Design strategies for solution-based metered dose inhalers

Published: 16-Jun-2017

In recent decades, pharma companies have developed a better understanding of the factors affecting the performance of solution-based metered dose inhalers (MDIs). Here, Dr David Lewis examines the mathematical tools available to guide the manipulation of device and formulation to improve performance

You need to be a subscriber to read this article.
Click here to find out more.

In a suspension MDI, the active pharmaceutical ingredient (API) is suspended in the delivered droplets, so there is not necessarily a direct link between droplet size and the size of the delivered API particles. In contrast, in solution MDIs, the API is completely and homogenously dissolved within the delivered droplets. This makes drug delivery a more predictable process and underpins the highly consistent dosing that such products deliver during their lifetime.

In the 1990s, the phasing out of chlorofluorcarbons (CFCs) as propellants in MDIs and their replacement by hydrofluoroalkanes (HFAs) led to a resurgence of interest in solution-based MDIs, and the associated growth of the knowledge base for this technology. As a result, we now have semi-empirical design equations for ethanol-HFA-based MDIs that enable performance matching and/or the reliable and rapid development of new products with well-defined performance characteristics. For solution-based MDIs, these include delivery efficiency, which is quantified via the metrics of fine particle fraction (FPF) and/or fine particle dose (FPD), and aerodynamic particle size distribution (APSD).

In this article, we examine the properties that influence MDI performance and the empirical formulae that allow the efficient manipulation of device geometry and formulation composition to meet defined drug delivery targets. The application of these relationships is illustrated via example data for ipratropium bromide (IpBr) formulations. The ability to predict and regulate performance substantially eases the development of solution MDIs for specific illnesses and, at the same time, focuses attention on strategies to advance the technology to further enhance clinical outcomes by, for example, improving patient acceptance and comfort.

Developing solution MDIs: a step-by-step approach

Characterising solubility: Creating an optimal formulation for a solution MDI is substantially aided, in the first instance, by measuring the solubility of the API within the core components of the formulation: the propellant (typically HFA) and the cosolvent (typically ethanol). These data provide a foundation for the effective manipulation of the composition of the formulation, in particular making it possible to determine the drug loading potential of the MDI. Maximising drug loading is often an important goal as it reduces the dose required for efficacy, thereby easing delivery requirements.

Example solubility data for IpBr (0.037 w/w solution measured at 4 °C) within an HFA, water and ethanol system is shown in Figure 1. Water has been included in the study to investigate the tolerance of drug solubility to inadvertent water ingress during manufacturing, storage or use. The results define a clear region wherein homogenous solutions can be obtained and indicate that solubility can be achieved provided that water content is controlled within the formulation.

Figure 1: Solubility phase diagram for IpBr (0.037 w/w solution at 4 °C) differentiates compositions wherein the system is a single phase, homogenous solution (°), from those where it exists as a two or three phase dispersion (.).<sup>2</sup>

Figure 1: Solubility phase diagram for IpBr (0.037 w/w solution at 4 °C) differentiates compositions wherein the system is a single phase, homogenous solution (°), from those where it exists as a two or three phase dispersion (.).2

Particle size distribution (PSD) dynamics/development: Although formulation composition is dependent on solubility, it is equally defined by the need to deliver a clinically efficacious dose at a respirable particle size. Characterising the processes behind the atomisation of a formulation — the dispersion and evaporation of the solution to leave residual drug particles — is challenging. However, because of the link between delivered drug particle size — the primary particle size of interest — and initial droplet size, substantial effort has been invested to understand them. Table I summarises the steps involved in the transformation of an atomised droplet into a residual particle, highlighting which component of the formulation defines behaviour at each point.

Table 1: Droplets propelled out of the MDI valve transform rapidly into far smaller residual drug particles via a process of flashing and evaporation.<sup>2</sup>

Table 1: Droplets propelled out of the MDI valve transform rapidly into far smaller residual drug particles via a process of flashing and evaporation.2

During the flashing and evaporation phase of plume development, the diameter (or volume) of each droplet within the plume decreases to a final residual value. Research has shown that this value is independent of the characteristics of the device and is defined purely by the non-volatile content (NVC) of the formulation.1 This means that the mass median aerodynamic diameter (MMAD) of the particles produced by an MDI can be predicted from the NVC of the formulation using an empirical equation, for any given propellant. These relationships are presented below for HFA 134a and HFA 227ea formulations.

Equation 1: MMAD134a = 2.31 (%w/w NVC)1/3

Equation 2: MMAD227ea = 3.26 (%w/wNVC)1/4

Quantifying the efficiency of drug delivery (FPD and FPF): The APSD of the delivered drug is important in defining MDI performance because it influences deposition behaviour in the lung, and consequently the ability of the drug to reach the target within the lung. Potential MDI drug delivery performance may be characterised by the metrics FPD — the dose (mass) of drug delivered that has an aerodynamic particle size of µm — and ex-valve FPF — the ratio of the FPD to the ex-valve metered dose. A high metered dose will not necessarily correspond to a high FPD if the device is inefficient in terms of dispersion performance.

Equation 3: FPF = 2.1 x 10-5 a-1.5 v-0.25 C134a3

Equation 3 describes the relationship linking the ex-valve FPF of the product with key characteristics of the device, such as actuator exit orifice (a, expressed in mm) and metered volume (v, expressed in μL), and of the formulation, namely propellant concentration (HFA-134a – C, %w/w).1 This was derived from the results of an extensive series of experiments in which the drug delivery performance of a wide range of MDI products was characterised, using an Andersen Cascade Impactor (ACI) to determine APSD and FPF (see Figure 2 for example data).

