Bridging the Gap: Conference Record [Abstract book, International Conference on AIDS (12th: 1998: Geneva, Switzerland)]

764 Abstracts 41107-41111

12th World AIDS Conference

Method: Sensitivity of Murex HIV-1.2.0 was assessed by comparing data for 20
seroconversion series, over 380 HIV-1 samples, over 200 HIV-2 samples and 59
HIV-O samples with that obtained using other assays. Comparative performance
of the assay was also assessed using 132 subtyped HIV-1 group M samples
(subtypes A, B, C, D, E, F) including 7 samples which could not be subtyped and
6 samples of mixed subtype. Specificity of the assay was assessed using 36,856
HIV-negative samples.
Results: Sensitivity of Murex HIV-1.2.0 was found to be at least equal to
the best available alternative assays on all groups of HIV positive samples (Seroconvertors, HIV-1/M, HIV-1/0 and HIV-2). Unlike alternative assays, Murex
HIV-1.2.0 gave signals of greater than 75% of the maximum recordable signal
with all 132 subtyped samples, demonstrating the excellent performance of the
assay on such HIV-1 group M variants. Specificity of Murex HIV-1.2.0 was also
found to be excellent. Murex HIV-1.2.0 detected 33/36,856 HIV-negative samples
as repeatable false positives, giving a specificity of 99.91%. This specificity is
equal to or significantly better than the routine specificities of the most sensitive
alternative assays available.
Conclusion: A new assay, Murex HIV-1.2.0, has been developed to detect
antibodies to HIV. The assay has class-leading: sensitivity to HIV-1 (including
group O samples and group M variants) and HIV-2; specificity; and in process
security features.
41107 Detection of HIV-1 subtypes by a new rapid
immunochromatographic test
Hiroyasu Arai1, S. Devare2, T. Kosaka1, K. Takeda1. 1Dainabot C., Ltd. 344
Minoridai, Matsudo, Chiba, 271, Japan; 2Abbot Lab., North Chicago, IL, USA
Objectives: A new immunochromatographic rapid test (Determine'm HIV-1/2) has
been developed for detection of antibodies to HIV-1 and HIV-2 in human whole
blood, serum, and plasma. The study objective was to demonstrate the excellent
performance of this test using a variety of HIV-1 subtype (A-G and group-O)
samples.
Method: This test uses a nitrocellulose strip with a capture site containing
HIV-1, HIV-2, and Group-O specific antigens. On the strip HIV-1, HIV-2, and
Group-O specific antigens are conjugated to Selenium colloid (Se). If a sample
contains anti-HIV-1 or anti-HIV-2 antibodies, the antibodies first react with the
antigen-Se colloid. As the antibody-antigen-Se complexes flow past the capture
site, the antibodies react with the antigens at the site forming a visible red line. In
the absence of anti-HIV-1 or anti-HIV-2 antibodies in the sample, no visible red line
forms at the capture site. The test also contains a procedural control site which
confirms assay validity by the formation of a visible red line. For serum/plasma
samples, 50 ul is placed on the sample application pad. For whole blood sample,
50 ul is placed on the pad followed by one drop of buffer. Results are evaluated
after 15 minutes.
Results: The test showed 100% sensitivity in testing of serum/plasma samples with antibodies to HIV-1 subtype A (113/113), subtype-B (31/31), subtype-C
(5/5), subtype-D (24/24), subtype-E (10/10), subtype-F (7/7), and subtype-G (3/3).
Sensitivity in HIV-2 (174/174) and Group-O (12/12) assays were also 100%. In
726 HIV negative serum/plasma and 400 HIV negative whole blood samples,
specificity were 99.7% and 100%, respectively.
Conclusion: This evaluation demonstrated the excellent performance of this
immunochromatographic test. The test detected all known HIV-1 subtypes, HIV-2
and Group-O samples. Unique features of the test include rapid results, ease of
use, whole blood capability and room temperature storage. We conclude that this
test is very suitable for developing countries.
41108 Validation study of HIV testing in urine samples
David Mabey1, John Changalucha2, B. West3, D. Ross3, R. Balira2, J. Todd3,
T. Boerma4. 1London Sch of Hygiene+Trop Medicine Keppel Street London
WCIE 7HT, England; 2NIMR, Mwanza, Tanzania; 3NIMR+LSHTM,
Mwanza+London, Tanzania+UK; 4KIT Amsterdam, Netherlands
Background and Objectives: Large-scale surveys of HIV status require noninvasive methods of specimen collection to maintain high compliance. It is possible
to test for HIV antigen in urine and saliva and reports on these tests give good
performance levels. This study was undertaken to see if it was feasible to use a
urine HIV test, GACPAT, for a population study in northern Tanzania, by comparing to HIVELISA on blood spot eluates and GACELISA on urines from the same
population.
Methods: Samples of urine and blood spots were collected from adults in a
population survey in Mwanza Region, Tanzania. Blood spot eluates were tested
by Vironostika HIV-Uniform II (Organon Teknika) and by Enzygnost Anti HIV1/1
(Behring Diagnostics). Urine specimens from 74 individuals with a positive blood
spot were tested with both GACPAT and GACELISA. GACPAT was performed on
214 urine specimens (50%) randomly selected from those with a negative result
and GACELISA was purposive performed on 117 of these 214 urines.
Results: For 288 samples with both a urine GACPAT and blood spot result, both
tests gave a positive result in 72 cases, with 4 discrepant results. Compared to

