dengue vaccines: herd immunity considerations · models of dengue transmission herd immunity...

IntroductionModels of Dengue Transmission

Herd Immunity ConsiderationsConclusions

Dengue Vaccines: Herd ImmunityConsiderations

Derek Cummings

Johns Hopkins Bloomberg School of Public Health

Fondation Merieux, October 27, 2010

1/33



Outline

1 Introduction

2 Models of Dengue Transmission

3 Herd Immunity ConsiderationsImmune enhancement of transmissionVaccine performanceAge effects of VaccinationBasic Reproductive Number

4 Conclusions

2/33



Dengue Virus

Mosquito-borne flavivirusFour antigenically distinctserotypesEstimated 50-100 millioninfections annuallyPresentation ranges fromasymptomatic to severe form(DHF)2 ◦ cases are 50 times aslikely to have DHF

3/33



Immune Enhancement

Apoptosis ofnaive T cells

Interferon γ

Antibody-dependentenhancement

Expansion oflow a�nitycross-reactiveT cells

Secretion ofvasoactivecytokines:TNFαInterleukin 2Interferon γ

Non-neutralising antibody

Fcγ receptor

Dengue virus

Platelet

Plasma leakageDengue virus and antibody complexing withplatelets and activating clotting cascade

Endothelium

Antigenpresentation

HLA-peptide

Monocyte

Memory T-cell

T-cellreceptor

Farrar, NEJM, 2009

4/33



Challenges of Vaccine Development

Cannot contribute to immunopathogenesisNeeds to generate neutralizing immunity to all fourserotypes

avoid virus-to-virus interference or flavivirus interference

Needs to create long-lasting immunity

5/33



Status of vaccine development

On hold

The candidate produced bySanofi Pasteur is soon tobegin phase III trialsThis candidate is yearsahead of others indevelopment

6/33



Questions

Can vaccinations lead to increased secondary infections?What are the characteristics of these vaccines?What is the impact of non-immunized individuals?

What will be the effect of the age shift on symptomatic andsevere cases?What fraction of the population will we have to vaccinate?

7/33



Dengue Transmission Model Structure

S E I

Vectorpopula/on

Age specific model with cross immunity, immune enhancementand SEI model of vectors

8/33



Vaccination

%''"%'Zij

!" #$" %$" #$&" %''"!" #$" %Z #$&" %'X Yi Ci Zi Yij

Vaccine provides all or nothing immunity to all or a subset ofserotypes with some probability

9/33



Immune enhancement of transmissionVaccine performanceAge effects of VaccinationBasic Reproductive Number

Results of hypothetical vaccination 1

R0 = 4, φ = [1.5,1.5,1.5,1.5], v = [0.5,0.5,0.5,0.5]

10/33





R0 = 4, φ = [1.5,1.5,1.5,1], v = [0,0,0,0.98]

11/33





R0 = 4, φ = [1.5,1.5,1.5,1.5], v = [0,0,0,0.6]

12/33




Model parameterization

0 5 10 15 20 25 30

1980 1985 1990 1995 2000 2005

0.00.2

0.40.6

0.8

Year

Inc. p

er 10

00

0.00.2

0.40.6

0.8

Year

Inc. p

er 10

00

0.05

0.0

0.1

Age

Model was parameterized to match dominant multiannualperiodicity of all serotype incidence data as well as proportionof incidence occurring at each age.

13/33




Hypothetical vaccines

Scenario 1: Single dose delivered at 1 year of age with 0.9independent probability of generating immunity to eachserotype reaching 90% of targeted children

Scenario 2: Three doses delivered at 12 months, 18 monthsand 24 months with 0.9 independent probability of gainingimmunity to each serotype at third dose reaching 90% oftargeted children

Scenario 3: Three doses delivered at 12 months, 18 monthsand 24 months with 0.1, 0.2 and 0.9 probability of gainingimmunity to each serotype at 1st, 2nd and 3rd doserespectively reaching 90% of targeted children

14/33




Reduction in incidence

Over 5 years Over 10 yearsScenario 1 29.0% 87.1%Scenario 2 18.1% 79.5%Scenario 3 22.9% 84.4%

15/33




Age as a risk factor for severe disease

2 4 6 8 10 12 14

0.4

0.6

0.8

1.0

1.2

1.4

Age (Years)

Odd

s of

DH

F

Using hospital based datafrom Thailand, we havefound a increasing odds ofDHF compared to DF uponsecondary exposure withage (Fried, PloS NTDS,2010)Vaccine will shift average toolder individuals, thus theirseverity will impact overallimpact of vaccine

