Society: ASGE

Introduction
Malignant colorectal polyps are defined as lesions that invade into the submucosa and represent early colorectal cancer (T1 by the TNM Classification of Malignant Tumors system). Management of these lesions depends on the depth of invasion which can be estimated endoscopically using various classification methods that assess lesion morphology and surface features. Superficially invasive lesions are generally amenable to endoscopic resection techniques whereas deeply invasive lesions are typically managed with surgical resection given the high risk of invasion beyond the submucosa. Here, we describe successful endoscopic resection of a deeply invasive colon lesion in a non-surgical candidate using hybrid endoscopic submucosal dissection (ESD) and endoscopic full-thickness resection (EFTR) with the full-thickness resection device (FTRD).

Case
A 71-year-old male with a history of chronic kidney disease stage IV, type II diabetes mellitus, peripheral arterial disease and below-the-knee amputation, and recently diagnosed pulmonary embolism was found to have a colon mass on imaging after presenting with acute on chronic anemia. Colonoscopy revealed a 30 mm polypoid lesion (Paris classification 0-Is) in the hepatic flexure with features concerning for submucosal invasion (NICE classification Type III), including superficial ulceration and oozing with minimal contact (Figure 1A). After multidisciplinary discussion, the patient was felt to be a poor surgical candidate and thus endoscopic resection using endoscopic submucosal dissection (ESD) was pursued in case the invasion of the cancer was limited to superficial submucosa.

Procedure
A circumferential incision surrounding the lesion was made and endoscopic submucosal dissection was performed (Figure 1B). During dissection, invasion into the muscularis propria was visualized (Figure 1C) and thus the decision was made to use the full-thickness resection device (FTRD) to complete the resection.

The grasping forceps were used to pull the lesion into the device cap and the clip was deployed. Given the presence of submucosal invasion, gentle intermittent suction was applied to ensure complete entrapment of the lesion into the FTRD cap. The snare was closed and the lesion was resected using electrocautery (Figure 1D). The final pathology report revealed T2 invasive adenocarcinoma (Figure 2) with negative resection margins (R0).

Conclusions
While deeply invasive colon lesions are traditionally removed surgically, endoscopic resection using hybrid ESD-EFTR may offer an alternative treatment approach in select patients who are non-surgical candidates. In this case, initial submucosal dissection allows for the lesion to be more pliable and hence more amenable to be completely pulled into the FTRD cap which allowed for complete endoscopic resection of a T2 adenocarcinoma.

Figure 1. A) A 30 mm polypoid lesion with overlying ulceration is seen at the hepatic flexure (Paris classification 0-Is, NICE Type III). B) An initial circumferential incision around the lesion is made for ESD. C) Lesion invasion into the muscularis propria is visualized during submucosal dissection. D) Final view of the resection site showing appropriate position of the clip.

Figure 1. A) A 30 mm polypoid lesion with overlying ulceration is seen at the hepatic flexure (Paris classification 0-Is, NICE Type III). B) An initial circumferential incision around the lesion is made for ESD. C) Lesion invasion into the muscularis propria is visualized during submucosal dissection. D) Final view of the resection site showing appropriate position of the clip.

Figure 2. Final pathology of the lesion shows T2 invasive adenocarcinoma with negative (R0) resection margins

Figure 2. Final pathology of the lesion shows T2 invasive adenocarcinoma with negative (R0) resection margins

Introduction
Comorbidities and hostile surgical fields make esophago-bronchial fistula surgery difficult. Mismatch between the distal stent flare and gastric conduit diameter limit fully covered metal stent (FCMS) use in esophago-gastro-bronchial fistulas. We successfully managed a case of esophago-gastro-mediastino-bronchial fistula in a patient with a history of esophagectomy by combining a FCMS, percutaneous endoscopic thoracotomy and a retrograde PEJ-G tube. This management approach changed the disease course from requiring hospice to being back in the community.

