SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Interferon (IFN)-α–based therapy for chronic hepatitis C is effective in fewer than 50% of all treated patients, with a substantially lower response rate in black patients. The goal of this study was to investigate the underlying host transcriptional response associated with interferon treatment outcomes. We collected peripheral blood mononuclear cells from chronic hepatitis C patients before initiation of IFN-α therapy and incubated the cells with or without IFN-α for 6 hours, followed by microarray assay to identify IFN-induced gene transcription. The microarray datasets were analyzed statistically according to the patients' race and virological responses to subsequent IFN-α treatment. The global induction of IFN-stimulated genes (ISGs) was significantly greater in sustained virological responders compared with nonresponders and in white patients compared with black patients. In addition, a significantly greater global induction of ISGs was observed in sustained virological responders compared with nonresponders within the group of white patients. The level of IFN-induced signal transducer and activator of transcription (STAT) 1 activation, a key component of the Janus kinase (JAK)-STAT signaling pathway, correlated with the global induction of ISGs and was significantly higher in white patients than in black patients. In conclusion, both treatment outcome and race are associated with different transcriptional responses to IFN-α. Because this difference is evident in the global induction of ISGs rather than a selective effect on a subset of such genes, key factors affecting the outcome of IFN-α therapy are likely to act at the JAK-STAT pathway that controls transcription of downstream ISGs. (HEPATOLOGY 2006;44:352–359.)

Various interferon (IFN)-α preparations, some in combination with ribavirin, are currently the only treatment for chronic hepatitis C virus (HCV) infection approved by the US Food and Drug Administration. Two recently developed pegylated IFNs (PEG-IFNs), interferon PEG-IFN alpha-2b and PEG-IFN alpha-2a, have resulted in improved efficacy. However, the sustained virological response rates for patients infected with HCV genotype 1 and with high viral load, the most common profile in this country, remains less than 50%.1, 2 The sustained virological response rate has been shown to be associated with multiple viral and host factors. In particular, the response rate in black patients is significantly lower than that seen in white patients.3–8

The underlying mechanisms that determine the efficacy of IFN-α in different patient populations are poorly understood. IFN initiates signals from the cell surface to the nucleus, resulting in the upregulated expression of multiple genes described as IFN-stimulated genes (ISGs).9, 10 It was recently reported that upregulated baseline expression of a set of ISGs in patients with chronic HCV infection11 or HCV and HIV coinfection12 predicted nonresponse to subsequent interferon therapy. However, variable treatment outcomes could also be mediated by the differential induction of ISGs after administration of IFN-α, either through the direct antiviral function or the immune regulatory function of ISG proteins.13, 14

Previously, we investigated ISG induction in peripheral blood mononuclear cells (PBMCs) treated with IFN15 and demonstrated that the overall patterns of ISG induction were remarkably similar in PBMCs treated ex vivo and in vivo. We hypothesized that different expression patterns of ISGs would be associated with variation in treatment outcomes after IFN-α therapy among individual patients, as well as with the different response rates to IFN-α therapy based on race. To test this hypothesis, we used microarrays to measure the pretreatment transcriptional responses of HCV-infected patients using PBMCs treated with IFN-α ex vivo. The microarray datasets were analyzed according to the outcome of subsequent treatment or race. In addition, we measured the activation of signal transducer and activator of transcription (STAT) 1, a key molecule of the Janus kinase (JAK)-STAT pathway that controls the transcription of downstream ISGs, in IFN-treated PBMCs and correlated the levels of STAT1 activation with the induction of ISGs measured with microarrays.

Patients and Methods

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Patients.

