MECP2 Duplication Syndrome: AI-Based Diagnosis, Severity Scale Development and Correlation with Clinical and Molecular Variables.

Background

MECP2 duplication syndrome (MDS) (MIM#300260) is a rare X-linked neurodevelopmental disorder. This study aims to (1) develop a specific clinical severity scale, (2) explore its correlation with clinical and molecular variables, and (3) automate diagnosis using the Face2gene platform.

Methods

A retrospective study was conducted on genetically confirmed MDS patients who were evaluated at a pediatric hospital between 2012 and 2024. Epidemiological, clinical, and molecular data were collected. A standardized clinical questionnaire was collaboratively developed with input from physicians and parents. Patient photographs were used to train Face2Gene.

Results

Thirty-five patients (0-24 years, 30 males) were included. Key features in males comprised intellectual disability (100%), hypotonia (93%), autism spectrum disorder (77%) and developmental regression (52%). Recurrent respiratory infections (79%), dysphagia (73%), constipation (73%) and gastroesophageal reflux (57%) were common. Seizures occurred in 53%, with 33% being treatment-refractory. The Face2Gene algorithm was successfully trained to identify MDS. A specific clinical severity scale (MECPDup) was developed and validated, correlating with the MBA (a scale developed for Rett syndrome). The MECPDup score was significantly higher in males (p < 0.001) and those with early death (p = 0.003). It showed significant positive correlations with age (p < 0.001) and duplication size (p = 0.044).

Conclusions

This study expands the understanding of MDS through comprehensive clinical and molecular insights. The integration of AI-based facial recognition technology and the development of the MECPDup severity scale hold promise for enhancing diagnostic accuracy, monitoring disease progression, and evaluating treatment responses in individuals affected by MDS.