Association of p-glycoprotein and bile salt export pump gene polymorphisms with advanced liver disease in hepatitis C virus infected patients

Abstract

BACKGROUND  Single-nucleotide polymorphisms (SNPs) can influence the hepatitis C virus (HCV) infection and progression. ABCB1-gene SNPs - c.1236C>T, c.2677G>T and c.3435C>T - are associated with drug efficacy, hepatotoxicity, and liver injury. ABCB11 c.1331T>C is associated with cholestasis and altered bilirubin levels, potentially worsening liver disease.

OBJECTIVE  To investigate the impact of ABCB1 and ABCB11 SNPs on disease progression in chronic HCV patients.

METHODS  A total of 232 HCV patients unresponsive to conventional therapy were analysed. Serum samples were genotyped by quantitative polymerase chain reaction (qPCR), and the genotype and allele-based analysis was performed using RStudio.

FINDINGS  Cirrhosis was present in 59.1% of patients, along with diabetes (28.4%) and hepatic steatosis (46.1%). The most frequent ABCB1 variant allele was c.3435C>T (36.8%), followed by c.1236C>T (30.8%) and c.2677G>T (26.7%). The ABCB11 c.1331CC genotype was observed in 31.3% of the cohort. Genotypes 1236TT and 2677TT and their alleles were associated with lower total cholesterol. 2677TT genotype and 2677T allele were associated with lower high-density lipoprotein. Patients with 1331CC genotype had higher aspartate aminotransferase levels, and the 1331CC genotype was a risk factor for cirrhosis. A fully variant combined allele (1236T and 2677T and 3435T and 1331C) was associated with higher alpha-fetoprotein and lower cholesterol.

MAIN CONCLUSIONS  ABCB1 and ABCB11 SNPs are associated with worse clinical outcomes in HCV, underscoring their relevance in disease monitoring.

Key words:
hepatitis C; ATP-binding cassette; ABCB1 ; ABCB11 ; drug resistance; genetic factors

Hepatitis C virus (HCV) is the leading cause of chronic hepatitis, with 50 million individuals estimated to be chronically infected with the virus worldwide and 1 million new infections occurring per year.¹ HCV-infected individuals are at high risk of developing liver disease that can progress to cirrhosis and hepatocellular carcinoma (HCC), which is the most common primary liver cancer and the main reason for liver transplantation.^2,3 During hepatocyte infection, biochemical alterations may occur, including increased serum levels of hepatic and canalicular membrane enzymes, such as aspartate aminotransferase (AST), alanine aminotransferase (ALT), and gamma-glutamyl transferase (GGT), which are associated with cholestasis and HCV-induced liver damage.⁴ Given that the liver plays an important role in lipid metabolism, the lipid profile may deteriorate in patients with more severe disease, including uncompensated cirrhosis, with reduced levels of cholesterol and high-density lipoprotein (HDL) ensuing.^5,6,7

Several viral and host factors are associated with increased susceptibility to HCV infection and liver disease progression. These include age, male sex, obesity, insulin resistance, type 2 diabetes, co-infection with hepatitis B virus or human immunodeficiency virus (HIV), hepatic steatosis,^8,9 and vitamin D deficiency.^10,11,12,13

ATP-binding cassette (ABC) transporters, which are present in several species, are transmembrane proteins that harvest energy from ATP hydrolysis to transport a variety of molecules, including drugs.¹⁴ One of the most-studied proteins in this family is P-glycoprotein, which functions as a drug efflux pump. P-glycoprotein is encoded by the ABCB1/multidrug resistance 1 (MDR1) gene, the single nucleotide polymorphisms (SNPs) of which have been associated with hepatotoxicity and resistance to multiple drugs (usually antineoplastic).^15,16,17

Three polymorphisms in the ABCB1 gene (c.3435C>T, c.2677G>T, and c.1236C>T) are associated with differences in its expression and in the activity of its translated protein. The SNP c.3435C>T (rs1045642), located in exon 26 of ABCB1, is a synonymous polymorphism (p.Ile1145Ile) that is associated with reduced protein expression.¹⁸ The c.2677G>T (rs2032582) variant in exon 21 is a non-synonymous polymorphism (p.Ala893Ser), resulting in an amino acid substitution located at the cytoplasmic region within the second transmembrane domain of the ABCB1 protein, potentially affecting its activity and substrate specificity.¹⁸ The c.1236C>T (rs1128503) genetic variant in exon 12, a synonymous polymorphism (p.Gly412Gly), is associated with changes in protein expression.¹⁸

The ABCB11 gene encodes the bile salt export pump (BSEP), another important protein in the ABC transporter family that is found mainly in the apical (canalicular) membrane of hepatocytes. This protein actively transports bile salts from the liver into the bile canaliculus, ensuring normal bile flow through the osmotic process.¹⁹ Polymorphisms in ABCB11 are associated with progressive familial intrahepatic cholestasis type 2 and an increased risk of HCC during early human development. Specifically, c.1331T>C (rs2287622) is a non-synonymous polymorphism (p.Val444Ala) located in exon 13 that is associated with cholestatic liver disease (mainly progressive intrahepatic cholestasis) and drug-induced cholestasis characterised by decreased or interrupted bile flow.^20,21,22,23

Genetic elements that confer a greater predisposition to liver problems can act as important regulatory factors in worsening the outcomes of patients with chronic hepatitis, especially those with HCV infection. The presence of the 3435T, 2677T, and 1236T variant alleles in ABCB1 can influence the plasma concentration and sustained virological response (SVR) of several HCV drugs, including telaprevir, sofosbuvir, and ribavirin (RBV).^24,25 The homozygous TT genotype of the ABCB1 polymorphism in exon 26 is associated with a 2.0-fold greater risk of non-response to drugs than the homozygous CC and heterozygous CT genotypes.²⁴ The c.1331T>C SNP in ABCB11 exerts an apparent influence on the pharmacokinetics of RBV and significantly elevates total bilirubin levels after antiviral therapy. The increase is greater in individuals with the 1331CC genotype than in those with 1331TT.²⁵

Investigation of the influence of various ABCB1 (c.1236C>T, c.2677G>T, and c.3435C>T) and ABCB11 (c.1331T>C) SNPs will enhance our understanding of the clinical evolution of patients carrying the variants, since this variants can aggravate hepatic tissue damage and cause cholestasis in predisposed individuals, thereby increasing the levels of biochemical markers of liver injury that should be monitored. Thus, this study aimed to investigate the genotypic and allelic profiles of c.1236C>T, c.2677G>T, c.3435C>T, and c.1331T>C as well as their relationships with the clinical and laboratory characteristics of patients with chronic HCV infection unresponsive to interferon (IFN) and/or RBV treatment.

SUBJECTS AND METHODS

Ethical aspects - This study was approved by the Ethics Committee of the Oswaldo Cruz Foundation (CAAE 34246914.4.1001.5248; number:2.927.747/18). The purpose of the study was explained to the participants. Confidentiality regarding patient identity and personal information was assured by highlighting that only researchers could access the information, which would be used solely for research purposes. The participants were then asked to sign an informed consent form.

Study sample and clinical data - In this retrospective, hypothesis-driven study, designed to investigate whether genetic variants in the ABCB1 and ABCB11 genes influence the disease progression in chronic HCV patients (primary hypothesis). For the purposes of this study, disease progression was defined as the presence of advanced fibrosis (grade F3) or cirrhosis (grade fibrosis F4), which constituted the primary outcome of the study. The secondary outcomes included laboratory and clinical parameters biologically linked to liver injury or metabolic processes potentially affected by ABCB gene polymorphisms, like lipid profile, liver enzymes, and metabolic biomarkers. Since the study includes both hypothesis-driven components (primary outcome) and exploratory components (secondary outcomes), we acknowledge the issue of multiple comparisons in the Discussion and clarify that exploratory associations should be interpreted with caution.

We analysed the medical records and laboratory test results of 232 patients with chronic HCV infection (genotypes 1a and 1b), pooled between November 2015 and November 2017 from the Liver Diseases Outpatient Clinic of Gaffrée and Guinle University Hospital (Federal University of the State of Rio de Janeiro — UNIRIO) in Rio de Janeiro, Brazil. All patients had been diagnosed with hepatitis as confirmed through serological and molecular tests. Clinical and laboratory information was obtained from the outpatient records and registered into our database by a specialised medical team.

The data obtained included age, sex, body mass index, diabetes status, serological levels of liver enzymes (AST, ALT, and GGT) and metabolic biomarkers [albumin, glucose, triglycerides, total cholesterol, low-density lipoprotein (LDL), HDL, bilirubin, alpha-fetoprotein (AFP)], cellular and haematological biomarkers (platelets, haemoglobin, leukocytes, and haematocrit), viral load, homeostatic model assessment of insulin resistance, hepatic steatosis, METAVIR activity and fibrosis/cirrhosis level, and histopathological examinations of liver biopsies. Reference values were adopted for each biomarker mentioned, following protocols described in the literature.^26,27

A simple random probability sampling form was used to ensure that the study sample was representative of the overall population, thereby guaranteeing the internal validity of the study. The minimum number of participants (n = 57) was determined using the equation n = z² × p × (1 - p)/e², where z is the confidence level based on a standard normal distribution (1.96 for 95% confidence interval), p is the expected prevalence (0.183 for HCV in the general population, because no data were available for the study sample), and e is the maximum acceptable error in the estimate (0.05).²⁸

The inclusion criteria were as follows: HCV-infected patients; older than 18 years of age; receiving follow-up care at the liver disease outpatient clinic; and having signed the informed consent form. The exclusion criteria were as follows: Patients co-infected with other hepatitis viruses (A and E); and haemolysed samples and/or samples with a volume of less than 200 μL.

Nucleic acid extraction and analysis of the ABCB1 and ABCB11 genes - The nucleic acid extraction from the serum sample was performed using the QIAamp RNA Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer's instructions. This extraction kit was selected due to its ability to simultaneously isolate DNA and RNA, allowing a single extraction step to be used for the analysis of HCV, an RNA virus, as reported in a previously published study,²⁹ as well as for the analysis of human genes performed in the present study.

Genotyping of polymorphisms in the ABCB1 and ABCB11 genes was performed using real-time polymerase chain reaction (qPCR) with the TaqMan Universal PCR Master Reagent Mix (Applied Biosystems, Foster City, CA, United States) and the TaqMan SNP Genotyping Assay kit (Applied Biosystems) for allele discrimination for each polymorphism, according to the manufacturer's instructions.

The TaqMan SNP genotyping assays used for each polymorphism were as follows: ABCB1 c.1236C>T (Assay ID: C___7586662_10), ABCB1 c.2677G>T (Assay ID: C_11711720C_30), ABCB1 c.3435C>T (Assay ID: C___7586657_20), and ABCB11 c.1331T>C (Assay ID: C__16182459_10). Each assay involved a pair of oligonucleotides (sense and antisense) and a pair of fluorophores (VIC and FAM)-labelled probes with distinct excitation/emission spectra to distinguish the alleles. Three genotypes for each polymorphism were determined based on the manufacturer's information regarding the excitation and emission spectra of each fluorophore used in the TaqMan SNP genotyping assays [Supplementary data (Table I)].