Blood spot, the GACPAT assay had a sensitivity of 97.3% (95% Cl: 91.3%-99.5%)
and specificity of 99.1% (95% Cl: 96.9%-99.8%). In the 191 subjects with both
urine GACELISA and blood spot result, 74 were positive by blood spot and 73 by
GACELISA, with 5 discrepant results. Compared to blood spot, the GACELISA
assay had a sensitivity of 96.0% (95% Cl: 89.4%-99.0%) and specificity of 98.3%
(95% CI; 94.5%-99.8%). The comparability of GACPAT versus GACELISA on
urine was excellent (kappa = 0.967).

Conclusion: There was excellent agreement between GACPAT and
GACELISA HIV tests on urine samples. The sensitivity and specificity of these
HIV tests on urine specimens are adequate for estimating the prevalence of
HIV on anonymized specimens in population surveys. However, neither test has
adequate sensitivity to be used as the basis for the screening of the HIV status
of individuals who will be given their HIV test result. This still requires a blood
specimen.
41109 | Predictors of HIV infection among professional blood
donors with Western Blot indeterminate results
Usha K. Baveja1, D. Chattopadhya2, R.K. Aggarwal3, V. Doda4, S. Kumari5,
J. Sokhey6. 1Consultant, AIDS NICD, DII/28 West Kidwai Kidwai Nagar, New
Delhi 110023; 2Joint Director, NICD, 3Assistant Director, NICD, 4R.M.L., New
Delhi; 5W.H.O. Searo, New Delhi; 6Director, NICD, Delhi 110054, India
Objectives: To find out if an epidemiological or serological characteristic recorded
at the time of initial HIV testing could be considered as a predictor for HIV-1 infection in Western Blot indeterminate (WBI) specimens and to evaluate testing
strategies for resolving WBI results.
Design: Prospective, among professional blood donors (PBD) with WBI result.
Methods: 72 HIV sero-reactive and 66 non-reactive PBD were subjected to
study of epidemiological profile, STD markers and follow up HIV-1 testing 6, 12, 24
and 48 weaks after to detect any seroconversion. Initial and follow up specimens
with WBI results were subjected to two EIAa as well as to line immunoassay
(LIA).
Results: Among the PBD HIV sero-reactive and WBI 90.3% were unmarried,
75% were heterosexually promiscuous, 36.1% had visited STD clinics compared
to the WBI and HIV non reactive donors (72.7%, 42.4% and 12.1% respectively;
P < 0.001). Prevalence of antitreponemal antibodies was higher in the former
group (16.7%) compared to latter group (1.5%, P < 0.002). Four out of 55 (7.3%)
PBD HIV seroreactive with WBI result seroconverted. LIA was found to be more
reliable test compared to combination of EIAs to determine HIV-1 infection in WBI
results at initial testing.
Conclusion: The history of heterosexual promiscuity, prevalence of STD markers, high OD values in screening EIA, 'P24' only pattern of bands in WB and
positivity by LIA have high predictive value for HIV-1 infection in specimens
showing WBI pattern of results at initial testing.
