Role of hepatic sonic hedgehog protein expression in the diagnosis of metabolic dysfunction-associated steatohepatitis

BACKGROUND Ballooned hepatocytes are a histological hallmark in the diagnosis of metabolic dysfunction-associated steatohepatitis (MASH). Identifying ballooned hepatocytes on routine stains is challenging. Cytokeratin 8/18 protein is diffusely expressed in normal hepatocytes but absent in ballooned hepatocytes. Conversely, sonic hedgehog (SHH) protein is absent in normal hepatocytes but present in ballooned hepatocytes. AIM To investigate the utility of immunostaining for positive SHH protein expression in ballooned hepatocytes in MASH. METHODS Clinicopathological data from hospitalized patients with metabolic dysfunction-associated steatotic liver (MASL) disease at the Second Hospital of Nanjing from January 2020 to November 2022 were analyzed. The Nonalcoholic Steatohepatitis Clinical Research Network scoring system was used. Post-staining, digitized images were acquired, and area quantification algorithms were used to quantify SHH expression. RESULTS A total of 190 MASL disease patients who underwent liver biopsy were enrolled in this study; 58.9% (112/190) had definite MASH, and 41.1% (78/190) had MASL. There were significant differences in body mass index (P < 0.001), diabetes (P < 0.01), metabolic syndrome (P < 0.02), and circulating M65 and M30 (P < 0.001), as well as aspartate aminotransferase (AST), alanine aminotransferase, glucose, uric acid, controlled attenuation parameter (CAP), liver stiffness measurement, and nonalcoholic fatty liver disease fibrosis score (P < 0.05). Serum M30 and M65 levels were almost three times greater in MASH than in MASL patients. Hepatic SHH expression correlated with circulating M65 (r = 0.346, P = 0.002) and circulating M30 (r = 0.471, P < 0.001), as did alanine aminotransferase (r = 0.490, P < 0.001), AST and CAP (r = 0.554, P < 0.001 and r = 0.432, P < 0.001, respectively). Hepatic SHH expression correlated with histological steatosis grade (r = 0.502, P < 0.001), ballooning hepatocytes (r = 0.496, P < 0.001), lobular inflammation (r = 0.450, P < 0.001), and fibrosis stage (r = 0.303, P = 0.006). Logistic modeling revealed diabetes, AST, CAP and hepatic SHH expression as independent predictors of MASH [defined as nonalcoholic fatty liver disease activity score ≥ 5: Odds ratio (OR) = 20.95, P = 0.043, OR = 1.044, P = 0.023, OR = 1.034, P = 0.008, and OR = 7.151, P = 0.017, respectively] and histological ballooning hepatocytes and circulating M30 as independent predictors of advanced fibrosis (defined as portal and pericellular fibrosis ≥ 2: OR = 6.440, P = 0.023, and OR = 1.012, P = 0.005, respectively). The Fleiss’ kappa increased interobserver agreement of assessment of ballooning using SHH immunostaining, from 0.65 to 0.85. CONCLUSION Hepatic SHH protein expression assisted in the diagnosis of MASH. SHH immunostaining may be useful for classifying and quantifying ballooned hepatocytes by artificial intelligence algorithms.