Geometry‐Related Technical Complications After
LVAD
Implantation: Impact on Device Dysfunction and Clinical Outcomes
Ali Saad Merzah, Ezin Deniz, Sara Knigge, Jasmin Hanke, Alina Zubarevich, Mohamad Dia, Chaikovska Sofia, Todurov Borys, Saeed Torabi, Bastian Schmack, Alexander Weymann, Zurab Darbaidze, Josephina Haunschild, Christian D. Etz, Max Biewer, Arjang Ruhparwar, Jan D. Schmitto, Günes Dogan ABSTRACT
Background
Geometry of LVAD implantation is a determinant of pump performance and hemocompatibility. Despite technically successful operations, avoidable geometric errors may produce characteristic imaging patterns and clinically relevant device dysfunction.
Objectives
To determine the frequency, diagnostic pathway, and clinical impact of preventable geometry‐related failure modes after HeartMate 3 implantation in a single‐center referral cohort, specifically inflow cannula malposition, outflow graft kinking/excessive length, and aortic anastomotic stenosis.
Methods
Retrospective, single‐center analysis of 21 consecutive HeartMate 3 recipients implanted between 2016 and 2025 at external centers and subsequently evaluated in our department. All patients underwent CT/CTA when device dysfunction was suspected. Failure modes were predefined and adjudicated using standardized imaging criteria. CT/CTA was the primary modality used to confirm geometric failure modes following triggers such as low‐flow alarms, hemolysis, and abnormal chest radiography.
Results
Among 21 HeartMate 3 recipients, inflow cannula malposition was identified in 7 patients, outflow graft kinking/excessive length in 12, and aortic anastomotic stenosis in 2.
Conclusions
In this single‐center referral cohort, geometric failure modes were identified in 15 of 21 patients (71.4%) after HeartMate 3 implantation and showed reproducible CT/CTA signatures. A standardized imaging‐driven diagnostic pathway may improve recognition of correctable geometric problems in patients referred for suspected device dysfunction; given the referral‐based design, these frequencies should not be interpreted as the true incidence in an unselected HeartMate 3 population.