Chronic hepatitis C: new diagnostic tools and therapeutic strategies Damen, M.

Transcription

1 UvA-DARE (Digital Academic Repository) Chronic hepatitis C: new diagnostic tools and therapeutic strategies Damen, M. Link to publication Citation for published version (APA): Damen, M. (1999). Chronic hepatitis C: new diagnostic tools and therapeutic strategies General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons). Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible. UvA-DARE is a service provided by the library of the University of Amsterdam ( Download date: 01 Jul 2018

2 Chapter 3 RELIABILITY OF THE THIRD GENERATION RECOMBINANT IMMUNOBLOT ASSAY FOR HEPATITIS C M.Damen, H.L.Zaaijer, H.T.M.Cuypers, H.Vrielink, C.L.vd Poel, H.W.Reesink, and P.N.Lelie Central Laboratory of the Netherlands Red Cross Blood Transfusion Service (CLB), Department of Viral Serology; and the Red Cross Blood Bank Amsterdam, Amsterdam, the Netherlands Transfusion 1995; 35:

3 Background: In a confirmatory laboratory, the second-generation recombinant immunoblot assay (RIBA-2) was replaced by the third-generation RIBA (RIBA-3) in March The aim of this validation study was to compare the sensitivity and specificity of RIBA-2 and RIBA-3 in a routine setting, by using a validated hepatitis C virus (HCV) RNA polymerase chain reaction to establish plasma viremia. Study design and methods: RIBA-2 testing was performed (March 1991-March 1993) in 593 HCV RNA-positive and 1498 HCV RNA-negative subjects. RIBA-3 testing was performed (March 1993-May 1994) in 220 HCV RNA-positive and 530 HCV RNAnegative subjects. All samples reacted for anti-hcv in enzyme-linked immunosorbent assay. Results: In HCV RNA-positive individuals, the sensitivity of RIBA-3 was significantly higher than that of RIBA-2 (99.5% vs. 93.3%, p=0.0005). This was not caused by inclusion of the NS5 antigen, but by a higher sensitivity of the antigens c33 and cloo (RIBA-2: 94.3% and 62.6%; RIBA-3: 99.5% and 88.6%). Replacement of the c22 and c 100 recombinant proteins by synthetic peptides significantly reduced nonspecific reactivity against these antigens (p<0.0001). Unfortunately, increased nonspecific reactivity against the modified c33 antigen and the new NS5 antigen canceled out this effect. Two-band reactivity occurred more often in nonviremic persons than in viremic persons (32.7% vs. 8.2%, po.0001). Risk factors for HCV infection were less frequently observed in 11 blood donors with two-band reactivity than in 6 blood donors with other positive RIBA-3 patterns (18% vs. 83%, p=0.03). Conclusions: The higher sensitivity of RIBA-3 significantly reduced the number of indeterminate test results in HCV RNA-positive persons. Confirmatory laboratories must be aware of the frequent occurrence of nonspecific, isolated reactivity and even nonspecific, twoband reactivity in anti-hcv enzyme-linked immunosorbent assay-reactive blood donors. 54

