Errata: Erratum Volume 26, Issue 2, 444, Article first published online: 20 March 2012
An abstract of this work was presented at the 27th Annual ACVIM Forum in Montreal, QC, Canada, June 2009.
Corresponding author: Jean-Pierre Lavoie, Département des Sciences Cliniques, Faculté de Médecine Vétérinaire, Université de Montréal, 3200 rue Sicotte, Saint-Hyacinthe J2S 7C6, QC, Canada; e-mail: firstname.lastname@example.org.
Background: Corticosteroids currently are the most effective pharmacological treatment available to control heaves in horses. Systemically administered corticosteroids have been shown to alter immune response in horses, humans, and other species. Aerosolized administration theoretically minimizes systemic adverse effects, but the effect of inhaled corticosteroids on immune function has not been evaluated in horses.
Objectives: To evaluate the effects of prolonged administration of inhaled fluticasone on the immune system of heaves-affected horses.
Animals: Heaves-affected horses were treated with inhaled fluticasone (n = 5) for 11 months or received environmental modifications only (n = 5).
Methods: Prospective analysis. Clinical parameters and CBC, lymphocyte subpopulations and function, and circulating neutrophil gene expression were sequentially measured. Primary and anamnestic immune responses also were evaluated by measuring antigen-specific antibodies in response to vaccination with bovine viral antigen and tetanus toxoid, respectively.
Results: No clinical adverse effects were observed and no differences in immune function were detected between treated and untreated horses.
Conclusions and Clinical Importance: The treatment of heaves-affected horses with inhaled fluticasone at therapeutic dosages for 11 months has no significant detectable effect on innate and adaptive (both humoral and cell-mediated) immune parameters studied. These results suggest that prolonged administration of fluticasone would not compromise the systemic immune response to pathogens nor vaccination in adult horses.
Heaves or recurrent airway obstruction is a common disease of horses stabled for extended periods. Susceptible horses develop lower airway obstruction and neutrophilic airway inflammation with inhalation of dust present in hay and bedding.1 Corticosteroids currently are the most effective pharmacological treatment available for the condition. However, systemic administration of corticosteroids to horses has been associated with several adverse effects, including adrenocortical suppression2,3 and dysfunction,4 laminitis,5,6 hepatopathy,6,7 muscle wasting,7 altered bone metabolism,8 and increased susceptibility to infection.9–11 Systemic corticosteroids also have been shown to affect the equine immune system by inducing transient peripheral neutrophilia and lymphopenia,12,13 changes in lymphocytes subpopulations and expression of activation markers,14 as well as by decreasing the antibody response to vaccination.15
Inhaled corticosteroids now are commonly used for the treatment of heaves in horses.16–20 This approach aims at achieving maximal concentrations of drug within the airways, while minimizing systemic adverse effects. In people, in whom inhaled corticosteroids are the first-line therapy for treatment of asthma and chronic obstructive pulmonary disease (COPD), prolonged administration of corticosteroids has been associated with adverse effects including reduction in bone density, cataracts, glaucoma, skin bruising,21 and modulation of cellular immunity at high doses.22 They also have been shown to induce a dose-dependent suppression of endogenous cortisol in horses2,23–25 suggesting that systemic effects may occur in this species. Thus, the objective of this study was to investigate the effects of long-term treatment of heaves-affected horses with inhaled fluticasone on innate and acquired (humoral and cell mediated) immune responses.
Materials and Methods
Ten heaves-affected mixed breed horses (6 mares and 4 geldings) aged 13–23 years (mean ± SD, 17.9 ± 2.9 years) weighing 410–535 kg (475 ± 38 kg) were used in this study. Horses were diagnosed with heaves on the basis of history, characteristic clinical presentation and worsening of clinical signs, pulmonary function, and bronchoalveolar lavage cytology after exposure to hay. All horses belonged to the research herd from the Respiratory Cellular and Molecular Biology Laboratory of the Université de Montréal, and were part of a larger study evaluating the reversibility of pulmonary remodeling. They were dewormed routinely and vaccinated annually (tetanus, West Nile, Eastern and Western equine encephalitis, equine herpes virus 1/4, influenza, and rabies vaccines). Horses were managed as a closed herd except for a few days in month 6 when they were hospitalized for thoracoscopy. Physical examination, CBC, and blood biochemistry profiles were performed before the study to exclude concomitant medical disorders. All experimental procedures were performed in accordance with the Canadian Council of Animal Care and approved by the University of Montréal Animal Care Committee.
