Search

Background The Impella (ABIOMED, Danvers, MA) is a temporary left ventricular support device, intended for short-term cardiac support for percutaneous procedures and cardiac shock. The Impella device is available in several sizes capable of producing different flows. It is FDA approved for up to 6 days of support. The device can provide significantly higher blood flow than an intra-aortic balloon pump and can be implanted percutaneously. We sought to describe potential complications and outcomes associated with extended Impella use for cardiogenic shock. Methods We retrospectively evaluated patients who underwent implantation of either the Impella CP (Impella 3.5) or Impella 5.0 from September 1, 2017 to September 1, 2019 at our institution. Hemolysis was defined as plasma free hemoglobin level > 50 mg/dL, a lactate dehydrogenase (LDH) level > 500 units/L, and bilirubin increase > 1 mg/dL. Results 58 patients underwent percutaneous insertion of an Impella device, with the indication of cardiogenic shock in 32 patients (55%). In these 32 patients, only one had an Impella 2.5 device, 21 had an Impella CP (66%), and 10 (31%) had an Impella 5.0. The average age at time of implantation was 55 ± 17 years, and 25 patients were male (78%). Ejection fraction was 17 ± 11.7%. The average duration of Impella support was 5.5 days (1, 16 days). Complications included hematoma at the time of insertion (3 patients), leg ischemia (2 patients), improper positioning (7 patients), acute renal failure (28 patients), and hemolysis (20 patients). Hemolysis developed in 12 patients (57%) with the Impella CP and in 8 patients (80%) with the Impella 5.0, with an average LDH of 1452 units/L. 8 patients were transitioned to veno-arterial (VA) extracorporeal membrane oxygenation (ECMO), with the Impella utilized for left ventricular venting. 9 patients (28%) eventually underwent durable left ventricular assist device (LVAD) implantation on average 7 ± 6.1 days after Impella, and two patients underwent heart transplantation, both 9 days following Impella insertion. 23 patients (72%) did not survive past the initial hospitalization. Conclusion The Impella is a percutaneous temporary left ventricular support device used to provide additional cardiac output in patients with cardiogenic shock. Serious complications can occur with extended support, including hemolysis, acute renal failure, and high mortality. Of our 28 patients who received an Impella, 11 were able to be successfully bridged to either LVAD or heart transplantation with the Impella.

Purpose The Stanford Integrated Psychosocial Assessment for Transplantation (SIPAT) score is a validated tool for determination of psychosocial and behavioral risk in patients who are undergoing evaluation for organ transplant. The total score is comprised of subsets A-D which scores patient's readiness level and illness management, social support system, psychological stability, as well as lifestyle and effect of substance abuse, respectively. Several centers use SIPAT scores to determine suitability of Left Ventricular Assist Device (LVAD) candidates, however SIPAT remains an unvalidated risk assessment tool in this population. The objective of this study is to assess whether SIPAT scores are associated with 1-year hospital readmissions in LVAD patients. Methods SIPAT scores were documented for 121 LVAD patients (mean age: 50.5 years, 36 % female, 56% black, 45% BTT) implanted at Emory University Hospital from 2010 to 2018. Zero inflated poisson regression models were used to examine the association of SIPAT scores and subscales with the number of hospital readmissions within the 1st year after LVAD implantation, adjusted for age, sex, race, INTERMACS profile, BTT versus DT status, eGFR, and albumin. Results The mean total SIPAT score was 7.07 ±7.76, with mean scores on the subscales as follows: SIPAT A:0.94 ± 2.05, SIPAT B:0.98 ± 2.47, SIPAT C:1.42 ± 2.41, SIPAT D:3.63 ± 3.74. In the first year after LVAD implantation, 259 total hospitalizations occurred in 106 patients (87.6% of total patients had at least 1 admission and there were 16 deaths). The total SIPAT score was associated with a higher number of 1-year hospital readmissions with an adjusted HR: 1.02 for 1-point increase in score (95% CI: 1.00-1.04, p=0.012). Specifically, SIPAT A (adjusted HR: 1.12( 95% CI: 1.06-1.21, p Conclusion In LVAD patients, the total SIPAT score, driven by subscales A and D, was associated with an increased risk of 1-year hospital readmissions following LVAD implantation. These findings suggest that these scores may be useful in risk stratifying LVAD patients and highlight the population of patients who require additional support.

