Clue by the Blue: Using Methylene Blue as an Adjunct to Facilitate Access and Therapeutic Upper Endoscopic Interventions

Methylene blue (MB) is a dye studied in various gastrointestinal organs, including the esophagus, stomach, small intestine, and colon. MB stains actively absorb tissues, such as small intestinal and colonic epithelium, and the absence of staining suggests metaplastic, neoplastic, or inflammatory changes while staining in other areas of the gastrointestinal tract suggests metaplastic changes.^[1] MB staining has been used in the detection of Barrett’s esophagus, early gastric cancer, colonic dysplasia, and intraepithelial neoplasia in patients with ulcerative colitis.^[1] However, its use in therapeutic upper-endoscopic interventions has not been reported widely in the literature. The use of endoscopic ultrasound (EUS)-guided methylene blue (EUS-MB) to facilitate endoscopic interventions was reported rarely in the literature.^[2–8] In this article, we presented two cases in which MB was used to facilitate endoscopic interventions. The patients consented to the publication of their respective cases.

A 67-year-old man with a history of metastatic bladder carcinoma status post left nephrectomy and adrenalectomy complicated by pancreatic tail injury and leak with retroperitoneal fluid collection presented to the hospital with poor oral intake, lightheadedness, and dizziness. Endoscopic retrograde cholangiopancreatography (ERCP) was done for pancreatic duct stent placement; however, cannulation of the pancreatic duct was unsuccessful after multiple attempts due to nonvisualization of the pancreatic duct opening, despite doing an extended needle knife sphincterotomy. The patient was referred to interventional radiology and underwent percutaneous drainage of the retroperitoneal collection. Three months after percutaneous drainage placement, the patient presented to the hospital and had his drain removed. Four days later, he presented to the emergency department with a new onset of fever, chills, malaise, and anorexia. He was noted to be hypotensive and tachycardic, but vital signs were otherwise within normal limits. Laboratory data showed a white blood cell count of 12 K/CUMM concerning retroperitoneal fluid recollection complicated by secondary infection. Computed tomography (CT) of the abdomen and pelvis revealed an abnormal attenuation adjacent to the tail of the pancreas and hemidiaphragm overlying the spleen, concerning infection and abscess.

The patient was diagnosed with sepsis and started on piperacillin-tazobactam. A new percutaneous drain was placed with purulent material found. The patient was again referred to gastroenterology and scheduled for ERCP with pancreatic duct placement and possible EUS with cyst gastrostomy. A guide wire was passed into the biliary tree. The cholangiogram was normal. Cannulation of the pancreatic duct was attempted multiple times through the major papilla but was unsuccessful. The minor papilla cannulation over a guidewire was also tried but was unsuccessful (Fig. 1A). Given the unsuccessful cannulation of the major and minor papilla, continuous pancreatic leakage, and the need for definitive treatment, EUS was performed and showed distant collection from both the gastric and duodenal walls to drain. EUS was used with an injection of contrast into the pancreatic duct using a 22-gauge slimline needle to obtain a pancreatogram. The pancreatogram revealed a stricture at the head of the pancreas with contrast extravasation at the tail. Then, 5-mL diluted MB (1 mL diluted by 9 mL of saline) was injected into the pancreatic duct to identify the pancreatic duct orifice. An ERCP scope was then used to access the second duodenum to trace the MB dye. Inspection of the major papilla identified the pancreatic duct orifice (Fig. 1B). A 0.025-inch guidewire was passed into the ventral pancreatic duct, which was deeply cannulated with a short-nosed traction sphincterotome. A 5-mm pancreatic duct stricture was found within the proximal pancreatic head, which was dilated with a 4-mm balloon dilator (Fig. 1C). A 5-Fr × 3-cm stent with a full external pigtail and a single internal flap was placed into the ventral pancreatic duct (Fig. 1D). The pancreatic duct was draining very well without any contrast shown within the pancreatic system on the final scout film. The patient was discharged home the next day with a percutaneous drain in situ and scheduled for another ERCP after 6 weeks for pancreatic stent removal. The patient did not follow up as he was transitioned to comfort care due to widespread metastasis.

