NR1I2 genetic polymorphisms and the risk of anti‐tuberculosis drug‐induced hepatotoxicity: A systematic review and meta‐analysis

Abstract Anti‐tuberculosis drug‐induced hepatotoxicity (ATDH) is a serious adverse drug reaction. Conflicting results have been obtained regarding the associations of nuclear receptor subfamily 1 group I member 2 (NR1I2) gene polymorphisms on susceptibility to ATDH. Therefore, we aimed to evaluate the associations using a systematic review/meta‐analysis approach. PubMed, Medline, Cochrane Library, Web of Science and SinoMed databases were searched for all eligible studies from inception to June 10, 2020. Pooled adjusted odds ratios (ORs) with 95% confidence intervals (CIs) were employed to evaluate the strength of the association between the NR1I2 polymorphisms and the risk of ATDH. Subgroup analysis was performed by region of origin, and meta‐regression were performed to detect potential sources of heterogeneity. A total of five case‐control studies involving 572 cases and 1867 controls were identified. Fourteen SNPs in the NR1I2 gene have been reported, and the most heavily studied SNPs were rs3814055 and rs7643645. The pooled estimates did not exhibit any significant associations between SNPs rs3814055 and rs7643645 and the risk of ATDH (rs3814055: dominant model, OR = 1.00, 95% CI: 0.82‐1.22, P = 1.00; recessive model, OR = 1.17, 95% CI: 0.76‐1.78, P = .48; rs7643645: dominant model, OR = 1.04, 95% CI: 0.64‐1.68, P = .89; recessive model, OR = 0.98, 95% CI: 0.65‐1.49, P = .93). Subgroup analysis obtained similar negative results in Chinese patients, and the diagnostic criteria of ATDH may be the source of heterogeneity. Based on the meta‐analysis described in this report, we did not observe any association between NR1I2 gene polymorphisms and ATDH susceptibility. However, this conclusion should be interpreted with caution due to the low number of studies and the relatively small sample size.


Tuberculosis (TB) is a chronic infectious disease caused by
Mycobacterium tuberculosis (Mtb); this disease remains a major cause of death and suffering worldwide, and its control is a global public health issue. 1 The Global Tuberculosis Report 2019 released by the World Health Organization (WHO) indicated that there were 10 million new cases and 1.5 million deaths in 2018, and China had the second largest number of new TB cases worldwide with 886,000 estimated new cases and an incidence of 61/100,000. 2 The treatment of TB with short-course chemotherapy, recommended by the WHO TB program, has remained largely unchanged for the past 40 years; specifically, with this treatment, a combination of isoniazid (INH), rifampicin (RIF), pyrazinamide (PZA), ethambutol (EMB), and streptomycin is administered for a period of 6-8 months. 3 Although chemotherapy has dramatically increased the effectiveness of TB control, achieving a treatment success rate of 85%, 2 a variety of adverse drug reactions (ADRs) may occur during long-term multidrug combination therapy, among which anti-tuberculosis drug-induced hepatotoxicity (ATDH) is the most common adverse event that necessitates therapy interruption. 4 The manifestations of ATDH may vary from asymptomatic elevations in the liver enzymes to fulminant liver failure. 5 ATDH is also a common cause of drug-induced liver injury (DILI) in China. 6 Although the pathogenic mechanism underlying ATDH has not been fully elucidated, a number of hypotheses on the pathogenesis of ATDH have been proposed, such as drug metabolism and transporter enzymes, immune response, oxidative stress, and mitochondrial dysfunction. 7,8 Among the first-line anti-TB drugs, INH and RIF have been demonstrated to cause hepatotoxicity. 9 Recently, based on a pregnane X receptor (PXR)-humanized mouse model, Li and colleagues made an outstanding and substantial contribution towards elucidating the mechanism of ATDH by determining that cotreatment with RIF and INH causes accumulation of the endogenous hepatotoxin protoporphyrin Ⅸ (PPⅨ) in the liver through PXR-mediated alteration of the heme biosynthesis pathway. 10 PPIX is ubiquitous in all living cells in small amounts as a precursor of heme, and accumulation of PPIX in live human cells can cause hepatobiliary damage, liver injury, and even liver failure. 11,12 The RIF-INH co-therapy caused the accumulation of PPIX through PXR-mediated transcriptional activation of both the cytochrome P450 (CYP450) and aminolevulinic synthase-1 (ALAS1) genes. 13 ALAS1 is the rate-limiting enzyme in the production of heme in the liver and is upregulated by RIF-INH therapy, both by direct transcriptional activation and through derepression of negative feedback due to the incorporation of heme into the CYP450 apoprotein. 10 INH was determined to downregulate ferrochelatase (FECH), the enzyme that converts PPⅨ to heme, and to cause PPⅨ accumulation and liver injury. 12 FECH inhibition and ALAS1 induction may exert a synergistic effect on PPIX accumulation. 14 All these findings help to elucidate the mechanism underlying RIF and INH co-therapy-induced liver injury, which may be applied to study the risk factors and genetic susceptibility to ATDH, as well as the prevention and control of liver injury. PXR, encoded by the nuclear receptor subfamily 1 group I member 2 (NR1I2) gene, is a ligand-dependent transcription factor that is involved in the gene network regulating the metabolism of exogenous and endogenous substances. 15 Based on the PubMed Gene website (www.ncbi.nlm.nih.gov/gene), the human NR1I2 gene on chromosome 3q13.33 comprises ten exons, spans approximately 38 kilobases, encodes 434 amino acids, and contains 11 054 single-nucleotide polymorphisms (SNPs) (GRCh38.p13, Chr3: 119780484-119818485).
Genetic polymorphism could affect gene transcription and the activity of proteins encoded, which in turn may lead to changes in the pharmacokinetic and pharmacodynamic behavior of a drug, observed as differences in drug trans port, drug metabolism, pharmacodynamic drug effects, and adverse events. 16 Similarly, NR1I2 genetic polymorphisms can affect the pharmacokinetics and therapeutic response to many drugs, such as irinotecan, tacrolimus and atazanavir. 17 Rana and colleagues found that some of the non-synonymous variants of PXR may have adverse physiological consequences owing to its influence on the expression levels and functional output of drug-metabolizing enzymes and transporters, 18 and the T-allele was associated with significantly greater transcriptional activity than the C-allele of SNP rs3814055. 19 To date, a number of studies have been conducted to investigate the potential association between NR1I2 genetic polymorphisms and the risk of ATDH, with inconsistent results being obtained. [20][21][22][23][24] For example, earlier studies from Indonesia showed that patients with the TT genotype at rs3814055 had a significantly increased risk of ATDH, 21 but another study from China reached the opposite conclusion, namely, the T allele of rs3814055 was associated with a decreased risk for ATDH. 20 However, a systematic review and meta-analysis can be used to pool results from different studies, thereby enhancing the precision of estimates of treatment effects. 25 To the best of our knowledge, no systematic review or meta-analysis has been undertaken to clarify the effect of these polymorphisms on the risk of ATDH. So, we aimed to evaluate the association between NR1I2 genetic polymorphisms and the risk of ATDH using a systematic review/meta-analysis approach, and provide more accurate conclusions regarding genetic susceptibility research on ATDH and identify areas that warrant further investigation.

