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

12th World AIDS Conference

Abstracts 32293-32296 579

most significant negative impact on the MVL reduction, resulting in only a 0.22
log reduction compared to a 1.06 log reduction when this mutation was absent (p
= 0.009); the RT codon 215 mutation rended towards significance (p = 0.052).
Conclusions: These data suggest that GART may be helpful when clinical
failure is present. A higher GART score and a mutation of PI codon 90 predicted
a poor response to therapy. A larger, prospective study will be needed to clarify
GART's role in clinical retrovirology.
32293 Viral resistance and viral load response to abacavir (ABC,
1592) in an AIDS dementia complex trial (CNAB 3001)
Randall Lanier', Justin McArthur2, J.H. Atkinson3, R. Price4, D. Clifford5,
D. Simpson6, G. Torres7, G. Sturge8. '5 Moore Drive, RTP, NC; 2Johns Hopkins
University, Baltimore, MD; 3University of California, San Diego, CA; 4San
Francisco General Hospital, San Francisco, CA; 5 Washington University Medical
Center, Saint Louis; MO; 6Mount Sinai Medical Center, New York, NY; 7Saint
Vincents Hospital, New York, NY; 8GlaxoWellcome, RTP NC, USA
Background: It has been shown that ABC is effective at lowering HIV-1 viral
load in antiretroviral therapy (ART) naive and therapy experienced patients at
300 mg BID. Previous studies in therapy experienced patients have indicated that
phenotypic viral resistance >8-fold higher than a wild-type control virus predicts
lack of viral load response to ABC. CNAB 3001 was an international, double-blind
study of ABC in AIDS dementia complex and examined the effect of 600 mg BID
in therapy experienced patients and provides the first opportunity to determine if
higher doses can overcome viral resistance in the clinic.
Methods: 99 patients were enrolled. Subjects were randomized to receive
their current therapy plus ABC 600 mg BID or placebo for 12 weeks. A subset of
patients agreed to undergo lumbar puncture at baseline, week 6 and week 12 to
determine efficacy of ABC at reducing viral load in cerebrospinal fluid (CSF). HIV
from plasma and CSF were genotyped (GT, ABI sequencing of the RT gene) and
phenotyped (PT, Antivirogram, Virco NV, Belgium resistance assay) at baseline
and at 12 weeks. Correlations between GT and PT with viral load (VL) response
were examined.
Results: GT and PT for most patients at baseline will be presented. Blinded
plasma VL analysis indicates that about 25% of patients saw a >0.5 log drop during the randomized phase. CSF VL analysis and all correlations will be available
by March. Preliminary blinded plasma VL data are presented below:

Baseline -2.6 log
Genotype plasma VL at
wks 0, 4 and 8
184 M (wild type, n = 8) 1/8 (13%)
184 V (n = 26) 4/26(15%)

0.5-1.0 log
drop in
plasma VL
0
2/26 (8%)

>1.0 log
drop in
plasma VL
1/8 (13%)
4/26 (15%)

Total responders
(>0.5 log
drop in plasma VL)
1/8 (13%)
6/26 (23%)

