Biopharmaceutical assessment of naloxone permeation through human respiratory epithelial tissues: A chromatographic-mass spectrometric approach with cloud based aerosol dosing and delivery

Abstract

Understanding naloxone permeation is important for optimizing nasal delivery and supporting comparative assessment of nasal drug products. In this study, a stability-indicating LC–MS/MS method was developed and validated for the simultaneous quantification of naloxone and its related impurities, naloxone N-oxide and noroxymorphone, in in vitro permeation test receptor media. The validated method was applied to characterize naloxone permeation following cloud-based aerosol dosing across a synthetic Nuclepore Track-Etched membrane and a differentiated human EpiAirway^™ mucociliary tissue model under finite-dose conditions. The analytical procedure demonstrated linearity over 0.25–20.0 ng/mL in Dulbecco’s Phosphate-Buffered Saline and Krebs–Ringer Bicarbonate Buffer, with acceptable accuracy and precision. No degradation products or additional impurities were detected in permeation samples, confirming the stability-indicating capability of the method. Naloxone exhibited rapid early-time permeation across the synthetic membrane, whereas transport across the epithelial tissue model was attenuated and plateaued, reflecting physiological barrier function. Integration of cloud-based aerosol delivery with a validated LC–MS/MS platform enables mechanistic evaluation of nasal naloxone permeation and provides a supportive in vitro framework for formulation characterization and comparative assessments.

1. Introduction

The opioid crisis remains a public health concern in the United States. In 2022, over 82,000 deaths resulted from opioid overdoses—more than 76 % of all drug-related fatalities—largely driven by synthetic opioids such as fentanyl, frequently in combination with stimulants like cocaine and methamphetamine [1]. Naloxone, a high-affinity μ-opioid receptor antagonist, serves as the frontline intervention for reversing opioid-induced respiratory depression. Naloxone’s capacity to displace opioid agonists at central nervous system (CNS) receptors allows rapid restoration of normal respiratory function; importantly, the actions of naloxone do not reverse CNS depression caused by non-opioid agents such as benzodiazepines [2,3]. Despite naloxone’s rapid action following intravenous or intramuscular administration, naloxone has a relatively short half-life ranging from 30 to 81 min which often necessitates repeated dosing when long-acting opioids are involved. Naloxone undergoes hepatic glucuronidation to naloxone-3-glucuronide, which is renally excreted, and retains a favorable hepatic safety profile even in patients with liver impairment [2]. In addition, low oral bioavailability due to first-pass metabolism limits naloxone’s efficacy when administered enterally [4].

To provide more accessible, non-invasive administration, intranasal naloxone products such as Narcan (4 mg), RiVive (3 mg), Kloxxado (8 mg), and generic naloxone hydrochloride (4 mg) have been developed and FDA-approved [5,6]. These devices are intended for use by non-medical responders and achieve relative systemic bioavailability of approximately 43–54 % compared to injectable forms, making them highly suitable for emergency opioid overdose situations [7,8]. Despite these advances in delivery, the stability of naloxone in formulations remain of interest. Oxidative, photo, and thermal degradation can produce impurities such as naloxone N-oxide and noroxymorphone that may counteract naloxone’s effects if present in meaningful amounts [9–11]. Consequently, appropriately sensitive and selective analytical procedures are necessary to detect and control these impurities during product development and throughout the product’s shelf life [12,13].

Developing suitably sensitive and selective analytical procedures involve multiple challenges. First, naloxone and its impurities are often present at trace concentrations, measured in picograms to low nanograms per milliliter, especially in in vitro permeation test (IVPT) receptor media. LC–MS/MS methods can achieve the necessary lower limits of quantitation (e.g., 0.01 ng/mL for naloxone, 0.25 ng/mL for naloxone N-oxide, and noroxymorphone in human plasma) [13], but adapting these methods to IVPT matrices like Dulbecco’s Phosphate-Buffered Saline (DPBS) and Krebs–Ringer Bicarbonate Buffer (KRBB) introduces new challenges. Specifically, such in vitro matrices may induce ion suppression or enhancement during electrospray ionization. To mitigate these matrix effects, procedures should incorporate matrix-matched calibration, thorough matrix-effect evaluation, and use stable isotope-labeled internal standards (ISTD) [14]. Furthermore, naloxone and its impurities share similar chemical structures, making chromatographic separation technically challenging. High-resolution reverse-phase columns (e.g., EC-C18, phenyl-hexyl) and optimized gradient elution are typically required to achieve the necessary discrimination [15].

