Inhaled fosamprenavir for laryngopharyngeal reflux: Toxicology and fluid dynamics modeling

Abstract

Objectives

Approximately 25% of Americans suffer from laryngopharyngeal reflux (LPR), a disease for which no effective medical therapy exists. Pepsin is a predominant source of damage during LPR and a key therapeutic target. Fosamprenavir (FOS) inhibits pepsin and prevents damage in an LPR mouse model. Inhaled FOS protects at a lower dose than oral; however, the safety of inhaled FOS is unknown and there are no inhalers for laryngopharyngeal delivery. A pre‐Good Lab Practice (GLP) study of inhaled FOS was performed to assess safety and computational fluid dynamics (CFD) modeling used to predict the optimal particle size for a laryngopharyngeal dry powder inhaler (DPI).

Methods

Aerosolized FOS, amprenavir (APR), or air (control) were provided 5 days/week for 4 weeks (n = 6) in an LPR mouse model. Organs (nasal cavity, larynx, esophagus, trachea, lung, liver, heart, and kidney) were assessed by a pathologist and bronchoalveolar lavage cytokines and plasma cardiotoxicity markers were assessed by Luminex assay. CFD simulations were conducted in a model of a healthy 49‐year‐old female.

Results

No significant increase was observed in histologic lesions, cytokines, or cardiotoxicity markers in FOS or APR groups relative to the control. CFD predicted that laryngopharyngeal deposition was maximized with aerodynamic diameters of 8.1–11.5 μm for inhalation rates of 30–60 L/min.

Conclusions

A 4‐week pre‐GLP study supports the safety of inhaled FOS. A formal GLP assessment is underway to support a phase I clinical trial of an FOS DPI for LPR.

Level of Evidence

NA.

Fosamprenavir is a candidate pepsin‐targeting therapy for laryngopharyngeal reflux (LPR). Local delivery by inhalation could reduce dosing and was shown effective at 1/20th of the oral dose. This study makes progress toward a fosamprenavir dry powder inhaler for LPR through toxicologic analysis and fluid dynamics modeling.

1. INTRODUCTION

Laryngopharyngeal reflux (LPR) is an inflammatory condition of upper aerodigestive tract tissues related to the backflow of gastroduodenal contents. The most prevalent symptoms of LPR include dysphonia, globus sensation, throat pain, odynophagia, accumulation of viscous mucus, throat clearing, and cough. ¹ LPR can dramatically impact the quality of life and lead to serious health consequences such as airway stenosis, reactive airway disease, and laryngeal cancer. ² , ³ , ⁴ , ⁵ , ⁶

LPR is estimated to affect 10%–30% of the US population, ¹ , ⁷ , ⁸ , ⁹ yet there is no gold‐standard medical therapy. Although acid‐suppressing proton pump inhibitors (PPIs) are the mainstay therapy for GERD, their efficacy for LPR is poor. ¹⁰ Multichannel intraluminal impedance pH‐monitoring (MII‐pH) has demonstrated that many episodes of LPR are nonacidic and gaseous, and that weakly or nonacidic LPR is often associated with persistent symptoms in acid‐suppressed patients. ¹¹ , ¹² , ¹³ , ¹⁴ These symptoms are alleviated by anti‐reflux surgery ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ and may be ameliorated by less invasive strategies that limit reflux occurrence or neutralize reflux constituents beyond acid (e.g., dietary and lifestyle modification and over‐the‐counter alginate products). ²² , ²³ , ²⁴ , ²⁵ It is therefore reasonable to assume that one or more nonacid constituents of refluxate are responsible for LPR symptoms. Among nonacid components of refluxate, the digestive enzyme pepsin is considered a predominant damaging agent, biomarker, and therapeutic target for reflux‐attributed diseases. ⁵ , ²⁶ , ²⁷ , ²⁸ , ²⁹ , ³⁰ , ³¹ , ³² , ³³ , ³⁴ , ³⁵ At the pH of the laryngopharynx (pH ~ 6.8), pepsin is enzymatically inactive, whereas studies in experimental models demonstrate that it is endocytosed and retained in acidic endosomes in which its enzymatic activity would be restored. Exposure to pepsin leads to inflammatory and carcinogenic alterations irrespective of pH in vitro and in vivo including altered transcriptomic profiles; promotion of apoptotic resistance, cell migration, anchorage‐independent growth, and glycolysis; and development of tumors in a hamster cheek model. ³¹ , ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰

