Insights into DPI sensitivity to humidity: An integrated in-vitro-in-silico risk-assessment

Abstract

Dry powder inhalers (DPIs) are known to be sensitive to humidity. Erroneous storage of these products by patients in high humidity environments, could present a potential risk. Therefore, an in-vitro-in-silico approach was utilized to assess the risk that patient erroneous storage might pose. For this two commercial DPIs containing lactose and budesonide (Easyhaler^® and Novolizer^®) were used. These were evaluated in respect to their physical solid-state and micrometric properties as well as their in-vitro aerodynamic performance. Testing was carried out at time 0, 14 and 28 days after storage at 60% RH and > 90% RH. Using in-silico modeling the potential impact of powder sensitivity to humidity on the biopharmaceutical performance of budesonide was evaluated. Results revealed that the physical and aerodynamic properties of the powders having a smaller carrier particle size and a higher amount of excipient fines were more notably affected. Use of in-vitro results as inputs for in-silico pharmacokinetic modeling showed that some changes in powder properties can have a potential impact on the pulmonary availability of budesonide. So, it appears that it is important to consider the impact that different product characteristics might have on the physical stability of powders against moisture and their subsequent biopharmaceutical performance.

Introduction

In recent years, dry powder inhalers (DPIs) have become a popular drug delivery system to the lung. When compared to other inhalers, DPIs possess certain patient advantages i.e. ease of handling, no need for hand-breath coordination and the absence of propellants [1,2]. The efficacy of DPIs is demonstrated by their ability to deliver reproducible fine particle doses (FPDs) of a given inhalable API to its site of action in the respiratory tract. Likewise, the FPD can be defined as the dose of an aerosolized drug with a particle size <5 μm [3,4]. Due to their small size, inhalable API particles possess high surface to mass ratios resulting in highly cohesive and poorly flowable powders [5]. So, different formulation strategies are used to achieve the desired FPDs. These include blending of the micronized drug with large excipient particles (carrier-based) or the use of particles (API and/or excipient) with tailored properties that result in improved flowability (carrier-free) [6,7]. However, certain parameters require special attention, in order to yield a stable DPI formulation. The intrinsic characteristics of the API need to be carefully considered (i.e. solubility, dissolution kinetics, lung permeability, pKa, etc.) as well as the intended pharmacokinetic (PK) and pharmacodynamic (PD) profile based on the targeted therapeutic effect [8,9]. Additionally, it is important to consider the physico-chemical properties of the API and carrier, such as their hygroscopicity, solid-state and particle size distribution (PSD) [10,11]. In particular, it is inevitable to characterize and understand how moisture-solid interactions of a hygroscopic solid can impact functional properties crucial to DPI powders such as flowability and dispersibility. The impact of moisture on DPI in-vitro aerodynamic performance has been the subject of various studies [[12], [13], [14], [15], [16], [17]]. For instance, it has been shown that budesonide inhalers, Easyhaler^® and Novolizer^® have different sensitivities towards humidity and this can result in distinct in-vitro aerodynamic performances of the two products [12]. Considering that 42% of patients incorrectly store their inhaler in high humidity environments such as bathrooms [12,18], it is important to understand how differences encountered in the in-vitro aerodynamic performance after storage at distinct relative humidities can correlate to the in-vivo efficiency of DPIs.

A reliable correlation between in-vitro drug release to the in-vivo profiles (in-vitro-in-silico correlation (IVIVC)) can significantly help to shorten DPI development timeline and improve product quality [19]. In this context, different in-silico methodologies have recently been recognized as an important branch of clinical drug development. Recently, the FDA has published a guideline supporting the use of physiologically based pharmacokinetic (PBPK) models in order to support decisions during the life cycle of a drug product [20]. In fact, this type of modeling approach has been widely investigated in recent years and successfully applied in the case of oral dosage forms [[21], [22], [23], [24]]. However, the development of reliable in-silico models for inhalation is still a challenge due to the lack of knowledge about the fate of drugs in the lung as well as the absence of predictive standardized in-vitro methodologies. (PB)PK models are expected to be able to accurately predict the exposure of the drug in the lungs, based in its solubility, dissolution, absorption, metabolism, distribution and elimination. Yet, in-vitro methods for the assessment of the DPIs, according to pharmacopeia, are focused on the testing of the aerodynamic particle size distribution (APSD) and dose uniformity. These are suitable for quality control, but can be over simplistic when replicating the in-vivo scenario.

In this study, an integrated in-vitro-in-silico assessment was used as a risk assessment approach to explore the influence of moisture on the different powder properties of two commercial DPIs and their predicted in-vivo performance. As already mentioned, moisture-induced changes in the in-vitro aerodynamic performance of Easyhaler^® and Novolizer^®, have been previously reported, so it was our objective to better understand the mechanisms behind them [12]. Budesonide was consider an appropriate model API for this study as its local metabolism in the lung is minimal and any changes in deposition can influence the systemic levels. Furthermore, budesonide presents dose-proportional pharmacokinetics in patients [25]. Therefore, is our hypothesis that different formulation and/or manufacturing parameters employed in producing these two budesonide DPIs can result in unique micro-/mesoscopic properties, surface characteristics and process inherited attributes that can lead to dissimilar responses to inappropriate storage conditions. Thus, to investigate this, the physical solid-state, PSD, nanoscale inner surface area and in-vitro APSD of formulations were assessed across different time points of storage at two distinct storage relative humidities (RHs). The different APSD profiles obtained over the period of storage were then used to investigate in-silico the influence of such changes on lung deposition and on the plasma concentration profiles of budesonide. The results are discussed postulating the possible mechanisms governing in-vitro changes and the potential risk to altered product performance.