(572c) Investigation of Carrier Type and Blending Parameters on the Performance of a Dual-Combination DPI: A Comparative in-Vitro-in-Silico Study

Powder drugs can be orally delivered to the lung using dry powder inhalers (DPIs). Two main types of DPI formulation exist namely: (i) carrier‑free, consisting of soft active pharmaceutical ingredients (APIs) agglomerates and/or excipients with tailored properties and (ii) carrier‑based, comprising a physical mixture of coarse carrier particles and APIs [1]. Most of the marketed formulations are carrier‑based and offer the advantage of improved fine particle fraction (FPF) delivered to the patient, i.e. the quantity of an aerosolized API with particle size below 5 µm able to reach the lung.

It is important to understand the complex interplay between different parameters affecting the DPI performance, e.g. selection of carrier particles, API load, powder blending conditions, which are crucial to guarantee a stable dry adhesive mixture. However, one has also to consider how the formulation, and its respective performance will be impacted by the device and associated dispersion forces. To yield adhesive mixtures, the carrier needs to be properly blended with fine API particles by creating API-carrier interaction forces [2]. These subsequently need to be broken, also with the help of the DPI device, to detach the API from the carrier and allow its deposition in the deep lung.

Physiologically based pharmacokinetic modelling (PBPK) can help to better understand in-vivo DPI performance in-silico. By using PBPK, it is possible to gain, a priori understanding on how formulation parameters can affect the API behavior in-vivo. Therefore, this study aimed to use a comparative in-vitro-in-silico approach in which we investigated (i) how different carriers (α‑lactose monohydrate (αLH) and mannitol (MAN) as potential alternative and (ii) blending parameters will impact the blend uniformity and aerosolization performance of formulations, as well as (iii) the impact of the device selection on the DPI performance. The data was used as an input in developed PBPK models and the potential impact on the predicted pharmacokinetic (PK) parameters studied.

Materials and Methods

Materials

Budesonide (BUD) and formoterol fumarate dihydrate (FF) were kindly provided by Chemo (Laboratorios Liconsa S.A., Spain) and Chiesi Farmaceutici S.p.A. (Italy), respectively. As carriers, the InhaLac® 230 (αLH, Meggle, Germany) and Parteck® M DPI (MAN, Merck, Germany) were used.

Blends preparation and characterization

The investigation of the impact of different carrier types and process parameters on the blending performance was performed by preparing αLH- and MAN-based blends of a model formulation (comprising 1 wt% of BUD and 0.02 wt% of FF) under different conditions. A total of 10 blends were prepared in a laboratory scale resonant acoustic mixer (LabRAM I ResonantAcoustic® Mixer, Resodyn^TM Acoustic Mixers, USA) using a 236 mL vessel using the following blending parameters: (i) 30 s at three acceleration levels (i.e. 30, 45 and 60 g) and (ii) 90 s at two acceleration levels (i.e. 30 and 60 g). The blend uniformity was evaluated by measuring the content of both APIs in 10 samples, taken within the powder bed at different positions. BUD and FF content was quantified using high performance liquid chromatography (HPLC). The blend uniformity was considered acceptable for RSD <10% and those blends were further analyzed for their aerodynamic performance.

Aerodynamic performance of blends

The impact of inhaler type on the aerosolization behavior was evaluated using three selected blends with acceptable blend uniformity. The assessment of aerodynamic performance was performed using the Next Generation Impactor (NGI; Copley Scientific, UK) comparing capsule‑based (Cyclohaler®) and reservoir‑based (Novolizer®) inhalers. All experiments were carried out in triplicates and at a flow of 60 L/min. For Cyclohaler®, formulations were loaded into 20 hydroxypropyl methylcellulose (HPMC) capsules size 3, whereas in the case of Novolizer®, 40 actuations sufficed.

PBPK model development and predictions

Based on the reported in-vivo data following inhalation of Symbicort Turbuhaler®[3], the PBPK models for FF and BUD were developed using MATLAB® (R2021, MathWorks, USA). The models were developed based on the previously published model by Boger et al[4]. To predict the deposition patterns of FF and BUD [5], their aerodynamic size distributions and spirometry profiles were used as inputs in Multiple-Path Particle Dosimetry (MPPD; Applied Research Associates, Inc.).

