Introduction

As obesity rates rise, both in the United States and in other parts of the world, surgical intervention that alters the gastrointestinal tract, for example the Roux-en-Y gastric bypass (RYGB), is becoming increasingly prevalent, and access to the distal GI tract is often a challenge. RYGB is effective in that it involves a long-limb bypass which causes maldigestion and malabsorption [1]. Endoscopic retrograde cholangiopancreatography (ERCP) in patients who have undergone RYGB surgery is problematic with success rates well below 70% [2], because of difficulties accessing the ampulla [3, 4]. To overcome such challenges, various approaches to accessing the biliary tract have been proposed in the literature. Percutaneous transhepatic cholangiography [5, 6], laparoscopic transgastric endoscopy [7], intraoperative transjejunostomy [8], and double–balloon-enteroscopy-assisted approaches [9–11] have been tried with variable success. Many of these methods enabling therapeutic ERCP are time-consuming, costly, invasive, and, in many cases, challenging. Single balloon-assisted ERCP, yet another alternative, might have higher rates of complications as compared to conventional ERCP, such as pancreatitis, as demonstrated by Wang et al. [12].

Transgastrostomy-ERCP (TG-ERCP) has been described in the literature [13, 14] and is regarded as a successful means of gaining access to the native ampulla for complex therapeutic intervention. The objective of this study was to evaluate the effectiveness of TG-ERCP in treating patients with complex pancreatico-biliary diseases and to evaluate procedure-related complications.

Methods

From December 2004 to July 2009, eleven patients (10 women) who had a remote RYGB presented to the University of Virginia Health System for therapeutic ERCP. The indications for all eleven patients included suspected gallstone pancreatitis (n = 4), ampullary/biliary strictures (n = 5), pancreas divisum (n = 1), and common bile duct clipping as a result of RYGB surgery (n = 1) (Table 1). Before proceeding with TG-ERCP, four failed standard ERCP, three failed at reaching the ampulla using a pediatric or adult colonoscope, two patients failed percutaneous transhepatic cholangiography, and one patient failed single-balloon-assisted ERCP.

Table 1 Patient characteristics
Full size table

These patients underwent TG-ERCP following these steps:

  1. (1)

    Open Surgical Distal Gastrostomy: Access to the peritoneal cavity was via an upper midline incision through the old upper midline scar. The distal stomach was identified and a 32-French Malecot gastrostomy tube was then inserted into the lumen of the stomach and secured with a double purse-string suture of #2-0 silk. The stomach was secured to the anterior abdominal wall with interrupted #2-0 silk sutures. This secured the tube in place and secured the stomach to the anterior abdominal wall circumferentially. The midline incision was then closed with running #1 PDS suture to close the fascia and peritoneum in a single layer. Subcutaneous tissues were irrigated with saline and the skin was closed with a skin stapler. Sterile dressings were then applied. The tube itself was secured to the skin with a 2-0 nylon suture and then sterilely dressed.

  2. (2)

    TG-ERCP: After the SG tract had been allowed to mature (median 45 days; Table 1), patients underwent ERCP. A guidewire was placed through the gastrostomy tube and coiled in the stomach under fluoroscopy. Subsequently, the gastrostomy tube was removed and consecutive bougie dilations (Savary-Gillard, Cook Medical, Winston-Salem, NC, USA) of the stoma were performed up to 36 French (Fig. 1). A ball-tipped catheter (Boston Scientific, Marlborough, MA, USA) was inserted through the working channel of the duodenoscope (TJV-160VF; Olympus America, Center Valley, PA, USA) and the guidewire that was coiled in the stomach was back-fed through the ball tip catheter. The duodenoscope was passed over the wire into the stomach and then advanced to the major papilla. Therapeutic ERCP was performed using standard devices, including sphincterotomes, retrieval balloons, and stents (Fig. 2).

    Fig. 1
    figure 1

    Bougie dilation over guide wire placed in the excluded stomach

    Full size image
    Fig. 2
    figure 2

    Endoscopic retrograde cholangiography performed through the gastrostomy track created

    Full size image

All procedures were performed by experienced pancreaticobiliary endoscopists who routinely performed more than 500 ERCPs annually. Patients were captured prospectively in our pancreatico-biliary database. This study was approved by our Institutional Review Board.

Results

Patient Characteristics

Eleven patients (10 women) median age 52 years, range 37–61 years, had previous RYGB. Indications for ERCP included gallstone pancreatitis (n = 4), one of which was complicated by pseudocyst formation requiring cystenterostomy (n = 1) (Figs. 3, 4, 5), ampullary/biliary stricture (n = 5), pancreatic divisum (n = 1), and iatrogenic bile duct clipping during gastric bypass (n = 1).

Fig. 3
figure 3

Mature pseudocyst identified on CT abdomen

Full size image
Fig. 4
figure 4

Fluoroscopic images of conventional cystogastrostomy performed though the gastrostomy track

Full size image
Fig. 5
figure 5

Resolution of pseudocyst confirmed by CT abdomen

Full size image

All patients underwent successful surgical placement of a Stamm gastrostomy. The median time between surgical Stamm gastrostomy placement and TG-ERCP was 45 days (range 29–135 days).