4 INTRODUCTION In 1989, the hepatitis C virus (HCV) was identified as the major cause of non-a, non-b hepatitis. 1 Since then, assays for the serological diagnosis of HCV infection have been developed and considerably improved. Replacement of the first-generation anti-hcv enzyme-linked imunosorbent assay (ELISA) by the second-generation anti-hcv ELISA provided a significantly increased sensitivity for detection of HCV infection. 2 Unfortunately, screening tests often generate false-positive results, especially in populations at low risk for HCV infection, such as blood donors. For confirmation of reactivity in the second-generation anti-hcv ELISA, a second-generation recombinant immunoblot assay (RIBA-2) was introduced in Evaluation of this assay showed the necessity of HCV RNA detection by polymerase chain reaction (PCR) assay to discriminate between specific and nonspecific indeterminate reactivity and between resolved and ongoing infection. 3 " 5 In 1993, a thirdgeneration anti-hcv ELISA and a third-generation RIBA (RIBA-3), both including the NS5 antigen, became available in Europe. RIBA-3 employs the following HCV antigens to visualize anti-hcv reaction patterns: recombinant c33 and NS5 proteins, and synthetic c 100 and c22 peptides. Earlier, we reported the value of RIBA-3 as compared to other second- and third-generation confirmatory immunoblot assays, using selected HCV RNA-positive and -negative plasma panels. 6 In the present study, we compared RIBA-3 to RIBA-2 in a routine setting. The sensitivity and specificity of the respective assays could be determined in a large number of blood donors and patients with and without viremia, as tested with a validated HCV RNA PCR assay. 78 MATERIALSAND METHODS Study populations From March 1991 to May 1994, we tested 2841 individuals for HCV RNA and RIBA reaction pattern in our confirmatory laboratory. This population consisted of 1879 unpaid Dutch blood donors and 962 hospital patients or outpatients suspected of being infected with HCV, who were not treated with interferon. The samples were all selected because of reactivity in various brands of anti-hcv ELIS As as detected at different laboratories. Most laboratories used a second-generation anti-hcv ELISA (without the NS5 antigen) during the study period. From March 1991 to March 1993, we tested 593 HCV RNA-positive samples and 1498 HCV RNA-negative samples by RIBA-2. From March 1993 to May 1994, we tested 220 HCV RNA-positive samples and 530 HCV RNA-negative samples by RIBA-3. For the composition of the study populations, we excluded follow-up samples from the same individual. During 1994, we studied risk factors for HCV infection in RIBA-3 positive blood donors from 2 Dutch blood banks (n=17). All donors were found to be anti-hcv positive in a third-generation anti-hcv ELISA. In-depth counseling for risk factors and liver disease and a standard physical examination were performed. In addition, we collected a follow-up sample to determine the HCV infection status in RIBA-3, HCV RNA PCR, and liver transaminase tests. The frequency of risk factors in subjects with two-band reactivity in RIBA-3 (n=l 1) was compared to that in subjects with three- or four-band reactivity (n=6). 5 5

5 Serologie assays RIBA-2. The RIBA-2 (Ortho Diagnostics, Raritan, NJ) detects antibodies against the recombinant HCV antigens c22, c33, cloo and The antigens are blotted as separate bands on a nitrocellulose strip, flanked by a weak-positive (1+) control band, a moderately positive (3+) control band, and the fusion protein superoxide dismutase (SOD). A visible band with an intensity between 1+ and 3+ is considered to be 2+, and a visible band with an intensity greater than 3+ is considered to be 4+, according to the manufacturer's instructions. An indeterminate result is defined as single antibody reactivity of 1+ or more or antibody reactivity against cloo and of 1+ or more. Antibody reactivity of 1+ or more against other combinations of two or more antigens is considered to be a positive result. If reactivity against SOD is visible, a positive result is considered to be indeterminate. Third generation strip immunoblot assay. A third-generation strip immunoblot assay for the detection of antibodies to HCV (RIBA HCV 3.0 SIA, Chiron Corporation, Emeryville, CA) detects antibodies to the recombinant proteins c33 and NS5, to the synthetic peptide c22, and to a mixture of cloo synthetic peptides. Antibody reactivity of 1+ or more against two or more antigens is defined as positive. Single antibody reactivity of 1+ or more is considered to be an indeterminate result, according to the manufacturer's instructions. Multiple anti-hcv reactivity combined with reaction to SOD is regarded as indeterminate. HCV RNA cdna PCR We used cdna PCR after extraction of HCV RNA from fresh-frozen plasma samples, to establish viremia. This HCV RNA cdna PCR and the method of plasma collection have been standardized and validated as described before. 7-8 The quality of the in-house PCR assay did not change during the study period, as was proved in each test run by weakpositive and negative control samples. Statistics We used the chi-square test with Yates's correction to compare the sensitivity (in PCRpositive subjects) and specificity (in PCR-negative subjects) of the various antigens in RIBA-2 and RIBA-3. In case of an expected number less than 5, we used Fisher's exact test with two-sided p value. We considered a p value < 0.05 as significant. RESULTS Sensitivity of RIBA-3 and RIBA-2 The RIBA-2 and RIBA-3 results in HCV RNA-positive individuals are shown in Table 1. Table 1. Sensitivity of RIBA-2 and RIBA-3 in HCV RNA-positive, anti-hcv ELISA-positive subjects Total PCR positive (n=813) RIBA positive RIBA indeterminate Number Percentage Number Percentage RIBA-2 RIBA-3 p= n=593 n= %' %* % 1 0.5% 56

