 ± 18.20) and rapid (20.47 ± 20.68) acetylators. Subgroup analysis by continent showed that the highest mean drug‐induced liver injury prevalence was in Asian slow acetylators (42.83 ± 27.61). The incidence of early treatment failure was decreased by genotype‐guided isoniazid dosing in one study. Traditional weight‐based dosing of isoniazid in most children and adults yielded therapeutic isoniazid levels (except for slow acetylators). Drug‐induced liver injury was more commonly observed in slow acetylators. Genotype‐guided dosing may prevent early treatment failure.
Study Highlights.
WHAT IS THE CURRENT KNOWLEDGE ON THE TOPIC?
Isoniazid (INH) metabolism is associated with NAT2 genetic polymorphisms, leading to toxicity, especially hepatotoxicity and treatment failure.
WHAT QUESTION DID THIS STUDY ADDRESS?
Conclusive data on the effect of NAT2 acetylation status on the pharmacokinetics, safety, and efficacy of INH in tuberculosis patients are limited. We aimed to update and clarify the effect of NAT2 phenotypes on the pharmacokinetics and clinical data of INH to improve future research on the pharmacokinetic INH modeling, leading to model‐informed precision dosing.
WHAT DOES THIS STUDY ADD TO OUR KNOWLEDGE?
The mean DILI prevalence in tuberculosis patients worldwide was the highest in the NAT2 slow acetylator based on observations of C max above the therapeutic level. Moreover, choosing an appropriate sampling timepoint for INH monitoring might be different for those with different NAT2 types.
HOW MIGHT THIS CHANGE CLINICAL PHARMACOLOGY OR TRANSLATIONAL SCIENCE?
Our findings pave the way for further pharmacokinetic modeling research and support NAT2 genotyping testing before receiving INH. Our data suggest a new strategy for blood sampling for INH monitoring in each NAT2 phenotype group.
INTRODUCTION
Tuberculosis (TB) remains a worldwide infectious disease that can lead to death. TB is caused by a deadly bacterium, Mycobacterium tuberculosis (MTB). 1 Isoniazid (INH) is a keystone drug for treating newly diagnosed TB. INH is a prodrug that requires an oxidative reaction to exert its antimycobacterial activity. 2 INH is primarily inactivated in the liver and intestine by acetylation, a process catalyzed by the phase II enzyme N‐acetyltransferase 2 (NAT2). 3 Polymorphisms in the NAT2 gene significantly influence the activity of the enzyme, leading to interindividual variation in drug metabolism and susceptibility to certain diseases such as TB. 4
NAT2 activity is classified into three main categories: rapid (RA), intermediate (IA), and slow (SA) acetylators. The SA type results in reduced NAT2 enzyme activity and slower INH metabolism compared with the RA or IA types, leading to higher INH plasma concentrations, prolonged exposure to the drug and an increased risk of INH toxicity, especially hepatotoxicity. 5 In contrast, the RA type results in enhanced NAT2 enzyme activity and faster INH metabolism compared with the SA or IA types, leading to lower INH plasma concentrations, a shorter exposure of MTB to INH, and an enhanced risk of treatment failure. 6 Administering an INH dosage tailored to the NAT2 genotype may optimize INH plasma levels, resulting in better efficacy and reduced liver injury. However, conclusive evidence of the effect of the NAT2 genotype on the PK, safety, and efficacy of INH for patients with TB has been limited. Therefore, this systematic review aimed to update and clarify data on the effect of NAT2 genotypes on the PK and clinical data of INH in patients with TB.
METHODS
Literature search
A systematic literature search was performed following the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses guidelines. A systematic search for research articles, in any language, published in PubMed, EMBASE, and Scopus as well as the Thai Journal Citation Index (TCI) and ASEAN Citation Index (ACI) was conducted. Papers were included until August 31, 2023, without filters and limits using the following search terms and Boolean operator: “isoniazid” AND “NAT2.” Only literature published in English was included in this systematic review.
Criteria for study selection
Duplicate records were automatically excluded by EndNote in the first round. If duplicate records were missed, they were manually detected by the authors in the second round. The authors screened the titles and abstracts following the selection criteria for further assessment. All chosen articles were read thoroughly to evaluate whether they were related to the objectives of this review. Then, the articles were evaluated by the authors again in the details to be the final articles included in this systematic review.
Eligibility criteria were determined based on the PICO elements of the review question, which include population, interventions, comparators, and outcomes. Studies were tentatively included if they reported an impact of NAT2 genotype on the PK, efficacy, or safety of INH usage in patients with TB, constituting the elements of outcomes. The PK measurement of INH encompassed plasma concentration at any time (C t) and the area under the curve at any time (AUCt). Additionally, the included studies were required to indicate the dosage regimen of INH and to mention the TB treatment guidelines used in the study, both of which were considered elements of interventions. The population elements analyzed in our review consisted of age, sex, country of study, and NAT2 acetylators.
Tentative studies were excluded if participants used INH for indications other than TB. In vitro and nonhuman studies with healthy volunteers were excluded. Reviews, systematic reviews, meta‐analyses, and book chapters were also excluded from this systematic review to avoid potential data duplication. Furthermore, full‐text articles that were not accessible by the Mahidol University Library and Knowledge Center were excluded.
Data extraction
A standardized form was developed in Microsoft Excel and used to collect the extracted data. The data consisted of: (1) characteristics of the study: author, publication year, country, study design, sample size, type of patients with TB (pulmonary and extrapulmonary), INH dosage, TB treatment regimen, INH quantification method, NAT2 genotyping method; (2) characteristics of the participants: population source, demographic data (age, weight, height, or body mass index (BMI)); (3) study outcomes: acetylator status (fast, intermediate, and slow), PK data (C t and AUCt), prevalence of drug‐induced liver injury (DILI), and early treatment failure (ETF). The outcome results were categorized into two groups, PK studies and safety and efficacy studies. The subjects were divided into subgroups by the age of the patients, children and young adolescents (age ≤ 15 years old) and adults and older adolescents (>15 years old), following WHO 2022 age group criteria, 7 the same criteria as that used in previous studies. 8 , 9 , 10 , 11 , 12 The criteria for DILI diagnosis were an elevation in liver function tests, aspartate aminotransferase (AST), or alanine aminotransferase (ALT) levels of more than fivefold of the upper limit of normal (ULN), AST or ALT levels of more than threefold of the ULN with symptoms such as nausea, vomiting, poor appetite, abdominal pain, or jaundice, or AST or ALT levels of more than threefold of the ULN with total bilirubin levels of more than twofold of the ULN. 13 , 14 The criteria for elevated liver enzymes were an AST or ALT more than twofold of the ULN. 13
Quality evaluation
To evaluate study quality, validation tools for PK studies were not available. However, we considered and obtained all studies that reported relevant PK parameters, including plasma concentrations or AUC, and the target population received INH with genotyping testing. To enhance the quality of the data extraction process, the recruited articles needed to precisely describe the quantitative method of determining the INH plasma level for PK studies and the NAT2 genotyping method for PK and safety studies.
