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

86 Abstracts 12363-12366

12th World AIDS Conference

showed significantly decreased susceptibility to MKC-442, and novel mutations
of T691 and F227L were identified by the DNA sequence analyses of their RT
genes. When the AZT-resistant mutant A018C was cultivated in the presence of
combined RTIs, even the three-drug combination of MKC-442+3TC+AZT could
not suppress the viral breakthrough during a 3-week cultivation. The breakthrough
virus showed lower susceptibility to MKC-442 and had a mutation of Y181C.
Conclusions: Three-drug combination (MKC-442+3TC+AZT) demonstrated a
more potent efficacy against wild-type HIV-1 strains than two-drug combinations.
However, the three-drug combination proved less efficient in inhibiting the viral
breakthrough from an AZT-resistant strain than that from an AZT-sensitive strain.
12363 Correlations between CD4 and RNA response in
antiretroviral-naive patients with very advanced disease
(trial ISS 047)
Vincenzo Fragola, Marco Floridia, R. Bucciardini, D. Ricciardulli, L.E. Weimer,
C.M. Galluzzo, M.F. Pirillo, M. Andreotti. Instituto Speriore Di Sanita'- Lab.
Virology, Viale Regina Elena, 299 - Rome, Italy
Objective: To evaluate the correlation between CD4 and RNA changes during
antiretroviral treatment in previously untreated patients with AIDS or CD4+ <
200/mm3.
Design: Data from a phase II, double blind, randomized multicenter trial comparing AZT + ddl + nevirapine (NVP) with AZT + ddl + placebo were used. CD4
(mean absolute changes from baseline) and RNA (mean log changes) responses
were considered and Pearson correlation coefficients were calculated analyzing
results at 8, 24, 40 and 48 weeks.
Results: The numbers of patients with contemporary RNA and CD4 values
available at week 8, 24, 40 and 48 were 55, 45, 35 and 31, respectively. Although
for both study treatments negative correlation coefficients were observed between
RNA and CD4 at all time points, the values observed were generally close to 0
and not statistically significant (coefficient range: -0.01 to -0.43), indicating only
a general trend for a negative correlation between RNA and CD4 response (RNA
decrease and CD4 increase).
When RNA changes were compared by treatment in patients with different CD4
responder status (non response: change <50 cells; response: change >50 cells),
RNA changes were similar among the treatment groups in patients with no CD4
response (at week 24, -1.16 log, AZT + ddl; -1.25 log, AZT + ddl + NVP) and different among CD4-responders, with more pronounced RNA decreases in patients
receiving the triple combination (log changes: -1.28 and -2.61, respectively).
However, patients with discordant RNA/CD4 responses were observed. To
evaluate the possible negative prognostic role of a discordant response, patients
with discordant CD4/RNA responses at week 24 were analyzed in terms of treatment failure at week 48. All patients with a decrease of at least 1.5-log in RNA
and no more than 50 CD4 cells gained at week 24 maintained RNA levels at
week 48 below -1.5 log from baseline (n = 6, RNA log decrease range 1.9-2.9).
Patients with the same characteristics of CD4 response but with an RNA decrease
between 1 and 1.5 log at week 24 had RNA at or above baseline value at week
48 (n = 2). The number of patients with CD4 response but no RNA response was
too low to be analysed.
Conclusions: RNA and CD4 responses, although showing a trend for a general
correlation, may behave independently in terms of dimension of response. Among
CD4-responders, triple combination induces a higher RNA response since week
24. Mean RNA log decreases at week 24 greater than 1.5 log in the presence of
no or low CD4 response were sustained at week 48. Wider samples are needed
to confirm these preliminary observations.
S12364 How low is low enough? Suppression of plasma viral
load below 20 copies/mL is necessary for long term
virologic response
Julio S.G. Montaner1, V. Montessori2, J.M. Raboud2, B. Conway3, P. Robinson4,
M. Myers4, D. Hall4. 1667 1081 Burrard Street Vancouver BC V6Z 146;
2Canadian HIV Trials Network, Vancouver BC; 3St Pauls Hospital, Vancouver
BC, Canada; 4Boehringer Ingelheim Pharmaceuticals, Ridgefield CT, USA
Background: The stated goal of antiretorviral therapy is to decrease plasma viral
load (pVL) below 400 to 500 copies/mL. More sensitive assays are now available
fore quantitation to 40 to 50 copies/mL.
Objective: The assess the relationship between duration of pVL suppression
and the pVL nadir attained.
Methods: PVL was measured every 4 weeks for 52 weeks among 150 participants in a controlled randomized trial of ZDV/NEV (n = 47), ZDV/ddl (n = 52) and

