Exploring the Pathogens in Primary Predominantly Antibody Deficiencies of Unknown Genetic Origin

Area of Science:

• Immunology
• Infectious Diseases
• Rare Diseases

Background:

• Primary antibody deficiencies (PADs) are rare, heterogeneous disorders increasing susceptibility to infections, autoimmunity, allergies, and malignancies.
• Delayed diagnosis in PADs leads to increased morbidity and mortality.
• Understanding infection patterns and predictors of bronchiectasis is key for early PAD diagnosis and improved outcomes.

Purpose of the Study:

• To identify infection patterns in patients with PADs.
• To analyze predictors of bronchiectasis presence at PAD diagnosis (BPAD).
• To facilitate earlier diagnosis and improve prognosis for PAD patients.

Main Methods:

• Analysis of infection types and pathogens in PAD patients without identified genetic origin from the ESID Registry.
• Use of generalized mixed-effects logistic regression to identify BPAD predictors.
• Evaluation of five machine learning classifiers for predictive performance.

Main Results:

• Recurrent respiratory infections were predominant in 861 PAD patients, with encapsulated bacteria being most common.
• Microbiological cultures were underutilized (5.9-12.3%).
• Predictors of BPAD included affected organ systems, serious bacterial infections, encapsulated bacteria, and CVID diagnosis; machine learning models showed moderate discrimination (ROC-AUC 0.679-0.746).

Conclusions:

• Recurrent encapsulated bacterial respiratory infections are characteristic of PAD, with significant underutilization of cultures.
• Generalized mixed-effects logistic regression effectively predicted BPAD.
• Clinicians should consider immunologic evaluation for patients with severe/recurrent bacterial infections, multi-system involvement, or repeated isolation of encapsulated bacteria.