Bridging to a Long‐Term Ventricular Assist Device With Short‐Term Mechanical Circulatory Support

Implanting short‐term mechanical circulatory support (MCS) devices as a bridge‐to‐decision is increasingly popular. However, outcomes have not been well studied in patients who receive short‐term MCS before receiving long‐term left ventricular assist device (LVAD) support. We analyzed outcomes in our single‐center experience with long‐term continuous‐flow (CF)‐LVAD recipients with pre‐implantation short‐term MCS. From November 2003 through March 2016, 526 patients (mean age, 54.7 ± 13.5 years) with chronic heart failure (mean ejection fraction, 21.7 ± 3.6%) underwent implantation of either the HeartMate II (n = 403) or the HeartWare device (n = 123). Before implantation, 269 patients received short‐term MCS with the TandemHeart, the Impella 2.5/5.0, an intra‐aortic balloon pump (IABP), venoarterial extracorporeal membrane oxygenation (VA‐ECMO), or the CentriMag. The short‐term MCS patients were compared with the CF‐LVAD–only patients regarding preoperative demographics, incidence of postoperative complications, and long‐term survival. The 269 patients received the following short‐term MCS devices: 57 TandemHeart, 27 Impella, 172 IABP, 12 VA‐ECMO, and 1 CentriMag. Survival at 30 days, 6 months, 1 year, and 2 years was 94.2, 87.2, 79.4, and 72.4%, respectively, for CF‐LVAD–only patients versus 91.0, 78.1, 73.4, and 65.6%, respectively, for short‐term MCS + CF‐LVAD patients (P = 0.17). Within the short‐term MCS group, survival at 24 months was poorest for patients supported with VA‐ECMO or the TandemHeart (P = 0.03 for both), and survival across all four time points was poorest for patients supported with VA‐ECMO (P = 0.02). Short‐term MCS was not an independent predictor of mortality in multivariate Cox regression models (hazard ratio = 1.12, 95% confidence interval = 0.84–1.49, P = 0.43). In conclusion, we found that using short‐term MCS therapy—except for VA‐ECMO—as a bridge to long‐term CF‐LVAD support was not associated with poorer survival.     

A Physiological Controller for Turbodynamic Ventricular Assist Devices Based on Left Ventricular Systolic Pressure

The current article presents a novel physiological feedback controller for turbodynamic ventricular assist devices (tVADs). This controller is based on the recording of the left ventricular (LV) pressure measured at the inlet cannula of a tVAD thus requiring only one pressure sensor. The LV systolic pressure (SP) is proposed as an indicator to determine the varying perfusion requirements. The algorithm to extract the SP from the pump inlet pressure signal used for the controller to adjust the speed of the tVAD shows robust behavior. Its performance was evaluated on a hybrid mock circulation. The experiments with changing perfusion requirements were compared with a physiological circulation and a pathological one assisted with a tVAD operated at constant speed. A sensitivity analysis of the controller parameters was conducted to identify their limits and their influence on a circulation. The performance of the proposed SP controller was evaluated for various values of LV contractility, as well as for a simulated pressure sensor drift. The response of a pathological circulation assisted by a tVAD controlled by the introduced SP controller matched the physiological circulation well, while over‐ and underpumping events were eliminated. The controller presented a robust performance during experiments with simulated pressure sensor drift.     

Changes in Total Cardiac Output and Oxygen Extraction During Exercise in Patients Supported With an HVAD Left Ventricular Assist Device

Following implantation of a left ventricular assist device (LVAD), acceptable functional performance is now being achieved; however, peak VO₂ and peak work load (watts) remain considerably limited. Maximum physical capacity is essentially dependent on generated cardiac output (CO) and arteriovenous oxygen difference (avDO₂). We investigated the changes in CO and avDO₂ during exercise in LVAD patients with an HVAD pump (HeartWare Inc., Framingham, MA, USA). Approximately 6 weeks after implantation, 20 patients (100% male, 60.8 ± 7.3 years old, BMI 25.7 ± 3.3) underwent a six‐minute walk test (6MWT), a cardiopulmonary exercise test (CPET), and noninvasive hemodynamic measurement. The mean six‐minute walking distance (6MWD) was 403 m (68% of predicted), and mean peak VO₂ was 10.9 mL/kg/min (39% of predicted). Mean total CO improved from 3.8 L at rest to 7.0 L at maximum exercise. The mean avDO₂ increased from 7.4 mL/dL (44% of oxygen content) at rest to 13.2 mL/dL (75% of oxygen content) at maximum exercise. There was a significant increase in both total CO (P < 0.01) and avDO₂ (P < 0.05) between rest and sub‐maximum exercise. As exercise levels increased, however, no further significant changes were achieved. Long‐term studies, especially in combination with exercise programs, would be desirable in order to observe the development of these parameters.     

