 |
|
|
| |
Efavirenz-based Regimens are Potent in Treatment-Naïve Patients Across a Wide Range of Pre-Treatment HIV-1 RNA and CD4 Cell Counts
|
| |
| |
Secondary Analysis of ACTG 5095
Reported by Jules Levin
Dan Kuritzkes substituting for Heather Ribaudo reported the oral talk at the Late Breaker oral session on Thursday Aug 17 at IAC 2006.
Summary by Authors:
After a median followup of 3 years;
--over 80% of patients had viral load <50 copies/ml
--median CD4 increase was 278 cells
Baseline viral load and CD4 count were not associated with treatment outcome. (In other words, regardless of what your viral load was at baseline or your CD4 count patients had the same responses).
These results demonstrate the potency of EFV-based regimens across a wide range of CD4 cell counts and viral load.
Addition of a third NRTI did not enhance responses in any CD4 or viral load subgroup
Note from Jules Levin: Some patients had >300,000 viral load at study entry and some patients had <50 CD4 count. These patients were just as likely to achieve <50 copies/ml and to achieve a high CD4 increase as patients who started therapy with low viral load or high CD4 count. In fact, patients with >300,000 copies/ml of viral load had the highest increase in CD4 count after 3 years on therapy, and it appeared on the graph that patients with <50 CD4 count at baseline achieved the greatest CD4 increase by 3 years. This is what can happen when using a potent regimen.
Kuritzkes provided this background.
Regimens including 2 NRTIs plus EFV are widely used as initial treatment for HIV-infected persons. Concern persists about virologic potency in advanced disease:
--high viral load (VL)
--low CD4 cell counts.
ACTG 5095: Study Design
Double-blind, randomized, placebo-controlled study of protease inhibitor-sparing regimens:
--AZT/3TC/abacavir (ABC) (3 drugs)
--AZT/3TC/+EFV (3 drugs)
--AZT/3TC/ABC+EFV (4 drugs)
Study population
--ARV-naïve HIV-1 infected patients
--HIV-1 RNA >400 c/ml; any CD4 cell count
--stratified by screening HIV-1 RNA; < or >100,000 c/ml
All subjects to be followed regardless of treatment change for 96 weeks after last randomization.
5095 History
1147 subjects randomized between March 2001 and November 2002.
AZT/3TC/ABC arm halted by DSMB in Feb 2003:
--AZT/3TC/ABC inferior to each of the 2 EFV-containing regimens (Gulick et al, NEJM 2004;350:1850-61)
Follow-up of EFV-containing arms extended until March 2005
--no difference between AZT/3TC+EFV and AZT/3TC/ABC+EFV over 3 years median follow-up
Gulick et al, JAMA 2006;296:769-81
CURRENT ANALYSIS: OBJECTIVES
To compare the effect of baseline VL and cd4 count on virological and immunological responses in EFV-treated subjects in A5095.
To assess whether adding a third NRTI improved responses in any subgroup.
Statistical Methods
Subjects were categorized into pre-determined subgroups by baseline VL and CD4 cell count.
Associations of VL and CD4 subgroups with risk of virologic failure used Kaplan-Meier and Cox proportional hazards models:
-virologic failure defined as confirmed VL>200 c/ml.
Associations with CD4 increases from baseline at weeks 48, 96, and 144 used normal linear models.
Analyses presented are intent-to-treat:
--as-treated analyses also performed.
Study Subjects
-- N=765 subjects enrolled in the 2 EFV arms.
-- Median follow-up of 144 weeks (about 3 years)
-- 19% women, 81% men
-- 21% Hispanic, 35% black, 41% white, 2% other
Time To Virologic Failure
By baseline VL - ITT
Kuritzkes said the Kaplan-Meier curve which he displayed showed that stratified by baseline viral load there was no difference between the different viral load groups.
Time to Virologic Failure
By baseline CD4 - ITT
On the Kaplan-Meier curve there appeared it be a slightly higher rate of failure by the highest CD4 group (>500). This was explained by patients who enrolled in the study with very high CD4 counts and decided later they did not want to be on therapy. The As-Treated analysis shows no difference in Time to Virologic Failure between the different CD4 groups.
Risk for Virologic Failure by baseline CD4 and VL
Kuritzkes displayed a graph table of the Cox Proportionate Hazard Model, and this showed no difference, no increase in Hazard Ratio, in Risk for Virologic Failure between the various VL groups (p=0.63) and CD4 groups (p=0.71). They used the high viral load group of 300,000 and the low CD4 group of <50 as the reference and no lower VL group nor any higher CD4 group had any less Risk of Virologic Failure. In other words, there was no difference for Risk of Virologic Failure between any VL group, high to low, or any CD4 group, low to high, using the Cox.
3-drug and 4-drug regimen had same Risk for Viral Failure Regardless of CD4 or Viral Load at Baseline
Kuritzkes displayed the graph-table reporting findings using the Cox Proportionate Hazard Model to compare the risk of failure for the 3-drug group (AZT/3TC/ABC) vs the 4-drug group (AZT/3TC/ABC+EFV) by the various VL & CD4 stratifications. There was no difference overall in risk for viral failure between the 3 and 4 drug regimens, and there was no difference in risk of virologic failure between the 4 vs 3 drug groups by the various VL stratas at baseline (p=0.43) or baseline CD4 count groups (p=0.99). In other words, the 4-drug and 3-drug regimens used in this study had the same risk of viral failure, the 4-drug & 3-drug regimens used in this study, were the same in risk for viral failure regardless of which CD4 or VL stratification at baseline.
Viral Load <50 c/ml
By baseline VL - ITT
During the first 24 weeks the line curves in the graph displayed there was a difference in the time it took to reach <50 between with the higher viral load groups taking longer to reach full suppression to <50, Kuritzkes said. The lower the viral load at baseline the quicker you breached <50 and the higher the VL the longer it took to reach <50. But, by week 48 there appeared to be no difference between the VL groups in proportion of patients who achieved below 50 copies and who maintained this throughout the followup period of 144 weeks.
Viral Load <50 c/ml
By baseline CD4 -ITT
There was no difference in time to reach <50 c/ml based on baseline CD4 count during the first 24 or 48 weeks nor throughout the followup period of 144 weeks.
CD4 Change from Baseline
By baseline VL - ITT
Patients who entered the study with >300,000 copies/ml of VL at baseline had the greatest increase in number of CD4 cells. They had an average increase of 350 cells from baseline to week 144. There was no difference between the other baseline VL groups in increase in CD$ counts throughout the 144 week followup period. Throughout the entire 144 week period there was a steady increase in CD4 count for all baseline VL groups, and there was an average increase of 250 CD4 cells for patients in the other baseline VL stratas.
CD4 Change from Baseline
By baseline CD4 - ITT
There was little difference in change in CD4 count over the course of the 144 week followup period between the baseline CD4 strata with CD4 increases at week 144 between 250 to 300 for all baseline CD4 groups. Although patients with <50 CD4s at baseline did appear to achieve the highest CD4 count at week 144, of 300. The patient group with >500 CD4s displayed less increases in CD$ counts because of the numbers of patients in this group who dropped out of the study because decided they did not want to be treated.
|
| |
|
|
|
|
|