Chemical stability is another concern. As naloxone is prone to oxidation, photo, and thermal degradation, sample handling should minimize artifact formation by implementing low-temperature storage, antioxidants, and protection from light [16]. Additionally, highly polar analytes like naloxone N-oxide and noroxymorphone are poorly retained on conventional reversed-phase columns and often require hydrophilic interaction chromatography for accurate quantification [10,17]. Moreover, most published analytical procedures are designed for plasma, urine, or liver microsome and may not be directly transferable to IVPT receptor media. Such transfer requires tailored sample preparation strategies, such as solid-phase extraction and matrix spiking to assure accurate, reproducible results [12,18]. These analytical requirements align with FDA and ICH M10 guidance, which require validated, stability-indicating analytical methods capable of quantifying analytes above specified thresholds [19]. Considering these analytical, regulatory, and formulation challenges, the current study aimed to validate a sensitive, selective, and stability-indicating LC–MS/MS method for the simultaneous quantification of naloxone, naloxone N-oxide, and noroxymorphone in IVPT receptor media (DPBS, KRBB) for biopharmaceutical assessment of naloxone permeation through human respiratory epithelial tissues via cloud-based aerosol delivery. This procedure was applied to permeation studies using both artificial membranes and a biologically relevant EpiAirway^™ mucociliary tissue model with dose delivery by the Vitrocell^® Cloud Alpha 12 system.

2. Materials & methods

2.1. Materials

Narcan (Naloxone HCl) nasal spray, 4 mg/0.1 mL was supplied by Emergent BioSolutions (Rockville, MD, USA). Reference standards (Naloxone, Naloxone N-oxide, Noroxymorphone, Naloxone-d₅), Formic Acid, and Krebs-Ringer Bicarbonate Buffer were purchased from MilliporeSigma (Rockville, MD, USA). Acetonitrile, and Ammonium Acetate were purchased from Fisher Scientific (Hampton, NH, USA). Methanol from Alfa Aesar (Haverhill, MA, USA), Water from Honeywell Burdick & Jackson (Muskegon, MI, USA), and Dulbecco’s Phosphate Buffered Saline (1X) with Calcium and Magnesium from Mediatech, Inc. (Manassas, VA, USA) were purchased to conduct the experiments. Nuclepore Track-Etched membrane was purchased from GE Healthcare Life Sciences (Marlborough, MA, USA). EpiAirway^™ mucociliary 3D tissue model and cell culture medium were purchased from Mattek Life Sciences (Ashland, MA, USA).

2.2. LC-MS/MS method

An Agilent ULTIVO triple quadrupole LC-MS/MS coupled with an Infinity II 1260 autosampler (Agilent Technologies, Inc., Santa Clara, CA, USA) was used for quantification of naloxone and its impurities in the permeation samples. A Poroshell 120 EC-C18 analytical column (2.1 ×50 mm, 1.9 μm; SN: 699675–902, Agilent Technologies, Inc., Santa Clara, CA, USA) and an InfinityLab Poroshell 120 EC-C18 guard column (2.1 mm, 1.9 μm; SN: 821725–940, Agilent Technologies, Inc., Santa Clara, CA, USA) were used for the chromatographic separation. An injection volume of 10 μL was used and the LC analytical column was maintained at a temperature of 30°C throughout the analysis. Water with 0.2 % Formic Acid (Mobile Phase A) and Methanol with 0.5 mM Ammonium Acetate (Mobile Phase B) were used as mobile phase that was pumped at a flow rate of 0.5 mL/min. Mobile phase composition for gradient elution are described in the supplemental section (Table S1). Autosampler temperature was maintained at 4°C and sample acquisition time was around 4.3 min. A mixture of Mobile Phase A and Mobile Phase B at a volume ratio of 95:5 was used as diluent for preparation of analytical standards and quality control (QC) samples. MS/MS detection was performed using positive ion electrospray (ESI+) in multiple reaction monitoring (MRM) mode. The MS/MS parameters applied to the ESI source were a capillary voltage of 4.5 kV, a desolvation gas (nitrogen) flow of 12 L/min, a desolvation gas temperature of 250 °C, a sheath gas flow rate of 11 L/min, a sheath gas temperature of 300 °C, and a nozzle voltage of 1.5 kV. The retention times (RT), average dwell times, m/z transitions, collision energy and fragmentor voltage for all compounds are shown in the supplemental section (Table S2). Data acquisition was performed using MassHunter acquisition software, version 1.2 (Agilent Technologies, Inc., Santa Clara, CA, USA). Data were analyzed using MassHunter quantitative analysis software, (version 10.2) and Qualitative analysis software (version 10.0) from Agilent Technologies, Inc. (Santa Clara, CA, USA).

2.3. Analytical procedure validation

The LC–MS/MS analytical procedure was validated for the simultaneous quantification of naloxone, naloxone N-oxide, and noroxymorphone in IVPT receptor media (DPBS and KRBB) in accordance with U.S. FDA and ICH M10 bioanalytical method validation guidance. Validation parameters included selectivity, sensitivity, linearity, accuracy, precision, stability, and reproducibility. Method selectivity was demonstrated by the absence of interfering peaks at the RT of the analytes and ISTD in blank DPBS and KRBB matrices. Calibration curves (CCs) were constructed using matrix-matched standards over a concentration range of 0.25–20.0 ng/mL, employing weighted linear regression (1/x), and exhibited correlation coefficients greater than 0.99 across all validation runs. Accuracy and precision were evaluated at the lower limit of quantification (LLOQ) and at low, mid, and high QC levels using replicate analyses within and across analytical runs. Intra-day and inter-day accuracy and precision values met acceptance criteria across both matrices. Analyte stability in DPBS and KRBB was confirmed under refrigerated storage conditions, with measured concentrations remaining within predefined acceptance limits relative to freshly prepared standards. System suitability testing (SST) was performed at the beginning of each analytical run to ensure consistent chromatographic performance and mass spectrometric response. A summary of validation results for all analytes in DPBS and KRBB is provided in Tables 1 and 2, with detailed numerical data presented in the Supplementary Information.