Statistical analysis
The subjects were analyzed separately in pediatric and adult groups. When the data were given as the median with a range or interquartile range (IQR), these values were standardized to the mean with SD via an appropriate method. 15 Briefly, Wan and colleagues proposed estimation methods to convert reported median values into estimated mean and SD. The conversion method provided an Excel spreadsheet with formulars to estimate the mean and SD from in various scenarios, including (1) if the article provided the median, range, and sample size, (2) the article provided the median, range, IQR, and sample size, (3) the article provided the median, IQR, and sample size.
The data were analyzed and statistically tested using GrapPad Prism v.6. The normality distribution of the data was determined with the Shapiro–Wilk normality test. Then, ANOVA was applied for normally distributed data, whereas the Kruskal–Wallis test and the Mann–Whitney test were applied for nonparametric data. The Kruskal–Wallis test was applied to compare differences in C max, AUC0–24, C2, C3, and the prevalence of DILI between the RA, IA, and SA groups. Dunn's test was employed for nonparametric multiple comparison analysis. The difference in the INH dose between children and adults was tested using the Mann–Whitney test. The p‐value <0.05 was considered as a statistical significance.
Assessment of reporting bias
No validated tools for PK studies were available to assess reporting bias.
RESULTS
We found 2162 articles including 258 articles from PubMed, 367 articles from EMBASE, 1529 articles from Scopus, and eight articles from the local databases TCI and ACI. After the exclusion of duplicate articles by automatic and manual methods, 1616 articles remained. The titles and abstracts of these articles were read and screened by the authors in EndNote, which categorized the type of article. Book chapters, reviews, conference abstracts, and nonfull‐text articles such as short communications and letters were excluded, and 892 items remained after these exclusions. The titles and abstracts of the remaining articles were read thoroughly and screened, and 718 records were excluded. The main reasons for exclusions were the studies that mentioned INH or NAT2 in the abstract but were not relevant to the objective of this systematic review (276 records), studies that were not performed in humans (114 records), studies that did not mention NAT2 acetylator status (121 records), and studies that were the type of NAT2 epidemiology without PK and safety data of INH (72 records). Therefore, 174 records were retained, and the full articles were obtained for final screening. After the final screening by a detailed reading of the full articles, the main reason for exclusion was that the studies did not stratify by NAT2 genotype (72 records). Finally, 90 articles were included in this systematic review. These were separated into two groups by the type of study, including PK studies (41 articles (Table S1)) and clinical (safety and efficacy) studies (53 articles (Table S2)). Moreover, four studies overlapped between PK and safety, and one study overlapped between efficacy and safety (Figure 1).
FIGURE 1.
Flow diagram of the study selection.
Demographic data
The demographic data of the patients in the studies comprised the number of patients, sex, NAT2 genotypes, age, weight, and BMI. These demographic data were collected for PK and safety studies separately. Moreover, PK studies were categorized by age into adults (26 articles) and children (15 articles), whereas safety studies (53 articles) were reported only in the adult population.
The total number of patients in the PK studies was 5183 adults and 1048 children. The number of male participants was higher than female participants. Seven articles did not provide sex information (Tables S1 and S2).
The percentage of SA adults (46.33%) was higher than that of IA (30.74%) and RA (22.64%) patients. The average age of the adults was 41.77 ± 8.26 years, whereas that of children was 4.25 ± 2.89 years. The average weight in adults was 56.88 ± 5.83 kg, whereas that in children was 12.54 ± 5.73 kg. The average BMI in adults was 21.00 ± 1.4 kg/m2 (Table 1).
TABLE 1.
Demographic data of patients in the included studies.
| Characteristics | PK studies (41 articles) | Safety studies (53 articles) | |
|---|---|---|---|
| Adults (26 articles) | Children (15 articles) | ||
| Number of patients, n | 5183 | 1048 | 23,924 |
| Sex, n (%) | |||
| Male | 1750 (33.77) | 522 (49.81) | 6664 (27.85) |
| Female | 1230 (23.73) | 461 (43.99) | 4293 (17.95) |
| N/A for sex | 2203 (42.50) | 65 (6.20) | 12,967 (54.20) |
| NAT2 phenotype, n (%) | |||
| Ultraslow acetylator | N/A | N/A | 1383 (5.82) |
| Slow acetylator | 1774 (46.33) | 382 (37.20) | 9678 (40.72) |
| Intermediate acetylator | 1177 (30.74) | 321 (31.25) | 7760 (32.65) |
| Rapid acetylator | 867 (22.64) | 306 (29.80) | 4281 (18.02) |
| Unknown | 11 (0.29) | 18 (1.75) | 664 (2.79) |
| Age (years), mean ± SD | 41.77 ± 8.26 | 4.25 ± 2.89 | 43.27 ± 8.92 |
| Weight (kg), mean ± SD | 56.88 ± 5.83 | 12.54 ± 5.73 | 57.91 ± 4.13 |
| BMI (kg/m2), mean ± SD | 20.26 ± 1.86 | N/A | 21.00 ± 1.40 |
Note: N/A: no data reported. Unknown: did not analyze or interpret NAT2 phenotypes.
Interestingly, an ultraslow acetylator type was also reported in some articles. The ultraslow acetylator was defined as a person harboring the genotypes *6/*6, *6/*7, and *7/*7. 16 These genotypes were interpreted as the SA type if the genotypes were categorized into a trimodal distribution. The percentage distribution of ultraslow and SA types in safety studies was 46.54%. This was close to the percentage of SA adults in the PK studies (46.33%). For the IA and RA groups, the percentage of both in the safety studies was consistent with those in the PK study. Therefore, our results show that the SA type was the most frequent among the NAT2 types. Overall, the average age, weight, and BMI are 43.27 ± 8.92 years old, 57.91 ± 4.13 kg, and 21.001 ± 1.4 kg/m2, respectively (Table 1).