Case Presentation
A 69-year-old man with history of esophageal cancer post-Ivor Lewis esophagectomy and chemoradiation, right lung partial pneumonectomy due to post-operative anastomotic leak who developed recurrent aspiration pneumonias and pulmonary edema requiring ICU admissions. These recurrent pneumonias and pulmonary edema over the prior 4 years were secondary to frequent collection of secretions in the persistent post-pneumonectomy cavity at the post-surgical site. He was given the option to go to hospice versus novel endoscopic intervention. He opted for the latter.

Endoscopy revealed a disrupted esophagogastric anastomosis around 15mm, and an esophago-gastro-mediastino-bronchial fistula in the upper third of the esophagus (Figure 1). Endoscopic management involved placing a percutaneous endoscopic thoracotomy tube, double lumen percutaneous endoscopic jejuno-gastric tube with a locking loop drain and FCMS across the defect (Figure 2). Endoscopic suturing approximated esophagogastric anastomosis.

A follow up endoscopy with gastrografin swallow was performed with observation under fluoroscopy (video). The retrograde drain suctioned contrast without spillage into the thoracic cavity. The patient was taught to aspirate from both tubes after drinking thin liquids for comfort.

The patient continues to do well clinically and is seen every three months for tube exchanges for the last 4 years. He has not had any pulmonary infections since the intervention. He manages mucus plugs or secretions in the lungs by connecting the thoracic tube to low intermittent suction, which improves oxygenation. He eats comfort foods by mouth and suctions them out. He maintains his weight through jejunostomy site feeding.

Conclusion
A patient with a complex refractory esophago-gastro-mediastino-bronchial fistula after Ivor-Lewis esophagectomy was treated with stenting, endoscopic thoractotomy and a double lumen PEJ-G tube with locking loop drain. This allowed for feeding via the first lumen of the PEJ tube and suctioning via the retrograde jejunogastric tube to prevent aspiration. Percutaneous thoracotomy tube with PEG bumper was used as a backup to suction any remaining contents that could have reached the post pneumonectomy space.

Figure 1. Demonstration of Communication Through Esophago-Gastric Anastomosis to the Post Pneumonectomy Cavity/ Dead Space

Figure 1. Demonstration of Communication Through Esophago-Gastric Anastomosis to the Post Pneumonectomy Cavity/ Dead Space

Figure 2. (a) Thoracotomy with PEG tube to drain cavity. (b) Retrograde double lumen percutaneous endoscopic jejunogastric tube inserted retrograde into esophagogastric anastomotic site. (c,d) Contrast injected into fully covered metal stent with locking loop drain to suction out any orally consumed material with suction off (c), and suction on (d).

Figure 2. (a) Thoracotomy with PEG tube to drain cavity. (b) Retrograde double lumen percutaneous endoscopic jejunogastric tube inserted retrograde into esophagogastric anastomotic site. (c,d) Contrast injected into fully covered metal stent with locking loop drain to suction out any orally consumed material with suction off (c), and suction on (d).

Introduction: Iron deficiency anemia (IDA) is a well-known complication after Roux-en-Y gastric bypass (RYGB). Many patients nonetheless undergo an evaluation to exclude alternate etiologies. We report a rare cause of gastrointestinal bleeding that was effectively managed with a unique modification of an established endoscopic technique.

