Hepatitis C Virus RNA and Antibody Response in the Clinical Course of Acute Hepatitis C Virus Infection MASSIMOPUOTI,~ ANTONELLA ZONARO,~ANTONELLA RAVAGGI,~ MARIA GWIA MAR IN,^ FILIPPOCASTELNUOV03 AND ELISABETTA cARIAN12 'Department of Znfectious Diseases, University of Brescia; 2Consorzw per le Bwtecnologie, Consiglio Nazionale delle Ricerche (CNR) and Znstitute of Chemistry, School of Medicine, University of Brescia; and 3Z Department of Infectious Diseases, Spedali Civili di Brescia, 25123 Brescia, Italy
Hepatitis C virus RNA, anti-hepatitis C virus immune response and biochemical markers of liver wury were investigated in 17 patients with acute non-A, non-B hepatitis. At the first observation, 1 to 3 wk from the clinical onset, all patients had hepatitis C virus RNA in their serum, and most (16 of 17) were positive for second-generation anti-hepatitis C virus enzyme immunoaesay. Follow-up serum samples were available for 10 patients. The rate of recombinant immunoblot aesay+onfirmed anti-hepatitis C virus enzyme immunoassay reactivities increased from 67% in the first 3 wk to 86%after 21 wk. ElevatedALT levels were associated with hepatitis C virus RNA positivity in most of cases, but the viral nucleic acid was also detected in sera with normal or slightly increased enzyme values. None of the single antibodies tested were related to hepatitis C virus RNA positivity or to the clinical phase of the infection. Therefore hepatitis C virus RNA determination might provide important additional information as compared with antihepatitis C virus markers, allowing earlier diagnosis, discrimination of active infection and, possibly, prognostic evaluation. (HEF+ATOLOGY 1992;16:877-881.)
reaction (PCR) technique, is presently the only tool for the determination of HCV viremia (6-10). Although some antibodies apparently correlate with infectivity and are developed early in the course of acute infection (4), PCR detection of HCV RNA remains the only way to differentiate active from past resolved infection. In addition, little information is available on the relationship of HCV viremia with antibody status and liver injury in the course of HCV infection. In this study we examined HCV RNA status in acute HCV infection with regard to anti-HCV immune response and to biochemical markers of liver injury.
Address reprint request3 to: Elisabetta Cariani, Consorzio per le Biotecnologie, Laboratorio di Biotecnologie, P.le Sped& C i a , 1, 25123 Brescia, Italy. 31/1139466
USed.
PATIENTS AND METHODS Patkr~t8. The Health Services of the district of Brescia,
Italy, were notified of 53 cases of acute NANBH in the period November 1989 to January 1991. T'hlrty-two patients were admitted to the Department of Infectious Diseases of the local general hospital in the same period for acute symptomatic NANBH. All patients were negative for serological markers of active HAV, HBV and HDV infection, tested with commercially available kits (Abbott Laboratories, North Chicago, IL, and Sorin Biomedica, Saluggia, Italy), for HBV DNA, assayed by Hepatitis C virus (HCV),a recently identified positive- radiological molecular hybridization (Abbott Genostics kit) stranded RNA virus (l), is the etiological agent of both and for IgM or increased IgG antibodies to cytomegalovirus or transfusion-associated and community-acquired non-A, Epstein-Barr virus (Flow Laboratories, Irvine, UK). Other non-B hepatitis (NANBH) (2, 3). Molecular cloning of noninfective causes of hepatitis (e.g., metabolic diseases, the HCV genome allowed the expression of recombinant alcohol abuse and hepatotoxic drugs) were excluded in all viral proteins and the development of diagnostic tech- patients. Four of the 32 patients were positive for anti-HIV niques aimed at the detection of antiviral antibodies antibodies and were excluded from this study to avoid possible (2-4). The first-generation assays, enabling the detection alterations in immune response. Acute-phase serum samples in the first 3 wk from the onset of clinical symptoms of antibodies against nonstructural viral proteins (e.g., colleded were available for 17 of 28 patients; their clinical profiles are ClOO and 5-1-1),were impaired by late seroconversion in shown in Table 1. acute infection and poor correlation with infectivity (5). HCV Serological Te8t8. Anti-HCV antibodies were assayed Second-generation anti-HCV assays, relying on the use in duplicate by enzyme immunoassay (EM 2; Ortho Diagnostic of both structural and nonstructural viral antigens, Systems, Raritan, NJ). Semiquantitative confirmation of provide an earlier and more specific appraisal of viral EM-positive samples was performed by the recombinant infection (4). However, the direct detection of the viral immunoblot assay (RIBA) for anti-HCV (4 Antigen RIBA, nucleic acid, carried out by the polymerase chain Ortho Diagnostic Systems). HCV RNA Determination. Serum HCV RNA was analyzed in all patients and in serum samples from healthy blood donors (negative controls). AU samples were stored at -80" C until Received December 6,1991;accepted May 17, 1992. HCV RNA was tested by a reverse transcription-nested polymerase chain reaction (PCR) assay with primers located in
877
878
HEPATOLOGY
PUOTI ET AL.
