Abstract
Background:
Hyperbilirubinemia is a predictor of severe drug-induced liver injury (DILI). Hepatobiliary ATP-binding cassette (ABC) transporters play an important role in the transportation of many drugs and bilirubin; however, little is known about these transporters and the risk of DILI. The aim of this study was to explore associations between genetic variations in important ABC transporters and susceptibility to DILI, with a particular focus on hyperbilirubinemia.
Methods:
A total of 200 patients with DILI and 200 healthy controls were enrolled as the training dataset. Another 106 patients with DILI were recruited as the validation dataset. They were genotyped for ABCB11 (BSEP) rs2287622, ABCB1 (MDR1) rs1128503, rs1045642, ABCB4 (MDR3) rs2230028, ABCC2 (MRP2) rs1885301, rs717620, rs2273697, rs3740066 and rs8187710 using polymerase chain reaction–based TaqMan genotyping assays.
Results:
There were no statistical differences in any of the nine ABC transporter single nucleotide polymorphisms between the DILI and control groups. However, in the DILI group, the patients with hyperbilirubinemia had a higher frequency of the ABCC2 rs717620 C/T and T/T genotypes than those without hyperbilirubinemia (44.2% vs 20.2%, p = 0.001). After adjusting for other confounding factors, the ABCC2 rs717620 T variant was still associated with an increased risk of hyperbilirubinemia (adjusted odds ratio [OR]: 3.83, 95% confidence interval [CI]: 1.73-8.48, p = 0.001). This association was confirmed by the validation dataset (adjusted OR: 3.92, 95% CI: 1.42-10.81, p = 0.015). We also found that the mortality group had higher frequencies of the ABCC2 (MRP2) rs717620 C/T and T/T genotypes than the survival group (50.0% vs 27.9%, p = 0.048).
Conclusion:
Carriage of the ABCC2 (MRP2) rs717620 T variant may increase the risk of hyperbilirubinemia and mortality in patients with DILI. Screening for this variant may help to prevent and mitigate drug-induced hyperbilirubinemia.
Keywords: ATP-binding cassette (ABC) transporters, ABCC2, Drug-induced liver injury, Hyperbilirubinemia, MRP2
1. INTRODUCTION
Drug-induced liver injury (DILI) is an important and challenging liver disease. Most cases of DILI ultimately resolve completely without residual liver injury; however, some cases may progress to acute liver failure (ALF), chronic liver injury, or cirrhosis, which may result in death or the need for liver transplantation.1
Hy’s law is a method used to predict the potential of a drug to cause a severe DILI and states that the occurrence of hyperbilirubinemia is an indicator of a poor outcome of DILI.2 Therefore, understanding the pathogenesis and risk factors of hyperbilirubinemia is crucial in the prevention of a serious DILI. Hyperbilirubinemia occurs when there is a disturbance in the homeostasis of bilirubin and bile acid caused by proteins involved in the synthesis, uptake, detoxification, and transport of bile acid and bilirubin. The superfamily of hepatobiliary ATP-binding cassette (ABC) transporters plays a crucial role in the disposition of bile acid, bilirubin and many drugs.3–9 Among them, bile salt export pump (BSEP, ABCB11), multidrug resistance 1 (MDR1, ABCB1), MDR3 (ABCB4), and multidrug resistance–associated protein 2 (MRP2, ABCC2) are the most important efflux transporters. Hereditary genetic variations of these ABC transporters may regulate transporter expression and may be related to many hereditary and acquired cholestatic liver diseases.3–9
A dysfunctional transporter gene may also determine an individual’s susceptibility to the occurrence of DILI.3–9 Recent studies have explored associations between these ABC genetic variations and DILI; however, the results have been inconsistent and inconclusive.10–25 The inclusion criteria for DILI, offending drugs, and ethnic populations were different in these studies, which may have affected the results. Therefore, the true relationship between genetic polymorphisms of ABC transporters and DILI remains controversial. The purpose of this study was to evaluate the influence of genetic variations in four major ABC transporters, ABCB11, ABCB1, ABCB4, and ABCC2, on the risk of DILI in Taiwanese patients, with a particular focus on hyperbilirubinemic DILI.
2. METHODS
2.1. Patients studied
A total of 200 patients with DILI were enrolled in the DILI case group, and 200 healthy adults in the control group. To verify associations between hyperbilirubinemic DILI and genetic variations of ABC transporters, a validation dataset including another 106 cases with DILI was also enrolled.
The inclusion criteria for the DILI cases were as follows: (1) an increase in serum alanine aminotransferase (ALT) or aspartate aminotransferase (AST) level greater than five times the upper limit of normal value (ULN), or an elevation in serum alkaline phosphatase (ALP) greater than twice the ULN during treatment with the drug, as suggested by the Drug-induced Liver Injury Network (DILIN)26; (2) serum ALT or AST level greater than five times the baseline value, or ALP level greater than twice the baseline level, in cases with elevated based line levels; and (3) any elevation of serum ALT, AST, or ALP associated with an increase in the level of serum total bilirubin (>2.5 mg/dL), in the absence of a prior diagnosis of liver disease, Gilbert’s syndrome, or evidence of hemolysis. The exclusion criteria for the DILI cases were as follows: (1) patients with possible acute or chronic viral hepatitis, such as positive serum hepatitis B virus surface antigen, IgM antibody to hepatitis B core antigen, IgM antibody to hepatitis A virus, antibody to hepatitis C virus, IgM antibody to Epstein-Barr virus, cytomegalovirus, and herpes simplex virus; (2) autoimmune liver diseases, such as autoimmune hepatitis, primary biliary cholangitis, and primary sclerosing cholangitis; (3) any other major hepatic or systemic diseases that may induce elevation of liver biochemical tests, such as alcoholic liver disease, nonalcoholic fatty liver disease, hepato-biliary stones or tumors, congestive heart failure, hypoxia, and bacteremia; (4) a RUCAM causality assessment score <5, indicating patients with the least possibility of having DILI;27 and (5) patients with incomplete clinical or laboratory data.
