Subclinical Hemorrhagic Cerebral Lesions Are Prevalent in Patients With Long Term Continuous-Flow LVAD Support

Abstracts S217 

5( 80) Histopathological Analysis of the Mitral Valve After Long-Term Mechanical Circulatory Support H. Hata ,1 T. Fujita,1 H. Ishibashi-Ueda,2 T. Nakatani,3 J. Kobayashi.1   1Cardiovascular Surgery, National Cerebral and Cardiovascular Center, Suita, Japan; 2Pathology, National Cerebral and Cardiovascular Center, Suita, Japan; 3Transplantation, National Cerebral and Cardiovascular Center, Suita, Japan. Purpose: In Japan, destination therapy-use of left ventricular assisting device (LVAD) is not allowed; however, the LVAD support period of bridge to transplant-use is extremely long. The objective of this study was to investigate the pathological change of the mitral valve after long-term LVAD support. Methods: Forty-one hearts removed at heart transplantation were investigated. Five patients had undergone mitral valve repair or replacement. Thirty-four patients were supported by extracorporeal pulsatile device (Nipro-LVAD) and 7 patients by implantable devices (DuraHeart in 2 patients, EVAHEART in 2 patients, Heartmate II in 1 patient, Jarvik 2000 in 1 patient, Novacor in 1 patient). We compared the histological change of the mitral valve with the results of echocardiogram. Results: The mean duration of LVAD support until heart transplantation was 961+/-307 (ranged from 416 to 1697, median 927) days. Although the mean left ventricular end diastolic/systolic diameter (from 73/66 to 61/55mm) and left atrial diameter (from 46 to 33mm) significantly decreased and mitral regurgitation improved after LVAD implantation, the mitral annulus was still dilated (mean circumference length was 11.1cm, ranged from 8.3 to 14.5cm) in the 36 patients without mitral valve operation. Mitral leaflet became thinner and showed myxomatous degeneration with loose collagen fiber in these patients. In the patients with previous mitral annuloplasty, the mitral annular circumference length was kept normally, mean of 8.1cm. Statistically, there was no close correlation between circumference length of mitral annulus and left ventricular dimension, LVAD support duration time, the grade of mitral regurgitation, cardiac function, or etiology of heart failure. Conclusion: Although significant reverse remodeling of left ventricle was observed after long-term LVAD support, mitral annulus dilation still remained. Concomitant mitral valve surgery with LVAD implantation may be the procedure of choice. 5( 81) Medium-Term Circulatory Support for Cardiogenic Shock in a Developing Country: Do We Need a Long-Term Device? M.A. Villavicencio ,1 V. Rossel,2 R. Larrea,1 J.P. Peralta,1 E. Larrain,1 J.S. Lim,3 E. Donoso,3 F. Gajardo,3 M. Hurtado.1  1Unidad de Trasplante, Clinica Davila, Santiago, Chile; 2Departamento de Medicina, Universidad

