Cardiac and Pulmonary Vascular Risk Stratification in Liver Transplantation

Introduction

Approximately 8,000 liver transplants are performed in the United States every year for patients with fulminant liver failure or end-stage liver disease [1]. The survival rate is greater than 90% and 70% for liver transplant recipients (LTRs) at 1-year and 5-years post-transplantation respectively [2]. Advancements in organ preservation techniques, antiviral agents, immunosuppression, and screening protocols will likely further improve survival rates by decreasing rates of hepatic disease and malignancy, which previously were common causes of death in LTRs [3, 4]. However, despite the improvement in short- and long-term survival after LT, there has been a marked rise in chronic disease burden and morbidity in this population [5]. Complications due to cardiac and pulmonary vascular diseases are a leading cause of morbidity and mortality after LT. In fact, cardiovascular disease (CVD) complications, defined in the literature to include arrhythmias, cardiac arrest, heart failure, stroke, myocardial infarction or thromboembolism, are now the leading cause of early (<1 year) mortality and the third leading cause of late (>1 year) mortality in liver transplant recipients (LTRs) [6, 7, 8]. CVD mortality after LT has increased by 50% since 2002 [9], and LTRs have a 64% increased risk of having a cardiovascular event (CVE) over 10 years compared to the general population [10]. Impressively, one in three LTRs will experience a CVE within 1 year of transplantation [7,8].

Reasons for the observed increased CVD risk in LT candidates are multifactorial. First, traditional cardiac risk factors such as hypertension, diabetes and hyperlipidemia [7,11,12] are all increasing in prevalence among LT candidates and are highly associated with post-LT CVEs both short- and long-term after transplant [8]. For example, one study showed that post-operative ischemic complications are 40% higher in patients with pre-LT insulin-dependent diabetes or known CAD than those without [12]. Second, liver-specific risk factors have also been associated with high rates of post-LT CVEs. NASH and alcohol-induced liver disease have both been shown to have an increased association with early major CVEs compared to other etiologies of ESLD [8]. This is particularly relevant as NASH cirrhosis is now the second most common and quickest-growing indication for liver transplantation, [13] and is projected to become the number one indication for LT within the next decade [13, 14]. Patients with NASH cirrhosis in particular have been shown to have a 4 times higher risk of CVE after LT as compared to patients with other causes of cirrhosis, such as cholestatic liver disease [7]. Third, the use of the Model for End-Stage Liver Disease (MELD) score to prioritize donor livers to the sickest patients generates a candidate pool with a high burden of critical illnesses and, as a result, a high prevalence of subclinical and clinical CVD [16]. Fourth, an aging candidate pool has resulted in a high burden of comorbid medical conditions, including CVD at the time of LT [8]. Fifth, the liver transplant operation itself is associated with significant stress on the cardiac and pulmonary vascular system which can unmask subclinical CVD [17]. Finally, post-transplant chronic immunosuppression can accelerate or contribute to de novo CVD [18]. All of these things in concert contribute to the high burden of CVD in LT candidates and recipients, and the incidence of CVD and its associated complications are expected to continue to rise over the next decade.

The purpose of this review is to discuss the approach to cardiac and pulmonary vascular disease risk stratification in LT candidates, with a specific focus on perioperative and short-term (<1 year) CVE risk [7,8]. We will address the epidemiology and available data supporting risk assessment for underlying arrhythmia, heart failure, stroke, coronary heart disease, valvular heart disease and portopulmonary hypertension in LT candidates. This article will also explore the prognostic tools available to risk stratify candidates based on their global risk for CVEs, which can be used to improve the candidate selection process and potentially improve allocation of scarce donor organs. Finally, while not the primary focus of this article, the emerging data on long-term CVD risk factors and complications will also be briefly discussed.

Arrhythmias

Arrhythmias are the most common CVD complication among LTRs within 30 days of LT [19]. Arrhythmias present prior to LT are indicative of underlying cardiac pathology that increases the risk of CVD complications following LT [20, 21]. The primary screening modality for underlying arrhythmia is with an electrocardiogram (ECG). Current AASLD/AST guidelines recommend that all LT candidates be screened with an ECG to assess for prevalent arrhythmia, and if identified, investigation into underlying causes and consultation with a cardiologist is recommended [22].

Atrial fibrillation, a common marker of underlying hypertension and cirrhotic cardiomyopathy (CCM) [17], is the most common pre-existing arrhythmia, and is present in 1–6% of LT candidates [8, 23] in contrast to 0.95–2.3% among the general population [8]. While conflicting data exists about the association between preexisting atrial fibrillation and liver graft dysfunction and mortality [8, 23], several studies have shown that pre-existing atrial fibrillation confers a higher risk of CV complications both intraoperatively and postoperatively [8, 23, 20]. For example, VanWagner et al. demonstrated that pre-existing atrial fibrillation was present in 33% of patients with major CVE within 90 days of LT compared to 7% of patients without a CVE [8].

