| Background and aims Esophagogastric variceal bleeding is a primary cause of death in cirrhotic patients.UGE(upper gastrointestinal endoscopy)is the gold standard for the diagnosis of EGV(esophagogastricvarices),but it is invasive.Recently,several studies have reported some non-invasive methods for the diagnosis of EGV.This study aimed to evaluate the diagnostic accuracy of the non-invasive methods,including serum markers and contrast-enhanced CT(computed tomography),for diagnosis of EGV in cirrhotic patients.Methods 1.All papers which evaluated the diagnostic accuracy of serum markers,including APRI(aspartate aminotransferasetoplatelet ratio index),AAR(aspartate aminotransferasetoalanine aminotransferase ratio),FIB-4,FI,King,Lok,Forns and Fibro Index scores,for EGV in cirrhotic patients were searched via Pub Med,EMBASE,CNKI,and Wanfang databases.The AUSROC(area under the summary receiver operatingcharacteristic curve),sensitivity,specificity,PLR(positive likelihood ratio),NLR(negativelikelihood ratio),and DOR(diagnostic odds ratio)were calculated.2.All patients who were consecutively admitted to our hospital and underwent UGE between January 2012 and June 2014 were eligible for this retrospective study to evaluate the diagnostic accuracy of serum markers,including APRI,AAR,FIB-4,Lok and Forns scores,for EV(esophageal varices)in cirrhotic patients.AUC(areas under curve),sensitivity,and specificitywere calculated.Subgroup analyses were performed according to the history of UGIB(upper gastrointestinal bleeding)and splenectomyor splenic artery embolization.The study quality was assessed according to the QUADAS (Quality Assessment of Diagnostic Accuracy Studies)tool.3.All papers which evaluated the diagnostic accuracy of contrast-enhanced CT for EGV in cirrhotic patients were searched via Pub Med and EMBASE databases.AUSROC,sensitivity,specificity,PLR,NLR,and DORwere calculated.Weperformed the subgroup analyses according to the location of EGV,CT technique,and study design.The study quality was assessed according to the QUADAS-2 tool.4.All cirrhotic patients who were consecutively admitted to our hospital and performedcontrast-enhanced CT and UGE between June 2012 and June 2014 were included for this retrospective study to evaluate the diagnostic accuracy of contrast-enhanced CT for EV.Two observers reviewed the contrast-enhanced CT scans to evaluate the presence and diameter of EV.AUC,sensitivity,and specificity were calculated.Results 1.Overall,12,4,5,0,0,4,3,and 1 paper was identified to explore thediagnostic accuracy of APRI,AAR,FIB-4,FI,King,Lok,Forns,and Fibro Index scores for EGV,respectively.The AUSROC of APRI,AAR,FIB-4,Lok,and Forns scores for the prediction of EGV were 0.6774,0.7275,0.7755,0.7885,and 0.7517,respectively;and those for the prediction of high-risk EGV were 0.7278,0.7448,0.7095,0.7264,and 0.653,respectively.The diagnostic threshold effects of FIB-4 and Forns scores for theprediction of EGV were statistically significant.The sensitivity/specificity/PLR/NLR/DOR of APRI,AAR,and Lok scores for theprediction of EGV were 0.6/0.67/1.77/0.58/3.13,0.64/0.63/1.97/0.54/4.18,and 0.74/0.68/2.34/0.4/5.76,respectively.The sensitivity/specificity/PLR/NLR/DOR of APRI,AAR,FIB-4,Lok,and Fornsscores for the prediction of high-risk EGV were 0.65/0.66/2.15/0.47/4.97,0.68/0.58/2.07/0.54/3.93,0.62/0.64/2.02/0.56/3.57,0.78/0.63/2.09/0.37/5.55,and 0.65/0.61/1.62/0.59/2.75,respectively.2.A total of 650 cirrhotic patients who performed serum markers and UGE were included in this study.According to the endoscopic results,81.4% patients had moderate-severe EV.In theoverall analysis,the AUC of these non-invasive scores for predicting moderate-severe EV and presence ofany EV were 0.506-0.6 and 0.539-0.612,respectively.In the subgroup analysis of patients without UGIB,their AUC for predicting moderate-severe EV and presence of any EV were 0.601-0.664 and 0.596-0.662,respectively.In the subgroup analysis of patients without UGIB and splenectomy or splenic artery embolization,their AUC for predicting moderate-severe EV and presence of any EV were 0.627-0.69 and 0.607-0.692,respectively.3.Seventeen papers which evaluated the diagnostic accuracy of contrast-enhanced CT for EGV were eligible.The overall study quality was modest.The AUSROC were 0.8975 and0.9495 for predicting any size and high-risk EGV,respectively.The sensitivity/specificity/PLR/NLR/DOR of CT for predicting any size and high-risk EGV were 0.87/0.8/3.67/0.18/22.7 and0.87/0.88/7.52/0.12/65.55,respectively.According to the location of EGV,the AUSROC was 0.9127 forpredicting any size gastric varices alone;and the AUSROC were 0.8958 and 0.9461 for predicting any sizeand high-risk EV alone,respectively.According to the CT technique,the AUSROC of MDCT(multi-detector CT)were 0.9047 and 0.949 for predicting any size and high-risk EGV,respectively;and the AUSROC of MDCT esophagograms for predicting any size and high-risk EGV were 0.8735 and 0.9664,respectively.In the subgroup analysis of prospective studies,the AUSROC were 0.9122 and 0.9507 for predicting any size and high-risk EGV,respectively.4.A total of 52 cirrhotic patients who performed contrast-enhanced CT and UGE were included in this study.According to the endoscopic results,13.5% of these patients did not have any EV,11.5% had mild EV,and 75% had moderate-severe EV.The AUC of contrast-enhanced CT scans for the diagnosis of EV was 0.835.The sensitivity and specificity was 95.56% and 71.43%,respectively.The AUC of contrast-enhanced CT scans for the diagnosis of moderate-severe EV was 0.821.When the cut-off value was 3.9mm,the sensitivity and specificity were 89.74% and 69.23%,respectively.Conclusions Serum markers had low to moderatediagnostic accuracy in predicting the presence of EGV in liver cirrhosis,which might notbe able to replace the utility of UGE for the diagnosis of EGV.Contrast-enhanced CT had high diagnostic accuracy for EGV in liver cirrhosis,which might be useful to decrease the use of UGE in clinical practice. |
PDF Full Text Request