The Impact of Coronary Artery Disease and Statins on Survival after Liver Transplantation

Abstract

Background:

Cardiovascular disease (CVD) is a major contributor to long-term mortality after liver transplantation (LT) necessitating aggressive modification of CVD risk. However, it is unclear how coronary artery disease (CAD) and development dyslipidemia following LT impacts clinical outcomes and how management of these factors may impact survival.

Methods:

Patients undergoing LT at Virginia Commonwealth University from January 2007 to January 2017 were included (N=495). CAD in all potential LT recipients (LTR) over the age of 50 years or risk factors was evaluated via coronary angiography. Impact of pre-LT CAD after transplantation was evaluated via survival analysis. Additionally, factors associated with new-onset dyslipidemia, statin use, and mortality were assessed using multiple logistic regression or Cox proportional hazards models.

Results:

The mean age of the cohort was 55.3±9.3 years at the time of LT, and median follow-up period was 4.5 years. CAD was noted in 129(26.1%) patients during the pre-LT evaluation. Presence or severity of CAD prior to LT did not impact post-LT survival. Dyslipidemia was present in 96 patients at LT, and 157 patients developed new onset dyslipidemia after LT. Statins were underutilized as only 45% of patients with known CAD were on therapy. In patients with new-onset dyslipidemia, statin therapy was initiated in 111(71%) with median time to initiation of statin therapy was 2.5 years. Statin use conferred survival benefit [HR: 0.25, 95%CI:0.12, 0.49], and was well-tolerated with only 12% of patients developing an adverse event requiring cessation of therapy.

Conclusion:

Pre-LT CAD did not impact survival after LT suggesting potentially a role of accelerated atherosclerosis that may not be captured on pre-LT testing. While statin therapy confers survival benefit, it is underused in LTR.

INTRODUCTION

Cardiovascular disease (CVD) is an important contributor to long-term mortality following liver transplantation (LT)(1–3). This in itself is not shocking, as CVD related mortality is the leading cause of morbidity and mortality in the United States.(4) However, unlike the general population, all potential LT recipients undergo vigorous pre-LT testing aimed at identifying and eliminating individuals deemed to be “high risk” for peri- and post-operative complications.(5) The cardiac evaluation prior to LT generally includes measure of cardiac function by echocardiography and assessment of coronary artery disease (CAD). Patients with heart failure with reduced ejection fraction are nearly universally excluded due to high intra-operative mortality(6). In contrast, while the practice for assessing for CAD is highly variable, patients with abnormal CAD assessment undergo coronary angiography, the gold standard for CAD assessment, and re-vascularization.(7) Those who are unable to be re-vascularized are eliminated from further LT consideration. Since not all patients receiving a LT undergo coronary angiography, it is unclear if CVD and mortality following LT represents an exacerbation of non-obstructive CAD present at the time of LT or if it results from accelerated atherosclerosis that LT recipients are at risk for.

The CVD risk is compounded further following LT as patients are at increased risk of developing dyslipidemia, which is an independent predictor of CVD mortality.(8) There are several mechanisms of post-LT dyslipidemia including corticosteroids, chronic exposure to immunosuppression, weight gain, and development of nonalcoholic fatty liver disease.(9–13) Prior studies have demonstrated that the prevalence of dyslipidemia is high following LT(8) and varies significantly based on etiology of chronic liver disease(1). Post-LT dyslipidemia is characterized by an increase in highly atherogenic small-dense low-density lipoprotein, which is negatively affected the type of immunosuppression and presence of hepatic steatosis.(9,10) Collectively, the presence of dyslipidemia and increase in atherogenic lipoprotein translates clinically into increased morbidity and mortality following LT.(14) However, the impact of dyslipidemia on future risk of mortality and more importantly, the effect of clinically treating dyslipidemia remains poorly defined. To address these limitations within the field, we conducted the following study to (1) link pre-LT CAD to post-LT mortality (2) evaluate the incident rates of dyslipidemia and (3) describe the management of post-LT dyslipidemia and the associated impact on mortality.

