Abstracts Vol. 1 [International Conference on AIDS (11th: 1996: Vancouver, Canada)]

Track B: Clinical Science

Tu.B.2169 - Tu.B.2174

simultaneously by standard serology (HIV p24 Ag) and DNA-HIV-PCR.The use of
Polymerase Chain Reaction (cellular HIV-DNA) to monitor HIV seronegative subjects does
not seem to be warranted nor cost/effective.
Francesco Castelli, MD, CI nica di Malattie Infettive e Tropicali, University of Brescia (Italy)
P le Sped all Civili n. 1-25124 Bre Scia (Italy)
Tel. + 39.30.394467 Fax [phone]
Tu.B.2169
EARLIER DIAGNOSIS OF HIV INFECTION BY SIMULTANEOUS DETECTION OF HIV
ANTIGEN AND ANTIBODY IGG TO HIV WITH ULTRASENSITIVE ENZYME
IMMUNOASSAY
Hashida S, Hashinaka K, Nishikata I, Ishikawa E. Department of Biochemistry, Miyazaki
Medical College, Kiyotake, Miyazaki 889- 16, Japan
Objective: To shorten the window period after- HIV infection, during which HIV antibodies
are not detected, hampering the diagnosis of HIV infection by conventional methods for
HIV antibodies.
Methods: Serum samples of ten HIV- I seroconversion serum panels were tested by an
ultrasensitive enzyme immunoassay (immune complex transfer enzyme immunoassay=
ICTEIA) for simultaneous detection of p24 antigen of HIV-I and antibody IgGs to reverse
transcriptase (RT) and p17, and the results were compared with those by conventional
methods.
Results: Signals by ICTEIA for simultaneous detection of p24 antigen and antibody IgGs to
RT and py17 became positive 6 to 42 days (two weeks on an average) earlier than the
detection of antibodies to HIV- I by conventional methods and remained strongly positive
even after levels of p24 antigen declined. By contrast, signals for p24 antigen alone became
positive 2 weeks (an average) earlier than those for antibodies to HIV-I but declined as levels of antibodies to HIV I rose, and signals for antibodies to HIV-I became positive only 2
weeks (an average) later than those for p24 antigen.
Conclusions: The simultaneous detection of p24 antigen and antibody IgGs to p l7 and RT
made possible both as early a diagnosis of HIV- I infection as the appearance of p24 antigen
in the circulation, shortening "the window period", and as reliable a diagnosis of the infection as that by the detection of antibodies to HIV I from the time of seroconversion until
late stages of the infection, since the serum level of antibody IgG to RT was high not only in
asymptomatic carriers but also in patients with AIDS-related complex and AIDS.
Dr Seiichi Hashida, Department of Biochemistry Miyazaki Medical College, Kiyotake,
Miyazaki 889 I 16, Japan Telephone:[phone] Fax:[phone] I
Tu.B.2170
ABC-WESTERN BLOT METHOD TO CONFIRM THE ANTIBODY AGAINST HIV- I IN
URINE SAMPLES
Ohya Hitomi*,Tsukano K, Ichikawa S*, Kawata K", Itoh A", Soda K*. *Kanagawa
Prefectural College of Nursing and Medical Technology, r "Yokohama City University School
of Medicine, Japan
Objectives:To improve the sensitivity of western blot method for confirming the antibody
against HIV-I in urine samples.
Methods: Thirty fiv e-pair specimens of serum and urine were collected from Hr IV I positive
patients.We used the ELISA method (Calypte HIV I test, Ohtsuka) for screening of antibody to HIV- I in unrine samples, and western blot method (LAV Blot I, Pasteur) to confirm.
Avidin Biotin complex kit (ABC kit,Vectastain) was used for improvement of western blot
method.
Results: On testing with usual western blot method, 30(85.7%) of 35 urine samples collected from patients with HIV- I were confirmed as positive. The remaining 5 urine samples
were indeterminate, though sera obtained from same patients were confirmed by usual
western blot test. Positive rates of antibody to GPI160, GP120 and GP4 I in urine sanmples
were 97.1%, 60% and 42.9% respectively and lower than those in serum specimens. By
using the ABC western blot method, the confirmative rate on urine samples increased to
100%. Anti-GPI60 band could detect in all urine samples.The positive rates of anti-GPI20
and anti GP4I in urine samples also improved as compared with the usual western blot
method, but were not similar to the results obtained from the serum specimens.
Discussions: The ABC-western blot method could confirm the HIV- I antibody in urine
samples. The mentioned results suggested that urine sample was useful for detection of
antibodies against HIVI, though it is necessary to improve the sensitivity of detection for
using the diagnosis.
Hitomi Ohya, Kanagawa Prefectural College of Nursin g and Medical Technology 50-I.,
Nakaocho,Asahi-ku,Yokohama 241,JapanTel: 81 45 361-6141 Fax: [phone]
Tu.B.2171
FACTORS ASSOCIATED WITH REPEATED HIV TEST IN MEXICO CITY.
del Rio Chiriboega Carlos, Uribe Z. P,Varela T C. Hernairdez T. C. CONASIDA (National
AIDS Council). Mexico.
Objective: To characterize the population that had more than one HIV test done asni well as
the reasons and factors related with this behavior.
Methods: We applied a questionnaire to all persons that returned for an HIV test in a two
year period (January 1994 to December 1995) at CONASIDA's AIDS Testng Centers in
Mexico City Descriptive analysis was done on sociodemo,raphic data, sexual behavior and
risk factors assocated with HIV infection.
Results: We analyzed 721 patients with more than one HIV test in a two year period,
69.6% were men and 30.4% women. 69.3% (n=500) have been tested twice, 16.4%
(n= 120) three times, 4.3% (n-3 ) four and 9.7% (n-70) five or more. 36.6% were men
having sex with men, 44.3% heterosexuals and 19% bisexuals.The main reasons related for
repeated HIV testing were:

