Endoluminal vacuum therapy (EVAC/EVT)

INTRODUCTION — The management of leaks and perforations of the gastrointestinal (GI) tract has been a burden for surgeons. This is especially difficult in the esophagus and rectum because of high morbidity and mortality rates. (See "Overview of gastrointestinal tract perforation".)

In the past, these cases have been managed primarily by surgical intervention; however, the evolution of alternative therapies such as endoscopic stent placement provides a less invasive method for control. Unfortunately, these less invasive therapies do not always succeed. As a result, other minimally invasive options to control leaks and perforations have been developed.

Negative pressure wound therapy has long been a treatment of choice for complex external wounds [1]. As the need for further options to control GI leaks has increased, it has been adapted for internal use. This has led to the development of endoluminal vacuum therapy (EVAC/EVT), which can be utilized to both avoid damage control surgeries as well as reduce morbidity and mortality rates. EVAC/EVT is still in its infancy, but the progress for its utilization is growing around the world. (See "Negative pressure wound therapy".)

This topic will cover the patient selection, technique, postoperative care, and outcomes of EVAC/EVT.

DEVICE AND PLACEMENT — EVAC/EVT delivers negative pressure to a leak site in the gastrointestinal (GI) tract through a nasogastric tube (NGT) terminating at a polyurethane sponge, termed endosponge (picture 1). The endosponge has the ability to collapse the cavity and create a seal for obtaining source control. The endosponge is delivered to the leak site endoscopically and placed in one of two ways:

●Intracavitary – The endosponge is placed in the leak cavity, outside of the mucosa-lined space. This will maximize contact between the endosponge and surround tissues that can granulate, and in turn, maximize the benefit of the negative pressure therapy.

●Intraluminal – The endosponge is placed within the GI tract on the mucosa and completely covers the leak site. The endosponge will seal off the leak by adhering to the adjacent mucosa.

In some instances, we have enlarged the opening to deliver the endosponge into the cavity. The inherent effect of the endosponge on nonepithelized tissue is where the granulation tissue arises. Intraluminal placement will only adhere the endosponge to the mucosa to obtain source control. Second intention healing does not occur at the endosponge/epithelial interface.

The pioneers of EVAC/EVT used a homemade device constructed from readily available GranuFoam dressing connected to a source of negative pressure (eg, Redovac) by an NGT. A manufactured device named endo-SPONGE is available in part of Europe [2-4].

EVAC/EVT should be performed by an endoscopist with extensive interventional endoscopic experience. This can be either a surgeon or gastroenterologist. One study demonstrated an average time for the procedure at 43 minutes and that proficiency for the procedure plateaued with respect to time at 10 procedures [5]. The same study indicated that procedures can be done in the operating room or the endoscopy suite, but the endoscopy suite cost was much lower. However, with critically ill patients, the operating venue should be chosen based on optimal patient care and safety rather than cost.

MECHANISM OF ACTION — EVAC/EVT has multiple benefits to aid in the healing of leaks and perforations of the gastrointestinal (GI) tract [6]:

●First, it is able to obtain and maintain source control. Intracavitary or intraluminal placement of the endosponge at a leak site allows for collapse and containment once negative pressure is applied.

●Second, adjacent tissue is pulled into the pores of the endosponge during active therapy. When the endosponge is removed, this causes debridement and eventual decontamination of the affected tissue.

●Lastly, use of EVAC/EVT encourages tissue reperfusion, encouraging eventual healing of a leak. Over time, the goal of therapy is for the cavity to contract and to reestablish continuity of the GI tract.

PATIENT SELECTION — Although in theory EVAC/EVT can be performed for any gastrointestinal (GI) leak that can be reached endoscopically barring a few technical limitations, in practice several clinical factors, as well as the experience of the provider, will determine its suitability.

Indications

●Ongoing leak in the GI tract. Leaks most commonly treated with EVAC/EVT are postoperative anastomotic leaks of the esophagus, stomach, colon, and rectum (eg, status post sleeve gastrectomy, esophagectomy, or low anterior resection). Leaks or perforations of other etiologies (eg, iatrogenic or Boerhaave tear) are also amenable to EVAC/EVT.

