A Novel Predictive Model for the Recurrence of Pediatric Alopecia Areata by Bioinformatics Analysis and a Single-Center Prospective Study

The study developed a predictive model for the recurrence of pediatric alopecia areata (AA) using bioinformatics and a prospective study with 80 children. It identified four key genes (CD8A, PRF1, XCL1, and BMP2) as biomarkers correlated with AA severity. These biomarkers were used in a logistic regression model, achieving high accuracy in predicting AA recurrence with an AUC of 0.854. The model offers a non-invasive method to forecast AA recurrence, potentially improving patient prognosis. However, limitations include the use of adult AA skin samples for bioinformatics analysis and the small sample size, suggesting the need for larger, multicenter studies for validation.