**4. Discussion**

The diagnosis of severe equine asthma (EA) is mainly based on history and clinical signs [2,3,6,25]. Collateral diagnostic investigations with additional tests, such as an airway endoscopy or lung function evaluation, can confirm and further characterize the diagnostic suspect of severe equine asthma. Among these additional tests, the cytological examination of bronchoalveolar lavage fluid (BALF) is still considered as a 'gold standard' in the diagnosis of severe lower airway inflammation in the horse [2,6,16,18].

However, if on the one hand the BALF cytology is able to disclose neutrophilic inflammation during the exacerbation of the disease [14], on the other hand, the inflammatory pattern completely normalizes during periods of remission, induced by antigen avoidance strategies [17,26]. Clinical signs are also largely reversible after treatment with inhaled or systemic corticosteroids, in absence of harmful environment [2,17]. Conversely, remodeling of the peripheral airways in the severe EA persists during the remission phases and it is related to residual airflow obstruction [26].

Airway remodeling includes increased airway smooth muscle mass, goblet cell hyperplasia/metaplasia, collagen and elastic fiber deposition within the lamina propria, peribronchiolar fibrosis, airway obstruction by mucus and inflammatory cells, airway adventitial inflammation [17,27]. Therefore, it is not ye<sup>t</sup> clear which role the above-mentioned remodeling process plays and how much it may affect airways inflammation and especially in maintaining disease during asymptomatic periods. Moreover, due to the absence of a significant association between lower airway inflammation detected with BALF and the degree of pulmonary dysfunction in equine asthma, the study of the structural small and large airway alterations has recently gained interest [2]. Identifying specific pathologic variables by endoscopic biopsy specimens could possibly facilitate the reaching of a diagnosis during the remission and monitoring response to treatment [17].

The histopathological characteristics of the bronchial mucosa and their evaluation in the diagnosis of severe EA are currently poorly characterized and very few information concerning the histological evaluation of endobronchial biopsies of the horse is present in the literature [17,18,28]. The main reasons for this lack consist of the fact that the transthoracic biopsy samples are a risky and expensive procedure, and therefore not carried out in clinical practice. The only studies existing in the literature for the histological evaluation of peripheral airway remodeling concern histological studies on lungs of horses during autopsy or slaughter [27,29]. On the contrary, the only method that can be used in life is an endobronchial biopsy sampling from the central airways, which is quite easy and repeatable. Nevertheless, the biopsies obtained from the large respiratory tract do not

specifically collect material from the areas affected by the pathology; the information is still very scarce in the course of EA [2]. If the effects were similar, you could then be able to make a diagnosis during the clinical remission phase of the severe EA. Furthermore, the studies carried out on BALF and on endobronchial biopsy specimens have focused on biomolecular type investigations gene expression [5,30–40]. Contrariwise, only very few papers deal with biopsies samples on paraffin sections for histological and immunohistochemical evaluations [5,17,18,20,37].

In our procedure, the difficulties in bioptic procedures encountered in the first sampling (EA-affected 3), suggested by the classic methodologies, which made the sampled material too superficial and unsuitable for evaluation, were largely solved using a manual technique that provided for access to the wall of the bronchus, by withdrawing the forceps towards the endoscope and at the same time by exerting a twist in such a way as to acquire a larger sample.

Bullone et al., 2016 first developed a 14-point histological scoring system that assesses histological variables of inflammation and remodeling of the large airways using endoscopic biopsy specimens [17]. This evaluation system, however, although considered a reliable tool for the assessment of airway obstruction caused by inflammation and remodeling in horses with severe forms of EA, still requires a standardized evaluation and is currently rarely used [3,18]. Their application of this scoring system in 20 horses (controls, exacerbation and remission of heaves) disclose a correlation between histology and the degree of airway obstruction clinically measured, but does not seem to discriminate horses with heaves in remission from control [17].

Using a classical histologic assessment for evaluation of endoscopic biopsies, Niedzwieds et al., 2018 also found no significant difference for any of the histological variables between control and study group with asthma, and concludes hoping for the use of a standardized scoring system, such that developed by Bullone et al., 2016, to differentiate horses with an asthma exacerbation from healthy horses [17,18].

