IgM antibody response to the hepatitis C virus core protein in intravenous drug users

VIROLOGY

IgM Antibody Response to the Hepatitis C Virus Core Protein in Intravenous Drug Users Pierluigi Toniutto, Edmondo Falleti, Vinicio Gasparini, Carlo Fabris, Sergio Gabriel Tisminetzky, Tiziana Lombardelli, Paola Pacco, Andrea Satta, and Mario Pirisi

To verify whether a solid-phase enzyme immunoassay for serum IgM antibodies to the hepatitis C virus (HCV) core protein (IgM anti-HCVcore) might be proposed as a surrogate test for serum HCV RNA, we studied 86 anti-HCV antibodypositive intravenous drug users. Serum HCV RNA was demonstrated by RT-PCR with primers derived from the 59 noncoding and the core region. IgM anti-HCVcore antibodies were found in 62/86 (72%) subjects; circulating HCV RNA was detected by the 59 noncoding assay in 53/86 samples (62%) and by the core region assay in 35/86 samples (41%). IgM anti-HCVcore reactivity was associated with core HCV RNA

seropositivity (p ,0.05) but not with 59 noncoding HCV RNA seropositivity (p 5 NS). Patients infected by HCV type 1a were more-often positive for IgM anti-HCVcore (p ,0.05) and for core HCV RNA (p 5 0.005) than patients infected by other HCV genotypes. IgM anti-HCVcore reactivity was significantly more common in subjects positive for core HCV RNA (p ,0.005) and in subjects aged .30 years (p ,0.05). In conclusion, the IgM anti-HCVcore assay frequently tests positive in intravenous drug users, particularly when infected by HCV 1a, but is not a surrogate of testing for serum HCV RNA. © 1999 Elsevier Science Inc.

INTRODUCTION

diagnosis of hepatitis C virus (HCV) infection (De Medina and Schiff 1995). In the presence of chronic liver disease, detection of anti-HCV antibodies is associated with circulating viral RNA in .90% of cases (Marin et al. 1994). This does not necessarily hold true for subjects with normal liver function tests found to be antiHCV-positive, for example at screening for blood donation. In these cases, the correct interpretation of a positive serologic test may require additional testing for serum HCV RNA (Alberti et al. 1992). In chronic viral hepatitis B and D, the detection of IgM antibodies to the core and delta antigens is correlated with viral replication, disease activity, and response to antiviral agents (Govindarajan et al. 1989; Scully et al. 1990; Smith et al. 1992). A specific IgM antibody response also can be documented in

Serologic tests that detect the presence of antibodies directed against a broad spectrum of structural and nonstructural antigens represent the mainstay for the From the Department of Experimental and Clinical Pathology and Medicine (PT, EF, VG, CF, PP, MP), University of Udine; International Centre for Genetic Engineering and Biotechnology (SGT), Trieste; Servizio per le Tossicodipendenze, (TL) Udine; Istituto di Patologia Medica (AS), University of Sassari, Italy. Address reprint requests to Dr. Mario Pirisi, Cattedra di Medicina Interna, Universita` degli Studi, Piazzale Santa Maria della Misericordia 1, 33100 Udine, Italy. This paper has been presented in abstract form at the 4th International Congress on Clinical Chemistry and Laboratory Medicine, Bled, Slovenia, 22–24 May 1997. Received 5 February 1998; revised and accepted 24 September 1998.

DIAGN MICROBIOL INFECT DIS 1999;33:69 –73 © 1999 Elsevier Science Inc. All rights reserved. 655 Avenue of the Americas, New York, NY 10010

0732-8893/99/$–see front matter PII S0732-8893(98)00135-7

P. Toniutto et al.

70 many patients with chronic hepatitis C (Brillanti et al. 1992; Quiroga et al. 1992). Recently, the presence of IgM antibodies to HCV antigens has been correlated with viral replication, histological activity, response to interferon therapy, and liver disease outcome (Quiroga et al. 1995; Yuki et al. 1995). Intravenous drug users (IVDUs) are one of the categories of patients with higher rates of HCV seropositivity (Esteban et al. 1989; van der Hoek et al. 1990). The interpretation of a positive anti-HCV test in these patients is not always straightforward. On the one hand, concomitant infection by other viruses, ethanol consumption, and multidrug use may all contribute to liver damage; on the other, liver function tests may be normal, as serum HCV RNA is reported to be undetectable in approximately half of anti-HCV positive IVDUs (McCruden et al. 1996; Simmonds et al. 1990). The availability of a fully automatic, inexpensive serologic test as surrogate of methods for HCV RNA detection would be particularly useful in these patients. Our aim has been to assess the significance of IgM antibodies to the HCV core protein (IgM antiHCVcore) in a group of IVDUs.

