Nebulized perflubron and carbon dioxide rapidly dilate constricted airways in an ovine model of allergic asthma

Female merino-cross farm-reared sheep aged 6 months were used in the experiments described here. After the sensitisation procedures (described below), sheep were transported to Animal House Facility at Veterinary Science, Parkville, Australia where they were kept in indoor pens for all the procedures described in this report. On arrival at the Animal House Facility all sheep were treated with an antihelminthic to ensure that they were parasite-free.

All experimental protocols used in this study were approved by the Animal Experimentation Ethics Committee at the University of Melbourne.

Acetyl-β-methylcholine chloride (MCh) was from Sigma-Aldrich, Stenhein, Germany and salbutamol sulphate solution was purchased from GSK Pty Ltd Australia. Gas mixture (12% CO₂ + 21% O₂+ balance N₂) was from CIG Pty Ltd Australia.

Characterization of perflubron delivered by the Sidestream nebulizer was conducted under cGMP conditions at SolAero Ltd laboratory. Methods to characterize perflubron aerosol were adapted from regulatory methods embodied in the European Nebulizer Standard [40]. The unique nature of perflubron required adaptation of methods including incorporation of trace quantities (5%) of a perflubron compatible non-volatile PFC which could be assayed gravimetrically to determine both aerosol output as well as particle size distributions. Aerosol output was determined using a breath simulation (500 mL at 15 cycles per minute) and collecting released aerosol representing the `inhaled aerosol’ on a pre-weighed 3M electrostatic filter, subsequently allowing volatile perflubron to dissipate and quantitatively weighing the non-volatile PFC component. The residual weight gain was then used to back calculate the original amount of deposited aerosol. Pre- and Post- weights of nebulizers provided the gross perflubron emitted, from which the analyzed aerosol proportion could be subtracted to estimate the proportion of perflubron in the vapor phase. Aerosol size was determined using a Marple 298X cascade impactor equipped with pre-weighed GF/A filters from which the non-volatile aerosol residue was similarly gravimetrically determined to quantify size fractions.

Sheep 6 months of age were subcutaneously sensitized with 50 μg house dust mite (HDM) extract (Dermatophagoides pteronyssinus; CSL Ltd, Parkville, VIC, Australia) mixed 1:1 in aluminum hydroxide [Al(OH)₃] adjuvant. Injections of 1 mL HDM / Al(OH)₃ were given on three occasions at 2 week intervals [38]. Levels of HDM-specific IgE (IgE-HDM) were determined by enzyme-linked immunosorbent assay (ELISA) of serum samples obtained from animals immediately prior to, and 7 days after, HDM sensitization [38],[41]. Sheep were only selected for use in the study if they had increased levels of HDM-specific IgE (IgE-HDM) ≥2-fold above pre-sensitization values.

To induce pulmonary sensitization to HDM, all sheep were given three inhalation challenges of aerosolised HDM (1 mg/ml in 5 mL of phosphate buffered saline). Lung function was recorded for one hour directly after the aerosol HDM challenge. HDM aerosols were generated with a medical nebulizer (Vitalair RapidNeb, Allersearch, Vermont, Australia). The nebulizer was connected to a dosimeter system consisting of a solenoid valve and a source of compressed air (20 psi). The output of the nebulizers was directed into a T-piece with one arm connected to the inspiratory port of a mechanical ventilator (Intermed Bear 2 – Adult Volume Ventilator, Bear Medical Systems, Inc., Riverside, CA). HDM challenges were delivered at a tidal volume of 500 mL and a rate of 20 breaths per minute for 10 minutes.

Data on lung function were obtained from all sheep using established techniques [42]. Briefly, pressures and flow were acquired from conscious unsedated sheep using a National Instruments Data Acquisition (DAQ) system with a Virtual Instrument programmed with Lab View® (National Instruments, Austin, USA). The sheep were restrained in a body sheath and head restraining harness tethered to a custom-made animal trolley. To measure oesophageal pressures, a balloon pressure catheter was advanced through one nostril into the lower esophagus. To assess tracheal pressures throughout the respiratory cycle a cuffed endotracheal tube was placed into the trachea and then a side-hole catheter was advanced 2 cm past the distal end of the endotracheal tube. The oesophageal and tracheal catheters were connected to differential transducers (Gaeltec 8 T, Gaeltec Ltd, Dunvegan, Scotland) to measure oesophageal pressure (P_oes) and airway pressures (P_aw) respectively. Transpulmonary pressure (P_l) was obtained from P_oes with respect to P_aw. Airflow was measured via a pneumotachograph (Hans Rudolph Inc, Kansas City, USA) attached to the proximal end of the endotracheal tube. Calibration and synchronization of the gas transducers and pneumotachograph were performed as described previously [43],[44]. Lung dynamics were determined by a modified Mead-Whittenberger method [45]. Breath by breath analysis of flow and pleural pressure (P_pl) was used to determine lung resistance (R_L) from an average of five breaths. Data from breaths which differed by more than 20% in volume and more than 10% of the breathing rate were rejected. Acquisition and detailed analysis of pulmonary physiological data were performed on customized Lab View software developed by Koumoundouros et al. [42].

