Correlation between Endoscopic and Angiographic Findings in Patients with Esophageal and Isolated Gastric Varices

Abstract

Background/Aim: The correlation between angiographic vascular patterns and endoscopic findings in portal hypertension is not sufficiently known, and knowledge of the vascular anatomy may contribute to an improvement in endoscopic embolization and transjugular retrograde obliteration procedures. We propose a new vascular map that should prove useful for this purpose. Methods: Between April 1985 and December 1997 we performed percutaneous transhepatic portography in a selected group of 75 patients (16 women and 59 men), aged 43–71 years, from whom informed consent was obtained. All patients had been diagnosed endoscopically as having either esophageal or isolated gastric varices. According to the Child-Pugh classification, class A, B, and C cirrhosis was seen in 19, 40, and 16 patients, respectively. We created a vascular map of esophageal and isolated gastric varices, based on the opacification of the portal venous collaterals on percutaneous transhepatic portography. We compared the patients in both variceal groups in terms of portal venous pressure, main blood supply, and drainage routes. Results: We found that the portal collateral system was divided into two systems: the portoazygos venous system and the portophrenic venous system. The former contributed to the formation of esophageal and cardiac varices and the latter to the formation of isolated gastric varices located at the fundus or at both the cardia and fundus. The left gastric vein participated as blood supply in 70% of the isolated gastric varices and in 100% of the esophageal varices (p < 0.01). The posterior gastric vein participated as blood supply in 70% of the isolated gastric varices and in 24% of the esophageal varices (p < 0.01). We classified the main blood drainage routes of isolated gastric varices functionally into three types: gastrorenal shunt (85%), gastrophrenic shunt (10%), and gastropericardiac shunt (5%). The portal venous pressure in patients with esophageal varices was 358 ± 66 mm H₂O, whereas in patients with isolated gastric varices it was 262 ± 44 mm H₂O (p < 0.01). Conclusion: We suggest that this new vascular map will be useful in endoscopic embolization and transjugular retrograde obliteration procedures for esophageal and isolated gastric varices.