MATERIALS AND METHODS Patients Eighty-six consecutive patients were enrolled in this study, which was conducted in accordance with the principles of the Declaration of Helsinki. The clinical and demographic characteristics of the patients, who were IVDUs attending a clinic for the prevention and treatment of drug dependence and had tested positive at a third-generation enzyme immunoassay for the detection of serum anti-HCV antibodies (Ortho Diagnostics, Raritan, NJ, USA), are summarized in Table 1. None of them was known to be a recent seroconverter. When they visited the clinic, blood samples were taken for both routine and research tests, and serum samples, prepared in aliquots and frozen within 2 h, were kept at 280°C until use (approximately 18 months later). Forty-five of these patients have been described previously (Toniutto et al. 1996).

HCV IgM Serology Serum IgM anti-HCVcore antibodies were detected by a solid-phase enzyme immunoassay (HCV IgM EIA 2.0, Abbott GmbH, Wiesbaden-Delkenheim, Germany). The optical densities of three negative and three positive control sera, included in each run, were read to verify the validity of the test and to calculate a sample-to-cutoff (S/CO) ratio. Retesting

TABLE 1 Clinical and Demographic Characteristics of Patients Variable

Value

Female (n/total) Age (years) mean 6 SD range Length of intravenous drug use (years) mean 6 SD range Injections $30/week (n/total) Anti-HIV-positive (n/total) HBsAg-positive (n/total) Bilirubin (mmol/l), mean 6 SD g-globulins (g/l) 6 SD Alanine aminotransferase* Persistently within normal range (n/total) Abnormal on #50% of determinations (n/total) Abnormal on .50% of determinations (n/total)

17/86 30.2 6 4.3 22–41 9.9 6 5.3 1–22 18/86 7/86 2/86 12 6 8 23.2 6 0.7

44/86 28/86 14/86

* Six determinations for each patient, on average; any value above the upper limit of reference range was considered abnormal.

in duplicate of the samples with S/CO ratios .1 was required for confirmation of IgM anti-HCVcore reactivity.

HCV RNA Determinations Serum HCV RNA determination was carried out by reverse transcriptase (RT)-polymerase chain reaction (PCR) with nested primers derived from the 59 noncoding region (outer antisense: 59-GTC CAC GGT CTA CGA GAC CT-39; outer sense: 59-GCC ATG GCG TTA GTA TGA GT-39; inner sense: 59-GTG CAG CCT CCA GGA CCC-39; and inner antisense: 39-CCG TGA GCG TTC GTG GGA TA-59), as described previously (Toniutto et al. 1996). Both negative (deionized ultrapure water and serum from antiHCV-negative healthy blood donors) and positive controls were used in each run. The sensitivity of this assay was 1 3 103 genomes per mL. The samples found HCV RNA-positive were submitted to a second RT-PCR to amplify sequences of HCV core RNA. A double PCR with two pairs of primers deduced from the core region (outer antisense: 59-ATG TAC CCC ATG AGG TCG GC-39; outer sense. 59-CGC GCG ACT AGG AAG ACT TC39; inner sense: 59-AGG AAG ACT TCC GAG CGG TC-39; inner antisense: 59-CAC GTA AGG GTA TCG ATG AC-39) was used, with slight modifications from a previously described method (Okamoto et al. 1992). The sensitivity of this assay was 5 3 103 genomes per mL.

IgM Anti-Hepatitis C Virus Core Protein in Drug Users TABLE 2 Association Between Anti-HCV IgM Reactivity and Virological Features

Statistical Analysis

Anti-HCV IgM Positive (n 5 62)

Negative (n 5 24)

39 30

14 5

15 4 6 2 12

1 5 4 1 3

Serum HCV RNA positivity 59 noncoding core HCV types HCV 1a HCV 1b HCV 2a HCV 2b HCV 3a

71

Statistical analysis of the data was performed by means of the BMDP/Dynamic statistical software package, release 7.0 (Statistical Software Ltd, Cork, Ireland). Shapiro and Wilk’s W test was applied to test normality of data (BMDP program 2D). When the data departed significantly from a normal distribution, they were transformed logarithmically. Numerical variables were compared between groups by Student’s t tests (two-tailed) (BMDP program 3D). The associations among ordinal variables were analyzed by Pearson’s chi-square tests (BMDP program 4F).

