Opportunities and Obligations: Vaccine Safety in the Genomics Era

1

• Opportunities and Obligations Vaccine Safety
  in the Genomics Era

Robert Davis, MD, MPH Director Immunization
Safety Office CDC
2

• Opportunities and Obligations Vaccine Safety
  in the Genomics Era

• History of vaccine safety issues
• Vaccine safety infrastructure
• New/future field of vaccine genomics
• Some examples
• Future opportunities

3
(No Transcript)
4
Disease Pre-vaccine Era
Year 1999
change
Diphtheria 206,939 1921 1 -99.99 Measles 89
4,134 1941 86 -99.99 Mumps 152,209 1968 35
2 -99.76 Pertussis 265,269 1934 6,031
-97.63 Polio (wild) 21,269 1952 0 -100.00 Rubel
la 57,686 1969 238 -99.58 Cong. Rubella
Synd. 20,000 (1964-5) 3 -99.98 Tetanus 1,56
0 1948 33 -97.88 Invasive Hib
Disease 20,000 1984 33 -99.83
5
Disease Pre-vaccine Era
Year 1999
change
Diphtheria 206,939 1921 1 -99.99 Measles 89
4,134 1941 86 -99.99 Mumps 152,209 1968 35
2 -99.76 Pertussis 265,269 1934 6,031
-97.63 Polio (wild) 21,269 1952 0 -100.00 Rubel
la 57,686 1969 238 -99.58 Cong. Rubella
Synd. 20,000 (1964-5) 3 -99.98 Tetanus 1,56
0 1948 33 -97.88 Invasive Hib
Disease 20,000 1984 33 -99.83
6
Disease Pre-vaccine Era
Year 1999
change
Diphtheria 206,939 1921 1 -99.99 Measles 89
4,134 1941 86 -99.99 Mumps 152,209 1968 35
2 -99.76 Pertussis 265,269 1934 6,031
-97.63 Polio (wild) 21,269 1952 0 -100.00 Rubel
la 57,686 1969 238 -99.58 Cong. Rubella
Synd. 20,000 (1964-5) 3 -99.98 Tetanus 1,56
0 1948 33 -97.88 Invasive Hib
Disease 20,000 1984 33 -99.83
7
Disease Pre-vaccine Era
Year 1999
change
Diphtheria 206,939 1921 1 -99.99 Measles 89
4,134 1941 86 -99.99 Mumps 152,209 1968 35
2 -99.76 Pertussis 265,269 1934 6,031
-97.63 Polio (wild) 21,269 1952 0 -100.00 Rubel
la 57,686 1969 238 -99.58 Cong. Rubella
Synd. 20,000 (1964-5) 3 -99.98 Tetanus 1,56
0 1948 33 -97.88 Invasive Hib
Disease 20,000 1984 33 -99.83
Vaccine Adverse Events 0 11,827

8
But.

• No vaccine is 100 percent safe.
• As more people are vaccinated, diseases decrease
  or even disappear
• But real - and perceived - vaccine side effects
  increase.
• Public concern about the safety of vaccines
• Decreased vaccination levels
• Disease epidemics
• Alternatively, high profile disasters shake
  public confidence in vaccine (and drug) safety
• Swine flu vaccine campaign GBS
• Rotavirus vaccine intussusception
• Vioxx myocardial infarction
• Lead to increased development costs, regulatory
  burden, and increased disease burden

9

• Whooping Cough Notifications and Vaccine
  Coverage, England and Wales 1960-93

10
Recent research supports safety of vaccine policy

• No increased risk for autism after MMR vaccine
• No increased risk for autism among children
  receiving high doses of thimerosal
• No increased risk for multiple sclerosis or optic
  neuritis after hepatitis B vaccine
• No increased risk for inflammatory bowel disease
  after MMR or MCV

11
Recent Research

• But, many parents think we are asking the wrong
  questions
• They dont really care about population based
  studies _if_ they think they are somehow at
  risk
• Recent finding Jan 2006 of autism gene from
  Utah study
• Likely no single gene dictates autism
• More likely that each gene - individually -
  raises the risk
• Parents want to know If my child has one or more
  risk genes, will the vaccines trigger autism (or
  some other problem)
• This is a very good question

12

• Parents want to know If my child has one or more
  risk genes, will the vaccines trigger autism (or
  some other problem)
• This fundamental idea will my genes modify the
  effect of an exposure - is the biggest question
  today in medication safety as well as in
  pharmaceutical development
• Are there some drugs that work better/are safer
  in some people

13
Personalized Genetic Medicine

• Personalized medicine personalized drug
  delivery under intense study by
  NIH/NHGRI/pharmaceutical industry
• Efficacy
• Beta blockers work better among carriers of
  specific genes (which also differ by race)
• Safety
• Albuterol may not work - and may even be harmful
  - among asthmatics with specific genetic
  variations

