Project

endocarditis; bacteremia; bacterial adherence; platelets; inflammation

Lead
Lay summary
Infective endocarditis (IE) is a deadly disease if not treated with antibiotics associated or not to surgery. Therefore, patients at risk were proposed antibiotic prophylaxis whenever they were exposed to medico-surgical procedures that might induce transient bacteremia. Yet, the incidence of IE has not decreased over the last decades. Case-control studies have demonstrated that the large majority of IE were not temporally related to medico-surgical procedures susceptible to induce high-grade transient bacteremia, but rather acquired during normal activities such as chewing or tooth brushing, which are responsible for recurrent bouts of low-grade bacteremia. Hence, a more global prevention of IE would require a more generalized approach. IE is traditionally studied in experimental models in which animals are inoculated i.v. with large bacterial numbers, in order to ensure 100% of infection. However, this might not be experimentally relevant to the human situation if IE is mostly acquired during recurrent low-grade bacteremia rather than after punctual high-grade bacteremia. Moreover, no integrative (or systems) assessment has been attempted to analyze the relative importance of host and microbial factors at the time of disease initiation, and especially not in a low-grade bacteremia realistic model. It is the very purpose of this proposal to fill this gap. We developed a setting allowing comparing the infectivity of punctual high-grade bacteremia (30 second i.v. bolus inoculation) to that of the progressive inoculation over 10 hours, which produces levels of bacteremia unable to induce IE per se. We will use this setting to set up a novel statistical regression model integrating "explanatory variables" (e.g. inoculum size, inoculation duration, bacterial adherence to matrix proteins, reactivity with platelets etc.) and "response variables" (e.g. infection dose, bacteremia level, vegetation counts, inflammatory profile). This analytical step will provide new insights into IE initiation, and help design "best-fit" interventions targeted at the most important factors involved in valve infection (e.g. adherence to host matrix protein, platelet aggregation etc.). Interventions will include pretreatment of animals with anti-adhesin antibodies or peptides, anti-aggregation therapy (with salicylate or ticlopidine), or modulation of the inflammation status if relevant in the statistical modeling. Comparing the analytical and intervention results in both the high-grade bolus and the low-grade continuous models of inoculation should enlighten us on the most appropriate experimental model to study the human disease and lifetime prevention strategies.

Direct link to Lay Summary

Last update: 21.02.2013

Active participation

Title	Year

Number	Title	Start	Funding scheme

Infective endocarditis (IE) is a deadly disease if not aggressively treated with antibiotics associated or not to surgery. Therefore, over the last 50 years patients at risk were proposed antibiotic prophylaxis whenever they were exposed to medico-surgical procedures that might induce transient bacteremia. Yet, the incidence of IE has not decreased over the last decades. This is explained by a changing epidemiology, and partly by the fact that antibiotic prophylaxis is poorly effective on a population scale. First, classical risk factors such as rheumatic heart disease have disappeared from industrialized countries, and are being replaced with new at-risk populations including intravenous drug users, elderly patients with valve sclerosis, and patients with healthcare-related conditions such as those carrying prosthetic valves or undergoing chronic hemodialysis. These patients are so widespread that it is impossible to predict which of them will develop IE and in which circumstance. Second, case-control studies have demonstrated that the large majority of IE were not temporally related to medico-surgical procedures susceptible to induce high-grade transient bacteremia, but rather during normal activities such as chewing or tooth brushing, which are responsible for recurrent bouts of low-grade bacteremia. Hence, a more global prevention of IE would require a more generalized approach.
IE is traditionally studied in experimental models in which animals with catheter-induced valve lesions are inoculated intravenously with large bacterial numbers, in order to ensure 100% of infection. This is convenient to study pathogenesis or the effect of various therapies. However, it might not be experimentally relevant to the human situation, if IE is mostly acquired during recurrent low-grade bacteremia rather than after punctual high-grade bacteremia. Moreover, in spite of many experimental data, no integrative (or systems) assessment has been attempted to analyze the relative importance of host and microbial factors at the time of disease initiation - and especially not in a low-grade bacteremia realistic model. Yet, such knowledge is critical to understand the dynamics of early IE initiation and help target more global preventive measures.
It is the very purpose of this proposal to fill this gap. We recently developed a setting (J. Entenza, unpublished) allowing comparing the infectivity of punctual high-grade bacteremia (30 second i.v. bolus inoculation) to that of the progressive inoculation over 10 hours, which produces levels of bacteremia unable to induce IE per se. These exploratory experiments (see Figure 1) show that progressive low-grade bacteremia is as infective as punctual high-grade bacteremia, provided it is applied long enough. Thus, the area-under-the-curve (AUC) of CFU/ml might be more relevant than peak bacterial concentration in the blood, confirming human case-control studies. We will use this setting to set up a novel statistical regression model integrating “explanatory variables” (e.g. inoculum size, inoculation duration, bacterial adherence to matrix proteins, reactivity with platelets etc.) and “response variables” (e.g. infection dose, bacteremia level, vegetation counts, inflammatory profile). Thus, we shall start to consider the following « saturated regression model » with 16 parameters:
(1) f(average response) = k1 + k2 B-inoc + k3 T-inoc + k4 Bact-adh + k5 Bact-platelet-react + k6 B-inoc*T-inoc + k7 B-inoc* Bact-adh + k8 B-inoc* Bact-platelet-react + k9 T-inoc* Bact-adh + k10 T-inoc* Bact-platelet-react + k11 Bact-adh* Bact-platelet-react + k12 B-inoc*T-inoc*Bact-adh + k13 B-inoc*T-inoc*Bact-platelet-react + k14 T-inoc*Bact-adh*Bact-platelet-react + k15 T-inoc*Bact-adh*Bact-platelet-react + k16 B-inoc*T-inoc*Bact-adh*Bact-platelet-react
This analytical step will provide new insights into IE initiation, and help design “best-fit” interventions targeted at the most important factors involved in valve infection (e.g. adherence to host matrix protein, platelet aggregation etc.). Interventions will include pretreatment of animals with anti-adhesin antibodies or peptides, anti-aggregation therapy (with salicylate or ticlopidine), or modulation of the inflammation status if relevant in the statistical modeling. Interventions will be run in both the high-grade bolus and the low-grade continuous models, as they might generate different results in these two conditions. Moreover, anti-inflammatory interventions will also take advantage of a newly developed IE model in mice, which will allow testing new hypotheses in transgenic animals, including the implication of the “danger-signaling” system NALP-3 (mice provided by Prof. Jurg Tshopp, University of Lausanne) and the blockage of its effector IL-1-beta with anti-IL-1 molecules (anakinra).
The implication of infective variables (e.g. bacterial adherence, platelet-reactivity etc.) might be more realistically studied in the low-grade condition than in “classical” bolus inoculation. Moreover, comparing the analytical and intervention results in both types of inoculation should enlighten us on the most appropriate experimental model to study the human disease, as well as reveal underestimated global and lifetime prevention strategies (e.g. anti-aggregation) as used in cardiovascular prevention.