Allergic rhinitis — also known as hay fever — is a very common condition that occurs as an inflammatory response to a person breathing in an allergen, such as dust or pollen. Symptoms include congestion, runny nose, sneezing and itching. One treatment option to relieve symptoms is a nasal corticosteroid that a person applies to their nostrils using a spray device.
In this series of simulations,i researchers in CDER and external collaborators studied how different factors may affect corticosteroid nasal deposition (where the drug particles land in the nose) and, relatedly, nasal distribution (where the drug particles ultimately ‘reside’). This information can be used to predict local and systemic (throughout the body) drug concentrations. For these simulations, the researchers examined three groups of factors — spray characteristics, human factors, and nasal anatomy — to better understand how these affect drug delivery for generic nasal drugs. To conduct the simulations, the researchers created three-dimensional reconstructions of the nasal cavity (inside of the nose) from computed tomography scans of one healthy person and one patient with allergic rhinitis.
Computational Simulations and Analysis
The researchers used computational fluid dynamics (CFD), which is a numerical modeling technique that makes predictions about fluid-flow scenarios, to simulate the flow and particle fields within the modeled nasal cavity for the selected nasal sprays. As mentioned, the researchers were looking at deposition and distribution of the medication.
The modeled nasal cavity included three parts of the nose: the squamous (closest to the nostril entrance), respiratory, and olfactory epithelia (the middle and upper nostril, respectively). The investigators then subdivided the respiratory epithelium region into three regions, including a target site at the base for the nasal sprays to “hit.” For the simulations of the nasal sprays in action, the researchers used ANSYS Fluent, a commercial CFD simulation software package that can model different scenarios by changing variables. The investigators ran five simulated tests for each variable and then conducted statistical analysis to determine if the parameters had a significant effect on the spray mass delivered to the target site.
To compare spray characteristics, the investigators performed computational simulations for three commercial nasal sprays used to treat allergic rhinitis, each with different active ingredients: fluticasone propionate, mometasone furoate, and budesonide. Specifically, the simulations looked at the different sizes and speeds of the spray droplet that each product made and the different “plume” shapes (the shape of the overall arrangement of droplets).
The researchers also examined human factors that could potentially affect the predicted deposition and distribution of the product inside the nose. These potential effects included the actuation force (or how strongly the person pushes the nozzle) and how this may produce different-sized droplets; airflow rate (air movement in the nostrils); cone angle (the direction of the droplets on the plume’s edges); nozzle subtraction (when the nozzle is removed from the nose); nozzle depth (how far the nozzle goes into the nostril); and nozzle position (where the spray release point is inside the nose).
To consider differences in nasal anatomy, the researchers modeled both the left and right nostril and conducted simulations of the spray nozzle inserted into each nostril. The rationale was that there are inherent differences in anatomy between the nostrils.
The researchers found that the simulations of the three spray products predicted that, on average, more than 94% of the medication went into the nasal cavity. At least 91% of all drug mass entered the nose. Overall, the different sprays had little effect on medication distribution.
Meanwhile, the actuation force had a significant effect on how much medication reached the target site in the respiratory epithelium in the “healthy” computational model but not the “rhinitic” (inflamed) model. The airflow rate and nozzle depth followed the same pattern. The cone angle and nozzle subtraction did not produce a significant effect on either the healthy or rhinitic model.
However, in the nozzle position study, the researchers found a significant effect on how much medication hit the target when investigators ran simulations for different spray release points. More specifically, reducing the nozzle insertion depth from 10 mm to 5 mm increased the target site dose by 2-fold in the healthy model. Meanwhile, 2-mm right-left shifts in the spray release point increased the target site dose by more than 10-fold in the healthy model. In the rhinitic model, reducing the nozzle insertion depth from 10 mm to 5 mm caused a relative change of about 20%, while 2-mm shifts in the spray release point increased the target site dose by more than 4-fold.
Nasal anatomy (simulations of sprays into the right or left nostrils) also had a significant impact on the medication hitting the target in both models. The facts that the nozzle position and different nasal anatomy were significant factors can help to understand the inherent variability in these medical products with regard to corticosteroid deposition and distribution.
In some cases, these findings from computational simulations mirrored findings from in vitro and in vivo studies. For example, the modest effect of spray characteristics and inhalation rate that this study showed is consistent with in vitro experiments. These consistent findings showcase the ability of computational models in corroborating results from other studies.
This study yielded several benefits for the development and approval of generic nasal drugs. Predictions suggested that human factors and nasal anatomy may explain much of the variability observed in human testing, which will help both industry and CDER when interpreting future in vivo data. Also, industry may choose to adopt the CFD method as an internal decision-making tool. Finally, the results helped CDER evaluate the clinical relevance of various spray characteristics. This may make generic drugs available faster by providing information on how similar these characteristics should be to demonstrate bioequivalence.
i Kimbell JS, Garcia GJM, Schroeter JD, et al. Nasal steroid spray simulations using measured spray characteristics in healthy and rhinitic nasal passages. Journal of Aerosol Science. 2023.