Jul
8
Written by:
Emily Sherman
7/8/2010 8:06 AM
If you have HIV, researchers need your help. Donate plasma and receive $400/donation for your time. Visit www.idonateplasma.com for details.
|
Rapid confirmation of HIV infection
Niel T. Constantine(1) and Fassil Ketema(1)
Objective: To investigate the utility of a prototype rapid confirmatory HIV assay which offers specific results in 5 min and has applications in a variety of important testing situations.
Methods: The performance of the rapid confirmatory assay was assessed with 849 blood samples, including serum, plasma, venipuncture whole blood, and peripheral blood collected via fingerstick. Included were over 700 HIV Western blot (WB)-confirmed antibody positive sera, and others which were classified as negative or indeterminate by WB.The analytic sensitivity of the rapid confirmatory assay was assessed using 13 HIV-1 seroconversion panels, and all results were compared to those of an FDA-licensed WB reference test.
Results: The rapid test exhibited 100% concordance with the reference test when testing HIV-1 WB-confirmed positive samples, and 92.3% concordance with samples having WB-inconclusive results.The sensitivity for confirming recent seroconversion was as good, or better than, the FDA-licensed HIV-1 WB in 10/13 panels. The rapid assay performed accurately with whole blood collected from fingerstick, and exhibited excellent precision and reproducibility.
Conclusion:We conclude that this rapid HIV confirmatory assay, the first of its kind, demonstrates proof of principle for the accurate confirmation of HIV-1 infection and offers important advantages in public health and clinical testing venues.
INTRODUCTION
Since their introduction in the late 1980s, rapid HIV assays have gained respect because of their continual
improvement and attractive features, as exemplified by the FDA licensure of a rapid HIV test in the USA1 and a number of approved rapid HIV tests in Europe.2 In addition to their use in laboratories which possess less than optimal infrastructure (e.g. unstable electricity), they have found applications as components of alternative HIV confirmatory strategies,3 in occupational exposure cases,4 and for women in labor,5 where results can be provided in a clinically relevant time frame to decrease transmission. More recently, rapid tests have been shown to offer important advantages in US public health clinics, where a large number of clients do not return for results.6 In such clinics, over 700 000 individuals, including over 1000 HIV-infected person who did not return to receive their results, would have been informed and counseled had rapid tests been used. This has important implications for providing intervention strategies to decrease
transmission and for enhancing the quality of life in infected individuals. Although the use of rapid screening assays allows for immediate counseling of HIV-negative individuals, persons having reactive screening results would still be required to return for supplemental testing. Also, persons are inappropriately given anti-HIV therapy based on reactive results from screening tests of source patients in occupational exposure cases and from women in labor, because results are later found to be false positives (N. T. Constantine, unpublished observations). The immediate availability of confirmatory results would provide for more effective testing, counseling, and treatment.
As with all HIV serologic screening strategies, rapid HIV assays produce a certain proportion of false-positive results. Recent reports have suggested that substantial numbers of false-positive results could occur in largescale testing situations,6,7 and that these tests may have a poor positive predictive value when testing pregnant women.8 Consequently, the use of HIV confirmatory assays such as Western blot (WB) or the indirect fluorescent assay in testing algorithms is considered the standard of care for diagnosing true infection and for the re-entry of ELISA-reactive individuals into the blood donor pool. However, current confirmatory assays are expensive, cumbersome to perform, require skilled laboratory staff to perform and correctly interpret results, and do not allow for rapid turnaround time for results.
Advances in the performance characteristics of rapid tests, improvement in their accuracy, and the availability of multiple recombinant and synthetic peptide antigens, impart rapid assays with a potential to simulate WB assays, line immunoassays (LIAs),9 or recombinant immunoblot assays (RIBAs),10 which incorporate separated antigens or distinct artificially produced antigens to identify specific antibodies to HIV.Accordingly, multiple specific antigens can be applied to membranes in a rapid test format, allowing for the differentiation of antibody reactivity. The three antigens of diagnostic significance in a WB are p24, gp41, and gp120/160.11 We report on an assessment of a rapid HIV confirmatory assay which incorporates separate HIV antigens of diagnostic significance, and discuss the advantages and applications of its use.
