| |
Pretreatment Intrahepatic CD8+ Cell Count Correlates with Virological Response to Antiviral Therapy in Chronic Hepatitis C Virus Infection
|
| |
| |
The Journal of Infectious Diseases 2003;188:1528-1532
Jan M. Vrolijk,1 Jaap Kwekkeboom,1 Harry L. A. Janssen,1 Bettina E. Hansen,1 Pieter E. Zondervan,2Albert D. M. E. Osterhaus,3 Solko W. Schalm,1 and Bart L. Haagmans3
Departments of 1Hepatology and Gastroenterology and 2Pathology, and 3Institute of Virology, Erasmus MC, University Medical Center Rotterdam, Rotterdam, The Netherlands
ABSTRACT. We analyzed the relationship between virological response and baseline immune factors in 17 patients chronically infected with hepatitis C virus (HCV) who received interferon-ribavirin therapy for 26 weeks. The number of intrahepatic CD8+ cells present in the portal tract before the start of treatment was found to be significantly higher in patients who responded to treatment than in nonresponders. The relationship between portal CD8+ cell counts and the response to therapy could be described by a logistic curve. Neither peripheral cytokine levels nor HCV-specific T cell reactivity in peripheral blood mononuclear cells showed a relationship to response to therapy.
The number of pretreatment CD8+ cells in the portal tract alone predicted response better than did either virus load or patient characteristics (multivariate logistic regression analysis including sex, age, genotype, stage of fibrosis, and virus load).
BACKGROUND
The ability of interferon (IFN) to inhibit hepatitis C virus (HCV) replication and to stimulate the clearance of infected cells is the basis of current standard antiviral therapy that uses IFN- and ribavirin. Most patients exhibit an initial response, with the loss of circulating HCV RNA. Unfortunately, complete virus clearance is not observed in all patients. Although the mechanism responsible for the eradication of HCV-infected hepatocytes is not well understood, there is consensus that this process is immune mediated. Revealing mechanistic studies have been hampered by the lack of representative laboratory models and by the difficulty of studying the intrahepatic compartment.
In peripheral blood, HCV-specific T lymphocytes are detected easily during acute HCV infection, exhibiting a broad reactivity toward different HCV antigens. In chronic hepatitis, HCV-specific T cell frequencies are low, and response to subsequent IFN- therapy is difficult to predict on the basis of this parameter. Higher frequencies of HCV-specific T cells may be found in the liver. Accordingly, liver biopsies may prove to be instrumental for the study of immune factors associated with response to subsequent IFN- therapy. The frequency of virus-specific intrahepatic T cells and absolute numbers of these cells present in the liver can possibly mirror the ability of the immune system to generate a sufficient antiviral response after the start of antiviral therapy. We examined baseline localization of diverse subsets of intrahepatic immune cells, by quantitative immunohistochemistry (CD4+, CD8+, and CD68+ cells), in pre- and end-of-treatment liver biopsy specimens and assessed causality to response to therapy for chronic HCV infection. For comparison, peripheral immune parameters, such as plasma interleukin (IL)10, IL-12, and IFN- levels and circulating HCV-specific T cells, were also determined.
Patients and Methods
Seventeen patients with chronic HCV infection who were assigned to receive IFN-2b (3 MU 3 times/week) and ribavirin (1000/1200 mg/day) (Schering-Plough) for 26 weeks in our center and from whom a pretreatment biopsy specimen was available were studied. Informed consent was obtained from all patients, and the institutional review board of the Erasmus MC approved the study.
All patients were treatment naive and HCV RNA positive, had biopsy-proven chronic HCV infection, and were seronegative for hepatitis B virus (HBV) and human immunodeficiency virus. None had clinical or biochemical evidence of other liver diseases. Blood samples for HCV RNA quantification were obtained at 0, 4, 12, and 26 weeks of therapy. Response to treatment was defined as HCV RNA levels <100 copies/mL at the end of therapy, determined by means of the Superquant assay (National Genetics Institute; limit of detection, 100 copies/mL).
Before treatment and during the last week of treatment, biopsies were performed percutaneously with a 14-gauge Tru-Cut needle.
Results
The median duration of infection was 14.5 years (range, 530 years); in 3 patients, the duration of infection could not be estimated. Median baseline virus load was 9.0 x 105 copies/mL (range, 8.0 x 104-3.5 x 107 copies/mL; mean, 5.3 x 106 copies/mL); 11 patients were infected with HCV genotype 1a/b.