HCV Genotyping

RESULTS

HCV strains were typed by dot blot hybridization of amplicons obtained by RT-PCR with nested primers derived from the 59 noncoding region (Toniutto et al. 1996). Briefly, 10 mL of the PCR products from each sample, after denaturation and incubation in a prehybridization solution, were spotted on five nitrocellulose filters and hybridized with genotype-specific 32 P-labeled oligonucleotide probes (Type 1: 59-CGC TCA ATG CCT GGA GAT-39; Type 2a: 59-CAC TCT ATG CCC GGC CAT-39; Type 2b 59-CAC TCT ATG TCC GGT CAT-39; Type 3: 59-CGC TCA ATA CCC AGA AAT-39; Type 4: 59-CGC TCA ATA GCC CGG AAA T-39). Hybridization was performed at 42°C for 2 h. After thorough washing of the filters, autoradiography was performed using Kodak X-OMAT AR films, with 2-h exposure at 280°C. To distinguish subtype 1a from subtype 1b, PCR products were digested with BstU I endonuclease. This restriction enzyme recognizes two restriction sites within HCV-1b products and one in HCV-1a products, thus generating three fragments (of 137 bp, 44 bp, and 30 bp) in the presence of HCV-1b and two fragments (of 167 bp and 44 bp) in the presence of HCV 1a. The digestion products can be visualized by electrophoresis on 12% acrylamide gels and ethidium bromide staining.

Fifty-three patients had detectable serum HCV RNA: 16 had infection by HCV genotype 1a, 9 by HCV 1b, 10 by HCV 2a, 3 by HCV 2b, and 15 by HCV 3a. Anti-HCV IgM antibodies were found in 62/86 (72%) patients. IgM anti-HCVcore reactivity was more common in patients positive for core HCV RNA ( p 5 0.020), whereas no association was observed between IgM anti-HCVcore reactivity and 59 noncoding HCV RNA ( p 5 0.696) (Table 2). Patients infected by HCV type 1a were found positive more often for IgM anti-HCVcore (15/16 vs. 24/37, p 5 0.029) and for core HCV RNA (15/16 vs. 20/37, p 5 0.005) than patients infected by other HCV genotypes. Abnormalities of alanine aminotransferases (ALT) were not associated with detection of serum HCV RNA. In the subsample of patients with abnormal ALT, similar to what occurred in the overall population, a positive result of the IgM anti-HCVcore test displayed a trend associated with a positive result of serum HCV RNA testing, although statistical significance was lost, probably because of the smaller sample size (Table 3). The detection of IgM anti-HCVcore antibodies was more common in patients .30 years of age (a cutoff value representing the mean age of the study population) in comparison with younger subjects (33/40 vs. 29/46, p 5 0.045). HIV-seropositive sub-

TABLE 3 Relationship Between Serum HCV RNA Positivity, Serum Alanine Aminotransferase (ALT) Elevation, and Anti-HCV IgM reactivity HCV RNA 59 noncoding

Abnormal ALT Of which anti-HCV IgM positive

Core

Positive (N 5 53)

Negative (N 5 33)

Positive (N 5 35)

Negative (N 5 51)

27

15

19

23

19

10

16*

13*

* p 5 0.053 by Pearson chi-square test.

72 jects tended to be more frequently IgM antiHCVcore-negative (4/24 vs 3/62, p 5 0.07). No differences were observed in either the rates of IgM anti-HCVcore reactivity or the IgM anti-HCVcore S/CO ratios when subjects were dichotomized according to gender.

