Search

Angiodysplasias (AGD) are common sites of bleeding in the gastrointestinal (GI) tract after Continuous Flow Left Ventricular Assist Device (CF-LVAD) implantation. We sought to investigate whether AGDs are formed as a result of LVAD physiology or preexist as a consequence of heart failure. Thirty-six subjects with HF reduced EF (HFrEF) underwent video capsule endoscopy (VCE) to assess for the presence of AGD. Fifty-three subjects without HF who underwent VCE for a nonbleeding indication formed a control group. The prevalence of AGD was significantly higher in the HFrEF compared to the non-HF controls (50% vs 13%, p = 0.0002). This association persisted after controlling for age and comorbidities. Within the HFrEF cohort, higher Ang2, NT-proBNP and BUN were associated with the presence of AGD. AGD in the GI tract are associated with HFrEF. This is the first description of a new pathology associated with HFrEF and adds to our understanding of CF LVAD associated GI bleeding.

The trends and outcomes of multiorgan heart-transplantation (HT) using hepatitis C virus (HCV) donors in the contemporary era are sparsely known. Using UNOS registry, 1322 adult multiorgan-HTs (n = 986 heart-kidney, n = 155 heart-lung, n = 181 heart-liver) between August-2015 and August-2020 were identified, of which 109 were performed using HCV-donors (n = 77 HCV nucleic-acid-amplification testing [NAT] positive irrespective of antibody status [HCV-viremic]; and n = 32 HCV Ab+/NAT-[HCV antibody + nonviremic]). The percentage of HCV-donors used for multiorgan-HT increased from 0% in 2015 to 14% in 2020 (p < 0.001), but there was wide variation across UNOS regions and center volumes. Recipients of multiorgan heart-kidney transplants from HCV-donors (n = 90) and HCV-naive (HCV Ab-/NAT-) donors (n = 896) had similar 1-year survival using unadjusted and adjusted Cox-proportional hazards-regression models including in propensity-score matched cohorts. Post-HT rates of cardiac-allograft-vasculopathy (5.4% vs 5.8%) and chronic-dialysis (7.3% vs 4.9%) at 1-year were also similar. Use of HCV-donors (HCV-viremic, HCV Ab+ nonviremic) for multiorgan-HT has increased significantly. Encouraging 1-year outcomes in heart-kidney recipients from HCV-donors should support further expansion of heart-kidney transplantation using HCV-donors.

BACKGROUND Hepatitis C virus (HCV) donors should be categorized as HCV-viremic (antibody [Ab] negative or positive/Nucleic Acid testing [NAT] positive) or HCV Ab+ nonviremic (Ab+/NAT–). Whereas recipients of hearts from HCV-viremic donors will develop viremia but can be cured of HCV shortly after transplant with direct-acting antivirals (DAAs), recipients of hearts from HCV Ab+ nonviremic donors are highly unlikely to become viremic or require DAAs. Given this important difference in risk, we assessed the utilization trends and post-heart-transplantation outcomes of HCV-naive (Ab–/NAT–), HCV-viremic, and HCV Ab+ nonviremic donor hearts. METHODS A total of 26,572 adult donors (≥18 years) with information on HCV Ab and NAT status were identified in the United Network for Organ Sharing registry between August 2015 and June 2018 for utilization rates. Adult heart transplant recipients of these donors were compared for primary graft failure (PGF) at 90 days and 1-year recipient survival. RESULTS A total of 96 HCV Ab+ nonviremic and 135 HCV-viremic adult donor hearts were transplanted during the study period. The utilization rates of both HCV Ab+ nonviremic (1.4%–23.4%) and HCV-viremic (0.7%–25.4%) donor hearts increased significantly approaching HCV-naive rates (29.04%). There was no significant difference in rates of PGF and 1-year survival between recipients in the 3 donor HCV groups. We also used (1:3) propensity score matching and found similar 1-year survival in different donor HCV groups (HCV-naive vs HCV Ab+ nonviremic, p = 0.59, and HCV-naive vs HCV-viremic, p = 0.98). CONCLUSIONS Recipients of HCV-viremic and HCV Ab+ nonviremic donor hearts had equivalent risk of PGF and 1-year mortality compared with recipients of HCV-naive donor hearts. Although only HCV-viremic organs require DAAs and the risk of coronary artery vasculopathy after treated HCV infection has not been defined, the utilization rates of both HCV Ab+ nonviremic and HCV-viremic adult donor hearts have increased at an equal pace now approaching HCV-naive rates.

BACKGROUND Hepatitis C (HCV) donors are rarely used for cardiac transplantation due to historically poor outcomes. In 2015, nucleic acid testing (NAT) for viral load was added to the routine work-up of organ donors, allowing for the distinction between subjects who remain viremic (HCV Ab + /NAT + ) and those who have cleared HCV and are no longer viremic (HCV Ab + /NAT – ). The American Society of Transplantation recently recommended that HCV Ab + /NAT – donors be considered non-infectious and safe for transplantation. We present our initial experience with such donors. METHODS All patients were counseled regarding donor HCV antibody (Ab) and NAT. Transplant recipients were tested post-transplant at 1 week and at 1, 3, and 6 months for HCV seropositivity and viremia. We also analyzed the UNOS database to determine the potential impact of widespread acceptance of HCV Ab + /NAT – organs. RESULTS Fourteen HCV Ab ‒ subjects received hearts from HCV Ab + /NAT – donors in 2017. Over a median follow-up of 256 (192 to 377) days, 3 patients developed a reactive HCV Ab, yet none had a detectable HCV viral load during prospective monitoring at any time. Analysis of the UNOS database for the calendar year 2016 revealed that only 7 (3%) of 220 HCV Ab + /NAT – donors were accepted for heart transplantation. CONCLUSIONS We have demonstrated the feasibility of utilizing HCV Ab + /NAT – donors for cardiac transplantation without recipient infection. A small percentage of recipients developed HCV Ab without evidence of viremia, possibly consistent with a biological false reactive test, as has been seen in other settings. Large-scale validation of our data may have a significant impact on transplantation rates.

Purpose Delay or complete lack of integration with the HeartMate 3 (HM3) driveline has led to an increased incidence of driveline infection (DLI). Negative pressure wound therapy (NPWT) may be helpful in decreasing the rate of infection by facilitating integration, therefore we evaluated the possible benefits of a single use NPWT system (Prevena) on DLI outcomes (Figure 1). Methods Retrospective chart review of primary HM3 patients implanted from January 2019 through October 2020. The INTERMACS 2016 definition of DLI was utilized. DLI outcomes were analyzed using Fisher's exact test. Freedom from DLI for the NPWT and non NPWT groups were analyzed using the Kaplan-Meier method. Results Within the study period there was a total of 40 primary implants. Twenty- two patients were implanted between January 2019 and October 2019 (non NPWT group) and 18 patients were implanted between November 2019 and October 2020 (NPWT group). In the non NPWT group, 3 patients developed DLI in the first six months (14 %). In the NPWT group, 1 patient developed DLI in the first six months (6 %) (p = 0.512, Figure 2).Average length of time to infection in the NPWT group was 146 days vs 69.3 days in the non NPWT group. Although exit site diameter tended to be smaller with the NPWT approach, overall integration and mobility of the driveline in its transcutaneous track was similar. Conclusion In a moderate size single center experience, use of a NPWT approach led to a decrease in DLI rate. Larger studies are needed to validate this finding.