NAT2 status impacted pharmacokinetic parameters and DILI prevalence
The patients were categorized by genotype, phenotype, and acetylation activities of the NAT2 enzyme into SA, IA, and RA groups. The PK parameters of INH in this review included C max, the concentration at the timepoint comprising 2 h (C2) and 3 h (C3), and AUC from 0 to 24 h (AUC0–24). The dosing of INH ranged from 4.7 to 18.3 mg/kg/day, and the average dose of INH was 5.98 ± 2.73 mg/kg/day (Table 2). The distributions of the INH dose and PK parameter data including C max, C2, C3, and AUC0–24 were tested for normality with the Shapiro–Wilk normality test, and the results showed that the INH dose and all PK parameters were nonparametric. Therefore, all PK parameters were tested using the Kruskal–Wallis test with Dunn's test for nonparametric multiple comparison analysis. The safety profiles in this review focused on the prevalence of DILI and elevated liver enzymes, which were classified as categorical data. Thus, the Kruskal–Wallis test with Dunn's test for nonparametric multiple comparison analysis was also employed for these data.
TABLE 2.
Pharmacokinetic parameters and safety data stratified by NAT2 phenotype in adults.
| Parameter a | NAT2 phenotype | ||
|---|---|---|---|
| SA | IA | RA | |
| Dose of INH (mg/kg/day) | 5.98 ± 2.73 (4.70–18.3) | ||
| C max (μg/mL) |
7.16 ± 4.85 (1.44–20.40) |
5.11 ± 2.78 (1.52–9.77) |
4.84 ± 3.60 (1.39–15.57) |
| C2 (μg/mL) |
6.82 ± 3.56 b (2.39–12.00) |
4.66 ± 2.69 (1.85–9.00) |
3.59 ± 3.01 b (0.77–10.63) |
| C3 (μg/mL) |
4.84 ± 1.31 b (4.03–6.79) |
2.14 ± 0.09 (2.08–2.20) |
1.57 ± 0.73 b (0.75–2.37) |
| AUC0–24 (μg h/mL) |
40.38 ± 33.42 b (17.67–119.83) |
15.48 ± 4.68 (9.87–20.00) |
19.46 ± 25.16 b (6.76–80.87) |
| Prevalence of elevated liver enzyme levels (%) |
41.79 ± 33.71 (14.93–100.00) |
14.71 ± 11.97 (0.00–33.33) |
12.11 ± 12.95 (0.00–33.33) |
| Prevalence of DILI (%) |
(0.00–75.76) |
19.487 ± 18.20 d (0.68–75.00) |
20.47 ± 20.68 c (0.00–100.00) |
| Time to DILI (days) | 26.07 ± 15.46 (14.00–55.20) | ||
Abbreviations: AUC0–24, area under the curve from 0 to 24 h; C max, maximum plasma concentration; C t , plasma concentration at t hour; DILI, drug‐induced liver injury; IA, intermediate acetylator; INH, isoniazid; RA, rapid acetylator; SA, slow acetylator.
All data expressed as mean ± SD (range).
Data were significant differences between SA and RA (p < 0.05).
Data were significant differences between SA and RA (p < 0.0001).
Data were significant differences between SA and IA (p < 0.001).
The normal therapeutic range of C max of INH in plasma ranged between 3 and 6 μg/mL. 17 The average C max of INH in the adult SA groups (7.16 ± 4.85 μg/mL) was not significantly different from that in the IA (5.11 ± 2.78 μg/mL) (p > 0.05) and RA groups (4.84 ± 3.60 μg/mL) (p > 0.05). The C2 or C3 of INH is used as a parameter to measure the safety and efficacy of INH in drug monitoring. The therapeutic range of C2 was the same as the C max of INH, whereas C3 ranged between 1 and 2 μg/mL. 18 The average C2 of INH in SA (6.82 ± 3.56 μg/mL) was higher than that of IA (4.66 ± 2.69 μg/mL) and RA (3.59 ± 3.01 μg/mL). The average C2 values of INH were significantly different among the SA, IA, and RA groups (p = 0.041) with a significant difference between the SA and RA types (p < 0.05). Similarly, the average C3 of INH in the SA group was 4.84 ± 1.31 μg/mL, higher than that in IA and RA groups, which were 2.14 ± 0.09 and 1.57 ± 0.73 μg/mL, respectively. The average C3 values of INH were significantly different among the SA, IA, and RA groups (p = 0.011), there was a significant difference between the SA and RA groups (p < 0.05). Another PK parameter that indicates the efficacy of INH is AUC0–24. The average INH AUC0–24 was 40.38 ± 33.42, 15.48 ± 4.68, and 19.46 ± 25.16 μg h/mL in the SA, IA, and RA groups, respectively (Table 2). The average INH AUC0–24 was significantly different among the SA, IA, and RA groups (p = 0.01), and the SA and RA groups were significantly different (p < 0.05).
The mean prevalence of DILI was 36.23 ± 19.84% in the SA group, which was significantly higher than that in the IA (19.49 ± 18.20%) (p < 0.001) and RA (20.47 ± 20.68%) groups (p < 0.0001). The mean prevalence of elevated liver enzyme levels was significantly different among the SA, IA, and RA groups (p = 0.042). The mean prevalence of elevated liver enzymes in the SA groups was 41.79 ± 33.71%, the highest value among the acetylator types. The time to DILI occurrence ranged from 14 to 55.2 days with an average of 26.07 ± 15.46 days. For early detection of DILI, we suggest monitoring liver function profiles at least 14 days after the first dose of INH administration (Table 2).
By comparing the effect of NAT2 phenotypes on PK parameters between adults and children, the average C max and AUC0–24 of INH in both was highest in the SA group. Moreover, the average INH C max in adults and children in the SA group was supratherapeutic, leading patients with slow NAT2 activity to have more toxicity from INH. As expected, the INH dose in children (9.40 ± 0.50 mg/kg/day) was significantly higher than that in adults (5.98 ± 2.73 mg/kg/day) (p = 0.0008). This result is consistent with the revised INH dose for children in the WHO guidelines for TB treatment 7 (Table 3). Safety profiles, including the prevalence of DILI and liver enzyme elevation, were not reported in children.
TABLE 3.
Pharmacokinetic parameters stratified by NAT2 phenotype in adults and children.
| Parameter a | Adults | Children |
|---|---|---|
| C max (range) (μg/mL) | ||
| SA |
7.16 ± 4.85 (1.44–20.40) |
6.43 ± 3.87 (2.20–16.86) |
| IA |
5.11 ± 2.78 (1.52–9.77) |
5.21 ± 2.30 (1.78–8.70) |
| RA |
4.84 ± 3.60 (1.39–15.57) |
4.72 ± 3.82 (0.60–16.00) |
| AUC0–24 (μg h/mL) | ||
| SA |
40.38 ± 33.42 (17.67–119.83) |
50.52 ± 30.97 (15.18–78.25) |
| IA |
15.48 ± 4.68 (9.87–20.00) |
21.43 ± 13.35 (9.78–36.00) |
| RA |
19.46 ± 25.16 (6.76–80.87) |
33.21 ± 30.15 (7.18–75.25) |
| Dose of INH (mg/kg/day) |
5.98 ± 2.73 b (4.70–18.30) |
9.40 ± 5.03 b (4.76–25.00) |
Abbreviations: AUC0–24, area under the curve from 0 to 24 h; C max, maximum plasma concentration; IA, intermediate acetylator; INH, isoniazid; RA, rapid acetylator; SA, slow acetylator.
