Ten tips for EUS-guided cystogastrostomy

By Shyam Varadarajulu, MD, Center for Interventional Endoscopy, Florida Hospital, Orlando, FL, USA

  1. What is the imaging modality of choice to evaluate a pancreatic fluid collection (PFC) prior to EUS-guided drainage?

The procedural technique and choice of endoprosthesis placement vary depending on the type of PFC being intervened: pseudocyst versus walled-off necrosis (WON). Although a computed tomography (CT) scanner is more widely available it does not quantify the degree of necrosis as precisely as a T2 weighted magnetic resonance imaging (MRI) scan. If an option is available, MRI should be the imaging modality of choice to evaluate a PFC.

  1. After a PFC has been identified for drainage, should a pancreatogram always be performed to confirm or exclude a pancreatic duct leak?

There are no data to support the routine performance of a pancreatogram in all patients with a PFC undergoing transmural drainage. In a recent meta-analysis of seven studies that included 551 patients with pancreatic fluid collections (PFCs), 48.8 percent underwent transmural drainage, 22.6 percent underwent transpapillary drainage and 25.6 percent underwent combined drainage. There was no significant difference in short-term treatment success rates between the three techniques:  79.9 percent, 80.8 percent and 87.9 percent, respectively. There were also no differences in rates of clinical success or recurrence among the cohorts. A substantial proportion (>55 percent) of patients with WON have disconnected pancreatic duct syndrome (DPDS) and placing a transpapillary stent is unlikely to be beneficial. On the other hand, in majority of patients with pseudocysts, the pancreatic duct is intact or only partially disrupted. When a bridging stent is placed across the duct leak in these patients, in conjunction with endoscopic transmural drainage, it restores ductal integrity and facilitates faster resolution of the pseudocyst.

  1. What are the key observations to be made at EUS prior to undertaking PFC drainage?

A) It is important to assess for the presence of a DPDS. When the distal pancreatic duct terminates into a PFC and cannot be traced any further (proximally), the patient likely has a DPDS (Figure). This is relevant because patients with DPDS tend to have a more progressive disease course that warrants repeat or adjunctive interventions.

B) Documentation of coexisting pancreatitis is vital. If an isolated PFC is observed but without coexisting changes of pancreatitis, a diagnosis of cyst neoplasm must be entertained.

C) It is critical to examine the pancreatic parenchyma proximal and distal to the PFC for the presence of a small hypoechoic mass, particularly in patients with unexplained pancreatitis. One in 80 patients referred for PFC drainage tend to have an underlying neoplasm. Finally, the point of maximal adherence/apposition of the gastric or duodenal wall to the PFC must be the chosen site for drainage. This is particularly important when draining an uncinate PFC. Although an uncinate PFC can be visualized from the stomach, once transmural dilation is performed via the gastric route, given the lack of adherence, the PFC tends to get smaller and ‘peels’ away from the gastric wall resulting in perforation. Therefore, uncinate PFCs should be preferentially drained via the trans-duodenal route.

  1. What type of stent should be placed – plastic or lumen-apposing metal stents (LAMS)?

No well-designed study has conclusively proven the superiority of one stent over the other. The only proven benefit for LAMS is the ease of placement which is important when treating very sick patients who cannot endure prolonged procedures and endoscopists not proficient in performing advanced interventions that involve multiple procedural steps.  For pseudocyst drainage, the plastic stent performs equally well as the LAMS. For WON, given the wider lumen, LAMS may be a better option as they tend to drain the necrotic contents better but this hypothesis remains to be proven.

  1. Are there specific circumstances where the placement of only plastic stents for PFC drainage is recommended?

In a patient with DPDS, a plastic stent can be left in situ indefinitely to drain the upstream

(disconnected) gland. An indwelling stent decreases the rates of PFC recurrence. Also, if a pseudoaneurysm is observed in vicinity of a PFC on CT, a plastic stent has less likelihood of causing friction and inducing delayed bleeding. Finally, if patient noncompliance is suspected, to minimize the likelihood of delayed adverse events, plastic stent placement is recommended.

