Ralph W. Niven

The influence of formulation on emission, deaggregation and deposition of dry powders for inhalation

Carrier and drug (“active”) particles containing different fluorescent dyes were studied by performing an experimental and theoretical analysis of the influence of formulation on the performance of dry powders for inhalation. The performance was evaluated based on the emission, deaggregation, and deposition of powders from a Spinhaler® into a model throat connected to a collection filter or an isokinetic sampling system and Andersen cascade impactor. Powders containing recombinant human granulocyte-colony stimulating factor/mannitol (GM) exhibited improved dispersion, more facile deaggregation and increased deposition in the model lung than those containing only mannitol (M). These results appear to be due to smaller interparticulate cohesive forces between GM particles as indirectly evidenced by surface indentations and bulk density measurements. Carrier particles (PEG 8000) improved total powder emission, but in many instances reduced the percent of active powder inhaled compared to formulations consisting of active powder alone. In formulations containing the carrier, deaggregation and deposition of active powders were highly dependent on the properties of the active powder and flow rate. The theoretical analysis of particle aerosolization criteria was limited to aerosolization of dry, uncharged powders. A relationship is derived relating flow rate of aerosolized powder to the smallest size of particle ogglomerate that can be expected to survive the turbulence generated by the inhaler/mouth/throat flow profile.

The Pulmonary Absorption of Aerosolized and Intratracheally Instilled rhG-CSF and monoPEGylated rhG-CSF

AbstractPurpose. The objective of this study was to highlight differences in the pulmonary absorption of a monoPEGylated rhG-CSF and rhG-CSF after intratracheal instillation and aerosol delivery. Methods. Male Sprague Dawley rats (250 g) were anesthetized and intratracheally instilled (IT) with protein solution or were endotracheally intubated and administered aerosol for 20 min via a Harvard small animal ventilator. A DeVilbiss “Aerosonic” nebulizer containing 5 ml of protein solution at ≈ 3 mg/ml was used to generate aerosol. The volume of protein solution deposited in the lung lobes was estimated to be ≈13 µl after delivery of Tc-99m HSA solutions. The PEGylated proteins consisted of a 6 kDa (P6) or 12 kDa PEG (PI2) linked to the N-terminus of rhG-CSF. rhG-CSF also was administered IT in buffers at pH 4 and pH 7 and in dosing volumes ranging from 100 to 400 µl. Blood samples were removed at intervals after dosing and the total white blood cell counts (WBC) were determined. Plasma was assayed for proteins by an enzyme immuno assay. Results. The plasma protein concentration v. time profiles were strikingly different for aerosol v. IT delivery. The Cmax values for rhG-CSF and P12 after aerosol delivery were greater than found after IT (Aerosol: 598 ± 135 (ng/ml) rhG-CSF; 182 ± 14 P12 v. IT: 105 ± 12 rhG-CSF; 65.9 ± 5 P12). Similarly, Tmax was reached much earlier after aerosol administration (Aerosol: 21.7 ± 4.8 (min) rhG-CSF; 168 ± 31 P12 v. IT: 100 ± 17 rhG-CSF; 310 ± 121 P12). Estimated bioavailabilities (Flung %) were significantly greater via aerosol delivery than those obtained after IT (Aerosol: 66 ± 14 rhG-CSF; 12.3 ± 1.9 P12 v. IT: 11.9 ± 1.5 rhG-CSF; 1.6 ± 0.1 P12). An increase in circulating WBC counts was induced by all proteins delivered to the lungs. The rate and extent of absorption of rhG-CSF was not influenced by the pH employed nor the instilled volume. Conclusions. Estimates of bioavailability are dependent upon the technique employed to administer drug to the lungs. Aerosol administration provides a better estimate of the systemic absorption of macromolecules.

