antigen detection for human immunodeficiency virusantigen detection for hiv 243 table 1-continued...

CLINICAL MICROBIOLOGY REVIEWS, JUly 1989. p. 241-249 Vol. 2, No. 30893-8512/89/030241-09$02.00/0Copyright © 1989. American Society for Microbiology

Antigen Detection for Human Immunodeficiency VirusDAVID J. HARRY, MYRA B. JENNINGS, JoANN YEE, AND JAMES R. CARLSON*

Department of Medical Pathology, S(chool of Medii4ne, University of Cahlij)rnia, Davis, California 95616

INTRODUCTION .................................... 241DEVELOPMENT OF HIV ANTIGEN ASSAYS .................................... 241COMMERCIAL HIV ANTIGEN ASSAYS .................................... 242CLINICAL LABORATORY APPLICATIONS .................................... 244

Diagnosis of Early HIV Infections .................................... 244Diagnosis of HIV Infection in Infants .................................... 244HIV Antigen Detection in Virus Culture .................................... 245HIV Antigen as a Predictive Marker .................................... 245Research Laboratory Applications .................................... 245

BIOLOGY OF HIV ANTIGENEMIA .................................... 246CONCLUSIONS .................................... 246ACKNOWLEDGMENT .................................... 247LITERATURE CITED.................................... 247

INTRODUCTION

The human immunodeficiency virus (HIV) was first iso-lated by F. Barre-Sinoussi and her colleagues at the PasteurInstitute in Paris in 1983 (6). The isolation procedure in-volved coculture of lymph node cells from a symptomaticperson with phytohemaglutinin-stimulated, peripheral bloodmononuclear cells from a healthy donor. The presence of aretrovirus, first designated lymphadenopathy-associated vi-rus and later officially called HIV, in culture was suggestedby the appearance of high levels of reverse transcriptase(RT) in culture supernatants after only 7 days and wasconfirmed by electron microscopy. Scientists at the NationalCancer Institute in the United States further developed theculture method to permit continuous production of HIV in ahuman T-cell lymphoma cell line (H9) (56). This proceduremade available large amounts of HIV antigen necessary fordevelopment and production of serological test reagents.The initial HIV serological tests became indispensable sci-entific tools to define the extent of the HIV infectionepidemic (9). Further, their use in screening blood donorshas virtually eliminated blood transfusion as a route of HIVtransmission (71). Several test formats are now in use for thedetection of anti-HIV antibody, including enzyme immu-noassay (EIA) (10, 58), Western blot (immunoblot) (15),indirect immunofluorescence assay (11), and radioimmuno-precipitation (40). Recently, more rapid, simple, and inex-pensive tests for anti-HIV antibody have been developed foruse in developing countries (12, 72).

Serological analysis of HIV-infected individuals has, untilrecently, been limited to the detection of anti-HIV antibod-ies. A new development is the direct detection of HIVantigen(s). The first methods described detected HIV anti-gen in cell culture supernatants, thus providing immunolog-ical confirmation of RT assay results (33, 48). The HIVantigen assays have been further refined to detect directlyHIV antigens in body fluids from infected individuals (30, 52,74). The HIV antigen assay has become an important labo-ratory test for diagnosis, prognosis, and research on HIVinfection. The intent of this review is to describe (i) thehistorical development of the HIV antigen assay and (ii) a

* Corresponding author.

technical comparison of selected commercially availableassays and (iii) their clinical and research applications.

DEVELOPMENT OF HIV ANTIGEN ASSAYS

Assays specific for HIV antigen were first described asadjuncts to HIV purification and characterization methods(4, 13). A competition radioimmunoassay for HIV p24 wasdeveloped to determine the antigenic cross-reactivity amongHIV type 1, human T-cell leukemia virus type I (HTLV-I),and HTLV-II p24 proteins isolated from culture superna-tants (62).

Later, indirect immunofluorescence assay methods, usinghuman anti-HIV antibodies to detect infected cells, werealso developed (46). The initial indirect immunofluorescenceassay methods for detecting infected cells in culture weretechnically difficult and required extensive absorption of thehuman polyclonal antibodies to prevent nonspecific bindingto cultured lymphocytes (46). Immunofluorescence assayswith anti-HIV polyclonal or monoclonal antibodies werealso used to detect the distribution of HIV in blood andtissue (17, 55, 56, 67, 68, 73).HIV antigen EIAs were first developed to facilitate detec-

tion of HIV in culture. Early methods for viral assay in cellculture were restricted to detection of RT activity (29, 60,64), with electron microscopy studies used as confirmation(6, 46). The routine use of RT as a laboratory test presentedseveral disadvantages (22). Since all human retrovirusespossess functionally identical RTs, the assay cannot discrim-inate among HIV and other retroviruses possibly present.Furthermore, the RT assay is technically very complex,requires a radioisotopic label, and cannot be convenientlyautomated, although a micromethod has now been described(63).The first HIV antigen EIA was described by McDougal et

al. in 1985 (48). High-titered anti-HIV immunoglobulin G(IgG) purified from a seropositive individual was used in asandwich EIA test format. The IgG was bound to the plasticsurface of a microtiter well as the capture antibody. HIVculture supernatant was added, followed, after incubationand washing, by incubation with an enzyme-labeled humananti-HIV antibody. If antigen was present, a "sandwich"'formed among the capture antibody, HIV antigen, and

241

on March 27, 2020 by guest

http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/

242 HARRY ET AL.

TABLE 1. Comparison of selected commercial HIV antigen assays'

Capture step Detection step

Manufacturer of Incubation Incubation

assay Antibody Matrix Specimen Vol Atbd(ml) Time Temp Antibody Time Temp(h) (OC) (min) (0C)

Abbott H anti-HIV Polystyrene beads S/P/CCS 0.2 16-20 15-30 R anti-HIV 240 40Coulter mc anti-p24 Microtiter well S/P/CCS 0.2 1 37 Biotin anti-HIV 60 37Du Pont pcR anti-p24 Microtiter well S/P/CCS 0.2 18 or 2' 22 or 37' Biotin-R anti-p24 120 37Genetic Systems mc anti-p24 Microtiter well S/P/CCS 0.2 18 or 2' 37 HRPO-H anti-HLV 120 37Cellular Products H anti-HIV Microtiter well S/P/CCS 0.2 18 or 2' 22 or 37' Biotin-mc anti-p24 120 37

' H anti-HIV, human anti-HIV; R anti-HIV, rabbit anti-HIV; mc anti-p24, monoclonal anti-p24; G anti-R, goat anti-rabbit; pcR anti-p24, polyclonal rabbitanti-p24; HRPO, horseradish peroxidase; Str-HRPO, streptavidin-horseradish peroxidase; OPD, o-phenvlenediamine: TMB, tetramethylbenzidine; S. serum; P.plasma; CCS, cell culture supernatant.

" Lowest recommended dilution of stock standard material.Rapid qualitative method.

enzyme-labeled detector or probe antibody. After the addi-tion of enzyme substrate, a positive reaction resulted incolor development. The sensitivity of the assay comparedfavorably with the RT assay, with the added advantage that20-fold less culture supernatant was needed, thus facilitatingthe miniaturization of cultures from flasks to cell culturetubes.

