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

68 Abstracts 12278-12281

12th World AIDS Conference

has potent antiviral activity and a favorable pharmacokinetic profile. The structure
of the enzyme-inhibitor complex should lead to a better understanding of the
structural basis for its tight binding properties and to its resistance profiles.
Methods: The three-dimensional crystal structures of JE-2147 and an analog
bound to both wild type and 184V mutant HIV protease were determined to
2.1-1.9 A resolution with R-factors of 18.6-19.0%.
Results: JE-2147 is structurally related to KNI-272 but contains a methylbenzyl
group in the P2' position that makes novel interactions with the wild type enzyme.
Two water molecules make bridging hydrogen bonds between the inhibitor and
the enzyme. The conformation of the P2' o-methylbenzyl substituent of JE-2147
in the complex with the 184V mutant differed from that of wild type. In addition,
rearrangements of the ST side chains from the mutant and wild type enzymes
were observed. An analog of JE-2147 that contains a P2' t-butyl substituent
shows increased levels of resistance to the 184V mutant protease, but did not
exhibit the structural rearrangements seen with JE-2147, in complexes with the
mutant enzyme, presumably because of the rigidity of the bulky t-butyl group.
Conclusions: The comparison of the crystal structures of JE-2147 and an
analog bound to wild type and 184V mutant HIV proteases indicate that molecular
flexibility is an important aspect in the design of inhibitors which are effective at
binding to both wild type and mutant HIV protease.
122781 Improved survival among men and women on triple
antiretroviral therapy
Robert Hogg, B. Yip, K.J.P. Craib, M.T. Schechter, M.V. O'Shaughnessy,
J.S.G. Montaner. BC Centre for Excellence in HIV/AIDS, 608-1081 Burrard St.
Vancouver, BC, V6Z 1Y6, Canada
Objective: To characterize survival following the initiation of double and triple
combination antiretroviral (ARV) therapy among HIV+ individuals with free access
to ARV medications.
Methods: In BC, antiretroviral therapies are distributed free of charge according
to specific therapeutic guidelines. Study subjects were antiretroviral naive, initially
prescribed any combination of 2 NRTIs including either d4T or 3TC (ERA-II) from
06/93-06/96 or 2 NRTIs and a PI (ERA-Ill) from 01/96-12/96. The outcomes in
this analysis were death and a primary diagnosis of AIDS. Statistical methods
followed the intent-to-treat principle with subjects being retained in their initial
treatment groups. Primary and secondary outcomes were examined from the
start of ARV therapy. Survival analyses were conducted using K-M methods.
Event free subjects were right censored at 30/06/97 and 12/31/97 for ERA-II and
ERA-III respectively or at last known contact.
Results: A total of 520 (376 ERA-II, 144 ERA-III) men and women were eligible
for this analysis. As of 12/31/97, there were a total of 44 deaths (41 ERA-II,
3 ERA-III), yielding a crude mortality rate of 8.5%. Product limit estimates of
the cumulative mortality rate at one year were 8.3% (~ 1.4%) and 2.1% (~
1.1%) for ERA-II and ERA-III subjects respectively (p = 0.004). After adjusting
for AIDS diagnosis, CD4 cell count, gender and age in a multivariate model,
ERA-II participants were 4.79 times (95% Cl: 1.47, 15.65; p = 0.009) more likely
to die than those in ERA-III after antiretroviral therapy initiation. In our second
multivariate model we restricted our analysis to 413 AIDS-free individuals. Again
after adjusting for the CD4 cell count, gender and age, we found ERA-II 3.78
times more likely to die or progress to AIDS (95% Cl: 1.15, 12.40; p = 0.028 than
ERA-III subjects.
Conclusion: Our analysis demonstrates, in a population-based study of participants enrolled in a province-wide anti-HIV treatment program, a significant
reduction in mortality and AIDS-free survival for those individuals initiated on
triple drug therapy. These results remain significant even after adjusting for a
number of salient characteristics including baseline AIDS, CD4 cell count, gender
and age. Overall, our results demonstrate the effectiveness of triple drug therapy
regimens outside the context of a large scale clinical trial.
S12279 1 Two-year durability of HIV-1 load suppression in patients
treated with indinavir who experience virus load declines
to <500 vRNA copies/mi
Daniel Holder, M. Shivaprakash, W.A. Schleif, J.A. Chodakewitz, J.H. Condra,
E.A. Emini. Merck Research Laboratories, 770 Sumneytown Pike, West Point,
PA, USA
Objectives: For patients on indinavir (IDV)-containing therapy, resistant virus
selection is, in general, slower in patients who succeed in suppressing virus to
<500 vRNA copies/mi. We assessed loss of virus suppression over time due to
resistant virus selection, in the population of patients who achieve such low viral
levels on IDV containing therapy.
Methods: The data were obtained from a number of IDV phase II studies (n =
106, initial therapies: 53 on IDV monotherapy, 13 on IDV + AZT, 13 on IDV + AZT
+ ddl, and 27 on IDV + AZT + 3TC). Patients whose initial therapy included IDV
and who demonstrated a decline in vRNA to <500 vRNA copies/mi before the
24th week of therapy and maintained the response for 12 weeks were included in
the analysis. Patients whose vRNA subsequently increased and remained above

