TY - JOUR
T1 - Validation of combined carcinoembryonic antigen and glucose testing in pancreatic cyst fluid to differentiate mucinous from non-mucinous cysts
AU - Gorris, Myrte
AU - Dijk, Frederike
AU - Farina, Arantza
AU - Halfwerk, Johannes B.
AU - Lekkerkerker, Selma J.
AU - Voermans, Rogier P.
AU - Wielenga, Mattheus C.
AU - Besselink, Marc G.
AU - van Hooft, Jeanin E.
AU - Hooijer, Gerrit K.
N1 - Funding Information: Marc G. Besselink has received research support from Intuitive Surgical, Medtronic, Ethicon and acted as proctor for Intuitive Surgical. Jeanin E. van Hooft has received research support from Cook Medical and acted as consultant for Cook Medical, Boston Scientific, Olympus, Medtronic and Abbvie. Rogier P. Voermans received a research grant and acted as consultant for Boston Scientific. Myrte Gorris, Frederike Dijk, Arantza Farina, Johannes B. Halfwerk, Gerrit K. Hooijer, Selma J. Lekkerkerker, and Mattheus C. Wielenga have no conflicts of interest or financial ties to disclose. Publisher Copyright: © 2023, The Author(s).
PY - 2023/5
Y1 - 2023/5
N2 - Background: More accurate diagnosis of mucinous cysts will reduce the risk of unnecessary pancreatic surgery. Carcinoembryonic antigen (CEA) and glucose in pancreatic cyst fluid (PCF) can differentiate mucinous from non-mucinous pancreatic cystic neoplasms (PCN). The current study assessed the value of combined CEA and glucose testing in PCF. Methods: Cross-sectional validation study including prospectively collected PCF from patients undergoing endoscopic ultrasonography-guided fine-needle aspiration (EUS-FNA) and pancreatic surgery. We performed laboratory measurements for CEA and glucose and measured glucose levels by a hand glucometer. Primary outcome was diagnostic accuracy evaluated by receiver operator curves (ROC), sensitivity, specificity, positive, and negative predictive value (PPV, NPV). Results: Overall, PCF was collected from 63 patients, including 33 (52%) with mucinous and 30 (48%) with non-mucinous PCN. Histopathology (n = 36; 57%), cytopathology (n = 2; 3%), or clinical and/or radiological diagnosis (n = 25; 40%) was used as reference standard. Combined CEA (cut-off ? 192?ng/ml) and laboratory glucose testing (cut-off ? 50?mg/dL) reached 92% specificity and 48% sensitivity, whereas either positive CEA (cut-off ? 20?ng/ml) or glucose testing (cut-off ? 50?mg/dL) showed 97% sensitivity and 50% specificity. Sensitivity and specificity were 80% and 68% for CEA ? 20?ng/mL versus 50% and 93% for CEA ? 192?ng/mL (the conventional cut-off level). Laboratory and glucometer glucose both reached 100% sensitivity and 60% and 45% specificity, respectively. None of the biomarkers and cut-offs reached a PPV exceeding 90%, whereas both glucose measurements had a NPV of 100% (i.e., high glucose excludes a mucinous cyst). Conclusion: Combined CEA and glucose testing in PCF reached high specificity and sensitivity for differentiating mucinous from non-mucinous PCN. Glucose testing, whether alone or combined with the new CEA cut-off (? 20?ng/mL), reached > 95% sensitivity for mucinous cysts, whereas only glucose reached a NPV > 95%. Graphical abstract: [Figure not available: see fulltext.].
AB - Background: More accurate diagnosis of mucinous cysts will reduce the risk of unnecessary pancreatic surgery. Carcinoembryonic antigen (CEA) and glucose in pancreatic cyst fluid (PCF) can differentiate mucinous from non-mucinous pancreatic cystic neoplasms (PCN). The current study assessed the value of combined CEA and glucose testing in PCF. Methods: Cross-sectional validation study including prospectively collected PCF from patients undergoing endoscopic ultrasonography-guided fine-needle aspiration (EUS-FNA) and pancreatic surgery. We performed laboratory measurements for CEA and glucose and measured glucose levels by a hand glucometer. Primary outcome was diagnostic accuracy evaluated by receiver operator curves (ROC), sensitivity, specificity, positive, and negative predictive value (PPV, NPV). Results: Overall, PCF was collected from 63 patients, including 33 (52%) with mucinous and 30 (48%) with non-mucinous PCN. Histopathology (n = 36; 57%), cytopathology (n = 2; 3%), or clinical and/or radiological diagnosis (n = 25; 40%) was used as reference standard. Combined CEA (cut-off ? 192?ng/ml) and laboratory glucose testing (cut-off ? 50?mg/dL) reached 92% specificity and 48% sensitivity, whereas either positive CEA (cut-off ? 20?ng/ml) or glucose testing (cut-off ? 50?mg/dL) showed 97% sensitivity and 50% specificity. Sensitivity and specificity were 80% and 68% for CEA ? 20?ng/mL versus 50% and 93% for CEA ? 192?ng/mL (the conventional cut-off level). Laboratory and glucometer glucose both reached 100% sensitivity and 60% and 45% specificity, respectively. None of the biomarkers and cut-offs reached a PPV exceeding 90%, whereas both glucose measurements had a NPV of 100% (i.e., high glucose excludes a mucinous cyst). Conclusion: Combined CEA and glucose testing in PCF reached high specificity and sensitivity for differentiating mucinous from non-mucinous PCN. Glucose testing, whether alone or combined with the new CEA cut-off (? 20?ng/mL), reached > 95% sensitivity for mucinous cysts, whereas only glucose reached a NPV > 95%. Graphical abstract: [Figure not available: see fulltext.].
KW - Carcinoembryonic antigen (CEA)
KW - Diagnostic accuracy
KW - Endoscopy: Endoscopic ultrasound diagnostic and therapeutic
KW - Glucose
KW - Hepatobiliary tract and spleen
KW - Pancreatic cyst fluid
KW - Pancreatic cystic neoplasms
UR - http://www.scopus.com/inward/record.url?scp=85146617804&partnerID=8YFLogxK
U2 - https://doi.org/10.1007/s00464-022-09822-6
DO - https://doi.org/10.1007/s00464-022-09822-6
M3 - Article
C2 - 36656409
SN - 0930-2794
VL - 37
SP - 3739
EP - 3746
JO - Surgical endoscopy
JF - Surgical endoscopy
IS - 5
ER -