Purpose Continuous flow left ventricular assist device (CF LVAD) supported patients with blood group O may wait longer for heart transplantation (HT), but it remains uncertain if during this waiting period such patients have greater morbidity and mortality from major adverse events (AEs). Methods We conducted a UNOS database analysis of all patients that were waitlisted for HT after CF LVAD implantation from January 2006 to March 2019. Major demographics, waitlist times, waitlist mortality and occurrence of major AEs were retrieved. Patients were categorized as blood group O and non-blood group O. Post-transplant survival was also retrieved. Waitlist and post HT survival were assessed by Kaplan Meier analysis and Cox proportional hazard ratios were calculated. AEs during device support were compared by logistic regression. Results Overall, 8,578 patients with CF LVADs were placed on the UNOS waitlist during the study period. As expected, Blood group O (n=4,115) had longer waitlist times (412 days, IQR: 102-576; non-O 250 days, IQR 39-323 days, p Conclusion Despite more major AEs during CF LVAD support, patients with blood group O experience similar wait list and post HT survival.

Purpose Right ventricular failure (RVF) remains a significant source of post-LVAD implantation morbidity despite advances in these mechanical circulatory support devices. The vasoactive inotropic score (VIS) is an emerging novel way to estimate total inotropic and vasotropic support, and we believe the post-operative VIS score could correlate with severe RVF post-LVAD implantation. Methods We performed a single-center retrospective study of 240 patients who received a continuous-flow durable LVAD between Jan 1st, 2006 through Dec 31st, 2017. Patients who were less than 18 years old, who underwent LVAD exchange, and who received a planned BiVAD were excluded. The VIS score was calculated as: dobutamine (mcg/kg/min) + 10 x milrinone (mcg/kg/min) + dopamine (mcg/kg/min) + 100 x epinephrine (mcg/kg/min) +100 x norepinephrine (mcg/kg/min) + 10,000 x vasopressin (mcg/kg/min). The VIS score at 6, 24, and 48 hours postoperatively were calculated and then the maximum VIS score within 48 hours for each patient used to stratify the cohort into quartiles. Severe RVF was defined as meeting clinical RVF and having inotropes >14 days after implant, RVAD placement during implant admission, or death from RVF during implant admission. Results In our 240-patient cohort, the VIS quartiles groups were 0-10, 11-15, 16-22, and 23-87. Patients in quartile 4 were older, had a higher CVP and INTERMACS level (Table 1). In our univariate analysis, quartile 4 was associated with severe RVF (OR 3.62, CI [1.37-9.56], p = 0.009). The multivariable model (which included age, INTERMACS level, total bilirubin over 2.5 mg/dL, RV dysfunction on echocardiogram, and gender) found that maximum VIS score within 48 hours for quartile 4 independently correlated with severe RVF (OR 3.04, CI [1.5 - 8.83], p = 0.04). The ROC-curve for this model had an AUC of 0.68. Conclusion Our results demonstrate that the highest VIS score within the first 48 hours after LVAD implantation correlates with severe RVF.

Purpose Recent data from single centers and registry analysis have demonstrated comparable one year survival after cardiac transplantation with HCV non-viremic donors. On the other hand, long-term outcomes remain unknown which is particularly relevant given the concern for accelerated CAV from HCV donors in the pre-DAART era. Therefore, in the current analysis we report our one year rates of CAV in recipients of HCV non-viremic organs. Methods A retrospective chart review was conducted of all transplants at Montefiore Medical Center between January 2017 and April 2019. Recipients were divided into 2 cohorts based on donor HCV status: non-HCV or HCV non-viremic (Ab+/NAT-). Donor and recipient characteristics as well as presence of CAV at one year assessment were collected. CAV was defined as intimal thickness >0.5mm by IVUS or OCT of the LAD when available or in cases without IVUS/OCT as Grade 1 or greater per the 2010 ISHLT CAV recommendations. Patients who received a HCV viremic organ, died before one year or did not have one year CAV assessment were excluded. Results During the study period, 97 transplants were performed. Breakdown by HCV status was: 70 non-HCV, 21 non-viremic, and 6 viremic transplants. 48 subjects in the non-HCV and 14 in HCV non-viremic categories reached one year and underwent CAV assessment. Characteristics associated with the development of CAV including recipient age, donor age, pre-transplant PRA, and rejection in the first year were not statistically different between groups. Rates of CAV were not different between recipients of non-HCV and HCV viremic organs at one year: 27% vs 14% (p=.32). At the time of presentation we anticipate data on presence of CAV on all 97 patients. Conclusion In this preliminary analysis, rates of CAV in recipients of HCV non-viremic organs were comparable to non-HCV donors. This finding provides further reassurance for the use of such donors in standard risk recipients.

Purpose Although various temporary mechanical circulatory support devices are utilized during acute myocardial infarction cardiogenic shock (AMI CS), comparisons of interdevice outcomes remain uncertain. We sought to determine patient characteristics and outcomes between venoarterial extracorporeal membrane oxygenation (VA ECMO) and Impella supported patients during AMI CS. Methods A single center retrospective study was conducted in consecutive patients with AMI CS from 02/12 to 09/19. Patients were grouped and analyzed by their initially assigned device as either VA ECMO or Impella. Device crossover was censored. Severity of illness was compared by sequential organ failure assessment (SOFA) and Vasoactive Inotropic Score (VIS). Thirty day survival and a composite of adverse neurologic (ischemic stroke and intracranial hemorrhage), thromboembolic and major bleeding events were compared between devices by Kaplan Meier analysis. Results One hundred and twelve patients with AMI CS underwent VA ECMO (n=57) or Impella (n=55) placement. Patients with VA ECMO had higher SOFA (15.0 IQR: 11.5-18.0 vs. 8.0 IQR: 5.0-13.0, p Conclusion Despite a higher severity of illness, AMI CS patients experienced similarly limited survival with VA ECMO or Impella. More adverse events occurred during VA ECMO support. Larger studies are warranted for optimal device selection during AMI CS.

