Journal of Viral Hepatitis, 2014, 21, 944–949
doi:10.1111/jvh.12281
Interferon–ribavirin therapy induces serum antibodies determining ‘rods and rings’ pattern in hepatitis C patients Cristina Novembrino,1 Alessio Aghemo,2 Chiara Ferraris Fusarini,1 Rita Maiavacca,1 Caterina Matinato,1 Giovanna Lunghi,1 Erminio Torresani,1 Mariangela Ronchi,1 Maria Cristina Garlaschi,1 Miriam Ramondetta3 and Massimo Colombo2 1Laboratory of Clinical Chemistry and Microbiology, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, Italy; 2First Division of Gastroenterology, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, Italy; and 3Transplantation Immunology, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, Italy Received February 2014; accepted for publication May 2014
SUMMARY. A cytoplasmic antigen associated to inosine-
5’-monophosphatedehydrogenase 2 eliciting specific antibodies (antirods and rings, RR) has been identified in patients with chronic hepatitis C who were exposed to pegylated interferon (PI) and ribavirin (RBV). The significance of anti-RR in these patients merits to be investigated. Sera from 88 chronic hepatitis C virus (HCV)-infected patients undergoing PI-RBV therapy were analysed for the presence of RR pattern by indirect immunofluorescence on HEp-2 substrate (Inova Diagnostics, San Diego, CA, USA). Anti-RR antibodies developed de novo in 32 patients independently of any demographic and virological feature, but with a significant association with cumulative exposure to
INTRODUCTION Hepatitis C virus (HCV) is a major cause of liver-related morbidity and mortality worldwide [1–3], yet the virus may also cause extra-hepatic manifestations related to autoimmune phenomena that are thought to result from a ‘molecular mimicry’ mechanism [4–7]. The standard of care treatment of HCV infection is the combination of PEG-interferon-a (PI) and ribavirin (RBV) which will ultimately activate a cascade of cellular signals leading to blocking viral protein synthesis and indirectly stimulating both innate- and virus-specific immune responses [8–11]. While HCV infection on itself is associated with an increased risk of autoimmune reactions that can be boosted by the immunomodulatory properties of interferon Abbreviations: ANA, antinuclear autoantibodies; HCV, hepatitis C virus; IMPDH2, inosine monophosphate dehydrogenase 2; IQR, interquartile range; NR, nonresponder; PI, pegylated interferon; RBV, ribavirin; REL, relapsers; RR, rods and rings. Correspondence: Cristina Novembrino, via F. Sforza, 35. 20122 Milan, Italy. E-mail: [email protected]
PI-RBV (P = 0.0089; chi-square test). RR pattern was significantly more frequent in relapsers than in patients achieving sustained virological response (56% vs 30%; P = 0.0282, chi-square test). Anti-RR titre ranged from 1:80 to 1:1280, but significantly declined following treatment cessation. Anti-RR develop de novo in a substantial proportion of patients exposed to PI-RBV in relation to the duration of treatment exposure. Further investigations are necessary to unravel the mechanisms leading to the formation of these autoantibodies. Keywords: autoantibodies, hepatitis C virus, inosine-5’monophosphatedehydrogenase 2, interferon, ribavirin.
[8], recently ribavirin-induced serum autoantibodies have been described in HCV patients undergoing PEG-interferona therapy resulting a typical cytoplasmic reactivity, named rods and ring (RR) at indirect immunofluorescence on HEp-2 cells [12–15]. While these novel cytoplasmic antigens were named after the structures resembling rods (3–10 lm in length) and rings (2–5 lm in diameter) [12–14], co-staining studies led to the identification of inosine monophosphate dehydrogenase 2 (IMPDH2) enzyme as a potential target antigen eliciting this novel class of serum autoantibodies [12]. In view of the potential interaction between anti-RR and the liver cell metabolism of ribavirin which is a decisive partner of PEG-interferon-a in the successful treatment of HCV, we aimed to evaluate the prevalence and the timing of serum anti-RR developing in HCV patients exposed to PI-RBV, trying to understand whether anti-RR seroconversion was associated with any clinical feature of hepatitis C or pattern of response to antiviral therapy.
MATERIALS AND METHODS The study population included 88 (47M/41F; aged 22–75 years) naive to treatment patients with chronic © 2014 John Wiley & Sons Ltd
Interferon–ribavirin therapy hepatitis C who, between 2008 and 2011, underwent PI-RBV therapy at the Liver Unit of the First Division of Gastroenterology, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico in Milan, Italy. Treatment consisted of once weekly subcutaneous injections of PEG-interferon-2a (Pegasys, Roche, Basel, Switzerland) 180 lg associated to oral ribavirin (Copegus, Roche, Basel, Switzerland) at a dose ranging from 800 to 1200 mg per day: 800– 1000 mg for patients weighting 75 Kg or less and 1200 mg for those with a body weight higher than 75 Kg. Duration of drug administration was according to HCV genotype: 24 weeks in case of viral genotypes other than 1 or 4, or 48 weeks for genotype 1 or 4. All the patients gave their written informed consent to receive therapy and agreement to scientific use of clinical data. The analytical evaluations of this retrospective longitudinal study were performed on frozen sera collected for HCV RNA quantification by real-time PCR (Abbott RealTime, Abbott Diagnostic; IL, USA) at baseline (T0), 12 weeks (T1), 24 weeks (T2) and 48 weeks (T3) after starting therapy. Patients were classified as sustained virological responders (SVR) if HCV RNA was undetectable 24 weeks after the end of the therapy, as nonresponder (NR) if HCV RNA was quantifiable during all the therapy duration and still after the end, as relapsers (REL) if HCV RNA was detectable after the end of therapy in patients with a previous virological response. Medical records were reviewed, and all patients were characterized at baseline for HVC genotype, viral load, modality of infection, hepatitis B virus co-infection, histological findings [16], biochemical data and presence of autoantibodies. Serum samples were analysed for anti-RR autoantibodies by indirect immunofluorescence on HEp-2 cells (INOVA Diagnostics, Inc.; San Diego, CA, USA) at a screening dilution of 1:80 with phosphate-buffered saline, according to standard indirect immunofluorescence protocol for antinuclear autoantibodies (ANA) testing; samples positive for anti-RR were serially diluted to determine the titre. All the indirect immunofluorescence analyses (screening and antiRR titre determinations) were performed using the same lot of slides and reagents. The slides were analysed by two different operators under 2009 and 4009 magnification on Nikon ECLIPSE 50i microscope coupled with Nikon Ds fi-1 camera for capturing images.
