The influence of formulation and spacer device on the in vitro performance of solution chlorofluorocarbon-free propellant-driven metered dose inhalers

The purpose of this study was to evaluate the hypothesis that spacer devices have limited effect on the in vitro fine particle dose emitted from solution metered dose inhalers containing different proportions of HFA134a [1,1,1,2-tetrafluoroethane] propellant. Two solution formulations (80% and 97.5% wt/wt HFA134a) were tested across the actuator alone, actuator plus Aerochamber, and Ace holding chamber. Particle size distributions were determined using laser diffraction (LD) and cascade impaction (CI). Multimodal particle size distributions were identified using LD. CI analyses were characterized by a major mode located at ≈0.5 μm. The fine particle dose emitted from the inhaler spacer combinations containing 97.5% HFA134a was independent of the device setup used. Fine particle doses were influenced by spacer setup in 80% HFA134a formulations, indicating different plume dynamics of low vapor pressure formulations. Sampling inlet deposition was approximately O when spacer devices were used with either formulation. When spacers were not used, sampling inlet deposition was increased significantly. However, inlet deposition with the 97.5% HFA134a formulation was significantly less than that of the 80% HFA134a formulation (≈25% of emitted dose compared with 69% respectively). Thus, high propellant concentration formulations appear to have more robust in vitro performance. This is particularly important given the preponderance of poor patient compliance that is associated with spacer use. High propellant concentrations had the advantage of fine particle doses that were independent of the device setup and significantly lowered sampling inlet deposition when no spacer was used.     

Factors Influencing Aerodynamic Particle Size Distribution of Suspension Pressurized Metered Dose Inhalers

Pressurized metered dose inhalers (pMDIs) are frequently used for the treatment of asthma and chronic obstructive pulmonary disease. The aerodynamic particle size distribution (APSD) of the residual particles delivered from a pMDI plays a key role in determining the amount and region of drug deposition in the lung and thereby the efficacy of the inhaler. In this study, a simulation model that predicts the APSD of residual particles from suspension pMDIs was utilized to identify the primary determinants for APSD. These findings were then applied to better understand the effect of changing drug concentration and micronized drug size on experimentally observed APSDs determined through Andersen Cascade Impactor testing. The experimental formulations evaluated had micronized drug mass median aerodynamic diameters (MMAD) between 1.2 and 2.6 μm and drug concentrations ranging from 0.01 to 1% (w/w) with 8.5% (w/w) ethanol in 1,1,1,2-tetrafluoroethane (HFA-134a). It was determined that the drug concentration, micronized drug size, and initially atomized droplet distribution have a significant impact in modulating the proportion of atomized droplets that contain multiple suspended drug particles, which in turn increases the residual APSD. These factors were found to be predictive of the residual particle MMAD for experimental suspension HFA-134a formulations containing ethanol. The empirical algebraic model allows predicting the residual particle size for a variety of suspension formulations with an average error of 0.096 μm (standard deviation of 0.1 μm).KEY WORDS: aerodynamic particle size distribution (APSD), formulation, pressurized metered dose inhaler (pMDI), suspension     

Determination of Reference Ultrasound Parameters for Model and Hydrofluoroalkane Propellants Using High-Resolution Ultrasonic Spectroscopy

The aim of this research was to determine the reference ultrasonic velocity (v) and attenuation coefficient (α) for 2H, 3H-perfluoropentane (HPFP), 1,1,1,2-tetrafluoroethane (HFA-134a) and 1,1,1,2,3,3,3-tetrafluoroethane (HFA-227) propellants, for the future purpose of characterising pressurised metered dose inhaler (pMDI) formulations using high-resolution ultrasonic spectroscopy (HRUS). Perfluoroheptane (PFH) was used as a reference material for HPFP. With its velocity and attenuation coefficient determined at 25 °C, HPFP was subsequently used as a reference for HFA-134a and HFA-227. It was found that there is a linear decline in ultrasonic velocity with an increase in temperature. As with HPFP, the ultrasonic velocity of HFA-134a and HFA-227 were successfully calculated at 25 °C. However, the difference in density and viscosity between reference and sample prevented accurate determination of reference attenuation coefficient for the hydrofluoroalkanes. With ultrasonic velocity alone, dispersion concentration and stability monitoring for experimental pMDI formulations is possible using HRUS. However, at this point in time measurement of particle size is not feasible.     