Purpose Increased hemoglobin A1C was reported to be a predictor of angiographically detected CAV. Our study examined the impact of glycemic control on myocardial blood flow (MBF) measured by 13N-ammonia positron emission tomography (PET) in transplant recipients. Methods We reviewed the records of patients who had heart transplant in the past 5 years at our center, and who had regadenoson-stress 13N-ammonia PET imaging. Three PET flow parameters were measured: stress MBF, myocardial flow reserve (MFR), and coronary vascular resistance (CVR = systolic blood pressure / stress MBF). Age, time from heart transplant to PET imaging, cold ischemia time during transplant, history of ≥ 2R cellular rejection, cholesterol level and A1C within 3 months of PET were recorded as potential risk factors. Correlations between PET and potential risk factors were analyzed (Pearson and Spearman's rank). Mann-Whitney test was used to compare PET flow between patients with A1C Results 57 patients (median age 57.6 years, 73% men) were identified. Median time from transplant to PET imaging was 25 months. Median stress MBF was 2.8 (2.3-3.2) ml/min/g, CVR was 46.7 (39.1-54.0) mmHg/ml/min/g, and MFR was 3.0 (2.6-3.6). A1C (median 6.4%) correlated significantly with stress MBF: r -0.395 (P 0.006), and with CVR: r 0.392 (P 0.006). Time from transplant to PET imaging showed weaker correlation with stress MBF: r -0.284 (p 0.032) and CVR: r 0.279 (p 0.035). Patients with A1C Conclusion Time from heart transplant and A1C level significantly correlated with MBF by 13N-ammonia PET. Patients with high HbA1C level had significantly lower MBF, suggesting that glycemic control may have an important role in prevention of CAV.

Purpose Machine learning (ML) techniques can improve predictive modeling over more traditional methods by identifying higher dimensionality and non-linear relationships between variables. We hypothesized that an AutoML algorithm would be superior to a logistic regression (LR) model for prediction of outcomes in HT. Methods The UNOS database was queried for patients receiving single-organ heart transplantation between 2006-2016. Pre-transplant variables for the donor and recipient were used for the prediction of one-year mortality or re-transplant. Auto machine-learning (AutoML) with stacking of Gradient Boosting Machine (GBM) derived algorithms was used to create a ML predictive meta-model. These results were compared to a traditional LR model using receiving operating characteristic (ROC) values. Results During this time period 18,612 patients with HT were identified. Observed one-year mortality or re-transplant was 11.5%. The AutoML derived model performed modestly (ROC=0.66, Figure 1) but showed improvement in outcome prediction over the LR model (ROC=0.62, Figure 2). Strongest predictive variables in the LR model were recipient bilirubin, creatinine, mechanical ventilation, donor age and ischemic time. The meta-model structure of AutoML precludes direct assessment of individual variable weight. Conclusion Using contemporary input from the UNOS database, AutoML meta-modeling outperformed LR for prediction of one-year outcomes in HT. Automation of predictive modeling using ML in HT is powerful, albeit limited by the “black box” effect on individual variables. AutoML in HT warrants further investigation.

Purpose Early acute cellular rejection (ACR) after orthotopic heart transplantation (OHT) is associated with adverse long-term outcomes. Data on whether the Mycophenolate mofetil (MMF) dose has any impact on rates of early ACR is lacking. The target MMF dose studied in clinical trials for immunosuppression with OHT has ranged from 2 to 3gm/day. In January of 2016 due to an increase in infectious complications, we modified our institutional policy whereby the target MMF dose was lowered from 3 gm/day to 2gm/day. In this analysis, we examined rates of early rejection stratified by target MMF dose. Methods We performed a retrospective analysis of consecutive OHT recipients at a single center from January 2015 to July 2019. Patients were divided into two groups based on the target MMF dose (3gm vs. 2 gm/day). The rates of ACR warranting a change in immunosuppressive regimen within 4 weeks of OHT (1R/1B or ≥ 2R) were compared between the two groups. Data on other factors that could impact early ACR such as age, gender, race, induction immunosuppression, and PRA was also collected. Univariate and multivariate-adjusted statistical analyses were performed. Results Of 169 OHT, mean age 53 years and 26% females, 67 patients were in the MMF 3gm/day group (Group 1) and 102 were in the 2 gm/day group (Group 2). Baseline characteristics were similar in both groups. The average daily dose of MMF for the first week after OHT was 2245 ± 724 mg in Group 1 and 1866 ± 525 mg in Group 2 (p Conclusion In this single-center study of OHT recipients comparing daily target MMF dose of 3gm with 2gm in the early post OHT period, there was a trend towards early ACR in the 2gm/day MMF dose group. This suggests that a higher MMF dose of 3gm/day, if tolerated, may help lower rates of early ACR. Further investigation in a larger sample is warranted to validate this finding.

Purpose Traditionally, heart transplant (HT) patients with cardiac allograft vasculopathy (CAV) are treated with a combination of calcineurin inhibitor (CNI) and sirolimus (SRL). Recent data suggest outcomes in patients with CAV may be improved by treatment with SRL and mycophenolate mofetil (MMF), without CNI. However, this CNI free regimen for HT patients with renal dysfunction resulted in more rejection. Here we evaluated the relationship between risk of rejection and SRL use in CAV. Methods This is a single center, retrospective study of HT patients with CAV, transplanted between 6/2003 - 10/2018. From 357 HT patients, 39 with CAV treated with SRL were identified. Rejection included ISHLT grading ≥ 1R/1B requiring therapy. Incidence of post SRL conversion rejection was analyzed for 2 groups: SRL/CNI and SRL/MMF. Results Of 39 patients identified, 11 received SRL/MMF and 28 received SRL/CNI. Six of 11 patients in the SRL/MMF group developed cellular rejection after transition to SRL, compared to 3 of 28 in the SRL/CNI group (55% vs 11%, p = 0.004; Fig. 1). Four of 6 patients on SRL/MMF who developed post SRL transition rejection, had a history of previous rejection. However, those who did not experience late rejection also had no history of early rejections (66% vs 0%, p=0.045). The average level of SRL was 10.5 ng/ml +/- 1.1 in the group with post SRL rejection and 7.6 ng/mL +/- 1.9 in the group without (p = 0.01). The average dose of MMF was 1834 +/- 376 mg daily in the group with post SRL rejection and 2000 +/- 0 mg daily in the group without (p = 0.78). Conclusion In HT patients with CAV treated with CNI free regimen, despite therapeutic levels of SRL, there is a significant increase in the risk of cellular rejection not seen in those on SRL/CNI. In those on SRL/MMF therapy who experienced post SRL rejection, there was a higher incidence of pre-conversion rejection. Reduced dose CNI/SRL immunosuppression regimens need to be considered in CAV patients, especially those with a history of previous rejection.

