JESREC score and mucosal eosinophilia can predict endotypes of chronic rhinosinusitis with nasal polyps

Abstract

ObjectiveRecently, JESREC score and mucosal eosinophil count have been used to diagnose eosinophilic chronic rhinosinusitis (ECRS) in Japan. However, it remains unknown whether the subtypes of CRS diagnosed by these criteria have different endotypes. In the present study, we investigated whether JESREC score and mucosal eosinophil count were appropriate for classification of CRS subgroups into endotypes. MethodsA cross-sectional study involving 71 consecutive patients with CRS with nasal polyps (CRSwNP) and 13 control patients was performed. Nasal polyp tissues from CRSwNP patients and uncinate process tissues from control patients were collected for analysis of inflammatory cells by immunohistochemistry and measurement of cytokines and chemokines by ELISA and quantitative real-time PCR. We compared the differences between subtypes according to JESREC score and mucosal eosinophil count and investigated the subgroups with different endotypes by cluster analysis and principal component analysis. ResultsIn the 71 CRSwNP patients, 9 patients had JESREC score <11 and mucosal eosinophil count <70/HPF (Group A), 20 patients had JESREC score ≥11 and mucosal eosinophil count <70/HPF (Group C), and 42 patients had JESREC score ≥11 and mucosal eosinophil count ≥70/high-power field (HPF) (Group D). Semiquantitative analysis of inflammatory cells showed that eosinophils, neutrophils, macrophages, mast cells, and basophils differed significantly between the subgroups. At the mRNA level, CLC, IL5, IL13, CCL11, CCL24, CCL26, POSTN, CSF3, and IL8 showed significant differences. At the protein level, eotaxin-2/CCL24, eotaxin-3/CCL26, and G-CSF had significant differences. Cluster analysis using gene expression levels in 55 CRS patients and 11 control patients revealed that the patients could be classified into five clusters. Cluster 1 (n=27) contained all patients with Group D. Cluster 2 (n=11) comprised all control patients. Cluster 3 (n=4) included mixed subtypes: one with Group A and three with Group D. Cluster 4 (n=7) and Cluster 5 (n=17) contained all patients with Groups A and C, respectively. Furthermore, the principal component analysis revealed that the subtypes had different characteristics. ConclusionCRS subtypes based on JESREC score and mucosal eosinophil count showed different inflammatory patterns, and unsupervised statistical analyses supported the classification that can predict endotypes. From these results, we concluded that the classification based on JESREC score and mucosal eosinophil count was useful for predicting CRS endotypes.