Therapeutic TG-ERCP

All eleven patients underwent biliary sphincterotomy and balloon sweep with stone/sludge extraction. Three patients underwent biliary stenting. One patient underwent pancreatic duct dilation with temporary stent placement. Three patients had ampullary biopsies. One patient underwent endoscopic transenteric pancreatic pseudocyst drainage and another had choledochoscopy using Spyglass (Boston Scientific).

Follow-up and Complications

Post procedure all patients were followed up in the clinic with a median follow up time of 42 days (range 7–123 days). Overall, three patients (27%) had post-surgical complications after placement of the gastrostomy tube. Two of the three patients developed surgical site infections (18%), which resolved with outpatient oral antibiotic therapy. One patient had mild bleeding (9%) that did not necessitate transfusion therapy (Fig. 6).

Fig. 6
figure 6

Post surgical (SG) complications

Full size image

Five patients (45%) developed post-ERCP complications (Fig. 7). Two of these patients (18%) had post-ERCP bleeding, considered to be mild and moderate [15]. One patient had a gastrointestinal bleed from TG-ERCP due to biliary sphincterotomy but did not require blood transfusion. The other patient presented with bleeding from the gastrostomy site the day after ERCP. This patient was transfused a total of two units packed red blood cells, and the bleeding site was found to be cutaneous and successfully cauterized with sliver nitrate.

Fig. 7
figure 7

Post ERCP complications

Full size image

Another two of the five patients had post-ERCP pancreatitis (18%). One patient with iatrogenic bile duct clipping intraoperatively, had severe post-ERCP pancreatitis using the Cotton classification [15], with subsequent abscess formation. A second was diagnosed with moderate post-ERCP pancreatitis [15].

One of the five patients (9%) sustained a contained gastric perforation [15] at the site of the endoscopic dilation. The patient was managed conservatively with IV antibiotics and total parenteral nutrition for two days before being able to start oral nutrition, with the PEG tube left in place to facilitate healing. She subsequently made a full recovery and returned to a regular diet.

Discussion

Bariatric surgery is an increasingly common procedure as obesity rates continue to rise in the United States. Consequently, it has become difficult managing these patients with surgically altered anatomy when they develop biliary or pancreatic diseases. Traditional peroral endoscopy for ERCP has a low success rate in patients who have had Roux-en Y gastric bypass. Wright et al. reported a success rate of 65–70% using a forward viewing pediatric or adult colonoscope followed by exchanging for a duodenoscope [2].

Other investigators have used pediatric colonoscopes and single or double-balloon enteroscopes to facilitate ERCP [12, 16] but these techniques also have their limitations. In particular, using a longer scope to perform therapeutic ERCP can be challenging, because the equipment currently available is suboptimal and does not enable the full array of therapeutic pancreatico-biliary intervention, for example pseudocyst drainage, choledochoscopy, or placement of large biliary stent [9, 11].

Percutaneous transhepatic cholangiography [5, 6] and laparoscopic transgastric endoscopy [7] are two techniques that have been described in the literature. The former is the current standard of patient care [6] yet, despite advances in technology, PTC is not reliably able to provide some therapeutic options, for example biliary or pancreatic sphincterotomy. The latter involves a more challenging single-stage procedure and may require conversion to open gastrostomy, as described by Gutierrez et al. [17], making it, at times, a multi-step procedure.

TG-ERCP was first described by Baron et al. [13] and seems to be a compromise of all the above-mentioned techniques. Comparable with our study, Gutierrez et al. [17] reported the successful use of surgical gastrostomy for RYGB anatomy to perform various techniques including ERCP. Their study demonstrated the use of a laparoscopic surgical gastrostomy as a single-step procedure with pancreaticobiliary intervention. Conversely, our study demonstrates a two-step procedure in which patients underwent an open gastrostomy with a maturation period before endoscopic intervention. Gutierrez et al. noted a 10% laparoscopic re-exploration rate with ultimate conversion to open gastrostomy [17]. Our series was able to avoid this conversion of access, and has the advantage of gastrostomy tract reusability for repeat ERCP in the future, which is important when treating benign conditions. It also enables the wider variety of endoscopic intervention that surgical gastrostomy allows, for example biliary and pancreatic stenting, ampullary biopsies, pseudocyst drainage, and choledochoscopy.

In our study, TG-ERCP has its own set of limitations. Primarily, the invasive approach for placement of the SG tube exposes the patient to the risk of post surgical complications, for example infections, bleeding, or gastric perforation. The other major limitation is the significant delay between SG tube placement and the therapeutic ERCP. While such delays could be limited to perhaps three weeks, even this time frame excludes emergency use of the technique. Thus, this technique could be reserved for stable patients with difficult anatomy having recurrent pancreatitis or biliary obstruction that have failed the various current endoscopic intervention requiring biliary or pancreatic intervention not amenable to PTC.

Overall, TG-ERCP enables therapeutic ERCP in patients that have undergone RYGB surgery. Although effective, this technique is associated with both surgically and endoscopically related complications and careful patient selection is indicated. The limitation of this technique is the time delay required for maturation of the gastrostomy tract. Given the paucity of data about this modality, further research is warranted to evaluate the safety and efficacy of this procedure as a means of facilitating ERCP in gastric bypass patients.