Pre-publication manuscript AIDS 1996, 10:

Transcription

1 Running head: Highly sensitive HIV-1 antigen test Heat-Mediated Immune Complex Dissociation and ELISA Signal Amplification Render Antigen p24 Detection in Plasma as Sensitive as HIV-1 RNA Detection by Polymerase Chain Reaction Jörg Schüpbach, Markus Flepp *, Dora Pontelli, Zuzana Tomasik, Ruedi Lüthy * and Jürg Böni From the Swiss National Center for Retroviruses, University of Zurich, and the * Division of Infectious Diseases, Department of Medicine, Zurich University Hospital, Zurich, Switzerland. Financial support: Swiss Federal Office of Public Health and Swiss National Science Foundation. Antigen test kits for the prospective part of this study were provided free of charge by Du Pont de Nemours GmbH, Bad Homburg, Germany. Requests for reprints or correspondance to: Jörg Schüpbach, M.D., Swiss National Center for Retroviruses, University of Zurich, Gloriastrasse 30, CH-8028 Zurich, Switzerland Tel. (+41) Fax. (+41) Informed consent was obtained from patients and the human experimentation guidelines of the Swiss Academy of Medical Sciences were followed in the conduct of this research. 1

2 Abstract Objective: To compare heat denaturation and acidification as immune complex dissociation (ICD) methods in adult HIV-1 infection and to increase the sensitivity by a signalampflication boosted HIV-1 p24 antigen ELISA which detects 0.5 pg/ml. Design: Prospective and retrospective blinded study of paired serum and plasma samples from 245 patients (112 of category A, 66 of B, 67 of C of the CDC 93 classification). Methods: Plasma and sera were prospectively tested for antigen with Du Pont's HIV-1 Core Profile ELISA using native, acidified, or heat-denatured material. Retrospective tests included batch analysis of heat-denatured samples for antigen by the signal-amplification boosted ELISA and for viral RNA by the Roche HIV Monitor. Results: With serum, native antigen was reactive in 26.5%. Acidification increased the rate to 53.1% (p.0001), but was inefficient at a CD4 + count 500/µl. Heat denaturation further elevated the rate to 67.8% (p.0007) and the use of plasma to 78.0% (p.008). The boosted ELISA, performed with plasma samples diluted 1:6, which eliminated the problem of heatinduced sample coagulation, was confirmed positive in 89.5% of serum and 97.8% of plasma samples. RNA was detected in 95.7%. Conclusion: Heat-mediated ICD combined with a signal-amplification boosted antigen ELISA detects HIV-1 expression as sensitively as a PCR kit for viral RNA, but at only a fraction of expenditure. The procedure uses a 50 µl plasma sample and should be fully automatable. Keywords: antigen p24, detection and quantification of virus expression, viral RNA load, adult HIV-1 infection 2

