Abstract Book Vol. 2 [International Conference on AIDS (14th: 2002: Barcelona, Spain)]

XIV International AIDS Conference

Abstracts WeOrB1338-WeOrB1341 31

Presenting author: Giorgio Barbarini, Clinic of Infectious and Tropical Diseases,
IRCCS S.Matteo University of Pavia, Via Taramelli, 27100 PAVIA, Italy, Tel.: [phone], Fax: [phone], E-mail: [email protected]
SWeOrB338 iQuantitation of HIV-1 subtypes by the VIDAS
PROBE qHIV test
P.J. Levasseur, D.W. Eustace, J.M. McCarty, J.L. Burg. bioMerieux, Inc.,
Rockland, United States
Background: The VIDAS PROBE qHIV test is currently under development for
determining HIV-1 viral loads. As this test has been anchored to the VQA subtype
B standard, the purpose of this study was to determine the sensitivity, reproducibility and quantitative accuracy of HIV-1 non-B subtypes.
Methods: Nineteen HIV-1 culture isolates, representing Group M subtypes A, B,
C, D, E, F, G and H and Group O (Boston Biomedica, Inc.), were serially diluted
to nominal values of 10000, 1000, 100, and 50 RNA copies/mL based on electron
microscopy (EM). HIV-1 RNA was isolated by specific target capture onto paramagnetic particles, amplified by automated TMA (transcription-mediated amplification) on the AMPstation instrument, and detected by an automated protocol on
the VIDAS instrument. A replicate size of 5 was tested for each subtype at each
level (2 operators, 3 days).
Results: Combining the results for all subtypes, the VIDAS PROBE qHIV test
showed 98.8% detection at 50 RNA copies/mL and 100% detection at >= 100
RNA copies/mL. The only nondetectable result occurred with one Group O sample where the other four replicates gave values of 18, 20, 16 and 19 RNA
copies/mL. The average precision, determined by the standard deviation of log10
transformed data, was 0.16, 0.20, 0.13 and 0.13 for inputs of 50, 100, 1000 and
10000, respectively. In general, quantitative accuracy was within 0.5 log10 of the
expected EM values although one C and two F subtypes showed systematic
quantitative bias. However, compared to expected values averaged from EM plus
two molecular methods, only the one C subtype had an average log10 difference
> +0.5 log10 (-0.59 log10).
Conclusions: These preliminary results indicate that the VIDAS PROBE qHIV
test under development has excellent sensitivity, precision, linearity, and accuracy for all major HIV-1 subtypes. In particular, the four Group O samples showed
similar results as the various Group M subtypes.
Presenting author: Larry Burg, bioMerieux, Inc., 1022 Hingham St., Rockland,
MA 02370, United States, Tel.: [phone], Fax: [phone], E-mail:
larry.burg @ na.biomerieux.com
WeOrB1 339 Use of dried blood spots for diagnosis and viral
quantification in perinatal HIV-1 infection in
Thailand
T Chaowanachan1, K. Chokephaibulkit2, T. Chotpitayasunondh3,
P. Wasinrapeel, A. Teeraratkul 1, B. Jetsawang 1, K. Neeyapun',
R.J. Simonds4, N.L. Young1. ' Thai MOPH-US Centers for Disease Control and
Prevention Collaboration, HIV/AIDS Collaboration, Dept Med. Sciences, Bldg. 2,
2nd Floor, Ministry of Public Health, Tivanon Road, Nonthaburi 11000, Thailand;
2Siriraj Hospital, Mahidol University, Bangkok, Thailand; 3Queen Sirikit National
Institute of Child Health, Bangkok, Thailand; 4Centers for Disease Control and
Prevention, Atlanta, United States
Background: To validate the use of field-site collected filter paper dried blood
spot (DBS) samples for HIV-1 diagnosis and viral load (VL) monitoring of infants
born to HIV-infected women in Thailand, where subtype E predominates.
Methods: 442 paired DBS and whole blood samples in EDTA (blood) were collected from 2-month old infants born to HIV-infected women. All blood samples
were tested using the modified Roche Amplicor HIV-1 DNA PCR test kit version
v.1.0 (reference assay). Paired DBS for the 34 DNA PCR positive blood samples, and the first 100 DNA PCR negative blood samples were tested with the
Roche Amplicor HIV-1 DNA PCR v.1.5 kit because the v.1.0 missed -15% of
non-B subtype DBS samples. Paired DBS and EDTA plasma samples were additionally tested using the qualitative Organon Teknika NucliSens RNA NASBA
kit. The 34 positive blood samples also had EDTA plasma (reference assay) and
DBS VL assays performed using the Roche Amplicor Monitor v.1.5 RNA PCR kit,
with DBS results hematocrit-corrected. DBS samples were silica-gel extracted for
both RNA assays. Reproducibility VL assays were set-up using 5 prepared DBS
samples (VL range 1,626-35,245 copies/mL), which were stored at 37~, 25~, -20~
and -37~ C, and retested at 1, 3, 6, 9 and 12 months.
Results: This table summarizes the results:
DNA PCR v.1.0 DNA PCR v.1.5 RNA NASBA RNA PCR VL
DBS+ Plasma+ DBS+ Plasma-DBS Plasma-DBS
+ Difference Correlation
Blood+ (n34) 34 34 34 GeoMean=0.11 log r=0.93
Blood- (n100) 0 0 0 N/A N/A

