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

12th World AIDS Conference

Abstracts 42190-42195 811

42190 Longitudinal study of CD8 T cell activation and viral load
in HIV-infected patients with >400 CD4 cell count
Martine Sinet', Fabrice Bouscarat', M. Levacher", M.C. Dazza', F. Chau',
B. Desforges1, M. Muffat-Joly', P.M. Girard2. 1INSERM Unite 13, Hop.
Bichat-Claude Bernard, 170 Bd. NEY, Paris; 2Dept. of Infectious Diseases, H6p.
Rothschild, Paris, France
Objectives: To investigate the temporal relationship between CD8+ lymphocyte
phenotypic alterations and plasma HIV RNA levels in the natural history of HIV
infection.
Patients and Methods: 33 HIV-infected patients with -400 CD4+ cells/1L
were studied prospectively during 3 years. The percentage of CD8+ lymphocytes
expressing activation markers (CD38, CD28, HLA-DR, CD45RO) was determined
by means of three-color flow cytometry. Plasma HIV RNA was quantified by bDNA
method (Quantiplex" Chiron, detection limit: 500 copies/mL).
Results: The mean duration of follow-up was 34 months (range 20-3) during
which patients had 4 to 9 visits including viral load and CD8+ activation marker
determinations. At the end of the study, a total of 202 samples had been analysed.
In 20 patients, an antiretroviral therapy was initiated during the study period, or
soon after. Initial levels of HIV RNA ranged from 2.7 to 5.6 log (mean 4.0), and
T cell counts/p L ranged from 399 to 1964 (mean 796) for CD4+ and from 376 to
3754 (mean 1435) for CD8+; HIV RNA was not significantly related to CD4+ cell
count, but correlations with the number of CD38+ or HLA-DR+ CD8+ cells were
significant. Time-dependent fluctuations of plasma HIV RNA paralleled those of
activated CD8+ cells (especially the CD38+CD28 CD8+ subset) and were patient-dependent. For the 33 patients, and before antiretroviral therapy, only mild
changes in HIV RNA were observed (mean: +0.025 log per year, range: 0.35
to +0.73), while CD4+ cell count decreased (mean: 102 cells per year, range:
558 to +24). When subgroups of patients are individualized according to CD8+
activation profile, the mean rate of CD4+ T cell loss is found significantly higher
in patients with important fluctuations of the CD38+CD28 CD8+ subset (159 vs
65 cells per year, P = 0.021).
Conclusion: Combined virologic and immunologic overtime survey of patients
appears informative in evaluating HIV disease progression. Marked fluctuations
of CD8+ activation indicate a drop in immune regulation of infection, as reflected
by an increased rate of CD4+ T cell loss.
42191 The CD38 subset of CD8 cells predicts progression to
AIDS in long-term surviving IDUs
Karin Froebel', M.P. Armitage2, S. Yang3, M. Struthers3, J. Whithelaw4,
S.M. Raab3. 'ICAPB, University of Edinburgh, King's Buildings, Edinburgh, EH9
3JN; 2Department of Medicine University of Edinburgh Edinburgh; 3Department
of Mathematics, Napier University Edinburgh; 4SNBTS, Edinburgh, UK
Background: Both CD38/CD8 (Giorgi et al J AIDS 6:904 1993) and
CD45RO/CD38/CD8 (Bofill et al AIDS 10:827, 1996) cells have been claimed
to add value to the CD4 count in predicting progression to AIDS. We therefore
compared the abilities of CD38/CD8 and CD45RO/CD38/CD8 cells to predict progression to AIDS in intravenous drug users (IDUs) who are long-term survivors.
Design: A prospective controlled study
Methods: The patients (n = 137) were members of the Edinburgh City Hospital
cohort (clinical director: Dr R Brettle). All had been infected by intravenous
drug injection and the majority had a known date of sero-conversion. The
percentages of CD4 cells, CD8cells, CD38/CD8 cells, CD45RO/CD8 cells and
CD45RO/CD38/CD8 cells were measured using 2- or 3-colour flow cytometry.
CD38 positive cells were gated on the CD45RO/CD8 subpopulation. Kaplan
Meier analysis and Cox proportional hazard regression analysis were used to
describe the survival distribution of each marker and to find the most predictive
markers.
Results: After 4 years of follow-up 25 had progressed to clinical AIDS. A
further 17 patients died after their last visit to hospital before progression to AIDS.
The analyisis showed that apart from CD45/CD8, all the markers are significantly related to disease progression individually. Among all the models fitted,
the model with two co-variates: high (44-81%) CD38/CD8 and low (0-19%)
CD4 is the best predictor of progression to AIDS (RR 18.9). The triple marker,
CD45RO/CD38/CD8, is not better than the dual marker CD38/CD8.
Conclusion: Measurement of CD38/CD8 cells adds to the value of the CD4
count in predicting the rate of progression to AIDS. Measurement of the CD45RO
subset of CD38/CD8 cells does not improve the predictive value. This study
extends the results of earlier work to patients infected with HIV by injecting drug
use, who are also long term survivors.
42192 Comparison of a PCR-based assay and the MT2
coculture method for the follow-up of syncytium inducing
strains in HIV patients under antiprotease therapy
Dolores Vaira2, C. Forceille2, C. Tsobo2, C. Gerard2, D. Sondag-Thull2,
B. Rentier1. 1Virologie fondamental et Immunologie; 2Laboratoir e Reference,

