Aucun titre de diapositive

1
THPE0167
Effects of nucleoside analogues versus ritonavir
boosted protease inhibitors on lipid levels
analysis of 12 clinical trials in 4231
antiretroviral naïve patients
Andrew Hill, Pharmacology Research Laboratories,
University of Liverpool, 70 Pembroke Place,
Liverpool L69 3GF, UK, e-mail microhaart_at_aol.com
Will Sawyer, South East Asian Research
Collaboration with Hawaii (SEARCH), Bangkok,
Thailand. Brian Gazzard, Chelsea and Westminster
Hospital, London.
Methods
The aim of this analysis was to estimate the
effect of different ritonavir boosted protease
inhibitors on lipids in naïve patients, using a
standardised format to report lipid elevations.
To be included in the analysis, clinical trials
needed to include (i) at least 50 antiretroviral
naïve patients studied for at least 48 weeks (ii)
use of a ritonavir boosted protease inhibitor
(using a total daily ritonavir dose of 100 -
200mg), in combination with a fixed pair of
nucleoside analogues, and (iii) with data
available on mean levels of total cholesterol,
triglycerides, LDL and HDL at baseline and at
Week 48. In multivariate analysis using inverse
variance weighting and analysis of covariance,
the relative effects on lipid parameters from the
protease inhibitors versus the nucleoside
analogues were assessed. The percentage rises in
lipid parameters in the atazanavir/ritonavir and
lopinavir/ritonavir arms of the CASTLE trial were
used in a sensitivity analysis. The analysis was
performed using Statistical Analysis Software
(SAS).
Results
Table 1 shows a summary of the 12 clinical trials
included in the analysis. Mean levels of the
four lipid parameters at baseline and week 48 are
shown in Table 2. The percentage of patients
using lipid lowering drugs was 4.5 in the SHARE
trial Elion 2007, 11 in the KLEAN trial Eron
2006, 5 in the GEMINI trial Walmsley 2007,
and 2.6 in the ABT-418 trial Johnson 2006.
In multivariate analysis, there was no
significant difference in lipid elevations
between either saquinavir, darunavir or
atazanavir when boosted with ritonavir. Also
there was no significant difference in lipid
elevations between fosamprenavir-ritonavir and
lopinavir- ritonavir. For the nucleoside
analogues, there was no significant difference in
lipid elevations seen between abacavir or
stavudine when given with lamivudine. Figure 1
shows the mean percentage rises in the four lipid
parameters, for the combined data from each
ritonavir-boosted protease inhibitor, stratified
for the use of either tenofovir/emtricitabine or
other nucleoside analogues. Summary results
from the multivariate analysis are shown in Table
3. For total cholesterol and triglycerides, the
rises were greater for trials of LPV/r and FPV/r,
relative to those using ATV/r, DRV/r or SQV/r
(plt0.001). However the use of either d4T or
abacavir as NRTIs also led to higher elevations
in total cholesterol and triglycerides, relative
to the use of tenofovir (plt0.001). For HDL and
LDL, there was no evidence for differences
between the boosted PIs. The confidence
intervals around estimates of change were wide
for both parameters, for each of the PIs.
There was also evidence for greater rises in
both LDL and HDL for patients taking either
abacavir/lamivudine or stavudine/lamivudine,
relative to those using tenofovir/emtricitabine
(plt0.001). This result was similar when all d4T
data was excluded from the analysis
(plt0.001). The comparison of results from the
multivariate analysis of the 12 trials versus the
equivalent lipids data from the CASTLE trial is
also shown in Figures 2-4. The mean percentage
rises in total cholesterol, triglycerides and LDL
were very similar for the atazanavir/ritonavir
and lopinavir/ritonavir arms of the CASTLE trial,
compared with the equivalent treatment groups in
the multivariate analysis. The mean rises in HDL
both the ATV/r and LPV/r treatment arms of the
CASTLE trial appeared higher than for the
multivariate analysis.
Conclusions
This meta-analysis of 12 clinical trials of
first-line HAART in 4231 antiretroviral naïve
patients suggests that both the choice of both
protease inhibitor and nucleoside backbone can
affect lipid elevations. Those receiving
lopinavir-ritonavir or fosamprenavir-ritonavir
showed significantly greater rises in total
cholesterol and triglycerides than use of the
other three ritonavir boosted proteases
saquinavir, darunavir or atazanavir. Rises in
LDL and HDL did not differ by choice of PI. The
use of either abacavir-lamivudine or
stavudine-lamivudine also led to significantly
greater rises in all four lipid parameters by
week 48 than the use of tenofovir-emtricitabine.
The lipid rises seen after 48 weeks of treatment
with TDF/FTC plus either DRV/r, ATV/r or SQV/r
in the meta-analysis were very similar to those
seen for TDF/FTC plus ATV/r in the CASTLE trial.
This observation is consistent with the head-to
head comparisons of lipids for DRV/r and ATV/r in
a study of healthy volunteers (C159).
Clinical Trials References
Abbott 418 Johnson et al. J Acquir Immune Defic
Syndr 2006, 43 153-160 Abbott 863 Walmsley et
al N Engl J Med 2001, 3462039-2046 Abbott-730
Gathe J et al. 15th CROI, February 2008, Boston,
USA abstract 775 ALERT Smith K et al. 4th IAS
Conference, Sydney, Australia, July 2007
abstract WEPEB023 ARTEMIS DeJesus E et al.
ICAAC, Chicago, USA September 2007 abstract
H-718b BMS-089 Malan N et al. 13th CROI,
Denver, Colorado, February 2006
abstract CASTLE Molina J et al. 15th CROI,
February 2008, Boston, USA abstract 37 GEMINI
Walmsley S et al. 11th EACS Conference, Madrid,
Spain, October 2007 abstract PS1/4 HEAT Smith
K et al. 15th CROI, Boston, USA, February 2008
abstract 774 KLEAN Eron J et al. Lancet 2006,
368 476-482, 2006. REDUCE Hicks C et al 11th
EACS Conference, Madrid, Spain, October 2007
abstract PS 7.01 SHARE Elion R et al. 4th IAS
Conference, Sydney, Australia, July 2007
abstract WEPEB033 SOLO  Gathe J et al. AIDS
2004, 18 1529-1537 TMC114-C159 Tomaka F et al.
HIV DART, Cancun, Mexico, December 2006.
abstract 85
XVII World AIDS Conference, Mexico City, August
2008 THPE0167