Introduction Left ventricular assist devices (LVADs) have undergone technological advances over the past several years, however, device thrombosis remains a feared complication. We present a case of a HeartWare HVAD presenting with pump thrombosis which failed medical therapy and was exchanged for an Abbott HeartMate 3 pump. Case Report The patient is a 38-year-old male with end stage-systolic heart failure due to a HIV associated cardiomyopathy who underwent HVAD implantation six years prior to presentation. The patient was in his usual state of health when he presented to the hospital due to a "high watt" alarm. He was asymptomatic. The patient, who was maintained on aspirin and warfarin, had a presenting INR which was subtherapeutic (1.2). Interrogation of the device logs showed an abrupt increase in power up to 9W with flows >10 L/min (baseline ∼4.5W and ∼5.5 L/min) (Figure). Computed tomography (CT) angiogram did not reveal an outflow graft thrombosis but did note pulmonary edema. The patient was diagnosed with pump thrombosis and was initially treated with two doses of alteplase 20 mg (Figure arrows). The power and flow normalized. Multiple lactate dehydrogenase (LDH) specimens were hemolyzed, but the LDH 3 days after admission ultimately was 3,550 U/L (baseline 268 U/L). Two days later, power spikes recurred. The patient was given one dose of alteplase 50 mg without success. The patient underwent emergent exchange to a HeartMate 3 LVAD via a re-do sternotomy. Surgery was notable for the epicardial ring of the HVAD being firmly incorporated at the left ventricular apex. To minimize trauma, the epicardial ring of the HeartMate 3 was sutured directly onto the epicardial ring of the HVAD. The Heartmate 3 was then inserted through both epicardial rings while the outflow graft was anastomosed to the remnant outflow graft of the HVAD. The patient did well post-operatively and was discharged on post-operative day number 15. Summary A thrombosed HVAD can be successfully managed with exchange to a HeartMate 3 LVAD.

Purpose Despite the recent advances in Left Ventricular Assist Device (LVAD) therapy, stroke remains a frequent and catastrophic complication. Hypertension (HTN) after LVAD implant has been associated with an increased risk of stroke, but the association between HTN before LVAD implant and stroke has never been studied. We hypothesized that LVAD patients with elevated blood pressure (BP) during outpatient visits before the LVAD implant may have an increased risk of stroke. Methods We retrospectively identified 175 patients on HeartMate II or HeartMate 3 LVAD support (age 54.3±13.1 y, 77% male, 35% ischemic etiology, 32% HeartMate 3) who had at least one outpatient BP measurement with a regular BP cuff in the 2 years preceding the LVAD implant. Average pre-implant BP was calculated based on outpatient visits. Stroke was defined based on clinical and radiographic evidence. Patients who had any event (death, stroke, device exchange, pump thrombosis, heart transplant) in the first 30 days post-implant were excluded. Results 17 patients (9.7%) had a stroke. Average pre-implant SBP was significantly higher in the patients who had a stroke after LVAD implant (105±13 vs 116±13 mmHg, p Conclusion High SBP in the years before LVAD implant is associated with an increased risk of post-implant stroke. Further studies are necessary to establish if high SBP before the implant may represent an early sign of vascular dysfunction that eventually leads to stroke.

Purpose The RESTAGE-HF multicenter study demonstrated high rates of myocardial recovery after combining LVAD support with HF pharmacological therapy and regular cardiac function monitoring protocol. We sought to investigate whether the time course of LVAD-mediated reverse remodeling impacts the post-explantation long-term sustainability of myocardial recovery. Methods We studied 36 HeartMate II LVAD patients with advanced non-ischemic dilated cardiomyopathy who completed the study protocol. LVAD explantations occurred when patients met the reverse remodeling criteria with the LVADs turned-down to minimum speed: LVEF >45%, left ventricular end diastolic diameter (LVEDD) 2.4 L/min/m2. We investigated whether meeting these criteria 12 months post-LVAD support impacted long-term sustainability of the myocardial structural and functional response or long-term survival free of Transplant (Tx) or LVAD. Results Nineteen patients (53%) met criteria and underwent LVAD explantation. Patients explanted at 12 months post-LVAD were 2 (10.5%), 9 (47.4%) and 8 (42.1%), respectively. There were no significant differences (p=0.80) in the survival free of Tx or LVAD at 1, 3, and 5 years post-LVAD removal between patients explanted at Conclusion In RESTAGE-HF multicenter study the time course of reverse remodeling during LVAD support did not appear to impact the long-term sustainability of the post-explantation favorable clinical response either in terms of long-term survival free of HF recurrence or stability of the improved cardiac structure and function.

Purpose Despite recent advances in the understanding of antibody-mediated rejection (AMR), our immunologic and histologic characterization remains imperfect. At present, robust markers of AMR severity are lacking. We examined the association between a readily available index - sum total mean fluorescence intensity (MFI) of all de-novo donor-specific antibodies (DSA) - and clinical outcomes. Methods A retrospective chart review of 146 consecutive heart transplant recipients from 2015-2018 was performed. AMR was defined as histologic (activated endothelial cells and intravascular macrophages) or immunopathologic (C4d) evidence on biopsy. DSA and MFI were determined via Luminex. Sum total MFI was calculated as the sum of the MFI of all class I and class II de-novo DSA at the time of initial presentation with AMR. Clinical outcomes through September 15, 2019 were retrieved for the endpoints of death or persistently depressed (>90 days) left ventricular ejection fraction (LVEF) Results Within the population of 146 patients, 46 patients developed de-novo DSAs. Of these, 15 met criteria for AMR and 31 did not. AMR patients had a significantly higher sum total MFI than those without clinical AMR. [38,296 ± 26,667 vs. 7,714 ± 6564; p Of AMR patients, all had at least one Class 2 DSA (13 out of 15 had DQ) while 4 of the patients did not have any Class 1 DSA. All patients were treated according to our institutional protocol with plasmapheresis/IVIG and bortezomib and/or rituximab. Over a mean follow up of 466 days ±356, 5 patients died, while an additional 3 patients had persistently reduced LVEF. Patients were dichotomized according to sum total MFI >30,000; 80% above this value had an adverse event vs. 33% below (p=0.05) Figure 1. The 2011 ISHLT pathologic grading system for AMR had no correlation to outcomes. Conclusion Sum total MFI, a simple and intuitive index, is a prognostic marker of clinical outcomes at the time of AMR. This index may help stratify patients for more aggressive treatment at the time of diagnosis.