500 vRNA copies/ml for any reason were classified as having a loss of viral
suppression. Loss of suppression was considered due to selection for resistant
virus if the circulating viral genotypes of such a patient contained IDV resistance
mutations.
Results: We estimate the percentage (with 95% confidence limits) over time
(measured from the point that the virus load was first determined to be below
500) of patients who lose virus suppression due to resistant virus selection to be

10% (4%, 15%), 12% (5%, 18%), and 20% (11%, 28%) at 48, 96, and 120 weeks,
respectively. Corresponding estimates of the percentage of patients with a loss of
virus suppression for any reason are 13% (6%, 19%), 22% (14%, 30%), and 32%
(22%, 41%) for 48, 96, and 120 weeks, respectively. Numerically, patients were
less likely to lose suppression if initial IDV therapy was given in combination with
AZT + 3TC (12% due to resistance selection, 16% overall at 108 weeks) than if
IDV was initially given as monotherapy (20% due to resistance selection, 31%
overall at 108 weeks), although the difference is not statistically significant.
Conclusion: These results demonstrate that the large majority of patients
whose virus load declines to below 500 vRNA copies/ml for at least 12 weeks,
as a result of therapy containing IDV, are likely to sustain the suppression for at
least two years.
S12280 1 Protease inhibitors are associated with declining AIDS
deaths in New York City (NYC)
Thomas Wong1, A. Reggy2, M.A. Chiasson3, R. Simonds3, V. Loo3, J. Heffess3.
1Columbia University; Columbia Presbyterian Med Center 622 W168 St. VC
12-225 New York, NY 10032; 2Centers for Disease Control Prevent, Atlanta, GA;
3New York City Department of Health, New York, NY, USA
Background: In NYC, the number of HIV/AIDS deaths dropped 30% between
1995 and 1996. We evaluated the association between combination antiretroviral
therapy with protease inhibitors (CPI) and mortality among AIDS patients.
Methods: In this population-based case-control study, a random sample was
selected from the NYC HIV/AIDS reporting system (HARS). Cases (n = 150)
and controls (n = 150), between the ages of 25 and 59, were diagnosed with
AIDS in NYC during 1994-1996 and proportionately sampled by year of AIDS
diagnosis. Women were oversampled (F:M = 1:2). Cases were those who died
of AIDS-related causes in 1996. Controls were not known to have died by the
end of 1996. Data were obtained for 1995-1996 from HARS, death certificates,
CD4 lab reporting system, and medical charts. Univariate and multivariate logistic
regression analyses were used to identify independent predictors of dying from
AIDS.
Results: Data are complete on 121 cases and 123 controls (total = 244).
Cases and controls were similar in age (mean, 40 vs. 39 yrs), sex (68% vs. 66%
male); proportion of Black or Hispanic (81% vs. 85%), of injecting drug users
(54% vs. 51%), and of uninsured (13% vs. 13%). Median CD4 count at AIDS
diagnosis was 30/pL for cases and 106//GL for controls (p < 0.001). In 1996, 12%
