Endoscopic color Doppler ultrasonographic evaluation of recurrent esophagogastric varices following endoscopic injection sclerotherapy

Abstract

Background: Endoscopic injection sclerotherapy (EIS) for treatment of esophagogastric varices is well established in Japan. However, varices may still recur unpredictably following EIS. We studied this problem using endoscopic color Doppler ultrasonography (ECDUS) and specifically examined esophagogastric blood flows. Methods: Prophylactic EIS was performed by intravariceal injection of 5% ethanolamine oleate (EO) in 49 patients with esophageal varices secondary to liver cirrhosis. No patient had documented hepatocellular carcinoma (HCC) before EIS, and patients who developed HCC during follow-up were excluded. We performed ECDUS before EIS, and 2 weeks and 2 years later. The esophagogastric intra- and extramural venous blood flows, including flow in the azygos vein, were compared between these observations. Gastric intramural blood flow and changes in extramural gastric blood flow, including the azygos vein flow, were scored. Dynamic computer tomography (CT), ultrasonography (US), and color Doppler-ultrasonography (CDUS) were also performed before EIS and 1 month following the procedure. Thereafter, patients underwent CT and US examinations every 6 months for 2 years to detect any development of porto-systemic shunts or HCC. Results: The average number of EIS procedure per patient was 3.1±0.8 (mean±SD), and the total amount of sclerosant injected was approximately 33.5±6.5 ml. The overall recurrence rate over the 2-year follow-up was 36.7%. The gastric intra- and extramural blood flows did not differ between those patients with or without major shunts before EIS. In patients with recurrent variceal formation, the gastric intramural blood flow score following EIS (2.1±0.5) was significantly higher than that in patients without recurrence (0.8±0.6) ( P<0.01). In addition, gastric extramural blood flow score following EIS (0.8±0.6) was significantly lower in patients with recurrence than that in those without recurrence (1.7±0.5) ( P<0.01). The same differences held after exclusion of patients with major shunts. The gastric intramural blood flow score in patients with recurrent variceal formation (2.1±0.4) was significantly higher than that in patients without recurrence score ( P<0.01). Moreover, gastric extramural blood flow score in patients with recurrent variceal formation (1.0±0.7) was significantly lower than in patients without recurrence (1.6±0.5) ( P<0.01). Conclusions: Two characteristics were observed in patients with recurrent cases of esophageal varices 2 weeks following EIS. The first was the maintenance of gastric intramural blood flow. The second was the absence of dilation of the gastric extramural blood vessels. These observations may enable us to predict the recurrence of esophagogastric varices using ECDUS within 2 weeks following EIS.