DISCUSSION The high rates of IgM anti-HCVcore seropositivity that we report here in IVDUs have been found in other categories of anti-HCV-positive subjects (Quiroga et al. 1995; Yuki et al. 1995). The reason that such a high proportion of anti-HCV-positive subjects maintain a measurable IgM anti-core response for a long time is unclear, as the usual pattern for an IgM response is to wane with time following a viral infection. In patients with hepatitis B, for example, the IgM anti-hepatitis B core antigen titre declines in the weeks after the acute infection, to reappear only during flareups of the disease (Smith et al. 1992). However, the mechanisms involved in liver injury during HCV infection might differ from those involved in hepatitis B. Indeed, it has been proposed that the IgM anti-HCV antibodies might play a role in the complement-mediated lysis of the HCVinfected hepatocytes and might neutralize circulating HCV in serum (Quiroga et al. 1992). We wondered whether the absence of a detectable IgM anti-HCVcore response might be accompanied by the absence of detectable serum HCV RNA. However, we could not demonstrate any association between IgM anti-HCVcore and viremia, as assessed by a highly sensitive RT-PCR with primers derived from the 59noncoding region. Interestingly, our data showed instead an association between measurable IgM anti-HCVcore response and positive results of the RT-PCR with nested primers derived from the core region. One possible explanation for this apparently discrepant result is that IgM anti-HCVcore antibodies may persist in subjects with higher viremic titres. Indeed, in the present study the RT-PCR assay for core HCV RNA was less sensitive than the RTPCR assay for 59noncoding HCV RNA. An association between viremic titre and IgM anti-HCVcore has already been suggested by others (Yuki et al. 1995). Higher viremic titres might also underlie the differences we found in IgM anti-HCVcore seropositivity according to age, because of the natural history of HCV infection (Gretch et al. 1994). However, against

P. Toniutto et al. this interpretation is that several IgM anti-HCVcorepositive sera tested HCV RNA negative at both assays; moreover, the sensitivities of the two RT-PCR assays, although different, are of the same order of magnitude. Alternatively, it can be hypothesized that many of the IgM anti-HCVcore-negative subjects might be infected by HCV variants with core nucleotide sequences having a high degree of heterogeneity in comparison with published isolates. On the one hand, this would result in a primer/template mismatch; on the other, to a mutated core gene sequence might correspond a core protein profoundly different from the recombinant HC-34 protein recognized by the IgM anti-HCVcore assay. Indeed, IVDUs differ from other groups at risk for HCV infection in having a different genotype distribution, with muchhigher prevalence of recently emerged genotypes, particularly HCV 3a (Silini et al. 1995a); moreover, isolates with genotypes different from HCV 1 are often negative at the core HCV RNA assay (Toniutto et al. 1996). According to this interpretation, IgM anti-HCVcore reactivity might represent a marker of exposure to HCV 1a. This might explain why it has been associated with poor response to interferon therapy (Quiroga et al. 1995; Yuki et al. 1995); in fact, patients infected by HCV types 2 and 3 are known to have higher response rates (Simmonds 1997). Finally, the peculiar genotype distribution found in IVDUs might explain why we did not observe differences in IgM anti-HCVcore seropositivity according to the presence of abnormalities of liver function tests, as reported in patients with chronic liver diseases. In fact, infection by the HCV genotypes that are common in IVDUs has been found to be associated with normal liver biochemistry (Silini et al. 1995b). In summary, the IgM anti-HCVcore assay, which frequently tests positive in IVDUs, cannot be used as a reliable surrogate test for serum HCV RNA. The lack of an association between IgM anti-HCVcore reactivity and HCV RNA seropositivity may derive from the peculiar distribution of HCV variants in this category of subjects.

The authors are indebted to Dario Liani for expert technical assistance, to Dr. Lorenza Ciani for cloning of the HCV core gene, and to Dr. Karlhans Endlich for his help in the preparation of the manuscript.

REFERENCES Alberti A, Morsica G, Chemello L, Cavalletto D, Noventa F, Pontisso P, Ruol A (1992) Hepatitis C viraemia and liver

disease in symptom-free individuals with anti-HCV. Lancet 340:697–698.

