- Email: [email protected]
Antibodies to hepatitis C virus in low-risk blood donors: Implications for counseling positive donors
GASTROENTEROCOGY 1991;101:1724-1727
Antibodies to Hepatitis C Virus in Low-Risk Blood Donors: Implications for Counseling Positive Donors HENRY H. HSU, MARITZA GONZALEZ, STEVEN K. H. FOUNG, STEPHEN M. FEINSTONE, and HARRY B. GREENBERG Departments of Medicine, Microbiology and Immunology, and Pathology, Stanford University Medical Center and Palo Alto Veterans Administration Medical Center, Palo Alto, California; and Laboratory of Hepatitis Research, Food and Drug Administration, Bethesda, Maryland
The significance of antibodies to hepatitis C virus (HCV) found by screening enzyme-linked immunoassay testing in a low-risk blood donor population is unclear. The rate of false positivity in this group as well as the usefulness of supplemental testing were examined by correlating the results of two screening enzyme-linked immunoassays (Ortho Diagnostics, Raritan, NJ, and Abbott Laboratories, North Chicago, IL) with supplemental antibody testing by the recombinant immunoblot assays (1 and 2) (fktho) and neutralization assay (Abbott). Polymerase chain reaction was used to detect HCV genomic RNA to confirm viremia. Among 11,243 volunteer donors who were screened for the presence of antibodies to HCV by enzyme-linked immunoassay, 60 (0.53%) sera were repeatedly reactive. Twenty-five of these 60 sera were available for further testing. Seven sera were reactive by both screening enzyme-linked immunoassays, as well as by both recombinant immunoblot and neutralization assays. Six of these seven sera had detectable HCV genomic RNA by polymerase chain reaction. Among the remaining 16 sera, none were reactive by either recombinant immunoblot assays, whereas two sera were reactive by the neutralization assay only. None of the 16 samples had detectable HCV genomic RNA. Five of the six sera with elevated aminotransferase levels were among the seven sera reactive by all immunoassays. It is concluded that there is a significant false positivity rate associated with screening enzymelinked immunoassay testing in a low-risk blood donor population. Supplemental testing correlates well with detection of hepatitis C genomic material by polymerase chain reaction- and identifies donors who are truly infected.
S
creening for antibodies to hepatitis C virus (HCV) has been implemented by blood banks to reduce the risk of posttransfusion hepatitis. Seroepidemiologic surveys indicate that between 0.3% and 1.4% of sera of low-risk volunteer blood donors are reactive when tested by the currently licensed screening enzyme-linked imrnunoassays (ELISA) for anti-HCV antibody (l-4). This represents potentially a large pool of asymptomatic individuals who may be harboring hepatitis C infection. However, the proportion of donors with false positive ELISA reactivity has not been clearly defined, and it is currently unclear how to reliably distinguish these individuals from donors who are truly infected. Three supplementary tests, the OrthoKhiron (Ortho Diagnostics, Raritan, NJ) recombinant immunoblot assays (RIBAs) 1 and 2 and the Abbott (Abbott Laboratories, North Chicago, IL) neutralization assay, are available to aid in discriminating between the true and false positive results obtained by initial screening ELISA. The usefulness of these additional antibody assays in differentiating between true and false positives has not been fully clarified. To better define the meaning of these assays in a low-risk volunteer blood donor population, we tested screening ELISA-reactive donors with the RIBA-1, RIBA-2, and neutralization assays. Hepatitis C virus genomic RNA was detected using polymerase chain reaction (PCR) to show hepatitis C viremia.
