Genetic polymorphisms of interleukin-1-beta in association with sustained response to anti-viral treatment in chronic hepatitis B in Chinese

Authors


Dr H. L.-Y. Chan, Department of Medicine and Therapeutics, 9/F Prince of Wales Hospital, 30-32 Ngan Shing Street, Shatin, Hong Kong.
E-mail: hlychan@cuhk.edu.hk

Abstract

Summary

Background

Interleukin-1β is a pro-inflammatory cytokine that may influence host defence against viral infection.

Aim

To investigate the impact of interleukin-1β gene polymorphism on the response to anti-viral treatment.

Method

Hepatitis B e antigen-positive chronic hepatitis B patients who have completed a randomized study of peginterferon alpha-2b and lamivudine combination vs. lamivudine monotherapy were included. Sustained responders were patients who had persistent hepatitis B e antigen loss and less than two occasions with hepatitis B virus DNA >100 000 copies/mL at any time up to week 76 post-treatment. Polymorphisms at interleukin-1β-511, -31 and -3954 and interleukin-1 receptor antagonist (RN) were studied.

Results

Eighty-eight patients were studied and 18 (20%) patients developed sustained response. Near complete linkage disequilibrium was observed between interleukin-1β-511 and -31 loci. After adjustment for the potential confounding effects of treatment allocation, hepatitis B virus genotype, pre-treatment alanine aminotransferase and hepatitis B virus DNA levels, genotype C/T at interleukin-1β-511 was found to be associated with higher sustained response than genotype C/C (adjusted odds ratio 10.4, 95% CI 1.1, 96.9, P = 0.040). The proportion of sustained responders tend to be higher among patients with allele T at interleukin-1β-511 (83%) than those without (70%) (P = 0.058).

Conclusion

High interleukin-1β production genotype at position -511 has a favourable response to anti-viral treatments.

Introduction

Chronic hepatitis B infection is the most common cause of liver cirrhosis and hepatocellular carcinoma in Asia. Treatment by nucleot(s)ide analogues is associated with potent viral suppression but substantial rates of relapse after cessation of treatment.1–3 Peginterferon with and without combination with lamivudine is associated with sustained response in approximately one-third of patients.4–6 Factors including low pre-treatment hepatitis B virus (HBV) DNA level, high alanine aminotransferase (ALT) levels and probably certain viral genotypes may be associated with more favourable treatment response.5, 6

Cytokines play an important role in the defence against viral infection through determination of the pattern of host immune response and direct viral suppression.7 Interleukin (IL)-1 is a pro-inflammatory cytokine with two polypeptides, IL-1α and IL-1β, and its activity is modulated by a natural competitive inhibitor, IL-1 receptor antagonist (IL-1RN).8 Genes encoding IL-1β have been mapped to chromosome 2q13–q21, and C/T single nucleotide polymorphism at IL-1β-31, -511, -3954 have been described. Specific genotypes at the IL-1β genes including T at allele -511, C at allele -31 and T at allele -3954 have been reported to be associated with increased production of IL-1β.9, 10 IL-1RN contains an 86 base pair variable number tandem repeat (VNTR) polymorphism in intron 2, of which allele 2 (the short allele) is associated with enhanced IL-1β production.11 In chronic hepatitis B, IL-1β gene polymorphisms associated with increased IL-1β production have been reported to be associated with lower HBV DNA and milder disease activity possibly related to the enhanced immunological clearance of the virus.12

To date, host factors that can predict treatment outcome of chronic hepatitis B are poorly defined. As IL-1β gene polymorphism may reflect the host immune status as well as the natural disease control, it may affect the virological response to anti-viral treatment. In this study, we aimed to investigate the impact of IL-1β gene polymorphism on the response to anti-viral treatment based on a cohort of chronic hepatitis B patients in a previously conducted clinical trial.