Treatments with nebulized S-1226, nebulized perflubron and nebulized salbutamol were generated with a Sidestream® nebulizer (Philips Respironics). The nebulizer was connected to a dosimeter system consisting of a solenoid valve and a source of compressed (20 psi driving pressure) gas mixture (12% CO₂, 21% O₂ and 67% N₂) for S-1226 or compressed medical air (21% O₂ and 79% N₂) for perflubron and salbutamol. Treatments were delivered at a tidal volume of 500 mL and a rate of 20 breaths per minute for 2 minutes. Treatments with CO₂ alone and medical air were performed following the same protocol but with an empty nebulizer.

Sheep (previously HDM sensitized) were challenged with HDM aerosols to induce early phase asthmatic responses. At the peak of these responses (characterised by an increase in lung resistance), the sheep were treated for two minutes with one of six treatment groups listed below. Lung resistance was monitored up to 20 minutes following cessation of treatment.

1- S-1226: Nebulized perfluorooctylbromide (perflubron) generated using a Sidestream® nebulizer (Philips Respironics) and driven by a compressed gas mixture containing 12% CO₂, 21% O₂ and balance N₂.

2- CO₂ alone: Inhalation of a gas mixture containing 12% CO₂, 21% O₂ and balance N₂.

3- Perflubron: Nebulized perflubron driven by compressed medical air.

4- Salbutamol (1 mg/ml): Nebulized salbutamol driven by compressed medical air.

5- Medical air: Inhalation of medical air

6- No treatment: room air

An unsensitised sheep was used in separate experiments to visually assess the airway response to inhaled S-1226. Real time video recordings were taken with a bronchoscope digital camera following methacholine challenge. To insure that the same airway was recorded throughout the procedure, the upper region of the airway was tattooed by injecting tissue dye using a transbronchial needle. The airway was then pre-contracted with an aerosol of methacholine (MCh) delivered via a Trudell Ltd Pty bronchoscope catheter, inserted into the biopsy port of the bronchoscope. Doubling doses of methacholine were nebulized directly into the lung segment (0.009, 0.018, 0.037 w/v) for 30 seconds with a 2 minute interval between methacholine doses until the airway was maximally constricted. The airway was then treated with nebulized S-1226 delivered through a Trudell Pty Ltd bronchoscope catheter and the effects video recorded in real time. If the airway re-constricted, second and third treatments were administered.

In each experiment several intervals were analysed: Baseline, HDM-induced early phase response, and three post-treatment periods (immediately, 1–10 minute and 10 to 20 minutes after treatment cessation). Means were calculated for each interval and treatment effect was expressed as percent decline of lung resistance.

Analysis of variance (ANOVA) was performed over each time interval to determine if any of the treatment groups performed better than any other. A two tailed Dunnett’s post-hoc test was performed. This involved separate analyses addressing two distinct questions. The first: is S-1226 superior to no treatment, medical air (negative control) or salbutamol (positive control). Secondly to determine which component of S-1226, CO₂ or perflubron contributed greater to its efficacy. Tests were performed using GraphPad Prism version 5.00 and R version 3.1.1 for windows. A p-value ≤ 0.05 was considered significant.

Results from characterization of nebulized perflubron are summarized in Table 1. The mean rate of perflubron delivered was 598 mg/minute, of which some 203 mg/minute consisted of aerosol with the remaining 395 mg/minute was estimated to be in vapor form - thus the majority (approximately 2/3 of the nebulized perflubron) was delivered in vapor form. The perflubron which was delivered in aerosol form demonstrated a relatively small particle size distribution, with a Mass Median Aerodynamic Diameter (MMAD) of 1.1 um, with a Geometric Standard Deviation (GSD) of 2.5 (Table 1). The perflubron aerosol size distributions obtained from the Sidestream nebulizer are shown in Figure 1.