Statistical analysis - All statistical analyses were performed using RStudio software (version 4.1.1, 2021). The significance level was set to a p-value of less than 0.05. Data are presented as frequencies and percentages for categorical variables and as means with standard deviations (SD) for continuous variables. The univariate analysis was performed using the chi-squared (χ²) test, Mann-Whitney U test, Student's t-test, or Wilcoxon's test to analyse categorical, parametric continuous, and non-parametric variables, as appropriate. Associations between individuals with genetic variants and those with wild-type genotype were evaluated against several potential predictor variables (body mass index, levels of fasting blood glucose, triglycerides, total cholesterol, LDL, HDL, AST, ALT, and GGT, degree of fibrosis, steatosis, and platelet count).

The comparisons between genotype groups and demographic variables (age, sex, BMI, diabetes, and others) were performed because these variables are well-established confounders in liver disease progression and in several secondary outcomes analysed in the study, including lipid profile, liver enzymes, and AFP levels. Assessing genotype distribution across these variables allows us to verify whether these factors may influence the associations observed and to ensure adequate characterisation of the cohort.

In the genotype-based analysis, individuals carrying the wild-type genotype were used as the reference group. Comparisons were performed separately between wild-type and heterozygous genotypes and between wild-type and mutant genotypes. For these analyses, continuous variables are presented as mean ± SD calculated based on the number of individuals in each genotype group. Heterozygous individuals were not duplicated within the same comparison. For allele-based analyses, the dataset was expanded to account for the two alleles carried by each individual, resulting in a total of 464 alleles. In this approach, continuous variables are presented as mean ± SD calculated based on allele counts, using alleles as the unit of analysis. Comparisons were performed between wild-type and variant alleles.

To determine whether the associations between the genetic polymorphisms and the clinical or laboratory parameters were independent of potential confounding factors, multivariate regression models were performed. For continuous outcomes, we employed multiple linear regression, whereas for categorical outcomes we used multinomial or logistic regression, as appropriate. The variables age, sex, BMI, and diabetes status were included as covariates in all multivariate models, as they are well-established predictors of lipid alterations, hepatic injury markers, and disease progression in chronic hepatitis C. These covariates were introduced simultaneously to adjust for their confounding effects and to estimate the independent contribution of each genotype or allele to the outcomes evaluated.

RESULTS

Clinical and epidemiological profiles - The clinical and epidemiological profiles of the study participants are described in Table I and stratified by the studied polymorphisms genotypes in Supplementary data (Table II). The mean age of the patients was 61.2 ± 9.8 years, with the number of individuals increasing progressively with an increase in the age range, and approximately 56.9% were females. Additionally, 32.2% were overweight and 16.5% were obese. A large minority of patients (46.1%) had steatosis, and the majority (59.1%) had F4 fibrosis/cirrhosis, whereas the frequency of HCC was low (2.2%). Approximately 12.1% (28/232) of the patients were co-infected with HIV.

TABLE I
Clinical and epidemiological profiles of patients with chronic hepatitis C virus (HCV) infection

Variable	Value (n = 232)
Age, yr	61.19 ± 9.8
Sex
Female	132 (56.9%)
Male	100 (43.1%)
Body mass index, kg/m2	26.38 ± 4.58
Diabetes
Yes	66 (28.4%)
No	159 (68.5%)
ND	7 (3.0%)
Albumin, mg/dL	4.04 ± 0.7
Glucose, mg/dL	106.7 ± 30.0
Triglycerides, mg/dL	118.6 ± 145.9
Total cholesterol, mg/dL	161.4 ± 32.1
Low-density lipoprotein, mg/dL	90.14 ± 44.8
High-density lipoprotein, mg/dL	56.0 ± 44.2
Aspartate aminotransferase, U/L	62.59 ± 49.5
Alanine aminotransferase, U/L	62.27 ± 51.7
AST/ALT ratio	1.07 ± 0.6
Platelets/L	163.07 × 103 ± 94.3 × 103
Gamma-glutamyl transferase, U/L	102.10 ± 103.7
Total bilirubin, mg/dL	0.85 ± 0.5
Haemoglobin, g/L	13.22 ± 2.0
Leukocytes, cells/mm3	5752.0 ± 3667.3
Haematocrit	40.31% ± 5.0%
Alpha-fetoprotein, ng/mL	17.90 ± 30.1
Elastography, Kpa	18.82 ± 12.1
Steatosis
Yes	107 (46.1%)
No	46 (19.8%)
ND	79 (34.0%)
Hepatocellular carcinoma
Yes	5 (2.2%)
No	227 (97.8%)
Fibrosis
F1	12 (5.1%)
F2	20 (8.6%)
F3	63 (27.2%)
F4 (Cirrhosis)	137 (59.0%)
HCV genotype
1a	121 (52.2%)
1b	111 (47.8%)

• AST: aspartate aminotransferase; ALT: alanine aminotransferase; ND: no data. Values are presented as n (%) or mean ± standard deviation (SD), as appropriate. Reference values: BMI: 18.5-25 = normal; 25-30 = overweight; > 30 = obesity; AST (U/L): 5-40 = normal; > 40 = elevated; ALT (U/L): 7-56 = normal; > 56 = elevated; GGT (U/L) male: 8-61 = normal; > 61 = elevated; GGT (U/L) female: 5-36 = normal; > 36 = elevated; Bilirubin (mg/dL): ≤ 1.2 = normal; Albumin (g/dL): 3.5-4.7 = normal; < 3.5 = low; > 4.7 = high; Cholesterol (mg/dL): ≤ 190 = normal; >190 = elevated; Platelets (1000/μL): 140-450 = normal; < 140 = thrombocytopenia; Glucose (mg/dL): ≤ 99 = normal; > 99 = elevated.^26,27

Frequency of polymorphisms - With regard to the ABCB1 SNPs (Figure), the frequencies of the CC, CT, and TT genotypes of c.1236C>T (rs1128503) were 49.6% (115/232), 32.2% (91/232), and 11.2% (26/232), respectively, whereas the frequency of the wild-type (C) allele was 69.2% and that of the variant allele (T) was 30.8%. For c.3435C>T (rs1045642), 15.5% (36/232) of the patients carried the TT genotype, 42.7% (99/232) carried the heterozygous genotype (CT), and 41.8% (97/232) carried the CC genotype. Moreover, the wild-type allele (C) is shown to be present in 63.2% of the samples and the variant allele (T) in 36.8%. For c.2677G>T (rs2032582), the frequency of the GG genotype was 55.6% (129/232), that of the TT genotype was 9.0% (21/232), and that of the heterozygous (GT) genotype was 35.4% (82/232), and the wild-type (G) and variant (T) alleles frequencies were 73.3% and 26.7%, respectively.

With regard to c.1331T>C (rs2287622) in the ABCB11 gene (Figure), 54.3% (126/232) of the patients carried the heterozygous genotype (TC), 14.7% (34/232) had the TT genotype, and 31.0% (72/232) carried the CC genotype. The wild-type allele (T) was present in 41.8% of the patients and the variant allele (C) in 58.2%. All genotype frequencies for the studied polymorphisms were in Hardy–Weinberg equilibrium based on the expected genotype frequencies calculated [Supplementary data (Table III)].

Figure:
Frequencies of the polymorphisms studied in patients with chronic hepatitis C virus (HCV) infection. These genotyping results were generated using Quant Studio 3. The graph is separated according to single-nucleotide polymorphism (SNP) and gene (ATP-binding cassette subfamily B member 1 - ABCB1, or member 11 - ABCB11), with purple representing the wild-type genotype (1236CC, 2677GG, 3435CC, or 1331TT), green the heterozygous genotype (1236CT, 2677GT, 3435CT, or 1331TC), and blue the variant genotype (1236TT, 2677TT, 3435TT, or 1331CC). Orange represents the frequency of the wild-type allele (1236C, 2677G, 3435C, or 1331T) and pink the frequency of the variant allele (1236T, 2677T, 3435T, or 1331C).

Association analysis - The four SNPs were evaluated in the genotype-based analysis for their effects on all studied biochemical and epidemiological variables, and the results are presented in Supplementary data (Table I). Table II includes only the variables that showed statistically significant differences according to genotype-based or allele-based analysis. In addition, only the SNPs that presented significant associations are displayed in this table, specifically ABCB1 2677G>T, ABCB1 1236C>T, and ABCB11 1331T>C. Regarding the ABCB1 c.3435C>T polymorphism, no statistically significant associations were observed with any of the clinical or laboratory parameters evaluated (data not shown).

In the univariable analysis (Table II), patients carrying the 2677TT genotype or 2677T allele had significantly lower levels of total cholesterol (p < 0.05) than those carrying the 2677GG or 2677GT genotypes or 2677G allele. Furthermore, significantly lower levels of HDL (p < 0.05) were associated with the 2677TT genotype and 2677T allele. In the multivariate analysis (Table II), a lower total cholesterol level continued to be observed among carriers of the 2677TT genotype [β = -32.16; 95% confidence interval (CI): -48.9 to -15.4; p < 0.01] and 2677T allele (β = -11.7; 95% CI: -19.2 to -4.3; p < 0.01). However, HDL levels were no longer significant (genotype: β = -16.6; 95%CI: -46.7 to 13.5; p > 0.1; allele: β = -3.01; 95% CI: -15.2 to 9.2; p > 0.1).

In the univariate and multivariate analysis, the total cholesterol level was significantly lower (p < 0.05) in carriers of the 1236TT genotype (β = -30.76; 95% CI: -45.5 to -15.9 p < 0.01) or 1236T allele (β = -11.9; 95% CI: -19.1 to -4.8; p < 0.01) than in individuals carrying 1236CC and 1236CT genotypes, or 1236C allele (Table II).

With regard to c.1331T>C, in the univariate analysis, the presence of the variant-type genotype (1331CC) was a risk factor for the development of liver cirrhosis (p < 0.05; OR: 2.40; 95% CI: 1.04-5.60) in this study samples compared with that of the wild genotype (1331TT). And carriers of the 1331CC genotype had higher AST levels (p < 0.05) than those in individuals carrying the 1331TT genotype (Table II). In the multivariate analysis, the association between the c.1331T>C variant and cirrhosis remained significant in the genotype-based analysis (adjusted OR = 3.05; 95% CI: 1.15 to 8.49), with the variant genotype showing a higher frequency of cirrhosis. Regarding AST levels, the association observed in the genotype-based analysis was no longer statistically significant after adjustment (β = 15.89; 95% CI: -3.9 to 35.8; p > 0.1), but in allele-based analysis the variant 1331C allele continued to show a significant association (β = 9.44; 95% CI: 0.73 to 18.2; p < 0.05) (Table II).

In the multivariate allele-based analysis (Table II), a positive diabetes status was independently associated with cholesterol levels, whereas sex and BMI were related to AST levels (p < 0.05). These findings highlight the importance of both genetic factors and biochemical/clinical variables as independent contributors to metabolic alterations.