of cases and 29% of controls were on CPI (Crude OR = 2.9, 95%CI = 1.4-6.0;
Mantel-Haenszel OR = 2.7, 95%CI = 1.3-5.0 stratified by CD4 count). A logistic
model was used to assess gender, age, race, health insurance, HIV risk, year of
AIDS diagnosis, CD4, and antiretroviral therapy. Independent predictors for dying
from AIDS were: non-use of CPI (adjusted OR = 3.1, 95%CI = 1.2-8.0), and low
CD4 count (adjusted OR = 1.7 for every 50 cell decrease, 95% CI = 1.3-2.0).
Conclusions: This population-based study suggests that AIDS death is strongly
associated with not using protease inhibitor containing combination antiretroviral therapy. This association supports the likelihood that the decline in NYC
AIDS deaths in 1996 resulted at least in part from increased use of combination
antiretroviral therapy with protease inhibitors.
S12281 1 Limited efficacy of nelfinavir after failure of other
protease inhibitors (PI)
Cecile I. Poggi', A. Lafeuillade2, G. Hittinger2, L. Chollet3, A. Rieu2,
M.P. Pradie3, N. Profizi3. Laboratorie Bilogie, C.H..TS. Toulon-La Seyne Sur
Mer; 2Hopital Chalucet, Toulon; 3Hopital De Toulon-La Seyne, La Seyne Sur
Mer, France
Objectives: to assess the antiviral response to nelfinavir used after failure of
2 reverse transcriptase inhibitors (RTI) + another PI, and to correlate it with
genotypic resistance.
Design: open-label study with monthly evaluation.
Methods: 20 patients (pts) with plasma RNA > 10 000 copies/ml (Amplicor
Roche) treated with 2 RTI + Saquinavir or Indinavir or Ritonavir were switched to
nelfinavir with continued 2 RTI (changing one of them when possible). Genotypic
resistance patterns were analysed using the LiPA (for RTI) and viral sequencing
(for PI) on plasma RNA.
Results: the overall duration of prior PI use was 16 ~ 2 months. At baseline,
mean CD4+ count was 166 ~ 29 x 106/L and mean plasma RNA was 5.20 ~ 0.14
log copies/mi. Mid term follow-up is available at present for the first 15 pts (mean:
4 + 1 months). No significant decrease in plasma RNA levels has been observed
in 12 pts. In 3 pts, levels decreased regularly after 2 (-2.1 log), 3 (-1.9 log) and 4
months (-2.7 log) of follow-up. To date, the protease gene has been sequenced
before nelfinavir in 9 pts. There were an average of 6 ~ 1 mutations known to be
associated with PI resistance: codons 10 (2 pts), 12 (1), 46 (4), 54 (2), 63 (9) 71
(7), 77 (3), 82 (3), 84 (2), 90 (6). After 4 ~ 2 months on nelfinavir, sequencing has
already been done in 7 pts. None had a D30N mutation. There were an average
of 8 ~ 2 mutations associated with PI resistance in non-responding pts compared
to 1.5 ~ 0.5 in those with decreased plasma RNA.

Conclusion: nelfinavir resistance has been linked to a D30N mutation. However, in this group of heavily pre-treated pts, salvage therapy with nelfinavir was
associated with a high rate of failure linked to other mutations selected by prior
PI use. Only 3 pts with less than 2 mutations able to confer cross-resistance to
this drug at baseline showed a decrease in plasma RNA levels in the months
following the switch to nelfinavir.