3 Introduction Antigen (Ag) p24 testing is widely used for the detection and quantification of HIV-1. In the presence of excess anti-p24 antibody most Ag is, however, hidden in immune complexes [1-3]. Immune complex dissociation (ICD) based on acidification [3-10] or on treatment with base [11] can improve Ag detection. The release of small Ag quantities from immune complexes is, however, inefficient if antibodies to p24 are present at very high titers. By contrast, boiling diluted serum or plasma for five minutes leads to the quantitative and irreversible release of immune-complexed HIV-1 p24 that can be readily measured with certain commercial assays that recognize heat-denatured antigen (HD-Ag) [12]. Application of the procedure to pediatric HIV-1 infection demonstrated its excellent specificity and a diagnostic sensitivity of 96%, which were both comparable to the detection of viral DNA by polymerase chain reaction (PCR). The quantification of HD-Ag in the first six months of life provided in addition an early opportunity to assess an infected child's prognosis [13]. We have also found that the sensitivity of the procedure could be further improved by combining the HD- Ag procedure with an ELISA signal amplification step. Application of this procedure to the testing of pediatric samples routinely referred to our laboratory resulted in a prospective sensitivity that was identical to that of DNA PCR, namely 38% at birth (n=13) and 100% in older children (n=62) [14 and in preparation]. The excellent results of heat denaturation with pediatric samples have meanwhile been fully confirmed and extended to African samples by others [15]. The present study was undertaken to assess the efficacy of ICD by heat and acidification in samples from HIV-1 infected adults, to compare the use of serum and plasma as a sample, to assess the problem of heat-induced sample coagulation and to find ways to overcome 3 it. The usefulness of a signalamplification boosted antigen ELISA of increased sensitivity in eliminating this problem was evaluated and an optimized procedure combining heat denaturation of plasma at a higher dilution and testing by the amplification-boosted ELISA was compared with a commercial PCR kit for the quantification of viral RNA. Materials and Methods: Patients and samples. Blood samples were collected from 245 adult HIV-1 infected individuals seen at the Division of Infectious Diseases, Department of Medicine of the Zurich University Hospital. All individuals participated in the prospective Swiss HIV Cohort Study [16] and were classifyed according to the 1993 classification of the Centers for Disease Control and Prevention [17]. From each patient, an EDTA-anticoagulated plasma and a serum sample were collected after obtaining informed consent. All samples were coded before they were brought to the testing site. All testing was performed blinded with respect to the patient's identity, stage of infection, or any other clinical or laboratory information. Antigen testing. All tests were based on the HIV-1 p24 core profile ELISA (Du Pont de Nemours, Bad Nauheim, Germany). The study involved a prospective branch, in which all testing was done in duplicate, and a retrospective branch involving single determinations. Prospective assays included testing the native undenatured serum (UD-Ag), testing after ICD by acidification with the Du Pont HIV-1 p24 Core Profile ADD (Acid Disruption Difference) kit (ADD- Ag), and testing after heat denaturation (HD-Ag). Assays for UD-Ag and ADD- Ag strictly followed the manufacturer's instructions, with the only exception that reactive samples were not confirmed with an additional neutralization assay. HD-Ag testing of serum or plasma samples was performed exactly as published [13]. All samples that were reactive, using a cut-

4 off corresponding to the mean of a negative triplicate control run on all plates plus three fixed standard deviations (established previously with 390 HIVnegatives [13]), were subjected to a confirmatory neutralization assay. Neutralization was performed with heatdenatured plasma diluted 1:3, using Du Pont's HIV-1 confirmatory assay reagents and procedure in combination with Du Pont's ELAST TM ELISA Amplification System which, in our hands, provides a sensitivity of 0.5 pg/ml [14]. In brief, the regular HD-Ag assay procedure was followed through the streptavidin horse radish peroxidase (HRP) reporter incubation and the 10-cycle washing step. Subsequently, the wells were incubated for 15 min at room temperature with 100 µl of the biotinyl-tyramide working solution, according to the manufacturer's instructions. Biotinyl-tyramide, whose phenolic group is rendered reactive through the enzymatic activity of the already bound streptavidin-hrp, binds covalently to electron-rich moieties on the solid phase, thus providing additional binding sites for streptavidin-hrp [18]. The wells were then washed in another 10- cycle step and incubated for 15 min with 100 µl of streptavidin-hrp diluted 1:3,000 in the PBS-BSA-Tween 20 buffer recommended by the manufacturer. After a final 10-cycle washing step, the customary ELISA substrate was added and the assay completed "as usual". In the retrospective analysis, all remaining samples were batch-screened by HD-Ag ELAST, at a sample dilution of 1:6. For reference, a total of 107 HIV-negative controls were tested in the same way. The Results Superiority of heat denaturation. In a prospective blinded study, viral p24 antigen concentrations were determined in 245 serum samples from adult HIV-1 infected patients. The distribution of these patients with respect to the CDC 93 classification is shown in Table 1. Antigen measurement was performed after no 4 cut-off of positivity for each assay plate was determined as the mean absorbance of 8 to 24 of these controls plus five standard deviations. Testing of heat-denatured samples yields considerably decreased absorbance readings for both antigenpositive and negative samples. Nevertheless, the procedure results in a net gain in sensitivity [12]. In agreement with this, the mean absorbance of the 107 controls distributed among four different experiments ranged from to and the standard deviations ranged from to The resulting cut-offs ranged from to and were thus much lower than those usually observed with other types of ELISA. Reactive samples with very low Ag p24 concentrations ( 2.0 pg/ml) were confirmed by neutralization, using HD-Ag ELAST at a sample dilution of 1:3. This resulted in a higher absorbance, so that neutralization effects could be assessed more safely. PCR for viral RNA. The HIV-1 RNA load was determined retrospectively in plasma samples that had been frozen at -70 C not later than 4 hours after drawing the sample and had never been thawed before testing. The Amplicor HIV Monitor (Roche Molecular Systems, Basel) was used and the manufacturer's instructions were strictly followed. Statistical evaluations. Frequencies were compared with 2x2 table tests. Quantitative distributions were compared by paired or unpaired t-test using the StatView 4.02 program for Macintosh (Abacus Concepts, Berkeley, California). treatment or after acid- or heat-mediated immune complex dissociation. The results of these procedures are shown in Table 1. HD-Ag was significantly superior to ADD-Ag (p.0007). ADD-Ag in turn was significantly superior to UD-Ag (p.0001); this effect was, however, restricted to patients with CD4 + cells <500/µl.