IgM Anti-Hepatitis C Virus Core Protein in Drug Users

Brillanti S, Masci C, Ricci P, Miglioli M, Barbara L (1992) Significance of IgM antibody to hepatitis C virus in patients with chronic hepatitis C. Hepatology 15:998– 1001. De Medina M, Schiff ER (1995) Hepatitis C: diagnostic assays. Sem Liver Dis 15:33–40. Esteban JI, Esteban R, Vialdomiu L, et al. (1989) Hepatitis C virus among risk groups in Spain. Lancet ii:294–297. Govindarajan S, Gupta S, Valinluck B, Redeker AG (1989) Correlation of IgM anti-hepatitis D virus (HDV) to HDV RNA in sera of chronic HDV. Hepatology 10:34–35. Gretch D, Corey L, Wilson J, dela Rosa C, Willson R, Carithers R Jr, Busch M, Hart J, Sayers M, Han J (1994) Assessment of hepatitis C virus RNA levels by quantitative competitive RNA polymerase chain reaction: High titer viremia correlates with advanced stage of disease. J Infect Dis 169:1219–1225. Marin MG, Bresciani S, Puoti M, Rodella A, Gussago A, Ravaggi A, Pizzocolo G, Albertini A, Cariani E (1994) Clinical significance of serum hepatitis C virus (HCV) RNA as marker of HCV infection. J Clin Microbiol 32: 3008–3012. McCruden EAB, Hillan KJ, McKay IC, Cassidy MT, Clark JC (1996) Hepatitis virus infection and liver disease in injecting drug users who died suddenly. J Clin Pathol 49:552–555. Okamoto H, Sugiyama Y, Okada S, Kurai K, Akahane Y, Sugai Y, Tanaka T, Sato K, Tsuda F, Miyakawa Y, Mayumi M (1992) Typing hepatitis C virus by polymerase chain reaction with type-specific primers: Application to clinical surveys and tracing infectious sources. J Gen Virol 73:673–679. Quiroga JA, Bosch O, Gonzalez R, Marriott E, Castillo I, Bartolome´ J, Carren˜o V (1992) Immunoglobulin M antibody to hepatitis C virus during interferon therapy for chronic hepatitis C. Gastroenterology 103:1285–1289. Quiroga JA, van Binsbergen J, Wang CY, Pardo M, Navas S, Trines C, Herrero M, Carren˜o V (1995) Immunoglobulin M antibody to hepatitis C virus core antigen: Correlations with viral replication, histological activity, and liver disease outcome. Hepatology 22:1635–1640.

73

Scully LJ, Brown D, Lloyd C, Shein R, Thomas HC (1990) Immunological studies before and during interferon therapy in chronic HBV infection: Identification of factors predicting response. Hepatology 12:1111–1117. Silini E, Bono F, Cividini A, Cerino A, Maccabruni A, Tinelli C, Bruno S, Bellobuono A, Mondelli MU (1995a) Molecular epidemiology of hepatitis C virus infection among intravenous drug users. J Hepatol 22:691–695. Silini E, Bono F, Cividini A, Cerino A, Bruno S, Rossi S, Belloni G, Brugnetti B, Civardi E, Salvaneschi L, Mondelli MU (1995b) Differential distribution of hepatitis C virus genotypes in patients with and without liver function abnormalities. Hepatology 21:285–290. Simmonds P (1997) Clinical relevance of hepatitis C virus genotypes. Gut 40:291–293. Simmonds P, Zhang LQ, Watson HG, Rebus S, Ferguson ED, Balfe P, Leadbetter GH, Yap PL, Ludlam CA (1990) Hepatitis C quantification and sequencing in blood products, hemophiliacs, and drug users. Lancet 336:1469–1472. Smith HM, Lau JYN, Daniels HM, Alexander GJM, Williams R (1992) Significance of IgM anti-HBc in chronic hepatitis B virus infection. J Med Virol 36:16–20. Toniutto P, Pirisi M, Tisminetzky SG, Fabris C, Chinellato E, Gerotto M, Falleti E, Ferroni P, Lombardelli T, Bartoli E, Baralle F (1996) Discordant results from hepatitis C virus genotyping by procedures based on amplification of different genomic regions. J Clin Microbiol 34:2382– 2385. van der Hoek JAR, van Haastrecht HJA, Goudsmith J, de Wolf F, Coutino RA (1990) Prevalence, incidence, and risk factors of hepatitis C virus infection among drug users in Amsterdam. J Infect Dis 162:823–826. Yuki N, Hayashi N, Ohkawa K, Hagiwara H, Oshita M, Katayama K, Sasaki Y, Kasahara A, Fusamoto H, Kamada T (1995) The significance of immunoglobulin M antibody response to hepatitis C virus core protein in patients with chronic hepatitis C. Hepatology 22:402–406.