SNOT-22–based clusters in chronic rhinosinusitis without nasal polyposis exhibit distinct endotypic and prognostic differences

Background: Endotypic and prognosticating features of chronic rhinosinusitis without nasal polyposis (CRSsNP) are poorly understood. Our objectives were to use an unbiased symptom-based approach to: (1) study symptoms, clinical and endotypic features; and (2) identify features predicating outcomes from endoscopic sinus surgery (ESS). Methods: Clinical, computed tomography (CT), histopathology, and 22-item Sino-Nasal Outcome Test (SNOT-22) data was collected on 146 adult CRSsNP patients who underwent ESS. Unsupervised network modeling of presurgical SNOT-22 scores was performed to classify symptom-based clusters. Subject characteristics and post-ESS SNOT-22 scores were compared between clusters. Results: Baseline characteristics of the subject population were as follows: females, 56.2%; revision ESS status in 35%; asthma prevalence, 32.6%; median Lund-Mackay CT score, 8; and median SNOT-22 total score, 43. Network mapping and unsupervised clustering of preoperative SNOT-22 scores revealed 4 clusters: (A) severely burdened with high scores in all 4 subdomains; (B) moderately burdened with high scores in the rhinologic subdomain; (C) moderately burdened with high scores in psychological-sleep subdomains; and (D) mildly burdened. The number of previous ESS and asthma prevalence differed significantly between clusters; CT scores were similar. Asthma burden and tissue eosinophilia were greatest in cluster A (p = 0.03). All groups showed significant improvement at 3 months post-ESS (p < 0.0001). At 6 months, patients in cluster C tended to worsen. Conclusion: SNOT-22–based network modeling of CRSsNP patients yielded 4 clusters with distinct features. Asthma prevalence and tissue eosinophilia were highest in the cluster with highest SNOT-22 scores. All patients showed significant improvement from ESS at 3 months; those with high sleep-psychosocial symptoms tended to show worsening at 6 months.