All data expressed as mean ± SD (range).
Data were significant differences between adults and children (p < 0.001).
We also analyzed the data by continent (Table 4 and Figure 2). The continents (and included countries) comprised: (1) Africa (Tunisia, Morrocco, Tanzania, Ethiopia, and South Africa), (2) the Americas (Canada, Mexico, Argentina, Brazil, Peru, and Venezuela), (3) Asia (China, Mongolia, Taiwan, Japan, South Korea, Thailand, Vietnam, Indonesia, India, Iran, and Turkey), and (4) Europe (Poland, Switzerland, Italy, Portugal, and Spain).
TABLE 4.
Subgroup analysis by continent in pharmacokinetic parameters and prevalence of DILI data stratified by NAT2 phenotype.
| Parameters stratified by continent | NAT2 phenotype | ||
|---|---|---|---|
| SA | IA | RA | |
| Africa | |||
| C max (μg/mL) a | 7.01 ± 6.91 | 2.73 ± 1.08 | 5.55 ± 5.09 |
| AUC0–24 (μg h/mL) a | 68.75 ± 72.24 | N/A | 45.43 ± 50.11 |
| DILI prevalence (%) a | 41.86 ± 22.96 | 36.60 ± 34.54 | 33.85 ± 47.61 |
| NAT2 phenotype distribution (%) | 60.16 | 21.12 | 18.28 |
| The Americas | |||
| C max (μg/mL) a | 7.20 ± 2.20 | 7.10 ± 3.00 | 3.90 ± 1.40 |
| AUC0–24 (μg h/mL) a | 26.10 ± 10.70 | 20.00 ± 9.10 | 7.50 ± 2.50 |
| DILI prevalence (%) a | 29.06 ± 18.89 | 15.12 ± 17.93 | 16.91 ± 13.35 |
| NAT2 phenotype distribution (%) | 41.94 | 31.17 | 26.81 |
| Asia | |||
| C max (μg/mL) a | 7.31 ± 3.65 | 5.85 ± 3.17 | 4.54 ± 2.54 |
| AUC0–24 (μg h/mL) a | 27.23 ± 10.96 | 13.30 ± 4.94 | 9.56 ± 3.71 |
| DILI prevalence (%) a | 42.83 ± 27.61 | 17.12 ± 10.56 | 19.04 ± 16.24 |
| NAT2 phenotype distribution (%) | 34.42 | 32.47 | 28.54 |
| Europe | |||
| C max (μg/mL) a | 7.09 b | 5.80 b | 3.39 b |
| AUC0–24 (μg h/mL) a | 38.86 ± 4.75 | 17.50 b | 14.34 ± 7.27 |
| DILI prevalence (%) a | 27.36 ± 22.96 | 18.94 ± 16.73 | 24.64 ± 22.04 |
| NAT2 phenotype distribution (%) | 58.93 | 33.92 | 6.24 |
Abbreviations: AUC0–24, area under the curve from 0 to 24 h; C max, maximum plasma concentration; DILI, drug‐induced liver injury; IA, intermediate acetylator; INH, isoniazid; N/A, no data reported; RA, rapid acetylator; SA, slow acetylator.
Data expressed as mean ± SD.
Data derived from one study from Poland (Table S1).
FIGURE 2.
N‐acetyltransferase 2 (NAT2) phenotype distribution (%), C max (μg/mL), and DILI prevalence (%) stratified by NAT2 status categorized by continent group. Red, purple, green, and yellow colors indicate the group of countries in Africa, Europe, The Americas, and Asia, respectively. C max, maximum plasma concentration; DILI, drug‐induced liver injury; IA, intermediate acetylator; RA, rapid acetylator; SA, slow acetylator.
The SA type was the most prevalent worldwide. In Africa, the frequency of SA was as high as 60%. Among SA carriers in Africa, DILI could have occurred in approximately 41% after the treatment of TB with a regimen including INH. The prevalence of DILI in Africa has risen by more than 30% for carriers of all acetylator types. Moreover, the average INH C max of carriers of the SA type in Africans (7.01 ± 6.91 μg/mL) was higher than the therapeutic target (3–6 μg/mL). The frequency of SA carriers in Europe, the Americas, and Asia was 58.93%, 41.94%, and 34.42%, respectively. Among SA carriers in Europe, the Americas, and Asia, the prevalence of DILI was 27.36 ± 22.96%, 29.06 ± 18.89%, and 42.83 ± 27.61%, respectively. The average INH C max of SA carriers among in all continents was approximately 7 μg/mL higher than the therapeutic range, leading to increased INH toxicity and DILI. The frequency of RA carriers in Europe (6.24%) showed the lowest proportion, whereas Asia (28.54%) had the highest proportion.
NAT2 status impacts INH efficacy
The INH efficacy is indicated by the incidence of ETF in each acetylator group. ETF was defined as a combined surrogate end point: (1) a positive sputum culture at the 8‐week examination or (2) no sign of improvement in the chest radiograph at 8 weeks for patients with a negative sputum culture at baseline. 6 Only one efficacy study was included a randomized controlled trial that aimed to investigate the dosing of INH between patients who received the standard dose (Std) of treatment and the pharmacogenetics (PGx)‐guided dose. The average Std doses in the SA, IA, and RA groups were 5.50 ± 0.80, 5.50 ± 0.90, and 5.50 ± 1.00 mg/kg/day, respectively, whereas the average PGx‐guided doses in the SA and RA groups were 2.60 ± 0.30 and 8.80 ± 1.30 mg/kg/day, respectively (Table 5). For the SA group, the ETF incidence for the Std dose was 22.20%, whereas in those given the PGx dose had no ETF despite that they received a lower INH dose than the Std group. For the RA group, the ETF incidence in PGx INH‐dosed patients (15.00%) was lower than that given the Std INH dose (39.50%) and lower than that in IA group (26.80%). Thus, giving a PGx‐guided INH dose may improve treatment efficacy by reducing the ETF.
TABLE 5.