In our study, we chose to use the same scoring system to Bullone et al., 2016 for histologic evaluation of biopsy samples [17]. Conversely to previously results, we found significant differences in histological scoring between subjects in remission and healthy control horses. These results demonstrate how, similarly to what happens for the remodeling of the peripheral airways, the progression and maintenance of the disease are preserved histologically even in the remission phase, and an altered immunoregulatory function of the airway epithelium in horses with severe equine asthma syndrome.

The most common mediators of airway inflammation involved in horse respiratory diseases include histamine, bradykinin, prostaglandins, leukotrienes, platelet factor and endothelin-1 [17,40,41]. Furthermore, another mediator with a physiological and pathological role in respiratory function is a neuropeptide called neurokinin A (NKA), a member of the tachykinin family [42–44]. Neurokinin A is involved in nonadrenergic/noncolinergic neurotransmitter mediation of excitatory type in the respiratory tract of rat and guinea pig in normal conditions [45,46]. The NKA effect manifests through the activation of NK-2 receptors; they contract the smooth muscle of the airways leading to narrowing of the lumen. An abnormal expression of NK-2 receptors has been associated in human beings with forms of asthma [47,48].

On the one hand, the role of NKA in bronchoconstriction and neurogenic inflammation in patients with asthma generated scientific attention and specific investigations on selective NKA receptor antagonists for therapeutic uses [49]. On the other hand, Neurokinin B has been shown to be a much less potent contractile agen<sup>t</sup> in vitro than NKA in both control and EA horses, probably due to fewer receptors or less affinity [20]. However, little information exists regarding the involvement of NKA receptors in the horse, moreover limited to the intestine [50] and to the lung of horses affected by EA [20]. Recently, the receptors of the capsaicin sensitive-sensory nerves stimulated by bacterial of lipopolysaccharides, capable of inducing neurogenic inflammation [23] as occur in human airways [48] have been investigated in equine bronchial tissue.

The research conducted by Venugopal et al., 2009 on healthy and EA-affected horses through a comparative method demonstrates how significantly the expression of NK-2 has increased in the bronchial epithelium and in the bronchial smooth muscle and lung vessels of EA-affected horses [20]. The results of this study demonstrate how much NK-2 receptors are upregulated in EA, and sugges<sup>t</sup> that NK-2 receptor antagonists may have some therapeutic effect in controlling the progression of airway hyperreactivity. Based on the theory that both NKA and NKB can cause concentration-dependent contractions in horse bronchial rings, and that this response is more evident in horses affected by severe equine asthma, the Authors hypothesized that the increase in the expression of NK-2 receptors may be a major cause of airway hyperreactivity in EA-affected horses.

The ex vivo study by Calzetta et al., 2018 confirms and extends the previous results by Venugopal et al., 2009 and suggests that the environmental exposure to bacterial of lipopolysaccharides may represent a crucial factor in modulating the bronchial responsiveness of severe equine asthma [20,23]. A chronic exposure to high concentration of inhaled endotoxin in fact has a deleterious effect in equine distal airways: it triggers dysfunctional airway smooth muscle (ASM) contractility due to the stimulation of capsaicin-sensitive sensory nerves, increased the release of NKA and the activation of NK2 receptors. Antagonizing NK2 receptors may have a beneficial impact on normalizing the contractility and on controlling clinical signs in severe equine asthma.

Our results reveal a significant variability in the immunohistochemical expression of NKA between control horses and horses with severe equine asthma in remission. However, contrarily to what was expected, NKA immunoreactivity appears to increase significantly after corticosteroid treatment. This demonstrates to what extent the clinical improvement associated with therapy does not correspond to a decrease in this marker of inflammation, but rather show a contrary trend. Therefore, this result could sugges<sup>t</sup> that, unlike cortisone drugs, there are potential bases for an efficacy of selective and non-selective NK2 receptor treatments.

T lymphocytes play a fundamental role in modulating the immune response during the pathogenesis of severe equine respiratory syndrome. Data of the literature suggests that lung T helper may be implicated through the secretion of Th-1 or Th-2 type cytokines [19,37,38,51–54]. Horses affected by severe equine asthma syndrome produce both of these cytokines, depending on the stage of the disease and the time when sampling is performed [55]. In addition, the expression of cytokines in the airway lymphocytes also appears to be influenced by the length of time in which the horses have manifested the disease clinically [55].