The Abbreviated Impactor Measurement (AIM) Concept: Part II—Influence of Evaporation of a Volatile Component—Evaluation with a “Droplet-Producing” Pressurized Metered Dose Inhaler (pMDI)-Based Formulation Containing Ethanol as Cosolvent

The abbreviated impactor measurement (AIM) concept is a potential solution to the labor-intensive full-resolution cascade impactor (CI) methodology for inhaler aerosol aerodynamic particle size measurement. In this validation study, the effect of increasing the internal dead volume on determined mass fractions relating to aerodynamic particle size was explored with two abbreviated impactors both based on the Andersen nonviable cascade impactor (ACI) operating principle (Copley fast screening Andersen impactor [C-FSA] and Trudell fast screening Andersen impactor [T-FSA]). A pressurized metered dose inhaler-delivered aerosol producing liquid ethanol droplets after propellant evaporation was chosen to characterize these systems. Measures of extrafine, fine, and coarse particle mass fractions from the abbreviated systems were compared with corresponding data obtained by a full-resolution ACI. The use of liquid ethanol-sensitive filter paper provided insight by rendering locations visible where partly evaporated droplets were still present when the “droplet-producing” aerosol was sampled. Extrafine particle fractions based on impactor-sized mass were near equivalent in the range 48.6% to 54%, comparing either abbreviated system with the benchmark ACI-measured data. The fine particle fraction of the impactor-sized mass determined by the T-FSA (94.4 ± 1.7%) was greater than using the C-FSA (90.5 ± 1.4%) and almost identical with the ACI-measured value (95.3 ± 0.4%). The improved agreement between T-FSA and ACI is likely the result of increasing the dead space between the entry to the induction port and the uppermost impaction stage, compared with that for the C-FSA. This dead space is needed to provide comparable conditions for ethanol evaporation in the uppermost parts of these impactors.     

Low-temperature micronization of a peptide drug in fluid propellant: Case study cetrorelix

Aim of this study was to elaborate an efficient method for the micronization of the decapeptide cetrorelix (a GnRH-antagonist), in order to obtain a microsuspension as basis for other pharmaceutical preparations, such as e.g. inhalation aerosols. A modified pearl-mill coupled with a cryostat was used for the micronization of cetrorelix in fluid propellant and operated under different conditions. The obtained cetrorelix suspensions were analyzed for particle size distribution, purity of cetrorelix, and for metal contamination through abrasion from parts of the mill. The method allowed an effective micronization of cetrorelix. The mean particle size of the initial cetrorelix lyophilizate bulk ware was reduced from 52.5 μ (Volume Mean Diameter, VMD) down to 14.9, 6.1 and 3.1 μm, respectively, respectively. The HPLC analysis of all cetrorelix suspensions after micronization did not show signs of decomposition as compared to the initial product. The elementary analysis of the suspensions performed by inductively coupled plasma mass spectrometry revealed a negligible amount of contaminants in the suspension (Zr=max. 0.6 ppm; Fe, Cr, Ni, Ba, below limit of quantification, i.e.<0.14 ppm). The only appreciable contaminant. Aluminum (Al=1.1 ppm), was derived from the mechanical capping of aluminum canisters prior to analysis. The Zr determination in the suspension of 0.6 ppm, is still considered to be negligible as compared to the legally tolerated limit of air contamination. By low-temperature micronization in fluid propellant, fine drug suspensions of cetrorelix for pMDIs can be directly manufactured in one-step procedure without destruction of the peptide structure and without appreciable product contamination. Published: July 12, 2001.     

Aerosol generation by metered-dose inhalers containing dimethyl ether/propane inverse microemulsions

Water soluble compounds were incorporated into metered-dose inhalers (MDIs) by using water-in-propellant lecithin microemulsions, in which dimethyl ether (DME) and propane acted as both continuous phase and propellant. Lecithin, water, and water soluble compounds were added to glass MDI containers, valves were crimped on, and propellants were added using a pressure burette. Aerosols were produced using commercially available actuators, and inertial impaction was used to determine the mass median aerodynamic diameter (MMAD), geometric standard deviation (GSD), and fine particle fraction (FPF) of the resulting aerosols. The DME/propane/lecithin, microemulsion MDIs generated aerosols with particle size distributions suitable for pulmonary delivery (eg, MMAD 3.1 μm, FPF 59% for DME with lecithin content 3%, water content 2.5% [wt/wt]). Increasing water concentration (up to 8% wt/wt) was correlated with a reduction in FPF. Freezing and rewarming had no adverse effect on MMAD, GSD, or FPF. Storage of microemulsion samples for up to 3 weeks did not adversely affect the MMAD, GSD, or FPF. This approach may enable the pulmonary delivery of water soluble therapeutic agents via MDIs.     