6 RIBA-2 was positive in 93.3 percent of the viremic subjects and RIBA-3 was positive in 99.5 percent (p=0.0005). As a result, the proportion of indeterminate results among HCV RNA-positive individuals was significantly lower in RIBA-3 than in RIBA-2. RIBA-3 was indeterminate in only 1 HCV RNA-positive patient (0.5%), who had isolated anti-c33 reactivity, whereas RIBA-2 was indeterminate in 40 HCV RNA-positive patients (6.7%) (32 had isolated anti-c22 and 8 isolated anti-c33 reactivities). In this study population of HCV RNA-positive, anti-hcv ELISA-positive subjects, we did not observe negative RIBA results. Table 2. Sensitivity of individual antigens in RIBA-2 and RIBA-3 in HCV RNA-positive, anti-hcv ELISA-positive subjects PCR positive (n : =813) Tested in RIBA-2 Tested in RIBA-3 Antigen (n=593) (n=220) p value c % 95.5% NS c % 99.5% cloo 62.6% 88.6% O % NS5 68.5% Table 2 compares the individual antigens' rates of reactivity in RIBA-2 and RIBA-3 in HCV RNA-positive subjects. Replacement of the recombinant cloo protein from yeast in RIBA-2 by the mixture of c 100 synthetic peptides in RIBA-3 increased the reaction rate significantly, from 62.6 to 88.6 percent (po.0001). The modified c33 antigen in RIBA-3 improved the sensitivity from 94.3 to 99.5 percent (p=0.002). The replacement of c22 recombinant protein by a mixture of peptides did not alter the sensitivity for core antibodies. In 68.3 percent of the viremic subjects, the new recombinant NS5 protein was reactive. The effect of the enhanced sensitivities of the c33 and the cloo proteins in RIBA-3 on the frequency of anti-hcv reactivities among HCV RNA-positive subjects is demonstrated in Table 3. Three-band reactivity against c22, c33, and cloo was significantly more frequent in RIBA-3 than in RIBA-2 (po.0001). Isolated anti-c22 reactivity disappeared on RIBA-3 in viremic subjects (p=0.001). Isolated anti-cloo and isolated anti-ns5 reactivities were not observed. The NS5 protein was the fourth reactive antigen in 61.1 percent of the viremic cases, the third antigen in 7.4 percent and the second antigen in none. Specificity of RIBA-3 Table 4 shows reactivity patterns of HCV RNA-negative subjects in RIBA-2 and RIBA-3. Replacement of the recombinant proteins c22 and c 100 by synthetic peptides in the RIBA-3 significantly decreased isolated reactivity against these proteins (po.0001). However, the modified c33 antigen doubled the percentage of isolated anti-c33 reactions (p=0.0001). Inclusion of the NS5 protein in the RIBA-3 revealed a new population of PCR-negative individuals with isolated anti-ns5 reactivity. 5 7