Intraoperative electrolyte imbalances are common during LT and can instigate cardiac dysfunction and arrhythmias. For example, the donor graft is commonly preserved in a high potassium solution, which contributes to arrhythmias and cardiac arrest, during reperfusion. In a retrospective study of 146 patients who underwent LT, Dec et al found that either ventricular tachycardia or fibrillation was the most common intraoperative cardiac complication (3.4%), but the fourth most common post-operative complication (2.7%) [4]. Donovan et al found that among 71 patients who underwent LT, 10 patients (14%) developed perioperative arrhythmias, with five patients developing atrial fibrillation and two developing ventricular arrhythmias [24]. Postoperatively, atrial fibrillation occurs in up to 43% of LTRs and is the most common contributing cause to rehospitalization within 90 days of LT [8]. In the same study, cardiac arrest occurred in 6% of recipients and was the fifth most common complication in this early 90-day period [8].

One potential risk factor for development of arrhythmia post-LT is QTc prolongation, which is present in 30–60% of ESLD patients [27]. QTc prolongation is a disorder of myocardial depolarization and occurs when a QT interval is greater than 0.45s in males and 0.47s in females [19, 17]. When QTc prolongation is present, the incidence of sudden cardiac death in ESLD patients is the same as that in patients without ESLD [25]. However, QTc prolongation may predispose LT candidates to ventricular arrhythmias [25]. Some studies have shown that QTc prolongation is not associated with increased mortality and it is not a contraindication for LT [12], while others conclude that the risk of post-LT CVEs and mortality is increased [19, 25, 26]. Current clinical practice guidance from the American Society of Transplantation (AST) recommends that patients with a prolonged QTc should be evaluated for reversible causes such as electrolyte derangements and medications [22]. Additionally, in patients with known dysrhythmias, perioperative hemodynamic monitoring is recommended with anticipated treatment readily available including anti-arrhythmic medications, transcutaneous or transvenous pacing, electrical cardioversion or defibrillation [22]. Family history should be obtained and the patient should be referred to a cardiologist if there is a positive family history of sudden cardiac death [28]. In the majority of ESLD patients, QTc prolongation will resolve on its own following LT [28].

Heart Failure and Cardiomyopathy

Studies have demonstrated that heart failure is the second most common early CVD complication post-LT, occurring in 28% of patients within 90-days of LT, and contributing to a quarter of all hospital admissions within the same timeframe [8]. Both heart failure and arrhythmias make up 50% of CVD complications within one year of LT [21], and postoperative HF is associated with mortality rates as high as 15% [29]. In the general population, the definition of heart failure with preserved ejection fraction (HFpEF) requires an ejection fraction (EF) >50%, while that of heart failure with reduced ejection fraction (HFrEF) requires an EF </=40% [30]. Notably, based on published clinical practice guidance from the AST, the absolute and relative contraindications to LT are EFs less than 40% and less than 50%, respectively [22]. However, there is little data to support these EF thresholds beyond clinical experience and center-specific risk tolerance. Pre-transplant LV dysfunction has been shown to predispose LTRs to increased risk of CV complications [6, 8]. Diastolic dysfunction prior to LT, specifically, has been associated with an increased risk of graft rejection and failure, post-LT systolic dysfunction, and mortality [31].

Post-operative HFpEF is more common than HFrEF in LTRs, and ranges from 3–43% depending on which echocardiographic consensus guidelines are used for the definition of diastolic dysfunction [9, 23, 32], with the 2016 American Society of Echocardiography (ASE) and European Association of Cardiac Imaging (EACI) guidelines being the most up to date [26, 33]. The prevalence of HFrEF post-LT is much lower, at just 2–7% (9, 23, 33]. This is likely due to the exclusion of ESLD patients with prevalent HFrEF from LT as discussed above [26].

LT candidates are screened for HF pre-operatively using echocardiography which can detect both HFpEF, which corresponds to diastolic dysfunction, and HFrEF, which corresponds to systolic dysfunction [34, 35]. The pre-operative risk assessment also includes an assessment for left ventricular outflow tract obstruction (LVOTO), which is inducible on dobutamine stress echocardiography in >40% of patients [36]. LVOTO can occur as a result of concurrent left ventricular hypertrophy (LVH) and hyperdynamic systolic function resulting in a significant outflow gradient. One study showed that a gradient of >36mmHg is significantly associated with intraoperative hypotension in transplant candidates and is a contraindication to transplantation in some centers [36]. Cardiac magnetic resonance (CMR) imaging can also be used to screen for or diagnose structural or functional cardiac abnormalities in patients with ESLD. CMR provides markers of abnormal cardiac structure and function that cannot be assessed with echocardiography. For example, CMR T1 and T2 tissue mapping can detect subendocardial edema and myocardial fibrosis, which are markers of myocardial remodeling described in patients with ESLD [37, 38, 39]. Functional testing, including the 6-minute walk test (6MWT) and cardiopulmonary exercise testing (CPET), may also be used pre-operatively to assess cardiopulmonary reserve in LT candidates [40]. In the 6MWT, the distance a LT candidate is able to walk in the allotted 6-minute period correlates with overall cardiopulmonary fitness. One study found that a distance of <250 meters was associated with a higher waitlist mortality, however, the correlation with posttransplant outcomes was not investigated [41]. CPET measures a participant’s anerobic threshold during a progressive exercise regimen [40]. An aerobic threshold >9mL/kg/min is an accurate predictor of early (90-day) post-transplant survival [42]. Of note, functional testing in LT candidates is often limited by the high prevalence of malnutrition, sarcopenia and severe deconditioning which may prohibit patient participation in this type of testing [40].

