Right Ventricular Assist Device Support After Continuous Flow Left Ventricular Assist Device Implantation: Predicting a Poor Predictor

Abstracts S237 Conclusion: Microvascular density and angiogenesis is higher following CF-LVAD support. Additional studies are warranted to investigate whether an increase in microvascular density represents a prerequisite for successful “myocardial recovery” while on CF-LVAD support.

6( 39) Effect of Continuous-Flow LVADs on Pulmonary Arterial Pressure: Comparison Between Axial and Centrifugal Flow Devices M.F. Jumean ,1 P.A. Quintero,1 A.R. Patel,1 D. DeNofrio,1 K.G. Warner,2 D.T. Pham,2 M.S. Kiernan.1  1Cardiology, Tufts Medical Center, Boston, MA; 2Cardiothoracic Surgery, Tufts Medical Center, Boston, MA. Purpose: Pulmonary hypertension (PH) is a risk factor for poor outcomes following cardiac transplantation. We sought to investigate whether unloading of the left ventricle and subsequent improvement in PH differs with implantation of axial versus centrifugal left ventricular assist device (LVAD). Methods: From April 2008 through January 2013, consecutive patients (pts) undergoing LVAD implantation at a single center were included. Pts undergoing RVAD or who died in the early post-operative period were excluded. Pre-operative and 3 month follow up trans-thoracic and trans-esophageal echocardiography studies were reviewed by two independent reviewers. Pulmonary artery systolic pressure (PASP) was estimated non-invasively based on American Society of Echocardiography guidelines. Results: A total of 125 pts underwent continuous flow LVAD placement; 12 were excluded due to concomitant RVAD placement or death prior to discharge. A total of 113 pts (mean age 55 (±12), 77% males, 38% had ischemic cardiomyopathy) were included with a median follow up of 138 days (IQ 55.3-290.3 days), 90 patients (79.5%) received a Heartmate II (HMII, Thoratec) while 23 (20.5%) underwent HVAD (Heartware) implantation. Mean preoperative PASP prior to LVAD was 45.2mmHg (±11.3) in HMII vs. 44.5 mmHg (±12.6) in HVAD pts (p= NS). At follow-up, mean PASP was 34.1mmHg (±10.9) in HMI and 24.1mmHg (±9.3) in HVAD patients (p= 0.002). Mean PASP decreased by 11.5mmHg in HMII group (±14.4, n= 77) vs. 18.7mmHg (±7.2, n= 12) in HVAD patients (p 0.01). Conclusion: Despite reported differences in the HQ curves and hemodynamic unloading between axial and centrifugal pump configurations, we found similar reductions in estimated PASP following LVAD implantation. Both types of continuous flow LVAVDs appear to improve pulmonary hypertension.

Pittsburgh, PA; 2Departments of Psychiatry, Epidemiology, & Biostatistic, University of Pittsburgh Medical Center, Pittsburgh, PA. Purpose: Right ventricular failure (RVF) after left ventricular assist device (LVAD) placement is often perilous. It has been hypothesized that in patients with advanced heart failure, RVF is linked to weight loss and measures of abnormal body composition, such as low body mass index (BMI). This postulate is examined in a cohort of LVAD supported patients. Methods: A retrospective analysis was performed of 114 continuous flow LVAD supported patients from 2004-2011. RVF was defined as need for right ventricular mechanical support or > 14 days of inotropic support postimplant. Three patient groups were identified based on the sample’s BMI distribution: lowest group (lowest quintile, BMI<  22.9, n= 23), intermediate group (middle quintiles, BMI 23 - 32.9, n= 68) and highest group (highest quintile, BMI >  33, n= 23). Kaplan-Meier analysis was used to compare the freedom from RVF; multivariate Cox- regression analysis was performed to determine risk factors for RVF. Results: The BMI groups (lowest, intermediate, highest) were similar with regards to all pre-operative characteristics except age (56y, 59y, 47y, p= 0.002), non-ischemic/idiopathic diagnosis (22%, 1%, 9%, p= 0.004), emergent LVAD implantation (0%, 4%, 17%, p= 0.032), and pre-implant heart rate ( 104, 85, 95, p= 0.001). Figure 1 shows that freedom from RVF was significantly worse for the lowest BMI group compared to the intermediate BMI group (p= 0.019) and the intermediate and highest BMI groups combined (p= 0.021). There was no difference between the intermediate and highest BMI groups (p= 0.500) or the lowest and highest BMI groups (p= 0.185). Cox- regression analysis revealed lowest BMI group (HR 2.519, p= 0.016, CI 1.187-5.347) and emergent device placement (HR 3.392, p= 0.028, CI 1.141-10.09) were independent risk factors for RVF. Conclusion: In LVAD patients, a low BMI is associated with heightened risk for RVF after device placement. Investigation of other potential markers of abnormal body composition may broaden our understanding of post-LVAD RVF.

