*2.6. Interpretation of Immunohistochemical Staining*

The immunohistochemical stainings were interpreted by two authors (EY Cho and D-H Kim) in a double-blinded fashion, and samples showing poor inter-rater reliability (κ < 0.20) were removed from data analysis. RUNX1 expression was considered positive when nuclear staining was present, and the intensity and proportion of positive nuclear staining was assessed for scoring. A composite score of RUNX1 protein expression was semi-quantitatively calculated by multiplying the proportion score of positive cells (0, absent; 1, 0%–10%; 2, 10%–50%; 3, 50%–80%; and 4, >80%) with the staining intensity score (0, none; 1, weak; 2, moderate; and 3, strong). A cutoff for the reduced expression of RUNX1 protein was determined by taking into account the distribution of the composite scores between normal and tumor tissues and by comparing both false-negative and false-positive rates at different cutoffs. RUNX1 expression was considered reduced in a tumor with a composite score less than two. To score for phospho-pRb (Ser-807/811) and Ki-67, positive staining was determined according to the percentage of positively stained nuclei.