METHODS
Samples
In total, 849 samples, including serum, plasma, heparinized whole blood, and fingerstick-derived whole blood, were used in the evaluation of the rapid test. These consisted of 718 HIV-1 WB-confirmed positives,
31 WB negatives, and 26 HIV-1 WB-indeterminate samples; all were repeatedly reactive by HIV-1/2 screening ELISAs. Twenty of the 718 HIV-1 WB-confirmed positive samples met the criteria for positivity but did not react to all viral antigens.The remaining 698 serum/ plasma/whole blood samples exhibited reactivity to all viral antigens, i.e. p18, p24, p32, gp41, p51, p55, p65, gp120, and gp160. Fingerstick whole-blood samples (n_20) were obtained from clinically confirmed HIVpositive (n_10) and HIV-uninfected individuals (n_10). In addition, 65 of the 849 samples had venipuncture whole-blood pairs. The analytic sensitivity of the rapid test was assessed using 74 samples from 13 seroconversion panels (Boston Biomedica, Inc., Bridgewater, MA, USA; BioClinical Partners, Inc., Franklin, MA, USA) and included the following well-characterized seroconversion panels: PRB-904, PRB-916, PRB-924, PRB-929, PRB-931, PRB-935, PRB-936, PRB-937, PRB-939, PRB-940, PRB-943, PRB-944, and donor 68106. Finally, a reproducibility investigation was conducted using three samples tested in replicates of five over a
4–day period, and 99 samples which were tested at two different time points.
Serum and plasma samples were obtained from the Clinical Immunology Laboratory at the University
of Maryland Medical System in Baltimore and from archived specimens from five international locations
(Ivory Coast, Trinidad, Thailand, Peru, and India). Fingerstick whole-blood samples were collected from
patients attending the AIDS Adolescent Clinic at the University of Maryland, Baltimore, and from other
volunteers after signing an IRB-approved consent form. Testing by the rapid confirmatory assay was conducted immediately at the site of collection for the fingerstick whole-blood samples. All samples were coded and unlinked to maintain confidentiality and were kept frozen at –20oC prior to testing. Table 1 lists the sample categories and numbers.
Tests and testing
The Quix HIV-1 Rapid Antibody Identification Test (Guardian Scientific, Inc., Columbia, MD, USA) is a
5–min, in vitro qualitative diagnostic assay designed to detect antibodies to individual proteins of HIV-1 in
whole-blood, serum, and plasma samples. Its intended use is to confirm the presence of antibodies to HIV-1 in specimens found to be repeatedly reactive by screening assays such as enzyme immunoassays or rapid tests. The Quix HIV-1 Rapid Antibody Identification Test incorporates one HIV recombinant protein corresponding to the core protein p24, and four synthetic peptides from the immunodominant regions of the envelope proteins gp41 and gp120. These proprietary antigens are immobilized on a test membrane as three distinct spots to allow the characterization of an individual’s immune response to the three specific viral antigens.
The Quix HIV-1 Rapid Antibody Identification Test consists of a disposable proprietary blood filter incorporated into a test cartridge. Following the addition of sample, the immobilized antigens on a porous membrane within the test device capture specific antibodies to HIV-1. After a wash step to remove nonspecific proteins, the presence of antibodies is revealed by the addition of a protein A–colloidal gold conjugate which binds to the adsorbed HIV antibodies, forming visually discernable red spots at the antigen sites on the membrane. A procedural control line, containing anti- IgG, binds nonspecific human immunoglobulins in the test sample; subsequent staining of this control line with
protein A–colloidal gold at this location indicates that the test has been performed correctly, that sample has been added, and that all reagents are working properly. The assay procedure consists of the separate addition into the test cartridge of two drops of proprietary buffer, one drop of serum, plasma, or whole-blood sample (45– 50 _L), and another two drops of the buffer, removal of the filter unit, and the addition of two drops of another proprietary buffer, two drops of protein A–colloidal gold conjugate, and two drops of the buffer.The criterion for a positive result by the test is the presence of reactivity to any two of the three viral-specific antigens p24, gp41, and gp120. Reactivity is defined as the appearance of a distinct red or pink circular spot. The absence of any reactivity to all three antigens, or the presence of any reactivity to any one of the three antigens of less intensity than to the corresponding antigens of the negative control, is considered a negative result. The presence of reactivity at only one of the three antigen