Purpose Outflow graft obstruction is a poorly described complication following continuous flow-left ventricular assist device (CF-LVAD) surgery. We sought to define the incidence of CF-LVAD outflow graft obstruction and assess clinical outcomes with a percutaneous treatment strategy. Methods From January 2012 to June 2019, 252 CF-LVAD patients were managed at our institution. Patients with suspected LVAD outflow graft obstruction underwent comprehensive evaluation with Doppler echocardiography, cardiac computed tomography with angiography and invasive hemodynamic assessment followed by percutaneous intervention. Results Fourteen patients (5.6%) developed hemodynamically significant LVAD outflow graft obstruction at a rate of 0.03 events per patient-year. Outflow graft obstruction presented a mean of 34±18 months after surgery. Patients presented with heart failure (HF; 86%), low LVAD pump flow (93%), or both (79%). LVAD flow declined by an average of 2.1±0.8L/min (p Conclusion LVAD outflow graft obstruction is a relatively common and underappreciated cause of recurrent HF and LVAD dysfunction. Outflow graft stenting can be achieved with low morbidity and provides a long-term solution to this complication.

Purpose Stereotactic body radiation therapy (SBRT) combined with electrocardiographic mapping has emerged as a potential noninvasive ablation technique for the treatment of refractory ventricular tachycardia (VT). The role of SBRT in patients with a left ventricular assist device (LVAD) is unknown. Methods Retrospective chart review was performed on all heart failure (HF) patients who underwent SBRT from January 1, 2018 to September 1, 2019. In this case series, we describe two patients who underwent SBRT after LVAD implantation. Results Of eleven heart failure patients who underwent SBRT, two had LVADs. The first was a 51 year old male with nonischemic cardiomyopathy and ventricular tachycardia refractory to medical therapy and stellate ganglion block. Given his intractable VT and cardiogenic shock, he underwent HeartMate 3 LVAD implantation as bridge-to-transplant (BTT). After LVAD implantation, he continued to have refractory VT despite subsequent VT ablations. CT cardiac mapping identified a 4 mm arrhythmic scar along the lateral wall of the right ventricle, and SBRT was performed with a single fraction of 25 Gy. Two weeks after SBRT, he experienced recurrent VT and eventually underwent orthotopic heart transplantation (OHT) four months after VT recurrence. The second patient was a 66 year old male with a history of nonischemic cardiomyopathy status post HeartWare LVAD as BTT who presented with recurrent VT three years following LVAD implantation. He was referred for SBRT after failure of antiarrhythmic drug (AAD) therapy and VT ablation. 4D-CT simulation defined a myocardial scar in the left ventricular outflow tract, and SBRT was performed with a single fraction of 25 Gy. His ventricular arrhythmias returned after four months, and additional AAD were utilized to suppress the arrhythmias until he underwent OHT two months later. The frequency of hospitalizations for VT did decrease following SBRT (Pt 1: 6 pre-SBRT vs. 3 post-SBRT, Pt 2: 4 pre-SBRT vs. 2 post-SBRT), but did not reach statistical significance (p= 0.12). Conclusion LVAD patients are prone to ventricular arrhythmias, and noninvasive SBRT may serve as an alternative therapy if AAD and traditional VT ablation have been unsuccessful, potentially keeping ventricular arrhythmias quiescent while awaiting heart transplantation. Further research is needed in this patient population.

Purpose Nearly 30% of heart transplant recipients develop de novo donor-specific antibodies post-transplant, leading to an increased risk of antibody mediated rejection, graft failure, and death. Given poor response rates to therapies targeting dnDSA once present, therapies preventing dnDSA altogether could impact transplant outcomes. It remains unclear if the use of induction therapy at the time of transplant mitigates the risk of dnDSA development. The present study attempts to address this question in a multicenter, retrospective analysis. Methods Multicenter, retrospective analysis of 319 heart transplant recipients from 4 participating centers in the U.S. The primary endpoint was the development of dnDSA. Results In the overall cohort, 206 of 319 (65%) patients received induction therapy at the time of transplant, with 200 (62%) receiving basiliximab and 6 (3%) receiving thymoglobulin. Overall 93 of 319 (29%) patients developed dnDSA post-transplant. The use of induction therapy reduced the risk of dnDSA (Kaplan Meier log rank p = 0.009, Figure 1). When assessing induction type, basiliximab reduced the risk of dnDSA compared to no induction, whereas thyroglobulin did not (Figure 2). However, in a multivariable Cox Regression model incorporating the use of an LVAD as BTT, the use of any induction was no longer statistically significant. Conclusion Although induction therapy, particularly with basiliximab, reduces the risk of dnDSA post-heart transplant in univariate analysis, this effect is no longer significant in a model incorporating LVAD as BTT.