DBS VL results were reproducible within 0.5 log of baseline at all 4 temperatures
for DBS with >10,000 copies/mL through 3 months.
Conclusions: Roche Amplicor DNA v.1.5 DBS, and NucliSens RNA plasma/DBS
results were qualitatively comparable to the reference Amplicor DNA v.1.0 blood
results. Roche Monitor RNA v.1.5 DBS VL results were quantitatively comparable to the reference Monitor RNA v.1.5 plasma VL results. DBS are reliable and
stable for HIV-1 DNA/RNA detection and VL load monitoring.

Presenting author: Nancy Young, HIV/AIDS Collaboration, Dept. Med. Sciences,
Bldg. 2, 2nd Floor, Ministry of Public Health, Tivanon Road, Nonthaburi 11000,
Thailand, Tel.: [phone], Fax: [phone], E-mail: [email protected]
SWeOrB1340 HIV-1 p24 antigen, an inexpensive alternative viral
load marker, is an independent correlate of CD4
T-cell changes in patients on successful long-term
antiretroviral therapy
J. SchOpbach1, J. B6ni1, Z. Tomasik', L. Bisset2, M. Fischer2, H. GOnthard2,
B. Ledergerber2, M. Opravil2. 1Swiss National Center for Retroviruses, Swiss
National Center for Retroviruses, University of Zurich, Gloriastrasse 30,
CH-8028 Zurich, Switzerland; 2Department of Internal Medicine/University
Hospital, Zurich, Switzerland
Background. Quantification of HIV-1 RNA by RT-PCR is widely used for evaluating antiretroviral therapies, but not affordable in resource-poor settings. We here
assess a possible inexpensive alternative, a modified p24 antigen test, in patients
receiving successful long-term antiretroviral therapy.
Methods. HIV-1 RNA and p24, as tested by an experimental procedure involving signal-amplification-boosted ELISA of heat-denatured plasma, were prospectively measured in 329 samples from 55 patients whose HIV-1 RNA had been
reduced to below 50 copies/ml and who continued to receive therapy. HIV-1 RNA
in PBMC was determined retrospectively Uni- and multivariate linear regression
analysis was used to correlate CD4 counts and their changes over time with the
counts or changes of other lymphocyte subsets, HIV-1 RNA, and p24.
Results. During a median follow-up of 504 days, CD4 counts increased by a median of 62 cells/year. Concentrations of lymphocytes and their subsets strongly
influenced the CD4 count. P24 was a significant, and among viral markers dominating, independent inverse correlate of both the CD4 count in a sample (P =
0.013) and its annual change in a patient (P < 0.0001). P24 retained similar significance even among 48 individuals whose viral RNA in plasma, apart from some
transient blips, always remained below 400 copies/ml. Importantly, p24 decreased
steadily in most patients (P < 0001) while a significant decrease of HIV-1 RNA in
plasma or PBMC was not observed.
Conclusions. P24 antigen is a significant independent inverse correlate of the
CD4 change observed in successfully treated patients. The residual virus production or occasional viral blips observed under therapy are clearly insufficient
for maintaining the levels of p24 over time. P24 thus provides additional, clinically