Impella 5.0 as a Second‐Line Mechanical Circulatory Support Strategy After Extracorporeal Life Support

The catheter‐based Impella 5.0 left ventricular assist device is a powerful and less invasive alternative for patients in cardiogenic shock. The use as second‐line therapy in patients with precedent extracorporeal life support (ECLS) has not been described before now. We analyzed our experience of consecutive patients treated with this alternative strategy. From April 2014 to December 2014, eight patients had been implanted as a second‐line option after ECLS support. The reason for the change from ECLS to Impella 5.0 was absence of cardiac recovery for primary weaning and complications of ECLS therapy. The mean time of ECLS support prior to Impella implantation was 12 ± 7 days. The implantation of the Impella 5.0/CP was technically successful in all patients, and the ECLS could be explanted in all eight patients who received Impella implantation as a second‐line treatment. The second‐line Impella 5.0 therapy resulted in two patients who turned into left ventricular assist device (LVAD) candidates, two primary weaning candidates, and four patients who died in the setting of sepsis or absent cardiac recovery and contraindications for durable LVAD therapy. Thereby, the overall hospital discharge survival as well as the 180‐day survival was 50% for Impella 5.0 implantations as second‐line procedure after ECLS. The latest follow‐up survival of this second‐line strategy after ECLS was three out of eight, as one patient died after 299 days of LVAD support due to sepsis. The use of Impella 5.0 constitutes a possible second‐line therapeutic option for those patients who do not show cardiac recovery during prolonged ECLS support or suffer from complications of ECLS therapy. This treatment allows additional time for decisions regarding cardiac recovery or indication for durable LVAD therapy.     

Meta‐Analysis of Time in Therapeutic Range in Continuous‐Flow Left Ventricular Assist Device Patients Receiving Warfarin

Continuous‐flow left ventricular assist devices (CF‐LVADs) prolong survival in advanced heart failure patients. Anticoagulation control is critical in CF‐LVAD patients due to increased thromboembolic and bleeding risk. We assessed the quality of INR control in CF‐LVAD patients measured by time in therapeutic range (TTR). We performed a systematic literature search of MEDLINE and SCOPUS through July 2017 to identify studies evaluating TTR in anticoagulated adult CF‐LVAD patients. Data on key characteristics and the TTR end point were then extracted from each study by two investigators using a standardized tool. Using a Hartung‐Knapp random effects model, a weighted mean TTR estimate with accompanying 95% confidence interval (CI) was calculated. Statistical heterogeneity was estimated using the I² statistic. Five published studies were included. All studies were single‐center, retrospective investigations that calculated TTR using the Rosendaal method. Sample sizes ranged from 11 to 115 patients (total of 270 patients) with durations of follow‐up ranging from 9 to 76 person‐years. On meta‐analysis, CF‐LVAD patients had a weighted mean TTR of 46.6% (95% CI: 36.0–57.3%, I² = 94%). This suggests that warfarin is difficult to manage in CF‐LVAD patients, which may contribute to high rates of bleeding and thromboembolic complications.     

Single‐Center Experience With HeartMate II Left Ventricular Assist Device Explantation

