Over-the-Scope-Clip Iatrogenic Ligation of the Gastroduodenal Artery: A Case Report

Introduction

The Ovesco Over-the-Scope-Clip® System (OTSC®, OVESCO Endoscopy AG, Tübingen, Germany) is a recently-introduced class of endoscopic clips intended to provide improved strength and tissue capture compared to conventional through-the-scope (TTS) endoscopic clips in the management of refractory gastrointestinal (GI) bleeding, perforations, anastomotic leaks, and fistulas. The OTSC was approved for use in Europe in 2009 and in the United States in 2010, and has 3 different sizes and 3 different clip designs, each of which can be mounted via an application cap at the distal end of the endoscope. Installing the OTSC on the endoscope is simple and is similar to attaching the band li-gator used for treating varices. The OTSC application cap depth can be adjusted to 3 or 6 mm depending on the clinical need. In general, the 6-mm cap length is ideal for grasping larger tissue; for example, when closing luminal defects/perforations. The 3-mm cap length has less tissue grasp, and is best suited for most other applications, like controlling GI bleeding [1]. Prior to the availability of OTSCs, complex gastrointestinal luminal defects and perforations were difficult to close with conventional TTS clips due to their inability to approximate the defect margins, necessitating surgical intervention. Additionally, at times, GI bleeds could not be adequately controlled with conventional endoscopic interventions, including TTS clips, resulting in the need for surgical or interventional radiology interventions [2]. Although the OTSC has been used to treat GI bleeding and bowel perforations for over a decade.[2], it has been relegated to second-line therapy, and it has only recently become considered a first-line hemostatic therapy for GI bleeding [3].

A review of the OTSC system found it to be a “safe and effective endoscopic therapy to manage gastrointestinal hemorrhage, perforations, anastomotic leaks, and defects created by endoscopic resections and for stent fixation.” This review reported an overall adverse event rate of 2.1% [4]. The most common reported adverse events were related to incorrect placement, defined as placement of the OTSC in a location other than the targeted defect and/or the disruption of nearby tissue/structures by the OTSC (27.5%), followed by perforation, bleeding, and infection. Most adverse events of the OTSC were related to poor precision of deployment [1]. Examples of incorrect placement included clip deployment on the tongue and endoscopist’s finger. Due to the powerful grasping ability of the OTSC, there has been documented grasping and disruption of nearby extraluminal tissues (eg, bowel, ureter) [1,4]. While the overall complication rate of OTSCs is generally low, the most common adverse event is inadvertent capture of extraluminal or adjacent tissue or structures, which can result in serious complications requiring the need for surgery or other interventions [1]. The potential complications related to the application of the OTSC therefore warrant very careful consideration of the surrounding anatomy prior to its deployment.

