Applications of Implantable Hemodynamic Monitoring in the Setting of Durable Mechanical Circulatory Support

Intravascular Pressure Monitoring in Patients with Advanced Heart Failure The use of wireless, battery-free, implantable, hemodynamic sensors to assess intravascular pressures has been demonstrated to improve clinical outcomes for patients with advanced heart failure and has now become incorporated into general clinical practice.1,2 Progress and advancement in this field has largely resulted from the investigational and clinical use of the CardioMEMS HF System (Abbott Laboratories, Abbott Park, IL).1,2 The CardioMEMS device is a wireless, battery-free, intravascular pressure sensor percutaneously positioned in the pulmonary artery (PA) to measure PA pressure (PAP) and is currently the only intravascular pressure sensor approved by the US Food and Drug Administration (FDA) for use in patients with heart failure.1,2 With the success of the CardioMEMS device, there has been significant development of a number of important competing technologies such as the Titan Wireless Implantable Hemodynamic Monitor (WIHM)3 (Integrated Sensing Systems, Inc., Ypsilanti, MI) and V-LAP system2 (Vectorious Medical Technologies, Tel Aviv, Israel). These systems permit direct assessment of left atrial (LA) pressure (V-LAP and Titan systems) or assessment of pressures within multiple intracardiac chambers (Titan system). Intravascular Pressure Monitoring in Patients Supported with Durable Mechanical Circulatory Support Devices There is significant interest and ongoing work to integrate wireless, battery-free, implantable hemodynamic sensors into the field of mechanical circulatory support (MCS), first as a platform to improve our understanding and management of hemodynamic changes in mechanical circulatory support device (MCSD)-supported patients and potentially as component of MCSDs themselves to serve as a signal input (hemodynamic pressures) into physiologic control algorithms to control pump speed. In this issue of the journal, two important articles address the potential of implantable hemodynamic sensors in these areas. The first article by Feldman et al.4 demonstrates the use of wireless implantable hemodynamic monitoring to improve timing of intervention with MCSD therapy and improve clinical outcomes after MCSD implant. The second article by Wang et al.5 validates the use of a pressure signal input in developing a physiologic control algorithm in a modeled biventricular support device application that maintains differential pump pressure across the left ventricular assist device (LVAD) and right ventricular assist device (RVAD) to achieve flow balancing within the system. To date, little clinical information exists on the use of intravascular pressure monitoring to affect clinical management in patients supported with a durable MCSD. However, recent studies have observed that approximately 20% to 30% of ambulatory patients supported with a durable LVAD have abnormal resting hemodynamics, suggesting that routine extension of serial, wireless, implantable hemodynamic monitoring to this group of patients would likely be of clinical benefit.6,7 Feldman et al.4 reported on a subgroup of 27 patients (15 treatment and 12 control patients) enrolled into “The CardioMEMS Heart Sensor Allows Monitoring of Pressure to Improve Outcomes in NYHA Class III Heart Failure Patients” (CHAMPION) trial who underwent LVAD implantation as part of their routine clinical care after randomization to the CardioMEMS device or control.1,2 These investigators importantly observed that use of serial hemodynamic monitoring with the CardioMEMS device (treatment group) was associated with: 1) a trend toward shorter length of time to LVAD implantation in the treatment group; 2) a reduction in degree of renal insufficiency at the time of LVAD implant in the treatment group; 3) a greater reduction in PA systolic pressure, PA diastolic pressure, and PA mean pressure in the treatment group compared with control after LVAD implant; and 4) a greater proportion of patients bridged from the LVAD to heart transplantation in the treatment group compared with control. Despite its success in reducing heart failure-related hospitalizations in advanced heart failure patients, intravascular pressure monitoring from a single location within the PA has potential limitations in general, and in its applications for use in the clinical management of LVAD patients, specifically. This technology represents an indirect assessment of left-sided cardiac filling pressures. Pulmonary artery pressure assessment is only an indirect measure of LA and LV pressures, and it may not accurately reflect LV preload in the presence of primary pulmonary vascular disease. Direct LA pressure monitoring also contributes important information regarding degree of mitral regurgitation, and there is direct correlation between mitral regurgitation and LA pressure. The enhanced LA pressure V-wave is generally not appreciated from the PAP waveform. LA pressure monitoring can also detect and discern different types of atrial arrhythmias. Further, the CardioMEMS technology does not permit multi-site monitoring of right- and left-sided cardiac chambers. A decline in PAP may reflect either a reduction in LV filling pressures or, potentially, a decrease in RV performance. Optimization of RV and LV unloading during LVAD support optimally requires knowledge of both left and right heart preload. LVAD speed adjustments aimed at reducing LV filling pressures as the only target goal could potentially have adverse consequences on RV systolic function, resulting in clinical RV failure or reduced systemic flow. Alternative approaches of intravascular pressure monitoring in LVAD patients include direct placement of pressure sensors within cardiac chambers. Recent first-in-human studies from Sweden3 have reported on implantation of the Titan WIHM in four patients undergoing concomitant implantation of a durable LVAD. The Titan device was implanted in the right superior pulmonary vein