Purpose The development of post-transplant diabetes mellitus (PTDM) is associated with an increased risk of acute rejection after kidney and liver transplantation, however, the outcomes after orthotopic heart transplant (OHT) have yet to be systematically studied. We sought to investigate whether there is any relationship between the development of PTDM and acute rejection after cardiac transplant. Methods This was a single center retrospective review of 100 consecutive OHT patients between May 2015 and May 2018. A total of 39 patients with a documented history of diabetes mellitus were excluded from the study. Patients were divided into 2 groups: those who developed PTDM and those who did not. PTDM was defined by a hemoglobin A1C (HBA1c) ≥ 6.5 and/or a fasting blood glucose (FBG) ≥ 126 at 6 months post-transplant. Rates of rejection (ISHLT grading ≥ 1R/1B - 3R/3B) were compared between the 2 groups. Results Of the 61 patients (mean age 54.5 ± 11.9), 12 (19.6%) developed persistent PTDM at 6 months. Nine of the 12 patients who developed PTDM at 6 months had rejection compared to 9 of 49 patients who did not develop PTDM (75 % vs 18.37 %; p= Conclusion PTDM is associated with an increased risk of acute cellular rejection within 6 months post-OHT. Larger studies are warranted to examine whether any causal relationship exists between the development of PTDM and acute rejection in heart transplant recipients.

Purpose Previous studies have demonstrated a correlation between increasing center volume and improved post heart transplant (HTx) survival. We performed a contemporary analysis to evaluate this relationship. Methods The UNOS database was analyzed to include adult (≥18 years) HTx recipients from years 2006 to 2016. Each center had to perform >65 transplants to be included. Centers were categorized based on annual transplant volume: 6-15, 16-25, 26- 39 and ≥40. Primary endpoint was one-year mortality or retransplantation. Survival analysis was performed using Kaplan-Meier method. Adjusted analysis were performed using Cox-regression modeling. Results Among 17,853 HTx recipients, distribution of patients was 32.1%, 30.8%, 14.5% and 22.6% in the order of increasing center Htx volume, respectively. Median recipient age was 56 years (IQR 47-63) and 75.5% were male. Primary endpoint of one-year mortality or retransplantation was similar across the centers irrespective of transplant volume (6-15/yr: 10.4%, 16-25/yr: 9.5%, 26-39/yr: 11.1%, ≥40/yr: 9.7%; χ2 p=0.10 and Log-rank p=0.11) as noted in figure 1. When adjusted for recipient and donor risk factors, the Cox-regression analysis showed that centers with ≥40 Htx/yr had a lower rate of one-year mortality/retransplantation (HR 0.86; 95% CI 0.75-0.98; p=0.029) compared to patients at centers with volume 6-15/yr. The two remaining categories 16-25/yr (HR 0.94; p=0.31) and 26-39/yr (HR 1.06; p=0.43) showed no difference in primary endpoint when compared to the lowest volume category (6-15 /yr). Conclusion There was no difference in baseline rate of mortality or retransplantation at 1-year among Htx centers irrespective of volume. Among all categories, only centers with ≥40 Htx/yr had an improved risk adjusted primary endpoint compared to the centers with lowest volume. The correlation between increasing center volume and improved survival appears to have weakened in this updated analysis among US transplant centers.

Purpose Despite the recent advances in Left Ventricular Assist Device (LVAD) therapy, gastrointestinal bleeding (GIB) remains a frequent and severe complication. A growing body of evidence suggests that the combination of hypoperfusion and hypoxia in the microcirculation may activate the angiogenesis cascade and promote the formation of arteriovenous malformations. We hypothesized that LVAD patients with concomitant significant restrictive or obstructive lung diseases may have an increased risk of GIB due to their chronic hypoxemic state. Methods We retrospectively identified 99 patients on HeartMate II LVAD support (age 54±13 y, 77% male, 36% ischemic etiology) who had a pulmonary function test (PFT) performed. Significant lung disease was defined as a moderate or severe obstructive/restrictive pattern detected during PFT and the concomitant use of inhaler therapy. GIB was defined as clinical evidence of GI bleeding and a decrease in hemoglobin of at least 1 g/dL. Results 11 patients (11%) had significant lung disease and 27 patients (27%) experienced a GIB. The risk of GIB was increased in patients with significant lung diseases (Hazard Ratio 2.59, 95% confidence interval 1.038 to 6.452). Freedom from GIB was significantly higher in patients without significant lung disease in Kaplan-Meier analysis (Figure). Conclusion Significant obstructive or restrictive lung diseases were associated with an increase in the risk of GIB. Further studies on the association of tissue hypoxia and activation of the angiogenesis cascade/bleeding diathesis in LVAD patients are warranted.

Purpose To address the high incidence of driveline malfunction in the HeartMate II® (HM2), a shield less, polyurethane -insulated, and modular cable driveline was introduced into the HeartMate 3® (HM3). Because these modifications resulted in changes to the physical properties of the driveline, we sought to evaluate the impact of these putative improvements on the incidence of driveline infection (DLI). Methods Retrospective chart review of all HM2 and HM3 between 2015 and 2019. DLI was defined according to the INTERMACS 2016 definition. Patients with DLI were reviewed for clinical course including diagnosis, treatment, and outcomes. Data were analyzed using Kaplan-Meier methodology. Results Over the study period, 93 HM2 and 68 HM3 were implanted. DLI was diagnosed in 16% of HM2 patients vs. 21.7% of HM3 patients (p=0.14, figure 1). Average time to DLI was 428.2 days in HM2 vs 370.3 days in HM3. EPPY was 0.14 for HM2 vs. 0.23 for HM3. HM3 DLI tended to be more severe as evidenced by need for surgical intervention 80% vs. 40% (p Conclusion In this single center experience, HeartMate 3® patients develop infections with shorter support times and required more aggressive interventions. Anecdotally, our team has noted a delayed or lack of integration of skin to polyurethane after implantation. In response, we have implemented the following: dual adjustable anchor, cadexomer iodine for excessive drainage, showering without dressings, suture removal prior to patient discharge, and incorporating picture instructions in driveline dressing kits to facilitate adherence to sterile technique. At the time of presentation of this abstract, we anticipate sharing our outcomes with these new techniques.

Purpose Telehealth has the potential to improve health care delivery by simplifying communication between providers and patients. This technology is well suited to LVAD patients because of the need for close monitoring and ease of access to the LVAD team for patients in remote areas. In this study, we describe our early experience with a novel telehealth platform that was designed specifically for LVAD patients. Methods In conjunction with Medocity Inc, we developed a platform that included daily VAD parameter logs, medication adherence, video-messaging, and instructional videos. We recruited 25 LVAD patients to utilize the platform on a daily basis. Patients were expected to login daily and enter their LVAD parameters, answer daily health questions, and enter medication compliance in the virtual pillbox. Patients also were given the option to review educational videos, message providers with concerns, as well as review their LVAD parameters to monitor for outliers. The primary outcome was patient satisfaction measured by a patient centered survey. Secondary outcomes were 30 day readmission and 30 day patient engagement. Results : 22 out of the 25 patients completed the end of survey study (1 refused, 2 wrong survey). Table 1 summarizes the responses; which notably includes over 90% of respondents stating that the application helps them feel more in control of their care. A sampling of comments: Being able to track my LVAD makes me more aware of the LVAD function. Very beneficial and easy access daily data records and allows faster communication from LVAD Coordinator I felt safe knowing that I was being monitored. Patient engagement was defined as: days logged into platform/days out of hospital. The average engagement was 71% for the first 30 days. There were 6 readmissions (23%) in the first 30 days. Our institutional 30 day readmission benchmark is 25%. Conclusion Telehealth applications can improve patient satisfaction through greater connectedness with their LVAD team. Its potential for reducing readmissions should be tested in a larger prospective study.

