Bronchial angiogenesis in horses with severe asthma and its response to corticosteroids

Abstract Background Severe asthma in horses is characterized by structural changes that thicken the lower airway wall, a change that is only partially reversible by current treatments. Increased vascularization contributes to the thickening of the bronchial wall in humans with asthma and is considered a potential new therapeutic target. Objective To determine the presence of angiogenesis in the bronchi of severely asthmatic horses, and if present, to evaluate its reversibility by treatment with corticosteroids. Animals Study 1: Bronchial samples from asthmatic horses in exacerbation (7), in remission (7), and aged‐matched healthy horses. Study 2: Endobronchial biopsy samples from asthmatic horses in exacerbation (6) and healthy horses (6) before and after treatment with dexamethasone. Methods Blinded, randomized controlled study. Immunohistochemistry was performed using collagen IV as a marker for vascular basement membranes. Number of vessels, vascular area, and mean vessel size in the bronchial lamina propria were measured by histomorphometry. Reversibility of vascular changes in Study 2 was assessed after 2 weeks of treatment with dexamethasone. Results The number of vessels and vascular area were increased in the airway walls of asthmatic horses in exacerbation (P = .01 and P = .02, respectively) and in remission (P = .02 and P = .04, respectively) when compared to controls. In Study 2, the differences observed between groups disappeared after 2 weeks of treatment with corticosteroids because of the increased number of vessels in healthy horses. Conclusions and Clinical Importance Angiogenesis contributes to thickening of the airway wall in asthmatic horses and was not reversed by a 2‐week treatment with corticosteroids.


| INTRODUCTION
Asthma in horses is an umbrella term encompassing various chronic inflammatory processes affecting the lower airways. 1,2 This syndrome is considered one of the most common conditions affecting adult horses. 3 Severe asthma in horses is associated with excessive mucus accumulation, airway hyperresponsiveness, and structural changes of the airway wall (remodeling) leading to airflow obstruction and poor performance. 4 This remodeling develops in response to chronic inflammation and includes epithelial hyperplasia, collagen deposition, and smooth muscle hypertrophy and hyperplasia. 5,6 These structural changes contribute to the impairment of lung function observed in asthmatic horses. [7][8][9] In asthmatic human patients, an association between chronic inflammation of the airways and the development of angiogenesis, a process defined as the formation of new vessels from preexisting ones, has been identified. 10,11 Moreover, a study reported increased airway mucosal blood flow in the airways of asthmatic humans when compared to normal subjects. 12 Similarities between asthma in humans and horses indicate that angiogenesis also may be present in affected horses. 5 Importantly, airway remodeling in severe asthma of horses is only partially reversible with current treatments (eg, inhaled and systemic corticosteroids, bronchodilators). 13,14 The stimulus leading to angiogenesis in asthmatic humans is unclear. 15 However, angiogenesis and chronic inflammation have been shown to be codependent. 16 Furthermore, hypoxic pulmonary tissues release angiogenic growth factors such as vascular endothelial growth factor (VEGF) that contribute to the migration of endothelial cells from preexisting vessels to the surrounding extracellular matrix (ECM) to form solid sprouts. Inflammatory cells such as macrophages, mast cells, and fibroblasts also can lead to an increase in vessel growth, 17 which may contribute to thickening of the bronchial wall. An in vitro study suggested that corticosteroids (CS) can reverse angiogenesis in horses, because they decrease the release of VEGF from equine pulmonary artery endothelial cells. 18 Although several studies have identified the presence of angiogenesis in humans with asthma, [19][20][21][22] with an increased number of subepithelial vessels and vascular area in lung tissue, 11,23 literature on lung angiogenesis in asthmatic horses is scarce. A 2012 study 24 showed an increased vascular density of superficial tracheal vessels in asthmatic horses, when evaluated by narrow-band imaging. However, no differences were observed at the level of the bronchi, conflicting with results obtained in humans. The authors attributed those findings to a lack of sensitivity of the imaging technique to assess vascular changes.
Because inflammation and hypoxemia are common findings in horses with severe asthma, our objectives were to evaluate the presence of bronchial angiogenesis using immunohistochemistry to quantitatively determine the presence of mature bronchial vessels in the airways of asthmatic and age-matched healthy horses, and to evaluate the angiogenic response to systemically administered CS. We hypothesized that an increased number of vessels would be observed in the airways of asthmatic horses when compared to controls, and that this process would be reversible by treatment with CS.   27 All horses were stabled in a barn in individual stalls with wood shavings, except for 4 horses in remission that were at pasture. Stabled horses were fed hay twice daily whereas those on pasture were not supplemented hay with the exception of 1 horse that also received hay pellets. Turnout was allowed daily for all of the stabled horses. Horses were maintained in remission status for a minimum of 3 months either by decreasing dust exposure by turning them out into pasture (n = 4), or by the administration of inhaled CS (n = 3, fluticasone 2500 μg/q12h for 3-4 months) before being sampled.

