Simultaneous Endoluminal Wound Vacuum and Stent Therapy for Large Esophageal Leaks After Esophagectomy

Abstract

Background

Anastomotic leaks from esophageal operations are serious life-threatening medical conditions that traditionally require invasive surgery. In this study, we discuss the outcomes of 3 patients who received minimally invasive simultaneous endoluminal wound vacuum therapy and an esophageal stent (sEV+S) treatment for uncontained esophageal leaks into the pleural space after esophagectomy.

Methods

Three patients received sEV+S placement between 2016 and 2024. In this technique, a black sponge sutured to the end of a drain is placed endoscopically into the leak cavity. An esophageal stent is then centered around the leak site.

Results

Patients were 54 to 71 years old. All 3 patients were men with esophageal cancer who underwent esophagectomy. The sEV+S placement was performed 2, 10, and 12 times. The total number of endoscopic procedures after the discovery of the leak was 3, 29, and 13. Length of time from leak to resolution was 16, 102, and 88 days. Hospital stay was 46, 113, and 105 days. All patients eventually returned to an oral diet (2 patients at discharge and 1 patient 23 days after discharge). A tracheoesophageal fistula developed in 1 patient requiring esophageal diversion with intercostal muscle flap repair, followed by eventual supercharged jejunal interposition graft reconstruction. Two patients were successfully treated with sEV+S placement. No patients required readmission, and no deaths occurred due to the procedures.

Conclusions

We demonstrate that sEV+S treatment can successfully treat large uncontained esophageal leaks after esophagectomy.

Visual Abstract

In Short.

• ▪A novel minimally invasive technique for treating large esophageal anastomotic leaks following esophagectomy is described.
• ▪Outcomes of the first patients who underwent simultaneous endoluminal wound vacuum and stent placement are presented.

Esophageal leaks after Ivor Lewis esophagectomy have been reported to be as high as 30% and can lead to sepsis, mediastinitis, and death if left untreated.¹ These leaks can be difficult to repair because the esophagus lacks serosa and a robust blood supply, making it prone to insufficient wound healing.² Currently, there are several treatment options for patients with esophageal anastomotic leaks. Initially, conservative management, including wide drainage, prolonged nothing by mouth status, and antibiotics/antifungals can be attempted or performed adjunctively if the patient is stable.

A first-line treatment option is endoscopic stents, which are effective in treating acute but not chronic leaks.³ Another possible option is T-tube placement, which has a low mortality rate but usually requires a prolonged management course because the esophagus is slow to seal the defect on its own.² A third option is endoscopic clips; however, they are mostly useful only for small and uncontaminated leaks. Endoscopic clips can also be used with fibrin sealants to enhance their efficacy.⁴

Larger perforations generally require reoperation with open or minimally invasive surgery to repair the esophagus with muscle flap or esophageal diversion.⁵ Esophageal repair with a muscle flap may fail, leading to eventual esophagostomy. These methods are often complex and highly morbid and can only be tolerated by healthier patients. Reversal of esophagostomy requires a reoperation with complex reconstruction using colon or jejunum. More recently, endoscopic vacuum-assisted wound closure has been demonstrated as an effective minimally invasive method to close esophageal leaks.⁶

Vacuum-assisted closure is commonly used to treat soft-tissue wounds. This technique removes fluid from the wound, reduces inflammation, reapproximates the wound edges, and stimulates granulation tissue to form.⁷ Endoluminal vacuum-assisted wound closure is emerging as an effective method to manage gastrointestinal leaks, and here we investigated its use alongside simultaneous stent placement. In addition, there is no single standardized technique for performing the endoluminal wound vacuum (EWV) procedure.⁸ In this report, we describe the methods and outcomes for 3 patients after simultaneous EWV treatment and stent (sEV+S) placement for large leaks after esophagectomy.

Material and Methods

Ethical Statement

The study was approved by the Houston Methodist Research Institute Institutional Review Board (approval number: PR000031398), and individual consent for this retrospective analysis was waived.

Procedures

A retrospective review was conducted of 3 patients with large uncontained esophageal leaks after esophagectomy for esophageal cancer (Figure 1A) between December 2016 and December 2024 to investigate the safety and efficacy of sEV+S placement. Large esophageal leaks were defined as radiographic evidence of contrast within a substantial pleural cavity on esophagram, chest roentgenogram, and computed tomography (CT) scan.

Figure 1.

