Egbe, Alexander C.

Abstract Background The landscape of heart transplantation (HT) has changed significantly with respect to patient selection, surgical techniques, and patient outcomes. We sought to investigate temporal trends in patient characteristics, waitlist and post-transplant outcomes after HT in the U.S. Methods We queried the national database of the United Network of Organ Sharing (UNOS) to identify adults listed for HT in the U.S. between 1991 and 2019. Patients were divided into four eras based on the three time points in which changes were made to the patient selection/allocation policy (1999, 2006, and 2018), and patient characteristics as well as waitlist and post-transplant outcomes were evaluated for each era. Results Between 1991 and 2019, a total of 95,179 patients were added to the waitlist for HT in the U.S. Compared to era 1, patients listed in era 4 were older (mean age: 52.4 years), more female (27.6%) and ethnic minorities (40%), and with higher-risk comorbidities (28.8% diabetes, 35.6% obese). Over the study period, there were 22,070 waitlist deaths and 61,687 transplants. Compared to the preceding era, there was significant decrease in waitlist mortality in the last 2 eras (e.g., sub-hazard ratio for era 4 vs era 3 =0.37, 95% CI=0.32–0.44). For each year, only 27.1% to 40.5% of those on the waitlist were transplanted. Among those who were transplanted, there was increase in the rates of in-hospital stroke (2.8% in era 1 to 3.7% in era 4), renal failure requiring dialysis (7.2% to 17.1%), and hospital length of stay (14 to 17 days), p-values<0.001 for all. However, this has not negatively impacted short-term survival when compared to the preceding era (1-year graft survival = 89.7% in era 4). Based on a projection model, we predict a 47% increase in living adult heart transplant recipients to to 44,366 in 2040. Conclusion There have been significant changes in the characteristics of patients listed for HT in the U.S., including an increasing proportion of high-risk co-morbidities. Although the transplant volume has increased, the wide supply-demand gap persisted. The last 2 changes in the allocation policy in 2006 and 2018 achieved their primary objective of reducing waitlist mortality. Funding Acknowledgement Type of funding sources: None.

Abstract Background There are limited data about outcomes after repair of native coarctation of aorta (COA) in adulthood. The purpose of this study was to describe the procedural outcomes, hemodynamic improvement, regression of LV hypertrophy and cardiovascular events in adults undergoing repair of native COA. Methods The primary outcomes were procedural complications, re-interventions, and hemodynamic improvement (residual COA gradient <20 mmHg) after COA repair. The secondary outcomes were improvement in the severity of hypertension, regression of left ventricular mass index (LVMI), and incidence of cardiovascular events (atrial fibrillation, heart failure hospitalization, and cardiovascular death) after COA repair. Clinical and imaging indices were assessed pre-intervention, and at 1-year (Y1) and 3-years (Y3) post-intervention. Results A total of 172 patients (age 38 [27–48]) underwent COA repair (surgical 161; transcatheter 11). There were no procedural deaths, and all patients had residual COA gradient <20 mmHg. One patients that received transcatheter stent therapy required re-dilation of stent at 12 months post-intervention. There as a reduction in the prevalence of patients requiring anti-hypertensive therapy from 73% (pre-intervention) to 59% and 64% at Y1 and Y3 respectively. However, 72% and 69% of the patients still had systolic blood pressure >130 mmHg (stage 1 and 2 hypertension) at Y1 and Y3 respectively. As compared to patients without hypertension at Y1, patients with stage 1 and stage 2 hypertension had less robust LVMI regression (% change in LVMI 12±5% vs 9±6% vs 5±3%, p<0.001). Hypertension (HR 1.16, 1.05–1.27) and LVMI regression (HR 0.86, 0.81–0.90) were independently associated with cardiovascular events. Conclusions Persistent hypertension was common after repair of native COA in adults. Hypertension (including stage 1 hypertension) was associated with suboptimal regression of LVMI and cardiovascular events. These results are concerning, and highlight the importance of early COA diagnosis and repair, and optimal medical therapy for hypertension after COA repair. Funding Acknowledgement Type of funding sources: None.

