Potential conflict of interest: Nothing to report.
Hepatitis C virus (HCV)/human immunodeficirency virus (HIV) coinfection poses a difficult therapeutic problem. Response to HCV-specific therapy is variable but might be influenced by host genetic factors, including polymorphisms of cytokine genes. Here, we studied whether interleukin-6 (IL-6) C174G gene polymorphism affects the response to antiviral treatment in HCV-infected HIV-positive subjects. We determined IL-6 genotypes in HIV-positive patients with acute (n = 52) and chronic (n = 60) hepatitis C treated with pegylated interferon-α. Two hundred ten HCV monoinfected, 197 HIV monoinfected, and 100 healthy individuals were studied as controls. Patients were classified into high and low producers according to IL-6 genotypes. Rates of sustained virological responses (SVRs) were compared between the IL-6 genotypes. Signal transducer and activator of transcription three phosphorylation was analyzed by Western blot in HCV core-transfected human hepatoma cell line (HUH7) cells. Distribution of IL-6 genotypes did not differ significantly between the study groups. SVR was achieved in 63% of HIV/HCV coinfected patients. Carriers of the IL-6 high producer (HP) genotype had significantly higher SVR rates than patients with an IL-6 low producer genotype (70.1% versus 52%; P < 0.002). This effect was seen in both HIV-positive patients with acute (74% versus 33%; P < 0.05) and chronic (66% versus 33%; P < 0.05) hepatitis C. Multivariate analysis confirmed IL-6 HP carriage as an independent positive predictor for SVR (Odd′s ratio 6.1; P = 0.004). This effect corresponds to the in vitro observation that in HCV core-transfected HUH7 cells, IL-6 overcomes the HCV core-mediated inhibition of STAT3 activation. Conclusion: Response rates to HCV-specific treatment are higher in HCV/HIV-positive patients carrying the IL-6 HP genotype, which might be because of IL-6 mediated STAT3 activation. (HEPATOLOGY 2007.)
Hepatitis C virus (HCV) is a major health problem in human immunodeficiency virus (HIV)-infected patients. In Europe and the United States approximately one-third of HIV-positive individuals are coinfected with HCV.1 Progression of hepatitis C in HIV-positive patients is characterized by more rapid progression toward cirrhosis and hepatocellular carcinoma,2 a shorter survival once HCV-related liver decompensation occurs,3 a higher rate of fulminant hepatic failure (without prior liver disease) in the context of highly active antiretroviral therapy,4 and a higher overall mortality rate5 compared with individuals with HCV monoinfection.1, 6 Thus, HCV-related morbidity and mortality have become an increasingly important health problem in HIV-infected patients, especially after the dramatic decline in morbidity and mortality caused by opportunistic diseases after the introduction of highly active anti-retroviral therapy in 1996.
Currently pegylated interferon alpha (peg-IFN-α) in combination with ribavirin represents the backbone of HCV-specific therapy. However, with interferon-based combination therapy sustained virological response (SVR) is achieved in only 50% of HCV monoinfected7, 8 and approximately 40% of HCV/HIV coinfected patients1 in clinical studies and may be even lower in clinical practice.9
The exact mechanism(s) for how interferon (IFN) mediates its anti-HCV activity are only incompletely understood but involve initiation of an antiviral state via activation of Janus kinase and signal transducers and activators of transcription (Jak-STAT) signaling pathways. Phosphorylation of STAT1 and STAT2 proteins is considered a key event in intracellular IFN signaling. However, Zhu et al.10 recently demonstrated that IFN-α also induced phosphorylation and activation of STAT3. In the liver, STAT3 exerts important roles concerning acute phase responses and protection against liver injury.10, 11 Furthermore, activated STAT3 has been shown to exhibit efficient anti-HCV activity in a HCV subgenomic replicon cell line.10 Thus, IFN-α antiviral activity also may involve activation of STAT3-signaling pathways and subsequent intracellular gene activation.12
Recent data suggest that HCV has developed strategies to counteract STAT3 activation as STAT3 expression is reduced in HCV-infected livers.13 Furthermore, the HCV core protein has been shown to prevent phosphorylation of STAT3,13–15 and defective Janus kinase and signal transducers and activators of transcription activation leads to resistance of HCV against IFN-α.16
In the liver, STAT3 is mainly activated by interleukin-6 (IL-6), a multifunctional cytokine that has been implicated in a variety of cellular functions including stimulation of hepatocytes to produce acute-phase proteins.17 In addition, IL-6 plays a role in liver regeneration and protection against hepatic injury.11
The IL-6 gene is polymorphic within its promotor region (C→G transition at position 174). Importantly, this polymorphism is associated with differences in the production of IL-6.18 Thus, individuals can be classified into high (genotypes IL-6 174 GG and 174 GC) and low (genotype IL-6 174 CC) IL-6 producers, respectively.18
Polymorphisms in the IL-6 gene have been linked to IL-6 serum levels and onset of inflammation in juvenile rheumatoid arthritis.18 However, the effect of the IL-6 polymorphism on response to HCV-specific treatment has not yet been addressed. Regarding the potentially pivotal role of host genetic factors for the outcome of anti-HCV treatment,19–22 the current study was designed to analyze the effect of IL-6 genotypes on response to HCV-specific treatment in both HCV/HIV coinfected and HCV monoinfected patients.
