Cystic fibrosis (CF) is an inherited multi-system disorder characterised by chronic airway infection, pancreatic insufficiency, elevated sweat chloride concentration, impaired fertility and hepatobiliary disease. The condition is due to mutations in the cystic fibrosis transmembrane conductance regulator (CFTR). Lack of CFTR function results in reduced fluid secretion and excessive fluid absorption, with a net effect of producing a thickening of the mucous component of the airway surface liquid. This causes plugging of the sub-mucosal glands and impairment of mucociliary clearance leading to infection, inflammation, and tissue damage resulting in bronchiectasis and a predisposition to infection with pathogens such as Pseudomonas aeruginosa.

P. aeruginosa is the most common and clinically important pathogen in patients with CF. The organism is a Gram-negative, non-fermentative, aerobic bacillus belonging to the family Pseudomonadaceae. Although classified as an aerobic organism P. aeruginosa is a facultative anaerobe which may allow it to survive within the relatively hypoxic lungs of patients with CF. It is ubiquitous within the environment and is particularly isolated from moist areas such as soil and water. The bacterium is frequently found in environmental reservoirs, such as the drains of hospital ward wash basins 1 and aerosols containing P. aeruginosa can be detected when opening taps 23. The isolation of P. aeruginosa from tap water follows contamination of the taps rather than the mains water supply 4.

Acquisition of P. aeruginosa infection in patients with CF can occur at any age. During early stages of P. aeruginosa infection in patients with CF the bacteria are not usually mucoid and can be cleared with aggressive antibiotic treatment 5. Once the patients lung becomes chronically infected the bacteria change to a mucoid phenotype and form biofilms. Most studies would suggest that 70–80% of patients are infected during the teenage years 6. A number of studies have also demonstrated an association between chronic P. aeruginosa infection and increased mortality 789. Acquisition of P. aeruginosa infection by patients with CF is usually from the environment, but recent studies have demonstrated patient to patient transmission of certain strains of P. aeruginosa 10111213. More worryingly, some of these epidemic strains are associated with increased morbidity and treatment burden when compared to non-epidemic strains 1415. While the precise method of transmission of epidemic strains of P. aeruginosa remains unclear, there is evidence that airborne transmission may be important 1617 and therefore a study was designed to investigate this further. The aims of the study were to: 1) examine the effects of varying environmental conditions on the survival of P. aeruginosa within artificially generated aerosols; and 2) identify differences in airborne survival of epidemic and non-epidemic strains under controlled environmental conditions.

The laminar flow apparatus used in this study enabled comparison of recovered concentrations of viable bacteria, during steady state nebulisation, under controlled environmental conditions. In addition to concentration data, the study produced data regarding the size distribution of the aerosol particles containing viable bacteria in the apparatus. Liquid particles suspended in an aerosol are categorized as being either droplets or droplet nuclei. Droplets are generally considered to be greater than 10 μm in diameter 21, whereas droplet nuclei are airborne particles from which most of the liquid has evaporated and therefore less than 10 μm in diameter 2223. Droplets tend to fall to the ground quickly, whereas droplet nuclei have a low terminal velocity and therefore can potentially remain airborne for a number of hours. It has been postulated that Gram negative bacteria such as P. aeruginosa can only survive in droplets and not in droplet nuclei, the implication being that respiratory droplet nuclei are not implicated in the transmission of this pathogen 24. The results of the laminar flow experiments suggest that this is not the case, as they demonstrate that P. aeruginosa can survive within droplet nuclei in artificially generated aerosols, with mucoid strains in particular surviving well in an aerosolised state. In so doing, we have demonstrated that aerial dissemination may be a plausible route of transmission for P. aeruginosa infection.

The Andersen 6-stage sampler is a standard piece of equipment and is probably the most widely used aerobiological sampler 25. While our experiments demonstrated that it is possible to recover P. aeruginosa from bioaerosols using an Andersen 6-stage impactor, it is important to remember that the concentration cultured using the sampler may not be a true representation of the actual concentration of bacteria within the air. It has been estimated that the quantity of bacteria cultured from an air sample represents only approximately 10% of the actual bacterial burden, with the remaining 90% in a viable but non-culturable state 26. Gram-negative bacteria are particularly vulnerable to the shear forces imposed during the sampling process. Consequently, it is generally thought that air samples in the clinical setting tend to underestimate the numbers of Gram-negative bacteria present in the air. Notwithstanding this, it is not known the extent to which these viable but non-culturable bacteria represent a potential risk for infection.

