Automated histological image analysis for the assessment of bleomycin-induced pulmonary fibrosis in rodents

Introduction: Bleomycin (BLM)-induced lung fibrosis is currently the main pre-clinical rodent model in use for the screening of potential therapeutics in Idiopathic Pulmonary Fibrosis (IPF). Typically, histological lung fibrosis scoring is observer-dependent and consequently subject to variability, precluding comparative studies between different laboratories. Thus, a reliable and observer-independent quantification of lung fibrosis is needed. Aims and Objectives: We developed automated image Analysis Protocol Packages (APPs) by using a flexible software (Visiopharm) to quantify fibrotic lung lesions and collagen deposition in digital images of lung sections from mice and rats treated with a double intratracheal administration of BLM. Methods: The first APP was set-up to automatically quantify the fibrotic tissue, detecting accumulation of connective tissue, excess of collagen deposition and cellular hyperproliferation in Masson's trichrome stained lung sections. The second APP was developed to specifically quantify collagen deposition on Mallory-Azan's trichrome stained slices. Results: We confirmed a significant increase in BLM-induced lung fibrosis and collagen deposition in both mice and rats at day 14 and 21 post-BLM, respectively, with a significant correlation with the Ashcroft score. Moreover, the antifibrotic effect by oral Nintedanib (60 mg/kg/day in mice and 100 mg/kg/day in rats from day 7 post-BLM) was clearly established. Conclusions: Overall, these data describe a novel observer-independent measure of BLM-induced lung fibrosis in rodents, which could be beneficial for the evaluation of potential preclinical drugs in the lung fibrosis field.