Polymer Selection Governs Nanocarrier Performance in Nose-to-Brain Delivery: A Unified Box–Behnken Comparison of Chitosan, Polycaprolactone and PLGA Systems for Two Antidepressants

Abstract

Carrier polymer is routinely chosen by convention rather than by evidence in nose-to-brain nanomedicine, and the absence of matched comparisons makes the consequences of that choice difficult to judge. This work addresses the deficit directly by developing four antidepressant-loaded nanoparticulate systems under a single, identical experimental framework. Viloxazine and venlafaxine were each encapsulated in a cationic polysaccharide (chitosan) and in a synthetic aliphatic polyester (polycaprolactone for viloxazine, poly(lactide-co-glycolide) for venlafaxine), giving the systems designated VXCN, VPNS, VFCN and VPLNP. Every system was optimised through a three-factor, three-level Box–Behnken design of seventeen runs, with particle size, entrapment efficiency and polydispersity index as responses and identical characterisation applied throughout.

All four quadratic models were statistically significant (F = 15.41–164.86; p ≤ 0.0008) with non-significant lack of fit, and confirmation batches prepared at the software-predicted optima agreed with prediction to within 4.8% for every response. The optimised carriers differed systematically by polymer class rather than by drug. Chitosan systems consistently produced the smallest particles (160.0 and 170.7 nm) with cationic or weakly anionic surfaces suited to mucoadhesion, but the lowest entrapment for viloxazine (70.1%). The polyesters delivered higher payloads — 78.12% for polycaprolactone and 83.61% for PLGA — with PLGA additionally furnishing the strongest electrokinetic stabilisation (−24.9 mV) at a size comparable to chitosan (172.0 nm). Polycaprolactone was the outlier, giving a particle size roughly twice that of the other three systems (323.7 nm) and falling outside the dimensional window generally regarded as favourable for transmucosal uptake.

Chitosan nanoparticles were converted into Carbopol-934 in-situ gels; batches VXCN-NPG4 and VFCN-G5 combined nasal-compatible pH (5.58 and 5.71), sol–gel transition within the physiological window (31.4 and 29.8 °C) and the highest drug content of their series. Across excised goat nasal mucosa, all optimised carriers markedly exceeded the corresponding plain drug dispersions, reaching 96% cumulative permeation at 24 h against 60% for unformulated drug, while histopathology confirmed that mucosal architecture remained intact. Formulations were physically stable over three months.

The study demonstrates that polymer identity, not drug identity, is the dominant determinant of nanocarrier quality attributes in this delivery route, and that the trade-off between mucoadhesive potential and encapsulation capacity is systematic and predictable. These findings provide a rational, statistically grounded basis for carrier selection in intranasal CNS formulation development.