All patients completed 26 weeks of IFN-2bribavirin therapy. Nine patients responded by the end of therapy (HCV RNA, <100 copies/mL), whereas, in 8 patients, HCV RNA remained detectable during the entire treatment period. No significant differences were observed with respect to pretreatment virus load or genotype between patients who responded to therapy and patients who did not. The average rate of virus load decline within the first 4 treatment weeks was 2.32 log (1.45 log in nonresponders vs. 3.32 log in responders; P = .015). The rate of decline in virus load was not different between genotype 1 and non-1 genotypes. In 4 patients, the quantitative polymerase chain reaction result for HCV RNA was negative after a 24-week treatment-free follow-up period.
No significant difference was observed between responders and nonresponders in baseline alanine aminotransferase levels (P = .321). HCV-specific lymphoproliferative responses were detected in 6 of 17 patients. However, baseline HCV-specific lymphocyte proliferation did not differ significantly between responders and nonresponders. In addition, other pretreatment peripheral markers of immune activity, such as amounts of circulating CD8+ T cells (P = .535) and plasma IL-12-p40 (P = .336), IL-10 (P = .321), and IFN- (P = .236) levels did not differ significantly.
Liver biopsy sections were evaluated histologically by determination of the Knodell score and the presence of CD4+, CD8+, and CD68+ cells. Knodell scores were similar in responders and nonresponders. CD8+ and CD68+ cells were present both in the portal tract and lobular region, but CD4+ cells were observed less frequently and were restricted mainly to the portal tracts. Amounts of CD68+ or CD4+ cells did not differ significantly between responders and nonresponders. Amounts of lobular CD8+ cells in pretreatment liver biopsy sections were similar in both groups (P = .815), but the amounts of CD8+ cells within the portal tracts were higher in responders, compared with nonresponders (P = .002). On the basis of univariate logistic regression modeling, the predicted probability of response to therapy as a function of the number of CD8+ cells in the portal tract in biopsy samples taken before treatment could be made. The number of pretreatment CD8+ cells in the portal tract alone predicted response better than did either virus load or patient characteristics (multivariate logistic regression analysis including sex, age, genotype, stage of fibrosis, and virus load). Amounts of intrahepatic CD8+ cells were not significantly different in responders and nonresponders in end-of-treatment biopsy samples. Reduced intrahepatic CD8+ cell counts in responders at the end of therapy might relate to the high relapse rate observed (5 of 9 patients experienced a virological relapse).
AUTHOR COMMENTS
IFN-alfa most likely has a pronounced direct antiviral activity in patients with HCV infection. The first dose-dependent phase of the HCV decrease curve, which lasts 24 h, can be explained assuming that IFN- mediates direct antiviral activity. The slower subsequent second phase is thought to be partly immune mediated. Indeed, we observed a significant correlation between pretreatment CD8+ cell counts and the decrease of virus titers during the first month of therapy (r = 0.514; P = .05). Both innate and virus-specific immune responses may further amplify HCV clearance during antiviral therapy, either because of their presence at the start of therapy or by IFN-ribavirinmediated activation. Several reports have demonstrated the appearance of HCV-specific T cell activity after treatment with IFN-alfa, even when baseline HCV-specific T cell reactivity was undetectable. We could not differentiate responders on the basis of lymphocyte stimulation test responses measured at the start of therapy. Rather, actual numbers of infiltrating CD8+ cells may have to be taken into account. Our observations are consistent with a study by Nelson et al., who demonstrated that the presence of HCV-specific cytotoxic T cells in the liver was associated with biochemical response to IFN-. CD8+ cells could be directly involved in the clearance of HCV from the liver during antiviral treatment through production of antiviral cytokines or the killing of infected hepatocytes. Alternatively, CD8+ cell count may correlate with other antiviral mechanisms implicated in the clearance of HCV during therapy.
The portal localization of CD8+ cells is of interest. As demonstrated in HBV infection, patients with high virus load were characterized by the presence of high amount of CD8+ cells located in the portal tracts, whereas relatively more CD8+ cells were found in the lobular region in cases of controlled HBV. It is possible that high CD8+ cell counts in the portal tract may benefit patients at later stages during therapy, when virus load is diminished and the migration of CD8+ cells into the lobular region may have been facilitated. Future studies need to further analyze frequency and localization of HCV-specific CD8+ cells in the liver.
In conclusion, the pretreatment CD8+ cell count in the liver, but not in peripheral blood, is higher in patients who respond to IFN-ribavirin therapy than in patients who do not. We suggest that immunohistological evaluation of biopsy samples may be considered with respect to factors reflecting a patient's immune system capacity to respond adequately to therapy. Moreover, these findings suggest that significant prognostic immune markers are to be found in the liver and should encourage further study of hepatic immune cells as important predictive factors.
|
|
| |
| |
|
|
|
