LVAD thrombolysis can be effective but carries a bleeding risk. Existing data suggest intraventricular (vs peripheral) delivery may reduce bleeding, but is associated with early recurrence. We have developed a novel hybrid thrombolytic approach (hTL) using an initial high dose LV delivery of alteplase followed by a maintenance LV infusion. We studied all patients with durable LVAD pump thrombosis (PT) treated with hTL in our institution between 1/2015 and 9/2019. We report the technique, procedural success and complications. We identified 10 patients with median (IQR) age 56 years (50, 64 years). All had undergone Heartware HVAD implantation. Duration of LVAD support preceding PT was 518 days (172, 920 days). All patients had significant increases in power and haemolysis markers, and were treated initially with heparin and tirofiban (n=9) or heparin alone (n=1). Time to hTL was 3 days (2, 5 days). hTL was done in the cardiac catheterisation laboratory after informed consent. Heparin and tirofiban were stopped in advance. An arterial sheath was placed in a femoral (n=5) or radial (n=5) artery. A pigtail catheter was advanced to the LV under fluoroscopic guidance. Alteplase was infused at 1mg/min and stopped when LVAD power returned to baseline. The pigtail catheter was left in situ, and the alteplase maintenance infusion continued at 1mg/hour to a weight-adjusted maximum. The sheath and catheter were then removed. Median total alteplase dose was 74mg (48, 90mg). There was acute normalisation of pump power in 8 patients, and 6 remained free of thrombosis at 1 year. Among these 6, 3 were bridged to transplant and 3 continue on LVAD support. In the 2 patients with recurrent thrombus (at days 8 and 37), 1 was managed with inotropes and urgent transplant, and 1 died. In the 2 acute non-responders, 1 went for urgent pump exchange and 1 died. There was no procedure-related bleeding or death. Two patients had vascular access site haematoma (both femoral access), and a further 3 patients had minor bleeding not requiring intervention. Three patients developed acute kidney injury probably related to haemolysis. We report a hTL protocol for LVAD thrombosis with 80% initial success, low rate of recurrent thrombosis, and no major bleeding. Interpretation is limited by small group size and retrospective design. These outcomes require comparison with other approaches, ultimately in a randomised trial.