A 60-year-old man with a history of atrial fibrillation status post pacemaker placement and Roux-en-Y gastric bypass at 40 years of age was recently admitted to the hospital with acute pancreatitis. CT of the abdomen and pelvis with contrast at this admission showed a prominent fluid collection surrounding the patient’s gastrojejunal anastomosis concerning gastric marginal ulcer contained perforation (Fig. 2). Esophagogastroduodenoscopy (EGD) was done and showed a normal gastrojejunal anastomosis. Given the intra-abdominal inflammatory fluid collection, the patient was scheduled for CT-guided fluid aspiration. Preprocedural imaging demonstrated 9.2 × 7.5–cm gas and fluid collection within the area of concern suspicious of abscess. The case was discussed with interventional radiology (IR), and given the concern of abscess on preprocedural imaging, it was recommended to treat the patient with a drain instead of aspiration. CT-guided drainage was done, and purulent fluid was shown, which was sent for cytology, gram stain, and culture. The gram stain and culture later grew Enterococcus Spp., coagulase-negative staphylococcus, and Candida albicans; cytology was negative for malignant cells. The patient was discharged on ciprofloxacin and metronidazole with an intra-abdominal drain in place. During follow-up with IR as an outpatient, the patient reported a high-drain output with a repeated CT scan showing persistent abscess that looked controlled with drainage. Given that, a fluoroscopic sinogram was done and showed contrast in the perigastric region with fistulous communication to the bypassed stomach and extending to the gastrojejunal anastomosis and the residual stomach suggesting fistula formation (Fig. 3). The patient was referred to gastroenterology and scheduled for upper EGD and EUS to internalize the drain. The EGD was done and showed erythema, erosion, and inflammation of the gastrojejunal anastomosis (Fig. 4A). A questionable 1- to 2-mm fistula orifice was found within the gastric pouch at the gastrojejunal anastomosis (Fig. 4B). A 3-mL diluted MB was injected into the questionable fistula orifice. The MB was seen promptly within the external drainage catheter and bag, which was placed by IR, demonstrating a connection with the gastric pouch through the fistula (Fig. 4C). A 0.035 guidewire was placed through the fistula and de-epithelization of the fistula tract was performed using a cytology brush (argon plasma coagulation was not used for this purpose because the patient had implanted gastric pacemaker). Over the guide wire, a 7-Fr × 3-cm double pigtail stent was placed through the fistula under EGD and EUS guidance because fluoroscopy was not available at the time of the procedure (Fig. 4D). The patient was in stable condition after the procedure and was discharged to continue outpatient follow-up IR and gastroenterology. Follow-up imaging showed resolution of the intra-abdominal fluid collection without any evidence of remaining stents through the gastric pouch.

