Automatic actuation

Paul Kippax, Julie Suman, and Dino Farina discuss the effect of automatic actuation parameters on the in vitro performance of nasal sprays.

Nasal delivery of a range of pharmaceutical actives, including vaccines, proteins and peptides, is currently the subject of much research in the pharmaceutical industry. A nasal spray formulation typically consists of an active ingredient dissolved or suspended in an aqueous medium, and is delivered to the nasal cavity using a metered spray pump actuated by the patient.

The pump action delivers an accurate, atomised dose that is associated with specific performance, such as the shape of the spray and the size of the emitted droplets. The spray droplet size and the overall spray plume shape may be important in defining the deposition pattern observed within the nasal cavity.1,2 As such, spray droplet sizing using in vitro techniques such as laser diffraction, along with accurate measurement of plume geometry and spray pattern using rapid imaging systems, can be used as a surrogate for US FDA in vivo bioequivalence studies.

driving forces

One of the most important issues that should be considered before commencing in vitro nasal spray characterisation is how to properly and realistically actuate the spray pump. Studies have shown that automated pump actuation results in a higher degree of reproducibility during testing compared with manual actuation.3 Therefore, automated nasal spray actuators have been developed to meet the needs of the industry. The FDA's 2003 draft guidance 'Bioavailability and Bioequivalence Studies for Nasal Aerosols and Nasal Sprays for Local Action'4 has been a driving force in the development of these systems, providing detailed advice regarding actuation station specifications and including recommendations for automated actuation 'for all comparative in vitro bioequivalence tests, to decrease variability in drug delivery due to operator factors'. The FDA also recommends that studies be carried out to confirm that the automated actuation station and its parameter settings provide a realistic model for manual actuation by a trained patient for all in vitro testing.

This article documents a study that was conducted to investigate the effects of automated actuator parameters on in vitro tests for nasal sprays with two placebo formulations. The derivation of representative actuation parameters is discussed and the effect of varying these parameters analysed.

Spray pump (device) performance was related to the properties of the formulation, in this case by considering the rheological properties of the suspension being atomised.

Tests were carried out using two commercially available pumps with spray volumes of 50ml (Pump A) and 100ml (Pump B). The behaviour of two different formulations was assessed - one a suspension containing 2% Avicel CL-611 (FMC Corporation, Philadelphia, PA) and the other a solution of 2% HPMC 2910 (Spectrum Chemical Mfg. Corp., Gardena, CA). The viscosities of both formulations were measured over the shear stress range 0.02 to 2 Pa at 25°C using a Bohlin Gemini Rheometer (Malvern Instruments).

The pumps were tested at each set of actuation conditions in triplicate. Each pump was weighed before and after actuation to measure spray weight, which was defined as being acceptable if within 15% of the theoretical spray weight of the pump.

simulation training

As mentioned, the FDA guidelines for bioequivalence studies stress the importance of using an automated actuation system that can accurately simulate a trained patient using the product. To determine an appropriate range of actuation settings, Image Therm Engineering's SprayVIEW Ergo NSP was used to measure and analyse the manual actuation of five trained volunteers, one male and four female, in the age range 30 to 45 years. In this analysis each pump was filled with water and actuated manually five times according to instructions in the manufacturers' information (package insert).

The SprayVIEW Ergo system automatically records the actuation velocity, acceleration, stroke length (distance travelled by the spray pump as it is compressed) and hold time (amount of time the pump is held it a compressed position) for manual actuation.5

From these results the average actuation parameters were calculated along with the extreme profiles relating to high- and low-velocity and acceleration (tables 1 and 2). These profiles were then programmed into the Image Therm SprayVIEW NSx nasal spray actuator, a system designed to accurately follow a user-specified pump actuation profile, as defined by the stroke length, hold time, velocity and acceleration. The SprayVIEW NSx actuator was used to control pump actuation for the droplet size, spray pattern and plume geometry measurements described below. In each case the average hold time and stroke length were constant.

The droplet size produced by the pump A and B for each of the specified actuation conditions was measured using Malvern Instruments' Spraytec - a laser diffraction instrument designed to measure accurately and reproducibly the particle size distribution data of sprays, including inhalers.

data acquisition

The instrument reported a cumulative volume distribution representing 10%, 50% and 90% of the distribution (Dv10, Dv50 and Dv90). The nasal spray tip was positioned 4 cm beneath the laser beam within the laser diffraction measurement zone.

The Spraytec can acquire data at a rate of one measurement every four milliseconds, allowing the changes in particle size during each pump actuation to be resolved and each phase of the actuation to be identified in accordance with the FDA guidance (figure 1). The median particle size (Dv50) delivered for the fully developed part of the plume was then calculated.