Fifty-eight patients with chronic HCV infection confirmed via qualitative HCV reverse-transcriptase polymerase chain reaction tests (Amplicor, Roche Molecular Diagnostics, Branchburg, NJ) were enrolled in the study, the goal of which was to examine the relationship between treatment outcome and ISG induction (Table 1). This patient population included 55 treatment-nav̈e patients who were studied prospectively prior to antiviral therapy. In addition, 3 nonresponders to PEG-IFN/ribavirin therapy were studied retrospectively after completion of therapy. Race was based on self-reporting, and patients were categorized as black (n = 24) or white (including Hispanic white) (n = 34). A pretreatment venous blood sample was collected using a cell preparation tube with sodium citrate (BD Vacutainer Systems, Franklin Lakes, NJ) for the isolation of PBMCs. Forty of the patients were subsequently treated according to standard clinical practice guidelines.16 Thirty-eight patients completed IFN-α therapy with long-term posttreatment follow-up. An end-of-treatment responder was defined as a patient with undetectable serum HCV RNA via Amplicor assay (≤50 IU/mL) at the end of treatment. A nonresponder (NR) was defined as a subject with detectable HCV RNA at the end of treatment. In patients with genotype 1 who failed to achieve early virological response (defined as at least a 2-log drop in viral load from baseline or undetectable HCV RNA) after 12 weeks of PEG-IFN/ribavirin treatment was discontinued according to standard treatment guidelines. These patients were also considered NRs. Sustained virological responders (SVRs) were end-of-treatment responders with undetectable HCV RNA in plasma 24 weeks after the end of treatment. Those end-of-treatment responders who became HCV RNA–positive 24 weeks after the end of treatment were defined as relapsers. For other experiments, 68 treatment-naïve patients with chronic HCV infection were enrolled for studies of IFN-α–induced STAT1 activation. The study protocols were approved by the institutional review boards at Stanford University, the University of Tennessee, and the University of California–San Francisco. All patients provided informed consent to participate in the study.

Table 1. Race and Treatment Outcome of Patients Enrolled in the Microarray Study
RaceTreatment OutcomeTreatment
IFN-α/RibavirinPEG-IFNPEG-IFN/Ribavirin
  • NOTE. See Patients and Methods for details. Numbers in parentheses indicate breakdown by HCV genotype (genotype 1, genotype non-1).

  • Abbreviations: IFN, interferon; PEG-IFN, pegylated interferon; NR, nonresponder; ETR, end-of-treatment responder; SVR, sustained virological responder; NA, not applicable.

  • *

    Including 3 retrospectively recruited patients who failed a previous PEG-IFN/ribavirin treatment.

  • Including 2 patients who discontinued treatment.

White (n = 34; age 40–60)NR, 13* 5 (4, 1)08 (6, 2)
 ETR, 21SVR, 135 (3, 2)1 (1, 0)7 (6, 1)
 Relapser, 81 (1, 0)07 (7, 0)
Black (n = 24; age 46–64)NR, 7 2 (1, 1)05 (5, 0)
 Untreated, 17 (16, 1) NANANA

Microarray Assay.

Aliquots of PBMCs, isolated within 2 hours after blood drawing, were incubated at 37°C for 6 hours in RPMI 1640 medium supplemented with 10% fetal calf serum at 106 cells/mL, with or without 200 U/mL of interferon alpha-2b (Schering, Kenilworth, NJ). Preparation of fluorescence-labeled targets and complementary DNA microarray hybridization was performed as previously reported.15 In brief, total RNA was extracted from PBMCs and used for complementary DNA synthesis via reverse-transcription. Complementary DNA was amplified via in vitro transcription.17 The amplified antisense RNA derived from PBMCs incubated with and without IFN-α was labeled with Cy5 and Cy3, respectively. The Cy3- and Cy5-labeled targets were mixed and hybridized to complementary DNA microarrays as previously described.17, 18 Microarrays were obtained from Stanford Functional Genomics Facility (Stanford, CA). Each microarray contained approximately 43,000 spots corresponding to approximately 38,500 distinct human sequence-verified genes.

Signal Detection and Processing.

The fluorescence signal on microarrays was acquired using a GenePix 4000b microarray scanner (Axon Instruments, Foster City, CA). The scanned images were processed using GenePix Pro 3.0 software. The data files were entered into the Stanford Microarray Database.19 For all subsequent analyses, we included only 9,048 array elements whose expression was well measured. We defined well-measured genes by having (1) a ratio of normalized signal intensity to background noise of more than 2 for either the Cy5 signal derived from IFN-stimulated PBMCs or the Cy3 signal derived from unstimulated PBMCs and (2) a pixel regression correlation of R less than 0.6 in at least 70% of all arrays.