The bronchiolar intraluminal accumulation of neutrophils usually accompanies the lymphocyte peribronchiolar infiltrate in severe EA anatomopathological findings [56], and takes place a few hours after environmental changes. The late phase of type I hypersensitivity plays a central role in a Th mixed response, mainly Th2 at the beginning and Th1 in the late phase of the process [35]. Both ways contribute to triggering inflammation and neutrophil recruitment in the airways by means of several cytokines, such as IL-8 [37,52,57]. In addition, a type III hypersensitivity reaction is able to partially explain the neutrophilic inflammation in the airways on EA-affected horses. However, the factors that initiate this process are still not fully explained [55].

Studies on horses with severe equine respiratory syndrome have revealed an increase in the gene expression of the pro-inflammatory cytokine IL-8 in bronchoalveolar cells [19,38,40,52,58,59], as well as an increase in IL-8 protein concentration in BALF [19,37,38,40,60–65].

It is well known that the main producers of IL-8 are the airway epithelial cells [27]. Ainsworth et al., 2006 demonstrated a 3- to 10-fold increase of IL-8 m-RNA in epithelial airway cells in severe equine asthma, and afterwards Berndt et al., 2007 confirmed how IL-8 mRNA expression in bronchial epithelial cells increases after exposure to stable dust, both in EA-affected subjects and in control animals [27,61]. These results were also confirmed

by another study [38], in which an increase in the level of IL-8 mRNA is observed both in BAL cells and in endobronchial biopsies during the crisis phases of the disease.

On the contrary, in the study by Pietra et al., 2011, although a significant overexpression of IL-8 and TNF α mRNA on BALF was observed in the group of treated horses, with a peak around the ninth day, a comparison between exposed and unexposed horses does not reveal significant differences [64]. According to the authors, this evidence could depend on the loss of uniformity of the sample, since the evolution of the pathology represents a continuum from healthy to pathological conditions.

Padoan et al., 2013, in a comparative research between clinical and endoscopic findings, cytological and cultural tests of BALF, histology of bronchial tissue and analysis of gene expression of inflammatory mediators in BALF and biopsies, found that six out of ten immunologically related genes (including IL-8) show a significant difference in expression between horses affected by severe equine asthma syndrome and the control group [40]. Specifically, however, IL-8 mRNA levels measured in biopsies are a hundred times lower than in BALF, and although an increase in its gene expression has been found in bronchial biopsies, this difference is not statistically relevant. Furthermore, the biopsies included in the study showed no significant differences in gene expression levels between EAaffected horses and controls. The lack of statistical significance between mRNA levels of inflammatory mediators in the respiratory epithelium could be caused by the small size of the samples or by the site where the biopsy is performed.

By using immunohistochemistry, Ainsworth et al., 2006 demonstrated IL-8 localization in the cytoplasm of epithelial cells in airway biopsies as well as in more recent times Tessier et al., 2017 observed a marked increase in IL-8 gene product determined by immunohistochemistry in bronchial cells in asthmatic, but not in non-asthmatic animals. In our study, we found significant differences between healthy and EA-affected in epithelial immunohistochemical expression, but none in IL-8 immunohistochemistry expression in the different stages of EA [5,37]. Unlike former suggestions in literature data on biomolecular BALF evaluation, however, this result appears to be aligned with the histological evaluations, which do not reveal any relevant difference in the inflammatory evaluation parameters in all phases.

Beyond the statistical significance of the results, the immunohistochemical observations reveal however interesting suggestions: the reactivity to NKA shows a trend to increase in intensity and to show same dysregulation characteristics (with the appearance of abnormal nuclear locations and variations in the mucosal distribution) in horses with severe equine syndrome compared to those of control. For IL-8, in horses with EA there is an intense positivity—often focal and localized to in single cells—which instead is not present in all the control subjects, where positivity is always widespread and of far weaker intensity. As far as these observations have diagnostic and/or immunopathogenic relevance for this disease in its various stages, it is not currently possible to determine it with certainty. Moreover, studies in this regard require a number of samples significantly higher and accompanied by biomolecular analyzes.