Aerodynamics and aerosol particle deaggregation phenomena in model oral-pharyngeal cavities

Abstract The results of numerical simulations of the aerodynamics and of solid aerosol deaggregation phenomena arising in the process of airflow through various model human oropharyngeal cavities are reported. Special attention is given to the relevance of these simulations to the inhalation of dry-powder therapeutic aerosols. Several two- and three-dimensional mouth and throat geometries (terminating just beyond the larynx) are considered. Cross-sectional area-averaged viscous stress values are numerically determined as a function of distance from the mouth opening. These values, ranging from approximately 10 to 500 dyn cm −2 , are compared with estimates of Van der Waals attractive forces per unit area of particle-particle contact so as to evaluate the ability of the flowing airstream to deaggregate aerosol particles that enter the mouth in an aggregated state (held together principally by Van der Waals attractive forces). Estimates of airstream viscous stress differ markedly depending on whether the geometry is two- or three-dimensional. Quantitative differences between flow in a 90°-bend model and an oropharyngeal geometry numerically reconstructed from a cast of a human mouth and throat are especially significant in regards to the ability of the airstream to break apart particle agglomerates. For all geometries it is observed that increasingly smaller particle agglomerates may potentially be separated as the airflow rate increases from 30 to 2001 min −1 . At the highest airflows, aggregated particles of diameter near to or even below 1 μm may potentially be separated by the airflow. If separation of particle agglomerates is to occur, it appears far more likely to take place in the throat than in the mouth. This is especially apparent for the more physiologically faithful oropharyngeal geometries considered.

Nebulization of liposomes iii. the effects of operating conditions and local environment

Multilamellar liposomes (MLV) of saturated phosphatidylcholine and dipalmitoyl phosphatidylglycerol (DPPG) (9:1 mole ratio) containing 5,6-carboxyfluorescein (CF) were prepared and extruded through 1.0-µm polycarbonate membranes. Diluted aqueous dispersions were aerosolized for a total of 80 min using a Collison nebulizer under a variety of conditions. The effects of air pressure, temperature, buffer osmotic strength, and pH on nebulized liposome dispersions were studied. Changes in air pressure produced large changes in the percentage release of CF and ranged from 1.3% (4 psig) to 88.2% (50 psig) after 80 min of nebulization. The temperature of the nebulizer dispersions dropped during experiments. The extent of the temperature drop varied according to the air pressure used and ranged from 5°C (4 psig) to 11°C (≥30 psig). The temperature of dispersions caused no increase in CF release until the gel-to-liquid crystalline transition temperature was exceeded (54.6°C), whereupon a 20% increase in leakage was observed after 80 min of nebulization. Aerosol mass output was relatively unaffected by the starting temperature of experiments when conducted within the ambient temperature range. Leakage from the liposomes was increased in hypotonic solution but decreased in hypertonic solutions. At a buffer pH of 2.85 the percentage leakage of CF was increased ≈18% compared to that at pH 7.2 and pH 10.75. Results show that the stability of liposomes composed of saturated phosphatidylcholine and DPPG (9:1 mole ratio) is affected by the operating and environmental conditions under which aerosolization takes place, with air pressure having the greatest effect.

Some Factors Associated with the Ultrasonic Nebulization of Proteins

Ultrasonic nebulization of lactate dehydrogenase (LDH) was investigated using a DeVilbiss “Aerosonic” nebulizer. The enzyme (8ml, 0.025mg/ml Na2HPO4, pH 7.0) was completely inactivated after 20 minutes of operation. However, the inactivation profile observed during ultrasonic nebulization was different from that previously observed using air-jet nebulization. At least two mechanisms are involved, one associated with heating and the other with aerosol production. By preventing heating of the nebulizer fluid during operation, the denaturation profile was dramatically altered. By additionally including 0.01% w/v Tween 80 or l%w/v PEG 8000, almost all activity was retained. Similar results were obtained by preventing aerosol production and heating. However, 100% of activity was lost when heating was allowed to occur without aerosol formation. The results demonstrate that cooling in conjunction with a surfactant is one approach that could be used to stabilize proteins to ultrasonic nebulization. However, cooling also significantly reduced solute output from the nebulizer. When operated at 10°C output was negligible. At 50°C the output was 5× greater than that found at room temperature. The median droplet size (µm) was not significantly influenced by the operating temperature of the nebulizer fluid (3.6 ± 0.4, 21°C; 3.9 ± 0.2, 50°C, p = NS (n = 6)) although the size distribution was noted to increase at the higher temperature.

