ISHLT2022 Roving Reporters – Reports from Advanced Heart Failure and Transplantation

Saturday, 30 April, 2022

Coronary Allograft Vasculopathy (CAV) impacts ~50% of patients at 10 years and is the leading reason for re-transplant. Prevention of its onset and progression with the use of aspirin and statin remain the mainstay of CAV therapy. The investigators of this study hypothesized that PCSK9i will significantly lower LDL and be well-tolerated in transplant patients with CAV.

In a phase II, open label, single center clinical trial, they investigated the impact of the PCSK9i evolocumab in heart transplant patients with CAV. Twenty-six heart transplant recipients, aged 19 to 80 years and with CAV diagnosed on coronary angiography, were enrolled. Patients with recent rejection or infection requiring IV therapy (within the past 3 months), acute liver dysfunction, renal dysfunction with GFR less than 20 ml/min, current or recent use of a PCSK9 inhibitor (within the past 3 months), and/or known allergy to evolocumab, were excluded.

The key findings of this study were:
  1. Evolocumab markedly reduced LDL in heart transplant patients with CAV
  2. No angiographic progression of CAV was observed during the study period, and a trend towards CAV regression was noted in the LAD
  3. Evolocumab exhibited an acceptable safety profile in this specialized patient cohort
– Summary by Rachna Kataria, MD

This session included talks focused on advances in microbiome research, highlighting the reciprocal relationships between the microbiome and other organs including the heart and lung.

Send In Your Samples: What Can Be Learned From Analyzing Your Microbiome?
Tereza Martinu, MD
, University of Toronto, Toronto, ON Canada
The microbiome is a collection of microbes that colonize the body at barrier surfaces and therefore: 1) is highly influenced by environmental and transplant-related factors; 2) significantly influences organ function and transplant-related outcomes (e.g., rejection); 3) has important immunomodulatory effects (e.g., Th17 & Treg).

Important areas of future focus in the study of the microbiome include: 1) Effects of specific microbial perturbations; 2) Timing; 3) Mechanisms; and 4) Virome and fungome (in addition to bacterial microbiome).

The possible clinical applications of the microbiome include: 1) its use as a potential diagnostic tool for personalized assessment and risk stratification; 2) its use for manipulation of the existing microbiome; and 3) its use in prevention and treatment.

Gut Microbiota and Metabolism of Immunosuppressive Drugs: One Size Does Not Fit All
Douglas L. Jennings, PharmD
, New York Presbyterian Hospital, New York, NY USA
Alterations in gut microbiome, as seen in patients with heart failure, persist even after LVAD and heart transplant. The obvious question, then, is whether these alterations impact the metabolism of standard immunosuppressive medications used by heart transplant patients. The short answer is, yes: very much so. Greater diversity of gut flora, higher abundance of individual taxa, and lower levels of inflammation and oxidative stress have all been associated with higher tacrolimus dosing requirements.

As for mycophenolate metabolism, beta-D-glucuronidase-producing bacteria were found to be associated with MMF-induced toxicities such as cytopenia. Preliminary data certainly raises concern for important interplay between gut microbiome and transplant immunosuppression. However, additional studies are warranted to determine the clinical applicability of these preliminary findings.

Can Transplantation or LVAD Change My Microbiome?
Melana Yuzefpolskaya, MD
, Columbia University, New York, NY USA
Circulating levels of pro-inflammatory cytokines increase with the severity of heart failure symptoms and reduction in functional capacity. However, targeting individual cytokines in heart failure has not yielded promising results. Hence, we should shift our focus to more upstream mediators of inflammation.

In heart failure, both decreased cardiac output and increased venous congestion can result in increased intestinal permeability, decreased gut microbiome, and decreased nutrient absorption. Worsening heart failure is associated with increasing severity of inflammation and depletion of essential gut microbiota. This reduction in gut microbiome diversity, endotoxemia, and increased TMAO levels persist even after LVAD and heart transplantation, and in fact is associated with increased risk and severity of post-LVAD infections and difficulty attaining target tacrolimus levels. As a result, there is increasing enthusiasm around investigating fecal microbial transplant (FMT) as a potential solution to this conundrum.

The Gut Microbiome and Heart Failure: Are We Ready for Therapeutic Trials Targeting Microbiome?
David Kaye, MD, PhD
, The Alfred Hospital, Melbourne, Australia
The role of gut dysbiosis (gut integrity + microflora) has been clearly implicated in many disorders, including cardiovascular diseases. Both the gut microflora and integrity of the gut wall, in particular its ability to prevent leakage of inflammatory cytokines, change in health and disease. In heart failure, both hypoperfusion and gut wall edema affect gut wall integrity.

