High-efficiency dry powder aerosol delivery to children: Review and application of new technologies

Highlights

• Key aerosol and device technologies are presented that maximize delivery to pediatric airways:

• Using these technologies gave <11% total loss from the nasal cavity to the third bifurcation branch.

• The aerosol grows from an MMAD of 1.5–3.2 μm, after 2 s residence time.

• 90% of aerosol size increase occurs in first ~0.6 s.

Abstract

While dry powder aerosol formulations offer a number of advantages, their use in children is often limited due to poor lung delivery efficiency and difficulties with consistent dry powder inhaler (DPI) usage. Both of these challenges can be attributed to the typical use of adult devices in pediatric subjects and a lack of pediatric-specific DPI development. In contrast, a number of technologies have recently been developed or progressed that can substantially improve the efficiency and reproducibility of DPI use in children including: (i) nose-to-lung administration with small particles, (ii) active positive-pressure devices, (iii) structures to reduce turbulence and jet momentum, and (iv) highly dispersible excipient enhanced growth particle formulations. In this study, these technologies and their recent development are first reviewed in depth. A case study is then considered in which these technologies are simultaneously applied in order to enable the nose-to-lung administration of dry powder aerosol to children with cystic fibrosis (CF). Using a combination of computational fluid dynamics (CFD) analysis and realistic in vitro experiments, device performance, aerosol size increases and lung delivery efficiency are considered for pediatric-CF subjects in the age ranges of 2–3, 5–6 and 9–10 years old. Results indicate that a new 3D rod array structure significantly improves performance of a nasal cannula reducing interface loss by a factor of 1.5-fold and produces a device emitted mass median aerodynamic diameter (MMAD) of 1.67 μm. For all ages considered, approximately 70% of the loaded dose reaches the lower lung beyond the lobar bronchi. Moreover, significant and rapid size increase of the aerosol is observed beyond the larynx and illustrates the potential for targeting lower airway deposition. In conclusion, concurrent CFD and realistic in vitro analysis indicates that a combination of multiple new technologies can be implemented to overcome obstacles that currently limit the use of DPIs in children as young as two years of age.

Introduction

Compared with other pharmaceutical aerosol approaches, dry powder inhalers (DPIs) offer a number of advantages in respiratory drug delivery including formulation stability, rapid dose administration, the potential for high inhaled doses, minimal cleaning and automatic coordination between inhalation and dose delivery (De Boer et al., 2017; Islam & Cleary, 2012; Smith et al., 2010; Weers et al., 2010, 2019). Despite these advantages, the use of DPIs with children is often challenging as they may not be able to perform adequate inhalation maneuvers (Devadason, 2006; Goralski & Davis, 2014; Lexmond, Hagedoorn, Frijlink, Rottier, & de Boer, 2017), and also due to high mouth-throat (MT) depositional loss, low lung delivery efficiency and often high intersubject variability (Below, Bickmann, & Breitkreutz, 2013; Devadason et al., 1997; Lindert, Below, & Breitkreutz, 2014; Ruzycki, Golshahi, Vehring, & Finlay, 2014). For commercial products, upper airway depositional losses of 80% or higher of the emitted dose are common for children in the age range of 5 up to approximately 14 years (Below et al., 2013; Devadason et al., 1997; Lindert et al., 2014; Ruzycki et al., 2014). Across a number of current dry powder inhalers (DPIs), pediatric lung delivery efficiencies are in a range of 5–30% of nominal dose, based on in vivo (Devadason et al., 1997) and realistic in vitro experiments (Below et al., 2013; Lindert et al., 2014; Ruzycki et al., 2014). Potentially more significant than the low dose delivery efficiency is the high intersubject variability associated with pediatric DPI use. For example, using the correlations for pediatric patients from Golshahi, Noga, Thompson, and Finlay (2011), applied to respirable aerosols at typical flow rates for a child (2.5 μm and 15 LPM), results in nasal cavity deposition fractions ranging from 3.4 to 24.3%. Furthermore, Lindert et al. (2014) reported a 2-fold difference in lung dose from a DPI based on inhalation waveform conditions even without varying the airway anatomy. Devadason et al. (1997) reported coefficients of variation near 100% for pediatric DPI use by 3 to 5-year-old children. As with adults, reducing extrathoracic depositional loss with an inhaler is expected to reduce variability in the lung delivery of the medication (Borgstrom, Olsson, & Thorsson, 2006).