14
Personalized Genetic Medicine

• Diagnostics
• The AmpliChip CYP450 test
• first microarray-based pharmacogenomic test in
  clinical setting.
• provides information on enzyme activity of the
  CYPC19 and CYP2D6 genes - genes that play
  particularly important roles in the metabolism of
  a large number of widely prescribed medications
• more accurate dosing safer dosing

15

• Personalized medicine personalized drug
  delivery under intense study by
  NIH/NHGRI/pharmaceutical industry
• Efficacy
• Beta blockers work better among carriers of
  specific genes (which also differ by race)
• Safety
• Albuterol appears to work less well, and may even
  be harmful, among asthmatics with specific
  genetic variations at the pnp gene
• Diagnositcs
• AmpliChip CYP450 test
• For vaccine Personalized approach is to
  understand which people are at risk for
• Vaccine adverse events
• Vaccine failure

16

• Pharmacogenomics vs. Vaccine-genomics

Pharmacogenomics Personalized therapies for
acute/chronic conditions Typically, response to
medication is observed, or can be measured Side
effects and adverse medication events common
medication discontinuation, illness, lawsuits
large incentive for personalizing
delivery Vaccine-genomics Personalized
vaccines to improve safety profile or vaccine
responsiveness Vaccine response rarely measured
in real world. Serious side effects or vaccine
AE very rare. Little economic incentive for
manufacturers to lead way for personalizing
delivery
17
Goals To understand the genetic variations that
predispose children, adolescents or adults to
vaccine adverse events or vaccine failure
18

• The typical study approach

Case-control study (rare outcome) Cases
children with seizures following MMR
vaccination Controls children vaccinated with
MMR who did not experience seizures Assess
genetic differences between cases and controls,
using either candidate genes or whole genome
approach Optimally identify a single
polymorphism or group of polymorphisms very
common in cases, uncommon in controls
19

• How would results be applied?

If able to identify a single polymorphism or
group of polymorphisms very common in cases yet
uncommon in controls (ie high RR for
disease) Assess predictive power of
polymorphism(s) when applied to population How
many people need to be identified excluded from
vaccination to prevent one seizure? Quantify
risks and benefits of excluding children/adults
from vaccination May be different depending on
vaccine, outcome, likelihood of exposure to wild
type disease, presence of herd immunity,
etc Ex MMR and seizures Smallpox
vaccine and myocarditis Study/identify risk
minimization processes Ex tylenol to prevent
febrile seizures vaccinating at different ages
not vaccinating, etc
20

• How do we create the system necessary for the
  optimal scientific study?

Needs System Basic science background Technolog
y Analytic capability Scientists Efficiencies
21

• How do we create the system necessary for the
  optimal scientific study?

System needs Need to have capacity to
ascertain rare events after vaccination On the
order of 1/1000 to 1/10,000 (or even
rarer) Cannot be done with premarketing or even
postmarketing clinical trials Option 1 VAERS
(Vaccine Adverse Events Reporting
System) Passively reported VAE Option 2
Population based setting Active identification
of VAEs possible Advantage full spectrum of
VAE unbiased ascertainment
22

• How do we create the system necessary for the
  optimal scientific study?

• Systems Vaccine Safety Datalink
• Began in 1991 as a collaborative project between
  CDC and four HMOs
• Group Health Cooperative, Seattle, WA
• Northwest Kaiser Permanente, Portland, OR
• Northern California Kaiser Permanente, Oakland
• South California Kaiser Permanente, Los Angeles
• Expanded in 2000 to include four more HMOs
• Harvard Pilgrim Health Care, Boston, MA
• HealthPartners, Minneapolis, MN
• Kaiser Permanente Colorado, Denver, CO
• Marshfield Clinic, Marshfield, WI
• Total over 10 million members

23

• Vaccine Safety Datalink (VSD)

Patient Characteristics (Birth records) (Census)
Health Outcomes (Hospital) (ER) (Clinic)
Vaccination Records
VSD Linked Analysis Database
24

• How do we create the system necessary for the
  optimal scientific study?

Needs System Basic science background Technolo
gy Analytic capability Scientists Other
Efficiencies
25

• How do we create the system necessary for the
  optimal scientific study?

Needs Basic science background Understanding
of pathways involved in potential VAEs Basic
disease pathogenesis Inflammation
pathways Immune response pathways Used to
identify potential candidate genes and candidate
gene pathways For many (if not most) of VAEs,
this is currently unknown Distinct from
medication AE related (for ex) to cyp450
pathway
26

• How do we create the system necessary for the
  optimal scientific study?

