Does the endoscopic view reflect the size of the gastric varix?

A 55-year-old man with a known history of HCV cirrhosis was admitted with hematemesis. His initial abdominal CT showed a huge gastric varix (A), with a maximum diameter of the varix of 8 mm (although from our experience, cross-sectional imaging is not a good indication of the volume), and CT portography also demonstrated a gastro-renal shunt (B). Eight ampules of N-butyl-2-cyanoacrylate (Histoacryl, B. Braun, Melsungen, Germany) were used (each 0.5 mL mixed with 0.8 mL Lipiodol, Guerbet, Villepinte, France). Each ampule requires about 30 seconds for injection and also about 60 seconds between injections for solidification. For several years we have been doing the injections under fluoroscopic guidance, but currently we rely on palpation of the varix for the assessment and confirmation of complete solidification. We also use postinjection CT portography in cases of nonexpulsion of Histoacryl and if the varix persists 6 months after injection. Postinjection CT portography in this case showed obliteration of the gastrorenal shunt (C), as is usually the case in our experience. We would like to highlight 3 points from our experience in similar cases. We have observed that the actual size of the varix cannot be accurately determined until it is fully obliterated by the tissue adhesive injection (thus rendering impractical the practice of risk categorizing the varices by size before injection; Figure D shows the varix before and after injection). We have also observed that reducing the incidence of recurrent variceal bleeding is related to the success of complete filling of the varix (which appears to be closely associated with obliteration of the gastrorenal shunt). Last, it appears that palpating the varix, ensuring no remaining soft parts, correlates well with gastrorenal shunt occlusion.