●Leak site accessible by an endoscope.

●For intracavitary placement of the endosponge, the leak site must be able to accommodate the bulk of the endosponge to be placed. Therefore, these sites need to be at least 1 cm in size to properly place the endosponge outside the lumen. (See 'Step 3: Endoscopic placement of the endosponge' below.)

Contraindications

●A leak or cavity that connects to the atmosphere (eg, enterocutaneous fistula). EVAC/EVT placed for such leaks will not be able to maintain a seal, which is required for source control. (See "Enterocutaneous and enteroatmospheric fistulas".)

●Visible vessel near or touching the endosponge.

●Patient inability to return at regular intervals for endosponge exchange. (See 'Step 5: Retrieval/exchange of the endosponge' below.)

Patient selection — Clinical factors including the patient's clinical stability and the size, chronicity, and location of the leak have a significant impact on the success rate of EVAC/EVT. Novice providers should select patients with favorable profiles, whereas the more experienced providers may tackle increasingly challenging cases.

●Patient clinical stability – In general, clinically unstable patients (eg, septic, hypotensive, rigid abdomen) should be treated with damage control surgery. EVAC/EVT should only be attempted in such patients by an experienced endoscopist/surgeon and/or when the patient is not a surgical candidate or refuses surgery (palliation) [7].

●Defect size – Large (>3 cm) mucosal defects are challenging for EVAC/EVT because the size of endosponge that can be inserted is limited by the luminal diameter of the esophagus or anus [7]. An undersized endosponge that does not occlude the defect will not achieve source control. However, an experienced endoscopist could potentially use a narrow but longer endosponge to seal off a large defect.

●Chronicity of the leak – The success rate of EVAC/EVT decreases with increasing chronicity of the leak [8,9]. The best result is achieved when EVAC/EVT is initiated within four to six weeks of the initial leak in the peritoneal cavity [10,11]. We favor treating chronic leaks (>90 days) surgically either immediately or after planned two-week EVAC/EVT therapy for debridement and source control [7].

●Fecal diversion for colorectal leaks – For colorectal leaks, continued passage of stool, luminal fluid, and air is a risk factor for EVAC/EVT failure by contaminating and obstructing the endosponge and disrupting the negative pressure suction in the cavity [12]. Although not mandatory, fecal diversion is recommended by experts [13]. Patients who already have fecal diversion from the initial operation are ideal candidates for EVAC/EVT if they develop a leak since EVAC/EVT can be performed without another operation. In a meta-analysis of 276 patients treated with EVAC/EVT, about a quarter required no diverting stoma, another quarter required permanent stoma, and the rest half required temporary diversion, which was subsequently reversed; 94 percent of the fecal diversions were performed during the initial operation [12].

●Location (accessibility) of the leak - Certain small bowel leaks are inaccessible by the endoscope. Additionally, certain leak sites in the foregut and hindgut may be technically accessible but difficult to access given their positions in relation to the endoscope within the lumen of the GI tract. As an example, a leak just beyond the gastroesophageal junction on the gastric side is very difficult to access because it requires the endoscope to be in a retroflexed position for placement and removal of the endosponge. A rendezvous technique has been developed, which uses the endoscope to pull the endosponge into position via an alternate access (eg, gastrostomy tube or percutaneous drain tract) other than a natural orifice such as the mouth [14,15].

A nomogram has been developed to aid in patient selection for EVAC/EVT of the foregut. This nomogram was developed using four major factors that determine the successful use for endoscopic therapy in GI leaks. These include age, body mass index, prior gastric surgery, and days until leak identification. The probability of success using EVAC/EVT can help the managing clinician decide on surgical intervention, using endoscopic therapy, or potentially transferring to a center well versed in leak management [16].

EVAC/EVT MANAGEMENT

Preoperative preparation — The EVAC/EVT procedure varies based on whether the leak is in the upper or lower gastrointestinal (GI) tract.