Beyond the physics of efficient aerosol formation and extrathoracic depositional loss, children also frequently have difficulty in using passive DPI devices that operate on negative inhalation pressure and are almost always designed for adults (Everard, 1996, 2004; Goralski & Davis, 2014). It is not until approximately 6 years of age that children can reliably perform a correct DPI inhalation maneuver with a passive device (Devadason, 2006; Goralski & Davis, 2014; Lexmond et al., 2017). A recent study by Lexmond et al. (2014) in children approximately 5–12 years old with an oral DPI simulator reported that 90% of inhalations were associated with oral obstruction, arising from the tongue and cheeks, during negative pressure inhalation against the variable resistance device. Lexmond et al. (2014) also observed that for 5-year-old subjects, only half of the trained children could successfully perform a passive DPI inhalation maneuver. Furthermore, between 10 and 44% of the trained pediatric subjects exhaled into the DPI simulator, which is expected to adversely affect aerosolization of the powder due to moisture exposure.

While the challenges of delivering dry powder aerosols to children are daunting, it is our opinion that both infants (0–2 years) and children (2–12 years) can significantly benefit from a properly designed DPI that is capable of reliable high-efficiency aerosol delivery to the lungs. Key problems related to DPI aerosol use in children can be characterized as: (i) exclusive use of oral administration with passive (adult) devices that operate on negative inhalation pressure, (ii) turbulence and jet momentum that result in high device and upper airway depositional aerosol loss, and (iii) implementation of relatively large and static aerosol sizes. Considering oral administration, it is generally held that oral aerosol inhalation produces significantly higher lung delivery efficiency compared with nose-to-lung delivery (Chua et al., 1994; Everard, Hardy, & Milner, 1993). While this is generally true, it is often not considered that for sufficiently small particles at low flow rates, oral or nasal deposition loss becomes similar and can generally be held below 5 or 10%. Furthermore, nose-to-lung delivery offers some advantages such as treating the entire respiratory airways and enabling use in infants and young children when coupled with positive-pressure devices. In contrast, considering pediatric inhalation through passive DPIs, the increased flexibility of pediatric upper airways and the oral obstructions observed by Lexmond et al. (2014) are expected to produce high depositional loss in vivo, potentially greater than predicted with realistic (but rigid) in vitro models. As the negative pressure generated by passive devices was observed to narrow the airways, it is reasonable to expect that a positive-pressure device would expand the flexible upper airways and thereby improve lung penetration of the aerosol. Use of a positive-pressure aerosol source may improve device reproducibility in the pediatric population.

It is well known that turbulent jets from DPIs significantly increase both device and extrathoracic depositional loss (DeHaan & Finlay, 2001, 2004; Longest, Tian, Delvadia, & Hindle, 2012; Longest, Tian, Walenga, & Hindle, 2012; Matida, DeHaan, Finlay, & Lange, 2003; Tian, Longest, Su, Walenga, & Hindle, 2011). As recently described by Weers et al. (2019), this explains why simply increasing device resistance (which decreases device flow rate) does not reduce MT deposition. Similarly, inhaling a smaller particle size without a properly designed inhaler will not necessarily improve lung delivery efficiency (Longest, Hindle, Das Choudhuri, & Xi, 2008).

Finally, new reportedly small particle DPIs do not produce a small “apparent aerosol diameter” based on extrathoracic depositional loss (Weers et al., 2019). For example, the recently available “extrafine particle” NEXThaler DPI has an impactor measured mass median aerodynamic diameter (MMAD) of 1.5 μm, but an apparent aerosol size of 4.7 μm based on human subject deposition data (Virchow et al., 2018). The reasons for the discrepancy between the measured MMAD and the apparent aerosol diameter are most likely turbulence and jet or spray momentum emanating from the device, as well as current pharmaceutical testing methods that exclude the induction port deposition fraction, which is often 50% or more of the aerosolized dose, in size determination. Moreover, a static aerosol size may not perform as well as an aerosol that is capable of entering the airways with a relatively small size and then increasing in size within the airways in a controlled manner (Hindle & Longest, 2010; Longest, Tian, Li, Son, & Hindle, 2012; Tian, Longest, Li, & Hindle, 2013).

