Potential conflict of interest: Dr. Newmann is a consultant for, is on the speakers' bureau of, and received grants from Roche, Human Genome Sciences, Merck. Dr. Pawlotsky is a consultant for and advises Roche. Dr. Zeuzem is a consultant for, advises, and is on the speakers' bureau of Roche. He on the speakers' bureau of Schering-Plough. Dr. Westin is on the speakers' bureau of Roche and Schering-Plough. Dr. Lagging is on the speakers' bureau of Roche, Abbott, and MSD.
High systemic levels of interferon-gamma-inducible protein 10 kDa (IP-10) at onset of combination therapy for chronic hepatitis C virus (HCV) infection predict poor outcome, but details regarding the impact of IP-10 on the reduction of HCV RNA during therapy remain unclear. In the present study, we correlated pretreatment levels of IP-10 in liver biopsies (n = 73) and plasma (n = 265) with HCV RNA throughout therapy within a phase III treatment trial (DITTO-HCV). Low levels of plasma or intrahepatic IP-10 were strongly associated with a pronounced reduction of HCV RNA during the first 24 hours of treatment in all patients (P < 0.0001 and P = 0.002, respectively) as well as when patients were grouped as genotype 1 or 4 (P = 0.0008 and P = 0.01) and 2 or 3 (P = 0.002, and P = 0.02). Low plasma levels of IP-10 also were predictive of the absolute reduction of HCV RNA (P < 0.0001) and the maximum reduction of HCV RNA in the first 4 days of treatment (P < 0.0001) as well as sustained virological response (genotype 1/4; P < 0.0001). To corroborate the relationship between early viral decline and IP-10, pretreatment plasma samples from an independent phase IV trial for HCV genotypes 2/3 (NORDynamIC trial; n = 382) were analyzed. The results confirmed an association between IP-10 and the immediate reduction of HCV RNA in response to therapy (P = 0.006). In contrast, pretreatment levels of IP-10 in liver or in plasma did not affect the decline of HCV RNA between days 8 and 29, i.e., the second-phase decline, or later time points in any of these cohorts. Conclusion: In patients with chronic hepatitis C, low levels of intrahepatic and systemic IP-10 predict a favorable first-phase decline of HCV RNA during therapy with pegylated interferon and ribavirin for genotypes of HCV. (HEPATOLOGY 2010.)
Interferon-gamma inducible protein 10 kDa (IP-10 or chemokine C-X-C motif ligand 10 [CXCL10]) is a chemotactic CXC chemokine of 77 amino acids length in its mature form.1, 2 IP-10 targets the CXCR3 receptor but unlike other CXC chemokines, IP-10 lacks chemotactic activity for neutrophils and instead attracts T lymphocytes, natural killer cells, and monocytes.2–6 IP-10 is produced by a variety of cells, including hepatocytes,7, 8 and has been implicated in the pathophysiology of multiple sclerosis,9, 10 diabetes mellitus,11, 12 and human immunodeficiency virus.13, 14 High systemic levels of IP-10 are found in a large proportion of patients with chronic hepatitis C, and baseline levels of IP-10 are reportedly elevated in patients infected with hepatitis C virus (HCV) genotypes 1 or 4 who do not achieve a sustained virological response (SVR) after completion of therapy.15 Furthermore, associations between higher baseline IP-10 levels and high viral load, high body mass index (BMI), and the presence of bridging fibrosis or cirrhosis have been reported.15 In difficult-to-treat patients infected with HCV genotype 1, cut-off levels of 150 and 600 pg/mL have yielded positive and negative predictive values for SVR of 71% and 100%, respectively.16
The cellular source of plasma IP-10 in chronic hepatitis C as well as the impact of IP-10 on the detailed kinetics of viral decline during therapy, however, remain unclear. We therefore assessed the expression of IP-10 messenger RNA (mRNA) in pretreatment liver biopsies from HCV-infected patients in relation to plasma levels of IP-10 and the association between both intrahepatic and plasma IP-10 and HCV RNA reduction during therapy, in particular regarding the first and second phase of viral decline.
BMI, body mass index; DITTO-HCV, Dynamically Individualized Treatment of Hepatitis C Infection and Correlates of Viral/Host Dynamics study; HCV, hepatitis C virus; IP-10, interferon-gamma inducible protein 10 kDa; mRNA, messenger RNA; RVR, rapid virological response; SVR, sustained virological response.