Impact of NAT2 phenotype on the efficacy of INH.
| Parameter | SA | IA | RA | ||
|---|---|---|---|---|---|
| Std | PGx | Std/PGx | PGx | Std | |
| Number of patients, n | 9 | 7 | 64 | 44 | 48 |
| Sex, n | |||||
|
Male (112 (65.12%)) |
4 | 4 | 44 | 25 | 35 |
|
Female (60 (34.88%)) |
5 | 3 | 20 | 19 | 13 |
|
Age (years), median (IQR) |
35.00 (35.00–43.00) |
48.00 (36.00–58.00) |
48.50 (34.00–60.00) |
50.50 (33.00–57.80) |
51.00 (30.00–60.20) |
|
Weight (kg), median (IQR) |
55.00 (52.80–59.80) |
59.50 (47.90–62.80) |
54.10 (50.10–62.20) |
50.00 (45.90–55.20) |
57.50 (47.60–63.80) |
|
Dose of INH (mg/kg/day), mean ± SD |
5.50 ± 0.80 | 2.60 ± 0.30 | 5.50 ± 0.90 | 8.80 ± 1.30 | 5.50 ± 1.00 |
|
Early treatment failure, n (incidence, %) |
2 (22.20) | 0 (0.00) | 15 (26.80) | 6 (15.00) | 17 (39.50) |
| Number of patients after 8 weeks of treatment, n | 9 | 7 | 56 | 40 | 43 |
Note: Data derived from Azuma et al. 6
Abbreviations: IA, intermediate acetylator; INH, isoniazid; IQR, interquartile range; n, number of patients; PGx, pharmacokinetic‐guided dose group; RA, rapid acetylator; SA, slow acetylator; Std, standard dose group.
DISCUSSION
In this systematic review, we revealed the impact of NAT2 genotype‐related acetylator metabolism on the dose, PK, efficacy, and safety of INH in patients with TB. The average INH given to patients with TB was consistent with the WHO 2022 and Thai 2021 guidelines for TB treatment, 4–6 mg/kg/day. 19 However, the maximum INH dose in one study in this review was 18.3 mg/kg/day, which was used for multidrug‐resistant TB treatment. 20 However, the average dose of INH in children was lower than that suggested by WHO 2022 7 (20 mg/kg/day, maximum 400 mg/kg/day) and Thai 2021 21 guidelines (10–15 mg/kg/day, maximum 300 mg/day or a high dose of INH at 15–20 mg/kg/day, maximum 900 mg/day). The average dose in children being lower than the suggested dose may be due to changes in dosing that have occurred from the previous (5 mg/kg/day) to the revised (10 mg/kg/day) guidelines since 2011. Additionally, the maximum INH dose in one study for children with TB (25.0 mg/kg/day) was employed in the 3HP regimen for latent TB (LTB) treatment. The 3HP regimen comprised INH and rifapentine (RPT) administered once weekly for 3 months. 22 Unfortunately, a weight‐based INH dose in this review could not be categorized by NAT2 phenotype status. However, we found only one study that reported the INH dosage usage by PGx‐guided regimen (2.5 mg/kg/day for SA, 5.0 mg/kg/day for IA, and 7.5 mg/kg/day for RA). 6
Globally, adult patients with TB who have the SA type have an average INH C max above the therapeutic level. As a consequence of this supratherapeutic level, the mean prevalence of DILI in the SA group was higher than that of the other phenotype groups. The highest average C max (20.4 μg/mL) in this review was in a study in which adult patients with TB received the 3HP regimen (900 mg of INH and RPT for weight ≥ 50 kg) for LTB treatment. Our results align with another systematic review, demonstrating that dose‐normalized INH C max in the RA and IA groups was significantly lower than that in the SA group of patients with TB. 23 However, our review surpasses the previous one in several aspects. Firstly, it provides more up‐to‐date information. Additionally, we included a greater number of recruited articles compared with the previous systematic review, and we gave further consideration to clinical issues, particularly DILI. Moreover, we meticulously categorized PK data and pharmacogenetics by continent worldwide, providing crucial information for the advancement of population PK modeling of INH. Population PK modeling holds promise as a potent tool for elucidating PK variability, largely attributed to NAT2 genotype in the disposition of INH. Recently, there has been a systematic review of population PK models, which reported that increasing INH clearance was approximately two‐ to threefold higher for NAT2 rapid acetylators compared with slow acetylators. This underscores the significance of the NAT2 genotype in the PK variability of INH disposition. 24
When comparing the INH C max between adults and children, INH C max in children in the SA group was higher than the therapeutic range, similar to that in SA adults. Despite the average INH dose in children being significantly higher than that in adults, the C max in children was comparable. This may be due to blood flow to the liver. In general, children have a larger liver and hepatic blood flow per body weight than adults, which results in increased hepatic substance clearance. 25 Moreover, Schaaf et al. 26 reported that INH elimination in children (≤13 years old) is faster than that in adults. Therefore, the INH dose given to children should be higher than that in adults. However, some studies indicated that NAT2 activity maturation in children in the RA group (4 months to 14 years) occurred after 4 years of age. 27 , 28 Immature NAT2 within the first 4 years of life might impair the enzyme activity, leading to an increase in INH plasma concentration with toxicity. Therefore, children aged <4 years should be closely monitored.
DILI is an important adverse effect in patients receiving INH. In terms of INH safety for different NAT2 phenotypes, our findings show the highest mean prevalence of DILI in the adult SA group, significantly higher than that in the IA and RA groups. The SD of the mean prevalence of DILI in each NAT2 phenotype was >10%, and the mean prevalence of DILI had a wide range from 0% to 100% because of the distribution of NAT2 phenotypes worldwide and the number of participants in NAT2 phenotype group. For example, Mthiyane et al. 29 studied patients with TB in South Africa and found that the frequency of SA was higher than that of RA. Only one patient with TB was recruited into the RA group, and this patient presented DILI, resulting in a 100% DILI prevalence. 29 In contrast, Azuma et al. 6 showed that the frequency of SA in Japanese patients with TB was lower than that of RA, with 16 participants in the SA group compared with 92 in the RA group. Moreover, TB patients in the SA group received PGx‐guided INH dose. No DILI was observed in the SA group of Japanese patients with TB. 6 The prevalence of DILI stratified by NAT2 phenotype in children was not reported because DILI is a rare condition in children. 30 , 31 DILI was reported in 6.9% of children who received INH and RIF, unstratified by NAT2 phenotypes. 30 Meta‐analyses have shown that SA is significantly related an approximately three‐times increased risk of INH‐induced hepatotoxicity compared with those in the IA and RA groups. 32 , 33 The molecular mechanism of INH metabolism by NAT2 in SA supports our results. The risk of INH‐induced hepatotoxicity increased in SA carriers given the Std INH dose, but those with INH in the supratherapeutic range had the highest DILI prevalence. Typically, INH is acetylated by NAT2 into acetylisoniazid, a nontoxic agent. Alternatively, INH can be slightly hydrolyzed by amidase into hydrazine, a hepatotoxic agent. 3 Therefore, the slow activity of NAT2 in SA carriers results in an increased INH plasma concentration in the supratherapeutic range. Then, the primary INH metabolic pathway may change to the hydrolysis pathway, leading to the accumulation of hydrazine. Hydrazine exposure can change the production and transportation of lipids in hepatocytes, leading to hepatocellular steatosis, 34 and inhibit complex I in the electron transport chain, causing mitochondrial oxidative stress. 35 Therefore, the NAT2 phenotype is a crucial factor related to DILI, especially for those in the SA group.