TABLE 1. Clinical protile of acute hepatitis C patients
R N B eapiet
Characteristic 4
O,D. 450
1
10
1
10
10
Value
I
1
10
1
10
nl
n2
Patients Mean age (range)
17 33 (17-68)
Sex M F Risk factor Transfusion Intravenous drug use Dialysis unknown HBV markers None HBcAb HBsAb Mean ALT peak (range)
11 6
4 8 1
4 15 1 1 1,910 (680-3,584)
(Stratagene, San Diego, CA) to establish the degree of sensitivity of the method. The RNA fragment corresponding to assay for HCV RNA. Reverse transcription-nested PCR products the positive strand of the viral genome was synthesized in uitro obtained from serial dilutions (from lo4 to lo-' copies) of HCV RNA by the RNA transcription kit (Stratagene). Serial dilutions of transcribed in uitro were analyzed by EB staining of agarose gel the in uitro-transcribed RNA were then reverse transcribed @ottom)and by the DEIA assay (top). Negative controls: nl = reaction with primer lA, and one tenth of the complementary DNA was mixture without nucleic acids; n2 = nucleic acids extracted from normal control sera. The cutoffvalue for the DEIA test (optical density amplified using the same protocol as above. DEIA. The DEIA test is a nonisotopic detection system (O.D.) 450 = 0.2) was derived by the mean + 3 S.D. of the results for that relies on the use of a mouse monoclonal antibody able 20 replicates of the PCR-negative controls. to discriminate double-stranded from single-stranded DNA (14). This test has already been applied to the analysis of the conserved 5' untranslated region, as previously described HBV-amplified (141, HCV-amplified (11)and HDV-amplified (11).Total RNA was extracted from 200 p1 of serum by the (15) fragments and, having an ELISA format, allows the guanidinium thiocyanate-phenol-chloroformprocedure (12). simultaneous analysis of a large number of samples. An Reverse transcription was performed on one third of the RNA HCV-specific oligonucleotide probe internal to the PCR volume with a HCV-specific primer located in the 5' product to be analyzed (5'-AGGTGAGTACACCGGAATTGC untranslated region of the viral genome (13) (lA, 5'- CAGGACGACCGGGTCCTTTCT-3', sense, position from GATGCACGGTCTACGAGACCTC-3', antisense, position -186 to -143) was immobilized through an avidin-biotin from - 1to - 21), and one tenth of the complementary DNA bridge on the wells of a microtiter plate and hybridized to was used for the first round of amplification (35 cycles; the denatured PCR product (20 pl) as previously described denaturation = 94" C for 1min; annealing = 45" C for 1min; (11).After the addition of the anti-double-stranded DNA elongation = 72" C for 2 min), performed with external antibody, enabling the detection of the hybridization event, primers 1A and 1B (5'-AACTACTGTCTTCACGCAGAA-3',and of a peroxidase-labeled antimouse Ig antibody, the resense, position from - 289 to - 269). In the second round, 3 pl action was read with a spectrophotometer at 450 nm. of the first reaction was amplified through 25 cycles using the same protocol as above and internal primers 2A (5'RESULTS GCGACCCAACACTACTCGGCT-3', antisense, position from - 70 to - 90) and 2B (5'-ATGGCGTTAGTATGAGTG-3', Sensitivity of HCVRNA Determination. In this study, sense, position from - 257 to - 240). One fifth (20 p1) of the HCV RNA was detected in serum samples by the PCR products was analyzed by agarose gel electrophoresis and combination of reverse transcription, nested PCR amethidium bromide (EB) staining. The specificity of the am- plification and nonisotopic hybridization (DEW test). plified fragments was further confirmed by a nonradioactive Serial dilutions of the RNA fragment amplified in the hybridization assay (DNA EIA [DEIAI). first step of nested PCR, transcribed in uitro from a Pre-PCR and post-PCR reactions were physically separated recombinant plasmid, were used as a template to to prevent contamination in the different steps of HCV RNA establish the degree of sensitivity of this assay. The amplification. Negative controls (reaction mixture without nested PCR products were examined by EB staining of nucleic acids or with nucleic acids extracted from normal control sera) were performed in each experiment, and results agarose gel and by the DEIA test (Fig. 1).The lower were only considered if consistent in at least three independent detection limit both by EB staining and by DELA was between lo2 and 10 RNA copies present in the initial experiments. The PCR product obtained with first-step primers 1A and sample used for reverse transcription. In fact, the 1B was cloned in the SmaI site of Bluescript plasmid vector sample containing 10' RNA copies was strongly positive
FIG.1. Sensitivity of the reverse transcription-nested PCR-DEIA
879
HCV RNA IN ACUTE INFECTION
Vol. 16, No. 4, 1992
TABLE 2. Relationship between HCV RNA, biochemical and immUnological parameters in the clinical course of acute
HCV infection
NV
=
Single antibodies
RIBA 2
“ m e from onset
(Wk)
Numberofsamples
ALT < 150% x N V
HCVRNA
EL42
R
I
NR
C22
ClOO
C33c
1-3 4-11 12-20 21-52
17
17 6
17 8 8 1
15 11 11 7
10 8 9
4
3
11 11
1 2
2 -
6
1
-
12 8 10 7
3 6 6 2
10 8 9 6
normal values; R
4 1
7 =
reactive; I
=
indeterminate; NR
=
6-1-1
5 7 6 4
not reactive
