Introduction Up to 10% of cirrhotic patients with ascites become refractory to diuretic therapy requiring either large volume paracenteses (LVP), transcutaneous intrahepatic portosystemic shunt (TIPSS) or liver transplantation (LT). In patients where TIPSS or LT is contraindicated, repeated LVP is the only option. LVP is an uncomfortable procedure requiring hospital visits and has risk of complications. The ALFA pump is a surgically implanted automated low flow pump system that transports fluid from the peritoneal cavity to the bladder and is an alternative treatment option for patients with refractory ascites. Our aim was to evaluate the early “real world” experience of the ALFA pump for the treatment of refractory ascites from centres in the UK.

Method Retrospective case note review of the outcomes of patients who had an ALFA pump implanted for refractory ascites outside trials at 4 NHS hospitals since Sept 2013.

Results ALFA pumps were implanted in 9 patients (6 male, 3 female; median age 74, range 41–81) at four centres (4 NCL, 2 UCL, 2 CAM, 1 QEHB). 7 patients had ascites related to cirrhosis (4 Child-Pugh B, 3 Child-Pugh C), 1 due to congestive cardiac failure and 1 with portal hypertension due to arterio-venous malformation. Prior to insertion of the ALFA pump the median frequency of paracenteses was every 10 days (range 7 – 20) and the median volume of ascites drained was 9 litres (range 8 to 12). ALFA pumps were successfully implanted in all patients. Requirement for paracenteses was substantially reduced post ALFA pump implantation with 7 patients no longer requiring any paracenteses. In one patient the pump blocked after 3 weeks and required reimplantation with a second pump that also blocked and the patient has returned to LVP. In another patient the pump blocked after 8 months and a second ALFA pump was successfully implanted. All 9 patients had ALFA pump related complications including 4 post-operative seromas, 3 bladder catheter migrations/occlusions, 5 site related infections, 6 acute kidney injury/electrolyte disturbance and 1 urinary incontinence. 5 of the 9 patients were alive with a functioning pump 6 months after implantation.

Conclusion The ALFA pump offers an alternative option for patients who have refractory ascites in whom TIPSS and LT are contraindicated. ALFA pump implantation substantially reduces the need for paracenteses in a cohort of patients who had previously required a significant number of LVP. ALFA pump related complications were frequent suggesting the need for careful patient selection in this high risk group. Outcomes of a randomised controlled trial are eagerly awaited.

Disclosure of interest None Declared.