Table 1.

Summary of analytical method validation for naloxone, naloxone N-oxide, and noroxymorphone in DPBS.

Analytical Procedure	Parameters Evaluated	Acceptance Criteria	Standards Evaluated	Days of Validation	Results
System Suitability	Replicates, RT RSD, Area RSD, and Symmetry RSD	n = 6, RSD < 2 %, RSD < 5 %, RSD < 10 %	SST	Every run	Compound	RT			Area			Symmetry
					Naloxone	0.09 %			1.37 %			4.83 %
					Naloxone N-oxide	0.09 %			2.46 %			4.78 %
					Noroxymorphone	0.18 %			3.31 %			4.83 %
Specificity	Interfering peaks	No interfering peaks	LLOQ, ISTD Blank and Mobile Phase Blank	Every run	Compound				Retention time (min)
					Naloxone				1.34
					Naloxone N-oxide				1.65
					Noroxymorphone				0.89
Linearity and Range	Replicates, Accuracy – LLOQ, Accuracy	n = 1, 20 % Nominal Concentration, 15 % Nominal Concentration	CC1-CC7	Every run	Compound	CCI					CC2 – CC7
					Naloxone	93.3 %–106.7 %					96.6 %–107.2 %
					Naloxone N-oxide	90.8 %–103.2 %					95.6 %–107.2 %
					Noroxymorphone	92.9 %–111.2 %					88.9 %–112.6 %
					Range: 0.25 ng/mL to 20 ng/mL
Accuracy (%) and Precision (%)	Replicates, Accuracy – LLOQ, Precision – LLOQ, Accuracy, and Precision	n = 5, 20 % Nominal Concentration, RSD < 20 %, 15 % Nominal Concentration, RSD < 15 %	LLOQ, LQC, MQC, and HQC	Validation Day 1, 2, and 3	Compound	LLOQ				LQC, MQC, & HQC
					Naloxone	95.7 %–116.5 %, 2.0 %–12.4 %				97.8 %–109.0 %, 2.8 %–8.8 %
					Naloxone N-oxide	96.2 %–113.6 %, 3.1 %–11.1 %				98.1 %–107.2 %, 2.4 %–7.9 %
					Noroxymorphone	97.6 %–118.5 %, 3.7 %–7.7 %				96.5 %–103.7 %, 2.5 %–8.5 %
Limit of Quantitation	Signal-to-noise ratio	≥ 10.0	LLOQ	Every run	Compound			Signal-to-noise ratio
					Naloxone			13.7
					Naloxone N-oxide			56.9
					Noroxymorphone			22.8
Stability	Replicates, Accuracy	n = 3, 15 % Nominal Concentration	LQC and HQC	N/A	Compound		LQC			HQC
					Naloxone		110.7 %			104.8 %
					Naloxone N-oxide		100.8 %			97.0 %
					Noroxymorphone		95.0 %			93.7 %

Table 2.

Summary of analytical method validation for naloxone, naloxone N-oxide, and noroxymorphone in KRBB.

Analytical Procedure	Parameters Evaluated	Acceptance Criteria	Standards Evaluated	Days of Validation	Results
System Suitability	Replicates, RT RSD, Area RSD, and Symmetry RSD	n = 6, RSD < 2 %, RSD < 5 %, RSD < 10 %	SST	Every run	Compound	RT			Area		Symmetry
					Naloxone	0.20 %			1.04 %		7.55 %
					Naloxone N-oxide	0.07 %			2.54 %		4.46 %
					Noroxymorphone	0.17 %			3.71 %		3.52 %
Specificity	Interfering peaks	No interfering peaks	LLOQ, ISTD Blank and Mobile Phase Blank	Every run	Compound				Retention time (min)
					Naloxone				1.34
					Naloxone N-oxide				1.65
					Noroxymorphone				0.89
Linearity and Range	Replicates, Accuracy – LLOQ, Accuracy	n = 1, 20 % Nominal Concentration, 15 % Nominal Concentration	CC1-CC7	Every run	Compound	CC1				CC2 – CC7
					Naloxone	103.9 %–112.6 %				94.6 %–102.7 %
					Naloxone N-oxide	96.0 %–105.8 %				95.4 %–104.5 %
					Noroxymorphone	109.5 %–117.4 %				87.2 %–102.7 %
					Range: 0.25 ng/mL to 20 ng/mL
Accuracy (%) and Precision (%)	Replicates, Accuracy – LLOQ, Precision – LLOQ, Accuracy, and Precision	n = 5, 20 % Nominal Concentration, RSD < 20 %, 15 % Nominal Concentration, RSD < 15 %	LLOQ, LQC, MQC, and HQC	Validation Day 1, 2, and 3	Compound	LLOQ				LQC, MQC, & HQC
					Naloxone	105.7 %–117.9 %, 3.5 %–5.2 %				95.7 %–104.4 %, 1.6 %–10.1 %
					Naloxone N-oxide	105.9 %–111.8 %, 4.8 %–6.3 %				94.3 %–102.8 %, 1.8 %–10.1 %
					Noroxymorphone	105.8 %– 114.8 %, 2.5 %–6.6 %				94.8 %–99.9 %, 2.1 %–10.1 %
Limit of Quantitation	Signal-to-noise ratio	≥ 10.0	LLOQ	Every run	Compound			Signal-to-noise ratio
					Naloxone			18.7
					Naloxone N-oxide			95.3
					Noroxymorphone			19.6
Stability	Replicates, Accuracy	n = 3, 15 % Nominal Concentration	LQC and HQC	N/A	Compound		LQC			HQC
					Naloxone		104.6 %			99.4 %
					Naloxone N-oxide		114.6 %			114.1 %
					Noroxymorphone		85.8 %			87.8 %