METHODS

All patients receiving a LT at Virginia Commonwealth University (VCU) are prospectively enrolled in a study evaluating post-LT outcomes, and the current study represents a retrospective analysis of this cohort. The sub-study was reviewed and approved by the VCU Institutional Review Board. The manuscript was reviewed and approved by all authors prior to submission.

Patient Population

The present analysis included all patients receiving a LT at VCU over a 10-year period between January 1, 2007, and January 1, 2017. Pre-LT CVD is assessed via echocardiography and either a cardiac stress test or coronary angiography. Coronary angiography is performed in all patients age 50 years or greater or presence of risk factors (diabetes, hypertension, dyslipidemia, obesity, family history, smoking history, and personal history of CAD). Those without risk factors and age <50 years have a non-invasive cardiac stress test and coronary angiography reserved for those with an abnormal stress test. The severity and distribution of CAD were defined according to the Coronary Artery Surgery Study (CASS).(15) The decision to intervene on coronary artery stenosis was made in a multidisciplinary approach as described previously.(7,16) Pediatric patients undergoing LT were excluded. Additionally, patients who were not transplanted at VCU medical center and later translocated to VCU were excluded as their pre-LT cardiac evaluation was not readily available. The data was collected pre-LT, at the time of LT and follow-up visits every six months. Unscheduled visits or hospitalizations were reviewed for all patients through review of medical records.

Definitions and Clinical Outcomes

Coronary assessment prior to LT included coronary angiography in patients over the age of 50, risk factors for CAD, and those with abnormal cardiac stress test as described previously.(7,16) CAD noted during the LT evaluation was characterized based on presence of any luminal stenosis of coronary arteries. CAD was further stratified as obstructive if >50% stenosis of any of the three major coronary vessels (right coronary artery, left anterior descending, or left circumflex artery).(15) Cardiac death was defined as death in the setting of myocardial infarction, ventricular arrhythmias, cardiogenic shock, or if the death was sudden and unexplained.(17)

Obesity was defined as body mass index ≥ 30kg/m².(18) Presence of diabetes was defined as use of diabetes medications, hemoglobin A1c >6.5% or elevated serum glucose (fasting > 126 mg/dL or random >200mg/dl).(19) Hypertension was defined as use of anti-hypertensive medications or systolic blood pressure≥130 or diastolic blood pressure≥80.(20) Dyslipidemia was defined as either (1) total cholesterol ≥200 mg/dL, (2) low-density lipoprotein cholesterol (LDL-C) ≥100 mg/dL in diabetics (3) LDL-C > 130mg/dL in non-diabetics or (4) High-density lipoprotein cholesterol (HDL-C) < 40mg/dL and 50 mg/dL in men and women, respectively.(21) Patients were considered to have dyslipidemia if they met these laboratory criteria at any follow-up visit. Patients with a pre-transplant diagnosis of dyslipidemia who remained on statin therapy were also considered to have dyslipidemia even if their lipid profile did not meet laboratory criteria. Guidelines to initiate and titrate lipid-lowering therapy has evolved over the course of the current study period,(21,22) thus to present data regarding statin use in this proof of concept study, appropriate initiation of statin therapy was considered if (1) serum LDL-C > 100 mg/dL in diabetics, (2) LDL-C >130 mg/dL in non-diabetics or (3) total cholesterol > 200mg/dL.(21) Finally, patients with a known history of CVD (myocardial infarction, CAD as documented on coronary angiography, stroke) were considered to be eligible for statin therapy irrespective of lipid profile.(23) Patients with CAD at LT and patients at increased risk for CVD events were considered eligible for aspirin therapy. Aspirin and statin use was determined by reviewing clinical notes (e.g., hepatology, primary care, transplant surgery, and endocrine), as well as a review of the prescriptions filled at the pharmacy. In patients not on statin therapy, medical records were reviewed to evaluate potential reasons for not initiating statin therapy. Adverse events linked to statin therapy include myosititis, myalagias, rhabomyolysis, rising serum transaminases (more than three times the upper limit of normal) and acute liver failure or injury and are quantified in the current study.(24)

Statistical Analysis:

Means and standard deviations (SD), or medians and interquartile range (IQR) in the presence of highly skewed distributions and frequencies and percentages are reported for each study variable at baseline. The prevalence of dyslipidemia at LT, 1, 3, and 5 years post-LT is reported. Factors associated with dyslipidemia were assessed with logistic regression models built to obtain crude and adjusted odds ratios (OR), and their corresponding 95% confidence intervals (95%CI), for age, gender, ethnicity, etiology of liver disease, obesity, hypertension, and diabetes. The generalized variance inflation factor (GVIF) was used to assess multicollinearity in the models. Cox proportional hazard models were built to evaluate the effect of the aforementioned variables on developing dyslipidemia after LT; crude and adjusted hazard ratios (HR) with their 95%CI were calculated. Locally estimated scatterplot smoothing (LOESS) smooth curves of the lipid panel (triglycerides, HDL-C, LDL-C, total cholesterol) were produced to evaluate the relationship between their pattern and the primary etiology of liver disease.

The pattern of statin use was described by reporting the proportions of users among those eligible for statin therapy at LT, 1,3, and 5 years after LT. Logistic regression models were used to assess association with post-LT statin therapy eligibility, and with post-LT statin therapy use. Crude and adjusted OR and 95%CI are reported. Kaplan-Meier analysis was used to describe the time until eligibility for statin therapy and the time until statin therapy initiation. Additionally, linear mixed-effects models were used to estimate the average impact of statin therapy and time on the serum liver enzymes, adjusting by age, gender, diabetes, and etiology of liver disease. Kaplan-Meier analysis was used to describe the survival time by statin therapy and by CAD presence. Cox proportional hazard models were built to evaluate the effect of these variables on survival time, adjusting by age, gender, ethnicity, etiology of liver disease, obesity, hypertension, and diabetes at LT. Adjusted HR and 95%CIs are reported. For all the models built in this analysis, after the full model was obtained, stepwise model selection was conducted, keeping variables with p values larger than 0.25, to determine the variables to hold on the final model, which had the minimum Akaike information criterion value. Statistical analyses were conducted in R (Version 3.5.3, R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

Patient Characteristics

A total of 495 patients underwent LT at VCU between January 2007 to January 2017, and the pre-LT demographics of the study cohort are summarized in Table 1. The mean age at LT was 55.3±9.3 years, and 72.3% were males. The most common etiology of chronic liver disease was hepatitis C (HCV) (n= 227 or 45.9%), followed by alcohol-related cirrhosis (n=79 or 16.0%), and nonalcoholic steatohepatitis (NASH) (n=78 or 15.8%). The prevalence of hypertension, diabetes, and obesity pre-LT was 48.5, 30.5, and 20.0% respectively. The median follow-up period was 4.5 years (IQR 2 - 5 years). Calcineurin inhibitors were the most commonly used immunosuppressants with tacrolimus serving as the main immunosuppressant in 347 (70%) LTR and cyclosporine in 119 (24%) of patients. Sirolimus was used in only 29 (5.9%) of patients. Coronary angiography was performed in 424 of 495 patients. The prevalence of any CAD and obstructive CAD was 26.1% (n=129) and 7.9% (n=39) respectively. Twenty-two patients required re-vascularization prior to LT. (Supplementary Figure 1).

Table 1.

Demographics and Clinical Characteristics of the Cohort (N=495)

Characteristics	N	%
Age, years (mean, sd)	55.3	9.3
Male	358	72.3
Race
Caucasian	361	75.7
African American	116	24.3
Liver Disease Etiology
Alcohol	78	15.8
Hepatitis C	227	45.9
NASH	79	16.0
Other	111	22.4
Cardiovascular Risk Factors
Diabetes	151	30.5
Dyslipidemia	96	20.3
History of CAD	7	1.4
Hypertension	240	48.5
Obesity	95	20.0
CAD found on Pre-LT evaluation
Any CAD	129	26.1
Obstructive CAD	39	30.2
Coronary intervention	22	17.1
Laboratory Test (mean, sd)
AST, U/mL	40.0	41.5
ALT, U/mL	69.2	75.1
MELD Score	16.6	9.3
Mortality
All-cause mortality	120	24.2
Cardiovascular mortality	29	5.9