by both methods, whereas the further 1:lO dilution resulted negative. HCV RNA and Antibody Detection in Clinical Samplee. Serum samples from 17 acute hepatitis patients were examined at the moment of hospitalization, 1to 3 wk from the onset of clinical symptoms. Follow-up serum samples collected from 10 patients during and after clinical illness for 16 to 52 wk (mean = 31 wk) were available for analysis. At the first observation, all samples contained HCV RNA sequences (Table 2). ALT levels were increased ( > 150% of normal values in all patients), and the second-generation EIA test was positive in 15 of 17 cases. However, when RIBA 2 was performed, EIA positivity was confirmed in only 10 samples; 4 samples were indeterminate, and 1 sample was negative. Concordant EIA 2-negative and RIBA 2-negative results were obtained in two HCV RNApositive cases. The analysis of follow-up serum samples revealed a progressive reduction in ALT levels in 8 of 10 patients. In most of the cases (8 of lo), a single ALT peak was observed, whereas two patients showed a biphasic ALT profile with two peaks at 3 wk and at 16 wk from one another, respectively. In two patients ALT remained more than 150% of normal value, and HCV RNA was positive at 19 wk and 25 wk from the clinical onset. AU the other patients had normal ALT levels and were negative for HCV RNA at the end of the follow-upperiod. The relationship between ALT values and HCV RNA status was not absolute during the clinical course of HCV infection. Even if HCV RNA positivity was always observed in samples with ALT levels more than 150%of normal value, the viral nucleic acid was also detected in samples with normal or slightly elevated ALT in the first 20 wk from the onset of clinical hepatitis (Table 2). The rate of EIA 2 positivity raised from 88% at the first observation to 100% on serum samples collected after 4 wk of illness. The percentage of RIBA 2-confirmed anti-HCV positivity was also increasing with the time of observation (67% at the first observation, 73% at 4 to 11wk from onset, 82% at 12 to 20 wk and 86% after 21 wk). The pattern of antibody response to the different anti-HCV antibodies was variable in single patients. At the first observation, anti-C22 and anti-C33c were more frequently observed (70.5%and 59% of samples) than anti-(3100 and anti-5-1-1 (18% and 29% of samples). During follow-up, all patients seroconverted to anti-C22
and anti-C33c, whereas three patients did not have anti-C100 and 5-1-1 develop. Reactivity against C22, ClOO and C33c disappeared in one, three and two patients, respectively, at variable intervals from onset (11 to 35 wk). Intermittent positivity of anti-HCV reactivities was also detected in some patients (Fig. 2). No sigmficant correlation was observed between HCV RNA positivity and reactivity against a specific antibody (p > 0.05 by x2 test with Yates’ correlation). DISCUSSION
The development of diagnostic assays for the detection of antibodies against HCV proteins greatly improved our epidemiological knowledge of HCV infection (1-3).The anti-HCV assays used, based on recombinant antigens corresponding to the viral nucleocapsid protein (C22) and nonstructural antigens encoded by NS3 (C33c) and NS4 (C100, 5-1-1) HCV genes, enable earlier diagnosis and, apparently, display better correlation with infectivity as compared with first-generation tests, which are only based on ClOO and 5-1-1 antigens (4). Because of the low titer of HCV particles in infected sera (3), the PCR is potentially the only test allowing the direct detection of circulating viral genomes (6-10).PCR-based techniques might therefore provide significant new information concerning the pattern of viral replication in the natural course of acute and chronic HCV infection. In this study HCV RNA, ALT levels and seroconversion to anti-HCV antibodies were monitored in acute NANBH patients. The viral nucleic acid was analyzed by a PCR protocol designed to obtain increased sensitivity and specificity. Nested PCR was performed with primers located in the conserved 5’ untranslated region of the viral genome (10, 131, and the specificity of the PCR results was confirmed by the DEIA assay, a nonisotopic detection system that provides the same information as hybridization with radioactively labeled probes (11,14, 15). The lower detection limit of this method is between lo2 and 10 copies/sample on RNA transcribed in vitro. Serum HCV RNA was present in all patients in the first sample tested, obtained 1to 3 wk after the onset of symptoms. Most of patients were also positive for anti-HCV tested by the second-generation EIA assay. However, in two cases, HCV RNA represented the only marker of HCV infection, and analysis by the RIBA 2 assay confirmed EIA results in only 10 of 15Em-positive patients. These observations are consistent with the
880
PUOTI ET AL.
HEPATOLOGY
Patient #I I
I
I
0
HCV RNA EIA c32 ClOO
+
c33c
10 20 30 40 Weeks from transfusion
50
Patient 13
Patient #l I
ALT 3ooo
I
’+ + + +
10
0
I
+
+ +a
c22
ClOO c33c 5-1.1
20
30
I
HCV R N A EIA
40
Weeks from onset
ALT
I
+
I
+
UUM __*_
I
I
---
+ -
I
+
-
10
I i
HCV RNA EIA c22 ClOO C33c 5-1-1
30
Weeks from onset
FIG.2. Serological, virological and biochemical course of HCV infection in three representative patients with acute hepatitis. Solid bars represent positive HCV RNA, and open bars represent negative HCV RNA. For single antibodies tested by RIBA horizontal bars indicate positive assay, and horizontal lines represent negative assay. The first sample obtained from patient 1(posttransfusionhepatitis) 7 wk after transfusion and 3 wk after the onset of symptoms was HCV RNA positive and EIA 2 positive, RIBA 2 indeterminate (anti-C22 reactive). Seroconversion to anti-C100, anti-C33c and anti-5-1-1 was observed 16 wk after transfusion. All antibodies except anti-C100 were s t i detectable at 56 wk. ALT levels fell to normal value 27 wk after transfusion, when HCV RNA was still positive, and remained in the normal limits during the subsequent 18 mo. HCV RNA was negative since 35 wk. Patient 2 (community-acquired hepatitis) was positive for HCV RNA, EIA 2 and RIBA 2 assays in correspondence with the ALT peak. ALT levels fell to normal value 13 wk after onset, but HCV RNA was still detected a t 21 wk. RIBA 2 was indeterminate since 17 wk. Seroconversion to anti-C100 and anti-5-1-1 was not observed during follow-up (39 wk). In patient 3 (community-acquired hepatitis), HCV RNA was the only marker detected 1wk after the onset of symptoms. EIA 2 and RIBA 2 were negative until 4 wk, and intermittent positivity of anti-C100 and anti-5-1-1was observed in subsequent samples. HCV RNA and increased ALT levels were persistently detected throughout the 25 wk of observation.