2.4. Preparation of solutions for method validation and application

Stock solutions of naloxone, naloxone N-oxide, and noroxymorphone were prepared in methanol at a nominal concentration of 0.1 mg/mL using certified reference materials. A combined working stock solution containing all three analytes was subsequently prepared and serially diluted with DPBS or KRBB, as appropriate, to generate matrix-matched calibration standards and QC samples. The calibration standards were prepared over a concentration range of 0.25–20.0 ng/mL in both DPBS and KRBB to encompass the expected concentrations in IVPT samples. QC samples were prepared independently at four concentration levels corresponding to the LLOQ, low, mid, and high QC levels. Naloxone-d₅ was used as an ISTD and was added at a constant concentration to all calibration standards, QC samples, and unknown samples prior to analysis. The System suitability samples containing naloxone, naloxone N-oxide, and noroxymorphone were prepared at representative concentrations within the calibration range and were analyzed at the beginning of each analytical run to verify chromatographic performance and mass spectrometric response. All calibration standards and QC samples were prepared using matrix-matched procedures and analyzed in accordance with FDA and ICH M10 bioanalytical method validation recommendations [19]. Supplementary Information

2.5. Naloxone permeation across Nuclepore Track-Etched membrane and EpiAirway mucociliary tissue model

The validated analytical procedure was applied to quantify naloxone and its impurities in permeation samples after dosing of naloxone hydrochloride nasal formulation (Narcan 4 mg/ 0.1 mL) to Nuclepore Track-Etched membrane and EpiAirway mucociliary tissue model.

For naloxone permeation across Nuclepore Track-Etched synthetic membrane, naloxone was aerosolized and deposited onto a synthetic Nuclepore^® Track-Etched polycarbonate membrane with a surface area of 4.91 cm² using the Vitrocell cloud alpha 12 system (VITROCELL Systems GmbH, Waldkirch, Germany). The deposition of naloxone from the commercial nasal spray formulation (Narcan^®, 4 mg/0.1 mL) was quantified using a quartz crystal microbalance (QCM), which served as a surrogate deposition surface within the cloud chamber. A total of 400 μL of naloxone nasal spray solution was nebulized using a vibrating mesh nebulizer (4–6 μm mesh pore size), which operates by high-frequency oscillation to generate uniform aerosol droplets. The nebulization process required approximately 76.5 s (at a nebulization rate of 0.2–0.8 mL/min) to achieve full deposition of the dose. Upon completion of dosing, the chamber lid was lifted after 7 min to allow the deposited layer to dry. The average deposited naloxone mass on the QCM was 57.61 μg/cm². Following aerosol deposition, in vitro permeation was assessed using a vertical in-line flow-through diffusion system (PermeGear, Inc., Hellertown, PA, USA), with the membrane acting as the donor interface. DPBS supplemented with Calcium and Magnesium ions was used as the receptor medium and was perfused tangentially across the underside of the membrane at a controlled flow rate of 200 μL/min. Receptor fluid samples were collected at the following time points: 5, 10, 20, 30, 45, 60, 90, and 120 min. Each sample was stored in amber glass vials (20 mL) at 4 °C to protect against light-induced degradation prior to analysis. Sample dilutions ranged from 1x to 20,000x, depending on the expected concentration. DPBS diluted in mobile phase (initial composition) was used as a blank control and injected alongside samples during LC–MS/MS analysis. Quantification of naloxone and its related impurities was based on the average of three replicates for each time point.