de Chile, Santiago, Chile; 3Unidad de Trasplante, Instituto Nacional del Tórax, Santiago, Chile. Purpose: Bridge-to-transplant long-term devices are commonly not affordable for our health system and usually they are not suitable for critically ill patients. Centrimag® may be an alternative for a one-step bridge-to-transplant surgical procedure. Methods: Retrospective review of 28 patients subjected to Centrimag® implantation. Twenty-three (82%) were male, mean age 42+13 years. Etiology of heart failure was ischemic in 13(46%), dilated cardiomyopathy in 8(29%), and other 7(25%).All were INTERMACS class I, 27(96%) had multiorgan failure, and in 26(93%) the indication was failure of medical management. Twelve (43%) had previous cardiac arrest with a mean arrest time of 21+17 min. Circulatory support was biventricular: 25(89%), univentricular 3(11%), and in 9(32%) an oxygenator was interposed until respiratory recovery. Results: Thirty-day post-implant survival was 79% (22 patients). Bridge-totransplant or recovery: 20 (71%) patients. Mean support time was 40 days, and 12(43%) were supported >  4 weeks. The longest support was 292 days. Eight (29%) died before been bridged. Post-implant complications: re-exploration for bleeding 10(36%), neurologic dysfunction 4(14%), pneumonia 10(36%), bacteremia 5(36%) and acute dialysis 8(29%). Kaplan Meier two-year survival: 62+10%. Mean follow-up time: 21 months with a total of 579 months. All alive patients were in NYHA functional class I with normal biventricular function. Conclusion: Centrimag® is useful to provide medium-term support for cardiogenic shock in a developing country. Support longer than 4 weeks is feasible. A bridge-to-transplant long-term device may not be needed in the setting of a waiting time less than 3 months. 5( 82) Subclinical Hemorrhagic Cerebral Lesions Are Prevalent in Patients With Long Term Continuous-Flow LVAD Support D. Yoshioka ,1 K. Toda,1 T. Nakamura,1 S. Miyagawa,1 Y. Yoshikawa,1 S. Fukushima,1 S. Saito,1 T. Saito,2 Y. Sawa.1  1Cardiovascular surgery, Osaka University Graduate School of Medicine, Suita, Osaka, Japan; 2Cardiovascular Surgery, Osaka University Graduate School of Medicine, Suita, Osaka, Japan. Purpose: Continuous-flow left ventricular assist device (cLVAD) provides sufficient clinical results for heart failure patients, however cerebrovascular accident (CVA) is still critical complications in patients with cLVAD and it is important to evaluate accurate etiology of CVA to decrease CVA in cLVAD patients. In this study, we evaluated the prevalence of asymptomatic small cerebrovascular complication during cLVAD support. Methods: Among 31 patients who underwent heart transplantation at Osaka University Hospital since 2011, 24 patients had continuous-flow LVAD support (DuraHeart: 16, HeartMateII: 3, HeartWare: 2, Jarvik2000: 3) of mean duration of 2.50 ± 0.95 years. In these 24 patients, excluding one patient who died of primary graft failure, we performed brain magnetic resonance imaging (MRI) before the first discharge after heart transplantation. The prevalence of symptomatic and asymptomatic cerebral lesions in these 23 patients was analyzed. An ischemic lesion was defined as high density signal in fluid-attenuated inversion recovery (FLAIR) -MR imaging, whereas a hemorrhagic lesion was defined as low density signal in T2-weighted gradient-echo imaging. Results: During cLVAD support, 2 patients had symptomatic cerebral infarction and 3 patients had symptomatic intracranial hemorrhage (2: intracerebral hemorrhage, 1: subarachnoid hemorrhage) which were detected by computed tomography scan. Among all 23 patients, 11 patients (48%) had silent ischemic lesion with FLAIR image (mean maximum size: 9.1 ± 6.5 mm, mean counts 1.6 ± 1.0). On the other hand, all 23 patients (100%) had silent cerebral microbleeds (CMBs) lesions (mean maximum size: 8.6 ± 6.2 mm, mean counts 13.7 ± 24.7) with T2-weighted gradient-echo imaging. Counts of CMBs was 0 ~4 in 9 patients, 5 ~ 10 in 5 patients, 11~20 in 4 patients, over 21 in 4 patients. Eight of 23 (35%) had CMBs larger than 5mm. All three patients who had bacteremia during cLVAD support had CMBs with a mean number of 17.0 counts which was significantly higher than patients without bacteremia. No patients had intracranial aneurysm which was detected by MR angiography.

S218

The Journal of Heart and Lung Transplantation, Vol 34, No 4S, April 2015

Conclusion: cLVAD patients had CMBs more frequently than small infarction. Careful attentions have to be paid for hemorrhagic cerebrovascular complications as well as ischemic complications. 5( 83) Device Geometry Does Not Predict Pump Thrombosis in HeartMate II Patients J.J. Han ,1 A.C. Gaffey,1 R. Sooppan,1 G. Hung,1 C.M. Venkataraman,1 E. Phillips,1 J.L. Howard,1 M.A. Acker,1 J. Eduardo Rame,2 P. Atluri.1   1Surgery, University of Pennsylvania, Philadelphia, PA; 2Cardiology, Hospital of the University of Pennsylvania, Philadelphia, PA. Purpose: Pump thrombosis (PT) is a common and potentially life-threatening complication among HeartMate II (HMII; Thoratec, Pleasanton, CA) patients. There have been efforts to correlate HMII geometry with higher risk of PT. The aim of this study was to test the validity of using HMII inflow cannula angle (ICA) and pump pocket depth (PPD) to predict PT. Methods: We performed a retrospective analysis of patients implanted with HMII left ventricular assist devices (LVAD) from January 2011 to March 2014 at our institution. Three blinded reviewers measured ICA and PPD from chest x-rays at post-implantation and most recent follow-up time points. The diagnosis of PT was visually confirmed upon device explantation. Results: HMII was implanted in 90 patients. Sixteen (20%) patients experienced PT. There was no statistical difference between PT and non-PT patients in their initial ICA (56.0° ± 10.1 vs. 54.6° ± 10.8, p = 0.63) and PPD (86.7 ± 24.9 mm vs. 81.1 ± 32.2 mm, p =  0.46). Prediction of PT using ICA and PPD by receiving operating characteristic (ROC) was poor (AUC= 0.54, 0.55, respectively) (Figure 1). Changes in HMII geometry were measured over 112.5 (IQR =  34.3-337.3) days. A decrease in PPD was observed (p =  0.0001). Initial ICA was a significant predictor of future angle change and suggested a convergence towards the median (55.3°) (ANOVA p = 0.005) (Figure 2). Conclusion: Post-operative ICA and PPD do not appear to predict PT in HMII patients in our experience. HMII geometry changes over time secondary to remodeling with a decrease in PPD and a convergence towards the median in ICA. Further investigation into the role of geometric VAD conformation post-implant may be warranted. 