earlier appearance of serum HCV RNA in comparison with antiviral antibodies during acute infection. In addition, because all EIA-positive, RIBA-indeterminate or nonreactive patients had serum HCV RNA and subsequently had RIBA positivity develop, the data suggest that EIA 2 is more sensitive than RIBA in the early phase of clinical hepatitis. Second-generation RIBA was reported to correlate better with viremia than the first-generation assays (4). These results show that this relationship is not absolute, at least in the clinical phase of acute hepatitis, further indicating that only HCV RNA testing may provide early and direct information on viral replication. This is also consistent with previous observations that demonstrated a delay of up to 12 wk between HCV RNA detection and anti-HCV appearance (8). The specificity of HCV RNA detection in acute hepatitis samples lacking serological markers or reac-
tivity with confirmation tests was supported by anti-HCV seroconversion in follow-up serum samples from the same patients. The pattern of antibody response was extremely variable, with earlier appearance, in most cases, of anti-C22 and anti-C33c. In some instances, antibodies were intermittently positive or disappeared at variable intervals during the follow-up. None of the tested antibodies appeared to correlate with the remission of hepatitis, thus excluding a possible protective or prognostic meaning of the anti-HCV reactivities assayed in this study. Because of the heterogeneous anti-HCV immune response, no clear correlation of single antibodies with the clinical stage or with viral replication could be seen. Although in acute-phase serum samples HCV RNA positivity was always associated with increased ALT levels, this relationship was not constant during follow-up because ALT normalization often did not
Vol. 16,No. 4, 1992
HCV RNA IN ACUTE INFECTION
88 1
Melpolder JC, Houghton M, Choo 5. Alter HJ, Purcell RH, Shih JW, Q-L, Kuo G. Detection of antibody to hepatitis C virus in prospectively followed transfusion recipients with acute and chronic non-A, non-B hepatitis. N Engl J Med 1989;321:14941500. 6. Weiner AJ, Kuo G, Bradley DW, Bonino F, Saracco G, Lee C, Rosenblatt J, et al. Detection of hepatitis C viral sequences in non-A, non-B hepatitis. Lancet 1990;335:1-3. 7. Garson JA, Tedder Rs, Briggs M, Tuke P, Glazebrook JA, Trute A, Parker D, et al. Detection of hepatitis C viral sequences in blood donations by “nested” polymerase chain reaction and prediction of infectivity. Lancet 1990;335:1419-1422. Jett B, Purcell RH. 8. Farci P, Alter HJ, Wong D, Miller RH, Shih JW, A long-term study of hepatitis C virus replication in non-A, non-B hepatitis. N Engl J Med 1991;325:98-104. 9. Cristiano K, Di Bisceglie AM, Hoofnagle JH,Feinstone SM. Hepatitis C viral RNA in serum of patients with chronic non-A, non-B hepatitis: detection by the polymerase chain reaction using 1991;14:51-55. multiple primer sets. HEPATOLDGY 10. Houghton M, Weiner A, Han J, Kuo G, Choo Q-L. Molecular biology of the hepatitis C viruses: implications for diagnosis, 1991;14: development and control of viral disease. HEPATOLOGY 381-388. 11. Imberti L, Cariani E, Bettinardi A, Zonaro A, Albertini A, Primi D. An immunoassay for specific amplified HCV sequences. J Virol Methods 1991;34:233-243. 12.Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroformextraction. Anal Biochem 1987;162:156-159. 13. Okamoto H, Okada S, Sugtyama Y, Yotsumoto S, Tanaka T, Yoshizawa H, Tsuda F, et al. The 5’ terminal sequence of hepatitis C virus genome. Jpn J Exp Med 1990;60:167-177. 14.Mantero G, Zonaro A, Albertini A, Bertolo P, Primi D. DNA enzyme immunoassay: general method for detecting producta of polymerase chain reaction. Clin Chem 1991;37:422-429. 15. Cariani E, Ravaggi A, Puoti M, Mantero G, Albertini A, Primi D. REFERENCES Evaluation of hepatitis delta virus RNA levels during interferon therapy by analysis of polymerase chain reaction products with a 1. Choo QL, Kuo G, Weiner AJ, Overby LR, Bradley DW, Houghton nonradioisotopic hybridization assay. HEPATOLOGY1992;15: M. Isolation of cDNA clone derived from a blood-borne non-A, 685-689. non-B viral hepatitis genome. Science 1989;244:359-362. 2. Kuo G,Choo Q-L, Alter HJ, Gitnick GL, Redeker AG, Purcell RH, 16.Zanetti A, Tanzi E, Zehender G, Magni E, Incarbone C, Zonaro A, Primi D, et al. Hepatitis C virus RNA in symptomless donors Miyamura T, et al. An assay for circulating antibodies to a major implicated in post-transfusion non-A, non-B hepatitis. Lancet etiologic virus of human non-A, non-B hepatitis. Science 1989; 1990;335:448. 244~362-364. 3. Choo Q-L, Weiner AJ, Overby LR, Kuo G, Houghton M. Hepatitis 17 Berman N, Alter HJ, Hishak KG, Purcell RH, Jones EA. The chronic sequelae of non-A, non-B hepatitis. Ann Intern Med C virus: the mqjor causative agent of viral non-A, non-B hepatitis. 1979;91:1-6. Br Med Bull 1990;46:423-441. 4. Van der Poel CL, Cuypers HTM, Reesink H W ,Weiner A J , Quan 18 Koretz RL, Stone 0, Gitnick GL. The long term course of non-A, non-B post-transfusion hepatitis. Gastroenterology 1980;79: S, di Nello R, van Boven JJP, et al. Confirmation of hepatitis C 893-898. virus infection by new four-antigen recombinant immunoblot asmy. Lancet 1991;337:317-319.
correspond to the clearance of the viral nucleic acid. This is consistent with previous reports on prospectively observed patients with posttransfusion hepatitis (8) and on HCV transmission by blood donors with normal ALT levels (16). The lack of a constant correlation between viral replication and elevated ALT has suggested the absence of a direct cytopathic effect of HCV, at least in some phases of infection, or the existence of extrahepatic sites of replication (8, 10). Alternately, the extremely high sensitivity of the PCR assay could allow the detection of low levels of viral replication that are not sufficient to determine ALT elevation. Eight of the 10 patients included in the follow-up study had normal ALT levels at the end of observation. Even if the follow-up period was too short to correctly estimate the rate of chronic evolution, our observation differs from other published data (17,181. This might be because of the specific characteristics of the study population, consisting only of hospitalized symptomatic patients. By contrast, previous data was primarily derived from prospectively studied posttransfusion hepatitis patients, including the asymptomatic cases. In conclusion, this study confirms the importance of HCV RNA testing for early diagnosis and direct appraisal of viremia. The variability of anti-HCV response indicates that HCV RNA determination may provide crucial information about the clinical phase of the infection.