sites is considered an indeterminate result. The pink or red procedural control line must be visible at a position above the three antigen sites for a result to be valid. Reference screening and confirmatory assays used for the evaluation of the Quix HIV-1Rapid Antibody Identification Test included the Genetic Systems HIV- 1/HIV-2 Peptide EIA (Redmond, WA, USA) or the Abbott HIVAB HIV-1/HIV-2 (rDNA) EIA (Abbott Park, IL, USA), the Abbott HIVAG Monoclonal p24 EIA, and the Bio-Rad NOVAPATH HIV-1 Immunoblot Assay (BioRad Laboratories, Hercules, CA, USA). All samples were tested with strict adherence to the manufacturers’s instructions. For the WB, reactivity to two of the three viral-specific antigens p24, gp41 and gp120/160 was classified as a positive result, whereas the absence of any reactivity was considered a negative result. Any sample exhibiting reactivity which did not meet the criteria for a positive or negative result was classified as indeterminate. Samples which produced discrepant results between the reference WB assay and the Quix HIV-1 Rapid Antibody Identification Test were retested in duplicate by the rapid assay, and 2/3 concordant results were considered definitive. This retesting was performed only to be confident that a technical error had not occurred which would falsely bias the true reactivity of the test being evaluated. When results were still discordant after repeat testing by the rapid test, a WB was again performed. Persistently discordant samples were further tested by the Abbott HIVAG monoclonal p24 EIA. A positive p24 antigen EIA result constituted a final classification of positive for that sample, while a negative p24 antigen result could not be used as a definitive result.
For the purpose of determining analytic sensitivity, results obtained with the Quix HIV-1 Rapid Antibody Identification Test were compared to those obtained by the providers of the seroconversion panels using the Bio-Rad NOVAPATH HIV-1 Immunoblot Assay as the reference. Qualitative results by each test for each of the bleeds were compared. Also, reactions were graded from 1_ (weak) to 4_ (strong) in order to give some insight into the intensity of reactions; _/_ reactions (possible reactivity) were noted but did not constitute a positive result for that antigen.
Precision and reproducibility of the Quix HIV-1 Rapid Antibody Identification Test were assessed using two methods: (1) three samples (strong HIV-1 positive, weak HIV-1 positive, and HIV negative) were each tested in replicates of five (precision) on 4 consecutive days (reproducibility) using the same lot of reagents and test cartridges to determine the intra-lot reproducibility of results, and (2) 99 samples were tested using two different lots of reagents and cartridges at different time periods to determine the inter-lot reproducibility of the assay.
RESULTS
Of the 775 different samples, excluding the seroconversion panels, the Quix HIV-1 Rapid Antibody Identification Test exhibited an overall qualitative concordance with WB of 99.2% (769/775). All WBconfirmed positive samples were also confirmed by the Quix rapid assay, yielding a 100% (718/718) sensitivity for the correct identification of positives. For 698 of these positives, the WB had exhibited reactivity to all viral antigens; the rapid assay showed reactivity to all three of its antigens. Of the remaining 57 samples, 31 were WB negative, of which the rapid assay correctly classified 27 samples; it produced four indeterminant results, showing a 1_ reactivity only to the p24 antigen with two sera and reactivity only to gp41 with two sera (1+ reactions). Twenty-six of the 57 samples were WB indeterminant, of which one produced a negative result and another a positive result (1_ p24 and 1_ gp41) by the rapid confirmatory assay. Of particular note, this latter sample exhibited a suspicious profile by WB; that is, it showed _/_ reactions to p32, p65, and gp160, but could not be resolved by p24 or HIV-2 testing. This resulted in a concordance of 92.3% for the rapid confirmatory assay as compared to the WB with indeterminant samples. Five of these total six discordant samples produced negative results for the presence of p24 antigen, while one had insufficient volume for testing. These samples were not tested for the presence of HIV RNA, because they were serum rather than plasma samples (as recommended), and had not been processed as required by the manufacturer of our test kits for RNA testing (plasma separation within 6 h).Table 1 shows the results for each category of samples, Table 2 illustrates the comparison of results by sample status, and Table 3 shows the reactions of the six samples which produced discordant results between the WB and the Quix HIV-1 Rapid Antibody Identification Test.