Needs System Basic science background Technolog
y Analytic capability Scientists Other
Efficiencies
27

• How do we create the system necessary for the
  optimal scientific study?

Needs Technology Analytic capability Technology
Use of 500K chips for SNP analysis becoming
more routine Could partner with producers of
chips (Affy Illumina etc) for cost,
individualized production etc Specimen
collection typically blood samples (buccal
swabs or other in future offer possibility of
remote/streamlined collection of specimens from
case/family) Data tracking one of major
challenges of Human Genome Project Will need
attention in any future endeavors for vaccine
genomics
28

• How do we create the system necessary for the
  optimal scientific study?

Needs Technology Analytic capability 500K
chips give information on 500,000 single
nucleotide polymorphisms Challenges typical
logistic regression analysis has 10-100
covariates (not 500K) 1. Running chips is a
specialized knowledge/capability 2. Need
mainframe computers for data storage and
analysis 3. Need advanced/new biostatistical
algorithms for fitting models 4. Almost
guaranteed to find more false than true
positives 5. Individual SNPs might not be as
important or illuminating as haplotypes
29

• How do we create the system necessary for the
  optimal scientific study?

Needs Analytic capability 1. Running chips is
a specialized knowledge/skill 2. Need
mainframe computers for data storage and
analysis Need to create this capability (ie
within CDC) or collaborate with academic
partners 3. Need advanced/new biostatistical
algorithms for fitting models Needs specialized
collaborations with biostatistical genetics and
genetic epidemiology
30

• How do we create the system necessary for the
  optimal scientific study?

Needs Analytic capability 4. Almost guaranteed
to find more false than true positives For
candidate genes can use standard approach For
non-candidate genes (a) assess strength and
consistency of association (b) assess biologic
plausibility (if possible) (c) replicate,
replicate, replicate 5. Individual SNPs might
not be as important or illuminating as
haplotypes
31

• How do we create the system necessary for the
  optimal scientific study?

Needs For identification of cases, selection of
controls, and enrollment Knowledge of
vaccine/schedule/adverse events Collaborative
network with organizations/populations of
interest Historically infectious disease
specialists epidemiologists For basic
science/gene pathways Immunologists/infectious
disease specialists Geneticists For
analysis Collaboration with partners with
capabilities to run samples Biostatisticians/gene
tic epidemiologists to analyze data
32

• How do we create the system necessary for the
  optimal scientific study?

How do we create the system necessary for the
optimal scientific study? Needs System Basic
science background Technology Analytic
capability Scientists Other Efficiencies
33

• How do we create the system necessary for the
  optimal scientific study?

Needs Efficiencies Consider moving away from
specific control groups. Option genotype 1000
people from each HMO and use that as a standard
control group for every study Expensive to begin
with, but saves cost savings and more efficient
in the long run
34

• How do we create the system necessary for the
  optimal scientific study?

How do we create the system necessary for the
optimal scientific study? Presently System
exists in integrated fashion (VSD) Basic science
background/scientific expertise needs
concentration/integration Technology/Analytic
capability available needs coordinated
approach Efficiencies needs evaluation
35
Vaccine Safety Case StudyRheumatoid Arthritis
and Hepatitis B Vaccine
36
Rheumatoid Arthritis Background

• Chronic autoimmune disease
• Population prevalence of 1-2 worldwide
• Over 7 million affected in U.S.
• lt50 5 year survival rate among most severely
  affected

37
Why study genetics vaccination with regard to
rheumatoid arthritis?

• RA HLA DR4 associated with increased
  susceptibility to disease
• Hepatitis B infection HLA DR3 associated with
  altered susceptibility
• Reports of RA among HB vaccine recipients HLA-DR
  types that either increase RA susceptibility or
  HBV response

38
Study Question

• Are there specific genes that predispose to
  Rheumatoid arthritis following hepatitis B
  vaccination?

39
Study design options

• Cohort Are rates of RA increased among vaccine
  recipients relative to non-HBV recipients?
• Specifically, are rates of RA particularly
  increased among persons with specific genetic
  polymorphisms (ie of HLA DR4) compared to those
  persons without such polymorphisms?
• Case-control Is HBV receipt over-represented
  among subjects with RA compared to controls?
• Specifically, is combination of HBV and certain
  polymorphisms over-represented among cases
  compared with controls?