Abbreviations used in this paper: CID, chimpanzee infectious dose: ELISA, enzyme-linked immuno~rbent assay: PCR, polymerase chain reaction; RIBA, recombinant immunoblot assay; SOD, superoxide dismutase. o 1991 by the American Gastroenterological Association 0016~5065/91/$3.00
December 1991
Methods Routine screening for HCV in all blood donations was instituted in May 1990 at the Stanford University Blood Center. All donations are tested by both licensed ELISAs (Ortho Diagnostics and Abbott Laboratories]. Between May and August 1990, there were 11,243 blood donations, 60 of which were repeatedly HCV ELISA reactive by either one or both screening assays. Sera were available on 25 of these donors for further testing. There was no discernible selection bias among sera available for further testing and sera unavailable. Surrogate markers [alanine aminotransferase (ALT) and anti-hepatitis B core (HBc) antibody] were routinely obtained on all sera using conventional assays. The 25 sera were subjected to supplemental testing with both the Ortho RIBA- and RIBA- and the Abbott neutralization assay. The Ortho RIBA- includes the following three recombinant antigens: HCV fusion polypeptides 5-l-l (synthesized in Escherichia coli) and CIOO-3(synthesized in yeast) and human superoxide dismutase (SOD) alone (synthesized in yeast), applied to a nitrocellulose strip. The lUBA-2 uses two additional HCV antigens (c33c and ~22-3) derived from different regions than the ClOO-3 region of the genome. The Abbott neutralization enzyme immunoassay uses an E. co&-derived recombinant antigen representing the Cl00 region of the HCV genome, thus, theoretically, eliminating cross-reactivity to yeast or human SOD. It is based on the principle that HCV antigens in solution may block (neutralize) antibodies from binding to the HCV ClOO-3 antigencoated bead. All testing was performed according to the manufacturers’ instructions. Hepatitis C virus RNA was detected in serum by PCR using nested primers as described in detail elsewhere (5). The primer pairs were selected to amplify a segment of the 5’ nontranslated region of the genome, an area whose nucleotide sequence is highly conserved among different HCV isolates. The outer primer pair consisted of the following sequence: 5’-GGCGACACTCCACCATAGATC-3’ and 5’-CATGGTGCACGGTCTACGAGACC-3’; the inner primers consisted of 5’-GGAACTACTGTCTTCACGCAGA-3’ and 5’-TCGCAAGCACCCTATCAGGCAG-3’. Briefly, nucleic acid was extracted from sera stored at -20°C using the guanidinium-phenol-chloroform method. The equivalent of 50 FL of serum was reverse-transcribed using avian myeloblastosis virus reverse transcriptase (Seikagaku, Rockville, MD] with 5 nm of the outer primer pair to initiate synthesis of complementary DNA. The outer primers were used for the first round of PCR (95°C for 1 minute, 50°C for 1 minute, and 72°C for 1 minute for 35 cycles in 50-FL volume). Five microliters of product was then subjected to a second round of PCR using the inner primer set under the same conditions. Ten microliters of amplified PCR product was visualized by ethidium bromide staining after electrophoresis in 2.5% agarose gels. A result was considered positive when a 259-base pair product corresponding to the segment amplified by the inner primer pair was clearly visualized. Negative controls were added during each step, and a result was considered valid only if these controls were negative throughout. All samples were tested under code without
HEPATITIS C TESTING IN BLOOD DONORS 1725
any information regarding the results of the serological assays. We determined the sensitivity of our PCR assay by correlating it with serum known to contain 106-lo7 chimpanzee infectious doses (CID] per milliliter. Our assay is able to detect the equivalent of 1 CID/mL.
Results Sera from 60/11,243 donors (0.53%) were found to be repeatedly reactive by one or both screening ELISAs. During this same period, there were 129 (1.1%) donors who had elevated serum ALT levels and 168 (1.5%) donors who were anti-HBc antibody positive. Among the 25 of the 60 positive sera available for further testing, 17 (68%) were reactive in the Ortho ELISA, whereas 20 of 25 (80%) were reactive in the Abbott ELISA; 12 of 25 (48%) were reactive by both tests (Table 1). Among these 12 sera, 7 were
reactive by all three supplemental assays (Table 1). Two additional sera were reactive by Abbott neutralization assay only. No additional samples were RIBA reactive. There was complete concordance between RIBA- and RIBA- testing. All the RIBA-2-positive samples showed reactivity to all recombinant HCV antigens (~22-3, c33c, and ClOO-3) on the nitrocellulose strip. Thus, 7 of 25 (28%) samples gave concordant results by all five serological tests, leaving 18 of 25 (72%) with discordant results, the majority of which were negative on supplemental testing. In this latter group, there were 5 of ‘18 (28%) samples that were reactive in both initial screening ELISAs. Among the 7 samples with concordant results, 6 had HCV genomic RNA detected by PCR (Table 1). Every one of the remaining 38 discordant samples were PCR negative. Six of the 25 samples had above normal serum ALT levels (range, 49-111 IU/mL); 5 of them were in concordant samples that were also PCR positive. One of 25 samples, also in the concordant group, was anti-HBc antibody positive. Overall, there was excellent concordance between the supplemental assays, surrogate marker abnormalities, and evidence of viremia as detected by HCV PCR.
