Modeling and Simulations of Olfactory Drug Delivery with Passive and Active Controls of Nasally Inhaled Pharmaceutical Aerosols

Summary

This manuscript reviews the modeling and simulations of different protocols to deliver medications to the olfactory region in image-based nasal airway models. Multiple software modules are used to develop the anatomically accurate nose model, generate computational mesh, simulate nasal airflows, and predict particle deposition at the olfactory region.

Abstract

There are many advantages of direct nose-to-brain drug delivery in the treatment of neurological disorders. However, its application is limited by the extremely low delivery efficiency (< 1%) to the olfactory mucosa that directly connects the brain. It is crucial to develop novel techniques to deliver neurological medications more effectively to the olfactory region. The objective of this study is to develop a numerical platform to simulate and improve intranasal olfactory drug delivery. A coupled image-CFD method was presented that synthetized the image-based model development, quality meshing, fluid simulation, and magnetic particle tracking. With this method, performances of three intranasal delivery protocols were numerically assessed and compared. Influences of breathing maneuvers, magnet layout, magnetic field strength, drug release position, and particle size on the olfactory dosage were also numerically studied.

From the simulations, we found that clinically significant olfactory dosage (up to 45%) were feasible using the combination of magnet layout and selective drug release. A 64 -fold higher delivery of dosage was predicted in the case with magnetophoretic guidance compared to the case without it. However, precise guidance of nasally inhaled aerosols to the olfactory region remains challenging due to the unstable nature of magnetophoresis, as well as the high sensitivity of olfactory dosage to patient-, device-, and particle-related factors.

Introduction

Drugs delivered to the olfactory region can bypass the blood-brain-barrier and directly enter the brain, leading to an efficient uptake and quick action onset of the drugs^1,2. However, conventional nasal devices such as nasal pumps and sprays deliver extremely low doses to the olfactory region (< 1%) via the nasal route^3,4. It is primarily due to the complicated structure of the human nose which is composed of narrow, convoluted passageways (Figure 1). The olfactory region locates above the superior meatus, where only a very small fraction of inhaled air can reach^5,6. Furthermore, conventional inhalation devices depend on aerodynamic forces to transport therapeutic agents to the target area⁷. There is no further control over the motions of particles after their release. Therefore, the transport and deposition of these particles predominately depend on their initial speeds and release positions. Due to the convoluted nasal passage as well as the lack of particle control, the majority of drug particles are trapped in the anterior nose and cannot reach the olfactory region⁸.

While there are many choices of nasal devices, those designed specifically for targeted olfactory delivery have rarely been reported^7,9. One exception is Hoekman and Ho¹⁰ who developed an olfactory-preferential delivery device and demonstrated higher cortex-to-blood drug levels in rats as opposed to using a nose drop. However, scaling the deposition results in rats to humans is not straightforward, considering the vast anatomical and physiological differences between these two species¹¹. Many limitations exist when using adapted versions of standard nasal devices for olfactory deliveries. One primary setback is that only a very small portion of medications can be delivered to the olfactory mucosa, through which the medications may enter the brain. Numerical modeling predicted that less than 0.5% of intranasally administered nanoparticles can deposit in the olfactory region^3,5. The deposition rate is even lower (0.007%) for micrometer particles¹². In order to make the nose-to-brain delivery clinically feasible, the olfactory deposition rate has to be significantly improved.

There exist several possible approaches to improve the olfactory delivery. One approach is the smart inhaler idea proposed by Kleinstreuer et al.¹³ As particles depositing in one region are mainly from one specific area at the inlet, it is possible to deliver particles to the target site by releasing them only from certain areas at the inlet. The smart delivery technique has been shown to generate a much more efficient lung delivery than conventional methods.^13,14 It is hypothesized that this smart delivery idea can also be applied in intranasal drug delivery to improve dosages to the olfactory mucosa. By releasing particles into different positions at the nostril opening and from different depths within the nasal cavity, improved olfactory delivery efficiencies and reduced drug waste in the anterior nose are possible.

Another possible method is to actively control the particle motion within the nasal cavity using a variety of field forces, such as electric or magnetic force. Electric control of charged particles has been suggested for targeted drug delivery to the human nose and lungs^15-17. Xi et al.¹⁸ numerically tested the performance of electric guidance of charged particles and predicted significantly improved olfactory doses. Similarly, guidance of ferromagnetic drug particles with an appropriate magnetic field also has the potential to target particles to the olfactory mucosa. Behaviors of inhaled agents, if ferromagnetic, can be altered by imposing appropriate magnetic forces¹⁹. Dames et al.²⁰ demonstrated that it is practical to target ferromagnetic particles to specific areas in mouse lungs. By packaging therapeutic agents with superparamagnetic iron oxide nanoparticles, the deposition in one lung of a mouse under the influence of a strong magnetic field was significantly increased compared to the other lung²⁰.