Case: A 65-year-old female s/p RYGB was referred for refractory IDA requiring multiple hospital admissions and blood transfusions. Capsule endoscopy revealed heme at 3 hours 30 minutes and a small bowel transit time of 5 hours 45 minutes. Anterograde double balloon enteroscopy (DBE) was performed, revealing a proximal ileal lobular lesion > 5 cm in diameter (1/2 of luminal circumference and involving several folds), with “strawberry-like” mucosa and active oozing (Figure 1). Argon coagulation (APC) was applied to bleeding foci and the lesion margin tattooed. Biopsies were consistent with a hemorrhagic lymphangioma. Cross-sectional imaging confirmed no finding that would mandate surgical intervention, prior to discussion of endoscopic management as a treatment strategy. The decision was then made to proceed with band ligation and auto-amputation as primary therapy for the bleeding lymphangioma. We first determined that the cap from the Speedband Super 7 System (M00542250, BOSTON SCIENTIFIC) was compatible with the enteroscope diameter (EN-580T, FUJIFILM). The 110 cm bander wire was then detached and the “hub end” was attached to the opposite end on a second bander system using a through-the-loop pull method (Figure 2A). The cap was mounted as per usual to the elongated wire, providing the required 200 cm channel length and allowing for the deployment of multiple bands over the expansile lesion. Subsequent DBE enabled five bands to be successfully deployed, capturing as much of the lymphangioma as possible (Figure 2B). No complications were observed. Surveillance DBE noted complete eradication of the lesion, with small scars and scattered insignificant lymphangiectasia along the periphery (Figure 2C). The patient’s anemia, blood transfusion requirements and iron infusion dependency resolved.

Discussion: Small bowel lymphangiomas are rare benign lesions associated with hemorrhage, intussusception, and enteropathy. Endoscopic therapies including mucosal resection for lesions < 2 cm, focal APC, and sclerotherapy have been anecdotally described. This case is unique in that it demonstrates auto-amputation via band ligation can be both safe and effective in treatment of larger hemorrhagic lymphangiomas. In addition, compatibility of standard multi-band devices with the double balloon enteroscope was demonstrated, which may profoundly impact the management of other small bowel lesions.

FIGURE 1. Large hemorrhagic lymphangioma in proximal ileum on double balloon enteroscopy.

FIGURE 1. Large hemorrhagic lymphangioma in proximal ileum on double balloon enteroscopy.

FIGURE 2A. Bander compatibility with the FUJI double balloon enteroscope. 2B. Band ligation of the lymphangioma. 2C. Follow up exam with eradication of the hemorrhagic lesion.

FIGURE 2A. Bander compatibility with the FUJI double balloon enteroscope. 2B. Band ligation of the lymphangioma. 2C. Follow up exam with eradication of the hemorrhagic lesion.

Management of RDFBIs is often quite challenging. It provides a sense of power and control to the patient, attracting attention and care; this potentially triggers repetitive similar behaviors.

A 35-year-old female with multiple prior admissions for recurrent deliberate foreign body ingestions (RDFBI), borderline personality disorder (BPD), post traumatic stress disorder (PTSD), presented with epigastric pain and nausea after ingesting multiple foreign bodies (FB). She has previously undergone innumerable endoscopies and laparotomies for removal of FB. Abdominal radiograph demonstrated FB in the stomach, presumed to be batteries and butter knife. Following endoscopic retrieval, she was admitted for psychiatric evaluation. While in-patient, she ingested batteries from television remote and nasal bridge holders of facemasks despite constant watch, requiring repeat endoscopies. She desired inpatient psychiatric care but multiple psychiatric facilities have declined to provide inpatient therapy; there is concern that this rejection might have contributed to self-injurious behavior. Queries and concerns were raised regarding ongoing urgent endoscopic interventions given that it seemed to offer little utility in ameliorating the underlying psychiatric problem. It was questioned whether repeated admissions and endoscopy could be potentiating her behavior, as it enabled the patient to get the attention she desired. A multidisciplinary team meeting involving Internists, Gastroenterologists, Emergency Room Physicians, Psychiatry, Ethics and Care Management was held. With the goal of ensuring that the needs of patient come first and providing care in empathetic and safe manner, while minimizing the burdens and risks of emergent endoscopies and prolonged hospitalization, an interdisciplinary care protocol was devised. This protocol delineates clear care pathway for when patients with RDFBIs present to our facility and are in need of either emergent or urgent endoscopy, in the presence or absence of suicidal ideation. Should the patient need admission, additional guidance on extra safety measures and stringent environmental risk assessment strategies are incorporated in the protocol. Since the inception of this protocol, frequency of out-of-hours endoscopies and urgent endoscopies for ingestion of FBs while remaining inpatient, have reduced. Although dedicated psychiatric inpatient rehabilitation is a crucial component to disrupt such self-injurious behavior, it unfortunately remains an unmet need in our healthcare system.