In vitro permeation of naloxone was also evaluated using the EpiAirway^™ mucociliary human epithelial tissue model, cultured on transwell inserts to simulate the nasal mucosal barrier. A dose of 150 μL of naloxone nasal spray solution (40 mg/mL) was nebulized directly onto the apical surface of the tissue using the Vitrocell^® Cloud Alpha 12 system. Nebulization was performed using a vibrating mesh nebulizer (4–6 μm pore size) over a duration of approximately 36.6 s, generating a uniform aerosol mist. Following deposition, the nebulized formulation was allowed to remain in contact with the tissue surface for 5 min, after which the chamber lid was lifted to facilitate air-drying of the deposited dose. A representative deposition mass of 12.95 μg/cm² was confirmed using the QCM under identical dosing conditions. Permeation was assessed by quantifying naloxone diffusion across the epithelial layer into the receptor compartment, which was filled with KRBB. At predetermined intervals (5, 10, 20, 30, 45, 60, 90, and 120 min), the entire base compartment was removed and replaced with fresh pre-warmed KRBB to maintain sink conditions. Receptor fluid samples were collected in amber glass vials (4 mL) and stored at 4 °C until analysis to minimize photodegradation and oxidative loss. Samples were collected in six replicates at each time point and analyzed to quantify naloxone and its related impurities. All samples were diluted 1,000-fold prior to injection into the LC–MS/MS system. Calibration standards of naloxone, naloxone N-oxide, and noroxymorphone were freshly prepared in the respective receptor matrix of KRBB at concentrations ranging from 0.25 to 20 ng/mL and analyzed concurrently with the permeation samples. Quantification of the analytes was performed using the validated LC–MS/MS method developed in-house.

2.6. Statistical analysis

The main permeation parameters were derived from flux–time and cumulative permeation–time profiles, including maximum flux (Jmax), time to maximum flux (Tmax), and cumulative permeation at the final sampling time (120 min). For each parameter, results are presented as mean ± SD with the number of replicates (n) specified for each model. Between-model comparisons (synthetic membrane vs EpiAirway^™ mucociliary tissue model) were performed using two-sided Welch’s t-test to accommodate unequal variances and unequal sample sizes, with statistical significance defined as p < 0.05. The area under the flux–time curve from 0 to 120 min (AUC_0–120) was calculated by trapezoidal integration for descriptive comparison.

3. Results and discussion

3.1. Development and optimization of the LC-MS/MS analytical method

The analytical protocol was systematically refined to simultaneously quantify naloxone, naloxone N-oxide, and noroxymorphone in IVPT receptor media (DPBS, KRBB), achieving trace-level sensitivity with appropriate reproducibility and throughput necessary for permeability and impurity studies under ICH/FDA guidelines [20]. Using ESI+, precursor and product ion transitions were optimized through direct infusion of standard solutions (naloxone, naloxone-d₅, naloxone N-oxide, and noroxymorphone). Only the major ion transitions were monitored (Figure S1 and Table S2). The cone voltage and collision energy (CE) were stepped across physiologically plausible ranges and tuned using Agilent MassHunter Optimizer software. The selected transitions yielded high signal-to-noise (S/N) and specificity, consistent with best practices in MRM optimization. Source parameters, including desolvation temperature (optimized around 250°C), nebulizer gas flow, and capillary voltage, were systematically adjusted to maximize ionization efficiency and minimize chemical noise, yielding a ~2–3 × increase in signal intensity [12,15]. Dwell times were set to balance sensitivity and data point density, targeting ≥ 12 points per peak based on 4.3 min retention windows [21].

Compared to fully porous particles, superficially porous (core–shell) Agilent Poroshell 120 EC-C18 columns offer higher efficiency and lower backpressure, closer to sub-2 μm performance while enabling faster flow rates [17]. We evaluated a 2.1 × 50 mm, 1.9 μm variant, achieving sharp, symmetric peaks (asymmetry <1.3) and complete baseline resolution (Rs > 1.5) for all analytes within ~4.3 min, an improvement over the 100 mm, 2.7 μm configuration, justifying its use in high-throughput IVPT applications.

A binary mobile phase system consisting of 0.2 % Formic Acid in Water (A) and 0.5 mM Ammonium Acetate in Methanol (B) was applied. Gradient optimization started at 5 % B (0 min), linearly ramped to 30 % B (1.5 min), then flushed to 95 % (2.0 min) B before re-equilibration. This gradient provided strong retention control for both polar impurities and amphiphilic naloxone, while preserving runtime efficiency. Injection volumes were capped at 10 μL (~5.8 % of the column’s theoretical 173 μL hold-up volume) to prevent column overload and preserve peak integrity. Post-run flow diversion effectively prevented matrix interference from entering the MS source [22].

These development efforts culminated in a robust and high-throughput LC–MS/MS analytical workflow characterized by a LLOQ of 0.25 ng/mL, adequate for detecting trace levels of naloxone and its impurities in IVPT receptor matrices. The method provided consistent RT for naloxone (≈ 1.33 min), naloxone N-oxide (≈1.65 min), and noroxymorphone (≈ 0.90 min), with suitable chromatographic performance. MRM selectivity was optimized through adjusted precursor–product ion transitions and fine tuning of source parameters. The use of a short Poroshell 120 EC-C18 column enabled fast and efficient chromatographic separation, with a total run time under 5 min. Moreover, the analytical procedure demonstrated strong retention time reproducibility and analytical stability over 2000 + injections, underscoring suitability for large-batch sample processing and long-term IVPT studies.