41110 1 HIV antibody avidity measurement: How to distinguish
primoinfected from old infected patients using one ELISA
test adapted to AXSYM automate
Christopher Payan1, A. Lemeur1, B. Evreur1, S. Kouyoumdjian1,
J.-M. Chennebault2, F. Lunel. 1 Laboratoire de Virologie Chu Angers, Angers;
2Service de Maladies Infectieuses Chu Angers, France
Objective: To determine the avidity of total HIV antibody (ab) in newly seropositive
patients.
Methods: 25 seropositive patients have been tested for HIV antibody avidity
with urea 12 M. 5 were known primoinfected patients (PI, less than 1 month), 4
recent infections (RI, less than 6 months), 8 known old infections (01, more than 1
year), and 8 unknown (UK). Dilution of serum (2 to 100), optimal concentration of
urea (8 to 16 M) and incubation time (5 to 60 min) with urea have been optimized
for HIV antibody testing upon AXSYM automate using HIV1-2 AXSYM kit (Abbott,
France). Some known PI and 01 samples were also confirmed using an ELISA
microplate system: Vironostika HIV Uniformll plus O (Organon Teknika). Each
urea treated sample was tested together with the non urea treated sample at the
same dilution in PBS buffer. Avidity index (Al) was calculated from the ratio of
urea signal to non urea signal and expressed in percentage.
Results: We found that a 5-diluted serum (40 f1 of serum) in urea at 12 M, during 10 min before testing in AXSYM gave similar results than a 21-diluted serum
in the ELISA after washing with urea at 8 M. The test was highly reproducible
with a standard variation of 5 to 14% upon 10 samples tested 3 times. The range
of Al in PI patients was 25 to 45% (35 + 6.3), in RI 60 to 70% (64 ~ 3.8), in
01 86-100% (95 + 5.6) and in UK 94-100% (95 ~ 2.0). Kinetics in 3 of the PI
patients showed an increase of Al from 28 to 94% in about 6 to 9 months. 62 to
68% Al were found around 3 to 4 months after contamination.
Conclusion: The HIV antibody avidity assay in the AXSYM automate may
help for the diagnosis of prime or recent infection in a newly seropositive patient,
compared to an old infected patient. The limit of 50% for PI and 70% for RI seems
reasonable. This test is very reproducible, easy to do and very quick (one sample
in 15 min).
41111| Identification of a new virus-associated epitope inducing
antibody responsible for the non-progression to AIDS
Pascale Galea1, C.R. Steinhart2, P. Willem3, N. Profizi4, C. Lecontel5,
J.C. Chermann5. 1 nserm Unite 322-Campus de Luminy-BP 33 163 Av. de

Luminy-13273 Marseille Cedex 09, France; 2Mercy Hospital, Miami, USA;
3South African Institute for Medical Research, Parktown, South Africa; 4Centre
Hospitalier Intercommunal, La Seyne Sur Mer; 5lnserm Unite 322, Marseille,
France
Background: the aim of the study was to identify a marker of non-progression
to AIDS in HIV-infected patient. By using R7V (a novel epitope present on HIV-1
surface and involved in infection) as a target in ELISA, we screened patients'