Purpose: The most accurate and widespread method for diagnosis of gastroesophageal (GE) varices nowadays is upper gastrointestinal endoscopy. The value of color Doppler ultrasonography (CDUS) in correlation with endoscopy has never been extensively studied for liver cirrhosis (LC) patients. Aim of the study was to determine value of CDUS detecting GE varices in different grades of those on upper gastrointestinal endoscopy.
Material and methods: During two years period 144 LC patients were examined with CDUS for presence of portal hypertension and portal-systemic collaterals (PC). Statements 'positive' when signs of GE varices were present or 'negative' when GE varices were not present were made by an experienced sonologist. PC feeding GE varices were subdivided into two groups. Group I - left gastric vein and esophageal varices, and group II - short gastric veins and gastric (fundal) varices. In all 144 patients the upper gastrointestinal endoscopy was performed and GE varices were graded into three classes: class F0 (absent) - 11 patients; class F1 (small) - 49 patients, class F2 (medium) 52 patients and class F3 (large) - 32 patients.
Results: In endoscopic class F0 there were no statements 'positive' and 11 statements 'negative'. In class F1 there were 32 'positive' and 17 'negative' statements - sensitivity 65%. In class F2 there were 45 'positive' and 7 'negative' statements - sensitivity 87%. In class F3 there were 30 'positive' and 2 'negative' statements - sensitivity 94%. Overall sensitivity was 80%.
Conclusions: CDUS showed excellent sensitivity for advanced size GE varices and good overall sensitivity of 80%. CDUS provided information about different tributaries of GE varices as left gastric vein and short gastric veins. However the sensitivity of CDUS in class F1 was low (65%) and upper gastrointestinal endoscopy was much more efficient.

Keywords: Color Doppler ultrasonography (CDUS), gastroesophageal (GE) varices, liver cirrhosis (LC), porto-systemic collaterals (PC)