Patients and methods

Patients and study design

Chinese chronic hepatitis B patients who have completed per protocol treatment in a randomized study of peginterferon alpha-2b (PegIntron®; Shering-Plough Corporation, Kenilworth, NJ, USA) and lamivudine (Zeffix®; GlaxoSmithKline, Middlesex, UK) combination vs. lamivudine monotherapy were included in this study.4 At screening, all patients had treatment-naïve chronic hepatitis B with positive hepatitis B e antigen (HBeAg), serum HBV DNA of at least 500 000 copies/mL and alanine transaminase (ALT) 1.3 to five times of upper limit of normal. Peginterferon alpha-2b was prescribed as subcutaneous injection at a dose of 1.5 μg/kg/week for patients with body weight <65 kg or 100 μg/week for patients with body weight more than 65 kg for 32 weeks. Liver biopsy was performed within 4 weeks before commencement of treatment. Histologic necro-inflammation was scored by Knodell scoring system (0–18) and liver fibrosis by Ishak scoring system (0–6). Lamivudine was administered as 100 mg oral daily for 52 weeks in both groups of patients. In patients receiving combination therapy, peginterferon alpha-2b was administered 8 weeks before the commencement of lamivudine, then combination of both treatments for 24 weeks followed by lamivudine monotherapy for a further 28 weeks.

After completion of treatment, all patients were followed up at week 4, 8, 16, 24, 32, 48, 52, and 12 weekly thereafter. Liver biochemistry was monitored at every post-treatment visit, HBeAg and antibodies to HBeAg (anti-HBe) were monitored at every visit from week 8 onwards, while HBV DNA was monitored at week 4, 8, 16, 24, 52 and then every 24 weekly. Hepatitis B surface antigen (HBsAg) was monitored yearly.

We defined sustained responders as patients who had persistent HBeAg loss and had less than two occasions with HBV DNA <100 000 copies/mL at any time up to week 76 of post-treatment follow-up. None of the patients developed virological relapse after week 76 post-treatment.13 Other patients who did not develop HBeAg loss after treatment or those who developed viral reactivation after initial HBeAg loss were defined as non-responders.

All patients gave written informed consent for this study. The study protocol has been approved by the local ethics committee.

Laboratory methods

Serum HBV DNA assay

Residual serum samples were stored in −80 °C freezer. HBV DNA was extracted by QIAGEN QIAamp DNA Mini Kit (QIAGEN Inc., Chatsworth, CA, USA) according to the instructions of the manufacturer. A TaqMan real-time polymerase chain reaction (PCR) assay was used to measure serum HBV DNA levels, as described previously.14, 15 This assay was standardized by serial dilution of EUROHEP genotype D HBV standard (from Dr KH Heerman, University of Goettingen, Goettingen, Germany) which contained 2.7 × 109 viral copies/mL. The range of HBV DNA detection was from 102 to 109 copies/mL with correlation coefficient of the standard curve routinely >0.99.

IL-1β polymorphism genotyping

Genomic DNA was extracted by QIAGEN QIAamp DNA Blood Mini Kit (QIAGEN Inc.) according to the instructions of the manufacturer. Three polymorphisms of IL-1β (−511C/T, −31C/T, +3954C/T) were analysed by 5° nuclease assay using TaqMan probes. Positions -511 and -31 were genotyped as described by El-Omar et al.16 PCR amplification was performed in a volume of 25 uL containing 1X Taqman Universal PCR Master Mix (Applied Biosystems, Foster City, CA, USA), 900 nm of each primer, 200 nm of each probe and 50 ng genomic DNA (Table 1). Position +3954 was determined by a validated TaqMan SNP Genotyping Assay – Assay number 001_0386 (Applied Biosystems). The assay was performed in a volume of 25 μL containing 1X Taqman Universal PCR Master Mix, 1X Assays-On-DemandTM SNP Genotyping Assay Mix (Applied Biosystems) and 50 ng genomic DNA. All amplification and detection were performed in ABI PRISM® 7000 Sequence Detection System (Applied Biosystems).