  1. When is endoscopic necrosectomy warranted?

An endoscopic necrosectomy is almost never warranted at index treatment unless there is no drainage of necrotic contents (the WON is predominantly composed of solid debris) after stent deployment. We routinely re-evaluate the patient at 72 hours post index intervention to determine the need for undertaking necrosectomy based on the following criteria: (a) non-resolution or new onset of (post-procedure) systemic inflammatory response syndrome (SIRS) or organ dysfunction, (b) persistence of WON (i.e., less than 25% decrease in WON size after index intervention) on follow-up imaging and (c) lack of resolution of presenting symptoms.

  1. Are there specific steps that can be adopted to maximize technical efficiency when performing necrosectomy?

A cap-fitted therapeutic gastroscope enables ‘sucking’ of the debris into the cap so that it can be dislodged easily into the stomach. Oval, 15-30mm, snares are helpful for removing non-adherent debris and braided snares (CaptivatorTM Boston Scientific) for removing adherent debris. The long-jaw alligator forceps enable removal of debris that cannot be snared. A net may be used at conclusion of the procedure for removal all non-adherent material but all liquid contents must be aspirated first, before deploying a net, or else the mesh gets ‘soaked’ and fails to open adequately. Gentamycin 120mg  admixed with 250 cc  of saline can be used intermittently for irrigation during the procedure and saline admixed with hydrogen peroxide (10% hydrogen peroxide  with  equal amonts of saline) may be used at conclusion to ‘sterilize’ the cavity. Use of hydrogen peroxide at beginning must be avoided as the froth obscures visualization of vasculature within the cavity.

  1. What is the minimum weight or age of a child who can undergo EUS-guided PFC drainage?

A therapeutic echoendoscope has an outer diameter of 14 mm and has a rigid tip that limits its use in very young children (less than 18 months of age). In young children, a symptomatic PFC induces luminal duodenal compression. If technical difficulty is encountered with the passage of a linear array echoendoscope, then a duodenoscope or therapeutic gastroscope can be used in conjunction with an ultrasound mini-probe to guide transluminal PFC drainage.

  1. If a patient has a persistent PFC after a transmural stent placement, is surgical cystgastrostomy indicated?

The most common reason for persistence of a PFC following transmural stent placement is ineffective drainage. A repeat intervention is likely to be successful in more than 75% of patients even if the collection is infected. There are several technical tips that help improve the outcome of reintervention: a) lavaging the collection and aspirating its contents prior to deploying additional stents, b) creating additional transmural tracts, c) placing LAMS for better drainage and d) adjunctive interventions such as endoscopic or percutaneous necrosectomy may be required.

  1. When should LAMS be removed?

Traditionally, plastic stents are removed 6-8 weeks post-placement. Given the faster resolution of PFCs following LAMS placement, the stents must be removed earlier. The downfall of not removing a LAMS in a timely manner is that the embedded stent can erode into an adjacent vasculature precipitating bleeding.

For WON, a follow-up imaging must be performed at three weeks. If the necrotic collection has resolved or is 3cm or less in size, the LAMS must be removed. If the collection is persistent or the size of the residual collection is 4cm or more, then close follow-up is recommended with LAMS removal once the desired clinical outcome is achieved (resolution of symptoms and WON measuring less than 3 cm in size).

For pseudocysts (no debris or <10-20% debris in the collection) that drain rapidly after LAMS deployment, a follow-up CT is recommended in 10-14 days and the stent must be removed if the fluid collection has resolved. A pseudocyst that is 6cm or less in size at presentation is likely to drain very rapidly after LAMS placement. In the opinion of this author, bleeding can occur even as early as 10 days in such cases. Therefore, for pseudocysts, particularly those measuring less than 6cm in size, a follow-up CT scan is recommended in 10 days.