Permutation Analysis.

Permutation analysis was used to evaluate the statistical significance of the number of ISGs with a greater average fold change (FC) in NRs compared with SVRs or in blacks compared with whites. The patients in the two groups were permutated a total of 1,000 times to estimate the frequency of events (i.e., permutations) in which the number of genes with a greater FC in one group over the other was equal to or smaller than that in NRs versus SVRs (or in blacks vs. whites). P values were determined by multiplying the frequency by 2. A P value of less than .05 was considered statistically significant.

IFN Stimulation Index Analysis.

To determine the IFN stimulation index (ISI) of each patient, the mean of log2(FC) for each of the 132 genes with an average FC >3 (ISGFC>3) among all 58 subjects was first calculated. Next, the individual log2(FC) of each ISGFC>3in each patient was normalized by dividing it by the mean of log2(FC) of the specific gene. Finally, the normalized log2(FC) of all ISGFC>3 from each patient was averaged to yield the ISI for that patient. Unpaired 2-tailed Student t tests were used for groupwise comparison of mean ISI. A P value of <.05 was considered statistically significant.

Flow Cytometric Assay for Phospho-STAT1.

PBMCs were incubated at 37°C for 15 minutes in RPMI 1640 medium supplemented with 10% fetal calf serum at 106 cells/mL, with or without 200 U/mL of interferon alpha-2b (Schering, Kenilworth, NJ). Phosphorylated STAT1 (P-STAT1) was detected using previously described methods.20 In brief, the cells were fixed with parafomaldehyde, permeabilized with methanol, and stained with Alexa Fluor 488-conjugated anti-PSTAT1 (BD Biosciences, San Jose, CA) and analyzed with a flow cytometer. The increase of P-STAT1 level was defined as the logarithm (base 2)-transformed ratio of the geometric mean of P-STAT1 signal intensity of cells incubated with IFN-α to that of cells incubated without IFN-α.

Results

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Global Transcriptional Response to IFN-α Is Greater in SVRs Compared With NRs.

To investigate the relationship between the patterns of ISG induction and the clinical outcome of IFN treatment, we studied a total of 58 chronic HCV patients (Table 1) with microarrays. The induction of ISG was defined as the FC of messenger RNA levels in PBMCs incubated ex vivo with IFN-α compared with cells incubated without IFN-α.

A total of 313 ISGs with an average FC >2 (ISGFC>2) (Supplemental Table; Supplementary material for this article can be found on the HEPATOLOGY website (http://interscience.wiley.com/jpages/0270-9139/suppmat/index.html) and 132 genes with an average FC >3 (ISGFC>3) were identified in the 58 patients. Using a 1-class SAM analysis,21 all the ISGFC>2 were identified as being significantly upregulated by IFN-α at a false discovery rate of <1%. Because these are the genes with the greatest IFN response and are likely to play important roles in the biological effects of IFN-α, we focused our subsequent analysis on these ISGs.

Forty of the enrolled patients were subsequently treated using standard clinical practice guidelines.16 Thirty-eight patients completed the course of IFN-α therapy with known long-term treatment outcome (Table 1), while two patients discontinued treatment. We calculated the average FC of all ISGFC>2 in NRs (n = 20) and SVRs (n = 13), respectively, without regard to race (Supplemental Table A). The average FC of each ISGFC>2 in the two groups is shown in Fig. 1A. Only 17 (5.4%) of the 313 ISGs had a greater average FC in the NR group, while 296 (95%) had a greater average FC in the SVR group (Fig. 1C). If the 3 retrospectively recruited NRs were excluded, only 9 of the 313 ISGs had a greater FC in the NRs (data not shown). Of note, if genes with a smaller FC were used for the comparison, the difference between the two groups was reduced. For example, when all 1,489 genes with an average FC >1.3 were compared (Fig. 1B), 436 (29%) had a greater FC in the NR group, with the vast majority (96%) of such genes falling within the FC range of 1.3 to 2. On the other hand, if genes with a greater FC were used for the comparison, the difference between the two groups was even greater. When 132 ISGFC>3 were compared, only 4 (3%) of the 132 genes had a greater FC in the NR group.

