Optimized Teucrium polium topical cream improves wound closure in a rat excisional model: A D-optimal design and RBF modelling approach

Chronic wounds require topical formulations that combine favorable physicochemical properties with measurable biological activity. In this study, a bio-based, oil-in-water cream containing Teucrium poliummethanolic extract was developed and optimized using a D-optimal design and a radial basis function (RBF) modeling approach. Three formulation variables, extract content (X1), cetyl alcohol (X2), and shea butter (X3), were evaluated in relation to two responses: spreadability (Y1) and wound closure on Day 7 (Y2). Predictive modeling was performed using an RBF network with leaveone- out cross-validation (LOOCV), which provided an internal estimate of predictive performance (R? training = 0.98-0.99; R? test = 0.95-0.96). Multi-response optimization identified an optimal formulation containing 8.0% extract, 1.5% cetyl alcohol, and 1.0% shea butter. Experimental confirmation yielded a spreadability of 54.8?0.9 cm? and 84.1?1.2% wound closure on Day 7. In the confirmatory in vivo comparison, a vehicle control, a positive control (Madecassol 1%), and selected formulations were followed through Day 14. F6, F11, and F18 showed the most favorable macroscopic wound-closure profiles, whereas F2 displayed a more moderate effect. These findings support the use of a multivariate optimization strategy for the rational development of bio-based topical creams, while the biological interpretation remains limited to wound closure under the tested conditions.