In another early assay, two noncompeting monoclonalantibodies with specificity for the major HIV core protein,p24, were used with a rabbit polyclonal anti-HIV antibody invarious combinations as capture and probe reagents todetect HIV in culture (33). This assay was as sensitive as RTfor the detection of HIV in culture. The p24 monoclonalantibodies were the first described that differentiated be-tween HIV strains on the basis of reactivities to coreproteins. Both monoclonal antibodies (MAb) were producedagainst HIV (HTLV-III) but one, MAb 22-6, reacted withthe prototype strains HIV (HTLV-III), HIV (lymphadenop-athy-associated virus type 1), and HIV (acquired immuno-deficiency syndrome [AIDS]-related virus type 2), while theother, MAb 22-3, reacted only with HIV (HTLV-III) andHIV (lymphadenopathy-associated virus type 1) and notwith HIV (AIDS-related virus type 2). By altering the testformat, e.g., using either MAb 22-6 or MAb 22-3 as captureantibody with the rabbit anti-HIV antibody as probe, it waspossible to identify several antigenically distinct HIV strainsfrom the same patient (7).These initial tests detected HIV antigens only in virus

culture supernatants; the sensitivities were not adequate forthe detection of the very low titers of HIV antigen present inbody fluids of most HIV-infected individuals. The develop-ment of a triple-antibody sandwich procedure, using anti-bodies primarily specific for the p24 antigen, provided theneeded amplification to detect HIV antigens in clinicalspecimens (30, 74). The p24 antigen is a predominant struc-tural protein for HIV and may be present in higher concen-trations than other viral antigens in clinical specimens. Forexample, a system composed of human anti-HIV antibody,rabbit anti-HIV p24, and enzyme-labeled goat anti-rabbitantibody detected between 50 and 100 pg of whole HIV orpurified p24 per ml, respectively, 1,000 times more sensitivethan the RT assay (74).

COMMERCIAL HIV ANTIGEN ASSAYS

Several HIV antigen EIAs are now commercially avail-able. Table 1 describes the specific methods for five such

EIA test kits: Abbott Laboratories (North Chicago, Ill.), DuPont Co. (Wilmington, Del.), Coulter Electronics, Inc. (Hi-aleah, Fla.), Genetic Systems Inc. (Seattle, Wash.), andCellular Products Inc. (Buffalo, N.Y.). All assays use asolid-phase antigen capture matrix composed of either poly-clonal or monoclonal anti-HIV antibodies, bound to eitherpolystyrene beads or microtiter plate wells. Test samples areincubated with the capture matrix. After incubation andwashing to remove unbound protein, the matrix is incubatedwith anti-HIV probe antibody reagent. If HIV antigen hasbound to the capture matrix during the first incubation, theHIV-specific antibody in the probe reagent will also bind,thus forming an immobilized, three-component, antibody-antigen-antibody sandwich. Use of capture and probe anti-bodies with different specificities ensures that both probeand capture antibody will not compete for the same antigenmoieties. A color development, or "indicator," step isperformed next. The commercial antigen EIAs vary mostsubstantially in this step. Four of the assays, Du Pont,Abbott, Coulter, and Cellular Products, use an enzyme-conjugated indicator reagent that specifically binds to theimmobilized sandwich. Genetic Systems consolidates theindicator and probe reagent into a single step by using anenzyme-conjugated anti-HIV antibody as probe reagent. Theenzyme used in all five assays is horseradish peroxidase.Three assays, Du Pont, Coulter, and Cellular Products, usebiotinylated probe antibody. In these three assays, theindicator reagent is composed of horseradish peroxidase-conjugated streptavidin. The four assays that use a horse-radish peroxidase-conjugated indicator reagent require anadditional incubation step in which specific binding occursbetween the immobilized sandwich and the horseradishperoxidase-conjugated indicator. Following a series ofwashes to remove unbound conjugate, the final assay stepfor all five tests is the addition of an enzyme substratereagent. Substrate turnover produces color that is propor-tional to the amount of HIV antigen present. Spectrophoto-metric absorbance measurements can thus be related toantigen concentration.

All kits incorporate a method for qualitative determinationof results, based on a comparison of sample absorbanceswith a "negative cutoff" or "reactive threshold" value. Thereactive threshold value is derived from the mean of repli-cate absorbance values observed for a negative controlmaterial supplied with the kit. Any sample with an absorb-ance exceeding the reactive threshold value is consideredreactive. Semiquantitative results may be obtained with

CLIN. MICROBIOL. REV.


http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/

ANTIGEN DETECTION FOR HIV 243

TABLE 1-Continued

Indicator step Data analysis

Incubation ConfirmationAntibodyy Conj tgsite Conjulgaite temp ( C ) S~hstratc Timc TcpConjugate Conan Sensitivity" (neutraliza-Anibdy Cojuat onjugateCtime(mm) tem SusrteC)n estiiy tion assay)time(min) tSubstrateTime Temp determination (pg/mil)

(min) C)

G anti-R HRPO 120 40 OPD 30 22 Semiquantitative 15 YesStr-HRPO 30 37 TMB 30 37 Quantitative 8 YesStr-HRPO 30 37 OPD 30 22 Quantitative 30 Yes

TMB 30 22 Semiquantitative No (repeat)Str-HRPO 15 37 TMB 30 25 Quantitative 25 No

Abbott and Genetic Systems assays, using an optionalstandard curve procedure. Serial dilutions of a stock stan-dard preparation are run in parallel with the control andexperimental specimens. A standard curve is derived byplotting absorbance data for the standard dilutions againstthe dilution factor, using graphical or linear/quadratic least-squares procedures. Comparison of the experimentalabsorbance with the standard curve allows semiquantitativedetermination of antigen concentration. The method is de-scribed as "semiquantitative," because the absolute antigenconcentration in the stock standard preparation is not statedin molar or concentration units, but in arbitrary units permilliliter, presumably standardized among succeeding testkit lots. Three test kits, Du Pont, Coulter, and CellularProducts, provide a quantitative method for determination ofp24 core antigen concentration. A standard curve is pre-pared by using serial dilutions of a stock solution containingp24 antigen at a known concentration. Standardization be-tween manufacturers for HIV antigen quantitation has notyet been achieved as evidenced by a recent report in whichAbbott standards used in the Du Pont assay were 4-foldlower than stated values and Du Pont standards in theAbbott test were 2.5- to 4-fold higher than stated values (S.Crowe, N. McManus, M. McGrath, and J. Mills, Abstr. 4thInt. Conf. AIDS 1988, vol. 2, abstr. 1616, p. 79). Thesediscrepancies may reflect differences in the avidity andspecificity of both capture and probe antibodies as well as inthe material used as the standard.

Reactive specimens should be confirmed. The preferredmethod of confirmation uses an antigen neutralization assayprovided with some kits. This procedure requires preincuba-tion of the sample with antibody to HIV. followed byrepetition of the antigen EIA. HIV antigen present in thesample will be sequestered in immune complexes during thepreincubation step and consequently will not be boundduring the capture phase of the antigen EIA. This results ina reduction in measured absorbance for the neutralizedspecimen in comparison with that of a non-neutralizedidentical sample run simultaneously. Initially reactive spec-imens are considered confirmed when an absorbance reduc-tion of at least 50% is observed for the neutralized sampleand an absorbance exceeding the reactive threshold value isobserved for the non-neutralized sample.Specimen requirements are similar for all commercial

antigen EIAs. Serum, plasma, and cell culture supernatantsmay all be tested. The use of ethylenediaminetetraaceticacid, sodium citrate, or heparin as anticoagulant does notappear to affect performance. All manufacturers discouragetesting of turbid, lipemic, or hemolyzed samples. Specimenstorage at 2 to 80C for up to 1 week and extended frozenstorage at -20 or -70'C are both permissible, althoughrepeated freezing and thawing is not recommended.

