DDX3X Syndrome: Clinical, Neuroimaging, AI-Assisted Facial Profiling and Genotype-Phenotype Correlations

Area of Science:

• Genetics and Genomics
• Neurodevelopmental Disorders
• Medical Diagnostics

Background:

• DDX3X syndrome (MIM#300958) is a neurodevelopmental disorder primarily affecting females, characterized by intellectual disability, language impairment, and specific neurobehavioral traits.
• Previous reports noted dysmorphic features, but a consistent facial phenotype and clear genotype-phenotype correlations remained unestablished for DDX3X syndrome.
• Establishing a recognizable facial phenotype and genotype-phenotype correlations is crucial for improving the diagnosis and understanding of DDX3X syndrome.

Purpose of the Study:

• To identify and characterize the facial phenotype associated with DDX3X syndrome in affected individuals.
• To explore potential genotype-phenotype correlations in DDX3X syndrome.
• To evaluate the utility of automated facial analysis software (Face2Gene) as a diagnostic aid for DDX3X syndrome.

Main Methods:

• An observational, ambispective, descriptive study was conducted on 9 patients (8 females) aged 0-18 years with a molecular diagnosis of DDX3X.
• Clinical data, standardized facial images, neurobehavioral assessments, neuroimaging, and molecular data were collected.
• Automated facial analysis was performed using Face2Gene after algorithm training; in silico analysis was conducted for variants of uncertain significance.

Main Results:

• Frequent facial features included thin upper lip (100%), tapered chin (89%), long uniform eyebrows (89%), short neck (89%), and long face (67%).
• Face2Gene identified DDX3X syndrome in 92% of cases within the top 5 suggestions, demonstrating its diagnostic utility.
• All females exhibited intellectual disability and language disorder; loss-of-function variants correlated with greater clinical severity.

Conclusions:

• This study suggests a recognizable facial phenotype for DDX3X syndrome, aiding in clinical identification.
• A potential genotype-phenotype correlation is supported, with loss-of-function variants linked to increased severity.
• Automated facial analysis shows promise as a valuable diagnostic tool for DDX3X syndrome.