Hepatitis C virus RNA, anti-hepatitis C virus immune response and biochemical markers of liver wury were investigated in 17 patients with acute non-A, non-B hepatitis. At the first observation, 1 to 3 wk from the clinical onset, all patients had hepatitis C virus RNA in their serum, and most (16 of 17) were positive for second-generation anti-hepatitis C virus enzyme immunoaesay. Follow-up serum samples were available for 10 patients. The rate of recombinant immunoblot aesay+onfirmed anti-hepatitis C virus enzyme immunoassay reactivities increased from 67% in the first 3 wk to 86%after 21 wk. ElevatedALT levels were associated with hepatitis C virus RNA positivity in most of cases, but the viral nucleic acid was also detected in sera with normal or slightly increased enzyme values. None of the single antibodies tested were related to hepatitis C virus RNA positivity or to the clinical phase of the infection. Therefore hepatitis C virus RNA determination might provide important additional information as compared with antihepatitis C virus markers, allowing earlier diagnosis, discrimination of active infection and, possibly, prognostic evaluation. (HEF+ATOLOGY 1992;16:877-881.)
reaction (PCR) technique, is presently the only tool for the determination of HCV viremia (6-10). Although some antibodies apparently correlate with infectivity and are developed early in the course of acute infection (4), PCR detection of HCV RNA remains the only way to differentiate active from past resolved infection. In addition, little information is available on the relationship of HCV viremia with antibody status and liver injury in the course of HCV infection. In this study we examined HCV RNA status in acute HCV infection with regard to anti-HCV immune response and to biochemical markers of liver injury.
Address reprint request3 to: Elisabetta Cariani, Consorzio per le Biotecnologie, Laboratorio di Biotecnologie, P.le Sped& C i a , 1, 25123 Brescia, Italy. 31/1139466
USed.
PATIENTS AND METHODS Patkr~t8. The Health Services of the district of Brescia,
Italy, were notified of 53 cases of acute NANBH in the period November 1989 to January 1991. T'hlrty-two patients were admitted to the Department of Infectious Diseases of the local general hospital in the same period for acute symptomatic NANBH. All patients were negative for serological markers of active HAV, HBV and HDV infection, tested with commercially available kits (Abbott Laboratories, North Chicago, IL, and Sorin Biomedica, Saluggia, Italy), for HBV DNA, assayed by Hepatitis C virus (HCV),a recently identified positive- radiological molecular hybridization (Abbott Genostics kit) stranded RNA virus (l), is the etiological agent of both and for IgM or increased IgG antibodies to cytomegalovirus or transfusion-associated and community-acquired non-A, Epstein-Barr virus (Flow Laboratories, Irvine, UK). Other non-B hepatitis (NANBH) (2, 3). Molecular cloning of noninfective causes of hepatitis (e.g., metabolic diseases, the HCV genome allowed the expression of recombinant alcohol abuse and hepatotoxic drugs) were excluded in all viral proteins and the development of diagnostic tech- patients. Four of the 32 patients were positive for anti-HIV niques aimed at the detection of antiviral antibodies antibodies and were excluded from this study to avoid possible (2-4). The first-generation assays, enabling the detection alterations in immune response. Acute-phase serum samples in the first 3 wk from the onset of clinical symptoms of antibodies against nonstructural viral proteins (e.g., colleded were available for 17 of 28 patients; their clinical profiles are ClOO and 5-1-1),were impaired by late seroconversion in shown in Table 1. acute infection and poor correlation with infectivity (5). HCV Serological Te8t8. Anti-HCV antibodies were assayed Second-generation anti-HCV assays, relying on the use in duplicate by enzyme immunoassay (EM 2; Ortho Diagnostic of both structural and nonstructural viral antigens, Systems, Raritan, NJ). Semiquantitative confirmation of provide an earlier and more specific appraisal of viral EM-positive samples was performed by the recombinant infection (4). However, the direct detection of the viral immunoblot assay (RIBA) for anti-HCV (4 Antigen RIBA, nucleic acid, carried out by the polymerase chain Ortho Diagnostic Systems). HCV RNA Determination. Serum HCV RNA was analyzed in all patients and in serum samples from healthy blood donors (negative controls). AU samples were stored at -80" C until Received December 6,1991;accepted May 17, 1992. HCV RNA was tested by a reverse transcription-nested polymerase chain reaction (PCR) assay with primers located in
877
878
HEPATOLOGY
PUOTI ET AL.