Statistical analysis Continuous variables were reported as median [interquartile range (IQR)]; categorical variables were described with number and proportions. Comparisons between groups were performed by Mann–Whitney test or Student’s t-test © 2014 John Wiley & Sons Ltd
945
(depending on data distribution) for continuous variables and by chi-square test or Fisher’s exact tests for categorical variables. Multivariate logistic regression analysis was used to evaluate predictors of anti-RR positivity and predictors of therapy response. All analyses were performed using software MedCalc [17]; P values < 0.05 were considered significant.
RESULTS Pretreatment demographic, clinical, biochemical and virological characteristics of the study population are reported in Table 1. The majority of patients had high pretreatment viral load (>8 UI/mL 9 105), mostly were infected by genotype 1b of HCV, whereas 17 patients (19%) were co-infected
Table 1 Demographic, clinical and biochemical features of the 88 hepatitis C virus-infected patients at baseline. Data are expressed as number (percentage) or as median [interquartile range] for continuous variables Parameters Demographics Males Age; years BMI; Kg/m2 HCV RNA Viral load 8 UI/mL 9 105 Viral genotype 1 and 4 Other Mode of HCV infection Sporadic Blood transfusion Injection–drug abuse HBV co–infection Biochemical data (Reference values) ALT (5–31 U/L) AST (5–32 U/L) GGT (5–36 U/L) AP (35–104 U/L) PLT (130–400 103/mmc) Histological characteristics Cirrhosis Autoantibodies ANA positive SMA and LKM
47 (53%) 54 (44.5–63.5) 25.4 (22.5–27.2) 22 (25%) 16 (18%) 50 (57%) 51 (58%) 37 (42%) 61 16 11 17 46 50 41 70 195
(69%) (18%) (13%) (19%) [5.2–116] [33.2–81] [26–80.5] [8.2–91.7] [45–235.5]
6 (7%) 21 (24%) 6 (7%)
ALT, alanine aminotransferase; AST, aspartate aminotransferase; GGT, gamma glutamyl transpeptidase; AP, alkaline phosphatase; PLT, platelet.
946
C. Novembrino et al.
with hepatitis B virus as determined by the co-presence of serum viral antigen. The grade of necro-inflammation was light or moderate in a majority of patients (95%) and severe in four patients only. Histologically, six patients (7%) had clear cut cirrhosis. Presence of serum autoantibodies, especially ANA, was evidenced in about a quarter of patients. Thyroid dysfunction was present in two patients. At baseline, none of the patients showed RR autoantibodies, which could be detected as cytoplasmic reactivity in serum samples collected from 32 patients (36%) following exposure to antiviral therapy (Fig. 1). No significant differences were found between anti-RRpositive and anti-RR-negative patients in terms of sex, age, biochemical data, genotype and load of HCV, rates of hepatitis B virus co-infection, liver disease staging and presence of nonorgan-specific autoantibodies. Although ribavirin dosing was higher in RR-positive than in RR-negative patients (1000 vs 800 mg; P = 0.0499; Mann–Whitney test), the potential predictor of drug dosing in eliciting anti-RR was ruled out by a multivariate logistic regression analysis controlling for other variable like sex, age, load and genotype of HCV. The frequency of anti-RR increased in parallel with therapy duration (P = 0.0089; chi-square test), with rates of 9%, 38% and 53% at week 12, 24 and 48, respectively (Fig. 2). Moreover, anti-RR were more frequent (thought not significantly) in patients infected with genotype 1 or 4 than in those infected with other genotypes (45% vs 27%; P = 0.13; chi-squared test). The same was true for anti-RR titre which ranged from 1:80 to 1:1280: in patients infected with HCV genotype 1 or 4, high titres (≥ 1:320) were more frequent than low titres (P = 0.039; chi-squared test) and, as shown in
Fig. 1 Indirect immunofluorescence on HEp-2 of antinuclear autoantibodies-negative serum from hepatitis C virus patient showing rods and rings pattern.
Fig. 2 Histogram showing the number of new cases of anti-rods and rings reactivity at 3 months (T1), 6 months (T2) and 12 months (T3) after the beginning of antiviral treatment.
Fig. 3 Box and dot plot graph comparing anti-rods and rings titre in patients divided on the basis of hepatitis C virus genotype. Fig. 3, the median titre was significantly higher in these patients than in those infected with genotype other than 1 or 4 (P = 0.03; Mann–Whitney test). Classifying the study population on the basis of virological response to anti-HCV therapy, 53 patients were SVR (60%), 27 REL (31%) and eight NR (9%). While the rates of NR were related to viral genotype (all the NR patients were infected with HCV with genotype 1 or 4, and in SVR and REL, viral genotypes were equally distributed), anti-RR pattern was significantly more frequently detected in REL than in SVR or NR (P = 0.0282; chi-square test). As shown in Fig. 4, in REL group, the prevalence of RR autoantibodies was 56%, while in SVR and NR was 30% and 12%, respectively. However, multivariate logistic regression analysis, including sex, age, viral load and genotype, interferon lambda polymorphism and © 2014 John Wiley & Sons Ltd
Interferon–ribavirin therapy
Fig. 4 Distribution of patients with rods and ring (RR)positive (grey) and RR-negative (black) in sustained virological responders (SVR), relapsers (REL) and nonresponders (NR).
