Endoscopic Ultrasonography-Guided Pancreatic Drainage

Surgery for pancreatic pseudocysts (PPCs) may be associated with high rates of complication and mortality. Since its introduction in the late 1980s, endoscopic drainage of PPCs has become established as an alternative nonsurgical approach. The obvious limitation of this technique is its relatively blind approach. The ideal approach for PPC puncture combines endoscopy with real-time endosonography using an interventional echoendoscope. Endoscopic ultrasonographic longitudinal scanners can be used for guidance of transmural punctures and drainage procedures, as well as to access a dilated pancreatic duct if the duct cannot be drained by conventional methods because of complete obstruction. In this chapter, we will review the indications, techniques and clinical algorithm for EUS-guided pseudocyst drainage and pancreaticogastrostomy.

The management of pancreatic pseudocysts (PPCs) has traditionally been surgical. Although highly effective, surgery may be associated with a complication rate of 35% and a mortality of 10%, a situation that has encouraged the development of nonsurgical approaches. Percutaneous puncture and aspiration under computed tomography (CT) or ultrasound guidance has been used, but aspiration alone has been found to be ineffective because of high recurrence rates of up to 71%. Continuous percutaneous drainage with indwelling catheters reduces the relapse rates, but may be associated with a complication rate ranging from 5% to 60%. Complications include fistula formation, infection, and bleeding.

Endoscopic transmural drainage of a PPC is an alternative nonsurgical approach. Since the first reports by Sahel and colleagues and Cremer and colleagues, endoscopic drainage of PPCs has become established. This procedure entails the creation of a fistulous tract between the PPC and the gastric lumen (cystogastrostomy) or duodenal lumen (cystoduodenostomy). Having established endoscopic access to the PPC, a nasocystic catheter or a stent can be placed for continuous drainage. The obvious limitation of endoscopic transmural drainage of a PPC is its relatively blind approach. The risk of perforation is particularly high when endoscopically visible intraluminal bulging is absent. A major risk of endoscopic cystoduodenostomy or cystogastrostomy is hemorrhage (6% of cases). The ideal approach for PPC puncture combines endoscopy with real-time endosonography using an interventional echoendoscope. Several investigators have described the use of endoscopic ultrasonography (EUS) longitudinal scanners for guidance of transmural puncture and drainage procedures. The same technique could be used to access a dilated pancreatic duct in cases where the duct cannot be drained by conventional endoscopic retrograde cholangiopancreatography (ERCP) because of complete obstruction.

EUS-guided drainage of pancreatic pseudocysts

Indications

PPCs are reported to complicate between 10% and 20% of patients with acute and chronic pancreatitis. Most of these PPCs are asymptomatic and do not require treatment. Spontaneous regression of PPCs is reported to occur in 7% to 60% of patients. The indication for PPC drainage will differ depending on whether the cyst develops in the setting of acute or chronic pancreatitis. For PPCs that complicate acute pancreatitis, drainage is indicated when pancreatitis fails to resolve with conservative measures. PPCs that are not associated with persistent pancreatitis should be kept under observation, as there is a high probability of spontaneous resolution. A 6-week observation period is generally recommended before considering decompression. Spontaneous regression after persistence of more than 6 weeks is considered by some to be unlikely. In fact, this cutoff time of 6 weeks is heavily doubted in the literature nowadays, and large pseudocysts (>4 cm) should be treated.

For PPCs complicating chronic pancreatitis, drainage is indicated to relieve symptoms associated with a space-occupying mass, including neighboring organ compression. Such patients have chronic cysts that remain unchanged over a period of months. Patients typically complain of a dull and constant pain, and may develop symptoms of gastric outlet obstruction or jaundice from bile duct compression.

Multiple or multiloculated PPCs sometimes cannot be adequately treated by an endoscopic approach, and warrant surgical resection. It should be remembered that an endoscopic approach contaminates the cyst and risks infection if the contents of the PPC cannot be completely drained.

Is EUS Necessary?