40
Flow chart for case identification, sample
collection, and data analysis.All persons ages
15 to 59 with continuous HMO membership from
1/1/95 to 12/31/99Computer Definition of
Possible RA CasesChart Review of Possible RA
CasesRheumatologist Review of Selected Cases
todetermine if chart review is adequate to
satisfy1987 ACR criteria for RA Data
analysis (initial Kaiser retrospective
study)Obtain permission from NCK personal
physicians to contact RA patientsPts. Invited
to Enroll in RA-HBV Genetic StudiesPts.
Consented for Study, Blood DrawnBlood Shipped,
Fed Express, o/n to AtlantaHLA and Hepatitis
Antibody TestingData Analysis(genetic
case-only substudy)
41
In a separate study of RA, Celera Diagnostics
identified replicated 22 SNPs in RA
patient samples includes R620W missense SNP in
PTPN22 VSD study has assessed the frequency
of these gene SNPs among cases in addition to
the HLA DR4 polymorphisms listed previously
Begovich et al. 2004 AJHG 75330
42
(No Transcript)
43
Power calculations for gene-environment study of
RA, hepatitis B vaccination, and
HLA-polymorphismsIfGenetic risk (HLA-DR4) OR
5 (conservative)Environmental risk (HBV) OR
2 (likely over-estimate)AndIf DR4 prevalence
is 15 (NCK population of caucasians, NA and
AA)Will have 80 power, alpha 0.05 to detect
interaction of 10 HB coverage Sample size
needed 2
200 5
100 10
50
44

• Opportunities and Obligations Vaccine Safety in
  the Genomics Era

Screen VSD data-sets yearly Identify
subjects/collect specimens on cases q yr
febrile seizures severe limb swelling q 5 yrs
arthritis prolonged crying q 10 yrs
encephalopathy GBS anaphylaxis w/high profile
situations ie intussusceptionGBS Run
genome-scans (500K chips or higher) on cases
Compare with standard age, HMO, race matched
controls  
45

• Opportunities and Obligations Vaccine Safety in
  the Genomics Era

Vision for the Future Screen VSD data-sets
yearly Identify subjects/collect specimens
on cases q yr febrile seizures severe limb
swelling q 5 yrs arthritis prolonged crying
q 10 yrs encephalopathy GBS
anaphylaxis w/high profile situations ie
intussusceptionGBS Run genome-scans (500K chips
or higher) on cases Compare with standard age,
HMO, race matched controls Assess findings for
_candidate_ genes Generate new set(s) of
potential candidate genes/pathways for next
iteration  
46
Opportunities and Obligations Vaccine Safety in
the Genomics Era Conclusions

• Study of vaccine genomics just beginning to get
  underway
• Evaluations of gene-environment interactions
  (HLA-HBV and RA MMR and FH epilepsy and febrile
  seizures) can be wrapped into large database
  infrastructure (US VSD Scandinavian
  population-based studies)
• Other studies not discussed today (ie HLA control
  of antibody response to specific vaccination) can
  be accomplished within the venue of prelicensure
  clinical trials

47
Opportunities and Obligations Vaccine Safety in
the Genomics Era Challenges on the horizon

• New vaccines
• Rotavirus
• HPV
• Acellular pertussis
• MMR-V, and many more
• Increased focus on adolescents and adults
  (meningococcal varicella etc)
• Different diseases/potential adverse events (ie
  autoimmune)
• Increasingly packed schedule
• Relatively unknown safety profile

48
Opportunities and Obligations Vaccine Safety in
the Genomics Era Vision into the Future

• CDC has a critical role for integrating genomics
  into vaccine safety
• Infrastructure (collaborations)
• Only CDC with VSD and VAERS able to identify
  subjects with rare AEs
• Scientists with the expertise in understanding
  adverse events
• Forge collaborations with genomics community
• Begin to understand how genetic variation
  underlies VAE
• Understand how to identify people at increased
  risk, and devise alternate immunization
  strategies

49
(No Transcript)
50
Is there evidence from the literature that
interactions between vaccination and subgroups
exist?MMR Vaccination and Febrile Seizures.
Evaluation of Susceptible Subgroups and Long-term
Prognosis JAMA Vol. 292 No. 3, July 21, 2004
Vestergaard, Hviid, Madsen et al
51

• It is known that the risk for febrile seizures is
  increased after MMR vaccination.
• Is this increase even higher among particular
  people
• personal or family history of seizures
• perinatal factors
• socioeconomic status.
• The latter 2 are environment, but the first
  (family history) gives some glimpse of genetic
  risk factors that might interact with environment
  (MMR vaccination)

52
Setting

• Population-based cohort study of all children
  born in Denmark 1/1/91 12/31/98 and who
  survived to 3 months of age total of 537,171
  children
• Unique personal ID allowed link to mother and
  father
• (similar database construction underway within
  VSD)

53
Vaccination

• National Board of Health maintains registry of
  MMR vaccinations administered in Denmark
• Collected info on vaccinations given 1/1/91
  12/31/99
• Jeryl-Lynn strain used/studied
• (recommended) Age of vaccine 15 months
• Exact date of vaccine not available imputed day
  of week as Wednesday.