TABLE II
Statistical analysis of significant variables associated with the single-nucleotide polymorphisms (SNPs) c.2677G>T and c.1236C>T in the ATP-binding cassette subfamily B member 1 (ABCB1) gene and c.1331T>C in the ABCB11 gene

Analysis by genotype				p-value (univariate)	Adjusted p-value (multivariate)	Analysis by allele		p-value (univariate)	Adjusted p-value (multivariate)
ABCB1 c.2677G>T SNP
	GG (n = 129)	GT (n = 82)	TT (n = 21)			2677G (n = 339)	2677T (n = 125)
Total cholesterol, mg/dL	166.7 ± 31.1	159.2 ± 30.4	139.1 ± 35.8	< 0.01	< 0.001 (β = -32.16; 95% CI: -48.9 to -15.9)	164.73 ± 30.9	152.68 ± 33.4	< 0.01	<0.01 (β = -11.7; 95% CI: -19.2 to -4.3)
High-density lipoprotein, mg/dL	57.0 ± 29.3	57.3 ± 62.1	42.0 ± 14.6	< 0.05	0.27 (β = -16.6; 95% CI: -46.7 to 13.5)	57.10 ± 40.3	52.91 ± 53.4	< 0.01	0.62 (β = -3.01; 95% CI: -15.2 to 9.2)
ABCB1 c.1236C>T SNP
	CC (n = 115)	CT (n = 91)	TT (n = 26)			1236C (n = 321)	1236T (n = 143)
Total cholesterol, mg/dL	165.9 ± 30.1	162.5 ± 30.0	137.1 ± 38.7	< 0.001	< 0.001 (β = -30.76; 95% CI: -45.5 to -15.9)	164.87 ± 30.0	153.70 ± 35.1	< 0.01	<0.01 (β = -11.9; 95% CI: -19.1 to -4.8)
ABCB11 c.1331T>C SNP
	TT (n = 34)	TC (n = 126)	CC (n = 72)			1331T (n = 194)	1331C (n = 270)
Cirrhosis				< 0.05 (OR: 2.40; 95% CI: 1.04–5.60)	< 0.05 (adjusted OR = 3.05; 95% CI: 1.15 to 8.49)			0.05	0.1 (adjusted OR = 1.51; 95% CI: 0.98 to 2.34)
Yes	16 (47.1)	72 (57.1)	49 (68.1)			90 (46.4)	100 (37.0)
No	18 (52.9)	54 (42.9)	23 (31.9)			104 (53.6)	170 (63.0)
Aspartate aminotransferase, U/L	59.6 ± 32.6	54.8 ± 37.8	77.9 ± 68.1	< 0.05	0.11 (β = 15.89; 95% CI: -3.9 to 35.8)	56.39 ± 36.0	67.02 ± 56.8	0.2	< 0.05 (β = 9.44; 95% CI: 0.73 to 18.2)

• ABCB1: ATP-binding cassette subfamily B member 1; ABCB11: ATP-binding cassette subfamily B member 11; CI: confidence interval; SNP: single-nucleotide polymorphism. The statistical analysis was performed by genotype and by allele for each polymorphism studied. Continuous variables are presented as mean ± standard deviation (SD). In genotype-based analyses, values were calculated per individual within each genotype group, using the wild-type genotype as reference. In allele-based analyses, values were calculated using allele counts as the unit of analysis. No significant results were found for the analysis of the c.3435C>T polymorphism; therefore, this SNP was not included in Table II. In the multivariate analysis, potential confounders (age, sex, body mass index, and diabetes) were incorporated into the model to control for their effects. *In the multivariate allele analysis, a positive diabetes status was independently associated with cholesterol levels, whereas sex and body mass index (BMI) were related to aspartate aminotransferase (AST) levels (p < 0.05).

In the univariate combined allele analysis (Table III), the presence of the fully variant combined allele (1236T, 2677T, 3435T, and 1331C; TTTC) was significantly associated with lower total cholesterol and higher AFP levels compared to other genotypes (p < 0.05). However, in the multivariate analysis, only the lower cholesterol levels (β = -10.84; 95% CI: -19.25 to -2.4; p < 0.05) remained associated with the fully variant combined allele (Table III). In addition, in the combined-allele multivariate analysis (Table III), cholesterol levels were also independently associated with diabetes, reinforcing the importance of both variables as independent contributors to metabolic alterations.

TABLE III
Analysis by combining allele of the studied variants

Analysis by combined allele			p-value (univariate)	Adjusted p-value (multivariate)
	Fully variant combined allele TTTC (n = 88)	Other allelic combination (n = 376)
Alpha-fetoprotein, ng/mL	19.07 ± 22.1	17.63 ± 31.5	< 0.05	0.69 (β = 2.23; 95% CI: -8.9 to 13.40)
Total cholesterol, mg/dL	153.33 ± 31.1	163.29 ± 32.1	< 0.05	< 0.05 (β = -10.84; 95% CI: -19.2 to -2.4)

• CI: confidence interval. The analysis was carried out by stratifying the 16 possible combined alleles into two groups: (1) alleles carrying the variant form for all four polymorphisms studied, or fully variant combined allele (1236T and 2677T and 3435T and 1331C; TTTC); and (2) composed of the remaining possibilities, that is, alleles presenting at least one wild-type allele among the four loci (1236C or 2677G or 3435C or 1331T). In the multivariate analysis, potential confounders (age, sex, body mass index - BMI, and diabetes) were incorporated into the model to control for their effects. *In the combined-genotype multivariate analysis, cholesterol levels were also independently associated with diabetes.

DISCUSSION

This study highlights the impact of ABCB1 and ABCB11 gene polymorphisms in patients with chronic HCV infection unresponsive to IFN and/or RBV therapy. The presence of polymorphism in ABCB1 (c.2677G>T, c.3435C>T, and c.1236C>T) and ABCB11 (c.1331T>C) is known to be associated with drug resistance, hepatotoxicity, and cholestasis as well as resistance to drugs used in the treatment of HCV-infected patients, such as telaprevir, sofosbuvir, and RBV.^{15,16,17,20-25} In this study, 59.0% of the patients had advanced liver diseases such as cirrhosis (F4) and 2.2% had HCC. Some studies have shown that the presence of these polymorphisms can directly or indirectly influence the clinical evolution of HCV-infected patients to cirrhosis and HCC.^{30,31,32,33,34}

Our results showed that, among ABCB1 polymorphisms, the 2677TT genotype occurred at the lowest frequency, being present in only 9.0% of the samples, whereas the 1236TT and 3435TT genotypes were observed in 10.8% and 15.5% of the cohort, respectively. In contrast, the 1331CC genotype of the ABCB11 was found in 31.0% of the samples. Overall, in the genotype-based analysis for these four polymorphisms, we observed a greater prevalence of heterozygote genotypes in the study samples, which may be a characteristic of a mixed population such as that in Brazil. All variants were in Hardy-Weinberg equilibrium, indicating that the observed genotype frequencies were consistent with those expected for a genetically stable sample. This supports the reliability of the genotyping procedures and suggests that no sample stratification or methodological bias significantly affected the distribution of these genotypes, thereby reinforcing the validity of the association analyses performed.

The study was primarily hypothesis-driven, focused on evaluating whether ABCB1 and ABCB11 variants influence the development of advanced fibrosis or cirrhosis. The assessment of secondary biochemical and metabolic parameters reflects the biological relevance of ABC transporters in hepatic injury and bile acid regulation. However, we recognise that the substantial number of variables included also introduces exploratory elements. Therefore, the potential for type I error is acknowledged, and the secondary findings should be interpreted as hypothesis-generating.

It was observed, in the univariate and multivariate genotype or allele-based analysis, that the 2677TT and 1236TT genotypes were an independent factor significantly associated with lower total cholesterol levels compared with their wild-type and heterozygous genotypes. Additionally, the presence of the variant allele (T) in the 2677 and 1236 polymorphisms correlated with lower serum levels of total cholesterol compared with the levels in patients with the wild-type alleles (2677G and 1236C). The worsening of HCV-associated liver disease is related to changes in the lipid profile, with significant reductions in total cholesterol, as the liver is responsible for lipid metabolism.^5,6,7 These changes in the lipid profile are usually mainly observed in individuals with decompensated cirrhosis and HCC. However, because the patients in this study had predominately compensated cirrhosis, the lipid profile decrease was unexpected. Therefore, the reductions in total cholesterol in these individuals may be closely related to the presence of 2677TT and 1236TT genotypes, which may be associated with a worse prognosis and a greater risk of developing cirrhosis and liver cancer.

In the univariate genotype and allele-based analysis, the 2677TT genotype and 2677T allele were associated with a lower serum HDL level, but not in the multivariate. The disappearance of the association in the adjusted analysis may reflect the stronger influence of metabolic comorbidities, such as diabetes, on lipid parameters, which can mask or override the modest genetic effect expected for this polymorphism.

In univariate and multivariate analysis of our cohort, the presence of the variant-type 1331CC genotype in ABCB11 was a risk factor for the development of liver cirrhosis (p < 0.05) as compared with the presence of 1331TT genotype, indicating that the variant genotype is a relevant factor in the development of advanced liver disease in patients with chronic HCV infection. Carriers of the 1331CC genotype also exhibited higher AST levels (p < 0.05) compared with individuals carrying 1331TT or 1331TC genotypes. In the multivariate allele analysis, the 1331T>C variant remained independently associated with AST levels. AST is an important biomarker of liver injury, with its higher serum levels indicating greater tissue damage.

The importance of the 1331CC genotype as a risk factor for progression to cirrhosis under chronic HCV infection is supported by other research findings, where its presence has been shown to increase plasma bile acid levels, consequently worsening fibrosis and increasing the risk of HCC in HCV-infected individuals.^31,32,33 Additionally, in our exploratory analysis [Supplementary data (Table II)], we observed higher levels of GGT in individuals carrying 1331CC genotype than in those carrying the wild-type genotype, although the difference was not statistically significant. Increased expression of GGT, an important canalicular membrane enzyme and a biomarker of liver damage, is associated with worse prognosis and reduced survival in patients with HCC.³⁵

In the exploratory analysis of clinical data [Supplementary data (Table II)], no clear pattern was detected among the polymorphisms. For ABCB1 c.2677G>T, higher AST, ALT, and bilirubin levels were observed mainly in individuals carrying the TT genotype (and, for ALT and bilirubin, also in GT carriers). In contrast, for ABCB1 c.1236C>T and c.3435C>T, lower AST and ALT levels were observed among CT and TT genotypes, while bilirubin levels tended to be lower in CT and higher in TT genotype carriers. For ABCB11 c.1331T>C, AST and ALT levels were lower in TC genotype carriers but higher in CC genotype, whereas bilirubin levels were higher in both TC and CC genotypes. However, these differences did not reach statistical significance, possibly because most of patients in the cohort had cirrhosis, a condition that independently elevates liver injury markers and may confound genotype-related effects. In addition, a higher frequency of HCC was observed among carriers of variant genotypes relative to wild-type individuals; nevertheless, given the small number of HCC cases in the sample (2.2%), this observation should be interpreted with caution and does not allow for definitive conclusions.

