Endoscopic stenting as salvage therapy for high-risk Boerhaave’s syndrome patients

Abstract

Background

Spontaneous oesophageal perforation (Boerhaave’s syndrome) is a life-threatening emergency with significant morbidity and mortality. Surgical repair remains a key component of management for selected patients; however, many high-risk patients are unsuitable for operative management. This article discusses how, in these high-risk individuals, minimally invasive strategies such as the use of covered self-expanding metal stents (SEMS) offer a valuable alternative to surgery. Stent therapy aims to achieve rapid control of oesophageal leakage, limit mediastinal contamination, and facilitate early sepsis control. When applied in appropriately selected patients, this approach may reduce procedural morbidity while providing effective source control.

Methods

This retrospective cohort study analysed all patients diagnosed with spontaneous oesophageal perforation between 2012 and 2024 at a tertiary referral centre. Patients deemed unfit for surgical intervention following comprehensive clinical assessment and acute multidisciplinary team (MDT) discussion—incorporating physiological status, comorbidities, Charlson Comorbidity Index (CCI), Pittsburgh score, markers of infection, and end-organ function—were managed with covered SEMS. Clinical, biochemical, and outcome data were reviewed.

Results

Of 102 patients with oesophageal perforation, 69 were diagnosed with Boerhaave’s syndrome. Twenty-two high-risk patients underwent endoscopic stenting with covered SEMS. The in-hospital mortality rate was 45%, predominantly from sepsis and multiorgan failure, while no deaths were directly stent-related. The 1-year survival rate was 50%.

Conclusions

Endoscopic stenting offers a feasible and effective salvage option for high-risk Boerhaave’s syndrome patients unsuitable for surgery. SEMS likely prevent otherwise inevitable fatalities in this cohort.

Highlight box.

Key findings

• In high-risk patients with spontaneous oesophageal perforation (Boerhaave’s syndrome) who were unsuitable for surgery, endoscopic placement of covered self-expanding metal stents (SEMS) was feasible and not associated with stent-related mortality.

• Despite a high overall in-hospital mortality reflecting severe baseline illness, SEMS provided effective source control and achieved a 1-year survival rate of 50% in this high-risk cohort.

What is known and what is new?

• Operative repair remains the established definitive treatment for Boerhaave’s syndrome, while conservative management is reserved for contained leaks; however, outcomes in non-operative, high-risk patients remain poorly defined.

• This study represents one of the largest single-centre series focusing exclusively on SEMS use in high-risk Boerhaave’s syndrome patients, demonstrating that stenting can avert otherwise inevitable mortality in patients unsuitable for surgery.

What is the implication, and what should change now?

• Endoscopic stenting should be considered a viable alternative therapeutic strategy for physiologically unstable or comorbid patients with Boerhaave’s syndrome within a multidisciplinary framework. Future management algorithms should formally incorporate SEMS for selected high-risk patients, and prospective multicentre studies are needed to refine patient selection and compare outcomes with other minimally invasive strategies such as endoscopic vacuum therapy.

Introduction

Spontaneous oesophageal perforation, first described by Hermann Boerhaave in the early 18^th century, is a life-threatening clinical emergency resulting from a transmural tear of the oesophageal wall due to a sudden increase in intra-oesophageal pressure in the absence of trauma or iatrogenic injury (1). Boerhaave’s seminal case involved a naval admiral who developed sudden, severe chest pain following forceful emesis, with post-mortem findings revealing mediastinal contamination secondary to oesophageal rupture (1).

Historically, this condition carried a universally fatal prognosis until the mid-20th century (2). In 1946, Barrett’s landmark review highlighted the nearly 100% mortality rate of untreated cases and strongly advocated surgical intervention (2). A year later, he reported the first successful surgical management of Boerhaave’s syndrome using a combination of thoracotomy and laparotomy (3).

Although a subset of patients can be managed conservatively—with broad-spectrum antibiotics and a nil per os (NPO) regimen—this approach is generally reserved for minor, well-contained leaks without systemic signs (4). Surgical repair remains the gold standard, particularly for large or uncontained perforations (5). Operative strategies typically involve thoracotomy with repair of the perforation over a T-tube, mediastinal and pleural drainage, and gastrointestinal (GI) decompression via gastrostomy, with nutritional support through jejunostomy (6,7). Trans-hiatal approaches via laparotomy have also been employed when the perforation is accessible from the hiatus, mediastinal contamination is minimal, and the patient is not fit for thoracotomy (8).