Before initiation of the study, all horses were stabled and fed hay for 1–3 months to induce clinical exacerbation of respiratory disease. They then were assigned to 2 groups with similar severity of airway obstruction based on results of lung function tests (maximal variation in transpulmonary pressure [ΔPL], 44 ± 10 cmH2O in untreated group; 43 ± 10 cmH2O in fluticasone group). The 2 groups also were similar in terms of mean age and sex distribution. After a stabling period necessary to document the bronchoconstriction, horses in the untreated control group (n = 5, age, 17.4 ± 3.6 years; 3 mares, 2 geldings) were kept outside on pasture for the duration of the study and received no medications. They received a complete pelleted feed twice a day to maintain body condition (no hay) and had free access to grass in the summer. Horses in the fluticasone group (n = 5, age, 18.4 ± 2.2 years; 3 mares, 2 geldings) remained stabled and were fed hay for the first 5 months of the study, and then were kept on pasture with the untreated horses (under the same conditions) for the remaining 6 months of the study. They were administered inhaled fluticasone propionatea from a metered-dose inhaler via a commercially available mask,b at a starting dose of 2,000 μg twice a day. The dose then was adjusted as needed to keep the horses asymptomatic (from 2,000 μg q24h to 3,000 μg q12h), but, from the 6th month of study until the end, all horses in the fluticasone group received inhaled fluticasone at 2,000 μg once a day, between 7:00 and 9:00 am. While stabled, horses were turned out in a paddock 2–4 hours each day. The study was initiated in the springtime and finished in winter of the following year.
The timeline of the study is outlined in Figure 1. Horses were observed daily during the whole study. Blood samples for CBC were collected in ethylenediaminetetraacetic acid tubes, cell counts were performed with an automated analysis system,c and blood smears were reviewed by a clinical pathologist. Heparinized blood samples were processed within an hour for gene expression analysis and kept on ice overnight for lymphocyte phenotyping and proliferation assays. Vaccination was done with a tetanus toxoidd and the infectious bovine rhinotracheitis (IBR) vaccine, which also contained bovine viral diarrhea (type I and II), parainfluenza-3 and syncytial bovine virus antigens, combined with an appropriate adjuvant.e The experimental vaccination protocol was initiated 2 months after the horses received their last annual vaccine booster injection. Blood samples were collected before the 1st vaccine injection and then sequentially for 4 months. Serum was separated, aliquoted, and stored at −80°C until analysis. All blood samplings were performed by jugular venipuncture in the morning (before fluticazone administration in the treated group). As part of the larger study evaluating the reversibility of pulmonary remodeling, lung biopsies were performed under thoracoscopic guidance 2 weeks before the beginning of the study, at approximately 6 months, and after the last CBC sampling at the completion of this study. Blood samples were collected either before or at least 1 month after the surgeries to prevent interference with the results.
Lymphocyte Phenotyping, Activation State, and Proliferation Assays
Peripheral blood mononuclear cells were isolated with Ficoll gradient centrifugation as described previously.26 Isolated cells were analyzed by flow cytometry (fluorescent activated cell sorting [FACS]) for lymphocyte antigen markers (cluster of differentiation [CD]4 clone HB61A, CD8 clone HT14A, B cell clone cz2.1).27,28 Cell expression of major histocompatibility complex (MHC) class II molecules (clone cz11) and lymphocyte function-associated antigen-1 (LFA-1) (orCD11a/CD18, clone cz3.2) also were measured as markers of lymphocyte activation. The secondary stage used was fluorescein isothiocyanate-conjugated F(ab') fragment goat anti-mouse immunoglobulin G (IgG) (heavy + light chains) antibody. Samples were analyzed on a FACScalibur flow cytometer equipped with a 488 μm argon laser by Cell Quest Analysis software. Leukocyte subpopulations were displayed in a dot plot and gated according to size based on forward light scatter, and according to granularity based on 90° side light scatter.29 A region was placed around lymphocytes, and data were collected on 10,000 gated cells. Results indicate percent positive cells and mean fluorescence intensity in the lymphocyte-gated area.
