Chronic pancreatitis (CP) is a progressive fibro-inflammatory disease whose early diagnosis remains challenging. Conventional cross-sectional imaging often fails to detect early parenchymal and ductal changes, while endoscopic ultrasound (EUS) with Rosemont criteria (RC) is operator-dependent and may not reliably differentiate early CP from minimal pancreatic changes. Endoscopic ultrasound–based shear wave elastography (EUS-SWE) has emerged as a promising quantitative tool to assess pancreatic stiffness.¹ This prospective study evaluated the feasibility, reproducibility, and diagnostic performance of EUS-SWE for CP in a Western cohort and explored its correlation with RC and the number of EUS features.

All consecutive adult patients undergoing biliopancreatic or upper GI EUS between April and September 2025 at our tertiary center were evaluated. Exclusion criteria included known or suspected pancreatic adenocarcinoma, previous pancreatic or gastric surgery, lack of informed consent, or disagreement between two expert endosonographers regarding RC classification. For each patient, demographic and clinical data were collected, RC were independently assessed by two endoscopists, and five SWE measurements were obtained in the pancreatic body. Only measurements with VsN > 50% were considered reliable. According to RC, patients were classified as CP (“consistent” or “suggestive” RC) or non-CP (“indeterminate” or “normal” RC). Technical feasibility of EUS-SWE was assessed. Statistical analyses evaluated the association between SWE values and RC (RC class and number of RC features), while diagnostic performance was assessed via ROC analysis.

Of 298 screened patients, 82 were included. The mean age was 63 years, and most participants were Caucasian. Based on RC, 18 patients (22%) were classified as CP and 64 (78%) as non-CP. EUS-SWE proved feasible in 96.5% of cases, with a mean acquisition time of 65 seconds. Reproducibility was acceptable, with a mean IQR/median of 21.2% and no significant differences between CP and non-CP groups. Mean shear wave velocity (Vs) values increased progressively across RC classes: 1.88 m/s in normal pancreas, 2.13 m/s in indeterminate cases, 2.62 m/s in suggestive CP, and 2.86 m/s in consistent CP. Mean Vs was significantly higher in CP compared with non-CP (2.67 vs 1.97 m/s, p < 0.001). SWE values correlated positively with both the number of RC features (rs = 0.32, p 0.006) and RC class (rs = 0.39, p < 0.001). ROC analysis yielded an AUC of 0.74, indicating moderate diagnostic accuracy. The optimal cut-off for CP diagnosis was 2.28 m/s, with 72% sensitivity, 75% specificity, and a likelihood ratio of 2.89.

In this prospective cohort, EUS-SWE was highly feasible, rapid, and reproducible. SWE values correlated with RC and differentiated CP from non-CP with moderate accuracy, supporting its potential as a complementary, objective tool for CP assessment. The lower diagnostic performance compared with previous Asian studies may reflect the lower prevalence of advanced CP, with a predominance of early or preclinical disease in our cohort. Larger multicenter studies are needed to validate SWE cut-offs in diverse populations, clarify potential confounders, and assess whether longitudinal SWE monitoring can predict disease progression, pancreatic insufficiency, or pancreatic cancer risk.