Endoscopic and fluoroscopic demonstration of the technique. (A) Computed tomographic scan demonstrates a large leak. (B) Endoscopic view of a partially treated large esophageal leak showing leak site and true lumen. (C) Fluoroscopy image of completed endoluminal wound vacuum and simultaneous esophageal stent. (D) Schematic of endoluminal wound vacuum and stent in place.

EWV therapy requires a device assembled from a drain, wound vacuum sponge, sutures, and tape. Once assembled, the EWV device is guided through the mouth and placed at the leak site within the esophagus using esophagoscopy and fluoroscopy. Then, the previous EWV is removed, and the leak site is evaluated using esophagogastroduodenoscopy (EGD) view (Figure 1B).

A 16F nasogastric tube (dual-lumen stomach tube) or 19F Blake drain is placed into the nostril and brought out through the mouth. A black wound vacuum sponge is cut into a wedge shape and sutured in place onto the tube. A nylon 2-0 loop is placed at the tip of the exposed end of the tube to assist with positioning the EWV drain. EGD with rat tooth forceps is used to snag the end of the secured nylon loop. Then, the EGD scope is used to intubate the cervical esophagus, delivering the EWV drain into the esophagus and directing it into the leak cavity. Next, a guidewire is placed into the distal gastric conduit under fluoroscopic guidance and a fully covered 23-mm × 105-mm or 23-mm × 155-mm esophageal stent is placed under fluoroscopic guidance and centered around the perforation site (Figure 1C).

Suction is applied with a portable wound vacuum machine or using wall suction at −100 mm Hg. The wound vacuum device and esophageal stent are exchanged multiple times until the healing is complete or failed. We prefer to replace the stent and EWV weekly; however, we are still trying to determine the best approach. After initial placement of EWV drain, we recommend CT chest without contrast to ensure the EWV drain is not near the airway or major vasculature. The full stent and EWV setup is depicted in the schematic (Figure 1D). Endoscopies and esophagrams are performed during each procedure in the operating room.

Results

Each patient’s demographics, esophageal cancer staging, neoadjuvant treatment, diet, and relevant comorbidities are listed in Table 1. Patients were an average age of 64 years (range, 54-71 years) at the time of esophagectomy. All patients were men and had cardiac, respiratory, and renal comorbidities. One patient had liver fibrosis and was a renal transplant recipient on long-term immunosuppression. The average time from esophagectomy to identification of the leak was 12.7 days. The major events timeline for each patient appears in Figure 2. Patient safety and efficacy outcomes after EWV and stent therapy are summarized in Table 2.

Table 1.

Patient Demographics, Esophageal Cancer Staging, Comorbidities, and Diet

Variables	Patient
	1	2	3
Sex	Male	Male	Male
Age at procedure, y	71	67	54
Esophageal cancer staging	Stage 1C (T2 N0 M0) moderately differentiated adenocarcinoma of distal esophagus, GEJ, and proximal stomach	Stage 1B (T1a N0 M0) moderately differentiated adenocarcinoma of GEJ	Stage 3A (T1b N1 M0) moderately differentiated adenocarcinoma of GEJ
Neoadjuvant regimen	None	None	3 cycles (fluorouracil, oxaliplatin, docetaxel)
Comorbidities	AFib, HLD, HTN, anemia, asthma, COPD, obesity-sleep apnea, type 2 DM, chronic benzodiazepine & opioid use	HTN, liver fibrosis, renal failure status after transplantation, immunosuppression, substance abuse	Aortic aneurysm, AVR, bicuspid AV, CHF, HLD, HTN, pacemaker, obesity-sleep apnea, renal cancer (nephrectomy), stage II CKD, testicular cancer (orchiectomy), RA, intellectual development disorder
Pre-sEV+S diet	Oral (full liquids) and J tube	J tube	J tube
Diet at time of leak	Nothing by mouth	Nothing by mouth	Nothing by mouth
Post-sEV+S diet (immediate)	J tube	J tube	J tube
Possible contributing factors to large leak	Advanced age, frailty, prior fundoplication	Immunosuppression, esophageal varices	Prior nephrectomy, morbid obesity, intra-abdominal adhesions

AFib, atrial fibrillation; AV, aortic valve; AVR, aortic valve replacement; CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; CKD, chronic kidney disease; DM, diabetes mellitus; GEJ, gastroesophageal junction; HLD, hyperlipidemia; HTN, hypertension; J tube, jejunostomy tube; RA, rheumatoid arthritis; sEV+S, simultaneous endoluminal wound vacuum and stent.

Figure 2.

Timeline of events for (A) patient 1, (B) patient 2, and (C) patient 3. (EWV, endoluminal wound vacuum; sEV+S, simultaneous endoluminal wound vacuum and stent.)