More recently, minimally invasive endoscopic techniques have gained increasing attention. Among these, endoscopic vacuum therapy (EVT) (9,10) and covered self-expanding metal stents (SEMS) (11) have emerged as promising non-operative alternatives. EVT requires multiple procedures under general anaesthesia every 72–96 hours and precludes oral feeding. Moreover, it has been reported to prolong hospital stay, posing significant logistical and cost-related challenges (12,13). While EVT has shown promise, these limitations make covered SEMS a more pragmatic option in many settings, particularly in resource-limited or high-throughput centres (11). However, stenting is not without its constraints; for example, SEMS may be technically unsuitable for proximal perforations located less than 20 cm from the incisors due to anchorage limitations (14).

Considering these practical factors, our institution has employed covered SEMS since 2012 for managing spontaneous oesophageal perforations in patients deemed unsuitable for surgery. Although the literature supports stenting as a viable alternative, most studies report on mixed populations, including iatrogenic and malignant perforations, making it difficult to extrapolate outcomes for Boerhaave’s syndrome alone. Additionally, there remains a paucity of data focusing on stent outcomes in high-risk patients—those with delayed presentations, significant comorbidities, or physiological instability. This retrospective cohort study evaluates outcomes of endoscopic stenting in high-risk patients with spontaneous oesophageal perforation (Boerhaave’s syndrome). It aims to provide insight into the utility of stenting in this particularly vulnerable group. Minimally invasive endoscopic therapies have demonstrated safety and feasibility in selected patients with oesophageal perforation. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-aw-2042/rc).

Methods

At our tertiary care centre, all patients diagnosed with oesophageal perforation are promptly evaluated and discussed within a multidisciplinary team (MDT) comprising anaesthetists, intensivists, and upper GI surgeons. The primary responsibility for managing patients with oesophageal perforation lies with the surgical team, who determine the most appropriate treatment pathway.

Patients suitable for conservative management—typically those with contained leaks without systemic sepsis—are treated with broad-spectrum antibiotics, close observation, and supportive care. Patients presenting with uncontained perforation and features of sepsis are offered surgical intervention. This may involve thoracotomy combined with laparotomy, or laparotomy alone via a rooftop incision, depending on the location and extent of the perforation. Patients deemed unsuitable for surgery due to comorbidities or physiological instability are taken to theatre for assessment by the on-call upper GI consultant and surgical team to determine the feasibility of endoscopic management.

During endoscopy, the site and suspected cause of the perforation are identified (Figure 1). The extent of communication with pleural or mediastinal cavities is assessed, followed by endoscopic lavage of these compartments using a gastroscope under direct vision. Bilateral intercostal chest drains are inserted to ensure adequate drainage. Subsequently, a covered SEMS is placed across the defect under combined endoscopic and fluoroscopic guidance; no routine endoscopic fixation or suturing is needed (Figure 2). Stent dimensions, described in the results section, refer to stent diameter at the flared ends and body; stent length is selected intraoperatively according to defect size and anatomical considerations. A nasogastric tube is also inserted for gastric decompression.

Figure 1.

Oesophageal perforation at endoscopy, communicating with the pleural cavity.

Figure 2.

Oesophageal stenting under combined fluoroscopic (left) and endoscopic guidance (right).

Post-procedure, patients are admitted to the intensive care unit (ICU), maintained in a 30-degree head-up position, and initiated on intravenous proton-pump inhibitors, broad-spectrum antibiotics, antifungal agents, and total parenteral nutrition. A computed tomography (CT) scan with oral contrast is performed once the patient is stable enough to be taken to the radiology department and able to swallow the contrast (Figure 3). If no leak is observed, oral intake is gradually resumed. The stent is typically retained for 4–6 weeks and removed endoscopically upon confirmation of complete mucosal healing confirmed using contrast-enhanced CT and/or endoscopic assessment (Figure 4).

Figure 3.

CT with oral contrast showing stent in place and no evidence of contrast leak. CT, computed tomography.