To study distribution of the IL-6 alleles and the potential effect of IL-6 genotypes on treatment outcome, 112 HIV/HCV coinfected white patients were enrolled into this cohort study. Within this group, 52 patients fulfilled at least 2 of the following 3 criteria within the last 4 months before diagnosis of HCV infection: (1) HCV seroconversion; (2) alanine aminotransferase > 350 IU with prior normal aminotransferases; (3) risk exposure to HCV (modified to reference23). These patients were therefore considered to represent acute hepatitis C.
HCV/HIV coinfected patients were treated with peg-IFN-α and ribavirin (800 mg/day) for 24 (HCV genotypes 2 and 3) or 48 weeks (genotypes 1 and 4).
As a control, 210 chronically HCV genotype 1 monoinfected patients were included to further analyze the impact of the IL-6 polymorphism in HIV-negative patients. HCV monoinfected patients were either treated with standard IFN-α (n = 60) or peg-IFN-α (n = 150) in combination with ribavirin.
Patients who were HCV RNA negative at the end of treatment and 6 months after the end of treatment were classified as sustained virological responders. All other patients were considered to be nonresponders.
Before antiviral treatment, patients were meticulously characterized with respect to HCV genotype, HCV viral load, HBV serology, alanine aminotransferase, gamma-glutamyltransferase, and HIV viral load.
In addition, 100 (HIV-negative, HCV-negative) persons and 197 HIV monoinfected patients were included in our study to analyze the distribution of the IL-6 genotypes (Table 1).
All subjects were tested negative for hepatitis B virus infection (hepatitis B surface antigen, anti-hepatitis B surface antigen, and anti-hepatitis B core antigen) as assessed by commercially available assays according to the manufacturer's instructions (Abbott, Wiesbaden, Germany).
Ethylene diamine tetraacetic acid blood samples were obtained from each patient to determine their cytokine genotypes. Liver function tests were determined by routine biochemical procedures. CD4+ and CD8+ T cell counts were assessed by routine flow cytometrical examinations on a FACScalibur (BD).
Informed consent was obtained from each patient before inclusion into the study, and the study conformed with the ethical guidelines of the Helsinki declaration as approved by the local ethics committees.
IL-6 genotypes were analyzed using the cytokine genotyping tray (One Lambda, CA) following the manufacturer's protocol, allowing classification into IL-6 high (174GG and 174GC) and low (174CC) producers.
Diagnosis of HCV Infection.
HCV antibodies were detected with a microparticle enzyme immunoassay (Axsym; Abbott) and confirmed by dot immunoassay (Matrix; Abbott). HCV RNA was detected with a nucleic acid purification kit (QIAamp Viral Kit; Qiagen, Hilden, Germany), followed by reverse transcription and nested polymerase chain reaction as described elsewhere.22 Quantitative determination of HCV load was done by branched DNA technology (Chiron, Emeryville, CA). HCV genotypes were determined by the Innolipa II line probe assay (Innogenetics, Zwijndrecht, Belgium).
Transient Transfection of HUH7 Cells.
HCV complementary DNA encoding the full-length core protein (amino acids 1-191) was polymerase chain reaction–amplified from HCV plasmid 4z+C5p#8 (HCV genotype 1a, kindly provided by M. Houghton, Chiron) and cloned into the pCMV/LIC expression vector (Invitrogen, Karlsruhe, Germany). Transient transfections of HUH7 cells were done using LipofectAMINE (Invitrogen, Karlsruhe, Germany), following the manufacturer's instructions. Expression of HCV core protein was controlled by Western blotting 48 hours after transfection.
Western Blot Analysis.
