Endosonographic evaluation of the rectum in children with extrahepatic portal venous obstruction.

Rectal endoscopic ultrasonography (REUS) was performed using an Olympus EU-M3 ultrasound fiberscope in 25 children with extrahepatic portal venous obstruction (median age of 8, range 3-16 years) and in eight control subjects (median age of 8, range 6-13). Rectal varices (tortuous or nodular distended veins seen beneath the mucosa) and portal hypertensive rectopathy were diagnosed endoscopically. At rectal endosonography, rectal varices were seen as rounded, oval, or longitudinal echo-free structures in the submucosa. Perirectal veins outside the rectal wall were also seen in REUS. Portal hypertensive rectopathy was endoscopically observed in nine (36%) patients. Rectal varices were detected by endoscopy in nine (36%) and by endosonographic examination in 19 (76%) patients with extrahepatic portal venous obstruction. The number (2, range 0-10) and size (3, range 0-4.5 mm) of submucosal veins seen on endosonography in patients were greater than in controls (0, range 0-1, and 0, range 0-2 mm; p < 0.001 for both). The size of perirectal veins was greater in patients than in controls (3.5, range 3-5 mm versus 2.5, range 2-3 mm; p < 0.001). Communicating veins between the submucosal and perirectal veins were seen in nine (36%) patients. Rectal endoscopic ultrasonography was superior to endoscopy in detecting the presence of rectal varices (76% versus 35%, p < 0.005). Our study suggests that endosonography is useful in detecting changes in the rectal and perirectal vasculature in patients with extrahepatic portal venous obstruction.