ALT, Alanine transaminase; AST, Aspartate transaminase; CAD, Coronary artery disease; MELD, Model for End-Stage Liver Disease; NASH, Nonalcoholic steatohepatitis

Dyslipidemia

At the time of LT, 96 patients (20.3%) had a diagnosis of dyslipidemia. Patients with NASH had the highest prevalence (N= 35 or 46.1%), followed by alcohol-related cirrhosis (N=15 or 19.7%), and HCV (N=24 or 11.4%) (p<0.001). Factors associated with dyslipidemia prior to LT included older age, NASH as an indication for LT, diabetes, hypertension, and CAD (Table 2). In an adjusted model, older age (OR 1.06, 95% CI: 1.01, 1.11, P=0.017), NASH as indication for LT (OR 2.95, 95%CI: 1.37, 6.56, P=0.007), and hypertension (OR 2.05, 95%CI: 1.13, 3.80, p=0.020) remained significantly associated with diagnosis of dyslipidemia (Table 2).

Table 2.

Association between Clinical Characteristics and Dyslipidemia at the Time of Liver Transplantation

Characteristics	Unadjusted	p-value	Adjusteda	p-value	Adjustedb	p-value
	OR (95% CI)		OR (95% CI)		OR (95% CI)
Age	1.06 (1.03, 1.10)	<0.001	1.05 (1.00, 1.10)	0.052	1.06 (1.01, 1.11)	0.017
Male	1.04 (0.64, 1.74)	0.885	1.10 (0.55, 2.28)	0.792
Ethnicity
Caucasian	1.00		1.00
African American	0.76 (0.43, 1.30)	0.331	0.97 (0.43, 2.08)	0.947
Etiology
Alcohol	1.00		1.00		1.00
HCV	0.50 (0.25, 1.02)	0.051	0.46 (0.22, 0.98)	0.040	0.44 (0.21, 0.93)	0.028
NASH	3.34 (1.66, 7.00)	<0.001	2.57 (1.14, 5.92)	0.024	2.95 (1.37, 6.56)	0.007
Hypertension	2.84 (1.78, 4.63)	<0.001	1.72 (0.93, 3.24)	0.090	2.05 (1.13, 3.80)	0.020
Diabetes	2.31 (1.46, 3.65)	<0.001	1.59 (0.86, 2.91)	0.138
CAD	2.22 (1.38, 3.54)	<0.001	1.60 (0.87, 2.90)	0.127

^a.Variables in the model: age, sex, ethnicity, etiology, obesity, hypertension, diabetes, and CAD.

^b.Variables in the model: age, etiology, obesity, and hypertension.

CAD, Coronary artery disease; HCV, Chronic Hepatitis C; OR, Odds ratio; NASH, Nonalcoholic steatohepatitis

The prevalence of dyslipidemia post LT at 1, 3, and 5 years was 32.5% (n=146), 46.8% (n=142), and 55.3% (n=115), respectively. The rate of missing lipid profile during the first two-years after LT consisted of 111 (22.4%) patients, however, of these 40 patients had died within 1 year post-LT. No systematic differences were noted among patients who had lipid profile performed within 2 years following LT and those who did not. In patients without CAD and dyslipidemia prior to LT, the annual incident rates of dyslipidemia are depicted in Figure 1. The median time from LT to developing dyslipidemia was 1.5 years (IQR 0.5, 3.0). The specific changes in serum total cholesterol, LDL-C, HDL-C, and triglycerides are delineated in Supplementary Figures 2A–D stratified to etiology of liver disease. Total cholesterol increased across all etiology of chronic liver disease over time, but the rate of change was the greatest within the first 2 years following LT. Similarly, serum triglycerides also rapidly increased across etiologies of chronic liver disease with time with the greatest increase in slope occurring within the first 2 years following LT. The change in serum LDL-C and HDL-C were similar across etiology of chronic liver disease.

Figure 1:

The Annual Incident Rates of Dyslipidemia after Liver Transplantation.