Purpose We have previously reported a high prevalence of gastrointestinal angiodysplasia (GIAD) in advanced HF patients undergoing LVAD evaluation. We now hypothesize that GIAD develop in relation to advancing heart failure and are associated with angiogenic dysregulation. Methods After informed consent, patients being evaluated for LVAD underwent video capsule endoscopy (VCE) for assessment of GIAD. Two control groups were formed: (1) NYHA Class 1 HF patients with LVEF angiogenesis was measured in both HF groups and a subset of paired pre and post LVAD patients. Results Median age and percent female of the advanced HF (n=17), NYHA 1 (n=4), non-HF controls (n=36) were as follows: (61, 61, 56; p = .789) and (18%, 0%, and 56%; p =0.020). Prevalence of GIAD was much higher in the advanced HF group as compared to the NYHA 1 and non-HF controls: 71% vs 25% vs 33% (p=0.009). Median Ang2 levels were numerically higher in the advanced HF as compared to the NYHA 1 group (5.4 vs 3.3ng/mL; p=0.138) although this did not reach statistical significance in this small sample. In 9 subjects with paired measurements, Ang2 increased from a mean of 6.7±3.5 to 11.3±7.6ng/mL (p=0.047) after LVAD implantation. Conclusion GIADs are already present in the majority of patients undergoing LVAD implantation and appear to be part of the pathophysiology of advanced heart failure. Angiogenic dysregulation predates LVAD implantation but is augmented in the setting of continuous flow LVAD support. Enrollment in this study is ongoing and at the time of abstract presentation, we anticipate a larger sample size.

Purpose VS in heart transplantation has been associated with adverse survival outcomes. We evaluated risk factors for and early outcomes of VS amongst CF-LVAD patients undergoing OHT. Methods Single center, retrospective cohort of all adult CF-LVAD BTT patients between 5/2013 to 5/2018. VS was defined as profound vasodilatory shock (SVR ≤ 800 dynes/s/cm−5) with a preserved cardiac index (CI ≥2.2 L/min/m2) requiring vasopressor support (vasopressor score ≥ 4) within 48hrs post OHT. Vasopressor score = (epinephrine dose x 100 μg/kg/min) + (norepinephrine dose x 100 μg/kg/min) + (vasopressin dose x 10000 units/kg/min) + (phenylephrine dose x 100 μg/kg/min). Pulsatility Index (PI) of HM II CF-LVAD at index admission was obtained as a surrogate marker of pulsatility. Results Eighty-four CF-LVAD patients underwent OHT during the study period. Median age was 56 years, 22 (26.2%) females, and 29 (34.5%) ischemic etiology. Eighteen (21.4%) patients developed VS after OHT. VS was associated with a higher median BMI (31.0 vs 27.3; p = 0.01) as compared to patients absent VS. Median duration of CF-LVAD support was longer in VS patients; 13.4 months vs 9.2 months, p=0.27. In 69 HMII patients, the mean PI at admission was lower in those who subsequently developed VS; 5.9 vs 6.4; p= 0.04. The rest of the recipient, donor and intraoperative characteristics were similar between groups. Multivariate logistic regression analysis revealed BMI (OR 1.34; 95% CI 1.11-1.63; p Conclusion VS after transplant in CF LVAD subjects is associated with low pulsatility and high BMI. In this single center experience, VS was not associated with worse outcomes.

Purpose High-Intensity interval training (HIIT) has been shown to induce reverse left ventricular (LV) remodeling in patients with heart failure. We tested whether this effect could be reproduced in patients supported with long-term LVAD. Methods Chronic LVAD pts prospectively completed 15 cycle ergometer based, supervised HIIT sessions. Cardiopulmonary exercise test (CPX) with full LVAD support and resting turn-down echocardiogram (HM II speed reduced to 6000 rpm) were performed before and after training. Training intensity was based on peak power output (PPO) attained during baseline CPX. Each session lasted 30 min: 3 min warm-up (50% PPO), six 30 sec challenges (100% PPO) and six interspersed 4 min recovery periods (40% of PPO). One reader, unaware of the study intervention, interpreted the turn-down echocardiograms. Two blinded independent readers identified oxygen uptake (VO2) at ventilatory threshold (VT) and peak exercise (pVO2) in all CPXs. Results Nine subjects were enrolled (6 male, 7 NICM, 7 BTT) mean ± SD age 47±16yrs. Time from heart failure diagnosis to LVAD 60±34 mos., time on LVAD 28±16 mos. Adherence to HIIT was high; one pt withdrew due to drive line malfunction and one voluntarily withdrew. No serious adverse events occurred during exercise. All patients had ICD in place. Four were on ACE-I/ARB, 9 on Beta-blocker and 5 on loop diuretics. Two had DM2, 4 had AF, 2 had prior CVA and 3 had prior pump exchange. Seven pts completed the protocol. HIIT significantly reduced LV end-diastolic (193±90 to 183 ±89 ml; p=0.0041) and end-systolic (133±63 to 128±63; p=0.04) volumes. LV size decreased in all pts. Submaximal aerobic capacity, VO2 at VT (7.0±1.5 to 9.0±1.6 ml.kg.min-1; p=0.02) and pVO2 (11.3±2.3 to 12.2±2.4 ml.kg.min-1 p=0.02) also improved after HIIT. Conclusion A short HIIT program guided by CPX induced reverse LV remodeling and improved aerobic capacity in pts with long-term LVAD support. Addition of HIIT to established recovery protocols warrants further investigation.