We recently identified the protease inhibitor fosamprenavir (FOS; prodrug of amprenavir, APR), an FDA‐approved therapy for HIV/AIDS, as a candidate therapy for LPR. FOS binds and inhibits pepsin in the low micromolar range and oral gavage of FOS at the manufacturer‐recommended dose for HIV prevented inflammatory damage in an LPR mouse model. ⁴¹ Inhaled FOS provided similar protection at 1/20th of the oral dose. ⁴¹ Although oral FOS has a good safety profile, topical administration at a lower dose should reduce the potential for side effects. Such topical administration may be achieved through a dry powder inhaler (DPI). DPIs were originally developed to treat asthma and chronic obstructive pulmonary disease (COPD), delivering active drugs in powdered form through the oral cavity into the lungs. Currently, no commercial DPI has been developed specifically for drug delivery to the laryngopharynx. Aerodynamic particle size distribution (APSD), inhalation rate, and airway diameter are key determinants of the regional doses of inhaled drugs along the airway. ⁴² , ⁴³ Current DPIs have drug particles with diameters smaller than 5 μm, which minimize drug deposition in the upper airway and maximize drug delivery to the lungs. The optimal particle size for drug delivery to the laryngopharynx has not been established yet. To develop aerosolized FOS for LPR, the toxicology of inhaled FOS requires evaluation and a device optimized for drug delivery to the laryngopharynx will need to be developed. Herein we evaluated inhaled FOS with a pre‐Good Lab Practice (GLP) toxicity study and investigated the optimal particle size of FOS as a dry powder for laryngopharyngeal delivery using computational fluid dynamics (CFD) simulations in a model of the adult respiratory tract.

2. METHODS

2.1. In vivo toxicology

Experiments were approved by the University of Minnesota (UMN) Institutional Animal Care and Use Committee (protocol 2102‐38881A) and utilized a previously validated LPR mouse model wherein mechanical wounding and instillation are used to directly injure the larynx. ⁴¹ The model reliably replicates epithelial alterations observed in patients with LPR and confers advantages over existing surgical animal models of reflux in which refluxate inconsistently reaches the laryngopharynx and the horizontally positioned upper digestive tract promotes liquid, as opposed to the gaseous refluxate typical of LPR. ⁷ , ⁴¹ , ⁴⁴ , ⁴⁵ , ⁴⁶ , ⁴⁷ , ⁴⁸ The sham group of the LPR mouse model, which employs mechanical laryngeal wounding and instillation with saline, was utilized to examine the safety of inhaled FOS and APR herein (see treatment schema in Figure 1). In 6‐week‐old female Jackson A/J mice (Jackson Laboratory, Bar Harbor, ME) acclimated 1 week post‐arrival, a laryngeal scratch was performed on days 2 and 9 by the same personnel to the anterior cricoid/tracheal cartilage and saline instilled 3 days/week for 4 weeks. FOS (Mylan Laboratories Limited, Hyderabad, Telangana, India) and APR (MedChemExpress LLC, Monmouth Junction, NJ) dry powder, and supermicrometer aerosols were generated and administered 5 days/week using a six‐port exposure chamber (Intox); air alone was administered as control. ⁴¹ Details are provided in the Supporting information, including a description of the dry powder aerosolization, particle size distribution measurements, and mass exposure calculations. The average aerosol concentrations were 0.020 mg/L FOS and 0.016 mg/L APR; given the respiratory minute volume of mice (0.020 L/min), the inhaled mass of FOS and APR were 1.1 and 0.96 mg/kg/day, respectively.

FIGURE 1.

In vivo treatment schema.

Mice were weighed, euthanized by CO₂, placed on their back, skin cut down the midline starting at the lower lip, and trachea exposed by separating fatty tissue containing salivary glands and submandibular lymph nodes. A small slit (~1 mm) was made on the exposed surface of the trachea. To collect bronchoalveolar lavage (BAL), a 1 mL syringe containing 0.5 mL DPBS (no calcium or magnesium, ThermoFisher Scientific, Waltham, MA) was connected to a dispensing tip (blunted syringe needle with a 45° bend), the tip inserted into the slit in the trachea, and a curved tip forceps used to grasp nearby fatty tissue, compressing trachea and extra tissue around the syringe tip to create a seal. The DPBS was slowly injected into the lungs, and then the syringe plunger was drawn back slowly to aspirate DPBS from the lungs. BAL was placed on ice. Whole blood was collected by severing major vessels in the subclavian space and pipetting pooled blood from the pleural cavity. Plasma was obtained by placing blood in EDTA tubes and incubating on ice. BAL and plasma were centrifuged to clear cells. Supernatants were collected. Plasma was diluted 1:10 in DPBS. Both were stored at −80 °C until analysis.

