Lessons Learned from the Step Study

1
Lessons Learned from the Step Study

• Susan Buchbinder, MD
• HIV Research Section
• San Francisco Department of Public Health
• International AIDS Society Meeting
• Mexico City
• August 4, 2008

2
MRKAd5 trivalent vaccine

• Vaccine 111 admixture of 3 Ad5 vectors
• Encoded transgenes codon-optimized,
  near-consensus clade B HIV-1 sequences
• Placebo vaccine dilution buffer without Ad5

3
Trial Design

• 3000 high-risk HIV uninfected men and women
• Initial study 1500 pts w/ Ad5 NAb lt200 (Dec
  2004)
• Modification additional 1500 w/ Ad5 NAb gt200
  (July 2005)
• Randomization stratified by Ad5 lt18, 19-200,
  201-1000, gt1000
• Primary hypotheses Ad5 lt200 subset
• Decrease in HIV acquisition and/or
• Lower viral load setpoint (3 months
  post-diagnosis)
• Secondary hypotheses Total population
• Same as primary (Ad5 lt200 and Ad5 gt200 combined)

4
Step Study sites
5
Planned Interim Analysis Results Ad5lt200 MITT
Analysis
HIV Acquisition
Early viral RNA
6
Incidence (95 CI) of HIV Infection MITT
population (males)
18 is the LOQ for the Ad5 titer assay includes
all HIV cases thru Oct 17, 2007
7
Variables included in univariate/multivariate
analyses

• Vaccine vs. placebo
• Baseline Ad5
• Circumcision (self-report)
• Age
• Race
• Region
• Baseline risk factors (previous 6 months)
• male sex partners
• Unprotected receptive anal sex
• Unprotected insertive anal sex
• Substance use
• Self-reported sexually transmitted infection

8
Variables included in univariate/multivariate
analyses

• Vaccine vs. placebo
• Baseline Ad5
• Circumcision (self-report)
• Age
• Race
• Region
• Baseline risk factors (previous 6 months)
• male sex partners
• Unprotected receptive anal sex
• Unprotected insertive anal sex
• Substance use
• Self-reported sexually transmitted infection

significant interaction with vaccine vs. placebo
9
Estimated Relative Risk of HIV Infection Vaccine
Placebo(95 CI)
Men with unknown circumcision status (49, 2.7)
were excluded from analyses. All analyses are
based on the Cox proportional hazards regression
model for time-to-event data.
10
Hazard of HIV Infection MITT population
(males)Crude estimation method using 26 week
intervals
Time interval of estimated HIV infection in weeks
relative to randomizationSummaries exclude 1
female infection (placebo group with Ad5 18).
MITT population includes all HIV cases diagnosed
after baseline.
11
Next Steps

• Ongoing follow-up of study volunteers
• All pts unblinded
• gt95 retention since results announced
• HVTN 504 to
• Evaluate longer-term effects
• Provide risk reduction counseling, linkage to
  prevention services
• Collect additional data and specimens
• Awaiting data on
• HSV-2 status, HLA typing, sexual network
  clustering
• Multivariate analysis with longitudinal risk
• Laboratory investigation of reasons for failure,
  control of viremia
• Opening up data, specimens to external
  investigators
• External Scientific Review Committee prioritizes
  laboratory studies, specimens

12
Important Scientific Lessons from Step

• Demonstrated utility of test-of-concept trials
• Quicker awareness of beneficial and adverse
  effects
• Recalibrated the NHP challenge model
• SHIV 89.6P not useful
• Consideration of genetic factors (MAMU-A01)
• New questions not easily addressed in NHP studies
• Importance of vector immunity, incl.
  tissue-specific responses
• Transmission across mucosal barriers
• Data and specimens to directly address failure to
  protect
• Need to develop validated, functional assays as
  correlates
• Exploration of breakthrough infections
• Intriguing leads in potential protection in
  subgroups

13
Important Community Lessons from Step

• Transparency is critical
• Prepare (as best you can) for the unexpected
• Communication w/ investigators, participants,
  public
• Each prevention trial impacts on all others
• Community is important at each stage of a trial
• Study volunteers are critical to our search for
  an HIV vaccine

14
Conclusions

• The Step Study was a pivotal trial for the HIV
  vaccine field
• Definitive results on vaccine efficacy reached 33
  months after first participant enrolled
• Important contributions that could only have been
  learned through clinical efficacy trials
• Lessons continue to be mined from trial data and
  specimens.

15
The Study Volunteers
Acknowledgments

• For their dedication and commitment in the search
  for an HIV vaccine

16
STEP Study Protocol Team

• Sarah Alexander
• Susan Buchbinder
• Gail Broder
• Lisa Bull
• Danny Casimiro
• Ann Duerr
• Cheryl Ewing
• Dan Fitzgerald
• Paula Frew
• Lori Gabryelski
• Peter Gilbert
• Tirzah Griffin
• Soyon Im
• Dale Lawrence

• Rosario Leon
• David Li
• Ellen MacLachlan
• Julie McElrath
• Devan Mehrotra
• Robin Mogg
• Dewayne Mullis
• Gabriela ONeill
• Mary Pleier
• Michael Robertson
• Steve Self
• Rosario Leon
• Amanda Vettori
• Steve Wakefield
• Amy Zhou