Purpose A lack of standardized protocols looking for cardiac recovery during CF LVAD support is partially responsible for low rates of device explant. LVAD wean assessments are time consuming and therefore often not performed. We sought to determine whether a baseline (full-support) echo could serve as a screening assessment to identify those individuals with explant potential who should undergo wean assessment. Methods A retrospective review of all LVAD weaning studies performed at Montefiore between 2007 and 2018 was conducted. LVAD weaning studies were performed with HMII at 6000 rpms and HVAD at 1800rpms which are the generally accepted “net-zero” flows for each device. For individuals who had multiple weaning assessments, only the most recent was analyzed. Only those studies with adequate windows as defined by clear endocardial definition in both apical 4 and 2 chamber views were included. LV diameters, volumes by Simpson's biplane method, LVEF and RV fractional area changes (FAC) were measured at the baseline LVAD speed and at weaned speed after 15minutes. Results Of the 57 patients who had weaning studies, 41 met the inclusion criteria with a mean age of 46 yrs, 41.5% ischemic etiology, 24% female, and a median duration of 222 days of LVAD support. During wean assessments, there was an increase in all measures of LV size: LVIDd - 5.83cms to 6.15cm, LVEDV - 191mL to 207mL (p=0.0001) and LVESV - 141mL to 148mL. There was a corresponding modest increase in LVEF from 31 to 34% (p=0.07). Fourteen (34%) had a LVEF >40% on baseline echo; 5 of these went on to normal LVEF (>50%) on wean study. Conversely, no patient with a LVEF Conclusion During LVAD wean there was evidence of LV re-loading as shown by an increase in LV volumes and a modest increase in LVEF. An LVEF >40% on baseline echo identified all patients with a normal LVEF during wean assessment and therefore may serve as an appropriate screening tool.

Purpose Although regions of stasis such as the left atrial appendage (LAA) may be potential sites of thrombus formation during atrial fibrillation (AF), its thrombotic contribution during high flow mechanical unloading is unknown. Moreover, registry studies only show a mixed association of AF and stroke during continuous flow left ventricular assist device (CF LVAD) support. Since a lower peak Left Atrial Appendage Emptying Velocity (LAAEV) of less than 40 cm/s during AF, mechanistically contributes to thrombosis and stroke in non-CF LVAD populations, we sought to evaluate the LAAEV in CF LVAD patients with and without AF. Methods We conducted a single center review of patients who had a measurement of the LAAEV by pulsed-waved Doppler by transesophageal echocardiography (TEE). To evaluate at the steady state CF LVAD speed, operating room studies were excluded. CF LVAD patients were categorized into groups with and without AF during TEE and LAAEVs were compared. Matched non-heart failure (HF) and HF patients with and without AF were included as controls. Results Thirty eight patients with a CF LVAD had a peak LAAEV measurement by TEE, of whom 15 were in AF. There were no differences in age (AF: 58±11 vs. no AF: 59±13 years, p=0.8) and gender (AF: 20% vs. no AF: 9% female, p=0.6). The peak LAAEV was significantly lower for CF LVAD patients in AF (33±18 cm/s) in comparison to those in sinus rhythm (SR) (60±20 cm/s, p Conclusion Despite mechanical unloading, the LAAEV remains reduced in patients with AF. Further studies are warranted to investigate the relation between reduced LAAEV and thromboembolism during CF LVAD support.

Purpose The pre-operative pulmonary artery pulsatility index (PAPi) is associated with severe right ventricular failure (RVF) after LVAD implantation. We set out to assess how the PAPi trends pre- to post-operatively in patients with and without severe RVF. Methods A single-center retrospective study of 230 patients who received a continuous-flow durable LVAD between 1/2006 and 9/2016. PAPi was defined as: [(PA systolic - PA diastolic) ÷ RA pressure]. Severe RVF was defined as meeting criteria for clinical RVF and having inotropes > 14 days after implant, inotropes re-started after 14 days of implant, RVAD placement during implant admission, and death from RVF during implant admission. The hemodynamics pre-implant and post-operatively at 6, 24, and 72 hours were examined. Results In our 230-patient cohort, 62 patients (27%) were found to have severe RVF. Patients with severe RVF had a lower pre-operative (p-value = 0.04), 6-hour post-LVAD (p-value Conclusion The PAPi declines significantly post-operatively in patients with severe RVF and without severe RVF at 6 and 24 hours. For patients with severe RVF, the PAPi significantly improves from 6 to 72 hours post-operatively.

Purpose Although intravenous anticoagulation is the mainstay medical therapy employed during CF LVAD thrombosis, the comparative impact of indirect thrombin inhibition with unfractionated heparin (UH) and direct thrombin inhibition with bivalirudin (BV) is unknown. Methods We conducted a single center review of all patients with a Heart Mate(HM) II who were admitted for suspected device thrombosis (SDT) from September 2011 to September 2018. Device thrombosis was suspected due to elevated outpatient lactate dehydrogenase (LDH), evidence of systemic emboli, or device alarms. After admission, patients were categorized into those receiving UH or BV. Crossovers were excluded. Freedom from device exchange was calculated with Kaplan Meier analysis. Results Twenty-three patients were admitted for SDT, of whom 13 received UH and 10 received BV. There were no differences in age (UH: 49.2 vs. BV: 46.9 years, p=0.60) and time from implant to SDT (UH: 10.3 vs. BV: 10.4 months, p=0.98). LDH remained elevated after UH (879, IQR: 755-1049 to 1028, IQR: 550-1438 U/L, p=0.34) but dropped with BV (839, IQR: 733-914 to 453, IQR: 352-465 U/L, p=0.002). During the follow up period, there was a lower likelihood of a device exchange in patients treated with BV (HR: 0.28, 95% CI: 0.09-0.89, p=0.04, figure 1). Conclusion In this sample of HM II patients, intravenous direct thrombin inhibition was more effective in treating device thrombosis.

Purpose Neutrophil to lymphocyte ratio (NLR), calculated from a CBC, is a marker of systemic inflammation and long-term risk in chronic illness, including heart failure (HF). Vasoplegia syndrome is a severe vasodilatory shock state after cardiac surgery. Patients undergoing heart transplantation (HTx) may be at an increased risk of vasoplegia due to inflammatory cytokine release secondary to HF. The role of NLR as a predictive marker for vasoplegia in patients undergoing HTx has not been studied. Methods Retrospective review of consecutive patients who underwent HTx from 7/2016 to 7/2018. Patients with conditions or treatments known to affect WBC count were excluded. Pre-HTx NLR was calculated from day of HTx and stratified by tertile. Vasoplegia was defined as vasopressor administration for > 24 hours to maintain mean arterial pressure > 65 mmHg for hypotension not attributed to other etiologies within 48 hours of HTx. The primary outcome was rates of vasoplegia between tertiles. Results 78 patients underwent HTx of which 70 met inclusion criteria. 18 patients had vasoplegia. Vasoplegia occurred in 8.7% (n=2) in the 1st tertile, 25% (n=6) in the 2nd tertile and 43.4% (n= 10) in the 3rd tertile (comparison 1st vs 3rd tertile, p=0.04) (Figure). In a multivariate analysis, adjusted for prior LVAD, patients in the 3rd tertile had higher rates of vasoplegia (adjusted OR 2.47, 95% CI 1.87-4.55) compared to the 1st. Mean NLR in patients without vasoplegia was 3.68±0.48 compared to 6.72±1.7 in patients with vasoplegia (p=0.019). There was no demographic or medical comorbidity difference other than hypothyroidism (p=0.04) between groups. Conclusion Vasoplegia is associated with elevated pre-HTx NLR compared to patients without vasoplegia. Chronic inflammation due to HF may play a role in the development of post-HTx vasoplegia. NLR is an inexpensive tool which clinicians may use pre-HTx to stratify which patients are at an increased risk of development of vasoplegia post-HTx.

