loading page

  • +2
  • Erik Sorensen,
  • Bartley Griffith,
  • Erika Feller,
  • Lynn Dees,
  • David Kaczorowski
Erik Sorensen
University of Maryland Medical System

Corresponding Author:[email protected]

Author Profile
Bartley Griffith
University of Maryland School of Medicine
Author Profile
Erika Feller
University of Maryland School of Medicine
Author Profile
Lynn Dees
University of Maryland Medical Center
Author Profile
David Kaczorowski
University of Pittsburgh Medical Center
Author Profile


Background: We previously demonstrated better inflow cannula (IFC) position and reduced pump thrombosis with a centrifugal-flow LVAD (CF-LVAD) compared to an axial-flow device. We hypothesized that implant technique and patient anatomy would affect CF-LVAD IFC positioning and that malposition would impact LV unloading and outcomes. Methods: Pre- and postoperative computed tomography (CT) scans were reviewed for patients with six-month follow-up. Malposition was quantified using angular deviation from an ideal line in two planes. IFC position was compared between conventional sternotomy (CS) and lateral thoracotomy-hemisternotomy (LTHS). The influence of LV end-diastolic dimension (LVEDD), body mass index (BMI), and CT-derived anatomy was determined. LV unloading was assessed by LVAD flow index (FI) and pre- to post-LVAD decrement in mitral regurgitation (MR) and LVEDD. Outcome measures were pump thrombus or stroke (PT/eCVA); 30-day and total heart failure-related readmissions (HFRAs); and survival free of surgery for LVAD dysfunction. Results: One hundred fourteen patients met criteria. Total malposition magnitude was higher for CS than LTHS (p=0.04). Midline-LV apex distance predicted lateral-plane malposition (p=0.04), while apex-LVOT angle predicted both anterior- (p=0.01) and lateral-plane (p=0.04) malposition. Lateral-plane malposition predicted decreased LVAD FI at three (p=0.03) and six (p=0.01) months. Total malposition magnitude predicted increased 30-day HFRAs (p=0.04), while lateral-plane malposition predicted more overall HFRAs (p=0.01). Malposition was not associated with PT/eCVA, changes in MR or LVEDD, or survival free of surgical revision. Conclusions: Patient anatomy and surgical technique were associated with CF-LVAD IFC malposition. In turn, malposition was associated with increased readmissions and decreased LVAD FI.