Figure 4.

Endoscopy at 12 weeks showing complete healing before (left) and after the removal of stent (right).

The current study is a retrospective cohort study conducted over a 12-year period [2012–2024], including all patients aged >16 years diagnosed with spontaneous oesophageal perforation (Boerhaave’s syndrome). All consecutive patients with radiologically or endoscopically confirmed spontaneous oesophageal perforation were included; iatrogenic or malignant perforations were excluded from the primary subgroup analysis.

Data extraction was performed by two independent reviewers to minimise selection and transcription bias. All eligible patients within the 12-year period were included; no formal sample-size calculation was required owing to the retrospective design.

Ethics

As this was a retrospective audit of routinely collected clinical data, formal ethical approval was not required as per local institutional policy. The study was registered as a clinical audit with the Clinical Governance and Audit Department of Royal Preston Hospital (audit registration No. UGI/SE/2023-24/07). Individual informed consent was waived due to the retrospective nature of the study and the use of anonymised data. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Statistical analysis

Descriptive and comparative analyses were conducted using SPSS v16. Continuous data are presented as median interquartile range (IQR); categorical variables as n (%). Missing data were excluded pairwise. No sensitivity analyses were performed owing to complete data capture.

Results

Over a 12-year period [2012–2024], a total of 102 patients with oesophageal perforation were managed at our institution. The aetiologies varied from iatrogenic to malignant perforations (Table 1). Sixty-nine of these cases were Boerhaave’s syndrome, accounting for 67% of all perforations.

Table 1. Aetiology and management of oesophageal perforation.

Cause of perforation	Conservative	Surgical	Stenting	Palliative	Total
Boerhaave’s syndrome	12	31	22	4	69
Iatrogenic	6	11	1	2	20
Foreign body	3	3	0	0	6
Cancer perforation	1	0	3	1	5
Cancer lung invasion	0	0	1	0	1
Corrosive intake	0	1	0	0	1
Total	22	46	27	7	102

Data are presented as number.

The male-to-female ratio in the Boerhaave’s syndrome group was 3:1. The median age was 66 years [mean: 63 years; standard deviation (SD): 18.7 years; 95% confidence interval (CI): 58.5–67.6]. The peak incidence occurred in the 70–80-year age group, with 18 individuals falling within this range (Figure 5).

Figure 5.

Age group distribution of the patients with Boerhaave’s syndrome.

Patients were classified as high risk for surgical intervention following MDT assessment based on physiological instability, delayed presentation (>24 hours), significant comorbidity burden [Charlson Comorbidity Index (CCI) ≥3], high Pittsburgh Perforation Severity Score, and anaesthetic risk [American Society of Anaesthesiologists (ASA) ≥ III]. Chronological age alone was not used as an exclusion criterion for surgery.

Management strategies

Management strategies for Boerhaave’s syndrome were categorised into four primary approaches:

1. Conservative management;

2. Surgical intervention;

3. Endoscopic stenting;

4. Palliative care for patients presenting in end-of-life scenarios (Table 1).

Conservative management was applied in 12 patients, typically those with well-contained perforations and no evidence of systemic sepsis. These patients received close clinical monitoring, broad-spectrum antibiotics, and supportive measures.

Surgical intervention was the most employed strategy, with 31 patients undergoing operative procedures—17 requiring combined thoracotomy and laparotomy, and 14 treated with laparotomy alone (Table 1).

Endoscopic stenting was performed in 22 patients who were deemed unfit for surgery due to advanced comorbidities, delayed presentation, or physiological instability. All procedures utilised covered SEMS (Danis Seal^®, 30 mm × 25 mm × 30 mm; Ella-CS, Hradec Králové, Czech Republic). Reported dimensions refer to stent diameter at the flared ends and body; stent length was selected intraoperatively according to defect size and anatomical considerations.

This stented cohort included 13 males and 9 females, with a median age of 71.5 years [range, 39–90 years; IQR: 22.75 years; first quartile (Q1) =55 years, third quartile (Q3) =77.75 years]. Most patients were classified as ASA grade III (median: 3; IQR: 3–3). The median CCI was 4 (IQR: 2; Q1 =3, Q3 =5), and the median Pittsburgh Perforation Severity Score was 8 (IQR: 3; Q1 =6.5, Q3 =9.5).