RR reactivity did not show a predictor significantly associated with virological response. No correlation between anti-RR titre and ANA, therapy response and viral load was highlighted. In 23 anti-RR-positive patients, antibody was tested 1–12 months post-treatment cessation: all but one patient remained seroposive showing a decline of serum titre in 80% of the cases.
DISCUSSION Despite a strict association has been repeatedly demonstrated between chronic infection with HCV and risk of autoimmune manifestations [5,18,19], a few studies only are available which investigate anti-RR reactivity in hepatitis C patients who have been exposed to PI-RBV. In this retrospective study, one-third of 88 patients with chronic hepatitis C treated with PI-RBV were found to develop anti-RR as consequence of antiviral therapy, at a frequency that was higher than previously reported in comparable populations [13,20]. Interestingly, a previous study in patients exposed to PI or RBV alone failed to show the onset of anti-RR [20]. The fact that in the present study, none of the pretreatment serum samples exhibited cytoplasmic reactivity for RR and that serum concentration of anti-RR in reactive patients tended to increase with therapy progression was taken as additional evidence that combined PI-RBV therapy did trigger patients’ reactivity. While our findings confirmed previous observations in similar patients [14,15,20], the present study based on sequential sampling of patients exposed to PI-RBV allowed us to evaluate the kinetics of anti-RR seroconversion with respect to treatment timing. © 2014 John Wiley & Sons Ltd
947
The study highlighted the highest rates (53%) of seroconversion occurring in patients exposed to PI-RBV for 48 weeks compared to patients receiving shorter treatment with PI-RBV depending on virological response and virus genotype. This is in agreement with findings by Keppeke and colleagues reporting a serum anti-RR peaking rate of 42% following 36 weeks of therapy [20]. Interestingly, we and others could not find any correlation between rates and kinetics of anti-RR in patients with chronic hepatitis C undergoing PI-RBV therapy and the relevant demographic features of our patient including co-infection with other viruses [13–15,20]. We acknowledge, however, that due the limited sample size of the study and the retrospective selection of patients, our assumptions should be cautiously taken, as other cofactors known to influence patients’ susceptibility to autoimmune reactions could not be measured. Further, the clear cut relationship between treatment duration, HCV genotype and virological response (most nonresponders have to interrupt treatment at week 12) did not allow to rule out spurious associations between antiRR, virus genotypes and virological response. Along this line, it is interesting to note that a correlation between serum anti-RR and interferon lambda polymorphism, which is known to be involved in response to drug [21], was not observed. An additional reason why our study was not qualified to investigate any association between anti-RR and a response to therapy was that NR patients were largely underrepresented (9%) and all infected by the more difficult to cure genotype 1 or 4 HCV. An intriguing finding was the higher rates of anti-RR among response categories like REL vs SVR patients, serum anti-RR antibodies being more frequently detected in the former (56%) rather than in the latter patients (30%). Interestingly, this finding confirmed previous observations in selected groups of HCV patients stratified by response to PI-RBV [14,15], reporting the lowest anti-RR frequency in patients achieving sustained virological response. In contrast, Keppeke and colleagues did not find any correlation between RR pattern and response to therapy [20]. In analogy with a previous report [20], serum anti-RR were still detectable weeks after the cessation of antiviral treatment, although at declining titre. While this observation lends further support to the immune stimulating effect of interferon in the promotion of serum anti-RR, we acknowledge that a post-treatment follow-up spanning from 4 to 48 weeks only was a too short period of time for anti-RR clearance to take place in all patients. One important question raised by our and previous study is whether the onset of serum anti-RR has any pathogenetic implication in the PI-RBV algorithm or merely reflects an inoffensive host reaction to the immunostimulatory activity of interferon.
948
C. Novembrino et al.
Co-staining and radioimmunoprecipitations studies identified IMPDH2 as the target antigen of anti-RR autoantibodies [12]. Indeed, this enzyme which is involved in the de novo cellular synthesis of guanosine is selectively inhibited by RVR which after cell internalization and conversion induces structural modifications of IMPDH2 causing normally masked epitopes that might participate to RR formation, to become accessible to the immune system. This hypothesis fits with the observation that RR reactivity develops in patients under combined therapy with PI-RBV only and not in those untreated or exposed to either drug [13–15,20]. In a recent paper, Carcamo and colleagues reported that the use of other IMPDH2 inhibitors could induce expression of RR in cultured cells, with different induction efficiency [12]. In particular, the mycophenolic acid, an immunosuppressing drug used to prevent rejection in organ transplanted patients, is a potent inhibitor of IMPDH2. In a previous study evaluating the frequency of RR pattern in consecutive samples tested for ANA screening, we reported the presence of anti-RR in liver-transplanted patients treated with mycophenolic acid [22], 80% of whom with previous HCV infection treated with standard therapy. On the contrary, evaluating a population of 53 renal-transplanted patients (7% infected with HCV) on the same immunosuppressant regimen, we observed RR pattern in none of the samples (unpublished data). These observations need to be externally validated, but they lend
support to hypothesis that the onset of anti-RR autoantibodies is surely a consequence of therapy, but it is a response to disturbances exclusively in hepatocyte synthetic pathways. The fact that anti-RR could be detected in a minority of patients exposed to PI-RBV raises the question of predisposing factors causing anti-RR production. None of the demographic, clinical and biochemical parameters evaluated in our study seemed to be actively involved. Indeed, anti-RR formation could be influenced by cellular IMPDH2/inhibitor ratio, with an over expression of the protein being able to contrast RR induction [12]. Studies evaluating possible genetic polymorphisms influencing enzymatic expression and activity might contribute to clarify this mechanism. In conclusion, despite limitations of this study, we provide conclusive evidence of the relationship which exists between anti-RR and HCV patients exposure to PI-RBV, whereas we tentatively offer a preliminary explanation for these novel cytoplasmic autoantibodies to reflect a drugrelated injury of the HCV-infected hepatocytes.
