dengue vaccination impact: beyond efficacy · dengue vaccination impact: beyond efficacy isabelle...
TRANSCRIPT
DENGUE VACCINATION IMPACT: BEYOND EFFICACY
Isabelle Delrieu, PhD & Bradford D. Gessner, MDAgence de Médecine [email protected] & [email protected]
Vaccinology 2016X International Symposium for Latin American ExpertsSeptember 20-22, Brasilia, Brazil
BURDEN CONCEPTSVE AND PH BURDEN MEASURES
DENGUE VACCINEBEYOND BURDEN
Vaccine Launches 2013
3
→ What drives new vaccines
introduction by governments?
4
“Disease burden has been consistently mentioned by
policymakers in countries to be the number one factor in
setting priorities for vaccines to be introduced into
immunization programs; the higher the burden, the more
attractive a potential addition to the immunization regime
of the country would be.”
Burden measure limitations
• Poor diagnostics: non-bacteremic Hib/Sp, typhoid
• Causal etiology gone at time of presentation: flu/viral ARI pathogens precipitating bacterial ARI
• Pathogen present but not causal: flu
• Lack of testing, poor specimen transport systems: all etiologies
• Limited health care access: all etiologies
5
BURDEN CONCEPTS
VACCINE EFFICACY AND PUBLIC HEALTH BURDEN MEASURES
DENGUE VACCINEBEYOND BURDEN
Definition of measures
• Vaccine effectiveness/efficacy (VE) = 1 – (Incidence[vaccinated] ÷ Incidence[unvaccinated])
• Vaccine preventable disease incidence (VPDI)= Incidence[unvaccinated] – Incidence[vaccinated]= Incidence[unvaccinated] X VE
• Number needed to vaccinate (NNV)= 100,000/VPDI/length of follow-up for VPDI
• Number prevented (nationally) (estimated!)= VPDI * (birth cohort/100,000) * years of follow-up for VPDI
7
Feikin, Scott, Gessner. Use of vaccines as probes to define disease burden. Lancet
2014;383:1762-70
VPDI, NNV and Cases prevented; VPDI per 100,000 CYO
Syndrome Etiology confirmed Clinical outcome
VE VPDI NNV Cases prev.
VE VPDI NNV Cases prev.
Gambia PCV radiological pneumonia
70% 140 357 216 37% 1300 38 2002
Indonesia, Hib, hospitalized meningitis
86% 16 3125 1516 22% 160 313 15,155
Kenya rotavirus, acute gastroenteritis
84% 3300 15 101,244 34% 19,000 3 582,920
8
Lancet 2005;365:1139-46; Lancet 2005;365:43-52; Vaccine 2012;30 (suppl 1):A52-60
Measures useful outside of developing country settings: acute gastroenteritis (AGE)
9
Study VE VPDI NNV Cases prev.
Finland (Vaccine 2012;31:176-82)
Confirmed inpatient AGE 80% 390 256 237
All cause inpatient AGE 54% 1070 93 651
Kenya (Vaccine 2012;30 Supp 1:A52-60)
Confirmed severe 84% 3300 15 101,244
Community severe AGE 34% 19,000 3 582,920
Public health impact can be greater in settings where vaccine efficacy is lower: acute gastroenteritis (AGE)
10
Study VE VPDI NNV Cases prev.
Severe rotavirus AGE (NEJM 2010;362:289-98)
South Africa 77% 4200 24 46,284
Malawi 49% 6700 15 42,813
Severe rotavirus AGE (Lancet2010;376:615-23)
Vietnam 64% 2200 26 55,425
Bangladesh 43% 3500 16 192,950
Etiology conf.
X-ray lobar
consolidation
Severe, not
X-ray confirmed
Mild
Unvaccinated VaccinatedCategory of
pneumonia
Impact of vaccine against categories of pneumonia (Lancet 2014;383:1762-70)
Target
etiology
Other
etiologies
Other
etiologies
Target
etiology
RTS,S VPDI against malaria-specific and all-cause hospitalization
(Lancet 2015;386:31-45).
