Cellsens analysis software

For measurement of maximal mucosal thickness and epithelial hyperplasia, sections were stained using hematoxylin and eosin (H&E) and analysed in a blinded manner for group assignment by pathologist of Asan Medical Center. Epithelial hyperplasia was scored on none (0) to severe (4) basis, as described.⁸⁵ Maximal mucosal thickness was assessed at the transition zone between respiratory and olfactory epithelium (CellSens analysis software, Olympus). Eosinophils were analysed after staining with Sirius Direct Red 80 (Sigma–Aldrich, 365548), and eosinophil infiltration into the sinonasal mucosal was measured by counting the cell number per high‐powered field, as previously described.³¹ Mast cells were analysed after acidic toluidine blue staining, as described.²¹

The specimens were embedded in paraffin, cut into 4 µm thick sections, and stained with hematoxylin and eosin. Sections from 3 different depths for each specimen were used to obtain a representative sample of the entire specimen. The evaluation was performed by calculating the quantitative score for synovitis as established by Krenn et al.^{[18 (link)]} The cell layer linings of the synovium were classified into the following 3 types: 1 layer of lining as normal, 2 to 3 layers with lacuna between the extracellular matrix and itself as mild, and more than 4 layers with eosinophil-rich cytoplasm as severe (Fig. 1). Almost all specimens showed a cell lining layer with a mixed grade. Therefore, 2 authors (I.T. and T.M.) measured the length of the normal and severe types and the total length of the synovial lining in each specimen using image analysis software (CellSens analysis software; Olympus, Tokyo, Japan). The percentages of normal and severe types in each section were calculated by dividing the length of the normal and severe areas by the total length. The intra- and inter-observer correlation coefficients for the lengths of the 2 types were 0.88 and 0.82, respectively.