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

804 Abstracts 42156-42159

12th World AIDS Conference

was ranging from 0.7 to 1.8 (mean = 1.6). For a given patient, constant Loglo
difference was observed across time. Those patients are infected by subtypes
that are not accurately amplified by the current Amplicor 1.0 assay. Following
Loglo differences were observed from samples of subtypes A (2.30, 0.87, 0.28,
0.88), subtypes C (0.14, 0.16), subtypes D (1.56, 0.28), subtypes E (0.76, 1.34,
1.72, 1.18, 1.39), subtypes F (0.27). Similar differences in viral load results were
observed when comparing 1.0 and 1.5 assay.
Conclusions: In anti-retroviral drug clinical trials conducted in Europe, incidence of HIV-1 subtypes other than B inaccurately amplified by Amplicor 1.0 is
currently around 4% and will probably raise. It is recommended to use the Roche
Amplicor 1.0 assay + "Add-in Primers" or the newest Roche Amplicor 1.5 assay
at first visit to accurately quantify all patients plasma viral load.
42156 Usefulness of HIV-RNA levels for monitoring HIV-1
positive individuals in community based settings
Angelos Hatzakis1, Giota Toulouhi1, V. Paparizos2, H. Sambatakou3,
T. Kordossis4, A. Karafoulidou5, M. Lazanas6, A. Antoniadou7. 1Athens
University Medical School, Athens; 2Andreas Syngros Hospital, Athens; 3Gen.
Hosp. Athens"G. Gennimatas", Athens; 4Dept. Pathophysiology Laikon Hospital,
Anthens; 5Hemophilia Centre Laikon Hospital, Athens; 6Tzanio Hospital, Athens;
7Dept. Internal Medicine Laikon Hospital, Athens, Greece
Background: CD4 counts and plasma HIV-RNA levels predict progression of HIV
infection in natural history studies of untreated individuals. However, the additional
to current CD4 count importance of HIV-RNA levels for monitoring HIV-1 positive
individuals in community-based settings, is not well established.
Methods: Three hundred and thirteen HIV-1 positive individuals AIDS-free (193
in CDC stage A and 120 in CDC stage B) have been enrolled in an ongoing
prevalent cohort study initiated in 1995. Clinical status, CD4 counts and plasma
samples are obtained every 3 to 6 months. HIV-RNA in plasma is quantitated
with the 2d generation b-DNA signal amplification method. The primary endpoints
were: a) documented clinical AIDS or death and b) change in CDC stage.
Results: At study entry 122 subjects were untreated and 191 on monotherapy
or long term combination antiretroviral therapy with reverse transcriptase inhibitors. Patients under protease inhibitors treatment were excluded. The median
value of patients' initial CD4 counts was 297 cells/pm3 (5-95% centiles:33-749)
and HIV-RNA levels 18720 eq/ml (5-95% centiles: <500-225900). Baseline HIVRNA levels were strongly related with disease stage or CD4 counts at entry. The
median follow-up time was 9.3 months ranging from 1 month to 2.4 years. For
clinical AIDS or death, both baseline HIV-RNA levels and CD4 counts were independent predictors of disease progression. Risk of AIDS or death was reduced
by 59% (95% CI: 17%-80%) for every 10-fold lower HIV-RNA level at baseline
(P = 0.014) and by 54% (95% Cl:32%-68%) for every 100 cells/pm3 higher CD4
count at baseline (P < 0.001). For change in CDC stage, CD4 count as well
as high levels of HIV-RNA (above 20000 eq/ml) were significant predictors of
disease progression (P = 0.013 and 0.047 respectively).
Conclusions: HIV-RNA levels in addition to CD4 count is important for monitoring HIV-1 individuals in community-based settings.
42157 Viral load testing in Ontario
Carol Major1, R. Remis2, T. Degazio3, R. Galli3, S. Read4, M. Fearon3. Ontario
Viral Load Eval Group; 'Ontario Ministry of Health, HIV Laboratory, 81
Resources Road, Etobicoke, Ontario; 2University of Toronto, Toronto, ON;
3Ministry of Health PHL, Toronto, ON; 4Hospital for Sick Children, Toronto, ON;
5AIDS Bureau Ontario Ministry of Health, Toronto, ON, Canada
Background: Since November 1996 HIV viral load testing has been made available to all diagnosed HIV infected persons residing in the Province of Ontario
(estimated n = 12,100), by the Public Health Laboratory System, Ministry of Health
at 5 sites. A working group is evaluating the impact of the HIV viral load testing
program.
Methods: We established a laboratory information system which captures
unique patient identifiers, sex, date of diagnosis, viral load result, CD4 count and
current antiretroviral therapy. Testing was done using the Chiron bDNA version
2.0 kit with a dynamic range of 500-800,000 RNA copies /ml(c/ml), using the
same lot of product at each site. SAS was used to carry out descriptive analyses
of the data.
Results: To October 31, 1997, 12,000 tests have been done representing 5,500
unique patients. We estimate that approximately 80% of patients on antiretroviral
therapy but only about 25% of patients not on therapy have had at least 1 viral
load testing. 31% have been tested twice, 30% 3 times and 23% 4 or more times.
Among those with at least 3 follow up tests (n = 1844), we observed the following
observations: the median viral load decreased from 5908 c/ml at Test 1 (T1), to
2172 c/ml at T2, and to below detection at T3; CD4 counts remained stable at
273 at T1, 269 at T2, and 290 at T3. The proportion not on therapy decreased
from 24% at T1 to 14% at T3 and the proportion on triple therapy increased from
40% at T1 to 59% at T3. Amongst patients on triple therapy (n = 1894), 49% had
viral load <500 c/ml and a further 20% had viral load between 500-4,999 c/ml at