ACKNOWLEDGEMENTS The authors acknowledge the technical support of Agnese Follesa, Cinzia Cortellaro, Maria Monasteri. No external funding was received for the present study.
REFERENCES 1 Antonelli A, Ferri C, Galeazzi M et al. HCV infection: pathogenesis, clinical manifestations and therapy. Clin Exp Rheumatol 2008; 26: S39–S47. 2 Rosen HR. Chronic hepatitis C infection. N Engl J Med 2011; 364: 2429–2438. 3 Ikeda K, Saitoh S, Koida I et al. A multivariate analysis of risk factors for hepatocellular carcinogenesis: a prospective observation of 795 patients with viral and alcoholic cirrhosis. Hepatology 1993; 18: 47–53. 4 Bogdanos DP, Choudhuri K, Vergani D. Molecular mimicry and autoimmune liver disease: virtuous intentions, malign consequences. Liver 2001; 21: 225–232. 5 Himoto T, Masaki T. Extrahepatic manifestations and autoantibodies in patients with hepatitis C virus infection. Clin Dev Immunol 2012; 2012: 871401. doi:10.1155/2012/ 871401. 6 Ramos-Casals M, Muñ oz S, Medina F et al. Systemic autoimmune diseases
7
8
9
10
in patients with hepatitis C virus infection: characterization of 1020 cases (the HISPAMEC registry). J Rheumatol 2009; 36: 1442–1448. Ramos-Casals M, Garcıa-Carrasco M, Cervera R et al. Th1/Th2 cytokine imbalance in patients with Sj€ ogren syndrome secondary to hepatitis C virus infection. Semin Arthritis Rheum 2002; 32: 56–63. Horner SM, Gale M. Intracellular innate immune cascades and interferon defences that control hepatitis C virus. J Interferon Cytokine Res 2009; 29: 489–498. Zhou S, Liu R, Baroudy BM, Malcom BA, Reyes GR. The effect of ribavirin and IMPDH inhibitors on hepatitis C virus subgenomic replicon RNA. Virology 2003; 310: 333–342. Khakoo S, Glue P, Grellier L et al. Ribavirin and interferon alfa-2b in chronic hepatitis C: assessment of possible pharmacokinetic and pharmacodynamic interactions. Br J Clin Pharmacol 1998; 46: 563–570.
11 Feld JJ, Nanda S, Huang Y et al. Hepatic gene expression during treatment with peginterferon and ribavirin: identifying molecular pathways for treatment response. Hepatology 2007; 46: 1548–1563. 12 Carcamo WC, Satoh M, Kasahara H et al. Induction of cytoplasmic rods and rings structures by inhibition of the CTP and GTP synthetis pathway in mammalian cells. PLoS ONE 2011; 6: 1–12. 13 Seelig HP, Appelhans H, Bauer O et al. Autoantibodies against inosine5’-monophosphate dehydrogenase 2 – Characteristics and prevalence in patients with HCV-infection. Clin Lab 2011; 57: 753–765. 14 Covini G, Carcamo WC, Bredi E et al. Cytoplasmic rods and rings autoantibodies developed during pegylated interferon and ribavirin therapy in patients with chronic hepatitis C. Anti Ther 2012; 17: 805–811. 15 Carcamo WC, Ceribelli A, Calise SJ et al. Differential reactivity to
© 2014 John Wiley & Sons Ltd
Interferon–ribavirin therapy IMPDH2 by anti-rods/rings autoantibodies and unresponsiveness to pegylated interferon-alpha/ribavirin therapy in US and Italian HCV patients. J Clin Immunol 2013; 33: 420–426. 16 Ishak K, Baptista A, Bianchi L et al. Histological grading and staging of chronic hepatitis. J Hepatol 1995; 22: 696–699. 17 Stephan C, Wesseling S, Schink T, Jung K. Comparison of eight computer programs for receiver-operatic characteristic analysis. Clin Chem 2003; 49: 433–439.
© 2014 John Wiley & Sons Ltd
18 Vergani D, Mieli-Vergani G. Autoimmune manifestations in viral hepatitis. Semin Immunopathol 2013; 35: 73–85. 19 Lawson E, Walker-Bone K. The changing spectrum of rheumatic disease in HIV infection. Br Med Bull 2012; 103: 203–221. 20 Keppeke GD, Nunes E, Gomes Ferraz ML et al. Longitudinal study of a human drug-induced model of autoantibody to cytoplasmic rods/rings following HCV therapy with ribavirin and interferon-a. PLoS ONE 2012; 7: e45392. doi:10.
949
1371/journal.pone.0045392. Epub 2012 Sep 24 21 Bellanti F, Vendemiale G, Altomare E, Serviddio G. The impact of interferon lambda 3 gene polymorphism on natural course and treatment of hepatitis C. Clin Dev Immunol 2012; 2012: 849373. doi: 10.1155/2012/ 849373. Epub 2012 Aug 27 22 Novembrino C, Maiavacca R, Matinato C et al. Prevalence of ‘Rings Rods’ autoantibodies on HEp-2 preparations. Biochim Clin 2013; 37: M035.