7 Table 3. Frequency of RIBA-2 and RIBA-3 reactivities in HCV RNA-positive, anti-hcv ELISApositive subjects PCR positive (n=813) clooand RIBA-2 RIBA-3(%NS5-positive)* c22 c (n=593) (n=220) %> 84.0 % + (61.1 %) %' 10.9 %* (5.5 %) % 4.1 % ( 1.4 %) % 0.5 % (0.5 %) %8 0 % 8 (0 %) % 0.5 % (0 %) % 0 % (0 %) Percentage of the 220 HCV RNA-positive samples showing NS5 reactivity in addition to the indicated RIBA pattern of reactivity. p< p< p=0.001 Table 4. Frequency of RIBA-2 and RIBA-3 reactivities among HCV RNA-negative, anti-hcv ELISA-positive subjects. PCR-negative, ELISA -positive (n = 2028) RIBA-2 RIBA-3 RIBA result (n = 1498) (n = 530) Negative 604 (40.3 %) 262 (49.4 %) Indeterminate 783 (52.3 %) 213 (40.2 %) c (26.4 %)* 48 (9.1 %)* c33 81 (5.4 %y 55 (10.4 %) + cloo and (20.5 %)* 59 (li.i %y NS5 51 (9.6 %) Positive 111 (7.4 %) 55 (10.4 %) 3-4 bands 65 (4.3 %) 37 (7.0 %) c22 and c33 28 (1.9 %) 4 (0.8 %) c22andcl00 10 (0.7 %) 5 (0.9 %) c33andcl00 8 (0.5 %) 8 (1.5 %) NS5 and c22 0 (0 %) NS5 and c33 0 (0 %) NS5andcl00 1 (0.2 %) p< ' p = * p< JS

8 Misclassification of two-band patterns as positive The frequency of RIBA-3 two-band reactivity in HCV RNA-negative subjects could be predicted from the frequency of the isolated reactivities in this population (Table 4). For example, the predicted frequency of c33 and cloo reactivity is 10.4% x 11.1% = 1.2% ; 1.2% x 530 = 6 cases. The actual frequency of this pattern, however, is 1.5 percent (8 cases). Similar results are obtained by predicting other two-band reactivities in RIBA-3, except those with NS5, which could be explained by the prior selection of the majority of the population through anti-hcv screening tests without NS5. Those findings suggest that two-antigen patterns in RIBA-3 may be due, at least in part, to combined nonspecific reactivity to two antigens. To examine this possibility further, we compared the frequency of two-band RIBA-3 positive reactions in individuals with and without detectable HCV RNA (Table 5). Two-antigen reactivity was observed more often in HCV RNA-negative subjects than in HCV RNA-positive subjects (32.7% vs. 8.2%, po.0001). The proportion of c33 and cloo and c22 and cloo reactivity was especially high in HCV RNA-negative persons (p<0.001). Table 5. Frequency of positive RIBA-3 reactivity in anti-hcv ELISA-positive subjects with and without detectable HCV RNA. RIBA-3 positive (n=: 274) PCR positive PCR negative NS5 C22 C33 C100 (n=2 19) (n=55) %* 40.0 %' % 23.6 % % 7.3 % % 1.8 % % % % 0 % % 1.8 % %' 9.1 %« % 1.8 % % 0 % % 0 % p = ' p = 0.001» p = Risk factors in blood donors with RIBA-3 two-band reactivity In 1994, we examined risk factors for HCV infection in 11 blood donors with two-band reactivity in RIBA-3 and 6 blood donors with three- or four-band reactivity. The respective two-band reactivities were NS5 and c33 (n=5), NS5 and cloo (n=3), c33 and cloo (n=l), and c22 and c33 (n=2). The two-band reactivity disappeared in the follow-up sample in 8 of 11 donors. In all 11 cases, the HCV RNA PCR was negative, liver transaminase values were normal, and no signs of liver disease were present. In 2 (18%) of 11 donors with two-band reactivity, a parenteral risk factor for HCV infection was present (one tattoo 40 years before 59

