Pancreatic cancer stands 3^rd in cancer-related death. Endoscopic ultrasound (EUS) plays a key role in characterizing and diagnosing solid pancreatic lesions. The literature suggests, with limited evidence, that ancillary EUS techniques like EUS-guided elastography, which assess tissue stiffness, potentially improve the differentiation of malignant lesions. In this single-center retrospective study, we analyzed the performance of EUS-guided elastography in characterizing solid pancreatic lesions compared with histopathological results.

Our team conducted a retrospective analysis of 508 consecutive patients who underwent Endoscopic Ultrasound (EUS) at the Institute of Pancreatic Diseases, Semmelweis University, between January and October 2025. We included 108 patients with suspected solid pancreatic tumors and divided them into two groups. The first group (54 patients) had elastography and histopathological results. The second group (54 patients) had histopathological results but no elastography. Demographic parameters, endoscopic data (elastography, ROSE, MOSE) were collected for both groups. Statistical analysis was performed using Microsoft Office Excel and GraphPad (Prism10) software. 

In the elastography group (EG), the median age was 70 years (95% CI: 66–74), compared to 68 years (95% CI: 61–71) in the non-elastography group (NEG), with similar mean ages between groups (69.09 vs. 65.07 years; 95% CI for the mean: 66.03–72.16 vs. 62.12–68.03). Regarding sex distribution, 38.89% of patients in the EG were male and 61.11% female, whereas in the NEG, males comprised 62.96% and females 37.04% (P=0.01; OR=0.37; 95% CI 0.17–0.80).

Pancreatic adenocarcinoma was the most frequent diagnosis in both groups, 57.4% in the EG and 66.7% in NEG (P>0.05). Median tumor size was 675 mm² (median dimensions: 27x25mm) in the EG and 900 mm² (median dimensions: 30x29mm) in NEG (P=0.1417). Tumors most frequently involved the pancreatic head in both groups (EG: 54.74%; NEG: 50.00%) (P=0.1448).

The use of MOSE and ROSE showed high prevalence in both groups. MOSE was applied in 96.36% of patients in the EG group and in 94.44% of patients in the NEG group. ROSE was utilized in 77.78% of cases in the EG group and in 72.22% of cases in the NEG group.

Among the 54 patients who underwent elastography (EG), 43 (79.62%) exhibited blue colour, while 11 (20.37%) demonstrated mixed colour. In blue-EG cohort, 34 (79.07%) had malignant lesions and 9 (20.93%) had benign lesions (P=0.69, OR=1.41, 95% CI 0.34–5.63). In mixed-EG cohort, 8 (72.73%) had malignant lesions and 3 (27.27%) had benign lesions (P=0.69, OR=1.41, 95% CI 0.34–5.63).

In EG, rebiopsy was required in 14 (25.93%) patients, 11 (78.57%) blue colour and 3 (21.42%) mixed colour (P>0.99, OR=0.91, 95% CI 0.19–3.65). Among the rebiopsy cases with blue colour, 5 (45.45%) were malignant and 6 (54.55%) were benign (P>0.99, OR=0.41, 95% CI 0.02–4.70), while among rebiopsy cases with mixed colour, 6 (75.00%) were malignant and 2 (25.00%) benign (P=0.25, OR=3.22, 95% CI 0.47–18.07).

In EG group, the sensitivity, specificity, positive predictive value, negative predictive value and likelihood ratio are 0.80 (0.66–0.90), 0.25 (0.08–0.53), 0.79 (0.64–0.88), 0.27 (0.09–0.56) and 1.07 respectively.

In the NEG, 46 (85.19%) had malignant lesions and 8 (14.81%) had benign lesions. Rebiopsy was necessary in 10 (18.52%) patients (95% CI 0.09–0.30). Among the rebiopsied patients, 7 (70.00%) cases were malignant and 3 (30.00%) cases were benign.

Rebiopsy data showed no statistically significant difference between EG (malignant lesions: 25.93%) and NEG (malignant lesions: 18.52%) groups. 

Our results suggest that elastography does not significantly improve overall sampling success or reduce the need for rebiopsy, which aligns with observations reported in the international literature regarding the use of elastography. However, because the study is a small, single-center retrospective analysis, a larger multicenter prospective randomized trials incorporating both quantitative and qualitative assessments are needed.