If guideline-directed medical therapy (GDMT) for HF is initiated prior to LT, it should be continued post-operatively. However, angiotensin converting enzyme inhibitors (ACEi) may need to be dose reduced for kidney dysfunction or electrolyte abnormalities that may be caused by calcineurin inhibitors or aldosterone antagonists [12]. In addition, while carvedilol is often the optimal beta-blocker in patients with ESLD [43], it should be used with caution post-LT due to its ability to reduce portal pressures which can lead to graft ischemia [43].

An important contributor to HF prevalence in LT candidates is the presence of cirrhotic cardiomyopathy (CCM). CCM is the most common risk factor for post-LT HF and is found in 40–50% of ESLD patients [44, 45]. CCM has long been characterized as hyperdynamic cardiac function due to low systemic vascular resistance and high cardiac output in the absence of known causes of cardiac disease (e.g., prior MI) [39]. The diagnostic criteria for CCM based on echocardiogram has recently been updated to require:

1. LVEF </= 50% or

2. Absolute global longitudinal strain of <18%

AND 3 or more of the following:

1. Septal e’ velocity <7cm/s

2. E/e’ ratio >15

3. Left atrial volume index (LAVI) >34 mL/m²

4. Tricuspid regurgitant (TR) velocity >2.8 m/second [39]

CCM is thought to be caused by myocardial hypertrophy, myocardial fibrosis, and subendocardial edema [17, 46]. The risk of developing CCM is in part dependent on the etiology of the ESLD, and is highest with cirrhosis caused by NASH, alcohol, HCV and hemochromatosis [22]. CCM has been shown to raise the risk of intraoperative complications and of developing HF post-operatively [47]. Given the demonstration of poor CV outcomes in patients with HF prior to LT, and the fact that the majority of ESLD patients have some component of CCM [44, 45], appropriate screening of LT candidates and close monitoring with repeated echocardiography post-LT are important. Based on recommendation from the Cirrhotic Cardiomyopathy Consortium, the recommended screening interval for CCM using echocardiogram in LT candidates is at minimum every 6 months prior to transplant and in intervals of 6, 12, and 24 months post-transplant in all patients with any degree of pre-transplant systolic or diastolic dysfunction based on the aforementioned criteria [39].

Stroke

Stroke is a CVD complication that is not well studied in liver transplantation. Little data exists on the prevalence of ischemic and hemorrhagic stroke in candidates prior to LT, however, hemorrhagic stroke may be more prevalent in patients with alcohol-related cirrhosis compared to other etiologies [48]. In a multivariable analysis of 32,810 LTRs, VanWagner et al reported that pretransplant stroke increased the likelihood of 30-day and 90-day major CVEs with an incidence risk ratio of 6.3 and 4.8, respectively [8]. The same study found that stroke occurred in 9% of those hospitalized within 90 days of transplant.

The risk of post-LT stroke is highest in recipients of older age and those with pretransplant diabetes, hypertension, hyperlipidemia [49]. Hypertension is a potentially modifiable risk factor for stroke and prevalent in up to 53.7% of LT candidates [50]. Hypertension prevalence increases after LT with up to 92% of LTRs having hypertension by 6 years post-transplant [50]. This is a result of several factors including direct calcineurin inhibition, nephrotoxicity from calcineurin inhibitors (particularly with the combination of cyclosporine and mTOR inhibitors), steroid use, pre-existing hypertension, older age, and both renal and systemic vasoconstriction [7, 50]. Adherence to blood pressure management guidelines has been shown to be inversely related to CVD complication rates and CVD mortality, however, those at highest risk for poor CV outcomes are less likely to have appropriate blood pressure management [50]. In addition to improved blood pressure control before and after LT, smoking cessation, statin use, and anticoagulation, if indicated, can also reduce the risk of stroke [8]. There is currently no guideline recommendation to perform routine imaging screening in LT waitlist candidates for stroke. However, in LT candidates and recipients who present with cognitive, sensory or motor dysfunction a high index of suspicion for possible stroke is required and diagnostic imaging should be pursued.