For the 13 seroconversion panels, the Quix HIV-1 Rapid Antibody Identification Test exhibited identical results to those of the Bio-Rad HIV-1 WB assay in five
panels (panels PRB 916, 929, 931, 937, and 944), and had a 1_ or greater reactivity at least one bleed earlier than the WB assay in five panels (panels PRB 904, 939, 940, 943, and donor 68106); in three panels, the rapid test produced reactivity one bleed later than did the WB assay (panels PRB 924, 935, and 936). For the panels in which discordance occurred, reactivity to the different antigens fluctuated fairly equally by the two tests for reactivity to p24, or to the envelope antigens gp41 or gp120/160; that is, there were no consistent differences by either of the tests. Similarly, when comparing quantitative reactivity to each antigen by each test near the time of seroconversion in panels where discordance was observed, results were equally variable (_/_ or 1_) for each of the tests. In two panels (68106 and PRB940), the rapid test exhibited a final positive result one or two bleeds prior to the WB, while the WB was positive prior to the rapid assay in one panel (PRB935).A comparison of reactivity of the two tests using seroconversion panels is shown in Table 4.
When assessing the ability of the Quix HIV-1 Rapid Antibody Identification Test to produce identical results with the same lot of reagents upon repeat testing, there was 100% reproducibility when testing three samples over a 4–day period, and 100% precision when testing samples in replicates of five on each of those days. Furthermore, when 99 samples were tested using two different lots of reagents, only two samples had reactivities which resulted in a change of classification between the two time periods. In one sample, the difference was produced by a _/_ reaction versus a negative result to the p24 antigen only.The other sample produced a 1+ reaction versus a negative reaction to the
p24 antigen only. These two samples, therefore, had changes in classification from indeterminate to negative; the Western blot results for these samples were indeterminate (_/_reactions to p18, p24 and p31) and negative, respectively.
Twenty whole-blood samples collected by fingerstick from 10 known HIV-positive individuals and from 10 healthy HIV-negative persons produced the expected results by the Quix HIV-1 Rapid Antibody Identification Test. Reactions were clear and similar to those produced when testing fresh or archived samples. Finally, the Quix Rapid Antibody Identification Test produced the expected and required reaction at the control line in all samples, indicating that the procedure was performed correctly and that each sample had been added.
DISCUSSION
Diagnostic tests for the detection of HIV infection have evolved from the classical ELISA and WB tests to new technologies and testing strategies that offer advantages in a variety of testing situations. Among the advances are: HIV screening and confirmatory tests using saliva,12 urine,13 and whole blood collected on filter paper,14 rapid tests which use serum, saliva, or whole blood,15–17 fourth-generation screening tests which detect HIV antibody and antigen simultaneously,18–20 the routine use of tests to detect p24 antigen21 or viral nucleic acids,22 and sensitive/less sensitive testing strategies23,24 which can be used to estimate the incidence of infection in order to target the most at-risk populations for implementation of intervention strategies. Furthermore, molecular tests were introduced which have the potential to quantify viral load,25 identify the genotype of isolates,26 and measure anti-drug resistance.27 Newer tests also incorporate additional antigens in order to detect escape variants of HIV,28 and some tests can actually identify the specific group or clade of HIV-1.29,30 Alternative testing algorithms,31 including alternative confirmatory strategies using rapid assays,3 can result in cost and time savings, thereby addressing important issues for developing countries.
The Quix HIV-1 Rapid Antibody Identification Test represents a new advance in HIV diagnostics and is the first such rapid assay developed specifically for the purpose of confirming HIV infection. Only one published
report suggests the use of a single rapid test (Genie, BioRad,CA, USA) as a confirmatory strategy,32 but this study had a sample size of only 50 positive samples, the test incorporates only gp41 of HIV-1, and the test was not specifically designed as a confirmatory test. However, we are aware of another simple, semirapid confirmatory test (Bionor HIV-1 & 2 Test, Bionor AS, Norway) which has demonstrated excellent preliminary results in our laboratory (N.T. Constantine and F. Ketema, unpublished observation).