Animal models have shown that gut microbiome can play a role as a mediator of cardiovascular (CV) physiology. Interestingly, human studies involving patients with HFpEF and those with hypertension have shown a striking decrease in bacteria that produce the short chain fatty acid butyrate, which is normally associated with reduced systemic blood pressure.

Key metabolic targets for the gut microbiome and that influence the CV phenotype include: dietary fiber, amino acids, bile acids, dietary choline, phosphatidylC, lysine, and carnitine. Simple “untargeted” interventions such as diet, fecal microbial transplant (FMT), and targeted interventions such as short chain fatty acid supplements and enzyme inhibition, may have some role in CV disease.
– Summary by Rachna Kataria, MD
Risky Business: Primary Graft Dysfunction and Risk Modeling
Javier Segovia, MD, PhD
, Hospital Puerta de Hierro, Madrid, Spain
Primary graft dysfunction is related with frequent, unexpected high lethality, resource consumption with poor results, and frustration amongst transplant physicians. There remains an unmet need for tools to predict, diagnose and treat PGD. Various risk scores used in heart transplantation are RADIAL score which includes recipient age > 60 years, DM, inotrope therapy, RAP > 10 mm hg, donor age > 30 years, ischemic time > 240 mins.

2013 ISHLT Consensus on PGD (Montreal) includes:
  1. PGD- left ventricle- includes left and biventricular dysfunction
    1. Mild PGD-LV- One of the following criteria- LVEF < 40%, or hemodynamics with RAP > 15 mm Hg, PCWP > 20 mm Hg, CI < 2.0 L/min/m2 lasting more than 1 hr requiring low-dose inotropes
    2. Moderate PGD-LV- One criterion from I, and 1 from II- (I) LVEF < 40%, or hemodynamics with RAP > 15 m Hg, PCWP > 20 mm Hg, CI < 2.0 l/min/m2, MAP < 70 mm Hg lasting more than 1 hr and One criteria from following- high dose inotropes or newly placed IABP
    3. Severe PGD- LV- Dependence on left or biventricular mechanical support including ECMO, LVAD, BiVAD or percutaneous LVAD
  2. PGD- right ventricle- includes right ventricular dysfunction alone. PGD-RV- Diagnosis requires either both I, ii, or iii alone:
    1. hemodynamics with RAP > 15 mm Hg, PCWP < 15 mm Hg, CI < 2.0 L/min/m2
    2. TPG < 15 mm Hg and/or PASP < 50 mm Hg, or
    3. Need for RVAD
Secondary graft dysfunction occurs when there is discernible cause for graft dysfunction (e.g., hyperacute rejection, pulmonary hypertension, known surgical complication). Donor risk factors include age, cause of death, trauma, cardiac dysfunction, inotropic support, and comorbidities: DM, HTN, downtime of cardiac arrest , drug abuse, LVH, valvular disease, Hormone treatment, CAD/wall motion abnormalities, sepsis, marginal donors, and hypernatremia.

Recipient risk factors include age, weight, mechanical support, congenital heart disease, multiple reoperations, LVAD explant, comorbidities like renal and liver dysfunction, ventilator dependence, multiorgan transplant, elevated PVR, allosensitization, infection and retransplant.

Procedural risk factors include ischemia time, donor recipient size mismatch, non-cardiac organ donation, experience of organ procurement and center volume, cardioplegic solution, increased blood transfusion requirement, elective vs emergency transplant.

Other scores used are PREDICTA score: Preoperative MCS, Recipient DM, Cardiopulmonary bypass time > 180 min, Implant time, donor Age. ABCE score-treatment with ACE/ARB/ARNI/MRA, treatment with amiodarone and BB, previous cardiac surgery and ischemic time.

PGD has more than 30% hospital mortality and 40% one-year mortality, but good long term outcomes. Predictive models to date have relative value. However, we must be aware of characteristics associated with PGD. The new models using AI may overcome some of the limitations of classical ones- PGD consortium and studies based on pathophysiology of PGD may find these potential risks.

Rise of Machines: PGD Risk Scoring Systems in the MCS Bridge to Transplant Population
Lauren Truby, MD
, Duke University Medical Center, Durham, NC USA
Despite ongoing improvements in waitlist outcomes, heart transplant recipients remain at high risk for early post-operative mortality. PGD is a devastating clinical event that is responsible for 60% of deaths in the first 30 days following heart transplantation. It is broadly defined as transient, acute failure of the new allograft to support recipient circulation. Despite recovery of biventricular function in 24-72 hours, it leads to multi-organ failure, prolonged intubation, and significant morbidity. There is increased recognition that patients bridged to transplant with LVAD are at increased risk of PGD.