54
Outcome ascertainment

• Incidence of first febrile seizure, recurrent
  febrile seizures, and subsequent epilepsy,
  gathered from National Hospital Registry which
  includes inpatient, outpatient and emergency
  department visits.
• For febrile seizures, excluded children with
  history of non-febrile seizures, cerebral palsy,
  head trauma, intracranial tumors, meningitis and
  encephalitis (this clarifies the phenotype of
  febrile seizure).
• Used ICD-10 coding schema

55
Evaluation of high risk strata (effect
modifiers)

• Personal history of febrile seizures
• Family history
• Birth characteristics and sociodemographic factors

56
Statistical Analysis

• Poisson regression. Entered at 3 mo of age.
  Censored at first diagnosis of FS, epilepsy,
  death, emigration, enceph/mening/head injury, or
  age 5 years or 12/31/99
• MMR vaccination analyzed as time varying
  covariate

57
Subgroup analysis tested for statistical
interaction

• Question Is the RR for children with family
  history (FH) of epilepsy higher than RR for
  children without family history of epilepsy
• Risk after MMR vaccination among children with
  FH
• Risk not after MMR vacc among children with FH
• Vs
• Risk after MMR vaccination among children
  without FH
• Risk not after MMR vacc among children without
  FH

58
Results

• 439,251 children (82) received MMR vaccination
• 17,986 children developed febrile seizures at
  least once
• RR for FS in the 2 wks after MMR vaccination
  2.75 (95 CI 2.55-2.97)
• RR did not vary significantly in the subgroups of
  children
• family history of seizures
• perinatal factors
• socioeconomic status

59
Subgroup analysis, continued

• RR was higher among the subgroup of children with
• family history of epilepsy
• RR of 4 following vaccination among subjects
  with family history
• RR of 2.7 after vaccination among subjects
• with family history

60
However, story is different when evaluated using
risk difference metric

• Among children with family history of epilepsy,
  vaccination adds 3.4 additional seizures per 1000
  persons vaccinated
• (Due to higher RR (4) acting on a lower
  baseline risk among non-vaccinated children
  (compared to children with personal history of
  FS)

61
Significant findings

• When viewed on a multiplicative scale, there is
  only interaction noted among children with a
  family history of epilepsy (RR of 4 following
  vaccination among those with family history of
  epilepsy vs RR of 2.7 among those without family
  history of epilepsy)
• However, viewed on an additive scale, the message
  is much different
• 1.6 additional seizures per 1000 children overall
• 19.5 additional seizures per 1000 among children
  with personal history
• 3.4 additional seizures per 1000 among children
  with FH epilepsy
• Additive scale is influenced by baseline risk,
  and give a more clinically relevant picture of
  the risk to the particular person

62
Example of rapid response to emerging public
health question

• New conjugate meningococcal vaccination licensed
  in January 2005
• Recommended for universal use among adolescents
  in February 2005
• Distribution commenced in March 2005. 
• By December, 2005, 7 reports of Guillain-Barre
  Syndrome within six weeks following Menactra
  administration had been received by the VAERS
  passive reporting system. 
• All cases were among 17-19 year olds, and
  occurred 11-31 days following vaccination. 
• Quickly mobilized VSD and studied 110,000
  vaccinees.
• Comparison with background rates indicated that
  the number of reported cases was expected
• MMWR publication notified public and medical
  community
• But did not result in additional reports
• Result
• Vaccine remained on market

63
Vision into the Future

• Three major goals
• 1) Establish broad input into the research agenda
  for Immunization Safety Office
• External advisory panel comprised of wide range
  of stakeholders including federal agencies, major
  medical organizations, FDA, CMS, NIH, DoD,
  parental groups to help shape vaccine safety
  research plan
• 2) Active surveillance of new vaccines
• Weekly updates from VSD and potentially other
  MCOs
• Real-time assessments of vaccine safety

64
Vision into the Future

• 3) Bring vaccine safety research into the
  genomics era
• Personalized medicine personalized drug
  delivery under intense study by
  NIH/NHGRI/pharmaceutical industry
• Understand which people/subgroups are at
  increased risk for
• Vaccine adverse events
• Vaccine failure

65
The Road to a Public Health Approach to
Pharmacogenomics will be a long and raucous
journey
66
Immunization Safety Office