TABLE 1. Clinical protile of acute hepatitis C patients
R N B eapiet
Characteristic 4
O,D. 450
1
10
1
10
10
Value
I
1
10
1
10
nl
n2
Patients Mean age (range)
17 33 (17-68)
Sex M F Risk factor Transfusion Intravenous drug use Dialysis unknown HBV markers None HBcAb HBsAb Mean ALT peak (range)
11 6
4 8 1
4 15 1 1 1,910 (680-3,584)
(Stratagene, San Diego, CA) to establish the degree of sensitivity of the method. The RNA fragment corresponding to assay for HCV RNA. Reverse transcription-nested PCR products the positive strand of the viral genome was synthesized in uitro obtained from serial dilutions (from lo4 to lo-' copies) of HCV RNA by the RNA transcription kit (Stratagene). Serial dilutions of transcribed in uitro were analyzed by EB staining of agarose gel the in uitro-transcribed RNA were then reverse transcribed @ottom)and by the DEIA assay (top). Negative controls: nl = reaction with primer lA, and one tenth of the complementary DNA was mixture without nucleic acids; n2 = nucleic acids extracted from normal control sera. The cutoffvalue for the DEIA test (optical density amplified using the same protocol as above. DEIA. The DEIA test is a nonisotopic detection system (O.D.) 450 = 0.2) was derived by the mean + 3 S.D. of the results for that relies on the use of a mouse monoclonal antibody able 20 replicates of the PCR-negative controls. to discriminate double-stranded from single-stranded DNA (14). This test has already been applied to the analysis of the conserved 5' untranslated region, as previously described HBV-amplified (141, HCV-amplified (11)and HDV-amplified (11).Total RNA was extracted from 200 p1 of serum by the (15) fragments and, having an ELISA format, allows the guanidinium thiocyanate-phenol-chloroformprocedure (12). simultaneous analysis of a large number of samples. An Reverse transcription was performed on one third of the RNA HCV-specific oligonucleotide probe internal to the PCR volume with a HCV-specific primer located in the 5' product to be analyzed (5'-AGGTGAGTACACCGGAATTGC untranslated region of the viral genome (13) (lA, 5'- CAGGACGACCGGGTCCTTTCT-3', sense, position from GATGCACGGTCTACGAGACCTC-3', antisense, position -186 to -143) was immobilized through an avidin-biotin from - 1to - 21), and one tenth of the complementary DNA bridge on the wells of a microtiter plate and hybridized to was used for the first round of amplification (35 cycles; the denatured PCR product (20 pl) as previously described denaturation = 94" C for 1min; annealing = 45" C for 1min; (11).After the addition of the anti-double-stranded DNA elongation = 72" C for 2 min), performed with external antibody, enabling the detection of the hybridization event, primers 1A and 1B (5'-AACTACTGTCTTCACGCAGAA-3',and of a peroxidase-labeled antimouse Ig antibody, the resense, position from - 289 to - 269). In the second round, 3 pl action was read with a spectrophotometer at 450 nm. of the first reaction was amplified through 25 cycles using the same protocol as above and internal primers 2A (5'RESULTS GCGACCCAACACTACTCGGCT-3', antisense, position from - 70 to - 90) and 2B (5'-ATGGCGTTAGTATGAGTG-3', Sensitivity of HCVRNA Determination. In this study, sense, position from - 257 to - 240). One fifth (20 p1) of the HCV RNA was detected in serum samples by the PCR products was analyzed by agarose gel electrophoresis and combination of reverse transcription, nested PCR amethidium bromide (EB) staining. The specificity of the am- plification and nonisotopic hybridization (DEW test). plified fragments was further confirmed by a nonradioactive Serial dilutions of the RNA fragment amplified in the hybridization assay (DNA EIA [DEIAI). first step of nested PCR, transcribed in uitro from a Pre-PCR and post-PCR reactions were physically separated recombinant plasmid, were used as a template to to prevent contamination in the different steps of HCV RNA establish the degree of sensitivity of this assay. The amplification. Negative controls (reaction mixture without nested PCR products were examined by EB staining of nucleic acids or with nucleic acids extracted from normal control sera) were performed in each experiment, and results agarose gel and by the DEIA test (Fig. 1).The lower were only considered if consistent in at least three independent detection limit both by EB staining and by DELA was between lo2 and 10 RNA copies present in the initial experiments. The PCR product obtained with first-step primers 1A and sample used for reverse transcription. In fact, the 1B was cloned in the SmaI site of Bluescript plasmid vector sample containing 10' RNA copies was strongly positive
FIG.1. Sensitivity of the reverse transcription-nested PCR-DEIA
879
HCV RNA IN ACUTE INFECTION
Vol. 16, No. 4, 1992
TABLE 2. Relationship between HCV RNA, biochemical and immUnological parameters in the clinical course of acute
HCV infection
NV
=
Single antibodies
RIBA 2
“ m e from onset
(Wk)
Numberofsamples
ALT < 150% x N V
HCVRNA
EL42
R
I
NR
C22
ClOO
C33c
1-3 4-11 12-20 21-52
17
17 6
17 8 8 1
15 11 11 7
10 8 9
4
3
11 11
1 2
2 -
6
1
-
12 8 10 7
3 6 6 2
10 8 9 6
normal values; R
4 1
7 =
reactive; I
=
indeterminate; NR
=
6-1-1
5 7 6 4
not reactive