| Search strategies
This systematic review and meta-analysis followed the PRISMA statement and guidelines. 26 The literature search was performed anti-tuberculosis drug-induced hepatotoxicity, genetic polymorphisms, meta-analysis, NR1I2, pregnane X receptor in the PubMed, Medline, Cochrane Library, Web of Science and SinoMed (Chinese Biomedical Literature Service System) databases from inception to June 10, 2020, for all relevant papers, and the search terms included "PXR" or "NR1I2" or "pregnane X receptor" or "nuclear receptor subfamily 1 group I member 2", "drug-induced liver injury" or "drug-induced hepatotoxicity" or "drug-induced hepatitis" or "drug-induced liver damage" or "drug-induced hepatic injury" or "toxic hepatitis," and "antituberculosis" or "anti-tuberculosis" or "antitubercular" or "tuberculosis treatment." The work was updated before the statistical analysis was performed to prevent the latest published relative report from being lost. The full search strategies for each database are listed in the Supplementary Material (Table S1).
All the records identified from the databases mentioned above have been imported into EndNote X8 (Thomson Reuters, New York, NY), and duplicate records have been deleted. Reviewers were divided into two groups that worked in parallel. The reviewers independently screened each record by title, keywords, and abstract against the eligibility criteria. Full texts were referred to when information in the records was inadequate for determination. Any disagreement between the two groups of reviewers was resolved by an additional reviewer. Manual searching was performed by reviewing the references of the included studies.

| Inclusion and exclusion criteria
The eligible studies included in the present study met the following criteria: (a) case-control studies to assess the association between NR1I2 polymorphisms and risk of ATDH were analyzed; (b) cases were TB patients with ATDH, while controls were TB patients without ATDH; (c) TB patients receiving first-line anti-TB drug treatment were investigated; (d) studies reported odds ratios (ORs) with 95% confidence intervals (CIs) for the risk of ATDH or sufficient data to estimate ORs and their 95% CIs; and (e) the language was restricted to English or Chinese.
The studies were excluded if (a) they were conference abstracts, protocols, or summaries; (b) the sample size for each group was less than 10; and (c) studies with duplicate data were reported in multiple studies by the same research group.

