Why, when and how should exposure be considered at the within-host scale? A modelling contribution to PRRSv infection.

Understanding the impact of pathogen exposure on the within-host dynamics and its outcome in terms of infectiousness is a key issue to better understand and control the infection spread. Most experimental and modelling studies tackling this issue looked at the impact of the exposure dose on the infection probability and pathogen load, very few on the within-host immune response. Our aim was to explore the impact on the within-host response not only of the exposure dose, but also of its duration and peak, for contrasted virulence levels. We used an integrative modelling approach of the within-host dynamics at the between-cell level. We focused on the porcine reproductive and respiratory syndrome virus, a major concern for the swine industry. We quantified the impact of exposure and virulence on the viral dynamics and immune response by global sensitivity analyses and descriptive statistics. We found that the area under the viral curve, an indicator of the infection severity, was fully determined by the exposure intensity. The infection duration increased with the strain virulence and, for a given strain, exhibited a positive linear correlation with the exposure intensity logarithm and the exposure duration. Taking into account the exposure intensity is hence necessary. Besides, representing the exposure due to contacts by a single punctual dose would tend to underestimate the infection duration. As the infection severity and duration both contribute to the pig infectiousness, a prolonged exposure of the adequate intensity would be recommended in an immuno-epidemiological context.

Modelling Salmonella transmission among pigs from farm to slaughterhouse: Interplay between management variability and epidemiological uncertainty.

Salmonella carriage and cutaneous contamination of pigs at slaughter are a major risk for carcass contamination. They depend on Salmonella prevalence at farm, but also on transmission and skin soiling among pigs during their journey from farm to slaughterhouse. To better understand and potentially control what influences Salmonella transmission within a pig batch during this transport and lairage step, we proposed a compartmental, discrete-time and stochastic model. We calibrated the model using pork chain data from Brittany. We carried out a sensitivity analysis to evaluate the impact of the variability in management protocols and of the uncertainty in epidemiological parameters on three model outcomes: prevalence of infection, average cutaneous contamination and number of new infections at slaughter. Each outcome is mainly influenced by a single management factor: prevalence at slaughter mainly depends on the prevalence at farm, cutaneous contamination on the contamination of lairage pens and new infections on the total duration of transport and lairage. However, these results are strongly affected by the uncertainty in epidemiological parameters. Re-excretion of carriers due to stress does not have a major impact on the number of new infections.

Integrative model of the immune response to a pulmonary macrophage infection: what determines the infection duration?

The immune mechanisms which determine the infection duration induced by pathogens targeting pulmonary macrophages are poorly known. To explore the impact of such pathogens, it is indispensable to integrate the various immune mechanisms and to take into account the variability in pathogen virulence and host susceptibility. In this context, mathematical models complement experimentation and are powerful tools to represent and explore the complex mechanisms involved in the infection and immune dynamics. We developed an original mathematical model in which we detailed the interactions between the macrophages and the pathogen, the orientation of the adaptive response and the cytokine regulations. We applied our model to the Porcine Respiratory and Reproductive Syndrome virus (PRRSv), a major concern for the swine industry. We extracted value ranges for the model parameters from modelling and experimental studies on respiratory pathogens. We identified the most influential parameters through a sensitivity analysis. We defined a parameter set, the reference scenario, resulting in a realistic and representative immune response to PRRSv infection. We then defined scenarios corresponding to graduated levels of strain virulence and host susceptibility around the reference scenario. We observed that high levels of antiviral cytokines and a dominant cellular response were associated with either short, the usual assumption, or long infection durations, depending on the immune mechanisms involved. To identify these mechanisms, we need to combine the levels of antiviral cytokines, including IFNγ, and IL10. The latter is a good indicator of the infected macrophage level, both combined provide the adaptive response orientation. Available PRRSv vaccines lack efficiency. By integrating the main interactions between the complex immune mechanisms, this modelling framework could be used to help designing more efficient vaccination strategies.

Fat-free mass predictions through a Bayesian Network enable body composition comparisons in various populations.

The respective contribution of fat-free mass (FFM) and fat mass to body weight (Wgt) is a relevant indicator of risk for major public health issues. In an earlier study, a Bayesian Network (BN) was designed to predict FFM from a DXA database (1999-2004 NHANES, n = 10,402) with easily accessible variables [sex, age, Wgt, and height (Hgt)]. The objective of the present study was to assess the robustness of these BN predictions in different population contexts (age, BMI, ethnicity, etc.) when covariables were stochastically deduced from population-based distributions. BN covariables were adjusted to 82 published distributions for age, Wgt, and Hgt from 16 studies assessing body composition. Anthropometric adjustments required a surrogate database (n = 23,411) to get the missing correlation between published Wgt and Hgt distributions. Published BMI distributions and their predicted BN counterparts were correlated (R(2) = 0.99; P < 0.001). Predicted FFM distributions were closely adjusted to their published counterparts for both sexes between 20 and 79 y old, with some discrepancies for Asian populations. In addition, BN predictions revealed a very good agreement between FFM assessed in different population contexts. The mean difference between published FFM values (61.1 ± 3.44 and 42.7 ± 3.32 kg for men and women, respectively) and BN predictions (61.6 ± 3.11 and 42.4 ± 2.76 kg for men and women, respectively) was <1% when FFM was assessed by DXA; the difference rose to 3.6% when FFM was assessed by bioelectric impedance analysis or by densitometry methods. These results suggest that it is possible, within certain anthropometric limitations, to use BN predictions as a complementary body composition analysis for large populations.

Body composition predicted with a Bayesian network from simple variables.

The relative contributions of fat-free mass (FFM) and fat mass (FM) to body weight are key indicators for several major public health issues. Predictive models could offer new insights into body composition analysis. A non-parametric equation derived from a probabilistic Bayesian network (BN) was established by including sex, age, body weight and height. We hypothesised that it would be possible to assess the body composition of any subject from easily accessible covariables by selecting an adjusted FFM value within a reference dual-energy X-ray absorptiometry (DXA) measurement database (1999-2004 National Health and Nutrition Examination Survey (NHANES), n 10 402). FM was directly calculated as body weight minus FFM. A French DXA database (n 1140) was used (1) to adjust the model parameters (n 380) and (2) to cross-validate the model responses (n 760). French subjects were significantly different from American NHANES subjects with respect to age, weight and FM. Despite this different population context, BN prediction was highly reliable. Correlations between BN predictions and DXA measurements were significant for FFM (R2 0·94, P < 0·001, standard error of prediction (SEP) 2·82 kg) and the percentage of FM (FM%) (R2 0·81, P < 0·001, SEP 3·73 %). Two previously published linear models were applied to the subjects of the French database and compared with BN predictions. BN predictions were more accurate for both FFM and FM than those obtained from linear models. In addition, BN prediction generated stochastic variability in the FM% expressed in terms of BMI. The use of such predictions in large populations could be of interest for many public health issues.