Chapter 17 Emerging Infectious Diseases
17.1 Overview and Learning Objectives
In this chapter, we will take a look at the evolutionary processes underlying the emergence of new Infectious Diseases.
The learning objectives for this chapter are:
- Understand the central mechanisms and drivers of emergence
- Assess the potential impact of different control strategies on potential future ID emergence
17.2 Introduction
The definition of emerging infectious disease (EID) is a bit vague, but in general a pathogen that is ‘new’ to a specific population can be considered an EID. A ‘new’ pathogen can mean a genuinely new one (e.g. SARS or MERS-CoV in humans), or a new variant of an existing pathogen (e.g. the 2009 H1N1 influenza that caused a pandemic, or drug resistant Staph Aureus). In each case, there is usually a mix of deterministic drivers (e.g. changes in climate, population density, antibiotic use) and random chance that lead to the emergence of new diseases.
17.3 Emergence and Evolution
The word ‘emergence’ describes what is observed, a new disease shows up for the first time (emerges) in a population. This emergence is almost always linked to evolutionary processes. For instance, a pathogen that is already in the human population evolves drug resistance. Or a new pathogen is introduced into the human population and then undergoes evolution to adapt to the new host. While one could define emergence as the phenomenon and evolution as the mechanism, in practice, you will often find these terms used interchangeably, and the exact meaning will depend on how the author defines it. In this chapter, we will discuss both the mechanisms of evolution and how they can lead to the emergence of new pathogens. The ability to understand and predict the emergence of new IDs is, of course, of vital importance, recent references providing additional information are (Brett, Drake, and Rohani 2017; Dibble et al. 2016).
17.4 Ecological drivers of evolution and emergence
Many external, ecological factors affect ID dynamics. Weather, climate, vegetation, the abundance and density of hosts and other species, nutritional status, and general living conditions are all important factors influencing human and animal ID. This is true both for short-term dynamics (e.g., weather patterns can affect disease incidence) as well as for the long-term evolutionary dynamics. The emergence of many new viral pathogens among humans (e.g., HIV, SARS, MERS) is likely caused by the closer contact humans have with the animals who are the main hosts of those diseases. Transmission in these situations of close human-animal contact is often called spillover. Similarly, as discussed in the previous chapter, measles and other diseases were only able to emerge and establish themselves among humans once human populations became large enough. As global warming will affect climate and as people become more affluent and urbanized, we can expect further changes in ID dynamics and evolution over the next decades - though predicting what changes exactly will occur, and which diseases might increase and decrease in importance is very difficult (Altizer et al. 2013).
17.5 Emergence dynamics
During the process of emergence, the new disease is initially - by definition - not present (or at least not detectable by our surveillance methods). Initially a small number of new hosts are infected by the new disease. The disease “bounces around” for a while with few transmission events and overall low numbers. If conditions are right (the local reproductive number is greater than 1), the disease might take off and spread. In contrast to deterministic models, a reproductive number that is larger than 1 is necessary, but not sufficient for the pathogen to produce an outbreak. An initial introduction can by chance be followed by extinction (i.e. an infected person recovers before they can infect someone else), even if R0>1. Thus, the concept derived from deterministic models, that there is no outbreak for R0<1 and one gets an outbreak for R0>1 needs to be modified for stochastic (arguably closer to the real world) settings. An R0<1 still means no outbreak, but now for R0>1, an outbreak is not guaranteed. Instead, there is a certain probability that an outbreak happens. We won’t go into details here (see e.g. (Matt J. Keeling and Rohani 2008)) but one finds that for an SIR-type stochastic model, the probability that an outbreak occurs, P, if started by a single infected is given by P=1-1/R0. Thus, as expected, the larger R0, the more likely an outbreak is to occur, but it’s not guaranteed. Similarly, one can show that if there are initially N infected individuals, the probability for an outbreak is P=1-1/RN0. Again, this makes intuitive sense, namely more infected individuals make it more likely that an outbreak occurs. For instance if the pathogen has an R0=2 and there are 10 initially infectious individuals, the probability for an outbreak is 99.9%. That means as soon as the first few individuals are infected, it is almost certain that the disease takes off.
Once a pathogen has spread enough that our surveillance systems notice it, we consider this new disease as having emerged in that population. The newly emerged (observed) disease can then either be brought under control relatively quickly (e.g. 2003 SARS outbreak), or with much effort (e.g. 2014 Ebola outbreak), or not be fully controlled but continue at a low level (e.g. MERS-CoV) or uncontrolled and spreading widely (e.g. 2009 influenza pandemic).
17.6 Modeling ID Emergence Dynamics
To study evolutionary dynamics, we will need to implement models that allow for changes in the system on top of the non-evolutionary dynamics of the ID. Unless we plan to model many different potential new variants, it is often easiest to pre-specify the number of variants we want to track. In the simplest form where we only track pathogen evolution, we might model the wild-type (normal, pre-existing) form of the pathogen and a single mutant that is different in some important characteristic, e.g., resistant to a drug or able to evade a vaccine. We would build a model with these compartments. We often also pre-specify the characteristics (i.e., the parameter values) for both the wild-type and mutant. We implement the process of mutant generation in the model and run the simulation. Under certain circumstances, we might see the mutant be generated and take over the population.
If we want to model many different mutants, maybe allowing for random, not pre-specified, differences in their fitness, and perhaps even allowing for host evolution, the models get quite a bit more complicated. They are not necessarily conceptually harder, but there is more bookkeeping and coding involved making it technically trickier.
17.7 Summary and Cartoon
This module provides a brief discussion of ID emergence.
17.8 Exercises
17.9 Further Resources
- A few nice overview articles are (M. E. J. Woolhouse 2002)