The characterization of nebulized perflubron demonstrated that the majority of the nebulized perflubron existed in the vapor form while the remaining proportion of aerosol existed as a relatively fine aerosol with an MMAD of 1.1 um. We infer that perflubron as vapor and aerosol, as well as CO₂ gas component of inhaled S-1226, would rapidly distribute throughout the respiratory tract.

Sensitized and HDM challenged sheep treated with S-1226 for 2 minutes during the peak of the early phase allergic response showed an immediate decline in lung resistance in all animals following HDM challenge (Figure 2).

Treatment with S-1226 compared to the other arms of the study are shown in Figure 3. Lung resistance declined by 50.1% ± 5.5, 47.2% ± 6.2 and 51.8% ± 8.3 (mean ± SE) immediately, 1 to 10 minutes and 10 to 20 minutes after treatment with S-1226 respectively. Treatment with 12% CO₂ resulted in declines in lung resistance of 36.9% ± 6.8, 20.6% ± 12.0 and 24.3% ± 28.4 (mean ± SE) immediately, 1 to 10 minutes and 10 to 20 minutes after treatment respectively. Treatment with perflubron + medical air resulted in declines in lung resistance of 24.8% ± 7.5, -1.3% ± 10.7 and -7.2% ± 5.1 (mean ± SE) immediately, 1 to 10 minutes and 10 to 20 minutes after treatment respectively. Treatment with salbutamol resulted in declines in lung resistance of 31.3% ± 6.8, 15.3% ± 7.1 and 14.3% ± 9.6 (mean ± SE) immediately, 1 to 10 minutes and 10 to 20 minutes after treatment respectively. Following inhalation of medical air, lung resistance declined by 23.1% ± 9.7, 16.0% ± 10.8 and 13.1% ± 17.5 (mean ± SE) immediately, 1 to 10 minutes and 10 to 20 minutes after treatment respectively. Lung resistance measured following HDM challenge with no treatment decreased by 1.9% ± 4.7, 24.4% ± 4.2 (mean ± SE) at 1 to 10 minutes and 10 to 20 minutes after treatment respectively.

Inhalation of nebulized S-1226 caused a rapid and sustained decrease in lung resistance during the early phase asthmatic response which was significantly greater than no treatment at 1–10 minutes (p = 0.002) and significantly greater than medical air immediately, 1–10 minutes and 10–20 minutes after treatment (p = 0.027, 0.018 and 0.039 respectively). The effect of S-1226 was significantly greater than salbutamol 1–10 minutes and 10–20 minutes after treatment (p = 0.025 and 0.046 respectively). S-1226 showed a significantly greater decline in lung resistance than perflubron alone immediately, 1–10 minutes and 10–20 minutes after treatment (p = 0.019, 0.002 and 0.012 respectively). Finally, the effect of S-1226 was stronger but not significantly different from CO₂ alone immediately, 1–10 minutes and 10–20 minutes after treatment (p = 0.329, 0.124 and 0.319 respectively). The CO₂ decline in lung resistance appeared shorter-lasting than the decline observed with S-1226.

The rapidity of the effect was demonstrated by video-bronchoscopy. MCh challenge resulted in rapid constriction of the segmental and subsegmental bronchi. Without treatment the airway narrowing lasted for at least 20 minutes before a slow natural recovery. Following treatment with a spray of S-1226 there was an immediate re-opening (within 5 seconds) of the MCh-constricted airway (Figure 4). A video recording of this treatment effect can be viewed online under Additional file 1. The effect was sustained during the delivery of S-1226 and up to 1 minute after treatment cessation. After that the airway began to re-constrict. Second and third treatments with S-1226 had identical effects.

Our study did not attempt to determine the mechanism of action of S-1226. The bronchodilatory effect of CO₂ has been demonstrated in many mammalian species including cats, rats dogs [19],[22] as well as sheep and humans However, the literature indicates that it might operate through several distinct mechanisms.

The first mechanism is through a direct physiologic relaxant effect of CO₂ on airway smooth muscle [20]–[22]. The way(s) in which CO₂ regulates bronchial smooth muscle tone is not fully understood. Recent in vitro data indicates that part of this effect is mediated by pH dependent and part by pH independent but epithelium dependent mechanisms [22]. Studies in vivo indicate the response (in rats) to inhaled CO₂ does not require adrenergic pathways [46]. Fisher and Hansen [27] showed in their study of exercise induced asthma that cholinergic pathways were not involved in the bronchodilatory effects of inhaled CO₂. Bronchodilator drugs that use β- adrenergic and cholinergic pathways form the current basis for treating obstructive lung disease.