relevant information not provided by HIV-1 RNA tests alone.
Presenting author: Jorg Schupbach, Swiss National Center for Retroviruses,
University of Zurich, Gloriastrasse 30, CH-8028 Zurich, Switzerland, Tel.: [phone], Fax: [phone], E-mail: [email protected]
WeOrB1 341 Evaluation of an ultrasensitive p24 antigen assay
(UPTA) as a possible surrogate marker of HIV-1
RNA in resource-poor settings
R. Respess', A. Cachafeiro2, D. Withum3, S. Fiscus2, D. Newman3,
I. Cabruja4, B. Branson3, O. Varnier5, T Dondero3. 1Centers for Disease
Control and Prevention, Centers for Disease Control and Prevention, 1600
Clifton Road, Mailstop A-12, Atlanta, GA 30333, United States; 2University of
North Carolina, Chapel Hill, NC, United States; 3Centers for Disease Control and
Prevention, Atlanta, GA, United States; 4PerkinElmer Life Sciences, Boston, MA,
United States; 5University of Genova, Genova, Italy
Background: The measurement of RNA viral load (VL) is useful for the clinical
management of HIV infection. Commercially available VL quantitation methods
are based on expensive nucleic acid-based amplification approaches requiring
skilled technicians, specialized equipment, and dedicated facilities, making their
use prohibitive for many resource-poor settings. A simple, sensitive, yet inexpensive HIV-1 quantitation based on modifications of a commercially available p24
antigen assay has been previously described, but has met with limited success in
transferability of laboratory methods.
Methods: To address this, a laboratory protocol, integrated test kit, and dedicated
reader software were developed and evaluated based on the modified p24 antigen assay. Anonymized sera from 39 newly diagnosed persons with HIV-1 subtype B were blinded, tested in duplicate by UPTA, and compared with previously
determined VL. Log 10 transformation of VL and the average of duplicate UPTA
test results were compared by Pearson product-moment correlation coefficient to
assess strength of statistical relationship.
Results: UPTA and VL test results were highly correlated overall (r = 0.87 [p <
0.001], r2 = 0.76). Correlation of UPTA and VL on specimens with VL < 5,000 (r =
0.94 [p < 0.0001], r2 = 0.88) and with VL > 5,000 (where higher values were truncated), (r = 0.64 [p < 0.001], r2 = 0.41) was significant in both groups. Correlation
of duplicate testing of same sample was high (r = 0.89 [p < 0.0001]).
Conclusions: The UPTA and VL results were highly correlated indicating that
UPTA may be a suitable surrogate for VL on HIV-1 subtype B specimens. UPTA
may be potentially useful as a simple, and less expensive approach to VL in

clinical staging, treatment monitoring, and pediatric diagnoses in resource-poor
settings. Studies to assess assay in sequential sera from persons on treatment,
and in non-B subtypes, are ongoing.
Presenting author: Richard Respess, Centers for Disease Control and Prevention, 1600 Clifton Road, Mailstop A-12, Atlanta, GA 30333, United States, Tel.:
[phone], Fax: [phone], E-mail: [email protected]