Effective Extracellular Volume Fraction Determined by Equilibrium Contrast-Enhanced CT for Differentiating Autoimmune...

Background/Objectives: The aim of this study was to determine whether extracellular volume (ECV) fraction as determined by contrast-enhanced computed tomography (CECT) can help distinguish between autoimmune pancreatitis (AIP) and pancreatic ductal adenocarcinoma (PDAC). Methods: Participants comprised 101 patients, including 20 diagnosed with AIP (AIP group), 42 with histologically confirmed PDAC (PDAC group), and 39 without pancreatic disease (healthy group). Contrast enhancement (CE) was calculated as CT attenuation in Hounsfield units [HU] on equilibrium-phase CECT-CT attenuation on pre-contrast CT. The ECV fraction was calculated by measuring the region of interest within the pancreatic region and aorta on pre-contrast and equilibrium-phase CECT. CT measurements were compared among groups. CE and ECV fractions were also compared for diffuse (n = 12) and focal or segmental types (n = 8). Focal- or segmental-type AIP was defined as the involvement of one or two pancreas segments. Diagnostic efficacy was evaluated through receiver operating characteristic (ROC) analyses. Results: CE and ECV fractions differed significantly between the groups (p < 0.001 each). CE was significantly higher in the AIP group (56.8 ± 7.9 HU) than in the PDAC group (42.3 ± 17.0 HU, p < 0.001) or healthy group (32.2 ± 6.1 HU, p < 0.001). ECV fraction was significantly higher in the AIP group (47.2 ± 7.3%) than in the PDAC group (31.7 ± 12.0%, p < 0.001) or healthy group (27.5 ± 5.4%, p < 0.001). In the AIP group, no significant differences in CE (56.7 ± 8.2 HU vs. 56.9 ± 8.1 HU; p = 1.000) or ECV fraction (48.0 ± 5.6% vs. 46.6 ± 8.4%; p = 0.970) were seen between diffuse and focal or segmental types. Areas under the ROC curve for differentiating AIP from PDAC were 0.78 for CE and 0.86 for ECV fraction, showing no significant difference (p = 0.083). Conclusions: ECV fraction might be clinically useful in differentiating AIP from PDAC.