The main question is: what is the best route for draining a PPC? To obtain a response, Kahaleh and colleagues have reported a prospective comparative study on the two techniques of endoscopic transmural drainage and EUS-guided drainage. A total of 99 consecutive patients underwent endoscopic management of PPC according to a predetermined treatment algorithm as follows. Patients with bulging lesions without obvious portal hypertension underwent endoscopic transmural drainage, and all remaining patients underwent EUS-guided drainage. Patients were followed prospectively, with cross-sectional imaging during visits to the clinic. The investigators compared short-term and long-term results (effectiveness and complications) at 1 and 6 months postprocedure. Forty-six patients (37 men) underwent EUS drainage and 53 patients (39 men) had endoscopic transmural drainage. There were no significant differences between the two groups regarding short-term success (93% vs 94%) or long-term success (84% vs 91%); 68 of the 99 patients completed 6 months of follow-up. Complications occurred in 19% of EUS patients versus 18% of endoscopic patients, and consisted of bleeding in 3, infection of the collection in 8, stent migration into the pseudocyst in 3, and pneumoperitoneum in 5 patients. All complications except one were able to be managed conservatively. No clear differences in efficacy or safety were observed between conventional and EUS-guided cystenterostomy. The choice of technique is likely best predicated by individual patient presentation and local expertise.

From the technical point of view, the EUS-guided approach has two crucial steps. The first is the identification of an optimal point to puncture, without intervening vessels and with a short distance between the cyst and the gut wall. Once this point is identified, the endoscope should be straightened as much as possible in a stable position. The second critical step is that once the puncture has been performed and the guide wire is curled inside the cyst cavity, the wall dilator must be introduced without losing the endoscope position and under ultrasonographic view. Once the dilator has been inserted through the parietal fistula the ultrasonographic view is no longer needed, and the dilation and stent insertion can be made under endoscopic view.

A Web-based survey concerning the technique of pancreatic collection drainage was sent to members of the American Society for Gastrointestinal Endoscopy (ASGE) in the United States and internationally. Of the 3054 endoscopists to whom the survey was sent, 266 (8.7%) replied; 198 performed pseudocyst drainage (103 [52%] ASGE members from the United States and 95 [48%] international members). The median of the total number of drainages per physician was 15 (range 1–364). The transgastric route was the most commonly used drainage route (65%). Transmural entry was performed using a needle-knife in 53% of transmural drainages that were not EUS-guided. The number of stents placed ranged from 1 to 5, and these remained in place for 2 to 30 weeks. A CT scan was used before drainage by 95% of all respondents. EUS imaging was used before drainage by 72 of 103 (70%) United States endoscopists compared with 56 of 95 (59%) international endoscopists ( P = .1). EUS-guided drainage was used by 56% of United States endoscopists compared with 43% of international endoscopists ( P = .06). The most common site of transmural entry for drainage of PPCs appears to be the transgastric route. Although CT is the most commonly used predrainage imaging modality, EUS is used before and during transmural drainage of pseudocysts in both the United States and abroad, particularly in academic medical centers. Use of EUS before or during drainage does not appear to differ significantly among endoscopists in the United States and internationally.

More recently, Varadarajulu and colleagues reported a randomized study to compare the rate of technical success between EUS and esophagogastroduodenoscopy (EGD) for transmural drainage of PPCs. Thirty patients were randomized to undergo pseudocyst drainage by EUS (n = 15) or EGD (n = 15) over a 6-month period. Except for the sex, there was no difference in patient or clinical characteristics between the two cohorts. Although all the patients (n = 14) randomized to EUS underwent successful drainage (100%), the procedure was technically successful in only 5 of 15 patients (33%) randomized to an EGD ( P <.001). All 10 patients who failed drainage by EGD underwent successful drainage of the pseudocyst on a crossover to EUS. There was no significant difference in the rates of treatment success between EUS and EGD after stenting, either by intention-to-treat (ITT) analysis (100% vs 87%; P = .48) or as-treated analysis (95.8% vs 80%; P = .32). Major procedure-related bleeding was encountered in 2 patients in whom drainage by EGD was attempted; one resulted in death and the other necessitated a blood transfusion. No significant difference was observed between EUS and EGD with regard to complications either by ITT (0% vs 13%; P = .48) or as-treated analysis (4% vs 20%; P = .32). Technical success was significantly greater for EUS than for EGD, even after adjusting for luminal compression and sex (adjusted exact odds ratio 39.4; P = .001). The investigators concluded that when available, EUS should be considered as the first-line treatment modality for endoscopic drainage of a PPC, given its high technical success rate.

A second randomized study was published in 2009 and came to a different conclusion, suggesting that EUS-guided PPC drainage is equivalent to EGD, and should be performed in nonbulging PPCs. A total of 60 consecutive patients with PPCs were randomly divided into two groups to undergo either EUS (n = 31) or EGD (n = 29). The technical success rate, complications, and short-term and long-term results were prospectively evaluated.