Pulmonary Delivery of Powders and Solutions Containing Recombinant Human Granulocyte Colony-Stimulating Factor (rhG-CSF) to the Rabbit

Two powder formulations (MMAD <4 µm) containing rhG-CSF were insufflated (IF) via an endotracheal tube at doses of 5, 75 or 500 µg/kg to New Zealand white rabbits. Doses of 5 and 500 µg/kg of solutions were administered by intratracheal instillation (IT), subcutaneous (SC) injection in the thigh and intravenous injection (IV) via the marginal ear vein. Blood samples were removed at regular intervals from an indwelling jugular catheter. Blood was analyzed directly for total white blood cell counts (WBC). Plasma was assayed for rhG-CSF by a specific ELISA. The distribution of radioactive dose in lung tissue was found after administering Tc99m HSA in solution or when incorporated into powders. The pharmacokinetics and pharmacodynamics were determined for all routes of administration. High dose IV concentration vs. time profiles declined biexponentially (t1/2 α = 0.6 ± 0.2 hrs, t1/2 β = 4.6 ± 0.2 hrs, n = 8). Clearance was dose dependent (11.6 ± 2.6 [500 µg/kg, n = 8] vs. 21.8 ± 3.3 ml/hr/kg [5 µg/kg, n = 5]). A normal systemic response was obtained after IF, indicating that rhG-CSF retains activity in the solid state. Dissolution and absorption of rhG-CSF from the powders were not rate limiting. The plasma concentration vs. time profiles peaked at similar times to those after IT (Tmax 1 -2 hrs) but were earlier than obtained after SC (Tmax 6-10 hrs). Powders were less efficiently dosed to the lung lobes after insufflation compared with instillates (14.7 ± 10.5 vs. 60.1 ± 10.6%), resulting in bioavailabilities ranging from 5 to 33%. Bioavailability after SC was 11.0 ± 7.0% and 95.3 ± 7.9% (n = 6) for the low and high doses, respectively.

The Influence of Sodium Glycocholate and Other Additives on the in vivo Transfection of Plasmid DNA in the Lungs

AbstractPurpose. A plasmid containing the luciferase ‘marker’ cDNA was constructed to test non viral gene delivery formulations in vivo. Methods. A scale up procedure was devised to produce up to gram quantities of plasmid. Sufficient quantities were generated to process and test the DNA with various additives and to generate a spray-dried powder formulation of the plasmid. Male Sprague-Dawley rats (250 g) were intratracheally instilled with 200–250 µl of solution containing 200 µg plasmid ± lipid [DC Chol:DOPE 1:1 molar (2mg/kg)] growth factors [KGF (10 mg/kg), EGF (5 mg/kg)], permeation enhancers [sodium glycocholate (0.01 to 10% w/v)), sodium deoxycholate (1% w/v), beta-cyclodextrin (1% w/v)], surfactant [Tween 80 (1% w/v)], a mucolytic [N-acetylcysteine (10% w/v)] and positively charged synthetic polymers [PVAVAM 6 and 14%]. Animals were sacrificed 24 hr post-dose and the lungs were assayed for luciferase using a chemiluminescent assay. Results. The relative ability of the materials to promote luciferase production in the lungs was permeation enhancer >> DNA alone ≥ lipid, mucolytic, surfactant, growth factor > polymer. Protein production in the lungs ranged from 10 times below the DNA control (≈16 pg) using the polymers (≈1.5 pg) to ≈125 times greater than the control using the permeation enhancer (≈2050 pg). The transfection capabilities of the majority of additives was low. The enhancing effects of sodium glycocholate were dose-dependent and perhaps associated with the critical micelle concentration. Although the bile salt was the most successful of the tested compounds, it resulted in significant mortality when used at concentrations greater than 1 % w/v. Conclusions. The results suggest that transfection can be significantly enhanced by additives such as NaGC but some toxicity may be unavoidable.

Pulmonary Absorption of Recombinant Methionyl Human Granulocyte Colony Stimulating Factor (r-huG-CSF) After Intratracheal Instillation to the Hamster

