Prospective Evaluation of Antiretrovirals in Resource Limited Settings

1
ProspectiveEvaluation ofAntiretrovirals
inResourceLimitedSettings
AIDS Clinical Trials Group Study A5175
2
Background Origins of PEARLS

• Data from randomized clinical trials of ARV
  efficacy come largely from resource rich areas of
  the world
• Antiretroviral (ARV) efficacy in resource-limited
  settings (RLS) could be affected by multiple
  factors including human genetics, culture,
  nutrition, and environment
• ACTG committee of multinational investigators
  began deliberations in 2002
• What to start?
• Common initial ARV regimen is thymidine
  analog3TCNNRTI
• Data on efficacy of alternative regimens in RLS
  needed
• PEARLS protocol finalized May 2004
• Primary Objective Demonstrate non-inferiority of
  once-daily PI- or NNRTI-containing regimens
  compared to a twice-daily ARV regimen for initial
  treatment of HIV-1 in diverse areas of the world

3
Study Design

• Step 1 (initial regimen) 111 randomization
• Arm 1A ZDV/3TC BID EFV QD
• Arm 1B ddI-EC QD FTC QD ATV QD
• Arm 1C TDF/FTC QD EFV QD
• Randomization stratified by country and screening
  plasma HIV RNA ( or lt 100K)
• Planned follow-up duration until 30 of
  participants have met the primary endpoint

4
Primary Endpoint

• Treatment failure (primary efficacy endpoint)
  defined by time from randomization to the first
  of the following (ITT analysis)
• Virological failure (confirmed plasma HIV-1 RNA
  1,000 c/mL at week 16 or later)
• HIV Disease Progression (new AIDS defining
  condition at least 12 weeks following study entry
  and not associated with IRIS)
• Death due to any cause

5
Study Population

• HIV-1-infected men and women gt 18 years of age
• Naïve to antiretroviral therapy (lt 7 days total
  SD NVP /- ZDV for pMTCT allowed)
• CD4 lt 300 in past 90 days
• 1,571 participants recruited from 12 sites in 8
  RLS countries (n1,361) and 31 US sites (n210)
  between May 2005 and August 2007

6
PEARLS Study Sites
7
Planned DSMB Review May 2008

• 5th regularly scheduled interim review of safety
  and efficacy data
• used pre-specified interim boundary of P0.002,
  and corresponding 99.8 confidence intervals
  about treatment comparison estimates
• Included all visits/specimens through March 9,
  2008
• Median follow-up 72 weeks
• Loss-to-follow-up 4 at 48 wks 6 at 96 wks
• No unexpected safety findings
• No conclusion could be reached about the
  comparison between Arms 1A and 1C
• Conclusive evidence that Arm 1B (ddI-ECFTCATV)
  is inferior to control Arm 1A (ZDV/3TCEFV) for
  the primary efficacy endpoint

8
DSMB Findings
1.67 (1.02, 2.75)
1.77 (1.04, 3.03)
3.00 (0.61,14.87)
0.99 (0.23, 4.26)
9
Participant Characteristics
10
Cumulative Probability of Treatment
FailureddI-ECFTCATV
11
Probability of FailureddI-ECFTCATV
12
Death and AIDS Progression Endpoints
ddI-ECFTCATV
13
Baseline Correlates of Treatment Failure
ddI-ECFTCATV
Exploratory, multi-covariate Cox regression that
included the following covariates Screening
CD4 cells Plasma HIV-1 RNA Gender, Age,
Race Baseline BMI Geographic location TB
history AIDS OI history Entry HBV surface
Ag Entry Albumin Self reported adherence
wks 1-12
14
Baseline Correlates of Treatment Failure
ddI-ECFTCATV
15
Conclusions

• The regimen ddI-ECFTCATV had significantly
  greater risk of treatment failure compared to
  ZDV/3TCEFV
• Participants currently taking ddI-ECFTCATV are
  being switched to an NNRTI-based study-provided
  regimen
• The relative efficacy of ddI-ECFTCATV varied by
  baseline factors including gender, age, CD4
  count, prior AIDS OI, prior TB and country

16
Perspectives, Caveats, and Future Studies

• 48 and 96 week treatment success rates for
  ddIFTCATV were 84 and 76
• Heterogeneity of treatment failure risk across
  countries should be interpreted cautiously
  analysis not corrected for multiple comparisons
• ddI-EC dose was based on weight and was
  administered on an empty stomach separately from
  ATV
• Pharmacology, drug resistance and adherence
  analyses are underway to help explain Arm 1B
  findings
• The comparison between Arms 1A and 1C is ongoing
  and follow-up of these arms continues without
  change
• Other PEARLS Study presentations MOPE0053,
  THPE0044, THPE0098, THPE0820

17
Acknowledgements

• PEARLS study participants (N1,571) and sites (N
  43)
• PEARLS Study Team (N 82)
• PEARLS Co-Chairs Tim Flanigan, James Hakim, N.
  Kumarasamy
• SDAC - Laura Smeaton, Victor DeGruttola
• ACTG Ops Ron Barnett, Barbara Brizz, Barbara
  Bastow, Laura Moran, Lara Hosey
• FSTRF Apsara Nair, Ann Walawander
• NIAID/DAIDS Karin Klingman, Edith Swann, Anna
  Martinez, Eva Smith
• Boehringer-Ingelheim Pharmaceuticals - Carolyn
  Conner, Marita McDonough
• Bristol-Myers Squibb Gary Thal, Jonathan Uy
• Gilead Sciences Jim Rooney, Audrey Shaw
• GlaxoSmithKline Keith Pappa, Elke Loeschel