5( 84) Left Ventricular Dimension Decrement Index Early After LVAD Implantation: A Novel Risk Marker for Late Pump Thrombosis E. Joyce ,1 G.C. Stewart,1 M. Hickey,1 J. Rivero,1 I. Gosev,2 M. Leacche,2 M.M. Givertz,1 G.S. Couper,2 M.R. Mehra.1  1Cardiovascular Division, Department of Medicine, Brigham and Womens Hospital, Boston, MA; 2Division of Cardiac Surgery, Brigham and Womens Hospital, Boston, MA. Purpose: Early risk stratification for pump thrombosis in continuous-flow left ventricular assist device (LVAD) therapy remains challenging. We hypothesized that the decrement in LV size prior to discharge after device implantation is associated with increased risk of pump thrombosis. Methods: In 65 patients (age 59±11 years, 79% men, 45% destination therapy) discharged alive without evidence of thrombosis or other device malfunction following first HeartMate II LVAD implantation (January 2011-June 2014), LV dimensions in end-diastole (LVIDd) and end-systole (LVIDs) were compared between pre-implantation and optimal set speed predischarge echocardiography. LV dimension decrement indices (preoperative dimension - optimal set speed dimension divided by preoperative dimension x100) for LVIDd [LVIDdDI] and LVIDs [LVIDsDI] were calculated. Patients were followed for occurrence of confirmed or suspected (lactate dehydrogenase > 1000IU/l or 2.5x upper limit of normal ± evidence of pump dysfunction) LVAD thrombosis for the duration of therapy through November 2014. Results: Incidence of pump thrombosis was 0.07 per-patient-year (n= 18 patients overall). Median time to event was 8 (IQR 2, 17) months. Baseline preoperative LVIDd (p= 0.73) and LVIDs (p= 0.84) were similar between patients with and without LVAD thrombosis. Mean set speed (9240±300 vs. 9190±380 rpms, p= 0.60), other ramp parameters, and change in right heart indices between preoperative and set speed studies did not differ between groups. However, LVIDdDI (18±13 vs. 25±11%, p= 0.04) and LVIDsDI (16±16 vs. 28±13%, p= 0.006) were significantly lower in the pump thrombosis compared to no thrombosis group. On ROC curve analysis, AUC for association with LVAD thrombosis was 0.67 (0.52-0.82, p= 0.04) for LVIDdDI and 0.69 (0.55-0.83, p= 0.02) for LVIDsDI. A cut-off value of ≤ 15% for both indices had sensitivities of 82% (LVIDdDI) and 84% (LVIDsDI) for predicting pump thrombosis (specificities 41% and 35% respectively). Conclusion: LV Dimension Decrement Index measured at optimized speed setting on predischarge echocardiography is predictive of late pump thrombosis. This novel marker reflects intrinsic limitations to optimization of the LVAD/patient interface and identifies patients in whom more intensive surveillance for pump thrombosis may be indicated. 5( 85) Change of Left Atrial Pressure, LAP Measured With a Wireless Implantable Pressure Sensor (Titan Sensor) During Echocardiographic RAMP-Test in HeartMate II Patients L. Hubbert ,1 J. Baranowski,2 B. Delshad,3 H. Ahn.3  1Department of Cardiology, and Department of Medical and Health Sciences, Linkoping University, Linkoping, Sweden; 2Department of Clinical Physiology and Department of Medical and Health Sciences, Linköping University, Linkoping, Sweden; 3Department of Thoracic and Vascular Surgery, and Department of Medical and Health Sciences, Linkoping University, Linkoping, Sweden. Purpose: At implantation and in the initial postoperative course patients undergoing implantation of a Heart Mate II™ (HMII) are routinely monitored by a standard fluid filled catheter measuring the left atrial pressure (LAP) which is of special interest when following the balance between preload of the pump and unloading of the LV. After discharge from the intensive care units, Echocardiography is used to estimate the filling pressure of the left ventricle and optimizing the pump speed. The purpose of this study was to evaluate a new device for long-term wireless monitoring of LAP and evaluate its correspondence with Echocardiographic RAMP-test and pulmonary wedge pressure 10 weeks after implantation of HertMate II and the device. Methods: A wireless microelectromechanical pressure sensor, (Titan™, ISSYS Inc. Ypsilanti, MI, USA) was placed in the left atrium (LA) while