3.2. Analytical procedure validation

The LC–MS/MS analytical procedure demonstrated robust and reproducible performance for the simultaneous quantification of naloxone, naloxone N-oxide, and noroxymorphone in IVPT receptor matrices (DPBS and KRBB). SST performed at the beginning of each analytical run confirmed consistent chromatographic and mass spectrometric performance, including stable RT, peak symmetry, and peak response. Across all validation runs, retention time variability remained low, and no evidence of carryover was observed following upper limit of quantification injections (Tables S3 and S4). Method selectivity was demonstrated by the absence of matrix-derived interferences at the RT of naloxone, naloxone N-oxide, noroxymorphone, and the ISTD in blank DPBS and KRBB samples. Representative chromatograms obtained at the LLOQ illustrate clear baseline separation and adequate signal-to-noise ratios for all analytes in both matrices (Figs. 1 and 2). Compound identity and analytical specificity were further supported by the use of optimized quantifier and qualifier MRM transitions monitored under dynamic MRM conditions (Figure S1 and Table S2). Matrix-matched CCs exhibited strong linearity over the validated concentration range of 0.25–20.0 ng/mL in both DPBS and KRBB [23]. Weighted (1/x) linear regression yielded correlation coefficients greater than 0.99 across all validation runs, with consistent back-calculated concentrations observed at each calibration level (Tables S5 and S6). The selected calibration range encompassed concentrations measured in IVPT receptor samples following aerosol dosing and subsequent dilution. Accuracy and precision were evaluated at the lower limit of quantification and at low, mid, and high QC levels in both matrices. Measured concentrations demonstrated acceptable agreement with nominal values across all analytes, with reproducible intra-day and inter-day performance. Accuracy results for naloxone, naloxone N-oxide, and noroxymorphone in DPBS and KRBB are summarized in Tables S7 and S8, while corresponding precision data are presented in Tables S9 and S10. Collectively, these results confirm reliable method performance across the validated concentration range (Tables S7–S10). The lower limit of quantification for naloxone, naloxone N-oxide, and noroxymorphone was established at 0.25 ng/mL in both DPBS and KRBB, with acceptable signal-to-noise ratios, accuracy, and precision demonstrated across replicate analyses (Tables S11 and S12). This level of sensitivity supports quantification of trace-level permeation and accommodates the dilution factors required for IVPT receptor samples. These results are comparable to those reported by Jiang et al. [24], who validated naloxone in mouse plasma with precision ≤ 6.4 % and LLOQ of 0.2 ng/mL, and Krieter et al. [25], who reported a validated range of 10.0 pg/mL – 10.0 ng/mL for nasal naloxone in human plasma with similar analytical performance. In contrast to these published methods, the current approach was specifically optimized and validated for IVPT receptor media (DPBS and KRBB), enabling trace-level quantification under finite-dose permeation conditions while maintaining selectivity in non-biological matrices. This matrix-specific adaptation is critical for accurately characterizing permeation profiles generated in flow-through IVPT systems.

Fig. 1.

Chromatographic separation of naloxone, naloxone-d₅ (ISTD), naloxone N-oxide, and noroxymorphone at their respective LLOQ concentrations in DPBS matrix.

Fig. 2.

Chromatographic separation of naloxone, naloxone-d₅ (ISTD), naloxone N-oxide, and noroxymorphone at their respective LLOQ concentrations in KRBB matrix.

Analyte stability in DPBS and KRBB was evaluated under refrigerated storage conditions to support the stability-indicating capability of the analytical procedure. Naloxone, naloxone N-oxide, and noroxymorphone remained stable over the assessed storage period, with measured concentrations remaining within predefined acceptance limits relative to freshly prepared standards (Tables S13 and S14). Consistent with these findings, naloxone N-oxide and noroxymorphone were not detected in IVPT permeation samples, as their formation was not expected under the controlled experimental conditions employed. This outcome confirms the chemical stability of naloxone during permeation testing and demonstrates that the analytical method could detect potential degradation products had they been present. A consolidated summary of validation performance for all analytes in DPBS and KRBB, including linearity, accuracy, precision, and sensitivity, is provided in Tables 1 and 2, with detailed numerical results presented in the Supplementary Information [26].

3.3. Naloxone permeation across Nuclepore Track-Etched membrane

The selection of the Nuclepore Track-Etched membrane was guided by its complementary capabilities for assessing nasal drug permeation. The Nuclepore Track-Etched membrane provides a well-defined, inert, and highly permeable synthetic substrate that enables mechanistic evaluation of formulation-driven factors, such as aerosol deposition, droplet spreading, wetting behavior, and dissolution-controlled release, under finite-dose conditions. Its simplicity and minimal barrier properties facilitate isolation of physicochemical drivers of early permeation kinetics; however, it does not recapitulate key biological features of the nasal epithelium, including mucus, cellular junctions, and active transport processes.

