Needle-based confocal laser endomicroscopy for diagnosis of pancreatic cystic lesions: a meta-analysis

Abstract

Background and aim The role of needle-based confocal laser endomicroscopy (nCLE) in the diagnosis of pancreatic cystic lesions (PCLs) remains controversial. This study aimed to systematically evaluate the diagnostic accuracy and adverse effects of nCLE in the detection of pathological subtypes in patients with PCLs. Material and methods We conducted a comprehensive literature search using MEDLINE, EMBASE, and Cochrane for identifying studies that reported the use of nCLE for PCLs diagnosis (dated prior to 10 October 2020). Studies with a sample size >10 were included. We used the QUADAS-2 criteria for quality evaluation. We first extracted the diagnostic rates and the information on adverse events (AEs) from the studies; then used STATA15.0 to calculate the variables, draw forest plots and summary receiver operating characteristic (SROC) curves; and finally, we completed subgroup analyses to explore the heterogeneity. Results Overall, 299 article titles were identified after an initial search, and ten studies with 547 individuals with PCLs were included in the analysis. The pooled sensitivity, specificity, positive likelihood ratio, and pooled negative likelihood ratio of nCLE in detecting gastric disorders were 90%, 96%, 20.4, and 0.11, respectively. The pooled sensitivity and specificity showed a substantial heterogeneity. An ROC curve was constructed with an area under the curve (AUC) of 0.94. The overall AEs rate of pancreatitis was 2.7%. Conclusions We showed that nCLE had a relatively high sensitivity and specificity for diagnosing PCLs with a relatively low rate of AEs occurring. We suggest that nCLE has good diagnostic accuracy for PCLs.