Endoluminal Vacuum Therapy for the Management of Boerhaave Syndrome: A Case Series

Boerhaave syndrome is described as a sudden increase in intra-luminal oesophageal pressure leading to transmural oesophageal perforation [1]. This uncommon condition carries a poor prognosis due to high associated morbidity and mortality. It is characterised by Mackler’s triad of vomiting, chest pain, and subcutaneous emphysema; however, up to a third of cases are clinically atypical often leading to delayed diagnosis [1]. Early intervention has been shown to provide the best possible clinical outcomes, as widespread contamination, distal obstruction, and sepsis are life-threatening complications of delayed treatment [1‒3].

Historically, open primary oesophageal repair was considered to be the gold standard treatment of Boerhaave syndrome, which focused on pleural and mediastinal debridement and spillage prevention [1, 2]. Oesophagectomy would also be carried out if there was extensive oesophageal necrosis preventing effective repair [1, 2].

Recently, endoscopic self-expanding metal stents (SEMSs) and endoluminal vacuum therapy (EVT) have been increasingly utilised to seal oesophageal anastomotic leaks [1, 3]. Endoscopic treatment can be particularly helpful in patients presenting early without widespread contamination or those who are considered unfit for surgical intervention [1, 3].

Herein, we report the successful use of the novel EVT, particularly the Eso-SPONGE in the treatment of 3 surgically unfit patients with Boerhaave syndrome in the Imperial College NHS Trust London, a tertiary upper gastrointestinal centre, receiving referrals from all around the country, treated between 2019 and 2021. The CARE Checklist has been completed by the authors for this case series, attached as online supplementary material (for all online suppl. material, see https://doi.org/10.1159/000540694).

EVT for oesophageal perforation involves the endoscopic placement of a polyurethane sponge (Eso-SPONGE) into the wound cavity using an overtube, which is then removed when sponge is in place, allowing it to unfold [4]. The sponge is then connected via a trans-nasal drain to continuous negative pressure (50–125 mm Hg) and is exchanged every 2–5 days endoscopically [4]. At our unit, the sponge exchanges were conducted as recommended, with some practical issues regarding facilitating sponge exchanges on weekends, leading to some variation in exact sponge exchange timings.

A previously fit 66-year-old female presented with an acute history of vomiting and chest pain. The initial chest X-ray led to a diagnosis of Boerhaave syndrome, showing bibasal consolidation and pneumomediastinum. Her computed tomography (CT) chest abdomen pelvis scan confirmed a large left-sided defect in the mid-portion of the oesophagus, with gas locules tracking into a large left basal collection shown in Figure 1a.

Having been deemed unfit for surgical intervention due to difficulties with ventilation and haemodynamic instability, she was initially stabilised on the intensive care unit (ICU) with the aid of radiologically placed chest drains and intravenous antibiotics. She was subsequently referred for EVT, using the Eso-SPONGE. Her first therapeutic endoscopy took place on day ten of admission, revealing a 5-cm defect in the oesophageal wall and an adjacent cavity. In addition to this, major vessels and the chest drains could be visualised endoscopically. The Eso-SPONGE was placed into the cavity and connected to suction (75 mm Hg). She remained on ICU for 3 months, and during this period, her Eso-SPONGE was exchanged 16 times, resulting in the resolution of sepsis and healing of the defect, as shown in Figure 1b. Sponge exchange was performed every three to 5 days, and she was successfully stepped down from ITU. Following a period of rehabilitation, she was discharged to her home, where she continued to make good recovery.

A frail 80-year-old woman was admitted to hospital with a history of vomiting followed by sudden onset backpain and epigastric tenderness. Her CT chest abdomen pelvis scan confirmed the diagnosis of Boerhaave syndrome, with a tear seen at the distal third of the oesophagus.

A CT scan performed on admission demonstrated ground glass opacities in the right upper zone suggestive of infection, increased bilateral pleural effusions, and widespread subcutaneous emphysema shown in Figure 2a, b. Having discussed management options with the patient, she decided to undergo endoscopic management due to concerns regarding the risk of oesophagectomy. She was started on total parenteral nutrition, a nasogastric (NG) tube was sited, and she was covered with broad-spectrum antibiotics and anti-fungals. In addition, a chest drain was inserted to manage the pleural effusions.

The patient’s endoscopic management included drainage as well as SEMS placement. Contrast swallow showed an ongoing leak from the left lateral aspect of the defect, and CT, cranial stent migration, and failure to control the extravasation as shown in Figure 2c, necessitating placement of a second stent.

Subsequent CTs after placement of the second stent demonstrated persistent oesophago-pleural fistulation, loculated left hydropneumothorax, right pleural effusion, and mild left lung atelectasis; therefore, an Eso-SPONGE was then considered. Under sedation, previous stents and suturing holding the stents in situ were removed, and an Eso-SPONGE was guided with a snare into the defect lumen, leaving an 80% intra-oesophageal defect. A clip was used to secure the sponge. The sponge was connected to continuous suction at 120–125 mm Hg. The NG tube was replaced, and repeat endoscopy and sponge exchange were scheduled 5 days later.

