Initial clinical experience with a new automated antitachycardia pacing algorithm: Feasibility and safety in an ambulatory patient cohort

Background: Antitachycardia pacing (ATP) in implantable cardioverter-defibrillators (ICD) decreases patient shock burden but has recognized limitations. A new automated ATP (AATP) based on electrophysiological first principles was designed. The study objective was to assess the feasibility and safety of AATP in ambulatory ICD patients. Methods and Results: Enrolled patients had dual chamber or cardiac resynchronization therapy ICDs, history of ≥1 ICD-Treated ventricular tachycardias (VT)/ventricular fibrillation episode, or a recorded, sustained monomorphic VT. Detection was set to ventricular fibrillation number of intervals to detect=24/32, VT number of intervals to detect≥16, and a fast VT zone of 240 to 320 ms. AATP prescribed the components and delivery of successive ATP sequences in real time, using the same settings for all patients. ICD datalogs were uploaded every ≈3 months, at unscheduled visits, exit, and death. Episodes and adverse events were adjudicated by separate committees. Results were adjusted (generalized estimating equations) for multiple episodes. AATP was downloaded into the ICDs of 144 patients (121 men), aged 67.4±11.9 years, left ventricular ejection fraction 33.1±13.6% (n=137), and treated 1626 episodes in 49 patients during 14.5±5.1 months of follow-up. Datalogs permitted adjudication of 702 episodes, including 669 sustained monomorphic VT, 20 polymorphic VT, 10 supraventricular tachycardia, and 3 malsensing episodes. AATP terminated 39 of 69 (59% adjusted) sustained monomorphic VT in the fast VT zone, 509 of 590 (85% adjusted) in the VT zone, and 6 of 10 in the ventricular fibrillation zone. No supraventricular tachycardias converted to VT or ventricular fibrillation. No anomalous AATP behavior was observed. Conclusions: The new AATP algorithm safely generated ATP sequences and controlled therapy progression in all zones without need for individualized programing.