Effect of Artificial Pulse and Pulsatility on Gastrointestinal Bleeding Rates after Left Ventricular Assist Device Implantation

Objectives: Gastrointestinal (GI) bleedings remain one of the most frequent complications after left ventricular assist device (LVAD) implantation. The HeartMate 3 (HM3; Abbott, USA) is a new-generation LVAD which is associated with several new technical features (e.g., fully magnetically levitated pump, artificial pulse, large pump gaps, and modular driveline). These benefits are supposed to lead to superior outcomes and reduced adverse events compared to its predecessor HeartMate II (HMII; Abbott, USA) and other comparable assist devices such as the HeartWare HVAD (Medtronic, USA). With this study, we investigated the effect of HM3 implantation on the occurrence of GI bleedings compared to other LVADs and focus on the aspect of pulsatility.

Methods: We retrospectively studied a patient cohort of 595 patients who were supported with left ventricular assist devices. A total of 170 patients were supported by HMII, 81 patients were supported by HM3, and 344 patients were supported by HVAD (HeartWare, Medtronic). Data were determined through retrospective examination of medical records. Exclusion criteria were biventricular assist devices and other types of assist devices as well as LVAD exchange and reoperative procedures.

Results: Out of 595 patients, 146 patients (24.4%) presented with gastrointestinal bleeding after LVAD implantation. In the multivariate analysis GI bleeding was significantly dependent on age but not on the BMI or gender of the patient. Statistically we were able to show that the occurrence of GI bleedings is significantly connected to general bleedings but not to cerebral hemorrhages. The subgroup analysis revealed that the occurrence of GI bleedings was higher in the HVAD (26%) and HMII (26%) groups compared to the HM3 patients (14%). Aortic valve opening, Lavare Cycle, and the Artificial Pulse feature showed decreased rates of GI bleeding. However, statistical significance could not be achieved.

Conclusion: The novel HeartMate 3 shows a promising adverse event profile with a reduced rate of gastrointestinal bleedings compared to its competitors. Also the effect of pulsatility and aortic valve opening might decrease the rate of gastrointestinal bleedings. However, statistical significance could not be achieved in this study. Thus, larger multicenter cohort analyses are needed.