The inclusion criterion for the controls was adults with normal serum levels of ALT, AST, ALP, and bilirubin on enrollment. The exclusion criteria for the controls were as follows: (1) chronic hepatitis B or C infection; (2) autoimmune liver diseases, such as autoimmune hepatitis, primary biliary cholangitis, and primary sclerosing cholangitis; and (3) any other major hepatic or systemic diseases that may affect liver biochemical tests, such as alcoholic liver disease, non-alcoholic fatty liver disease, hepato-biliary stones or tumors, congestive heart failure, hypoxia, and bacteremia.
The types of DILI were classified according to the DILIN guidelines.26 This is characterized based upon the relative elevations of serum ALT and ALP (R ratio of ALT to ALP, both expressed in multiples of the ULN) into hepatocellular (R ≥ 5), mixed (R = 2-5), or cholestatic (R ≤ 2).
Hyperbilirubinemia was defined as a serum total bilirubin level >2.5 mg/dL as suggested by the DILIN.26 ALF was defined as prolongation of prothrombin time, international normalized ratio >1.5 or presence of any degree of hepatic encephalopathy according to the European Association for the Study of the Liver.28
The study protocol was approved by the Institutional Review Board of Taipei Veterans General Hospital (No. 2017-01-007B).
2.2. Determination of ABC genotypes
Venous blood was obtained from each subject, and DNA was extracted and deposited in the Biobank of Taipei Veterans General Hospital. Samples and controls were genotyped for the ABCB11 (BSEP) c.1331T>C polymorphism (rs2287622), ABCB1 (MDR1) c.1236T>C (rs1128503), c.3435T>C (rs1045642), ABCB4 (MDR3) c.1954A>G (rs2230028) and ABCC2 (MRP2) c.-1549A>G (rs1885301), c.-24C>T (rs717620), c.1249G>A (rs2273697), c.3972C>T (rs3740066), and c.4544G>A (rs8187710). These candidate single nucleotide polymorphisms (SNPs) were selected from previous publications which disclosed associations with DILI and hereditary cholestatic liver diseases.3–25 Genotyping was performed using TaqMan genotyping assays (Thermo Fisher Scientific Inc., Waltham, MA, USA). TaqMan polymerase chain reactions (PCR) were performed according to the manufacturer’s standard protocol. Briefly, 20 ng genomic DNA was mixed with the supplied 2× TaqMan Universal PCR Master Mix No AmpErase UNG and 20× TaqMan Assay Mix to a final volume of 5 µL in a 384-well plate. Each sample underwent 40 amplification cycles on a GeneAmp PCR System 9700 (Thermo Fisher Scientific Inc.). Fluorescent signals of the two probes were analyzed using an ABI PRISM® 7900HT Sequence Detection System (Thermo Fisher Scientific Inc.). The genotypes were determined automatically using Sequence Detection Software (Thermo Fisher Scientific Inc.).
2.3. Statistical analysis
Expected gene frequencies were calculated from respective single allele frequencies according to the Hardy-Weinberg equation using the chi-square goodness-of-fit test. Differences in genotype distributions were compared between the DILI cases and controls, three DILI types, and cases with or without hyperbilirubinemia using the chi-square test with or without Yates’ correction. Bonferroni correction was applied for multiple comparisons. Means were compared using the Student’s t test or ANOVA. Odds ratios (ORs) and confidence intervals (CIs) were estimated using logistic regression analysis. All analyses were performed using SPSS version 17.0 (SPSS Inc, Chicago, IL, USA), and a p < 0.05 was considered to be statistically significant.
3. RESULTS
There were no statistical differences in age and gender among the controls, DILI groups in the training dataset, and DILI groups in the validation dataset (Table 1). There were also no statistical differences in the frequency of three DILI types, ALF, and mortality rates between the DILI training dataset and validation dataset. A total of 86 (43%) patients in the DILI training dataset and 21 (19.8%) patients in the DILI validation dataset had hyperbilirubinemia. Antituberculosis drugs (ATDs) were the most common type of drug that induced liver injury in both the DILI training and validation datasets, which included the standard four combination therapy of isoniazid, rifampicin, pyrazinamide, and ethambutol (Table 2).
Table 1.
Basic characteristics of controls, DILI patients in training dataset and DILI patients in validation dataset
| Controls (n = 200) | DILI training (n = 200) | DILI validation (n = 106) | p | |
|---|---|---|---|---|
| Age, y | 56.9 ± 12.6 | 59.2 ± 12.4 | 58.6 ± 12.7 | 0.179 |
| Gender (F/M) | 95/105 | 91/109 | 49/57 | 0.922 |
| Peak ALT, U/L | 30.4 ± 7.7 | 779.3 ± 967.1 | 530.8 ± 497.9 | <0.001 |
| Peak AST, U/L | 33.8 ± 6.5 | 784.7 ± 945.6 | 539.3 ± 523.6 | <0.001 |
| Peak total bilirubin, mg/dL | 1.2 ± 0.3 | 4.1 ± 5.5 | 3.0 ± 4.0 | <0.001 |
| Peak ALP, U/L | 80.4 ± 14.8 | 157.4 ± 155.0 | 123.2 ± 91.4 | <0.001 |
| Albumin, g/dL | 5.0 ± 0.4 | 4.2 ± 0.8 | 4.7 ± 0.7 | <0.001 |
| Creatinine, mg/dL | 1.0 ± 0.2 | 1.6 ± 0.6 | 1.2 ± 0.6 | <0.001 |
| Platelet (103/µL) | 289.4 ± 62.8 | 208.9 ± 88.7 | 212.5 ± 64.8 | <0.001 |
| DILI type (H/M/C) | ... | 158/25/17 | 85/13/8 | 0.990c |
| Hyperbilirubinemiaa | 0 | 86 (43.0%) | 21 (19.8%) | <0.001c |
| ALFb | 0 | 37 (18.5%) | 14 (13.2%) | 0.237c |
| Mortality | 0 | 24 (12%) | 11 (10.4%) | 0.813c |
Data were expressed as mean ± SD or numbers. Reference value: ALT: 0-40 U/L, AST: 5-45 U/L, total bilirubin: 0.2-1.6 mg/dL, ALP: 10-100 U/L, albumin: 3.5-5.5 g/dL, platelet: 150-400 × 103/µL, international normalized ratio: 0.8-1.2.
