Infection, reinfection, and vaccination under suboptimal immune protection: epidemiological perspectives
Introduction
Mathematical epidemiological models for the dynamics of microparasite infections that induce lifelong immunity have been extensively developed (Kermack and McKendrick, 1927; Anderson and May, 1991; Grenfell et al., 2001) and used as predictive tools to assist in the design of control programmes (Osborne et al., 2000). This class of infections is viral and usually occurs during childhood (if there is no vaccination occurring), which is indicative of high transmissibility.
Such effective immunity is observed in infections such as measles, mumps and rubella (MMR) but this is unusual. More common is the occurrence of several reinfections throughout life. In similar vein to characterization of a vaccine response (McLean and Blower, 1993), susceptibility to reinfection following a primary infection can be attributed to a combination of two factors: immune protection may wane over time (temporary immunity), or immunity may not be fully protective (partial immunity). Both effects are likely due to a combination of host insufficiency in acquiring and maintaining specific immunity (e.g. tuberculosis—Vynnycky and Fine, 1997; pertussis—Hethcote, 1999; van Boven et al., 2000) and of pathogen ability to generate antigenic diversity thereby avoiding immune recognition (e.g. influenza—Hay et al., 2001; Earn et al., 2002; respiratory syncytial virus—Cane, 2001).
The specific within-host mechanisms allowing recurrent infections are not explicit in this paper. Rather, we construct a series of simple models to investigate the epidemiological consequences of varying the duration and degree of immune protection. The models are systematically analysed and reveal two main outcomes: the duration of immunity has a crucial impact on potential inter-epidemic periods; the degree of immune protection determines a reinfection threshold in transmission responsible for a steep increase in the prevalence of infection. The models are then extended to analyse the effects of temporary and partial immunity on the global impact of vaccination programmes.
Mathematical models have long been associated with the planning of vaccination programmes. The main contributions have been the prediction of the vaccination coverage necessary to eradicate infections (Nokes and Anderson (1992), Nokes and Anderson (1993)), and the illustration of the post-vaccination dynamics if coverage does not meet the eradication threshold (McLean, 1995a). McLean and Blower (1993) have also investigated the consequences of vaccine failures, but under the assumption that immunity induced by natural infection was fully protective. Here this assumption is relaxed to accommodate the recurrence of infections. In the case of temporary protection, we reach the same basic conclusion that the waning of immunity is a major obstacle to disease eradication even if individuals are protected until very late in their lifetime. In the case of partial protection, our conclusions are fundamentally different due to reinfection threshold. Below the reinfection threshold, levels of infection are low and vaccination impact is high. The transmission dynamics are essentially described by the susceptible–infected–recovered (SIR) framework. Populations where transmissibility is above the reinfection threshold are of greater concern. They sustain very high endemicities and vaccination programmes are unable to increase natural immunity further. The transmission dynamics are essentially susceptible–infected–susceptible (SIS). The transition between the two regimes happens over a short range in transmission coefficient, indicating high sensitivity to changes.
Throughout this paper we assume that birth and death rates are equal ensuring a constant host population size, the duration of infection is typically two orders of magnitude lower than the host lifetime, and the periods of infection and infectiousness coincide. Vaccination programmes are implemented at birth. These assumptions simplify the model analysis and do not induce unwanted effects.
Section snippets
SIR and SIS models
The essence of infectious disease dynamics can be represented by simple systems of ordinary differential equations (Anderson and May, 1991; Diekmann and Heesterbeek, 2000). The host population is assumed homogeneous at birth and differentiation occurs as a result of infection experience. The population is assumed to mix homogeneously and transmission is according to the mass-action principle. Hosts are divided into three proportions: susceptible (S), infectious (I), and recovered (R). Under the
Temporary immune protection
Pertussis is a highly contagious infection of the respiratory tract where recurrence has been attributed to the waning of immunity (Hethcote, 1999; van Boven et al., 2000). Control by vaccination is less successful than for other childhood diseases, and recent trends are for overall increase, and occasionally strong epidemic outbreaks.
The temporary immunity model is represented diagrammatically in Fig. 1(c). Our assumption is that, upon infection, individuals develop an immune response that is
Temporary immune protection and vaccination
We proceed to illustrate the impact of mass vaccination at birth under the temporary immunity model. The model, which assumes that protection induced by the vaccine is equivalent to protection acquired in response to natural infection, is formalized by the system of equationswhere the new parameter, v, represents coverage of the vaccination programme. Fig. 7 illustrates the effect of vaccination in this system when α=0.0015. In this case, the average
Partial immune protection and vaccination
Here, we illustrate the impact of mass vaccination at birth in the case where both infection and vaccination induce partial protection. First, we consider the simplest situation of a vaccine that induces as much protection as a previous infection. Then we analyse the impact of more efficacious vaccines.
Discussion
Since the majority of pathogen infections involve reinfection, the dynamic consequences of any intervention on rates of (re)infection should be considered. We suggest that most infections can be classified according to the relative significance of partial and temporary immunity in creating susceptibility to reinfection (Fig. 5). Temporary immunity is, to a large extent, determined by the propensity for antigenic change, but also alteration in host status (e.g. immunosuppression due to other
Acknowledgements
We thank Francisco Dionisio and Isabel Gordo for comments on previous versions of this manuscript. This research was initiated while MGMG was a Wellcome Trust Fellow at the University of Warwick, and continued with funding from The Gulbenkian Foundation, and FCT of Portugal. We thank the Wellcome Trust for grant support for LJW.
References (32)
- et al.
Ecology and evolution of the flu
Trends Ecol. Evol.
(2002) - et al.
Exploring the evolution of diversity in pathogen populations
Trends Microbiol.
(2001) Simulations of pertussis epidemiology in the United Stateseffects of adult booster vaccinations
Math. Biosci.
(1999)- et al.
Seasonally forced disease dynamics explored as switching between attractors
Physica D
(2001) After the honeymoon in measles control
Lancet
(1995)- et al.
Imperfect vaccines and herd immunity to HIV
Proc. R. Soc. London B
(1993) - et al.
A novel TB vaccine; strategies to combat a complex pathogen
Immunol. Lett.
(2003) - et al.
Waning immunity and sub-clinical infection in an epidemic modelimplications for pertussis in The Netherlands
Math. Biosci.
(2000) - et al.
Modelling epidemics caused by respiratory syncytial virus (RSV)
Math. Biosci.
(2001) - et al.
Infectious Diseases of HumansDynamics and Control
(1991)
The dynamics of cocirculating influenza strains confering partial cross-immunity
J. Math. Biol.
Improving vaccines against tuberculosis
Immunol. Cell Biol.
Molecular epidemiology of respiratory syncytial virus
Rev. Med. Virol.
Trypanosoma evansi in Indonesian buffaloesevaluation of simple models of natural immunity to infection
Epidemiol. Infect.
Mathematical epidemiology of infectious diseasesmodel building, analysis and interpretation. Mathematical and Computational Biology
Ecological and immunological determinants of influenza evolution
Nature
Cited by (140)
The reinfection threshold, revisited
2023, Mathematical BiosciencesInteractions between hosts affect virus competition mechanism within an infectious strain
2023, Chaos, Solitons and FractalsPreventive control strategy on second wave of Covid-19 pandemic model incorporating lock-down effect
2022, Alexandria Engineering JournalEpidemiology of influenza virus reinfection in Guangxi, China: a retrospective analysis of a nine-year influenza surveillance data: Characteristics of influenza virus reinfection
2022, International Journal of Infectious Diseases