ThP3.3 - Prophylactic Endoluminal Vacuum Therapy – towards an era of reducing leak related morbidity after elective oesophago-gastric resection

Abstract Aims Uncontrolled sepsis following leak from oesophago-gastric surgery is a devastating condition with increased morbidity & mortality affecting longer-term cancer-free survival in patients with cancer resections. We hypothesised that endoluminal vacuum therapy (EVT), previously used to treat leaks, could be used in prophylactic settings to provide source-control, prevent mediastinitis and minimise sepsis related morbidity from leaks. Methods An ad-hocEVT device was created using VACgranufoam (KCI)and a nasogastric-tube(16/18Fr), and placed intraluminally across the repair/anastomosis in select high-risk cases. Continuous negative pressure(75–125mmHg) was applied. Assessment of the repair-site/anastomosis was performed by radiology, endoscopy, clinical assessment or a combination following cessation of EVT. Results 27 patients had prophylactic EVT; M:F ratio was 3:1 and the median age was 68 years (range: 41-82years). 26surgeries were for cancer, while 1 was benign. The rationale for prophylactic EVT included patient related (10), technical factors (16) or both (1). The median duration of prophylactic EVT was 7 days (range: 4-14days). The median length-of-stay was 19.7 days (range: 6-96 days), the two longest admissions were unrelated to the anastomosis(prolonged chyle-leak, and long-term-sequelae of life-threatening bleed). One-patient developed a full-thickness microleak at endoscopic-evaluation, yet remained well with no systemic signs of sepsis, successfully managed with luminalEVT for a further 6days in a ward-based environment. Though prophylacticEVT did not prevent the leak, it significantly minimised the systemic hit that traditionally results from leak and associated mediastinal sepsis. Conclusions ProphylacticEVT is a useful adjunct for management of patients with a high risk for leak, though not a substitute for good surgical technique. There are no short-term complications, yet the additional morbidity, mortality and length-of-stay associated with a fully-fledged leak is avoided. Randomised-controlled-trials are required to further standardise selection criteria, assess potential patient benefit and longer-term impact on survivorship.