Coronary Artery Disease

Surgical and pharmacological advancements in the management of coronary artery disease (CAD) have significantly lowered the burden of CV morbidity and mortality in the general public [51]. The prevalence of CAD among ESLD patients is the same or greater than for the general population [52], ranging from 2.5–27% [50, 51]. LT candidates with known CAD are at greater risk for intraoperative complications during LT surgery [5]. Roughly 25% of LT candidates with traditional CAD risk factors (such as hypertension, hyperlipidemia, and diabetes) have moderate CAD stenosis [52], defined as >50% stenosis, and those with 2 or more traditional CAD risk factors are more likely to have obstructive CAD [53, 54]. Those with severe stenosis had a higher risk of mortality despite coronary revascularization [56]. Even in the absence of severe stenosis, multivessel CAD is associated with increased mortality and postoperative hemodynamic instability [56]. LT candidates with NASH cirrhosis are more likely to have traditional risk factors for CAD [57], and both NASH and renal dysfunction predispose to critical coronary stenosis [57, 58]. Conflicting data exists about whether NASH cirrhosis is an independent CAD risk factor over traditional CAD risk factors [58, 59, 60].

The utility of stress testing as a means of CAD risk stratification in LT waitlist candidates is debated [61]. Exercise stress testing has poor predictive value in the ESLD population as patients are often too debilitated to reach their target heart rate [24], and therefore, pharmacologic stress testing using dobutamine, dipyridamole, or adenosine is often used. Plevak et al suggested that the stress imaging modality of choice be dobutamine stress echocardiography (DSE) [62]. However, studies have shown that subsequent cardiac catherization does not correlate with imaged wall motion abnormalities [63]. Williams et al showed a lack of correlation between positive DSE and intraoperative CVEs [65], and another study demonstrated that stress echocardiography failed to identify LT candidates at high cardiac risk for LT surgery. [56, 61]. Thus, DSE has very poor sensitivity, as low as 13%, and low negative predictive value (75%) for the detection of obstructive CAD [24, 64, 65, 66]. Furthermore, other studies have demonstrated pre-LT DSE to have a positive predictive value of only 27% and a specificity of 87% for determining adverse CVEs within 30-days post-LT [56]. Therefore, DSE is not useful for risk stratification for obstructive CAD or CVEs in patients with ESLD [63]. The chronic vasodilatory state of ESLD patients makes pharmacologic vasodilator testing unreliable as well [66], with one study showing that even high-risk patients were identified as low-risk in a myocardial perfusion study (MPS) [61]. A meta-analysis explored the utility of DSE and MPS in CAD detection compared to the gold standard of coronary catheterization among LT candidates [67]. The pooled sensitivity was 28% and 61% and specificity was 82% and 74% for DSE and MPS respectively.

Given the poor performance of noninvasive stress testing in the detection of CAD among ESLD patients [12], coronary angiography, either invasive or noninvasive is the gold standard for detecting CAD in this population [22]. Several studies have proposed that patients above the age of 45 and with 2 or more traditional CAD risk factors undergo invasive coronary catheterization (Figure 1) [12, 68]. However, CT angiography is emerging as a noninvasive alternative to invasive coronary catheterization [12]. Anatomical coronary assessment with coronary CTA is not limited by the hemodynamic abnormalities of ESLD, and CT provides the additional advantage of simultaneously obtaining a coronary artery calcium score (CACS), which in itself adds prognostic information [69, 70]. CACS > 400 is associated with increased risk for underlying obstructive CAD in LT [71]. Furthermore, CAC scoring has been shown to be predictive of CAD requiring revascularization and 1-month post-LT complications [71] Finally, coronary CTA recently demonstrated superior sensitivity for evaluation of obstructive CAD compared to other noninvasive modalities before kidney transplantation [72].

Figure 1.

Proposed Screening Algorithm for Coronary Artery Disease Risk

* Tissue doppler echocardiography with myocardial deformation imaging according to American Society for Echocardiography guidelines

**CAD risk factors include hypercholesterolemia, hypertension, diabetes, current/prior tobacco use, prior CVD, or left ventricular hypertrophy. Abbreviations: ECG, electrocardiogram; TTE, transthoracic echocardiogram; CAD, coronary artery disease; RFs, risk factors

When examining how best to mitigate the risk of atherosclerotic CVD (ASCVD) complications post-LT, it is important to recognize three facts. Firstly, not all candidates undergo coronary angiography and therefore some may have undetected subclinical CAD going into LT surgery. Secondly, LTRs are at risk of accelerated atherosclerosis following transplant as demonstrated in several studies [73, 74, 75], and are therefore less likely to form collaterals leading to an increased risk of CAD mortality long-term. And finally, LTRs are at an increased risk of developing dyslipidemia post-LT as a result of immunosuppressive agents (Figure 2). Therefore, more studies are needed to determine the degree to which LTRs’ risk of CAD-related mortality is a result of undiagnosed subclinical CAD versus post-LT risk factors [76].

Figure 2.