ALF = acute liver failure; ALP = alkaline phosphatase; ALT = alanine aminotransferase; AST = aspartate aminotransferase; C = cholestatic type; DILI = drug-induced liver injury; H= hepatocellular type; M = mixed type.
aHyperbilirubinemia was defined as serum total bilirubin >2.5mg/dL.
bALF was defined as prolongation of prothrombin time, international normalized ratio > 1.5, or presence of any degree of hepatic encephalopathy.
cCompared between DILI training dataset and validation dataset 2 groups.
Table 2.
Category of incriminated drugs for drug-induced liver injury in this study
| Category of drugs | Training dataset (n = 200) | Validation dataset (n = 106) |
|---|---|---|
| Anti-tuberculosis drugs | 100 (50.0%) | 46 (43.4%) |
| Nonsteroidal anti-inflammatory drugs | 18 (9.0%) | 11 (10.4%) |
| Statins | 16 (8.0%) | 7 (6.6%) |
| Antibacterial drugs | 15 (7.5%) | 9 (8.5%) |
| Anticonvulsant drugs | 11 (10.5%) | 8 (7.6%) |
| Hypouricemic drugs | 9 (4.5%) | 5 (4.7%) |
| Antifungal drugs | 7 (3.5%) | 2 (1.9%) |
| Antineoplastic drugs | 6 (3.0%) | 5 (4.7%) |
| Antiarrhythmic drugs | 5 (2.5%) | 2 (1.9%) |
| Hormone-related drugs | 5 (2.5%) | 4 (3.7%) |
| Antipsychotic drugs | 4 (2.0%) | 2 (1.9%) |
| Others | 4 (2.0%) | 5 (4.7%) |
All of the ABC transporter genotypes were in Hardy-Weinberg equilibrium. There were no statistical differences in any of the nine SNPs of ABC transporters between the DILI training dataset and control group (Table 3). Furthermore, there were no statistical differences in these SNPs among the hepatocellular, mixed, and cholestatic types of DILI (Table 4). However, the patients with hepatocellular DILI had higher frequencies of hyperbilirubinemia and ALF compared with the patients with the other types.
Table 3.
The distribution of genetic variations of ABC transporters in DILI cases and controls based on training dataset
| DILI (n =200) | Control (n =200) | p | |
|---|---|---|---|
| ABCB11 | |||
| rs2287622 | |||
| CC/CT/TT | 103/80/17 | 105/78/17 | 0.978 |
| ABCB1 | |||
| rs1128503 | |||
| CC/TC/TT | 27/87/86 | 26/87/87 | 0.988 |
| rs1045642 | |||
| CC/TC/TT | 74/92/34 | 72/90/31 | 0.851 |
| ABCB4 | |||
| rs2230028 | |||
| AA/GA/GG | 185/15/0 | 189/11/0 | 0.417 |
| ABCC2 | |||
| rs1885301 | |||
| AA/AG/GG | 9/55/136 | 16/70/114 | 0.058 |
| rs717620 | |||
| CC/CT/TT | 139/55/6 | 117/72/11 | 0.060 |
| rs2273697 | |||
| AA/AG/GG | 3/38/159 | 2/36/162 | 0.863 |
| rs3740066 | |||
| CC/CT/TT | 133/58/9 | 113/72/15 | 0.099 |
| rs8187710 | |||
| GG/GA/AA | 199/1/0 | 200/0/0 | 0.317 |
No statistical difference between 2 groups in all genetic variations.
ABC = ATP-binding cassette; DILI = drug-induced liver injury.
Table 4.
The difference of genetic variations of ABC transporters, hyperbilirubinemia, acute liver failure, and mortality among three types of drug-induced liver injury based on training dataset
| Hepatocellular (n = 158) | Mixed (n = 25) | Cholestatic (n = 17) | p | |
|---|---|---|---|---|
| ABCB11 | ||||
| rs2287622 | 83/60/15 | 13/11/1 | 7/9/1 | 0.685 |
| CC/CT/TT | (52.5/38.0/9.5) | (52.0/44.0/4.0) | (41.2/52.9/5.9) | |
| ABCB1 | ||||
| rs1128503 | 22/69/67 | 2/12/11 | 3/6/8 | 0.868 |
| CC/TC/TT | (13.9/43.7/42.4) | (8.0/48.0/44.0) | (17.6/35.3/47.1) | |
| rs1045642 | 58/73/27 | 8/14/3 | 8/5/4 | 0.559 |
| CC/TC/TT | (36.7/46.2/17.1) | (32.0/56.0/12.0) | (47.1/29.4/23.5) | |
| ABCB4 | ||||
| rs2230028 | 147/11//0 | 22/3/0 | 16/1/0 | 0.651 |
| AA/GA/GG | (93.0/7.0/0.0) | (88.0/12.0/0.0) | (94.1/5.9/0.0) | |
| ABCC2 | ||||
| rs1885301 | 8/43/107 | 1/6/18 | 0/6/11 | 0.832 |
| AA/AG/GG | (5.1/27.2/67.7) | (4.0/24.0/72.0) | (0.0/35.3/64.7) | |
| rs717620 | 109/43/6 | 18/7/0 | 12/5/0 | 0.799 |
| CC/CT/TT | (69.0/27.2/3.8) | (72.0/28.0/0.0) | (70.6/29.4/0.0) | |
| rs2273697 | 2/31/125 | 1/4/20 | 0/3/14 | 0.819 |
| AA/AG/GG | (1.3/19.6/79.1) | (4.0/16.0/80.0) | (0.0/17.6/82.4) | |
| rs3740066 | 105/46/7 | 16/7/2 | 12/5/0 | 0.882 |
| CC/CT/TT | (66.5/29.1/4.4) | (64.0/28.0/8.0) | (70.6/29.4/0.0) | |
| rs8187710 | 158/0/0 | 25/0/0 | 16/1/0 | 0.083 |
| GG/GA/AA | (100.0/0.0/0.0) | (100.0/0.0/0.0) | (94.1/5.9/0.0) | |
| Hyperbilirubinemia | 78 (49.4) | 8 (32.0) | 0 (0.0) | <0.001* |
| Acute liver failure | 35 (22.2) | 2 (8.0) | 0 (0.0) | 0.029* |
| Mortality | 23 (14.6) | 1 (4.0) | 0 (0.0) | 0.090 |
ABC = ATP-binding cassette.