Discussion The overall prevalence of antibody to HCV by ELISA in our low-risk blood donor population is comparable with that found in other reported blood bank series (4,6). A high percentage of ELISA positives gave discordant results, i.e., they were unconfirmed by supplemental testing. These discordant results likely represent false positive ELISAs. Comparable results were reported by Weiner et al. who noted that 37 of 67 (55%) of Ortho ELISA-reactive samples were not confirmed by RIBA (‘7). Van der Poe1 et al.
1726
GASTROENTEROLOGY Vol. 101. No. 6
HSU ET AL.
Table 1. Results of Screening and Supplementary Assays and HCVPCR Screening assay
Donors with serologically concordant results
Donor no.
Ortho ELISA
1
+ + + + + + +
2 3 4” 5 6 7
Donors with serologically discordant results
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Supplementary
Abbott ELISA
+ + + + + + + + + +
-
-
+ + + + + + + + + + + + +
RIBA-
RIBA-
assay Neutralization
-
-
-
IND
IND
IND -
-
-
IND
IND -
-
+ + IND -
-
Serum ALT
HCV PCR
77 111 103 NL 53 71 NL
Present Present Present Present Present Absent
NL NL NL NL NL NL NL NL NL 49 NL NL NL NL NL NL NL NL
Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent
IND, indeterminate; NL, within normal limits ( < 45 IU/mL). “Anti-HBc antibody positive.
showed that only 6 of 34 (18%) of recipients of one or more ELISA-reactive blood products developed posttransfusion hepatitis, indicating that most ELISApositive units do not transmit HCV (6). Among samples that were concordant, there was excellent agreement between supplementary testing, surrogate markers, and HCV PCR in verifying HCV infection. Because these blood products were excluded from use, we cannot show the potential infectivity of these units. However, Garson et al. have shown that the ability to detect HCV RNA by PCR is a good predictor of infectivity (8). Only one of their six ELISA-reactive donors transmitted non-A, non-B hepatitis: this was also the only donor in whom HCV RNA was detected by PCR. Thus, it seems likely that the donors in the concordant group are infected with hepatitis C. Of interest in this group is the single donor who did not have detectable HCV RNA. Whereas it is possible that the sera of this donor had false positive reactivity to the assays, we suspect it more likely that this individual either had a level of viremia below the detectability of our PCR assay or his infection had cleared and he was no longer viremic. These serological and PCR results suggest that the majority of our low-risk donors with positive results by screening ELISA probably do not have HCV infec-
tion. If this sample is representative of the donor population as a whole, the true prevalence of HCV infection may be significantly lower than that predicted by ELISA testing alone. The RIBA-1, RIBA-2, and neutralization assays appear to correlate well with the ability to detect HCV RNA by PCR; it seems that all are useful confirmatory assays. A larger sample size will be required to determine if these assays are of equivalent specificity, although there were two discordant samples that were reactive by neutralization assay but not by RIBA or HCV PCR. The high proportion of concordant samples that were viremic indicates that the anti-HCV antibodies detected by these immunologic assays correlate with ongoing infection. Using RIBA-2, antibodies against the recombinant HCV antigens ~22-3 and c33c as well as (X00-3, were detected in sera containing HCV genomic RNA, indicating that these antibodies also correlated with ongoing infection and did not appear to be associated with clinical immunity. The general implementation by blood banks of HCV antibody testing among low-risk blood donors means that there will be many asymptomatic individuals who will be seroreactive when- tested with the currently licensed HCV ELISA screening tests. Based on these results, physicians should not diagnose hepati-
HEPATITIS C TESTING IN BLOOD DONORS
December 1991
tis C infection in low-risk blood donors on the basis of screening ELISA testing alone. We recommend that clinical evidence suggestive of liver disease be sought in addition to the use of supplemental serological assays. If the RIBA and/or neutralization assay confirms reactivity, and particularly if there is clinical suspicion of hepatitis, the donor should be counseled that there is a high likelihood that he is infected and will need to be followed for evidence of chronic hepatitis. However, if the confirmatory assays are nonreactive and there is no clinical evidence suggestive of hepatitis, these donors can be reassured that they are unlikely to have HCV infection. At present it is not clear whether any further testing or follow-up is indicated in those individuals with normal liver enzyme and negative supplementary serological testing results. References Janot C, Courouce AM, Maniez M. Antibodies to hepatitis C virus in French blood donors (letter). Lancet 1989;2:796-797. Kuhn1 P, Seidl S, Stangel W, Beyer J, Sibrowski W, Flik J. Antibody to hepatitis C virus in German blood donors (letter). Lancet 1989;2:324. Sirchia G, Almini D, Bellobuono A, Giovanetti AM, Marconi M, Mercuriali F, Mozzi F, Parravicini A, Pizzi M, Zanuso F.
