Endoscopic vacuum therapy for nonhealing cavity with fistulous tract after peroral endoscopic myotomy for Zenker’s diverticulum

Video

Video 1

Endoscopic vacuum therapy for nonhealing cavity with fistulous tract after peroral endoscopic myotomy for Zenker diverticulum.

Introduction

Several recent studies have evaluated a cost-effective modified endoscopic vacuum therapy (EVT) for GI transmural defects; this method appears to be safe with high clinical and technical success rates. Its advantages over commercially available EVT systems include its low cost and easy insertion, the longer interval between EVT system exchanges, and reduced tissue ingrowth.^1,2

Case

A 72-year-old woman with a history of obstructive sleep apnea managed with continuous positive airway pressure (CPAP), and GERD presented with dysphagia primarily for solids, globus sensation, and intermittent cough over the past year. The initial video fluoroscopic swallow study (VFSS) revealed a Zenker’s diverticulum (2.5 $\times$ 1.5 cm), for which she underwent a successful peroral endoscopic myotomy (Z-POEM) procedure. However, a postprocedural esophagogram taken on the following day showed extraluminal contrast, indicative of a perforation. A CT scan was also performed on day 1 that showed pneumomediastinum, likely related to the recent Z-POEM and CPAP use for obstructive sleep apnea. Although the gastroenterology team advised discontinuing CPAP use, the patient relied on it because of a nighttime drop in oxygen saturation after the procedure. The patient was asymptomatic and hemodynamically stable and was managed conservatively with prophylactic antibiotics (ampicillin-sulbactam 1.5 g, intravenous [IV], 4 times per day). The use of nonsteroidal anti-inflammatory drugs was withheld, and in case of pain, IV acetaminophen or morphine was used. Given the stable clinical status, attempts were made to gradually advance the diet from nothing by mouth (NPO) to soft solids. Following multiple failed attempts to advance the diet, a follow-up VFSS was conducted, which showed an air-filled fistula at the level of C4-6. A subsequent upper endoscopy demonstrated nonclosure of the mucosal incision at the site of the prior Z-POEM, with a poorly healing cavity notable for exposed adventitia, friable granulation tissue, and a visible fistulous tract opening at the base of the cavity (Fig. 1). An attempt at primary closure was made via a 12-mm over-the-scope clip (OTSC) (Ovesco, Tübingen, Germany). OTSC was preferred, as other tools, including X-tack (Apollo Endosurgery, Austin, Tex, USA), are hard to use in narrow lumens. Following this, the patient was kept NPO and received central parenteral nutrition via a peripherally inserted central catheter line, and CPAP was discontinued to prevent worsening of the fistulous tract. However, a subsequent esophagogram showed extraluminal contrast at the level of the newly placed clip, suggestive of persistent fistula, which led to the decision to perform EVT.

Figure 1.

Endoscopic images. A, Nonclosure of the mucosal incision at the site of the prior peroral endoscopic myotomy for Zenker’s diverticulum, with a poorly healing cavity and a visible fistulous tract opening at the base of the cavity. B, Complete closure of the fistulous tract and the absence of necrotic tissue.

Procedure

The modified EVT used in this case involved wrapping a gauze around the fenestrated part of a 16F nasogastric tube, securing it with 2.0 sofsilk suture. An antimicrobial incise drape was used to cover the gauze, and a 16-gauge needle was used to make multiple punctures in the tube. The assembly was connected to a wall suction to assess its functionality. During endoscopy, a persistent poorly healing transmural cavity was observed in the proximal esophagus. The OTSC was fragmented and removed, and the EVT was inserted through the nares, advanced under endoscopic guidance to the level of the cavity, and was positioned alongside the defect; it was not pushed into the fistulous tract (Video 1, available online at www.videogie.org). The device was then connected to wall suction at −150 mm Hg, with endoscopic confirmation of compression of the esophagus and adherence against the cavity. The EVT tubing was then secured at the nares, and the device was maintained on continuous suction at −150 mm Hg. Based on the literature, the optimal suction pressure used in EVT is 125 to 175 mm Hg at continuous moderate intensity. The selection of the suction pressure is dependent on the size of defect, the depth, presence of fluid collections, and the patient’s tolerance. Although the location of GI defect is a factor to consider, it is not the sole determinant of the appropriate suction pressure.³ In our case, the EVT was well tolerated by the patient. Although she felt uncomfortable in her throat and had throat pain, this was managed by IV acetaminophen or morphine.

Outcome

The patient was maintained NPO and received central parenteral nutrition via a peripherally inserted central catheter line; the EVT remained attached to the continuous wall suction set at −150 mm Hg. After 4 days, an endoscopy was performed to observe the progress of healing, and based on the observations, it was decided to replace the EVT at that same time. A week later, the patient underwent another endoscopy, which showed complete closure of the fistulous tract and the absence of necrotic tissue (Fig. 1). Therefore, the EVT was removed. A follow-up VFSS showed no fluoroscopic evidence of esophageal leak or fistula. Soon after, the patient tolerated peroral intake and was discharged.

Conclusion

Z-POEM is a safe and effective endoscopic technique for Zenker’s diverticulum management. However, adverse events like transmural defects may occur.⁴ In this case, we showcased the novel application of EVT for successfully managing a post–Z-POEM transmural defect refractory to primary closure. We believe that modified EVT could be used as first-line treatment for fistula management after Z-POEM. Also, ensuring that no positive pressure is used on the defect after the procedure is of utmost importance.

Disclosure

Dr Khashab is a consultant for Boston Scientific, Medtronic, Olympus, GI Supply, Pentax, and Apollo Endosurgery and receives royalties from UpToDate and Elsevier. All other authors disclosed no financial relationships relevant to this publication.

Supplementary data

Video 1

Endoscopic vacuum therapy for nonhealing cavity with fistulous tract after peroral endoscopic myotomy for Zenker diverticulum.