Introduction The Stanford Integrated Psychosocial Assessment for Transplantation (SIPAT) score assesses psychosocial and behavioral risk in patients who are undergoing evaluation for heart transplant. The SIPAT score is associated with adverse outcomes including increased rates of rejection, admissions and infections post-transplant. Many Stage D heart failure patients will undergo SIPAT testing, with some receiving an LVAD as bridge to transplant (BTT) and other as destination therapy (DT). However, the role of SIPAT scores in left ventricular assist devices (LVAD) patients remains unclear based on a limited number of studies. The objective of this study assesses whether SIPAT scores were associated with increased mortality. Methods SIPAT scores were documented by a Licensed Master Social Worker for 137 LVAD patients (mean age: 50.0±13 years 35.8% female, 59.8% black)including BTT and DT (57.7%) implanted at Emory University Hospital from 2010 to 2018. The total SIPAT score is comprised of 4 subscales that represent different psychosocial domains: A - patient readiness and illness management, B - social support system, C - psychological stability and psychopathology, and D—lifestyle and substance abuse. Multivariable Cox regression models were used to examine the association of SIPAT scores with mortality. Results During a follow-up period of 1.58 years [IQR 0.8-2.87], 53 patients (38.7%) died. The mean total SIPAT score was 7.45 ±8.06, and 92.0% of patients classified as excellent-good candidates (total SIPAT score ≤20). There was no association between total SIPAT scores and mortality in the total cohort. However, the SIPAT B subscale was associated with mortality in men after adjusting for age, race, albumin, and renal function (SIPAT B *sex interaction P=0.009). Men who died had a higher mean score on the SIPAT B subscale than those who did not (2.00 ± 3.65 vs. 0.63 ± 1.63, p=0.05). Sex-stratified Cox models demonstrated that the SIPAT B subscale was associated with mortality in men (adjusted HR: 1.39, 95% CI 1.11-1.75, p=0.004); but not in women (p=0.4). Conclusions In patients with LVAD, the SIPAT B subscale was associated with an increased risk of mortality in men. These findings suggest that social support may play an even greater role for male patients with LVAD as compared to females. More research is needed to determine which factors clearly define the association between social support and mortality in the LVAD population.

Purpose The 2018 Revised UNOS Heart Allocation System was proposed to 1) address the rapid increase in status 1A candidates by re-classifying them into groups of decreasing acuity, and 2) reflect current prognosis for left ventricular assist device (LVAD) patients. The present study was designed to validate the rationale for the new guidelines by descriptively exploring their effect on patients listed for heart transplant (HT) at our center. Methods We retrospectively evaluated patients listed for HT at Emory University Hospital from 2011 to 2017. Patients were re-classified into the 6-tier Revised UNOS Heart Allocation System. Poisson regression was used to compare outcome rates (transplant and waitlist (WL) removal for death/deterioration) between tiers. Results There were 214 patients (mean age: 50.14 ± 11.9 yrs, 83% male, 57% black) listed for HT during the study period, 97 (45%) as status 1A and 117 (55%) as status 1B. Patients listed as status 1A would be re-classified as tiers 2 (12 [12.4%], 3 (83 [85.6%]), and 4 (2 [2.1%]), and those listed as status 1B would be re-classified as tiers 2 (5[4.3%]) and 4 (112 [95.7%]). Outcome rates by new tier: Patients re-listed as tier 2 had the highest rate of WL death/deterioration (p Conclusion The 2018 Revised UNOS Heart Allocation System accurately prioritizes those status 1A, 1Ab and 1B patients at the highest risk for WL mortality/deterioration in our cohort.