For the analysis by combined alleles, the alleles profile was categorised into two groups to optimise and facilitate understanding of the statistical analysis results, since there were 16 combinations. The first group comprised alleles carrying the variant form at all four loci (1236T and 2677T and 3435T and 1331C; TTTC); called of fully variant combined allele. The second group comprised the other possible combination of alleles; that is, alleles carrying at least one wild-type allele (1236C or 2677G or 3435C or 1331T) among the four polymorphisms studied. Consequently, we observed that the presence of the variant TTTC combined allele was significantly associated with a lower level of total cholesterol compared with the effects of the other alleles, in univariate and multivariate analysis. Corroborating the other results of this study that showed the importance of these polymorphisms as independent factors in the expression of the lipid profile and consequently the clinical evolution of patients with chronic HCV infection.

Moreover, we observed a higher serum level of AFP in the group with the fully variant combined allele (TTTC) in the univariate analysis, however, this association did not remain significant after adjustment for clinical confounders in the multivariate model. AFP is an important biomarker for diagnosing and monitoring the emergence of tumours, especially HCC. In patients with chronic HCV infection, especially with advanced stages of fibrosis/cirrhosis (present in 59.0% of our study group), the presence of elevated AFP may mean a greater risk of developing HCC, even after achieving SVR.³⁶ However, the loss of significance after adjustment suggests that the variant allele alone may not independently influence AFP levels, and that the observed elevation is more likely explained by the burden of liver disease rather than by a direct genetic effect. Nonetheless, the descriptive trend observed in the univariate analysis highlights the need for future studies with larger sample sizes and better stratification by fibrosis stage to further clarify whether specific genetic profiles could modulate AFP dynamics in chronic HCV infection.

This study highlights the roles of ABCB1 and ABCB11 polymorphisms in liver disease progression in patients with chronic HCV infection. The presence of these four SNPs was either directly (1331TT genotype) or indirectly (via lower total cholesterol and HDL and higher AST and AFP levels) related to the development of liver cirrhosis, which is a risk factor for HCC.

Cholestasis is characterised by the decrease or interruption of bile flow to the duodenum, where changes in alkaline phosphatase, GGT, and bilirubin levels can be observed.¹⁹ HCV infection can lead to intrahepatic cholestasis, and the presence of the c.1331T>C SNP is also a risk factor associated with several cholestatic diseases.^{19,20,21,22,23} In our study, we observed that individuals with the 1331CC variant genotype had higher levels of GGT and total bilirubin than those carrying the wild-type genotype 1331TT, albeit the difference was not statistically significant [Supplementary data (Table I)].

After the sample collection period, the patients received direct-acting antivirals (DAAs), and 99.3% achieved SVR. This high SVR rate suggests that the studied ABCB1 and ABCB11 genetic polymorphisms were not associated with impaired treatment response in the era of interferon-free regimens. From a public health perspective, broad access to DAA therapy must be implemented as a key strategy for achieving HCV elimination. However, although DAAs have been available since 2014, ensuring universal access for all HCV-infected individuals remains a major challenge, particularly in low- and middle-income countries without universal healthcare policies, due to the high cost of these medications. In addition, the challenge of HCV elimination also depends on expanded access to diagnostic testing, especially early diagnosis. Currently, most individuals are diagnosed only decades after infection, when the liver is already compromised with advanced fibrosis or cirrhosis.

Finally, these results support the well-established stability of SVR with DAA therapy, regardless of host-related factors, reinforcing the need to expand access to INF-free regimens independently of the genetic background of infected individuals. It is also important to emphasise that achieving SVR reduces but does not eliminate the risk of developing HCC in patients with advanced liver disease (fibrosis F3 and cirrhosis), given that chronic inflammation and tissue damage are already established. Therefore, continued clinical surveillance remains essential even after the SVR. In this context, the presence of ABCB1 and ABCB11 polymorphisms may contribute to this clinical outcome; however, further studies are needed to evaluate their impact in the post-SVR setting under DAA therapy.

One limitation of this study was the older age of patients. Given that the participants were unresponsive to conventional treatment (IFN and/or RBV), many would have been infected with HCV for many years. Thus, the results reflected the effects on an older cohort and may be different in a younger cohort. Moreover, because of this long period of chronic HCV infection, most patients had advanced liver diseases, such as grade 3 fibrosis and cirrhosis. Consequently, this limited our knowledge regarding these SNPs in individuals with mild or moderate fibrosis since the presence of these polymorphisms can influence faster disease evolution. In addition, due to the large number of variables analysed, our findings should be interpreted as hypothesis-generating and require confirmation in future studies with greater statistical power.

Finally, we recommended that the patients with advanced fibrosis and cirrhosis be stringently monitored, even after DAA treatment, owing to the risk of developing HCC despite achieving SVR. However, many of the individuals had discontinued follow-up with the medical team, mainly because of the Coronavirus disease 19 (COVID-19) pandemic that occurred between 2020 and 2023. This situation has made it challenging to access further information on their post-treatment clinical evolution.

In conclusion - Studies have shown that polymorphisms in the ABCB1 gene are associated with an increased risk of HCC and may predict significant changes in the pathological characteristics of the disease and influence patient response to anticancer treatment.^37,38,39 Polymorphisms in the ABCB11 gene can cause intrahepatic cholestasis and increased inflammation, negatively impacting liver cancer prognosis.⁴⁰ However, there is no consensus regarding the relationship between these polymorphisms and HCC.^41,42 Further studies are needed, especially with a focus on chronic HCV infection, because an SVR with DAAs does not exclude the risk of HCC in individuals with advanced cirrhosis, which must be continuously monitored.

To the best of our knowledge, this is the first study to evaluate the influence of ABCB1 and ABCB11 polymorphisms on the clinical and laboratory characteristics of patients with chronic HCV infection in Brazil. Given the known association of these SNPs with cholestatic damage and their impact on the efficacy and safety of antiviral drugs, it is important to establish their associated risks in the context of chronic HCV infection to understand their influence on the clinical evolution and clinical features of affected patients. Moreover, these genetic variants may serve as prognostic factors of liver damage.

SUPPLEMENTARY MATERIALS

Supplementary material

Supplementary material

ACKNOWLEDGEMENTS

To Ana Galha, always available and essential in the collection, processing, and organisation of the samples.

• Financial support: CAPES [No. 88887.623119/2021-00 (master's scholarship to Campos LB)], CNPq, FAPERJ.

DATA AVAILABILITY

The datasets generated and/or analysed during the current study cannot be made publicly available due to ethical and privacy restrictions involving human participants. However, the data is available from the corresponding author upon reasonable request.