Truby et al sought to review current evidence for the association between BTT LVADs and PGD, and to discuss the early evidence for PGD risk associated with tMCS in the context of the new allocation system. The RADIAL score predicts incident PGD, and was developed and validated in Europe. Their definition was similar to ISHLT’s: hemodynamic compromise necessitating high dose inotropes of MCS. There was very low utilization of pre-transplant IABP or LVAD with poor generalizability to the United States’ heart transplant landscape. BTT LVAD is associated with early post-transplant mortality and in a propensity matched cohort, the major cause of one-year cardiovascular mortality in propensity matched cohort was post graft failure.

Various predictors of ISHLT severe primary graft dysfunction in BTT patients is Creatinine, CVP/PCWP ratio, use of amiodarone, and more than one year of continuous flow LVAD support. Clinical features of high risk BTT recipients include recipient age, country, pre-transplant hemodialysis, non-ischemic CMP, and pre HT-CVP/PCWP ratio. Pre-HT VA-ECMO support increases the risk of PGD.

There are many questions that remain unanswered, but in conclusion, PGD remains a substantial clinical challenge following HT, and clinical risk factors alone fail to capture the spectrum of PGD risk in the rapidly changing landscape of transplant. Its underlying mechanism remains poorly understood. Precision medicine approaches hold great potential for the identification of novel biomarkers of PGD risk in the donor and recipient, and have lent insight into potential mechanisms that warrant additional investigation.

Danger Zone: How to Risk Stratify DCD Donors for Primary Graft Dysfunction
Simon Messer, MD
, Royal Papworth Hospital, Cambridge, UK
Risk stratifying DCD donor hearts – normothermic regional perfusion, functional warm ischemic time, perfusate lactate predicting ECMO, Coronary vascular resistance

The Papworth series NRP showed 100% one-year survival, with direct procurement and perfusion. DPP0 showed 86% survival with 18% ECMO with DPP, Sydney had 31% ECMO use. Normothermic regional perfusion: NYU did 7 DCD, no ECMO with 100% survival. Vanderbilt did 15 DCD hearts with no ECMO with 100% survival.

Serum lactate is a highly sensitive and specific predictor of post cardiac transplant outcomes using the organ care system. But lactate during ex situ heart perfusion does not predict the requirement for MCS following DCD heart transplants.

Mohite et al noted that the diastolic pressure and the coronary vascular resistance fell after 30 minutes of organ care system perfusion in the donor hearts that did not suffer PGD, whereas it remained high in the donor hearts that suffered PGD. In conclusion, NRP may reduce the risk of severe PGD in DCD heart transplant recipients. Prolonging the functional warm ischemic time beyond 30 mins may be associated with higher one-year mortality. Prolonged asystole to cardioplegia may be associated with severe PGD. Lactate did not seem to predict the requirement for ECMO post DCD heart transplant. Coronary vascular resistance may be a useful tool in predicting severe PGD in future.

Demystifying Vasoplegia After Heart Transplant: Predictor or Consequence of Primary Graft Dysfunction?
Fabiana Marcondes-Braga MD, PhD
, Heart Institute Hospital, Sao Paolo, Brazil
Vasoplegia may be defined as severe hypotension (MAP 50 mm Hg), low systemic vascular resistance, or normal or high cardiac index (CI > 2.5 L/min/m2) in the first 48 hours after HT. It is refractory to vasopressor therapies (IV norepinephrine > 0.5 g/kg/min), and happens immediately after cardiac surgery or within 6-48 hours of weaning cardiopulmonary bypass. Mechanisms include increased systemic inflammation and endothelial dysfunction that leads to persistent hypotension, reduced systemic vascular resistance with normal or increased cardiac output.

Risk factors of vasoplegia after heart transplant include mechanical circulatory support pre-HT, prolonged cross clamp time, prolonged cardiopulmonary bypass time, and large transfusion requirements. Its treatment includes vasopressors to improve vascular tone and restore an adequate perfusion pressure with catecholamines, vasopressin, and methylene blue.

Hostile environment of the recipient results in multiple proinflammatory cytokines and overproduction of NO and other vasodilators. PGD pathogenesis involves:
  1. Brain death release of NE-> Calcium overload-> impaired myocardial contractility
  2. Low levels of T3/cortisol/insulin leading to depression of myocardial contractility
  3. Reperfusion causing calcium overload -> impaired contractility
  4. Older donor heart-> susceptible to ischemic injury
  5. High PVR in recipient
  6. Inflammatory response refractory to conventional vasopressors
Treatment includes low dose of inotropes; high dose of inotropes; or IABP, ECMO, LVAD and BiVAD. Vasoplegia is a risk factor for mortality after HT. However, in patients with no PGD, vasoplegia did not have much impact on outcomes.