Purpose Previous reports have suggested an association between pre-transplant mechanical circulatory support (MCS) and an increased risk for post-transplant de novo donor specific antibodies (dnDSA). However, it is unclear if specific MCS devices pose a greater risk for dnDSA. The present study seeks to better understand the risk of dnDSA posed by a variety of MCS devices in a multicenter, collaborative study. Methods Multicenter, retrospective analysis of 319 heart transplant recipients from 4 U.S. centers between 2011 - 2017. The primary outcome was the development of post-transplant dnDSA. Results 145 of 319 (45%) patients were supported with durable MCS devices prior to transplant, including 47 Heartware (HVAD), 73 Heartmate II (HM2), and 25 total artificial hearts (TAH). MCS patients had a higher risk of dnDSA compared to those transplanted without mechanical support (37% vs. 23%, p = 0.006; Kaplan Meier log rank p Conclusion Pre-transplant MCS is associated with higher risk for dnDSA. Similar risk is seen regardless of device type; however, the risk associated with TAH was not significantly increased compared to no MCS.

Women are more likely to die from heart failure (HF) than men, but are less likely to receive heart transplant (HT) or left ventricular assist device (LVAD). We sought to investigate differences in HT/LVAD eligibility based on sex. We identified all patients evaluated at Emory University for HT/

A shared understanding of medical conditions between patients and their health care providers may improve self-care and outcomes. In this study, the concordance between responses to a medical history self-report (MHSR) form and the corresponding provider documentation in electronic health records (EHRs) of 19 select co-morbidities and habits in 230 patients with heart failure were evaluated. Overall concordance was assessed using the κ statistic, and crude, positive, and negative agreement were determined for each condition. Concordance between MHSR and EHR varied widely for cardiovascular conditions (κ = 0.37 to 0.96), noncardiovascular conditions (κ = 0.06 to 1.00), and habits (κ = 0.26 to 0.69). Less than 80% crude agreement was seen for history of arrhythmias (72%), dyslipidemia (74%), and hypertension (79%) among cardiovascular conditions and lung disease (70%) and peripheral arterial disease (78%) for noncardiovascular conditions. Perfect agreement was observed for only 1 of the 19 conditions (human immunodeficiency virus status). Negative agreement >80% was more frequent than >80% positive agreement for a condition (15 of 19 [79%] vs 8 of 19 [42%], respectively, p = 0.02). Only 20% of patients had concordant MSHRs and EHRs for all 7 cardiovascular conditions; in 40% of patients, concordance was observed for ≤5 conditions. For noncardiovascular conditions, only 28% of MSHR-EHR pairs agreed for all 9 conditions; 37% agreed for ≤7 conditions. Cumulatively, 39% of the pairs matched for ≤15 of 19 conditions. In conclusion, there is significant variation in the perceptions of patients with heart failure compared to providers’ records of co-morbidities and habits. The root causes of this variation and its impact on outcomes need further study.

Background Incremental value of echocardiography over clinical parameters for outcome prediction in advanced heart failure (HF) is not well established. Methods and Results We evaluated 223 patients with advanced HF receiving optimal therapy (91.9% angiotensin-converting enzyme inhibitor/angiotensin receptor blocker, 92.8% β-blockers, 71.8% biventricular pacemaker, and/or defibrillator use). The Seattle Heart Failure Model (SHFM) was used as the reference clinical risk prediction scheme. The incremental value of echocardiographic parameters for event prediction (death or urgent heart transplantation) was measured by the improvement in fit and discrimination achieved by addition of standard echocardiographic parameters to the SHFM. After a median follow-up of 2.4 years, there were 38 (17.0%) events (35 deaths; 3 urgent transplants). The SHFM had likelihood ratio (LR) χ 2 32.0 and C statistic 0.756 for event prediction. Left ventricular end-systolic volume, stroke volume, and severe tricuspid regurgitation were independent echocardiographic predictors of events. The addition of these parameters to SHFM improved LR χ 2 to 72.0 and C statistic to 0.866 ( P P = .019, respectively). Reclassifying the SHFM-predicted risk with use of the echocardiography-added model resulted in improved prognostic separation. Conclusions Addition of standard echocardiographic variables to the SHFM results in significant improvement in risk prediction for patients with advanced HF.

