Label-free quantification of topical drug permeation in skin using sparse spectral sampling stimulated Raman scattering imaging.

Significance: Stimulated Raman scattering (SRS) imaging has been used in evaluating topical drug product permeation. However, it has been mainly applied to a small group of active pharmaceutical ingredients (API) that have unique molecular bands which generate a prominent SRS signal against the background signal. The requirement of such unique molecular bands has been a large hurdle to generalizing SRS imaging to study a wide range of pharmaceutical molecules. Aim: To overcome this barrier, an imaging method based on sparse spectral sampling stimulated Raman scattering (S4RS) microscopy and multivariate analysis was developed in this study. Approach: Rather than relying on a single unique and strong peak, the API was specifically resolved using combined information from multiple wavenumbers across its spectrum. The workflow of applying this method to study APIs is demonstrated using metronidazole as an example. The sparse vibrational bands were first selected using Elastic-Net. Next, the metronidazole signal was unmixed from the compound signal of the skin-drug product mixture using multivariate curve resolution-alternating least squares (MCR-ALS). Results: The developed method demonstrated good capability in identifying the metronidazole signal in standard metronidazole-containing samples, including polymer films and skin samples treated with and without the drug product. This method enabled evaluation of the cutaneous pharmacokinetics (cPK) of metronidazole in Conclusions: The developed method enables the study of APIs that lack a unique prominent peak, thereby bridging SRS imaging to a broader range of drug products.