Copyright of Journal of Viral Hepatitis is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
doi:10.1111/jvh.12281
Interferon–ribavirin therapy induces serum antibodies determining ‘rods and rings’ pattern in hepatitis C patients Cristina Novembrino,1 Alessio Aghemo,2 Chiara Ferraris Fusarini,1 Rita Maiavacca,1 Caterina Matinato,1 Giovanna Lunghi,1 Erminio Torresani,1 Mariangela Ronchi,1 Maria Cristina Garlaschi,1 Miriam Ramondetta3 and Massimo Colombo2 1Laboratory of Clinical Chemistry and Microbiology, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, Italy; 2First Division of Gastroenterology, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, Italy; and 3Transplantation Immunology, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, Italy Received February 2014; accepted for publication May 2014
SUMMARY. A cytoplasmic antigen associated to inosine-
5’-monophosphatedehydrogenase 2 eliciting specific antibodies (antirods and rings, RR) has been identified in patients with chronic hepatitis C who were exposed to pegylated interferon (PI) and ribavirin (RBV). The significance of anti-RR in these patients merits to be investigated. Sera from 88 chronic hepatitis C virus (HCV)-infected patients undergoing PI-RBV therapy were analysed for the presence of RR pattern by indirect immunofluorescence on HEp-2 substrate (Inova Diagnostics, San Diego, CA, USA). Anti-RR antibodies developed de novo in 32 patients independently of any demographic and virological feature, but with a significant association with cumulative exposure to
INTRODUCTION Hepatitis C virus (HCV) is a major cause of liver-related morbidity and mortality worldwide [1–3], yet the virus may also cause extra-hepatic manifestations related to autoimmune phenomena that are thought to result from a ‘molecular mimicry’ mechanism [4–7]. The standard of care treatment of HCV infection is the combination of PEG-interferon-a (PI) and ribavirin (RBV) which will ultimately activate a cascade of cellular signals leading to blocking viral protein synthesis and indirectly stimulating both innate- and virus-specific immune responses [8–11]. While HCV infection on itself is associated with an increased risk of autoimmune reactions that can be boosted by the immunomodulatory properties of interferon Abbreviations: ANA, antinuclear autoantibodies; HCV, hepatitis C virus; IMPDH2, inosine monophosphate dehydrogenase 2; IQR, interquartile range; NR, nonresponder; PI, pegylated interferon; RBV, ribavirin; REL, relapsers; RR, rods and rings. Correspondence: Cristina Novembrino, via F. Sforza, 35. 20122 Milan, Italy. E-mail: [email protected]
PI-RBV (P = 0.0089; chi-square test). RR pattern was significantly more frequent in relapsers than in patients achieving sustained virological response (56% vs 30%; P = 0.0282, chi-square test). Anti-RR titre ranged from 1:80 to 1:1280, but significantly declined following treatment cessation. Anti-RR develop de novo in a substantial proportion of patients exposed to PI-RBV in relation to the duration of treatment exposure. Further investigations are necessary to unravel the mechanisms leading to the formation of these autoantibodies. Keywords: autoantibodies, hepatitis C virus, inosine-5’monophosphatedehydrogenase 2, interferon, ribavirin.
[8], recently ribavirin-induced serum autoantibodies have been described in HCV patients undergoing PEG-interferona therapy resulting a typical cytoplasmic reactivity, named rods and ring (RR) at indirect immunofluorescence on HEp-2 cells [12–15]. While these novel cytoplasmic antigens were named after the structures resembling rods (3–10 lm in length) and rings (2–5 lm in diameter) [12–14], co-staining studies led to the identification of inosine monophosphate dehydrogenase 2 (IMPDH2) enzyme as a potential target antigen eliciting this novel class of serum autoantibodies [12]. In view of the potential interaction between anti-RR and the liver cell metabolism of ribavirin which is a decisive partner of PEG-interferon-a in the successful treatment of HCV, we aimed to evaluate the prevalence and the timing of serum anti-RR developing in HCV patients exposed to PI-RBV, trying to understand whether anti-RR seroconversion was associated with any clinical feature of hepatitis C or pattern of response to antiviral therapy.
MATERIALS AND METHODS The study population included 88 (47M/41F; aged 22–75 years) naive to treatment patients with chronic © 2014 John Wiley & Sons Ltd
Interferon–ribavirin therapy hepatitis C who, between 2008 and 2011, underwent PI-RBV therapy at the Liver Unit of the First Division of Gastroenterology, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico in Milan, Italy. Treatment consisted of once weekly subcutaneous injections of PEG-interferon-2a (Pegasys, Roche, Basel, Switzerland) 180 lg associated to oral ribavirin (Copegus, Roche, Basel, Switzerland) at a dose ranging from 800 to 1200 mg per day: 800– 1000 mg for patients weighting 75 Kg or less and 1200 mg for those with a body weight higher than 75 Kg. Duration of drug administration was according to HCV genotype: 24 weeks in case of viral genotypes other than 1 or 4, or 48 weeks for genotype 1 or 4. All the patients gave their written informed consent to receive therapy and agreement to scientific use of clinical data. The analytical evaluations of this retrospective longitudinal study were performed on frozen sera collected for HCV RNA quantification by real-time PCR (Abbott RealTime, Abbott Diagnostic; IL, USA) at baseline (T0), 12 weeks (T1), 24 weeks (T2) and 48 weeks (T3) after starting therapy. Patients were classified as sustained virological responders (SVR) if HCV RNA was undetectable 24 weeks after the end of the therapy, as nonresponder (NR) if HCV RNA was quantifiable during all the therapy duration and still after the end, as relapsers (REL) if HCV RNA was detectable after the end of therapy in patients with a previous virological response. Medical records were reviewed, and all patients were characterized at baseline for HVC genotype, viral load, modality of infection, hepatitis B virus co-infection, histological findings [16], biochemical data and presence of autoantibodies. Serum samples were analysed for anti-RR autoantibodies by indirect immunofluorescence on HEp-2 cells (INOVA Diagnostics, Inc.; San Diego, CA, USA) at a screening dilution of 1:80 with phosphate-buffered saline, according to standard indirect immunofluorescence protocol for antinuclear autoantibodies (ANA) testing; samples positive for anti-RR were serially diluted to determine the titre. All the indirect immunofluorescence analyses (screening and antiRR titre determinations) were performed using the same lot of slides and reagents. The slides were analysed by two different operators under 2009 and 4009 magnification on Nikon ECLIPSE 50i microscope coupled with Nikon Ds fi-1 camera for capturing images.
