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

50 Abstracts 12193-12197

12th World AIDS Conference

The CD4 cells count decreased in the first four months (media -50 cells) and
they was increased later. None of the patients had any AIDS-defining illness and
they remain with non detectable plasma viral load.
Conclusion: People with fever, lymph node enlargement and malaise, with
or without rash, post-exposure to a known HIV infected source or not, must
be suspected to have an APHI and must be treated with ART as soon as the
diagnosis is confirmed.
121931 Immunological and virological effects of combination
antiretroviral therapy: Primary HIV infection (PHI) vs
asymptomatic, established infection (AEI)
Andrew Carr1, J. Zaunders2, P. Cunningham2, G. Kaufmann2, A. Kelleher2,
D.A. Cooper2. Victoria Street Sydney 2010; St. Vincent Hospital, Sydney,
Australia
Objective: To determine if combination therapy commenced during PHI would
have greater anti-HIV or immunological benefit than when commenced during
AEI.
Methods: We treated 12 AEI patients (mean 3 years infection) with CD4 counts
>250 cells/pl and 16 symptomatic PHI patients with two nucleoside analogues
and indinavir for 12 months. Comparisons were made for HIV RNA load and
kinetics, CD4+ and CD8+ T lymphocytes (including CD45RA/RO, CD62L, CD28,
CD38, HLA-DR subsets), HIV antibody concentration (by anti-p24 ELISA and
immunoblots), beta-2 microglobulin and adherence to therapy.
Results: HIV RNA responses were non-significantly less in the PHI group, but
fell to less than 50 copies/ml plasma in most patients. Mean CD4 count increases
were 182 and 185 cells/mm3, respectively. CD8 counts declined significantly to
near normal levels in PHI patients but were unchanged in AEI patients. Naive
CD4 (CD45RA+CD62L+) cells increased in both groups but more so in PHI pts.
Activated (CD38+, HLA-DR+) CD8 cells were higher at baseline in PHI patients
and, although declining in both groups, remained higher in AEI patients at all time
points. Similar changes for CD8+CD28-cells were seen. Adherence rates were
similar between the two groups.
Conclusion: Dual nucleoside analogue + indinavir therapy led to a more normal immune phenotype after 12 months when commenced during PHI than when
commenced during AEI.
112194 T lymphocyte dynamics in primary HIV infected patients
treated with HAART
Nicolas Sachsenberg, G.A. Schockmel, L. Perrin. Laboratory of Virology,
University Hospital, Geneva, Switzerland
Background: Rapid changes in the concentrations in CD4 and CD8 T cells are
observed during primary HIV infection (PHI). This likely reflects virus-T-cell-interactions susceptible to be modified by antiviral treatment.
Methods: Changes in CD4 and CD8 T cells including naive (CD45RA) and
memory (CD45RO) subsets and their respective proliferation rates were assessed
in 9 PHI patients on HAART.
Results: At baseline, high proliferation in CD8 T cells (19.1 ~ 14.6% Ki-67+;
uninfected controls: 0.9 4 0.4%) and to a lesser extent in CD4 T cells (4.7 ~ 5.0%;
controls: 0.9 ~ 0.4%) was observed. This was associated with a rapid increase in
CD4 T cells, mainly of memory phenotype. At week 24, CD4 levels were 627 ~
255 (baseline: 336 ~ 150); the increase in CD4 memory cells (baseline: 168 ~ 76;
24 weeks: 312 ~ 147) was more marked than that of the naive subset (baseline:
96 ~ 77; 24 weeks: 144 ~ 102). This was associated with higher proliferation
rates in the memory subset (baseline: 6.5 ~ 2.6%; 24 weeks: 3.0 ~ 1.0%) than in
the naive subset (baseline: 2.4 ~ 1.2%; 24 weeks 0.8 ~ 0.5%). In contrast to CD4
T cells, CD8 T cell levels remained stable, despite high proliferation rates in the
memory subset, likely due to massive apoptosis. During treatment, proliferation
rates of CD8 cells decreased in parallel with viremia, reaching values in the
normal range at 6 months (1.3 ~ 0.6% Ki-67+; viremia <100 copies/ml).
Conclusions: In PHI, high proliferation rates in CD8 and to a lesser extent CD4
T cells were observed pre-treatment. During HAART, high proliferation rates were
associated with a rapid increase in CD4 T cells, but not CD8 T cells. Proliferation
rates were higher in CD4CD45RO T cells than in CD4CD45RA T cells, and this
translated into a more rapid increase. After 24 weeks, CD4 and CD8 proliferation
rates dropped to values in the normal range.
12195 Neurological symptoms during primary HIV infection
correlate with high levels of HIV-RNA in cerebral spinal
fluid
Giuseppe Tambussi1, A. Gori2, B. Capiluppi1, C. Ballota2, L. Papagno2,
A. Saracco1, A. Lazzarin1. 1San Raffaele Scientific Institute, Division of
Infectious Diseases Via Stamira D'Ancona, 20 20127 Milan; 2Clinic of Infectious
Diseases, L. Sacco Hospital Milan, Italy

Background: Neurological involvement occurs in about 40% of patients with
serological evidence of primary HIV infection (PHI). Clinical manifestations may
range from mild signs to a severe picture of meningoencephalitis.
Methods: The present analysis involves 21 patients with diagnosed symptomatic PHI. The following variables were considered for statistical analysis:
plasma and cerebral spinal fluid (CSF) viral load (VL) and time (T) between referred onset of symptoms and lumbar puncture, calculated in days. Correlations
between plasma and CSF VL, and between plasma/CSF VL and T were analyzed.