5 The superiority of HD-Ag was predominantly due to a higher detection of antigen in the clinical categories A and B. In particular, HD-Ag was significantly more effective than ADD-Ag in patients with CD4 500/µl. Results of the two methods did not significantly differ for category C, and for the patients with a CD4 + cell count of /µl the higher detection rate of HD-Ag just missed significance. All three antigen detection methods showed a clear increase of reactivity in correlation with CD4 + cell decrease, but the situation was more complicated when the CD4 + cell decrease within a given clinical category was considered. For example, there was no increase in reactivity by UD-Ag and ADD- Ag from A2 to A3 and none by HD-Ag from B2 to B3. There was a steady increase by UD-Ag and ADD-Ag with respect to clinical progression, while with HD-Ag the percentage even fell from category B to C. Confirmation of HD-Ag reactive samples by neutralization with plasma and HD-Ag ELAST (see Methods) was positive in 96.2 %. Negative neutralization was restricted to 6 samples with low absorbance/cut-off (A/C) ratios (all 3.3). Superiority of testing plasma instead of serum. Testing of plasma instead of serum for HD-Ag yielded 191 (78%) reactives, which was significantly higher than the 67.8% obtained with serum (Table 1). The difference was attributable to improved detection in all categories, although it was only significant in those with the highest sample numbers, i.e. category A and CD4 + cells 200/µl. Sample coagulation. A important practical problem of ICD procedures is coagulation of the sample, which occurs in both acidand heat-mediated ICD. However, it is in almost all cases possible to transfer the heat-coagulated material to the assay plate and to continue the test. In contrast to previous findings based on a limited number of materials [12], in the present study heat-induced coagulation occurred less frequently with plasma than with serum (30.6% vs. 68%; p<.0001). Coagulation had, however, no influence upon the quality of the test result: positives were similarly frequent among coagulating and non-coagulating samples (75% vs. 79%, not significant, Chi-square=0.4). Nevertheless, since coagulation renders manual testing cumbersome and presents an insurmountable obstacle to automated testing, we evaluated whether the problem could be solved with a higher sample dilution. This required, however, that a detection procedure of increased sensitivity was used in order to compensate for the dilution effect. 5