Results and Discussion

Powders used for inhalation can either contain a single drug or a combination of two or more APIs. In this study, we selected a combination product of BUD and FF in the same ratio as in commercial products (i.e. Symbicort). Evaluation of blend uniformity was performed following each blending experiment. For αLH containing blends, a blending time of 30 s in combination with an acceleration of 60 g resulted in satisfactory homogeneity (Figure 1). For MAN based blends instead, homogeneity was achieved following 90 s of blending at both acceleration rates (i.e. 30 and 60 g). Interestingly, different blending trends of BUD and FF were observed when comparing αLH and MAN as carriers. When using αLH as carrier, the homogeneity of FF was improved by higher blending times and accelerations, contrary to BUD where de-mixing was observed. For mixtures containing MAN, a higher mixing time was necessary to evenly distribute cohesive FF particles, independently of the acceleration level, while BUD homogeneity was dependent on both time and the acceleration levels. The observed effect can be attributed to a higher surface area of MAN, hence for the same mass of API, longer time was needed to evenly distribute it over the bigger carrier particles. Additionally, the mixing behavior and efficiency might be impacted by difference in loading of the two API.

Ideally, the DPI device will allow the patients to generate the preferred airflow e.g. 60 L/min and will deliver constant and high FPF to the patients. Generally, two main types of DPI devices are single-dose (e.g. capsule-based) and multi-dose (e.g. reservoir). We selected Cyclohaler® (capsule) and Novolizer® (reservoir) for our study. Three blends with satisfactory homogeneity (i.e., one αLH mixture obtained after blending at 60 g for 30 s and two MAN mixtures prepared at 30 and 60 g for 90 s) were further evaluated in terms of their aerodynamic performance. Interestingly, for all blends a higher FPF was seen when the reservoir‑based inhaler was used, i.e. the FPF values were 30% for the αLH‑based blend and 40% for the MAN ones. For capsule‑based systems, the FPF was in fact around 23‑24% for both αLH‑ and MAN‑based blends.

The mass mean aerodynamic diameter (MMAD) of the delivered particles was around 1 µm for MAN containing blends compared to the αLH ones, where it was around 2 µm, for both inhalers. These observed differences in aerodynamic performance were further evaluated by use of PBPK models. It is expected that different deposition patterns could potentially impact the in-vivo performance of the combination formulation. The models showed that distinct changes in API deposition can occur in case when different carriers are used, resulting in different bioavailability of the two drugs. Additionally, the selection of the device showed to be a critical parameter that impacting the performance of the formulations.

Conclusion

The present study highlights the importance of a proper selection of carrier type, blending process parameters and inhaler device for the design of an efficient DPI combination product. Compared to αLH, the selection of MAN as carrier in combination with a higher blending time and a reservoir based type of inhaler resulted in the improved aerodynamic performance of BUD and FF. Utilization of PBPK models demonstrated how a combined in-vitro-in-silico approach can be used to understand a priori the impact of different formulation approaches on the in-vivo behavior of DPIs.

References

[1] D. Traini, Inhalation Drug Delivery. 2013.

[2] W. Kaialy, “On the effects of blending, physicochemical properties, and their interactions on the performance of carrier-based dry powders for inhalation — A review,” Adv. Colloid Interface Sci., vol. 235, pp. 70–89, 2016.

[3] S. Lähelmä et al., “Equivalent lung dose and systemic exposure of budesonide/formoterol combination via easyhaler and turbuhaler,” J. Aerosol Med. Pulm. Drug Deliv., vol. 28, no. 6, pp. 462–473, 2015.

[4] E. Boger and O. Wigström, “A Partial Differential Equation Approach to Inhalation Physiologically Based Pharmacokinetic Modeling,” CPT Pharmacometrics Syst. Pharmacol., vol. 7, no. 10, pp. 638–646, 2018.

[5] Á. Farkas et al., “Numerical simulation of emitted particle characteristics and airway deposition distribution of Symbicort® Turbuhaler® dry powder fixed combination aerosol drug,” Eur. J. Pharm. Sci., vol. 93, pp. 371–379, 2016.