●For upper GI tract leaks, the procedure should be done under general anesthesia with endotracheal intubation. It is important to protect the airway given the size of the endosponge that will be passing through the oropharynx and by the larynx. Once intubation has occurred, the patient positioning is up to the endoscopist, whether supine or decubitus.

●For lower GI tract leaks, anesthesia and positioning is up to the endoscopist.

●Patients undergoing EVAC/EVT are generally already receiving broad-spectrum antibiotics for the leak or perforation.

●Although EVAC/EVT is an endoscopic procedure, it is usually performed under general anesthesia. As such, standard mechanical and pharmacologic venous thromboembolic prophylaxis should be provided per institutional protocols. (See "Prevention of venous thromboembolic disease in adult nonorthopedic surgical patients".)

Operative technique — The EVAC/EVT procedure can be divided into five steps [7]:

Step 1: Diagnostic endoscopy — Initially, the distance of the leak from the accessed orifice (mouth or anus) is noted to determine if the leak is amenable to endoscopic therapy. Once the leak is identified, the cavity is irrigated and debrided endoscopically as much as possible.

Next, additional features, including size of the perforation, level of contamination, and size of the cavity associated with the leak, must be noted (picture 2). This will determine the size of the endosponge and whether its placement will be intracavitary or intraluminal. We prefer to place the endosponge into the cavity (intracavitary) for better debridement. However, when the size of the perforation is <1 cm, endosponges are generally placed intraluminally because the GranuFoam cannot be compressed to <8 mm [17].

Step 2: Constructing the endosponge — There is no commercially available endosponge in the United States; the endosponge needs to be created by the surgeon/endoscopist. There are several ways to create it, but most reports use a polyurethane GranuFoam provided for use externally by KCI, but not the polyvinyl WhiteFoam, which does not compress as much. The GranuFoam is cut to about 6 cm in length and no larger than 3 cm in width. The width is limited by the lumen of the GI tract (anus or upper esophageal sphincter and esophageal lumen) (picture 3). The endosponge should be tailored to the size of the cavity or leak site. In parts of Europe, there is a manufactured device available called endo-SPONGE that can be used.

The endosponge is attached to tubing to provide the negative pressure externally. This is usually a nasogastric tube (NGT) of varying size.

●For upper GI leaks, the size of the endosponge plus the NGT is usually too large to pass through the nasopharynx. Thus, it is essential to place the tube through the nose and pull it through the mouth prior to attaching the endosponge to allow it to function as a normal NGT.

●For lower GI tract leaks, the endosponge/NGT complex can be directly inserted transanally.

The endosponge is secured to the tubing with a strong permanent suture. Keeping the endosponge attached to the tubing is extremely important both for maintaining a continuous seal and for ease of retrieval. A second suture is placed at the distal end of the endosponge to permit an endoscopic grasper to carry it through the GI tract (picture 4). This can be done with the same permanent suture and with an air knot placed through both the endosponge and the tip of the tubing for stability and strength (the NGT traverses the entire length of the endosponge in a tunnel made with scissors).

The air-sump port in the NGT must be occluded to prevent air to enter the endosponge. This can be accomplished by reversing the sump valve or simply occluding the port with tape.

Step 3: Endoscopic placement of the endosponge — The endoscope is used to carry the endosponge through the GI tract with a rat tooth grasper placed through the working channel of the endoscope grasping the air knot suture. Once the desired location is reached (picture 5), the suture securing the endosponge is released from the grasper under direct visualization. Negative pressure should only be applied after the endosponge has been placed at the ideal location (detailed in next step). Once the negative pressure is applied, the endoscope is removed gently. Pulling the endoscope prior to that may dislodge or displace the endosponge into the wrong location. Prior to removing the endoscope completely from the mouth, one should ascertain that there is no coiled tubing in the oropharynx. The NGT is then secured to the nose or nasal septum with a bridle or tape.

Step 4: Application of negative pressure — Since there is no specific device to provide negative pressure for EVAC/EVT, a portable device such as the KCI wound VAC system is commonly used. Adapting the tubing to the negative pressure device should be done in a closed system. No definitive recommendations have been made for pressure, but continuous pressure settings between 125 and 175 mmHg seem appropriate.