*p < 0.05.
The patients with hyperbilirubinemia had higher frequencies of ABCC2 (MRP2) rs717620 C/T and T/T genotypes than those without hyperbilirubinemia (44.2% vs 20.2%, p = 0.001, Table 5), and the difference was still statistically significant after Bonferroni correction (significance was defined as p < 0.006). There were no statistical differences in any of the genotypes between the patients with and without ALF (Table 6). However, the mortality group had marginally higher frequencies of ABCC2 (MRP2) rs717620 C/T and T/T genotypes than the survival group (50.0% vs 27.9%, p = 0.048, Table 7). In the patients with ATD-induced liver injury, there were no statistical differences in the frequency of hyperbilirubinemia in any of the genotypes. Furthermore, if the DILI group was separated into subgroups of ABCC2 rs717620 with or without the T variant, there were no statistical differences in ALT and AST levels between these two subgroups (814.8 ± 988.6 vs 763.7 ± 960.7 U/L, p = 0.735 and 742.1 ± 830.3 vs 803.4 ± 994.3 U/L, p = 0.652, respectively).
Table 5.
The distribution of genetic variations of ABC transporters in drug-induced hyperbilirubinemia based on training dataset
| Hyperbilirubinemiaa (n = 86) | Non-hyberbilirubinemia (n = 114) | p | |
|---|---|---|---|
| ABCB11 | |||
| rs2287622 | 47/30/9 | 56/50/8 | 0.375 |
| CC/CT/TT | (54.7/34.9/10.4) | (49.1/43.9/7.0) | |
| ABCB1 | |||
| rs1128503 | 7/38/41 | 20/49/45 | 0.136 |
| CC/TC/TT | (8.1/44.2/47.7) | (17.5/43.0/39.5) | |
| rs1045642 | 34/39/13 | 40/53/21 | 0.744 |
| CC/TC/TT | (39.5/45.3/15.2) | (35.1/46.5/18.4) | |
| ABCB4 | |||
| rs2230028 | 77/9/0 | 108/6/0 | 0.167 |
| AA/GA/GG | (89.5/10.5/0) | (94.7/5.3/0) | |
| ABCC2 | |||
| rs1885301 | 7/25/54 | 2/30/82 | 0.075 |
| AA/AG/GG | (8.1/29.1/62.8) | (1.8/26.3/71.9) | |
| rs717620 | 48/34/4 | 91/21/2 | 0.001* |
| CC/CT/TT | (55.8/39.5/4.7) | (79.8/18.4/1.8) | |
| rs2273697 | 2/18/66 | 1/20/93 | 0.570 |
| AA/AG/GG | (2.3/20.9/76.8) | (0.9/17.5/81.6) | |
| rs3740066 | 50/30/6 | 83/28/3 | 0.066 |
| CC/CT/TT | (58.1/34.9/7.0) | (72.8/24.6/2.6) | |
| rs8187710 | 86/0/0 | 113/1/0 | 0.384 |
| GG/GA/AA | (100/0/0) | (99.1/0.9/0) | |
Data were expressed as number (%).
ABC = ATP-binding cassette.
*p < 0.05.
aHyperbilirubinemia was defined as serum total bilirubin >2.5 mg/dL.
Table 6.
The distribution of genetic variations of ABC transporters in drug-induced ALF based on training dataset
| ALFa(n = 37) | Non-ALF(n = 163) | p | |
|---|---|---|---|
| ABCB11 | |||
| rs2287622 | 22/10/5 | 81/70/12 | 0.149 |
| CC/CT/TT | (59.5/27.0/13.5) | (49.7/42.9/7.4) | |
| ABCB1 | |||
| rs1128503 | 3/19/15 | 24/68/71 | 0.431 |
| CC/TC/TT | (8.1/41.7/43.6) | (14.7/43.0/39.5) | |
| rs1045642 | 9/23/5 | 65/69/29 | 0.087 |
| CC/TC/TT | (24.3/62.2/13.5) | (39.9/42.3/17.8) | |
| ABCB4 | |||
| rs2230028 | 34/3/0 | 151/12/0 | 0.550 |
| AA/GA/GG | (91.9/8.1/0.0) | (92.6/7.4/0.0) | |
| ABCC2 | |||
| rs1885301 | 2/9/26 | 7/46/110 | 0.868 |
| AA/AG/GG | (5.4/24.3/70.3) | (4.3/28.2/67.5) | |
| rs717620 | 21/14/2 | 118/41/4 | 0.157 |
| CC/CT/TT | (56.8/37.8/5.4) | (72.4/25.2/2.5) | |
| rs2273697 | 0/7/30 | 3/31/129 | 0.706 |
| AA/AG/GG | (0.0/18.9/81.1) | (1.8/19.0/79.1) | |
| rs3740066 | 26/9/2 | 107/49/7 | 0.770 |
| CC/CT/TT | (70.3/24.3/5.4) | (65.6/30.1/4.3) | |
| rs8187710 | 37/0/0 | 162/1/0 | 0.815 |
| GG/GA/AA | (100/0.0/0.0) | (99.4/0.6/0.0) | |
Data were expressed as number (%). No statistical difference between 2 groups in all genetic variations.
ABC = ATP-binding cassette; ALF = acute liver failure.
Table 7.
