Image analysis of platelet derived growth factor receptor-beta (PDGFRβ) expression to determine the grade and dynamics of myelofibrosis in bone marrow biopsy samples Running title: PDGFRβ expression in myelofibrosis – image analysis

Background: Myelofibrosis (MF) is characterized by accumulation of stromal cells and extracellular matrix. Progression of fibrosis is an important clinical issue and monitoring is required for new therapeutic approaches. Currently the quantification is based on semiquantitative evaluation of reticulin silver stained slides. We recently reported that platelet derived growth factor receptor beta (PDGFRβ) expression in fibroblasts is a useful marker of stromal activation. PDGFRβ expression based scores represent significant differences in different myelofibrosis grade which provides optimal source of quantification. In this study slide based measurements were performed in order to support correlations of PDGFRβ expression with MF grade. Methods: Scanned image tiles from 79 bone marrow samples (BM) with different myelofibrosis grades were evaluated for PDGFRβ-related IHC parameters. Following the determination of immunopositive (brown component) and total area (region of interest) of the BM, PDGFRβ related image parameters were defined and evaluated in comparison to the classical reticulin based grading. Results: Eight PDGFRβ expression related image parameters showed excellent correlation with the MF grade (correlation coefficient ranging between 0.79 and 0.83) and with PDGFRβ score (0.76-0.87). Despite the significant sample heterogeneity the parameters showed significant differences between fibrotic and non-fibrotic cases and between mild and advanced fibrosis. Distribution of values within a particular specimen emphasizes the heterogeneity of bone marrow involvement which may cause difficulties in semi-quantitative methods. Page 2 of 31 John Wiley and Sons, Inc. Cytometry: Part B Clinical Cytometry