For proliferation assays, lymphocytes were isolated by a combined carbonyl iron and Ficoll method, and labeled with carboxyl-fluorescein diacetate, succinimidyl ester (CFSE) as described before.26 Briefly, cells were resuspended in Roswell Park Memorial Institute (RPMI) cell culture medium with 10% fetal calf serum (FCS), antibiotics/antimycotics and mercaptoethanol solution. In another set of experiments, medium was prepared with 10% autologous serum instead of FCS. Cell suspensions were incubated with or without addition of pokeweed mitogen (PWM) or concanavalin A (ConA). Cells were harvested at 96 hours and tested for CFSE fluorescence with flow cytometry (FL1). A decrease in CFSE mean fluorescence was considered proportional to the division of fluorescence dye to daughter cells at each cell division. Results were analyzed using the percentage of cells with lower fluorescence than the control nonstimulated cells.
Gene Expression by Peripheral Blood Neutrophils
Neutrophil Isolation and Culture Conditions
Peripheral blood neutrophils were isolated using immunomagnetic selection (magnetic-activated cell sorting) as reported previously.30 Briefly, neutrophils were retrieved from the leukocyte-rich supernatant by sequential incubation with primary monoclonal antibodyf and secondary rat anti-mouse IgM antibody conjugated to paramagnetic microbeadsg before being loaded on a ferromagnetic LS separation column.g Cytospin slides were preparedh and stained with Protocol Hema 3i for differential counting of >400 cells to assess neutrophil purity. Viability was determined by trypan blue exclusion.
Purified neutrophils were suspended at 5 × 106 cells/mL in culture medium RPMI 1640j supplemented with 10% heat inactivated low-endotoxin FCS,j 4 mM L-glutamine,j 100 U/mL penicillin, and 100 μg/mL streptomycin.j The cells were incubated in 6-well suspension plates for 5 hours at 37°C, 5% CO2 in the presence of 100 ng/mL lipopolysaccharide (LPS) from Escherichia coli 0111:Bk and 10 nM fMLPk to induce proinflammatory cytokine gene expression.30 Unstimulated neutrophils were used as control (resting). At the end of incubation time, 106 cells per test were homogenized in TRIzol Reagentl and stored at −80°C until RNA extraction. Complementary DNA (cDNA) samples derived from LPS-stimulated neutrophils with or without dexamethasone (10−6 M)31 were used to assess the expression of glucocorticoid (GC)-responsive genes (GC receptor and glutamine synthetase).
RNA extraction, reverse-transcription (500 ng total RNA), and real-time PCR were performed as described previously32 with the Rotor-Gene Real-Time Centrifugal DNA Amplification System 3000.m Primers pairs were as follows (5′→3′): interleukin (IL)-8.S (sens) CTTTCTGCAGCTCTGTGTGAAG and IL-8.AS (anti-sens) GCAGACCTCAGCTCCGTTGAC; tumor necrosis factor-alpha (TNF-α).S CTTGTGCCTCAGCCTCTTCTCCTTC and TNF-α.AS TCTTGATGGCAGAGAGGAGGTTGAC; GC receptors (GCR).S TCATTAAGCTCCCCTGGCAGAGAA and GCR.AS ATTGAGAGTGAAACGGCCTTGGAC; glutamine synthetase (glut. Synt.).S ACTGGATTCCACGAAACCTCCAAC and glut. synt.AS GCTGCAAGTCTAGTCCGCTTAGTT; Ubiquitin.S TAGCAGTTTCTTCGTGTCCGT and Ubiquitin.AS TGTAATCGGAAAGAGTGCGG. GCR primers were designed to amplify the precursor (unspliced) transcript encoding both the α and β GCR isoforms, which are both down-regulated by GC treatment.33 All primers spanned exon-intron boundaries in order to prevent amplification of genomic DNA. Values were normalized using ubiquitin expression as reference gene.
Dosage of Serum-Specific IgG
Serum titers of tetanus toxoid antigen-specific IgGb were determined by use of ELISAs as described previously.34 Serum titers of IBR antigen-specific IgG were determined with a commercial ELISA test kit designed for the detection of IBR antibodies in bovine serum,n in which a secondary anti-equine monoclonal antibody (CVS3928) was substituted to the anti-bovine 1 provided in the kit.