REFERENCES

• ¹ WHO - World Health Organization. Hepatitis C. 2024 [cited 2025 Jul 8]. Available from: https://www.who.int/news-room/fact-sheets/detail/hepatitis-c
  » https://www.who.int/news-room/fact-sheets/detail/hepatitis-c
• ² Mysore KR, Leung DH. Hepatitis B and C. Clin Liver Dis. 2018; 22(4): 703-22.
• ³ Petruzziello A. Epidemiology of hepatitis B virus (HBV) and hepatitis C virus (HCV) related hepatocellular carcinoma. Open Virol J. 2018; 12: 26-32.
• ⁴ Calvaruso V, Craxì A. Implication of normal liver enzymes in liver disease. J Viral Hepat. 2009; 16(8): 529-36.
• ⁵ Akkiz H, Carr BI, Guerra V, Donghia R, Yalçın K, Karaoğullarından U, et al. Plasma lipids, tumor parameters and survival in HCC patients with HBV and HCV. J Transl Sci. 2021; 7(3): 10.15761.
• ⁶ Weigand K, Peschel G, Grimm J, Müller M, Höring M, Krautbauer S, et al. HCV infection and liver cirrhosis are associated with a less-favorable serum cholesteryl ester profile which improves through the successful treatment of HCV. Biomedicines. 2022; 10(12): 3152.
• ⁷ Trefts E, Gannon M, Wasserman DH. The liver. Curr Biol. 2017; 27(21): R1147-51.
• ⁸ Westbrook RH, Dusheiko G. Natural history of hepatitis C. J Hepatol. 2014; 61(Suppl. 1): S58-68.
• ⁹ Afsari A, Lee E, Shokrani B, Boortalary T, Sherif ZA, Nouraie M, et al. Clinical and pathological risk factors associated with liver fibrosis and steatosis in African-Americans with chronic hepatitis C. Dig Dis Sci. 2017; 62(8): 2159-65.
• ¹⁰ Kim TH, Yun SG, Choi J, Goh HG, Lee HA, Yim SY, et al. Differential impact of serum 25-Hydroxyvitamin D3 levels on the prognosis of patients with liver cirrhosis according to MELD and child-pugh scores. J Korean Med Sci. 2020; 35(19): e129.
• ¹¹ Trépo E, Ouziel R, Pradat P, Momozawa Y, Quertinmont E, Gervy C, et al. Marked 25-hydroxyvitamin D deficiency is associated with poor prognosis in patients with alcoholic liver disease. J Hepatol. 2013; 59(2): 344-50.
• ¹² Yousif MM, Sadek AMEM, Farrag HA, Selim FO, Hamed EF, Salama RI. Associated vitamin D deficiency is a risk factor for the complication of HCV-related liver cirrhosis including hepatic encephalopathy and spontaneous bacterial peritonitis. Intern Emerg Med. 2019; 14(5): 753-61.
• ¹³ Estrabaud E, Vidaud M, Marcellin P, Asselah T. Genomics and HCV infection: progression of fibrosis and treatment response. J Hepatol. 2012; 57(5): 1110-25.
• ¹⁴ Liu X. ABC family transporters. Adv Exp Med Biol. 2019; 1141: 13-100.
• ¹⁵ Fukunaga K, Nakagawa H, Ishikawa T, Kubo M, Mushiroda T. ABCB1 polymorphism is associated with atorvastatin-induced liver injury in Japanese population. BMC Genet. 2016; 17(1): 79.
• ¹⁶ Qu KK, Zhang CN, Dong LX, Wang SS, Zhang ZD, Zhang L. Association of ABCB1 polymorphisms with lipid homeostasis and liver injury response to atorvastatin in the Chinese population. Can J Physiol Pharmacol. 2020; 98(1): 15-22.
• ¹⁷ ElFayoumi RI, Hagras MM, Abozenadaha A, Gari M, Abosoudah I, Shinawi T, et al. The influence of polymorphisms in the drug transporter, ABCB1 on the toxicity of glucocorticoids in Saudi children with acute lymphoblastic leukaemia. Pharmacol Rep. 2019; 71(1): 90-5.
• ¹⁸ Fung KL, Gottesman MM. A synonymous polymorphism in a common MDR1 (ABCB1) haplotype shapes protein function. Biochim Biophys Acta. 2009; 1794(5): 860-71.
• ¹⁹ Telbisz Á, Homolya L. Recent advances in the exploration of the bile salt export pump (BSEP/ABCB11) function. Expert Opin Ther Targets. 2016; 20(4): 501-14.
• ²⁰ Hsu YC, Chen HL, Wu MZ, Liu YJ, Lee PH, Sheu JC, et al. Adult progressive intrahepatic cholestasis associated with genetic variations in ATP8B1 and ABCB11. Hepatol Res. 2009; 39(6): 625-31.
• ²¹ 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(4): 411-25.
• ²² Dixon PH, van Mil SW, Chambers J, Strautnieks S, Thompson RJ, Lammert F, et al. Contribution of variant alleles of ABCB11 to susceptibility to intrahepatic cholestasis of pregnancy. Gut. 2009; 58(4): 537-44.
• ²³ 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(1): 47-60.
• ²⁴ Timucin M, Alagozlu H, Ozdemir S, Ozdemir O. Association between ABCB1 (MDR1) gene polymorphism and unresponsiveness combined therapy in chronic hepatitis C virus. Hepat Mon. 2013; 13(4): e7522.
• ²⁵ Abdelkawy KS, El-Haggar SM, Ziada DH, Ebaid NF, El-Magd MA, Elbarbry FA. The effect of genetic variations on ribavirin pharmacokinetics and treatment response in HCV-4 Egyptian patients receiving sofosbuvir/daclatasvir and ribavirin. Biomed Pharmacother. 2020; 121: 109657.
• ²⁶ Lala V, Zubair M, Minter DA. Liver function tests. In: StatPearls. Treasure Island: StatPearls Publishing; 2026.
• ²⁷ Rosenfeld LG, Malta DC, Szwarcwald CL, Bacal NS, Cuder MAM, Pereira CA, et al. Reference values for blood count laboratory tests in the Brazilian adult population, National Health Survey. Rev Bras Epidemiol. 2019; 22(Suppl. 02): E190003.SUPL.2.
• ²⁸ MS/SVS - Ministério da Saúde/Secretaria de Vigilância em Saúde. Boletim epidemiológico de hepatites virais 2024. Brasília: Ministério da Saúde; 2024 [cited 2025 Jul 8]. Available from: https://www.gov.br/aids/pt-br/central-de-conteudo/boletins-epidemiologicos/2024/boletim-epidemiologico-hepatites-virais-2024/view
  » https://www.gov.br/aids/pt-br/central-de-conteudo/boletins-epidemiologicos/2024/boletim-epidemiologico-hepatites-virais-2024/view
• ²⁹ Campos LB, de Almeida NAA, de Santana CG, Barbosa ENP, Horta MAP, Amendola Pires M, et al. Before direct-acting antivirals for hepatitis C virus: evaluation of core protein R70Q and L/C91M substitutions in chronically infected Brazilian patients unresponsive to IFN and/or RBV. Viruses. 2023; 15(1): 187.
• ³⁰ Zavaglia C, Silini E, Mangia A, Airoldi A, Piazzolla V, Vangeli M, et al. Prognostic factors of hepatic decompensation and hepatocellular carcinoma in patients with transfusion-acquired HCV infection. Liver Int. 2014; 34(7): e308-16.
• ³¹ Lei JH, Yang X, Xiao XQ, Chen Z, Peng F. A preliminary investigation on single nucleotide polymorphism rs2287622 of bile salt export pump gene in patients with chronic hepatitis C virus infection in Hunan, China. BMC Gastroenterol. 2017; 17(1): 42.
• ³² Iwata R, Baur K, Stieger B, Mertens JC, Daly AK, Frei P, et al. A common polymorphism in the ABCB11 gene is associated with advanced fibrosis in hepatitis C but not in non-alcoholic fatty liver disease. Clin Sci (Lond). 2011; 120(7): 287-96.
• ³³ Besheer T, Arafa M, El-Maksoud MA, Elalfy H, Hasson A, Zalata K, et al. Diagnosis of cirrhosis in patients with chronic hepatitis C genotype 4: role of ABCB11 genotype polymorphism and plasma bile acid levels. Turk J Gastroenterol. 2018; 29(3): 299-307.
• ³⁴ De Mattia E, Cecchin E, Polesel J, Bignucolo A, Roncato R, Lupo F, et al. Genetic biomarkers for hepatocellular cancer risk in a caucasian population. World J Gastroenterol. 2017; 23(36): 6674-84.
• ³⁵ Carr BI, Akkiz H, Bag HG, Karaoğullarından U, Yalçın K, Ekin N, et al. Serum levels of gamma-glutamyl transpeptidase in relation to HCC human biology and prognosis. J Transl Sci. 2021; 7(3): 10.15761/jts.
• ³⁶ Semmler G, Meyer EL, Kozbial K, Schwabl P, Hametner-Schreil S, Zanetto A, et al. HCC risk stratification after cure of hepatitis C in patients with compensated advanced chronic liver disease. J Hepatol. 2022; 76(4): 812-21.
• ³⁷ Wang ZC, Liu LZ, Liu XY, Hu JJ, Wu YN, Shi JY, et al. Genetic polymorphisms of the multidrug resistance 1 gene MDR1 and the risk of hepatocellular carcinoma. Tumour Biol. 2015; 36(9): 7007-15.
• ³⁸ Baldissera VD, de Mattos AA, Coral GP, de Araujo FB, Marroni CA, Brandão ABM, et al. Evaluation of the C3435T polymorphism in the MDR1 gene in patients with hepatocellular carcinoma. Ann Hepatol. 2012; 11(6): 899-906.
• ³⁹ Hu W, Huang S, Dong L, Yu C, Li C, Zhang J. MDR1 gene polymorphism correlated with pathological characteristics and prognosis in patients with primary hepatocellular carcinoma receiving interventional therapy. Anticancer Drugs. 2019; 30(3): 233-40.
• ⁴⁰ Vitale G, Mattiaccio A, Conti A, Turco L, Seri M, Piscaglia F, et al. Genetics in familial intrahepatic cholestasis: clinical patterns and development of liver and biliary cancers: a review of the literature. Cancers (Basel). 2022; 14(14): 3421.
• ⁴¹ Okubo H, Ando H, Ishizuka K, Morishige JI, Ikejima K, Shiina S, et al. Impact of genetic polymorphisms on the pharmacokinetics and pharmacodynamics of lenvatinib in patients with hepatocellular carcinoma. J Pharmacol Sci. 2022; 148(1): 6-13.
• ⁴² Chang Q, He ZL, Peng YC, Duan SG, Dai YX, Zhao XH. A meta-analysis of MDR1 polymorphisms rs1128503 and rs1045642 and susceptibility to hepatocellular carcinoma. J Int Med Res. 2019; 47(7): 2800-9.

Edited by

• Handling editor:
  Marcelo Ribeiro-Alves | https://orcid.org/0000-0002-8663-3364

FIRST REVIEW ROUND - REVIEWERS' COMMENTS

About the reviewer

REVIEWER #1

a) Adequacy of the abstract;

Considering the elements highlighted in the abstract's structure, the content appears appropriate and accurately reflects the methodology and results described in the study

b) Originality and importance of the contribution for the development of the field of study. And relevance of:

Regarding originality, I note that the topic has been widely discussed in literature, particularly in relation to genetic polymorphisms in ABC genes and their impact on Hepatitis C severity. However, this study brings a relevant contribution by analyzing a Brazilian cohort and correlating the polymorphism data with clinical and laboratory findings. The results are thoroughly discussed, including the study's limitations. I consider this work relevant and of interest to the field.

c) Methodology, results and discussion;

In terms of methodology, the study presents a well-defined and clearly described approach. The results of the polymorphism analysis using the PCR technique are thoroughly explored, both in the narrative and in the accompanying figures.

The discussion of the results is well developed, as it effectively correlates the data obtained from the different haplotypes evaluated in the study with the laboratory findings related to lipid profiles and hepatic parameters, including alpha-fetoprotein as a biomarker for hepatocellular carcinoma.

However, in the discussion of the results, it is not clear why the ABCB1 gene SNP -3435C>T was not included in Table IV. If no significant difference was observed, this should have been explicitly stated in the text. I also believe that the data presented in Table III should be discussed in greater depth. Furthermore, Supplementary Table 1 provides material for several additional analyses that were not explored in the manuscript.

d) References;

OK.

e) Figures and tables.

It is recommended that the authors verify the formatting of the tables, as some (e.g., Table IV) appear to be cropped in the version provided to the reviewers, limiting proper visualization.

REVIEWER #2

The manuscript "Association of p-glycoprotein and bile salt export pump gene polymorphisms with advanced liver disease in hepatitis C virus infected patients" reports an observational retrospective study aimed at investigating the impact of ABCB1 and ABCB11 SNPs on disease progression in chronic HCV patients. Unfortunately, the manuscript presents important limitations in both study design and statistical analysis. Below, I list major comments that should be addressed before the manuscript can be considered for publication.

Study design: The description of the study design is insufficient. The primary outcome is not clearly defined, and the central hypothesis remains unclear. The recruitment of 232 HCV-infected participants and the extraction of clinical and demographic data from medical records are described, but the rationale for the comparisons performed is not adequately justified. The decision to compare ABCB1 and ABCB11 genotype frequencies across all extracted variables, including demographic factors such as sex and age, needs to be supported by a clear analytical framework.

2. Statistical analysis / Results:

Table III: HWE analysis must be performed using absolute genotype counts (observed and expected).

Table IV: Multivariate models should be applied to control for confounding variables. For example, cholesterol levels are influenced by factors such as age and sex. It should be made clear whether the reported associations remain significant after adjustment. Additionally, the table should be presented in landscape format, as some columns were hidden in the PDF version and could not be reviewed.

Table V represents a "combination of genotypes", and not a haplotype analysis. The authors compared individuals homozygous for all alternative alleles with those carrying at least one wild-type genotype. In fact, as ABCB1 and ABCB11 are located on different chromosomes, their variants cannot be combined in haplotypes.

Editorial Decision and Recommendations

Thank you for submitting your manuscript "Association of p-glycoprotein and bile salt export pump gene polymorphisms with advanced liver disease in hepatitis C virus infected patients" to Memórias do Instituto Oswaldo Cruz. I have carefully evaluated the manuscript along with the reviewers' comments. While the study addresses an important topic in hepatology and genetics, several significant concerns must be addressed before the manuscript can be considered for publication.

Endorsement and Expansion of Reviewer Comments

I fully endorse the reviewers' assessments and would like to reinforce and expand on several critical points:

Study Design and Methodological Concerns

The reviewers correctly identified that the study design lacks sufficient clarity regarding the primary outcome and central hypothesis. The manuscript states it aims "to investigate the impact of ABCB1 and ABCB11 SNPs on disease progression in chronic HCV patients," but fails to:

Clearly define what constitutes "disease progression" as the primary endpoint

Justify the selection of specific variables for analysis among the many clinical parameters collected

Present a clear analytical framework that would explain why certain variables were prioritized over others

The retrospective nature of the study is acceptable, but the authors must explicitly state whether this was a hypothesis-driven investigation or an exploratory analysis. If exploratory, appropriate corrections for multiple comparisons must be applied and clearly documented.

Statistical Analysis Limitations

Reviewer 2 correctly identified several statistical issues that require attention:

Hardy-Weinberg Equilibrium analysis: Table III should present HWE analysis using absolute genotype counts (observed vs. expected), not percentages. The current presentation obscures the actual statistical test performed.

