Novel Chronic Rhinosinusitis (CRS) Subtypes Based on Clinical Features and Response to Treatment.

Abstract

Current classification of chronic rhinosinusitis (CRS) is based primarily on the absence or presence of nasal polyposis. To (1) examine novel CRS classifiers based on clinical characteristics and (2) examine whether these classifiers predict response to treatment. Between 2004 and 2020, patients seen initially at Massachusetts General Hospital meeting the CRS definition (N = 942 cohort) were analyzed retrospectively for clinical characteristics. Patients seen through 2016 to 2020 (N = 312 cohort) were analyzed longitudinally for their response to treatment, including antibiotics; oral steroids; dilute budesonide rinses; instillation of budesonide, itraconazole, gentamicin, or tobramycin; and biologic treatment. Treatment responses and an "end-of-treatment" physician global assessment were recorded on a 0 to 3 scale. Results were analyzed by CRS subgroups and novel subset classifiers. Principal-component analysis was used to further interrogate CRS classifiers. Eosinophilic mucin was present in 134 patients, including 40 (4.2%) with allergic fungal rhinosinusitis and 94 (10%) with eosinophilic mucin rhinosinusitis. Relative to patients with CRS with nasal polyps but without eosinophilic mucin rhinosinusitis, patients with eosinophilic mucin rhinosinusitis have a higher burden of illness and a poorer response to topical steroids but comparably good responses to dupilumab. Patients with POLYPOID_CRS, characterized by intrasinus polypoid tissue without frank nasal polyposis, have a mixed infectious and T_H2 inflammatory phenotype with a disproportionately high burden of bacterial infection and a 65.8% prevalence of tissue eosinophils. Patients with CRS with a history of recurrent acute rhinosinusitis or CRS with Staphylococcus aureus or Gram-negative nonacute sinusitis pathogen infection ("CRS with pathogens") have a higher burden of illness and poorer response to treatment. Historic features associated with "poor" responders include Gram-negative nonacute sinusitis pathogen infection, oral steroid use more than every 4 months, and 2 or more previous surgeries. Principal-component analysis confirmed a strong inverse relationship between both bacterial infection and number of previous surgeries with response to treatment. Overall, medical treatment reduced the frequency of sinus surgery compared with historic frequencies and was beneficial with end-of-treatment physician global assessment scores of 2 or higher in 78.5% of patients. Novel classifiers of CRS can be defined on the basis of clinical characteristics and response to medical treatment.