6( 41) Right Ventricular Assist Device Support After Continuous Flow Left Ventricular Assist Device Implantation: Predicting a Poor Predictor S.K. Singh ,1 R. Hernandez,2 J. Anand,1 A. Baldwin,2 E. Sandoval Martinez,2 W.E. Cohn,2 O.H. Frazier,2 H.R. Mallidi.1  1Surgery, Baylor College of Medicine, Houston, TX; 2Centre for Cardiac Support, Texas Heart Institute, Houston, TX.

6( 40) Can Body Mass Index Identify Patients at Risk for Right Ventricular Failure after LVAD Placement? J.K. Bhama ,1 M.A. Dew,2 J.J. Teuteberg,1 J.A. Hayanga,1 L. Huffman,1 S. Bansal,1 M. Simon,1 C.A. Bermudez,1 T. Richards,1 R.L. Kormos.1   1Heart & Vascular Institute, University of Pittsburgh Medical Center,

Purpose: Demographic, perioperative and late clinical outcomes in patients who require right ventricular assist device (RVAD) support after left ventricular assist device (LVAD) implantation, has not been well described. Predictive tools of RVAD after LVAD are based on univariate models from combined pulsatile and continuous flow (CF) VAD patients. Multivariable models have been inconsistent in predictors elucidated. We aimed to review contemporary CF LVADs with long term follow-up, to determine the predictors of RVAD need after CF LVAD, and the impact on mid and long term survival.

S238

The Journal of Heart and Lung Transplantation, Vol 32, No 4S, April 2014

Methods: A retrospective review of a single center was done. All HeartMate II CF LVADs (2003-2012) were identified. Patients were stratified by whether they received a RVAD after LVAD implant. Descriptive univariate statistics, multivariate regression and survival analysis were performed. Results: We identified 294 patients; 26 (8.8%) required a RVAD. Mean follow up was 2.5 years (maximum 8 years). Late survival was worse in those who received a RVAD (15% vs 76%, p< 0.001). Receiving an RVAD was an independent predictor of mortality (HR 2.4, 95% CI 3.7-40, p< 0.001). Patients who required a RVAD had similar demographics and preoperative characteristics, however more likely to be: bridge to transplant (69% vs 46%, p= 0.03), intubated pre-operatively (31% vs 12%, p= 0.008), on temporary circulatory support preop (27% vs 8%, p= 0.09) and had higher preoperative bilirubin and AST. RVAD patients more often required redo sternotomy (69% vs 39%, p= 0.003). In-hospital outcomes included longer ventilator support (30+50 vs 9+17 days, p< 0.001). Multivariable regression analysis found significant independent predictors of getting a RVAD after LVAD implant to be redo sternotomy (HR 1.5, 95% CI 1.3-9.3, p= 0.002) and pre-operative temporary circulatory support (HR 1.3, 95% CI 1.2-9.6, p= 0.01). Conclusion: This is the largest series of patients with contemporary CF LVADs and long term follow-up. RVAD requirement after LVAD purports very poor early, mid and long term survival. Redo sternotomy and pre-operative temporary circulatory support were independent predictors of RVAD requirement, which along with univariate variables identified above, may be used to identify and stratify this high risk cohort. 6( 42) Isolated RVAD Implantation - First Results from the EUROMACS Registry A.M. Bernhardt ,1 T.M. De By,2 H. Reichenspurner,1 T. Deuse.1   1Department of Cardiovascular Surgery, University Heart Center Hamburg, Hamburg, Germany; 2EUROMACS, Berlin, Germany. Purpose: Once, right ventricular dysfunction occurs, it is difficult to treat and associated with high mortality. Temporary short-term devices as right ventricular assist devices (RVAD) are available and have been used for temporary right heart support. In some patients RV function does not recover and long-term devices are needed. Recently, isolated RVAD implantation with HeartWare HVAD devices has been reported. However, isolated implantation on the right side is rare and still an off-label use for this device. Therefore we analysed data from the EUROMACS database in which procedures and outcomes data for patients receiving mechanical circulatory support are registered. This registry started in 2012 with the goal of running a European registry for data from patients with mechanical circulatory support systems and first results are analyzed. Methods: Until September 2013 data of 7 patients (mean age 60.6 ± 7.8 years, 100% male) with an isolated RVAD were submitted to the EUROMACS registry. All seven patients received a HeartWare HVAD as an RVAD. Two patients (28.6%) were in INTERMACS class 1, three patients (42.9%) were in INTERMACS class and two patients (28.6%) in INTERMACS class 3. Device strategy was rescue therapy in six patients (85.7%) and destination therapy in one patient (14.3%). Indications for RVAD placement was acute myocardial infarction in two (28.6%), failure to wean from cardiopulmonary bypass in two (28.6%) and post-cardiotomy syndrome in three patients (42.9%). Intra- and postoperative results of the EUROMACS registry were analysed. Results: Inflow cannulas were implanted into the RA in five patients (71.4%) and in the RV in two patients (28.6%). Mean time in OR was 315.7 ± 212.3 min. Mean RVAD flow was 4.6 ± 1.1 l/min at a pump speed of 2560± 628 rpm and a power of 3.9 ± 1.3 W. Four patients (57.1 %) survived the first 30 days. During follow-up one patient died after 44 days due to multiorgan failure. In the surviving three patients, one patient was transplanted after 419 days, one device was explanted after 40 days due to pump thrombosis and recovered RV function, and one patient is still on the device after 494 days. Conclusion: In this very specific and sick patient cohort within the EUROMACS registry, isolated RVAD implantation is associated with acceptable results. Further studies on this subject are necessary using MCS registries such as EUROMACS or IMACS.