Abstract Background Certain factors such as left ventricular (LV) geometry and loading conditions affect the validity and reliability of LV ejection fraction (LVEF) as a true measure of LV contractility. LV global longitudinal strain (LVGLS) is less sensitive to these factors, and it has superior prognostic performance in patients with acquired heart disease. The purpose of this study was to determine the clinical implications of using LVGLS (instead of LVEF) as the measure of LV systolic function in adults with Ebstein anomaly given the inherent abnormalities of LV geometry and preload in this population. Methods Retrospective cohort study of 673 adults with Ebstein anomaly (2003–2018). We hypothesized that LVGLS had a stronger correlation with heart failure indices and transplant-free survival compared to LVEF. Results Compared to LVEF, LVGLS had stronger correlations with cardiac index (r=0.46 vs r=0.21, p=0.007), glomerular filtration rate (r=0.57 vs r=0.19, p<0.001), and NT-proBNP (r=−0.64 vs r=−0.41, p=0.01). Of 673 patients, 514 (76%) had normal LV systolic function (LVGLSNormal-LVEFNormal), 87 (13%) had subclinical LV dysfunction (LVGLSLow-LVEFNormal) and 66 (10%) had overt LV dysfunction (LVGLSLow-LVEFLow). Compared to the overt LV dysfunction group, the subclinical LV dysfunction group had similar 10-year transplant-free survival (64% vs 63%, p=0.6), but were less likely to be on heart failure therapy (12% vs 82%, p<0.001). LVGLS (but not LVEF) was the independent predictor of transplant-free survival Conclusions LVGLS provided more robust risk stratification and prognostication than LVEF in patients with Ebstein anomaly, and patients with low LVGLS had reduced transplant-free survival regardless of LVEF. The use of LVGLS (rather than LVEF) as the measure of LV systolic function has important clinical implications with regards to initiation of medical and surgical therapies. Funding Acknowledgement Type of funding sources: None.

AbstractAimsThe prognostic implication of left atrial (LA) dysfunction and left ventricular diastolic dysfunction (LVDD) in patients with coarctation of aorta (COA) is unknown. The purpose of this study was to determine whether LA dysfunction and LVDD were associated with mortality in COA patients.Methods and resultsThis is a retrospective review of adults (age ≥18 years) with repaired COA that underwent transthoracic echocardiogram (2000–18). LVDD was determined using the 2016 guidelines for LV diastolic function assessment, and LA dysfunction was assessed using LA reservoir strain. Of 721 patients, LV diastolic function could be determined in 635 (88%); and 414 (65%) had no LVDD, while 146 (23%), 53 (8%), and 22 (4%) had Grade I/II/III LVDD, respectively. The mean LA reservoir strain was 39 ± 11%, and patients were divided into quartiles: top quartile (reference group), mild LA dysfunction, moderate LA dysfunction, and severe LA dysfunction. Grade III LVDD (but not Grades I and II) was associated with death/transplant. On the other hand, there was an incremental risk of death/transplant across LA strain quartiles: mild LA dysfunction [hazard ratio (HR) 1.16, 1.04–2.06], moderate LA dysfunction (HR 1.75, 1.27–3.58), and severe LA dysfunction (HR 3.49, 1.88–7.16). Of 86 patients with indeterminate diastolic function, there was a trend towards a lower 5-year transplant-free survival in patients with LA dysfunction vs. normal LA function (83% vs. 91%, P = 0.06).ConclusionLA dysfunction (but not LVDD) was associated with incremental risk of mortality and thus can be used for prognostication in all patients including those with indeterminate diastolic function.