thumbnail image

Figure 1. Groupwise comparison of average FC of messenger RNA levels of ISGs. (A) SVR versus NR comparison of 313 ISGFC>2. (B) SVR versus NR comparison of average FC for 1,488 genes with an average FC >1.3. (C) Summary of SVR versus NR comparison results. (D) Summary of black versus white comparison results. The purple bars are the number of ISGFC>2. The green bars are the number of ISGFC>2 with significantly different FC between the compared groups identified by Student t test (P < .05). P values were determined via permutation analysis (see Patients and Methods). FC, fold change; SVR, sustained virological responder; NR, nonresponder.

Download figure to PowerPoint

If the observed FC difference between the SVR and NR groups was completely random due to intrinsic variability (or noise) of the microarray assay, we would expect a roughly even distribution of genes with a greater average FC in either group. To evaluate the statistical significance of the apparently skewed distribution in which only 17 of the 313 ISGFC>2 had a greater average FC in one of the two groups, a permutation analysis (see Patients and Methods) was performed that yielded a P value of .044. A significantly skewed distribution (P = .002) was also observed if only patients infected with genotype 1 were considered (Fig. 1C, lower bars). These results suggest a systematic difference in transcriptional response between the NR and SVR groups, with a greater induction of ISGs in SVRs than in NRs.

Student t tests were used to compare the FC of each ISG among NRs and SVRs. A substantial portion (44% for patients infected with all genotypes and 61% for genotype 1–infected patients) of ISGFC>2 had significantly different FC (P < .05 in individual t test) between these two groups, and all were greater in the SVR group. These results are displayed as green bars in Fig. 1C. If the apparent significant genes (P < .05) were identified completely by chance, only 16 (313 × 0.05) such genes would be expected, and they could have a greater FC in either group.

The groupwise comparisons depicted in Fig. 1 are based on the average FC of each individual ISG across all subjects in a specific group. To further evaluate the significance of the observed differences in the global transcriptional responses, we conducted an independent statistical analysis based on the ISI, a parameter we developed to measure the overall induction level of the set of 132 ISGFC>3(see Patients and Methods). This parameter created a single metric that allowed us to quantify the global transcriptional response to IFN of each individual subject.

We performed a groupwise comparison of ISI between SVRs and NRs. The average ISI of the SVR group was significantly greater than that of the NR group for patients infected with all genotypes as well as for those infected with genotype 1 (Table 2). This result confirmed the significant difference in the overall induction of ISGs between SVRs and NRs shown in Fig. 1. Taken together, our analyses demonstrate a global difference in the transcriptional response of ISGs, rather than specific differences in a subset of ISGs, in SVRs versus NRs.

Table 2. Groupwise Comparison of IFN Stimulation Index
GroupAll GenotypesGenotype 1
nMean ISI (SD)P Value*nMean ISI (SD)P Value*
  • Abbreviations: ISI, interferon stimulation index; NR, nonresponder; SVR, sustained virological responder.

  • *

    Student t test.

NR200.92 (0.33) 150.94 (0.31) 
SVR131.27 (0.23).001101.31 (0.19).001
Black240.89 (0.32) 220.92 (0.29) 
White311.16 (0.24)<.001271.17 (0.24).002
White NR131.04 (0.24) 101.04 (0.26) 
White SVR131.27 (0.23).021101.31 (0.19).017
White NR131.04 (0.24) 101.04 (0.26) 
Black NR70.69 (0.38).06150.75 (0.33).14

Global Transcriptional Response to IFN-α Is Greater in White Patients Compared With Black Patients.

