The study aimed to determine the usefulness of the elastography in the diagnosis of malignancy of solid pancreatic tumors. There were 123 patients (F/M; 51/72, aged 62 ± 14) enrolled into the study with the diagnosis of pancreatic masses. Malignant pancreatic adenocarcinoma was identified in 78 patients and an inflammatory mass corresponding to chronic pancreatitis in the remaining 45 patients. The mass elasticity of a tumor (A-elasticity) and a reference zone (B-elasticity) and the B/A strain ratio were measured. All these elastographic parameters differed between groups and correlated significantly with malignancies (r = 0.841; r = -0.834; r = 0.487, respectively). Receiver operating characteristic (ROC) analysis showed that A-elasticity between 0.05% and 0.14% alone, as well as the B/A strain ratio between 7.87 and 18.23 alone, enabled the recognition of all malignant pancreatic tumors with 100% sensitivity and ≥ 97.8% specificity. Surprisingly, B-elasticity alone also was helpful in recognizing malignant tumors (71% sensitivity, 80% specificity, 0.74 accuracy, and 0.792 area under the curve), although it appeared worse than A-elasticity and B/A strain ratio (p < 0.001). In multivariable regression analysis, A-elasticity identified 89.5% of malignancies (p < 0.001). A-elasticity and B-elasticity were the only significant independent factors influencing the tumor identification (r2 = 0.927; p < 0.001). The assessment of tumor elasticity appears sufficient to identify malignant tumors of the pancreas.
Keywords: Elastography; Endosonography; Pancreatic malignancy; ROC analysis; Strain ratio; Tumor elasticity.