| Study 2
The experimental protocol and characteristics of the horses studied have been reported previously. 28 In brief, subsegmental

| Immunohistochemistry
Collagen IV, a major basement membrane protein, was used to outline the endothelial basement membrane. 10 Two complete airways (Study 1) and 2 endobronchial biopsy samples (EBB-Study 2) were studied for each horse. Equine bronchi specimens initially snapfrozen in liquid nitrogen and maintained at À80 C were thawed in phosphate-buffered saline (PBS 1Â) at room temperature and cut into 0.5 cm samples. Tissue samples then were fixed for 24 hours in 4% formaldehyde and embedded in paraffin blocks until analyzed.
All of the blocks were stored at room temperature before use. Histologic sections of 5 μm thickness were used for immunohistochemistry. Enzymatic antigen retrieval was performed using pepsin

| Histomorphometric analysis
Histologic sections were evaluated on a 20Â magnification and digitized using an automated upright microscope system with a FLIR camera (Life Science Research, Leica DM4000-B, Buffalo Grove, IL, USA); (FLIR Integrated Imaging Solutions, Inc, Richmond, BC, Canada) and specialized software for image acquisition (Panoptiq 5.0, ViewsIQ, Inc, Vancouver, BC, Canada). The area between the epithelial basement membrane and a maximum distance of 150 μm in the lamina propria was analyzed in each region of interest (ROI), as previously described. 11,23 Six ROI with an area of 1 mm 2 were F I G U R E 1 Representative image of a bronchus from an asthmatic horse (A) and a control horse (B) stained with collagen IV highlighting in brown the basement membrane of blood vessels in the airway lamina propria. The smooth muscle cells are also diffusely stained by collagen IV. BM, basement membrane; ECM, extracellular matrix; EP, epithelium; SM, smooth muscle. Scale bar 100 μm F I G U R E 2 Example of the tracings for the calculation of the vascular area and the number of vessels in the airway lamina propria. A white line was firstly used to calculate basement membrane length (black arrows), and secondly, it was used to outline the extracellular matrix area (red arrows). Luminal circumference of each vessel included in a region of interest is outlined in red (vascular area). Each vessel was manually assigned a number for vessel count. Scale bar 100 μm F I G U R E 3 (A) Pulmonary resistance (R L ) and (B) pulmonary elastance (E L ) measured in control (n = 7) and asthmatic horses in exacerbation (n = 7) and remission (n = 7). The error bars indicate the standard deviation of the mean. Significantly higher R L and E L values were observed in horses in exacerbation when compared with control (P < .0001, P = .0008, respectively) and remission (P < .0001, P = .0012, respectively) F I G U R E 4 Histomorphometric results of Study 1. A, Total number of bronchial vessels per μm 2 of extracellular matrix (ECM) in control and asthmatic horses in exacerbation and remission. An increased number of vessels in horses in exacerbation (P = .007) and remission (P = .02) when compared to control horses was observed. B, Vascular area of control and asthmatic horses in exacerbation and remission. An increased vascular area was observed in horses in exacerbation when compared to the control (P = .02) and remission horses (P = .04). C, Total epithelial basement membrane length values of control and asthmatic horses in exacerbation and remission. No differences were observed between the 3 groups. D, Mean bronchial vessel size of control and asthmatic horses in exacerbation and remission. No differences were observed between the 3 groups. The error bars indicate the standard deviation of the mean randomly selected using number randomization software (https:// www.random.org) to analyze blood vessels as described in previous quantitative studies. 23