6 A simple solution to this problem was found in a commercially available signal amplification system, ELAST TM, with which the HD-Ag method was coupled (see Methods). The increased sensitivity of this procedure permitted the assay to be run at a sample dilution of 1:6, which completely prevented coagulation. After heating, all samples could be easily transferred to the assay plate and after the first incubation step they were without any difficulty removed from the plate by an automated plate washer. HD-Ag ELAST is as sensitive as PCR for viral RNA. For evaluation of the sensitivity of the HD-Ag ELAST, all remaining plasma samples were systematically tested with this method. Serum was used, if plasma was no longer available. A total of 239 samples including 182 plasma and 57 serum samples were tested. Of these, 117 were randomized for additional testing for viral RNA in plasma, using aliquots that had been frozen at 70 C within 4 hours after withdrawal and had never been thawed. submitted to confirmatory neutralization at a dilution of 1:3. Of 19 such samples, 17 (89.5%) were confirmed. Thus, a total of 178 plasma samples (97.8%) were antigenpositive (Table 2). The two non-reactive plasma samples were both from individuals of category A1, one with a CD4 + count of 1090/µl and 755 RNA copies/ml and the other with 810 CD4 + cells/µl; material for RNA determination was not available for this patient. One of the two reactive plasma samples negative by confirmation was from an A3 patient with a CD4 + cell count of 310/µl and 2781 viral RNA copies/ml, while the other was stage B2 with a CD4 + cell count of 240/µl and RNA not measured. Sensitivity of HD-Ag ELAST in plasma was lowest (85.7%) in the 14 patients with a CD4 + count 500/µl. By contrast, all but two of the 168 patients with counts below 500/µl were positive (98.8%). Serum samples had again a significantly lower sensitivity of 51/57 (89.5%; p.01) Plasma Serum HD-Ag ELAST absorbance/cut-off below detection limit RNA copies/ml Fig. 1 Comparison of viral RNA determination by polymerase chain reaction and antigen p24 determination by heat-denatured antigen ELAST (HDAg ELAST). Plasma and serum samples are marked differently. Screening of the 182 plasma samples by HD-Ag ELAST yielded a reactive result in 180 specimens (98.9%). All reactive samples with 2.0 pg/ml antigen that were still available at sufficient quantity were PCR for viral RNA among 117 samples randomized from all stages of infection was positive in 95.7%. There were no differences between the positivity rates of HD-Ag ELAST and PCR, although the latter was positive with all of 6 samples with CD4 + cells 500/µl. This possible advantage was, however, far from being significant and fully compensated by some 6

7 negative results in the more advanced stages of infection. In order to broaden the comparison for samples from the earlier stages of infection, for which an increased sensitivity of PCR might be expected, all remaining samples from patients with CD4 + cells 300/µl were tested for RNA, which increased the total to 147. Analysis of these samples yielded 141 (95.9%) positives by PCR and 140 (95.2%) reactives by HD-Ag ELAST (Fig. 1). Six of the 7 samples (86%) that were negative by HD-Ag ELAST were sera, although the sera's share of the total was only 27%. After subtraction of a reactive, but nonneutralizable sample, 105 (98.1%) of the 107 plasma samples were positive by HD- Ag ELAST, but only 34 (85%) of the 40 serum samples (p.005). Serum samples exhibited significantly lower antigen absorbance/cut-off rates than plasma samples (p<.0001), as demonstrated in Fig. 2. Thus, the results of HD-Ag testing of these two materials were confirmed. reactivity (absorbance/cut-off) UD-Ag serum n=245 ADD- Ag serum n=245 HD-Ag serum n=245 HD-Ag plasma n=245 HD-Ag ELAST serum n=57 HD-Ag ELAST plasma n=182 HD-Ag ELAST negative controls n=107 Fig. 2 Overview on the effects achieved by the various measures used to improve antigen detection. The box plot rendition of the reactivity of each sample is a percentile-based analysis, in which the five horizontal lines represent, from bottom to top, the 10th, 25th, 50th, 75th and respectively 90th percentile and outrunners are plotted individually. UD-Ag, undenatured antigen; ADD-Ag, acid-disruption difference antigen; HD-Ag, heat-denatured antigen; HD-Ag ELAST, heat-denatured antigen combined with the ELAST enzyme-linked immunosorbent assay amplification system. Fig. 2 also summarizes and quantitates the improvements achieved with the different antigen detection procedures. From the testing of native serum by the regular assay to the testing of plasma samples by HD-Ag ELAST the median absorbance/cut-off ratios were increased by about two orders of magnitude. Discussion Our analysis of 245 paired serum and plasma samples from adult patients at different stages of HIV-1 infection significantly demonstrates that heat denaturation is a more efficient method for immune complex dissociation than acidification, which is in turn superior to the testing of native samples (Table 1, Fig. 2). The finding that plasma yielded higher rates of Ag-positives (Table 1), as well as higher absorbance/cut-off values (Fig. 2) and antigen concentrations (not shown) than serum clearly indicates that the sample of choice for antigen detection is plasma, and not serum. A previous study of plasma samples from pediatric HIV-infection demonstrated a sensitivity of 96% for HD-Ag, which was equal to that of PCR for viral DNA [14]. The sensitivity of HD-Ag with samples from the mothers of these children was, however, significantly lower. Application of the identical procedure to the testing of adults in the present study yielded again a markedly lower rate of positives (78% with plasma), thus confirming that the antigen concentrations in adult HIV-1 infection are lower than in pediatric infection. The present study shows that in adult infection heat denaturation in combination with a conventional antigen test is inferior to PCR. However, in combination with the signal amplification system, which lowered the detection limit to about 0.5 pg/ml, the same diagnostic sensitivity as with a commercial kit for HIV RNA quantification was achieved (Table 2, Fig. 1). At the same time, the sample dilution was increased to a degree, at which heat coagulation presented no longer a problem. The present procedure should thus be 7