Table 1.   Primers and probe sequences used for interleukin (IL)-1β and IL-1 receptor antagonist (RN) genotyping
IL-1β-511
 IL1–1β-511-C5′ FAM – TGT TCT CTG CCT CGG GAG CTC TCT G – BHQ 3′
 IL-1β-511-T5′ JOE – CTG TTC TCT GCC TCA GGA GCT CTC TGT C – BHQ 3′
 IL-1β-511-F5′– TCC TCA GAG GCT CCT GCA AT – 3′
 IL-1β-511-R5′– TGT GGG TCT CTA CCT TGG GTG – 3′
IL-1β-31
 IL-1β-31-C5′ FAM – TCG CTG TTT TTA TGG CTT TCA AAA GCA G – BHQ 3′
 IL-1β-31-T5′ JOE – CCT CGC TGT TTT TAT AGC TTT CAA AAG CAG A – BHQ 3′
 IL-1β-31-F5′– CCC TTT CCT TTA ACT TGA TTG TGA – 3′
 IL-1β-31-R5′– GGT TTG GTA TCT GCC AGT TTC TC – 3′
 IL-1β-3954
 IL-1β-3954-T5′ FAM – TCC CAT GTG TCA AAG A – MGB 3′
 IL-1β-3954-C5′ VIC – TGT GTC GAA GAA GAT – MGB 3′
 IL-1β-3954-F5′– CCT AAA CAA CAT GTG CTC CAC ATT – 3′
 IL-1β-3954-R5′– ATC GTG CAC ATA AGC CTC GTT A – 3′
IL-1RN
 IL-1RN-F5′ FAM – CTC AGC AAC ACT CCT AT - 3′
 IL-1RN-R5′ - TCC TGG TCT GCA GGT AA – 3′

IL-1RN polymorphism genotyping

The 86 base pair VNTR polymorphism of IL-1RN intron 2 was assessed by PCR-based method. The VNTR region was first amplified by primers IL1RN-F and IL1RN-R (Table 1). The 5° end of IL-1RN-F was labelled with 6-FAM. Amplification was carried out in a final volume of 25 μL containing 1X PCR Buffer II, 0.2 mm dNTPs, 3 mm MgCl2, 0.8 mm of each primer, 1 unit of AmpliTaq Gold Polymerase (Applied Biosystems) and 50 ng of genomic DNA. Thermal cycling was performed in 96-Well GeneAmp® PCR System 9700 (Applied Biosystems). The amplicons were of five different sizes depending on the number of repeats: 240 bp = 2 repeats (allele 2), 325 bp = 3 repeats (allele 4), 410 bp = 4 repeats (allele 1), 500 bp = 5 repeats (allele 3) and 595 bp = 6 repeats (allele 5).17 The amplicon size was detected by GeneScan technology using ABI Prism® 3100 Genetic Analyzer (Applied Biosystems) for detection and GeneMapper® software (Applied Biosystems) for analysis. GeneScanTM-1000 ROXTM standard (Applied Biosystems) was used as size standard which was first diluted in fivefolds with Hi-Di FormamideTM (Applied Biosystems). The PCR products were diluted in 20-folds with autoclaved deionized water. One microlitre of diluted PCR product was mixed with 9 μL of diluted size standard. This mixture was then denatured at 95 °C for 5 min and placed on ice immediately until sample loading.

Positive controls of all genotypes were included in each assay for assay verification. Direct sequencing was performed for randomly selected samples to reconfirm genotyping results.

Statistical analysis

Analysis of the genotype data was performed using PowerMarker software version 3.23 (http://statgen.ncsu.edu/powermarker/index.html).18 Hardy–Weinberg equilibrium was assessed by chi-squared test or Fisher's exact test as appropriate. D′ and r2 were calculated to evaluate linkage disequilibrium for all pairwise single nucleotide polymorphism combinations. Haplotypes and their frequencies were inferred using the PowerMarker software, which utilized expectation-maximization algorithm to estimate haplotype frequencies.

