Consistency of color-deconvolution for analysis of image intensity of alpha smooth muscle actin-positive myofibroblasts in solid multicystic ameloblastomas

Abstract

ABSTRACT Ameloblastoma is an odontogenic tumor with a slow, locally aggressive growth pattern and multiple clinico-histologic types. The number of stromal myofibroblasts within ameloblastoma often is correlated with growth and aggressiveness. Color-deconvolution to separate different colors of immunostained tissues is a promising approach to quantifying myofibroblasts in tumors such as ameloblastoma. We investigated the reliability of the color-deconvolution method using cross-sectional design to evaluate alpha-smooth muscle actin (α-SMA)-positive myofibroblasts in solid multicystic ameloblastoma. Formalin fixed tissues of eight cases of solid multicystic ameloblastoma were immunostained for α-SMA using the horseradish peroxidase-diaminobenzidine (HRP-DAB) method. Color-deconvolution using ImageJ software was used to quantify the staining intensity of brown DAB α-SMA stained myofibroblasts. Color-deconvoluted images of brown DAB stained tissues exhibited distinct morphological features of solid multicystic ameloblastoma with α-SMA stained myofibroblasts distributed abundantly adjacent to the ameloblastoma epithelial islands. The computed image intensity of α-SMA stained myofibroblasts was quantitatively similar among the different ameloblastoma samples. A combination of color-deconvolution and α-SMA staining of myofibroblasts is a useful diagnostic tool for evaluating histologic differentiation and growth pattern of ameloblastoma.