ZDV/ddl/NEV (n = 51). Duration of suppression was defined as the time between
pVL nadir and remained at least one log below the baseline pVL.
Results: Median baseline and nadir pVL were 4.5 and 1.8 log10 copies/ml
respectively. The median duration of pVL suppression was 231 days, 14 days
or 1 day for patients whose pVL nadir was <20 copies/mL, 20-400 copies/mL
or > 400 copies/mL. At baseline, 101 patients had pVL < 50,000, 42 had pVL
50,000-250,000 and 7 had pVL > 250,000. In a multivariate model controlling for
treatment and baseline pVL, a pVL nadir <20 copies/mL was highly statistically
significant (RR = 10.0, p =.0001) while a pVL nadir 20-400 copies/mL was not
related to duration of suppression (RR = 1.5, p =.11). Baseline CD4 cell count
and maximum change in CD4 cell count were not statistically significantly related
to duration of pVL suppression.
Conclusions: Our results clearly illustrate the importance of achieving plasma
viral load <20 copies/mL (rather than 500 copies/mL) in order to maximize the
durability of the treatment effect.
112365 Predictors of duration of virologic suppression in a
meta-analysis of two trials of ZDV/ddl vs ZDV/ddl/NVP
Janet Raboud1, J.S.G. Montaner2, S. Rae2, R. Bucciardini3, M. Floridio3,
S. Vella3. 1620-1081 Burrard Street, Vancouver, BC V6Z 146; 2Canadian HIV
Trials Network, Vancouver, Canada; 3lstituto Superior Di Sanita, Rome, Italy
Objectives: To determine the relationship between plasma viral load (pVL) nadir
and duration of virologic suppression in a meta-analysis of two randomized trials
of ZDV/ddl vs ZDV/ddl/NVP.
Methods: Original data from 170 antiretroviral naive patients with baseline
CD4 and pVL in INCAS and ISS047 trials were pooled together in a single
database and analysed together. 87 patients received ZDV/ddl and 83 received
ZDV/ddl/NVP. The duration of suppression was defined to be the time that pVL
remained below 5000 copies/mL. The pVL nadir was dichotomized according to
whether or not it was below 400 copies/mL. Patients who missed >28 days of
any study therapy were defined to be non compliant.
Results: 107 patients achieved a nadir <400 copies/mL. The median durations
of suppression were 279 and 1 days for patients with a pVL nadir below and above
400 copies/mL. In a multivariate proportional hazards model, the pVL nadir (RR
= 0.14, p <.001) was highly statistically significant, after adjusting for treatment
(RR = 1.61, p =.05), compliance (RR = 2.31, p <.001), and baseline pVL (RR =
1.80, p <.01). Baseline CD4 cell count was not statistically significantly related
to duration of pVL suppression.
0.6 -8 1
QOS
0.4
1 O.2 1 'L I-ha=dir > 400 c Pies;mL
S-------0----L__
" 0.0._. __L_____1 _ _______
0 8 16 24 32 40 48
Number of Weeks
Conclusions: Our results illustrate the importance of suppressing pVL below
400 copies/mL in order to maximize the durability of the treatment effect. Further
work is currently underway to retest samples with pVL < 400 copies/ml with
a more sensitive assay to find the optimal level of suppression necessary to
maximize pVL response.
123661 Combination anti-HIV interactions and resistance profile
of the nonnucleoside RT inhibitor (+)-calanolide A
Robert Buckheit1, J. Russell1, V.F. Boltz1, L.A. Pallansch1, Z.Q. Xu2, M. Flavin2.
1Southern Research Institute, 431 Aviation Way, Frederick, MD; 2Sarawak
MediChem Pharmaceuticals, Lamont, IL, USA
Objectives: The anti-HIV agent (+)-calanolide A exhibits properties which renders it unique among the class of NNRTIs. Thus, we believe that (+)-calanolide A
may be useful in the therapy of AIDS patients as one component of an antiviral
drug cocktail, supplementing existing protease inhibitor cocktails. Calanolide A
is presently being evaluated in a Phase I/Il clinical trial under the direction of
Sarawak MediChem Research, Inc. In order to best define how the compound
might eventually be used in the therapy of AIDS patients, we performed an extensive evaluation of the resistance profile and combination anti-HIV interactions
of (+)-calanolide A in vitro.
Methods and Results: We have performed in vitro assays using both established and fresh human cells to evaluate the sensitivity profile of (+)-calanolide
A against resistant strains, selected for drug resistant viruses and evaluated the
interaction of (+)-calanolide A in combination with other anti-HIV agents. We have
found that (+)-calanolide A is unique in its primary selection for virus strains

bearing a T1391 amino acid change in the RT. This change was observed using
both laboratory derived and clinical virus strains. Amino acid changes L1001,
K103N and Y188C also have a negative impact on the activity of (+)-calanolide A.
The compound exhibits enhanced sensitivity to strains bearing the Y181C amino
acid change. In two and three drug combination assays, (+)-calanolide A has
interacted synergistically with nucleoside analogs and the NNRTI UC781. Additive
interactions were found with most NNRTIs and the protease inhibitors. In three

1.0o

S0.&
o.e
o U.
a.- n0

H

Il pVL Nadir
L --20 cxpieL'mL
|- ---20-40 co pies-m L
- -->400 copiemL
" "-- -' I'-I-- -... 1_.___
0 8 16 24 32 40 48
N u rn ber of Wee ks

o0.