A valuable experimental setup to model exposure to Legionella’s aerosols generated by shower-like systems
Introduction
Members of the genus Legionella (Gram negative bacilli) are ubiquitous in natural and anthropogenic aquatic ecosystems. These bacteria are responsible for severe pneumonia, which may be fatal in 30% of cases when considering nosocomial infections. L. pneumophila is, by far, the most frequent species associated with Legionnaires’ disease (LD). Legionella is now the number one cause of reported waterborne disease in the United States. The latest epidemiological data showed an increase in LD worldwide (Centers for Disease Control and Prevention, 2018; InVS, 2018) with the identification of new sources of contamination by Legionella aerosol dispersion (i.e., car washing stations (Baldovin et al., 2018), street cleaning trucks (Valero et al., 2017), aerosols from biologic wastewater treatment plants (Loenenbach et al., 2018), reclaimed water used for spray irrigation (Pepper and Gerba, 2018), etc.). To assess the risk of infection of LD, researchers have used the quantitative microbial risk assessment (QMRA) method to provide models for aerosol dispersion (Thomas W. Armstrong and Haas, 2007; T. W. Armstrong and Haas, 2007; Buse et al., 2012; Hamilton and Haas, 2016; Schoen and Ashbolt, 2011). However, these risk models are only based (i) on the fluidic characteristics of water systems that could generate bioaerosols and (ii) on the infectious doses extrapolated to humans from animal experiments using inhalation, intraperitoneal injections or tracheal instillation for infection of the animal model.
In previous works, we used a medical nebulization device to generate Legionella aerosols. Thus, we extensively characterized the bioaerosols produced by a vibrating-mesh nebulizer in terms of aerodynamic features and airborne Legionella emitted dose (Allegra et al., 2016). We also experimentally assessed the deposition of these bacteria in the pulmonary region using a 3D replica of the upper airways (Pourchez et al., 2017) and using an ex vivo ventilated porcine lung as an innovative human-like respiratory model (Perinel et al., 2018). We showed that the developed experimental setup can be used to mimic bacterial inhalation by an anatomical model of the respiratory tract to assess the Legionella dose reaching the thoracic region for a given bioaerosol source. Therefore, an interesting further step would consist of demonstrating that nebulizers can be used to satisfactorily mimic bioaerosol exposure during showering events. Indeed, if cooling towers are the most significant sources of Legionella outbreaks at a community level, spa pools and showers from public facilities are the most significant sources of Legionella nosocomial outbreaks (K. A. Hamilton et al., 2018).
For this purpose, we compared (in terms of aerodynamic size distribution and bacteria concentration) two different technologies of bioaerosol generation: a vibrating-mesh nebulization system and a jet nebulizer. The jet nebulizer has been in continuous development since medicinal aerosol delivery started in the 19th century. Today, the majority of jet nebulizers are inexpensive devices operated via compressed gas. The gas passes through a small aperture in the nebulizer to collect and atomize the liquid. The aerosol generated by atomization contains large and small droplets and is driven to a baffle. Large droplets are impacted by the baffle, while small droplets are transported out of the nebulizer. In contrast, the vibrating-mesh nebulization system is a recent technology for aerosol generation using an annular piezo element to produce mesh vibration to push the liquid through the mesh. Holes in the mesh have a conical structure, with the largest cross-section of the cone in contact with the liquid. The mesh deforms into the liquid side, thus pumping and loading the holes with liquid. This deformation on the other side of the liquid-drug reservoir ejects droplets through the holes. We hypothesize that with these two technologies, our experimental setup will be able to reproduce many different shower-like systems on the market place.
For adaptation of our previous model to showering facilities, the original replica of the human upper airways (El Merhie et al., 2015; Leclerc et al., 2014) was placed in an experimental sealed enclosure mimicking a shower cubicle and connected to a filter mimicking the thoracic region (Perinel et al., 2018; Pourchez et al., 2017). A respiratory pump was used to fit the breathing parameters corresponding to adult male physiology at rest: breathing rate of 15 breaths per minute and tidal volume of 500 mL (Gradon and Marijnissen, 2003; NF EN 13544-1, 2002). The results obtained using this experimental setup were compared with the MMAD and the emitted dose of bioaerosols collected by the Coriolis® Delta air sampler in the lab’s shower that is routinely used by the staff. The lab shower will be our reference in this work and hereafter referred to the “real” shower.
