Image analysis of liver collagen using sirius red is more accurate and correlates better with serum fibrosis markers than trichrome

Background: Collagen proportional area (CPA) determined by quantitative digital image analysis better quantifies liver fibrosis than histological stage; however, its clinical use has been limited by non-standardized methods. Aim: This study aimed to compare CPA obtained using different staining methods, magnifications and biopsy sizes. Methods: Two hundred and forty-nine patients with chronic hepatitis C who had a liver biopsy and serum fibrosis markers performed were included. CPA was measured either using a sirius red (CPAs) or a trichrome (CPAt) stain. Results: CPAs measured at 209 and 409 magnifications generated similar outcomes with interclass correlation (ICC) coefficient of 0.98. Compared with trichrome, sirius red staining had much less variation with an ICC coefficient of 0.99 for slides stained in the same batch and 0.92 in different batches. Mean CPAs was higher than mean CPAt by 3.53%, P \u3c 0.001. Morphological analysis found that sirius red detected delicate fibrous septa and spurs better than trichrome. Both CPAs and CPAt correlated well with Metavir stage, whereas CPAs had better ability to detect cirrhosis with the area under ROC curve of 0.95. Overall CPA had superior correlation with serum markers of fibrosis in Metavir F2–F4 than that in F0–F1 and CPAs correlated better with serum fibrosis markers than CPAt in Metavir F0–F1. Multivairate analysis found that HA, a2-macroglobulin, platelet count and albumin were independently correlated with CPAs and only HA was independently correlated with CPAt. Conclusions: Sirius red staining for CPA determination was more accurate and reliable for quantifying hepatic collagen compared with trichrome staining