Effects of Immunosuppressive Classes on Components of the Metabolic Profile

Abbreviations: CNI, Calcineurin Inhibitors (e.g., tacrolimus, cyclosporine); MTOR Inhibitors, Mammalian Target of Rapamycin (e.g., everolimus)

As a result of improved ASCVD screening, interventions, and medical management for LT candidates prior to LT, CAD is increasingly becoming a later-term (>1 year) CVD complication [77], constituting up to 39.8% of CV complications among patients over roughly a 10-year period post-LT [26]. Management of traditional risk factors (e.g., hypertension, diabetes mellitus and hyperlipidemia), exacerbated by immunosuppressive agents, is necessary to secure the long-term health of LTRs. In addition to traditional risk factors, post-transplant metabolic syndrome and renal dysfunction, both influenced by chronic immunosuppressive agents, also need to be continually addressed in the post-transplant period in order to reduce long-term ASCVD complications.

Valvular Heart Disease

Aortic stenosis (AS) is the most common stenotic valvular lesion in adults though specific prevalence in LT candidates is unknown [40]. AS can result in systemic hypoperfusion and the associated left ventricular hypertrophy can be arrhythmogenic [78]. Mitral regurgitation (MR), tricuspid regurgitation (TR) or both have been found in 28% of all LTRs [79]. The effect of these valvular disorders on post-LT CV outcomes remains unclear, however they appear to be of greatest significance intra- and perioperatively. For example, TR can result in venous congestion and lead to right ventricular heart failure, thereby causing hypoperfusion to the new graft [40]. For this reason, LT candidates should be screened preoperatively for valvular dysfunction, and pretransplant repair in appropriate candidates is recommended in severe cases [40].

The treatment options for AS, in particular, have been well studied. These options include surgical aortic valve repair (SAVR) and transcatheter aortic valve replacement (TAVR). A retrospective cohort study performed by Peeraphatdit et al showed comparable in-hospital (1.8% vs 2%) and 30-day mortality (3.2% vs 4.6%) among 105 cirrhotic patients undergoing TAVR and SAVR respectively [80]. However, when stratified by MELD scores, the SAVR group showed improved median survival among patients with a MELD <12 compared to the TAVR group (4.4 years vs 2.8 years). There was no difference in survival between TAVR and SAVR in patients with MELD >/=12 [80].

Severe valvular disease in LT candidates necessitates close hemodynamic monitoring, and repair is recommended in severe cases prior to transplantation. MELD scores, shown to be an independent predictor of survival in the repair of aortic valves [80], may be of utility in determining optimal management of tricuspid and mitral valve disorders as well.

Portopulmonary Hypertension

Portopulmonary hypertension (PoPH) is a disorder that occurs in 5–10% of LT candidates, where pulmonary vasculature constriction leads to remodeling [81]. By definition, patients with PoPH also have pulmonary hypertension which can lead to right-sided HF resulting in adverse CV outcomes [22].

Findings of elevated pulmonary artery systolic pressure (PASP) on echocardiography may be indicative of PoPH. While no consensus has been reached on a cutoff value for PASP that should trigger further invasive testing, a threshold of PASP >45mmHg has been suggested [82]. If PASP is >45mmHg, right heart catheterization (RHC) findings diagnostic for PPH are:

1. Elevated mean pulmonary arterial pressure (mPAP) >/=25mmHg,

2. Pulmonary vascular resistance (PVR) of > 240 dynes·s per cm−5, and

3. Low pulmonary capillary wedge pressure (PCWP) <=/=15mmHg [12, 22].

mPAP of 25–34mmHg is consistent with mild PoPH, while mPAP >/=35mmHg is consistent with moderate to severe PoPH. One study showed that LT candidates with PoPH and mPAP >/=35mmHg had a 50% mortality rate, which increased to 100% for those with mPAP of 50mmHg or greater [83]. If left untreated, the median survival for all patients with PoPH approaches 15 months, with a 5-year survival of just 14% [84]. In contrast, for those treated with pulmonary vasodilators, median survival increases to 46 months and 5-year mortality increases to 45% [84] LT is possible in those with mild PoPH, however, those with moderate to severe PoPH must undergo a trial of vasodilator therapy with prostaglandins, phosphodiesterase inhibitors or endothelin receptor antagonists in an attempt to lower pulmonary pressures prior to LT [85]. If mPAP is successfully lowered to <35mmHg and PVR <400 dynes.sec.cm⁻⁵, MELD exception points can be granted to the LT candidates [86]. A study was done on outcomes after LT among 11 patients with moderate-severe PoPH, in whom adequate mPAP reduction was achieved with vasodilator therapy prior to LT. Interestingly, the one- and five-year survival following LT was 91% and 67% respectively, and nine of the eleven patients were off vasodilator therapy at 9.2 months post-LT [87]. In another smaller study, four patients with PoPH who responded to vasodilator therapy with epoprostenol were successfully transplanted. Of the four, two were weaned off epoprostenol within 8 months post-LT, while two remained on oral vasodilators at the time of publication, corresponding to 9–18 months post-transplant [88]. These studies demonstrate that pharmacologic therapy can permit safe LT in patients with PoPH. However, intraoperative challenges exist with PoPH, and cardiac output, which is a surrogate marker of RV function, is a predictor of operative survival [84, 89]. Therefore, inotropes and prostacyclins may need to be administered during the procedure, and consultation with PH specialists is recommended [22]. Importantly, there are no data to support the concept that PoPH (treated or untreated) should be an indication for LT [90].