Laboratory and clinical findings

On presentation, the median C-reactive protein (CRP) was 262.5 mg/L (IQR: 148 mg/L; Q1 =185 mg/L, Q3 =333 mg/L), while the median white blood cell (WBC) count was 15.5×10⁹/L (range, 10×10⁹/L–34×10⁹/L; IQR: 18×10⁹/L; Q1 =11×10⁹/L, Q3 =29×10⁹/L).

The distal oesophagus was the most common site of perforation. The median perforation size was 3 cm (IQR: 2 cm; Q1 =2.5 cm, Q3 =4.5 cm).

Regarding underlying pathology:

• ❖ 13 patients (59%) had no identifiable oesophageal disease;

• ❖ 4 (18%) had benign oesophageal strictures;

• ❖ 2 (9%) had oesophagitis;

• ❖ 1 (5%) had an oesophageal diverticulum; perforation occurred adjacent to but not within the diverticulum;

• ❖ 2 (9%) had combined stricture with oesophagitis.

Outcomes

The median ICU stay was 20 days (range, 3–52 days; IQR: 19 days; Q1 =11 days, Q3 =29 days), and the median total hospital stay was 48.5 days (range, 24–127 days; IQR: 56.5 days; Q1 =31.5 days, Q3 =88 days).

Complications were classified according to the Clavien-Dindo system (Table 2). Stent-related complications occurred in 3 cases (13.6%), all due to stent migration requiring revision.

Table 2. Complication severity distribution (n=22).

Grade	Definition	Cases	Percentage (%)	Clinical examples
Grade 0	No complications	5	22.7	Uneventful recovery
Grade II	Pharmacological treatment†	2	9.1	Blood transfusions
Grade IIIb	Intervention under GA	3	13.6	Stent revision
Grade IVb	Life-threatening (ICU care)	2	13.6	ARDS, ventilator dependence, CNS complications
Grade V	Death	10	45.5	Sepsis, multi-organ failure

ARDS, acute respiratory distress syndrome; CNS, central nervous system; GA, general anaesthesia; ICU, intensive care unit.

The in-hospital mortality rate was 45%, with the most common cause of death being progressive sepsis leading to multi-organ failure. Importantly, none of the deaths were directly attributable to stent-related complications.

In terms of survival, the in-hospital survival rate was 55.5%, and the 1-year survival rate was 50%. At the time of data retrieval, with a median follow-up of 49.4 months (range, 7–138 months), no additional late mortality was observed.

Discussion

In our series, spontaneous oesophageal perforation—commonly referred to as Boerhaave’s syndrome—accounted for 67% of cases, consistent with the condition being a leading cause of non-iatrogenic perforations. The reported incidence of Boerhaave’s syndrome varies widely in the literature. In a multinational study using the Pittsburgh Oesophageal Perforation Severity Score, approximately 40% of oesophageal perforations were attributed to Boerhaave’s syndrome, identifying it as the most common underlying cause in that cohort (15). An English study by Markar et al. (16) reported spontaneous perforation in 81.9% of cases, while iatrogenic causes accounted for only 5.9%, underscoring the predominance of spontaneous aetiology in certain populations. This variation may reflect differences in referral patterns, diagnostic thresholds, and the prevalence of endoscopic procedures or trauma-related injuries across institutions.

Our demographic data revealed a strong male predominance, with a male-to-female ratio of 3:1 in the Boerhaave subgroup, consistent with prior literature citing ratios of 2:1 to 5:1 (17). This is likely attributable to behavioural risk factors such as higher alcohol intake and forceful vomiting, more common among men. The peak age of presentation in our study (70–80 years) was slightly higher than the traditionally reported 40–60 years range (18,19). This may reflect improved longevity, heightened diagnostic vigilance in the elderly, and referral bias towards tertiary centres managing complex cases.

Anatomically, the lower third of the oesophagus was most frequently affected, in keeping with classical descriptions of Boerhaave’s syndrome (1). This region is particularly vulnerable due to the anatomical transition zone at the gastro-oesophageal junction, where elevated intraluminal pressure during emesis predisposes to rupture (20).