• Vaccine Safety Datalink (VSD) project
• Collaborative project with comprehensive medical
  and immunization histories of over 7.5 million
  people.
• Studies health problems among vaccinated people
  compared with unvaccinated people.
• Currently using in-depth evaluations to study the
  safety of thimerosal in vaccines, and risk for
  autism following vaccination.
• Vaccine Adverse Event Reporting System (VAERS)
• An early-warning passive surveillance system to
  detect problems related to vaccines.
• Clinical Immunization Safety Assessment (CISA)
  Network
• Provides in-depth, standardized clinical
  evaluations for individuals with unusual or
  severe vaccine adverse events to understand
  virologic, immunologic and genetic markers for
  post-vaccination adverse events.
• Brighton Collaboration
• A global collaboration to standardize case
  definitions and the study of vaccine reactions,
  providing a common vocabulary for vaccine
  safety research,
• Vaccine Acceptance and Risk Perception (VARP)
• Scientific study of interventions that increase
  vaccine acceptance
• Vaccine Technology
• Development of safer vaccines and delivery
  (needle-free jet injectors)

67

• Importance of Vaccine Safety

• Higher standard of safety expected of vaccines
• "First do no harm" (primum non nocere)
• Moral duty public health clinical medicine
• Vaccinees generally healthy (vs. ill for drugs)
• Vaccinations universally recommended or mandated
• Lower risk tolerance search for rare reactions
• vaccine lt 1/100,000 vs. drug 1/1 - 1/1000
• Studies of rare events
• More costly and difficult
• Less likely to be definitive
• Has, up until now, largely ignored issues of
  subgroup susceptibility

68

• Lines of Evidence Suggesting Causality

• Temporal association
• illness follows exposure
• cases cluster within definable time after
  vaccination
• Biologic plausibility
• animal models
• tissue culture or other models
• Specificity
• unique clinical picture
• unique laboratory result
• Epidemiologic evidence
• risk of illness gt expected by chance

69

• IOM Vaccine Safety Report, 1991-94 Conclusions

Pert Rub DT/Td/T Meas Mump
Polio Hep B Hib Categ 1 Autism

Neuropathy Transverse No

Residual myelitis
(IPV) Evidence
seizure
Thrombocytopenia

Anaphylaxis
Categ 3 Infantile spasms
Encephalopathy
Early onset Hib Favors
Hypsarrythmia Infantile spasm (DT)

(conjugate) Rejection Reye syndrome
SIDS (DT)

SIDS
Categ 4 Acute Chronic GBS
Anaphylaxis GBS (OPV)
Early onset
Favors encepha- arthritis

Hib 18m Accept
lopathy

(unconjugated
Shock

PRP)
Categ 5 Anaphylaxis Acute Anaphylaxis
Thrombocytopenia Polio (OPV) Anaphylaxis
Establish Protracted arthritis
Anaphylaxis (MMR) Death (OPV) Causal
Crying
Death
70

• Traditional Vaccine Safety Studies Pre-Licensure

• Laboratory
• Animals
• Humans
• Phases I gross toxicity (N 10)
• Phase II dosing range/ reactogenicity (N
  10-100)
• Phase III efficacy ( preliminary safety) (N
  1000-10,000)
• Advantages
• Close, detailed follow-up
• Randomized, placebo-controls gt causality
  assessment easy
• Disadvantage
• Poorly detected reactions rare, delayed onset,
  subpopulations
• No standard case definition for "safety"

71

• Traditional Vaccine Safety Studies Post-Licensure

• Traditional tools
• passive surveillance (spontaneous reporting
  system)
• ad hoc controlled epidemiologic studies
• New tools
• Phase IV trials "linked" to licensure of new
  vaccine
• Large-Linked Database (LLDB) in HMO population
• N 10,000
• pre-organized LLDB's
• ongoing safety monitoring
• controlled epidemiologic studies
• "enhanced" passive surveillance as "registry"

72
Traditional Passive Surveillance (e.g., VAERS)

• Strengths
• National in scope
• Timeliness
• Relative low cost
• Weaknesses
• Under-, biased reporting
• Complexity
• multiple "exposures" "outcomes"
• detect "new" change "known" AE's
• mix of causal and coincidental events
• Generally unable to assess causality

73

• Institute of Medicine (IOM) Reports on Vaccine
  Safety

• "Many gaps and limitations" in current knowledge
  research capacity
• Infrastructure for vaccine safety surveillance
  inadequate
• Needed Population laboratory under active
  surveillance

74
Vaccine Safety Datalink

• Began in 1991 as a collaborative project between
  CDC and four HMOs
• Group Health Cooperative, Seattle, WA
• Northwest Kaiser Permanente, Portland, OR
• Northern California Kaiser Permanente, Oakland
• South California Kaiser Permanente, Los Angeles
• HealthPartners, Minneapolis
• Harvard Pilgrim Health Plan, Boston
• Kaiser Colorado, Denver
• Marshfield Clinic, Wisconsin
• Total over 6 million members

75
Advantages of HMOs for Health Research

• Identifiable (large) population
• incidence rates and attributable risks
• Computerized data bases
• Cost data
• Integrated systems
• Infrastructure