Thus, we see that vasoplegia and PGD have common physiological pathways. Many risk factors of both are similar and include pre-transplant MCS, prolonged cross clamp and cardiopulmonary bypass times, and large transfusion requirements. Both are associated with poor prognosis. Vasoplegia may not have much impact on outcomes in absence of PGD. The nature of interplay between vasoplegia and PGD is not completely known.

The Age of Artificial Intelligence (AI): Novel Primary Graft Dysfunction Risk Scores Using Machine Learning
Yasbanoo Moayedi, MD, FRCPC
, University Health Network, Toronto, ON Canada
Artificial intelligence is simply defined as a computer system that is able to perform tasks that normally require human intelligence. It involves machine learning, both supervised and unsupervised, and can be classified further. AI increases discriminatory power.

Calibration is defined as the ability of the model to assign an average risk of outcome accurately to a population. When observed outcome is more than estimated, it’s underestimating risk, and when it is less than estimated risk, it is overestimating risk. Overfitting is the failure to learn rules that are generalizable; ML algorithms learn rules that perform well on training data but fail on test data. Explainability is a debate over using ‘black-box’ algorithms with 99% accuracy vs 80% decision tree with recognizable features.

PGD is a heterogeneous entity, has high dimensional data, with unknown distributions as current risk scores may not capture complexity, and the data is time-varying. “There is huge potential for machine learning to transform healthcare, but going from ‘code to clinic’ is the hard part.” Clinicians need to learn AI but don’t need to be experts. PGD is a high value problem for machine learning. The PGD consortium that includes 12 international centers provides a first opportunity to identify and predict PGD using more granular data. To achieve standard of care, ML algorithms need to be generalizable and vigorously tested.
– Summary by Anju Bhardwaj, MD

Changing Landscape of Adult Congenital Heart Disease: Need for Advanced Therapies
Ari Cedars, MD
, Johns Hopkins University, Baltimore, MD USA
As the number of individuals with adult congenital heart disease (CHD) increases, their early surgical correction and hearts with defects could fail, needing replacement strategies. Adults with CHD die young, and the number of their hospitalizations is increasing. Heart failure (HF) remains the leading cause of mortality in these patients, and there is no pharmacological therapy available for them.

The most common underlying causes of death differed by lesion severity. Those with severe lesions most commonly died from underlying CHD, whereas those with non-severe disease more commonly died from non‐CHD causes. Because of the unique features of HF in CHD as outlined, these patients should be evaluated and managed by or in consultation with cardiologists and cardiac surgeons with expertise in CHD, ideally at a center with expertise in both CHD and HF. Providers should have a thorough knowledge of an individual patient’s anatomy and physiology, which requires a thorough review of all surgical and procedural records.

How to Identify Advanced Heart Failure in Adult Congenital Heart Patients: Challenges in Identifying HF in ACHD
Rose Tompkins, MD
, Cedars-Sinai Medical Center, Los Angeles, CA USA
The number of adult patients with congenital heart disease is increasing. Heart failure is the leading cause of late mortality in adult congenital heart disease. There is no universally accepted definition for ACHD HF, as their hearts are never structurally normal, they exhibit nonspecific signs and symptoms, and most ACHD patients have high NT-proBNP levels.

The course of HF in ACHD is different than acquired HF with a different age of onset, gradual onset of symptoms, and they never return to baseline—unlike ACHD HF, where the baseline is never normal, but they return to baseline amidst events. The staging of HF in ACHD is also not straightforward due to the vast heterogeneity of ACHD patients. Their NYHA class is unreliable and baseline VO2 abnormal; also, there are challenges in assessment of progressive ventricular dysfunction per echocardiogram. Furthermore, ACHD patients often under-report symptoms until they are quite advanced. Approaching HF management in ACHD is also different, as there is no guideline directed medical therapy in ACHD HF.

If despite optimization there is progressive decline, early referral for OHT/advanced therapies is recommended, ideally at a center with ACHD expertise. It is noteworthy that there are many challenges with regards to listing, waitlist time, transplant surgery and most patients do not get transplanted. We may need frank and honest goals of care discussions sand palliative care team involvement. Timely intervention impacts long term outcomes.

When to Consider Heart/Liver Transplant in Fontan Patients
Sharon Chen, MD, MPH
, Stanford University, Stanford, CA USA
All Fontan patients are seen regularly by a transplant hepatologist irrespective of their health status in a multidisciplinary clinic setting with serial assessments over time. One should have regular joint selection and monthly review meetings with the liver transplant team. Final intraoperative assessment of the liver at the time of transplant includes the decision to proceed made after gross inspection +/- biopsy.

