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

12th World AIDS Conference

Abstracts 42170-42175 807

42170 Cerebrospinal fluid (CSF) versus plasma HIV RNA levels
in patients with AIDS-related neurological diseases
Pedro Cahn1, Carlos Zala2, H. Salomon3, J. Zirulnik1, V. Hodara3, M. Vazquez1,
C. Ochoa1, R. Glancszpigel4. 'Infectologia-Hospital Fernandez, Angel Peluffo
3932 (1181), Buenos Aires; 2Fundacion Huesped Buenos Aires; 3University of
Buenos Aires, Buenos Aires; 4Edyabe SA Buenos Aires, Argentina
Objectives: To obtain information with regard to CSF and plasma HIV RNA levels
from HIV infected individuals with AIDS-related neurological diseases.
Patients and Methods: Paired CSF and plasma samples were prospectively
collected from HIV infected subjects who underwent a LP for a diagnostic work-up.
HIV RNA concentrations in plasma and CSF were determined by bDNA (Chiron
2.0). ARV therapy was initiated or changed within 1 week of LP.
Results: From Aug 97 to Jan 98, 20 paired CSF and plasma samples were
obtained from 15 patients presenting with Toxo encephalitis (n = 1), Cryptococcal
meningitis (n = 5), PML (n = 4), TB encephalitis (n = 1) and AIDS related dementia
(n = 4). HIV RNA was detected in CSF from 12 out of 15 patients. Median plasma
and CSF HIV RNA copies/ml were 4.27 and 3.11 respectively. The Spearman
rank correlation coefficient and Pearson coefficient between plasma and CSF HIV
RNA copies/mi were 0.275 (p = 0.23) and 0.183 (p = 0.51). Five patients agreed
to have a second LP within 8 weeks of initiation or change ARV therapy. In all of
them, changes in CSF HIV RNA levels paralleled changes in plasma HIV RNA
levels upon initiation or change ARV therapy.
Conclusion: In AIDS related neurological diseases, CSF HIV RNA levels
may not correlate with ma HIV RNA levels. Although discordance between CSF
and plasma HIV RNA levels suggests compartmentalisation of viral production,
changes in plasma and CSF HIV RNA appears to follow initiation or change of
ARV therapy.
42171 Correlation of HIV-1 subtype and quantitative PCR RNA
assays in HIV-1 infected children
Monika Kaimll, U. Dietrich2, M. Funk1, R. Linde', J. Cinatl3, W. Kreuz1.
SKinderklinik, Universitat Frankfurt, 60590 2Frankfurt; Georg-Speyer-Haus,
Paul-Ehrlich-Str. 42 60590 Frankfurt; 3Universitat Sklinik Frankfurt, Virologie
60590 Frankfurt, Germany
Issue: The high prevalence of non-B subtypes in Romanian and African HIV-1
infected children treated in Germany indicates subtype determination for better
interpretation of the viral load.
Project: 14 perinatal infected children, 6 from Romania and 8 from Africa, age
1-14 years (median 6.2 years), were investigated prior to antiretroviral therapy.
HIV-1 subtype was determined by V3 based differential serotyping and direct
sequencing of the V3 domain. A quantitative RNA PCR assay (Amplicor version
I, Roche) was used in first line. If a low viral load in contrast to loss of CD4 cell
count or clinical progression was detected a b-DNA-PCR (Chiron) assay was
carried out additionally.
Results: 4 patients had decreasing CD4 cell count (<500/p1) and low viral
load (<10000 copies/ml, Amplicor). Subtype C and E was found in African and
Romanian children as well. In patients with low copy numbers (RNA-PCR, Amplicor) significant higher amount of virus (-1 log) could be detected by using
b-DNA-PCR assay (Chiron).
Conclusion: Determination of the HIV-1 subtype seems to be necessary in
African and Romanian children in order to interpret discrepancies between viral
load and clinical progression or low number of CD4 cells.
42172 Quantification of HIV-1 proviral DNA
Shirley Kwok', J. Sninsky', C. Chistopherson', J.F. Krowka2, S. Haynes2.
'Roche Molecular Systems, 1145 Atlantic Ave., Alameda, CA; 2California Dept.
of Health Services, Berkeley, CA, USA
Background: The level of proviral HIV DNA in peripheral blood ononuclear cells
(PBMC's) is often reported as copies/106 PBMC's or copies/p/g DNA. PBMC
counts are typically determined using a variety of methods, including hemocytometer counts and the concentration of DNA by spectrophotometric absorption.
Objective: To determine if Hoescht dye can be used to accurately quantify
total genomic DNA in a crude cell lysate and circumvent the need to perform cell
counts or purifications of the extracted DNA.
Methods: A dilution series of 8E5 cells, each cell containing a single copy of
proviral HIV-1 DNA, was prepared in uninfected PBMC's and extracted with a lysis
buffer containing proteinase K. The number of 8E5 cells and PBMC's delivered
into each extraction was determined by hemacytometer count. To determine the
amount of DNA in the lysate, an aliquot of the extract was diluted into the dye
and measured against a known standard on a fluorimeter. Fifty microliters of the
lysates were amplified for 30 cycles with the AMPLICOR HIV-1 MONITOR v1.5
primers in the presence of a DNA quantitation standard. Amplified products were
quantified on microwell plates using the AMPLICOR HIV-1 MONITOR format.
Results: The PCR-determined 8E5 copies were compared to a) the theoretical