Step 5: Retrieval/exchange of the endosponge — Like other external wound dressings, the endosponge needs to be changed frequently. The ideal interval for exchange is still a matter of ongoing research, but reports recommend somewhere between three and five days. This leaves the endosponge in place long enough to provide a reasonable interval break from procedures but short enough to prevent too much tissue ingrowth into the endosponge.

Removal of an endosponge should be done under endoscopic vision. The endoscope itself is used to dislodge the granulation tissue circumferentially at the tissue/endosponge interface. Pulling on the NGT without such a step may disconnect the NGT from the endosponge and make retrieval of the endosponge much more difficult.

Supportive care — Patients are usually kept in the hospital for the duration of the EVAC/EVT treatment. Routine care includes endoscopic reevaluation/endosponge change at regular intervals, ongoing nutritional support, and drainage of the peritoneal or pleural space (when necessary). Fecal diversion may be necessary to treat colorectal leaks.

Additional procedure to decontaminate peritoneal/pleural cavity — Although EVAC/EVT is capable of achieving source control in most, if not all, patients, those with free perforation into the peritoneal or pleural cavity may have already incurred significant contamination of those spaces that cannot be fully debrided by EVAC/EVT alone. Such patients may require additional procedures, such as percutaneous drainage by interventional radiology, or even surgery to wash out the peritoneum or pleural space [7]. Such procedures can be performed before or after the initiation of EVAC/EVT at the discretion of the endoscopist/surgeon. Once source control is secured internally by EVAC/EVT, the peritoneal/pleural drains should be removed to promote healing.

Nutritional support — Wound healing ultimately depends on nutritional support, although the optimal approach varies between upper and lower GI tract leaks.

●Upper GI leaks – Patients with upper GI tract leaks cannot tolerate oral nutrition; nutritional supplementation during EVAC/EVT therapy relies upon total parenteral nutrition (TPN) or tube feeds.

•TPN is favored given the patients will achieve their nutritional goals regardless of cessation to tube feeds for potentially upcoming procedures. A nutritionist should be following the patient care as tube feedings are often held for repeated endoscopic procedures.

•Tube feeds should be administered to the small bowel via an extension gastrojejunostomy tube or a surgically placed jejunostomy tube. Gastric tube feeds can be pulled into the endosponge by the negative pressure, causing it to clog or contaminate the area. Percutaneous endoscopic gastrostomy (PEG) tube placement is contraindicated in the presence of an esophageal or gastric leak, but if one is already placed, it can be converted to a feeding gastrojejunostomy and be used for distal feeding. It would be counterproductive to place a PEG/gastrostomy tube on suction while the EVAC/EVT is ongoing. Two sources of negative pressure may make it difficult for the endosponge to maintain a seal at the leak site.

●Lower GI leaks – Patients with lower GI tract leaks can take normal oral nutrition, but the patient needs to have a diverting stoma proximal to the leak site. Luminal flow of the fecal contents will disrupt the function of the endosponge, causing it to be dislodged and malfunction. The stoma can be reversed shortly after (within one month of) healing. EVAC/EVT has been credited with a higher rate of stoma reversal than other techniques of leak control.

Staff education — At most hospitals, EVAC/EVT is still a new procedure, and as such, equipment can be mishandled by staff not trained in its intended use. Measures must be taken to ensure that the NGT used for EVAC/EVT is not used for tube feeds or medication administration. Also, the negative pressure application needs to be continuous rather than intermittent. Nursing staff should also avoid disconnecting the tubing or turning off therapy when patient is mobilizing.

Termination of EVAC/EVT therapy — EVAC/EVT therapy is usually terminated when the cavity size can no longer be reduced or is too small to accommodate an endosponge. Early surgical closure of the defect is possible for colorectal leaks. Failure of EVAC/EVT may prompt surgical intervention.