Statistical analysis was performed by Statistical Package for Social Science (version 11.0; SPSS, Chicago, IL, USA). Continuous variable were expressed as mean ± s.d. or median (range) as appropriate. Categorical variables were compared by Pearson chi-squared test and continuous variables were compared by Student's t-test or Mann–Whitney U-test as appropriate. Analysis of variance (anova) and Kruskal–Wallis test were used to investigate the association of different IL-1β gene polymorphisms with baseline clinical variables as appropriate. Logistic regression model was used to adjust the confounding effect of baseline variables to determine the independent effect of IL-1β gene polymorphism on sustained response. Factors including treatment allocation (combination treatment vs. lamivudine monotherapy), HBV genotype (C vs. B), ALT and log HBV DNA levels were included in logistic regression analysis. All tests were two-sided. Statistical significance was taken as P < 0.05.

Results

Clinical characteristics

Among one hundred patients who have been randomized into the previous clinical trial comparing combination of peginterferon and lamivudine vs. lamivudine monotherapy, 88 patients were included in this study. Other patients were excluded because of premature termination of treatment (n = 4), negative HBeAg at baseline (n = 1) and failure to obtain consent (n = 7). The post-treatment follow-up was 126 ± 30 weeks and all patients had at least 76 weeks of post-treatment follow-up. Eighteen of 88 (20%) patients developed sustained response (Table 2). One sustained responder on combination treatment had HBsAg seroclearance. Non-responders tend to be older (P = 0.038) with male predominance (P = 0.031) when compared with the sustained responders. Patients who received combination treatment had significantly higher rate of sustained response than those who received lamivudine monotherapy (P = 0.006). Otherwise there was no statistical difference in the HBV genotype distribution, ALT levels, log HBV DNA levels and histologic necro-inflammation and fibrosis between the two patient groups at baseline.

Table 2.   Baseline clinical characteristics of patients
 Sustained responder (n = 18)Non-sustained responders (n = 70) Overall (n = 88)
  1. HBeAg, hepatitis B e antigen; HBV, hepatitis B virus; ALT, alanine aminotransferase.

Age (year)29 ± 834 ± 1033 ± 10
Male8 (44%)50 (71%)58 (66%)
HBeAg positive18 (100%)70 (100%)88 (100%)
Log HBV DNA (copies/mL)7.9 ± 0.728.08 ± 1.128.04 ± 1.05
ALT (IU/L)173.5 (57–461)124.5 (36–1179)129.5 (36–1179)
Genotype (B:C:mixed)5:10:323:44:328:54:6
Necro-inflammation score6 (1–12)5 (1–12)6 (1–12)
Fibrosis score1 (0–5)1 (0–6)6 (0–6)
Combination treatment14 (78%)29 (41%)43 (49%)

Polymorphisms in IL-1β genes

The distribution of genotypes for the IL-1β loci and IL-1RN was in Hardy–Weinberg equilibrium. Overall, C/C, C/T and T/T genotypes of IL-1β-511 occurred in 24 (27%), 46 (52%) and 18 (21%) patients. Near complete linkage disequilibrium was observed between IL-1β-511 and -31 loci with linkage disequilibrium coefficient of 0.93. The predominant haplotypes of IL-1β-511/-31 were C-T (52%) and T-C (47%). Eighty-six (98%) patients had genotype C/C at IL-1β-3954 while C/T (two patients, 2%) was only a minor genotype. At IL-1RN, 72 (82%) patients had genotype 1/1, 15 (17%) patients had genotype 1/2 and one (1%) patient had genotype 1/4.

Association of IL-1β gene polymorphisms with baseline characteristics

Genotype T/T at IL-1β-511 was associated with higher ALT level than genotype C/C (Table 3). Similarly, genotype C/C at IL-1β-31 was associated with higher ALT level than genotype T/T. Heterozygous C/T genotype at IL-1β-511 and -31 were not associated with higher ALT levels. There was no association between IL-1β-3954 and IL-1RN polymorphisms with ALT levels.

Table 3.   Association of interleukin (IL)-1β and IL-1 receptor antagonist (RN) genotypes and baseline alanine aminotransferase (ALT) and hepatitis B virus (HBV) DNA levels
 nALTPP (Kruskal–Wallis)Log HBV DNAPP (anova)
  1. NA, not applicable due to too small number of patients.

