Epidemiological consequences of immune sensitisation by pre-exposure to vector saliva

Abstract

Blood-feeding arthropods-like mosquitoes, sand flies, and ticks-transmit many diseases that impose serious public health and economic burdens. When a blood-feeding arthropod bites a mammal, it injects saliva containing immunogenic compounds that facilitate feeding. Evidence from Leishmania, Plasmodium and arboviral infections suggests that the immune responses elicited by pre-exposure to arthropod saliva can alter disease progression if the host later becomes infected. Such pre-sensitisation of host immunity has been reported to both exacerbate and limit infection symptoms, depending on the system in question, with potential implications for recovery. To explore if and how immune pre-sensitisation alters the effects of vector control, we develop a general model of vector-borne disease. We show that the abundance of pre-sensitised infected hosts should increase when control efforts moderately increase vector mortality rates. If immune pre-sensitisation leads to more rapid clearance of infection, increasing vector mortality rates may achieve greater than expected disease control. However, when immune pre-sensitisation prolongs the duration of infection, e.g., through mildly symptomatic cases for which treatment is unlikely to be sought, vector control can actually increase the total number of infected hosts. The rising infections may go unnoticed unless active surveillance methods are used to detect such sub-clinical individuals, who could provide long-lasting reservoirs for transmission and suffer long-term health consequences of those sub-clinical infections. Sensitivity analysis suggests that these negative consequences could be mitigated through integrated vector management. While the effect of saliva pre-exposure on acute symptoms is well-studied for leishmaniasis, the immunological and clinical consequences are largely uncharted for other vector-parasite-host combinations. We find a large range of plausible epidemiological outcomes, positive and negative for public health, underscoring the need to quantify how immune pre-sensitisation modulates recovery and transmission rates in vector-borne diseases.

Fig 1. Schematic of vector-borne disease dynamics model when host immunity can be pre-sensitised through vector saliva pre-exposure.

Susceptible hosts (H_S) may become pre-sensitised ($H_S^{\prime }$) when bitten by either a susceptible vector (V_S), an exposed but non-infectious vector (V_E), or an infectious vector (V_I) if parasite transmission is unsuccessful. Susceptible hosts that are pre-exposed to vector saliva remain sensitised until the protective status is lost over time. Susceptible pre-sensitised and naïve hosts ($H_S^{\prime }$ and H_S respectively) can become infected ($H_I^{\prime }$ and H_I respectively) when bitten by an infectious vector. Upon recovering from infection, the host gains immunity against future infections (H_R), but that immunity wanes over time. Further details can be found in Methods. Infection routes of hosts are shown in thick black; pre-sensitisation routes in grey; infection of vectors in thin black. The movement between classes is shown in solid lines and the interactions that lead to that movement are shown by dashed lines.

Fig 2. Interventions targeting vector survival, such as insecticide spraying, increase the likelihood of immune pre-sensitisation through pre-exposure to vector saliva.

Shown are the equilibrium abundances in the ODE model of (a) susceptible (V_S; blue), exposed (V_E; purple) and infectious (V_I; red) vectors, and (b) infected hosts that are not pre-exposed (H_I; red) and that are pre-exposed ($H_I^{\prime }$; pink) to vector saliva. Here, the x-axis is the daily rate of vector mortality imposed by vector control. Pre-sensitised and naïve infected hosts are assumed to have identical recovery rates (γ_H = γ_H′ = 60⁻¹ per day) and transmission probabilities (T_HV = T_H′V = T_VH = T_VH′ = 0.5). Note that the force of infection from vectors, rT_HVV_I, is proportional to the abundance of infectious vectors, and the rate of immune pre-sensitisation through vector saliva pre-exposure, rP_HV(V_S + V_E + (1 − T_HV)V_I), is roughly proportional to the abundance of susceptible vectors (notice that V_S is at least one order of magnitude larger than V_E or V_I).

Fig 3. Increasing vector mortality can elevate pre-sensitised and total infection cases when pre-exposure to vector saliva prolongs the time to recovery.

Shown are (a) the abundance of infectious vectors, (b) naïve infected hosts, and (c) pre-sensitised infected hosts in the ODE model. The total infection cases in the host population (the sum of b and c) can increase with vector mortality, since pre-sensitised infections are more abundant (note the difference in scale between b and c). The dashed grey line shows the result when pre-exposure has no effect (i.e. γ_H′ = γ_H). Shown in cool colours are the results when pre-sensitisation causes decreased recovery rates with the strength of pre-sensitisation reflected in the intensity of blue: recovery rates of pre-sensitised infected hosts equaling $\frac{1}{2}$, $\frac{1}{3}$, and $\frac{1}{5}$ th of the recovery rate of naïve infected hosts. Conversely, the effects of pre-exposure as increased recovery rate are shown in warm colours reflecting 2, 3, and 5 times the recovery rate of naïve infected hosts.

Fig 4. Both ordinary (ODE) and delay (DDE) differential equation models point to the possibility of an adverse consequence of moderate vector control interventions.

Shown are the percentage of infection cases relative to the pre-intervention level predicted by (a) the ODE model and (b) the DDE model plotted against the intensity of intervention-driven vector mortality. The colour keys are as described in Fig 3.

Fig 5. The severity and detectable signs of the adverse interaction between vector saliva pre-exposure and increased vector mortality are likely to be influenced by the effect of saliva pre-sensitisation in reducing the rate of recovery and the assumption about the variability in parasite development.

Shown are (a) the level of intervention-driven vector mortality (“effort”) required to achieve a 90% reduction in the number of infected hosts and (b) the percentage of the pre-intervention number of infected hosts after a small reduction (25%) in vector lifespan due to vector control. The x-axis denotes the assumed effect of saliva pre-exposure on recovery as the proportional reduction in recovery rate.