16/33




Age changes in Thailand

DHFIncidenceinThailand

Inc.per100

0

Average age of medicallyattended dengue illness inThailand has doubled in thelast 15 yearsDuring this period, theincidence of DHF hasremained fairly constant

17/33




Age changes associated with Demographic Changes

The Impact of the Demographic Transition on Dengue inThailand: Insights from a Statistical Analysis andMathematical ModelingDerek A. T. Cummings 1*, Sopon Iamsirithaworn 2, Justin T. Lessler 1, Aidan McDermott 3, RungnapaPrasanthong 2, Ananda Nisalak 4, Richard G. Jarman 4, Donald S. Burke 5, Robert V. Gibbons 4

1980 1990 2000

0.06

0.10

0.14

Year

λλ i

a

Mean λλi,1985−2005

Freq

uenc

y

0.05 0.07 0.09 0.11

010

b●

% change in λλi,1985−2005

Freq

uenc

y

−70 −50 −30 −10 0

015

●

c

Recent increases in agespecific incidence linked todemographic changesAging of population hascontributed to a halving ofthe force of infection and adoubling of the mean age ofDHF cases

18/33




Changes in demography

19/33




Results of age specific simulations with changes indemography

Inc.per1000Second.Infec/ons

Year

AgeClass

20/33




Age changes associated with Demographic Changes

The Impact of the Demographic Transition on Dengue inThailand: Insights from a Statistical Analysis andMathematical ModelingDerek A. T. Cummings 1*, Sopon Iamsirithaworn 2, Justin T. Lessler 1, Aidan McDermott 3, RungnapaPrasanthong 2, Ananda Nisalak 4, Richard G. Jarman 4, Donald S. Burke 5, Robert V. Gibbons 4

1980 1990 2000

0.06

0.10

0.14

Year

λλ i

a

Mean λλi,1985−2005

Freq

uenc

y

0.05 0.07 0.09 0.11

010

b●

% change in λλi,1985−2005

Freq

uenc

y

−70 −50 −30 −10 0

015

●

c

Recent increases in agespecific incidence linked todemographic changesAging of population hascontributed to a halving ofthe force of infection and adoubling of the mean age ofDHF cases

21/33




Role of Immune Individuals in Population

Immune individuals provide blood mealsAn increase in the proportion of immune individualsreduces contact between infectious and susceptibleindividualsReduction in contact between infectious and susceptibleindividuals reduces the hazard that each susceptibleindividual experiencesAnother example of herd immunity effects

22/33




Calculating the Basic Reproductive Number fromForce of Infection

Force of infection can be used to estimate the fraction ofthe population susceptible to infectionApply hazard to population to estimate fraction susceptible.Use theoretical results that 1/(fraction susceptible)estimates basic reproductive number1 = R = sR0

1s = R0

23/33




Estimates of Basic Reproductive Number

R0 associated with more rainfall, higher temperatures andlarger population density

24/33




Estimates of Basic Reproductive Number

Spatial variation in R0 suggests that 78% of the country wouldhave to be targeted compared to 88% of the country without

spatially targeting vaccine

25/33



Conclusions

We find through simulation that vaccines could conceivablyincrease transmission but only vaccines that would not beconsidered for licensureDemographic changes have lead to herd immunity effectsreducing the hazard that young people experienceWe find spatial heterogeneity in the basic reproductivenumber across ThailandSuggests that optimal vaccination strategies may bespatially heterogeneous

26/33



Thank you

27/33



At equilibrium 1=R=sR0

1/s=R0

I

S I

S

28/33



Cumula&

vepropo

r&on

ofinciden

ceofd

enguebyageinBangkok

29/33



Using Age data to estimate the Force of Infection

Cumula&

vepropo

r&on

ofinciden

ceofd

enguebyageinBangkok

30/33



Estimating the Force of Infection

31/33



Fitting models to cumulative age specific incidence

Cumula&

vepropo

r&on

ofinciden

ceofd

enguebyageinBangkok

32/33



Estimates of the Force of Infection

33/33



Partnership with WHO, v2V and Sanofi Pasteur

Organizing a meeting in late summer of modeling groups,epidemiologists and vaccine experts.

Working meeting meant to apply multiple modeling approachesto assess Sanofi Pasteur’s candidates

34/33

dengue vaccines: herd immunity considerations · models of dengue transmission herd immunity...

Documents