IL-1β- 511
 C/C24122.5 (36–276)Referent0.257.9 ± 1.1Referent0.29
 C/T46127 (57–1179)0.277.9 ± 1.10.88
 T/T18182 (76–436)0.0288.4 ± 0.90.20
IL-1β-31
 T/T23119 (36–276)Referent0.287.9 ± 1.1Referent0.35
 C/T45125 (57–1179)0.317.9 ± 1.10.90
 C/C20182 (76–436)0.0208.3 ± 0.900.24
IL-1β-3954
 C/C86129.5 (36–1179)ReferentNA8.0 ± 1.0ReferentNA
 C/T2153.5 (71–236)0.809.3 ± 0.430.077
IL-1RN
 1/172131 (36–1179)Referent0.498.1 ± 1.0Referent0.10
 1/215130 (68–329)0.727.6 ± 1.10.067
 1/41124NA6.95NA

Polymorphism 1/1 IL-1RN tend to have higher baseline log HBV DNA than polymorphism 1/2 but the difference just fell short of statistical significance (Table 3). The number of patients with C/T at IL-1β-3954 was too small to determine its clinical significance. There was no association between IL-1β-511 and -31 with HBV DNA levels.

Eighty-two of 88 patients had liver histology results available for evaluation. There was no association between the polymorphism at any of the IL-1β foci or IL-1RN with histologic necro-inflammation and fibrosis (Table 4).

Table 4.   Association of interleukin (IL)-1β and IL-1 receptor antagonist (RN) genotypes with baseline histologic necro-inflammatory and fibrosis scores
 NNecroinflammatory score PP (Kruskal–Wallis)Fibrosis score PP (Kruskal–Wallis)
  1. NA, not applicable due to too small number of patients.

IL-1β-511
 C/C245 (1–11)Referent0.811 (0–5)Referent1.00
 C/T415 (1–11)0.891 (0–6)0.98
 T/T177 (1–12)0.551 (0–4)0.94
IL-1β-31
 T/T235 (1–11)Referent0.921 (0–5)Referent0.93
 C/T405 (1–11)0.701 (0–6)0.80
 C/C195 (1–12)0.871 (0–4)0.98
IL-1β-3954
 C/C815 (1–12)ReferentNA1 (0–6)ReferentNA
 C/T19NA4NA
IL-1RN
 1/1675 (1–11)Referent0.401 (0–6)Referent0.53
 1/2146 (1–11)0.391 (0–6)0.29
 1/412NA1NA

Association of IL-1β gene polymorphism with sustained response

On univariate analysis, the proportions of different genotypes at various IL-1β foci and IL-1RN were comparable between the sustained responders and the non-responders (Table 5). After adjustment of baseline factors including treatment allocation, HBV genotype, ALT and HBV DNA levels, genotype C/T at IL-1β-511 was found to be associated with higher sustained response (Tables 5 and 6). The proportion of sustained responder also tend to be higher among patients with allele T at IL-1β-511, i.e. C/T and T/T genotypes (83%) than those patients without (70%) (P = 0.058). There was no association between IL-1β-3954 and IL-1RN gene polymorphism with sustained response. Fifty per cent of haplotype T-C at IL-1β-511/-31 vs. 44% of haplotype C-T at IL-1β-511/-31 had sustained response (odds ratio 0.76; 95% confidence 0.36, 1.61; P = 0.47), and the association did not reach statistical significance after adjustment for treatment allocation, HBV genotype, baseline ALT and log HBV DNA levels (adjusted odds ratio 0.53; 95% confidence interval 0.22, 1.30; P = 0.17).

Table 5.   Association of interleukin (IL)-1β and IL-1 receptor antagonist (RN) genotypes and haplotypes with sustained response
 SR (n, %)No SR (n, %)OR (95% CI) PAOR (95% CI) P
  1. SR, sustained response; OR, odds ratio; CI, confidence interval; AOR, adjusted odds ratio [adjusted for treatment allocation, hepatitis B virus (HBV) genotype, baseline alanine aminotransferase and log HBV DNA].