Western blot was performed following standard protocols. In brief, cell lysates were dissolved in sodium dodecyl sulfate sample buffer and applied to a 12% polyacrylamide slab gel. After separation by electrophoresis, proteins were transferred to a nitrocellulose membrane (Schleicher and Schuell, Germany) by electroblotting. The membrane was placed in blocking buffer (5% nonfat dry milk dissolved in 150 mM NaCl, 50 mM Tris, pH 7.5, and 0.02% Tween-20) for 1 hour. Next, the membrane was incubated with antibody for 1 hour at room temperature using the following antibodies: anti-STAT3 and anti-Phospho-STAT3 (Cell Signaling) and anti-HCV core. The membrane was washed 3 times with phosphate-buffered saline/Tween-20, and incubated with horseradish peroxidase–labeled goat anti-rabbit IgG (Zymed, Germany) at a dilution of 1:4,000 in phosphate-buffered saline/Tween-20. Then it was washed 3 times more in phosphate-buffered saline/Tween-20. Finally it was soaked in enhanced chemiluminescent detection reagent and exposed to Hyperfilm MP (Amersham, Germany) for a period of 10 seconds.
Distributions of GNB3 genotypes were analyzed with software designed by Strom and Wienker (http://ihg.gsf.de). Treatment response rates in patients with different cytokine genotypes were analyzed using 2 × 2 contingency tables. For statistical comparisons between the groups, chi-squared statistics, Fisher's exact test, and Mann-Whitney U test were used as appropriate. To determine the effect of the IL-6 genotype in comparison to HCV load and genotype, the main established predictors of response, we stratified our patient data as HCV genotype 1 versus non-1, and HCV RNA ≤ 2.5 million copies/ml versus >2.5 million copies/ml.8, 9 These stratified parameters were analyzed together with IL-6 producer status in a forward conditional stepwise logistic regression model using SVR as outcome variable. P values less than 0.05 (2-sided) were regarded as significant. All calculations were performed on a personal computer with SPSS 12.0 software (SPSS, Chicago, IL).
One hundred twelve patients with HCV/HIV coinfection were enrolled into this study, including 52 individuals who fulfilled the criteria of acute hepatitis C and 60 subjects with chronic HCV infection. Patients received peg-IFN-α plus body weight–adapted doses of ribavirin for 24 (genotypes 2 and 3) and 48 (genotypes 1 and 4) weeks, respectively. Both cohorts (acute and chronic hepatitis C) differed significantly in terms of sex distribution, alanine aminotransferase, and gamma glutamyltransferase levels as well as HCV loads (Table 2).
Table 2. Characteristics of Acute and Chronically HCV Coinfected Patients
Overall, the SVR was 63% (60/112). Coinfected subjects with an acute hepatitis C had higher SVR rates as compared with chronically infected patients (67% versus 58%). However, these differences did not reach statistical significance.
In addition, 210 patients with chronic hepatitis C genotype 1 monoinfection were enrolled and stratified into a standard IFN-α group (n = 60) and a peg-IFN-α therapy group (n =150). As expected, sustained response rates were better in patients receiving peg-IFN-α [SVR: 70/150 (46.6%)] than in patients receiving standard IFN-α [SVR: 22/60 (36.6%); P = 0.12].
Distribution of IL-6 genotypes in HIV/HCV coinfected patients [43GG (38.4%), 45GC (40.2%), 24CC (21.4%)] was comparable to that seen in healthy individuals [29GG (29%), 47GC (47%), 24CC (24%)], HIV monoinfected patients [71GG (36%), 90GC (45.7%), 36CC (18.3%)], and HCV monoinfected patients [70GG (33.3%), 103GC (49.1%), 37CC (17.6%)], respectively. All distributions were in accordance with the Hardy-Weinberg equilibrium. Furthermore, no differences in genotype distribution could be observed between HIV-positive patients with acute [22GG (42.3%), 22GC (42.3%), 8CC (15.4%)] and chronic hepatitis C [23GG (38.3%), 22GC (36.7%), 15CC (25%)], although chronically HCV-infected patients encompassed a slightly higher proportion of IL-6 low producers (25% versus 15.4%, NS). Finally, no significant differences could be detected for demographic variables, route of infection, liver enzymes, HCV genotype, HCV, or HIV load between the IL-6 genotypes in HCV/HIV coinfected patients (Table 3).
Table 3. Distribution of IL-6 Genotypes Among HCV/HIV Coinfected Patients
Number of cases (number/total in %).
MSM; men who have sex with men.
eNA, not analyzed.
fND indicates < 200,000 copies/ml.