To achieve maximum sensitivity, three assays, Du Pont,Abbott, and Genetic Systems, require an overnight incuba-tion period for the capture phase. Completion of the probeand indicator phases generally requires an additional 2 to 5 h,establishing a realistic turnaround time of approximately 24h for the complete assay. An additional 24 h is needed tocomplete confirmatory neutralization or repeat procedures.Thus. for most kits, a confirmed positive result wouldrequire at least 2 days, using the most sensitive test protocol.The Coulter and Cellular Products quantitative assays re-quire only 1- and 2-h incubations, respectively, at 370Cduring the capture phase, thus considerably shortening theturnaround time. Incubation times for the remaining stepsare equivalent to those of the other assays. Du Pont andGenetic Systems both offer a rapid procedure for qualitativedetermination that allows completion of the assay in 5 and2.5 h, respectively.Based on list price, the cost per bead or well for the four

assays are as follows: Abbott, $8.00 per bead; Du Pont,$4.79 per well; Genetic Systems, $6.00 per well; Coulter,$5.50 per well; and Cellular Products, $2.35 per well. Thecost per reportable test will be considerably higher, sinceaccount must be taken of standards, controls, replicatesamples, and confirmatory determinations on initially reac-tive specimens.

Sensitivity limits for the four assays detailed in Table 1vary between 10 and 30 pg/ml when controls are tested.However, comparison of relative sensitivity is difficult dueto the diversity of antigenic materials used for controls,different antibody specificities, and lack of a uniform refer-ence standard. Several reports have compared the sensitiv-ity of a select group of commercial assays, but the availablepublished data are not sufficient to draw conclusions aboutthe relative performance of all commercially available HIVantigen ETAs (34, 36; Crowe et al., Abstr. 4th Int. Conf.AIDS 1988; F. Barin, A. Courouce. A. Maniez, and M.Rouziouz, Abstr. 4th Int. Conf. AIDS 1988, vol. 2, abstr.1618, p. 80).The HIV antigen ElAs are very specific. Nonspecific

reactivity has not been demonstrated with a large group ofhuman viruses, including HTLV-I, cytomegalovirus, herpessimplex virus (30); poliovirus type 1. adenovirus type 5,varicella-zoster virus (22); and numerous cell culture lines.However, limited cross-reactivity for HIV type 2 has beenreported (D. Paul, M. Knigge, M. Kennedy, D. Mack, and J.Leibowitch, Abstr. 4th Int. Conf. AIDS 1988, vol. 2, abstr.1647, p. 87). Clinical evaluation of the Abbott HIV antigenassay in large low-risk populations has demonstrated greatspecificity. False-positive results were not detected in>250,000 blood donors when the neutralization test wasused to confirm repeatedly reactive tests (U. Backer, F.Weinauer. A. G. Gathof, E. Gossrau, J. Eberle, and F.

VOL. 2, 1989


http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/

244 HARRY ET AL.

Deinhardt, Abstr. 4th Int. Conf. AIDS 1988, vol. 2, abstr.7757, p. 364). The repeatedly reactive rate of 0.18% wasentirely attributable to false-positive results as confirmed byusing a supplementary neutralization assay. New HIV-infected individuals were not identified by HIV antigenscreening in this low-risk population, indicating that theprevalence of HIV infection may be too low to warrant theaddition of this test as a routine screening procedure at blooddonor centers.

If it is assumed that all seropositive individuals are in-fected with HIV, the sensitivity of the HIV antigen assay indetecting HIV infection directly in patient specimens variesfrom 4% in asymptomatic seropositive subjects to 70% inAIDS patients (38). Of 22 seropositive adult and pediatricpatients with AIDS, 19 had detectable amounts of antigen intheir sera, in contrast to 1 of 13 asymptomatic seropositiveindividuals (30). Antigen positivity rates in cerebrospinalfluid demonstrated a similar trend, being 55 and 0%, respec-tively, for AIDS and asymptomatic patients.The apparent lack of sensitivity of the HIV antigen assay

in detecting HIV infection, particularly in asymptomaticindividuals, limits its use as a screening assay for HIVinfection in low-risk groups, but there are other usefulclinical and research applications for this test. Furtherstudies may better define the sensitivity, specificity, andclinical utility of this new type of laboratory test. It will alsobe especially important to establish quality assurance pro-cedures to control intra- and interlaboratory proficiency inthe performance of the HIV antigen EIAs (40; F. B.Hollinger, J. W. M. Gold, L. Meyer, C. van de Horst, R.Makuch, W. Parks, and the NIH-NIAID AIDS ProgramTreatment Branch, Abstr. 4th Int. Conf. AIDS 1988, vol. 2,abstr. 1618, p. 82).

CLINICAL LABORATORY APPLICATIONSInfection with HIV may result in an acute viral syndrome

that mimics influenza or mononucleosis with headache,fever, and malaise (1). After the acute phase, during whichseroconversion may occur, an asymptomatic period of vari-able duration follows. Eventually, after a mean latent periodof about 8 years, virtually all HIV-infected individualsdevelop laboratory and clinical signs of HIV disease, withprogression to AIDS as the ultimate outcome (50).The clinical laboratory diagnosis of HIV infection is

routinely based on the use of anti-HIV serological tests (9)and virus isolation (46). Other laboratory markers for HIVdisease progression are nonspecific and include enumerationof T cells, serum immunoglobulin, and P2-microglobulinlevels (50, 75). The HIV antigen assay has improved thelaboratory diagnosis of HIV infection by providing a moresensitive marker than routine antibody screening during theearly stages of infection and for confirming neonatal infec-tion in the presence of passively transferred maternal anti-HIV IgG (49). In addition, HIV antigen assays may be usefulin the diagnosis of persistent infection in children whobecome seronegative but who remain infected (8). The HIVantigen assay has also improved diagnosis of HIV infectionby culture methods by eliminating the need for the RT assayand by increasing the test sensitivity and decreasing the timefor results (22).The detection of HIV antigen in serum, plasma, or cere-

brospinal fluid after seroconversion is a useful indicator ofHIV disease progression, offering a simple and more specificalternative to the nonspecific laboratory tests. The HIVantigen assay can also be used to monitor the efficacy ofantiviral drug therapy (14, 19, 35, 59).

Diagnosis of Early HIV Infections

HIV antigen has been detected in serum or plasma in theearly stages of HIV infection even before seroconversion ina small percentage of infected asymptomatic subjects (3, 30)and in those persons with an acute HIV-associated illness(27, 39, 69, 70). However, shortly after seroconversion, mostHIV-infected individuals become HIV antigen negative orare only transiently positive (69).The HIV antigen assay should be very useful for the

diagnosis of acute HIV infection. However, the generalbenefit of this assay for screening asymptomatic high-risk orlow-risk populations will depend on the length of the"window" period between infection and seroconversion andon the incidence of infection in the population tested. Forexample, the HIV antigen assays will be less useful if thewindow is short and the incidence of infection is low.