Conclusions: 3 he main reasons for repeating an HIV test were: I. lack of condom use
despite adequate information, particularly among bisexual men, 2. Having an HIV infected
partner and 3 anxiety of being infected, particularly among homosexual ien and heterosexuals who had multiple sex partners and inconsistent condom use Educationl strategies
for bisexual men must consider the fact that anxiety of being infected is not present even
though they know they are at risk because they don't use any protective methods.
del Rio Chriboga Carlos, Calz. de Tlalpan 4585 do Piso. Col. Toriello Guerra, Deleg. -lalpan
Mexico City Mexico (5-25) 528.4086/528.4848 Fax: [phone]
Tu.B.2172
HIV-I SUBTYPING IN THAILAND USING GENOMIC AND SEROLOGIC
TECHNIQUES
Gaywee, lariyanarat*, Artenstein AW**, VanCott TC",Trichavaro R*, Suk c hamnnong A,
Amlee P*, Souza MS*, McCutchan FEY, *Carr JK, ***Markowitz L, Michael RA ',
Nitayaphan 5*. *AFRIMS, Bangkok,Thailand; ' VVRAIR, Rockville, MD:y + hHMJFAMM,
Rockville, MD
Objectives: To determine HIV I subtypes in infected Thai subjects using heteroduplex
mobility assay (HMA), differential polymerase chain reaction (PCR) and peptide enzymelinked immunosorbent assay (V3-ELISA); and to compare the sensitivity and specificity of
differential PCR and V3-ELISA using HMA as the reference standard
Methods: Pasired peripher al blood mononuclear cells (PBMC) and sera were collected from
38 HIV antibody-positive subjects enrolled in a natural history study in Thailand. HIV.subtyp
ing was conducted by differential nested PCR and HMA using DNA extracted from PBMC,
and byV3-ELISA using sera; differential PCR employs primer pairs which distinguished HIV
subtype E from subtype B.TheV3-ELISA utilises peptides specific for HIV subtypes E and B.
Results: HMA subtyping classified 32 (84%) cases as being infected with HIV subtype F and
6 with subtype B (16%). All samples were subtyped by differential PCR and there was 100%
concordance with the HMA. Compared to the molecular virologic assays, the V3 ELISA correctly predicted 75% (24/32) of samples as be n, subtype E and 100% (6/6) as subtype B.
Six samples that molecularly subtyped as E were repeatedly dually reactive by serotyping
and 2 were serologically nontypable due to overall low levels of V3 antibody.
Conclusions:The predominant HIV subtype in the cohort is subtype E. Differential PCR
was the more specific and sensitive for HIV subtyping than the V3-El ISA. However; serotyp
ing is less rigorous with respect to sampling requirements, specimen processing and techni
cal requirements; thus appearing more practical for use in a "fi eld" environment. Both assays
are suitable in areas such as Thailand, where to date, only two HIV subtypes are reported
to co- circulate. Differential PCR should be used to complement the V3-FL ISA.- Ihe broader
subtyping range of HMA allows its use as a confrmatory and surveillance technique in
instances where neither differential PCR nor the V3-ELISA gives ai definitive result.
Jariyanarat Gaywee. Dept. of Retrovirology AFRIMS, 315/6 Ravithi Road,. Bangkok I0400,
Thailand Telephone: [phone], Fax: [phone]
Tu.B.2173
SENSITIVITY AND SPECIFICITY OF PCR, IGA AND P24 ANTIGEN FOR EARLY DIAGNOSIS OF HIV-I INFECTION IN BREASTFED INFANTS
Coberly Jacquelinem, Ruff A*, Kline R*, Desormeaux J* Quinn T', Kacergis J*, Sunnmer S
