Pathogen adaptation under imperfect vaccination: implications for pertussis

1
Pathogen adaptation under imperfect vaccination
implications for pertussis
Michiel van Boven1, Frits Mooi2,3, Hester de
Melker3Joop Schellekens3 Mirjam
Kretzschmar31Wageningen University/Utrecht
University2Utrecht University/Academic Hospital
Utrecht3National Institute of Public Health
the Environment
2
Pertussis, basic facts

• gram-negative bacterium
• first described 1540 !
• first isolated 1906 by Bordet and Gengou
• main species in the genus Bordetella B.
  pertussis, B. parapertussis, and B.
  bronchiseptica
• B. pertussis and B. parapertussis mostly human
• B. bronchiseptica dogs, pigs, sheep
• Bp and Bpp limited survival outside the host
• Bb prolonged starvation resistance
• Bp and Bpp infections severe in unvaccinated
  infants, usually mild in adolescents and adults

3
Pertussis vaccination

• before 1940 a leading cause of infant death
• nowadays very low mortality rates in developed
  countries
• Dutch vaccination program started in 1953
• vaccine killed whole-cell (Tohama)
• vaccination coverage 96
• up to 2002 vaccination at age 3,4,5, and 11
  months
• since 2002 vaccination at age 2,3,4, and 10
  months
• since 2002 booster with subunit vaccine at 4
  years
• 2006 replacement of whole-cell vaccine by
  subunit vaccine
• subunit vaccines 1-5 components (e.g., ptx,
  pertactin, fha)

4
Pertussis trend in the Netherlands
5
Age distribution of cases before and after 1996
6
Distribution of cases by vaccination status
7
Virulence genes of B. pertussis
8
Phase modulation in the bordetellae
9
Questions

• What is the contribution of circulation in
  unvaccinated infants to the overall circulation
  of pertussis?
• How does the infection incidence depend on period
  of immunity after vaccination or infection?
• How will the pathogen population evolve in
  response to vaccination?

10
Model structure
Central idea there is a difference between
infection in immunologically naïve individuals
(primary infection) and infection in
individuals whose immune system has been primed
(secondary infection)
11
Model parameters
12
Population dynamical analysis invasion

• herd immunity cannot always be achieved (McLean
  and others)
• the reproduction ratio increases with p if
• for the default parameter values, Rp increases
  with p if secondary infections are 7 more
  transmissible than primary infections

13
Population dynamical analysis endemicity
14
Evolutionary adaptation
Adaptation of B. pertussis to vaccination occurs
in two ways(1) the pathogen population may
evolve to become polymorphic (2) the
pathogen may evolve higher or lower levels
of virulence gene expression
15
Scenarios

• B. pertussis can increase (or decrease) its
  efficiency in immunologically naïve individuals
  by increasing (decreasing) the expression of
  virulence genes. On the other hand, increased
  expression of virulence genes results in a
  stronger immune response in primed individuals.
• B. pertussis can evolve to circumvent the
  immunity induced by vaccination. However, strains
  that circumvent the vaccination induced immune
  response have reduced fitness.

16
Evolutionary invasion analysis

• fitness measure the growth rate ?(y,x) of a
  mutant strain characterized by a variable y
  in a resident pathogen population
  characterized by a variable x
• the selection gradient
• ESS condition
• maximum condition
• convergence condition

17
1. virulence gene expression

• In the first example, the parameters f1 and f2
  are molded by selection.
• For this scenario, the ESS condition reads

18
1. virulence gene expression
19
1. virulence gene expression
trade-off
20
2. immune evasion

• In this example, the parameters sV and a are
  supposed to be molded by selection, and the
  ESS condition reads

21
2. immune evasion

• Suppose that a resident strain is present that
  cannot infect individuals in class V (gv0)
• The infectious period of the resident strain is
  days.
• A mutant strain that is fully able to infect
  individuals in class V (i.e. gv0) can
  invade if its infectious period is not
  shorter than days.
• If the period of protection after vaccination is
  ten years (instead of five), the mutant can
  invade the infectious period is not shorter
  than days.

22
Pathogen adaptation summary of results

• For realistic parameter values primary
  susceptibles constitute only a small fraction of
  the population, while secondary susceptibles
  abound. Consequently, pertussis circulation
  depends mainly on (unnoticed) infections in
  children, adolescents and adults.
• The pathogen is more likely to adapt to
  efficiently exploit secondary susceptibles than
  to efficiently exploit primary susceptibles.
• Pertussis strains that evade the immunity induced
  by vaccination can only invade if they incur no
  or a modest fitness cost.

23
Tests and open questions

• How long does immunity, against infection and
  against disease, last after infection and
  vaccination?
• Are there systematic differences between strains
  found in countries with high vaccination
  coverage and strains found in countries with low
  vaccination coverage?

24
The optimal amount of antiviral control

• Michiel van Boven1, Don Klinkenberg1, Franjo
  Weissing2, Hans Heesterbeek1
• 1Faculty of Veterinary Medicine, Utrecht
  University
• 2Theoretical Biology, University of Groningen

25
Main question What is the optimal amount of
costly (i.e. potentially lethal) antiviral
therapy when faced with a virulent pathogen
that can kill the host?
26
Two perspectives

• the public health officer maximize the
  performance of the population
• the individual maximize your own performance
  given the actions of those around you

27
Objective functions

• life expectancy, L(y,x)
• probability to be alive after T years, L(y,x,T)
• perceived risk, L(y, I(x), V(x))

28
Model structure
µ background mortality ? recovery rate ?
antivirals induced mortality ? antiviral
control rate a infection induced mortality s
non-compliance rate ? force of infection
29
1. Life expectancy at the endemic equilibrium

• pathogen absent
• no antiviral control
• no individual differences
• rare type ?y in a resident population ?x

30
Endemic pathogens, life expectancy as objective
function
31
Endemic pathogens, life expectancy as objective
function
32
Endemic pathogens, limited time horizon
33
Endemic pathogens, limited time horizon
34
Outbreak situations, limited time horizon
?