Statistical analysis Continuous variables were reported as median [interquartile range (IQR)]; categorical variables were described with number and proportions. Comparisons between groups were performed by Mann–Whitney test or Student’s t-test © 2014 John Wiley & Sons Ltd
945
(depending on data distribution) for continuous variables and by chi-square test or Fisher’s exact tests for categorical variables. Multivariate logistic regression analysis was used to evaluate predictors of anti-RR positivity and predictors of therapy response. All analyses were performed using software MedCalc [17]; P values < 0.05 were considered significant.
RESULTS Pretreatment demographic, clinical, biochemical and virological characteristics of the study population are reported in Table 1. The majority of patients had high pretreatment viral load (>8 UI/mL 9 105), mostly were infected by genotype 1b of HCV, whereas 17 patients (19%) were co-infected
Table 1 Demographic, clinical and biochemical features of the 88 hepatitis C virus-infected patients at baseline. Data are expressed as number (percentage) or as median [interquartile range] for continuous variables Parameters Demographics Males Age; years BMI; Kg/m2 HCV RNA Viral load 8 UI/mL 9 105 Viral genotype 1 and 4 Other Mode of HCV infection Sporadic Blood transfusion Injection–drug abuse HBV co–infection Biochemical data (Reference values) ALT (5–31 U/L) AST (5–32 U/L) GGT (5–36 U/L) AP (35–104 U/L) PLT (130–400 103/mmc) Histological characteristics Cirrhosis Autoantibodies ANA positive SMA and LKM
47 (53%) 54 (44.5–63.5) 25.4 (22.5–27.2) 22 (25%) 16 (18%) 50 (57%) 51 (58%) 37 (42%) 61 16 11 17 46 50 41 70 195
(69%) (18%) (13%) (19%) [5.2–116] [33.2–81] [26–80.5] [8.2–91.7] [45–235.5]
6 (7%) 21 (24%) 6 (7%)
ALT, alanine aminotransferase; AST, aspartate aminotransferase; GGT, gamma glutamyl transpeptidase; AP, alkaline phosphatase; PLT, platelet.
946
C. Novembrino et al.
with hepatitis B virus as determined by the co-presence of serum viral antigen. The grade of necro-inflammation was light or moderate in a majority of patients (95%) and severe in four patients only. Histologically, six patients (7%) had clear cut cirrhosis. Presence of serum autoantibodies, especially ANA, was evidenced in about a quarter of patients. Thyroid dysfunction was present in two patients. At baseline, none of the patients showed RR autoantibodies, which could be detected as cytoplasmic reactivity in serum samples collected from 32 patients (36%) following exposure to antiviral therapy (Fig. 1). No significant differences were found between anti-RRpositive and anti-RR-negative patients in terms of sex, age, biochemical data, genotype and load of HCV, rates of hepatitis B virus co-infection, liver disease staging and presence of nonorgan-specific autoantibodies. Although ribavirin dosing was higher in RR-positive than in RR-negative patients (1000 vs 800 mg; P = 0.0499; Mann–Whitney test), the potential predictor of drug dosing in eliciting anti-RR was ruled out by a multivariate logistic regression analysis controlling for other variable like sex, age, load and genotype of HCV. The frequency of anti-RR increased in parallel with therapy duration (P = 0.0089; chi-square test), with rates of 9%, 38% and 53% at week 12, 24 and 48, respectively (Fig. 2). Moreover, anti-RR were more frequent (thought not significantly) in patients infected with genotype 1 or 4 than in those infected with other genotypes (45% vs 27%; P = 0.13; chi-squared test). The same was true for anti-RR titre which ranged from 1:80 to 1:1280: in patients infected with HCV genotype 1 or 4, high titres (≥ 1:320) were more frequent than low titres (P = 0.039; chi-squared test) and, as shown in
Fig. 1 Indirect immunofluorescence on HEp-2 of antinuclear autoantibodies-negative serum from hepatitis C virus patient showing rods and rings pattern.
Fig. 2 Histogram showing the number of new cases of anti-rods and rings reactivity at 3 months (T1), 6 months (T2) and 12 months (T3) after the beginning of antiviral treatment.
Fig. 3 Box and dot plot graph comparing anti-rods and rings titre in patients divided on the basis of hepatitis C virus genotype. Fig. 3, the median titre was significantly higher in these patients than in those infected with genotype other than 1 or 4 (P = 0.03; Mann–Whitney test). Classifying the study population on the basis of virological response to anti-HCV therapy, 53 patients were SVR (60%), 27 REL (31%) and eight NR (9%). While the rates of NR were related to viral genotype (all the NR patients were infected with HCV with genotype 1 or 4, and in SVR and REL, viral genotypes were equally distributed), anti-RR pattern was significantly more frequently detected in REL than in SVR or NR (P = 0.0282; chi-square test). As shown in Fig. 4, in REL group, the prevalence of RR autoantibodies was 56%, while in SVR and NR was 30% and 12%, respectively. However, multivariate logistic regression analysis, including sex, age, viral load and genotype, interferon lambda polymorphism and © 2014 John Wiley & Sons Ltd
Interferon–ribavirin therapy
Fig. 4 Distribution of patients with rods and ring (RR)positive (grey) and RR-negative (black) in sustained virological responders (SVR), relapsers (REL) and nonresponders (NR).