The purpose of this study was to describe procedural outcomes, hemodynamic improvement, regression of left ventricular (LV) mass hypertrophy, and cardiovascular. The primary outcomes were procedural complications, reinterventions, and hemodynamic improvement after coarctation of aorta (COA) repair. The secondary outcomes were improvement in the severity of hypertension, regression of LV mass index, and incidence of cardiovascular events (atrial fibrillation, ventricular tachycardia, heart failure hospitalization, and cardiovascular death) after COA repair. Secondary outcomes were assessed only in patients with isolated COA who had clinical and imaging follow-up at 1 year and 3 years postintervention. Of 172 patients that underwent COA repair (surgical 161; transcatheter 11), there were no procedural deaths, and all patients had residual COA gradient <20 mm Hg. Of 128 patients that met criteria for secondary outcomes assessment, 39 (36%) had a reduction in the intensity of antihypertension therapy, and cardiovascular events occurred in 16 (13%) patients. There was no significant reduction in the overall prevalence of hypertension (stage 1 and stage 2) over time (78% versus 70% versus 73%, P =0.4 at baseline, 1 year and 3 years). Postintervention hypertension (both stage 1 and 2) were independent risk factors for suboptimal left ventricular mass index regression and cardiovascular events. Persistent hypertension was common after repair of native COA in adults and was associated with suboptimal left ventricular mass index regression and cardiovascular events. These results suggest that optimal blood pressure control with medical therapy after COA repair may result in improved clinical outcomes.

Background: Prior studies have shown that hospital case volume is not associated with survival in patients with out-of-hospital cardiac arrest (OHCA). However, how case volume impact on survival for in-hospital cardiac arrest (IHCA) is unknown Methods: We queried the National Inpatient Sample (NIS) in the U.S. 2005-2011 to identify cases in which in-hospital CPR was performed for IHCA. Years > 2011 were excluded due to redesign of NIS that precludes case volume calculation. Restricted cubic spine was used to evaluate the association between hospital annual CPR volume and survival to hospital discharge. Further analysis was conducted with case volume in quartiles. Results: Across more than 1000 hospitals in NIS, we identified 137,466 cases (mean age 67, 45% female) of IHCA for which CPR was performed over the study period. Median [Q1, Q3] case volume was 68 [34, 100]. Compared to those in the 1 st quartile (q1) of case volume, hospitals in the 4 th quartile (q4) tends to have younger patients (mean=66 vs 68 yrs), higher comorbidities (median Elixhauser score=4 vs 3), large bed size (89 vs 37%), and in low income areas (37 vs 32%). After adjustment for patient and hospital factors, higher hospital case volume was inversely associated with lower rates of survival (Fig1). Risk-adjusted rates of survival from q1 through q4 were: 28.1%, 25.5%, 25.4% and 23.7%. We arrived at similar results when analysis was stratified by age, gender and hospital bed size (not shown) except that the difference across quartiles was more pronounced among small bed size hospitals (survival rate=27.4% in q1 vs 8.1% in q4). However, among those who survived to hospital discharge, nonroutine home discharge was higher among patients in q1 (73.9%) vs q4 (69.6%) Conclusion: Unlike OHCA, hospital case volume is inversely associated with survival to hospital discharge in patients undergoing CPR for IHCA. Hence, quality measures to reduce the incidence of CPR/IHCA have the potential to reduce in-hospital mortality

Introduction: Bioprosthetic valve thrombosis (BPVT) is increasingly being recognized as a cause of prosthetic valve dysfunction within the first years after surgical implantation. However, in the absence of changes in clinical status, current guidelines recommend cardiac imaging only after 10 years post bioprosthetic valve implantation . Hypothesis: We hypothesized that an increase in prosthetic valve gradient is detectable months before a formal diagnosis of symptomatic BPVT. Methods: Eighty-three patients diagnosed with BPVT and treated medically at our institution between 2002 and 2017 were retrospectively identified. Transvalvular gradients recorded on TTE immediately post implantation, at time of clinical BPVT diagnosis, and within a 12-month period prior to BPVT diagnosis were available in 45 patients: 18 for the aortic position, 11 for mitral, and 16 for tricuspid. When more than one study was conducted in the 12 months’ timeframe, the latest value was used in the analysis. Data are reported as mean (95% CI). Results: For all valve positions, an increased gradient was detectable within 12 months prior to formal BPVT diagnosis: mean increase vs. baseline of 17 (10-23) mmHg, p < 0.0001 for aortic prostheses; 5 (2-8) mmHg, p = 0.004 for mitral prostheses, and 2 mm Hg (1-3) mmHg, p = 0.003 for tricuspid prostheses. The figure presents the evolution of prosthetic valve gradients (normalized to the baseline, immediately post-implantation value) in patients with aortic (left), mitral (middle), and tricuspid (right) valve . Conclusions: Elevated gradients precede formal clinical diagnosis of BPVT. Closer monitoring of transvalvular gradients may prove useful for BPVT diagnosis before patients become symptomatic. With recent evidence supporting the efficacy of anticoagulant therapy in BPVT, earlier diagnosis may lead to an increased rate of response and better clinical outcomes.