4.
5.
6.
7.
8.
1727
Prevalence of hepatitis C virus antibodies in Italian blood donors. The Italian Cooperative Group. VOX Sang 1990;59:2629. Stevens CE, Taylor PE, Pindyck J, IChoo QL, Bradley DW, Kuo G, Houghton M. Epidemiology of hepatitis C virus. A preliminary study in volunteer blood donors. J.4MA 1990;263:49-53. Cristiano K, Di Bisceglie A, Hoofnagle J, Feinstone S. Hepatitis C viral RNA in serum of patients with chronic non-A, non-B hepatitis: detection by the polymerase chain reaction using multiple primer sets. Hepatology 1991;14:51-55. van der Poe1 CL, Reesink HW, Schaasberg W, Leentvaar KA, Bakker E, Exe1 OP, Lelie PN. Infectivity of blood seropositive for hepatitis C virus antibodies. Lancet 1990;335:558-560, Weiner AJ, Truett MA, Rosenblatt J, Han J, Quan S, Polito AJ, Kuo G, Choo Q, Houghton M, Aqius C. HCV testing in low-risk population (letter). Lancet 1990;336:695. Garson JA, Tedder RS, Briggs M, Tuke P, Glazebrook JA, Trute A, Parker D, Barbara J, Contreras M, Aloysius S. Detection of hepatitis C viral sequences in blood donations by “nested” polymerase chain reaction and prediction of infectivity. Lancet 1990;335:1419-1422.
Received January 28,199l. Accepted July 30,1991. Address requests for reprints to: Henry Hsu, M.D., Gastroenterology Division, Palo Alto Veterans Administration Medical Center, 154-C, 3801 Miranda Avenue, Palo Alto, California 94304. Supported by a National Institutes of Health Training grant (DK07056), a Veterans Administration Merit Review grant, and a National Institutes of Health Digestive Disease Center grant (DK38707).
Antibodies to Hepatitis C Virus in Low-Risk Blood Donors: Implications for Counseling Positive Donors HENRY H. HSU, MARITZA GONZALEZ, STEVEN K. H. FOUNG, STEPHEN M. FEINSTONE, and HARRY B. GREENBERG Departments of Medicine, Microbiology and Immunology, and Pathology, Stanford University Medical Center and Palo Alto Veterans Administration Medical Center, Palo Alto, California; and Laboratory of Hepatitis Research, Food and Drug Administration, Bethesda, Maryland
The significance of antibodies to hepatitis C virus (HCV) found by screening enzyme-linked immunoassay testing in a low-risk blood donor population is unclear. The rate of false positivity in this group as well as the usefulness of supplemental testing were examined by correlating the results of two screening enzyme-linked immunoassays (Ortho Diagnostics, Raritan, NJ, and Abbott Laboratories, North Chicago, IL) with supplemental antibody testing by the recombinant immunoblot assays (1 and 2) (fktho) and neutralization assay (Abbott). Polymerase chain reaction was used to detect HCV genomic RNA to confirm viremia. Among 11,243 volunteer donors who were screened for the presence of antibodies to HCV by enzyme-linked immunoassay, 60 (0.53%) sera were repeatedly reactive. Twenty-five of these 60 sera were available for further testing. Seven sera were reactive by both screening enzyme-linked immunoassays, as well as by both recombinant immunoblot and neutralization assays. Six of these seven sera had detectable HCV genomic RNA by polymerase chain reaction. Among the remaining 16 sera, none were reactive by either recombinant immunoblot assays, whereas two sera were reactive by the neutralization assay only. None of the 16 samples had detectable HCV genomic RNA. Five of the six sera with elevated aminotransferase levels were among the seven sera reactive by all immunoassays. It is concluded that there is a significant false positivity rate associated with screening enzymelinked immunoassay testing in a low-risk blood donor population. Supplemental testing correlates well with detection of hepatitis C genomic material by polymerase chain reaction- and identifies donors who are truly infected.