By a repeated measures analysis of variance (ANOVA) and by a sequential Bonferroni's procedure to adjust comparison-wise α levels, peripheral blood leukocytes, lymphocytes, neutrophils, eosinophils, basophils, and monocytes were compared at each time points between treated and nontreated horses, and within each group values at each time point were compared with values at baseline. A repeated measures ANOVA was used to analyze log 10 transformed anti-tetanus and anti-IBR IgGb titers. Analysis was done by SAS v. 9.2.o
Unpaired Student's t tests were used to compare proportions of CD4+ and CD8+ peripheral blood T cells, CD4+/CD8+ ratio, expression of LFA-1 and CMH II on lymphocytes and monocytes, and proliferation of lymphocytes after stimulation between the 2 groups at 9 months. Analysis was done by Statview v. 5.0.o
For gene expression by peripheral blood neutrophils, data were log 10 transformed to normalize distribution. One-way ANOVA and Bonferroni's multiple comparison posthoc tests were used to assess the in vitro effect of dexamethasone on GC-sensitive genes in cDNA samples from stimulated neutrophils. Where gene expression by neutrophils isolated from fluticasone-treated and untreated horses was compared, 2-way repeated measures ANOVA followed by Bonferroni's posthoc comparisons was used to evaluate the effect of fluticasone treatment and cell stimulation. Analysis was done by Graphpad Prism 5 software.p
Inhaled fluticasone administration was well tolerated by all horses, and administration took approximately 4 minutes. Airway obstruction present at baseline was resolved for more than 2 months when immune function parameters were studied. During the 11 months of the study, 3 horses developed clinical disorders, 2 in the fluticasone group and 1 in the untreated group. Approximately 6 months after the beginning of the study, a fluticasone-treated gelding had a suspected episode of postthoracoscopy telogen effluvium diagnosed on the basis of clinical findings and histological examination of skin biopsies. Another treated horse developed a corneal ulcer believed to be traumatic, which resolved with topical antibiotic treatment. Finally, a horse in the untreated group developed a warm and painful swelling of the left front leg, associated with a moderate peripheral blood neutrophilia (12.5 × 109/L; reference range, 5.5–12.5 × 109/L) and hyperfibrinogenemia (5 g/L; reference range, 1–4 g/L) 5 months after the beginning of the study. The problem resolved with a 5-day course of penicillin and phenylbutazone. The latter episode began 2 weeks before the 6-month CBC and vaccination.
Differential White Blood Cell Count
All values remained within reference ranges at all time points (Fig 2). The only significant change observed was a decrease in monocyte count in untreated horses between baseline and 1 month (P= .0045; Fig 2D).
Peripheral Blood Lymphocyte Phenotyping, Activation State, and Proliferation Assays
No significant differences between fluticasone and untreated groups were observed in the proportions of CD4+ T cells (P= .45), CD8+ T cells (P= .89), and B cells (P= .83), and in the CD4+/CD8+ T-cell ratio (P= .68; Fig 3A–D). There was no significant difference in the proportion of lymphocytes expressing MHC class II molecule (90.5 ± 3.1 and 86.3 ± 6.4% in the fluticasone and untreated group, respectively, P= .22) and LFA-1 molecule (96.4 ± 0.7 and 95.1 ± 2.6%, P= .30). The mean fluorescence intensity also was comparable between the 2 groups for MHC class II (Fig 3E) and LFA-1 (Fig 3F).
The lymphocyte proliferation assays revealed no significant difference between fluticasone-treated and untreated horses with either PWM or Con A stimulation after 4 days of incubation in the presence of either FCS (Fig 3G and 3H) or autologous serum (results not shown).
Peripheral Blood Neutrophil Gene Expression
The purity and viability of isolated neutrophils was 98.96 ± 0.74 and 98.44 ± 1.15% (mean ± SD), respectively. An 8-month treatment period with fluticasone did not cause differences in IL-8 and TNFα mRNA expression in resting and stimulated neutrophils (Fig 4A), nor did it decrease GC receptor mRNA expression or up-regulate glutamine synthetase in resting and stimulated neutrophils when compared with untreated horses (Fig 4B). Conversely, significantly increased glutamine synthetase (P < .05) and decreased GCRs expression (P < .01) was quantified in cDNA samples from in vitro dexamethasone-treated neutrophils (not shown, n = 3), performed as a positive control.