Section snippets
Description of the “real” shower experiments used as a reference
The shower analysed in this study is located on the ground floor of the building of the Center for Health Engineering of Mines Saint-Etienne (CIS, Saint-Etienne, France). This shower is regularly used by the academic staff and postgraduate students at lunchtime after sporting activities.
Three independent experiments were carried out. The concentration of bacteria in the water network was determined by sampling 1 L of water from the first flow and 1 L during a 30-min showering event respectively
Quantification of the exposure to airborne bacteria during 1 min in the shower as a reference
The technical features of the “real” shower are presented in Fig. 2. As determined by qPCR, the total bacteria concentration in the water network was 3.3 ± 1.0 × 104 bacteria.L−1. Assuming that the aerosols are evenly distributed in the shower space, when extrapolating the numbers, we found that the mean number of bacteria collected per minute by the Coriolis® Delta air sampler was 8.6 ± 3.2 × 102. As the shower cubicle has a volume of 1.125 m3, the Coriolis® system (300 L min−1) can collect
Discussion
QMRA provides a very useful methodology to manage the risk of infections by foodborne or some waterborne pathogens with simple cases of contamination. Legionella is currently a well-documented human waterborne pathogen, but as shown by Bentham and Whiley (Bentham and Whiley, 2018; Whiley et al., 2014), persistent and indeterminable uncertainties remain regarding Legionella epidemiology, frequency of exposure, aerosol characterization (quantity, size, content of Legionella), strain infectivity,
Conclusion
- •
Viable and cultivable aerosolized Legionella bacilli were detected as early as the 262 nm cut-off diameter of DLPI, confirming that Legionella bacilli are approximately 1–2 μm long and 0.3–0.9 μm wide and can reach the deep lung.
- •
The Legionella physiological states were not modified by the nebulization process or by its transportation through the 12-stage DLPI or the anatomical model.
- •
After a 1-min shower (used as a reference in this study), contaminated by 3 × 104 Legionella cells/L, 22 viable
Funding information
This study was supported by a grant from Jean Monnet University, France.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
We would like to thank: (i) Dr. C. Ginevra from the French Reference Center for Legionella for providing us with the GFP-modified strain. (ii) M. Landraud, P. Bourrassaud, A. Moravia, Z. Reynes, M. Pareux and M. Escot for the validation of the model during their internships and (iii) the teaching laboratory of Technical University Institute of Saint-Etienne for providing the environmental bacteria strains. Finally, we would like to dedicate this paper to the memory of Serge Riffard.
References (48)
- et al.
Assessing the risk of Legionnaires’ disease: the inhalation exposure model and the estimated risk in residential bathrooms
Regul. Toxicol. Pharmacol.
(2013) - et al.
Legionellae in engineered systems and use of quantitative microbial risk assessment to predict exposure
Water Res.
(2012) - et al.
Viability and infectivity of viable but nonculturable Legionella pneumophila strains induced at high temperatures
Water Res.
(2019) - et al.
Evaluation of exposure to brevundimonas diminuta and Pseudomonas aeruginosa during showering
J. Aerosol Sci.
(2017) - et al.
Impact of acoustic airflow nebulization on intrasinus drug deposition of a human plastinated nasal cast: new insights into the mechanisms involved
Int. J. Pharm.
(2011) - et al.
Droplet distribution and airborne bacteria in an experimental shower unit
Water Res.
(2018) - et al.
Health risks from exposure to Legionella in reclaimed water aerosols: toilet flushing, spray irrigation, and cooling towers
Water Res.
(2018) - et al.
Using quantitative microbial risk assessment and life cycle assessment to assess management options in urban water and sanitation infrastructures: opportunities and unresolved issues
Microb. Risk Anal., Quantitative Microbial Risk Assessment of Reclaimed Water
(2017) - et al.
Impact of acoustic airflow on intrasinus drug deposition: new insights into the vibrating mode and the optimal acoustic frequency to enhance the delivery of nebulized antibiotic
Int. J. Pharm.
(2015) - et al.
Assessing sinus aerosol deposition: benefits of SPECT-CT imaging
Int. J. Pharm.
(2014)