Comparison of management approaches

Surgical repair remains the mainstay of treatment for oesophageal perforation, particularly when contamination is extensive or presentation is delayed (5,20). In our cohort, surgery was the most common approach (31 patients). Combined thoracic and abdominal approaches were employed when contamination was extensive in the mediastinum, while trans-hiatal access was used for lower perforations without thoracic spread (21).

However, many patients in our series were poor surgical candidates due to age, comorbidities, or physiological instability. Patients managed with endoscopic stenting demonstrated substantially higher comorbidity burden, physiological derangement, and delayed presentation compared with surgically managed patients, precluding meaningful risk-adjusted comparison. Formal comparative statistical analysis was therefore not performed due to significant baseline heterogeneity. Twenty-two patients underwent endoscopic stenting, representing a high-risk non-operative cohort with a median age of 71.5 years, a median CCI of 4, and a Pittsburgh score of 8—figures denoting significant physiological compromise. Both scores have been validated as independent predictors of mortality in oesophageal perforation (22).

Inflammatory markers were markedly elevated (median CRP: 262.5 mg/L; WBC: 15.5×10⁹/L), consistent with systemic sepsis, another key prognostic determinant (23). In this setting, SEMS provided a minimally invasive alternative, allowing for rapid source control and avoidance of major surgery. Similar findings have been reported in national and multicentre analyses supporting the role of SEMS in benign oesophageal perforation (24).

Mortality and outcomes

Our in-hospital mortality rate of 45% appears higher than rates reported in general perforation cohorts, but it reflects the extreme risk profile of our patients. For instance, Thornblade et al. (24) reported a 12% 6-month mortality rate among stented patients in a large national cohort (mean age 49 years, CCI ≥2 in 42%). Their cohort excluded patients over 64 years, whereas the median age in our study was 71.5 years. The higher CCI and Pittsburgh score in our group indicate a markedly sicker population, explaining the disparity. The prolonged ICU and hospital stay observed reflects the severity of sepsis, requirement for organ support, and stepwise recovery in a physiologically compromised cohort. ICU duration was influenced by local protocols requiring resolution of sepsis, nutritional optimisation, and ventilatory weaning prior to step-down care.

Crucially, stenting in our study served as salvage therapy for patients who were otherwise non-operative candidates. Historically, untreated Boerhaave’s syndrome carried nearly 100% mortality (3), so despite the high numerical mortality, SEMS likely prevented otherwise fatal outcomes.

Strengths and limitations

This is one of the largest single-centre series evaluating SEMS specifically in spontaneous oesophageal perforation (Boerhaave’s syndrome) among high-risk patients. Consistency in management by experienced upper GI surgeons, standardised MDT protocols, and comprehensive data collection lend reliability to our findings.

Nevertheless, the retrospective design carries the inherent risk of selection and documentation bias. Absence of a matched surgical control group limits comparative interpretation. Certain variables, such as the timing of symptom onset and mediastinal contamination severity, were incompletely documented. Moreover, long-term functional outcomes (e.g., dysphagia, quality of life) were not assessed. As a single-centre experience, generalisability is limited to similar tertiary settings managing complex, physiologically unstable patients.

Conclusions

Endoscopic stenting using SEMS represents a feasible and effective salvage option for spontaneous oesophageal perforation (Boerhaave’s syndrome) in patients unsuitable for surgery.

Despite high in-hospital mortality reflecting severe baseline illness, SEMS likely prevented otherwise inevitable fatalities. Our findings support the role of stenting in high-risk, non-operative patients when applied within a multidisciplinary, protocol-driven framework.

Further prospective multicentre studies are warranted to define optimal patient selection criteria and compare SEMS outcomes against other emerging minimally invasive techniques, such as EVT.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. As this was a retrospective audit of routinely collected clinical data, formal ethical approval was not required as per local institutional policy. The study was registered as a clinical audit with the Clinical Governance and Audit Department of Royal Preston Hospital (audit registration No. UGI/SE/2023-24/07). Individual informed consent was waived due to the retrospective nature of the study and the use of anonymised data. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Data Sharing Statement

Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-aw-2042/dss

jtd-18-02-132-dss.pdf ^{(68.8KB, pdf)}

DOI: 10.21037/jtd-2025-aw-2042