IL-1β- 511
 C/C3 (17)21 (30)Referent Referent 
 C/T12 (67)34 (49)2.5 (0.62, 9.8)0.2010.37 (1.11, 96.96)0.040
 T/T3 (17)15 (21)1.4 (0.25, 7.9)0.704.1 (0.31, 55.9)0.28
IL-1β-31
 T/T3 (17)20 (29)Referent Referent 
 C/T10 (56)35 (50)1.9 (0.47, 7.7)0.377.9 (0.85, 73.6)0.088
 C/C5 (28)13 (21)2.2 (0.46, 10.8)0.328.1 (0.73, 90.8)0.069
IL-1β-3954
 C/C18 (100)68 (97)Referent Referent 
 C/T0 (0)2 (3)Undefined1.00Undefined1.00
IL-1RN
 1/116 (89)56 (80)Referent Referent 
 1/22 (11)13 (19)0.54 (0.11, 2.6)0.450.82 (0.14, 5.1)0.84
 1/40 (0)1 (1)Undefined1.00Undefined1.00
IL-1β-511
 C/C3 (17)21 (30)Referent Referent 
 C/T and/T/T15 (83)49 (70)2.1 (0.56, 8.2)0.278.3 (0.93, 73.5)0.058
IL-1β-31
 T/T3 (17)20 (29)Referent Referent 
 C/T and/C/C15 (83)50 (71)2.0 (0.52, 7.7)0.317.97 (0.90, 70.95)0.063
Haplotype -511/-31
 T-C18 (50)64 (46)Referent Referent 
 C-T16 (44)75 (54)0.76 (0.36, 1.6)0.470.53 (0.22, 1.3)0.17
 C-C2 (6)1 (1)7.1 (0.61, 82.9)0.126.9 (0.49, 98.1)0.15
Table 6.   Logistic regression table of IL-1β-511 polymorphism and other baseline factors for sustained response
FactorAOR (95% CI)P-value
  1. AOR, adjusted odds ratio; CI, confidence interval; ALT, alanine aminotransferase.

Model 1
 Combination treatment (vs. lamivudine)7.2 (1.2, 31.9)0.010
 Genotype C (vs. B)2.1 (0.48, 8.6)0.33
 Log HBV DNA at baseline (copies/mL)0.70 (0.38, 1.3)0.25
 ALT at baseline (IU/L)1.0 (0.99, 1.0)0.81
 IL-1β-511 C/T (vs. CC)10.4 (1.1, 96.9)0.040
 IL-1β-511 T/T (vs. C/C)4.1 (0.31, 55.9)0.28
Model 2: (C/T + T/T vs. CC)
 Combination treatment (vs. lamivudine)7.2 (1.7, 31.0)0.009
 Genotype C (vs. B)1.8 (0.44, 7.4)0.41
 Log HBV DNA at baseline (copies/mL)0.65 (0.35, 1.2)0.18
 ALT at baseline (IU/L)1.0 (0.99, 1.0)0.82
 IL-1β- 511 C/T and T/T (vs. C/C)8.3 (0.93, 73.5)0.058

Discussion

In this study, we have demonstrated the relationship between IL-1β gene polymorphism and the biochemical severity of hepatitis as well as the virological response to anti-viral treatment among HBeAg-positive chronic hepatitis B patients. High IL-1β production genotypes (i.e. T at IL-1β-511) tend to be associated with higher baseline ALT levels and higher sustained virological response rates independent of the baseline clinical status. As IL-1β is an important pro-inflammatory cytokine, our findings support the role of inflammation in the viral clearance of chronic hepatitis B.