Serum Anti-Tetanus IgGb Titers
Detectable serum anti-tetanus toxoid IgGb titers were present in all horses before vaccination (Fig 5). Booster vaccination at 6 months (Time 0 on Fig 5) resulted in development of an antigen-specific IgGb response (P < .0001) similar in both group of horses (P= .66).
Serum Anti-IBR IgG Titers
As expected, all horses had negative IBR titers before vaccination. Vaccination with a bovine multivalent vaccine including IBR antigens at 6 and 7 months resulted in development of anti-IBR IgG (P < .0001) of a similar magnitude (P= .77) in both groups of horses (Fig 6).
Prolonged administration of corticosteroids may be required to control airway obstruction in heaves-affected horses when appropriate environmental dust control is not implemented. Because of the adverse effects that have been observed with the oral or injectable use of corticosteroids, treatment duration is usually short, from days to weeks. Although better tolerated than when systemically administered in human patients, inhaled corticosteroids have nevertheless been associated with adverse effects, including a decrease in the immune response22 and increased susceptibility to infection when administered over extended periods.35 The risk of infection was increased especially in elderly patients and in those with severe airway obstruction, 2 key features of heaves in horses. In the present study, no adverse alterations of the immune system or clinical adverse effects were observed over an 11-month period of administration of fluticasone propionate in horses with heaves.
Innate immunity is responsible for the initial response to infectious agents, and it was evaluated here by measuring peripheral blood leukocyte count and neutrophil transcriptional response to bacterial products ex vivo (refer to Fig 1 for measurement timeline). No significant alterations in these parameters were observed with fluticasone treatment. Neutrophil and monocyte counts remained within reference ranges at all time points and did not significantly vary over time in the fluticasone group. These results are in contrast with the transient peripheral blood neutrophilia observed after a single systemic administration of corticosteroids.12–14,36 However, and in agreement with our findings, inhaled beclomethasone did not alter differential white cell count over a 22-month treatment course in asthmatic children.37 The decrease in peripheral monocyte count in the untreated group at 1 month was unexplained.
Because factors other than cell number can affect innate immune function, we also investigated possible down regulation of proinflammatory cytokine production by neutrophils after 8 months of treatment. We found no differences in IL-8 and TNFα mRNA baseline expression, and observed an appropriate response with LPS and fMLP stimulation in vitro. These results contrast with the inhibition of equine neutrophil respiratory burst and LPS-induced TNFα and IL-8 gene transcription by dexamethasone observed in vitro.31 Accordingly, neither of the GC-responsive genes assessed (GC receptors and glutamine synthetase) were altered in peripheral blood neutrophils from fluticasone-treated horses in this study. These results support minimal if any exposure of circulating neutrophils to fluticasone and any active metabolite.