the most recent test.
Conclusions: Although viral load testing is accessed by most patients on
antiretroviral therapy, most of those not on therapy have not been tested. This
requires further investigation into barriers to testing and/or to accessing the health
care system for some individuals with HIV infection. The observed shifts to triple
therapy and reduction in viral load are encouraging. Centralized viral load testing

and complete data collection facilitates effective evaluation of advances in the
care of persons with HIV infection.
S42158 Evaluation of two second generation HIV-1 viral load
assays with improved lower detection limits
Richard Galli1, Carol Major2, G. Strunc2, T. Degazio2, C. Swantee2, M. Fearon2.
Otario Ministry of Health, HIV Laboratory, 81 Resources Road Etobicoke,
Ontario; 2Ontario Ministry of Health, PHL, Toronto, ON, Canada
Objective: To assess the performance characteristics of two ultra-sensitive nucleic acid amplification methods for the quantitation of HIV-1 RNA in archived
plasma specimens.
Design: Cross sectional, random sample study.
Methods: The assays assessed in this study were: The NucliSens HIV-1 QT
assay, sample input volume 2.0 ml., dynamic range of detection 40-10,000,000
copies/ml (Organon Teknika), and the Amplicor HIV-1 Monitor Ultrasensitive assay, sample input 0.5 ml., dynamic range 50-50,000 copies/ml. (Roche). A total
of 97 frozen, leftover plasma specimens from anti-retroviral treated and untreated
HIV infected adult patients with original viral loads ranging from <500 to 60,700
copies/mil were tested by all methods, and plasma RNA levels were compared
between the two second generation assays and to the viral load test of record
(Quantiplex 2.0 bDNA Assay, Chiron Corp.). Correlation of plasma RNA levels to
therapy category was also examined.
Results: Of the 64 specimens below detection with the bDNA assay, 37 (57.8%)
were also below detection on either of the ultra-sensitive assays. Of these, 34
patients were on triple therapy which included at least 1 protease inhibitor. The
remaining 27/64 specimens had between 50 and 4,482 copies/ml with either of
the second generation assays. Between assay variations of up to 2.3 log were
observed between Quantiplex 2.0 bDNA and the second generation assays on 33
specimens in the quantifiable range, with Pearson correlation coefficients ranging from 0.77-.86. Correlation of measurable RNA levels below 500 copies/ml
between the second generation assays was high. In general, plasma RNA levels
obtained with the Roche Ultrasensitive assay > NucliSens > Quantiplex bDNA
2.0.
Conclusions: Second generation plasma RNA quantitation assays having detection limits down to 50 copies/ml, or less are ideal for baseline and follow-up
monitoring of HIV infected patients. Although inter-assay variation was <0.3 log
for most specimens in this study, interpretation of longitudinal studies in individual
patients using multiple assay types must recognize the potential for highly variable
inter-assay plasma RNA levels.
42159 The roles of HIV RNA levels and CD4 counts when
assessing all-cause mortality
Veronica Miller1, Caroline Sabin2, C. Rottman1, A.N. Phillips2, A.M. Hill3,
T. Leder1, P. Gute1, S. Staszewski1. 'Zentrum der Inneren Medizin,
Theodor-Stern-Kai 7 60590 Frankfurt, Germany; 2Department of Primary Care
and Population Sciences, Royal Free Hospital School of Medicine, London;
3Glaxo-Wellcome Research, Greenford, England
Background: In early stages of HIV infection the HIV RNA level is one of the most
important independent markers for assessing prognosis. However, at later stages
of disease, when considering all-cause mortality as an endpoint, the relative roles
of HIV RNA levels and CD4 counts are less clear.
Methods: Information was collected on all individuals in the Frankfurt HIVCohort who had been followed since January 1995 and had regular HIV RNA
and CD4 measurements performed monthly. RNA measures were performed
using the Roche Amplicor assay. Statistical analyses were carried out using both
a person-years approach and Cox regression methods. Patients were followed
from the 1st January 1995, or the date of their first HIV RNA measurement.
Patients were followed to the time of death or the end of 1997.
Results: 1511 patients, followed for a total of 1747 patient-years, were included
in the study. These patients contributed a total of over 17,000 RNA measures and
16,000 CD4 counts for analysis (median of 10 CD4 counts and 10 RNA levels
per patient). Over the follow-up period, 59 patients died: at the time of death CD4
counts ranged from 1 to 1252 (median:21) cells/mm3 and RNA levels ranged
from undetectable to 6.52 (median:4.98) log (copies/ml). RNA levels were undetectable in 10 of the patients who had died (15%). Whilst the latest HIV RNA level
was predictive of survival in univariate analyses when fitted as a time-dependent
covariate (relative hazard (RH) 1.68 per log higher, 95% CI 1.39 to 2.04, p =
0.0001), it became non-significant after adjustment for the CD4 count (RH 1.12,
95% CI 0.92 to 1.35, p = 0.26). The CD4 count itself remained highly significant
(RH 1.81 per 1092 lower, 95% CI 1.65 to 1.99, p = 0.0001).
Conclusions: In this cohort of HIV-infected individuals, the CD4 count was
a stronger predicter of death due to all causes over the short term (in the next
month or so) than the HIV-RNA level. These results do not undermine the view
that the HIV RNA level is a strong long-term predictor of death.

42160 1 Development of quantitative assays using the 5'
nuclease technology
Shirley Kwok, S.D. Lu, S. Tsang, S.Y. Kao, J. Weiss. Roche Molecular Systems,
1145 Atlantic Ave., Alameda, CA, USA
Objectives: To develop automated viral quantitative assays with the requisite sen