Our experiments demonstrated that the solution used for nebulisation is critical for the survival of the bacteria within aerosols. The use of hypotonic solutions produced a significant reduction in the ability of the bacteria to survive within the aerosol. Inoculation of bacteria into a hypotonic solution will result in an osmotic shock. However, in our experiments we found that this was not sufficient to kill the bacteria, as the control counts from the nebuliser remained stable pre- and post-nebulisation; it may however be that the osmotic shock is enough to leave the bacteria in a weakened state so that subsequent survival in an aerosol is significantly reduced. The reduction in survival may also have been due to the increase in the boiling temperature that occurs with the addition of a solute to a solution. An increase in the boiling temperature will result in a reduced rate of evaporation and hence protect the bacteria from the lethal effects of desiccation. The solution which allowed for the greatest survival of bacteria was 10% FBS. This solution will not only protect the bacteria from osmotic shock and desiccation, but also contains nutrients that will allow the bacteria to continue metabolising whilst suspended in the aerosol. In many respects this represents the case for bioaerosols produced by patients with CF, who will tend to liberate microorganisms in aerosols containing salts and proteinaceous material.

The environmental conditions of the aerosol are important for survival of bacteria within droplet nuclei. Bacteria are more rapidly desiccated in conditions of high temperature or low relative humidity. The data presented here confirms that the survival of P. aeruginosa is adversely effected by a small increase in temperature. However, increased humidity did not statistically improve the survival of P. aeruginosa within the aerosol. Both of these effects may have potential clinical implications as it would suggest patients with CF should avoid cool and damp air conditions in an effort to minimise infection with P. aeruginosa. Consideration should be made to the environmental air conditions when CF units are been designed. Potentially the use of warm and dry air may reduce the risk of cross-infection by P. aeruginosa between patients with CF. It would also seem reasonable to advise patients with CF to avoid the use of devices such as humidifiers as they may promote the survival of P. aeruginosa within aerosols.

The presence of a mucoid phenotype seems to be particularly important for the survival of P. aeruginosa with an aerosol. Isolates of the Liverpool, Leeds Paediatric and Unique CF strains expressing both a mucoid and non-mucoid phenotype were studied in the laminar flow apparatus. For all the strains the presence of a mucoid phenotype resulted in significantly increased concentrations in the aerosols. The mucoid strains over-produce the exopolysaccaride alginate which has been implicated in the persistence and pathogenesis of chronic P. aeruginosa infection in patients with CF 27. The alginate may also improve the survival of the P. aeruginosa in aerosols by providing a protective barrier which reduces the lethal effects of desiccation upon the bacteria.

The data presented from this study provides support for the hypothesis that airborne dissemination may be important for the cross-infection of P. aeruginosa between patients with CF. There have been two studies that have demonstrated that patients with CF can produce aerosols containing viable P. aeruginosa 1617 as well as a study demonstrating that Burkholderia cepacia complex can be disseminated from patients with CF during physiotherapy 28. A further study only published in abstract form examining the frequency of bacterial shedding in the out-patients setting demonstrated that respiratory pathogens were more frequently cultured from the hands and airborne droplets of patients with CF than from environmental surfaces or equipment 29. The sampling at the Liverpool CF unit demonstrated that the outbreak strain P. aeruginosa could be cultured from the air 3 hours after a patient had left the area. To survive this length of time suspended within an aerosol the bacteria must be in a droplet nuclei, because larger droplets would have fallen to the floor due the effect of gravity.

Although this study is limited in that the aerosol residency time for the laminar flow apparatus was relatively short at only just over 1 minute, this was however long enough for the bacteria to travel 4 m – conclusively demonstrating that the bacteria were transported in droplet nuclei and not in droplet form. The concentrations of P. aeruginosa used in the present study were lower than those commonly found in sputum samples from patients with CF which can contain approximately 10⁷ CFU g^-1 30. Although it is theoretically possible that the inhalation of just a few bacteria might result in colonisation, the dose of P. aeruginosa required to cause pulmonary infection in patients with CF is not known

There appears to be little difference in airborne survival between the epidemic and non-epidemic strains of P. aeruginosa expressing a non-mucoid phenotype. However the expression of the mucoid phenotype does seem important for survival within droplet nuclei. This suggests that particular emphasis should be made on segregating individuals free from P. aeruginosa from those with chronic infection who are more likely to be infected with mucoid strains of P. aeruginosa.

Finally, our study suggests that environment appears to influence bacterial survival and optimal conditions for reducing airborne transmission of P. aeruginosa between patients could include warming and drying the air within clinical areas and avoidance of humidification devices.

CF: Cystic fibrosis; CFTR: Cystic fibrosis transmembrane conductance regulator; FBS: Foetal bovine serum; NCIMB: National Collection of Industrial and Marine Bacteria; RH: Relative humidity.

IJC undertook the studies, performed the statistical analysis and drafted the manuscript. LAF participated in the design of the study and assisted in undertaking the studies, CBB, MD, and DGP participated in the design and coordination of the study. All authors read and approved the manuscript.