Optimization of bioethanol production during simultaneous saccharification and fermentation in very high-gravity cassava mash

Hydrolysis and fermentation conditions for production of ethanol from very high-gravity cassava mash by Saccharomyces cerevisiae during simultaneous saccharification and fermentation (SSF) processing were optimized using a statistical methodology. During the first part of the study, Placket–Burman design (PBD) was used to study 19 factors that could potentially influence ethanol production. Gravity, particle size, initial pH, and fermentation temperature were identified as key factors that significantly increased final ethanol concentration. The main and interaction effects of these factors were subsequently evaluated based on a quadratic equation generated by central composite design (CCD) using response-surface methodology (RSM). Under the optimized very high-gravity conditions, the final ethanol concentration obtained from experiment increased from 8.21% (wt.%) to 15.03% (wt.%) and was in good agreement with model prediction. By employing two other commercial Saccharomyces strains, similar results were obtained under the same optimized condition. Therefore, we conclude that final ethanol concentration, ethanol productivity (V _P/max), glucose utilization (Y _G/s, Y _P/s), and fermentation efficiency (η _f) were enhanced or maintained under the optimized condition of 40% gravity, 390 μm particle size, initial pH 5.5, and 27°C fermentation temperature.     

Evaluation of the TSI aerosol impactor 3306/3321 system using a redesigned impactor stage with solution and suspension metered-dose inhalers

The purpose of this research was to evaluate a redesigned impactor stage for the TSI Model 3306 Impactor Inlet with nozzles adjusted to obtain a target cut-point of 4.7 μm. It has been determined that the previous cut-point used in the Model 3306 was nominally closer to 4.14 μm, thus potentially impacting the characterization of aerosol mass. The reassessment of the Model 3306 was performed on 4 solution and 2 suspension metered-dose inhaler (MDI) formulations. The redesigned impactor stage resulted in a 5% to 6% increase in aerosol mass when compared with the previous impactor stage for the products Ventolin-HFA, Proventil-HFA, and 2 cyclosporin solution formulations with high ethanol concentrations (15% wt/wt). For the formulations with low ethanol concentrations (3% wt/wt), minimal differences were observed between the 2 cut-points. In addition, this study reevaluated the requirement of a vertical inlet extension length when using the TSI 3306/3321 system with the redesigned cut-point. It was shown that the use of a 20-cm extension provides mass and aerosol size distributions that are comparable to the Andersen 8-stage Cascade Impactor, for both solution and suspension MDIs. This work indicates that the TSI 3306/3321 system is suitable for preformulation studies of both suspension and solution MDI systems. Published: March 10, 2006     

Measurement of resuspended aerosol in the Chernobyl area

During anthropogenic activities, such as agricultural soil management and traffic on unpaved roads, size distribution measurements were performed of atmospheric particulate radionuclides at a site in the Chernobyl 30-km exclusion zone. Analysis of cascade impactor measurements showed an increase of the total atmospheric radioactivity. In the cases of harrowing by a tractor and traffic on unpaved roads, a common shape of the size distribution was found with two maxima, the first in the 2–4 μm range, the second in the 12–20 μm range. The size distributions were compared to measurements during wind-driven resuspension. Particle number concentration measurements with an Aerodynamic Particle Sizer showed a dynamic dependence of the particle concentration in different size ranges on anthropogenic action. The increase of the mean concentration was for the large particles more than one order of magnitude higher than for fine particles during anthropogenic enhanced resuspension. From the measurement of the mass concentration, the radioactive loading could be estimated. An enrichment of radionuclides on resuspended particles (compared to soil particles) was found, with the highest enrichment for large particles. Micrometeorological considerations showed that large particles may frequently be subject to medium range transport. The dry deposition velocity was measured; the mean value of 0.026 m s^–1±0.016 m s^–1 is typical for 6–9 μm diameter particles. Received: 17 February 1998 / Accepted in revised form: 17 June 1998     