Lack of multivariate analysis: The univariate analyses presented in Tables IV and V cannot establish independent associations between the SNPs and clinical parameters. Given that factors such as age, sex, BMI, and diabetes status significantly influence lipid profiles and liver parameters, multivariate regression models controlling for these confounders are essential. For example, the association between 2677TT genotype and lower cholesterol levels may simply reflect age or diabetes differences between genotype groups rather than a direct genetic effect.

Multiple comparisons issue: With the large number of variables analyzed (28+ parameters across multiple SNPs), the risk of type I errors (false positives) is substantial. The authors must either apply appropriate corrections (e.g., Bonferroni, Benjamini-Hochberg) or explicitly acknowledge this limitation and frame their findings as hypothesis-generating rather than conclusive.

Misleading "haplotype" terminology: Reviewer 2 is absolutely correct that Table V does not represent a true haplotype analysis. Since ABCB1 (chromosome 7) and ABCB11 (chromosome 2) are located on different chromosomes, they segregate independently and cannot form haplotypes. The authors should reframe this as a "combined genotype analysis" or "genetic risk score" rather than using the technically incorrect term "haplotype."

Presentation and Interpretation Issues

Missing SNP in Table IV: As Reviewer 1 noted, the c.3435C>T polymorphism is discussed in the text but absent from Table IV without explanation. If no significant associations were found, this should be explicitly stated.

Clinical relevance considerations: The manuscript notes that "after the sample collection period, the patients received direct-acting antivirals (DAAs), and 99.3% achieved SVR." This important finding significantly impacts the interpretation of results, as it suggests these polymorphisms did not affect response to modern therapy. The authors should more thoroughly discuss the clinical relevance of their findings in the context of current DAA-based treatment paradigms.

Table formatting: Several tables (particularly Table IV) appear cropped in the current version, making proper evaluation difficult. All tables should be reformatted for clarity, with Table IV presented in landscape orientation as suggested.

AUTHORS' RESPONSE TO THE REVIEWERS

Manuscript ID: MIOC-2025-0173

Title: Association of p-glycoprotein and bile salt export pump gene polymorphisms with advanced liver disease in hepatitis C virus infected patients

We would like to thank you for your valuable contributions, constructive suggestions and detailed comments that significantly enriched the quality of this work. We inform you that all requests and recommendations were met, and the main changes are highlighted in yellow in the manuscript.

Reviewer #1

Reviewer comments:

a) Adequacy of the abstract: Considering the elements highlighted in the abstract's structure, the content appears appropriate and accurately reflects the methodology and results described in the study.

Response: We appreciate the reviewer's positive assessment. We are pleased that the abstract was considered appropriate and consistent with the study's methodology and results.

b) Originality and importance of the contribution for the development of the field of study. And relevance of: Regarding originality, I note that the topic has been widely discussed in literature, particularly in relation to genetic polymorphisms in ABC genes and their impact on Hepatitis C severity. However, this study brings a relevant contribution by analyzing a Brazilian cohort and correlating the polymorphism data with clinical and laboratory findings. The results are thoroughly discussed, including the study's limitations. I consider this work relevant and of interest to the field.

Response: We thank the reviewer for recognizing the relevance and contribution of our study.

c) Methodology, results and discussion: In terms of methodology, the study presents a well-defined and clearly described approach. The results of the polymorphism analysis using the PCR technique are thoroughly explored, both in the narrative and in the accompanying figures. The discussion of the results is well developed, as it effectively correlates the data obtained from the different haplotypes evaluated in the study with the laboratory findings related to lipid profiles and hepatic parameters, including alpha fetoprotein as a biomarker for hepatocellular carcinoma. However, in the discussion of the results, it is not clear why the ABCB1 gene SNP -3435C>T was not included in Table IV. If no significant difference was observed, this should have been explicitly stated in the text. I also believe that the data presented in Table III should be discussed in greater depth. Furthermore, Supplementary Table 1 provides material for several additional analyses that were not explored in the manuscript.

Response: We thank the reviewer for the constructive feedback and positive comments on our methodology and discussion. The ABCB1 3435C>T was not included in Table IV because no statistically significant associations were identified. To improve clarity, we have now explicitly stated this information in the Results section, and we clarify in the table legend that Table IV presents only variables and SNPs with significant differences. In Results: "Table IV includes only the variables that showed statistically significant differences according to genotype or allele. In addition, only the SNPs that presented significant associations are displayed in this table, specifically ABCB1 2677G>T/A, ABCB1 1236C>T, and ABCB11 1331T>C." And in Table IV Legend: "No significant results were found for the analysis by allele for the c.1331T>C polymorphism; therefore, this SNP was not included in Table IV." In the revised manuscript, we also have expanded the discussion of the results presented in Table III (Hardy-Weinberg equilibrium). Discussion: "All variants were in Hardy–Weinberg equilibrium, indicating that the observed genotype frequencies were consistent with those expected for a genetically stable population. This supports the reliability of the genotyping procedures and suggests that no population stratification or methodological bias significantly affected the distribution of these SNPs, thereby reinforcing the validity of the association analyses performed." In the revised manuscript, we included in discussion more exploratory analysis with the data presented in supplementary table S1.

d) References: OK

Response: We appreciate the reviewer's positive evaluation regarding the references.

e) Figures and tables: It is recommended that the authors verify the formatting of the tables, as some (e.g., Table IV) appear to be cropped in the version provided to the reviewers, limiting proper visualization.

Response: We thank the reviewer for noting this issue. We have carefully reviewed and corrected the formatting of all tables in the revised version.

Reviewer #2

Reviewer comments:

The manuscript "Association of p-glycoprotein and bile salt export pump gene polymorphisms with advanced liver disease in hepatitis C virus infected patients" reports an observational retrospective study aimed at investigating the impact of ABCB1 and ABCB11 SNPs on disease progression in chronic HCV patients. Unfortunately, the manuscript presents important limitations in both study design and statistical analysis. Below, I list major comments that should be addressed before the manuscript can be considered for publication.

Response: We appreciate the constructive comments and have revised the manuscript accordingly to address the reviewer's concerns. In the following sections, we provide detailed responses and indicate the specific changes made to the text.

Study design: The description of the study design is insufficient. The primary outcome is not clearly defined, and the central hypothesis remains unclear. The recruitment of 232 HCV-infected participants and the extraction of clinical and demographic data from medical records are described, but the rationale for the comparisons performed is not adequately justified. The decision to compare ABCB1 and ABCB11 genotype frequencies across all extracted variables, including demographic factors such as sex and age, needs to be supported by a clear analytical framework.

Response: We thank the reviewers for their thorough evaluation and constructive comments regarding the study design. In the revised manuscript, we have clarified both the central hypothesis and the analytical framework guiding the study. First, we now explicitly define "disease progression" as our primary outcome, operationalized as the presence of advanced fibrosis (F3) or cirrhosis. Second, we have expanded the Methods section to clearly state that the study was hypothesis-driven, based on prior evidence suggesting that ABCB1 and ABCB11 polymorphisms may influence hepatic injury. Accordingly, we justified the selection of the clinical and laboratory variables analyzed as secondary outcomes, as these parameters (e.g., ALT, AST, GGT, AFP, lipid profile) are biologically linked to liver damage and metabolic alterations. We also justified the comparisons between genetic polymorphisms and demographic variables. This analysis was performed because these variables are well established confounders in liver disease progression and secondary outcomes. We revised the manuscript to present a more explicit analytical framework, explaining the rationale for prioritizing specific variables and how these relate to the functional consequences of the polymorphisms studied. Finally, to address the reviewer's concern about the number of statistical comparisons, we now acknowledge this limitation and describe the approach used.

2. Statistical analysis / Results: Table III: HWE analysis must be performed using absolute genotype counts (observed and expected).

Response: We thank the reviewer for this important observation. In the revised manuscript, Table III has been updated to present the Hardy–Weinberg Equilibrium (HWE) analysis using absolute genotype counts (both observed and expected), as recommended. The chi-square values and corresponding p-values have also been included to ensure full transparency of the statistical test performed.

Table IV: Multivariate models should be applied to control for confounding variables. For example, cholesterol levels are influenced by factors such as age and sex. It should be made clear whether the reported associations remain significant after adjustment. Additionally, the table should be presented in landscape format, as some columns were hidden in the PDF version and could not be reviewed.

Response: We thank the reviewer for this very important observation and suggestion. In the revised manuscript, we performed a multivariable analysis that included the relevant confounding variables. The methodology used for this analysis is specified in the Methods section. The results of the multivariate analysis are described in the Results section, under Statistical Analysis, and are presented in Tables IV and V. Finally, we reformatted Table IV in landscape orientation to prevent loss of information and improve readability.

Table V represents a "combination of genotypes", and not a haplotype analysis. The authors compared individuals homozygous for all alternative alleles with those carrying at least one wild-type genotype. In fact, as ABCB1 and ABCB11 are located on different chromosomes, their variants cannot be combined in haplotypes.

Response: We thank the reviewer for this important clarification. We agree that Table V does not represent a true haplotype analysis, since ABCB1 (chromosome 7) and ABCB11 (chromosome 2) segregate independently. Accordingly, and following the reviewer's suggestion, we have replaced the term "haplotype analysis" with "combined genotype analysis" throughout the manuscript, including in the Abstract, Results, Discussion, and in Table V.

• peer review recommendation: accept

History

• Received
  12 July 2025
• Accepted
  19 Mar 2026

REVIEWERS' COMMENTS

About the reviewer

REVIEWER #1

Dear Editor, upon careful consideration of the comments provided by Reviewer 2, I identified several aspects that had not been fully addressed in my initial assessment. These points of revision clearly contribute to improving data transparency and strengthening the overall robustness of the manuscript. In conclusion, after addressing all comments, the authors have incorporated the suggested modifications, and I consider the manuscript suitable for publication.

REVIEWER #2

The manuscript entitled "Association of p-glycoprotein and bile salt export pump gene polymorphisms with advanced liver disease in hepatitis C virus infected patients" is relevant, since genetic association analyses with the presented outcome (and other related phenotypes) are still scarce for the Brazilian population.

After reviewing the latest version, I agreed with all the reviewers' and editor's comments. However, I will add some suggestions and have some questions that I will point out below.

Review all gene names (abbreviations): italicized.

The word mutation/mutated is no longer commonly used when discussing association studies. The most commonly used term is variant or genetic variant (including for monogenic traits), although the term polymorphism is still used. I suggest reviewing these terms throughout the text, tables, and figures.

Check where the term "population" appears. "Sample" or "cohort" would be better.

The main point that needs to be reviewed relates to the statistical analysis of association. Depending on how the analyses were performed, there will be changes in the writing of the abstract, results and discussion, as well as the tables.

In Table IV, it was not clear how the means/standard deviations were obtained. Were heterozygotes counted in both groups? I suggest performing an analysis using dominant and recessive models, to always have a group as a reference and not repeat heterozygotes in the analyses.