S
creening for antibodies to hepatitis C virus (HCV) has been implemented by blood banks to reduce the risk of posttransfusion hepatitis. Seroepidemiologic surveys indicate that between 0.3% and 1.4% of sera of low-risk volunteer blood donors are reactive when tested by the currently licensed screening enzyme-linked imrnunoassays (ELISA) for anti-HCV antibody (l-4). This represents potentially a large pool of asymptomatic individuals who may be harboring hepatitis C infection. However, the proportion of donors with false positive ELISA reactivity has not been clearly defined, and it is currently unclear how to reliably distinguish these individuals from donors who are truly infected. Three supplementary tests, the OrthoKhiron (Ortho Diagnostics, Raritan, NJ) recombinant immunoblot assays (RIBAs) 1 and 2 and the Abbott (Abbott Laboratories, North Chicago, IL) neutralization assay, are available to aid in discriminating between the true and false positive results obtained by initial screening ELISA. The usefulness of these additional antibody assays in differentiating between true and false positives has not been fully clarified. To better define the meaning of these assays in a low-risk volunteer blood donor population, we tested screening ELISA-reactive donors with the RIBA-1, RIBA-2, and neutralization assays. Hepatitis C virus genomic RNA was detected using polymerase chain reaction (PCR) to show hepatitis C viremia.
Abbreviations used in this paper: CID, chimpanzee infectious dose: ELISA, enzyme-linked immuno~rbent assay: PCR, polymerase chain reaction; RIBA, recombinant immunoblot assay; SOD, superoxide dismutase. o 1991 by the American Gastroenterological Association 0016~5065/91/$3.00
December 1991
Methods Routine screening for HCV in all blood donations was instituted in May 1990 at the Stanford University Blood Center. All donations are tested by both licensed ELISAs (Ortho Diagnostics and Abbott Laboratories]. Between May and August 1990, there were 11,243 blood donations, 60 of which were repeatedly HCV ELISA reactive by either one or both screening assays. Sera were available on 25 of these donors for further testing. There was no discernible selection bias among sera available for further testing and sera unavailable. Surrogate markers [alanine aminotransferase (ALT) and anti-hepatitis B core (HBc) antibody] were routinely obtained on all sera using conventional assays. The 25 sera were subjected to supplemental testing with both the Ortho RIBA- and RIBA- and the Abbott neutralization assay. The Ortho RIBA- includes the following three recombinant antigens: HCV fusion polypeptides 5-l-l (synthesized in Escherichia coli) and CIOO-3(synthesized in yeast) and human superoxide dismutase (SOD) alone (synthesized in yeast), applied to a nitrocellulose strip. The lUBA-2 uses two additional HCV antigens (c33c and ~22-3) derived from different regions than the ClOO-3 region of the genome. The Abbott neutralization enzyme immunoassay uses an E. co&-derived recombinant antigen representing the Cl00 region of the HCV genome, thus, theoretically, eliminating cross-reactivity to yeast or human SOD. It is based on the principle that HCV antigens in solution may block (neutralize) antibodies from binding to the HCV ClOO-3 antigencoated bead. All testing was performed according to the manufacturers’ instructions. Hepatitis C virus RNA was detected in serum by PCR using nested primers as described in detail elsewhere (5). The primer pairs were selected to amplify a segment of the 5’ nontranslated region of the genome, an area whose nucleotide sequence is highly conserved among different HCV isolates. The outer primer pair consisted of the following sequence: 5’-GGCGACACTCCACCATAGATC-3’ and 5’-CATGGTGCACGGTCTACGAGACC-3’; the inner primers consisted of 5’-GGAACTACTGTCTTCACGCAGA-3’ and 5’-TCGCAAGCACCCTATCAGGCAG-3’. Briefly, nucleic acid was extracted from sera stored at -20°C using the guanidinium-phenol-chloroform method. The equivalent of 50 FL of serum was reverse-transcribed using avian myeloblastosis virus reverse transcriptase (Seikagaku, Rockville, MD] with 5 nm of the outer primer pair to initiate synthesis of complementary DNA. The outer primers were used for the first round of PCR (95°C for 1 minute, 50°C for 1 minute, and 72°C for 1 minute for 35 cycles in 50-FL volume). Five microliters of product was then subjected to a second round of PCR using the inner primer set under the same conditions. Ten microliters of amplified PCR product was visualized by ethidium bromide staining after electrophoresis in 2.5% agarose gels. A result was considered positive when a 259-base pair product corresponding to the segment amplified by the inner primer pair was clearly visualized. Negative controls were added during each step, and a result was considered valid only if these controls were negative throughout. All samples were tested under code without
HEPATITIS C TESTING IN BLOOD DONORS 1725
any information regarding the results of the serological assays. We determined the sensitivity of our PCR assay by correlating it with serum known to contain 106-lo7 chimpanzee infectious doses (CID] per milliliter. Our assay is able to detect the equivalent of 1 CID/mL.