As much as physicians should respect patient autonomy and not deny care, it is crucial to recognize self-destructive behavioral patterns, and take measures to break that cycle by adopting a more comprehensive approach focusing on the underlying psychiatric condition, incorporating and maximizing counseling and behavioral therapies.

Abdominal radiograph and upper endoscopy demonstrating ingested metal blade in the stomach

Abdominal radiograph and upper endoscopy demonstrating ingested metal blade in the stomach

Abdominal radiograph and endoscopic images demonstrating ingested batteries and nasal bridge holder of face masks

Abdominal radiograph and endoscopic images demonstrating ingested batteries and nasal bridge holder of face masks

Background & Aims
Complications like esophago/gastro-respiratory track fistulas (EGRFs) are a source of considerable morbidity & mortality due to frequent aspirations & malnutrition. Treatment is challenging due to alternating pressure gradients which limit healing. Surgical management is not always feasible and recurrence after conventional methods like clipping is common. Therefore, novel techniques for management are required. Our aim was to test spiration valve system (SVS) deployment in the management of refractory EGRFs as proof of principle in a series of 5 patients from July 2020 to Feb 2022

Technique
Our representative case is a 59 year old patient with a chronic esophagobronchial fistula who had been treated previously with an over the scope clip which was unsuccessful.
SVS has been used to successfully treat post pneumonectomy air leak in patients with emphysema. This one-way umbrella shaped valve is made up of a Nitinol (nickle-titanium) framework and a polyurethane membrane. Its 6 struts help appose it against the airway and due to its flexible nature, it expands and contracts during breathing.
We used this device in 5 different patients but herein we describe the technique that we have found to be most successful.
Unlike its conventional deployment with a bronchoscope in respiratory tree, to prevent esophageal leak into the airway, its deployment needs an innovative approach to be placed on the esophageal side. Its delivery system is not long enough to pass through a conventional EGD scope. To overcome this limitation, we positioned an EGD scope on the esophageal side & a bronchoscope on the airway side. Under fluoroscopic guidance, a guidewire was passed through the fistula and retrieved using a bronchoscopy forceps. Next, a steerable sheath(TourGuide Steerable Sheath, Medtronic, MN, USA) was introduced over a guidewire. Once its placement across the fistula was confirmed with the distal tip of the sheath in the airway, the inner stylet & guidewire were removed. The SVS deployment catheter containing a loaded valve was then introduced inside the steerable sheath. Under both fluoroscopic & direct visualization using the EGD scope, the sheath followed by the SVS deployment catheter were retracted & the valve deployed with the umbrella struts positioned to occlude the entrance to the fistula on the esophageal side. Adequate occlusion was then confirmed by injecting contrast on the esophageal side under fluoroscopy

Results
Out of the 5 patients treated with SVS, 4(80%) had resolution of defect at median follow up of 164(IQR 113.3-268.8)days. Only 1 patient had an unresolved fistula but reported an improvement in symptoms. Details in table

Conclusion
SVS is an effective way to manage EGRFs especially where conventional methods have either failed or were not feasible. Trials on more cases are required to see the efficacy of this novel approach

Table- Patient demographics, previous interventions and outcomes

Table- Patient demographics, previous interventions and outcomes

Figure 1<br /> a- Fistula leak<br /> b- Leak resloved after placing a SVS

Figure 1
a- Fistula leak
b- Leak resloved after placing a SVS

Introduction: Roux-en-Y gastric bypass is one of the most common bariatric surgeries in the US, with over 40,000 procedures annually. Although rare, anastomotic leaks and fistula formation can occur and are associated with significant morbidity and mortality. We present a case of a Roux-en-Y gastric bypass complicated by severe gastro-jejunostomy stricture causing a completely blind gastric pouch with resultant anastomotic leak and fistula formation treated endoscopically with lumen opposing metal stent and endoscopic suturing.