In the analysis of combined genotypes, it was not clear in the table how this combination of genotypes occurred. For example, see Discussion: "The first group comprised individuals with a 100% mutant genotype (TTTC); that is, those who carried the mutant genotype of all four SNPs (1236TT, 2677TT, 3435TT and 1331CC). The second group comprised patients with other possible combinations of genotypes; that is, individuals who had the wild-type genotype for at least one of the SNPs studied." If this is what was done, N should be at most 21 (due to the lower frequency of the TT genotype for G2677T). Table V (n=88 for 100% mutant genotype) suggests that an analysis was performed regarding the presence of at least one mutant "allele" (TTTC). In this case, how were the averages calculated for the two groups, given that we are talking about the number of "alleles" (n=88 and 376) and not the number of individuals?

Table IV: For the polymorphism c.1331T>C, the numbers (N) are wrong: N=270 is in relation to the 1331C allele and N=194 is in relation to the 1331T allele.

Legend: "No significant results were found for the analysis by allele for c.1331T>C..." Shouldn't the SNP be c.3435C>T? On the third line: "The result presented is for comparison with the analysis by genotype." This sentence should be deleted because allele analyses were also performed.

AUTHORS' RESPONSE TO THE REVIEWERS

Name of Journal: Memórias do Instituto Oswaldo Cruz Manuscript ID: MIOC-2025-0173 Title: Association of p-glycoprotein and bile salt export pump gene polymorphisms with advanced liver disease in hepatitis C virus infected patients We would like to thank the reviewer for the valuable contributions, constructive suggestions, and careful evaluation of the manuscript, which significantly improved its clarity, transparency, and overall quality. We are pleased to inform you that all comments and recommendations have been fully addressed in the revised version of the manuscript, and the main changes are highlighted (Text Highlight Color) in yellow.

Reviewer #1

Reviewer comment: Dear Editor, upon careful consideration of the comments provided by Reviewer 2, I identified several aspects that had not been fully addressed in my initial assessment. These points of revision clearly contribute to improving data transparency and strengthening the overall robustness of the manuscript. In conclusion, after addressing all comments, the authors have incorporated the suggested modifications, and I consider the manuscript suitable for publication.

Response: We thank the reviewer for the careful re-evaluation of our manuscript and for considering the additional points raised by Reviewer 2. We appreciate the recognition that addressing these comments contributed to improving data transparency and strengthening the overall robustness of the study. We are grateful for the positive assessment and for considering the revised manuscript suitable for publication after the incorporation of the suggested modifications.

Reviewer #2

Reviewer comments: The manuscript entitled "Association of p-glycoprotein and bile salt export pump gene polymorphisms with advanced liver disease in hepatitis C virus infected patients" is relevant, since genetic association analyses with the presented outcome (and other related phenotypes) are still scarce for the Brazilian population. After reviewing the latest version, I agreed with all the reviewers' and editor's comments. However, I will add some suggestions and have some questions that I will point out below.

a) Review all gene names (abbreviations): italicized.

Response: We thank the reviewer for the careful observation regarding gene nomenclature. All gene names (ABCB1 and ABCB11) were reviewed and are now consistently formatted in italics throughout the manuscript, following international standards. All suggested alterations were incorporated throughout the manuscript.

b) The word mutation/mutated is no longer commonly used when discussing association studies. The most commonly used term is variant or genetic variant (including for monogenic traits), although the term polymorphism is still used. I suggest reviewing these terms throughout the text, tables, and figures.

Response: Throughout the manuscript, the terms "mutation" and "mutant" were replaced by "variant", "variant genotype" or polymorphism, in accordance with current recommendations for genetic association studies. All suggested alterations were incorporated throughout the manuscript.

c) Check where the term "population" appears. "Sample" or "cohort" would be better.

Response: We thank the reviewer for this important suggestion. As recommended, the terminology "population" was carefully revised and replaced throughout the manuscript with "study sample" or "cohort," as appropriate, to better reflect the nature of our study. All suggested alterations were incorporated throughout the manuscript.

d) The main point that needs to be reviewed relates to the statistical analysis of association. Depending on how the analyses were performed, there will be changes in the writing of the abstract, results and discussion, as well as the tables.

Response: The manuscript was carefully revised to standardize and clarify the terminology used to describe genetic association analyses. The text now consistently distinguishes between genotype- and allele-based analyses and avoids ambiguous terminology. Corresponding revisions were implemented throughout the Abstract, Results, Discussion, and tables to ensure full coherence between the statistical analyses performed and their interpretation. All suggested alterations were incorporated throughout the manuscript.

e) In Table IV, it was not clear how the means/standard deviations were obtained. Were heterozygotes counted in both groups? I suggest performing an analysis using dominant and recessive models, to always have a group as a reference and not repeat heterozygotes in the analyses.

Response: We thank the reviewer for this important comment regarding the calculation of means and standard deviations and the analytical approach used in article. In the genotype-based analysis, each patient was considered as a single observational unit and categorized according to their genotype. The wild-type genotype was used as the reference group, and comparisons were performed between wild-type versus heterozygous and wild-type versus mutant genotypes. In this approach, means and standard deviations were calculated per individual, and heterozygotes were not counted more than once. In the allele-based analysis, the unit of analysis was the allele rather than the individual. Accordingly, the dataset was expanded so that each individual contributed two alleles, allowing comparisons between wild-type and variant alleles. In this case, means and standard deviations were calculated based on allele counts, and not on the number of patients. We have revised the Methods section and the table legends to explicitly describe these analytical strategies and clarify how summary statistics were obtained, thereby ensuring transparency and avoiding misinterpretation.

f) In the analysis of combined genotypes, it was not clear in the table how this combination of genotypes occurred.

Response: We thank the reviewer for this important observation. We acknowledge that the terminology used in the previous version was inaccurate. The analysis presented does not correspond to a combined genotype (individual-based) analysis, but rather to a combined allele analysis. In this analysis, the unit of observation was the allele, not the individual. Accordingly, the groups were defined as follows: Group 1: alleles carrying the variant form for all four polymorphisms studied (1236T and 2677T and 3435T and 1331C); Group 2: alleles presenting at least one wild-type allele among the four loci (1236C or 2677G or 3435C or 1331T). We have corrected the terminology throughout the manuscript, tables, and figure legends, replacing "combined genotype analysis" with "combined allele analysis", and we have clarified the definition of the comparison groups to avoid further misunderstanding.

g) ABSTRACT: The sentences "Most frequent ABCB1 SNP was c.3435C>T (TT; 15.5%), followed by c.1236C>T (TT; 11.2%) and c.2677G>T (TT; 9.0%). ABCB11 c.1331T>C was found in 31.3%." should be revised.

Response: We thank the reviewer for the suggestion. The Abstract was revised accordingly, and the frequencies were corrected to report variant alleles as recommended: "The most frequent ABCB1 variant allele was c.3435C>T (36.8%), followed by c.1236C>T (30.8%) and c.2677G>T (26.7%). The ABCB11 c.1331CC genotype was observed in 31.3% of the cohort."

h) Sponsorships: "Concelho": CONSELHO.

Response: We thank the reviewer for pointing out this typographical error. The word 'Concelho' has been corrected to 'CONSELHO' in the revised manuscript.

i) INTRODUCTION: In some locations c.2677G>T, in others c.2677G>T/A.

Response: We thank the reviewer for pointing out this inconsistency. The entire manuscript was carefully reviewed, and the nomenclature was standardized to c.2677G>T throughout. All suggested alterations were incorporated throughout the manuscript.

j) MATERIAL AND METHODS: Suggestion: "Extraction and analysis of the ABCB1 and ABCB11 genes" should be changed to "DNA Extraction and analysis of the ABCB1 and ABCB11 genes". Wouldn't the kit used for extraction be the QIAamp DNA Mini Kit, and not the QIAamp RNA Mini Kit?

Response: We thank the reviewer for this important observation. The extraction kit used in this study does not separate genetic material by type (RNA versus DNA); therefore, it allows the simultaneous extraction of both nucleic acids from the same biological sample. As hepatitis C virus is an RNA virus and constitutes the basis of this study, the QIAamp RNA Mini Kit was intentionally used, as it enables downstream analyses of viral RNA as well as host genomic DNA from the same extracted material. According to the manufacturer (QIAGEN), this kit is suitable for the extraction of both RNA and DNA. In addition, we have revised the wording as suggested, and the section now reads "Nucleic Acid extraction and analysis of the ABCB1 and ABCB11 genes" in the revised manuscript.

k) Suggestion: "genes was performed using real-time polymerase chain reaction with the TaqMan Universal PCR Master Reagent Mix (Applied Biosystems, Foster City, CA, United States) and the TaqMan SNP Genotyping Assay kit (Applied Biosystems) for allele discrimination for each polymorphism, according to the manufacturer's instructions."

Response: We thank the reviewer for the suggestion. The wording was revised as recommended.

l) Table 1 is not necessary and could be a supplementary table, with the assay number for each polymorphism.

Response: We thank the reviewer for this valuable suggestion. The manuscript was revised accordingly: Table 1 was removed from the main text and included as Supplementary Table S1.

m) RESULTS: Suggestion: "Statistical analysis" for "Association analysis". Page 10: Line 29: "Furthermore, significantly lower levels of HDL (P < 0.05) were associated with the 2677TT mutation and 2677T allele." To: "Furthermore, significantly lower levels of HDL (P < 0.05) were associated with the 2677TT GENOTYPE and 2677T allele."

Response: We thank the reviewer for the suggestion. The section title was revised from "Statistical analysis" to "Association analysis," and the wording was corrected to replace "mutation" with "genotype," as recommended.

t) Table II: Remove "n (%)" from the title. Not all variables are represented by "n (%)", some, for example albumin, are represented by the mean +/- standard deviation. Therefore, the second column of the table should not have the title "Frequency".

Response: We thank the reviewer for this suggestion. The term "Frequency" was replaced with "Value." The table legend was revised to clarify that data are presented as n (%) or mean ± standard deviation, as appropriate.

u) Table III: Not necessary in the manuscript. Normally in this type of study, only Hardy-Weinberg equilibrium is mentioned in the text.

Response: We thank the reviewer for this comment. To improve transparency, Table III was not removed but relocated to the Supplementary Material (Table S3). The main results regarding Hardy–Weinberg equilibrium were maintained and clearly described in the main text.

v) Table IV: For the polymorphism c.1331T>C, the numbers (N) are wrong: N=270 is in relation to the 1331C allele and N=194 is in relation to the 1331T allele. Legend: "No significant results were found for the analysis by allele for c.1331T>C..." Shouldn't the SNP be c.3435C>T?

Response: We thank the reviewer for these important observations. The allele frequencies for the c.1331T>C polymorphism, which were previously reversed, have been corrected in Table IV. In addition, the suggested changes to the legend were implemented: the SNP notation was revised, and the sentence referring to comparison with genotype analysis was removed.