Several studies have attempted to determine the length ofthe window phase. In a cohort of 741 homosexual men, 35seroconverted during the study period and HIV antigen wasdetected in 14% before and in 17% after seroconversion (30).The elapsed time between antigen detection and seroconver-sion was 3 months, and the antigen titer fell below detectablelimits within 1 month of seroconversion. Other reported timeintervals between infection and seroconversion vary be-tween 1 and 5 months (3, 39), while one recent reportsuggests a longer window period (57).

Diagnosis of HIV Infection in Infants

Standard serological diagnosis of HIV infection in high-risk neonates and young children is difficult due to thepassive transfer of maternal anti-HIV IgG (49) and thepresence of virus in the absence of autologous anti-HIVantibodies following the loss of maternal antibody (8). Clas-sically, the detection of specific IgM antibodies has proveduseful in differentiating between the specific response of theinfant to an infection and passively transferred maternalantibody (25). However, in the case of HIV infection, IgMassays have not been helpful (26). A recently developed andpotentially better assay for detecting neonatal antibodymeasures in vitro production of HIV-specific antibody byperipheral blood lymphocytes in culture (5). Currently, HIVculture and HIV antigen assay are the most specific diagnos-tic tests for HIV infection in neonates and young children (8,49).

Since maternal anti-HIV antibody may be transferred tothe fetus in utero without the concomitant transfer of HIV(49), the early detection of anti-HIV antibody from periph-eral blood is not a definitive diagnosis for HIV infection ofthe fetus or neonate. Detectable titers of maternal antibodycan persist for a considerable period. Among 71 neonatesborn to seropositive mothers, the median age for disappear-ance of maternal anti-HIV antibody was 11 months; 75% ofall children lost antibody by 1 year (49). When HIV antigenassay was applied to samples from 20 of the seropositivechildren, only 55% were positive, confirming infection.Presumably, the other 45% were false-positive antibodyassays due to the presence of residual maternal antibody. Inanother study, HIV antigen assay was positive when serumfrom each of six seropositive pediatric AIDS patients wastested (30).The incidence of HIV-infected neonates and young chil-

dren who remain seronegative for extended periods is muchhigher than in the adult population. The HIV antigen EIAhas particular value in confirming HIV infection in these



http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/


groups. Of 85 children with HIV infection, ranging in agefrom 2.5 months to 4 years, 9 (11%) were seronegative byanti-HIV EIA (8). However, all nine were positive by HIVantigen EIA. By using Western blot and a more sensitiveEIA with recombinant antigens, antibody was detected infour of nine of these children, indicating that only five weretruly antibody negative. Lange and Goudsmit (42) alsoreported on one of nine HIV-infected children with antigen-emia and absent anti-HIV antibody titer.

In conclusion, detection of HIV antigens may be useful inthe specific diagnosis of HIV infections in infants withmaternal anti-HIV IgG. When used in combination withmore sensitive anti-HIV antibody tests, HIV antigen testsmay confirm HIV infection in seronegative children born toseropositive mothers.

HIV Antigen Detection in Virus Culture

Initially, isolation of HIV in culture was the only methodfor the diagnosis of HIV infection. The development of HIVantibody tests (9) soon supplanted culture for screening.Results of early efforts to use HIV culture methods to assessantiviral drug efficacy were ambiguous. HIV isolation stud-ies failed to provide endpoints that predicted therapeuticefficacy in zidovudine (azidothymidine) drug trials- becausethe method lacked sufficient sensitivity to detect the subtleinhibitory effect of the drug (23). As more effective drugs aredeveloped, however, HIV isolation may prove to be moreuseful for testing them. At present, virus isolation is mostuseful for the diagnosis of infection in neonates and for theidentification of new HIV strains.

Virus culture has not been widely used because theprocedure is difficult. Fresh human peripheral blood mono-nuclear cells are the most sensitive target cells and are noteasily obtained. Most important, however, is the need forfrequent RT assays to detect HIV in culture. The use of RTas a marker for HIV production suffers from (i) technicaldifficulty, (ii) lack of specificity (i.e., it cannot distinguishamong retroviruses), and (iii) long incubation time, requiring1 to 3 weeks of cocultivation of lymphocytes from the patientwith phytohemagglutinin-stimulated normal lymphoblastsfor the detection of HIV (22).

Recently, HIV antigen assays have been compared withRT assay for the detection of HIV in culture (22, 37, 44, 51,68, 74). The majority of reports have shown that the HIVantigen assay is more sensitive than RT, yielding positiveresults on more specimens and at an earlier time in theculture process. In one study (22), RT activity first appearedin 34 cultures 6 to 16 days after initiation of culture. Theaverage time for detection of HIV by RT was 9.6 days. Incontrast, p24 was detectable in culture supernatants by HIVantigen assay within 3 to 9 days, with an average detectiontime of 5.9 days.The relative sensitivity of HIV antigen EIA is 100- to

1,000-fold greater than that of the RT assay in detecting virusin cell culture supernatants (22, 32, 51, 74). If performancestandards and reproducibility (41) can be established for thistest, it offers a suitable substitute for RT in HIV cultureapplications.

HIV Antigen as a Predictive Marker

A strong correlation has been established between theonset of antigenemia and the progression of disease afterseroconversion (2, 19, 24, 29, 30, 38, 42, 45, 47, 50, 53, 61).This suggests the use of HIV antigen EIA as a prognostic

index in assessing disease progression. In a group of 96hemophiliacs, a much higher incidence of AIDS was ob-served in those individuals who were positive for HIVantigen (2). Moreover, antigen titers tended to increase withdisease progression. In eight individuals who initially testedantigen positive, the antigen titer increased almost threefoldover approximately 3 years and all patients developed clin-ical complications of active HIV infection. Another study of32 patients with AIDS-related complex (ARC) found asignificant correlation between HIV antigen positivity orabsence of anti-p24 antibody or both and progression toAIDS (38). Within a 23-month period, 23% of ARC patientswho were initially antigen positive progressed to AIDS,whereas only 8% of those who were initially antigen negativedid so. Of 34 seropositive men, 8 of 16 with antigenemiadeveloped AIDS or ARC, compared with only 1 of 18without antigenemia (53).

In a recent study of 281 men who were seropositive atentry, 46 (16%) were initially HIV antigen positive (50). Overthe next 3 years, 59% of men initially positive for HIVantigen progressed to AIDS versus 15% of those who wereinitially HIV antigen negative. The presence of p24 antigenwas positively correlated with reduced T-helper lymphocytelevels (T4 < 400 x 106/liter) as a predictive variable. Fur-thermore, HIV antigen was a better predictor of diseaseprogression than loss of anti-p24 antibody. Other usefulpredictors of disease progression were elevated serum levelsof P32-microglobulin, packed cell volume of <40, proportionof T4 lymphocytes of <25%, and absolute number of lym-phocytes being s200 x 106/liter. The implication of thisstudy is that simple and more straightforward laboratorytests such as HIV antigen EIA, 32-microglobulin, andpacked cell volume may be as useful as the more difficultT-cell assays for the prediction of HIV disease progression.HIV antigen test results may also serve as a useful marker

for predicting the efficacy of antiviral drug therapy. Acorrelation has been established between reduction in anti-genemia in zidovudine-treated HIV antigen-positive subjectsand resultant beneficial clinical outcome (14, 23). HIV anti-gen detection has demonstrated significant reduction inantigen levels in serum, plasma, and cerebrospinal fluid ofsubjects treated with zidovudine (14, 17, 18, 35, 59). In drugefficacy studies, zidovudine has been compared with acyclo-vir (18) and suramin (20), and only zidovudine was effectivein reducing antigenemia.The HIV antigen assay is an effective laboratory tool for

assessing antiviral drug activity in clinical studies. Onepotential limitation is the low prevalence of antigenemiaobserved in asymptomatic, HIV-infected populations (4, 30,38), which may limit the use of the HIV antigen assay to drugstudies involving ARC and AIDS patients in whom theprevalence of detectable antigenemia is much higher. Eithermore sensitive HIV antigen assays need to be developed ortests measuring other parameters, such as 32-microglobulinlevels, may prove to be important adjuncts to this type ofstudy (50, 75; M. Jacobson, D. Abrams, J. Wilber, P.Volberding, J. Mills, R. Chaisson, and A. Moss, Abstr. 4thInt. Conf. AIDS 1988, vol. 2, abstr. 3646, p. 178).