Studies on the interaction of water with ethylcellulose: Effect of polymer particle size

The purpose of this research was to investigate the interaction of water with ethylcellulose samples and assess the effect of particle size on the interaction. The distribution of water within coarse particle ethylcellulose (CPEC; average particle size 310 μm) and fine particle ethylcellulose (FPEC; average particle size 9.7 μm) of 7 cps viscosity grade was assessed by differential scanning calorimetry (DSC) and dynamic vapor sorption analysis. The amounts of nonfreezing and freezing water in hydrated samples were determined from melting endotherms obtained by DSC. An increase in water content resulted in an increase in the enthalpy of fusion of water for the two particle size fractions of EC. The amount of nonfreezable water was not affected by the change in particle size at low water contents. Exposure of ethylcellulose to water for 30 minutes is sufficient to achieve equilibration within the hydrated polymer at 47% wt/wt water content. The moisture sorption profiles were analyzed according to the Guggenheim-Anderson-de Boer (GAB) and Young and Nelson equations, which can help to distinguish moisture distribution in different physical forms. The amount of externally adsorbed moisture was greater in the case of FPEC. Internally absorbed moisture was evident only with the CPEC. In light of these results, and explanation is offered for the success of FPEC in wet-granulation methods where CPEC was not successful.     

Evaluation of preseparator performance for the 8-stage nonviable Andersen impactor

The preseparator of an Andersen impactor with different coating treatments for a range of particle-size distributions was evaluated. Limited theoretical simulations constrained by simplifying assumptions of the airflow fields in the preseparator and upper stages of an 8-stage Andersen impactor were used to reveal low-velocity and high-pressure regions for potential deposition. These regions were then sampled in subsequent particle deposition experiments. Disodium fluorescein aerosols were sampled with different coating treatments of the preseparator floor. Particles collected at impactor stages determined particle size distributions. Stage deposition was compared between different preseparator treatments (buffer and silicon oil). Collection efficiency in the preseparator followed the pattern buffer >silicon oil >untreated. Statistical differences (P>0.05) were noted in collection efficiency of large particles (45 μm-75 μm) in the preseparator. The mass median aerodynamic diameters and geometric standard deviations showed some statistical differences when different preseparator treatments for large particles were used; therefore, preseparator coating was shown to influence performance and thereby estimates of particle size by intertial impaction.     

A New Rapid On-Line Imaging Method to Determine Particle Size Distribution of Granules

The purpose of this research was to study the feasibility of the new image analysis method in the particle size determination of the granules. The method is capable of forming a three-dimensional topographic image of a sample surface from a digital picture. In the method, a flat granule bed surface was illuminated from three different directions, using the three primary colors (red, green, and blue). One color picture was taken by a digital camera, after which a topographic image of the object surface was constructed. The particle size distribution was then calculated from the image data. The particle size analysis method was tested both off-line and on-line. Off-line particle size measurement results determined by the image analysis method corresponded quite well to those of sieve analysis in the size fraction range 250–1,000 μm. In on-line application, images were successfully retrieved and median granule size trend could be calculated and followed during fluid bed granulations.     

Chemometric evaluation of pharmaceutical properties of antipyrine granules by near-infrared spectroscopy

The purpose of this research was to apply near-infrared (NIR) spectroscopy with chemometrics to predict the change of pharmaceutical properties of antipyrine granules during granulation by regulation of the amount of water added. The various kinds of granules (mean particle size, 70–750 μm) were obtained from the powder mixture (1 g of antipyrine, 6 g of hydroxypropylcellulose, 140 g of lactose, and 60 g of potato starch) by regulation of the added water amount (11–19 wt/wt%) in a high-speed mixer. The granules were characterized by mean particle size, angle of repose, compressibility, tablet porosity, and tablet hardness as parameters of pharmaceutical properties. To predict the pharmaceutical properties, NIR spectra of the granules were measured and analyzed by principal component regression, (PCR) analysis. The mean particle size of the granules increased from 81 μm to 650 μm with an increase in the amount of water, and it was possible to make larger spherical granules with narrow particle size distribution using a high-speed mixer. Angle of repose, compressibility, and porosity of the tablets decreased with an increase of added water, but tablet hardness increased. The independent calibration models to evaluate particle size, angle of repose, and tablet porosity and hardness were established by using PCR based on NIR spectra of granules, respectively. The correlation coefficient constants of calibration curves for prediction of mean particle size, angle of repose, tablet porosity, and tablet hardness were 0.9109, 0.8912, 0.7437, and 0.8064, respectively. It is possible that the pharmaceutical properties of the granule, such as mean particle size, angle of repose, tablet porosity, and tablet hardness, could be predicted by an NIR-chemometric method.     