The distribution of genetic variations of ABC transporters in drug-induced mortality based on training dataset
| Mortality (n = 24) | Survival (n = 176) | p | |
|---|---|---|---|
| ABCB11 | |||
| rs2287622 | 12/8/4 | 91/72/13 | 0.293 |
| CC/CT/TT | (50.0/33.3/16.7) | (51.7/40.9/7.4) | |
| ABCB1 | |||
| rs1128503 | 2/13/9 | 25/74/77 | 0.486 |
| CC/TC/TT | (8.3/54.2/37.5) | (14.2/42.0/43.8) | |
| rs1045642 | 7/15/2 | 67/77/32 | 0.195 |
| CC/TC/TT | (29.2/62.5/8.3) | (38.1/43.8/18.2) | |
| ABCB4 | |||
| rs2230028 | 22/2/0 | 163/13/0 | 0.697 |
| AA/GA/GG | (91.7/8.3/0.0) | (92.6/7.4/0.0) | |
| ABCC2 | |||
| rs1885301 | 1/5/18 | 8/50/118 | 0.725 |
| AA/AG/GG | (4.2/20.8/75.0) | (4.5/28.4/67.0) | |
| rs717620 | 12/10/2 | 127/45/4 | 0.048* |
| CC/CT/TT | (50.0/41.7/8.3) | (72.2/25.6/2.3) | |
| rs2273697 | 0/4/20 | 3/34/139 | 0.764 |
| AA/AG/GG | (0.0/16.7/83.3) | (1.7/19.3/79.0) | |
| rs3740066 | 17/6/1 | 116/52/8 | 0.889 |
| CC/CT/TT | (70.8/25.0/4.2) | (66.0/29.5/4.5) | |
| rs8187710 | 24/0/0 | 175/1/0 | 1.000 |
| GG/GA/AA | (100/0.0/0.0) | (99.4/0.6/0.0) | |
Data were expressed as number (%).
*p < 0.05.
ABC = ATP-binding cassette
In univariate analysis, ABCC2 rs717620 T variant, ALT, AST, ALP, and albumin were found to be associated with the risk of hyperbilirubinemia in the DILI training dataset. In multivariable analysis, ABCC2 T variant was the most significant factor after adjusting for ALT, AST, ALP, albumin, age, and gender (adjusted OR: 3.83, 95% CI: 1.73-8.48, p = 0.001, Table 8).
Table 8.
Logistic analysis of ABCC2 rs717620 T variant and the risk of hyperbilirubinemia in drug-induced liver injury
| Odds ratio | 95% confidence interval | p | |
|---|---|---|---|
| Training dataset (n = 200) | |||
| Univariate analysis | |||
| ABCC2 rs717620 T variant (n = 61) | 3.13 | 1.68-5.86 | 0.001* |
| Peak ALT (n = 200) | 1.01 | 1.00-1.01 | 0.001* |
| Peak AST (n = 200) | 1.01 | 1.00-1.01 | 0.001* |
| Peak ALP (n = 200) | 0.99 | 0.99-1.00 | 0.012* |
| Albumin (n = 200) | 0.18 | 0.11-0.31 | 0.001* |
| Age (n = 200) | 1.02 | 0.99-1.05 | 0.084 |
| Gender (male = 109) | 0.62 | 0.35-1.08 | 0.093 |
| Multivariate analysis | |||
| ABCC2 rs717620 T variant (n = 61) | 3.83 | 1.73-8.48 | 0.001* |
| Peak ALT (n = 200) | 1.00 | 0.99-1.00 | 0.904 |
| Peak AST (n = 200) | 1.01 | 1.00-1.01 | 0.186 |
| Peak ALP (n = 200) | 0.99 | 0.99-1.00 | 0.006* |
| Albumin (n = 200) | 0.21 | 0.12-0.37 | 0.001* |
| Age (n = 200) | 1.03 | 0.99-1.06 | 0.090 |
| Gender (male = 109) | 0.57 | 0.27-1.19 | 0.135 |
| Validation dataset (n=106) | |||
| Univariate analysis | |||
| ABCC2 rs717620 T variant (n = 26) | 3.92 | 1.42-10.81 | 0.001* |
| Peak ALT (n = 106) | 1.01 | 1.00-1.01 | 0.001* |
| Peak AST (n = 106) | 1.01 | 1.00-1.01 | 0.008* |
| Peak ALP (n = 106) | 0.99 | 0.99-1.01 | 0.813 |
| Albumin (n = 106) | 0.22 | 0.10-0.46 | 0.001* |
| Age (n = 106) | 1.02 | 0.98-1.06 | 0.335 |
| Gender (male = 57) | 0.58 | 0.22-1.52 | 0.266 |
| Multivariate analysis | |||
| ABCC2 rs717620 T variant (n = 26) | 4.91 | 1.36-17.73 | 0.015* |
| Peak ALT (n = 106) | 1.00 | 0.99-1.00 | 0.302 |
| Peak AST (n = 106) | 1.00 | 1.00-1.01 | 0.681 |
| Peak ALP (n = 106) | 0.99 | 0.99-1.00 | 0.440 |
| Albumin (n = 106) | 0.27 | 0.12-0.62 | 0.002* |
| Age (n = 106) | 1.05 | 0.99-1.11 | 0.083 |
| Gender (male = 57) | 0.66 | 0.20-2.23 | 0.506 |
*p < 0.05.
In the validation dataset of another 106 cases with DILI, the patients with the ABCC2 rs717620 T variant also had a higher risk of hyperbilirubinemia (adjusted OR: 4.91, 95% CI: 1.36-17.73, p = 0.015, Table 8). In addition, serum albumin level was associated with hyperbilirubinemia in the univariate and multivariable analyses of both datasets. Although serum ALT, AST, and ALP levels were associated with the risk of hyperbilirubinemia in the univariate analysis, there was no statistical significance in the multivariable analysis.
With regard to mortality, the ABCC2 rs717620 T variant was associated with mortality in the univariate analysis (Table 9), although this association was not found in the multivariable analysis after adjustments for other variables. However, serum albumin level was associated with mortality in the univariate and multivariable analyses of both datasets.
Table 9.
