Prevalence of 'deep' rectal varices in patients with cirrhosis: an EUS-based study.

Background Endoscopic ultrasound (EUS) permits identification of dilated veins of the intrinsic rectal venous system (deep varices) in portal hypertension. The aim of this cross-sectional study was to assess the prevalence of, and risk factors for, deep rectal varices, using EUS. Methods A cohort of 96 patients with cirrhosis was studied. Both routine rectoscopy and rectal EUS were performed. Deep varices were assessed as peri-rectal and para-rectal. A three-grade scale was used to assess the size of the deep varices. Results On rectoscopic examination, congestive rectopathy was seen in 11(11%) patients, blue veins in 36(38%) and rectal varices in 13 (14%). On EUS, deep varices were seen in 49(51%) subjects. Small peri-rectal varices were seen in 29 (30%) patients and large peri-rectal varices in 10 (10%). Small para-rectal varices were seen in 22 (23%) patients and large para-rectal in 13 (14%). Of the 83 patients without rectal varices on rectoscopy, 39 (47%) had varices detectable with EUS. Patients with large deep varices had more advanced liver disease and significantly thicker rectal wall. They did not differ from other patients in terms of the other analysed factors. Conclusion EUS permits identification of deep rectal varices in a large proportion of patients without detectable varices on rectoscopy. Unlike in previous study, we found that the presence of large deep rectal varices correlates with the degree of liver failure and thickness of rectal wall but not with the grade of portal hypertension in the oesophagus or the stomach. The clinical significance of these varices is uncertain and requires further study.