Background Impaired renal function portends poor heart failure (HF) outcomes. The Seattle Heart Failure Score (SHFS), a multimarker risk assessment tool, however does not incorporate renal function. In this study, we assessed the incremental value of renal function over the SHFS in patients with advanced HF on contemporary optimal treatment. Methods Blood urea nitrogen (BUN), serum creatinine (sCr), BUN/sCr ratio, and estimated glomerular filtration rate were assessed in survival models with SHFS as the base model among 443 patients with HF (52 ± 12 years, male 68.5%, white 52.4%, ejection fraction 0.18 ± 0.08). Incremental value of renal function was assessed by changes in the likelihood ratio χ 2 and the area under the receiver operating characteristic curves for 1-, 2-, and 3-year event prediction. Results During a median follow-up of 21 months, 108 (24.5%) of 443 patients had an event (death [n = 92], urgent transplantation [n = 13], or ventricular assist device implantation [n = 3]). All renal parameters individually were associated with outcome (BUN, P P P = .006; and estimated glomerular filtration rate, P = .006); however, only BUN was an independent predictor of events in multivariable analyses. Addition of BUN improved the predictive ability of SHFS (Δlikelihood ratio χ 2 5.03, P = .025); however, the increase in the area under the receiver operating characteristic curve was marginal (year 1, 0.786 to 0.791; year 2, 0.732 to 0.741; year 3, 0.745 to 0.754; all P > .2). Conclusion Among the various renal function parameters, BUN had the strongest association with outcomes in patients with advanced HF. However, the incremental value of renal function over the SHFS for risk determination was marginal.

Decisions regarding cardiac transplantation listing are difficult in patients with heart failure who have relatively discordant peak exercise oxygen consumption (Vo(2)) and cardiac index (CI) values. One hundred five patients with heart failure who underwent cardiopulmonary exercise testing and right-sided cardiac catheterization for transplantation evaluation were studied. Patients were divided into 4 groups on the basis of peak Vo(2) and CI: group 1, Vo(2)or = 12 ml/min/kg, CIor = 1.8 L/min/m(2) (n = 30); group 2, Vo(2)or = 12 ml/min/kg, CI1.8, L/min/m(2) (n = 27); group 3, Vo(2)12 ml/min/kg, CIor = 1.8 L/min/m(2) (n = 25); and group 4, Vo(2)12 ml/min/kg, CI1.8 L/min/m(2) (n = 23). Groups were compared for event-free (death or ventricular assist device) survival. The overall CI was 1.9 + or - 0.4 L/min/m(2) and peak Vo(2) was 12.4 + or - 2.8 ml/min/kg; values in the 4 groups were as follows: group 1, peak Vo(2) 14.7 + or - 2.1 ml/min/kg, CI 2.2 + or - 0.3 L/min/m(2); group 2, peak VO(2) 14.2 + or - 1.3 ml/min/kg, CI 1.5 + or - 0.2 L/min/m(2); group 3, peak Vo(2) 10.2 + or - 1.3 ml/min/kg, CI 2.1 + or - 0.3 L/min/m(2); and group 4, peak Vo(2) 9.7 + or - 2.0 ml/min/kg, CI 1.6 + or - 0.2 L/min/m(2). After a median follow-up period of 3.7 years, 28 patients (26.0%) had events. Event-free survival was 96%, 95%, 96%, and 79% for 6 months (p = 0.04); 88%, 81%, 90%, and 73% for 12 months (p = 0.09); 88%, 73%, 85%, and 65% for 18 months (p = 0.11); and 83%, 73%, 79%, and 53% for 24 months (p = 0.06) for groups 1 to 4, respectively. Median survival was 5.1, 3.0, 3.9, and 2.6 years, respectively, in groups 1 to 4 (p = 0.052). In conclusion, almost half the patients had relatively discordant peak Vo(2) and CI measurements. Patients with lower peak Vo(2) values but relatively preserved CI values had survival comparable to post-transplantation survival, whereas those with low CI but preserved Vo(2) had a lower survival rate. These results suggest that the former group may be safely monitored on medical therapy, whereas the latter may benefit from early listing.