to provide direct measurement of LA pressures. A ramp-speed test was performed in three subjects, postoperatively. Left atrial pressures obtained from the Titan WIHM system were measured concomitantly with systemic blood pressure measured with Doppler, pump parameters, LV end-diastolic diameter, degree of mitral regurgitation, aortic opening and cusps closure rate, and LA diameter. There were significant correlations between sensor pressure, pump speed, and LV end-diastolic and end-systolic dimensions. In one subject, the sensor identified a rise in the LA pressure of ≈10 mm Hg during home monitoring. Further subject evaluation revealed evidence of hemolysis leading to a diagnosis of suspected pump thrombosis of the LVAD despite a normal echocardiographic examination. At the time of operation for LVAD exchange, evidence of thrombus was verified within the pump.3 Importantly, this technology lends itself to implantation of sensors in multiple cardiac chambers to simultaneously measure right- and left-sided pressures. Important insights into the physiologic interactions of LV unloading and RV performance during rest and exercise and during prolonged LVAD support could be ascertained from such measurements. More precise targets of LV unloading could be determined with simultaneous measurements of right-sided pressures. Limitations with intracardiac sensors include a systemic thromboembolic risk that is not present with pressure monitoring from the PA, and sensors would be less accessible for explant in the setting of device infection. The Application of Implantable Hemodynamic Sensors in Physiologic Control Algorithms for Mechanical Circulatory Support Devices Durable MCSD support is associated with adverse physiologic sequelae that, in large part, represent the limitations of the current design of clinical MCSDs operating in a “fixed speed” mode algorithm that cause diminished arterial pulse pressure.8–11 Further, this operating algorithm requires frequent provider interaction and assessment with clinical, echocardiographic, or right heart catheterization data to determine appropriate pump speeds necessary for clinical optimization. Changes in pump flow in continuous-flow rotary pumps can occur without changes in pump speed (“fixed speed” mode; reliance on the Starling mechanism to alter pump output) or occur as a consequence of changes in pump speed based upon an algorithm-based adjustment from a signal input (physiologic flow adjustments).12 The absence of autonomous speed control in MCSDs eliminates the potential for creating pulsatility matching that of the normal physiologic state or significantly hinders the ability of current MCSDs to respond to widely varying changes in preload and afterload conditions. Although, current clinical MCSDs possess hydrodynamic properties (Starling mechanism) that permit changes in pump output to altering preload and afterload conditions, this response, under some circumstances, is not appropriate as the hydrodynamic properties of continuous-flow rotary pumps are less sensitive to changes in preload (i.e., small changes in pump output relative to large changes in preload) and more sensitive to changes in afterload (i.e., large changes in pump flow relative to small changes in afterload) when compared with the native heart.13 Thus, more appropriate changes in pump output are limited without alteration of pump speed. This “uninformed” mode of speed operation limits continuous-flow rotary pumps from autonomously and appropriately responding to widely varying changes in loading conditions and precludes: 1) optimizing preload in the setting of increased physiologic demand, 2) maintenance of adequate pump flows in times of high afterload (i.e., hypertension), 3) maintenance of adequate LV volumes in times of low preload (i.e., prevention of “suction events”), and 4) appropriate flow balancing in applications with biventricular MCSDs. Given these limitations in continuous-flow rotary pump designs, there is a critical need to develop speed management algorithms into future MCSD designs to achieve “intelligent” pump designs that autonomously control pump speed. Pressure signals from implantable hemodynamic sensors represent an important source for input data. This approach is highlighted by Wang et al.5 who modeled the use of control algorithm incorporating two proportional integral controllers for LVAD and RVAD that were designed to maintain differential pump pressure across the LVAD and RVAD to provide left–right balance and physiologic flow. This mode used pressure sensors for the measurement of VAD inlet and outlet pressures to calculate differential pump pressure for left and right devices. This BiVAD control algorithm successfully diminished LVAD flow in response to augmentation in pulmonary vascular resistance (reduced preload for the LVAD) and RVAD flow rates with augmentation in vena cavae resistance (reduced preload for the RVAD). The BiVAD control algorithm simultaneously reduced LVAD and RVAD flow rates when transitioning from exercise to rest and prevented suction during a rapid increase in pulmonary vascular resistance. Although the integration of pressure sensors into MCSDs to provide input into speed algorithms represents one source of many potential types of signal inputs, there are a number of important limitations in using wireless, implantable pressure sensors. The issue of sensor drift and reliability is an important concern with this technology. Drift in sensor performance would adversely impact the safety profile of pressure-based speed algorithms. In addition to drift, the reliability of current sensors to perform over years as is expected with an LVAD is unknown. The use of wireless, implantable hemodynamic sensors is being integrated into the clinical realm of advanced heart failure and in patients supported with durable MCSDs. This technology will likely provide immediate benefit with respect to patient management and clinical outcomes and could possibly contribute to unique control algorithms for MCSDs in the near future.