In order to overcome the primary limitations associated with dry powder aerosol delivery to children, as well as many limitations that are also present in adults, a number of technologies have recently been developed or re-discovered and substantially extended. These technologies include:

• Nose-to-lung aerosol administration with sufficiently small particles (Bass, Boc, Hindle, Dodson, & Longest, 2019; Golshahi et al., 2011)

• Use of active positive-pressure DPIs (Farkas, Bonasera, Bass, Hindle, & Longest, 2020; Farkas et al., 2018b, Farkas et al., 2018c)

• Patient interfaces that reduce turbulence and jet momentum effects without substantially increasing particle depositional loss (Bass & Longest, 2020; Farkas et al., 2020; Longest, Son, Holbrook, & Hindle, 2013)

• Highly dispersible spray-dried powder formulations (Chan, 2006; Vehring, 2008; Vehring, Foss, & Lechuga-Ballesteros, 2007; Weers & Miller, 2015) that change size within the airways (Hindle & Longest, 2012; Son, Longest, & Hindle, 2012)

The following sections review recent progress in each of these areas. The use of concurrent CFD and in vitro analysis is then summarized, which has enabled rapid development of these new technologies. Following this review, concurrent analysis is implemented to explore the simultaneous application of these new technologies in order to improve the trans-nasal delivery of dry powder aerosols to children with cystic fibrosis across an age range of 2–10 years old.

A number of previous studies have characterized the nasal deposition of ambient inhaled particles in pediatric nasal models (Golshahi et al., 2011; Golshahi & Finlay, 2012; Golshahi, Noga, & Finlay, 2012; Golshahi, Tian, Noga, & Finlay, 2013; Xi, Berlinski, Zhou, Greenberg, & Ou, 2012; Xi, Si, Kim, & Berlinski, 2011; Xi, Si, Zhou, Kim, & Berlinski, 2014). The term ambient particles refers to an aerosol that is inhaled from the environment without jet or spray momentum, as typically occurs with a pharmaceutical aerosol generation device, and without a patient interface, such as a mouthpiece (MP) or nasal cannula (NC). From these studies, it is well known that particles with a sufficiently small impaction parameter (d_a²Q – where d_a is the particle aerodynamic diameter and Q is the inhalation flow rate) produce low depositional loss in nasal airways, even for infants and children. For example, the d_a²Q impaction parameter for a 1.7 μm aerosol delivered at 15 LPM is approximately 40 μm² L/min, which, based on the study of Golshahi et al. (2011) across 14 pediatric nasal models (4–14 years old), corresponds to nasal deposition in the range of approximately 0–12%. In contrast, employing a typical DPI aerosol size of 6 μm at a flow rate of 15 LPM results in a predicted nasal deposition range of 25–75% based on the data of Golshahi et al. (2011).

The high depositional losses of ambient particles are consistent with most nebulized drug delivery studies employing pharmaceutical aerosols with a droplet size that is approximately 5 μm or greater and a mask or nasal cannula interface. For a mesh nebulized aerosol, El Taoum, Xi, Kim, and Berlinski (2015) reported lung doses of 0–3% of nebulized drug in an anatomically realistic nasal cavity model of a 5-year-old child that included cyclic respiration and a variety of patient interfaces. Other bench-top testing studies that have considered nebulized aerosol delivery with a mask interface have also reported approximately 10% or less lung delivery efficiency in pediatric subjects (Lin, Harwood, Fink, Goodfellow, & Ari, 2015; Smaldone, Sangwan, & Shah, 2007). Ari, Harwood, Sheard, Dailey, and Fink (2011) considered pediatric trans-nasal aerosol delivery with a nasal cannula interface without a nasal model and also reported lung delivery efficiency of approximately 10% and below (relative to total dose).

While it is relatively well known that smaller particle size can significantly improve the delivery of pharmaceutical aerosols to infants and children, this approach has not been widely applied. One exception is the recent in vitro study of Bass, Farkas, and Longest (2019) in infants, which demonstrated high efficiency lung delivery of a pharmaceutical aerosol using a mesh nebulizer, custom mixer-heater (Spence, Longest, Wei, Dhapare, & Hindle, 2019), and excipient enhanced growth (EEG) (Hindle & Longest, 2012) aerosol formulation. The mixer-heater device was used to heat the gas stream to a temperature that remained safe for direct inhalation and reduced the aerosol mass median aerodynamic diameter (MMAD) to approximately 1.5 μm, indicating dry particle formation. A streamlined nasal cannula interface was used to further reduce device system and nasal airway depositional loss (Longest, Golshahi, & Hindle, 2013). In vitro experiments and corresponding computational fluid dynamics (CFD) simulations demonstrated >90% delivery efficiency of the nebulized dose to a tracheal filter (Bass, Boc, et al., 2019).