In patients with continuous flow left ventricular assist devices (CF‐LVADs) myocardial recovery is uncommon. Given the heterogeneity of the population implanted and low incidence of recovery, the discovery of native left ventricular (LV) recovery and criteria for explantation of CF‐LVAD system is not clearly determined. We sought to analyze the characteristics of the patients who underwent CF‐LVAD explantation at our institution. Prospectively collected data on patients supported with CF‐LVADs were reviewed retrospectively. Patients who underwent CF‐LVAD explants were identified and their characteristics were analyzed with a focus on patient presentation and determinants of explantability. From November 2006 to June 2014, 223 patients (181 male, 42 female) underwent implantation of HeartMate II LVAD. Seven female (16.7%) and one male (0.6%) patients were explanted (P < 0.001). Mean age was 43 ± 9 years and etiology for cardiomyopathy was ischemic in three (37.5%) patients, nonischemic in four (50%) patients, and mixed in the one (12.5%) male patient of the cohort. Five (62.5%) patients presented acutely with significant hemolysis, and were found to have LV improvement as well as reduced, absent, or reversed diastolic flow velocities on echocardiography. Overall, mean lactate dehydrogenase level before explantation was 1709 ± 1168 U/L compared to the mean baseline level of 601 ± 316 U/L (P = 0.048). Mean LV ejection fraction (LVEF) improved from 17 ± 7% preimplant to 56 ± 11% pre‐explantation (P < 0.001). Median number of days on CF‐LVAD support was 870 (interquartile range, 209–975) while mean duration of follow‐up after the CF‐LVAD explantation was 276 ± 240 days. Mean LVEF dropped from 46 ± 19% postexplantation to 34 ± 10% during the most recent follow‐up (P = 0.015). At our institution, patients who underwent LVAD explants were predominantly women with nonischemic cardiomyopathy. Clinical evidence of hemolysis and echocardiographic evidence of reduced or absent diastolic flow velocities were common findings in these patients. Over time, patient's native LV function declined in the absence of LVAD (after LVAD explantation). Significant challenges remain in predicting LV recovery and identifying those individuals who have recovered myocardial function significant enough to be explanted.     

Derivation of Indices of Left Ventricular Contractility in the Setting of Continuous‐Flow Left Ventricular Assist Device Support

It is important to accurately monitor residual cardiac function in patients under long‐term continuous‐flow left ventricular assist device (cfLVAD) support. Two new measures of left ventricular (LV) chamber contractility in the cfLVAD‐unloaded ventricle include I_Q, a regression coefficient between maximum flow acceleration and flow pulsatility at different pump speeds; and K, a logarithmic relationship between volumes moved in systole and diastole. We sought to optimize these indices. We also propose RI_Q, a ratio between maximum flow acceleration and flow pulsatility at baseline pump speed, as an alternative to I_Q. Eleven patients (mean age 49 ± 11 years) were studied. The K index was derived at baseline pump speed by defining systolic and diastolic onset as time points at which maximum and minimum volumes move through the pump. I_Q across the full range of pump speeds was markedly different between patients. It was unreliable in three patients with underlying atrial fibrillation (coefficient of determination R² range: 0.38–0.74) and also when calculated without pump speed manipulation (R² range: 0.01–0.74). The K index was within physiological ranges, but poorly correlated to both I_Q (P = 0.42) and RI_Q (P = 0.92). In four patients there was excellent correspondence between RI_Q and I_Q, while four other patients showed a poor relationship between these indices. As RI_Q does not require pump speed changes, it may be a more clinically appropriate measure. Further studies are required to determine the validity of these indices.     

Impact of Residual Mitral Regurgitation on Right Ventricular Systolic Function After Left Ventricular Assist Device Implantation

Significant mitral regurgitation (MR) is thought to decrease after left ventricular assist device (LVAD) implantation, and therefore repair of mitral valve is not indicated in current practice. However, residual moderate and severe MR leads to pulmonary artery pressure increase, thereby resulting in right ventricular (RV) dysfunction during follow‐up. We examined the impact of residual MR on systolic function of the right ventricle by echocardiography after LVAD implantation. This study included 90 patients (mean age: 51.7 ± 10.9 years, 14.4% female) who underwent LVAD implantation (HeartMate II = 21, HeartWare = 69) in a single center between December 2010 and June 2014. Echocardiograms obtained at 3–6 months and over after implantation were analyzed retrospectively. RV systolic function was graded as normal, mild, moderate, and severely depressed. MR (≥moderate) was observed in 43 and 44% of patients at early and late period, respectively. Systolic function of the RV was severely depressed in 16 and 9% of all patients. Initial analysis (mean duration of support 174.3 ± 42.5 days) showed a statistically significant correlation between less MR and improved systolic function of RV (P = 0.01). Secondary echocardiographic analysis (following a mean duration of support of 435.1 ± 203 days) was also statistically significant for MR degree and RV systolic dysfunction (P = 0.008). Residual MR after LVAD implantation may cause deterioration of RV systolic function and cause right‐sided heart failure symptoms. Repair of severe MR, in selected patients such as those with severe pulmonary hypertension and depressed RV, may be considered to improve the patient's clinical course during pump support.     