The results of the Quix HIV-1 Rapid Antibody Identification Test were compared against those of an FDA-licensed WB assay with the objective of determining concordance for the correct identification of antibodies in serum or whole blood of truly infectedindividuals. The distribution of test samples was chosen to simulate the population normally tested by our WB algorithm; that is, all samples are ELISA positive, with the majority (93%) being WB positive, 3% being indeterminate, and 4% being negative. The test accomplished
the goal for which it was designed: the sensitivity for confirming infection was 100% with HIV-1 WBconfirmed positive samples.The Quix rapid test was also presumably capable of detecting HIV non-B clades, as was evidenced by the correct classification of 25 HIVpositivesamples from geographic areas where non-B clades predominate. The Quix rapid test showed a concordance of 92% (24/26) with the WB test when testing samples which were classified as indeterminant by WB. We could not, however, conclusively rule out these samples as being from truly infected individuals. One sample which was repeatedly reactive by the Quix rapid assay and the ELISA, and exhibited a suggestive WB profile (_/_reactions to p32, p65, and gp160), could not be resolved by the p24 antigen assay. Conversely, one sample which was classified as indeterminant by the WB, based on reactivity to a non-viral specific component near p65, was negative by the Quix rapid test and the p24 antigen assay, suggesting that this sample may have been from a noninfected person. There were four samples which were indeterminate by the Quix HIV-1 Rapid Antibody Identification Test because of reactivity to p24 or gp41, but which were classified as negative by WB. Discordant results (indeterminate versus negative) by different WB assays are not uncommon (N. T. Constantine and F. Ketema, unpublished observations), and discordant results (indeterminate versus positive) by WB can be noted when testing seroconversion panels.33 Furthermore, WB assays may be negative during early seroconversion while screening tests are reactive.33 Therefore, it is difficult to determine which test is yielding the most accurate results.Testing for p24 antigen did not provide any useful information, and RNA testing for resolution could not be performed. We did not evaluate the specificity of the Quix rapid confirmatory test using HIV-negative samples (screening test negatives), because it is well known that confirmatory assays should not be used in a screening mode. Such specificity studies can only be accomplished using ELISA-positive samples that are subsequently found to be false positives. Our evaluation used only ELISA-positive samples, which represent the same sample group that should be tested by confirmatory assays. Thus, an assessment of the utility of a new confirmatory test must be performed in comparison with reference confirmatory assays on samples targeted for confirmatory testing. It would not be appropriate to evaluate the specificity of a confirmatory assay using ELISA-negative samples.
For the purpose of looking at the concordance of the Quix HIV-1 Rapid Antibody Identification Test with the WB, a large number of seroconversion panels was used. Although there were slight differences between the two assays qualitatively and quantitatively when testing the seroconversion panels, they were comparable in their ability to detect early infection; that is, each assay exhibited a similar ability to detect early infection with several panels, a variability that is also common between different WB assays. Because the purpose of a confirmatory assay is to detect established infection, slight differences in reactivity during the seroconversion period are not unexpected or worrying. The Quix HIV- 1 Rapid Antibody Identification Test produced reactive results as early as, or earlier than, the WB in a majority of panels.
The rapid test performed excellently with fingerstick blood specimens, although our sample size was too small (n_20) to allow us to draw definitive conclusions; however, these preliminary results indicate a potential for use with this testing medium. The use of fingerstick specimens has several advantages, including cost savings based on the elimination of blood-drawing supplies, the need for phlebotomists, and the need to discard blood units, which are drawn prior to testing in many developing countries,34 as well as being applicable in facilities where venipuncture is not feasible. The ability of this rapid HIV confirmatory test to perform successfully with fingerstick specimens is not surprising, since recent reports have shown accurate results of HIV rapid screening tests when testing blood collected by fingerstick. 17
It is clear that a rapid confirmatory assay has important utility, particularly with the availability of successful therapeutic strategies, which should be initiated within 2 h of exposure, i.e. for occupational exposure4 or as soon as possible prior to delivery for the interruption of vertical transmission from women in labor.5 In these applications, it is desirable to institute therapy only in truly infected individuals; however, we have noted an almost 1% false-positive result rate with the SUDS HIV-1 rapid test in our institution, sometimes resulting in therapy that was inappropriately administered based on the rapid screening test result, because some individuals were found subsequently to be uninfected (unpublished observations). Translating this rate to the large number of occupational exposure cases and women in labor without known HIV