Organs were collected in 10% neutral buffered formalin for pathology including the tongue, larynx, trachea, esophagus, mediastinal tissue (containing thymus and tracheal lymph nodes), heart, lungs, liver, both kidneys, and head for nasal turbinates. Tissues in fixative were transported to UMN Comparative Pathology Shared Resource for analysis. Tissues were embedded in paraffin, sectioned at 5 μm thickness, and stained with hematoxylin and eosin. Slides were reviewed by two board‐certified veterinary pathologists who were not blinded to treatment groups.

Toxicity markers were assessed via Luminex assay by Eve Technologies (Calgary, Alberta, Canada). BAL was analyzed using Mouse High Sensitivity T cell 18‐Plex Discovery Assay and plasma‐EDTA was analyzed using Mouse Cardiovascular Disease Panel 2 9‐Plex Assay. Samples were run in duplicate in each assay and means were compared across groups by Student's t‐test. Owing to the skewness of the data, the Mann–Whitney test was used to compare the groups. Results were considered statistically significant for p < .05.

2.2. CFD simulations

A model of a healthy adult was built from a magnetic resonance imaging scan of a 49‐year‐old female with normal airway anatomy and no pathology as verified by an ear‐nose‐throat surgeon (J.B.; Figure 2A). The anatomical sites (oral cavity, pharynx, larynx, trachea, and main bronchi) were mapped on the walls to quantify the regional doses. CFD simulations of airflow and particle transport were performed for constant inhalation rates of 30, 45, and 60 L/min in ANSYS Fluent™ 14.0 (ANSYS Inc., Lebanon, New Hampshire), representing typical inhalation rates of adult patients using DPIs.

FIGURE 2.

(A) 3D model of a healthy adult airway built from MRI scan of a 49‐year‐old female. (B) Comparison of extrathoracic deposition fraction (DF) predicted by CFD for inhalation rates of 30, 45, and 60 L/min with experimental data from Stahlhofen et al. (C) Laryngopharyngeal deposition fraction predicted by the CFD simulations and fitted with Equation (S3) in the Supporting information. (D) Log‐normal mass distribution of two hypothetical DPIs (DPI #1: d ₅₀ = 10 μm, σ _g = 1.3; DPI #2: d ₅₀ = 20 μm, σ _g = 2). (E) Estimated dose delivered to the laryngopharynx by the two hypothetical DPIs at an inhalation rate of 30 L/min.

The laryngopharyngeal deposition fraction (i.e., the fraction of inhaled particles that deposit at the oropharynx and larynx) was plotted against the impaction parameter (IP, units of μm² L/min), a parameter that accounts for the effects of both particle size and inhalation rate; it is defined as $\mathrm{IP}={d_{\mathrm{a}}}^{2}Q$, where $d_{\mathrm{a}}$ is the particle aerodynamic diameter in μm and Q is the inhalation rate in L/min. The CFD simulations were validated by comparing the predicted extrathoracic dose (i.e., oral cavity, oropharynx, larynx, and trachea) with in vitro experiments reported by Stahlhofen et al., showing good agreement (Figure 2B). ⁴² The laryngopharyngeal deposition fraction was estimated assuming that DPIs generate aerosol clouds with a log‐normal particle size distribution characterized by its mass median aerodynamic diameter $d_{50}$ (i.e., half of the aerosol mass is contained in particles with $d_{\mathrm{a}}\le d_{50}$) and geometric standard deviation ${\sigma }_{\mathrm{g}}$, a measure of the width of the particle size distribution. Further details regarding CFD simulations and analysis methods may be found in Supporting information.