Patients and Methods
Between February 2001 and November 2003, 270 patients (180 men and 90 women) were recruited in a phase III, open-label, randomized, multicenter trial conducted by the DITTO-HCV (Dynamically Individualized Treatment of Hepatitis C Infection and Correlates of Viral/Host Dynamics) study group at nine centers in France, Germany, Greece, Israel, Italy, Netherlands, Spain, Sweden, and Switzerland, as previously reported.17 All patients were adults, had compensated liver disease, were treatment-naïve for hepatitis C, and fulfilled the following inclusion criteria: a positive test for anti-HCV antibody, an HCV RNA level >1000 IU/mL, and two serum alanine aminotransferase values above the upper limit of normal within 6 months of treatment initiation. A total of 264 patients had pretreatment plasma available for IP-10 analysis, and 73 of these patients had liver biopsies from which RNA could be extracted for further evaluation (baseline characteristics are shown in Table 1).
Table 1. Baseline Characteristics
DITTO-HCV Patients with Baseline Liver Biopsy for IP-10 (n = 73)
DITTO-HCV Patients with Baseline Plasma for IP-10 (n = 264)
NORDynamIC Patients with Baseline Plasma for IP-10 (n = 359)
Data presented as
mean (standard deviation), †median (range), or
A total of 60 of 73 liver biopsies could be retrieved for histological evaluation.
A total of 226 of 264 liver biopsies could be retrieved for histological evaluation.
A total of 331 of 359 liver biopsies could be retrieved for histological evaluation.
ALT, alanine aminotransferase; ULN, upper limit of the normal range.
In addition, we evaluated pretreatment plasma samples in 382 treatment-naïve patients infected with HCV genotype 2/3 at 31 centers in Denmark, Finland, Norway, and Sweden who were randomized to 12 or 24 weeks of peginterferon α-2a 180 μg/week plus ribavirin 800 mg/day during the NORDynamIC trial.18 A total of 359 of these patients had pretreatment plasma available for IP-10 analysis.
All patients in the DITTO-HCV trial were initially treated for 6 weeks with 180 μg peginterferon α-2a administered subcutaneously once weekly (Pegasys; F. Hoffmann-La Roche, Basel, Switzerland) and ribavirin orally twice daily (Copegus; F. Hoffmann-La Roche) at a total daily dose of 1000 mg for patients weighing less than 75 kg and 1200 mg daily for those patients above 75 kg. After 6 weeks of therapy, half of the patients were randomized based on their viral kinetic classification to receive individualized therapy or to continue on standard combination therapy for a total of 48 weeks. Thus, only reductions of HCV RNA during these first 6 weeks are included in the analyses in this study.
In the NORDynamIC trial, patients were randomized at study entry to either 12 or 24 weeks of therapy with 180 μg peginterferon α-2a administered subcutaneously once weekly (Pegasys; F. Hoffmann-La Roche, Basel, Switzerland) and ribavirin twice daily (Copegus; F. Hoffmann-La Roche) at a total daily dose of 800 mg daily.
Genotyping of HCV was performed using the Inno-LiPA HCV II assay (Innogenetics NV, Ghent, Belgium).
HCV RNA Quantification.
HCV RNA was determined by reverse transcription polymerase chain reaction (RT-PCR) using Cobas Amplicor HCV Monitor version 2.0 (Roche Diagnostics, Branchburg, NJ), and quantified on days 0, 1, 4, 7, 8, 15, 22, and 29, at the end of treatment, and 24 months after the completion of treatment in the DITTO-HCV Trial.
In the NORDynamIC trial, HCV RNA was determined by RT-PCR of plasma using Cobas AmpliPrep/Cobas TaqMan HCV Test (Roche Diagnostics, Branchburg, NJ) on days 0, 3, 7, 8, and 29, week 8, week 12, at the end of treatment, and 24 weeks after the completion of therapy.
Classification of Treatment Outcome.
Patients were classified as having achieved a rapid virological response (RVR) if HCV RNA was undetectable (<50 IU/mL in the DITTO-HCV trial) in plasma on treatment day 29, and were classified as having an SVR if HCV RNA was undetectable in plasma 24 weeks after the completion of therapy.
