Intervention or not, post-colonoscopic pneumoperitoneum is well-known, though it occurs more frequently post-intervention, be it stenting or polypectomy. Typically, the presumed micro perforations are not found, such that we often attribute it to trans-colonic transfer due to raised intraluminal pressure. Anecdotally, it seems to be less prevalent when using CO2 instead of air, which is logical - the gas probably just gets absorbed. The situation is similar to our initial fears following lab experiments with laparoscopic pneumoperitoneum in the early 90s; stenting is definitely the way forward acutely; seedling remains a theoretical risk, but the risk is no longer theoretical if the tumour perforated due to obstruction. The preventative measures include less sedation, less distension, not persevering if there is no progress and, it goes without saying, discussing possible scenarios with the patient. After all, the incidence of perforation of a malignant stricture is not insignificant anyway such that sometimes a patient might decide to go straight to surgery or, for an attempted stenting and proceed in theatre.

I have changed my practice with the deployment of such stents to exclude the use of any gas, and have adopted water exchange as a technique to avoid such complications.

While I have no direct measurement to prove this, it seems self evident that use of a gas results in greater intraluminal pressure than that achieved with water, and gas readily spreads throughout the colon, whereas the expansive force of water is more localized.

In support of these ideas,  cecal perforation in the setting of not having achieved cecal intubation is well described.

The available devices in the United States have the additional limitation of all being woven - such devices are demonstrated to have little conformabilty (See Isayama) and all exert force to become straight - a particularly undesirable trait for this space.

Never dilate such stents, as the risk for perforation increases dramatically.

Using water exchange and combining fluoroscopy with endoscopy can result in a highly efficacious procedure.

Now we just need better devices

I am from a small/medium size hospital in Japan and we perform 4 to 5 colonic stenting procedures per year for colorectal cancer patients. We encountered only one case of abscess formation due to micro-perforation but as far as pneumoperitoneum (without inflammation) is concern, if the evaluation using abdominal X-ray is considered sufficient, then we have no experience. We do not routinely perform abdominal CTs after procedure unless there is something wrong in the clinical course of patients. My search of the literature reveals that although the frequency of perforation is documented (up to 5% depending on the type of stent used and date of publication), there is no record of the frequency of pneumoperitoneum (asymptomatic). How was pneumoperitoneum detected in your three cases who I assume were asymptomatic? Is it routine in your hospital for abdominal CT to be taken after the procedure? 

Dear doctors

Thanks for giving me several instructions for the question.

Actually, I make it a rule to perform CT scan routinely after the stent placement even if patients have no symptoms.  In our cases, pneumoperitoneum was followed regularly by CT scan although there was no symptoms (except abdominal distention) or no evidence of peritoneal irritation. Pneumoperitoneum seemed to improve by a conservative therapy.  I think asymptomatic pneumoperitoeum is more than 5% if CT scan is routinely performed after the stent placement. I have to add that we did not use carbon dioxide insufflation in these patients. Thanks for sharing your experiences! 

I must say, I agree with Dr Wong. Close monitoring is indicated, but, even if the incidence of an asymptomatic pneumoperitoneum were 10%, I would find routine post-procedure CT difficult to justify, considering management is based on the clinical findings. The scenario in oesophageal stenting is much different, where spontaneous resolution is not the norm and one has to look for pneumomediastinum proactively.