Next, we performed subgroup analysis categorized by continent, including Africa, the Americas, Asia, and Europe. The results showed that most patients with TB worldwide have the SA type. Furthermore, patient TB with SA type on each continent had an average INH C max (>7 μg/mL) above the therapeutic range (3–6 μg/mL), and the mean prevalence of DILI ranged from 27.36% to 42.83%. These results imply that, globally, patients with TB with the SA type have a higher risk of INH toxicity, especially hepatotoxicity. In 2008, Sabbagh et al. 36 studied the worldwide distribution of NAT2 diversity and reported that the SA type was predominant in Africa and Europe. The results of Sabbagh et al. were consistent with the results of this review. Unfortunately, NAT2 data for Australia were limited. However, a study in 1993 reported that the NAT2 phenotype in Western Australia was 63.3% RA and 36.7% SA. 37
One study indicated that C2 was clinically related to C max, which provides a way to monitor INH levels. 17 However, it should be considered that the C2 might be lower than the C max, especially in the IA and RA types, this issue should be considered. 38 Thus, NAT2 genotyping before blood sampling may benefit the interpretation of INH monitoring. Apart from C2, some studies have recommended the use of C3 for INH monitoring. 39 Typically, the therapeutic range of C3 is between 1 and 2 μg/mL. 40 Our results showed that the average C3 in the IA and RA groups given the Std INH dose ranged within therapeutic concentrations, whereas the average C3 in the SA group was >2 μg/mL. Patients who had an INH C3 > 2 μg/mL might be at risk of INH toxicity, especially hepatotoxicity. Therefore, our findings show that adults in the SA group given the Std INH dose may be at the risk of toxicity because the average C3 in this group was higher than the therapeutic range. Ben Fredj et al. 41 reported that an INH C3 > 2 μg/mL was associated with an increase in the neurotoxic adverse effects of INH, whereas an INH C3 > 3.69 μg/mL was related to INH‐induced hepatoxicity. In terms of INH monitoring, C3 may be suitable for routine monitoring. Alshaikheid et al. 42 found that using the Bayesian approach, a one timepoint concentration for a limited sampling strategy including C3, could be sufficient to estimate the INH AUC0–24 in routine INH monitoring. 42 Choosing an appropriate sampling timepoint for INH monitoring might be different for each NAT2 activity group. In general, the time to maximum concentration (T max) of INH ranges from 0.6 to 2.0 h depending on NAT2 phenotype. The T max of INH in SA carriers is slower than in RA carriers. 43 , 44 If a patient with TB harbors the RA type, the C3 of INH may be an appropriate parameter for monitoring because the C2 of INH in the RA group may be lower than INH C max leading to inappropriate interpretation. Conversely, INH C2 might be appropriate for patients with TB with the SA or IA types because the INH C2 in these groups is close to the same level as C max. Thus, NAT2 genotyping may be advantageous for determining the blood sampling time for INH monitoring.
Another PK parameter, AUC0–24, could be used to monitor the efficacy of INH in TB treatment. Pasipanodya et al. 45 mentioned that patients who had an INH AUC0–24 < 52 μg h/mL had poor long‐term clinical outcomes such as microbiologic failure or relapse or death. Conversely, Cojutti et al. 46 indicated that an INH AUC0–24 < 55 μg h/mL was associated with a lower risk of hepatotoxicity. Our results in this systematic review indicate that the AUC0–24 was <52 μg h/mL in all NAT2 phenotype carriers. Therefore, our findings on AUC0–24 might imply that people with all acetylator types should be closely monitored to prevent poor outcomes. Apart from AUC0–24, AUC0−∞ (target ≥ 10.52 μg h/mL) might be used as a parameter for efficacy monitoring associated with the maximum early bactericidal activity (EBA90). 47 The EBA90 is the ability to kill approximately 90% of rapidly metabolizing bacilli in the sputum of microscopy smear‐positive patients with pulmonary TB during the first 2 days of treatment. 47 However, our results did not report AUC0−∞ because there were insufficient data on this parameter.
In terms of efficacy, only one study was included in this review. Azuma et al. 6 indicated the incidence of ETF in patients who received Std and PGx‐guided doses stratified by NAT2 phenotype. 6 The results suggested that the PGx‐guided INH doses may improve the efficacy of TB treatment by reducing the incidence of ETF in the RA group.
The strengths of this systematic review were that we focused on how NAT2 phenotypes affected PK parameters and the prevalence of DILI. Second, we compared the INH dose and PK parameters in adults and children. Additionally, this is the first article to illustrate the global C max and DILI prevalence of the TB population by NAT2 phenotype distribution. Fourth, we assessed C2 and C3 data, which are common sampling timepoints in clinical practice for INH monitoring. Finally, we analyzed the data as the mean and SD which are more convenient parameters to compare in each group.
The limitations of this systematic review were as follows: First, we excluded non‐English articles that may have reported pertinent information. Second, the ethnicity of the patients was not reported for most of the studies in our systematic review; thus, we stratified the population of each study by countries and continents. Third, accessibility was limited for some full‐text articles that might have included data from countries such as the USA and Australia.
CONCLUSIONS
Traditional weight‐based dosing of INH in most children and adults yielded therapeutic INH levels, except for carriers of the SA NAT2 type. DILI was more commonly observed among SA carriers, which comprised the largest population of patients with TB worldwide. Genotype‐guided dosing may relieve this issue and prevent ETF. C2 or C3 may be appropriate parameters for the routine monitoring of INH safety.
FUNDING INFORMATION
No funding was received for this work.
CONFLICT OF INTEREST STATEMENT
The authors declared no competing interests in this work.
Supporting information
Table S1
Table S2
ACKNOWLEDGMENTS
The authors would like to thank Enago (www.enago.com) for the English language review.