The amount of naloxone permeated at each time point was analyzed using the validated LC-MS/MS analytical method. Naloxone was detected in the permeation samples with no interferences or ghost peaks from the excipients of the drug product (Figure S2). Fig. 3 show the flux profile and cumulative amount of naloxone permeated across the Nuclepore Track-Etched synthetic membrane in DPBS as a function of time over 120 min. These permeation profiles revealed a rapid and dissolution-driven transport dynamic following aerosol deposition. As illustrated in Fig. 3A, the permeation flux peaked sharply at 5 min post-application, reaching values exceeding 100,000 ng/cm²/min, followed by a steep decline approaching baseline by 30 min. The corresponding cumulative permeation data (Fig. 3B) showed that the total amount of naloxone permeated plateaued at approximately 0.130 mg/cm² by the 30-min mark, indicating near-complete depletion of the available dose from the membrane surface. Statistically, the peak flux displayed considerable variability (mean ~112.6 μg cm⁻² min⁻¹ with an RSD of 49.5 %), as did the cumulative permeation values (mean ~0.13 mg cm⁻² with an RSD of 46.3 %) (Fig. 3). This variability is consistent with the finite-dose, deposition-driven nature of aerosol-based permeation across artificial membranes, where early-time transport is governed by droplet deposition, surface wetting, and rapid film formation. Under these conditions, minor inter-replicate differences in deposited droplet distribution and spreading behavior can disproportionately influence the initial burst phase and, consequently, peak flux and cumulative permeation. Importantly, this variability reflects inherent experimental and model-related factors rather than analytical imprecision, as method validation demonstrated acceptable accuracy and precision. Despite this variability, the overall permeation trend was consistent across replicates.

Fig. 3.

A) Flux of naloxone permeated (ng/cm²/min) B) Cumulative amount of naloxone permeated per surface area of permeation (mg/cm²) in DPBS; n = 3.

These permeation profiles across the synthetic Nuclepore Track-Etched membrane may suggest a dissolution-limited, concentration gradient-driven passive diffusion model. Initially, naloxone was aerosolized as a liquid dispersion using the Vitrocell Cloud Alpha 12 system and subsequently dried on the membrane surface, forming a solid concentrated film. Upon exposure to the aqueous receptor medium, rapid dissolution of this solid film occurred, creating a steep concentration gradient across the membrane. The initial high flux can be explained by Fick’s First Law, where the diffusion rate is directly proportional to the concentration difference between donor and receptor compartments [27]. As naloxone dissolved and diffused, the concentration gradient diminished, resulting in the observed decline in flux and eventual plateauing of cumulative permeation.

These in vitro findings reflect certain key aspects of early pharmacokinetic behavior observed following intranasal administration of naloxone, where peak plasma concentrations typically occur within 15–30 min due to rapid absorption across the highly vascularized nasal epithelium [7,25]. While naloxone nasal sprays deliver the drug as liquid droplets directly to the nasal mucosa, the current IVPT setup involves an intermediate drying step, creating a solid film of naloxone that requires dissolution before diffusion. Despite this methodological difference, the IVPT model replicates essential physiological processes such as aerosol deposition, thin-layer hydration, and passive membrane diffusion, offering valuable mechanistic insights into nasal permeation. While promising, the current IVPT method employing synthetic membrane still requires further optimization and refinement to confirm its biopredictability. These refinements may include incorporation of more physiologically relevant mucosal models, precise control of dosing uniformity, and expansion of permeation time profiles to capture complete drug release kinetics.

3.4. Naloxone permeation across EpiAirway mucociliary tissue model

In contrast to the Nuclepore membrane, EpiAirway^™ model is a differentiated human airway epithelial tissue that incorporates mucociliary function, epithelial barrier integrity, and physiological hydration, providing greater biological relevance for nasal permeation assessment. While this model may reflect in vivo-like barrier properties and may offer acceptable reproducibility relative to synthetic membranes, it is less suited for dissecting formulation-dependent deposition and wetting phenomena and is subject to higher complexity and cost. Together, these models enable a tiered evaluation approach, in which the synthetic membrane supports mechanistic screening, and the tissue-based system provides physiologically relevant confirmation of permeation behavior.

Fig. 4 show the flux profile and cumulative amount of naloxone across EpiAirway^™ mucociliary tissue model to a receptor medium of KRBB as a function of time over 120 min. This permeation profile may reveal a physiologically relevant and diffusion-regulated transport behavior. As shown in Fig. 4A, the flux of naloxone peaked sharply at 5 min, reaching approximately 13.2 μg/cm²/min, before rapidly declining. The cumulative permeation profile (Fig. 4B) plateaued by approximately 45 min at 0.016 mg/cm², suggesting near-complete depletion of available drug from the apical surface. Variability across time points remained moderate, with the peak flux showing a relatively higher standard deviation (RSD), and cumulative permeation values displaying a moderate RSD at plateau. The permeation profile demonstrates rapid transport of naloxone across the biological membrane model, suggesting high permeability immediately following aerosol deposition. The EpiAirway^™ model, composed of differentiated human tracheal/bronchial epithelial cells cultured at the air–liquid interface, possesses tight junctions, cilia, and mucus production, thereby replicating key barrier and clearance mechanisms of the human nasal epithelium [28,29]. The rapid attainment of peak flux reflects efficient dissolution and diffusion of naloxone through the hydrated mucus and epithelial cell layers, while the plateau reflects gradual loss of donor concentration and the cumulative resistance of the tissue to continued diffusion.