Particles were assumed to be spherical and ranged from 150 nm to 30 µm in diameter. The governing equation is²¹:
(1)

The above equation describes the motion of a particle governed by drag force, gravitational force, Saffman lift force ²², Brownian force for nanoparticles, and magnetophoretic force if placed in a magnetic field. Here, v_i is the particle velocity, u_i is the flow velocity, τ_p is the particle response time, C_c is the Cunningham correction factor, and α is the air/particle density ratio. To effectively guide the intranasally administered drugs to the olfactory region, it is necessary for the applied magnetophoretic forces to overcome both the particle inertia and gravitational force. In this study, a composite of 20% maghemite (γ-Fe₂O₃, 4.9 g/cm³) and 80% active agent was assumed, which give a density of approximate 1.78 g/cm³ and a relative permeability of 50. The selection of γ-Fe₂O₃ was due to its low cytotoxic. Iron (3+) ions are widely found in human body and a slightly higher ion concentration will not cause significant side-effects²³.

Protocol

The MRI images were provided by the Hamner Institutes for Health Sciences and the usage of these images was approved by the Virginia Commonwealth University institutional review board.

1. Image-Based Nasal Airway Preparation

1. Acquire magnetic resonance (MR) images of a healthy non-smoking 53-year-old male (weight 73 kg and height 173 cm) that consist of 72 coronal cross-sections spaced 1.5 mm apart spanning the nostrils to the nasopharynx⁴.
2. Open Imaging Program (e.g., MIMICS)
   1. To import images, click "File", "Import images". Select the MR images and click "Ok".
   2. To construct the 3-D model, click "Segmentation", then "Threshold" to set the grey scale range between -1020 and -500. Click "Segmentation", "Calculate 3D".
   3. Click "Segmentation" and "Calculate polylines". Select the 3-D body, and click "Ok" to generate the polylines that define the solid geometry. Export the polylines as an IGES file.
3. Open Model Development Software (e.g., Gambit)
   1. Click "File", "Import", "IGES" to import the IGES file into the program. Click "Edge command button" on the right panel; click "Create Edge" and select "NURBS" to reconstruct smooth contours.
   2. Click "Face command button", then click "Form face". Select "Wireframe" to build a surface from edges. Continue to build all surfaces that cover the whole airway. Retain the nasal anatomical details such as the uvula, epiglottal fold, and laryngeal sinus (Figure 1). Click "File", "Export" "IGES" to export the nasal airway model.
4. Open Meshing Software (e.g., ICEM CFD)
   1. Click "File", "Import Geometry", "Legacy" and "STEP/IES" to import the nasal airway model. Click "Create Parts" to divide the airway surfaces into five different regions: nasal vestibule, nasal valve, turbinate region, olfactory, and nasopharynx.
   2. To generate computational mesh inside the airway, click "Mesh", "Global Mesh Setup". Specify the maximum mesh size as 0.1 mm and click "Apply".
   3. To add a body-fitted mesh in the near-wall region, click "Compute Mesh", "Prism Mesh". Specify the number of layers as 5 and the expanding ratio as 1.25 and click "Apply".

2. Passive Control of Particles

1. Vestibular Intubation: Front vs. Back
   1. Open Model Development Software to develop the nasal model with front vestibular intubation. Click "Volume", then "Move/copy" to change the location of the nebulizer catheter 5 mm into the vestibule from nostril tip. Click "injection" to release 60,000 particles (150 nm) into the nostril.
   2. Open the fluid simulation software (e.g., ANSYS Fluent) to compute particle deposition rates inside the nose. To compute the airflow field inside the airway, select the laminar flow model by clicking "Define", "Models", "Viscous"; chose "Laminar" under "Viscous model".
   3. Select the "Discrete Phase Model" to track particle motions. Check "Saffman Lift Force" under "Discrete Phase Model". Click "Report", then choose "Sample Trajectories"; select "nasal" under "Boundaries" and click "Compute" to find the number of particles deposited in the predefined olfactory region. Calculate the deposition rate as the ratio of the amount of deposited particles to the amount of particles entering the nostrils.
   4. Repeat steps 2.1.2 for 1 µm particles.
   5. Follow the step 2.1.1, insert the spray nozzle 5 mm into the vestibule from the back of the nostril. Repeat steps 2.1.2, and 2.1.3 to compute deposition rate for 150 nm particles. Repeat step 2.1.4 for 1 µm particles (back-intubation).
2. Deep Intubation
   1. Follow procedure 2.1.1 to insert the nebulizer catheter right beneath the olfactory region. Release 60,000 submicron particles (150 nm) from the nebulizer.
   2. Use fluid and simulation software to compute particle deposition rates inside the nose on both total and local basis by following similar procedures as listed in 2.1.2. Repeat this procedure for 1 µm particles.
   3. Repeat the above procedures while exercising breathing-holding and exhalation, respectively. Click "Define", then "Boundary Conditions" to open the boundary condition panel. Specify zero velocity at the two nostrils for breathing-holding. Specify vacuum pressure (200 Pa) at the nostrils and zero pressure at the outlet for exhalation.

