Proposing a novel technique to exclude the left ventricle with an assist device: insights from 4-dimensional flow magnetic resonance imaging

Eur J Cardiothorac Surg. 2016 Sep;50(3):439-45. doi: 10.1093/ejcts/ezw092. Epub 2016 Apr 22.

Abstract

Objectives: Left ventricular assist device (LVAD) implantation has become an effective treatment for end-stage heart failure patients. However, LVAD support disrupts the natural intracavitary blood flow path through the heart, introducing flow patterns potentially associated with thrombosis especially around the inflow cannula. We present initial insights into in vitro 4D flow magnetic resonance imaging (MRI)-based analysis of flow patterns and an enzyme-activated milk test in a porcine left ventricular (LV) model during LVAD support, in a standard implant technique and in combination with a novel cone-shaped LV sparing prosthesis.

Methods: 4D MRI of an LVAD-supported ex vivo porcine heart failure model was performed in a mock circulatory loop with an LVAD flow of 4.5 l/min. Two separate sets of hearts with an MR-compatible LVAD inflow cannula were used: one with a standard inflow via the pulmonary veins into the LV, one sparing the LV with the cone-shaped prosthesis sutured to the mitral annulus and the LVAD inflow cannula sparing the LV cavity. In correlation, an enzyme-activated milk test was performed to reveal locations of disturbed flows. Data were analysed visually and by performing a case-by-case analysis of flow patterns and enzyme activation locations.

Results: Using 4D flow MRI, the standard implant technique showed markedly altered flow patterns with flow velocity reduction in the LV cavity especially around the inflow conduit. The average flow velocity in the LV cavity was 12.8 ± 2.9 m/s. In contrast, the LV-sparing technique with a cone-shaped prosthesis showed faster and more organized flow of 36.3 ± 7.2 m/s on average to the LVAD inflow cannula. Compared with the standard technique, no recirculating blood flow components were detected. The enzyme-activated milk test confirmed the location and extent of helical and vertical flow patterns.

Conclusions: For the first time, intracardiac flows in an LVAD setting have been shown in vitro based on a mock porcine ex vivo heart failure model by applying 4D flow MRI. Differences in flow characteristics of the same model using a prosthesis sparing the LV cavity are encouraging to further investigate their potential to reduce potential LVAD-associated side-effects, such as intracavitary clotting and its sequelae.

Keywords: 4D flow; Heart replacement; LVAD; Thrombosis.

MeSH terms

  • Animals
  • Heart Atria / diagnostic imaging
  • Heart Atria / physiopathology
  • Heart Diseases / diagnostic imaging
  • Heart Diseases / etiology
  • Heart Failure / surgery
  • Heart Ventricles / diagnostic imaging
  • Heart Ventricles / physiopathology
  • Heart-Assist Devices* / adverse effects
  • Imaging, Three-Dimensional / methods
  • Magnetic Resonance Imaging / methods
  • Models, Cardiovascular*
  • Pilot Projects
  • Prosthesis Design
  • Prosthesis Implantation / adverse effects
  • Prosthesis Implantation / instrumentation
  • Prosthesis Implantation / methods*
  • Sus scrofa
  • Thrombosis / diagnostic imaging
  • Thrombosis / etiology