number of 8E5 cells in each sample based on cell counts and b) to the expected
copies based on the fraction of 8E5 DNA in the total lysate as determined by
Hoescht dye. A conversion factor of 150,000 cells for 1 microgram of DNA was
used. Good correlation was observed between the number of 8E5 quantified
experimentally and the numbers expected based on cell counts and Hoescht dye.
The assay has an analytical sensitivity of 10 copies and 3-log dynamic range.
Conclusion: Hoescht dye can be used to accurately measure total DNA in
the cell lysates. This approach offers several advantages over cell counts. It

eliminates the need to perform the cell counts which can be tedious as well
as variable. Hoescht dye measures the amount of DNA released, whereas cell
counts assume complete lysis of the cells. Finally, the procedure is simple to
use.
S42173 Impact of drug non-compliance and the frequency of viral
load testing on outcomes, costs and patterns of therapy
Anke Richter', K.N. Simpson2, J.A. Mauskopf1. Research Triangle Institute, PO
Box 12194, RTP North Carolina; 2Chiron Diagnostics Corp., 460 Horton St.,
Emerville, CA, USA
Background: There is currently no consensus on the optimal frequency of viral load testing. Drug noncompliance rates, the ability of viral load testing to
detect this, and different testing frequencies have a strong impact on disease
progression, costs, and patterns of therapy.
Method: A Monte Carlo simulation was performed, tracking, for each individual,
the expected RNA level, CD4 count, AIDS status, survival status, therapy, drug
cost, testing cost, and medical care cost, every month for 15 years. The simulation
compares these outcomes under four scenarios in which 0%, 20%, 40%, or 60%
of the population may randomly choose to go on a drug holiday during their first
therapy. These four scenarios are run for fast, average, and slow progressors, as
determined by their viral load set point, as well as different frequencies of viral
load testing. Choice of combination therapies and CD4 cell decline rates follow
the NIH Guidelines. The cost of drugs was taken from the Redbook 1997 where
available. The efficacy of each combination therapy regimen was estimated from
published clinical trials. A discount rate of 3% is applied.
Results: The Monte Carlo simulations were performed for 5,000 individuals
na've to treatment. Base case simulations at 3 month testing intervals and at
least one month drug holiday show a decrease in life years for patient populations
in which more individuals take drug holidays. There is also a sizable difference in life expectancy between the different types of progressors, with slower
progressors living longer. Similar results are seen in terms of quality-adjusted
life years (QALYS), where compliant patient populations and slow progressors
have higher QALYS. Non-drug medical care costs are lower but total costs are
higher for patient populations that are have high compliance rates and are slow
progressors.
Conclusion: Rates of patient compliance affect life expectancy, QALYS, and
medical care costs, as does the speed of patient progression. Compliant individuals and slow progressors live longer, have better QALYS, and have lower
medical care costs. Overall costs are higher for compliant patients and for slow
progressors due to the extended time they are taking medications. Increasing
the frequency of viral load testing increases testing costs but decreases time and
money spent on ineffectual therapy.
42174 The change of HIV-1 proviral DNA copy number in
peripheral mononuclear cells during anti-HIV therapies
N.T. Takata, Teruhisa Fujii. Hiroshima Univ. Med. Hospital, Hiroshima, Japan
Objectives: We simultaneously measured HIV-1 proviral DNA copy number in
peripheral mononuclear cells and serum HIV-1 RNA level (viral load) of HIV-1
infected patients. And then, we analyzed the relationship between their changes
and their anti- HIV therapies.
Methods: Eighteen patients with HIV-1 infection were examined for 6 months,
and 16 patients recieved medication of HIV drugs. Peripheral mononuclear cells
were separated from freshly drawn whole blood with EDTA. Next, genomic DNA
were extracted by using the standard method. After a HIV-1 RNA transcript encompassing the SK462-SK431 gag primer -binding regions were amplified using
a quantitative PCR method, the products were hybridized on microwell plates
coated with specific oligonucleotide probes and were quantified compared with
the quantitation standard. Serum viral load was measured with Amplicore Monitor
method (Roche Diagnostic Systems).
Results: The proviral DNA copy number in peripheral mononuclear cells ranged
from 0 to 4000 copies per 106 cells and the serum viral load ranged from <400 to
3.7 x 105 copies per ml (n = 55). The proviral DNA level generally changed less
than the serum viral load. In 18 patients, although 6 patients had the changes
of serum viral load more than 0.5 log from the baseline during the observation
period, their proviral DNA levels had not changed significantly. In patients with
HIV infection, their proviral DNA levels had no relation to their clinical stages,
their CD4 counts or their anti- HIV therapies, as well as their serum viral loads.
Conclusion: Although serum viral load dramatically decreased by the adequate
anti- HIV therapies, the proviral DNA level did not changed so much.
|42175 Plasma HIV-1 RNA quantification: Comparative
evaluation between NASBA and other commercial assays
Ines Zapiola', L. Mammana', B. Aquino', L. Sarro', M.A. Bartellini2,
G.R.M. Muchinikk, J. Soto3. 1 Virologia Idefi Aguero 2117, 1425 Buenos Aires;
2SPM, Buenos Aires; 3Cedeac, Mar del Plata, Argentina

Objective: To evaluate the performance in the measurement of plasma HIV-1
RNA using NASBA and other two commercial kits.
Materials and Methods: Blood samples with EDTA were obtained from 96
HIV-1 infected patients. Plasma was separated by centrifugation and stored at
-70 C until processing. HIV-1 RNA was determined by NASBA HIV-1 RNA QT
(lower detection limit: 400 c/ml) and bDNA QUANTIPLEX HIV-1 RNA (lower