Rapid diagnosis of mucinous cystic pancreatic lesions by on-site cyst fluid glucometry

Introduction Available intracystic biomarkers show a limited accuracy for characterizing cystic pancreatic lesions (CPL). Glucose is an attractive alternative due to its availability, low cost and the possibility of on-site quantification by glucometry. Aim To evaluate the diagnostic accuracy of on-site glucometry from samples obtained by EUS-FNA in the differential diagnosis between mucinous from non-mucinous CPL. Methods Retrospective, multicentre, cross-sectional study of patients who underwent EUS-FNA of a CPL. A derivation and a validation cohorts were evaluated. Intracystic glucose was quantified by on-site glucometry and colorimetry in the lab. Final diagnosis was based on surgical specimens or global evaluation of clinical and imaging data, cytology and intracystic CEA. Diagnostic accuracy was based on Receiver Operating Curve (ROC) curve analysis. Intraclass correlation coefficient (ICC) between on-site and lab glucose levels was calculated. Results Seventy two patients were finally analysed (40 in the derivation cohort and 32 in the validation cohort). Intracystic glucose levels by on-site glucometry was 12.3 ± 28.2 mg/dl for mucinous CPL and 103.3 ± 58.2 mg/dl for non-mucinous CPL, p < 0.001. For an optimal cut-off point of 73 mg/dl, on-site glucose had a sensitivity, specificity, and positive and negative predictive value for the diagnosis of mucinous CPL of 0.89, 0.90, 0.94, 0.82 respectively in the derivation cohort, and 1.0, 0.71, 0.91, 1.0 respectively in the validation cohort. Correlation of on-site and lab glucose quantification was very high (ICC = 0.98). Conclusion On-site glucometry is a feasible, accurate and reproducible method for the characterization of CPL after EUS-FNA. It shows an excellent correlation with laboratory glucose values. Registration number: 2019/612.