Logistic analysis of ABCC2 rs717620 T variant and the association of drug-related mortality in drug-induced liver injury
| Odds ratio | 95% confidence interval | p | |
|---|---|---|---|
| Training dataset (n =200) | |||
| Univariate analysis | |||
| ABCC2 rs717620 T variant (n = 61) | 2.59 | 1.09-6.16 | 0.031* |
| Peak ALT (n = 200) | 1.01 | 1.00-1.01 | 0.003* |
| Peak AST (n = 200) | 1.00 | 1.00-1.01 | 0.007* |
| Peak ALP (n = 200) | 0.99 | 0.99-1.01 | 0.362 |
| Albumin (n = 200) | 0.23 | 0.12-0.44 | 0.001* |
| Age (n = 200) | 1.00 | 0.97-1.04 | 0.956 |
| Gender (male = 109) | 0.46 | 0.19-1.10 | 0.080 |
| Multivariate analysis | |||
| ABCC2 rs717620 T variant (n = 61) | 1.88 | 0.68-5.19 | 0.224 |
| Peak ALT (n = 200) | 1.00 | 0.99-1.01 | 0.392 |
| Peak AST (n = 200) | 1.00 | 0.99-1.01 | 0.934 |
| Peak ALP (n = 200) | 0.99 | 0.98-1.01 | 0.340 |
| Albumin (n = 200) | 0.27 | 0.13-0.54 | 0.001* |
| Age (n = 200) | 0.99 | 0.96-1.03 | 0.722 |
| Gender (male = 109) | 0.52 | 0.20-1.38 | 0.189 |
| Validation dataset (n=106) | |||
| Univariate analysis | |||
| ABCC2 rs717620 T variant (n = 26) | 0.66 | 0.13-3.26 | 0.607 |
| Peak ALT (n = 106) | 1.01 | 1.00-1.01 | 0.001* |
| Peak AST (n = 106) | 1.01 | 1.00-1.01 | 0.001* |
| Peak ALP (n = 106) | 0.99 | 0.99-1.01 | 0.702 |
| Albumin (n = 106) | 0.17 | 0.07-0.41 | 0.001* |
| Age (n = 106) | 1.03 | 0.98-1.09 | 0.301 |
| Gender (male = 57) | 0.45 | 0.12-1.65 | 0.230 |
| Multivariate analysis | |||
| ABCC2 rs717620 T variant (n = 26) | 0.41 | 1.36-17.73 | 0.442 |
| Peak ALT (n = 106) | 1.00 | 0.99-1.01 | 0.183 |
| Peak AST (n = 106) | 1.00 | 0.99-1.01 | 0.717 |
| Peak ALP (n = 106) | 0.99 | 0.99-1.01 | 0.214 |
| Albumin (n = 106) | 0.14 | 0.04-0.49 | 0.002* |
| Age (n = 106) | 1.05 | 0.98-1.13 | 0.188 |
| Gender (male = 57) | 0.50 | 0.08-3.17 | 0.460 |
*p < 0.05.
4. DISCUSSION
Hepatic transporters may influence the disposition of drugs and are associated with the development of DILI. A dysfunctional transporter gene may also determine an individual’s susceptibility to the occurrence of severe liver injury. In this study, we showed that ABC transporter genetic variations may play a role in DILI, based on the finding that the ABCC2 (MRP2) rs717620 T variant was associated with an increased risk of hyperbilirubinemia and mortality in patients with DILI.
ABC transporters are a superfamily of enzymes that are responsible for the uptake and transport of many xenobiotics, lipids, bile acids, and bilirubin.3–9 Among them, ABCB11 (BSEP) is responsible for the transport of taurocholate and other cholate conjugates from hepatocytes to the bile canaliculi. In humans, the activity of this transporter is the major determinant of bile formation and bile flow. Mutations of the encoded ABCB11 gene have been associated with progressive familial intrahepatic cholestasis (PFIC) type 2. The major genetic polymorphism of ABCB11 is c.1331T>C, V444A (rs2287622), which has been associated with lower BSEP expression levels.11,14–16,18–22,24,25 In Switzerland, Lang et al11 first demonstrated an association between the rs2287622 genetic polymorphism of ABCB11 and drug-induced cholestasis. In addition, Meier et al,14 also from Switzerland, reported that this variant could increase the susceptibility to contraceptive-induced cholestasis, and Ulzurrun et al23 from Spain reported that this genetic polymorphism could increase the risk of hepatocellular DILI. However, in a study conducted in France, Roustit et al20 did not find an association between ABCB11 genetic variation and bosentan-induced liver toxicity. Similarly, Kagawa et al24 did not find an association between this variant and drug-induced cholestasis in a study conducted in Japan. In addition, Chen et al25 did not find an association between this genetic polymorphism and ATD-induced liver injury in a study conducted in China. Consistent with these studies, we did not find an association between this variant and the susceptibility to DILI including hepatocellular and mixed and cholestatic types in the present study.
ABCB1 (MDR1) is an important cell membrane protein that pumps many foreign substances out of cells.3–9 The major genetic polymorphisms of ABCB1 include c.1236T>C (rs1128503) and c.3435T>C (rs1045642). In the present study, we did not find associations between these genetic variants and the risk of DILI, which is consistent with previous Korean and Spanish studies.17,23
ABCB4 (MDR3) is also an important export enzyme associated with PFIC type 3.3–9 The major genetic polymorphism of ABCB4 is c.1954A>G (rs2230028), and Alfirevic et al12 reported that this genetic variant was not associated with tacrine-induced liver damage in a study conducted in the UK. Ulzurrun et al23 also reported that ABCB4 genetic polymorphisms did not increase the risk of DILI in a study conducted in Spain. Likewise, we found no association between this variant and the susceptibility to DILI in the present study.
ABCC2 (MRP2) is expressed in the canalicular part of hepatocytes and functions in biliary transport.3–9 Several genetic variations of this gene may affect its activity, including c.-1549A>G (rs1885301), c.-24C>T (rs717620), c.1249G>A (rs2273697), c.3972C>T (rs3740066), and c.4544G>A (rs8187710). Some of these variants have been associated with Dubin-Johnson syndrome.3–9 In a study conducted in the UK, Daly et al10 showed that the ABCC2 rs717620 polymorphism was associated with the susceptibility to diclofenac-induced hepatotoxicity. Choi et al13 also reported that the rs717620 polymorphism of this gene was related to the risk of DILI in a study conducted in Korea. However, Kim et al17 and Ulzurrun et al,23 from Korea and Spain, respectively, disproved relationships between genetic variations of this gene and DILI. In the present study, we did not find an association with any of the five major SNPs of this gene. However, although ABCC2 rs1885301 and rs717620 did not significantly differ between the DILI and control groups in the present study, there seemed to be trends of associations (p = 0.058 and 0.060, respectively, Table 3). A further large-scale study is needed to elucidate the associations between ABCC2 genotypes and susceptibility to DILI.