Category: 24. Myocardial Function/Heart Failure—Clinical Nonpharmacological TreatmentSession-Poster Board Number: 1160-25Authors: Catherine Norton, Vasiliki Georgiopoulou, Andreas Kalogeropoulos, Lucy Fike, Grigorios Giamouzis, Sonjoy Laskar, Robert Cole, Andrew Smith, Wilson W.H. Tang, Sandra Dunbar, Javed Butler, Emory University School of Medicine, Atlanta, GA, Cleveland Clinic Foundation, Cleveland, OH Background: Cumulative adherence with self-care recommendations and association with outcomes is not well described in heart failure (HF) patients.Methods: We used self-report to evaluate adherence to eight HF self-care recommendations (exercise, medications, alcohol and smoking habits, diet, weight and symptom monitoring) among 286 patients with HF (age, 56±11.6 years; 34.3% female; 46.2% black). Adherence was defined as optimal (overall ≥80%) or ideal (≥80% adherence to each recommendation). Outcomes included death or transplant or ventricular assist device placement; rates of emergency department visits, hospitalizations, and length of stay; health status using the Kansas City Cardiomyopathy Questionnaire.Results: Mean follow-up was 525±295 days. Adherence to individual recommendations ranged from 89% for medication to 26% for exercise. Optimal adherence was reported by 34% of patients whereas only 11% indicated ideal adherence. Education was the only sociodemographic variable associated with adherence (odds ratio [OR] 1.15; 95% confidence interval [CI] 1.05-1.25 for optimal; and OR 1.15; 95% CI 1.02-1.30 for ideal adherence per year of education). Patients with optimal or ideal adherence had better clinical outcomes (Table); however, only ideal adherence was associated with better quality of life.Conclusions: In this HF cohort, better adherence with self-care recommendations was associated with improved clinical outcomes. However, adherence was suboptimal for most patients. Optimal AdherenceDeath/Left Ventricular Assist Device/Transplant, % 8.0 9.2 0.73All cause hospitalizations, per 1000 patient-days 2.8 2.0 0.07HF hospitalizations, per 1000 patient-days 1.2 0.9 0.12Emergency department visits, per 1000 patient-days 1.3 0.8 0.02Hospital length of stay, per 1000 patient-days 13.8 8.8 0.06Hospital length of stay - HF only, per 1000 patient-days 8.5 5.5 0.13Kansas City Cardiomyopathy Questionnaire Overall Summary Score 65.2±23.2 67.9±24.3 0.37Ideal AdherenceDeath/Left Ventricular Assist Device/Transplant, % 8.3 9.4 0.83All cause hospitalizations, per 1000 patient-days 2.7 1.5 0.09HF hospitalizations, per 1000 patient-days 1.2 0.4 0.08Emergency department visits, per 1000 patient-days 1.2 0.8 0.33Hospital length of stay, per 1000 patient-days 13.6 3.0 0.02Hospital length of stay - HF only, per 1000 patient-days 8.3 0.8 0.04Kansas City Cardiomyopathy Questionnaire Overall Summary Score 65.1±23.5 74.3±23.5 0.04

Patients with heart failure (HF) are hospitalized over a million times annually in the United States. Hospitalization marks a fundamental change in the natural history of HF, leading to frequent subsequent rehospitalizations and a significantly higher mortality compared with nonhospitalized patients. Three-fourths of all HF hospitalizations are due to exacerbation of symptoms in patients with known HF. One-half of hospitalized HF patients experience readmission within 6 months. Preventing HF hospitalization and rehospitalization is important to improve patient outcomes and curb health care costs. To implement cost-effective strategies to contain the HF hospitalization epidemic, optimal schemes to identify high-risk individuals are needed. In this review, we describe the risk factors that have been associated with hospitalization risk in HF and the various multimarker risk prediction schemes developed to predict HF rehospitalization. We comment on areas that represent gaps in our knowledge or difficulties in interpretation of the current literature, representing opportunities for future research. We also discuss issues with using HF readmission rate as a quality indicator.