P-values refer to genotype distribution (chi-square test)
In HCV/HIV coinfected patients, SVR rates differed significantly between carriers of a low and high producer genotype (33% versus 70%; P = 0.002) (Fig. 1A). Moreover, allele frequencies were significantly different between patients who achieved an SVR and those who did not (Armitage′s trend test: P = 0.015; odds ratio = 1.97), with a higher frequency of C alleles in the group of nonresponders than in responders (52% versus 33%; P = 0.008; odds ratio = 2.23; 95% confidence interval: 1.2-4.7).
In a subgroup analysis, these data could also be confirmed when HIV-positive patients with acute or chronic hepatitis C were analyzed separately. In the case of acute HCV infection, carriers of a high producer genotype were significantly more likely to achieve an SVR than the patients with a low producer genotype (74% versus 33%; P = 0.045). This was also true in the subgroup of HIV(+) patients with chronic hepatitis C (66% versus 33%; P = 0.035) (Fig. 1B).
When HCV(+)/HIV(+) patients were compared with respect to HCV genotypes, patients with genotype 1 infection had a significantly lower SVR rate as compared with patients infected with HCV genotypes non-1 infection (56% versus 80%, P = 0.02, odds ratio: 3.24, 95% confidence interval: 1.2-8.8) (Table 4).
Table 4. Logistic Regression Analysis Describing the Effect of Different Variables on SVR in HCV/HIV Coinfection
OR (95% CI)
The regression model was obtained by a forward conditional stepwise regression analysis.
Reference group; OR odds ration; CI confidence interval; NA not applicable.
Parameter was removed from the final model by the regression analysis.
Importantly, an effect of IL-6 genotypes on treatment outcome was only seen in the subgroup of patients with HCV genotypes 1 (P = 0.002), whereas SVR rates were not affected by the IL-6 genotype in patients infected with other HCV genotypes (Fig. 2).
Next, we performed a univariate analysis (Table 4) to identify the relative contributions of possibly confounding factors (viral load, age, sex, HCV genotype) other than IL-6 on outcome of HCV therapy in HCV/HIV-positive patients.
In the subgroup of HIV(+) patients with chronic hepatitis C, both HCV genotype (HCV genotype 1 versus non-1) and the IL-6 polymorphism were significantly associated with response to treatment (Table 4), whereas in acute hepatitis C only IL-6 genotype was a predictor of response.
Finally, we analyzed our data in a stepwise forward conditional regression model (Table 4). When coinfected patients were analyzed as a combined group using acute versus chronic hepatitis C as an additional variable, IL-6 genotype remained a strong predictor for SVR, whereas HCV load and age were removed from the final regression model. When patients with acute and chronic hepatitis C were analyzed in separate regression models, the IL-6 high producer genotype was confirmed as an independent prognostic factor of SVR in the group HIV-positive individuals with acute hepatitis C.
Next, we analyzed whether the IL-6 genotype also affects treatment outcome in HCV monoinfected patients. Based on our data in HIV/HCV coinfection, only patients with a HCV genotype 1 monoinfection were included. Overall, SVR was achieved in 22 of 60 (36.6%) and 79 of 151 (46.6%) of patients treated with standard and pegylated interferon combination therapy, respectively.
Response rates in patients treated with standard IFN-α were also significantly affected by the IL-6 genotype: 10 of the IL-6 174GG carriers achieved an SVR (62.5%) compared with 9 responders in IL-6 174CG (31%) and 3 responders in CC patients (20%) (P = 0.034). In HCV genotype 1, monoinfected patients treated with pegylated interferon/ribavirin carriers of a high producer genotype showed a tendency toward better SVR rates than patients with a low producer genotype. However, this difference did not reach statistical significance (Fig. 3).
To understand a possible mechanism underlying the effect of IL-6 genotypes on outcome of interferon therapy, we transiently transfected HUH7 hepatoma cells with HCV core protein (Fig. 4A).
At day 2 posttransfection, cells were cultured with 500 IU/ml IFN-α for 24 hours followed by analysis of total STAT3 and phosphorylated STAT3 proteins. Confirming previous studies, we found that the amount of phosphorylated STAT3 was markedly reduced when HCV core-transfected cells were exposed to IFN-α, whereas total amount of STAT3 proteins was not altered. Addition of IL-6 in concentrations corresponding to serum levels found in carriers of a low (1.5 pg/ml) or high (3 pg/ml) producer genotype,18 respectively, resulted in a dose-dependent restoration of STAT3 phosphorylation (Fig. 4B). Addition of IL-6 to the interferon-stimulated cells did not alter the total amount of STAT 3 proteins (data not shown).