Results Sera from 60/11,243 donors (0.53%) were found to be repeatedly reactive by one or both screening ELISAs. During this same period, there were 129 (1.1%) donors who had elevated serum ALT levels and 168 (1.5%) donors who were anti-HBc antibody positive. Among the 25 of the 60 positive sera available for further testing, 17 (68%) were reactive in the Ortho ELISA, whereas 20 of 25 (80%) were reactive in the Abbott ELISA; 12 of 25 (48%) were reactive by both tests (Table 1). Among these 12 sera, 7 were
reactive by all three supplemental assays (Table 1). Two additional sera were reactive by Abbott neutralization assay only. No additional samples were RIBA reactive. There was complete concordance between RIBA- and RIBA- testing. All the RIBA-2-positive samples showed reactivity to all recombinant HCV antigens (~22-3, c33c, and ClOO-3) on the nitrocellulose strip. Thus, 7 of 25 (28%) samples gave concordant results by all five serological tests, leaving 18 of 25 (72%) with discordant results, the majority of which were negative on supplemental testing. In this latter group, there were 5 of ‘18 (28%) samples that were reactive in both initial screening ELISAs. Among the 7 samples with concordant results, 6 had HCV genomic RNA detected by PCR (Table 1). Every one of the remaining 38 discordant samples were PCR negative. Six of the 25 samples had above normal serum ALT levels (range, 49-111 IU/mL); 5 of them were in concordant samples that were also PCR positive. One of 25 samples, also in the concordant group, was anti-HBc antibody positive. Overall, there was excellent concordance between the supplemental assays, surrogate marker abnormalities, and evidence of viremia as detected by HCV PCR.
Discussion The overall prevalence of antibody to HCV by ELISA in our low-risk blood donor population is comparable with that found in other reported blood bank series (4,6). A high percentage of ELISA positives gave discordant results, i.e., they were unconfirmed by supplemental testing. These discordant results likely represent false positive ELISAs. Comparable results were reported by Weiner et al. who noted that 37 of 67 (55%) of Ortho ELISA-reactive samples were not confirmed by RIBA (‘7). Van der Poe1 et al.
1726
GASTROENTEROLOGY Vol. 101. No. 6
HSU ET AL.
Table 1. Results of Screening and Supplementary Assays and HCVPCR Screening assay
Donors with serologically concordant results
Donor no.
Ortho ELISA
1
+ + + + + + +
2 3 4” 5 6 7
Donors with serologically discordant results
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Supplementary
Abbott ELISA
+ + + + + + + + + +
-
-
+ + + + + + + + + + + + +
RIBA-
RIBA-
assay Neutralization
-
-
-
IND
IND
IND -
-
-
IND
IND -
-
+ + IND -
-
Serum ALT
HCV PCR
77 111 103 NL 53 71 NL
Present Present Present Present Present Absent
NL NL NL NL NL NL NL NL NL 49 NL NL NL NL NL NL NL NL
Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent Absent
IND, indeterminate; NL, within normal limits ( < 45 IU/mL). “Anti-HBc antibody positive.