Risk Scores for Prediction of Global CV Risk in LTRs

Risk prediction for cardiac complications can be divided between general risk assessment tools and surgical cardiac risk calculators (Figure 3).

Figure 3.

Available Cardiac Risk Assessment Tools.

*studied in the LT population

Abbreviations: LT-specific, liver transplantation-specific

General Risk Assessment Tools

The Framingham Risk Score (FRS) predicts the 10-year risk of fatal or nonfatal coronary events using a multivariable risk prediction algorithm [91]. This score has been evaluated in LT candidates and studies show an inverse relationship between candidates with high FRS (FRS>20) and median survival at 1, 3, and 5 years post-LT [92]. The C-statistic for FRS ranges from 0.76–0.79, and displays moderate discrimination for CVEs and mortality among LT candidates [92].

The Reynold Risk Score (RRS) builds on the traditional CV risk factors used in the FRS by also adding high sensitivity C-reactive protein and parental family history of premature CHD [93]. While this risk score has been shown to be more predictive of CVD than FRS in the general population [93], it has not been applied to LT candidates and so its utility in this population has yet to be verified.

The American College of Cardiology (ACC) and American Heart Association (AHA) developed the pooled cohort equations (PCE) risk calculator to estimate the 10-year and lifetime risk of atherosclerotic cardiovascular disease (ASCVD) events [94]. The purpose of this calculator is to guide therapeutic interventions (e.g., statins, antihypertensive medications) to lower the risk of CVEs. However, patients with cirrhosis and LTRs were not included in the studies used to derive these equations and thus, the predictive ability of the PCE in our patient population is unknown. Other risk assessment tools include the European Systemic Coronary Risk Evaluation Project (SCORE) and the German Prospective Cardiovascular Munster Study (PROCAM) which have limited predictive power in waitlist candidates [95].

The utility of standard risk algorithms as a predictive tool is limited in the LT population, particularly those with subclinical disease [96]. The majority of CVD complications post-LT are non-coronary in nature [6], however, these risk scores are intended to identify the risk of events related to CAD. These tools are also yet to be calibrated to LT candidates [12, 82] and may be inaccurate in patients with subclinical CVD. Troponin levels, a marker of cardiac myocyte injury and death, have been shown to correlate with risk of post-LT CV events in subclinical or “asymptomatic” patients regardless of concomitant comorbidities. One study showed that a pre-transplant troponin I elevation (>0.07ng/ml) was associated with de novo CVD in post-transplant patients [77], while another showed that troponin elevation (>0.1ng/ml) was associated with higher 30-day post-LT mortality [97]. The I isoform has been shown to be more reliable in patients with renal dysfunction than the T isotope and may be a useful risk stratification tool in ESLD patients [97].

Surgical Cardiac Risk Calculators

Perioperative cardiac events are a leading cause of death following noncardiac surgery [98]. As such, risk stratification prior to surgery allows for cardiac optimization and is an important part of the informed consent process [99].

The Revised Cardiac Risk Index is a cardiac risk tool that uses 6 variables (history of ischemic heart disease, HF, stroke, insulin-dependent diabetes, CKD, high-risk surgery) to risk stratify patients into either low, medium or high-risk categories for perioperative cardiac complications in noncardiac surgeries. This risk calculator was incorporated into the ACC/AHA and ESC/ECA guidelines for perioperative cardiac risk assessment and management [100]. Park et al investigated the combined predictive ability of the RCRI and MELD and found that patients with both higher MELD and RCRI scores had higher rate of cumulative cardiac events [101]. Predictive ability was increased in models that used both RCRI and MELD compared to models using MELD alone, suggesting significant potential utility if adapted to the ESLD population [101].

Despite its popularity, the RCRI was validated on just 4315 patients [96]. In contrast, the MICA (myocardial infarction and cardiac arrest) risk calculator developed by Gupta et al was validated on a cohort of over 400,000 patients [100]. MICA is an interactive risk calculator and uses 5 predictors (type of surgery, dependent functional status, abnormal creatinine, American Society of Anesthesiologists’ class, and increasing age) to provide a probability estimate of MICA in a patient [100]. With a C statistic of 0.88, MICA has superior discriminative and predictive abilities compared to the RCRI although neither have been studied in the LT population [100].

The American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) developed a nationally validated, risk-adjusted, outcomes-based program that aims to improve the quality of surgical care. The ACS NSQIP developed a risk calculator to predict 30-day post-operative risks using 21 variables, including American Society of Anesthesiology class and functional status [102]. The C statistics for this calculator range from 0.806 to 0.99 and was 0.895 for cardiac complications. Despite its good discriminative ability, it has questionable validity among LT candidates given the likely small proportion of LT candidates among the national sample [102].