Temporary Right Ventricular Mechanical Support in High‐Risk Left Ventricular Assist Device Recipients Versus Permanent Biventricular or Total Artificial Heart Support

Early planned institution of temporary right ventricular assist device (RVAD) support with the CentriMag (Levitronix LLC, Waltham, MA, USA) in left ventricular assist device (LVAD) recipients was compared with permanent biventricular assist device (BVAD) or total artificial heart (TAH) support. Between 2007 and 2011, 77 patients (age range: 25–70 years) with preoperative evidence of biventricular dysfunction (University of Pennsylvania score >50; University of Michigan score >5) were included. Forty‐six patients (38 men; median age 54.5 years, range: 25–70 years) underwent LVAD placement combined with temporary RVAD support (group A); in 31 patients (25 men; median age 56.7 years, range: 28–68 years), a permanent BVAD or TAH implantation (group B) was performed. Within 30 days, 12 patients from group A (26.08%) and 14 patients from group B (45.1%) died on mechanical support (P = 0.02). Thirty patients (65.2%) in group A were weaned from temporary RVAD support and three (6.5%) underwent permanent RVAD (HeartWare, Inc., Framingham, MA, USA) placement. A total of 26 patients (56.5%) were discharged home in group A versus 17 (54.8%) in group B (P = 0.56). Three patients (8.5%) received heart transplantation in group A and six (19.3%) in group B (P = 0.04). In group A, 90‐day and 6‐month survival was 54.3% (n = 25) versus 51.6% (n = 16) in group B (P = 0.66). In group A, 1‐year survival was 45.6% (n = 21) versus 45.1% (n = 14) in group B (P = 0.81). The strategy of planned temporary RVAD support in LVAD recipients showed encouraging results if compared with those of a similar permanent BVAD/TAH population. Weaning from and removal of the temporary RVAD support may allow patients to be on LVAD support only despite preoperative biventricular dysfunction.     

左房-主动脉转流辅助循环在心脏手术中的应用

为评估左房主动脉转流作为左心辅助循环在心内直视术后顽固性左心衰竭的应用效果，为２７例术后出现顽固性左心衰竭的病人，经左心房辅助泵主动脉建立左心辅助循环，结果２３例脱离人工心肺机，２０例痊愈出院。死亡原因主要为顽固性心衰、心律失常、出血、多器官功能衰竭等。结果：左心辅助疗效肯定，具有良好的临床使用价值     

Use of Left Ventricular Assist Device (HeartMate II): A Singapore Experience

Recent advances in medical and device therapies in heart failure have improved the survival of patients with heart failure. However, due to the limited availability of suitable heart donors, left ventricular assist devices (LVADs) have become an important tool as a bridge‐to‐heart transplantation for patients with refractory heart failure in Singapore. We report our experience with the HeartMate II (HMII) LVAD (Thoratec Corporation, Pleasanton, CA, USA) as a bridge‐to‐heart transplant in our center from 2009 to 2012. This was a retrospective review of 23 consecutive patients who underwent HMII LVAD implantation in our center between May 2009 and December 2012. All patients were classified as Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) levels 1 to 3 and underwent LVAD implantation as a bridge‐to‐heart transplant. There were 17 male and 6 female patients. The mean age was 43.6 years old (range 14 to 64). The etiologies of heart failure included ischemic heart disease [8], idiopathic dilated cardiomyopathy [11], viral myocarditis [2], and chemotherapy‐induced cardiomyopathy [2]. Nine patients were INTERMACS level 1, 12 patients level 2, and two patients level 3. All patients successfully underwent HMII LVAD implantation. There was no mortality within the first 30 postoperative days. Postoperative complications included stroke with full neurological recovery (21.7%), mediastinal infection (21.7%), cardiac tamponade or mediastinal collection requiring reopening of the chest (39.1%), cardiac arrhythmia (13.0%), and pump thrombosis with pump replacement (4.3%). All patients were discharged from hospital after LVAD implantation. Three patients experienced driveline infections during outpatient follow‐up. There were 19 readmissions due to the following conditions: sub‐therapeutic anticoagulation (13.0%), gastrointestinal bleeding (13.0%), suspected pump thrombosis (13.0%), transient ischemic attack (8.7%), arrhythmia (8.7%), congestive cardiac failure due to severe aortic regurgitation (8.7%), right ventricular failure (4.3%), rhabdomyolysis (4.3%), and hematuria (4.3%). Post‐LVAD implantation, 20 patients were functionally New York Heart Association (NYHA) class I, while 3 reported NYHA III symptoms. Three patients were successfully bridged to heart transplantation. One patient was successfully explanted 11 months after LVAD implantation. There were two mortalities during the follow‐up period. The average duration of LVAD support was 522 days (range 47 to 1316 days). The HeartMate II LVAD has proven to be effective in our Asian population. Driveline infection rate remains low even in the tropical hot, humid climate in Singapore. With more patients ending up on extended periods of LVAD support, increased emphasis in the detection and management of long‐term complications of ventricular assist devices will be needed.     