The Image Therm SprayVIEW NSP image analysis system was used to quantify the shape of the emitted spray in terms of spray pattern and plume geometry measurements. Each pump used the specified actuation conditions as defined in Tables 1 and 2. The SprayVIEW NSP system uses a sheet laser light source to illuminate droplets passing through a specific plane within the spray plume. Rapid image acquisition, followed by analysis of the images obtained, allows each of the parameters specified in the FDA's guidance document to be assessed objectively.

spray patterns

Illuminating the plume horizontally allows the spray pattern to be assessed (figure 2). By time-averaging spray pattern images taken during an entire actuation cycle, an overall characteristic spray pattern can be obtained. From the time-averaged image, a variety of measurements can be made of the spray pattern including the maximum (Dmax) and minimum (Dmin) diameters and ovality ratio (Dmax/Dmin), as recommended by the FDA. Ovality ratio is used to provide a quantitative measure of shape, such that an ovality ratio of one is equivalent to a circle. In this study the spray pattern was assessed at the same distance away from the pump nozzle as was used for the laser diffraction study (4cm).

Illuminating the plume in a vertical plane allows the plume geometry to be assessed (figure 3). This was carried out using a single 'snapshot' image collected for the fully developed part of the plume (between 94ms and 143ms after the start of actuation, depending on the formulation being tested). From this image the plume geometry characteristics of the spray were measured including plume angle and plume width at the spray pattern distance (4cm) as recommended by the FDA.

The results obtained for each of the pumps and formulations are summarised in tables 3 to 6.

The spray weight delivered during actuation was in all cases found to be within 15% of the theoretical value, being on average 50.9mg for pump A and 103.3mg for pump B across both formulations. This shows that for each actuation profile the metering chamber in the pump was emptied.

While the spray weight was unaffected, the droplet size was found to vary depending on the chosen actuation parameters and the formulation. In general, the droplet size decreased as the velocity and acceleration was increased. The size reported for the 2% HPMC formulation was much coarser than for the 2% Avicel formulation, showing incomplete droplet break-up in the case of HPMC.

The percentage change in the mean particle size for each of the different actuation profiles was also much more pronounced for HPMC.

The droplet size distributions produced by the 2% Avicel formulation were within the range of performance expected for a commercially viable nasal spray product.

The performance of the 2% HPMC formulation was less robust compared with the Avicel formulation, as confirmed by the spray plume and geometry measurements. These show incomplete break-up of the liquid exiting the nasal spray nozzle, as evidenced by a narrow plume angle and irregular plume pattern. The pattern also changes significantly when the actuation profile is changed.

In contrast, the Avicel formulation produces a more uniform spray plume with lower ovality ratios and larger plume angles and is less sensitive to changes in actuation profile.

The above behaviour is, in part, related to the rheological properties of the formulations.

The 2% Avicel suspension displays a high viscosity that reduces as shear rate is increased (figure 4). This behaviour has always been viewed as advantageous: sedimentation within suspension-based formulations is prevented by the high viscosity at low shear rates, whereas the effective viscosity at higher shear rates is low, yielding a reasonable spray plume. In contrast, the 2% HPMC formulation appears to be Newtonian over the measured range.

significant advantages

However, the results here show that building a correlation between atomisation performance and the product viscosity is more complex that was first thought, as the magnitude of the viscosity of the HPMC formulation is similar to that of Avicel at high shear.

This suggests that the shear viscosity may not be a good predictor of atomisation performance in this case. Instead, the extensional viscosity (i.e. the behaviour of the solutions under tension rather than under shear) may better explain the results obtained above. This will be dependant on the molecular weight and conformation of each of the polymers. Nevertheless, maintaining the shear viscosity of the drug product will be important to overall product performance and stability.

In this study the group of volunteers used to generate the actuation parameters was not diverse: they were a trained group with a narrow age range. This means that the range of actuator velocities and accelerations measured is likely to be narrow.

A group of naïve, paediatric or geriatric volunteers is likely to generate a wider range of actuation velocities and accelerations, meaning that the resulting variability in actuation profiles will be much wider.

While it is clear that automated actuation offers significant advantages over manual actuation, care needs to be taken over the selection of actuation parameters used in nasal spray testing.

The results clearly show that even over the relatively narrow range of conditions studied here, these parameters have a noticeable impact on spray dynamics, with the magnitude of any changes being related to the specific formulation. It is therefore important that the actuation parameters are chosen to be a realistic mimic of the patient to characterise accurately and reproducibly in vitro spray performance.