Developing efficient predictive models for the diagnosis of VEXAS syndrome

Area of Science:

• Hematology
• Genetics
• Immunology

Background:

• VEXAS syndrome is an emerging autoinflammatory disease characterized by somatic mutations in the UBA1 gene.
• Diagnosis can be challenging due to overlapping symptoms with other conditions.
• Identifying reliable diagnostic markers is crucial for timely intervention.

Purpose of the Study:

• To identify clinical and laboratory features associated with UBA1 mutations.
• To develop predictive models for efficient VEXAS syndrome diagnosis.
• To aid clinicians in identifying patients with pathogenic UBA1 mutations.

Main Methods:

• Retrospective analysis of 144 patients undergoing UBA1 mutation testing.
• Evaluation of clinical and laboratory features at symptom onset and time of testing.
• Development of predictive models using the LASSO method.

Main Results:

• Skin rash, chondritis, and monocytopenia were significant features at both timepoints.
• Macrocytosis, uveitis, and pulmonary disease were associated at time of testing.
• LASSO models showed good to excellent discriminatory capacity (AUCs 0.81-0.92).

Conclusions:

• Multi-feature models effectively differentiate VEXAS syndrome cases.
• These models can help overcome diagnostic challenges and lack of consensus on key features.
• The developed models offer potential for earlier and more accurate VEXAS syndrome diagnosis.