(375q) Optimising Powder Properties for DPI Capsule Filling Performance

Most single dose and multi-unit dose Dry Powder Inhalers (DPI) dispense the Active Pharmaceutical Ingredients (API) alongside a suitable carrier from a pre-metered capsule or blister pack. Due to the small particle size of the API (typically less than 5 microns) they tend to be highly cohesive so an excipient, for example lactose monohydrate, is required as a transport medium for the API. As the drug formulation is ejected from the inhaler the carrier is stripped away as the API continues to the deep lung. The selection of the carrier excipient will not only influence the ejection of the powder from the inhaler, and subsequent detachment of the API, but also the filling of capsule and blister packs during the manufacturing process.

The pre-metered dose required for DPIs varies considerably depending on the API and medical condition, from less than 1mg to tens of milligrams. In order to dispense small quantities, dosator technology is routinely used, with modern dosators capable of filling up to 180,00 capsules per hour. The operating principle of dosators is relatively simple, the dosator is inserted into the powder bed, forcing powder into the cavity, the powder is then compacted via the dosator pin, locking the powder into the cavity. The dosator (including the powder plug) is then retracted from the powder bed, and excess powder is then removed during the doctoring stage before the powder plug is ejected into the capsule, blister pack or vial.

Due to the way dosators operate they are more sensitive to the physical properties of the powder compared with other filling mechanisms. As the powder is dispensed by volume rather than mass the bulk density must be uniform throughout the powder bed, the powder should also flow into and fill the voids formed by the dosator, finally the powder should exhibit sufficient cohesion/compressibility to from a stable plug during retraction, doctoring and ejection. Dosators are therefore suited to a limited range of powders necessitating the need to match the powder properties to the dosator.

In order to further investigate this relationship, 5 grades of lactose with varying particle sizes were processed through a lab-scale dosator (3P Innovation, Warwick, UK) using outlets of progressively smaller size. The single arm dosator is capable of delivering up to 450 doses per hour, and in conjunction with a rotating powder bed, conditions the powder prior to and during operation using a blade and pin system, ensuring a uniform bed height. For each dosator-lactose combination, the standard deviation across 30 runs was measured for a target doses of 50 mg. The target relative standard deviation (RSD) was 2%. The powders were also evaluated using an FT4 Powder Rheometer (Freeman Technology Ltd, Tewkesbury, UK) to quantify dynamic flow, bulk and shear properties.

The performance varied significantly depending on the combination of lactose and dosator outlet size. Optimum performance was not related to particle size, and instead correlated with dynamic flow properties. Powders with low Specific Energy (SE) and Aerated Energy (AE) values were demonstrated to have the best performance across all four dosators, however as the dosator outlet size decreased sensitivity to aeration became less influential while the impact of inter-particular friction and inter-locking increased.

With the largest dosator outlet – Dosator 1, optimal performance was demonstrated by lactose grades 3 and 4 which both generate low AE and low to mid SE values, whereas powders with high AE values - Lactose 1 and 2 and high SE values - Lactose 5 all presented substandard filling behaviour suggesting that dosator 1 is influenced by both the AE and SE. At the other end of the spectrum, Dosator 4 was less sensitive to the AE, with Lactose 1, 2 and 3 generating acceptable performance despite high AE values, whereas Lactose 4 and 5, both of which presented low AE values but high SE values generated higher than acceptable RSD values indicating that this dosator is much more sensitive to mechanical forces.

The larger dosators allows for a greater interaction with air at the outlet, increasing inter-particle spacing which in turn reduces interparticular interactions, as such there is a stronger correlation with AE. With smaller outlets, bridging/arching across the outlet due to friction and mechanical interlocking, as quantified by SE, dominate.

These results demonstrate that even minor variations in powder properties can result in substandard filling performance, as such it’s important that powders are matched to the dosator. Powder rheology, and dynamic flow testing in particular, can support the development of more easily processed DPI formations and the selection of equipment that will consistently deliver the correct dose.