Research Laboratory Applications

The HIV antigen assay has been applied to determineendpoint titrations of infectious HIV in culture (48, 65; E.Keddie, J. R. Higgins, J. R. Carlson, M. B. Jennings, andN. C. Pedersen, Abstr. Annu. Meet. Am. Soc. Microbiol.1987, T-38, p. 321). An obvious advantage of the HIV

VOL. 2, 1989


http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/

246 HARRY ET AL.

antigen assays is their requirement for much smaller vol-umes of test sample as compared with the RT assay. This hasmade it possible to miniaturize the culture system, allowingtesting of multiple replicate samples for calculating 50%tissue culture infectious dose of the HIV preparation usedfor inocula. Because the HIV antigen assays are optimizedfor the detection of p24 (a conserved region of the HIVgenome), an endpoint of infectivity may be determined fordifferent HIV isolates in any susceptible cell line or in freshperipheral blood mononuclear cells. Assays in which infec-tivity endpoints are established by HIV antigen assay havebeen used to detect antibody neutralization of infectivity andinactivation of HIV by both chemical and physical methods.

Caution should be exercised when using HIV antigenassays in bioassays for HIV inactivation because residualnoninfectious inactivated HIV may be detected in the testinoculum (Keddie et al., Abstr. Annu. Meet. Am. Soc.Microbiol. 1987). Therefore, an increase in HIV antigen titerover time, indicating active replication, must be detected incultures to confirm the presence of virus replication follow-ing inactivation procedures.

BIOLOGY OF HIV ANTIGENEMIA

The biology of HIV antigen expression and modulation ofanti-p24 antibody titers during infection can be described bytwo possible models. The first would propose that, afterinfection, HIV begins replication in the host and expressionof virus is detected before seroconversion as HIV antigen inperipheral blood or cerebrospinal fluid. Then, due to unde-termined host or virus factors, HIV becomes latent orexpressed at negligible levels until later in the diseasecourse, when replication is again activated and viral proteintiters increase with formation of immune complexes thatlead to depletion of antibody titers, particularly anti-p24.Another model proposes that continuous low-level produc-tion of HIV occurs during the incubation period followingprimary infection (3). The reduction in initial antigenemiafollowing seroconversion is caused by immune complexformation that continues during the seropositive asympto-matic phase of the infection. Progression to ARC or AIDS isassociated with an increase in antigen production, eventuallyleading to a condition of antigen excess, resulting in antigen-emia and reduction of p24 antibody titer.

Differentiating between the two alternatives will requireanswering two key unresolved questions: (i) what causesrapid loss of antigenemia following seroconversion? and (ii)what initiates the antigenemia with disease progression?The onset of antigenemia following primary infection

occurs within approximately 1 to 5 months. Following sero-conversion, the HIV antigen level quickly falls, usually toundetectable levels, where it remains in clinically asympto-matic seropositive individuals. Onset of ARC or AIDS maybe accompanied by detection of antigenemia. A steadilyincreasing antigen titer and decrease in anti-p24 antibodytiter have been strongly correlated with a progression ofdisease and poor prognosis (43, 50).

In support of the latent HIV model, Von Sydow et al. (69)observed HIV antigen-containing immune complexes duringthe phase of decreasing antigen titer immediately followingseroconversion. However, no complexed antigen was de-tected for months following loss of detectable antigenemia,suggesting low or absent synthesis of HIV in the host.Others (42, 52), however, have reported contrasting results,indicating the presence of antigen-antibody complexes ininfected individuals with or without the presence of anti-p24

antibodies, although immune complexes were more com-monly found in those subjects without detectable anti-p24levels (42). Further studies are necessary to resolve thequestion of whether reduced viral production or masking ofantigen by antibodies is primarily responsible for the loss ofantigenemia following seroconversion.

Several lines of evidence support the view that, during lateinfection, the reappearance of antigenemia and decrease inanti-p24 antibody are related to an increase in viral replica-tion, rather than selective inhibition of antibody production.It is difficult to envision an immune regulatory mechanismthat selectively inhibits only one among the several antibod-ies to HIV produced. Furthermore, it has been observed thatsome patients with high antigen levels also lack antibody top55, the undegraded gag precursor polyprotein from whichp24 is ultimately processed (2, 28). Thus, the reduction inantibody appears not to be totally selective for anti-p24.Preliminary studies also suggest that an increased synthesisof p24 antigen is related to increased HIV replication, sincea strong correlation exists between plasma HIV antigen titerand the ability of plasma to infect normal peripheral bloodlymphocytes in vitro with HIV (21). Correlation of HIVantigen detection and polymerase chain reaction for HIVribonucleic and deoxyribonucleic acids may help to resolvethese questions raised about the biology of HIV antigenemia(31).

CONCLUSIONS

Development of EIA procedures for the direct determina-tion of HIV antigen has been of significant benefit in bothclinical and research arenas. The commercial availability ofHIV antigen assays has permitted HIV antigen determina-tions to be performed on a routine basis.The HIV antigen EIA is an effective substitute for the

more costly and complex RT assay for detecting HIVinfection in cell culture, and it has facilitated the develop-ment of microculture methods. Wider implementation ofHIV antigen EIA for in vitro virus identification shouldresult in more widely available, less costly, and more rapidculture determinations. The use of simple and straightfor-ward laboratory tests such as ,32-microglobulin and packedcell volume, in conjunction with HIV antigen EIA, may beas useful as the more difficult T-cell assays for predictingHIV disease progression.

In clinical applications, the predictive value of HIV anti-gen ETA testing of serum or plasma is highly dependent onthe prevalence of HIV infection in the target human popu-lation. In low-prevalence populations (e.g., blood donors),the routine use of HIV antigen EIA as a screening proceduredoes not appear to be justified. In high-risk asymptomaticpopulations, the HIV antigen EIA should prove increasinglyuseful in diagnostic and prognostic applications, which in-clude (i) early detection of infection in seronegative individ-uals who are members of a high-risk group; (ii) confirmationof HIV infection in children, in whom the prevalence ofseronegative HIV-infected individuals may be higher than inthe adult population; (iii) diagnosis of HIV infection inneonates of seropositive mothers; (iv) use of prognosticindex during the later stages of infection; and (v) assessmentof therapeutic efficacies of various antiviral drugs.The interrelationship between HIV antigen and anti-HIV

antibody titers during disease progression appears to bequite complex. Disappearance of initial antigenemia follow-ing seroconversion and later reappearance, frequently ac-companied by a reduction in anti-p24 antibody titer, may be



http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/


explained by at least two different models. One proposesthat varying levels of viral expression modulate antigentiters, while the other proposes that selective suppression ofanti-p24 antibody during late infection results in reappear-ance of antigenemia. Currently available studies offer con-flicting evidence in support of both models, and definitiveconclusions must await further studies.