status, it is not unreasonable to expect thousands of false-positive results and inappropriate institution of therapy. Also, rapid tests have substantial utility when used in public health clinics, where a large number of positive individuals never return for their initial screening test results6 and a large number of positive individuals who do return still require confirmatory testing. Even when a rapid HIV screening test is performed in this latter case, persons with reactive results would still be required to revisit clinics following testing of their samples by confirmatory assays, which usually require days to weeks. Furthermore, confirmatory assays are generally not used in developing countries, because they are expensive, and instrumentation often cannot be supported. Consequently, alternative confirmatory strategies have been sought and are recommended by the World Health Organization in certain testing situations.31 These costsaving strategies include the use of two rapid HIV assays, two ELISAs, or a combination of both tests
performed in tandem. For these reasons, and in the many testing facilities which use rapid assays because of the absence of stable electricity or the unavailability of instrumentation, a rapid confirmatory assay would offer more effective testing. Our evaluation of this new-generation rapid HIV test, capable of confirming HIV infection, shows proof of principle that a confirmatory assay can be successfully constructed in a rapid test configuration. This is not surprising, because the test principle and antigens are similar to the widely used LIA and RIBA tests, and because rapid HIV tests have been shown to be accurate. We conclude that the Quix HIV-1 Rapid Antibody Identification Test, a 5–min confirmatory assay for correctly classifying HIV infection, has demonstrated the ability to accurately confirm HIV-1 infection on samples which have been found to be reactive by HIV screening tests. However, further studies should be conducted to assess this assay’s specificity by testing a large number of samples which are verified to have false-positive results by screening tests. Our present study did not evaluate this rapid confirmatory test’s ability to replace HIV screening assays, but rather evaluated it as an alternative (after using a screening test) to existing confirmatory tests, which are more rigorous to perform, more expensive, and less applicable in developing countries. The Quix HIV-1 Rapid Antibody Identification Test exhibits acceptable reproducibility, is equivalent to FDA-licensed WB assays in analytic sensitivity, has the potential to use fingerstick blood, and can provide rapid turnaround time for results. Although this rapid confirmatory assay has a number of important applications, the most important uses are in public health clinics, where client return rates are less than desirable, and for occupational exposure cases or for women in labor, where the use of antiretroviral therapy may be instituted appropriately. In these instances, a rapid confirmatory assay could be used following repeatedly reactive results by a rapid screening test to increase the predictive value of a reactive result with respect to it being from a truly infected person, while yielding final results in a much shorter time than current HIV-testing algorithms that use WB assays.
Thus, after 15 years of evolution of HIV diagnostics, new tests with attractive features continue to be introduced. This addition of a rapid HIV confirmatory test offers important advantages that will assist in addressing several of the outstanding issues that are important for effective testing with appropriate counseling, and the correct institution of treatment in a clinically relevant time frame.
ACKNOWLEDGMENT
The authors wish to express their appreciation to Dr Lydia Temoshok for her editorial comments.
REFERENCES
1. Auxter S. Bringing rapid HIV tests to market more quickly. Clin Lab News 2000; 26(1):10.
2. Simon F, Ly TD, Baillou-Beaufils A, et al. Sensitivity of screening kits for anti-HIV-1 subtype-O antibodies.AIDS 1994; 8:1628–1629.
3. Stetler HC, Granade TC, Nunez CA, et al. Field evaluation of rapid HIV serologic tests for screening and confirming HIV-1 infection in Honduras. AIDS 1997; 11:369–375.
4. Centers for Disease Control and Prevention. Guidelines for the management of health care worker exposures to HIV and recommendations for postexposure prophylaxis. MMWR 1998; 47:3.
5. Wade NA, Birkhead GS,Warren BL.Abbreviated regimens of zidovudine prophylaxis and prenatal transmission of the human immunodeficiency virus. N Engl J Med 1998; 339:1409–1414.
6. Centers for Disease Control and Prevention. Update:HIV counseling and testing using rapid tests United States 1995. MMWR 1998; 47:211–215.
7. Kane B. Rapid testing for HIV: why so fast? Ann Intern Med 1999; 131:481–483.
8. Auxter S. The movement toward expedited maternal and newborn HIV testing. Clin Lab News 1999; 25(1):8–9.
9. Fransen K, Pollet DE, Peeters M, et al. Evaluation of a line immunoassay for simultaneous confirmation of antibodies to HIV-1 and HIV-2. Eur J Clin Microbiol Infect Dis 1991; 10:939–946.
10. Constantine NT. Serologic tests for the retroviruses: approaching a decade of evolution. AIDS 1993; 7:1–13.
11. Centers for Disease Control and Prevention. Interpretations and use of the Western blot assay for serodiagnosis of human immunodeficiency virus type 1 infections. MMWR 1989; 38:1–7.
12. Food and Drug Administration. Licensed/approved HIV, HTLV and hepatitis tests. Updated August 29, 2000.Available from: www.fda.gov/cber/products/testkits.htm.