Conventional approach — EVAC/EVT therapy is finished when the leak site is completely sealed. This means that luminal contents will remain within the lumen lined with either epithelium or granulation tissue without leaking into other tracts, sinuses, or fistulas, and there is no further contamination. Endoscopically, the leak site will appear as a shallow wound bed with possibly a small amount of outpouching, but the continuity of the GI tract has been reestablished (picture 6). The length of EVAC/EVT use varies with the organ system and is also affected by the length of time the patient has had the leak [18].

Early surgical closure (for colorectal leak) — In patients with a colorectal leak, transanal suturing of the defect after a few sponge exchanges may further improve outcomes over the conventional approach to EVAC/EVT [19,20]. In a meta-analysis of 29 studies, early surgical closure was associated with fewer endosponge changes (4 versus 10) and a higher eventual intestinal continuity rate (82 versus 65 percent) compared with EVAC/EVT only [21].

Early surgical closure of colorectal leaks can be performed when the abscess cavity is covered with granulation tissue and the tissue edges are compliant enough to be reapproximated with the aid of an anal retractor or one of the endoscopic transanal platforms [13,22]. (See "Transanal endoscopic surgery (TES)", section on 'TES platforms'.)

Failure — The reported failure rate of EVAC/EVT ranges from 0 to 33 percent for esophagogastric leaks [23] and 17 to 19 percent for colorectal leaks [12,24]. EVAC/EVT has failed when [7]:

●Source control has been lost. Typically, the patient will become unstable, febrile, or tachycardic, prompting reevaluation and possibly definitive surgical intervention. Generally, this indicates the endosponge has migrated out of position. It would be up to the clinician to determine whether the patient should be reassessed endoscopically for potential repositioning of the endosponge, or to abandon EVAC/EVT therapy and proceed to definitive surgical management.

●There is failure to progress. With EVAC/EVT, patients are typically examined endoscopically twice a week. If the wound fails to continue to debride, contract, and revascularize, then surgical intervention can be considered.

OUTCOMES — As EVAC/EVT has flourished as a tool to help manage gastrointestinal tract leaks, many reviews, systematic reviews, and meta-analyses have been published to assess its effectiveness [23,25]. These are complex patients and there will always be adverse events and mortality, but many studies lack comparison with other endoscopic management modalities.

Esophagogastric leaks — The earliest assessments of mortality and effectiveness in foregut leaks using EVAC/EVT date back to its early use. In a 2013 study comparing EVAC/EVT therapy with stent placement for intrathoracic esophageal leaks, the mortality rate was much lower with EVAC/EVT than with stent placement (15 versus 25 percent) [26]. In the same study, stricture rate was also lower with EVAC/EVT therapy (9 versus 28 percent).

In another 2013 study comparing EVAC/EVT with conservative management, stent placement, and surgical diversion in patients with anastomotic leaks after esophagectomy, the severity of disease was assessed with Acute Physiology and Chronic Health Evaluation (APACHE) II scores, and the most severe patients either underwent surgical diversion or EVAC/EVT therapy. Once again, mortality of EVAC/EVT was lower compared with surgery (12 versus 50 percent) [27]. Stent placement was done in a less-acute cohort, and its mortality rate was 42 percent.

More recent systematic reviews and meta-analyses that compared EVAC/EVT with other endoscopic modalities for foregut leaks include:

●A 2021 systematic review and meta-analysis of esophagectomy and gastrectomy leak patients showed that the overall complication rate favors EVAC/EVT over SEMS (self-expanding metal stents; 9 versus 31 percent) [28]. Mortality also favored EVAC/EVT (9 versus 20 percent).

●Another 2020 meta-analyses of esophageal leaks showed that short-term complications, long-term complications, and mortality rates (11 versus 22 percent) all favored EVAC/EVT over SEMS [29].

●In another 2021 meta-analysis of esophageal leaks, adverse events were equal with EVAC/EVT and SEMS at 17 percent [30]. Stricture development remained lower in the EVAC/EVT group when the subset analysis was done, which is consistent with other studies [26,31].

Bariatric leaks — As EVAC/EVT therapy's use in bariatric surgery complications continues to show promise, one study compared EVAC/EVT with SEMS use in patients with sleeve leak [32]. They were able to show that only one death occurred in the EVAC/EVT group, and only one bleeding complication occurred in each group. No stricture issues were noted in the study.