SIDA-CHU-Universitd de Linge, Tour de Pathologie, Sart Tilman, Liege, Belgium
Background: The aim of the study was to investigate the suitability of a PCRbased assay for HIV-phenotyping (syncytium SI or non-syncytium inducing NSI)
directly from peripheral blood lymphocytes (PBL) for the follow-up of patients
under antiprotease treatment.
Patients and Methods: 15 patients were selected according to the presence of
a phenotypic SI strain at baseline just before antiprotease treatment. A constant

or transitory phenotype switch (SI - NSI) was observed in these patients during
treatment.
Viral phenotype was determined by cocultivation of 1.106 patients PBL with
MT2 cells (Koot et al. AIDS 1992, 6:49-54)
Viral genotype was analyzed on frozen PBMC by a selective amplification of
sequences in the V3 loop domain using primers recognizing SI-specific DNA
sequences (Fouchier et al. Clinical Microbiology, Apr. 1995, 906-911).
Results: At baseline of antiprotease treatment, the SI-specific DNA sequences
could be amplified in all patients (15/15) suggesting that there was a full correlation between SI phenotype and V3 genotype. Among the 15 patients, 7 have been
tested by PCR after a phenotype switch occurred during treatment. Although such
a phenotype switch was observed, the SI specific DNA sequences could still be
amplified. According to the results obtained so far, the phenotype switch is not
confirmed by a genotypic modification of the baseline strain.
Conclusions: As suggested by Ercoli et al. (AIDS 1997, 1211-1217) antiprotease therapy might be associated with a selection of less pathogenic viral strains
(NSI strains). Our results indicate that there is an apparent discordance between
the two typing methods and that this PCR based typing method seems not to be
appropriate for the follow up of patients under antiprotease therapy.
42193 1 Comparison between helper T-lymphocyte counts
obtained by zymmune and flow cytometry methods
Ellen Taylor, K.L. Eckert, S.G. Weng, E. Ramirez, J.B. Margolick. Johns Hopkins
University SHPH, 627 N Washington Street, Balto, MD 21205, USA
Objective: To compare results of the coventional flow cytometry method of determining helper t-lymphocyte counts against the Zymmune assay, a technologically
simpler and less expensive method.
Methods: Matched specimens collected in an outpatient research study were
sent to two laboratories, laboratory 1 assayed the specimens using conventional
flow cytometry method and laboratory used the Zymmune assay. Flow cytometry specimens phenotopic analysis was done using two color flow cytometry
of specimens stained using ammonium chloride lysis of whole blood and staining with fluorescein and isothiocyanate and phycoerythrin-conjugated antibodies.
The Zymmune method utilizes monoclonal antibodies attached to magnetic and
fluorescent particles to masure the absolute count of helper t-lymphocytes. Whole
blood is incubated with a mixture of fluorescent and magnetic beads in the well
of a microplate plate.
Results: A total of twenty-five samples were analyzed. Distribution the counts
approximated a normal curve for both methods. The two test measurents showed
a high level of correlation, p = 0.912 amd were not statistically significant different
when the two values were compared, p --.5 using paired t-test.
Conclusion: The results of this comparison indicate that the Zymmune assay
may be an appropriate alternative for determining helper t-lymphocyte (CD4)
counts.
42194 CD4 counts as a surrogate marker in Indian HIV
population
Shashank Joshi1, A.K. Deshpande2. 1Clinical Research Associate, Medicine
Dept., GMC & SIRJJH, Munbai; 2Professor & Head Dept., GMC & SIRJJH, India
Background: To study in Indian HIV patients the role of CD4 as a surrogate
marker.
Methods: 40 consecutive HIV positive patients at Department of Medicine,
Grant Medical College and Sir. J.J. Group of Hospitals were studied. CD4 counts
were measured by FACS counter. CD4 counts were compared with CDC clinical
class A, B, C. 57.5% were in Class A, 15% in Class B and 25% in Class C. In
Class A 30.4% had CD4 counts -500, 26.1% between 499-200 and 43.5% -:200.
In Class B 33.33% were between 499-200 and 66.64% <200. In Class C 9.1%
had CD4 count -500, 9.1% had CD4 count between 499-200 and 81.8% had
CD4 - 200. It can be clearly seen that these is a definate clinico-immunological
dissociation and needs further collaboration with viral load.
Conclusion: It appears that in Indian HIV positive patients, CD4 counts need
validation. Due to ethnic origin, low BMI, associated malabsorption, tuberculosis
and tropical disease it may be difficult to use it as a surrogate marker as used
in the West. Validation of CD4 counts in developing countries is essential in
presence of confounders especially as it is expensive.
230*/42195 Patterns of phenotypic and genotypic
cross-resistance among protease inhibitors in over
1,000 clinical HIV-1 isolates
Kurt Hertogs', B. Larder2, J. Mellors4, V. Miller5, S. Kemp3, F. Peeters',
R. Pauwels2. 1 Virco NV Belgium, Gen. Dewittelaan 1184 B-2800 Mechelen;
2Tibotec NV, Mechelen, Belgium; 3 Virco UK, Cambridge, UK; 4 University of
Pittsburgh, Pittsburgh PA, USA; SUniversity Of Frankfurt, Frankfurt, Germany
Objectives: Different patterns of cross-resistance among HIV protease inhibitors

have been reported. We therefore studied the frequency of phenotypic and
genotypic resistance and cross-resistance in plasma-derived recombinant HIV-1
isolates obtained from patients treated with different reverse transcriptase (RT)
and/or protease inhibitors.
Methods and Results: Generation of recombinant viruses and subsequent
susceptibility testing was performed using the Antivirogram" method (Hertogs et
al. Antimicrob. Agents Chemother. 1998 (42) 2: 269-276). Seven concentrations