Surarak T, Chumnumwat S, Nosoongnoen W, Tragulpiankit P. Efficacy, safety, and pharmacokinetics of isoniazid affected by NAT2 polymorphisms in patients with tuberculosis: A systematic review. Clin Transl Sci. 2024;17:e13795. doi: 10.1111/cts.13795
REFERENCES
- 1. Kalscheuer R, Palacios A, Anso I, et al. The Mycobacterium tuberculosis capsule: a cell structure with key implications in pathogenesis. Biochem J. 2019;476(14):1995‐2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Laborde J, Deraeve C, Bernardes‐Genisson V. Update of antitubercular prodrugs from a molecular perspective: mechanisms of action, bioactivation pathways, and associated resistance. Chem Med Chem. 2017;12(20):1657‐1676. [DOI] [PubMed] [Google Scholar]
- 3. Erwin ER, Addison AP, John SF, Olaleye OA, Rosell RC. Pharmacokinetics of isoniazid: the good, the bad, and the alternatives. Tuberculosis. 2019;116:S66‐S70. [DOI] [PubMed] [Google Scholar]
- 4. Gutierrez‐Virgen JE, Pina‐Pozas M, Hernandez‐Tobias EA, et al. NAT2 global landscape: genetic diversity and acetylation statuses from a systematic review. PLoS One. 2023;18(4):e0283726. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Sotsuka T, Sasaki Y, Hirai S, Yamagishi F, Ueno K. Association of isoniazid‐metabolizing enzyme genotypes and isoniazid‐induced hepatotoxicity in tuberculosis patients. In Vivo. 2011;25(5):803‐812. [PubMed] [Google Scholar]
- 6. Azuma J, Ohno M, Kubota R, et al. NAT2 genotype guided regimen reduces isoniazid‐induced liver injury and early treatment failure in the 6‐month four‐drug standard treatment of tuberculosis: a randomized controlled trial for pharmacogenetics‐based therapy. Eur J Clin Pharmacol. 2013;69(5):1091‐1101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. WHO . WHO Consolidated Guidelines on Tuberculosis: Module 5: Management of Tuberculosis in Children and Adolescents. WHO Guidelines Approved by the Guidelines Review Committee; 2022. [PubMed] [Google Scholar]
- 8. Verhagen LM, Coenen MJ, López D, et al. Full‐gene sequencing analysis of NAT2 and its relationship with isoniazid pharmacokinetics in Venezuelan children with tuberculosis. Pharmacogenomics. 2014;15(3):285‐296. [DOI] [PubMed] [Google Scholar]
- 9. Mukherjee A, Velpandian T, Singla M, Kanhiya K, Kabra SK, Lodha R. Pharmacokinetics of isoniazid, rifampicin, pyrazinamide and ethambutol in Indian children. BMC Infect Dis. 2015;15(1):126. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Pouplin T, Bang ND, Toi PV, et al. Naïve‐pooled pharmacokinetic analysis of pyrazinamide, isoniazid and rifampicin in plasma and cerebrospinal fluid of Vietnamese children with tuberculous meningitis. BMC Infect Dis. 2016;16(1):144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Shah I, Jadhao N, Mali N, Deshpande S, Gogtay N, Thatte U. Pharmacokinetics of isoniazid in Indian children with tuberculosis on daily treatment. Int J Tuberc Lung Dis. 2019;23(1):52‐57. [DOI] [PubMed] [Google Scholar]
- 12. Panjasawatwong N, Wattanakul T, Hoglund RM, et al. Population pharmacokinetic properties of antituberculosis drugs in Vietnamese children with tuberculous meningitis. Antimicrob Agents Chemother. 2021;65(1):e00487‐20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Ho HT, Wang TH, Hsiong CH, et al. The NAT2 tag SNP rs1495741 correlates with the susceptibility of antituberculosis drug‐induced hepatotoxicity. Pharmacogenet Genom. 2013;23(4):200‐207. [DOI] [PubMed] [Google Scholar]
- 14. Aithal GP, Watkins PB, Andrade RJ, et al. Case definition and phenotype standardization in drug‐induced liver injury. Clin Pharmacol Ther. 2011;89(6):806‐815. [DOI] [PubMed] [Google Scholar]
- 15. Wan X, Wang W, Liu J, Tong T. Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range. BMC Med Res Methodol. 2014;14:135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Nicoletti P, Devarbhavi H, Goel A, et al. Genetic risk factors in drug‐induced liver injury due to isoniazid‐containing antituberculosis drug regimens. Clin Pharmacol Ther. 2021;109(4):1125‐1135. [DOI] [PubMed] [Google Scholar]
- 17. Prahl JB, Johansen IS, Cohen AS, Frimodt‐Møller N, Andersen ÅB. Clinical significance of 2 h plasma concentrations of first‐line anti‐tuberculosis drugs: a prospective observational study. J Antimicrob Chemother. 2014;69(10):2841‐2847. [DOI] [PubMed] [Google Scholar]
- 18. Alshaikheid M, Chaabane A, Ben Fredj N, et al. Limited sampling strategy for predicting isoniazid exposure in patients with extrapulmonary tuberculosis. J Clin Pharm Ther. 2020;45(3):503‐512. [DOI] [PubMed] [Google Scholar]
- 19. WHO consolidated guidelines on tuberculosis: Module 4: Treatment ‐ Drug‐susceptible tuberculosis treatment. WHO Guidelines Approved by the Guidelines Review Committee; 2022. [PubMed] [Google Scholar]
- 20. Masiphephethu MV, Sariko M, Walongo T, et al. Pharmacogenetic testing for NAT2 genotypes in a Tanzanian population across the lifespan to guide future personalized isoniazid dosing. Tuberculosis. 2022;136:102246. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Division of Tuberculosis . National Tuberculosis Control Programme Guideline, Thailand 2021. Division of Tuberculosis; 2021. [Google Scholar]
- 22. Phaisal W, Jantarabenjakul W, Wacharachaisurapol N, et al. Pharmacokinetics of isoniazid and rifapentine in young pediatric patients with latent tuberculosis infection. Int J Infect Dis. 2022;122:725‐732. [DOI] [PubMed] [Google Scholar]
- 23. Hong BL, D'Cunha R, Li P, et al. A systematic review and meta‐analysis of isoniazid pharmacokinetics in healthy volunteers and patients with tuberculosis. Clin Ther. 2020;42(11):e220‐e241. [DOI] [PubMed] [Google Scholar]