Fig. 4.

A) Flux of naloxone permeated (ng/cm²/min) B) Cumulative amount of naloxone permeated per surface area of permeation (mg/cm²) in KRBB; n = 6.

Compared to the synthetic Nuclepore Track-Etched membrane used in the DPBS-based model, the EpiAirway^™ tissue introduces a significantly more complex and biorelevant barrier function. Quantitative comparison of key permeation parameters indicated that the EpiAirway^™ model exhibited a markedly lower maximum flux relative to the synthetic membrane condition (Jmax ~1.13 × 10⁴ vs ~1.03 × 10⁵ ng/cm²/min), while Tmax occurred at an early time point (~5 min) in both systems, consistent with an initial deposition-driven burst followed by rapid decline. Cumulative permeation at 120 min was also lower in the EpiAirway^™ model (~0.016 mg/cm²) compared with the synthetic membrane condition (~0.131 mg/cm²). When evaluated using Welch’s t-test, differences in Jmax and cumulative permeation showed a consistent directional trend but did not reach statistical significance (p ≈ 0.106 and p ≈ 0.085, respectively), reflecting high variability in the synthetic membrane dataset and limited replicate number. The mean flux-based AUC_0–120 further supported reduced overall permeation across the mucociliary tissue model (~1.09 × 10⁵ vs ~9.29 × 10⁵ ng/cm²), consistent with the expected barrier properties and mucociliary interface of the biological system. Nonetheless, while both systems demonstrated rapid onset of flux within 5 min, the overall cumulative amount permeated across the synthetic membrane (0.130 mg/cm²) was approximately eightfold higher than that across the EpiAirway tissue (0.017 mg/cm²). This disparity highlights the influence of cellular tight junctions, mucus, and tissue architecture in attenuating permeation, a feature that is absent in inert synthetic membranes. Such attenuation is physiologically meaningful, as the nasal epithelium serves as both an absorptive and protective barrier against xenobiotics [30,31]. These findings may offer valuable insight into the mechanism of nasal drug availability. The EpiAirway^™ model allows for accurate simulation of mucociliary clearance, epithelial resistance, and active transport barriers that determine systemic absorption of nasally administered drugs like naloxone. This model may be particularly advantageous over synthetic membranes for evaluating formulation performance under biorelevant barrier conditions and may be coupled with pharmacokinetic modeling to estimate in vivo drug input rates [32,33].

It is noteworthy that throughout the IVPT experiments conducted under controlled conditions, no detectable levels of naloxone degradation products or impurities were observed in any of the receptor samples, indicating the chemical stability of naloxone during the permeation process. Nonetheless, both the Nuclepore Track-Etched membrane and EpiAirway^™ mucociliary tissue model, when integrated with biorelevant aerosol delivery via the Vitrocell^® Cloud Alpha 12 system and a validated, sensitive stability-indicating LC–MS/MS method, establish a comprehensive and mechanistically grounded in vitro platform for characterizing nasal naloxone permeation. While the synthetic membrane offers a high-throughput, dissolution-focused assessment of drug release and transport potential, the EpiAirway^™ model more closely recapitulates the structural and functional barrier properties of the human nasal mucosa, accounting for tight junctions, mucus, and ciliary function. This dual-model approach enables robust evaluation of both the rate (flux) and extent (cumulative amount) of naloxone permeation, which are critical for demonstrating BE of systemically acting nasal drug products, particularly in settings where clinical endpoint studies are not feasible. The EpiAirway^™ model may also support the development of pharmacokinetic modeling strategies that improve understanding of systemic exposure following nasal delivery. While the present study does not establish a direct in vitro–in vivo correlation, IVPT-derived parameters such as early-time flux and cumulative permeation provide mechanistic insight into formulation-dependent release and epithelial barrier interactions that are relevant to rapid systemic absorption of naloxone. Although additional refinement of the IVPT method—such as improved discriminatory capability for formulation and process changes, enhanced dosing reproducibility, longer-duration sampling, and increased biological relevance may further strengthen its biopredictive utility, this experimental framework may hold promise as a complementary tool for formulation optimization and assessments of post-approval changes for nasal naloxone products [34].

4. Conclusion

A sensitive, selective, and stability-indicating LC–MS/MS method was developed and validated for the quantification of naloxone, naloxone N-oxide, and noroxymorphone in IVPT receptor media. Application of the method enabled biopharmaceutical assessment of naloxone permeation following cloud-based aerosol dosing across a synthetic Nuclepore Track-Etched membrane and a differentiated human EpiAirway^™ mucociliary tissue model. Distinct permeation behaviors were observed, with rapid transport across the synthetic membrane and attenuated permeation across the epithelial tissue model, reflecting differences in physicochemical and biological barrier properties. Integration of cloud-based aerosol delivery with a validated LC–MS/MS platform enables mechanistic evaluation of nasal naloxone permeation and provides a supportive in vitro framework for formulation characterization and comparative assessments.

Appendix A. Supporting information

Supplementary data associated with this article can be found in the online version at doi:10.1016/j.jpba.2026.117366.