ACKNOWLEDGMENT

This research was supported in part by funds provided by theCalifornia Universitywide Task Force on AIDS.

LITERATURE CITED1. Adler, M. W. 1987. Range and natural history of infection. Br.

Med. J. 294:1145-1147.2. Allain, J.-P., Y. Laurian, D. Paul, F. Verroust, M. Leather, C.

Gazengel, D. Senn, M.-J. Larrieu, and C. Bosser. 1987. Long-term evaluation of HIV antigen and antibodies to p24 and gp4lin patients with hemophilia. N. Engl. J. Med. 317:1114-1121.

3. Allain, J.-P., D. Paul, Y. Laurian, and D. Senn. 1986. Serolog-ical markers in early stages of human immunodeficiency virusinfection in hemophiliacs. Lancet ii:1233-1236.

4. Allan, J., J. Coligan, T. Lee, M. McLane, P. Kanki, J. Groop-man, and M. Essex. 1985. A new HTLV-III/LAV encodedantigen detected by antibodies from AIDS patients. Science230:810-813.

5. Amadori, A., C. Giaquinto, F. Zacchello, A. deRossi, G.Faulkner-Valle, and L. Chieco-Bianchi. 1988. In vitro productionof HIV-specific antibody in children at risk of AIDS. Lanceti:852-854.

6. Barre-Sinoussi, F., J. C. Chermann, F. Rey, M. T. Nugeyre, S.Chamaret, J. Gruest, C. Dauguet, C. Axler-Blin, F. Vezinet-Brun, C. Rouzioux, W. Rosenbaum, and L. Montagnier. 1983.Isolation of a T-lymphotropic retrovirus from a patient at riskfor acquired immune deficiency syndrome (AIDS). Science220:868-871.

7. Bolton, V., N. Pedersen, J. Higgins, M. Jennings, and J. Carlson.1987. Unique p24 epitope marker to identify multiple humanimmunodeficiency virus variants in blood from the same indi-viduals. J. Clin. Microbiol. 25:1411-1415.

8. Borkowsky, W., D. Paul, D. Bebenroth, K. Krasinski, T. Moore,and S. Chandwani. 1987. Human-immunodeficiency-virus infec-tions in infants negative for anti-HIV by enzyme-linked immu-noassay. Lancet i:1168-1171.

9. Carlson, J. R., M. L. Bryant, S. H. Hinrichs, J. Yamamato, N.Levy, J. Yee, J. Higgins, A. Levine, P. Holland, M. Gardner, andN. Pedersen. 1985. AIDS serology testing in low- and high-riskgroups. J. Am. Med. Assoc. 253:3405-3408.

10. Carlson, J. R., S. H. Hinrichs, J. Yee, M. B. Gardner, and N. C.Pedersen. 1986. Evaluation of three commercial screening testsfor AIDS virus antibodies. Am. J. Clin. Pathol. 86:357-359.

11. Carlson, J. R., J. Yee, S. H. Hinrichs, M. L. Bryant, M. B.Gardner, and N. C. Pedersen. 1987. Comparison of indirectimmunofluorescence and Western blot for the detection ofanti-human immunodeficiency virus antibodies. J. Clin. Micro-biol. 25:494-497.

12. Carlson, J. R., J. L. Yee, E. J. Watson-Williams, M. B. Jennings,S. C. Mertens, M. B. Gardner, J. Ghrayeb, and R. J. Biggar.1987. Rapid, easy, and economical screening test for antibodiesto human immunodeficiency virus. Lancet ii:361-362.

13. Casey, J., Y. Kim, P. Andersen, K. Watson, J. Fox, and S.Devare. 1985. Human T-cell lymphotropic virus type III: immu-nologic characterization and primary structure analysis of themajor internal protein, p24. J. Virol. 55:417-423.

14. Chaisson, R., J.-P. Allain, and P. Volberding. 1986. Significantchanges in HIV antigen level in the serum of patients treatedwith azidothymidine. N. Engl. J. Med. 315:1610-1611.

15. Consortium for Retrovirus Serology Standardization. 1988. Sero-logical diagnosis of human immunodeficiency virus infection byWestern blot. J. Am. Med. Assoc. 260:674-679.

16. Corey, J. M., B. Ohlsson-Wilhelm, N. Sheaffer, M. Steck, M.

Eyster, and F. Rapp. 1987. Detection of human immunodefi-ciency virus-infected lymphoid cells at low frequency by flowcytometry. J. Immunol. Methods 105:71-78.

17. deGans, J., J. Lange, M. Derix, F. deWolf, J. Eeftinck-Schatten-kerk, S. Danner, B. deVisser, P. Cloud, and J. Goudsmit. 1988.Decline of HIV antigen levels in cerebrospinal fluid duringtreatment with low-dose zidovudine. AIDS 1:37-40.

18. deWolf, F., J. Goudsmit, J. DeGans, R. Coutinho, J. Lange, P.Cload, P. Schellekens, A. Fiddian, and J. Noordaa. 1988. Effectof zidovudine on serum human immunodeficiency virus antigenlevels in symptom-free subjects. Lancet i:373-376.

19. deWolf, F., J. Goudsmit, D. Paul, J. Lange, C. Hooijkaas, P.Schellekens, R. Coutinho, and J. Noordaa. 1987. Risk of AIDSrelated complex and AIDS in homosexual men with persistentHIV antigenaemia. Br. Med. J. 295:569-572.

20. Eeftinck-Schattenkerk, J., S. Danner, J. Lange, D. Paul, C. vanBoxtel, F. Miedema, P. Schellekens, and J. Goudsmit. 1988.Persistence of human immunodeficiency virus antigenemia inpatients with acquired immunodeficiency syndrome treated witha reverse transcriptase inhibitor, suramin. Arch. Intern. Med.148:209-211.

21. Falk, L., D. Paul, A. Landay, and H. Kessler. 1987. HIVisolation from plasma of HIV-infected persons. N. Engl. J.Med. 316:1547-1548.

22. Feorino, P., B. Forrester, C. Schable, D. Warfield, and G.Schochetman. 1987. Comparison of antigen assay and reversetranscriptase assay for detecting human immunodeficiency virusin culture. J. Clin. Microbiol. 25:2344-2346.

23. Fischl, M., D. Richman, M. Grieco, M. Gottlieb, P. Volberding,0. Laskin, J. Leedom, J. Groopman, D. Mildvan, T. Schooley,G. Jackson, D. Durack, M. Phil, D. King, and the AZT Collab-orative Working Group. 1987. The efficacy of AZT in thetreatment of patients with AIDS and ARC: a double-blind,placebo-controlled trial. N. Engl. J. Med. 317:185-192.

24. Forster, S., L. Osborne, R. Cheingsong-Popov, C. Kenny, R.Burnell, D. Jeffries, A. Pinching, J. Harris, and J. Weber. 1987.Decline of anti-p24 antibody precedes antigenaemia as a corre-late of prognosis in HIV-1 infection. AIDS 4:235-240.

25. Fucillo, D., I. Shekarchi, and J. Sever. 1986. Rapid viral diag-nostics, p. 493-496. In N. R. Rose, H. Friedman, and J. L.Fahey (ed.), Manual of clinical laboratory immunology, 3rd ed.American Society for Microbiology, Washington, D.C.