76
VSD Analytic Approach

• Screening analyses (automated data)
• preliminary assessment of vaccine-outcome
  associations
• In-depth studies (chart reviews, interviews)
• validate outcomes (and dates)
• verify vaccination history (and dates)
• additional risk factor or clinical information

77
Selected Findings from VSD Studies

• No increased risk of chronic arthropathy among
  women receiving rubella vaccine
• No increased risk of aseptic meningitis after
  Jeryl-Lynn mumps vaccine (in U.S. MMR)
• Risk of clinical events after 2nd MMR greater at
  10-12 than at 4-6 years of age

78
Selected Current VSD Studies

• Neonatal and infant mortality
• Wheezing and asthma
• Timing of vaccination and type 1 diabetes
• MMR vaccine and IBD
• Hepatitis vaccination and risk of MS
• Possible sequelae of thimerosal in vaccines
• Rotavirus vaccine and intussusception

79
Diversity of Research Projects

• VSD database and infrastructure allow a wide
  range of studies in addition to vaccine safety
• vaccination coverage
• cost studies of different policies or strategies
• clinical trials
• descriptive epidemiology

80
Research in MCOs Caveats

• Identifiable (large) population
• enrollment and disenrollment
• Computerized data bases
• not developed for research
• Dynamic
• Cost data
• may not be generalizable
• Integrated systems
• not true of many health plans
• Infrastructure
• not in place in all health plans
• research infrastructure often not present

81
Research in Managed Care Organizations
Conclusions

• Managed care is the dominant health care delivery
  system in the U.S.
• MCOs provide great opportunities for
  population-based health research
• Immunization research in MCOs allows timely and
  efficient monitoring of vaccine safety

82
The Public Health Approach to Pharmacogenomics

• Goal Personalized delivery of therapeutics that
  accounts for the genetic variation of the patient

83
The Public Health Approach to Pharmacogenomics

• Goal Personalized delivery of therapeutics that
  accounts for the genetic variation of the patient
• Thesis
• Gene-based diagnostic tests are very powerful
• Have distinctive risk/benefit profiles
• May have unintended effects
• Therefore, the default for gene-based diagnostic
  tests and for pharmacogenetics should be
• RCTS
• Good observational studies
• A requirement linked to licensure

84
The Public Health Approach to determine the real
world effectiveness of pharmacogenomics and
monitor its applications

• Proper Public Health Use of Genetic Tests and
  Pharmacogenomics
• How do we get from here to here?
• Identification of Appropriate use of
• gene-disease genetic testing
• association
• Not all outcomes research means the same thing
• Evidence Integrating Surveillance
• Of Evidence
• Effectiveness

85
The Public Health Approach to determine the real
world effectiveness of pharmacogenomics and
monitor its applications

• Evidence Integrating Surveillance
• Of Evidence
• Effectiveness
• What is evidence?
• The basic science approach
• Variations in cyp450 polymorphisms are common
• Medications for cardiovascular disease and
  depression are among the most comonly used, and
  are a significant cost driver
• Use of these medications may produce adverse
  effects and/or difficulties in obtaining proper
  therapeutic index.
• Polymorphisms of the cyp450 pathway plays a role
  in responsivenes
• The basic science approach addresses the evidence
  about how cyp450 and medications interact to
  affect responsiveness

86
The Public Health Approach to determine the real
world effectiveness of pharmacogenomics and
monitor its applications

• Evidence Integrating Surveillance
• Of Evidence
• Effectiveness
• What is evidence?
• The public health approach
• Polymorphisms of the cyp450 pathway plays a role
  in responsiveness
• Do these polymorphisms affect measurable clinical
  outcomes?
• Increased morbidity/mortality?
• Increased costs of health care?
• Decreased quality of life?
• The public health approach asks
• Given that cyp450 pathway and medications work
  together to affect responsiveness, does it
  matter?
• Is there a better way to deliver rx for some
  people?

87
The Public Health Approach to determine the real
world effectiveness of pharmacogenomics and
monitor its applications

• Evidence Integrating Surveillance
• Of Evidence
• Effectiveness
• The public health approach
• In addition to wanting to know what happens at
  the overall population level, the PH approach
  also wants to know what happens in these
  subpopulations
• With drug interactions i.e. CV and other
  medications
• Elderly i.e. diminished cardiac function
• Pediatrics i.e. different disease?
• Different ethnic groups i.e. gene-gene-drug
  interactions

88
The Public Health Approach to determine the real
world effectiveness of pharmacogenomics and
monitor its applications

• Evidence Integrating Surveillance
• Of Evidence
• Effectiveness
• The public health approach
• How would we go about collecting information on
  measurable clinical outcomes (morbidity/mortality)
  in a diverse population (elderly, children,
  different ethnicities)?
• Observational studies
• Randomized clinical trials (RCTs)
• Large practical trials (PCTs)