Objectives The aim of this study was to validate the Seattle Heart Failure Model (SHFM) in patients with advanced heart failure (HF). Background The SHFM was developed primarily from clinical trial databases and extrapolated the benefit of interventions from published data. Methods We evaluated the discrimination and calibration of SHFM in 445 advanced HF patients (age 52 ± 12 years, 68.5% male, 52.4% white, ejection fraction 18 ± 8%) referred for cardiac transplantation. The primary end point was death (n = 92), urgent transplantation (n = 14), or left ventricular assist device (LVAD) implantation (n = 3); a secondary analysis was performed on mortality alone. Results Patients were receiving optimal therapy (angiotensin-II modulation 92.8%, beta-blockers 91.5%, aldosterone antagonists 46.3%), and 71.0% had an implantable device (defibrillator 30.4%, biventricular pacemaker 3.4%, combined 37.3%). During a median follow-up of 21 months, 109 patients (24.5%) had an event. Although discrimination was adequate (c-statistic >0.7), the SHFM overall underestimated absolute risk (observed vs. predicted event rate: 11.0% vs. 9.2%, 21.0% vs. 16.6%, and 27.9% vs. 22.8% at 1, 2, and 3 years, respectively). Risk underprediction was more prominent in patients with an implantable device. The SHFM had different calibration properties in white versus black patients, leading to net underestimation of absolute risk in blacks. Race-specific recalibration improved the accuracy of predictions. When analysis was restricted to mortality, the SHFM exhibited better performance. Conclusions In patients with advanced HF, the SHFM offers adequate discrimination, but absolute risk is underestimated, especially in blacks and in patients with devices. This is more prominent when including transplantation and LVAD implantation as an end point.

Purpose Clinical scores for right ventricular failure (RVF) risk after left ventricular assist device implantation (LVAD) were developed in mostly pulsatile-flow LVAD populations, whereas most patients currently receive continuous-flow LVADs. Methods and Materials We evaluated the Michigan, Penn, and Utah scores and multiple pre-operative echocardiographic parameters for 90-day RVF prediction in 69 patients who underwent HeartMate-II (N=58) or HeartWare (N=11) implantation in 2007-2011. Results Baseline characteristics are presented in Table 1 . Overall 18 patients (26.1%) experienced RVF (defined as ≥48h NO use; multi-organ failure due to RVF; inotropes for ≥14 days post-LVAD; or re-institution of inotropes). Among clinical scores, only Michigan predicted RVF (OR per point: 1.29; 95% CI: 1.04, 1.59; P=0.02) but discrimination was modest (C=0.64, P=0.093; sensitivity: 50%; specificity: 78.4%). Penn (C=0.50; P=0.99) and Utah (C=0.52; P=0.78) scores did not discriminate RVF. Among echocardiographic parameters, only a smaller left atrial diameter (OR per -10 mm: 2.36; 95% CI: 1.01, 5.45; P=0.045) was associated with RVF and, when added to, improved discrimination of Michigan score to C=0.76 (P Conclusions RVF risk prediction scores developed for pulsatile LVADs are suboptimal for patients with continuous-flow LVADs. Baseline clinical and echocardiographic characteristics (N=69) Patients with RVF (N=18) Patients without RVF (N=51) P Age, yrs 55.2±9.4 48.9±13.9 0.078 Male, % 61.1 47.1 0.56 Ischemic HF, % 27.8 29.4 1.00 Intra-aortic balloon pump pre-op, % 38.9 25.5 0.37 Inotropes pre-op, % 100 96.1 1.00 Pressors pre-op, % 11.1 5.9 0.60 Bilirubin, mg/dL 1.6±0.9 1.6±1.1 0.77 AST, U/L 58 (35, 127) 30 (23, 39) 0.002 Creatinine, mg/dL 2.0±0.7 1.6±1.0 0.014 Cardiac index, L/m 2 1.75±0.50 1.75±0.43 0.86 RV stroke work index, mmHg a l/min 0.52±0.26 0.49±0.24 0.50 LV ejection fraction, % 18.8±4.8 19.1±5.1 0.54 Left atrial diameter, mm 47±6 52±8 0.034 RV fractional area change, % 30.0±6.0 27.8±8.0 0.14

Alterations in Flow and Shear Stress in the Thoracic Aorta with a Continuous-Flow Left Ventricular Assist Device D. Gupta, M. Piccinelli, T. Passerini, A. Veneziani, L. Brewster, S.R. Laskar, D.Q. Nguyen, J.D. Vega, W.R. Taylor. Cardiology, Emory University School of Medicine, Atlanta, GA; Cardiothoracic Surgery, Emory University School of Medicine, Atlanta, GA; Mathematics, Emory University, Atlanta, GA. ●.