12
-40
-20
0
20
40
60
80
100
120
Malaria All-cause Malaria All-cause Malaria All-cause Malaria All-cause
No booster Booster No booster Booster
5 to 17 month old 6 to 12 weeks
VP
DI p
er
10
00
vac
cin
ee
s
BURDEN CONCEPTS
VACCINE EFFICACY AND PUBLIC HEALTH BURDEN MEASURES
DENGUE VACCINE BEYOND BURDEN
14
*All data based on 2 year VE estimates
NEJM 2015;373:1195-206 and NEJM 2015;372:113-23.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
All serotypes Serotype 1 Serotype 2 Serotype 3 Serotype 4 All serotypes
Confirmed dengue Hospitalizeddengue
Vac
cin
e e
ffic
acy
Vaccine efficacy against confirmed dengue in children 9-16 years of age in Latin America*
15
0
200
400
600
800
1000
1200
1400
1600
1800
2000
All serotypes Serotype 1 Serotype 2 Serotype 3 Serotype 4 All serotypes
Confirmed dengue Hospitalized dengue
VP
DI p
er 1
00
,00
0 p
er y
ear
Vaccine preventable disease incidence for confirmed dengue in Latin America*
*All data based on 2 year VE estimates
NEJM 2015;373:1195-206 and NEJM 2015;372:113-23.
16
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%C
on
firm
ed
ro
tavi
rus
ho
spit
aliz
atio
n
Gas
tro
ente
riti
s h
osp
ital
izat
ion
fro
m a
ny
cau
se
Vac
cin
e se
roty
pe
inva
sive
pn
eum
oco
ccal
dis
ease
Co
nso
lidat
ed C
AP
Rad
iolo
gica
lly c
on
firm
ed
CA
P
Susp
ecte
d C
AP
Inva
sive
Hib
dis
ease
Rad
iolo
gica
lly c
on
firm
ed
pn
eum
on
ia
Co
nfi
rme
d h
osp
ital
ized
sev
ere
Co
nfi
rme
d h
osp
ital
ized
den
gue
All
con
firm
ed
den
gue
Rotavirus LatinAmerica. NEJM2006;354:11-22
Pneumococcus Latin America. PLoS Med2014;11:e10011657
Hib Chile. Ped InfectDis J 1999;18:1060-4
Dengue. NEJM 2015
Vac
cin
e ef
fiac
y
Vaccine efficacy for dengue compared to other vaccines studied and used in Latin America
→ Comparable VE for etiologically confirmed disease
17
0
200
400
600
800
1000
1200
1400
1600
1800
2000
Co
nfi
rme
d r
ota
viru
s h
osp
ital
izat
ion
Gas
tro
ente
riti
s h
osp
ital
izat
ion
fro
man
y ca
use
Vac
cin
e se
roty
pe
inva
sive
pn
eum
oco
ccal
dis
ease
Co
nso
lidat
ed C
AP
Rad
iolo
gica
lly c
on
firm
ed
CA
P
Susp
ecte
d C
AP
Inva
sive
Hib
dis
ease
Rad
iolo
gica
lly c
on
firm
ed
pn
eum
on
ia
Co
nfi
rmed
ho
spit
aliz
ed s
ever
e
Co
nfi
rme
d h
osp
ital
ized
den
gue
All
con
firm
ed d
engu
e
Rotavirus LatinAmerica. NEJM2006;354:11-22
Pneumococcus Latin America. PLoS Med2014;11:e10011657
Hib Chile. Ped InfectDis J 1999;18:1060-4
Dengue. NEJM 2015
VP
DI p
er 1
00
K p
er y
ear
Vaccine preventable disease incidence (VPDI) for dengue compared to other vaccines studied and used in
Latin America
→ Dengue vaccine against syndromic disease?