Reasons that small particle aerosols may not commonly be used for pharmaceutical aerosol delivery to children include: (i) low dose delivery rates, (ii) difficulty in generating the small aerosol size, and (iii) high potential to exhale the dose. Dry powder inhalers can frequently be used to rapidly generate and deliver high aerosol doses (Farkas et al., 2015, Farkas et al., 2018a; Young et al., 2013), but typically have a relatively large apparent aerosol diameter with high extrathoracic losses (Newman & Busse, 2002; Weers et al., 2019). The previous study of Farkas et al. (2020) demonstrated the generation and delivery of a small particle dry powder aerosol through a nasal interface and pediatric NT model at relatively high efficiency. As described further below, the positive-pressure air-jet DPI with a spray-dried powder formulation effectively generated a small aerosol size (approximately 1.7 μm) (Farkas et al., 2020). The device actuation speed was fast (<5 s) resulting in a high dose delivery rate (Farkas et al., 2020). Furthermore, an EEG particle formulation was used to reduce the potential for exhalation of the spray-dried aerosol and to enable targeted drug delivery (Tian et al., 2013; Tian, Hindle, & Longest, 2014).

The vast majority of DPIs on the market are passive devices, which form an aerosol under negative pressure in response to a user's inhalation through the device. In contrast, active devices use an energy source external to the user to form the aerosol. As reviewed by Longest, Farkas, Bass, et al. (2020), positive-pressure active devices implement an external gas source to aerosolize the powder, which can be supplied by an air-syringe, manual ventilation bag, or compressed air electromechanical system. Depending on the volume of gas used, these DPIs can be classified as high (≥200 ml) or low (<200 ml) actuation air-volume (AAV) devices. Active devices are often perceived as having the disadvantage of increased complexity and cost due to the requirement for an external gas source. However, a significant advantage of positive-pressure devices may be their ability to deliver dry powder aerosol during invasive and non-invasive mechanical ventilation (Farkas et al., 2018c, 2018a; Feng, Tang, Leung, Dhanani, & Chan, 2017; Okuda, Tang, Yu, Finlay, & Chan, 2017; Pornputtapitak, El-Gendy, Mermis, O'Brein-Ladner, & Berkland, 2014; Walenga, Longest, Kaviratna, & Hindle, 2017), and their ability to administer both the aerosol and a full inhalation breath, which can be beneficial in administering dry powder aerosol to infants (Howe, Hindle, Bonasera, Rani, & Longest, 2020; Laube, Sharpless, Shermer, Sullivan, & Powell, 2012) and young children (Farkas et al., 2020).

Considering dry powder aerosol delivery to pediatric subjects, positive-pressure DPIs that deliver the aerosol and a full inhalation breath can overcome a number of previously observed limitations. First, use of a consistent positive-pressure gas source to form and deliver the aerosol should significantly reduce inter and intra-subject variability in drug delivery, especially if extrathoracic depositional loss can also be reduced. Secondly, positive-pressure operation provides the option of oral or nasal lung delivery of the aerosol. Potential advantages of trans-nasal delivery include administering pharmaceutical aerosol to infants and children that are too young to use a mouthpiece (approximately 2–3 years old) and the ability to treat the nasal and lung airways simultaneously. Thirdly, positive-pressure gas delivery will expand rather than collapse the extrathoracic airways, which should improve lung delivery of the aerosol. Providing a known volume of gas delivery can be used to assist with deep lung inhalation and expansion of constricted or obstructed tracheobronchial airways, thereby enabling improved targeting of the deep lung regions and delivery to diseased airways. Finally, positive pressure aerosol delivery requires forming a sealed connection with the lungs via the extrathoracic region. This sealed system prevents the user from exhaling through the powder containment region, which can degrade powder performance, and can be used to encourage a brief breath-hold to improve lung retention of the aerosol.