Following CTs demonstrated a decrease in the size of the collection, the sponge was intermittently changed, and a radiologically inserted percutaneous gastrostomy with jejunal extension halted the need for total parenteral nutrition and NG tube. Three months later, CT and contrast swallow demonstrated no large volume leak. The distal oesophagus was distorted and slightly narrowed; however, no significant stricturing was demonstrated. The patient was successfully discharged to her home with analgesia, continuing her regular medications, eating and drinking normally.

A 72-year-old Caucasian man presented to the Accident and Emergency Department with epigastric pain and surgical emphysema of the face, neck, and chest after an episode of vigorous vomiting precipitated by a food allergy. No pyrexia or tachycardia was reported. The patient’s medical history includes hypertension, stage 1 bowel cancer, and asthma. Non-contrast CT thorax showed spontaneous oesophageal rupture, pneumomediastinum, extensive subcutaneous emphysema, low-density bilateral pleural effusions, and bilateral lower lobe consolidations with an increasing volume, shown in Figure 3a–c. These findings confirmed the diagnosis of Boerhaave syndrome.

Upon admission, the patient preferred more conservative measures; hence, minimally invasive endoscopic stenting with a fully covered SEMS was used. The patient was admitted to a high dependency unit and underwent intubation and ventilation.

Stenting ultimately proved unsuccessful due to the presence of ongoing oesophageal leak, with a right para-oesophageal collection measuring 5 × 3 × 9 cm despite stent exchange. Furthermore, the patient experienced recurrent episodes of sepsis with bilateral pleural effusions, mediastinitis, pericardial effusion, and fluid overload.

The patient was therefore referred to receive an Eso-SPONGE that was changed every 3–5 days and was established on nil by mouth and jejunostomy feeding. The Eso-SPONGE was removed after 6 months of hospital stay when a water-soluble swallow performed showed smooth passage of contrast from oropharynx into oesophagus. He was discharged with a home-feeding regimen and soft diet. He has since made an excellent recovery and returned to normal function.

We present our experience with EVT for the treatment of Boerhaave syndrome, in which full defect closure was achieved in 3 patients who were either unfit or declined surgical intervention. The first patient received EVT as a primary intervention, whereas the 2 remaining patients had EVT as a secondary treatment after failure of SEMS to achieve full defect closure. This case series adds to the evidence supporting the use of EVT as a novel life-saving treatment utilised as either a primary therapeutic intervention or second-line intervention where conservative and endoscopic options fail.

Although this case series describes the experience of a single centre with a limited sample size of 3 patients, this study supports the existing evidence that EVT is indeed a safe and effective option in the management of oesophageal perforations. Though, it is important to highlight that our patients had prolonged hospitalisations and complications. However, these were managed successfully, and overall, treatment with EVT led to complete defect closure.

There is limited evidence surrounding the use of EVT for other indications apart from anastomotic leaks, including Boerhaave syndrome. The first experience of successful use of EVT in Boerhaave syndrome was published in 2008 by Wedemeyer et al. [5], which described successful EVT treatment after failure of surgery and SEMS in 2 patients, hence emphasising the importance of EVT as a safe endoscopic treatment option for oesophageal perforations. Alakkari et al. [6] reported the first UK experience of EVT for the management of oesophageal anastomotic leak and Boerhaave syndrome. In the case of anastomotic leaks, they found EVT results in faster healing and shorter lengths of hospital stay. In their study, Alakkari et al. [6] reported one case of spontaneous oesophageal perforation treated with EVT [6].

Our experience with EVT is mirrored by multiple studies, including the aforementioned (shown in Table 1) that have demonstrated its success as a promising alternative treatment to life-threatening oesophageal perforations [7‒9]. Despite the significant comorbidities of patients with Boerhaave syndrome and other oesophageal defects in our case series, and of other authors, EVT was well tolerated [7, 10, 11].

The use of EVT has evolved over time; for instance, in Bludau et al. [12] initial experience of EVT with 14 patients with various oesophageal perforations demonstrated that a combination of EVT and endoscopic stenting achieved higher closure rates of defects. In contrast, the most recent experience of EVT in the study by Bludau et al. [12] reaffirms some of the benefits of EVT when used in isolation and demonstrates the potential for EVT to be used as an initial treatment option where SEMS may be inadequate [14]. This study mirrors our small, yet similar experience of EVT.

Moreover, Brangewitz et al. [10] showed that successful wound closure was independently associated with EVT, the overall closure rate was also significantly higher in the EVT group (84.4%) compared with the SEMS group (53.8%), concluding that EVT is an effective endoscopic treatment option with higher effectiveness than stent placement. Similarly, Schneiwind et al. [11] showed that EVT was associated with superior outcomes and lower mortality (12%) compared to surgical intervention (50%, p = 0.01) and stenting (83%, p = 0.0014) among systemically ill patients matched by APACHE II scores. Additionally, in Laukotter et al. [13] study of 52 patients, EVT was successful as 94.2% of patients healed with EVT alone, negating the need for any secondary interventions.