Development of a Direct Mechanical Left Ventricular Assist Device for Left Ventricular Failure

Abstract: We have developed a direct mechanical left ventricular assist device (DMLVAD) for severe left ventricular failure. The DMLVAD was attached to the left ventricle and compressed the heart by a pneumatic driving unit. In a mock circulation model with an extracted non-beating heart, a cardiac output (CO) of 1.93 L/min was obtained at a driving pressure of 200 mm Hg. In a canine left ventricular failure model induced by injection of sodium hydroxide into the myocardium, the systolic arterial pressure, systolic left ventricular pressure, maximum LV dP/dt, peak flow, and CO increased by 21, 24, 58, 144, and 37%, respectively. The mean left atrial pressure also decreased by 15% when the DMLVAD was driven. These effects were most prominent when the mean left atrial pressure was over 15 mm Hg, and the driving pressure was over 100 mm Hg. Compression at late systole was more effective in obtaining greater CO. We suggest that the DMLVAD could be an optional circulatory assist device for patients with left ventricular failure awaiting heart transplantation.     

Pulse Oximeter Derived Blood Pressure Measurement in Patients With a Continuous Flow Left Ventricular Assist Device

Currently, blood pressure (BP) measurement is obtained noninvasively in patients with continuous flow left ventricular assist device (LVAD) by placing a Doppler probe over the brachial or radial artery with inflation and deflation of a manual BP cuff. We hypothesized that replacing the Doppler probe with a finger‐based pulse oximeter can yield BP measurements similar to the Doppler derived mean arterial pressure (MAP). We conducted a prospective study consisting of patients with contemporary continuous flow LVADs. In a small pilot phase I inpatient study, we compared direct arterial line measurements with an automated blood pressure (ABP) cuff, Doppler and pulse oximeter derived MAP. Our main phase II study included LVAD outpatients with a comparison between Doppler, ABP, and pulse oximeter derived MAP. A total of five phase I and 36 phase II patients were recruited during February–June 2014. In phase I, the average MAP measured by pulse oximeter was closer to arterial line MAP rather than Doppler (P = 0.06) or ABP (P < 0.01). In phase II, pulse oximeter MAP (96.6 mm Hg) was significantly closer to Doppler MAP (96.5 mm Hg) when compared to ABP (82.1 mm Hg) (P = 0.0001). Pulse oximeter derived blood pressure measurement may be as reliable as Doppler in patients with continuous flow LVADs.     

Minimally Invasive Left Ventricular Assist Device Explantation After Cardiac Recovery: Surgical Technical Considerations

The new generation of left ventricular assist devices (LVADs) has enabled minimally invasive surgical procedures for implantation. Herein we present two alternative approaches for minimally invasive LVAD explantation following cardiac recovery, avoiding a sternotomy and improving patient safety.     

Remote Monitoring of Left Ventricular Assist Device Parameters After HeartAssist‐5 Implantation

