Abbreviations and Acronyms

CVP

central venous pressure

EUROMACS

European Registry for Patients With Mechanical Circulatory Support

INTERMACS

Interagency Registry for Mechanically Assisted Circulatory Support

IVS

interventricular septum

LVAD

left ventricular assist device

RV

right ventricular

RVAD

right ventricular assist device

Introduction

Left ventricular assist devices (LVADs) have become an increasingly common therapeutic option for patients with advanced heart failure. Unfortunately, the right ventricle is not afforded the same benefit, and right ventricular failure (RVF) is a major cause of morbidity in patients with LVADs. Early intraoperative diagnosis and management of RVF after LVAD implantation is therefore crucial to improve patient outcomes.

Current Opportunities for Improvement

Right ventricular failure after LVAD implantation is common, occurring in approximately one-third of patients with third-generation LVADs.¹ However, there are significant challenges in the diagnosis and prediction of RVF.

The first problem is the lack of a universally accepted definition and classification for RVF.^2,3 Two leaders in this arena, the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) and the European Registry for Patients with Mechanical Circulatory Support (EUROMACS), have different definitions for RVF. Furthermore, individual studies, such as the pivotal MOMENTUM-3 trial,⁴ may use their own definitions for RVF. The absence of standardized definitions limits reproducibility and hinders valid comparisons of data.²

A second challenge is that RVF remains difficult to predict. Numerous clinical, hemodynamic, and echocardiographic parameters have been proposed for prediction of RVF, including elevated central venous pressure (CVP >15 mm Hg), elevated ratio of right-sided to left-sided filling pressures (CVP/pulmonary capillary wedge pressure >0.63), pulmonary artery pulsatility index (<1.85), and various echocardiographic parameters. Various models have been developed with these parameters. Unfortunately, the best prediction models perform little better than a coin toss for predicting RVF when subjected to independent validation.⁵

Pathophysiology of RVF

Right ventricular failure results from a multitude of factors after LVAD implantation. First, patients presenting for LVAD placement may have preexisting right ventricular (RV) dysfunction. Second, RV contractility is negatively affected by the loss of interventricular independence as soon as the pericardium is opened,⁶ by the leftward shift of the interventricular septum (IVS),^7,8 by impairment of the oblique contraction of the IVS due to the physical presence of the LVAD inflow cannula,⁸ and by ventricular mechanical dyssynchrony,⁹ all of which occur with LVAD implantation. Furthermore, increases in systemic cardiac output from the LVAD increase venous return, leading to volume overload of the right ventricle.

The beleaguered RV is also subject to additional perioperative stresses, such as volume loading from blood transfusions to treat anemia and coagulopathy, exposure to cardiopulmonary bypass, air embolism, hypotension leading to hypoperfusion, and ischemic time if aortic cross-clamping is required for concomitant surgical procedures.

Early Diagnosis and Management of RVF

The intraoperative diagnosis of early RVF should rely on numerous streams of information, including the patient monitor (eg, blood pressure, CVP, and cardiac output); the LVAD console screen (which can provide information about estimated LVAD flows and, in extreme cases of RVF, show suction alarms); and, most importantly, intraoperative transesophageal echocardiography. Clinical signs, such as hypotension with elevated CVP, may be seen in RVF but are nonspecific. Therefore, transesophageal echocardiography becomes the most important tool for assessment of RV function. Unlike diagnostic echocardiographic studies, which show function at a single time point, intraoperative echocardiographic findings are dynamic. The intraoperative transesophageal echocardiography examination should be recursive with a continual assessment and reassessment of RV size and RV function (using parameters such as the tricuspid annular systolic plane excursion, RV fractional area change, and RV s′ velocity) as well as severity of tricuspid regurgitation and position of the IVS.

Initial management of RVF has several facets. The first goal is to optimize preload. This may be achieved with aggressive diuresis, continuous venovenous hemodialysis, and, if the cardiopulmonary bypass cannulas are in place, direct removal of volume to the venous reservoir. Second is augmentation of RV contractility with inotropic agents such as epinephrine, dobutamine, and milrinone. Right ventricular afterload may be reduced by using pulmonary vasodilators, and though evidence for this is only based on case series,¹⁰ this is the author's universal practice. Another modifiable parameter is LVAD pump speed, which can be decreased to improve the position of the IVS. Finally, for severe and/or refractory RVF, several right ventricular assist devices (RVADs) are available. Options for mechanical support of the RV include the Impella RP (Abiomed); the PROTEK Duo (LivaNova); or a centrifugal pump, such as Centrimag (Abbott) or RotaFlow (Getinge), which connects to the right atrium and pulmonary artery. Venoarterial extracorporeal membrane oxygenation is also an option.

For patients who have a struggling right ventricle after LVAD implantation, decisions about the necessity and timing of RV mechanical circulatory support as opposed to medical management alone are important. It should be noted that need for RVAD is associated with elevated mortality following LVAD placement,³ and long-term options are limited. Available guidelines unfortunately do not delineate specific criteria for intraoperative RVAD placement,^10,11 and data supporting 1 approach over another are extremely limited. However, when RVAD support is indicated, guidelines recommend expeditious decision-making to avoid irreversible end-organ complications.¹⁰ Given these considerations, a proposed intraoperative decision algorithm is presented in Figure 1.

View Figure

• Download as Powerpoint
• Download Figure

Future Directions

Despite an improvement in our understanding of RVF over the past several decades, it remains a common complication following LVAD implantation and comes with considerable consequences. There is a need for a standardized definition of RVF and improved predictive models that may help with patient selection. Finally, we need improved treatment options for durable biventricular support. Work is ongoing at The Texas Heart Institute in this regard, the results of which this author looks forward to with great anticipation.