3. RESULTS

3.1. In vivo toxicology

One mouse from each control and FOS groups died; death in the control group suggests the cause was stress of procedures rather than the FOS drug, consistent with the absence of other signs of drug‐specific toxicity. All mice, regardless of the treatment group, had a variable amount of predominantly perivascular to less frequently peribronchiolar and subpleural inflammatory foci. This varied from predominantly lymphoplasmacytic, to a mixture of lymphocytes, plasma cells, granulocytes, and histiocytes. Granulocytes were present in mice from each treatment group. One mouse from the control group had rare, mild, and predominantly mononuclear myocarditis. Other findings in other organs were thought to be background and/or incidental lesions. Cytokines (granulocyte‐macrophage colony‐stimulating factor, GM‐CSF; interferon gamma, IFN‐γ; monocyte chemoattractant protein‐1, MCP‐1; tumor necrosis factor‐alpha, TNFα; IL‐8 homologs keratinocyte‐derived cytokine (KC), lipopolysaccharide‐induced CXC chemokine (LIX), and macrophage inflammatory protein 2‐alpha (MIP‐2); and IL‐1α, 1β, 2, 4, 5, 6, 7, 10, 12p70, 13, and 17A) were not significantly elevated in BAL of FOS or APR groups relative to control (Table 1). LIGHT (TNF superfamily member 14), Oncostatin M, phosphatidylinositol‐glycan biosynthesis class F protein‐2 (PIGF‐2), and soluble cluster of differentiation 40 ligand (sCD40L) were undetectable in plasma‐EDTA; remaining cardiotoxicity markers (chemokine ligand 16 (CXCL16), Endocan, Follistatin, Troponin I, and Troponin T) demonstrated no elevation in FOS or APR groups relative to control (Table 2).

TABLE 1.

Cytokine expression in bronchoalveolar lavage.

	Mean (IQR) pg/mL			p‐value
	Control	APR	FOS	APR vs. control	FOS vs. control
GM‐CSF	4.05 (4.05–6.15)	2.40 (0.00–5.25)	5.72 (0.00–6.89)	.36	>.99
IFN‐γ	2.09 (1.79–4.23)	2.21 (0.91–19.74)	1.68 (1.47–2.96)	.93	.83
IL‐1α	38.46 (31.15–56.09)	29.50 (3.25–57.24)	42.52 (0.00–48.92)	.85	>.99
IL‐1β	2.21 (1.82–3.44)	2.40 (1.13–2.96)	2.59 (2.46–3.20)	.52	.83
IL‐2	8.95 (8.00–11.88)	6.76 (5.72–9.98)	9.47 (5.89–15.29)	.65	.83
IL‐4	0.58 (0.50–0.87)	0.23 (0.08–0.28)	0.56 (0.16–0.64)	.0081	.53
IL‐5	3.77 (3.49–5.06)	2.80 (2.35–3.53)	5.54 (3.41–6.22)	.083	>.99
IL‐6	4.79 (2.90–6.55)	4.98 (1.95–8.14)	5.76 (5.07–6.83)	.78	.40
IL‐7	1.33 (1.13–2.04)	1.33 (1.13–1.74)	2.56 (1.03–2.56)	.85	.83
IL‐10	2.25 (2.01–3.09)	1.59 (1.41–2.25)	2.37 (2.01–2.61)	.12	.92
IL‐12p70	3.40 (2.89–4.64)	1.69 (1.27–3.40)	2.10 (0.00–2.63)	.46	.094
KC	111.29 (79.34–157.69)	71.37 (66.25–102.87)	171.74 (93.37–186.73)	.083	.68
LIX	0.00 (0.00–13.58)	2.07 (0.00–7.12)	9.15 (6.77–13.58)	>.99	.34
MCP‐1	52.45 (33.44–65.21)	20.28 (18.51–22.04)	40.45 (26.71–50.59)	.053	.68
MIP‐2	145.72 (132.56–159.20)	139.18 (109.02–163.28)	154.93 (135.37–169.09)	.71	.83
TNF‐α	1.93 (1.47–2.65)	0.50 (0.00–2.70)	2.18 (0.32–2.32)	.31	.83

Abbreviations: GM‐CSF, granulocyte‐macrophage colony‐stimulating factor; IFN‐γ, interferon gamma; IL, interleukin; KC, keratinocyte‐derived cytokine (IL‐8 homolog); LIX, lipopolysaccharide‐induced CXC chemokine (IL‐8 homolog); MCP‐1, monocyte chemoattractant protein‐1; MIP‐2, macrophage inflammatory protein 2‐alpha (IL‐8 homolog); TNF‐α, tumor necrosis factor‐alpha.

TABLE 2.

Cardiotoxicity marker expression in plasma.