Of 18 patients with Fontan listed for HLT, five received heart-only based on intra-operative assessment. All patients did well, though one is being closely monitored for progressive liver disease. All 13 patients who received HLT had pathology confirmed cirrhosis. As per a report from FOSTER study, HLT is better than HT for certain patients. It was a multicenter registry of adult Fontan patients referred for transplant. Over 130 patients, around 40 had HLT.

The survival benefit of HLT over HT, especially in those with more severe fatty liver disease. Liver surveillance is needed for all Fontans, not just at the time of transplant consideration, but should continue for those who receive heart-only transplants. Earlier transplant consideration is needed for many Fontan patients with > 10% Fontans seen for 1st heart failure consultation die or are declined for transplant within 30 days. FALD can progress during long wait times. There may be a potential role for VAD to bridge from heart-liver to heart only.

Durable Mechanical Circulatory Support for Adult Congenital Heart Disease
Fabrizio De Rita, FRCS
, Freeman Hospital, Newcastle-upon-Tyne, UK
In adults with CHD, HF remains the leading cause of morbidity and mortality, accounting for > 20% hospital admissions. Details of patient and device selection are critical to understand this very complex group. However, there is a very low level of evidence based on case series and case reports with regards to patient selection, who should perform implantation, and when, and on what type of device.

MCS in failing systemic RV- chronic dysfunction and dilatation leads to elevated end diastolic pressure, valvular dysfunction, and poor compliance of baffles. Three quarters of those catharized for HF have post capillary pulmonary hypertension. Adults with biventricular circulation with a systemic right ventricle develop premature HF. These young, high-risk patients have disproportionately poor access to advanced therapies. A subaortic VAD offers a high chance of bridge to heart transplant. There are certain intra-operative surgical considerations for these patients. The role of tricuspid valve replacement during implantation of a VAD for failing systemic RV to improve hemodynamics is the key question. It is associated with significant reduction in pulmonary wedge- and mean pulmonary artery pressure and subsequent improvement in subpulmonic left ventricular function with no obvious reverse remodeling of systemic RV. Therefore, it may not only expedite reduction in transpulmonary gradient but also prevent progressive subpulmonic LV failure.

MCS in failing Fontan physiology- Proportion of Fontan patients will develop Fontan failure, that is largely non-ventricular function related failure. Ultimate therapy is heart transplantation, but most of the patients are poor candidates for heart transplantation. Question remains if MCS can reverse end organ dysfunction or reduce morbidity/mortality on the waiting list for/after the transplant. Berlin Heart Fontan Cannula project focuses on RVAD as relief from venous congestion, to create effective access to the venous circulation, and reduce intra-operative risk. REGIVE study is actively enrolling patients.
– Summary by Anju Bhardwaj, MD
To Bx or not to Bx? That used to be the question.

But now a growing array of non-invasive options, which can be employed on various schedules and in various combinations, have spawned countless possible strategies for rejection surveillance after heart transplant (HT). Lest we become paralyzed by choice à la Hamlet, we might recall Falstaff’s credo that “the better part of valor is discretion.”

But what constitutes “discretion” in the context of 21st century post-HT surveillance? This was a prevailing question at the Saturday symposium Revisiting the Status Quo: Changing Paradigms in Routine Heart Transplant Rejection Surveillance. Angeline Leet, MD, of Alfred Hospital in Melbourne, Australia offered an astute answer: “The question we should be asking ourselves on a regular basis is ‘why are we performing this surveillance test?’... as each has associated costs, complications, and limitations in availability and expertise.” She argued that any test detecting early, subclinical rejection is “worth it” only if it prompts a significant change in prognosis or a modification in therapy. Speakers at today’s symposium shared this pragmatic view, focusing not just on how reliably a given test might predict the results of endomyocardial biopsy, but how it would change management. Such a focus demands a patient-specific approach, in lieu of “one size fits all” protocols.

Shelley Hall, MD, of Baylor University Medical Center in Dallas, TX USA, opened the symposium with an expert overview of the molecular modalities available, in her talk Beyond the Surface: Non-Invasive Diagnostics for Rejection Surveillance. Katharina Wassilew, MD, DScmed, MHBA, of Rigshospitalet in Copenhagen, Denmark offered further scrutiny in her talk, Do Non-Invasive Diagnostic Results Correlate with Histopathological Diagnoses Based on Endomyocardial Biopsy Interpretation?, while reminding us that even the “gold standard” (i.e., biopsy) has its limitations. Emanuele Cozzi, MD, PhD, of Padua University in Padova, Italy, echoed the importance of the “patient-specific” approach to diagnosing rejection—particularly in the interpretation of DSA measurements among highly-sensitized patients—in her talk Diagnosis and Clinical Implication of HLA and Non-HLA DSA in Daily Practice.