3. Active Control: Magnetophoretic Guidance

1. Test in a Two-Plate Channel
   1. Open magnetic particle tracking software (e.g., COMSOL). Click "Geometry", and "Rectangle" to build the two-plate channel. Click "Rectangle" to build the magnets around the two-plate channel.
   2. Compute the particle trajectories and deposition rate. Click "Model 1", "Laminar flow" and "Inlet 1"; specify the inlet velocity as 0.5 m/s. Click "Model 1", "Magnetic Fields", and "Magnetic Flux Conservation", specify the strength of the three magnets (1 × 10⁵ A/m).
   3. Click "Model 1", "Particle Tracking for Fluid Flow", and "Particle Properties"; specify the particle diameter (15 µm), density (1.78 g/cm³). Click "Inlet" to release 3,000 particles. Click "Magnetophoretic Force", specify particle relative permeability (50). Click "Compute".
   4. To find how many particles depositing in the selected area, click "Results", "1D Plot Group" and "Plot". Calculate the deposition rate as the ratio of the amount of particle deposited in certain area to the amount of particles entering the geometry.
   5. To adjust the magnet strength, click "Model 1", then "Magnetic Fields"; choose "Magnetic Flux Conservation", and change the magnet strength under "Magnetization". Increase the magnet strength by an increment of 1 × 10⁴ A/m and click "Compute".
   6. Repeat this procedure until the appropriate magnets arrangement was obtained for effective drug delivery to the olfactory region.
2. Test in the 2-D Idealized Nose Model
   1. Apply the magnetic strengths obtained in 3.1 into a 2-D nose model by putting three magnets 1 mm above the nose. Click "Model 1", "Geometry 1" to specify the size and position of the magnet. Click "Model 1", "Particle Tracking for Fluid Flow", "Inlet" to release 3,000 particles into the left nostril. Click "Particle Properties" to specify the particle size as 15 µm.
   2. Simulate the particle trajectories and subsequent olfactory delivery efficiencies by following similar procedures as listed in 3.1.2.
   3. Adjust the magnet layout and strength to improve olfactory delivery efficiency. To adjust the magnet size and position, click "Model 1", then "Geometry 1"; choose the magnet of interest, change the values of width, depth, height or x, y, z. Follow 3.1.5 to adjust the magnet strength.
3. Test in the 3-D Anatomically Accurate Nose Model
   1. Import the 3-D nasal airway model into Magnetic Particle Tracking software. Follow the procedure 3.2.1, put four magnets 1 mm above the nose and release 3,000 particles of 15 µm in diameter from one selected point only.
   2. Use Magnetic Particle Tracking software to track particle trajectories and compute olfactory delivery efficiencies by following similar procedures as listed in 3.2.1 - 3.2.3.
   3. Following 3.2.3, adjust the magnet layout and strength in the 3D model to improve the targeted delivery to the olfactory region.
   4. Test particle size ranging from 1 - 30 µm to find the right particle size for optimal magnetophoretic guidance to the olfactory region.

Results

Control Case:
Figure 3 displays the airflow field and particle deposition in the nasal airway with standard nasal devices. It clearly shows that airflow from the front nostril is ventilated to the upper passage and airflow from the back nostril is directed towards the nasal floor (Figure 3A). Aerosol particles are observed to move faster in the median passages and slower near the walls, forming an aerosol front in the mean flow direct...

Discussion

A coupled image-CFD method was presented in this study that incorporated the image-based model development, quality meshing, airflow simulation, and magnetic particle tracking. Multiple software modules were implemented to this aim, which included functions of segmentation of medical images, reconstruction/meshing of anatomically accurate airway models, and flow-particle simulations. Using this numerical method, performances of three intranasal delivery protocols were tested and compared. Compared to in vitro ex...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
MIMICS 13	Materialise Inc, Ann Arbor, MI	MR image segmentation
Gambit	ANSYS Inc, Canonsburg, PA	Model development
ANSYS ICEMCFD	ANSYS Inc, Canonsburg, PA	Meshing
ANSYS Fluent	ANSYS Inc, Canonsburg, PA	Fluid and particle simulation
COMSOL Multiphsics	COMSOL Inc, Burlington, MA	Magnetic particle tracing