6( 43) Banding of Right Ventricular Assist Device (RVAD) of Outflow Conduit: Is It Really Necessary With Current Devices? C. Lo ,1 S. Gregory,2 M. Stevens,3 D. Murphy,4 S. Marasco.1  1Sir Charles James Officer Brown Department of Cardiothoracic Surgery, Alfred Hospital, Prahran, Victoria, Australia; 23Innovative Cardiovascular Engineering and Technology Laboratory, The Prince Charles Hospital, Brisbane, Queensland, Australia; 3Innovative Cardiovascular Engineering and Technology Laboratory, The Prince Charles Hospital, Brisbane, Queensland, Australia; 4Department of Intensive Care, Alfred Hospital, Prahran, Victoria, Australia. Purpose: Current available implantable centrifugal ventricular assist devices (VADs) are designed to pump against left-sided afterload. For right ventricular (RV) support, VADs are adapted for the lower pressures by restricting outflow cannula diameter, simulating systemic afterload. Previous studies were thus designed around this assumption, assessing the degree of outflow cannula restriction required to simulate left-sided afterload. In this in vitro study, we observed the use of dual Heartware HVAD™ (Framingham, MA) devices in BiVAD configuration on a mock-circulation loop (MCL), varying outflow cannula configuration to determine its suitability for dual device support and the optimal settings. Methods: A Windkessel MCL including systemic and pulmonary circulations was used for this study. Physiological parameters and settings were based on those established in previous studies. Severe LV failure was simulated and supported using a Heartware HVAD device at 3000 RPM to maintain 5L/min aortic flow and mean aortic pressure of 80mmHg. Severe RV failure was simulated and supported using a second VAD, with unrestricted 10mm diameter outflow cannulae of 20, 40 and 60cm lengths and a 20cm length cannula restricted to 5mm diameter. In each test, pulmonary vascular resistance (PVR) was successively increased from 40 to 560 Dynes.s.cm-S in 40 Dynes.s.cm-S intervals, and flow dynamics were measured when a device speed sufficient to maintain balanced systemic and pulmonary flow rates was established. Results: HVAD speeds were maintained within manufacturer’s recommendations. It was observed that RVAD flow at the outlet (RVADQ) was consistently higher in the higher resistance settings. Power consumption and pressure gradient across the outflow conduit were also significantly higher. However, there appeared to be sufficient residual RV capacity to maintain equivalent pulmonary artery flow (PAQ) across all tested outflow resistance settings of the RVAD. Conclusion: In these simulations, the Heartware HVAD device as an RVAD was able to safelymaintain physiological PAQ without outflow cannula restriction in supporting a severely compromised RV. While results of this type of study cannot be directly translated to clinical use, our results suggest that further testing of the viability this device as an RVAD without outflow restriction is warranted. 6( 44) RV Fractional Area Change at 6 Months But Not at Pre-Implant Is Associated with Late Indices of Functional Capacity and Renal Function: A Case for the Importance of Late RV Dysfunction in Long Term MCS J. Rame ,1 E.Y. Birati,1 V. Ramjee,1 E. Laverty-Wilson,1 T. Plappert,1 J. Donnelly,1 J. Wald,1 P. Atluri,2 Y. Woo,2 M. Acker,2 J. Kirkpatrick.1   1Cardiology, Hospital of the University of Pennsylvania, Philadelphia, PA; 2Hospital of the University of Pennsylvania, Philadelphia, PA. Purpose: Right heart failure (RHF) post - LVAD implantation continues to present a great challenge in the perioperative period, however few studies address the importance of right ventricular function (RVF) in the later phases of mechanical support. Given the evidence for recoupling of RVF after successful mechanical unloading with LVAD, we hypothesized that late RVF, as measured by RV Fractional area change (RVFAC) would be associated with clinically important indices of end-organ viability. Methods: In this retrospective single center study, RVF was quantitatively analyzed by a core lab in paired pre-implant and post-implant (6 month to 1 year) echocardiograms in patients who underwent LV-continuous flow device support as Destination Therapy or Bridge to Decision. Long term data on 6 MW (meters) and renal function (MDRD GFR) were retrospectively