A diagnostic model for predicting type 2 nasal polyps using biomarkers in nasal secretion

Area of Science:

• Otorhinolaryngology
• Immunology
• Medical Diagnostics

Background:

• Accurate prediction of type 2 chronic rhinosinusitis with nasal polyps (CRSwNP) is crucial for effective treatment selection.
• Currently, a reliable non-invasive method for diagnosing type 2 CRSwNP is lacking.

Purpose of the Study:

• To develop and validate a non-invasive diagnostic model for type 2 CRSwNP.
• To identify optimal methods for collecting nasal secretions for biomarker analysis.

Main Methods:

• Optimized nasal secretion (NasSec) collection using polyvinyl alcohol (PVA) nasal packs.
• Defined Th2-low and Th2-high CRSwNP clusters using tissue biomarkers in 142 patients.
• Developed machine learning models (logistic regression, decision tree) to predict Th2-high CRSwNP using NasSec biomarkers.

Main Results:

• PVA nasal packs showed superior performance for NasSec collection based on IL-5 levels.
• Logistic regression and decision tree models achieved high AUCs (0.92 and 0.90) for predicting Th2-high CRSwNP.
• A decision tree model using IL-5 levels, blood eosinophil count, and asthma status was selected for clinical use.

Conclusions:

• A decision tree model integrating NasSec biomarkers and clinical features accurately defines type 2 CRSwNP.
• This non-invasive approach can guide therapeutic decisions in clinical practice.
• The model aids in distinguishing between Th2-low and Th2-high CRSwNP phenotypes.