showed that only 6 of 34 (18%) of recipients of one or more ELISA-reactive blood products developed posttransfusion hepatitis, indicating that most ELISApositive units do not transmit HCV (6). Among samples that were concordant, there was excellent agreement between supplementary testing, surrogate markers, and HCV PCR in verifying HCV infection. Because these blood products were excluded from use, we cannot show the potential infectivity of these units. However, Garson et al. have shown that the ability to detect HCV RNA by PCR is a good predictor of infectivity (8). Only one of their six ELISA-reactive donors transmitted non-A, non-B hepatitis: this was also the only donor in whom HCV RNA was detected by PCR. Thus, it seems likely that the donors in the concordant group are infected with hepatitis C. Of interest in this group is the single donor who did not have detectable HCV RNA. Whereas it is possible that the sera of this donor had false positive reactivity to the assays, we suspect it more likely that this individual either had a level of viremia below the detectability of our PCR assay or his infection had cleared and he was no longer viremic. These serological and PCR results suggest that the majority of our low-risk donors with positive results by screening ELISA probably do not have HCV infec-
tion. If this sample is representative of the donor population as a whole, the true prevalence of HCV infection may be significantly lower than that predicted by ELISA testing alone. The RIBA-1, RIBA-2, and neutralization assays appear to correlate well with the ability to detect HCV RNA by PCR; it seems that all are useful confirmatory assays. A larger sample size will be required to determine if these assays are of equivalent specificity, although there were two discordant samples that were reactive by neutralization assay but not by RIBA or HCV PCR. The high proportion of concordant samples that were viremic indicates that the anti-HCV antibodies detected by these immunologic assays correlate with ongoing infection. Using RIBA-2, antibodies against the recombinant HCV antigens ~22-3 and c33c as well as (X00-3, were detected in sera containing HCV genomic RNA, indicating that these antibodies also correlated with ongoing infection and did not appear to be associated with clinical immunity. The general implementation by blood banks of HCV antibody testing among low-risk blood donors means that there will be many asymptomatic individuals who will be seroreactive when- tested with the currently licensed HCV ELISA screening tests. Based on these results, physicians should not diagnose hepati-
HEPATITIS C TESTING IN BLOOD DONORS
December 1991
tis C infection in low-risk blood donors on the basis of screening ELISA testing alone. We recommend that clinical evidence suggestive of liver disease be sought in addition to the use of supplemental serological assays. If the RIBA and/or neutralization assay confirms reactivity, and particularly if there is clinical suspicion of hepatitis, the donor should be counseled that there is a high likelihood that he is infected and will need to be followed for evidence of chronic hepatitis. However, if the confirmatory assays are nonreactive and there is no clinical evidence suggestive of hepatitis, these donors can be reassured that they are unlikely to have HCV infection. At present it is not clear whether any further testing or follow-up is indicated in those individuals with normal liver enzyme and negative supplementary serological testing results. References Janot C, Courouce AM, Maniez M. Antibodies to hepatitis C virus in French blood donors (letter). Lancet 1989;2:796-797. Kuhn1 P, Seidl S, Stangel W, Beyer J, Sibrowski W, Flik J. Antibody to hepatitis C virus in German blood donors (letter). Lancet 1989;2:324. Sirchia G, Almini D, Bellobuono A, Giovanetti AM, Marconi M, Mercuriali F, Mozzi F, Parravicini A, Pizzi M, Zanuso F.
4.
5.
6.
7.
8.
1727
Prevalence of hepatitis C virus antibodies in Italian blood donors. The Italian Cooperative Group. VOX Sang 1990;59:2629. Stevens CE, Taylor PE, Pindyck J, IChoo QL, Bradley DW, Kuo G, Houghton M. Epidemiology of hepatitis C virus. A preliminary study in volunteer blood donors. J.4MA 1990;263:49-53. Cristiano K, Di Bisceglie A, Hoofnagle J, Feinstone S. Hepatitis C viral RNA in serum of patients with chronic non-A, non-B hepatitis: detection by the polymerase chain reaction using multiple primer sets. Hepatology 1991;14:51-55. van der Poe1 CL, Reesink HW, Schaasberg W, Leentvaar KA, Bakker E, Exe1 OP, Lelie PN. Infectivity of blood seropositive for hepatitis C virus antibodies. Lancet 1990;335:558-560, Weiner AJ, Truett MA, Rosenblatt J, Han J, Quan S, Polito AJ, Kuo G, Choo Q, Houghton M, Aqius C. HCV testing in low-risk population (letter). Lancet 1990;336:695. Garson JA, Tedder RS, Briggs M, Tuke P, Glazebrook JA, Trute A, Parker D, Barbara J, Contreras M, Aloysius S. Detection of hepatitis C viral sequences in blood donations by “nested” polymerase chain reaction and prediction of infectivity. Lancet 1990;335:1419-1422.
Received January 28,199l. Accepted July 30,1991. Address requests for reprints to: Henry Hsu, M.D., Gastroenterology Division, Palo Alto Veterans Administration Medical Center, 154-C, 3801 Miranda Avenue, Palo Alto, California 94304. Supported by a National Institutes of Health Training grant (DK07056), a Veterans Administration Merit Review grant, and a National Institutes of Health Digestive Disease Center grant (DK38707).