1Universitätsklinikum Essen Klinik für Kardiologie und Angiologie Essen, Deutschland
Introduction
Rhythm control by pulmonary vein isolation (PVI) in patients with atrial fibrillation (AF) improves morbidity and mortality. Recent studies showed that this also applies to patients with terminal heart failure. However, critical ill patients bear a high risk for intraprocedural hypotension during deep sedation, which is associated with increased periprocedural complications and mortality. Therefore, optimal intraprocedural monitoring is warranted to ensure patient safety.
Aims
We aimed to explore the potential of a novel noninvasive hemodynamic monitoring (NIHM) technology using artificial intelligence (AI) to predict future events of hypotension during PVI and investigate its feasibility in combination with different ablation modalities.
Methods
NIHM was conducted using the novel Acumen IQ technology ® (Edwards Lifescience, Irvine, CA, USA) in consecutive patients undergoing PVI. The primary outcome was the time to occurrence of a hypotension prediction index (HPI) >85, predicting a mean arterial pressure (MAD) <65 mmHg. The secondary outcome was the time to MAD <65mmHg. Additionally, individual causes leading to an HPI >85 were evaluated. Furthermore, interferences between noninvasive monitoring and different ablation modalities were assessed.
Results
In total 21 patients were monitored using this NIHM technology during PVI-procedures for AF ablation between August and November 2023. Patients were 38% (8/21) male and 62% (13/21) female and had a median age of 70 years. In all patients a primary endpoint event occurred. Time to HPI >85 was 23 + 13 min. In 19 patients (91%) the primary endpoint event was followed by the secondary endpoint event. Time to MAP <65 mmHg was 32 + 12 min. Hypotension was caused by hypovolemia in 63% (12/19), reduced peripheral vascular resistance in 21% (4/19) and reduced left ventricular inotropia in 16% (3/19), as assessed by the hemodynamic monitoring system. In 8 cases (38%) PVI was performed by radiofrequency ablation followed by 8 patients undergoing pulsed field ablation (38%) and 5 patients received cryoballoon ablation for PVI (24%). In general, the NIHM technology was feasible with all ablation modalities. However, during pulsed field ablation NIHM was shortly interrupted after energy release for a median of 40 + 20 seconds.
Conclusion
During PVI, prediction of hypotension using AI during NIHM is feasible and compatible with standard ablation modalities. This bears the potential to improve patient safety during complex procedures in future.