Case Report

We present a case of an OTSC complication involving disruption of nearby tissue/structures. A 71-year-old man presented with 6 months of epigastric abdominal pain and 2 weeks of hematemesis. He had multiple comorbidities, including alcohol use disorder, hypertension, atrial fibrillation only on aspirin and not on other anticoagulation, chronic back pain, long-term tobacco use, and chronic obstructive pulmonary disease. His home medications included daily aspirin, amlodipine, hydrochlorothiazide, an albuterol inhaler, and diclofenac, which he used twice a day for back pain. He denied any additional over-the-counter medication use and reported consuming up to 4 alcoholic drinks per day. Upon presentation, his vital signs revealed a blood pressure of 103/49 with a pulse of 147 beats per minute. His respiratory rate was 23 and his pulse oximetry was 97% on room air. Physical examination revealed a thin male in no apparent distress. His mucus membranes were dry, his oropharynx was clear with no sublingual jaundice, and his sclera were anicteric. His lungs were clear to auscultation, and his heart sounds revealed tachycardia. His abdomen was soft and nondistended with normal bowel sounds and mild epi-gastric tenderness to palpation. His skin was warm and dry with no yellow hue. His neurologic examination was normal. Rectal examination revealed black stool. Laboratory analysis revealed an initial hemoglobin level of 5.5 g/dL with a normal white blood cell count and a platelet count of 262×103/uL. His international normalized ratio was 1.5 with an elevated prothrombin time of 17.4 seconds. Basic electrolytes were normal, creatinine was 1.37 mg/dL, and the blood urea nitrogen level was 53 mg/dL. Liver enzymes showed a total bilirubin of 0.6 mg/dL, albumin of 2.7 g/dL, total protein of 4.7 g/dL, alkaline phosphatase of 102 U/L, aspartate aminotransferase of 52 U/L and an alanine aminotransferase of 31 U.L. Lipase level was normal at 18 U/L. A computed tomography scan (CT scan) of the abdomen and pelvis revealed no evidence of free fluid, but significant stranding near the duodenal bulb with thickening of the duodenal bulb. The pancreas had signs of chronic atrophy and there were no varices noted. The patient was resuscitated with intravenous fluids and packed red blood cells, started on an intravenous proton pump inhibitor, and then taken for an upper endoscopy. This initial endoscopy revealed esophagitis, and an oozing duodenal ulcer with 2 areas of visible vessel. The ulcer was treated with hemostatic spray, which controlled the bleeding. Hemoglobin increased to 8 g/dL and remained stable for 48 hours. The patient then had recurrent hematochezia with associated hypotension and drop in hemoglobin to 7.1 g/dL. He received further packed red blood cells and a repeat upper endoscopy revealed an oozing 10 mm cratered ulcer with an adherent clot at the first portion of the duodenum. Due to poor visualization from bleeding, the ulcer was again treated with hemostatic spray, which slowed the bleeding but did not control it. Due to continued upper GI bleeding, the patient was transferred to our tertiary care center for further management. On arrival to the tertiary facility, his blood pressure was 111/58 with a pulse of 99 beats per minute. Laboratory data revealed a hemoglobin of 6.2 g/dL and the remainder of his laboratory data were overall unchanged from his admission laboratory data. He was transfused with further packed red blood cells and an urgent upper endoscopy was performed. The endoscopy was significant for a large 5 cm circumferential ulcer at the duodenal bulb with a protuberant vessel (Figure 1). A single OTSC was deployed on the visible vessel (Figure 2). He remained hemodynamically stable overnight, with a rise in hemoglobin to 8.7 g/dL. However, the next day, he developed recurrent hematochezia with a drop in hemoglobin to 7.3 g/dL and associated hypotension. Given that he had undergone multiple endoscopic interventions, with persistent bleeding, Interventional Radiology was consulted for endovascular management of the bleeding. Angiography showed the OTSC had been deployed across the midportion of the gastroduodenal artery (GDA) from the duodenal lumen, causing a direct vascular injury and leaving a short stump at the origin of the GDA, preventing advancement of catheters and wires into the distal GDA where the suspected bleeding was occurring (Figures 3, 4).

With no way to reach the source of bleeding from an ante-grade approach, the GDA was accessed from a retrograde approach via the superior mesenteric artery. With a microcatheter positioned in the distal GDA, immediately inferior to the OTSC, the vessel and associated inferior pancreaticoduodenal branches were embolized with detachable AZUR CX Peripheral

Microcoils (Terumo, Somerset, NJ) 3–6 mm in diameter. After this was completed, additional coils were used to embolize the GDA stump superior to the OTSC in a “sandwich technique” [5] to successfully occlude all blood flow around the OTSC. Post-embolization angiography revealed no residual flow into the GDA from the celiac or superior mesenteric artery (Figure 5). Following the procedure, he was monitored for any further complications related to his endoscopic and endovascular interventions. His hemoglobin stabilized quickly after the endovascular intervention and the remainder of his laboratory test results, including lipase, remained normal throughout his hospitalization. He was ultimately discharged from the hospital without further complications. A CT scan was obtained 8 weeks later for lower quadrant abdominal pain, with findings of stable chronic atrophy of the pancreas, and no post-embolization complications such as bowel ischemia, residual bleeding, or pseudoaneurysm. A large stool burden was noted and thought to be the source of the lower quadrant abdominal pain. No other acute abnormalities were noted.