Halsey N. *Johns Hopkins University SHPH, Baltimore, MD, USA; C**enters for
Development and Health, Port au Prince, Haiti
Objectives: To determine the sensitivity and specificity of PCR, IgA and p24 antigen test
performed on cord blood and at I, 3, and 6 months of age for detecting HIV infection in
breastfed infants.
Methods: HIV I seropositive women were identified at delivery and their infants were fol
lowed prospectively for up to 18 months. Cord blood and infant blood dr awn at 1,3 iand 6
months was tested for HIV I: DNA by PCR, IgA by Western blot, aind p24 antigen by EIA
following acid hydrolysis. Infants were designated HIV I infected if they had a positive HIVI
IgG Wb at 15 to 18 mos; 2 positive PCRs at >1 mo of age with no subsequent negative
tests,or a I positive PCR at> Imo of age and a positive IgA or p24 at>3 os of aIge with
no subsequent negative tests.
Results: 368 woman/infant pairs were enrolled and HIV infection status was definitively
determined in 258.The specificity of each test exceeded 90% at each age. Sensitivity lof
PCR averaged 80%.The sensitivity of IgA and p24 increased to a peak of 63% and 57%
respectively at 6 mos. Of uninfected infants, 10% had >1 positive PCR. 3% had 2 positive
IgA and I I% had > positive p24 test.
Conclusions: Sensitivity varied by age.The specificity of these tests was high but false posi
tives did occunr.The apparent false positive tests could have been due to transient infection.
None of these tests alone can be reliably used at a singe age for early diagnosis of HIV
infection.
JS Coberly Johns Hopkins Univ., 6 15 N Wolfe Street #55 15, Baltimore, MD. USATelephone:
410 955 7767 Fax: [phone] email: icoberly)phnet.sph.jhu.edui
Tu.B.2174
FIRST CHECKTM HIV I-2,A TWO-STEP, FIVE MINUTE,WHOLE BLOOD
IMMUNOCHROMATOGRAPHIC ASSAY TO DETECT HIV 1/2 ANTIBODIES
See, Darryl M*, Tilles JG*, Flynn T*, Evanoff PA *, Morris HT. University of Calfornlia,
Irvine, Irvine, CA, USA "Worldwide Medical Corporation, Irvine, CA, USA
Objective: To evaluate a new, two step, five minute, immunoassay device which detects HIV
I/2 antibodies in whole blood, serum or plasma.
Methods: A) Samples: Ninety eight HIV positive patients were enrolled in liii il iI
evaluate a rapid immunoassay test. Each patient consented to the collection of whole
blood, serum and plasma specimens. In separate studies, 100 plasma sanples fom IIV neg
ative individuals collected at a regional blood center and a commercial panel continn 7
HIV 2 specimens fom AfSnca were evaluated. B) Analysis: One drop (40u ) of sample i
added to a hand-held test device, followed by fhe addition of 3 drops (50uL) of deve op
ment buffet.Within 5 minutes, the sample combines with colloidal gold particles and inmm
bilizes HIV 1/2 antigens to form a distinct red line in a test window A second window acts
as a "procedural control ". A single control line with no line in the test window is regarded
as "nonreactive".Two lines are regarded as "reactive", and a positive test line and negative
control line as "indeterminant".

so
o
Oen
0
u
C
<
c
O
I)
isa
i._
0
c
O
u
C
a)
c
0
C
a4
2
294

Factor

Homosexual men Bisexual men Heterosexuals
n=264 n=137 women(219) & men(101)

No condom use
HIV infected partner
Anxiety
Occupational risk
Without risk

38.6% (102)
27.3%(72)
34% (90)

69.3% (95)
30.6% (42)

32.8% (105)
34.3% (1 10)
27.8% (89)
1.9% (6)
3.1% (10)