9 and one blood transfusion 8 years before). Of the 6 donors with three- or four-band RIBA-3 reactivities, 5 were positive in the HCV RNA PCR. In addition, 5 (83%) of these 6 had a parenteral risk factor for HCV infection (intravenous drug abuse, 4 donors; blood transfusion, 1 donor). The difference in the presence of risk factors in blood donors with two-band reactivity in RIBA-3 and those with three- or four-band reactivity is significant (p=0.03). DISCUSSION Our study shows that replacement of RIBA-2 by RIBA-3 significantly reduced the number of indeterminate results among HCV RNA-positive subjects. Similar results have been observed by other investigators The increase in sensitivity from 93.3 percent in RIBA-2 to 99.5 percent in RIBA-3 was the result of the higher sensitivity of the c33 and cloo antigens in RIBA-3. Introduction of the NS5 antigen did not increase the sensitivity of RIBA-3, as other studies corroborate The limited sensitivity of the NS5 antigen is also demonstrated in seroconversion studies, in which anti-c33 and anti-c22 usually were the first detectable antibodies. 614 ' 9 In none of the HCV RNA-positive subjects in our study was the presence of anti-ns5 essential for a positive result in RIBA-3. Nonetheless, addition of the NS5 antigen resulted in a third or fourth reactive antigen in 68.5 percent of the viremic subjects. In our study population, introduction of the NS5 antigen produced a high proportion (9.6%) of isolated reactivities in HCV RNA-negative subjects. This proportion has increased further in our laboratory because third-generation anti-hcv ELISAs, containing the NS5 antigen, are now used more widely in the Netherlands (results not shown). The change from recombinant antigens in RIBA-2 to synthetic peptides in RIBA-3 led to a reduction in isolated c22 and c 100 reactivity. This result may also be influenced by the continuous removal from the donor population of blood donors with indeterminate reactivities. In spite of this, we found that the modified c33 antigen in RIBA-3 led to a significant increase of PCR-negative samples showing indeterminate c33 reactivity. In our study, we found a significantly higher proportion of two-band reactivity among HCV RNA-negative subjects than among HCV RNA-positive subjects. From our results (Table 4), we can deduce that, in subjects with three- or four-band reactivity in RIBA-3, the chance of being HCV RNA-positive is 84.4 percent, whereas in persons with two-band reactivity, it is 50 percent (po.0001). If we leave the c22 and c33 pattern out of consideration, this chance is even lower (30%). These results can be explained by a higher prevalence of resolved HCV infections among the limited number of subjects with two-band reactivity. However, a substantial number of the two-band reactivities, except perhaps that for c22 and c33, might be nonspecific. Among a limited number of blood donors with two-band reactivity, risk factors for HCV infection were present in only 18 percent of the cases, whereas in blood donors with three- or four-band reactivity, risk factors were present in 83 percent (p=0.03). These results demonstrate that two-band reactivities in RIBA-3 should be interpreted with caution. We recommend that these persons be tested for HCV RNA by PCR to prevent incorrect counseling. The quality of the next generation of RIB As may improve if a sensitive recombinant envelop-2 (E-2) antigen could be incorporated. A preliminary study of recombinant envelope-2 protein indicates that this antigen has a high sensitivity and specificity. 20 Application of a confirmatory immunoblot in which all antigenic domains of HCV are incorporated, in combination with the use of more stringent interpretation criteria for a positive result, should further increase the reliability of the assay. 60