Our group has recently developed a positive-pressure air-jet DPI concept for efficient aerosol generation and delivery to adults, children, and infants. As described in previous studies (Boc, Farkas, Longest, & Hindle, 2018; Farkas et al., 2018a, 2018b, 2018c), the air-jet DPI implements a small diameter inlet airflow passage, aerosolization chamber, and small diameter outlet aerosol flow passage. Positive-pressure gas passes through the inlet airflow passage and forms a high-speed turbulent jet within the aerosolization chamber (Longest & Farkas, 2019). Secondary flow velocities formed by the high-speed jet are used to initially fluidize the powder. As the fluidized powder enters the high speed jet region, additional powder deaggregation occurs (Longest & Farkas, 2019). The small diameter outlet orifice serves to both help form the secondary velocities and allow passage of sufficiently deaggregated particles out of the aerosolization chamber. Using this approach, AAVs of 10 ml and lower have been shown to effectively aerosolize 10 mg powder masses in devices that were designed to be integrated with a ventilation system, which required a small AAV so as to not increase the ventilation volume (Boc et al., 2018; Farkas et al., 2018a, 2018b, 2018c). For pediatric drug delivery, Farkas, Hindle, Bass, and Longest (2019) developed a positive-pressure pediatric air-jet DPI that was operated with a ventilation bag or compressed gas supply with 750 ml of air, in order to aerosolize the powder and provide a full inhaled breath for a 5-year-old child. The AAV selected for 5-year-old children was based on adult inhalers typically being tested at 50–75% of total lung capacity (TLC). For a 5-year-old child, typical TLC is 1.55 L (ICRP, 1994), such that the 750 ml AAV is at the lower end of the 50–75% TLC range used for adults. Using a highly dispersible spray-dried formulation (Son, Longest, Tian, & Hindle, 2013), the best case pediatric air-jet DPI produced an aerosol MMAD <1.75 μm and a fine particle fraction (<5 μm) ≥90% based on emitted dose. Actuation with the ventilation bag enabled lung delivery efficiency through the nasal and oral interfaces to a tracheal filter of 60% or greater, based on loaded dose. In both oral and nose-to-lung administrations, extrathoracic depositional losses were <10% (Farkas et al., 2019). Effective use of the positive-pressure device requires training for the caregiver to actuate the device correctly and for the patient to allow the device to inflate their lungs during use.

Computational fluid dynamics (CFD) studies of aerosolization within the air-jet DPI have revealed some interesting characteristics. At both high and low AAVs, increasing turbulence increases emitted dose (which is advantageous), but also increases MMAD (which is typically detrimental for efficient lung delivery) (Bass, Boc, et al., 2019; Longest & Farkas, 2019; Longest, Farkas, Bass, & Hindle, 2019). The direct relationship between internal device turbulence and MMAD is a unique characteristic of the air-jet system as most other aerosol generation units are assumed to have the opposite behavior. This behavior was attributed to a two stage aerosolization process of initial fluidization of the powder followed by turbulent deaggregation of fluidized agglomerates (Longest & Farkas, 2019). Excess turbulence was viewed to fluidize the powder too rapidly leaving less time for secondary turbulent deaggregation. Provided that sufficient emitted dose can be maintained, the air-jet DPI therefore performs better with lower flows and less turbulence (Longest & Farkas, 2019; Longest, Farkas, et al., 2019), which are ideal characteristics for efficient aerosol administration to infants and children (Bass, Farkas, & Longest, 2019). Furthermore, devices tend to produce a direct linear relationship between emitted dose and MMAD, i.e., higher emitted dose is directly proportional to higher MMAD (Bass, Farkas, & Longest, 2019). CFD and in vitro aerosol characterization can be used to identify and select device designs with beneficial emitted dose and MMAD relationships (Bass, Farkas, & Longest, 2019; Howe et al., 2020).

DPIs typically employ high turbulence and small diameter flow passages leading to the mouth-throat region in order to deaggregate dry powder formulations and form an inhalable aerosol (Coates, Chan, Fletcher, & Raper, 2005, 2006; Coates, Fletcher, Chan, & Raper, 2004; Coates, Chan, Fletcher, & Chiou, 2007; DeHaan & Finlay, 2001; Fenton, Keating, & Plosker, 2003; Shur et al., 2012). While some inhalable dose fraction can be formed with this method, depositional losses in the device and mouth-throat (MT) region are typically high (DeHaan & Finlay, 2001; Ilie, Matida, & Finlay, 2008; Longest, Tian,; Delvadia, Longest, & Byron, 2012; Longest, Tian, Walenga, & Hindle, 2012; Matida et al., 2003; Tian, Longest, Su, Walenga, & Hindle, 2011), due to increased impaction deposition and turbulence dispersion. Within the DPI, different structures and airflow passage designs are implemented to generate the turbulence and particle aggregate break-up mechanisms that are needed to deaggregate the powder (Coates et al., 2004, 2006, 2007; Shur et al., 2012; Wong et al., 2010, 2011a, 2011b). Longest, Son, Holbrook, and Hindle (2013) implemented a combination of in vitro analysis and CFD simulations to evaluate eight different aerosolization units including a standard constricted tube, impaction surface, 2D mesh, inward radial jets, and newly proposed 3D rod arrays. It was determined that for a set amount of input energy (applied as a negative pressure drop across the device) a 3D rod array with unidirectional rods was most effective at aerosolizing the powder.

