Improving Ex Vivo Nasal Mucosa Experimental Design for Drug Permeability Assessments: Correcting Mucosal Thickness Interference and Reevaluating Fluorescein Sodium as an Integrity Marker for Chemically Induced Mucosal Injury.

: Ex vivo nasal mucosa models provide physiologically relevant platforms for evaluating nasal drug permeability; however, their application is often limited by high experimental variability and the absence of standardized methodologies. This study aimed to improve experimental design by addressing two major limitations: the confounding effects of mucosal thickness and the questionable reliability of fluorescein sodium (Flu-Na) as an integrity marker for chemically induced mucosal injury. : Permeability experiments were conducted using porcine nasal tissues mounted in Franz diffusion cells, with melatonin and Flu-Na as model compounds. Tissues of varying thickness were collected from both intra- and inter-individual sources, and a numerical simulation-based method was employed to normalize apparent permeability coefficients (Papp) to a standardized mucosal thickness of 0.80 mm. The effects of thickness normalization and chemically induced damage were systematically evaluated. : Thickness normalization substantially reduced variability in melatonin Papp, particularly within same-animal comparisons, thereby improving statistical power and data reliability. In contrast, Flu-Na exhibited inconsistent correlations across different pigs and failed to reflect the expected increase in permeability following isopropyl alcohol (IPA)-induced epithelial damage. These results suggest that the relationship between epithelial injury and paracellular transport may be non-linear and not universally applicable under ex vivo conditions, limiting the suitability of Flu-Na as a standalone marker of mucosal integrity. : The findings highlight the importance of integrating mucosal thickness correction into standardized experimental protocols and call for a critical reassessment of Flu-Na in nasal drug delivery research.