Recombinant methionyl human granulocyte colony stimulating factor (G-CSF), a molecule of 18.8 kDa, has been shown to induce a systemic response after delivery by aerosol. In this work, rate and extent of absorption as well as the response were determined after bolus administration of solutions by intratracheal instillation (IT). The protein was quantified using a specific ELISA and the biological response was assessed by monitoring the increase in numbers of circulating white blood cells (WBC). A dose–response curve was obtained after IT, subcutaneous injection (SC), and intracardiac injection (IC) of 100 µL of a nominal dose ranging from 1 to 1000 µg/kg G-CSF (n = 5). WBC numbers were determined 24 hr postadministration. Absorption and clearance kinetics were determined after IT and IC of 500 µg/kg protein over a 24-hr time period (n = 5). The response of the lung to G-CSF was monitored by WBC counts and differentials in lung lavage fluid. 73.6 ± 10.5% (n = 7) of the IT dose reached the lung lobes. The response to single doses of G-CSF by IT or SC was similar, with WBC numbers increasing over 4× baseline at the higher doses. Absorption from the lung was rapid and did not follow first-order kinetics. Clearance after the IC dose was described by a biexponential equation (α = 1.41, β = 0.24 hr−1). Peak serum levels were obtained ≈1–2 hr after IT. The bioavailability was 45.9% of the administered dose and 62.0% of the dose reaching the lung lobes. These results indicate that G-CSF is rapidly absorbed from the lung. Pulmonary delivery via the airways has promise as an alternative to parenteral injection.

Some functional aspects of air-jet nebulizers

Mass median aerodynamic diameters (MMADs, /xm), flow rates (1 air/min) and mass outputs (ml/min and μ1/1 air) were determined over a range of air pressures (psig) for a variety of clinically used nebulizers. Nebulizers were also modified to determine the rate at which fluid was aspirated from the nebulizer cup into the nebulizer jets (ml/min and μ1/1 air). Only a small fraction of this fluid emerges as aerosol that can be inhaled. The ratio of mass output to fluid uptake from the reservoir is termed mechanical efficiency. The stability of the enzyme lactate dehydrogenase (LDH) to nebulization at two different starting volumes was examined in the 3-jet Collison nebulizer. The results are discussed with respect to the measured operating characteristics. In all the nebulizers tested, MMADs were inversely related (e.g. 6.4 to 3.8 μm at 10 and 40 psig, respectively, for the ‘Cirrus’ nebulizer), whereas mass outputs (e.g. 0.11 to 0.35 ml/min at 10 and 40 psig for the ‘Misty’) and flow rates (e.g. 4.5 to 10.4 l/min at 10 and 40 psig for the Misty) were directly related, to the driving air pressure. MMADs and mass output are measures of how well nebulizers atomize liquid into droplets of respirable sizes. Aspiration rates varied substantially between the nebulizers and ranged from 9 ml/min (5 psig) to 28 ml/min (40 psig) for the DeVilbiss No. 45 whereas the Collison 3-jet nebulizer aspirated liquid at 118 ml/min (5 psig) to 333 ml/min (40 psig). All nebulizers exhibited an exponential decline in the aspiration rate of liquid expressed in terms of the throughput of air (ml/1 air). Rate constants ranged from 0.6 to 21.6 (× 10−2 min/1 air) in the DeVilbiss No. 45 and Airlife Misty, respectively. These values give an indication of how the aspiration of fluid is affected by the air flow. Nebulizer mechanical efficiency was never > 2% (vents closed); this parameter is inversely related to the probable number of times that a drug will be aerosolized within the nebulizer. LDH was inactivated by nebulization. The degree of inactivation was greater at a 10 ml starting volume than at a 40 ml starting volume. The results indicate that monitoring the functional characteristics of nebulizers may be helpful in the assessment of drug stability.

Protein nebulization II. Stabilization of G-CSF to air-jet nebulization and the role of protectants

Abstract Air-jet nebulization induces aggregation and degradation of granulocyte-colony stimulating factor (G-CSF). The air-water interface as a site for destabilization of G-CSF is investigated and possible mechanisms of stabilization by polyethylene glycol (PEG) and Tween 80 are proposed. Equilibrium surface tension measurements show that G-CSF is highly surface-active, but dynamic surface tension measurements made using a maximum bubble pressure surface tensionometer show that G-CSF and surfactants have little surface activity in the interval between formation and destruction of aerosol generated within a nebulizer. Bubbling of air through solutions of 4 mg/ml G-CSF for up to 1 h resulted in ≈ 10% aggregation of total protein suggesting that destabilization at the air-water interface occurs. Polyethylene glycol (PEG) and Tween 80 are effective in reducing aggregation and degradation of G-CSF but their respective modes of protection differ. Protection is improved by increasing the PEG molecular weight and concentration indicating that PEG may influence stability through steric effects. In contrast, the protective effects of Tween 80 appear to relate to the micellar properties and not to the surface activity of the surfactant.