• peer review recommendation: accept

History

• Received
  12 July 2025
• Accepted
  19 Mar 2026

EDITORIAL DECISION AND RECOMMENDATIONS

About the reviewer

Editorial Decision and Recommendations

Dear Authors,

Thank you for the opportunity to review the revised version of your manuscript (MIOC-2025-0173.R2). I would like to commend you for the substantial effort put into addressing the previous comments. The clarification of the study design—specifically the explicit definition of the primary outcome (advanced fibrosis/cirrhosis) versus secondary outcomes—and the distinction between hypothesis-driven and exploratory components have significantly strengthened the manuscript's methodological framework. Furthermore, the correction of the "haplotype" terminology to "combined allele analysis" and the expanded discussion on the clinical relevance in the context of DAA therapy are excellent improvements that enhance the scientific accuracy and impact of the study.

However, upon careful comparison of the revised text with the specific concerns raised in the First Round of Review, there are two critical points that remain partially unaddressed. To ensure the manuscript meets the high standards required for publication, I request that you address the following:

1. Explicit Statement Regarding the c.3435C>T Polymorphism Reviewer 1 (Round 1) Comment c) noted that it was unclear why the ABCB1 gene SNP -3435C>T was not included in Table IV. The reviewer specifically requested that if no significant difference was observed, this should be explicitly stated in the text.

Current Status: In the revised Results and Abstract, you report associations for c.1236C>T, c.2677G>T, and c.1331T>C, as well as the combined allele analysis. However, the c.3435C>T polymorphism is conspicuously absent from the summary of findings without an explanation.

Recommendation: Please add a sentence in the Results section and/or a footnote in the table explicitly addressing this SNP.

Suggested addition (Results): "Regarding the ABCB1 c.3435C>T polymorphism, no statistically significant associations were observed with any of the clinical or laboratory parameters evaluated (data not shown)."

2. Presentation of Multivariate Analysis Results Reviewer 2 (Round 1) and the Editorial Decision strongly emphasized the necessity of multivariate regression models to control for confounding variables (age, sex, BMI, diabetes) and requested that the results be presented to show if associations remain significant after adjustment.

Recommendation: To fully satisfy the reviewer's concern regarding confounding factors, you must report the results of the multivariate analysis for the significant findings. Table IV: Please update Table IV to include columns for the multivariate analysis (e.g., adjusted Odds Ratios, Beta coefficients, and P-values) alongside the descriptive statistics. Alternatively, if Table IV is intended for descriptive data only, create a new table for the adjusted associations. Text Revision: Ensure the Results text specifies whether the associations held up after adjustment.

Suggested revision (Results): "Genotypes 1236TT and 2677TT and their alleles were associated with lower total cholesterol even after adjusting for age, sex, BMI, and diabetes status (P < 0.05)." (If the adjustment changed the significance, this must also be stated).

By addressing these two points, you will provide the transparency regarding negative findings (3435C>T) and the statistical robustness (multivariate adjustment) that were requested in the initial review.

Best regards,

Dr. Marcelo Ribeiro Alves

Handling Editor

Memórias do Instituto Oswaldo Cruz

AUTHORS' RESPONSE TO THE REVIEWERS

Name of Journal: Memórias do Instituto Oswaldo Cruz Manuscript ID: MIOC-2025-0173 Title: Association of p-glycoprotein and bile salt export pump gene polymorphisms with advanced liver disease in hepatitis C virus infected patients

Dear Dr. Marcelo Ribeiro Alves,

We sincerely thank you for the careful handling of our manuscript and for coordinating the review process. We are also grateful to the reviewers for their valuable contributions, constructive suggestions, and thorough evaluation, which significantly improved the clarity, transparency, and overall quality of our work. We are pleased to inform you that all comments and recommendations have been carefully addressed in the revised version of the manuscript.

Comments: Dear Authors, Thank you for the opportunity to review the revised version of your manuscript (MIOC-2025-0173.R2). I would like to commend you for the substantial effort put into addressing the previous comments. The clarification of the study design—specifically the explicit definition of the primary outcome (advanced fibrosis/cirrhosis) versus secondary outcomes—and the distinction between hypothesis-driven and exploratory components have significantly strengthened the manuscript's methodological framework. Furthermore, the correction of the "haplotype" terminology to "combined allele analysis" and the expanded discussion on the clinical relevance in the context of DAA therapy are excellent improvements that enhance the scientific accuracy and impact of the study. However, upon careful comparison of the revised text with the specific concerns raised in the First Round of Review, there are two critical points that remain partially unaddressed. To ensure the manuscript meets the high standards required for publication, I request that you address the following:

1. Explicit Statement Regarding the c.3435C>T Polymorphism Reviewer 1 (Round 1) Comment c) noted that it was unclear why the ABCB1 gene SNP -3435C>T was not included in Table IV. The reviewer specifically requested that if no significant difference was observed, this should be explicitly stated in the text.

Response: We thank you for highlighting the importance of ensuring full transparency regarding the ABCB1 c.3435C>T polymorphism. We would like to clarify that, in the previous revision, we had already specified that this SNP was not included in the main table because no statistically significant associations were observed. This information was provided both in the Results section and in the legend of Table II. To further enhance clarity and avoid any possible ambiguity, we have now added the following explicit sentence to the Results section: "Regarding the ABCB1 c.3435C>T polymorphism, no statistically significant associations were observed with any of the clinical or laboratory parameters evaluated (data not shown)." We believe this addition ensures complete transparency regarding the negative findings and fully addresses the concern raised.

2. Presentation of Multivariate Analysis Results Reviewer 2 (Round 1) and the Editorial Decision strongly emphasized the necessity of multivariate regression models to control for confounding variables (age, sex, BMI, diabetes) and requested that the results be presented to show if associations remain significant after adjustment.

Response: We thank you for emphasizing the importance of clearly presenting the multivariate regression results. We would like to clarify that, in the previous revision, a column containing the P values from the multivariate analyses had already been added to Tables II and III (formerly Tables IV and V). These adjusted P values correspond to multivariate models controlling for age, sex, body mass index (BMI), and diabetes status. In addition, the legends of both tables already included the statement: "In the multivariate analysis, potential confounders (age, sex, body mass index, and diabetes) were incorporated into the model to control for their effects." To further enhance clarity and ensure complete transparency regarding the adjusted analyses, we have now complemented the tables by adding the corresponding beta coefficients (β) and 95% confidence intervals to Tables II and III. These measures were also explicitly incorporated into the Results section, alongside the adjusted P values that were already reported. We believe these additions provide a more comprehensive presentation of the magnitude and direction of the associations and fully address the concern regarding multivariate adjustment and control of potential confounding factors.

Handling Editor's comments

Dear Dr. Barros:

Thank you for your thoughtful and comprehensive responses to the reviewers' comments and for the careful revisions incorporated into the manuscript (MIOC-2025-0173.R3). I appreciate the transparency added regarding the ABCB1 c.3435C>T polymorphism and the enhanced presentation of multivariate analyses. These modifications substantially improve the clarity and interpretability of your findings.

I am pleased to inform you that the manuscript is accepted for publication in Memorias do Instituto Oswaldo Cruz, in the certainty that the specific details regarding the presentation of statistical results pointed out below will be carefully addressed by the authors. These adjustments are essential to ensure the accuracy and clarity of the findings presented.

METHODOLOGICAL CONSIDERATION ON STATISTICAL REPORTING

Upon detailed review of the revised statistical reporting, I noted a methodological consideration regarding the presentation of results for binary outcomes. Specifically, for the analysis of cirrhosis (a dichotomous variable) in Table II, logistic regression was appropriately employed; however, the results are currently reported as beta coefficients with 95% CIs. For logistic regression models, the conventional and more interpretable measure of association is the Odds Ratio (OR) with its corresponding 95% CI, as this expresses the relative effect size compared to the reference category in a clinically meaningful manner.

Explanation for conversion: In logistic regression, beta represents the log-odds coefficient. The odds ratio is obtained by exponentiating beta: OR = exp(beta). Similarly, the 95% CI for the OR is calculated by exponentiating the lower and upper bounds of the beta coefficient's CI.

INCONSISTENCIES FOUND

I also observed inconsistencies in the reported values for the ABCB11 c.1331T>C polymorphism and cirrhosis (genotype-based analysis). You reported "beta = -1.12; 95%CI: 0.14 to 2.14".

1. Confidence Interval: The reported CI (0.14 to 2.14) does not contain the point estimate beta = -1.12, suggesting a possible typographical error in the CI bounds.

2. Direction of Effect: A negative beta (-1.12) indicates a protective effect (OR < 1). However, your text states that the "variant genotype showing a higher frequency of cirrhosis," which implies a risk factor (OR > 1). This suggests a potential error in either the sign of the coefficient or the reference category used in your model.

Please revisit your statistical output to confirm the correct beta estimates and confidence intervals before applying the exponentiation transformation.

CALCULATIONS AND SUGGESTED REVISIONS

1. Results section text – Genotype-based analysis for Cirrhosis:

Assuming the intended CI for beta was approximately -2.14 to -0.14 (symmetric around -1.12), the conversion would be: OR = exp(-1.12) approx. 0.33; 95% CI_OR = [exp(-2.14), exp(-0.14)] approx. [0.12, 0.87]

Replace: "the association between the c.1331T>C variant and cirrhosis remained significant in the genotype-based analysis (CC: beta = -1.12; 95%CI: 0.14 to 2.14; p<0.05), with the variant genotype showing a higher frequency of cirrhosis." with "the association between the c.1331T>C variant and cirrhosis remained significant in the genotype-based analysis (CC: adjusted OR = 0.33; 95%CI: 0.12–0.87; p<0.05), with the variant genotype showing a higher frequency of cirrhosis."

(Note: Please verify if the OR is indeed < 1, as this implies protection. If the variant is a risk factor, the beta should be positive.)

2. Table II – ABCB11 c.1331T>C SNP, cirrhosis, genotype-based analysis: Replace the phrase "< 0.05 (beta = -1.12; 95%CI: 0.14 to 2.14)" with "< 0.05 (adjusted OR = 0.33; 95%CI: 0.12–0.87)".

3. Table II – ABCB11 c.1331T>C SNP, cirrhosis, allele-based analysis: Current reporting: 0.1 (beta = 0.41; 95%CI: -0.02 to 0.85). OR = exp(0.41) approx. 1.51; 95% CI_OR = [exp(-0.02), exp(0.85)] approx. [0.98, 2.34]. Replace: "0.1 (beta = 0.41; 95%CI: -0.02 to 0.85)" with "0.1 (adjusted OR = 1.51; 95%CI: 0.98–2.34)".

IMPORTANT CLARIFICATION

For continuous outcomes (e.g., total cholesterol, HDL, AST), the presentation of beta coefficients with 95% CIs from linear regression models is statistically appropriate and requires no modification. Specifically, the result for AST levels reported in the text as "beta = 9.44; 95%CI: 0.73 to 18.2" refers to a continuous outcome and is correct as stated.

These adjustments will align your manuscript with standard reporting guidelines for regression analyses in clinical research and ensure the highest quality for your publication.

Sincerely,

Dr. Marcelo Ribeiro Alves

Handling Editor

Memórias do Instituto Oswaldo Cruz

• peer review recommendation: accept

History

• Received
  12 July 2025
• Accepted
  19 Mar 2026

Publication Dates

• Publication in this collection
  20 July 2026
• Date of issue
  2026

History

• Received
  12 July 2025
• Accepted
  19 Mar 2026