(T1330-07-38) Stimulated Raman Scattering (SRS) Microscopy and Deep Learning: Novel Pharmacokinetic Approach for Evaluation of Topical Bioequivalence

Purpose: The bioequivalence (BE) of pharmaceutical formulations is typically demonstrated by assessing the pharmacokinetics (PK) of a generic product relative to a reference listed drug (RLD) product. For drugs intended to act systemically, BE is normally assessed by estimating PK metrics from drug concentrations in the plasma; however, for topical products applied to the skin, it has been historically challenging to quantify the drug concentration at the target site of action in the skin, and therefore the applicability of the PK-based approach has been limited. Topical BE is often assessed using clinical end-point studies, which are known to be expensive and time consuming. The development of new methods capable of determining the amounts of drug at or near the site of action, i.e., in the stratum corneum (SC), viable epidermis, or deeper cutaneous tissues, has been a key area of research in the recent past. Stimulated Raman scattering (SRS) is a non-destructive optical method based on the coherent Raman scattering phenomenon. SRS signal detection has minimal background interference and the signal intensity generated is linear with the molecular concentration. These properties enable label-free quantitative visualization of drug uptake to the skin as a function of depth at rapid imaging rates. Here, we present a novel approach for evaluating local tissue amounts of drug following topical application to the skin based on SRS imaging and subsequent data processing by deep learning.

Methods: In the present study, two commercially available formulations containing tazarotene (Taz) topical cream, 0.1% were evaluated. The formulations examined were Tazorac^® (topical) cream and Tazarotene (topical) cream (from Taro Pharmaceuticals), which are the RLD and approved generic drug product, respectively. A third treatment group consisting of Tazorac^® (topical) cream was additionally used to evaluate the data from the RLD against itself. Finite doses of the formulations (5µL/cm²) were applied to human skin ex vivo, the dose was occluded, and drug permeation was monitored over 7 hrs by SRS imaging. The tissue was immediately placed on a glass bottom imaging dish within an incubation chamber (32°C) and XYZTλ data were acquired such that the frequency of the laser system was tuned at pre-defined time intervals to target either lipid (2864 cm^-1) or drug (1590 cm^-1) resonant features, based on the multi-window sparse spectral sampling SRS methodology.¹ Data processing was performed via deep learning using a U-Net convolutional neural network for image feature extraction and automated analysis of cutaneous PK parameters, i.e., area under the curve (AUC), time to maximum concentration (t_max) and maximum concentration (C_max) for lipid- and water-rich areas of the SC.²

Results: Penetration of Taz to the skin was evident for all formulations evaluated. The concentration-time profiles indicated that the RLD delivered similar amounts of drug across the skin compared with both itself and the generic product. The PK values of AUC and C_max were also found to be similar for these formulations.

Conclusion: The findings of this proof-of-concept study demonstrate that SRS is a powerful tool that allows reliable assessment of drug in the skin. The proposed approach also enables the imaging of drug distribution across the tissue in real time, which can additionally provide insights into the permeation pathways. Studies are ongoing with additional drug products, including alternate formulations to examine the sensitivity and robustness of SRS and further explore this method as a novel cutaneous PK-based approach for evaluation of BE.

References: [1] Pence, I. J., Kuzma, B. A., Brinkmann, M., Hellwig, T., and Evans, C. L., “Multi-window sparse spectral sampling stimulated raman scattering microscopy,” Biomedical Optics Express 12(10), 6095–6114 (2021).
[2] Feizpour, A., Marstrand, T., Bastholm, L., Eirefelt, S., and Evans, C. L., “Label-free quantification of pharmacokinetics in skin with stimulated raman scattering microscopy and deep learning,” Journal of Investigative Dermatology 141(2), 395–403 (2021).

Acknowledgements: We thank the helpful discussion and proofreading by members of the Evans Group. This project is supported by the Food and Drug Administration (FDA) of the U.S. Department of Health and Human Services (HHS) as part of a financial assistance award [FAIN] totaling $1,500,000 with 100 percent funded by FDA]/HHS. The contents are those of the author(s) and do not necessarily represent the official views of, nor an endorsement, by FDA/HHS, or the U.S. Government.