by both methods, whereas the further 1:lO dilution resulted negative. HCV RNA and Antibody Detection in Clinical Samplee. Serum samples from 17 acute hepatitis patients were examined at the moment of hospitalization, 1to 3 wk from the onset of clinical symptoms. Follow-up serum samples collected from 10 patients during and after clinical illness for 16 to 52 wk (mean = 31 wk) were available for analysis. At the first observation, all samples contained HCV RNA sequences (Table 2). ALT levels were increased ( > 150% of normal values in all patients), and the second-generation EIA test was positive in 15 of 17 cases. However, when RIBA 2 was performed, EIA positivity was confirmed in only 10 samples; 4 samples were indeterminate, and 1 sample was negative. Concordant EIA 2-negative and RIBA 2-negative results were obtained in two HCV RNApositive cases. The analysis of follow-up serum samples revealed a progressive reduction in ALT levels in 8 of 10 patients. In most of the cases (8 of lo), a single ALT peak was observed, whereas two patients showed a biphasic ALT profile with two peaks at 3 wk and at 16 wk from one another, respectively. In two patients ALT remained more than 150% of normal value, and HCV RNA was positive at 19 wk and 25 wk from the clinical onset. AU the other patients had normal ALT levels and were negative for HCV RNA at the end of the follow-upperiod. The relationship between ALT values and HCV RNA status was not absolute during the clinical course of HCV infection. Even if HCV RNA positivity was always observed in samples with ALT levels more than 150%of normal value, the viral nucleic acid was also detected in samples with normal or slightly elevated ALT in the first 20 wk from the onset of clinical hepatitis (Table 2). The rate of EIA 2 positivity raised from 88% at the first observation to 100% on serum samples collected after 4 wk of illness. The percentage of RIBA 2-confirmed anti-HCV positivity was also increasing with the time of observation (67% at the first observation, 73% at 4 to 11wk from onset, 82% at 12 to 20 wk and 86% after 21 wk). The pattern of antibody response to the different anti-HCV antibodies was variable in single patients. At the first observation, anti-C22 and anti-C33c were more frequently observed (70.5%and 59% of samples) than anti-(3100 and anti-5-1-1 (18% and 29% of samples). During follow-up, all patients seroconverted to anti-C22
and anti-C33c, whereas three patients did not have anti-C100 and 5-1-1 develop. Reactivity against C22, ClOO and C33c disappeared in one, three and two patients, respectively, at variable intervals from onset (11 to 35 wk). Intermittent positivity of anti-HCV reactivities was also detected in some patients (Fig. 2). No sigmficant correlation was observed between HCV RNA positivity and reactivity against a specific antibody (p > 0.05 by x2 test with Yates’ correlation). DISCUSSION
The development of diagnostic assays for the detection of antibodies against HCV proteins greatly improved our epidemiological knowledge of HCV infection (1-3).The anti-HCV assays used, based on recombinant antigens corresponding to the viral nucleocapsid protein (C22) and nonstructural antigens encoded by NS3 (C33c) and NS4 (C100, 5-1-1) HCV genes, enable earlier diagnosis and, apparently, display better correlation with infectivity as compared with first-generation tests, which are only based on ClOO and 5-1-1 antigens (4). Because of the low titer of HCV particles in infected sera (3), the PCR is potentially the only test allowing the direct detection of circulating viral genomes (6-10).PCR-based techniques might therefore provide significant new information concerning the pattern of viral replication in the natural course of acute and chronic HCV infection. In this study HCV RNA, ALT levels and seroconversion to anti-HCV antibodies were monitored in acute NANBH patients. The viral nucleic acid was analyzed by a PCR protocol designed to obtain increased sensitivity and specificity. Nested PCR was performed with primers located in the conserved 5’ untranslated region of the viral genome (10, 131, and the specificity of the PCR results was confirmed by the DEIA assay, a nonisotopic detection system that provides the same information as hybridization with radioactively labeled probes (11,14, 15). The lower detection limit of this method is between lo2 and 10 copies/sample on RNA transcribed in vitro. Serum HCV RNA was present in all patients in the first sample tested, obtained 1to 3 wk after the onset of symptoms. Most of patients were also positive for anti-HCV tested by the second-generation EIA assay. However, in two cases, HCV RNA represented the only marker of HCV infection, and analysis by the RIBA 2 assay confirmed EIA results in only 10 of 15Em-positive patients. These observations are consistent with the
880
PUOTI ET AL.
HEPATOLOGY
Patient #I I
I
I
0
HCV RNA EIA c32 ClOO
+
c33c
10 20 30 40 Weeks from transfusion
50
Patient 13
Patient #l I
ALT 3ooo
I
’+ + + +
10
0
I
+
+ +a
c22
ClOO c33c 5-1.1
20
30
I
HCV R N A EIA
40
Weeks from onset
ALT
I
+
I
+
UUM __*_
I
I
---
+ -
I
+
-
10
I i
HCV RNA EIA c22 ClOO C33c 5-1-1
30
Weeks from onset
FIG.2. Serological, virological and biochemical course of HCV infection in three representative patients with acute hepatitis. Solid bars represent positive HCV RNA, and open bars represent negative HCV RNA. For single antibodies tested by RIBA horizontal bars indicate positive assay, and horizontal lines represent negative assay. The first sample obtained from patient 1(posttransfusionhepatitis) 7 wk after transfusion and 3 wk after the onset of symptoms was HCV RNA positive and EIA 2 positive, RIBA 2 indeterminate (anti-C22 reactive). Seroconversion to anti-C100, anti-C33c and anti-5-1-1 was observed 16 wk after transfusion. All antibodies except anti-C100 were s t i detectable at 56 wk. ALT levels fell to normal value 27 wk after transfusion, when HCV RNA was still positive, and remained in the normal limits during the subsequent 18 mo. HCV RNA was negative since 35 wk. Patient 2 (community-acquired hepatitis) was positive for HCV RNA, EIA 2 and RIBA 2 assays in correspondence with the ALT peak. ALT levels fell to normal value 13 wk after onset, but HCV RNA was still detected a t 21 wk. RIBA 2 was indeterminate since 17 wk. Seroconversion to anti-C100 and anti-5-1-1 was not observed during follow-up (39 wk). In patient 3 (community-acquired hepatitis), HCV RNA was the only marker detected 1wk after the onset of symptoms. EIA 2 and RIBA 2 were negative until 4 wk, and intermittent positivity of anti-C100 and anti-5-1-1was observed in subsequent samples. HCV RNA and increased ALT levels were persistently detected throughout the 25 wk of observation.