RR reactivity did not show a predictor significantly associated with virological response. No correlation between anti-RR titre and ANA, therapy response and viral load was highlighted. In 23 anti-RR-positive patients, antibody was tested 1–12 months post-treatment cessation: all but one patient remained seroposive showing a decline of serum titre in 80% of the cases.
DISCUSSION Despite a strict association has been repeatedly demonstrated between chronic infection with HCV and risk of autoimmune manifestations [5,18,19], a few studies only are available which investigate anti-RR reactivity in hepatitis C patients who have been exposed to PI-RBV. In this retrospective study, one-third of 88 patients with chronic hepatitis C treated with PI-RBV were found to develop anti-RR as consequence of antiviral therapy, at a frequency that was higher than previously reported in comparable populations [13,20]. Interestingly, a previous study in patients exposed to PI or RBV alone failed to show the onset of anti-RR [20]. The fact that in the present study, none of the pretreatment serum samples exhibited cytoplasmic reactivity for RR and that serum concentration of anti-RR in reactive patients tended to increase with therapy progression was taken as additional evidence that combined PI-RBV therapy did trigger patients’ reactivity. While our findings confirmed previous observations in similar patients [14,15,20], the present study based on sequential sampling of patients exposed to PI-RBV allowed us to evaluate the kinetics of anti-RR seroconversion with respect to treatment timing. © 2014 John Wiley & Sons Ltd
947
The study highlighted the highest rates (53%) of seroconversion occurring in patients exposed to PI-RBV for 48 weeks compared to patients receiving shorter treatment with PI-RBV depending on virological response and virus genotype. This is in agreement with findings by Keppeke and colleagues reporting a serum anti-RR peaking rate of 42% following 36 weeks of therapy [20]. Interestingly, we and others could not find any correlation between rates and kinetics of anti-RR in patients with chronic hepatitis C undergoing PI-RBV therapy and the relevant demographic features of our patient including co-infection with other viruses [13–15,20]. We acknowledge, however, that due the limited sample size of the study and the retrospective selection of patients, our assumptions should be cautiously taken, as other cofactors known to influence patients’ susceptibility to autoimmune reactions could not be measured. Further, the clear cut relationship between treatment duration, HCV genotype and virological response (most nonresponders have to interrupt treatment at week 12) did not allow to rule out spurious associations between antiRR, virus genotypes and virological response. Along this line, it is interesting to note that a correlation between serum anti-RR and interferon lambda polymorphism, which is known to be involved in response to drug [21], was not observed. An additional reason why our study was not qualified to investigate any association between anti-RR and a response to therapy was that NR patients were largely underrepresented (9%) and all infected by the more difficult to cure genotype 1 or 4 HCV. An intriguing finding was the higher rates of anti-RR among response categories like REL vs SVR patients, serum anti-RR antibodies being more frequently detected in the former (56%) rather than in the latter patients (30%). Interestingly, this finding confirmed previous observations in selected groups of HCV patients stratified by response to PI-RBV [14,15], reporting the lowest anti-RR frequency in patients achieving sustained virological response. In contrast, Keppeke and colleagues did not find any correlation between RR pattern and response to therapy [20]. In analogy with a previous report [20], serum anti-RR were still detectable weeks after the cessation of antiviral treatment, although at declining titre. While this observation lends further support to the immune stimulating effect of interferon in the promotion of serum anti-RR, we acknowledge that a post-treatment follow-up spanning from 4 to 48 weeks only was a too short period of time for anti-RR clearance to take place in all patients. One important question raised by our and previous study is whether the onset of serum anti-RR has any pathogenetic implication in the PI-RBV algorithm or merely reflects an inoffensive host reaction to the immunostimulatory activity of interferon.
948
C. Novembrino et al.
Co-staining and radioimmunoprecipitations studies identified IMPDH2 as the target antigen of anti-RR autoantibodies [12]. Indeed, this enzyme which is involved in the de novo cellular synthesis of guanosine is selectively inhibited by RVR which after cell internalization and conversion induces structural modifications of IMPDH2 causing normally masked epitopes that might participate to RR formation, to become accessible to the immune system. This hypothesis fits with the observation that RR reactivity develops in patients under combined therapy with PI-RBV only and not in those untreated or exposed to either drug [13–15,20]. In a recent paper, Carcamo and colleagues reported that the use of other IMPDH2 inhibitors could induce expression of RR in cultured cells, with different induction efficiency [12]. In particular, the mycophenolic acid, an immunosuppressing drug used to prevent rejection in organ transplanted patients, is a potent inhibitor of IMPDH2. In a previous study evaluating the frequency of RR pattern in consecutive samples tested for ANA screening, we reported the presence of anti-RR in liver-transplanted patients treated with mycophenolic acid [22], 80% of whom with previous HCV infection treated with standard therapy. On the contrary, evaluating a population of 53 renal-transplanted patients (7% infected with HCV) on the same immunosuppressant regimen, we observed RR pattern in none of the samples (unpublished data). These observations need to be externally validated, but they lend
support to hypothesis that the onset of anti-RR autoantibodies is surely a consequence of therapy, but it is a response to disturbances exclusively in hepatocyte synthetic pathways. The fact that anti-RR could be detected in a minority of patients exposed to PI-RBV raises the question of predisposing factors causing anti-RR production. None of the demographic, clinical and biochemical parameters evaluated in our study seemed to be actively involved. Indeed, anti-RR formation could be influenced by cellular IMPDH2/inhibitor ratio, with an over expression of the protein being able to contrast RR induction [12]. Studies evaluating possible genetic polymorphisms influencing enzymatic expression and activity might contribute to clarify this mechanism. In conclusion, despite limitations of this study, we provide conclusive evidence of the relationship which exists between anti-RR and HCV patients exposure to PI-RBV, whereas we tentatively offer a preliminary explanation for these novel cytoplasmic autoantibodies to reflect a drugrelated injury of the HCV-infected hepatocytes.