Preparation of biodegradable microspheres and matrix devices containing naltrexone

In this study, the use of biodegradable polymers for microencapsulation of naltrexone using solvent evaporation technique is investigated. The use of naltrexone microspheres for the preparation of matrix devices is also studied. For this purpose, poly(L-lactide) (PLA) microspheres containing naltrexone prepared by solvent evaporation technique were compressed at temperatures above the Tg of the polymer. The effect of different process parameters, such as drug/polymer ratio and stirring rate during preparation of microspheres, on the morphology, size distribution, and in vitro drug release of microspheres was studied. As expected, stirring rate influenced particle size distribution of microspheres and hence drug release profiles. By increasing the stirring speed from 400 to 1200 rpm, the mean diameter of microspheres decreased from 251 μm to 104 μm. The drug release rate from smaller microspheres was faster than from larger microspheres. However, drug release from microspheres with low drug content (20% wt/wt) was not affected by the particle size of microspheres. Increasing the drug content of microspheres from 20% to 50% wt/wt led to significantly faster drug release from microspheres. It was also shown that drug release from matrix devices prepared by compression of naltrexone microspheres is much slower than that of microspheres. No burst release was observed with matrix devices. Applying higher compression force, when compressing microspheres to produce tablets, resulted in lower drug release from matrix devices. The results suggest that by regulating different variables, desired release profiles of naltrexone can be achieved using a PLA microparticulate system or matrix devices.     

Aerodynamic particle size analysis of aerosols from pressurized metered-dose inhalers: Comparison of andersen 8-stage cascade impactor,next generation pharmaceutical impactor,and model 3321 aerodynamic particle sizer aerosol spectrometer

The purpose of this research was to compare three different methods for the aerodynamic assessment of (1) chloroflurocarbon (CFC)-fluticasone propionate (Flovent), (2) CFC-sodium cromoglycate (Intal), and (3) hydrofluoroalkane (HFA)-beclomethasone dipropionate (Qvar) delivered by pressurized metered dose inhaler. Particle size distributions were compared determining mass median aerodynamic diameter (MMAD), geometric standard deviation (GSD), and fine particle fraction <4.7 μm aerodynamic diameter (FPF_{<4.7 μm}). Next Generation Pharmaceutical Impactor (NGI)-size distributions for Flovent comprised finer particles than determined by Andersen 8-stage impactor (ACI) (MMAD=2.0±0.05 μm [NGI]; 2.8±0.07 μm [ACI]); however FPF_{<4.7 μm} by both impactors was in the narrow range 88% to 93%. Size distribution agreement for Intal was better (MMAD=4.3±0.19 μm (NGI), 4.2±0.13 μm (ACI), with FPF_{<4.7 μm} ranging from 52% to 60%. The Aerodynamic Particle Sizer (APS) undersized aerosols produced with either formulation (MMAD=1.8±0.07 μm and 3.2±0.02 μm for Flovent and Intal, respectively), but values of FPF_{<4.7 μm} from the single-stage impactor (SSI) located at the inlet to the APS (82.9%±2.1% [Flovent], 46.4%±2.4% [Intal]) were fairly close to corresponding data from the multi-stage impactors. APS-measured size distributions for Qvar (MMAD=1.0±0.03 μm; FPF_{<4.7 μm}=96.4% ±2.5%), were in fair agreement with both NGI (MMAD=0.9±0.03 μm; FPF_{<4.7 μm}=96.7%±0.7%), and ACI (MMAD=1.2±0.02 μm, FPF_{<4.7 μm}=98%±0.5%), but FPF_{<4.7 μm} from the SSI (67.1%±4.1%) was lower than expected, based on equivalent data obtained by the other techniques. Particle bounce, incomplete evaporation of volatile constituents and the presence of surfactant particles are factors that may be responsible for discrepancies between the techniques.     