Discussion

The most recent American Gastroenterology Association (AGA) clinical practice update on endoscopic therapies for non-variceal upper gastrointestinal bleeding (NVUGIB) was published in 2020 [6]. The update was an expert review of available endoscopic modalities, including traditional mechanical therapies (eg, TTS clips), thermal therapies (eg, hemostatic forceps, heater probes, and bipolar/monopolar catheters), and diluted epinephrine, as well as treatment positioning for the newer modalities (eg, hemostatic powders and OTSCs). The update notes that there have been no superiority trials comparing the ability to control NVUGIB with traditional mechanical therapy versus thermal therapy, so best practice is to attempt to control NVUGIB at the endoscopist’s discretion based on the clinical scenario. It is, however, recommended to avoid epinephrine monotherapy, as there is a high risk of rebleeding in that situation. Epinephrine, if needed, should be relegated to only providing initial hemostasis to allow for further intervention with mechanical or thermal therapies. Dual therapy with epinephrine and either mechanical or thermal therapy is not superior to properly applied mechanical or thermal therapy alone. All lesions bleeding at time of endoscopy, as well as nonbleeding lesions at high risk for rebleeding, such as visible vessels, should be treated. Hemostatic powders should be considered only as salvage therapy in cases of active bleeding with poor visualization and inability to apply mechanical or thermal therapies, or as definitive therapy for bleeding malignancies. Finally, OTSCs should be considered second-line therapy in NVUGIB that is refractory to previously attempted traditional mechanical or thermal therapies but can also be considered as first-line therapy in clinical scenarios where the endoscopist predicts traditional therapies will be ineffective. The American College of Gastroenterology (ACG) released their society’s updated guidelines for NVUGIB in 2021, which essentially mirrored the AGA recommendations, to include the recommendation for OTSCs be used in patients with recurrent GI bleeding after previous successful endoscopic therapy [7].

The OTSC is generally considered a second-line therapy when traditional therapy has proven ineffective at controlling GI bleeding but is increasingly being utilized as first-line therapy. Although studies are ongoing to determine its efficacy as a primary intervention for GI bleeding, it has been described as an effective first-line therapy option for patients on anticoagulation, poor surgical candidates, or for patients with large bleeding visible vessels not amenable to traditional endoscopic therapy. The AGA noted that since there are few randomized controlled trials (RTC) comparing OTSCs to TTS clips for primary GI bleeding, they currently cannot recommend a shift to push the OTSC to first-line therapy [6]. There have been no RCTs comparing first-line use of OTSCs to thermal therapy. The few RCTs that do exist comparing OTSCs to TTS clips as first-line therapy have differing results. One large RCT found that use of OTSC as primary treatment for GI bleeding was not associated with a significant decrease in 30-day rebleeding rates when compared with traditional TTS clips [8], while 2 others concluded that OTSCs were superior to TTS clips when used as first-line therapy to prevent rebleeding [9,10]. Both the AGA and the ACG recommendations recognize that RCTs are ongoing evaluating the positioning of OTSCs as first-line therapy for GI bleeding, which could change future AGA and ACG recommendations; however, the societies suggest certain clinical scenarios may justify positioning OTSCs as first-line therapy for GI bleeding even now [6,7]. Those situations would include large ulcer beds with large visible vessels that likely would not be amenable to standard mechanical or thermal therapies [6].

Our patient had failed 2 attempts at controlling the GI bleeding, and the protuberant visible vessel within the large duodenal ulcer was ultimately treated with an OTSC. Although the clip was able to control the bleeding ulcer, the iatrogenic extraluminal vascular injury to the GDA caused by the intraluminal OTSC deployment then caused further bleeding. Additionally, the OTSC’s extraluminal ligation of the midportion of the GDA increased the complexity of performing endovascular embolization, for which surgery may have been the only other alternative.

The iatrogenic injury to our patient’s GDA by the OTSC is noteworthy because injury to the GDA, either natural or iatrogenic, can result in significant morbidity and mortality. For example, it is well established that pseudoaneurysm rupture of the GDA can be rapidly fatal, with a mortality rate of 40%. Such bleeding can be missed endoscopically [11]. Coil embolization to the GDA is generally very safe. However, it has rarely resulted in cases of severe ischemic and hemorrhagic gastrointestinal necrosis, and even more rarely, severe ischemia to the head of the pancreas leading to ischemic pancreatitis [12]. While these complications fortunately did not occur in this case related to the iatrogenic extraluminal injury to the GDA by the OTSC deployment in the duodenal lumen, it highlights that extraluminal structures, like vascular structures and even other organs, beyond the gastrointestinal lumen must be considered prior to deployment of OTSCs.

Conclusions

While the OTSC system is generally considered safe and effective, it has been documented to rarely have complications, some of which are related to injury to nearby tissue or structures [1,4], which occurred in our case. The potential complications related to the application of the OTSC therefore warrant careful consideration of surrounding anatomy prior to its deployment. As this device becomes more commonly used in the endoscopy suite, it is important to share the variety of applications as well as potential pitfalls that may be encountered in the clinical setting that impact not only endoscopists and patients, but other specialties as well.