Luttikhold et al. [16] conducted a multicentre retrospective cohort study examining the use of EVT in oesophageal perforations. The success rate was 89%, with EVT failing in 3 patients: two deceased during EVT (septic embolic stroke, pulmonary embolism) and one esophagectomy due to a persisting defect. Two adverse events occurred: one iatrogenic defect expansion during sponge exchange and one haemorrhage during sponge removal.

Despite the evident success of EVT in the treatment of oesophageal perforations, there are several disadvantages. The need for regular endoscopic assessment of defect closure following EVT is a concern, as this has shown to increase hospital stay [10]. Additionally, repeat endoscopic procedures may necessitate the use of general anaesthesia and return to theatre [10]. However, this is in fact what makes EVT unique compared to other treatment options, as the regular endoscopic interventions provide more chances at visualising wound cavity and therefore encourage regular drainage, preventing life-threatening sepsis [13]. On the other hand, it would be interesting to see how early introduction of EVT, particularly that of the Eso-SPONGE, affects rates of reoperation when used as first-line treatment in surgically unfit patients.

Other possible complications range from sponge dislocation to minor bleeding following sponge exchange and stricture formation [9, 13]. In Mastoridis et al. [9] study, strictures occurred in 33% of patients which required endoscopic balloon dilatation. However, these patients had post-surgical anastomotic leaks and not Boerhaave syndrome, and it is still unclear whether EVT alone is responsible for stricture formation as these patients had complicated oesophageal pathology which may have increased the risk of strictures. Therefore, this study’s findings should be interpreted with caution. Incidentally, in a study comparing EVT to stenting, anastomotic strictures were reported in 9.4% of the EVT group versus 28.2% of the stent group (p < 0.005) [10]. Authors also argue that some strictures seen in the EVT group were most likely unrelated to EVT itself as they occurred closer to previously placed stents and away from defects being treated [11]. This along with other studies suggests that EVT is indeed superior to stenting as a choice of treatment of upper GI defects [10, 14]. Therefore, the use of EVT before stenting may provide better treatment outcomes.

Chon et al. [17] described an innovative use of a hybrid SEMS to treat an oesophageal anastomotic leak which combined EVT and classic SEMS in a patient with previous total gastrectomy. The concept arose after limited evidence regarding superiority of one technique over the other. Though larger, randomised studies are required to substantiate the efficacy of this method, this case report sheds light on another potential therapeutic option for oesophageal perforations and anastomotic leaks.

Another concern with EVT is the time and costs associated. As demonstrated by Table 1 and by this case series, there is a wide range of intervention time scales and hospital stays of up to 6 months, especially if patients necessitate ICU support. Therefore, it is vital to consider the potential costs of this, especially in public health systems which necessitate cost-effectiveness analyses to implement interventions into national guidelines. Arguably, SEMS has been used in the past to seal oesophageal leaks and does not require as stringent monitoring or sponge exchange. However, as demonstrated by Brangewitz et al. [10] SEMS is susceptible to migration, blockage, and is not as effective at sealing the cavity as EVT, with lower closure rates.

Though there is indisputable scarcity of evidence in the literature regarding the use of EVT in the treatment of oesophageal perforations, this case adds to the existing evidence base highlighting the efficacy of the Eso-SPONGE and superiority to stenting. Undoubtedly, a prospective multicentre RCT is needed to substantiate the Eso-SPONGE’s true clinical benefit in treating Boerhaave syndrome. Hence, we support the suggestions made by Mastoridis et al. [9] regarding the need for establishing a national registry of EVT use in the treatment of oesophageal wall defects, and we acknowledge the interim results of a multicentre registry published by Richter et al. [15]. This will allow for additional evidence-based guidelines to be implemented, as publication bias towards successful treatment cannot be excluded.

In conclusion, despite the challenges associated with EVT, it is still one of the most promising among treatment methods for Boerhaave syndrome, particularly in cases which may warrant EVT as first-line endoscopic treatment, rather than attempting stenting first, as patients could have been arguably treated faster and more efficiently using the Eso-SPONGE alone. However, a cost-effectiveness analysis is warranted, alongside a blinded RCT to investigate the true benefit of EVT in Boerhaave syndrome.

Written informed consent was obtained from the patients for publication of this case series and any accompanying images. Ethical approval is not required for this study in accordance with local or national guidelines.

The authors have no conflicts of interest to declare.

Infrastructure support for this research was provided by the NIHR Imperial Biomedical Research Centre (BRC).

Dr. Daniella Soussi and Dr. Batool Helmi Ahmad Alharahsheh involved in data interpretation, literature review, and writing and overview of the manuscript. Mr. Piers Robert Boshier, Dr. Jonathan Hoare, and Mr. Christopher John Peters involved in the care of the 3 patients and overview of the manuscript. Dr. Natalie Direkze involved in the care of the 3 patients, sponge changes, and overview of the manuscript. Dr. Robert Thomas involved in the radiological interpretation of the imaging of all 3 patients and overview of the manuscript. Dr. Sophie Stevens involved in the care of the 3 patients, data interpretation, and writing and overview of the manuscript.

All data generated or analysed during this study are included in this article. Further enquiries can be directed to the corresponding author, D.S.