The rate of technical success of the drainage was higher for EUS (94%, 29/31) than for EGD (72%, 21/29; P = .039) in ITT analysis. In cases where EGD failed (n = 8) because the PPCs were nonbulging, a crossover was made to EUS, which was successfully performed in all these patients. Complications occurred in 7% of the EUS group and in 10% of the EGD group ( P = .67). During short-term follow-up, PPC resolution was achieved in 97% (28/29) in the EUS group and in 91% (19/21) in the EGD group ( P = .565). Long-term results analyzed on a per-protocol basis showed no significant difference in clinical outcomes between EUD (89%, 33/37) and conventional transmural drainage (86%, 18/21, P = .696).

Recently, a large study on EUS-guided PPC drainage showed a low rate of complications. Of 148 patients who underwent EUS, perforation was encountered at the site of transmural stenting in 2 (1.3%, 95% confidence interval [CI] 0.41–4.76) patients with a PPC in the uncinate. Other complications included bleeding in 1 (0.67%, 95% CI 0.16–3.68), stent migration in 1 (0.67%, 95% CI 0.16–3.68), and infection in 4 (2.7%, 95% CI 1.09–6.73) patients. Bleeding occurred in a patient with underlying acquired factor VIII inhibitors, there was stent migration in a patient who underwent drainage via the gastric cardia, and infection occurred in 2 patients with pseudocysts and in 2 with necrosis. Whereas 2 patients who developed postprocedural infection and 1 with stent migration were managed endoscopically, both perforations required surgery. Surgical debridement was performed in 2 patients who developed infection, with a successful outcome in one and death from underlying comorbidity in another.

In addition to its safety and therapeutic success rate, EUS also allows a diagnostic evaluation of the pancreatic cystic lesions. Thus, based on the EUS findings, the management plan is changed in 5% to 9% of patients because EUS identifies other cystic lesions misdiagnosed as pseudocysts.

Which Technique and Accessories are Best?

EUS-guided PPC drainage should be performed under propofol anesthesia with endotracheal intubation to avoid regurgitation in the fluoroscopy suite, with the patient in the left lateral or prone position. The patient should receive broad-spectrum antibiotics during and after the procedure to reduce the risk of PPC infection. A CT scan should be performed immediately before the intervention. CT is less invasive than endoscopy and gives information about important anatomic details (eg, varices, arterial pseudoaneurysms, multiple cysts or extended necrosis, ascites, large or atypically located gall bladder, pleural effusion).

The individual steps are ( Fig. 1 ):

1.

Locate the cyst and the contact zone between the gastric or duodenal wall and the cyst wall.
2.

Doppler assessment of the stomach or duodenal wall is performed for interposed vessels. Doppler ultrasonography is now mandatory before cyst drainage.
3.

Having determined the optimal site for puncture, the PPC is punctured using a 19-gauge fine-needle aspiration needle or the new Access 19-gauge needle (Cook Medical, Bloomington, IN, USA), which prevents damage to the Teflon part of the 0.0035-in guide wire. A sample of the cyst contents is aspirated and submitted for biochemical, cytologic, and tumor marker (eg, carcinoembryonic antigen) analysis. If infection is suspected, a sample should be sent for Gram stain, as well as culture and sensitivity.
4.

Contrast filling of the PPC is performed under fluoroscopy to document the size and anatomic boundaries of the cyst. Communication of the cyst with the pancreatic duct may be seen. Filling of the cyst can also be verified by EUS, seen as a visible streamline effect.
5.

The tract is dilated using an 8-mm balloon over the wire or the 8.5F or 10F cystostome. The main advantage of the cystostome is to create a large cystenterostomy, due to the diffusion of the cautery at the level of the puncturing tract.
6.

A chronic cyst with clear liquid contents can be drained with 2 7F or 8.5F double-pigtail stents. An infected cyst mandates irrigation by nasocystic catheter or 2 10F double-pigtail stents, and a nasocystic drain can be placed (see Fig. 1 ). The nasocystic catheter can be removed after 2 or 3 days after a CT examination showing a resolution of the PPC. Pancreatic cysts complicating necrotizing pancreatitis can be managed endoscopically, but require aggressive irrigation and drainage over an extended period of time.