The following case studies illustrate CFD technology applied to three classes of drug-delivery devices: inhalers, intravitreal
implants, and transdermal patches.
Respiratory drug delivery.
Inhalation technology is extensively used to treat obstructive airway diseases, such as asthma and chronic obstructive pulmonary
disease. The development cost for a new inhalation drug is high (1). One reason for the high cost is the extremely difficult
and expensive methods used to evaluate drug and device effectiveness in terms of deliverability and deposition. The cost is
especially high for dry powder inhalers (DPIs) because powder management is complicated by the special physics of the powder
and agglomeration effects (2). After experiments, computer simulations can assist in examining powder behavior by providing
the ability to look at individual forces acting between the particles. This visual can aid in the understanding of effects,
such as agglomeration, de-agglomeration, and particle deposition in the patient.
One specific tool involves particle analysis (ANSYS fluent, ANSYS), which includes a module to treat particles. The module
considers the interaction between the air flow and the aerosol particles as well as particle-to-particle and particle-to-wall
interactions. To understand and describe the transition from a powder pile at rest to a fully dispersed inhalable aerosol,
a researcher must consider various aspects of particle physics (3, 4). The ANSYS model includes soft-sphere collisions (spring–damper–model),
Coulomb forces, van der Waals forces, and surface-friction forces (static, dynamic, and rolling friction) for each particle.
Slipstream effects can be considered for small particles in the vicinity of larger carrier particles. Particle rotation is
tracked and transferred between particles during collisions as a function of friction forces. Particle agglomerates can be
detected by their number and size to analyze the effectiveness of the dispersion. Figures 1a–1c show a powder pile inside
a capsule that is dispersed by an incoming air flow. The particle module is under continual development.
Figures 1a and 1b
Ocular drug delivery.
Macular degeneration, retinitis pigmentosa, and diabetic retinopathy are examples of age-related eye diseases that can benefit
from sustained release approaches to drug delivery. The primary routes for delivering drug to the eye are systemic, topical,
periocular (e.g., conjuctival), and intraocular (e.g., intravitreal injections or implants) (5). Topical drops, gels, and
other vehicles suffer from the limitations of poor solute flux through the corneal surface and losses to fluid drainage and
tear-fluid turnover. As a result, high drug concentrations are used to compensate for the losses. Conjunctival and intravitreal
routes do not have the same drawbacks, so these vehicles use sustained-release systems to maintain drug concentrations at
the back of the eye for extended periods of time, typically months. Simulation can provide an understanding of transient and
spatial distribution of the drug throughout the eye, thus giving detailed information about what factors affect drug distribution
on the target surface.
Figure 2: Eye model created sing ANSYS simulation software.
Drug transport in the eye is driven by the convection of aqueous humor, elimination, and diffusion. Convection dominates transport
in the anterior chamber, whereas diffusion typically dominates in the posterior portions of the eye. A complete eye model
includes both chambers because drug may be lost to sinks in either chamber. The modeling effort, therefore, begins with establishing
a three-dimensional model of aqueous humor flow and elimination. First is the construction of an idealized version of a human
eye (see Figure 2). Material properties for the fluids are taken from literature sources as are boundary conditions for the
inflow of aqueous humor, scleral, and corneal boundary pressures, and intraocular pressure (IOP) (6, 7). Figure 3 shows results
for the baseline flow pattern and intraocular pressure (IOP) on the symmetry plane of the eye. With the convection patterns
established, a bolus of drug is added to the model to examine drug delivery. A bolus located in the center of the eye near
the retinal surface is ideal for mass transport, but may obstruct the patient's field of view. A key question addressed by
simulation, therefore, is how a peripheral bolus location impacts drug concentrations on the retinal surface. Figure 3c shows
the asymmetrical delivery pattern, central depletion, and drug penetration into the retinal, choroidal and scleral tissues
characteristic of peripherally placed drug boluses and implants.
Figure 3: Results of a drug-delivery model, showing (from left to right): intraocular pressure, speed countours, and distribution
Transdermal drug delivery.
Transdermal drug-delivery systems (or patches) are a preferred platform for systemic drug delivery because of improved patient
compliance and uniformity of plasma drug concentration over longer periods of time, relative to oral-dosage forms. The patch
itself is constructed of a backing, drug depot, adhesive film, and liner. The drug depot may be a reservoir, polymer or adhesive
matrix (8). A permeation enhancer is typically included in the formulation to increase the drug-diffusion coefficient in the
skin or the partition coefficient at the patch–skin interface (9). The challenge of patch design is to deliver therapeutically
relevant amounts of drug across a patient's skin and into the bloodstream.
Figure 4: Model geometry used in transdermal-patch simulations.
The patch model shown in Figure 4 is representative of a reservoir system. The components essential to modeling drug transport
are the drug reservoir and skin layer. The capillary bed is located at the far edge of the skin region. An infinite sink boundary
condition is most commonly used to model the driving force of drug through the patch and skin. This condition assumes that
the concentration of free drug in plasma is so low that the actual concentration can be ignored. It is more accurate, however,
to include a boundary condition that accounts for the pharmacokinetics of each drug. As part of a broader human body modeling
initiative, ANSYS has implemented a three-compartment pharmacokinetic model representing absorption, distribution, metabolism,
and elimination of drug using standard rate-dependent models (10). The three compartments are the plasma, target, and other
tissues. Figure 5 shows the rise and fall of drug concentration in plasma after a series of 24-hour patch applications.
Figure 5: Evolution of free drug concentration in plasma over days.
The above examples illustrate key ways to model today's drug-delivery devices. Simulation software enables users to construct
(or import) anatomical structures, model physiologic transport processes, and incorporate physical and physiological phenomena.
The results can reduce testing time and cost in drug development.
Marc Horner, PhD, is lead technical services engineer in healthcare, and Dr. –Ing
Ralf Kroeger is a senior CFD engineer, both at ANSYS.