Transcript Large Simple Trials of Vaccine Safety

EXPERIENCE WITH
BRIDGING STUDIES IN THE
CENTER FOR BIOLOGICS
EVALUATION & RESEARCH
Susan S. Ellenberg, Ph.D.
Office of Biostatistics and Epidemiology
Center for Biologics Evaluation &
Research, U.S. FDA
Kitisato University-Harvard School of Public
Health Symposium
October 28, 2003
1
ACKNOWLEDGEMENTS
• A. Dale Horne, Dr. P.H., Chief,
Vaccines Evaluation Branch, Division
of Biostatistics
• Karen L. Goldenthal, M.D., Director,
Division of Vaccines and Related
Biological Applications
2
BRIDGING STUDIES IN
VACCINE DEVELOPMENT
• Bridging studies are commonly performed
in vaccine research
• Outcomes usually evaluated
– Immune responses
– Safety-related events
• Bridging studies may evaluate
–
–
–
–
–
Effect of manufacturing change
Effect of formulation change
Effect of dose/schedule change
(Effect of other vaccines given concomitantly)
Validity of assumption that observed effect can
be generalized to other populations
3
IMMUNE RESPONSE
• Vaccines work by stimulating the
development of protective antibodies that
can protect vaccinated individuals when
they are exposed to disease-causing
bacteria and viruses
• “Immune response” refers to the rise in
the relevant antibody levels
• In many (but not all) cases, immune
response is a reliable surrogate for clinical
efficacy (protection from disease)
4
DESIGN OF VACCINE
BRIDGING STUDIES
• For studying effects of changes in
manufacturing, composition or method of
administration, randomized noninferiority
studies are generally conducted
• Examples:
– Randomize participants to receive vaccine
manufactured with original process or with new
process
– Randomize participants to receive vaccine on
original schedule or on new schedule
• Analyze results to determine if immune
responses are similar in randomized groups
5
RATIONALE FOR
RANDOMIZED BRIDGING
STUDIES
• Changes made in manufacturing, or
administration, could possibly affect
vaccine efficacy or safety
• Consistency of immune responses expected
to predict consistency of prevention of
clinical disease
• Occasionally, bridging studies identify
problems with new approach
– Combining acellular pertussis vaccine with Hib
vaccine had adverse effect on Hib immune
responses for some products
6
SAFETY EVALUATION
• Safety is always assessed in bridging
studies, but usually without formal
“bridging” criteria
• Changes in manufacturing are of particular
concern
– Worst case: Cutter incident, 1950’s
– Cutter was one of several manufacturers of
Salk polio vaccine, using killed virus
– Changes in manufacturing process for Salk polio
vaccine resulted in inadequate killing process;
Cutter vaccine caused hundreds of polio cases
7
BRIDGING ACROSS
POPULATIONS
• These bridging studies differ from those
to assess changes in manufacturing,
formulation or delivery
• Not possible to randomize country/ethnic
group!
• Generally done as non-randomized but
controlled studies, comparing immune
responses in the region where clinical
efficacy was demonstrated, to these
outcomes in a different region
8
RATIONALE FOR BRIDGING
ACROSS POPULATIONS
• Inefficient to do multiple large
clinical efficacy trials of new vaccines
• Often clinical efficacy trials are
possible only in certain regions
– Disease endemic in limited areas
– Existing vaccines in some areas
• High likelihood that vaccine effective
in one population will also be effective
in other populations
9
EVIDENCE FOR “ETHNIC
FACTORS” IN VACCINE
EFFECTS
• Haemophilus influenzae b (PRP-D)
– Vaccine shown to be highly effective in
Finland
– Poor efficacy shown in Alaskans
– Possible explanations:
• Risk of Hib disease higher in Alaska
• Occurs at earlier ages
• Stronger and more rapid immune response
required for protection in Alaska
10
APPROACH TO BRIDGING
STUDIES FOR VACCINE
LICENSURE IN THE U.S.
• Often not feasible to evaluate clinical
efficacy in the U.S.
• Licensed vaccines still needed for
travelers, military, certain
occupational categories
• Usual approach: perform clinical
efficacy study where disease rate is
relatively high, then “bridge” to U.S.
population with single-arm study of
new vaccine
11
DESIGN OF POPULATION
BRIDGING STUDIES
• Comparison of immune responses is
fundamental objective
• Primary outcomes:
– Per cent “responders” (immune response
above threshold predicting clinical
protection from disease)
– Ratio of geometric mean concentration
of antibodies
• Study designed to have good power to
rule out important difference in
parameters of immune response
• Safety outcomes also measured
12
OTHER
CONSIDERATIONS
• Since study must be observational,
try to make study in U.S. as similar as
possible to that in country where
primary study was performed
–
–
–
–
–
–
Manufacturing considerations
Age of participants
Concomitant vaccines administered
Schedule and route of administration
Surveillance for adverse events
Timing of blood draws for response
assessment
13
VACCINES LICENSED IN
U.S. WITH NON-U.S.
EFFICACY DATA
(NOT COMPLETE LIST)
• Acellular pertussis-containing
vaccines (DTaP)
• Oral polio vaccine
• Typhoid Vi Polysaccharide
• Japanese encephalitis
• Hepatitis A
14
COUNTRIES PROVIDING
PRIMARY CLINICAL EFFICACY
DATA FOR U.S. LICENSURE
•
•
•
•
•
•
Sweden
Italy
Finland
Germany
Indonesia
Thailand
•
•
•
•
•
•
South Africa
Nepal
England
Chile
Japan
Soviet Union
15
KEY CONSIDERATION:
STRAINS OF
INFECTIOUS AGENT
• Prevalent strains may differ from country
to country
• If vaccines protecting against all strains
prevalent anywhere cannot be developed,
different vaccines may have to be
developed in different regions; in such
cases, would not try to “bridge”
• Not routinely an issue, but likely will be
important for HIV vaccines in future
16