13. Desai S, Bates H, Michalski FJ. Detection of antibody to HIV-1 in urine. Lancet 1991; 337:183–184.
14. Farzadegan H, Quinn T, Frank B. Detecting antibodies to human immunodeficiency virus in dried blood on filter papers. J Infect Dis 1987; 55:1073–1074.
15. Constantine NT, Callahan JD, Watts DM. Retroviral testing: essentials for quality control and laboratory diagnosis. Boca Raton, FL: CRC Press, 1992.
16. Saville R, Constantine NT, Depaola L,Wisnour C, Falker W. Evaluation of HIV-1/2 rapid/simple assays to detect HIV antibodies in oral fluids. J Clin Lab Anal 1997; 11: 63–68.
17. Arai H, Petchclai B, Khupulsup K, Kurimura T,Takeda K. Evaluation of a rapid immunochromatographic test for detection of antibodies to human immunodeficiency virus. J Clin Microbiol 1999; 37(2):367–370.
18. Weber B, Fall EH, Berger A, Doerr HW. Reduction of diagnostic window by new fourth-generation human immunodeficiency virus screening assays. J Clin Microbiol 1998; 36(8):2235–2239.
19. Gurtler L, Muglbacher A, Michl U, et al. Reduction of the diagnostic window with a new combined p24 antigen and human immunodeficiency virus antibody screening assay. J Virol Methods 1998; 75:27–38.
20. Saville RD, Constantine NT, Cleghorn FR, et al. Fourth generation enzyme-linked immunosorbent assay for the simultaneous detection of human immunodeficiency antigen and antibody. J Clin Microbiol 2001; 39:2518–2524.
21. American Association of Blood Banks. HIV-1 antigen test implementation guidance. AABB Assoc Bull 1996; No. 92–2.
22. Food and Drug Administration. Guidance for industry application of current statutory authority to nucleic acid testing of pooled plasma. Draft guidance. Updated March 17, 2001. Available from: www.fda.gov/cber/gdlns/ ppnat.txt.
23. Janssen RS, Satten GA, Stramer SL, et al. New testing strategy to detect HIV-1 infection for use in incidence estimates and for clinical and prevention purposes. JAMA 1998; 280:42–48.
24. Constantine NT, Cafarella T, Cleghorn F. Abstract PS18. In: Abstracts of the 15th Annual Meeting of the Association of Public Health Laboratories, 2000.
25. Mellors JW, Rinaldo CR, Gupta P, et al. Prognosis in HIV- 1 infection predicted by the quantity of virus in plasma. Science 1996; 272:1167–1170.
26. Myers RA, Szuch WF, Patel JD, Joseph JM. Abstract P18. In:Abstracts of the Conference on the Laboratory Science of HIV, 1998.
27. Hirsch MS, Conway B, Daquila RT, et al. Antiretroviral drug resistance testing in adults with HIV infection. JAMA 1998; 279:1984–1991.
28. Gurtler L, Zekeng L, Simon F, et al. Reactivity of five anti- HIV-1 subtype O specimens in six different anti-HIV screening ELISAs and three immunoblots. J Virol Methods 1995; 51:177–184.
29. Constantine NT, Zekeng L, Sangare A, et al. Diagnostic challenges for rapid HIV assays: performance using HIV- 1 group O, group M, and HIV-2 samples. J Human Virol 1997; 1:46–52.
30. Cheingsong-Popov R, Callow D, Beddows S, et al. Geographical diversity of HIV-1: serologic reactivity to env epitopes and relationship to neutralization. J Infect Dis 1992; 165:256–261.
31. Tamashiro H, Maskill W, Emmanuel J, et al. Reducing the cost of HIV antibody testing. Lancet 1993; 342:87–90.
32. Chan EL, Sidaway F, Horsman GB. A comparison of the Genie and Western blot assays in confirmatory testing for HIV-1 antibody. J Med Microbiol 1996; 44:223–225. 33. Boston Biomedica, Inc. Seroconversion panels data sheets. Boston Biomedica, Inc.:West Bridgewater, Massachusetts, 1997.
34. Constantine NT. Intervening in blood supply and use systems. In: Gibney L, DiClemente J, Vermund S, eds. Preventing HIV in developing countries: biomedical and behavioral approaches. New York: Plenum, 1999: 71–85
Source: http://www.seracare.com/Portals/0/Papers/Rapid%20confirmation%20of%20HIV%20infection.pdf