Prediction of oesophageal varices in hepatic cirrhosis by simple serum non-invasive markers: Results of a multicenter, large-scale study

Abstract

Preliminary data suggest that non-invasive methods could be useful to assess presence of oesophageal varices (OV) in cirrhotic patients. We aimed to further investigate simple serum non-invasive markers for diagnosing and grading OV. A retrospective set of 510 cirrhotics and a prospective set of 110 cirrhotics were enrolled consecutively in five centers. Platelets, AST-to-ALT ratio, AST-to-platelet-ratio index, Forns' index, Lok index, Fib-4, and Fibroindex were measured within 2 months from upper endoscopy, taken as a gold standard. Performance was expressed as sensitivity, specificity, positive, and negative predictive values (PPV, NPV), accuracy, and area under the curve (AUC). A combination of Lok index (cutoff=1.5) and Forns' index (cutoff=8.8) had 0.80 AUC (0.76-0.84, 95% CI), and high NPV (>90%) to exclude clinically relevant OV, defined as large OV or small OV with red signs or in Child-Pugh C cirrhosis. By applying this combination, upper endoscopy would have been avoided in 1/3 of our cirrhotics. Large OV could be excluded with 96% NPV by Lok index (cutoff=1.5). A combination of Lok index (cutoff=0.9) and Forns' index (cutoff=8.5) predicted presence of any grade OV with good performance: 0.82 AUC (0.76-0.88, 95% CI), 88% PPV. Serum non-invasive markers may be useful as a first line tool to identify cirrhotic patients in which the risk of clinically relevant OV is trivial, and to reduce the number of upper endoscopies. However, we are still far from the possibility of replacing upper endoscopy by simple serum non-invasive markers in the vast majority of patients.