Although several left ventricular assist devices (LVADs) have been used widely, remote monitoring of LVAD parameters has been available only recently. We present our remote monitoring experience with an axial‐flow LVAD (HeartAssist‐5, MicroMed Cardiovascular, Inc., Houston, TX, USA). Five consecutive patients who were implanted a HeartAssist‐5 LVAD because of end‐stage heart failure due to ischemic (n = 4) or idiopathic (n = 1) cardiomyopathy, and discharged from hospital between December 2011 and January 2013 were analyzed. The data (pump speed, pump flow, power consumption) obtained from clinical visits and remote monitoring were studied. During a median follow‐up of 253 (range: 80–394) days, fine tuning of LVADs was performed at clinical visits. All patients are doing well and are in New York Heart Association Class‐I/II. A total of 39 alarms were received from three patients. One patient was hospitalized for suspected thrombosis and was subjected to physical examinations as well as laboratory and echocardiographic evaluations; however, no evidence of thrombus washout or pump thrombus was found. The patient was treated conservatively. Remaining alarms were due to insufficient water intake and were resolved by increased water consumption at night and summer times, and fine tuning of pump speed. No alarms were received from the remaining two patients. We believe that remote monitoring is a useful technology for early detection and treatment of serious problems occurring out of hospital thereby improving patient care. Future developments may ease troubleshooting, provide more data from the patient and the pump, and eventually increase physician and patient satisfaction. Despite all potential clinical benefits, remote monitoring should be taken as a supplement to rather than a substitute for routine clinical visits for patient follow‐up.     

Effects of Left Ventricular Assist Device on Cardiac Function: Experimental Study of Relationship between Pump Flow and Left Ventricular Diastolic Function

The left ventricular assist device (LVAD) with centrifugal pump has two characteristics. One is a pump flow wave of the centrifugal pump, consisting of the pulsatile flow of the native heart and the nonpulsatile flow of the centrifugal pump. The other is that the centrifugal pump fills from the native heart not only in the systolic phase, but also in the diastolic phase. In the case of the apex outlet LVAD with centrifugal pump, blood flows from the left atrium through the left ventricle to the pump. Pump flow is regulated by preload, and preload is regulated by diastolic hemodynamics. The aim of this study is to analyze the relationship between pump flow and the diastolic hemodynamics of the native heart. Ten anesthetized intact pigs were studied after placement of an LVAD. Data were recorded with the LVAD off (control) and the LVAD on. The assist rate was changed to 25%, 50%, and 75%. The indexes of left ventricular (LV) diastolic function included LV myocardial relaxation (time constant of isovolumic pressure decay [Tau] and maximum negative dP/dt [LV dP/dt min]) and LV filling (peak filling rate [PFR], time to peak filling rate [tPFR], and diastolic filling time [DFT]). Stroke volume decreased significantly in 75% assist. LV end-systolic pressure decreased significantly in 50% and 75% assist. LV end-diastolic volume decreased as assist rate increased, but there were no significant changes. Stroke work decreased significantly in 50% and 75% assist. LV dP/dt min decreased significantly in 50% and 75% assist. Tau prolonged as assist rate increased, but there were no significant changes. DFT shortened significantly in 75% assist. PFR increased significantly in 75% assist. tPFR shortened significantly in 50% and 75% assist. In this study, LV relaxation delayed as an increasing of pump assist rate, but it suggested a result of reduction of cardiac work. Also, it was suggested that LVAD increases the pressure difference between the left atrium and the left ventricle in the diastolic phase. This phenomenon is due to the filling of the left ventricle. In this study it was suggested that as pump assist rate increases, it is more effective to keep cardiac function in the diastolic phase.     

Systematic Review of Phosphodiesterase‐5 Inhibitor Use in Right Ventricular Failure Following Left Ventricular Assist Device Implantation

Our aim was to identify relevant literature supporting the use of phosphodiesterase‐5 (PDE5) inhibitors in patients with persistent pulmonary hypertension with signs of postprocedural right ventricular (RV) dysfunction following left ventricular assist device (LVAD) implantation. We searched MEDLINE, SCOPUS, and Web of Science from inception through November 27, 2014 for citations evaluating patients with end‐stage heart failure necessitating LVAD, continuous and pulsatile, who received a PDE5 inhibitor to prevent RV failure. Outcomes of interest included changes in mean pulmonary artery pressure, pulmonary vascular resistance, central venous pressure, cardiac index, and mean arterial pressure. Results are presented qualitatively. Four citations (n = 83 patients) were included. These included a single case report, two retrospective case series, and a prospective open‐label study with a historical control. All four studies utilized the PDE5 inhibitor sildenafil with various doses for up to 3 months. Sildenafil routinely reduced mean pulmonary artery pressures as soon as 90 min after administration. Reductions in pulmonary vascular resistance were also seen shortly after the procedure and maintained through 12–15 weeks. While one study saw improvements in postoperative central venous pressures, another did not. Evidence supporting PDE5 inhibitor use to attenuate RV failure in patients requiring an LVAD is weak.