The Cardiovascular Risk in Orthotopic Liver Transplantation (CAR-OLT) score is a CVD risk prediction model that was derived in LT candidates and uses pre-transplant demographic, social and clinical variables to estimate the risk of a major CVEs in the first year post-transplantation [9]. The model utilizes a point-based system for twelve weighted covariates, namely preoperative recipient age, sex, race, employment status, education status, history of hepatocellular carcinoma, diabetes, heart failure, atrial fibrillation, pulmonary or systemic hypertension, and respiratory failure (available at www.carolt.us) [9]. The model defines major CVEs as myocardial infarction, cardiac revascularization, heart failure, atrial fibrillation, cardiac arrest, pulmonary embolism or stroke within 1 year after LT [9]. This score has a good discriminative ability with a C statistic of 0.78 and is unique in that it estimates risk of CVD complications specifically after LT [103]. Therefore, it is a potential prognostic tool in assessing a patient’s risk of a CVE after LT and informing risk-based discussions. However, it has not been prospectively validated in LT and whether closer monitoring or risk factor optimization in patients with high CAR-OLT risk alters post-LT outcomes is unknown.

Risk Factors for Long-Term (>1 year) CVD Complications in LTRs

Traditional cardiac risk factors, namely hypertension, diabetes, and hyperlipidemia, play an important role in the development of long-term CVD complications [8], which are the third-leading cause of long-term mortality in LTRs, as previously stated. Post-transplant diabetes mellitus has been shown to confer the greatest long-term risk for major CVEs in LTRs, with cumulative incidences of 13% and 27% at 5 and 10 years respectively [104]. Additionally, studies have shown that hypertension can develop in 65–80% of previously normotensive LTRs [105]. Hypertension is often missed in LTRs when assessed solely by in-office blood pressure checks alone which detected 2.5 times fewer cases when compared to 24-hour ambulatory blood pressure monitoring, in part due to loss of normal nocturnal decrease in blood pressure following organ transplantation [105]. This failure to capture hypertensive patients prevents providers from managing CV risk factors appropriately among LTRs. Additionally, in a single-center retrospective cohort of 495 LTRs, Patel et al demonstrated the accelerated development and progression of traditional CVD risk factors post-OLT. In particular, they found that the prevalence of dyslipidemia prior to LT was 20% and increased to 55% at 5-years post-LT [106].

Hypertension, diabetes and hypertriglyceridemia (>150mg/dL) are three of five components of posttransplant metabolic syndrome (PTMS), a significant risk factor for long-term CVD complications. The remaining two components of PTMS are HDL level (<40mg/dL) and BMI >30kg/m2, however, only three are required to make the diagnosis [7]. The prevalence of PTMS following OLT is 2–3 times higher than the rates of metabolic syndrome in the general population, and ranges from 40–58% [107]. Among LTRs, the risk of having a CVE is 3–4 times higher in those with PTMS compared to those without, although there is no difference in all-cause mortality [10]. Immunosuppressive agents also contribute to PTMS given their tendency to cause metabolic derangements [7, 10]. By identifying and managing traditional risk factors early in the posttransplant course, the prevalence of PTMS, and therefore the prevalence of CVD events, can be significantly reduced [10]. Overall, there is a small amount of data on the risk factors for long-term CVD complications compared to perioperative and early CVEs in LTRs. However, application of evidence-based guideline-directed medical therapy has been demonstrated to reduce CVEs and long-term mortality in LTRs. For example, among 602 LTRs at a large tertiary care network, achieving a blood pressure <140/90 mmHg within the first year following liver transplant was associated with a 35% lower hazard of CVEs and a 42% lower hazard of mortality compared to LTRs with BP >= 140/90 mmHg [50]. In another retrospective study, Patel et al. demonstrated that statin use for secondary prevention of ASCVD was both safe and associated with a 75% lower hazard of mortality compared to non-statin use in LTRs [76]. Although more research studies are needed to better characterize long-term CVD morbidity and mortality in LTRs these studies highlight an important practice gap between actual and optimal medical management for CVD in LTRs [73]. Future prospective studies that utilize mixed-methods approaches in which the underlying reasons for a lack of guideline adherence are elucidated and that focus on primary prevention of ASCVD are needed in order to optimize care for LTRs.

Conclusion

Due to the scarcity of donor organs, LT recipients should be appropriately selected and their CV risk factors appropriately monitored and managed as CVD represents a major contributor to short and long-term adverse outcomes in this population (Table 1). Pre-operative cardiac assessments are intended to 1) identify and manage CV risk factors pre-operatively, or 2) to exclude highest risk candidates from transplantation by screening for absolute contraindications, namely significant obstructive coronary artery disease, severe heart failure or severe portopulmonary hypertension. As outlined above, the majority of CVD complications in the first year post-LT are non-coronary in origin and mostly consists of atrial fibrillation and heart failure. Prospective studies are required to determine whether aggressive risk factor management both prior to and following LT is beneficial in reducing the rates of post-LT CVD complications.