Objective Bioprosthetic valve thrombosis (BPVT) is increasingly recognised as a major cause of prosthetic dysfunction in the first years postimplantation. How early abnormal gradients can be detected prior to diagnosis and how fast they normalise with anticoagulant therapy is unknown. We set forth to (1) evaluate patterns of increase in gradients prior to BPVT diagnosis and (2) characterise time-course of response to anticoagulation. Methods Patients treated with warfarin for BPVT (1999–2019) with clinically significant reduction of mean gradients (≥25%) were identified retrospectively. Recovery was defined as gradient decrease ≥50%, to postimplantation or to normal-range gradients per position, model and size. Time-to-BPVT (implantation—BPVT diagnosis), potential diagnostic delay (first abnormal gradient by position, model and size—BPVT diagnosis) and time-to-recovery (BPVT diagnosis—complete resolution) were recorded. Results 77 patients were identified; 32 (42%) aortic (23 surgical—12 porcine, 11 pericardial; 9 transcatheter); 24 (31%) mitral; 21 (27%) tricuspid. Median time-to-BPVT was 24, 21 and 10 months, respectively. Potential diagnostic delay was median 21 months for aortic, 4 months for mitral, but 0 for tricuspid. Recovery was significantly faster in mitral than aortic (median 2.5 vs 4.8 months, p=0.038) and tricuspid (median 5.9 months, p=0.025) positions. Porcine aortic valves responded faster than pericardial aortic valves (median 2.9 vs 20.3 months, p=0.004). Conclusion Gradients start to increase months before the clinical BPVT diagnosis. Recovery is faster in mitral and surgical aortic porcine valves; a longer warfarin trial seems indicated in tricuspid and surgical aortic pericardial valves.

Background: Warfarin is known to be effective in the treatment of bioprosthetic valve thrombosis (BPVT). However, the time-course of response to anticoagulation has not been fully characterized. Hypothesis: We hypothesized that response to warfarin can differ depending on valve position. Methods: We retrospectively identified 45 patients with BPVT diagnosis between 2002 and 2019 that involved either the mitral (MV) or the aortic valve (AV) prostheses, all with eventual echocardiographic resolution with warfarin. BPVT involved AV in 26 (57.8%) patients and MV in 19 (42.2%). Mean gradients (MGs) were tracked starting from BPVT diagnosis until recovery, which was defined as MGs decreasing to initial post-implantation MGs, to 50% below MGs at BPVT diagnosis, or to < 20 mmHg in AV and < 5 mmHg in MV. Kaplan Meier analysis was used to compare recovery time between MV and AV. Duration is reported relative to the time of BPVT diagnosis. Results: Recovery was faster in MV than in AV patients treated with warfarin, with median time of 3.8 months in the AV and 1.8 months in the MV group, p < 0.001 (Figure 1). At 1 month follow up, 15.4% recovered in AV group and 21.1% in MV, p = 0.605. At 3 months, 38.5% recovered in the AV and 78.9% in the MV, p = 0.009. At 6 months, 57.7% of AV group recovered while 100% recovered in the MV group, p = 0.001. The longest time for recovery was 30 months in 1 AV patient. Conclusion: Recovery using warfarin tended to be faster in BPVT involving the MV compared with the AV. In our study, all patients with MV BPVT recovered by 6 months, while this only happened in half the patients with AV BPVT. This suggests that a longer duration of warfarin trial may be needed in the case of AV disease. This might also suggest that the pathogenesis may differ between the two valves.