26. Gaetano, C., G. Scano, M. Carbonari, G. Giannini, I. Mez-zaroma, F. Aiuti, L. Marolla, A. M. Casadei, and E. Carapella.1987. Delayed and defective anti-HIV 1gM response in infants.Lancet i:631.

27. Gaines, H., J. Albert, M. Von Sydow, A. Sonnerborg, F. Chiodi,A. Ehrnst, 0. Strannegard, and B. Asjo. 1987. HIV antige-naemia and virus isolation from plasma during primary HIVinfection. Lancet i:1317-1318.

28. Gill, J., J. Menitove, P. Anderson, J. Casper, S. Devare, C.Wood, S. Adair, J. Casey, C. Scheffel, and R. Montgomery. 1986.HTLV-II1 serology in hemophilia: relationship with immuno-logic abnormalities. J. Pediatr. 108:511-516.

29. Goudsmit, J., J. Lange, D. Paul, and G. Dawson. 1987. Antigen-emia and antibody titers to core and envelope antigens in AIDS,AIDS-related complex, and subclinical human immunodefi-ciency virus infection. J. Infect. Dis. 155:558-560.

30. Goudsmit, J., D. A. Paul, J. Lange, H. Speelman, J. Noordaa, H.Van Der Helm, F. de Wolf, L. Epstein, W. Krone, E. Wolters, J.Oleske, and R. Coutinho. 1986. Expression of human immuno-deficiency virus antigen (HIV-Ag) in serum and cerebrospinalfluid during acute and chronic infection. Lancet ii:177-180.

31. Hart, C., T. Spira, J. Moore, J. Sninsky, G. Schochetman, A.Lifson, J. Galphin, and C.-Y. Ou. 1988. Direct detection of HIVRNA expression in seropositive subjects. Lancet ii:596-599.

32. Healy, D., J. Jowett, F. Beaton, W. Maskill, and I. Gust. 1987.Comparison of enzyme immunoassay and reverse transcriptaseassay for detection of HIV in culture supernatants. J. Virol.Methods 17:237-245.

33. Higgins, J. R., N. C. Pedersen, and J. R. Carlson. 1986.Detection and differentiation by sandwich enzyme-linked im-munosorbent assay of human T-cell lymphotropic virus type

VOL. 2, 1989


http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/

248 HARRY ET AL.

III/lymphadenopathy-associated virus- and acquired immuno-deficiency syndrome-associated retrovirus-like clinical isolates.J. Clin. Microbiol. 24:424-430.

34. Hoffman, A., B. Banapour, and J. Levy. 1985. Characterizationof the AIDS-associated retrovirus reverse transcriptase andoptimal conditions for its detection in virions. Virology 147:326-335.

35. Jackson, G., D. Paul, L. Falk, M. Rubenis, J. Despotes, D. Mack,M. Knigge, and E. Emeson. 1988. Human immunodeficiencyvirus (HIV) antigenemia (p24) in the acquired immunodeficiencysyndrome (AIDS) and the effect of treatment with zidovudine(AZT). Ann. Intern. Med. 108:175-180.

36. Jackson, J., K. Sannerud, F. Rhame, and H. Balfour, Jr. 1988.Evaluation of two commercial tests for human immunodefi-ciency virus antigens in culture supernatant fluid. Am. J. Clin.Pathol. 89:788-790.

37. Jackson, J., K. Sannerud, F. Rhame, R. Tsang, and H. Balfour.1987. Comparison of reverse transcriptase assay with the Retro-Tek viral capture assay for detection of human immunodefi-ciency virus. Diagn. Microbiol. Infect. Dis. 7:185-192.

38. Kenny, C., J. Parkin, G. Undershill, N. Shah, B. Burell, E.Osborne, and D. Jefferies. 1987. HIV antigen testing. Lanceti:565-566.

39. Kessler, H., B. Blaauw, J. Spear, D. Paul, L. Falk, and A.Landay. 1987. Diagnosis of human immunodeficiency virusinfection in seronegative homosexuals presenting with an acuteviral syndrome. J. Am. Med. Assoc. 258:1196-1199.

40. Kitchen, L. W., F. Barin, J. L. Sullivan, M. F. McLane, D. B.Brettler, P. H. Levine, and M. Essex. 1984. Aetiology of AIDS-antibodies to human T-cell leukemia virus (type III) in hemo-philiacs. Nature (London) 312:367-369.

41. Kruppenbacher, J., T. Mertens, and H. Eggers. 1988. Reproduc-ibility of HIV antigen test. Lancet i:298.

42. Lange, J., and J. Goudsmit. 1987. Decline of antibody reactivityto HIV core protein secondary to increased production of HIVantigen. Lancet i:448.

43. Lange, J., D. Paul, H. Huisman, F. DeWolf, H. Van Den Berg, R.Coutinho, S. Danner, J. Noordaa, and J. Goudsmit. 1986.Persistent HIV antigenaemia and decline of HIV core antibodiesassociated with transition to AIDS. Br. Med. J. 293:1459-1462.

44. Lee, M., K. Sano, F. Morales, and D. Imagawa. 1988. Compa-rable sensitivities for detection of human immunodeficiencyvirus by sensitive reverse transcriptase and antigen captureenzyme-linked immunosorbent assays. J. Clin. Microbiol. 26:371-374.

45. Lefrere, J., D. Salmon, A. Courouce, P. Lambia, J. Fine, C.Doinel, and C. Salmon. 1988. Evolution towards AIDS inHIV-infected individuals. Lancet i:1220-1221.

46. Levy, J. A., and J. Shimabukuro. 1985. Recovery of AIDS-associated retroviruses from patients with AIDS or AIDS-related conditions from clinically healthy individuals. J. Infect.Dis. 152:734-738.

47. Mayer, K., L. Falk, D. Paul, G. Dawson, A. Stoddard, J.McCusker, S. Saltzman, M. Moon, R. Ferriani, and J. Groop-man. 1987. Correlation of enzyme-linked immunosorbent assaysfor serum human immunodeficiency virus antigen and antibod-ies to recombinant viral proteins with subsequent clinical out-comes in a cohort of asymptomatic homosexual men. Am. J.Med. 83:208-212.

48. McDougal, J. S., S. P. Cort, M. S. Kennedy, C. D. Cabridilla,F. M. Feorino, D. P. Francis, D. Hicks, U. S. Kalyanraman, andL. S. Martin. 1985. Immunoassay for the detection and quanti-tation of infectious human retrovirus, lymphadenopathy-associ-ated virus (LAV). J. Immunol. Methods 76:171-183.

49. Mok, J., A. DeRossi, A. Ades, C. Giaquinto, I. Grosch-Worner,and C. Peckham. 1987. Infants born to mothers seropositive forhuman immunodeficiency virus. Lancet i:1164-1167.

50. Moss, A., P. Bacchetti, D. Osmond, W. Krampf, R. Chaisson, D.Stites, J. Wilber, J.-P. Allain, and J. Carlson. 1988. Seroposi-tivity for HIV and the development of AIDS or AIDS relatedconditions: three year follow-up of the San Francisco GeneralHospital cohort. Br. Med. J. 296:745-750.

51. Newell, A., M. Malkovsky, D. Orr, D. Taylor-Robinson, and A.

Dalgleish. 1987. Antigen test versus reverse transcriptase assayfor detecting HIV. Lancet ii:1146-1147.