| Statistical analysis
Heterogeneity among groups for age and weight was analyzed by 1-way analysis of variance (ANOVA) and with the exact Chi-squared test for sex. For the first study, a 1-way ANOVA was used to examine differences among the 3 groups for R L and E L and histomorphometric traits. This analysis was followed by contrasts between pairs of means, adjusting the alpha level downward using the Benjamini-Hochberg F I G U R E 5 Histomorphometric results of Study 2. A, Total number of bronchial vessels per μm of epithelial basement membrane in control and asthmatic horses in exacerbation. A statistically significant increase in the number of vessels was observed in horses in exacerbation (P = .002) when compared to control horses at baseline. No significant differences were observed between control and exacerbation groups (P = .35) after 2 weeks of treatment because of an increase in the number of vessels in control horses (P = .01) after 2 weeks of treatment with dexamethasone. The number of vessels of horses in exacerbation remained unchanged. B, Mean bronchial vessel size of control and asthmatic horses in exacerbation. No significant differences were observed between groups at baseline (P = .1) or at week 2 (P = .24). C, Total epithelial basement membrane length values of control and asthmatic horses in exacerbation. No significant differences were observed between groups at baseline (P = .3) or at week 2 (P = .12). The error bars indicate the standard deviation of the mean procedure. Values obtained in the second phase of the study were compared between 2 groups, control and exacerbation, using a 2-sample t test. Values from the different groups (Study 2) at each time point were compared using a 2-way repeated-measures ANOVA. All statistical calculations were performed using computer software (GraphPad Prism 7.0, GraphPad Software, Inc, La Jolla, California). Statistical significance was set at P ≤ .05.

| Animals
Sixteen mares and 5 geldings were studied, with a weight (mean ± SD) of 504 ± 74.5 kg and age of 16.8 ± 5.1 years. No association was observed between sex and group (P = .3). Likewise, mean age (P = .14) and mean weight (P = .35) did not differ significantly among diagnosis groups (exacerbation, remission, control; P > .05).

| Lung function
As expected because they were inclusion criteria, significantly higher R L and E L were observed in asthmatic horses in exacerbation when compared with controls (P < .01) and remission (P < .01; Figure 3A,B).

| Histomorphometric analysis
The number of vessels corrected by the surface area of ECM was significantly increased in asthmatic horses in either exacerbation (P = .01; mean difference, À140 vessels/mm 2 ) or remission (P = .02; mean difference, À120 vessels/mm 2 ) when compared to control horses ( Figure 4A). Furthermore, vascular area (measured by tracing the area internal to the vascular endothelial membrane of each vessel) was significantly increased in asthmatic horses in exacerbation when compared to controls (P = .02; mean difference, À1180 μm 2 ) or to horses in remission (P = .04; mean difference, +747.3 μm 2 ; Figure 4B). No difference was observed among groups in epithelial basement membrane length (P = .86; Figure 4C) or mean vessel size (vascular area divided by number of vessels; P = .4; Figure 4D).

| Animals
Six mares and 6 geldings were evaluated, with a weight of 493.8 ± 51.6 kg and age of 14.3 ± 6.2 years. Additional details for Study 2 are reported elsewhere. 28

| Lung function
As previously described, 28 the significant differences in R L observed at baseline between horses in exacerbation and healthy controls were lost after 2 weeks of treatment with dexamethasone.

