A classification of chronic rhinosinusitis with nasal polyps based on structured histopathology

Aims In chronic rhinosinusitis with nasal polyps (CRSwNP), tools based on objective evidence, such as histopathology, are needed to assist clinical decision‐making. The main aim of this exploratory investigation was to determine whether structured histopathology could be used to classify CRSwNP in homogeneous histological clusters. Methods and results A cohort of 135 CRSwNP patients was assessed, on the basis of clinicopathological features: allergic fungal rhinosinusitis (17 patients); non‐steroidal anti‐inflammatory drug‐exacerbated respiratory disease (19 patients); intrinsic asthma (18 patients); extrinsic asthma (21 patients); allergy (21 patients); histologically eosinophilic (22 patients); and histologically non‐eosinophilic (17 patients). For structured histopathology, we considered: the degree of inflammation; eosinophil count; eosinophil aggregates; neutrophil infiltration; goblet cell hyperplasia; basement membrane thickening; fibrosis; hyperplastic/papillary changes; squamous metaplasia; mucosal ulceration; and subepithelial oedema. Cluster analysis identified four distinct sets of cases. On discriminant analysis, the global error rate was 1.48%, and the stratified error rates were 4.34%, 0%, 0%, and 0% for clusters 1, 2, 3 and 4, respectively. Cluster 1 was characterised by infrequent fibrosis (<4.5% of cases). Cluster 2 mainly featured neutrophil infiltration in 100% of cases, hyperplastic/papillary changes in 70% of cases, and fibrosis in 65% of cases. Cluster 3 showed fibrosis in 100% of cases. Cluster 4 showed hyperplastic/papillary changes in 100% of cases, and fibrosis in 92% of cases. Conclusions This study shows that cluster analysis can identify different histotypes among CRSwNP patients. The next step will be to investigate, in a larger series, the clinical (e.g. prognostic) implications of identifying such homogeneous clusters of patients with CRSwNP on the basis of their structured histopathology.