International Journal of Biomedicine (Dec 2024)
Expression of Androgen, Estrogen, and Progesterone Receptors in the Skin of Patients with Severe Acne and the Assessment of Their Predictive Potential Using Artificial Intelligence Methods
Abstract
Background: In recent years, the study of acne pathogenesis has continued, and genetic regulation, innate and adaptive immune responses, and the contribution of several endocrinologic mechanisms have been reported. A sebaceous-hair follicle (SHF) is known to be regulated by sex hormones, including androgens, estrogens, and progesterone. At present, there is evidence of the involvement of sex hormones in the pathogenesis of acne. Still, there is no data on immunohistochemical (IHC) analysis of sex hormone receptor expression in the skin of severe acne (SA) in the available literature. There is also no IHC analysis of predictors of acne development. The present study aimed to determine and analyze the immunohistochemical expression of androgen, estrogen, and progesterone receptors in the skin of SA patients and to evaluate their predictive potential using artificial intelligence methods. Methods and Results: This prospective, open, non-randomized, single-center comparative study was conducted between 2019 and 2023. The study included 53 patients (the main group [MG]) with SA and 21 apparently healthy individuals (the comparison group [CG]). Participants were between 15 and 46 years old (the median age was 22.0 years). An IHC study was performed in the Pathomorphological Department at the NMRC PHOI, named after Dmitry Rogachev (Moscow, Russia). Skin samples from patients were collected by punch biopsy. To perform the IHC study, we used the mouse monoclonal anti-androgen receptor (AR) antibody (Clone AR441, diluted 1:50; DAKO, Denmark), mouse monoclonal anti-estrogen receptor (ER) antibody (Clone 1D5, RTU; DAKO, Denmark), and mouse monoclonal anti-progesterone receptor (PR) antibody (Clone PgR636, RTU; DAKO, Denmark). For each marker under study, positivity was determined in three compartments: epidermal keratinocytes, dermal fibroblasts, and sebocytes of sebaceous glands. Quantitative AR, ER, and PR expression assessment was performed using QuPath image analysis software according to the manufacturer's protocol. Statistical analysis was performed using Python 3.11. The predictive potential of the studied IHC markers was assessed using mathematical modeling methods and artificial intelligence. Our study revealed for the first time a significant increase in the AR expression in epidermal keratinocytes and dermal fibroblasts in SA patients compared to healthy individuals, with no significant difference in the AR expression in sebocytes of sebaceous glands and overall positivity. A significant ER overexpression in dermal fibroblasts of SA patients with no significant differences for other studied compartments and overall positivity was also found. Moreover, the studied compartments had no statistically significant differences in progesterone receptor expression. Conclusion: Mathematical modeling methods using artificial intelligence have made it possible to establish for the first time that the AR expression in dermal fibroblasts is a significant predictor of severe acne.
Keywords