	Median (IQR) pg/mL			p‐value
	Control	APR	FOS	APR vs. control	FOS vs. control
CXCL16	16.39 (0.00–26.45)	0.00 (0.00–0.00)	0.00 (0.00–0.00)	.14	.072
Endocan	0.00 (0.00–0.00)	0.00 (0.00–0.00)	0.00 (0.00–0.00)	.36	.42
LIGHT	ND	ND	ND
Follistatin	0.00 (0.00–129.25)	0.00 (0.00–39.88)	0.00 (0.00–261.45)	.75	>.99
Oncostatin M	ND	ND	ND
sCD40L	ND	ND	ND
PIGF‐2	ND	ND	ND
Troponin I	67,639.82 (128.20–111,665.77)	76,276.77 (59,003.83–275,983.99)	108,149.33 (42,504.33–249,926.09)	.52	.40
Troponin T	3927.90 (0.00–4403.37)	5568.25 (3327.45–20,113.41)	9584.82 (2484.50–15,301.53)	.23	.21

Abbreviation: ND, not detected.

3.2. CFD modeling

The CFD simulations predicted that a monodisperse aerosol with aerodynamic diameter d _a = 11.5 μm would provide a laryngopharyngeal deposition fraction of 0.76 (i.e., 76% of the DPI dose would deposit in the laryngopharynx) for an inhalation rate of 30 L/min (corresponding to IP = 3964 μm² L/min) (Figure 2C). Similarly, for inhalation rates of 45 and 60 L/min, the optimal aerodynamic diameter to maximize laryngopharyngeal deposition was predicted to be 9.4 and 8.1 μm, respectively. Assuming that FOS particles in a DPI formulation have a hypothetical density of ρ _p = 1400 kg/m³, the optimal geometric diameter to maximize laryngopharyngeal deposition computed from the aerodynamic diameter (see Equation S2 in the Supporting information) would be 6.9–9.7 μm for inhalation rates of 30–60 L/min based on the relationship between aerodynamic diameter and geometric diameter (see Supporting information). ⁴⁹

However, DPIs generate polydisperse aerosols. To illustrate how the APSD affects the DPI dose delivered to the laryngopharynx, we considered two hypothetical DPIs with different mass median diameters (d ₅₀) and geometric standard deviations (σ _g). DPI #1 (d ₅₀ = 10 μm, σ _g = 1.3) has a narrower APSD with d ₅₀ near the aerodynamic diameter that maximizes laryngopharyngeal deposition, while DPI #2 (d ₅₀ = 20 μm, σ _g = 2) has a wider APSD with d ₅₀ shifted from the optimal value (Figure 2D). For an inhalation rate of 30 L/min, we estimated that 58.1% of the dose from DPI #1 would deposit in the laryngopharynx, while only 19.6% of the dose from DPI #2 would deposit in the laryngopharynx (Figure 2E). After accounting for the log‐normal distribution of pharmaceutical aerosols, we estimated that the optimal mass median aerodynamic diameter for laryngopharyngeal deposition would be d _a = 9.1 μm at an inhalation rate of 30 L/min and assuming a geometric standard deviation of σ _g = 1.3. These results illustrate that the particle size distribution generated by the DPI has a major impact on the dose delivered to the laryngopharynx.

4. DISCUSSION

The hypothesis that inhibitors of peptic activity and/or receptor antagonists hold potential for treatment of LPR is gaining acceptance, with over a decade of supportive research, lending toward the production of novel therapeutic options. ²⁶ , ²⁸ , ³⁴ , ⁵⁰ Proof‐of‐concept that peptic inhibition reduces aerodigestive tract damage has been established in animal models of LPR. ⁴¹ , ⁵¹ Local delivery by inhalation confers the benefit of reduced dosing as verified in our LPR mouse model. ⁴¹ Although oral FOS has a reasonably good safety profile, that of inhaled FOS has not yet been characterized. Further, commercial inhalers for laryngopharyngeal delivery do not currently exist. Therefore, the development of aerosolized FOS for LPR will require its toxicologic assessment and design, development, and testing of an inhaler optimized for laryngopharyngeal delivery.

Standard GLP assessment required for registration of an investigative new drug with the Food and Drug Administration involves a 28‐day toxicologic trial in rodent and non‐rodent models. Herein, treatment of equal duration produced no toxicity in a rodent model as indicated by organ pathology, BAL inflammatory cytokines, and plasma cardiotoxicity markers. This preliminary assessment bodes well for formal GLP assessment which is currently underway to support an FDA‐regulated phase I, randomized clinical trial to assess the safety and toxicity of inhaled FOS as a dry powder for LPR.