Dr. Leet offered other “candid” insights in her talk, Say Cheese: Use of Imaging in Rejection Surveillance. Surveying the relevant literature, she noted that global longitudinal strain as assessed by two-dimensional speckle tracking exhibits excellent sensitivity for diagnosis of early rejection; however, other echocardiographic markers (e.g., ejection fraction, circumferential strain, ventricular thickness) perform poorly. Cardiac MRI offers the theoretical advantage that it can characterize areas of myocardium not accessible via endomyocardial biopsy. This has been borne out by data showing that T1 relaxation time is predictive of biopsy-negative rejection and that T2 relaxation time predicts not only rejection but adverse outcomes (Miller et al 2019).

Finally, Eugene DePasquale, MD, of USC in Los Angeles, CA USA, offered a practical synthesis of these findings in Putting It All Together: Integrating Approaches for Rejection Assessment. One exciting area of integration is in the combined use of gene expression profiling and donor-derived cell free DNA—analyses by his group and others indicate that their combination performs better than either test alone. The molecular approach has replaced endomyocardial biopsy as the routine mode of rejection surveillance at USC. “To Bx or not to Bx” remains a question, but one asked less frequently—and only “for cause”—at many HT centers.
– Summary by Brian Wayda, MD

Cell-free donor-derived DNA (cfdNA) has become a powerful marker of myocardial damage with a high negative predictive value for the detection of rejection. This oral abstract session provided new insights into the use and implications of donor derived cell-free DNA (dd-cfdDNA) in heart transplant recipients.

Rationale and Design of a Randomized Controlled Trial of Donor-Derived Cell-Free DNA to Detect Rejection in Cardiac Transplantation (DETECT)
Josef Stehlik, MD, MPH, University of Utah School of Medicine, Salt Lake City, UT USA
The investigators presented the rationale and study design of a multi-center, prospective, randomized controlled DETECT trial (NCT05081739). The primary objective of this RCT will be to examine if rejection surveillance of heart transplant recipients with dd-cfDNA using the Prospera test is non-inferior to rejection surveillance with endomyocardial biopsy (EMB) and histology in the first post-transplant year.

Six hundred patients will be randomized 1:1 to the study group (dd-cfDNA) or the control group (EMB). The study duration is estimated to be approximately four years, with up to three years spent enrolling the participants followed by one year follow up for each participant. All patients aged 18 years or greater, who are listed for transplant and willing to provide written informed consent, will be enrolled. Key exclusion criteria will include multiple solid organ or tissue transplant; prior history of organ or cellular transplant; pregnancy; hemodynamic instability; and planned use of commercial or investigational cfDNA or GEP assays.

The primary endpoint will be a composite of treated rejection, graft dysfunction, re-transplantation, or death at 12 months after transplant. Secondary endpoints will include individual components of the primary endpoint; rejection with hemodynamic compromise; de novo donor specific antibodies; cardiac allograft vasculopathy; HR-QOL; and number of EMB. Core labs will be utilized: pathology, echocardiography, DSA, and protein biomarkers, transcriptomics.

Study results are anticipated to directly impact clinical practice in rejection surveillance, and get us closer to using dd-cfDNA in clinical decision making beyond rejection surveillance.

Should We Be Comforted by a “Negative” Endomyocardial Biopsy? Risk of Future Events with Donor Derived Cell Free DNA in the Setting of Histologic Quiescence
Jeffrey Teuteberg, MD, Stanford University, Stanford, CA USA
The investigators sought to examine the significance of elevated levels of donor-derived cell free DNA (dd-cfDNA) in the setting of a “negative” endomyocardial biopsy (EMB). The study reviewed patients from the multi-center Surveillance Using HeartCare Outcomes Registry (SHORE) database who had dd-cfDNA measured within 30 days of a negative EMB (Grades 2R/AMR1 or higher rejection) within the first 100 days of heart transplant (HT). These patients were followed for the subsequent development of significant rejection and de novo donor-specific antibodies (dnDSA) over the next 365 days.

The study included a total of 648 HT recipients with a mean age of 57, 74% male, 64% white, 60% of which had PRA < 1% and had a total of 982 paired biopsies with a median dd-cfDNA of 0.05% for those with a Grade 0R and 0.06% for Grade 1R biopsy. The dd-cfDNA was measured a median of 112 days post-transplant for Grade 0R and 109 days post-transplant for Grade 1R.

