Hepatocellular cancer tumor growth and metabolic risk factors.

Abstract

162 Background: Understanding hepatocellular cancer (HCC) tumor growth patterns is important for determining optimal treatment strategy. However, growth of untreated HCC tumors is unclear and prior studies are limited by older imaging technologies and small samples. We evaluated HCC growth patterns and predictors in a cohort of patients with serial modern computed tomography (CT) and magnetic resonance imaging (MRI). Methods: Two radiologists retrospectively evaluated CT and MRI imaging for 56 patients with 83 HCC tumors with at least two images. Tumor Volume (TV) = 4/3* π *average short axis (X)*average long axis (Y)*X. To assess the relative change in size, normalized tumor volume (NTV) = TV/ baseline TV. Growth curves were fitted using an exponential growth model for tumor >2 observations. Univariate analysis explored clinical and pathologic predictors of NTV. A multivariate model was built by backward elimination of significant univariate variables. Results: Among 56 patients: 77% male, 96% cirrhotic, 50% Child-Pugh Class (CPC) B, 59% Hepatitis C, 32% obese, 37% diabetic, and 12% used metformin. Mean baseline and end TV were 98.8 cm3 and 213.0 cm3, respectively. Mean follow-up=8.3 months and mean number of serial images=3. Exponential TV growth was observed on visual inspection. Diabetes predicted lower NTV (i.e. slower TV growth) in univariate and multivariate analysis (Table). Obesity decreased the impact of diabetes. Conclusions: Diabetic patients had slower growing tumors than non-diabetics, even when clinical and pathologic factors were controlled. Tumor growth fit an exponential pattern. Further study of the impact of metabolic conditions on HCC tumor growth may provide insights into growth factors and identify potential treatment targets. [Table: see text]