Abstract
BACKGROUND/AIMS: Soft pancreatic texture is a key risk factor for postoperative pancreatic fistula (POPF). However, preoperative quantification remains difficult. Intraoperative palpation is the current standard for assessing pancreatic hardness; however, this technique is not feasible in robot-assisted surgery. We prospectively evaluated whether endoscopic ultrasound strain elastography (EUS-SE) could predict pancreatic hardness. METHODS: Nine consecutive patients scheduled for pancreatectomies underwent preoperative EUS-SE. The strain ratio (SR) values from the nontumorous parenchyma were compared using intraoperative palpation (soft vs. hard). A receiver operating characteristic (ROC) analysis was used to determine the optimal SR cutoff. The associations between computed tomography (CT) attenuation, fibrosis, and postoperative outcomes were examined. RESULTS: Six pancreases were classified as soft, and three pancreases were classified as hard. The median SR was identical in both groups (6.0; interquartile range [IQR] of 4.7-7.7 vs. 5.5-6.3; p = 1.000). The ROC-derived SR cutoff of 7.0 yielded an area under the curve of 0.500, indicating no discriminatory ability. The SR was not correlated with CT attenuation, the fibrosis status, or POPF. No adverse events occurred during EUS-SE. CONCLUSION: In this pilot study, preoperative EUS-SE did not predict intraoperative pancreatic textures. Strain-based elastography alone appears to be insufficient for hardness stratification and should be complemented by more quantitative modalities.