The use of EUS-MB helped achieve successful intervention in both cases, which enabled drainage of the pancreatic duct and clearing of persistent infection in the first case and gastric staple line fistula detection and repair in the second case. Selective cannulation of the pancreatic duct during ERCP can be challenging, even for experienced endoscopists, because of the inability to locate the pancreatic duct orifice. Cannulation of the pancreatic duct at the major papilla has demonstrated success rates of 90–98%, whereas cannulation at the minor papilla is more challenging, with failure rates ranging from 5–10%.^[2] Unsuccessful and prolonged ERCP subjects the patient to post-ERCP pancreatitis and delays therapy. In our case, cannulation of the major and minor papillae was unsuccessful after multiple attempts because of the inability to identify the pancreatic duct orifice. Locating the pancreatic orifice can be challenging in patients with an ampullary tumor, a history of ampullectomy, pancreatic orifice stenosis secondary to endoscopic sphincterotomy or surgical sphincteroplasty, and pancreatic duct stricture or surgical deformity.^[2] In our case, the patient had a pancreatic injury that might have contributed to the difficulty of localizing the pancreatic orifice. The use of EUS-MB to perform cholangiopancreatography, followed by ERCP, was reported in rare cases in the setting of prior ERCP failure due to difficulty locating the pancreatic orifice.^[2–7] Consiglieri et al^[2] performed a study to determine the safety and the effectiveness of MB-guided cholangiopancreatography. The study included 11 patients with a prior ERCP failure due to difficulty locating the pancreatic orifice. All the cases had successful cannulation, but one was complicated by peripancreatic abscess formation attributed to precut and anticoagulation therapy, a known uncommon complication of EUS and ERCP.^[2] The use of EUS-MB was also reported by Munzor et al^[3] in two cases after the initial ERCP failed to locate the pancreatic orifice. The first patient had a Whipple resection for chronic pancreatitis, and the second patient had ampullary adenoma status post ampullectomy and pancreatic divisum. Our first case was similar to those previously reported cases in which ERCP initially failed due to the challenging anatomical location of the pancreatic orifice and then succeeded after EUS-guided MB injection with no reported complications.

The use of MB with advanced endoscopic procedures has been reported rarely in the literature. Wang et al^[8] reported using EUS-MB to guide the creation of a EUS-guided gastrojejunal anastomosis and the insertion of a double-flared, fully covered metal stent after an unsuccessful attempt to pass the endoscope through malignant duodenal stenosis. A guidewire was advanced, and MB was mixed with normal saline and injected through the catheter to dilate the proximal jejunum. EUS was then used to visualize the dilated bowel near the stomach, and the bowel was punctured. The aspiration of the MB confirmed the location of the proximal jejunum.^[8] This is similar to our second case, where EUS-MB was used concurrently with an advanced endoscopic procedure to assist in locating and confirming the desired anatomical place. In our case, the fistula connected the gastric pouch and the gastrojejunal anastomosis. This highlights the novel use of MB in advanced endoscopy and opens the door for more research on other potential benefits in variable procedures. Moreover, there is a reported benefit of intraoperative MB injections in laparoscopic radical gastrectomies in preventing a missed anastomotic leak.^[9] In the literature, the use of oral MB has been reported to confirm the absence of leakage from postsurgical anastomoses in the upper gastrointestinal tract after stomach and esophageal surgeries if there was no blue dye present in the postsurgical drain. However, this is not applicable in the lower gastrointestinal tract as contact of MB dye with the high bacterial load for an hour or more will reduce the MB to leucomethylene blue and result in a change of dye color that would lead to false-negative results.^[10] Moreover, the oral MB challenge postoperatively has been found comparable to the upper gastrointestinal series in ruling out leaks after gastric bypass surgeries. This offers a cheaper alternative with similar efficacy and no radiation exposure.^[11]

MB-directed biopsies during EGD have also been studied, showing an improved detection rate of intestinal dysplasia and Barrett’s esophagus. Of note, in one study, MB-directed biopsies identified more cancers using fewer biopsies compared with the traditional jumbo random biopsy technique. This method not only enhances cancer detection but also proves to be more cost-effective.^[12,13] The use of MB with advanced endoscopic procedures has been reported rarely in the literature. More research is needed regarding the various uses of MB in advanced endoscopy and to compare it with other traditional methods.

MB is a cost-efficient, safe, versatile, and accessible diagnostic tool that can be used in surgeries and procedures to improve diagnostic and surgical outcomes. However, it is uncommonly used in advanced endoscopy. This article highlights the use of EUS-MB to guide ERCP when pancreatic anatomy is challenging and difficult to access, as well as its role in gastric fistula detection and repair, especially when fluoroscopy is not available. These findings pave the way for further research into more applications of MB in advanced endoscopy.