Since patients with DILI and hyperbilirubinemia are associated with a high mortality rate and are a serious concern for the pharmaceutical industry and healthcare providers,2 we intended to explore the association between ABC transporter genetic polymorphisms and the risk of hyperbilirubinemia in this study. We found that the ABCC2 rs717620 T variant increased the risk of hyperbilirubinemia in the patients with DILI, and this association was verified by our validation dataset. This finding may be because the rs717620 T variant is associated with lower ABCC2 (MRP2) activity and interferes with the disposition and excretion of drug metabolites and bilirubin. Choi et al13 demonstrated that the ABCC2 rs717620 T variant reduced MRP2 promoter activity by approximately 40%.
The association between the ABCC2 rs717620 T variant and drug-related mortality in this study was tenuous and was demonstrated in chi-square and univariate analyses, but not in the multivariable analysis. This may be due to the limited number of cases of mortality in this study.
Previous studies have reported associations between age and gender with the risk of DILI.29–31 In the present study, multivariable logistic regression was used to lessen the impact of these confounding factors, and the association between the ABCC2 rs717620 genetic variation and hyperbilirubinemia in DILI remained after these adjustments.
Although the results of this study should be confirmed by other studies with a larger sample size and more diverse ethnic population, it still proposes a potential pharmacogenetic approach for the surveillance and prevention of severe DILIs. To mitigate serious cases of DILI, avoiding the co-administration of drugs that are substrates of ABCC2, regularly and closely monitoring liver function, and decreasing the dose of drugs may be considered.
The main limitation of this study is that most of our DILI cases had the hepatocellular type, especially in the patients with ATD-induced liver injury, which may have affected the significant association between transporter genetic variants and the cholestatic type of liver injury.
In conclusion, in the present study, we showed that people with the ABCC2 (MRP2) rs717620 T variant may have a higher risk of hyperbilirubinemia and mortality if they have DILI. Screening for this important ABC transporter variant may be beneficial in the prevention and mitigation of severe DILI.
ACKNOWLEDGMENTS
This study was supported by the grant from the Ministry of Science and Technology, Taiwan (MOST 106-2314-B-075-054) and the grand from Taipei Veterans General Hospital (V109C-180). All samples for this study were obtained from the Biobank of Taipei Veterans General Hospital. The authors would like to acknowledge the support by the Biobank of Taipei Veterans General Hospital.
Footnotes
Conflicts of interest: The authors declare that they have no conflicts of interest related to the subject matter or materials discussed in this article.
References
- 1.Chitturi S, Farrell GC. Schiff ER, Maddrey WC, Sorrell MF, editors. Drug-induced liver disease. In: Schiff’s diseases of the liver 2012, p. 11th ed. Hoboken: Wiley-Blackwell; 703–82. [Google Scholar]
- 2.Robles-Diaz M, Lucena MI, Kaplowitz N, Stephens C, Medina-Cáliz I, González-Jimenez A, et al. ; Spanish DILI Registry; SLatinDILI Network; Safer and Faster Evidence-based Translation Consortium. Use of Hy’s law and a new composite algorithm to predict acute liver failure in patients with drug-induced liver injury. Gastroenterology 2014;147:109–18.e5. [DOI] [PubMed] [Google Scholar]
- 3.Pauli-Magnus C, Meier PJ. Hepatobiliary transporters and drug-induced cholestasis. Hepatology 2006;44:778–87. [DOI] [PubMed] [Google Scholar]
- 4.Stieger B. Role of the bile salt export pump, BSEP, in acquired forms of cholestasis. Drug Metab Rev 2010;42:437–45. [DOI] [PubMed] [Google Scholar]
- 5.Stieger B, Geier A. Genetic variations of bile salt transporters as predisposing factors for drug-induced cholestasis, intrahepatic cholestasis of pregnancy and therapeutic response of viral hepatitis. Expert Opin Drug Metab Toxicol 2011;7:411–25. [DOI] [PubMed] [Google Scholar]
- 6.Nicolaou M, Andress EJ, Zolnerciks JK, Dixon PH, Williamson C, Linton KJ. Canalicular ABC transporters and liver disease. J Pathol 2012;226:300–15. [DOI] [PubMed] [Google Scholar]
- 7.de Lima Toccafondo Vieira M, Tagliati CA. Hepatobiliary transporters in drug-induced cholestasis: a perspective on the current identifying tools. Expert Opin Drug Metab Toxicol 2014;10:581–97. [DOI] [PubMed] [Google Scholar]
- 8.Köck K, Ferslew BC, Netterberg I, Yang K, Urban TJ, Swaan PW, et al. Risk factors for development of cholestatic drug-induced liver injury: inhibition of hepatic basolateral bile acid transporters multidrug resistance-associated proteins 3 and 4. Drug Metab Dispos 2014;42:665–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Yang K, Battista C, Woodhead JL, Stahl SH, Mettetal JT, Watkins PB, et al. Systems pharmacology modeling of drug-induced hyperbilirubinemia: differentiating hepatotoxicity and inhibition of enzymes/transporters. Clin Pharmacol Ther 2017;101:501–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Daly AK, Aithal GP, Leathart JB, Swainsbury RA, Dang TS, Day CP. Genetic susceptibility to diclofenac-induced hepatotoxicity: contribution of UGT2B7, CYP2C8, and ABCC2 genotypes. Gastroenterology 2007;132:272–81. [DOI] [PubMed] [Google Scholar]