Response to HCV therapy depends on both viral and host factors. Among the host factors, cytokines, which are crucially involved in the regulation of antiviral immune responses, are likely candidates to affect the response to antiviral therapy.19–22 For instance, the production of inappropriate levels of tumor necrosis factor alpha and interleukin-10 have been reported to contribute to viral persistence and to affect response to treatment.24–26
Here, we analyzed the effect of the IL-6 C174G polymorphism on response to HCV-specific treatment in patients with HIV coinfection. We identified the IL-6 high producer genotype as independent predictor of sustained virological response to HCV-specific treatment with pegylated interferon in HCV/HIV coinfected patients, irrespective of whether patients were acutely or chronically infected with hepatitis C virus. Performing a subgroup analysis, we found that this effect of the IL-6 genotype was restricted to patients infected with the difficult-to-treat HCV genotype 1. In patients with a HCV genotype 1 monoinfection, the IL-6 high producer genotype only predicted treatment outcome when standard interferon was used for treatment. A comparable trend was observed in patients treated with pegylated interferon but failed to reach statistical significance.
Thus, our data suggest that the IL-6 genotype affects treatment outcome when therapy is conducted under unfavorable circumstances (HCV/HIV coinfection plus difficult-to-treat HCV genotype or HCV genotype 1 monoinfection and treatment with standard interferon).
At the moment, why the more potent pegylated interferon widely compensated the negative effect of the IL-6 174 CC genotype in HCV monoinfection but not in HCV/HIV coinfection remains unclear. Both study groups (HCV monoinfected and HCV/HIV coinfected) differed significantly with respect to sex distribution and risk factors for acquisition of HCV infection, respectively. However, IL-6 genotypes were equally distributed in men and women and across the presumed routes of transmission of HCV infection. Furthermore, regression analysis of the 2 study groups did not identify “hemophilia,” “MSM,” or sex as an independent predictor of response to treatment (Table 2 and data not shown). Thus, differences in the study populations were unlikely to account for the diverse effect of the IL-6 genotype.
Alternatively, the more potent combination therapy with pegylated interferon might induce a shift of the cytokine profile including altered secretion of IL-6. In this case, changes in the cytokine milieu due to administration of pegylated interferon might compensate the negative impact of the IL-6 low producer genotype. In the case of coinfected patients, therapy with pegylated interferon/ribavirin might not be sufficient to induce an efficient secretion of antiviral cytokines due to the HIV-mediated impairment of the immune system. This could explain why the IL-6 genotype affects response to pegylated interferon in coinfected but not in HCV monoinfected patients.
Recently, Zhu et al.12 demonstrated that phosphorylation and activation of STAT3 is involved in the antiviral activity of IFN-α. Both in vitro and in vivo studies suggest that HCV has evolved strategies to counteract STAT3 activation. Larrea and colleagues13 demonstrated that STAT3 expression is reduced in HCV-infected livers. Furthermore, the HCV core protein has been shown to prevent phosphorylation of STAT3,13–15 which has been associated with resistance of HCV against IFN-α.16
In line with these reports, our in vitro studies confirmed impaired STAT3 activation in HCV core-transfected HUH7 cells. We further demonstrated that IL-6 can overcome HCV core-induced inhibition of STAT3 activation in a dose-dependent manner. Of note, intrahepatic STAT3 is mainly mediated by interleukin-6.17 The IL-6 genotype is associated with different IL-6 serum levels. Thus, our data are in line with a model in which the higher levels of IL-6 in patients with an IL-6 high producer genotype are associated with enhanced STAT3 activation and thus overcome the negative effect of HCV core, preventing resistance to IFN-α therapy. Further studies are warranted to test this concept in the in vivo situation, including analysis of intrahepatic STAT3 activation and IL-6 levels.
Whether the IL-6 polymorphism also affects spontaneous recovery from HCV infection in HIV-positive subjects remains to be clarified. However, the finding that HCV/HIV coinfected subjects displayed a similar IL-6 genotype distribution as healthy controls argues against a significant impact on the natural course of HCV infection. This suggestion is in line with previous reports demonstrating that neither IL6 serum levels nor the IL-6 genotype affects outcome of hepatitis C.27–30
In conclusion, our data emphasize the importance of genetic factors for the response to HCV-specific therapy, which my become unmasked in conditions of immunosuppression such as occurs in HCV/HIV coinfected patients.