Case Description: A 58-year-old female presented to an outside facility for revision gastric bypass surgery following antecedent weight gain. POD 7 was complicated by worsening abdominal pain, with computed tomography demonstrating multiple intra-abdominal fluid collections, concerning for anastomotic leak. Endoscopic evaluation demonstrated a strictured gastrojejunal opening that was dilated to 10 mm. She was managed conservatively for several weeks with external drain placements and TPN prior to transfer to our facility. Gastroenterology was consulted and esophagogastroduodenoscopy (EGD) was performed. While EGD demonstrated evidence of gastric pouch, there was no discernable evidence of a gastrojejunosotomy, suggestive of a now completely excluded gastric pouch (Figure 1a). Along the anterior wall of the pouch was a 2 x 2 mm defect. Fluoroscopic evaluation demonstrated forward flow into the peritoneum and down a paracolic gutter consistent with an open perforation (Figure 1b). Given the apparent complete gastric obstruction, the decision was made to attempt gastrogastric fistula creation via endoscopic ultrasound. A 19g EUS needle was inserted into the remnant stomach and aspiration for fluid was performed. Following infusion of 500 cc of normal saline, a 20 x 10 mm lumen opposing metal stent was deployed into the remnant gastric lumen creating a patent gastrogastrostomy (GG). Following GG creation, the decision was made to use endoscopic sutures to repair the perforation. One suture was placed in a running fashion with cinches on both ends across the perforation with excellent tissue approximation. Subsequent upper gastrointestinal follow through demonstrated no evidence of ongoing leak and forward flow of contrast through the recently recreated GG LAMS. Repeat endoscopy one month later demonstrated persistent closure of the previously noted perforation.

Discussion: This case demonstrates a rare complication of gastric bypass surgery, a completely blind gastric pouch resulting in anastomotic leak and fistula formation. While these complications are associated with a high morbidity and mortality, this case demonstrates how novel endoscopic techniques such as lumen opposing metal stent placement and endoscopic suturing can be used successfully to prevent further complications.

Esophageal fully covered self-expanding metal stents (FCSEMS) are used for treatment of malignant and benign strictures, and for esophageal tears or perforation. Stent fixation with endoscopic suturing (OverStitch, Apollo) minimizes migration. Here we describe a previously unreported complication related to entanglement of the needle driver with the stent proximal end during suturing and explore rescue options.
Case 1: A 69-year-old male with esophageal adenocarcinoma s/p chemoradiation had an esophageal stent placed due to dysphagia. He was seen one month later for management of stent migration. The stent was removed easily. The esophageal lumen was narrow and less compliant due to prior radiation so a through-the-scope (TTS) 18mm x 80mm FCSEMS (Taewoong) was deployed. One suture was anchored with the OverStitch device mounted on a double channel therapeutic (2T) upper endoscope. With placement of a second suture, the stent got caught and a diagnostic upper endoscope was used to visualize the stent trapped in the needle driver between the metal loop and the mounting bracket. Neither opening the needle driver nor retraction released the stent. Both scopes and the stent were removed; outside the patient, forceful handle separation opened the needle driver and released the stent. A new stent was placed without issue.
Case 2: A 52-year-old female with oral squamous cell carcinoma s/p resection, chemoradiation, and PEG placement was seen for gastrocutaneous fistula closure after PEG removal. An Over the Scope Clip (OTSC, Ovesco) was used for fistula closure, but with passage of the scope and OTSC, a 7cm longitudinal laceration of the esophageal wall was noted. A TTS 18mm x 140mm FCSEMS (Taewoong) was deployed; two sutures were used for fixation. After completion of the second suture, entanglement of the stent similar to case 1 was noted. During attempts to free the needle driver from the stent, the mounting bracket dislodged. The 2T scope was withdrawn and a diagnostic upper endoscope was used to visualize the trapped stent. APC was applied to the stent, freeing it from the mounting bracket.
These cases highlight a complication not previously reported. In the needle driver “open” position, there is not enough space between the needle and the mounting bracket to allow stent wire to get caught. We suspect that radiation changes reducing esophageal wall compliance and the use of a flexible stent (Taewoong) contribute to this complication.
In one case, forceful opening of the handle once outside the patient resulted in releasing the stent. In the second, APC of the stent wires was used to free the stent. In both cases, a diagnostic upper endoscope was needed to diagnose, and in one case, perform the rescue maneuver. Alternatives to consider are the use of a non-TTS stent or other fixation devices such as TTS suturing (X-Tack, Apollo) or Stentfix (Ovesco).