Inhaled carbon dioxide (5, 10 and 15%) has been shown to reduce total lung resistance induced by intravenous histamine, acetylcholine and serotonin in dogs, cats and rats [19]. Inhaled CO₂ has also been shown to reduce total lung resistance induced by pulmonary artery occlusion in cats and dogs [19]. All species studied responded to CO₂ but there were small differences in response to the different broncho-constricting agents between species.

Studies by Ingram and colleagues [47] suggest that the bronchodilatory effect of CO₂ may be independent of blood PCO₂. They studied the effect of varying airway and arterial PCO₂ on airway tone in dogs. They showed that low airway PCO₂ combined with high arterial PCO₂ resulted in a large increase in airway resistance [47]. These findings indicate that CO₂ may have unique properties when delivered through the airways and supports the previously cited in vitro work that airway epithelium plays a part in the relaxant response [22]. Finally the rapidity of the response as seen in the sheep video indicates that neural mechanisms are involved. The non-adrenergic-non cholinergic (NANC) pathways are the best candidates for this effect. The NANC system has both broncho-constrictive and bronchodilator components [48]. CO₂ alone or S-1226 may activate the bronchodilatory pathways or inhibit the broncho-constrictive pathways. Further work is required to address these questions.

The second mechanism may enhance delivery of CO₂ to the airways. It could do this in several ways; first sustained release of CO₂ dissolved in perflubron droplets as the droplets evaporate may explain the temporal synergism observed with the combination treatment. Second, because much of S-1226 is in gaseous form, it will be delivered rapidly into obstructed airways by diffusion and possibly through collateral ventilation.

The third mechanism whereby this combination treatment may work is by facilitating penetration and lubrication of mucus plugs, thus enhancing muco-ciliary clearance. We recently reported, using an in vitro system, that mucin plug clearance was significantly enhanced by perflubron [49].

S-1226 may also lower the surface tension in the inflamed airways. It has been shown that the normal airway is lined by a thin film of endogenous surfactant with a low surface tension [50]. This film is important for maintaining small airway patency at low lung volumes [51] and is abnormal in asthma [52],[53].

Finally, carbon dioxide and perfluorocarbons may also be beneficial for their anti-inflammatory properties. Perflubron given in vitro or during partial liquid ventilation exerts an anti-inflammatory effect on alveolar cells [54], decreases white blood cell counts, [54],[55] and the pro-inflammatory cytokines IL-1 and IL-6. Perflubron also increases release of the anti-inflammatory chemokine IL-10 [54].

Like with most experimental systems, large animal models of allergic airway disease have both strengths, and also limitations, which need to be recognised when evaluating new therapies for asthma [56]–[58]. For example, clinically efficacious anti-asthma drugs such as leukotriene LTE4 receptor antagonists, cromolyn and corticosteroids have all been successfully tested in the sheep model of asthma [59]. The success of these drugs is in contrast to the testing of platelet activated factor (PAF) antagonist therapies in sheep. PAF antagonists were shown to alleviate allergic airway responses in sheep [60], however the positive outcome with these treatments in sheep did not translate to humans giving negative results in clinical trials [61]. Nevertheless, sheep are an appropriate species to investigate the efficacy of the bronchodilator reported here for a number of important reasons. Firstly, the sheep respiratory system is at least comparable to the human system with respect to lung size, lung anatomy and morphology, and the pharmacological and physiological parameters of the ovine lung [57]–[59]. Pertinent to this study, as the mechanism of the bronchodilator most likely involves both the airway smooth muscle and the intact airway epithelium [22], is that the arrangement and morphology of ovine airway epithelium [62] and airway smooth muscle down the tracheo-bronchial tree [57] is functionally similar to that observed in human lungs. Thus, the bronchodilator interactions we report here for the sheep model could reasonably translate to the human lung. The other major asset of the sheep model for examining bronchodilator effectiveness relates to the docile nature of sheep, which allows experiments to be ethically conducted in unsedated and unanaesthetised animals [59],[63]. These anaesthetic and/or sedation agents, if administered as is necessary for behavioural reasons in most other animal species modeling asthma, may otherwise dampen, and thus mask, the bronchodilator responses.

In summary, we provide evidence that treatment with inhaled S-1226 delivered as an aerosol/gas mixture may play an important role in the treatment of acute asthma attacks. We show that perflubron combined with 12% CO₂ is a potent airway smooth muscle relaxant in this ovine model of allergic asthma. The response to the combination treatment was rapid and exceeded treatment with medical air (negative control) and salbutamol (positive control) in both amplitude and temporal components. Further work is required to fully evaluate the mechanism(s) of action of this novel therapy.