To evaluate a possible alteration of the acquired immune system, we first studied the lymphocyte count at 0, 1, 6, and 11 months. In contrast with the transient lymphopenia observed in horses after a single systemic administration of corticosteroids,12–14,36 no significant changes in lymphocyte counts were observed with fluticasone in the present study. Similar to our findings, no change in lymphocyte counts was observed in beclomethasone-treated children.37 We further evaluated the peripheral blood lymphocyte subpopulation distribution (CD4+ and CD8+ T cells, and B cells) and the expression of cell surface molecules (MHC class II and LFA-1) after 9 months of fluticasone inhalation, and found no changes in those parameters. Equine lymphocytes constitutively express the MHC class II and the integrin LFA-1 molecules. The function of MHC class II in lymphocytes is unknown, but expression levels have been associated with lymphocyte maturation.38 The expression of integrins is up-regulated when cells are activated, and they facilitate cell-to-cell interaction for costimulation. Our results contrast with those reported in horses and human subjects after systemic administration of corticosteroids. In horses, for example, a decrease in total lymphocyte count and CD4+/CD8+ ratio and an increase in the expression of LFA-1 in leukocytes was observed for 48 hours after administration of a single 0.025 mg/kg IV dose of dexamethasone.36 Similarly, in human subjects, prednisolone and dexamethasone (PO or IV) cause a decrease in total lymphocyte, T (CD4+ and CD8+), and B cell counts, with the CD4+ T cell distribution being more severely affected.39,40 Studies on the effects of administration of inhaled corticosteroids on cell-mediated immunity in human and animal subjects led to conflicting results. A 22-month period inhaled beclomethasone in asthmatic children did not alter lymphocyte subpopulation distributions37 nor did several weeks of treatment with fluticasone in healthy dogs,41 or flunisolide in healthy and asthmatic cats.42 Contrary to the effects observed after administration of a single dose43,44 and our results, inhaled fluticasone administered for 4 weeks to healthy volunteers caused a decrease in activated CD4+ and CD8+ T cells.22 Duration of treatment, relatively higher dosage in the study in humans, and health status of the subjects could explain these differences. We further measured the proliferative capacity of lymphocyte in response to mitogens. Proliferation was similar in treated and untreated horses at 9 months, whether lymphocytes were exposed to ConA (T-cell specific) or to PWM (B and T cells). This response was independent of the presence of autologous horse serum, which could have promoted a more sustainable effect comparable to the in vivo condition. These results are in agreement with studies in human and animal subjects after inhaled or systemic corticosteroid administration.13,14,22,37 Only prednisolone PO has been shown to induce a transient (<24 hours) decrease in phytohemaglutinin-induced lymphocyte proliferation.39
Humoral immunity was investigated by measuring the primary response to an unknown antigen to horses (IBR antigen) and the anamnestic response to tetanus toxoid. Both groups responded with a similar increase in titers against these 2 antigens, despite 6 months of treatment with fluticasone. This finding contrasts with the almost complete abrogation of IgGa and IgGb response to a bovine viral vaccine observed with dexamethasone (0.2 mg/kg IM, twice a week) in horses.15 However, these results are in agreement with the normal IgG vaccinal response of COPD and asthma patients when treated with inhaled corticosteroids,45–47 and after the administration of a single dose of dexamethasone to healthy horses.36
To the authors' knowledge, no direct effect of the horse's environment on its systemic immune system has been established. However, from the 6th month of the study, all horses were out in paddocks or pastures. Thus, neutrophil gene expression measurement and lymphocyte function tests were performed when both groups of horses had been out in pasture for 2 and 5 months, respectively. These time points were chosen in order to prevent a possible effect of variable environments on the parameters studied.
In summary, this study shows that long-term treatment of heaves-affected horses with inhaled fluticasone at the therapeutic dosage has no detectable effect on the innate and acquired humoral and cell-mediated-immune parameters studied. These results indicate that this treatment would not preclude the use of vaccines in heaves-affected horses.
aFlovent, Glaxo Wellcome, Mississauga, ON, Canada
bEquine Aeromask, Trudell Medical International, London, ON, Canada
cAdvia 120 Hematology System, Siemens Healthcare Diagnostics, Deerfield, IL
dTetanus Toxoid, Serial No.: 1630104B, Fort Dodge Wyeth Animal Health, IA
eTriangle 4 + Type II BVD, Serial No.: 178191A, Fort Dodge Wyeth Animal Health
f#DH24A, VMRD Inc, Pullman, WA
gMACS, Miltenyi Biotec, Auburn, CA
hCytospin 2, Shandon, Southern Instruments, Sewickley, PA
iFisher Scientific, Ottawa, ON, Canada
jGIBCO, Invitrogen, Burlington, ON, Canada
kSigma-Aldrich, Oakville, ON, Canada
mCorbett Research, Montreal Biotech, Montreal, QC, Canada
nInfectious Bovine Rhinotracheitis (IBR-Ab) Svanovir, Savnova Biotech Ab, Uppsala, Sweden
oSAS Institute Inc, Cary, NC
pGraphPad Software Inc, La Jolla, CA
The authors thank Josiane Lefebvre-Lavoie for technical assistance.
The study was supported by a grant from the Canadian Institutes of Health Research (208855), Allergen NCE, le Fonds du Centenaire, and the Clinical Research Pfizer Fund.