Current commercial DPIs maximize pulmonary delivery by generating plumes of fine particles of 1–4 μm to bypass the upper airways. ⁵² A prior CFD study estimated that in a monodisperse aerosol (i.e., consisting of particles of a single size), 8–10 μm particles would provide maximal laryngeal deposition, ⁵³ which is in good agreement with the optimal aerodynamic diameter of 8.1–11.5 μm predicted in this study. However, although monodisperse aerosols may be generated under controlled laboratory conditions, DPIs are polydisperse. To our knowledge, this is the first study to examine how the APSD influences laryngopharyngeal drug delivery. After accounting for the log‐normal distribution of pharmaceutical aerosols, we estimated that the optimal mass median aerodynamic diameter for laryngopharyngeal deposition would be d _a = 9.1 μm at an inhalation rate of 30 L/min which is slightly smaller than the optimal aerodynamic diameter of 11.5 μm for a monodisperse aerosol. Our analysis of two hypothetical DPIs with moderately different APSD revealed a threefold difference in the laryngopharyngeal dose. These results illustrate that DPIs for laryngopharyngeal drug delivery may need to be redesigned as compared to DPIs for pulmonary drug delivery to generate plume properties that maximize drug delivery to the target site.

Limitations of the study include unmasked evaluation of organ pathology. While this increases risk of bias, negative organ pathology was corroborated by absence of drug‐related effects on inflammatory and cardiotoxicity markers in BAL and plasma. In addition, as with any experimental observation, caution should be exercised when translating in vivo findings from a limited number of animals to the clinical situation. Animal toxicology studies will continue to be critical for demonstrating the safety of investigational new drugs until superior models are substantiated as acceptable alternatives; however, differences between mouse and human physiology should be kept in mind when evaluating the implications of these data. Limitations of CFD analysis include the small sample size (n = 1), use of a simplified turbulence model in the steady‐state CFD simulations, lack of consideration of mucociliary clearance, and drug physicochemical properties (e.g., solubility). Future research to address these limitations could include CFD simulations in a larger cohort to investigate interindividual variability and the development of a physiologically based pharmacokinetic model of drug dissolution, transport, and absorption to account for mucociliary clearance and dissolution rates. Such future work may determine the optimal APSD and inhalation rate (which is associated with the airflow resistance of the device) to maximize drug delivery to the laryngopharynx. These data would inform the design of the FOS DPI and patient instructions for use regarding optimal breathing technique. Finally, whereas the most prevalent symptoms of LPR (globus sensation, throat clearing, and hoarseness) primarily involve the laryngopharynx, LPR produces symptoms affecting more proximal regions of the aerodigestive tract. A randomized placebo‐controlled trial using a validated symptom questionnaire as an outcome measure will ultimately yield the best assessment of the efficacy of a laryngopharyngeal FOS DPI for the resolution of the various symptoms associated with LPR.

5. CONCLUSION

FOS is a candidate LPR therapeutic with a lower effective dosage required by inhalation versus ingestion. To develop inhaled FOS for LPR, the safety of inhaled FOS was verified over 4 weeks in a rodent model, and the optimal particle size for DPI delivery to the laryngopharynx was estimated via CFD. The results indicate a good safety profile and feasibility of a laryngopharyngeal FOS DPI and support a formal preclinical GLP assessment and phase I clinical trial in LPR patients.

FUNDING INFORMATION

This work was funded by the Dr. Ralph and Marian Falk Medical Research Trust of the Bank of America, N‐Zyme Biomedical Inc., and the Department of Otolaryngology and Communication Sciences, Medical College of Wisconsin, Milwaukee, WI, USA.

CONFLICT OF INTEREST STATEMENT

N.J. is a co‐founder, Chief Scientific Officer, and an investor in N‐Zyme Biomedical. N.J. is an inventor on International Patent Application PCT/US2021/027758, Aerosolized formulations of HIV protease inhibitors for the treatment of airway reflux, filed on April 16, 2021, and International Patent Application: PCT/US2023/071204, Sustained‐release oral fosamprenavir for the treatment of reflux, filed on July 28, 2023. T.S. is an investor in N‐Zyme Biomedical. Other authors have no financial relationships or conflicts of interest to disclose.

Supporting information

Data S1: In vivo toxicology aerosolization methods.

Computational fluid dynamics (CFD) methods.

Lesnick A, Samuels TL, Seabloom D, et al. Inhaled fosamprenavir for laryngopharyngeal reflux: Toxicology and fluid dynamics modeling. Laryngoscope Investigative Otolaryngology. 2024;9(1):e1219. doi: 10.1002/lio2.1219