Inflammatory tumor in pancreas: a novel complication after endoscopic injection of cyanoacrylate for gastric fundal varices

Abstract

To the Editor: Bleeding from gastric fundal varices (FV), a complication of portal hypertension resulting from cirrhosis, can be massive and remains a major therapeutic challenge. We read with great interest the article by Sato et al.1 in the Journal of Gastroenterology, about inflammatory tumor of the tail of the pancreas as a late complication after endoscopic ablation of FV with n-butyl2-cyanoacrylate (NBCA). Their endoscopic treatment was performed twice. We speculate that they used 2 ml of NBCA diluted with Lipiodol (at a ratio of 1.4 : 0.6) per injection. Did they performed two treatment sessions, or were two injections given at one treatment session? We strongly agree with the use of NBCA for achieving hemostasis of bleeding from FV.2 However, because most FV are associated with a large gastrorenal shunt,3 the value of NBCA for treating large (>12-mm-diameter) FV without active bleeding is controversial, due to the potential for systemic complications related to embolization through the gastrorenal shunt.4 Therefore, after achieving initial hemostasis with NBCA, portal hemodynamics should be evaluated to decide the most appropriate subsequent therapeutic strategy. Balloon-occluded retrograde transvenous obliteration (BRTO) is a recently developed method for treating FV associated with a gastrorenal or gastrophrenic shunt; the method employs ethanolamine oleate, a common sclerosant.5 B-RTO achieves excellent prevention of recurrent bleeding with few major complications (such as fever, hemoglobinuria, and worsening of esophageal varices), even in patients with poor liver function.6 The main limitation in using B-RTO in an emergency setting is the requirement for temporary control of bleeding with or without NBCA. Recently, multi-detector row computed tomographic angiography has been able to provide excellent visualization of draining vessels such as gastrorenal or gastrophrenic shunts.7 When FV are not associated with catheterizable draining veins, additional endoscopic obliterative therapy may be justifiable. For the elective treatment of FV, the optimal ratio of NBCA to Lipiodol has not yet been determined. A ratio of more than 70% means that hardening starts to occur in the injection catheter.3 Accordingly, NBCA-Lipiodol mixture with a high ratio must be injected rapidly, and distal embolization is likely to occur. Eradication of the feeding veins, as well as the varices, is necessary for the prevention of recurrent bleeding.8 Slow injection of an NBCA-Lipiodol mixture (with a ratio of more than 40% and less than 70%3) until visualization of the feeding vein occurs under fluoroscopic monitoring9 might be useful to avoid the novel complication reported by Sato et al.1 Akio Matsumoto, Kengo Takimoto, Yuuki Yamauchi, Masashi Kuchide, and Hideto Inokuchi Department of Gastroenterology, Takeda General Hospital, 281 Ishida Moriminami-cho, Fushimi-ku, Kyoto 601-1495, Japan

DOI: 10.1007/s00535-004-1641-1

Cite this paper

@article{Matsumoto2004InflammatoryTI, title={Inflammatory tumor in pancreas: a novel complication after endoscopic injection of cyanoacrylate for gastric fundal varices}, author={Akio Matsumoto and Kengo Takimoto and Yuuki Yamauchi and Masashi Kuchide and Hideto Inokuchi}, journal={Journal of Gastroenterology}, year={2004}, volume={40}, pages={854-854} }