10 REFERENCES 1 Choo QL, Kuo G, Weiner AI, Overby LR, Bradley DW, Houghton M. Isolation of a cdna clone from a blood-borne non-a, non-b viral hepatitis genome. Science 1989; 244: Bresters D, Cuypers HTM, Reesink HW, Schaasberg WP, Poel CL vd, Mauser-Bunschoten EP, et al. Enhanced sensitivity of a second generation ELISA for antibody to hepatitis C virus. Vox Sang 1992;62: Bresters D, Zaaijer HL, Cuypers HTM, Reesink HW, Winkel IN, Exel Oehlers PJ v, et al. Recombinant immunoblot assay reaction patterns and hepatitis C virus RNA in blood donors and non-a, non-b hepatitis patients. Transfusion 1993;33: Zaaijer HL, Cuypers HTM, Reesink HW, Lelie PN. Should the cloo antigen be removed from HCV antibody assays? Vox Sang 1994;66: Busch MP, Tobler L, Quan S, Wilber JC, Johnson P, Polito A, et al. A pattern of and c only on hepatitis C virus (HCV) recombinant immunoblot assay does not reflect HCV infection in blood donors. Transfusion 1993;33: Zaaijer HL, Vrielink H, Exel-Oehlers PJ van, Cuypers HTM, Lelie PN. Confirmation of hepatitis C infection: a comparison of 5 immunoblot assays. Transfusion 1994; 34: Zaaijer HL, Cuypers HTM, Reesink HW, Winkel IN, Gerken G, Lelie PN. Reliability of polymerase chain reaction for detection of hepatitis C virus. Lancet 1993;341: Cuypers HTM, Bresters D, Winkel IN, Reesink HW, Weiner AJ, Houghton M, et al. Storage conditions of blood samples and primer selection affect the yield of cdna-polymerase chain reaction products ofhepatitis C virus. J Clin Microbiol 1992; 30: Zaaijer HL, Cuypers HTM, Reesink HW, Lelie PN, Vrielink H, Poel CL van der. New immunoblot resolves indeterminate results for antibody to hepatitis C virus. Transfusion 1994;34: Garcia-Samaniego J, Enriguez A. Soriano V Gutierrez M, Baquero M, Munoz F. Third-generation recombinant immunoblot assay to confirm hepatitis C virus-indeterminate serological samples. Vox Sang 1993;64: Leon P, Lopez JA, Domingo CJ, Elola C, Echevarria JM. Serological studies and hepatitis C virus RNA in serum samples from blood donors with indeterminate results in second generation recombinant immunoblot assay. Vox Sang 1994;66: Garcia-Samaniego J, Soriano V, Silva E, Enriquez A, Munoz F, Gonzales-Lahoz J, et al. Significance of HCV RIBA-2 indeterminate results in high-risk individuals: assessment by a new third-generation RIBA assay and PCR. Vox Sang 1994;66: Tobler LH, Busch MP, Wilber J, Dinello R, Quan S, Polito A, et al. Evaluation of indeterminate c22-3 reactivity in volunteer blood donors. Transfusion 1994;34: Uyttendaele S, Claeys H, Mertens W, Verhaert H, Vermylen C. Evaluation of third generation screening and confirmatory assays for HCV antibodies. Vox Sang 1994;66: Dow BC, Follett EAC, Jordan T, McOmish F, Davidson J, Gillon J, et al. Testing of blood donations for hepatitis C virus. Lancet 1994;343: Vernelen K, Claeys H, Verhaert H, Volckaerts A, Vermylen C. Significance of NS3 and NS5 antigens in screening for HCV antibody. Lancet 1994;343: Couroucé AM, Bouchardeau F, Girault A, Marrec N Le. Significance of NS3 and NS5 antigens in screening for HCV antibody. Lancet 1994;343: Goffin E, Pirson Y, Cornu C, Jadoul M, Ypersele de Strihou C van. Significance of NS3 and NS5 antigens in screening for HCV antibody. Lancet 1994;343: Couroucé AM, Le Marrec N, Girault A, Ducamp S, Simon N. Anti-hepatitis C virus (anti-hcv) seroconversion in patients undergoing hemodialysis: comparison of second- and third generation anti-hcv assays. Transfusion 1994: 34: Zaaijer HL, Vallari DS, Cunningham M, Lesniewski R, Reesink HW, Poel CL vd, et al. E2 and NS5: New antigens for detection ofhepatitis C virus antibodies. J Med Virol 1994; 44:

11 62