Liver biopsies were obtained from patients in the DITTO-HCV trial within 12 months prior to inclusion in the study, and liver biopsy samples were processed for both histological evaluation (≥1.5 cm) and for RNA analysis (≥1 cm). The biopsy material for RNA analysis was immediately immersed in RNAlater (Ambion, AMS Technology, Cambridgeshire, UK) and stored at −70°C until assayed. In total, RNA from 72 liver biopsies could be retrieved and evaluated.
Histological Evaluation of Liver Biopsies.
Only biopsies with a length exceeding 1.5 cm and containing more than six portal tracts were evaluated. In total, liver biopsies from 228 infected patients in the DITTO-HCV trial were retrieved and evaluated. For each biopsy, a hematoxylin-eosin stain and a Sirius Red stain were centrally staged and graded by two independent observers experienced in pseudo-numerical scoring of liver biopsies in a blinded fashion according to the Ishak protocol.19 Equivocal issues were debated after the independent scores were noted, and a consensus score was obtained. In addition, steatosis was graded as follows: absent = 0, less than 30% of hepatocytes involved = 1, 30%–70% of hepatocytes involved = 2, and >70% of hepatocytes involved = 3.20
IP-10 mRNA Quantification in Liver Biopsies.
Total RNA was isolated from liver biopsies using the RNeasyMini Kit (Qiagen) and subsequently treated with deoxyribonuclease I. RNA integrity was assessed using RNA 6000 nanochips with an Agilent 2100 Bioanalyzer. First-strand complementary DNA was synthesized from 500 ng purified RNA using the SuperScript II Ribonuclease H(−) reverse transcriptase (Invitrogen) and random hexadeoxynucleotides. For real-time PCR, the Human SYBR Green QuantiTect Primer Assay for IP-10 (CXCL10, cat. no. QT01003065) was used (Qiagen). Primers matching the highly conserved 5′ untranscribed region of the different HCV genotypes (forward: 5′-AGCGTCTAGCCATGGCGT-3′; reverse: 5′-GGTGTACTCACCGGTTCCG-3′) were from TIB MolBiol. Reactions were performed using a 7900HT Real-Time PCR System (Applied Biosystems) and all samples were assayed in triplicate. Optical data obtained were analyzed using the default and variable parameters available in the Sequence Detection Systems software (SDS, version 2.2.2; Applied Biosystems). Expression level of target gene was normalized using as endogenous control genes the eukaryotic translation elongation factor 1 alpha 1 (forward: 5′-AGCAAAAATGACCCACCAATG-3′; reverse: 5′-GGCCTGGATGGTTCAGGATA-3′) and the beta glucuronidase (forward: 5′-CCACCAGGGACCATCCAAT-3′; reverse: 5′-AGTCAAAATATGTGTTCTGGACAAAGTAA-3′). The expression level of IP-10 mRNA in the first of the 72 liver biopsy evaluated was chosen as the reference, and assigned 1.0 arbitrary units (AU).
IP-10 Quantification in Plasma.
Quantification of IP-10 was performed using Quantikine (R&D Systems, Minneapolis, MN), a solid-phase enzyme-linked immunosorbent assay, on plasma samples obtained during the week prior to the start of therapy. All samples were stored at −70°C until assayed.
Individual characteristics between groups were evaluated using the Wilcoxon-Mann-Whitney U-test, and Spearman's rank correlation coefficient rs test was utilized to evaluate relationships between variables. Viral reduction was evaluated using Kaplan-Meier cumulative survival plots displaying the proportion of patients attaining serum HCV RNA level <50 IU/mL during the initial 6 weeks of therapy (during which time all patients received the same treatment). The log-rank test was used for comparison between the Kaplan-Meier plots with patients grouped as to whether they had intrahepatic IP-10 mRNA levels above or below the median level (i.e., 0.8 arbitrary units). All statistical analyses were performed using StatView for Macintosh (version 5.0; SAS Institute Inc., Cary, NC). All reported P values are two-sided, and P values < 0.05 were considered significant.
The treatment studies were approved by ethical committees, and conformed to the guidelines of the 1975 Declaration of Helsinki. Informed consent was obtained from each patient included in this study.
