Endoluminal vacuum therapy (EVT) as a treatment option for upper gastrointestinal (UGI) tract leaks;10-years' experience in a specialist oesphagogastric centre

ABSTRACT

Introduction: Upper gastrointestinal (UGI) leaks present a significant management challenge, especially in patients with delayed presentation and established sepsis. Traditional treatment strategies such as surgery are associated with high mortality rates. Endoluminal vacuum therapy (EVT) is an emerging treatment option which may reduce morbidity and mortality compared to traditional treatments in this group of patients. We report the outcomes for patients with UGI leaks treated with EVT in a tertiary UK hospital over a 10-year period. Aims & Methods: Between April 2011 and April 2021, 66 patients with UGI leaks from different causes were treated with EVT using an ad-hoc endoluminal vacuum device (EVD). The EVD was constructed using a piece of open cell foam sutured around the distal end of a nasogastric tube, and placed endoscopically either through the perforation and into the extra-luminal leak cavity or intraluminally depending on the morphology of the leak cavity. Continuous negative pressure (125 mmHg) was applied. Endoscopic re-evaluation of the leak cavity with a change of the EVD was performed every 48-120h depending on the patients clinical condition. Patients were fed enterally, and treated with broad-spectrum antibiotics and anti-fungal medication until healing was complete. Information related to treatment and outcome was recorded prospectively. Results: Patients had a median age of 67 years (range 25-92), and mean Apache II score of 21 (range 6-36) at the time of leak diagnosis. Fifty-one (77%) leaks were oesophageal, 12 (18%) gastric, 2 (3%) duodenal, and 1 (2%) pharyngeal. The cause of the leak was anastomotic in 24 patients (36%), iatrogenic in 21 patients (32%), spontaneous in 20 patients (30%), and traumatic in 1 patient (2%). The median number of EVD changes required to heal the leak was 6 (range 1-27), and the median length of hospital stay was 44 days (range 1-196). Successful resolution of the leak occurred in 58 (88%) patients. Eight (12%) patients died during treatment. There were no complications related to insertion of the EVD. Nine (14%) patients had complications during treatment which required further intervention including bleeding in 4 patients (6%), a cerebrovascular accident in 1 patient (2%), a pulmonary embolism in 1 patient (2%), a myocardial infarction in 1 patient (2%), and COVID-19 infection in 2 patients (3%). Following resolution of the leak, one patient (2%) developed a stricture which required endoscopic dilation. Conclusion: EVT is a safe and effective treatment for UGI leaks, and can be used successfully to treat a disparate range of leak causes in critically unwell patients. Further studies are required to develop a standardized procedure to improve the ease with which EVT can be delivered, and enable broader adoption of EVT for this group of patients.