| Data extraction
An extraction form was designed to extract data, and the following information was extracted from each study if available: (a) study characteristics: first author, publication year, study design, inclusion and exclusion criteria, and sample size; (b) population characteristics: sex, mean/median age of total subjects, treatment regimen used, diagnostic criteria of ATDH, method of causality assessment, and covariates adjusted; and (c) adjusted ORs with 95% CIs under different genetic models reported by the original study or allele frequencies in ATDH cases and controls. The data extraction procedure was also implemented independently by the two parallel groups of reviewers.
Any disagreement was resolved by an additional reviewer. All data were directly taken from the included studies, and no further information was obtained by consulting the authors.

| Quality assessment
Study quality was assessed by the following revised criteria according to Little's recommendations 27 to determine potential bias and its effect on summary results. These criteria included seven items: (a) scientific design, (b) definite inclusion of study population, (c) explicit information on study population, (d) explicit diagnostic criteria on ATDH, (e) genetic detection method, (f) correct statistical analysis and (g) logical discussion of study bias. Each item can be judged as "yes" (low risk of bias) or "no" (high risk of bias). One point was awarded if an item was judged as low risk of bias. An overall quality scoring was generated, with a maximum score of 7 points, and a score > 4 was defined as a study of high quality. 28

| Study identification and characteristics
A total of 106 relevant articles were identified after an initial screening, and 5 case-control studies, which consisted of 572 ATDH cases and 1867 controls, were included. [20][21][22][23][24] The flow chart of the included and excluded studies is shown in Figure 1. There were 14 SNPs in the NR1I2 gene that have been reported previously (ie, rs3814055, rs7643645, rs13059232, rs2461823, rs3814057, rs6785049, rs12488820, rs1523127, rs2276707, rs2461825, rs2472677, rs2472682, rs3732357, and rs3732360), and the most heavily studied SNPs were rs3814055 and rs7643645. and the Chinese Society of Hepatology. 24 Only one study 24 performed a causality assessment using the updated Roussel Uclaf Causality Assessment Method (RUCAM). 30 The average quality score was 6.2, which demonstrated that the methodological quality was better.

| D ISCUSS I ON
The present study investigated the genetic association between the 14 polymorphisms of NR1I2 genes and susceptibility to ATDH.
Our study was the first to gather all the case-control studies per-

0.69
Abbreviations: 95%CI, 95% confidence interval; ATDH, anti-tuberculosis drug-induced hepatotoxicity; NR1I2, nuclear receptor subfamily 1 group I member 2; OR, Odds ratio; SNPs, single-nucleotide polymorphisms. CI: 1.052-2.502, P = .029). 24 The difference between those studies may be attributed to multiple factors, such as different diagnostic criteria of ATDH, the use of causality assessment, and the adjusted covariates ( Table 1). The meta-regression analysis showed that the diagnostic criteria of ATDH might be the most likely potential sources of heterogeneity between studies in dominant model analysis ( Of course, we must also note that the sample size included in the present meta-analysis study is relatively low. Although adjusted ORs and 95% CIs were used to evaluate the pooled effect, the covariates adjusted by different studies are not consistent. All of these factors may affect the results of this meta-analysis, and studies with larger sample sizes are required to assess the association between NR1I2 gene polymorphism and ATDH susceptibility more comprehensively.

F I G U R E 4 Forest plot of the relation between SNP rs7643645 (dominant model) and the risk of ATDH with the random effects model
In this study, for the first time, we gathered all published articles regarding NR1I2 genetic polymorphisms and the risk of ATDH and increased the sample size to achieve more accurate results. However, this meta-analysis had several limitations. First, the included studies were limited, and small sample sizes limit the power of analysis results.
Moreover, the limited number of studies hindered the adequate exploration of the source of heterogeneity by subgroup analysis. Second, all studies included in this meta-analysis were identified from selected databases, and publication bias may have occurred. Since fewer than nine studies were included, a publication bias test was not performed.
Finally, there is a lack of clarity on some of the issues regarding the uniform criteria used for the diagnosis of ATDH and causality assessment, which may have impacted the true summary effect.

| CON CLUS ION
Based on the present meta-analysis, we did not detect any association between NR1I2 gene polymorphisms and ATDH susceptibility. However, this conclusion should be interpreted with caution due to the low number of studies and the relatively small sample size. More investigation on the association between ATDH and NR1I2 gene polymorphisms is warranted to obtain a reliable conclusion.

E THI C S APPROVAL
Not required.

D I SCLOS U R E S
The authors declare that they have no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data generated and analyzed in the study are available from the corresponding author upon reasonable request.