Endoscopy 2020; 52(S 01): S139
DOI: 10.1055/s-0040-1704430
ESGE Days 2020 ePoster Podium presentations
Pancreatic cysts 11:00 – 11:30 Thursday, April 23, 2020 ePoster Podium 3
© Georg Thieme Verlag KG Stuttgart · New York

DIAGNOSTIC ACCURACY OF INTRACYSTIC GLUCOSE AS COMPARED WITH CEA FOR THE DIAGNOSIS OF MUCINOUS PANCREATIC CYSTIC LESIONS: A META ANALYSIS

G Capurso
1   Pancreas Translational and Clinical Research Center San Raffaele Scientific Institute IRCCS Vita Salute San Raffaele University, Pancreato-Biliary Endoscopy and Endosonography Division, Milan, Italy
,
G Rossi
1   Pancreas Translational and Clinical Research Center San Raffaele Scientific Institute IRCCS Vita Salute San Raffaele University, Pancreato-Biliary Endoscopy and Endosonography Division, Milan, Italy
,
MC Petrone
1   Pancreas Translational and Clinical Research Center San Raffaele Scientific Institute IRCCS Vita Salute San Raffaele University, Pancreato-Biliary Endoscopy and Endosonography Division, Milan, Italy
,
PG Arcidiacono
1   Pancreas Translational and Clinical Research Center San Raffaele Scientific Institute IRCCS Vita Salute San Raffaele University, Pancreato-Biliary Endoscopy and Endosonography Division, Milan, Italy
› Author Affiliations
Further Information

Publication History

Publication Date:
23 April 2020 (online)

 

Aims Differential diagnosis between mucinous(M) and non-mucinous(NM) pancreatic cystic lesions(PCLs) is often difficult and both false positive and negative results can lead to clinically relevant overtreatmentand undertreatment. Dosage of intracystic Carcinoembryonic antigen(CEA) with a cut-off>192 ng/ml suggests M-PCL diagnosis, but its sensitivity is limited. Recently, it has been reported that low levels of intracystic glucose diagnose M-PCLs with high accuracy, but data are limited and heterogeneous.

We aimed to perform a meta-analysis to gather data on sensitivity, specificity and accuracy of intracystic glucose as compared with CEA for M-PCLs diagnosis.

Methods A computerized bibliographic search was performed on Pubmed. Pooled effects were calculated using a random-effects model and expressed in terms of pooled sensitivity and specificity and OR(95%CI) for accuracy for M-PCLs diagnosis. Heterogeneity was evaluated by I2and publication bias by Begg-Mazumdar test and funnel plot visual inspection.

Results 4cohort studies were included(3 US,1portuguese), for total 319 patients(207 M, 112 NM). Employed glucose cut-off were 50 in 3studies and 66 mg/dl in1; all studies employed 192 ng/ml CEAcut-off. Employed gold-standard for diagnosis in the 4studies was a composite of surgical pathology and cytology. Glucose dosage showed higher pooled sensitivity and lower heterogeneity(92%; I2=0%) compared to CEA(69.4%; I2=64%), while specificity was higher for CEA(75.6%glucose vs 92%CEA) with lower heterogeneity(I2=89% glucose vs 69%CEA). Glucose dosage was very close to be significantly superior in terms of pooled accuracy(OR 1.96; 95% CI 0.97-3.94; p=0.057; I2=53%). There was no publication bias at Begg-Mazumdar test and funnel plot.

Conclusions Our meta-analysis suggests that intracystic dosage of glucose is more sensitive, but CEA more specificfor M-PCLs diagnosis, with glucose overall better accuracy. Given the heterogeneity of results and relative low number of investigated patients considering the high PCLs prevalence, more studies are needed to define if combined use of both markers with different cut-offs is needed to increase accuracy significantly.