Purpose The American Society of Transplantation has recently called for better risk stratification of donor hearts based on hepatitis C virus (HCV) antibody (Ab)/NAT status, advised to consider NAT negative (non-viremic) hearts noninfectious, and underscored the need for more research before HCV NAT+ (viremic) hearts can be used routinely. We examined early trends in utilization rates and outcomes of HCV donor hearts based on Ab and NAT status. Methods Between 8/2015 and 6/2018, we identified 29,225 donors with HCV Ab/NAT status and organ disposition in UNOS (for utilization trends). After excluding multi-organ and re-transplants, 7260 adult heart transplantation (HT) recipients were used to analyze baseline donor/recipient characteristics and outcomes in different donor HCV Ab/NAT categories. Results During the study period, the acceptance rates for HCV Ab+/NAT- donor hearts increased from 1.4% to 23.4%, Ab+/NAT+ from 0.6% to 26.0% and Ab-/NAT+ from 0% to 9.1% (p Conclusion Even though infectious risk starkly differs and long-term outcomes are unclear in the era of DAAs, we found a significant and near identical increase in utilization of both HCV viremic (NAT+) and non-viremic (Ab+/NAT-) donor hearts. Early survival was similar to recipients of HCV Ab-/NAT- donor hearts.

Purpose Early identification of right ventricular failure (RVF) after LVAD implantation is critical to reducing morbidity and mortality. The pre-operative pulmonary artery pulsatility index (PAPi) is associated with severe right ventricular failure (RVF) after LVAD implantation. We set out to assess if the PAPi in the immediate post-operative period is also predictive of severe RVF. Methods A single-center retrospective study of 230 patients who received a continuous-flow durable LVAD between 1/2006 and 9/2016. PAPi was defined as: [(PA systolic - PA diastolic) ÷ RA pressure]. Severe RVF was defined as meeting criteria for clinical RVF and having inotropes > 14 days after implant, inotropes re-started after 14 days of implant, RVAD placement during implant admission, or death from RVF during implant admission. The hemodynamics pre-implant and post-operatively were documented. A multivariate analysis for predictors of severe RVF was performed, and ROC curves were created. Results In our 230-patient cohort, 62 patients (27%) were found to have severe RVF. Patients in the severe RVF group had higher INTERMACS profile, were more likely to receive pre-operative vasopressors, have lower total bilirubin and lower PAPi. The multivariable model (which included age, INTERMACS level, creatinine over 1.5 mg/dL, total bilirubin over 2.5 mg/dL, and gender) found that a PAPi 16 at 6 hours post-op (OR 3.1 [1.1, 8.6], p = 0.03, n = 208) were significant predictors of severe RVF. The pulmonary artery (PA) pressure at 6 hours and the CVP, PAPi, and PA pressure at 24 hours were not predictive of severe RVF in our model. A PAPi 16 (ROC c-statistic 0.71) at 6 hours post-op. Conclusion The PAPi at 6 hours is a significant predictor of severe RVF and adds value to standard hemodynamic measurements.

Purpose CAV has an immunological component and continues to limit long term outcomes in heart transplantation (HT). Smaller single center studies have suggested that induction with r-ATG may reduce development of CAV. We evaluated the association of induction with r-ATG, Basiliximab (BxB), or ‘no induction’ and CAV in HT recipients, traditionally considered not highly sensitized [pre-transplant panel reactive antibody (PRA) class 1 and class 2 ≤ 10%]. Methods Between 6/2004 and 03/2015, we identified 4,654 adult HT recipients in UNOS who either had induction with r-ATG, BxB or ‘no induction’ at the time of HT, and had information on pre-HT PRA levels and CAV. Donor age above 55 years or structural abnormalities, multi-organ or repeat transplants were excluded. The 3 groups: r-ATG (n=644), BxB (n=991), no induction (n=3019) were compared for baseline donor/recipient characteristics and CAV. Results Overall, pre-transplant PRA 1 and 2 levels were 0.5±1.7% and 0.3±1.3% respectively. Compared to no induction or BxB, HT recipients in r-ATG group were slightly younger (53 vs 55 vs 56 yrs), more black race (23% vs 16% vs 22%), less Status 1A (49% vs 59 vs 62%). BxB group had more recipients with creatinine >1.5 (24% vs 23% in r-ARG vs 17% no-induction) (all p Conclusion Although rates of CAV were higher in our specific cohort than previously reported for the overall HT population, r-ATG was associated with lower CAV in HT recipients who are traditionally considered to be not highly sensitized.

Purpose Pre-transplant amiodarone and left ventricular assist device (LVAD) has been shown to be independently associated with severe PGD; however, the interaction between both exposures has not been examined. We evaluated the relationship of pre-transplant amiodarone, left ventricular assist device and severe PGD. Methods Single center, retrospective cohort, adult (age > 18) OHT patients between 2006-2018. Severe PGD was defined as left or right or biventricular failure requiring mechanical circulatory support within 24 hours of OHT. Pre-transplant amiodarone use was within 3 months preceding OHT. Patients were grouped by presence (denoted +) or absence (denoted -) of LVAD and amiodarone. Results A total of 257 patients were transplanted during the study period, mean age 53.8 years, 73 (28.4%) females, 162 (63%) BTT and 90 (35%) patients had amiodarone use. There were 21 (8.2%) patients with severe PGD. The prevalence of severe PGD was highest in LVAD+/Amiodarone+ group (20.3%), followed by LVAD+/Amiodarone - (6.8%), LVAD-/Amiodarone+ (3.2%) and LVAD-/Amiodarone- (1.6%), p=0.0008 (Figure 1). Multivariate logistic regression analysis of pre-operative risk factors revealed LVAD (OR 5.95; 95% CI 1.23-28.52; p=0.03), amiodarone (OR 3.52; 95% CI 1.26-9.87; p=0.02), recipient diabetes (OR 3.36; 95% CI 1.11-10.21; p=0.03) and recipient hypertension (OR 0.22; 95% CI 0.08-0.61; p=0.04) were independently associated with severe PGD. Additive interaction analysis of amiodarone and LVAD was inconclusive due to the small sample size [relative excess risk due to interaction 8.13; 95% CI -10.37-26.2; p=0.39, attributable proportion 0.66; 95% CI 0.19-1.13; p=0.01 and synergy index 3.61; 95%CI 0.49-26.32; p=0.21]. Conclusion LVAD and amiodarone are associated with severe PGD, however the interaction of both exposures requires further analysis with a larger cohort.