The rate of virological response to IFN-α therapy was previously found to be much lower in black patients compared with white patients. To examine if this racial difference in IFN response was associated with a differential transcriptional response to IFN, we first compared the average FC of ISGFC>2between all prospectively studied patients self-identified as black (n = 24) and white (n = 31, excluding 3 NRs to previous treatment), regardless of treatment outcome (Table 1). The average FC values of these two racial groups are listed in Supplemental Table A, while the results of the comparison are presented in Fig. 1D. Only 21 (7%) of these genes had a greater FC in blacks when patients infected with all genotypes were considered (P = .028, permutation analysis). A similarly skewed distribution was observed in patients infected with genotype 1 (P = .018, permutation analysis). Student t tests identified the majority of ISGFC>2(57% for patients infected with all genotypes and 61% for genotype 1–infected patients) to have significantly different FC between black and white patients, and all of these genes were induced to a greater extent in white patients.

We also conducted a groupwise comparison of ISI between blacks and whites (Table 2). The average ISI of white patients is significantly greater than that of black patients. Similar results were obtained when patients infected with HCV genotype non-1 were either included or excluded. This result confirmed a significantly greater global induction of ISGs in whites compared with blacks (Fig. 1).

To separately examine the effects of treatment outcome and race on global ISG induction, we conducted the following groupwise comparison of ISI: white SVRs versus white NRs and white NRs versus black NRs (Table 2). The average ISI of the white SVRs was significantly greater than that of the white NRs. The average ISI of the white NRs was greater than that of the black NRs, although the difference was not statistically significant, which is likely due to the reduced sample size for the black NR group. Taken together, these results suggest that both treatment outcome and race are associated with different transcriptional response to IFN-α.

Comparison of Downregulated Genes in Different Patient Groups.

To explore the difference in the transcription pattern of genes downregulated by IFN in different patient groups, we applied a similar ISI analysis based on a set of 174 genes with an average FC <0.67. Significant differences were not detected in the overall suppression of these genes between SVRs and NRs (P = .82) or between whites and blacks (P = .75).

Comparison of JAK-STAT Pathway Activity in Black and White Patients.

Previous results suggest a difference in the activity of IFN signal transduction pathway between blacks and whites, as well as between NRs and SVRs. To explore this issue further, we measured the level of IFN-α–induced tyrosine P-STAT1, a key measure of activation of the JAK-STAT pathway that controls the transcription of downstream ISGs. PBMCs from 23 chronic hepatitis C patients were incubated with or without IFN-α for 15 minutes, followed by a single cell-based flow cytometric assay for P-STAT1,20, 22, 23 and microarray assay for ISI, which reflects the overall induction of ISGs. A significant positive correlation was observed between ISI and P-STAT1 (Fig. 2), indicating that the greater global transcriptional responses to IFN we observed are associated with higher levels of IFN-α–induced P-STAT1.

thumbnail image

Figure 2. The global induction of ISGs correlated with phospho-activation of STAT1. PBMCs from chronic hepatitis C patients (n = 23) were incubated with or without IFN-α (200 U/mL). The increase of P-STAT1 (ΔP-STAT1) was determined via flow cytometric assay after 15 minutes of incubation. The ISI was determined via microarray analysis after 6 hours of incubation. ISI, interferon stimulation index; rS, Spearman's correlation coefficient; P-STAT1, phosphorylated STAT1.

Download figure to PowerPoint

Next, we compared the levels of P-STAT1 between black and white patients with chronic HCV infection (Fig. 3). The levels of IFN-α–induced P-STAT1 in PBMCs from white patients were significantly higher than those seen in black patients. Taken together, these results confirm that the greater overall induction of ISGs in the white patients was associated with a greater activity of the JAK-STAT pathway.

thumbnail image

Figure 3. Comparison of P-STAT1 levels in black and white patients with chronic HCV infection. PBMCs from treatment-naïve black and white patients were incubated with or without IFN-α for 15 minutes, followed by a flow cytometric assay to measure levels of P-STAT1. (A) Representative histograms of P-STAT1 signal intensity in PBMCs of a black patient and a white patient. (B) Groupwise comparison of STAT1 phospho-activation (ΔP-STAT1) between black and white patients (Student t test). Black patients: genotype 1, n = 8; genotype non-1, n = 3. White patients: genotype 1, n = 8; genotype non-1, n = 2. P-STAT1, phosphorylated STAT1; IFN, interferon.