In unadjusted analysis, factors associated with the development of dyslipidemia post-LT included male gender, HCV as the indication for LT, hypertension, diabetes, and the presence of CAD. In the adjusted model, subjects with HCV as the indication for LT, were less likely to develop dyslipidemia post LT (HR 0.54 95%CI: 0.36, 0.83, P=0.004), while those with pre-LT CAD were more likely (HR 1.56, 95%CI: 1.08, 2.27, P=0.019) (Table 3).

Table 3.

Association between Clinical Characteristics and New-Onset Dyslipidemia Following Liver Transplantation

Characteristics	Unadjusted	p-value	Adjusteda	p-value	Adjustedb	p-value
	HR (95% CI)		HR (95% CI)		HR (95% CI)c
Age	1.02 (0.99, 1.03)	0.074	1.00 (0.99, 1.03)	0.831
Male	0.66 (0.48, 0.92)	0.014	0.67 (0.44, 1.02)	0.065	0.68 (0.46, 1.01)	0.059
Ethnicity
Caucasian	1.00		1.00
African American	0.89 (0.61, 1.31)	0.557	0.99 (0.61, 1.62)	0.977
Etiology
Alcohol	1.00		1.00		1.00
HCV	0.53 (0.35, 0.81)	0.003	0.56 (0.36, 0.88)	0.012	0.54 (0.36, 0.83)	0.004
NASH	1.63 (0.98, 2.71)	0.058	1.37 (0.78, 2.39)	0.269	1.48 (0.88, 2.47)	0.137
Obesity	1.11 (0.75, 1.62)	0.607	0.94 (0.60, 1.46)	0.779
Hypertension	1.51 (1.10, 2.07)	0.010	1.14 (0.77, 1.70)	0.506
Diabetes	1.46 (1.05, 2.02)	0.011	1.14 (0.76, 1.70)	0.530
CAD	1.80 (1.29, 2.53)	<0.001	1.54 (1.03, 2.31)	0.038	1.56 (1.08, 2.27)	0.019

^a.Variables in the model: age, sex, ethnicity, etiology, obesity, hypertension, diabetes, and CAD pre-LT.

^b.Variables in the model: sex, etiology, and CAD.

CAD, Coronary artery disease; HCV, Chronic Hepatitis C; HR, Hazards ratio; NASH, Nonalcoholic steatohepatitis

Pattern of Statin and Aspirin Use

Prior to LT, only 14.5% (n=16) with any CAD were on statin therapy, and the proportions of patients on statin was slightly higher in patients with obstructive CAD (n=11 or 33.3%). After LT, 374 patients (75.6%) were eligible for statin therapy during the study period; however, only 46.8% (n=175) of the eligible patients received statin therapy. A deeper analysis of the patients eligible for statin therapy, 18.4% (n=18) of the patients with any CAD were on statin therapy at 1-year, 45.7% (n=32) at 3-years, and 40.0% (n=12) at 5-years. In patients with obstructive CAD at LT who underwent revascularization, 50% were on statin therapy at 1-year.

In patients with new onset dyslipidemia, the median time to initiation of statin therapy from diagnosis of dyslipidemia was 2.5 years (IQR 0.5, 4.5) (Figure 2). Initiation of statin therapy following LT in patients with CAD is depicted in Figure 2. The likelihood of initiating statin therapy was higher among diabetics with OR of 1.58 (95% CI 1.00, 2.51, P=0.049) and African Americans with OR of 1.80 (1.05, 3.09, P=0.03). In contrast, males (OR 0.45, 95% CI 0.28, 0.71, P<0.001), and patients with HCV (OR 0.46, 95% CI 0.24, 0.86, P=0.02) were less likely to be started on statin therapy. Finally, LTR started on statin therapy had lower aspartate transaminase (AST) values than eligible patients that did not start statin therapy (Adjusted difference 6.0 U/mL, 95%CI −3.1, 15.1 U/mL).

Figure 2:

Initiation of Statin Therapy in Patients with CAD and New-Onset Dyslipidemia

After initiation of statin therapy, 26.3% (n=46) of patients did not have a follow-up lipid profile assessed within 2 years. In patients with a repeat lipid profile after initiation of statin therapy, 60% (n=77) were adequately titrated to either high-dose statin therapy or goal LDL-C level. The intensity of statin therapy was increased in 19.2% (n=10) of patients with a sub-optimal reduction in serum LDL-C. Finally, the majority of LT recipients started on statin therapy received moderate-intensity statin therapy (Figure 3).