Despite negative histology on EMB, those with a cfDNA >= 0.20% were at significantly higher risk for the development of significant rejection (14.3% v 5.2%, p<0.01) and dnDSA (11.3% v. 6.8%, p<0.01) over the subsequent year. The authors concluded that the use of dd-cfDNA may be a better method to determine true quiescence and called into question the utility of the EMB as the gold standard for cardiac allograft monitoring.

Prognostic Implications and Characteristics of Low dd-cfDNA Results in Heart Transplant Patients with Biopsy Proven Rejection
Roopa Rao, MD, Indiana University, Indianapolis, IN USA
In contrast to the previous abstract, the investigators of this study examined the characteristics and prognostic implications of low level dd-cfDNA in patients with histological evidence of acute rejection. atients enrolled in the Surveillance HeartCare Outcomes Registry (SHORE) with biopsy proven acute cellular rejection (ACR) and antibody mediated rejection (AMR) who had low levels of dd-cfDNA (<0.15%) were evaluated. Patients with ACR 1R were excluded. Only those patients who had a dd-cfDNA level within 30 days of biopsy were included. Sixty-one patients (male 78%, median age 48 years) had low dd-cfDNA and endomyocardial biopsy evidence of ACR ≥2R and/or AMR ≥1R. Mean duration of the biopsy from the time of transplant was 196.31 days. ACR was seen in 36 patients (35=2R, 1=3R) and AMR was seen in 27 patients (21=pAMR 1(H+), 6=pAMR2). 2 patients had both ACR and AMR. Mean ejection fraction (EF) at the time of the rejection was mostly preserved. 2 patients died 330±31days after the initial diagnosis of rejection. None of the patients developed graft dysfunction at one year follow up. The authors concluded that patients with low dd-cfDNA and biopsy-confirmed rejection typically had lower grades of rejection and maintained preserved left ventricular function on echocardiography. As such, consideration of dd-cfDNA in treatment and immunosuppression management requires ongoing evaluation.

Absolute Quantification of Donor Derived Cell Free DNA in Heart Transplant Patients
Paul J. Kim, MD, UC San Diego Health, San Diego, CA USA
The authors investigated the performance of absolute quantification to detect acute rejection (AR) in heart transplant recipients. They measured both dd-cfDNA fraction (%) and absolute quantity (copies/mL) using a clinically available SNP-based massively multiplexed PCR dd-cfDNA assay.

Out of 447 samples collected from 150 heart-transplant only patients, 29 had AR and 418 did not. Both dd-cfDNA fraction and absolute quantity were significantly higher in samples with acute rejection (AR) compared to samples with no AR. Median dd-cfDNA fraction for AR was 0.62% (IQR: 0.09-1.43) and significantly higher than no AR at 0.04% (0.01-0.11, p < 0.01). Median absolute dd-cfDNA for AR was 78.8 copies/mL (5.3-125.6) and significantly higher than no AR at 2.7 copies/mL (1.37-6.90, p=1.67e-09).

The results suggest that the absolute quantification of dd-cfDNA may increase the accuracy to discriminate rejection when compared with dd-cfDNA expressed as a percentage of total cfDNA.

Impact on Donor Derived Cell Free DNA (dd-cfDNA) of Procurement Using Paragonix SherpaPak™ (SP) Cardiac Transport System versus ICE Transportation
Johanna van Zyl, PhD, Baylor Scott & White Research Institute, Dallas, TX USA
The authors aimed to determine if differences exist in dd-cfDNA values when using SherpaPak (SP) compared to historical ice transportation methods. They performed a retrospective review of 50 single organ heart transplant (HT) recipients who had heart care measurements within 90 days post-transplant. Twenty-nine recipient hearts were transported on ice and 21 with SP. Recipient and donor characteristics were similar with expected shorter mean and maximum total ischemic times on ice compared to SP. Median dd-cfDNA level was the highest within the first 27 days in both groups and were comparable up to 90 days. In SP patients, long ischemic times had higher dd-cfDNA compared to short ischemic times (p=0.01). This relationship was not observed in ice patients (p=0.41). The overall cohort with long ischemic times, regardless of transport method, had higher levels of dd-cfDNA within the first 90 days (p=0.012). The authors concluded that patients with long ischemic times are more likely to have higher dd-cfDNA vs short ischemic times. Further studies are needed to confirm these preliminary findings.