Table 1.

Cardiovascular Outcomes, Modalities for Screening and their Limitations, and Areas for Future Research

	CV Outcome	Screening Modalities	Thresholds for Increased CV Risk	Limitations	Research Gaps
	Arrhythmias	12-Lead ECG	Any Abnormal Heart Rhythm QTc Prolongation >0.45s in males. >0 47s m females	Abnormalities may not be captured during single ECG	Rate vs. rhythm control and anticoagulation management in atrial fibrillation and associations with clinical outcomes
	Heart Failure and Cardiomyopathy	Functional Testing (6MWT, CPET) 12-Lead ECG 20 Echocardiography DSE Cardiac MR (CMR)	6MWD <250m Anaerobic threshold <9 mi/min/kg Decreased exerase tolerance Systohc dysfunction Global longitudinal strain < −18% Dtastobc dysfunction inducible LVOTO Chamber enlargement Myocardial edema or fibrosis	Functional testing may be limited to ambulatory patients Renal dysfunction may precude gadolinium administration m CMR	Clinical significance of impaired excitation contraction coupling on ECG Role of CMR for screening
	Stroke	-	-	No modality recommended for stroke screening - imaging with CT/MRI recommended only for diagnosis	Prevalence of ischemic and hemorrhagic stroke in LT candidates and recipients
	Coronary Artery Otsease	Non-lnvasive Stress Testing (DSE. MPS) Non-Invasive Coronary Evaluation (CCTA, CACS) Invasive Coronary Evaluation (Coronary Catheterization)	Wan motion abnormalities Perfusion defects CACS >400 HU	Patients may not achieve maximum chronotropy with pharmacologic stress testing MPS of limited utility given chronic vasodilatory state in ESLD CCTA. CACS limited in obese patients and those unable to slay still or following required breathing maneuveurs Risk of contrast-induced nephropathy with CCTA CACS and coronary catheterization	The impact of subchnical CVD vs accelerated ASCVD development on post-LT CVD complications
	Valvular Hean Disease	2D Echocardiography	Valvular stenosis or insufficiency	Visualization may be limited with 2D echocardiography, and may require transesophageal echocardtogarphy (TEE)	Outcomes following valvular repair in mitral and tricuspid regurgitation Effects of valvular disorders on CV outcomes, beyond penoperatrve period Role of MELD in selection of appropriate therapy
	Portopulmonary Hypertension	2D Echocardiography	Elevated RVSP/PASP >45mmHg	Volume optimization important prior to echocardiography

Abbreviations: ECG, electrocardiogram; 6MWT, 6-minute walk test; 6MWD, 6-minute walk distance; CPET, cardiopulmonary exercise testing; AT, anerobic threshold; 2D Echocardiography, 2-Dimensional echocardiography; CAD, coronary artery disease; RFs, risk factors; DSE, dobutamine stress echocardiography; MPS, myocardial perfusion scintigraphy; CCTA, coronary CT angiography; CACS, coronary artery calcium score; CCM, cirrhotic cardiomyopathy; LVEF, left ventricular ejection fraction; LVOTO, left ventricular outflow tract obstruction; ESLD, end-stage liver disease; MELD, model for end-stage liver disease; LT, liver transplantation; RVSP, right ventricular systolic pressure; PASP, pulmonary artery systolic pressure

Key Points:

• Cardiovascular disease complications are the leading cause of early mortality among liver transplant recipients

• Traditional cardiac risk factors, such as diabetes, hypertension and hyperlipidemia, are increasing in prevalence among liver transplant candidates and are highly associated with post-liver transplant cardiac events

• The majority of cardiovascular disease complications post-liver transplantation are non-ischemic

• Atrial fibrillation has been shown to be the most common cause of early (within 90 days) major cardiovascular events among liver transplant recipients

• The Cardiovascular Risk in Orthotopic Liver Transplantation (CAR-OLT) score is a CVD risk prediction model that uses pre-transplant demographic, social and clinical variables to estimate the risk of a major CVEs in the first-year post-transplantation

Synopsis:

Cardiovascular disease complications are the leading cause of early (short-term) mortality among liver transplant recipients. The increasingly older candidate pool has multiple comorbidities necessitating cardiac and pulmonary vascular disease risk stratification of patients for optimal allocation of scarce donor livers. Arrhythmias, heart failure, stroke, and coronary artery disease are common pre-transplant cardiovascular comorbidities and contribute to cardiovascular complications post-liver transplantation. Valvular heart disease and portopulmonary hypertension present intra-operative challenges during liver transplantation surgery. To assist with global cardiovascular risk prediction, the CAR-OLT score estimates the risk of cardiovascular complications in liver transplant candidates within the first-year post-transplant.