A strong association was observed between intrahepatic expression of IP-10 mRNA and plasma IP-10 levels (Fig. 1; rs = 0.707; P < 0.0001). Moreover, intrahepatic IP-10 mRNA was significantly associated with the time-point of the first undetectable HCV RNA sample while on treatment, as previously reported for plasma IP-10,15 when patients were grouped as having liver IP-10 mRNA expression above or below the median, i.e., 0.8 arbitrary units (Fig. 2; P = 0.015 for all patients, P = 0.017 for patients infected with genotype 1 or 4 and P = 0.028 for patients infected with genotype 2 or 3; log-rank test). Similarly, intrahepatic IP-10 mRNA below the median was significantly associated with an RVR (defined by undetectable plasma HCV RNA on treatment day 29) in patients infected with genotype 1 or 4 virus (P = 0.035; Mann-Whitney U-test) but not in those carrying genotype 2 or 3 (P = 0.36); these results are coherent with those previously reported for plasma IP-10.16
The reduction of HCV RNA during treatment with peginterferon and ribavirin characteristically follows two phases: an initial phase (“first-phase decline”) referring to the viral decline during the first day or days of therapy, and a slower second phase which is usually defined as the viral decline from the second to the fourth week. These phases are assumed to reflect the antiviral action of interferon (first phase) and the elimination of infected hepatocytes (second phase). Knowledge about the impact of IP-10 on the different phases of viral elimination thus may be helpful in understanding the mechanisms of IP-10–related resistance to therapy.
Figure 3 shows the association between intrahepatic IP-10 mRNA or plasma IP-10 and the first day decline in HCV RNA, i.e. the reduction observed during the first 24 hours after the initiation of peginterferon/ribavirin therapy for all genotypes. Regardless of genotype, patients who achieved more than 1.0 log10 reduction (i.e., the median reduction) of HCV RNA during the first day of therapy had a significantly lower degree of intrahepatic IP-10 mRNA expression than those who did not (genotype 1/4 median 0.58 versus 1.30, P = 0.01; genotype 2/3 median 0.53 versus 1.48, P = 0.02; Mann-Whitney U-test) as well as significantly lower plasma IP-10 (genotype 1/4 median 190 versus 315 pg/mL, P = 0.0008; genotype 2/3 median 158 versus 350 pg/mL, P = 0.002). This association remained significant whether or not the absolute reduction of HCV RNA was greater than the median 1.4 log10 the first 4 days of therapy (genotype 1/4 median 202 versus 315 pg/mL, P = 0.001; genotype 2/3 median 177 versus 266 pg/mL, P = 0.03), as well as if the maximum reduction of HCV RNA the first 4 days of therapy was greater than the median 1.5 log10 or not (genotype 1/4 median 193 versus 319 pg/mL, P < 0.0001; genotype 2/3 median 178 versus 345 pg/mL, P = 0.03). Sixty-three of 266 patients (24% of all patients, 31% of genotype 1/4 and 7% of genotype 2/3) had a rebound of HCV RNA between treatment day 1 and 4.
The association between lower plasma IP-10 and a pronounced reduction of HCV RNA during the first day of therapy remained significant regardless of whether patients had significant fibrosis (i.e., bridging fibrosis or cirrhosis, referring to Ishak stages 3–6), steatosis, or BMI above 25 kg/m2 (Fig. 4). Patients with fibrosis had significantly higher baseline plasma IP-10 than those without fibrosis (median 340 versus 188 pg/mL, P < 0.0001), and patients with steatosis had significantly higher IP-10 levels than those without (median 266 versus 203 pg/mL, P = 0.03), as did those with BMI ≥ 25 kg/m2 as compared to with < 25 kg/m2 (median 289 versus 194 pg/mL, P = 0.007).
In contrast, no significant association was observed between the intrahepatic IP-10 mRNA expression or plasma IP-10 and the reduction of HCV RNA between treatment days 8 and 29, i.e., the second-phase decline in HCV RNA,21 although a nonsignificant trend toward greater reduction was noted in patients having lower intrahepatic and plasma levels of IP-10 for genotypes 1 or 4 (Fig. 5). Ten (4%) patients had a flat partial response (eight with genotype 1 and two with genotype 4) and 22 (8%) were null responders (17 with genotype 1, one with genotype 2, and four with genotype 4).17 Even if these patients were removed from the analysis, or if the analysis was restricted to only these patients, no significant association was noted between the intrahepatic IP-10 mRNA expression or plasma IP-10 and the reduction of HCV RNA between treatment days 8 and 29. Similarly, no association was noted between the reduction of HCV RNA between treatment day 29 and week 6, i.e., a third phase of HCV viral decay that is sometimes observed and is reported to reflect the intrinsic death rate of infected cells,21 for patients with a flat partial response or null response (P = 0.78).