32295 | Long term viral suppression using quadruple drug
combination therapy in treatment experienced HIV/AIDS
patients, following initial protease failure
Clayton Barbour, II. Desert AIDS Project 750 South Vella Road, Palm Springs,
California, 92264, USA
Objective: To evaluate the magnitude and durability of HIV viral suppression
using combination therapy with ritonavir, saquinavir, and two nucleoside reverse
transcriptase inhibitors, in patients failing triple drug therapy.
Patient Population: 32 HIV infected patients with CD4 cell counts <250. All
patients had HIV viral loads >5000 or evidence of disease progression, despite
4 months of triple drug therapy consisting of 2 nucleoside reverse transcriptase
inhibitors and 1 protease inhibitor (28 saquinavir, 2 ritonavir, and 2 indinavir
failures).
Regimen: Combination of ritonavir 600 mg. BID + saquinavir 400 mg. BID,
and two nucleoside reverse transcriptase inhibitors. Breakdown of nucleoside
combinations; zidovudine + lamivudine 20%, didanosine + stavudine 30%, and
stavudine + lamivudine 50%.
Methods: Safety monitoring, HIV RNA-PCR viral loads, and CD 4 lymphocyte
cell counts were obtained at baseline, weeks 2, 4, 8, 12, 16, and thereafter
followed every 8 weeks. Six patients were withdrawn secondary to adverse side
effects/Laboratory toxicity, and 3 were withdrawn secondary to non-adherence to
medication.
Results: The mean CD4 cell count at baseline was 79 cells/mm3 (Range
9-36), and mean HIV-RNA viral load was 4.86 log1o copies/ml (Range 3.29-.82).
At the 48 week point 87% of the 23 patients remaining in the study have achieved
and maintained a reduction of HIV-RNA viral load below 400 copies/ml (70% with
HIV-RNA viral load below the limits of detection at 25 copies/ml). Mean increase
in CD 4 cell count of 184 cells/mm3 at 48 weeks. 2 of 2 initial ritonavir failures,
2 of 2 initial indinavir failures, and 8 of 28 initial saquinavir failures experienced
failure with dual protease (ritonavir/saquinavir) therapy.
Conclusion: Long term follow-up reveals the combination of ritonavir and
saquinavir with 2 nucleoside reverse transcriptase inhibitors to be safe in the
majority of patients, however significant clinical and laboratory toxicity can occur.
Screening of potential patients for existing hepatic disease can identify individuals
with increased risk for toxicity. The data indicates that prior failure of regimens
containing ritonavir or indinavir may preclude success of salvage therapy with
ritonavir/saquinavir dual protease therapy.
32296 Patients harbouring both the multinucleoside analogue
resistance mutation Q151M and the AZT resistance
mutation T215Y/F carry these two mutations on different
HIV-1 quasispecies
Anne-Mieke Van Damme1, K. Van Laethem2, V. de Vroey2, J. Desmyter2, E. de
Clercq2, J.-C. Schmit3, T. Harrer4, L. Ruiz5, B. Clotet5, S. Sprecher6,
P. Hermans7. 1Rega Institute Minderbroedersstraat T10 B-3000 Leuven; 2Rega
Institute & University Hospitals Leuven; 6lnstitut Pasteur Brussels; 7Centre
Hospitalier Universitaire St. Pierre, Belgium; 3Laboratoire de Retrovirologie CRP
Sante Luxembourg, Luxembourg; 4Dept. of Internal Medicine III University of
Erlangen-Nurnberg Erlangen, Germany; 5Retrovirology Laboratory "IRSI
CAIXA" Hospital Universitari Badalona, Spain
Background: We and others have recently described a new set of mutations
respon-sible for high-level cross-resistance against all presently used nucleoside
analogue reverse transcriptase (RT) inhibitors, with as marker mutation Q151M
(Schmit et al., J. Infect. Dis. 174: 962-968, 1996). The mutation T215Y/F, together
with several other mutations, is responsible for the classical pattern of resistance
to AZT (zido-vudine). Using selective PCR (ARMS), we observed that some
patients have both the Q151M and the T215Y/F as a mixed population with the
wild type (WT).
Aim: The study intended to verify whether patient virus strains can harbour
both Q151M and T215Y/F, or whether these mutations are located on different
virus strains.
Methods: Q151M positive patients from an European screening program using
the 151 ARMS were evaluated for the presence of the T215Y/F using the 215
ARMS. The RT gene of patients carrying mutant virus at both positions was
cloned using the TOPO TA Cloning kit (Invitrogen). The initial PCR product and
several clones of each patient were sequenced.
Results: Four Q151M positive patients also carried T215Y or T215F. In three
of these patients, direct sequencing either identified the 215, or the 151 mutation.
For all patients, the presence of both mutations was confirmed by cloning and
sequencing of the clones. Cloned virus carried either the 215 mutation together
with other classical AZT-related mutations or the 151 mutation together with other
multinucleoside analogue resistance mutations, but never both sets of mutations
in one and the same virus quasispecies.
Conclusions: Selective PCR is more reliable than direct PCR sequencing to
detect mixtures of mutants at positions 151 and 215 of the HIV-1 RT. Patients
can carry both the classical AZT resistance mutations including T215Y/F and
the multinucleoside analogue resistance mutations including Q151M. However,

these mutations are found on different HIV-1 quasispecies.

Conclusions: Patients in CNAB 3001 had extensive ART experience. ApproxFurther analysis of the virology data will examine the effect of ABC administered
at 600 MG BID in experienced patients. Unblinded results will be examined
to determine relationships between virologic response and neuropsychological
performance.
| 32294 Pre-existent mutations of the HIV protease gene were
associated with therapeutic failures in patients on a
saquinavir/ritonavirnar therapy
Philip Cunningham1, G.R. Kaufmannt, D. Sayer2, D. Smith1, M. French2,
D.A. Cooper3. 1Centre for Immunology St. Vincent's Hosp, Sydney; 2Dept. Clin.
Immunology Royal Perth Hosp, Perth; 3Nat'l Centre for HIV Epidemiology,
Australia
Introduction: Protease inhibitors significantly improved HIV therapy. Nevertheless, we observed a therapeutic failure rate of 51% in an observational cohort of
HIV positive patients on saquinavir, ritonavir and two nucleosides.
Objective: To analyze the prevalence of mutations of the HIV protease gene
in this cohort at baseline, which may have influenced virological outcome.
Methods: HIV protease genotype was determined by dye labelled primer DNA
sequencing from viral RNA amplified from plasma samples of 17 patients. Patient
characteristics at baseline were: age: 40 ~ 7y, CD4+ T lymphocytes: 177 ~ 106
cells/mm3, viral load: 5.19 ~ 0.47 logio cop/ml. 8/17 patients were pretreated with
saquinavir, 2 with ritonavir and 16 were nucleoside experienced. After a therapy of
48 weeks 7/17 patients had undetectable viral load, 7 responded only transiently
to the treatment and 3 were non responders.
Results: 16/17 subjects demonstrated at least one point mutation (L63P: 14
patients, A71 V/T: 7, L101: 5, L90M: 5, V771: 3, V82A: 3, 154V: 2, M361: 1 and
G48V: 1). Multiple mutations were found in 10 individuals, of which 6 were pretreated with protease inhibitoinhibitors (6 mutations in 2 patients, 4 in 2, 3 in 2 and 2 in
4). Of the 6 saquinavir pretreated patients, 3 showed mutation L90M and both
ritonavir pretreated individuals had mutation V82A. 5 patients were successfully
treated despite mutations at positions 63 and 63 + 71, but therapy failed in all
subjects with L101, V82A or L90M. Three non-responders showed mutations
G48V+L63P+V71 I+V82A, L101+L63P and L101+A71T+V771.
Conclusion: We observed a high prevalence of mutations of the HIV protease
gene before commencing the combined saquinavir/ritonavir therapy. Pretreatment
with saquinavir and ritonavir may have led to mutations at position 90 and 82,
which were associated with treatment failures. Mutations at position 63 and 71
did not appear to reduce the efficacy of the saquinavir/ritonavir therapy.