While the air-jet DPI improved MMAD with lower internal turbulence, a potential disadvantage is the small diameter jet of high velocity aerosol exiting the device, which can lead to unnecessary impaction loss in the patient interface and extrathoracic airways. Farkas et al. (2020) observed relatively low MT deposition with an adult air-jet DPI (<10% of emitted dose), but unexpected powder deposition on the back of the throat as a result of the high turbulence jet. The CFD study of Bass & Longest (2020) explored pediatric patient interfaces that could reduce the effect of the high-intensity turbulent jet that exits the air-jet DPI. Internal structures within the interface that were considered included non-smooth surfaces, rapid and stepped expansions, impaction surfaces and various 3D rod array designs. CFD results revealed that a combination of a 3D rod array with a rapidly expanding interface in the region of the rod array best dissipated the turbulent jet while minimizing depositional loss in the mouthpiece (Bass & Longest, 2020). For oral aerosol administration, the optimal flow passage compared with previous design candidates reduced device, mouthpiece, and mouth-throat deposition efficiencies by factors of 8-, 3-, and 2-fold, respectively (Bass & Longest, 2020). For nose-to-lung aerosol administration, the optimal flow pathway compared with previous designs reduced device, nasal cannula, and nose-throat deposition by 16-, 6-, and 1.3-fold, respectively (Bass & Longest, 2020).

Following the CFD study of Bass et al. (2020), Farkas et al. (2020) considered pediatric oral aerosol delivery with a realistic in vitro MT airway model using an air-jet DPI and MP interface, which included a 3D rod array to improve secondary break-up of the aerosol and dissipate the turbulent jet before entering the MT region. A new vertical aerosolization chamber was also considered that was expected to be less sensitive to larger powder mass loadings. Devices were loaded with 10 mg doses of a spray dried formulation and actuated with positive pressure using a flow rate of 10–20 L/min and an air volume of 750 ml consistent with a 5-year-old child. Inclusion of the 3D rod array in the MP was shown to further reduce the aerosol size to an MMAD of <1.7 μm without significantly increasing aerosol loss in the device. Best case device and MP combinations produced <2% MT depositional loss and >70% lung delivery efficiency (based on loaded dose) in a realistic in vitro pediatric MT geometry. Aerosol delivery with a 3D rod array and trans-nasal delivery, which is the focus of the current study, has not been previously considered experimentally.

Considering the small airway diameters and relatively low inhalation volumes involved with pediatric aerosol delivery, highly dispersible powder formulations are important. Spray drying techniques are one of the most flexible and practical methods to generate large quantities of highly dispersible powder formulations (Chan, 2006; Vehring, 2008; Vehring et al., 2007; Weers & Miller, 2015). These spray-dried formulations typically require the application of particle formulation engineering techniques to make them highly dispersible, including the formation of large porous particles (Edwards, Ben-Jebria, & Langer, 1998; Edwards et al., 1997), PulmoSpheres™ (Geller, Weers, & Heuerding, 2011; Weers & Miller, 2015), or inclusion of dispersion enhancers (Feng et al., 2011; Son et al., 2013). Using these approaches, previous studies have reported the formation of small-particle dry powder aerosols as needed for efficient pediatric drug delivery (Edwards et al., 1997, 1998; Son et al., 2013; Weers et al., 2015). Nevertheless, the use of a static particle size for efficient lung delivery of an aerosol has limitations. For example, particles that are small enough to avoid extrathoracic deposition may also lack sufficient inertia for targeting deposition in the tracheobronchial airways, or may be exhaled. Increasing particle size results in higher extrathoracic depositional loss and a net reduction in lung delivery. As demonstrated by Walenga and Longest (2016), conventional MDI and DPI devices deposit approximately 1% of the inhaled dose within the entire small tracheobronchial region of adult lungs.

To address the limitations associated with the delivery of static aerosol sizes, the concept of controlled condensational growth has recently been developed (Hindle & Longest, 2010, 2012; Longest & Hindle, 2010, 2011). In this approach, an aerosol is delivered to the respiratory tract with a sufficiently small size to minimize device and upper airway deposition. Droplet size increase through condensational growth allows for retention of the aerosol, which without growth would likely be exhaled. Techniques to produce the required size increase include enhanced condensational growth (ECG) and excipient enhanced growth (EEG). In the ECG approach, the aerosol is delivered with air saturated with water vapor a few degrees above body temperature, which creates supersaturated relative humidity conditions in the lungs to foster condensational growth of the droplets (Hindle & Longest, 2010; Tian, Longest, Su, & Hindle, 2011). With EEG, formulated particles contain a combination of a drug and a hygroscopic excipient, and the natural relative humidity in the lungs provides the water vapor source for aerosol size increase (Hindle & Longest, 2012; Longest & Hindle, 2011). Combination particles that contain both the therapeutic agent and a hygroscopic excipient, in order to generate aerosol size increase in the airways, are referred to as EEG formulations (Hindle & Longest, 2012; Son et al., 2013).

