The current study characterized fungal detection and identification from airway samples of horses. A high proportion of samples were positive for fungi by both culture and cytology, with phagocytosis of fungal elements commonly observed. In addition, there was very little concordance between fungi identified in the respiratory tract and in the hay fed. Finally, high variability of distribution and type of fungal elements in sample sites, of phagocytosis, and between culture and cytology was observed.
Prevalence
The prevalence of fungal detection by cytology in the tracheal wash (69.4% [95% CI .5788-.8083]) in the current study was similar to previously described (79% [95% CI .7612-.8202]) (Dauvillier et al., 2019). In contrast, the prevalence of positive fungal culture in the tracheal wash (82.3% [95% CI .7275-.9177]), with a predominance of Aspergillus and Penicillium, was considerably higher than previously described (up to 55% [95% CI .5125-.5846]) (Dauvillier et al., 2019). No standardized method for mycological examination from respiratory samples has yet been established, and pre-analytical methodology differed between studies (Borman et al., 2010; Denning et al., 2014). In the current study, samples were transported under refrigerated conditions in sterile tubes, while temperature during transport was not mentioned in a previous study (Dauvillier et al., 2019), where Sabouraud agar tubes were used. Transport at room temperature in the mentioned study may have altered fungal growth if analysis was delayed, potentially leading to lower detection rates (Stewart and Cuming, 2015).
Interestingly, phagocytosis of fungal elements was observed in three-quarters of TW samples and most of BALF samples, with spores more often phagocyted than hyphae. Similarly to bacteria, finding fungal elements on cytology has been reported as more suggestive of true infection when phagocyted by neutrophils or macrophages (Bain, 1997; Cian et al., 2015). Additionally, while a small amount of degenerate neutrophils may be normally found in respiratory secretions due to normal clearance from the body (Bain, 1997), a larger amount of degenerate neutrophils has been considered a positive indicator of bacterial infection (Jocelyn et al., 2018). To our knowledge, there is to date no published investigation on degenerate neutrophils related to fungal infections. In the current study, only 3 horses (1 TW, 2 BALR) were identified with degenerate neutrophils, irrespectively to the fungal culture of the same sample.
A high prevalence of fungi was found in the current study, which was similar between horses with and without prior clinical suspicion of respiratory disease. Prevalence was higher in the TW than BALF, both by culture and cytology, and may be in favor of environmental contamination rather than true infection. Further studies are warranted in order to determine whether detecting fungi in the respiratory tract of horses is a risk factor for respiratory disease such as equine asthma, as being previously suggested (Dauvillier et al., 2019).
Samples and modalities
Hay is recognized as one of the primary sources of inhaled fungal particles in horses, and has been used as an initiating/exacerbating agent for eliciting severe equine asthma (Pirie et al., 2003). However, in the current study, only a few fungi found in hay fed were also cultured from corresponding respiratory samples. These findings suggest that hay might not have been the primary source of molds/yeasts detected from the lower respiratory tract of these horses. Testing straw from each yard for mold content, in addition to hay, would have been relevant but was not performed in the current study for financial reasons.
There was only slight agreement between fungal culture and cytology in both TW and BALF, while the prevalence was not different. This is consistent with previous studies showing poor agreement between fungal culture and cytology from the TW, with no correlation between a positive mycology and detection of fungal growth on cytology (Dauvillier et al., 2019).
To our knowledge, this is the first report of pooled BALF cytology from both lungs compared to each lung separately with a focus on fungi detection. There was only moderate agreement between individual (left and right lung) BALF samples. As a consequence, blind sampling of BALF, often leading to wedging of the BAL tube in the right caudal lung, may lead to false negative results for fungal elements on cytology, as also previously demonstrated for neutrophils and hemosiderophages (Couetil and Thompson, 2020; Depecker et al., 2014). In addition, fungal detection by cytology on pooled BALF was significantly different from combined samples from both lungs.
A very high proportion of samples in this study were found to be positive (91.9% in TW, 37.1% in BALF) by fungal culture and/or cytology, rendering questionable the clinical significance of the presence of fungi in the lower respiratory tract of horses, particularly when found in the TW. It has been suggested that to be cytologically significant, large numbers of fungi should be visualized (Stewart and Cuming, 2015). In the current study, only presence/absence (and not relative quantification) was considered, possibly contributing at least partially to the high proportion of positive samples observed. Establishment of a pathological cut-off, both for cytology and mycology, as previously described for bacterial infections and inflammatory diseases (Bain, 1997; Denning et al., 2014; Richard et al., 2010; Richard and Maillard, 2009), could help in differentiating between true fungal infection and environmental contamination.
Limitations
Seasonal variation has been found in the concentration of respirable particles in horse barns (Kutasi et al., 2011; Nardoni et al., 2005), with a significant influence on mEA subtypes based on BALF cytology (Hansen et al., 2018; Secombe et al., 2015). An effect of season on the presence of fungi in the respiratory tract of horses was not reported to the authors’ knowledge, and horses were then indifferently sampled throughout the year in this study.
Sedation protocols were elected at the clinician’s discretion based on the horse’s clinical condition and competition constraints. Antitussive effects of sedatives on BALF cytology results have been previously studied, with no influence of butorphanol on BALF volume collected or neutrophil proportions (Westermann et al., 2005). While fungal elements were not included in these previous investigations, no significant influence of sedatives has currently been expected. Furthermore, the prevalence of detecting fungi by culture or cytology, both on TW and BALF, was not significantly different between referred cases and horses in the field, despite the systematic use of butorphanol for clinical horses and the absence of its use for non-clinical horses.
A total volume of 500 mL of sterile saline per lung was infused, according to recommendations for cytology (Robinson and Chairperson, 2001). However, studies have shown that increasing the volume of fluid instilled for BAL may lead to cytological variations (Couëtil et al., 2016; Orard et al., 2016). Additionally, dilution of sputum had a clear effect on yeast recovery in human patients, with fewer patients found positive with a more dilute sputum (Pashley et al., 2012). Whether infusing a smaller amount would have led to increased retrieval of fungal elements on cytology or culture from BALF is currently unknown.
While endoscopes were efficiently disinfected between horses, a possible BALF contamination due to the passage of the endoscope through the nasal cavities could not be ruled out. However, this was considered unlikely considering the low prevalence of fungi recorded from BALF by either culture or cytology, as well as the observation of horses with negative BALF / positive TW for fungal detection.
Delays from sampling to analysis is another concern. In the current study, all samples were systematically processed within 24 hours. All samples were kept refrigerated (+ 4°C) until processing, as recommended in order to prevent in vitro microbial proliferation and cellular degeneration.(Cian et al., 2015) In addition, signs of fungal proliferation (germinating spores, branching hyphae) were observed in only one sample.