8 usable for fully automated testing. The higher dilution may in some cases even contribute to the increase in sensitivity, since a prozone effect observed with certain samples at a dilution of 1:3 may thereby be eliminated [15]. Virus expression, responsible for the pathogenicity of HIV infection, can be assessed in two ways, on the level of RNA and protein. Most virus is produced in the lymphatic organs, where it can not be routinely assessed [19, 20]. Therefore, the amount of viral RNA in plasma has become an increasingly used surrogate marker for the total viral load [21-27]. At the same time, quantification of p24 as another possible virus load marker has become less attractive, due to a clearly inferior sensitivity of the currently used acidification-based ICD procedures [28]. Nevertheless, antigen testing has important practical advantages compared with tests for viral RNA. The antigen is considerably more stable and less affected by variation in time and the physical conditions during the transport to the laboratory. The sample preparation is much simpler and the test can be performed with equipment already available in most laboratories, at only a fraction of the time, work and cost of a molecular test. In addition, the sensitivity of antigen tests may be less affected by genomic diversity than that of a sequencebased molecular tests. For example, an HIV-1 subtype O isolate in culture supernatant was detectable by antigen test (though less efficiently than mainstream isolates), but PCR with conventional primers failed completely [29]. All this considered, we expect that the highly sensitive procedure presented here should renew the interest in antigen testing as an easy and affordable method for detecting and quantifying virus expression. It is clear that tests for viral antigen and for RNA detect different things. The viral RNA found in plasma is of necessity associated with a particle that offers protection against degradation by RNase. In contrast, viral antigen may or may not be particle-associated; the free or immune complex bound forms may originate from destroyed virus particles, destroyed infected cells, or cells that produce and release viral proteins independent of particles. Extensive further studies are thus needed in order to assess the role and value of this improved antigen detection procedure in the management of HIV infection. Acknowledgements: The authors wish to thank Soksimon Kaing for excellent technical assistance and Mike Winiger, Liliane Clausen, Monika Hürlimann and Simona Malnatti for collecting the blood specimens. We are also indebted to Dr. Richard Cone for critical reading of the manuscript. References: 1. de Wolf F, Goudsmit J, Paul DA et al.: Risk of AIDS-related complex and AIDS in homosexual men with persistent HIV antigenaemia. Br Med J 1987, 295: Pedersen C, Moller-Nielsen C, Vestergaard BF, Gerstoft J, Korgsgaard K, Nielsen JQ: Temporal relation of anigenaemia and loss of antibodies to core antigens to development of clinical disease in HIV infection. Br Med J 1987, 295: Lange JM, Paul DA, de Wolf F, Coutinho RA, Goudsmit J: Viral gene expression, antibody production and immune complex formation in human immunodeficiency virus infection. AIDS 1987, 1: Kageyama S, Yamada O, Mohammad SS et al.: An improved method for the detection of HIV antigen in the blood of carriers. J Virol Methods 1988, 22: Mathiesen T, Sundqvist VA, Albert J, Ohlsson E, Wahren B: Acid-hydrolysis of serum samples to increase detection of HIV antigen. J Virol Methods 1988, 22:

9 6. Von Sydow M, Gaines H, Sönnerborg A, Forsgren M, Pherson PO, Strannegard O: Antigen detection in primary HIV infection. Br Med J 1988, 296: Kestens L, Hoofd G, Gigase PL, Deleys R, van der Groen G: HIV antigen detection in circulating immune complexes. J Virol Methods 1991, 31: Nishanian P, Huskins KR, Stehn S, Detels R, Fahey JL: A simple method for improved assay demonstrates that HIV p24 antigen is present as immune complexes in most sera from HIV-infected individuals. J Inf Dis 1990, 162, Ascher D, Roberts C, Fowler A: Acidification modified p24 antigen capture assay in HIV seropositives. J AIDS 1992, 5: Miles SA, Balden E, Magpantay L, et al.: Rapid serologic testing with immune-complex-dissociated HIV p24 antigen for early detection of HIV infection in neonates. N Engl J Med 1993, 328: Hyman J, Lockwood D, Holody T, Youngbar P: Base dissociation assay (BDA), a simplified method for immune-complex dissociation and detection of HIV p24. IX International Conference on AIDS/IV STD World Congress. Berlin, June 1993 [abstract PO-A ]. 12. Schüpbach J, Böni J: Quantitative and sensitive detection of immune-complexed and free HIV antigen after boiling of serum. J Virol Methods 1993, 43: Schüpbach J, Böni J, Tomasik Z et al.: Sensitive detection and early prognostic significance of antigen p24 in heat-denatured plasma of human immunodeficiency virus type 1-infected infants. J Inf Dis 1994, 170: Schüpbach J, Tomasik Z, Böni J, Kind C, and the Swiss Neonatal HIV Study Group: Sensitive, reliable and cost-efficient diagnosis of pediatric HIV-1 infection by antigen p24 testing of plasma after heatmediated immune complex dissociation. FEMS Symposium on Recent Advances in the Diagnosis of Viral Diseases, Istanbul, July 1995 [abstract OP11]. 15. Lyamuya E, Bredberg-Raden U, Massawe A et al.: Performance of a modified HIV-1 p24 antigen assay for early diagnosis of HIV-1 infection in infants and prediction of mother to infant transmission of HIV-1 in Dar Es Salaam, Tanzania. J AIDS, 1996, in press. 16. von Overbeck J, Egger M, Smith GD et al.: Survival in HIV infection: do sex and category of transmission matter? AIDS 1994, 8: Centers for Disease Control and Prevention: 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR Morb Mortal Wkly Rep 1992,41: Bobrow MN, Harris TD, Shaughnessy KJ, Litt GJ: Catalyzed reporter deposition, a novel method of signal amplification. Application to immunoassays. J Immunol Methods 1989, 125: Pantaleo G, Graziosi C, Demarest JF et al.: HIV infection is active and progressive in lymphoid tissue during the clinically latent stage of disease. Nature 1993, 362: Embretson J, Zupancic M, Ribas JL et al.: Massive covert infection of helper T lymphocytes and macrophages by HIV during the incubation period of AIDS. Nature 1993, 362: Piatak M Jr, Saag MS, Yang LC et al.: High levels of HIV-1 in plasma during all stages of infection determined by competitive PCR. Science 1993, 259: Wei X, Ghosh KG, Taylor ME et al.: Viral dynamics in human immunodeviciency virus type 1 infection. Nature 1995, 373: Ho DD, Neumann AU, Perelson AS, Chen W, Leonard JM, Markowitz M: Rapid turnover of plasma virions and CD4 lymphocytes in HIV-1 infection. Nature 1995, 373: Cao Y, Ho DD, Todd J et al.: Clinical evaluation of branched DNA signal amplification for quantifying HIV type 1 in human plasma. AIDS Res Hum Retroviruses 1995, 11: Henrard DR, Daar E, Farzadegan H et al.: Virologic and immunologic characterization of symptomatic and asymptomatic primary HIV-1 infection. J Acquir Immune Defic Syndr Hum Retrovirol 1995, 9: Mellors JW, Kingsley LA, Rinaldo CR et al.: Quantitation of HIV-1 RNA in plasma predicts outcome after seroconversion. Ann 9

10 10