52. Paul, D. A., L. Falk, H. Kessler, R. Chase, B. Blaauw, D.Chudwin, and A. Landay. 1987. Correlation of serum HIVantigen and antibody with clinical status in HIV-infected pa-tients. J. Med. Virol. 22:357-363.

53. Pedersen, C., C. Mollernielsen, B. Vestergaard, J. Gerstoft, K.Krogsgaard, and J. Nielsen. 1987. Temporal relation of antige-naemia and loss of antibodies to core antigens to development ofclinical disease in HIV infection. Br. Med. J. 295:567-569.

54. Perkovic, D., J.-P. Chausseau, N. Lapointe, J. Michaud, S.Garzon, C. Strykowski, C. Tsoukas, N. Gilmore, H. Goldman,M. Gornitsky, M. Popovic, and J.-M. Dupuy. 1986. Detection ofHTLV-lII/LAV antigens in peripheral blood lymphocytes frompatients with AIDS. Arch. Virol. 91:11-19.

55. Perkovic, D., M. Gornitsky, D. Ajdutrovic, J. M. Dupuy, J. P.Chausseau, J. Michaud, N. Lapointe, N. Gilmore, C. Tsoukas, G.Zwadle, and M. Popovic. 1987. Pathogenicity of HIV in lym-phatic organs of patients with AIDS. J. Pathol. 152:31-35.

56. Popovic, M., M. G. Sarngadharan, E. Read, and R. C. Gallo.1984. Detection, isolation, and continuous production of cyto-pathic retroviruses (HTLV-III) from patients with AIDS andpre-AIDS. Science 224:497-500.

57. Ranki, A., M. Krohn, J.-P. Allain, G. Franchini, S.-L. Valle, J.Antonen, M. Leuther, and K. Krohn. 1987. Long latency pre-cedes overt seroconversion in sexually transmitted human im-munodeficiency virus infection. Lancet ii:589-593.

58. Reesnik, H. W., P. N. Lelie, J. G. Huisman, W. Schaasberg, M.Gonsalves, C. Aaij, I. N. Winkel, J. A. Van Der Does, A. C.Hekker, J. Desmyter, and J. Goudsmit. 1986. Evaluation of sixenzyme immunoassays for antibody against human immunode-ficiency virus. Lancet ii:483-486.

59. Reiss, P., J. Lange, C. Boucher, S. Danner, and J. Goudsmit.1988. Resumption of HIV antigen production during continuouszidovudine treatment. Lancet i:421.

60. Rey, M. A., B. Spire, D. Formont, F. Barre-Sinoussi, L. Mon-tagnier, and J. C. Chermann. 1984. Characterization of the RNAdependent DNA polymerase of a new human T-lymphotropicretrovirus (lymphadenopathy associated virus). Biochem.Biophys. Res. Commun. 121:126-133.

61. Rollag, H., S. Evensen, S. Froland, and A. Glomstein. 1987. Thesignificance of HIV antigen and HIV antibodies specific againstenvelope and core proteins in serum of HIV-infected hemophil-iacs. Eur. J. Haematol. 39:353-355.

62. Sarngadharan, M., L. Bruch, M. Popovic, and R. Gallo. 1985.Immunological properties of the Gag protein p24 of the acquiredimmunodeficiency syndrome retrovirus (human T-cell leukemiavirus type IlI). Proc. Natl. Acad. Sci. USA 82:3481-3484.

63. Spira, T., L. Bozeman, R. Holman, D. Warfield, S. Phillips, andP. Feorino. 1987. Micromethod for assaying reverse tran-scriptase of human T-cell lymphotropic virus type III/lymphad-enopathy-associated virus. J. Clin. Microbiol. 25:97-99.

64. Steckelberg, J., and F. Cockerill. 1988. Serologic testing forhuman immunodeficiency virus antibodies. Mayo Clin. Proc.63:373-380.

65. Steimer, K. S., G. Van Nest, D. Dina, P. J. Barr, P. A. Luciw,and E. T. Miller. 1987. Genetically engineered human immuno-deficiency envelope glycoprotein gpl20 produced in yeast is thetarget of neutralizing antibodies, p. 236-241. In R. Chanock(ed.), Vaccines 87. Cold Spring Harbor Laboratories, ColdSpring Harbor, N.Y.

66. Tenner-Racz, K., P. Racz, M. Bofill, A. Schultz-Meyer, M.Dietrich, P. Kern, J. Weber, A. Pinching, F. Veronese-Dimarzo,M. Popovic, D. Klatzmann, J. Gluckman, and G. Janossy. 1986.HTLV-Ill/LAV viral antigens in lymph nodes of homosexualmen with persistent generalized lymphadenopathy and AIDS.Am. J. Pathol. 123:9-15.

67. Van de Perne, P., A. DeClerco, J. Cogniaux-Leclerc, D.Nzaramba, J. P. Butzler, and S. Sprecher-Goldberger. 1988.Detection of HIV p17 antigen in lymphocytes but not epithelialcells from cervicovaginal secretions of women seropositive forHIV: implications for heterosexual transmission of the virus.Genitourin. Med. 64:30-33.



http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/


68. Viscidi, R., H. Farzadegan, F. Leister, M. Francisco, and R.Yolken. 1988. Enzyme immunoassay for detection of humanimmunodeficiency virus antigens in cell cultures. J. Clin. Micro-biol. 26:453-458.

69. Von Sydow, M., H. Gaines, A. Sonnerborg, M. Forsgren, P.Pehrson, and 0. Strannegard. 1988. Antigen detection in pri-mary HIV infection. Br. Med. J. 296:238-240.

70. Wall, R., D. Denning, and A. Amos. 1987. HIV antigenaemia inacute HIV infection. Lancet i:566.

71. Ward, J., S. Holmberg, J. Allen, D. Cohn, S. Critchley, S.Kleinman, B. Lenes, 0. Ravenholt, J. Davis, M. Quinn, and H.Jaffe. 1988. Transmission of human immunodeficiency virus(HIV) by blood transfusions screened as negative for HIVantibody. N. Engl. J. Med. 318:473-478.

72. Watson-Williams, E. J., J. L. Yee, J. R. Carlson, S. C. Mertens,

P. Holland, L. Sinor, and F. V. Plapp. 1988. Solid phase red celladherence immunoassay for anti-HIV-1: a simple, rapid andaccurate method for donor screening. Transfusion 28:184-186.

73. Wigdahl, B., R. Guyton, and P. Sarin. 1987. Human immuno-deficiency virus infection of the developing human nervous

system. Virology 159:440-445.74. Wittek, A. E., M. A. Phelan, M. A. Wells, L. K. Vujcic, J. S.

Epstein, H. C. Lane, and G. V. Quinnan. 1987. Detection ofhuman immunodeficiency virus core protein in plasma by en-

zyme immunoassay. Ann. Intern. Med. 107:286-292.75. Zolla-Pazner, S., D. William, W. El-Sudr, M. Marmov, and R.

Stahl. 1984. Quantitation of P2-microglobulin and other immunecharacteristics in a prospective study of men at risk for acquiredimmune deficiency syndrome. J. Am. Med. Assoc. 251:2951-2955.

VOL. 2, 1989


http://cmr.asm

.org/D

ownloaded from

http://cmr.asm.org/

antigen detection for human immunodeficiency virusantigen detection for hiv 243 table 1-continued...

Documents