The new approach of controlled condensational growth has been successful at improving lung delivery for orally-administered aerosols (Hindle & Longest, 2010; Son et al., 2013; Tian et al., 2013; Tian, Longest, Su, & Hindle, 2011) and for nose-to-lung delivery with nebulizer generated aerosols (Golshahi, Tian, et al., 2013; Longest, Tian, & Hindle, 2011) based on CFD simulations and realistic in vitro experiments. Considering nose-to-lung delivery with a nebulizer, Golshahi, Tian, Noga, & Finlay, 2013 previously demonstrated that both EEG and ECG approaches reduced cannula and nasal depositional losses by an order of magnitude and delivered approximately 80% of the loaded dose to the lungs with steady flow. Using an aerosol mixer-heater system (Longest, Walenga, Son, & Hindle, 2013) combined with a mesh nebulizer, Golshahi, Longest, Azimi, Syed, and Hindle (2014) demonstrated that synchronizing the aerosol delivery with patient breathing was important to achieve high efficiency aerosol delivery (>70%) past the nose and to the lung in an adult high flow nasal cannula (HFNC) system. Tian et al. (2014) reported CFD simulations of nose-to-lung administered EEG aerosols through the conducting tracheobronchial (TB) airways and found minimal nasal depositional loss, substantial droplet growth, and a significant dose enhancement to the lower TB region of 40-fold compared with marketed inhalers.

As reviewed or highlighted in multiple publications, both CFD and realistic in vitro experiments can be used to better understand and significantly improve multiple aspects of respiratory drug delivery (Byron et al., 2010; Carrigy, Ruzycki, Golshahi, & Finlay, 2014; Delvadia, Hindle, Longest, & Byron, 2013; Delvadia et al., 2012; Delvadia, Longest, Hindle, & Byron, 2013; Delvadia, Wei, Longest, Venitz, & Byron, 2016; Longest, Bass, et al., 2019; Longest & Holbrook, 2012; Ruzycki, Javaheri, & Finlay, 2013; Wei et al., 2018; Wong, Fletcher, Traini, Chan, & Young, 2012). Considering CFD, useful insights and design optimizations can be made in the areas of particle engineering (Longest, Farkas, Hassan, & Hindle, 2020), aerosol dispersion (Bass, Farkas, & Longest, 2019; Longest & Farkas, 2019; Longest, Farkas, et al., 2019; Wong et al., 2011b; Wong et al., 2012), device and interface design (Coates et al., 2007; Coates, Chan, Fletcher, & Raper, 2005; Coates et al., 2006; Coates et al., 2004; Coates, Fletcher, Chan, & Raper, 2005; Hindle & Longest, 2013; Longest & Hindle, 2009b; Shur et al., 2012) and airway transport and deposition (Lambert, O'Shaughnessy, Tawhai, Hoffman, & Lin, 2011; Longest, Tian, Delvadia, & Hindle, 2012; Longest, Tian, Walenga, & Hindle, 2012; Tian et al., 2014; Tian et al., 2013; Tian, Longest, Su, Walenga, & Hindle, 2011). Similarly, realistic in vitro experiments can be conducted to capture extrathoracic depositional loss, aerosol size distribution entering the lungs, transport through the upper tracheobronchial airways, and transport within lower airway segments including the alveolar region (Berg & Robinson, 2011; Berg, Weisman, Oldham, & Robinson, 2010; Byron et al., 2010; Delvadia, Hindle, Longest, & Byron, 2013; Delvadia et al., 2012; Delvadia, Longest, Hindle, & Byron, 2013; Olsson, Borgstrom, Lundback, & Svensson, 2013; Ruzycki et al., 2014). In the area of pediatric aerosol delivery, Carrigy et al. (2014) reviewed both CFD and realistic in vitro experiments.