Colorectal leaks — Colorectal surgery leaks have also seen a significant rise in use. The demonstration of adverse events is less prevalent, but some studies have illustrated these outcomes. Unfortunately, these review studies don't have comparison data to use of endoscopic therapies.

●A 2021 systematic review showed anastomotic stricture in two studies at 8 and 18 percent [33]. Ongoing peritonitis or sepsis occurred in two studies at 7 and 14 percent. Bleeding complications were noted in two studies at 3 and 6 percent. Mortality was not reported in this study.

●A 2019 systematic review reported an overall complication rate of 14 percent in 276 patients. This included two patients with bleeding (<1 percent), 13 patients with stenosis (5 percent), and 10 patients who died from metastatic disease, not EVAC/EVT therapy (4 percent) [12].

SUMMARY AND RECOMMENDATIONS

●Device and placement – Endoluminal vacuum therapy (EVAC/EVT) is a minimally invasive method of managing leaks and perforations of the gastrointestinal (GI) tract. EVAC/EVT delivers negative pressure to a leak site in the GI tract through a nasogastric tube (NGT) terminating at a polyurethane GranuFoam sponge, termed endosponge. The endosponge is delivered to the site of the leak endoscopically and can be placed outside (intracavitary) or inside (intraluminal) the GI lumen. (See 'Introduction' above and 'Device and placement' above.)

●Mechanism of action – EVAC/EVT promotes healing of leaks/perforations of the GI tract by maintaining source control, debriding wounds, and encouraging tissue reperfusion. (See 'Mechanism of action' above.)

●Indications – EVAC/EVT can be used to treat any GI leak that is accessible by endoscope. Currently, EVAC/EVT is most commonly used to treat esophageal anastomotic leaks, gastric leaks after bariatric surgery (sleeve gastrectomy), and colorectal anastomotic leaks. (See 'Indications' above.)

●Contraindications – EVAC/EVT is contraindicated when the cavity connects to the atmosphere (eg, enterocutaneous fistula), when there is visible vessel nearby, or when there is a lack of reliable follow-up. (See 'Contraindications' above.)

●Patient selection – Patients who are clinically unstable (eg, hypotensive, septic, rigid abdomen) should be treated surgically rather than with EVAC/EVT. Additionally, a defect size >3 cm, a chronic leak of more than four to six weeks, absence of a proximal diverting stoma (in colorectal leaks), and certain unfavorable locations of the leak/cavity may make EVAC/EVT difficult or reduce the success rate. (See 'Patient selection' above.)

●Operative techniques – The steps of EVAC/EVT include diagnostic endoscopy to find the leak and note its characteristics, construction of the appropriate endosponge, endoscopic placement of the endosponge, application of negative pressure, and endosponge retrieval/exchange. (See 'Operative technique' above.)

We generally place an intracavitary endosponge when the defect is ≥1 cm in size and place an intraluminal endosponge when the defect is <1 cm. Leaks >3 cm in size are difficult to heal with EVAC/EVT, and therefore are better treated surgically. (See 'Step 3: Endoscopic placement of the endosponge' above.)

●Supportive care – Patients are usually kept in the hospital for the duration of EVAC/EVT. Besides endoscopic reevaluation/endosponge change at regular intervals, supportive care includes ongoing nutritional support and drainage of the peritoneal or pleural space (when necessary). Fecal diversion may be necessary for a colorectal leak. (See 'Supportive care' above.)

●Termination of EVAC/EVT – EVAC/EVT therapy is usually terminated when the cavity size can no longer be reduced or is too small to accommodate an endosponge. Early surgical closure of the defect is possible for colorectal leaks. Failure of EVAC/EVT may prompt surgical intervention. (See 'Termination of EVAC/EVT therapy' above.)

●Outcomes – EVAC/EVT has been associated with lower mortality, complication, and stricture rate compared with endoscopic stenting in retrospective studies. Nevertheless, various techniques may be used in tandem, rather than in isolation, to heal GI leaks and limit potential complications. (See 'Outcomes' above.)