A covariance analysis, using neurological symptoms as the grouping variable,
was then performed. Differences in mean log plasma and CSF VL in the group
with and without neurological symptom were tested by T-test.
Results: Neurologic symptoms (NS) were present in 11 patients (52%), ranging
between severe and persistent headache to clinical signs suggestive of meningitis. Five out of 11 patients underwent hospitalization because of severity of
neurological symptoms; among the other six subjects, the most relevant neurological sign was headache, which lasted more than 7 days. The mean plasma VL
log was 5.5 in patients with NS and 5.1 and in patients without NS (p < 0.03).
The mean CSF VL was 4.1 log in the NS group and 2.6 log in patients without
NS (p < 0.00001). There was no correlation between T and CSF VL, either if we
considered the entire population (r = -.10) or the NS and no-NS groups (r = -.15
and r = -.28) separately.
Conclusions: Neurological involvement is a common and often underestimated feature during PHI. There is a strong relationship between neurological
symptoms and CSF viral load. Plasma viral load alone does not correlate or
predict the CNS involvement. In patients with PHI an early treatment with drugs
with high penetration in the central nervous system must be considered.
S12196 Persistent or relapsing oral candidiasis as an indicator of
HIV disease in children
Gareth Tudor-Williams', Shunmay Yeung2, D. York3, R. Booy1, D. Wilkinson4.
'Dept. of Paediatrics, St. Mary's Hospital, Paddington, London W2 1NY UK;
2Hlabisa Hospital Kwazulu/Natal; 3 Virology Dept., King Edward Hospital,
Durban; Centre for Epidemiological Research, SA, Hlabisa, South Africa
Background: Confirming the diagnosis of HIV infection in children is difficult in
remote or resource-poor settings; access to appropriate laboratory tests may
be limited, and the existing clinical case definition (CCD) lacks sensitivity and
specificity.
Aim: To assess response of oral candidiasis (OC) to treatment, as a diagnostic
indicator of HIV infection in children.
Methods: Setting: HIabisa Hospital in rural KwaZulu/Natal. Study population:
Consecutive admissions to the paediatric ward of children >3 months of age with
OC. OC was graded (1 = mild, 3 = severe) on admission. Children were treated
with Nystatin and reviewed three times (Day 2-5, Day 6-10, and Day 11-30). A
coded admission blood sample was analysed for HIV antibodies and, for infants
<18 months, proviral DNA.
Findings: 74 children (mean age 15 months) were studied, of whom 41 were
HIV infected. The severity of oral thrush on admission was not predictive of HIV
status (58% of the HIV infected vs. 46% of the uninfected children had Grade 3
OC). Nutritional status also was not predictive (54% of HIV infected vs. 52% of
uninfected were severely malnourished). Ten children died during follow-up: six
were HIV infected. Nine of 45 children had Grade 3 OC at third follow-up, and all
9 were HIV infected (p = 0.006, PPV and specificity in this small cohort nominally
= 100%). All of the HIV uninfected children had responded to treatment by 3rd
follow-up. Half of the HIV infected children cleared OC by 3rd follow-up, thus
the sensitivity was low (50%). WHO CCD would have misclassified 15% of the
uninfected children, and would have had a PPV of 67% in this cohort.
Conclusion: Primary health care workers should be trained to observe the
child's response to treatment, in addition to the presence of oral candidiasis at
presentation. Non responders or those that relapse with Grade 3 OC at Day 11-30
are very likely to be HIV infected. This finding had a greater positive predictive
value than the WHO CCD.
S12197 Treatment of primary HIV infection with nelfinavir,
zidovudine, and lamuvidine
Frederick Hecht1, D. Smith2, D. Cooper2, R. Anderson3, A. Peterson3
J.O. Kahn4. 1995 Potrero Avenue, San Francisco, California 94110; 3Agouron
Pharmaceuticals, Inc., La Jolla, CA; 4University of California, San Francisco, CA,
USA; 2University of New South Wales, Australia
Objective: To determine the efficacy of nelfinavir (NFV), zidovudine (AZT), and
lamuvidine (3TC) in treating primary HIV infection.
Methods: Patients were diagnosed with primary HIV infection based on presence of HIV-1 RNA with a negative or indeterminant HIV-1 antibody test, or documented seroconversion within 6 months of enrollment. Patients received open
label treatment with NFV, AZT, and 3TC. HIV-1 RNA was measured by Chiron
bDNA testing, supplemented by Roche Ultrasensitive PCR. Lymph node biopsies
have been obtained and quantitation of lymph node viral burden will be performed.
Results: 13 patients have been enrolled from 2/97 to 1/98, with median follow-up of 26 weeks. At baseline, median HIV-1 RNA was 77,720 (range <500 to
>1,600,000), and median CD4 count was 541/11 (range 308-930). Ten pts have
received at least 12 weeks of treatment. All 10, including the one with an initial
HIV-1 RNA > 1.6 million, reduced HIV-1 RNA levels to <500 copies/mi within 12
weeks of starting treatment. Two pts have discontinued treatment; the remaining
8 who have continued treatment for more than 12 weeks have all remained

<500 copies/ml of HIV RNA. Declines to undetectable levels on the Ultrasensitive
HIV-1 RNA assay (typical limit of detection <50 copies/ml) have been slower. Of
5 pts on with results available after 6 months or more on treatment, 4 became
undetectable on this test within 6 months of starting treatment while one took 8
months to become undetectable on this assay.
Conclusions: Even with high initial HIV-1 RNA levels, treatment of pts with
primary HIV infection with NFV, AZT, and 3TC appears to be highly effective at
lowering HIV-1 RNA to undetectable levels and maintaining virologic control.