- 24. Thomas L, Raju AP, Chaithra MSS, et al. Influence of N‐acetyltransferase 2 (NAT2) genotype/single nucleotide polymorphisms on clearance of isoniazid in tuberculosis patients: a systematic review of population pharmacokinetic models. Eur J Clin Pharmacol. 2022;78(10):1535‐1553. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Ginsberg G, Hattis D, Sonawane B, et al. Evaluation of child/adult pharmacokinetic differences from a database derived from the therapeutic drug literature. Toxicol Sci. 2002;66(2):185‐200. [DOI] [PubMed] [Google Scholar]
- 26. Schaaf HS, Parkin DP, Seifart HI, et al. Isoniazid pharmacokinetics in children treated for respiratory tuberculosis. Arch Dis Child. 2005;90(6):614‐618. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Pariente‐Khayat A, Rey E, Gendrel D, et al. Isoniazid acetylation metabolic ratio during maturation in children. Clin Pharmacol Ther. 1997;62(4):377‐383. [DOI] [PubMed] [Google Scholar]
- 28. Keller GA, Fabian L, Gomez M, Gonzalez CD, Diez RA, Girolamo GD. Age‐distribution and genotype‐phenotype correlation for N‐acetyltransferase in Argentine children under isoniazid treatment. Int J Clin Pharmacol Ther. 2014;52(4):292‐302. [DOI] [PubMed] [Google Scholar]
- 29. Mthiyane T, Millard J, Adamson J, et al. N‐acetyltransferase 2 genotypes among Zulu‐speaking South Africans and isoniazid and N‐acetyl‐isoniazid pharmacokinetics during antituberculosis treatment. Antimicrob Agents Chemother. 2020;64(4):e02376‐19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Devarbhavi H, Karanth D, Prasanna KS, Adarsh CK, Patil M. Drug‐induced liver injury with hypersensitivity features has a better outcome: a single‐center experience of 39 children and adolescents. Hepatology. 2011;54(4):1344‐1350. [DOI] [PubMed] [Google Scholar]
- 31. Dijmarescu I, Guta OM, Brezeanu LE, Dijmarescu AD, Becheanu CA, Pacurar D. Drug‐induced hepatitis in children: the experience of a single center in Romania. Children (Basel). 2022;9(8):1136. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Du H, Chen X, Fang Y, et al. Slow N‐acetyltransferase 2 genotype contributes to anti‐tuberculosis drug‐induced hepatotoxicity: a meta‐analysis. Mol Biol Rep. 2013;40(5):3591‐3596. [DOI] [PubMed] [Google Scholar]
- 33. Shi J, Xie M, Wang J, Xu Y, Liu X. Susceptibility of N‐acetyltransferase 2 slow acetylators to antituberculosis drug‐induced liver injury: a meta‐analysis. Pharmacogenomics. 2015;16(18):2083‐2097. [DOI] [PubMed] [Google Scholar]
- 34. Richards VE, Chau B, White MR, McQueen CA. Hepatic gene expression and lipid homeostasis in C57BL/6 mice exposed to hydrazine or acetylhydrazine. Toxicol Sci. 2004;82(1):318‐332. [DOI] [PubMed] [Google Scholar]
- 35. Ramachandran A, Visschers RGJ, Duan L, Akakpo JY, Jaeschke H. Mitochondrial dysfunction as a mechanism of drug‐induced hepatotoxicity: current understanding and future perspectives. J Clin Transl Res. 2018;4(1):75‐100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Sabbagh A, Langaney A, Darlu P, Gerard N, Krishnamoorthy R, Poloni ES. Worldwide distribution of NAT2 diversity: implications for NAT2 evolutionary history. BMC Genet. 2008;9:21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Ilett KF, Chiswell GM, Spargo RM, Platt E, Minchin RF. Acetylation phenotype and genotype in aboriginal leprosy patients from the north‐west region of Western Australia. Pharmacogenetics. 1993;3(5):264‐269. [DOI] [PubMed] [Google Scholar]
- 38. Ungcharoen U, Sriplung H, Mahasirimongkol S, et al. The influence of NAT2 genotypes on isoniazid plasma concentration of pulmonary tuberculosis patients in southern Thailand. Tuberc Respir Dis (Seoul). 2020;83(Supple 1):S55‐S62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Toure A, Cabral M, Niang A, et al. Prevention of isoniazid toxicity by NAT2 genotyping in Senegalese tuberculosis patients. Toxicol Rep. 2016;3:826‐831. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Fredj NB, Romdhane HB, Woillard JB, et al. Population pharmacokinetic model of isoniazid in patients with tuberculosis in Tunisia. Int J Infect Dis. 2021;104:562‐567. [DOI] [PubMed] [Google Scholar]
- 41. Ben Fredj N, Gam R, Kerkni E, et al. Risk factors of isoniazid‐induced hepatotoxicity in Tunisian tuberculosis patients. Pharmacogenomics J. 2017;17(4):372‐377. [DOI] [PubMed] [Google Scholar]
- 42. Alshaikheid M, Ben Fredj N, Hannachi I, et al. Development of a population pharmacokinetic model and Bayesian estimators for isoniazid in Tunisian tuberculosis patients. Pharmacogenomics J. 2021;21(4):467‐475. [DOI] [PubMed] [Google Scholar]
- 43. Huerta‐García AP, Medellín‐Garibay SE, Salazar‐González RA, et al. Anthropometric and genetic factors associated with the exposure of rifampicin and isoniazid in Mexican patients with tuberculosis. Ther Drug Monit. 2019;41(5):648‐656. [DOI] [PubMed] [Google Scholar]
- 44. Lee J, Choi G, Yang S, et al. Development of a limited sampling strategy for the estimation of isoniazid exposure considering N‐acetyltransferase 2 genotypes in Korean patients with tuberculosis. Tuberculosis. 2021;127:102052. [DOI] [PubMed] [Google Scholar]
- 45. Pasipanodya JG, McIlleron H, Burger A, Wash PA, Smith P, Gumbo T. Serum drug concentrations predictive of pulmonary tuberculosis outcomes. J Infect Dis. 2013;208(9):1464‐1473. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Cojutti P, Duranti S, Isola M, et al. Might isoniazid plasma exposure be a valuable predictor of drug‐related hepatotoxicity risk among adult patients with TB? J Antimicrob Chemother. 2016;71(5):1323‐1329. [DOI] [PubMed] [Google Scholar]
- 47. Donald PR, Parkin DP, Seifart HI, et al. The influence of dose and N‐acetyltransferase‐2 (NAT2) genotype and phenotype on the pharmacokinetics and pharmacodynamics of isoniazid. Eur J Clin Pharmacol. 2007;63(7):633‐639. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1
Table S2