Purpose: The purpose of this study was to evaluate the degree of fatty infiltration (FI) of the pancreas using area-based assessment on computed tomography (CT) (CT area-based assessment) in its correlation and agreement/concordance with histopathology-based assessment. Furthermore, we examined whether CT area-based assessment was better than CT attenuation index-based assessment.
Materials and methods: We retrospectively evaluated the degree of FI of the pancreas in 37 pancreatic cancer patients who had undergone preoperative CT and pancreaticoduodenectomy. The degree of FI of the pancreas was examined by histopathology-based assessment using surgical resection samples, and CT area-based and CT attenuation index-based assessments.
Results: Mean values of pancreatic FI measured by area-based assessment on unenhanced CT and by histopathology-based assessments were 14.4 ± 23.2 % (range 0-77.7 %) and 16.2 ± 17.4 % (range 0.2-60.2 %), respectively. Area-based assessment on unenhanced CT showed higher correlation and concordance with histopathology-based assessment, demonstrating a Spearman correlation coefficient of 0.78 (P < 0.0001) and a Kendall's tau-b coefficient of 0.69 (P < 0.0001). For CT attenuation index-based assessment, the corresponding values were -0.66 (P < 0.0001) and -0.39 (P = 0.008), respectively.
Conclusions: To order/categorize subjects according to the degree of FI of the pancreas, CT area-based assessment is suggested to be better than CT attenuation index-based assessment.
Keywords: Computed tomography; Fatty infiltration; Histopathology; Pancreas; Pancreatic cancer.