89
Public Health Approach to the real world
effectiveness of pharmacogenomics

• Evidence Integrating Surveillance
• Of Evidence
• Effectiveness
• Observational (cohort or case-control) studies
• Cyp450
• Depression
• (Rate of) good outcome (Rate of) bad outcome
• () anti-dep a b
• () anti-dep - c d
• Cyp450 --
• Depression
• (Rate of) good outcome (Rate of) bad outcome
• () anti-dep a b
• () anti-dep - c d

90
Public Health Approach to the real world
effectiveness of pharmacogenomics

• Observational (cohort or case-control) studies
• Cyp450
• Depression
• (Rate of) good outcome (Rate of) bad outcome
• () anti-dep a b
• () anti-dep - c d
• Cyp450 --
• Depression
• (Rate of) good outcome (Rate of) bad outcome
• () anti-dep a b
• () anti-dep - c d
• Advantage Data is easily available
  (relatively)
• Comparison by gene group is relatively
  unbiased
• Disadvantage Sample size limitations when
  stratifying additionally by elderly, by
  children, by other medications, by ethnic groups,
  etc.

91
Public Health Approach to the real world
effectiveness of pharmacogenomics

• Evidence Integrating Surveillance
• Of Evidence
• Effectiveness
• Randomized Clinical Trials allow you to enroll
  based on gene status
• cyp450 Asthma
• (Rate of) good outcome (Rate of) bad outcome
• () anti-depl a b
• () anti-dep - c d
• cyp450 -- Asthma
• (Rate of) good outcome (Rate of) bad outcome
• () anti-depl a b
• () anti-dep - c d

92
Public Health Approach to the real world
effectiveness of pharmacogenomics

• Evidence Integrating Surveillance
• Of Evidence
• Effectiveness
• Randomized Clinical Trials allow enrollment based
  on gene status
• Problems with generalizability and sample size
  requirements has led to concept of Large
  Practical Clinical Trials
• Objective To enroll many (gt100,000) people in
  trial randomized at patient (or clinic/provider)
  level
• Will allow for head to head comparisons of
  commonly used medications
• For pharmacogenomics, can study not only
• does statin A work better than statin B, but
  also
• are there haplotypic groups whereby statin A
  works best for haplotypic group A, while statin
  B works best for haplotypic group B?

93
Public Health Approach to the real world
effectiveness of pharmacogenomics

• Evidence Integrating Surveillance
• Of Evidence
• Effectiveness
• Large Practical Clinical Trials
• Head to head comparisons of commonly used
  medications
• Can study not only does statin A work better
  than statin B, but also
• are there haplotypic groups whereby statin A
  works best for haplotypic group A, while statin
  B works best for haplotypic group B?
• Utilizing the natural experiments among large
  numbers
• Can also study these genetic differences in drug
  effectiveness among risk groups (elderly,
  pediatrics, etc)
• Can look at interactions with other genes, other
  medications
• Advantage studies looks at drug, gene and system
  effects
• Diadvantage very expensive to do properly, even
  with observational data

94
Public Health Approach to the real world
effectiveness of pharmacogenomics

• Evidence Integrating Surveillance
• Of Evidence
• Effectiveness
• RCTs or quasi-experimental designs
• What type of system is necessary to get evidence
  integrated into practice?
• NEEDS
• Network of Researchers Organizations IRBs Data
• Clinical researchers MCOs
• Health care researchers BCBS/United Standards EMR
  development
• Biostatisticians Medicare/aid
• VA

95
Public Health Approach to the real world
effectiveness of pharmacogenomics

• Evidence Integrating Surveillance
• Of Evidence
• Effectiveness
• Safety
• Vaccine model
• VAERS reporting
• VSD (population denominator based
  collaborative project)
• Future registry
• buccal swabs for DNA
• candidate gene generation
• Pharmaceutical model
• AERS reporting
• CERT and other population based collaborative
  projects
• Future? registry
• buccal swabs for DNA

96
Vision into the Future Vaccine safety research
in the Era of Genomics

• CDC has a critical role for integrating genomics
  into vaccine safety
• Infrastructure (collaborations)
• Only CDC with VSD and VAERS able to identify
  subjects with rare AEs
• Scientists with the expertise in understanding
  adverse events
• Forge collaborations with genomics community
• Begin to understand how genetic variation
  underlies VAE
• Understand how to identify people at increased
  risk, and devise alternate immunization
  strategies

97
How do we create the system necessary for the
optimal scientific study? Needs System Basic
science background Technology Analytic
capability Scientists Efficiencies