Figure 1. Diagrams demonstrating the small open space where stent interstices can become entrapped in the mounted OverStitch (Apollo) device with the needle driver in the “open” position (red arrow), most apparent with an oblique view shown in panel B

Figure 1. Diagrams demonstrating the small open space where stent interstices can become entrapped in the mounted OverStitch (Apollo) device with the needle driver in the “open” position (red arrow), most apparent with an oblique view shown in panel B

Figure 2. Endoscopic view of Case 1; stent trapped on needle driver in the “open” position (red arrow)

Figure 2. Endoscopic view of Case 1; stent trapped on needle driver in the “open” position (red arrow)

Background

Pericardial effusion in the context of malignancy is usually related to radiation or chemotherapy. In rare cases, cardiac tamponade can be the first manifestation of esophageal cancer. We present a case in which esophageal squamous cell carcinoma (SCC) created an esophago-pericardial fistula that was treated with an esophageal stent.

Case description

A 60-year-old gentleman presented with progressive dysphagia to solids and liquids for 5 months. He presented for outpatient endoscopic ultrasound for abnormal esophageal thickening found on computed tomography (CT) of his abdomen. He was found hypotensive and tachycardic with complaints suggestive of possible food impaction from the night prior. Given abnormal vitals, patient was admitted. A CT scan revealed a large necrotic mass in the posterior mediastinum and a large pericardial effusion. Point of care ultrasound revealed tamponade physiology and underwent pericardiocentesis with removal of >700 mL of purulent pericardial fluid. Broad antibiotics were started. An upper gastrointestinal (GI) series showed luminal narrowing in the distal esophagus with contrast extravasation into a large cavity along the right lateral distal esophagus (Figure 1). Esophagogastroduodenoscopy (EGD) revealed a large fungating circumferential mass in the distal esophagus at 29 cm with a widely patent fistula to the mediastinum. Under direct endoscopic and fluoroscopic guidance, a 23 mm x 155 mm partially covered stent was placed in the esophagus, covering the fistula opening. A repeat upper GI series post-stent revealed no evidence of contrast extravasation (Figure 2). The patient was able to tolerate a diet afterward. Biopsies of the mass were positive for poorly differentiated SCC of esophageal origin with extensive necrosis. The patient was discharged to hospice with a pericardial drain in place with appropriate follow-up.

Discussion

Esophageal SCC is a rare cause of esophago-pericardial fistula that has a high mordity and mortality. Any communication between the esophageal lumen and pericardium is a nidus for bacterial infection of the pericardial sac. The first step in treatment should be pericardial drainage and antibiotics, followed by closure of the fistula. In our case, the patient presented with atypical symptoms of cardiac tamponade. Diagnosis was confirmed quickly and endoscopic therapy was successful. The patient tolerated the procedure well, without stent migration, and was able to tolerate oral nutrition within 3 days of the stent placement. We hope to add to the literature in which stent placement is an effective therapeutic measure for esophageal fistula in the setting of malignancy and educate GI physicians on the potential for cardiac tamponade in these cases.

Figure 1: Active extravasation of contrast into pericardial space

Figure 1: Active extravasation of contrast into pericardial space

Figure 2: Post esophageal stent placement without extravasation of contrast

Figure 2: Post esophageal stent placement without extravasation of contrast

Brian Lee

Appearances