ACKNOWLEDGEMENTS The authors acknowledge the technical support of Agnese Follesa, Cinzia Cortellaro, Maria Monasteri. No external funding was received for the present study.
REFERENCES 1 Antonelli A, Ferri C, Galeazzi M et al. HCV infection: pathogenesis, clinical manifestations and therapy. Clin Exp Rheumatol 2008; 26: S39–S47. 2 Rosen HR. Chronic hepatitis C infection. N Engl J Med 2011; 364: 2429–2438. 3 Ikeda K, Saitoh S, Koida I et al. A multivariate analysis of risk factors for hepatocellular carcinogenesis: a prospective observation of 795 patients with viral and alcoholic cirrhosis. Hepatology 1993; 18: 47–53. 4 Bogdanos DP, Choudhuri K, Vergani D. Molecular mimicry and autoimmune liver disease: virtuous intentions, malign consequences. Liver 2001; 21: 225–232. 5 Himoto T, Masaki T. Extrahepatic manifestations and autoantibodies in patients with hepatitis C virus infection. Clin Dev Immunol 2012; 2012: 871401. doi:10.1155/2012/ 871401. 6 Ramos-Casals M, Muñ oz S, Medina F et al. Systemic autoimmune diseases
7
8
9
10
in patients with hepatitis C virus infection: characterization of 1020 cases (the HISPAMEC registry). J Rheumatol 2009; 36: 1442–1448. Ramos-Casals M, Garcıa-Carrasco M, Cervera R et al. Th1/Th2 cytokine imbalance in patients with Sj€ ogren syndrome secondary to hepatitis C virus infection. Semin Arthritis Rheum 2002; 32: 56–63. Horner SM, Gale M. Intracellular innate immune cascades and interferon defences that control hepatitis C virus. J Interferon Cytokine Res 2009; 29: 489–498. Zhou S, Liu R, Baroudy BM, Malcom BA, Reyes GR. The effect of ribavirin and IMPDH inhibitors on hepatitis C virus subgenomic replicon RNA. Virology 2003; 310: 333–342. Khakoo S, Glue P, Grellier L et al. Ribavirin and interferon alfa-2b in chronic hepatitis C: assessment of possible pharmacokinetic and pharmacodynamic interactions. Br J Clin Pharmacol 1998; 46: 563–570.
11 Feld JJ, Nanda S, Huang Y et al. Hepatic gene expression during treatment with peginterferon and ribavirin: identifying molecular pathways for treatment response. Hepatology 2007; 46: 1548–1563. 12 Carcamo WC, Satoh M, Kasahara H et al. Induction of cytoplasmic rods and rings structures by inhibition of the CTP and GTP synthetis pathway in mammalian cells. PLoS ONE 2011; 6: 1–12. 13 Seelig HP, Appelhans H, Bauer O et al. Autoantibodies against inosine5’-monophosphate dehydrogenase 2 – Characteristics and prevalence in patients with HCV-infection. Clin Lab 2011; 57: 753–765. 14 Covini G, Carcamo WC, Bredi E et al. Cytoplasmic rods and rings autoantibodies developed during pegylated interferon and ribavirin therapy in patients with chronic hepatitis C. Anti Ther 2012; 17: 805–811. 15 Carcamo WC, Ceribelli A, Calise SJ et al. Differential reactivity to
© 2014 John Wiley & Sons Ltd
Interferon–ribavirin therapy IMPDH2 by anti-rods/rings autoantibodies and unresponsiveness to pegylated interferon-alpha/ribavirin therapy in US and Italian HCV patients. J Clin Immunol 2013; 33: 420–426. 16 Ishak K, Baptista A, Bianchi L et al. Histological grading and staging of chronic hepatitis. J Hepatol 1995; 22: 696–699. 17 Stephan C, Wesseling S, Schink T, Jung K. Comparison of eight computer programs for receiver-operatic characteristic analysis. Clin Chem 2003; 49: 433–439.
© 2014 John Wiley & Sons Ltd
18 Vergani D, Mieli-Vergani G. Autoimmune manifestations in viral hepatitis. Semin Immunopathol 2013; 35: 73–85. 19 Lawson E, Walker-Bone K. The changing spectrum of rheumatic disease in HIV infection. Br Med Bull 2012; 103: 203–221. 20 Keppeke GD, Nunes E, Gomes Ferraz ML et al. Longitudinal study of a human drug-induced model of autoantibody to cytoplasmic rods/rings following HCV therapy with ribavirin and interferon-a. PLoS ONE 2012; 7: e45392. doi:10.
949
1371/journal.pone.0045392. Epub 2012 Sep 24 21 Bellanti F, Vendemiale G, Altomare E, Serviddio G. The impact of interferon lambda 3 gene polymorphism on natural course and treatment of hepatitis C. Clin Dev Immunol 2012; 2012: 849373. doi: 10.1155/2012/ 849373. Epub 2012 Aug 27 22 Novembrino C, Maiavacca R, Matinato C et al. Prevalence of ‘Rings Rods’ autoantibodies on HEp-2 preparations. Biochim Clin 2013; 37: M035.
Copyright of Journal of Viral Hepatitis is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