As highlighted by Longest, Bass, et al. (2019), our group frequently employs the simultaneous or concurrent application of both CFD analysis and realistic in vitro experiments, which is illustrated in Fig. 1. In this approach, both techniques are applied simultaneously in order to take advantage of each method's strengths and to minimize each method's weaknesses. Briefly, realistic in vitro experiments are used to assess initial and key prototype performance of both aerosol formation and lung delivery efficiency as a benchmark. CFD models are developed of key system aspects, such as mouthpiece aerosol transport and extrathoracic depositional loss. The CFD models are first validated with the initial in vitro data. Once validated, the CFD models are then used to generate valuable insights into system transport characteristics and explore design alternatives. Best performing model designs or strategies are then produced and tested experimentally to verify system performance improvements. In some instances, quantitative relations are formed between CFD-predicted parameters and experimental critical quality attributes, such as device emitted dose and MMAD (Bass, Farkas, & Longest, 2019; Bass & Longest, 2020; Hindle & Longest, 2013; Longest & Hindle, 2009b; Longest, Son, et al., 2013; Longest & Farkas, 2019; Longest, Farkas, et al., 2019). For example, in the quantitative concurrent analysis of Longest, Bass, et al. (2019), aerosol dispersion parameters were used to capture both emitted dose and aerosol size for a series of air-jet DPIs. These or similar dispersion parameters where then used to optimize DPI performance for low AAV (Longest, Farkas, et al., 2019) and high AAV (Bass, Farkas, & Longest, 2019) applications based on CFD analysis. Optimized devices were then prototyped and tested experimentally in each respective study. This concurrent analysis approach has been applied to the development of dry powder formulations for high dispersion (Longest, Farkas, Hassan, & Hindle, 2020), powder aerosolization within DPIs (Bass, Farkas, & Longest, 2019; Longest & Farkas, 2019; Longest, Farkas, et al., 2019), aerosol transmission through inhalers (Hindle & Longest, 2013; Longest & Hindle, 2009b), and aerosol delivery strategies (Longest & Hindle, 2012b; Longest, Tian, Li, et al., 2012; Tian et al., 2014; Tian et al., 2013; Tian, Longest, Su, & Hindle, 2011).

In order to illustrate the combined use of new technologies for high efficiency aerosol administration described above and concurrent CFD and realistic in vitro analysis, a case study is considered based on nose-to-lung dry powder aerosol administration to children with cystic fibrosis (CF). Specifically, the objective of this case study is to use multiple new techniques simultaneously in order to overcome the primary limitations associated with poor dry powder aerosol administration to children and enable high efficiency trans-nasal DPI use in this population, based on concurrent CFD and realistic in vitro analysis. Techniques used to improve lung delivery efficiency of the dry powder aerosol include nose-to-lung administration in subjects as young as 2-years-old, use of a positive-pressure active DPI, implementation of patient interfaces that improve aerosol deaggregation and dissipate the flow field, and controlled condensational growth of the aerosol within the airways. Dry powder aerosol administration is assessed through nose-throat (NT) and upper tracheobronchial (TB) models of children in the age ranges of 2–3, 5–6, and 9-10 years-old. The realistic in vitro upper TB airway models are based on airway scans of pediatric subjects that have CF with moderate lung damage. As previously implemented (Longest et al., 2015; Longest, Tian, Li, et al., 2012), the upper TB airway models are enclosed in an aerosol growth chamber that is sized to represent aerosol residence time in the lungs and connect to a Next Generation Impactor (NGI) for aerosol sizing in the experiments.

The delivery system evaluated includes a newly developed positive-pressure air-jet DPI, nose-to-lung patient interface with a 3D rod array used to dissipate the turbulent jet leaving the aerosol generation unit, and an excipient enhanced growth (EEG) powder formulation. Potential pediatric-CF drug delivery applications include the administration of inhaled antibiotics, anti-inflammatories, mucus clearance agents, medications that address airway constriction, and potential corrective gene therapies. Consistent with the concurrent analysis approach, in vitro experiments are implemented to benchmark system performance in terms of device emitted dose, system delivery efficiency, extrathoracic and upper TB depositional loss and aerosol size increase through the growth chamber for the 5-6 year-old airway model. Once validated, the CFD model is used to illustrate regional aerosol deposition, aerosol size increase, and lung delivery efficiency across multiple pediatric age ranges. In contrast with the expected 5–10% lung delivery efficiency for pediatric subjects with commercial systems (Below et al., 2013; Devadason, 2006; Devadason et al., 1997; Lindert et al., 2014), the concurrent CFD and realistic in vitro analysis shows that the implementation of multiple new technologies increases the lung delivery efficiency to >70% of the loaded dose, with little effect of pediatric age and the potential for targeted deep lung delivery through aerosol size increase.