Download figure to PowerPoint

To examine whether the variable activity of the JAK-STAT pathway in white patients compared with black patients (Fig. 3) was due to different levels of STAT1 protein, we isolated PBMCs from groups of black and white patients with chronic HCV infection and measured the levels of STAT1 protein directly in the cells incubated with or without IFN for 15 minutes (Fig. 4). Significant differences were not detected between the groups of blacks versus whites, suggesting that the racial difference in the JAK-STAT pathway is not due to different levels of STAT1 protein in these two populations.

thumbnail image

Figure 4. Levels of STAT1 protein in PBMCs from black and white patients with chronic HCV infection. PBMCs from IFN treatment-naïve patients were incubated with or without IFN-α (200 U/mL) for 15 minutes. Whole cell lysates (1 million cells/well) were resolved with SDS-PAGE, transferred onto a polyvinylidene fluoride membrane, and analyzed via Western blotting using anti-STAT1 or anti-actin (both from Lab Vision, Fremont, CA) as the primary antibodies and peroxidase-labeled anti-mouse antibodies as the secondary antibodies. The signals were developed with ECL reagent and viewed on a film. The signal intensity of specific bands was determined with a densitometer. The level of STAT1 in each patient was normalized to that of actin. (A) Western blotting of STAT1 and actin for 4 representative black patients and white patients. (B) Groupwise comparison of normalized levels of STAT1 between black patients and white patients (Student t test). Black patients: genotype 1, n = 10; genotype non-1, n = 2. White patients: genotype 1, n = 9; genotype non-1, n = 3. IFN, interferon; STAT1, signal transducer and activator of transcription 1.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

In this study, we demonstrated that clinical outcome of IFN therapy in chronic hepatitis C patients is associated with a transcriptional response to IFN. Patients who achieved long-term clearance of HCV after IFN-α therapy had a greater overall transcriptional response to IFN-α than patients who failed IFN-α treatment. Furthermore, we demonstrated that the overall induction of ISGs was greater in white patients than in black patients. Of interest, an oligonucleotide microarray study of the in vivo transcriptional responses in PBMCs of patients receiving PEG-IFN/ribavirin therapy indicated that a greater global transcriptional response was associated with an early decrease in HCV RNA levels at day 3 to day 28 (the last time point measured) of therapy compared with patients showing little or no decrease in HCV RNA levels (M. Taylor, personal communication and M. Taylor et al. Distinct differences in gene response between patients resistant to interferon/ribavirin therapy and those showing early virus clearance. The 11th International Symposium on Hepatitis C Virus and Related Viruses. Oct. 3-7, 2004, Heidelberg, Germany). Although this study differs from ours in several aspects, including strategies for examining PBMC transcriptional responses (in vivo versus ex vivo responses and 1-7 days vs. 6 hours of treatment) and the time point for treatment outcomes (early decrease in viral load up to day 28 vs. long-term elimination of virus 6 months after the end of treatment), the results of the two studies complement and support each other. Both studies found that a greater global response to IFN was associated with viral clearance. Together, these findings have important implications for the mechanisms underlying the different outcomes of IFN-α therapy.

First, our findings imply that the variable virological responses to IFN are associated with intrinsic characteristics of the IFN signal pathway in different individuals. Given that our study involved ex vivo exposure of PBMCs to IFN-α, the differences we observed are independent of any variations in uptake and metabolism of injected IFN-α due to factors such as body habitus that could potentially affect treatment outcome. Second, because the enhanced induction of ISGs in SVRs or whites is a global effect rather than a selective effect on a subset of specific ISGs, the relevant key factors—either genetic, or environmental, or both—affecting treatment outcome are likely to act at an early step in the IFN signal transduction cascade to influence the transcription of all ISGs, rather than on the transcription of a limited number of individual ISGs. The differences in transcriptional response were observed in PBMCs, which we have used as a surrogate for the primary target cell of HCV: the hepatocyte. Given that the IFN signal pathway is widely shared among different cell types, our findings with PBMCs are hypothesized to reflect the major characteristics of the overall pattern of ISG induction in the liver. Future studies conducted on hepatocytes could test this hypothesis directly.