In agreement with previous studies,15, 16 patients infected with genotype 1/4 who achieved SVR had significantly lower baseline levels of plasma IP-10 than those who did not (median 190 versus 321 pg/mL, P < 0.0001). No significant association between IP-10 and SVR was noted for patients infected with HCV genotype 2/3, but it should be noted that only 9 (12%) of these patients failed to achieve SVR in the DITTO-HCV study.
To corroborate the relationship between first-phase viral decline and IP-10, we analyzed pretreatment plasma samples from an independent phase IV treatment trial for treatment-naïve patients infected with HCV genotype 2/3 (the NORDynamIC trial; n = 382).18 In this trial, the first plasma sample during therapy was drawn on day 3. Patients who achieved more than 0.7 log10 reduction/day (i.e., the median reduction) of HCV RNA during the first 3 days of therapy had significantly lower pretreatment IP-10 levels (P = 0.0057; Fig. 6). In agreement with the results from the DITTO-HCV trial, no significant association was observed between plasma IP-10 and the reduction of HCV RNA between treatment days 8 and 29 in this trial (P = 0.16).
The first-phase decline in HCV RNA during combination therapy, i.e., the rapid viral elimination during the first day or days of treatment, is assumed to result from the blocking of the production or release of virions, and thus primarily reflects the antiviral effectiveness of interferon.21, 22 This phase of viral elimination is reportedly dependent on the fibrosis stage, steatosis grade, gamma-glutamyl transpeptidase level, and homeostasis model assessment of insulin resistance (HOMA-IR) score, but independent of baseline viral load or alanine aminotransferase levels.23 Similarly, the first-phase viral decline predicts the rate of the slower second-phase decline,24 as well as the final treatment outcome.23, 25 In contrast, the second-phase decline, which is typically defined as the slower viral elimination during the second to fourth week of therapy,21 is assumed to reflect the death rate of infected cells and, speculatively, immune-mediated clearance of residual infection. This phase of viral decay has been demonstrated to be inversely correlated with baseline viral load and low baseline alanine aminotransferase levels.22
In this light, our finding that baseline IP-10 levels were strongly associated with the decline of HCV RNA during the first phase of viral decline for all genotypes, and not so with the second phase was unexpected considering previous suggestions that IP-10 concentrations mirror the degree of local chemokine signaling in HCV-infected hepatocytes aimed at recruiting CXCR3-expressing mononuclear cells to the infected liver,16 which intuitively should exert its effect on the second-phase viral reduction. Our findings tentatively suggest that IP-10, or related interferon-induced gene products, impinge upon the antiviral or immediate immune-activating properties of interferon. Although this hypothesis should be the subject of further study, it is of note that nonresponders to HCV therapy reportedly have a high expression of interferon-stimulated genes, including those encoding for IP-10, in pretreatment liver biopsies, thus supporting the idea that ongoing transcription of interferon-stimulated genes may affect the therapeutic efficacy of interferon.26
In this study, we also demonstrate that plasma concentrations of IP-10 mirror the levels in the HCV-infected liver, as indicated by a significant association between pretreatment expression of intrahepatic IP-10 mRNA and plasma IP-10. Previous immunohistochemical studies of liver biopsies have revealed that hepatocytes, but not liver-infiltrating mononuclear cells, express IP-10 in ongoing HCV infection.8 The strong correlation observed between liver and plasma IP-10 in this study implies that IP-10 in chronic HCV infection originates from hepatocytes, although it cannot be formally excluded that the intrahepatic IP-10 mRNA may also originate from other intrahepatic cells.
In conclusion, we propose that (1) there is a strong association between pretreatment expression of intrahepatic IP-10 mRNA and plasma concentrations of IP-10, indicating that HCV-infected hepatocytes are likely the primary source of plasma IP-10 in chronic HCV infection, (2) the intrahepatic expression of IP-10 mRNA predicts HCV viral kinetic response, (3) low IP-10 levels both in the liver and plasma before the onset of treatment are associated with a pronounced first-phase reduction of HCV viral loads for all viral genotypes, and (4) low pretreatment systemic IP-10 levels are associated with SVR for genotype 1/4. The role of IP-10 in interferon resistance, in particular regarding its potential role in the immediate actions of interferon, warrants further investigation.
We thank Marie-Louise Landelius and Ulla Gingsjö for expert technical assistance.