| Histomorphometric analysis
An increased number of vessels was observed in asthmatic horses in exacerbation (P = .002; 95% confidence interval [CI], 0.01 to 0.03) in comparison to control horses at baseline. This difference was lost after 2 weeks of treatment with CS (P = .35; 95% CI, À0.01 to 0.02; Figure 5A).   19 Angiogenesis in turn can contribute to maintaining or amplifying inflammation in diseases such as asthma by transporting cells, oxygen, and nutrients to the site of inflammation. 16,31 Our study provides the first evidence of angiogenesis occurring in bronchial tissues of the central airways in asthmatic horses. The increased number of vessels in asthmatic horses, either in exacerbation or remission, when compared to control horses, and the increased vascular area in horses in exacerbation are in agreement with observations in children and adults with mild to moderate asthma. 11,23,32 Both groups of asthmatic horses had an increased number of vessels, but increased vascular area was observed only in horses in exacerbation when compared to controls. This absence of difference between the remission and control groups could be a consequence of these vessels being smaller in size although mean vessel size was not statistically different among the 3 groups or also could be related to the administration of inhaled CS for a minimum of 3 months before sample collection in 3 of the horses in the remission group. In addition, mean vessel size is considered to be an estimate, because correction for vessel dilatation was not performed because of the difficulty of measuring the diameter of the vessels in a large percentage of the vessels (because of folding and elongation of vessels), which could have contributed to the differences observed. Epithelial basement membrane length was used as a control variable to ensure that both groups had similar areas analyzed. This was confirmed by the absence of differences between groups in both Study 1 and Study 2. Because of the differences in morphology between the bronchial samples used in Study 1 (main bronchi) and Study 2 (endobronchial biopsy samples), the number of vessels in the first study was corrected by the area of ECM analyzed, whereas in Study 2 this variable was corrected by the epithelial basement membrane length.
Conventional treatment for asthma in horses and humans includes antigen avoidance and administration of CS and bronchodilators. [33][34][35] However, these treatments only partially reverse the airway remodeling observed in asthmatic horses. 13,36 Contrary to what we had hypothesized, the number of vessels in asthmatic horses did not decrease after treatment. Furthermore, an increased number of vessels was observed in control horses after 2 weeks of treatment with dexamethasone. Studies in human asthmatics have shown decreased angiogenesis with CS. [37][38][39] However, these apparent discrepancies may be a result of species differences. Indeed, a recent study comparing the responses of mice and horses to glucocorticoids in an in vitro aortic ring model of angiogenesis suggested an upregulation of proangiogenic pathways in the horses, whereas it was inhibited in mice. 40 Vessel outgrowth was evident after 3 days and was maintained for 1 week. Furthermore, upregulation of proangiogenic growth factor VEGF signaling was observed in equine vessels exposed to cortisol. 40 This finding might explain why in our study, healthy horses had an increase in the number of vessels with CS. Nevertheless, these results contrast with another study in which dexamethasone inhibited the expression of VEGF by equine pulmonary artery endothelial cells stimulated with interleukin-4, 18 a cytokine shown to be upregulated in horses with asthma, 41 suggesting that the angiogenic response to CS also may be agonist dependent.
Furthermore, the increased number of vessels after treatment was seen only in healthy horses rather than asthmatic horses. This could be a result of bronchial vessels already being augmented in comparison to baseline in horses with asthma, which may suggest that the capability of vessels to undergo angiogenesis might have reached a plateau or that angiogenesis is not decreased but only controlled by CS administration.
Immunohistochemistry allows identification and localization of molecules of interest by the immunolabeling of specific antigenantibody reactions. 42,43 Although manipulations and assessment of results can be challenging using this technique, 44,45 it is considered the gold standard method when quantifying bronchial vasculature. 10 In our study, the basement membrane protein, collagen IV, was used as a biomarker for evaluating the presence of angiogenesis. 10,11 This antibody highlights the vessel by staining the basement membrane supporting the endothelium. 11 This marker is considered superior to endothelial biomarkers such as platelet endothelial cell adhesion molecule-1 (PECAM-1) and factor VII, 20,47 because it is considered selective for the vascular basement membrane. Lymphatic vessels also might be positively stained depending on the epitopes of collagen IV targeted by the antibody. However, the staining likely would be weak in lymphatics because of their incomplete basement membrane and absence of pericytes. 46 Furthermore, the antibody used in our study was reported to be specific for blood vessels in humans. 47 A limitation of our study was the absence of milder clinical cases of asthma, and therefore results may not be directly applicable to horses with milder disease. Furthermore, our study was performed on a small number of animals. The strict experimentally controlled environmental conditions under which horses with severe asthma and control horses were maintained during the study necessitated limiting the number of animals required to reach statistical differences. Nevertheless, the study may not have had the power to detect smaller changes in some of the variables that were evaluated.
In conclusion, our study determined that in addition to pulmonary structural changes, an increase in vascularity occurs in horses with severe asthma. These results may offer a new opportunity for treatment by targeting angiogenesis and possibly decreasing inflammation and remodeling of the asthmatic airways. Furthermore, treatment with CS increased angiogenesis in healthy horses, whereas angiogenesis was maintained in asthmatic horses suggesting that conventional treatment does not decrease the number of bronchial vessels observed in the ECM. Lastly, the proangiogenic effects of CS observed in our study may have implications for diseases of horses associated with increased cortisol secretion or when CS are administered chronically.

CONFLICT OF INTEREST DECLARATION
Authors declare no conflict of interest.

OFF-LABEL ANTIMICROBIAL DECLARATION
Authors declare no off-label use of antimicrobials.

HUMAN ETHICS APPROVAL DECLARATION
Authors declare human ethics approval was not needed for this study.