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

36 Abstracts 12124-12128

12th World AIDS Conference

1996-1997, 157 (88.7%) from common number of patients concerned to IDUs
and 20 (11.3%) to heterosexual group.
Results: At 140 (79.1%) from 177 patients are revealed antibodies to HCV, that
is confirmed in the appropriate tests. Antibodies to HCV were determined at 131
(83.4%) from 157 IDUs patients compare to 9 (45.0%) among 20 patients from
heterosexual group. HBsAg alone was detected only at 4 (20.0%) from heterosexual patients and was not detected among IDUs. HBsAg and HCV antibodies
in tandem were determined at 26 (16.5%) IDUs and 7 (35.0%) heterosexuals.
Conclusion: Prevalence of HCV among IDUs group is much higher, than
among heterosexual group at Ukrainian HIV-infected patients. It is possible, that
such dissemination HCV among IDUs in a near future will result in reduction
of a free-symptoms stage, faster transition to AIDS and increase the cost of
treatment. All these factors in aggregate with substantial growth of HIV-infected
IDUs number (4360 registered cases in 1996, 7950 cases in 1997) can appear
excessive burden for the budget of Ukrainian public health service.
S507*/12124 HIV serum RNA, CD4 cell numbers and risk for
clinical endpoints in a case-cohort study of ACTG
175
David Katzenstein1, S. Kim2, M. Hughes4, S. Hammer3, V. Degruttola2. 1CFAR,
Stanford University, Stanford, CA; 2Harvard School of Public Health, 3Harvard
Medical School, Boston, MA, USA; 4London School, Hygiene & Tropical
Medicine, London, UK
Objectives: To assess the relative contribution measurement of HIV RNA and
CD4 cells to clinical outcomes in patients receiving nucleoside therapies.
Design: Retrospective, case-cohort analysis of 457 subjects in the AIDS Clinical Trials Group 175 trial.
Methods: CD4 cells and HIV serum RNA (Amplicor) were compared among
245 cases (AIDS/death) and a cohort of 212 subjects (non-progressing) weighted
for antiretroviral experience, CD4 cell numbers and drug assignment. Proportional hazards models including baseline, 8, 20 and 56 week HIV RNA and CD4
cells, changes and normalized area under the curve-baseline (NAUCMB) were
constructed. Hazard ratios (HR) were calculated for AIDS/death per 100 CD4
cells/mm3, per 10 percent (%) CD4 and per log1i decreases in RNA.
Results: At baseline, CD4 cell numbers, % and HIV RNA were significantly
associated with HRs of 0.62, 0.57 and 0.40 respectively, (p < 0.01). Over the
56 weeks of therapy a lower NAUCMB for RNA was associated with an 78%
reduction in risk, higher CD4 numbers with a 64% reduction and CD4% was not
independently significant. In landmark analyses, after 56 weeks, baseline levels
of HIV RNA levels and CD4 numbers remained significant, and 56 week studies
were significantly associated with subsequent endpoints.
Conclusions: Higher HIV serum RNA levels are independently associated with
increased risk for AIDS and death among asymptomatic subjects with 200-500
CD cells and these associations persist with nucleoside antiretroviral therapies.
A reduction of HIV RNA or increase in CD numbers significantly reduce risk of
clinical disease and increase survival. RNA levels and CD numbers predict the
risk of subsequent clinical events and may be used to individualize therapy.
12125 Proton MR spectroscopy findings throughout the course
of HIV
Robert Lenkinski1, Kim Cecil1, I. Frank2, L. Desiderio1, D. Kolson3,
F. Gonzalez-Scarano3, I. Matozza2. Radiology 1 Silverstein, Univ of Penn 3400,
Spruce ST Philadelphia, PA 19104; 2lnfectious Diseases Univ of Pennsylvania,
Philadelphia, PA; 3Neurology Univ of Pennsylvania, Philadelphia, PA, USA
Background: While standard neurological, psychiatric and neuropsychological
tests provide a limited measure of viral influence in HIV, magnetic resonance
spectroscopy (MRS) may provide a non-invasive biochemical measure of the
metabolic alterations observed in a given region of the brain. This work seeks to
assess metabolic changes in vivo with proton MRS throughout the course of HIV
infection and AIDS.
Methods: Two MRI/MRS protocols are employed to examine patients at the
distinct stages of HIV infection. All examinations were performed on a GE SIGNA
1.5T MR scanner. Our initial protocol samples three regions (right and left centrum
semiovale, thalmus) in the brain. Asymptomatic HIV-positive patients are studied
every six months from the timepoint of initial diagnosis to three years out. Our
initial cohort of patients numbered 92, with 65 patients remaining in the study at
this time. Fifty-nine patients have completed their sixth visit. A second protocol
examines patients that are asymptomatic as well as symptomatic. This method
uses a technique to sample 0.36 cc regions of the frontal lobe. This method provides pure cortical gray matter and selected white matter regions. Forty patients
have been studied with this protocol.
Results: During the initial stage of infection, the ratios of amino acids compared
to creatine are significantly elevated compared to normal control subjects. We
find that patients who were asymptomatic, having no clinical evidence of AIDS
dementia complex, had increased levels of myo-lnositol (ml) in white matter.
Patients who were symptomatic, with diagnosed AIDS dementia complex (ADC),

had decreased levels of N-acetyl aspartate (NAA) in gray matter when compared
to normal controls and asymptomatic patients.
Conclusion: The spectroscopic protocols can follow patients throughout the
course of HIV infection and the stages of ADC (Figure 2). These results suggest
that changes in metabolism can be detected prior to neuronal damage. MRS may
identify patients who would benefit from drug therapies.

