LAUSR

Because, as this study shows, most ganglia are lateral to the celiac axis, the most appropriate intervention for the control group in a randomized trial is bilateral CPN. Until a study comparing CGN with bilateral CPN is published, the superiority of either method remains uncertain. In this issue of Gastrointestinal Endoscopy, Kappelle et al took on the daunting task of finding out what actually happens during attempted EUS-guided celiac plexus neurolysis (CPN) and EUS-guided ganglia neurolysis (CGN) in a cadaver model. In the former, the neurolytic agent is injected around the ganglia, whereas in the latter, it is injected in the ganglia. There are multiple reasons to suspect that targeted ganglia injection is unnecessary or inappropriate: (1) if there are multiple ganglia, many ganglia would need injection (not just 1 or 2), (2) if the ganglia are very small, there is not much space to inject anything anyway, (3) several controlled trials have shown that not targeting ganglia (CPN) is quite effective (so it would be hard to show significant clinical superiority for CGN), (4) the data supporting the superiority of CGN over CPN are flawed, and (5) CPN is painful, but sticking the needle directly into the ganglia is even worse! (Unpublished observation.) Several findings of this study appear to confirm the futility of CGN, but owing to the limitations of the cadaver model, it leaves some cogent issues unresolved. Regarding the number of ganglia, it appears clear that the number of ganglia visible by EUS is small (generally 1 or 2), and they are seen mostly (66%) on the left side or anterior to the celiac axis (30%). Therefore, right-sided ganglia are visible in only 4% of cases. However, in the cadavers, the actual number of ganglia identified was generally twice as high (3-5), and approximately 50% (1-2 per cadaver) were found on the right side (despite generally being invisible by EUS). This is potentially useful information. Presumably, if EUS views only a fraction of the existing ganglia, and if pain relief correlates with the number of ganglia exposed to sclerosant (which makes sense but is unproven), then a technique that safely spreads the drug beyond visible ganglia would be preferable. The study also confirms that the celiac ganglia are quite small, that targeted injection of even small amounts of sclerosant is difficult (high resistance was experienced during direct injection), and that sclerosant diffuses well beyond the capsule of the targeted ganglia.