Relationship of Noninvasive Detection of Allograft Rejection and Injury (Donor-Derived Cell Free DNA and Gene Expression Profiling) and Tissue-Based Molecular Microscopic Diagnosis After Heart Transplantation
Dae Hyun Lee, MD, University of South Florida, Tampa, FL USA
The authors sought to examine the relationship between dd-cfDNA and gene expression profiling (GEP) and molecular microscopic transcript (MMDx) in heart transplant (HTx). A total of 108 patients were included in this retrospective analysis. Using MMDx, there were 9 patients with ACR, 39 with AMR, 2 with mixed rejection (AMR/ACR), and 58 with no rejection. Elevations in dd-cfDNA occurred in the ACR, AMR, mixed rejection groups, when compared to no rejection (Median: 0.24%, 1.20%, 2.90%, 0.12%, respectively, P<0.001). However, GEP was not different between different groups of MMDx. The AUC for dd-cfDNA and rejection by MMDx was 0.89 [95% C.I.: 0.83-0.96]. These findings suggest that non-invasive quantitative detection of dd-cfDNA correlates well with the tissue molecular microscopic transcript in both AMR and ACR in heart transplant patients. Future studies should examine how the addition of dd-cfDNA/GEP and MMDx would complement the diagnosis and prognosis of allograft injury in HTx.
– Summary by Rachna Kataria, MD
Time to trust our gut? (as a marker of post-transplant outcomes)

Saturday’s oral abstract session, More Than Meets the Gut: Nutrition and the Microbiome in Heart Transplantation, featured cutting-edge research exploring the potential role of the gut microbiome in predicting (and perhaps influencing) post-heart transplant (HT) management and outcomes.

It opened with two abstracts presented by Mark Dela Cruz, MD, of the University of Chicago, titled The Gut Microbiome as a Marker of Early Cardiac Allograft Injury and The Gut Microbiome and Tacrolimus Dosing in the Peri-Heart Transplant Period. Both analyses utilized a single-center, prospective cohort (n = 38 and 32, respectively) in which stool samples were collected within the first two weeks post-HT. Overall within-sample (i.e., “alpha”) microbial diversity was not associated with markers of injury (i.e., Allosure and Allomap scores) or tacrolimus metabolism; however, a look at specific bacterial taxa was more revealing. Specifically, greater abundances of the phyla Bacteroides and Firmicutes were predictive of lower Allosure and Allomap scores, respectively; both were associated with slower Tacrolimus metabolism. These signals were far from subtle; slow Tacrolimus metabolizers had double the abundance of Bacteroides (39% vs. 19%), and similar ~20% absolute differences were seen for the other observed associations.

The humble, ubiquitous Bacteroides retained the spotlight in the presentation that followed: Association of Gut Bacterial Beta Glucuronidase Activity with Mycophenolate Mofetil (MMF) Induced Cytopenia, by Lorenzo Braghieri, MD, of Columbia University Medical Center in New York, NY USA. Stool abundance of β-D-glucuronidases-producing (GUS) bacteria—and of the specific GUS species Bacteroides dorei and Bacteroides cellulosilyticus—were each associated with higher rates of leukopenia after HT in their cohort (n=45). That GUS bacteria are known to influence the metabolism of MMF—a leading cause of post-HT cytopenia—suggests that this association is causal.

For those of us still hesitant to “trust our gut,” further compelling evidence was offered by Joseph Spinner, MD, of Baylor College of Medicine in Houston, TX USA in his presentation Alterations to the Intestinal Microbiome Are Associated with Post-Heart Transplant Outcomes in Children. In their sample of 105 children undergoing HT, there were significant compositional differences in the gut microbiome both 1) pre- vs. post-HT, and 2) in those with vs. without adverse post-HT events.

All three presenters acknowledged the “chicken vs. egg” dilemma posed by their findings, as articulated by Dr. Spinner: “Whether these compositional alterations are a cause or effect, or if they can be reversed, remains to be elucidated.” Indeed, we can now only speculate as to the mechanism of these (fairly convincing) links between gut microbiota and clinical events after HT. Nonetheless, the authors’ painstaking efforts to construct these novel cohorts and conduct sophisticated analyses are commendable. And while a mechanistic understanding may be elusive, perhaps the use of gut microbial markers in prognostication and clinical management (e.g. in guiding MMF dosing) is not as far off?

Additional abstracts presented in today’s session explored the importance of metabolism in the HT context on a more macroscopic scale. In Metabolic Syndrome in Heart Transplantation: An Underestimated Risk Factor?, Veronica Ferrara, MS, of University of Udine in Udine, Italy presented a rigorous analysis that answered the titular question with a resounding “yes.” However, low BMI seems much less influential, as shown by Jignesh Patel, MD, PhD, of Cedars-Sinai Medical Center in Los Angeles, CA USA in his analysis Severely Underweight: Is It a Risk Factor for Poor Outcomes in Post-Heart Transplant Patients?
– Summary by Brian Wayda, MD