- 11.Lang C, Meier Y, Stieger B, Beuers U, Lang T, Kerb R, et al. Mutations and polymorphisms in the bile salt export pump and the multidrug resistance protein 3 associated with drug-induced liver injury. Pharmacogenet Genomics 2007;17:47–60. [DOI] [PubMed] [Google Scholar]
- 12.Alfirevic A, Mills T, Carr D, Barratt BJ, Jawaid A, Sherwood J, et al. Tacrine-induced liver damage: an analysis of 19 candidate genes. Pharmacogenet Genomics 2007;17:1091–100. [DOI] [PubMed] [Google Scholar]
- 13.Choi JH, Ahn BM, Yi J, Lee JH, Lee JH, Nam SW, et al. MRP2 haplotypes confer differential susceptibility to toxic liver injury. Pharmacogenet Genomics 2007;17:403–15. [DOI] [PubMed] [Google Scholar]
- 14.Meier Y, Zodan T, Lang C, Zimmermann R, Kullak-Ublick GA, Meier PJ, et al. Increased susceptibility for intrahepatic cholestasis of pregnancy and contraceptive-induced cholestasis in carriers of the 1331T>C polymorphism in the bile salt export pump. World J Gastroenterol 2008;14:38–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Morgan RE, Trauner M, van Staden CJ, Lee PH, Ramachandran B, Eschenberg M, et al. Interference with bile salt export pump function is a susceptibility factor for human liver injury in drug development. Toxicol Sci 2010;118:485–500. [DOI] [PubMed] [Google Scholar]
- 16.Dawson S, Stahl S, Paul N, Barber J, Kenna JG. In vitro inhibition of the bile salt export pump correlates with risk of cholestatic drug-induced liver injury in humans. Drug Metab Dispos 2012;40:130–8. [DOI] [PubMed] [Google Scholar]
- 17.Kim SH, Kim SH, Lee JH, Lee BH, Kim YS, Park JS, et al. Polymorphisms in drug transporter genes (ABCB1, SLCO1B1 and ABCC2) and hepatitis induced by antituberculosis drugs. Tuberculosis (Edinb) 2012;92:100–4. [DOI] [PubMed] [Google Scholar]
- 18.Ulzurrun E, Stephens C, Crespo E, Ruiz-Cabello F, Ruiz-Nuñez J, Saenz-López P, et al. Role of chemical structures and the 1331T>C bile salt export pump polymorphism in idiosyncratic drug-induced liver injury. Liver Int 2013;33:1378–85. [DOI] [PubMed] [Google Scholar]
- 19.El Sherrif Y, Potts JR, Howard MR, Barnardo A, Cairns S, Knisely AS, et al. Hepatotoxicity from anabolic androgenic steroids marketed as dietary supplements: contribution from ATP8B1/ABCB11 mutations? Liver Int 2013;33:1266–70. [DOI] [PubMed] [Google Scholar]
- 20.Roustit M, Fonrose X, Montani D, Girerd B, Stanke-Labesque F, Gonnet N, et al. CYP2C9, SLCO1B1, SLCO1B3, and ABCB11 polymorphisms in patients with bosentan-induced liver toxicity. Clin Pharmacol Ther 2014;95:583–5. [DOI] [PubMed] [Google Scholar]
- 21.Rodrigues AD, Lai Y, Cvijic ME, Elkin LL, Zvyaga T, Soars MG. Drug-induced perturbations of the bile acid pool, cholestasis, and hepatotoxicity: mechanistic considerations beyond the direct inhibition of the bile salt export pump. Drug Metab Dispos 2014;42:566–74. [DOI] [PubMed] [Google Scholar]
- 22.Garzel B, Yang H, Zhang L, Huang SM, Polli JE, Wang H. The role of bile salt export pump gene repression in drug-induced cholestatic liver toxicity. Drug Metab Dispos 2014;42:318–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Ulzurrun E, Stephens C, Ruiz-Cabello F, Robles-Diaz M, Saenz-López P, Hallal H, et al. Selected ABCB1, ABCB4 and ABCC2 polymorphisms do not enhance the risk of drug-induced hepatotoxicity in a Spanish cohort. PLoS One 2014;9:e94675. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kagawa T, Hirose S, Arase Y, Oka A, Anzai K, Tsuruya K, et al. No contribution of the ABCB11 p.444A polymorphism in Japanese patients with drug-induced cholestasis. Drug Metab Dispos 2015;43:691–7. [DOI] [PubMed] [Google Scholar]
- 25.Chen R, Wang J, Tang S, Zhang Y, Lv X, Wu S, et al. Role of polymorphic bile salt export pump (BSEP, ABCB11) transporters in anti-tuberculosis drug-induced liver injury in a Chinese cohort. Sci Rep 2016;6:27750. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Fontana RJ, Watkins PB, Bonkovsky HL, Chalasani N, Davern T, Serrano J, et al. ; DILIN Study Group. Drug-Induced Liver Injury Network (DILIN) prospective study: rationale, design and conduct. Drug Saf 2009;32:55–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Danan G, Benichou C. Causality assessment of adverse reactions to drugs–I. A novel method based on the conclusions of international consensus meetings: application to drug-induced liver injuries. J Clin Epidemiol 1993;46:1323–30. [DOI] [PubMed] [Google Scholar]
- 28.European Association for the Study of the Liver. EASL Clinical Practical Guidelines on the management of acute (fulminant) liver failure. J Hepatol 2017;66:1047–81. [DOI] [PubMed] [Google Scholar]
- 29.Huang YS, Chern HD, Su WJ, Wu JC, Chang SC, Chiang CH, et al. Cytochrome P450 2E1 genotype and the susceptibility to antituberculosis drug-induced hepatitis. Hepatology 2003;37:924–30. [DOI] [PubMed] [Google Scholar]
- 30.Chalasani NP, Hayashi PH, Bonkovsky HL, Navarro VJ, Lee WM, Fontana RJPractice Parameters Committee of the American College of Gastroenterology. ACG Clinical Guideline: the diagnosis and management of idiosyncratic drug-induced liver injury. Am J Gastroenterol 2014;109:950–66; quiz 967. [DOI] [PubMed] [Google Scholar]
- 31.Björnsson ES. Epidemiology and risk factors for idiosyncratic drug-induced liver injury. Semin Liver Dis 2014;34:115–22. [DOI] [PubMed] [Google Scholar]