Table 2.

Safety and Efficacy Outcomes of Endoluminal Would Vacuum and Stent Therapy

Variable	Patient
	1	2	3
Total days of sEV+S therapy (# of exchanges)	4 (2)	33 (10)	54 (12)
EWV single-therapy exchanges (pre-sEV+S, post-sEV+S)	0	19 (18, 1)	1 (1, 0)
Stent single-therapy exchanges (pre-sEV+S, post-sEV+S)	1 (0, 1)	0	1 (0, 1)
Time from leak identification to resolution, d	16	102	88
Conversions to open procedure	Yes	No	No
Return to oral diet at discharge	No	Yes	Yes
Time to oral diet after discharge, d	23	…	…
Hospital length of stay
Total, d	46	113	105
After last stent was removed, d	19	7	1
Vital status at last follow-up	Alive	Alive	Alive
Discharge location	Long-term acute care center	Home	Skilled nursing facility
Readmission	No	No	No
Complications	Yesa	No	No
Mortality	Yesb	No	No

EWV, endoluminal wound vacuum; sEV+S, simultaneous endoluminal wound vacuum and stent.

^aA tracheoesophageal fistula developed.

^bPatient died 2 years later of progression of metastatic disease (unrelated to esophageal leak or sEV+S therapy).

No patients in the study experienced bleeding, disseminated infection, or mortality related to the procedure. A tracheoesophageal fistula developed in the first patient that may have been secondary to the EWV. He required a right thoracotomy to repair the tracheoesophageal fistula with an intercostal muscle flap, followed by an esophageal diversion and gastric conduit resection. This patient later underwent a supercharged jejunal interposition graft to restore continuity of the digestive tract. He eventually died 2 years later of other comorbidities.

Although a tracheoesophageal fistula developed in 1 patient, 2 patients had appropriate healing of the leak with healthy granulation tissue. Patients underwent an average of 14 EWV exchanges and an average of 8.3 stent exchanges. Two patients returned to oral intake at discharge, and the remaining patient tolerated an oral diet within 1 month after discharge.

Comment

Large esophageal leaks into the pleural space after Ivor Lewis esophagectomy are challenging to manage, leaving patients with limited minimally invasive treatment options. In addition, many patients with esophageal leaks are poor candidates for surgery given their baseline comorbidities and acute sepsis from the leak. Risk factors for leaks after esophagectomy include anastomotic location, cardiovascular and renal comorbidities, inadequate perfusion of the gastric conduit, tobacco use, steroids, malnutrition, obesity, lymph node dissection status, and serum albumin levels.⁹ This was consistent in our cohort of patients, who had cardiovascular disease, kidney disease, and obesity-related sleep apnea.

Invasive surgery may be effective in treating some patients with esophageal leaks; however, the high probability of esophageal discontinuity and the subsequent need for reconstruction are major drawbacks for operative management. Although EWV therapy and stent placement alone are effective minimally invasive options for management, their effectiveness is limited by the size of the esophageal defect. This was demonstrated in the case of patient 2, whom we were initially unsuccessful in managing with EWV alone. We believe that the size of the defect hindered the EWV from generating adequate negative pressure and effectively preventing contamination of the leak cavity.

By adding the esophageal stent, a seal was created for the EWV within the leak cavity and thus allowed for the desired negative pressure for effective therapy. The stent essentially acts as the adhesive drape used in negative pressure wound therapy for soft-tissue wounds. The stent additionally diverted saliva and gastric reflux away from the defect and leak cavity, which prevented further contamination while the EWV provided source control. We believe these patients could tolerate oral intake earlier with the stent in place; however, our preference was to keep the patients nothing by mouth to decrease risk of aspiration.

Slight modifications were made to the described technique for each patient depending on the location, size, and healing of the esophageal leak. Our first attempt at sEV+S placement resulted in a complication with a fistula to the airway. However, the use of this technique provided temporary control of the esophageal leak and resulted in resolution of sepsis, allowing the patient to eventually become a candidate for invasive surgery. After our experience with the first patient, we now require CT imaging after index sEV+S placement to ensure that neither the drain nor stent abuts major mediastinal structures such as the airway or aorta.

Our experience indicates that sEV+S placement can be effective and may be attempted as a potential rescue procedure. We propose a treatment algorithm starting with a stent alone and then considering EWV or sEV+S for patients with persistent, large esophageal leaks. Although this is our early experience with a novel technique of combining EWV and stent therapy, more investigation is needed to further define its indications, identify ideal candidates for therapy, and compare outcomes with other management strategies.