Figure 2: Observed ex-valve FPF compared with ex-valve FPF predicted from Equation 3 for ethanol based HFA 134a solution formulation, related to MDI orifice diameter. Actuator orifice diameter = 0.22–0.42 mm, metering volume 50 µL.<sup>2</sup>

Figure 2: Observed ex-valve FPF compared with ex-valve FPF predicted from Equation 3 for ethanol based HFA 134a solution formulation, related to MDI orifice diameter. Actuator orifice diameter = 0.22–0.42 mm, metering volume 50 µL.2

The formulation-device combinations tested varied in terms of API (entity and dose), HFA 134a content, level of non-volatile additives, ethanol content, the diameter of the actuator and metering valve volume. This suggests that the derived equation will hold for the vast majority of solution MDI products.

Example study: predicting the performance of IpBr solution MDIs

To illustrate and test the effectiveness of the developed design equations, they were used to predict the performance of IpBr formulations/products. Referencing the solubility information presented in Figure 1, an HFA-134a-based solution containing 13% w/w ethanol and 1% w/w glycerol (as a non-volatile additive, see Equation 1) was formulated for testing. Table II displays the experimental results generated using a nominal formulation dose of 20 µg, 40 µg, 80 µg and 160 µg, delivered via a device fitted with a 0.30 mm actuator (Bespak BK 630).

Table II: Comparing observed and calculated values for FPD, FPF and MMAD for IpBr formulations delivered with a Bespak BK630 actuator shows that the developed design equations reliably predict MDI performance. Mean ± SD (*ratio of fine particle dose to metered dose). Obs. = Experimental Observation, Calc. = Calculated using equations.

Table II: Comparing observed and calculated values for FPD, FPF and MMAD for IpBr formulations delivered with a Bespak BK630 actuator shows that the developed design equations reliably predict MDI performance. Mean ± SD (*ratio of fine particle dose to metered dose). Obs. = Experimental Observation, Calc. = Calculated using equations.

For each dose, the observed FPD, FPF and MMAD values, as measured by multistage cascade impaction, are compared with those generated using Equations 1 and 3 as presented above. The results show a high degree of agreement between the experimental and the predicted values. With these design equations in place, optimising a product to meet clinical goals and/or to improve cost-effectiveness becomes far easier — a tabletop exercise rather than a trial and error experimental study. Figure 2 shows the impact of changing the actuator orifice diameter (using four different orifice diameters: 0.22, 0.30, 0.33 and 0.42 mm) on FPD. Like the previous experiment, the calculated values of FPD, FPF and MMAD were found to be in close agreement with the values derived from experimental observations, with FPD decreasing as actuator orifice diameter increases.

These results show how alternative FPD values can be delivered from this single packaged formulation to achieve the most cost-effective solution for the delivery of a clinically efficacious dose. For example, if we take a FPF to be 30%, then to achieve a FPD of approximately 75 µg, we can either use a metered dose of 260 µg delivered using the standard 0.30mm actuator or a smaller dose — 160 µg coupled with a smaller 0.22 mm actuator orifice — depending on which is found to be the most clinically effective, safest and least expensive overall option.

A platform for progress

Beyond faster product development and cost optimisation, the high degree of understanding of MDI behaviour encapsulated by the preceding design procedures provides a secure foundation to focus on improving the patient’s experience of using MDIs. Reduction of undesirable oropharyngeal drug deposition may be achieved with add-on devices such as spacers and holding chambers, but also by careful formulation and device design such that drug delivery is focused upon the intended targeted region.

For example, Equation 3 indicates that reducing cosolvent content below 5% w/w, so that HFA content can be pushed up to >95%, can result in an ex-valve FPF of more than 70%. A reduction in drug loading would be necessary to ensure solubility in the face of reduced cosolvent concentration, but the potential prize is highly efficient drug delivery with conventional hardware.

This approach is exemplified by the reformulation of a Pulmicort, a CFC suspension-based product that delivers the API Budesonide. The example reference product delivers a FPD of 8 µg from a metered dose of 50 µg (FPF = 17%). At low ethanol concentrations, an HFA 134a-based alternative can deliver an FPF of 82%, thereby providing a 9 µg FPD from an 11 µg metered dose. This is clearly an attractive proposition in terms of drug usage that delivers substantially reduced oropharyngeal drug deposition, an added attraction for the patient. Such performance can be readily achieved, even without advances in device technology, simply by using the knowledge available to exploit the performance of current hardware.

In conclusion

Exploiting the full potential of a technology relies in the first instance on developing a comprehensive understanding of how it works. The performance of solution MDIs is arguably better understood than that of any other orally inhaled product as evidenced from the empirical equations presented here, which enable the prediction of MMAD and ex-valve FPF from the geometry of a device and the composition of the formulation. These equations allow product developers to manipulate characteristics such as actuator orifice diameter, NVC and cosolvent content to control the efficiency of drug delivery. Such correlations support the reformulation and development of products that combine optimised cost-efficiency with high patient acceptance, as a result of low oropharyngeal deposition, potentially leading to “warmer” and “softer” drug delivery.

References

  1. D.A. Lewis, et al., “Theory and Practice with Solution Systems,” in R.N. Dalby, et al. (Eds.), Respiratory Drug Delivery 2004. Volume 1 (DHI Publishing, River Grove, IL, USA, 2004): pp 109–115.
  2. D.A. Lewis, “A Priori Design of Metered Dose Inhalers,” Respiratory Drug Delivery Asia 1, 187–198 (2016).

You may also like