Article Text

Download PDFPDF

Management of gastric fundal varices associated with a gastrorenal shunt
  1. Department of Gastroenterology
  2. Takeda General Hospital
  3. 28-1 Ishida Moriminami-cho
  4. Fushimi, Kyoto, Japan
  5. Second Department of Internal Medicine
  6. Osaka Medical College, Takatsuki, Osaka, Japan
  1. A Matsumoto, Department of Gastroenterology, Takeda General Hospital, 28-1 Ishida Moriminami-cho, Fushimi, Kyoto, 601–1495, Japan.akio_m{at}
  1. R JALAN
  1. P C HAYES
  1. Institute of Hepatology
  2. University College London Medical School
  3. London, UK
  4. Liver Unit, Royal Infirmary of Edinburgh
  5. Lauriston Place, Edinburgh EH3 9YW, UK
  1. Dr R Jalan.r.jalan{at}

Statistics from

Editor,—We read with great interest the article by Jalan and colleagues (OpenUrlCrossRefPubMedWeb of Science) on the clinical position of transjugular intrahepatic portosystemic stent-shunt (TIPSS). This procedure is a useful method of reducing portal pressure by creating a portosystemic shunt in the liver. They suggested that TIPSS can be a successful treatment for bleeding gastric fundal varices (FV) unresponsive to pharmacological and endoscopic therapy. However, Sanyalet al reported that TIPSS was ineffective for FV associated with a large gastrorenal shunt, even when the hepatic venous pressure gradient falls below the critical bleeding threshold of 12 mm Hg.1

The behaviour of varices at different sites seems to differ.2 Therefore, FV should be treated on the basis of their haemodynamics. FV arise from the dilation of short or posterior gastric veins and are frequently associated with a large gastrorenal shunt that decompresses the portal system.3 Balloon occluded retrograde transvenous obliteration (B-RTO) is a novel radiological treatment for FV that was developed by Kanagawa and colleagues.4 This procedure involves insertion of a balloon catheter into a gastrorenal shunt via the femoral or internal jugular vein. It is similar to TIPSS but less invasive. The therapeutic effect of B-RTO is excellent without major complications, even for patients with poor liver function.4 ,5 However, there have been few controlled trials of this technique.6

Patients with bleeding from FV have a high risk of dying from an episode of variceal bleeding or from liver failure, even when TIPSS is successful in stopping acute bleeding.7 Hence patients with high risk FV should preferably undergo prophylactic treatment. Although the risk factors for the first episode of bleeding from FV are still not clear, Kim et al determined the one year probability of bleeding as a function of all possible combinations of two endoscopic variables (variceal size and the presence of red spots) for patients in Child's class A, B, or C.8 According to their classification, FV with a one year probability of bleeding (16%) can be considered as high risk varices.

As TIPSS seems to be ineffective for FV associated with a gastrorenal shunt, β blockers or nitrates (which are widely used to treat high risk oesophageal varices) may also be ineffective for primary prophylaxis of bleeding from FV. Accordingly, prophylactic B-RTO may be justifiable due to its simplicity and safety.

Because the gastrorenal shunt tends to be occluded after B-RTO,4 ,5 however, the long term effect of this procedure on portal haemodynamics needs to be evaluated.

Although a prospective randomised study comparing B-RTO with TIPSS for the prevention of bleeding or rebleeding from FV is still needed, we hope that B-RTO will become a firstline treatment for high risk FV associated with a gastrorenal shunt in the near future.



Editor,—We thank Matsumoto and colleagues for their interest in our paper. They suggest that transjugular intrahepatic portosystemic stent-shunt (TIPSS) is ineffective for the management of bleeding from fundal varices and given the haemodynamic characteristics of fundal varices, the appropriate treatment for bleeding from them is balloon occluded retrograde transvenous obliteration (B-RTO). They quote Sanyal's paper1-1 as evidence in support of their suggestion that TIPSS is unlikely to be useful in the setting of fundal varices. Sanyal et al reported their experience of TIPSS in 12 patients who underwent this procedure for gastric varices and in six patients these varices did not disappear on follow up. The aim of treatment of bleeding varices is firstly to control bleeding and secondly to prevent rebleeding. In the paper by Sanyal et al, no data were provided about how many patients bled from gastric varices in the follow up period compared with those who rebled with oesophageal varices.1-1However, our previous study1-2 and that of Chau and colleagues1-3 clearly show that post-TIPSS bleeding from either oesophageal or gastric varices is a function of portal pressure and has little to do with whether bleeding is from oesophageal or gastric varices. Both Stanley and colleagues1-2 and Chau and colleagues1-3 compared the outcome of TIPSS insertion for variceal bleeding from oesophageal or gastric varices. In the study by Stanley et al, 106 patients (oesophageal varices 74; gastric varices 32) underwent TIPSS for variceal bleeding and during follow up the rates for variceal rebleeding were similar in both groups and there was no difference in survival. In the study by Chau et al, 112 patients (oesophageal varices 84; gastric varices 28) with variceal bleeding underwent TIPSS for uncontrolled variceal bleeding. Bleeding was controlled in all patients after TIPSS except for one in each group. Twenty four per cent of patients in the oesophageal varices group and 29% in the gastric varices group rebled during follow up. Most early rebleeding (within seven days after TIPSS) was related to oesophageal ulceration secondary to previous sclerotherapy. Rates of mortality were similar in both groups. These results suggest that emergency TIPSS is equally effective in the control of gastric fundal variceal bleeding compared with oesophageal variceal bleeding.

Matsumoto et al also suggest that there is likely to be a place for B-RTO in the primary prophylaxis of bleeding from fundal varices and that pharmacological agents have no place in their management. Again, the data for their suggestion do not exist in the literature. We think that it is extremely difficult to suggest failure of pharmacological therapy for primary prophylaxis of fundal varices based on the assumption that portal pressure changes are unlikely to be important in the management of fundal varices.

The data in the literature do not support either of the points that have been suggested by Matsumoto et al. Although data on the use of B-RTO for the treatment of fundal varices are exciting, we look forward to randomised controlled clinical trials comparing TIPSS with B-RTO.


  1. 1-1.
  2. 1-2.
  3. 1-3.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.