Factor Analysis-Based Quantitative Endotyping Improves Associations With CRS Cross-Sectional and Longitudinal Outcomes.

Abstract

Background: Chronic rhinosinusitis (CRS) is a heterogeneous inflammatory disease characterized by persistent sinonasal inflammation. There is increasing interest in endotype-based classification, which categorizes CRS based on underlying inflammatory pathways. We applied factor analysis to facilitate continuous endotype assignment to CRS patients and assess cross-sectional and longitudinal CRS outcomes.

Methods: We prospectively enrolled 203 patients (52 controls, 88 CRSsNP, and 63 CRSwNP) undergoing endoscopic nasal or sinus surgery (ESS). Middle meatal mucus biomarkers were analyzed pre-ESS (V0) and 6-12 months post-ESS (V1). Factor analysis was performed to identify latent factors. Factor scores were generated, and statistical analyses were conducted to assess correlations with radiographic (Lund-Mackay [LM]) and patient-reported (SNOT-22, CRS-PRO) outcomes measured at V0, V1, and V2 (1.5-5 years post-ESS).

Results: Four factors were identified: Type 1 (T1), Type 2 (T2), Type 3 (T3), and macrophage-associated (M). CRSwNP patients had higher T2 and M factor scores than CRSsNP and controls. T2, T3, and M factor scores demonstrated stronger or equivalent associations with radiographic and patient-reported outcomes compared to individual biomarkers. Following ESS, median T2 and M factor scores significantly declined, while T1 and T3 remained stable. V0 T1 and T2 were weakly associated with long-term (V2) radiographic and symptom scores. V1 factor scores were more consistently predictive of long-term (V2) outcomes, with T2, T3, and M demonstrating modest correlations with radiographic severity and CRS-PRO.

Conclusions: Factor analysis identifies distinct, quantifiable patterns of inflammation in CRS, offering improved associations with cross-sectional and longitudinal outcomes compared to individual biomarkers.