In Asia where chronic hepatitis B is usually transmitted perinatally, most patients have an initial phase of immune tolerance for approximately 20–30 years before immune clearance.19 The immune clearance phase is characterized by an increased serum ALT level accompanied by reduction of viral load. In chronic hepatitis B, the HBV specific T-cell response is impaired.20, 21 The success of anti-viral treatment particularly interferon-based treatment relies on an enhancement of the host immune response against HBV.22 Enhancement of immune response may be accompanied by an elevation of ALT level indicating increased hepatic inflammation.23, 24 Patient with higher pre-treatment ALT levels tend to have more favourable response to lamivudine treatment and patients who have ALT flares during interferon treatment also tend to have higher rates of response.25–27 One possible outcome of enhanced hepatic necro-inflammation is clearance of covalently closed circular HBV DNA together with the infected hepatocytes, which might be an important mechanism to ensure sustained virological response.28, 29

In the present study, we studied a cohort of HBeAg-positive chronic hepatitis B patients previously recruited in a randomized controlled study of combination peginterferon and lamivudine treatment vs. lamivudine monotherapy.4 These patients are an ideal cohort to investigate sustained virological response as all of them have undergone systematic follow-up for at least 76 weeks after stopping treatment. None of the sustained responders developed virological relapse after week 76.13 The distribution of IL-1β gene polymorphism in this cohort is comparable with that described in China and Japan among healthy non-hepatitis individuals.12, 30, 31 Polymorphism at IL-1β-511 and -31 are at almost complete linkage disequilibrium. C/T polymorphism at IL-1β-3954 and the short allele, allele 2, at IL-1RN are relatively uncommon in the studied population.

We have demonstrated C/C genotype of IL-1β-511 and T/T genotype of IL-1β-31 are associated with higher baseline ALT levels. This result supports previous experimental findings that these IL-1β gene polymorphisms are associated with increased production of the pro-inflammatory cytokine IL-1β.9 However, we fail to show any association between IL-1β gene polymorphism with histologic disease severity. As our study only included HBeAg-positive chronic hepatitis B patients with elevated ALT levels, the liver histology among patients with HBeAg-negative inactive disease cannot be evaluated.

There is a trend for T at IL-1β-511 to associate with higher sustained virological response after adjusted for imbalance of the potential baseline confounders. Anti-viral treatment (combination treatment vs. lamivudine alone), HBV genotype, ALT and HBV DNA levels have been reported previously to influence the treatment response.4–6 In this study, most sustained responders received combination treatment. There was also imbalance in the distribution of IL-1β genotype at -511 with respect to the baseline ALT levels. High IL-1β genotypes and hence hepatic inflammation seem to be an important element for viral clearance by anti-viral treatment.

Our study has several limitations. The major limitation is the small sample size as restricted by the patient recruitment in the previous clinical trial. This limitation imposes difficulty to investigate the significance of uncommon IL-1β gene polymorphisms such as IL-1β-3954 and IL-1RN. Secondly, patients were either treated with combination treatment or lamivudine alone. On multivariate analysis, combination therapy is the most important predictor of sustained response (Table 6). Because of our small sample size and low event rate (sustained response) particularly among lamivudine monotherapy group, it will be difficult to study the role of IL-1β gene polymorphism in the two treatment groups separately. Our results are therefore derived from two heterogeneous groups of patients and can only suggest the easy-to-treat IL-1β genotypes in general but not to a specific treatment modality. Future studies with larger sample size including patients on a specific treatment will be required to confirm our findings. Thirdly, there is no placebo group in our study. We cannot exclude the possibility that high IL-1β production genotypes are associated with a higher rate of spontaneous HBeAg seroconversion. Previous studies in chronic hepatitis C suggest that high IL-1β production genotypes are associated with poorer long-term outcome and higher risk of hepatocellular carcinoma.31, 32 On the contrary, a study in China suggested that high IL-1β production genotypes are associated with lower viral load and milder disease in chronic hepatitis B.12

In conclusion, we have demonstrated polymorphisms at IL-1β-511 and -31 affect hepatic inflammation and treatment response in HBeAg-positive chronic hepatitis B. High IL-1β production genotypes at these two foci tend to have higher rates of sustained response independent of the baseline disease status. Whether this pro-inflammatory cytokine gene polymorphism influences the natural history and development of hepatocellular carcinoma in chronic hepatitis B warrants further study.

Acknowledgement

The work described in this paper was partially supported by a grant from the Research Grant Council of the Hong Kong Special Administrative Region to H.L.Y. Chan (Project number CUHK 4304/04M).

Ancillary