Strain ratio vs strain histogram for the elastographic evaluation of non-calcific chronic pancreatitis (CP) ‒ A prospective, single-centre, comparative study.

Abstract

Introduction: Diagnosis of chronic pancreatitis (CP) at early stages remains challenging. Endoscopic ultrasound (EUS) guided strain elastography (SE) has improved the diagnostic yield of CP. Aim of our study was to evaluate the diagnostic yield of SE quantification methods, strain ratio (SR) and strain histogram (SH), in detecting non-calcific CP.

Methods: Prospective, comparative study including patients undergoing EUS-guided SE for non-calcific CP evaluation. Procedures were performed with linear echoendoscopes. Number of EUS criteria according to Rosemont classification were evaluated. SH analysis targeted a pancreatic body area, while SR compared this area (A) to a soft extrapancreatic reference area (B), with B/A as the final ratio. Data are shown as percentages and mean (95%CI) and analyzed using ANOVA, linear regression and Spearman test. Diagnostic accuracy of SR and SH were evaluated using the Rosemont classification as reference. STARD criteria for studies was followed.

Results: 284 patients were included (mean age 49.8 years, range 17-85; 142 males). Among them, 43 (15.1%) had a normal pancreas, 110 (38.7%) had indeterminate findings, and 131 (46.1%) had suggestive CP findings. A SR>2.42 showed 91.7% sensitivity and 88.1% specificity (ROC=0.968). An SH<116.1 demonstrated 88.0% sensitivity and 97.7% specificity (ROC=0.964). EUS criteria correlated with pancreatic fibrosis severity by SR (r=0.785, p<0.0001) and SH (r=-0.742, p<0.0001). Age correlated with EUS criteria (r=0.127, p=0.03), but not with SE.

Conclusions: The quantification of the degree of pancreatic fibrosis by SR and SH during pancreatic EUS-guided SE exhibit high diagnostic accuracy for CP and are not influenced by age.