Abundance of picophytoplankton in the halocline of a meromictic lake, Lake Suigetsu, Japan

Numerous (0.5 to 4.8 × 10⁵ cells/ml), small phytoplankton (smaller than 0.5–1 × 1–2 μm in cell size, picophytoplankton) were distributed in the halocline (depth 2–12 m, 4–14 practical salinity units) of the saline meromictic lake, Lake Suigetsu (35°35′ N, 135°52′ E), located in the central part of the coast of Wakasa Bay along the Japan Sea in Fukui Prefecture, Japan. Vertical distribution of phytoplankton revealed that the maximum number of picophytoplankton was always observed near or a little deeper than the oxic-anoxic boundary layer (depth 5–6 m); they were dominant phytoplankton in the water layer deeper than the oxic-anoxic boundary from July to late September 2005. Spectral analysis of autofluorescence emitted from the particle fractions smaller than 5 μm measured with a spectrofluorometer and from individual cells measured with a microscope photodiode array detector revealed that the major component of picophytoplankton was phycoerythrin-rich, unicellular cyanobacteria (picocyanobacteria). Eukaryotic phytoplankton about 2.5 μm in diameter were also found, but the numbers were low. Fluorescence intensity of chlorophyll a at 685 nm (room temperature) emitted from the particle fractions smaller than 5 μm was increased by the addition of 3-(3,4-dichlorophenyl)-1,1-dimethylurea. These observations indicated that at least some picophytoplankton had a functional photosystem II in the halocline where sulfide, the potential inhibitor of oxygenic photosynthesis, was always present. The large abundance together with their physiological potency suggest that picophytoplankton are one of the important primary producers in the halocline of Lake Suigetsu. Electronic Supplementary Material Supplementary material is available for this article at and is accessible for authorized users.     

Formulation of a dry powder influenza vaccine for nasal delivery

The purpose of this research was to prepare a dry powder vaccine formulation containing whole inactivated influenza virus (VIIV) and a mucoadhesive compound suitable for nasal delivery. Powders containing WIIV and either lactose or trehalose were produced by lyophilization. A micro-ball mill was used to reduce the lyophilized cake to sizes suitable for nasal delivery. Chitosan flakes were reduced in size using a cryo-milling technique. Milled powders were sieved between 45 and 125 μm aggregate sizes and characterized for particle size and distribution, morphology, and flow properties. Powders were blended in the micro-ball mill without the ball. Lyophilization followed by milling produced irregularly shaped, polydisperse particles with a median primary particle diameter of ≈21 μm and a yield of ≈37% of particles in the 45 to 125 μm particle size range. Flow properties of lactose and trehalose powders after lyophilization followed by milling and sieving were similar. Cryo-milling produced a small yield of particles in the desired size range (<10%). Lyophilization followed by milling and sieving produced particles suitable for nasal delivery with different physicochemical properties as a function of processing conditions and components of the formulation. Further optimization of particle size and morphology is required for these powders to be suitable for clinical evaluation. Published: March 10, 2006     

Factors Affecting the Stability and Performance of Ipratropium Bromide; Fenoterol Hydrobromide Pressurized-Metered Dose Inhalers

The aim of the study was to investigate the factors affecting the stability and performance of ipratropium bromide and fenoterol hydrobromide in a pressurized-metered dose inhaler (pMDI). A factorial design was applied to investigate the effects of three parameters (propellant, water, and ethanol) on the performance of 27 designed formulations of a solution-based pMDI. The formulations that contained a hydrofluoroalkane (HFA) propellant lower than 72% v/v and an ethanol concentration higher than 27% v/v remained as clear solutions. Nine formulations that contained the HFA propellant higher than 74% v/v precipitated. The results indicated that it was not only the HFA propellant content of the formulations that was related to the formulation instability but also ethanol content. Only six formulations from the 18 formulations, that did not precipitate, produced drug contents that were within the acceptable range (80–120%). These six formulations generated aerosols with mass median aerodynamic diameters (MMAD) of approximately 2 μm with a fine particle fraction (FPF; particle size, <6.4 μm) between 45% and 52%. The MMAD and FPF did not change significantly after 6 months of storage (P > 0.05).KEY WORDS: fenoterol hydrobromide, ipratropium bromide, pressurized metered dose inhaler, stability