I12126 The role of HIV-1 phenotype in viral pathogenesis and its
relation to viral load, and CD4+ T-cell count
Bernd Kupfer1, R. Kaiser1, J.K. Rockstroh2, B. Matz', K.E. Schneweis1.
Ilnstitute of Medical Microbiology, Sigmund Freud Str. 25, Bonn; 2Department of
Medicine, Bonn, Germany
Objectives: To determine the predictive value of HIV-1 phenotype in peripheral
blood mononuclear cell (PBMC)-coculture, and to examine the relation between
viral phenotype, viral load and CD4+ T cell count.
Design: In study A 132 HIV-1 infected individuals were analysed retrospectively for the relation between the result of their initial HIV-cultivation in peripheral
blood mononuclear cell (PBMC) -coculture and survival rate six years afterwards.
In study B 176 patients were analysed since 1994 for markers of HIV disease
progression.
Methods: HIV-1 phenotype was determined by PBMC-cocultivation, viral load
by NASBA'm HIV-RNA QT System, and CD4+ T cell count by flow cytometry.
Results: Study A: The percentage of survival for patients with initial negative
virus culture was significantly higher (95%) than for patients with NSI-isolates
(78%) and SI-isolates (21%) (P < 0.05 and P < 0.0001, respectively). When
SI-phenotype was subdivided into "moderately cytopathogenic" and "highly cytopathogenic" significant differences in the rate of survival between these subgroups could be observed (45% vs. 14%; P < 0.05). Study B: Progression from
negative virus culture to the isolation of NSI-variants was associated with increasing viral load (P < 0.0001) but did not affect CD4+ T cell count significantly (P >
0.07), whereas the switch from NSI- to SI-virus was accompanied by significant
decline of CD4+ T cells (P < 0.0001) but no change in viral load (P > 0.21).
Conclusions: Isolation and phenotyping of HIV represents an additional striking
predictive marker for progression of HIV-infection.
S121271 HIV-RNA and CD4+ decline without therapy: Reference
values to gauge effectiveness of antiretrovirals
Alvaro Muhoz1, R.H. Lyles1, J.B. Margolick', J.V. Giorgi2, Y. Chen', J.P. Phair3,
J.W. Mellors4. 'Johns Hopkins School of Public Health, Room E7008, 615 N
Wolfe Street, Baltimore, MD; 2UCLA School of Medicine, Los Angeles, CA;
3Northwestern University Medical School, Chicago, IL; 4 University of Pittsburgh,
Pittsburgh, PA, USA
Objectives: To fully characterize the relation between baseline viral load (HIVRNA) and decline of CD4+ lymphocytes using longitudinal data collected prior
to the availability of antiretroviral therapies in a large cohort. To provide 95%
confidence limits for expected CD4+ decline at different baseline HIV-RNA and
CD4+ cell counts, whereby values less than the lower limits will be indicative of
therapeutic effectiveness in treated populations.
Methods: Baseline HIV-RNA was measured by bDNA assay (version 2.0) in
stored samples collected around September 1985 in the Multicenter AIDS Cohort
Study (MACS). Semiannual CD4+ measurements collected up to July of 1988
formed the basis for estimation of decline of CD4+ lymphocytes. We used randomized regression methods for statistical analysis: CD4+ decline was modeled
as a function of baseline values of HIV-RNA (fixed effect) and CD4 cell count
and was allowed to vary across individuals (random effects). Confidence intervals were obtained by computationally intensive resampling methods (bootstrap)
without need for strong assumptions.
Results: The study population was composed of 1604 HIV positive MACS
participants who contributed 7515 CD4+ measurements prior to July 1988 excluding those taken under reported antiretroviral usage. Using selected values of
baseline HIV-RNA and CD4+ cell count in fitted models, expected loss of CD4+
cells per year (95% C.I.) was:
Baseline CD4 Baseline HIV-RNA:
(cells/uL) 1,000 copies/ml 5,000 copies/ml 50,000 copies/ml
350 19.4 (2, 36) 44.3 (28, 61) 69.6 (60, 80)
500 27.8(12, 44) 51.6 (43, 60) 87.1 (77,97)
750 41.9 (31, 53) 63.8 (52, 75) 116.2 (102, 130)
Conclusions: The expected rate of CD4+ decline is strongly related to baseline values of both viral load and CD4+ cell count. The summary models provide
expected CD4 declines and 95% confidence intervals that are useful in gauging
the effectiveness of antiretroviral therapies at the population level across the full
range of HIV-RNA and CD4 values. Results presented here illustrate the role
that observational (cohort) studies can play in conjunction with clinical trials to
assess the full benefit that therapies may provide to populations of HIV infected
individuals.
12128 Detection of HIV RNA in plasma below quantifiable limits
associated with poor virologic outcomes
Christopher D. Pilcher, D.A. Wohl, Z.A. Beatty, D. Nickisher, J.J. Eron. CB #7030
547 Burnett Womack UNC-Chapel Hill NC, USA

Background: The University of North Carolina (UNC) is a tertiary care center
caring for over 1,000 HIV-1 infected persons. UNC uses the Roche Monitor"
system for HIV-1 RNA quantitation, with a lower limit of quantifiability of 400
copies/ml HIV-1 RNA and a variable limit of detection lower than 400 copies. The
clinical significance of the distinction between HIV-1 RNA levels "below quantifiable limits", but detected (BQL) and "below detectable limits" (BDL) is not known
and was examined in a retrospective study.