The best characterized signal transduction pathway triggered by IFN-α is the JAK-STAT pathway. Binding of IFN to its cognate transmembrane receptors on the cell surface triggers dimerization of the receptor subunits and activation of the receptor-associated JAKs TYK2 and JAK1. The activated JAKs phosphorylate and activate STAT1 and STAT2, two of the members of the signal transducer and activator family of proteins. P-STAT1 and P-STAT2 assemble in a complex with another protein, IRF9, to form the heteromeric transcription factor ISGF3, which translocates into the nucleus and binds to cis-acting DNA elements in ISG promoters, resulting in their upregulated transcription.9, 10, 24 In the present study, we found that global ISG induction significantly correlated in individual patients with STAT1 activation in response to IFN-α (Fig. 2) and that the levels of STAT1 activation were higher in whites compared with blacks (Fig. 3), which is consistent with the greater global induction of ISGs and higher response rate to IFN therapy in white patients. Taken together, these findings support the hypothesis that the differential activity of the JAK-STAT pathway directly affects the outcome of IFN-α therapy.

In this study, we defined ISI as a quantitative measurement for the overall level of ISG induction detected via microarray analysis in individual subjects and used it to compare transcriptional response between different patient groups (Table 2). An advantage of ISI is that it allows correlation analysis to compare the global transcriptional response with other quantitative parameters of individual subjects, such as the levels of STAT1 activation. This approach will be useful for future studies of the relationship of the IFN transcriptional response to other host factors associated with the virological response to treatment.

Several viral and host factors, such as race, HCV genotype, viral load, and body weight, have been shown to be predictors of IFN-α treatment outcomes.25 The underlying mechanisms for these associations are generally unknown. In the present study, we found that the overall transcriptional response of ISGs differs between black and white patients. Future studies should be designed to elucidate the effect of each viral and host factor on the IFN signal transduction and ISG induction by assembling proper subject groups controlled for these factors, including patients infected with HCV genotype non-1. Such studies will help clarify the biological mechanism through which these factors affect the efficacy of IFN-α.

Although the PBMC samples were treated with interferon alpha-2b in the present study, a potential limitation is that the enrolled patients received different IFN treatments, including IFN/ribavirin, PEG-IFN, and PEG-IFN/ribavirin (Table 1), which have somewhat different response rates for certain patient categories. This occurred because the treatment recommendation changed during the course of this study (PEG-IFN was approved by the US Food and Drug Administration). Combination therapy with PEG-IFN and ribavirin has a response rate comparable to standard IFN/ribavirin for patients infected with HCV genotypes 2 or 3 but is more effective for genotype 1–infected patients with high viral loads.1, 2, 16 Most studies found a response rate of <15% on retreatment with PEG-IFN/ribavirin in NRs to previous IFN/ribavirin treatment.26 This suggests that only a small fraction of the NRs treated with IFN/ribavirin in our study might become SVRs if treated with PEG-IFN/ribavirin instead. Such patients would be misclassified in our study. However, there were only 4 patients in our white NR group who were infected with HCV genotype 1 and treated with IFN/ribavirin. Of note, the overall FC pattern of NR versus SVR did not change when these 4 patients were excluded from the NR group (data not shown). Therefore, it is unlikely that inclusion of these patients changed the findings or conclusions of the study.

In conclusion, we found a greater global ISG induction to IFN therapy for chronic hepatitis C in SVRs and white patients compared with NRs and black patients, respectively. These results suggest that the key factors determining outcomes of IFN-α therapy in chronic HCV-infected patients act at an upstream site of the IFN signal transduction cascade, or the JAK-STAT system. This was further supported by our direct analysis of P-STAT1. Additional studies will be required to determine if there is a genetic or other basis for the differences observed between patients with different treatment outcomes.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

We thank B. Cunningham, A. Kennedy, Dr. C. Riely, and Dr. J. Fleckenstein for assistance with enrolling the study subjects.

References

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supplementary material for this article can be found on the H EPATOLOGY website ( http://interscience.wiley.com/jpages/0270-9139/suppmat/index.html ).

FilenameFormatSizeDescription
jws-hep.21267.doc575KTable A. Average fold-change of ISGs in different patient groups

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.