Pancreatic fibrosis assessment and association with pancreatic cancer: comparison with the extracellular volume fraction
H Fukui,H Onishi,T Ota,A Nakamoto,T Tsuboyama,Y Enchi,D Yamada,Y Takeda,S Kobayashi,Y Fukuda,H Eguchi,T Matsui,M Tatsumi,N Tomiyama
DOI: https://doi.org/10.1016/j.crad.2024.08.007
Abstract:Aims: To compare the iodine washout rate (IWR) from multiphasic contrast-enhanced computed tomography (CT) with the extracellular volume fraction (fECV) for assessing pancreatic fibrosis and its association with pancreatic cancer. Materials and methods: The study included 51 individuals (33 men; median age: 69 years; 21 with pancreatic cancer, 30 with other diseases) who underwent multiphasic contrast-enhanced CT and histological evaluation for fibrotic changes in pancreas. The histological pancreatic fibrosis fraction (HPFF) was assessed on Azan-stained sections. Pancreatic parenchymal enhancement values were measured to calculate IWR and fECV. Statistical methods, such as Spearman's rho and Mann-Whitney U-test, were used. Linear regression models using IWR and fECV were constructed to predict HPFF, with the performance expressed as root mean squared error (RMSE) and Akaike information criterion (AIC). Results: HPFF correlated with all CT parameters at the estimated transection line, strongest for IWRPPP-EP (r=-0.69, P<0.01). HPFF and fECV values were higher in the pancreatic cancer group than in controls (30% vs. 12.5%, P<0.01; 40.3% vs. 33.0%, P<0.01), whereas IWR values were lower (IWRPPP-EP: 43.3% vs. 55.0%, P<0.01; IWRPVP-EP: 25.0% vs. 33.5%, P<0.01). Linear regression models combining IWRPPP-EP + fECV and IWRPVP-EP + fECV were superior for predicting HPFF, with lower RMSE (9.23-9.35) and AIC (379.38-380.72) values than models with IWR or fECV alone. Conclusion: IWRPPP-EP, IWRPVP-EP, and fECV were reliable biomarkers for noninvasively assessing pancreatic fibrosis and were associated with pancreatic cancer risk. Linear regression combining these variables showed enhanced predictive accuracy for pancreatic fibrosis.