18
0
100
200
300
400
500
600
700
Confirmed hospitalizedsevere
Confirmed hospitalizeddengue
All confirmed dengue
NN
VNumber needed to vaccinate (NNV) to prevent cases of dengue
659
236
27
BURDEN CONCEPTS
VACCINE EFFICACY AND PUBLIC HEALTH BURDEN MEASURES
DENGUE VACCINE
BEYOND BURDEN
Age distribution
Sp Hib Rotavirus Malaria Dengue
<5 year disease ++++ ++++ ++++ ++++ ++/+++
<5 year severity/sequelae
++++ -- -- ++++ +
5+ year disease ++/+++ + -- ++++ ++++
5+ year severity/sequelae
+++ + -- + +
20
Sequelae/mortality
Sp/Hibmeningitis
Sp/Hibpneumonia
Rotavirus Malaria Dengue
Cognitive (MR, dev delay, learning disability, language)
++++ -- -- +++ --
Sensory (hearing, vision)
++++ -- -- -- --
Physical (CP, seizures) ++++ -- -- +++ +
Stunting ? ? + +++ --
Case fatality ratio ++++ +++ + ++ +
21
E.g., in US, big cost driver for Hib was long-term care and
institutionalization for meningitis sequelae
Indirect/replacement/rebound effects
Sp Hib Rotavirus Malaria Dengue
Indirect +++ ++++ ++ -- ?
Replacement +++ + (so far)
-- -- ?
Rebound -/+ (without booster)
+ (without boosterin some settings)
-- +++ (depends on transmission)
?
22
Work in different directions:
• Indirect effects can greatly increase immunization efficiency and public
health value
• Replacement can completely negate immunization efficiency
• Rebound shifts disease to older age; generally beneficial
Immunization program issues
23
Sp Hib Rotavirus Malaria Dengue
Fits with current childhood schedule
+++ +++ +++ -- +
Duration of immunity (with booster)
+++ +++ Lessrelevant
-/+ ??
Variable geographic distribution within affected countries
-- -- -- ++ +++
Health system impact
Sp Hib Rotavirus Malaria Dengue
Outbreak potential + -- -- + +++
May overwhelm clinical resources
+ -- ++ ++ +++
Requires other intensive + expensive interventions
+ + ++ +++ +++
Increasing incidence in absence of vaccine
-- -- -- -- +++
Political dimension + + + +++ +++
24
Equity
Sp Hib Rotavirus Malaria Dengue
Differences in infection by population
+ + -- ++ ++
Differences in severe disease by population
+ + ++ ++ +
Differences in mortality by population
++ ++ +++ +++ +++
25
SUMMARY• Safety and efficacy just the start of assessing public health value of vaccine • Burden is the foundation of decision making
– Incidence– VPDI– NNV– Cases prevented– Sequelae– Mortality
• Other key issues– Vaccine characteristics– Programmatic concerns– Health system impact– Equity
• All of these features contribute to models estimating public health value of vaccine
• Based on clinical trial data, dengue vaccine should have similar public health impact as other vaccines currently used in Latin American public health programs.
26
28
Vaccine probe studies:
3 distinct insights into the epidemiology of vaccine-preventable diseases:
1. can estimate absolute burden of disease incidence that is preventable by a
vaccine. The effect of a vaccine can be measured against different disease
manifestations as well as health-system endpoints such as health-care visits or
drug use.
2. can measure the contribution of a specific pathogen targeted by vaccination to a
broad clinical syndrome.
3. Vaccines can be used to investigate the causal chain in disease pathogenesis.
For example, if an infectious disease is hypothesised to cause a specific form of
cancer, a trial of an effective vaccine against the infection could test the
hypothesis
epidemiological methods of vaccine probe studies do not differ from those of
vaccine efficacy or effectiveness studies: RCT and non-randomised designs which
measure disease incidence before and after vaccine introduction. The probe approach
can be incorporated into the design of a vaccine efficacy study (eg, the Indonesian
Hib vaccine study) or it can be applied retrospectively or prospectively to studies
designed primarily to measure vaccine efficacy
vaccine probe studies may add evidence to potential H impact of a vaccine with low
VE by demonstrating impact against non specific syndromes (fever) or outcomes
(hosp, AB use, visit)