earlier appearance of serum HCV RNA in comparison with antiviral antibodies during acute infection. In addition, because all EIA-positive, RIBA-indeterminate or nonreactive patients had serum HCV RNA and subsequently had RIBA positivity develop, the data suggest that EIA 2 is more sensitive than RIBA in the early phase of clinical hepatitis. Second-generation RIBA was reported to correlate better with viremia than the first-generation assays (4). These results show that this relationship is not absolute, at least in the clinical phase of acute hepatitis, further indicating that only HCV RNA testing may provide early and direct information on viral replication. This is also consistent with previous observations that demonstrated a delay of up to 12 wk between HCV RNA detection and anti-HCV appearance (8). The specificity of HCV RNA detection in acute hepatitis samples lacking serological markers or reac-
tivity with confirmation tests was supported by anti-HCV seroconversion in follow-up serum samples from the same patients. The pattern of antibody response was extremely variable, with earlier appearance, in most cases, of anti-C22 and anti-C33c. In some instances, antibodies were intermittently positive or disappeared at variable intervals during the follow-up. None of the tested antibodies appeared to correlate with the remission of hepatitis, thus excluding a possible protective or prognostic meaning of the anti-HCV reactivities assayed in this study. Because of the heterogeneous anti-HCV immune response, no clear correlation of single antibodies with the clinical stage or with viral replication could be seen. Although in acute-phase serum samples HCV RNA positivity was always associated with increased ALT levels, this relationship was not constant during follow-up because ALT normalization often did not
Vol. 16,No. 4, 1992
HCV RNA IN ACUTE INFECTION
88 1
Melpolder JC, Houghton M, Choo 5. Alter HJ, Purcell RH, Shih JW, Q-L, Kuo G. Detection of antibody to hepatitis C virus in prospectively followed transfusion recipients with acute and chronic non-A, non-B hepatitis. N Engl J Med 1989;321:14941500. 6. Weiner AJ, Kuo G, Bradley DW, Bonino F, Saracco G, Lee C, Rosenblatt J, et al. Detection of hepatitis C viral sequences in non-A, non-B hepatitis. Lancet 1990;335:1-3. 7. Garson JA, Tedder Rs, Briggs M, Tuke P, Glazebrook JA, Trute A, Parker D, et al. Detection of hepatitis C viral sequences in blood donations by “nested” polymerase chain reaction and prediction of infectivity. Lancet 1990;335:1419-1422. Jett B, Purcell RH. 8. Farci P, Alter HJ, Wong D, Miller RH, Shih JW, A long-term study of hepatitis C virus replication in non-A, non-B hepatitis. N Engl J Med 1991;325:98-104. 9. Cristiano K, Di Bisceglie AM, Hoofnagle JH,Feinstone SM. Hepatitis C viral RNA in serum of patients with chronic non-A, non-B hepatitis: detection by the polymerase chain reaction using 1991;14:51-55. multiple primer sets. HEPATOLDGY 10. Houghton M, Weiner A, Han J, Kuo G, Choo Q-L. Molecular biology of the hepatitis C viruses: implications for diagnosis, 1991;14: development and control of viral disease. HEPATOLOGY 381-388. 11. Imberti L, Cariani E, Bettinardi A, Zonaro A, Albertini A, Primi D. An immunoassay for specific amplified HCV sequences. J Virol Methods 1991;34:233-243. 12.Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroformextraction. Anal Biochem 1987;162:156-159. 13. Okamoto H, Okada S, Sugtyama Y, Yotsumoto S, Tanaka T, Yoshizawa H, Tsuda F, et al. The 5’ terminal sequence of hepatitis C virus genome. Jpn J Exp Med 1990;60:167-177. 14.Mantero G, Zonaro A, Albertini A, Bertolo P, Primi D. DNA enzyme immunoassay: general method for detecting producta of polymerase chain reaction. Clin Chem 1991;37:422-429. 15. Cariani E, Ravaggi A, Puoti M, Mantero G, Albertini A, Primi D. REFERENCES Evaluation of hepatitis delta virus RNA levels during interferon therapy by analysis of polymerase chain reaction products with a 1. Choo QL, Kuo G, Weiner AJ, Overby LR, Bradley DW, Houghton nonradioisotopic hybridization assay. HEPATOLOGY1992;15: M. Isolation of cDNA clone derived from a blood-borne non-A, 685-689. non-B viral hepatitis genome. Science 1989;244:359-362. 2. Kuo G,Choo Q-L, Alter HJ, Gitnick GL, Redeker AG, Purcell RH, 16.Zanetti A, Tanzi E, Zehender G, Magni E, Incarbone C, Zonaro A, Primi D, et al. Hepatitis C virus RNA in symptomless donors Miyamura T, et al. An assay for circulating antibodies to a major implicated in post-transfusion non-A, non-B hepatitis. Lancet etiologic virus of human non-A, non-B hepatitis. Science 1989; 1990;335:448. 244~362-364. 3. Choo Q-L, Weiner AJ, Overby LR, Kuo G, Houghton M. Hepatitis 17 Berman N, Alter HJ, Hishak KG, Purcell RH, Jones EA. The chronic sequelae of non-A, non-B hepatitis. Ann Intern Med C virus: the mqjor causative agent of viral non-A, non-B hepatitis. 1979;91:1-6. Br Med Bull 1990;46:423-441. 4. Van der Poel CL, Cuypers HTM, Reesink H W ,Weiner A J , Quan 18 Koretz RL, Stone 0, Gitnick GL. The long term course of non-A, non-B post-transfusion hepatitis. Gastroenterology 1980;79: S, di Nello R, van Boven JJP, et al. Confirmation of hepatitis C 893-898. virus infection by new four-antigen recombinant immunoblot asmy. Lancet 1991;337:317-319.
correspond to the clearance of the viral nucleic acid. This is consistent with previous reports on prospectively observed patients with posttransfusion hepatitis (8) and on HCV transmission by blood donors with normal ALT levels (16). The lack of a constant correlation between viral replication and elevated ALT has suggested the absence of a direct cytopathic effect of HCV, at least in some phases of infection, or the existence of extrahepatic sites of replication (8, 10). Alternately, the extremely high sensitivity of the PCR assay could allow the detection of low levels of viral replication that are not sufficient to determine ALT elevation. Eight of the 10 patients included in the follow-up study had normal ALT levels at the end of observation. Even if the follow-up period was too short to correctly estimate the rate of chronic evolution, our observation differs from other published data (17,181. This might be because of the specific characteristics of the study population, consisting only of hospitalized symptomatic patients. By contrast, previous data was primarily derived from prospectively studied posttransfusion hepatitis patients, including the asymptomatic cases. In conclusion, this study confirms the importance of HCV RNA testing for early diagnosis and direct appraisal of viremia. The variability of anti-HCV response indicates that HCV RNA determination may provide crucial information about the clinical phase of the infection.