Figure 3:

Pattern of Statin Use after Liver Transplantation

A total of 286 LTR were eligible for aspirin therapy, however only 117 (40.9%) LTR were on aspirin during the study duration. In unadjusted and adjusted regression models, history of hypertension was positively associated with aspirin utilization (OR 1.91, 95% CI: 1.09, 3.42, p=0.03). While, African Americans and patients with NASH were more likely to use aspirin, this did not reach statistical significance (Supplementary Table 1).

Adverse effects associated with statin therapy

Statin therapy was interrupted in 53 (30.3%) of patients. Of these, 28 (16.0%) had temporary discontinuation of statin therapy, while 4 (2.2%) patients stopped statin therapy due to medical non-compliance without a reported adverse event. Twenty-one patients (12%) had documented an adverse event to statin therapy, and muscle related complications were the most common complication occurring in 11 (6.3%) patients. A mild increase in serum AST levels occurred in patients started on statin therapy when compared to baseline, however, similar increase rate was observed in AST levels when comparing patients with and without statin therapy (P=0.52). There was a trend towards a mild increase in serum alanine transaminase (ALT) in patients on statin therapy when compared to those not on statin therapy (adjusted difference 16.1 U/mL, 95%CI: 1.6, 30.6 U/mL, P=0.08) but this too failed to reach statistical significance. No patients developed liver failure or acute liver injury from statin use.

Mortality

Over the study duration, 120 (24.2%) patients died with a median time to death from LT of 2.1 years (IQR: 0.5, 4.6). The three major causes of post-LT deaths included malignancy (23.3%, n=28), CVD (19.2%, n=23) and infections (17.5%, n=21). After adjusting for age, gender, ethnicity, primary etiology of liver disease, obesity, hypertension, and diabetes at LT no statistically significant association was observed between mortality and presence of CAD (HR 0.91, 95%CI 0.54 , 1.55) or severity of CAD (HR 1.11, 95% CI: 0.52 , 2.36) (Figure 4A). Statin therapy had a protective effect on survival, it was associated with reduced mortality with HR of 0.25 (95% CI, 0.12, 0.49, P<0.001). The beneficial effect of statin on survival persisted even after adjusting for age, gender, ethnicity, etiology of liver disease, obesity, hypertension, diabetes, and CAD presence at LT (Figure 4B). The association between mortality and statin use was further assessed by evaluating the interaction between CAD and statin use, which was not found to be significant [HR 0.27, 95% CI 0.06, 1.17].

Figure 4.

A: Survival Curves by CAD Presence and Severity at LT

B: Survival Curves by Use of Statin Therapy in Eligible Patients.

Aspirin use (irrespective of statin use) did not impact overall survival in unadjusted analysis (HR 0.82, 95% CI 0.51, 1.34) or after adjusting age, gender, ethnicity, etiology of liver disease, obesity, hypertension, diabetes, and CAD presence at LT (adjusted HR 0.69, 95%CI: 0.38, 1.25). In patients on statin, aspirin use was associated with no survival benefit (HR 0.31 95%CI: 0.06, 1.55). Finally, no evidence of an interaction between aspirin and CAD on overall survival [HR 1.73, 95%CI: 0.52, 5.75] was noted after adjusting by age, gender, ethnicity, etiology of liver disease, obesity, hypertension, and diabetes.

DISCUSSION:

A major milestone after LT is 1-year survival, which has steadily increased over the past few decades; however, long term survival remains significantly lower when compared to age- and gender-matched controls in non-LT population.(25) Efforts to improve long-term mortality require a granular understanding of key contributors to post-LT mortality, particularly cardiovascular disease, which is one of the leading contributors to long-term mortality following LT(1,3). In the present study, we confirmed the importance of CVD associated mortality following LT but also noted that the presence and severity of CAD at the time of LT did not impact overall or cardiovascular mortality. As all potential LT candidates with obstructive CAD undergo revascularization prior to LT, the natural history of CAD in LT is therefore altered. It is possible that the current study was not adequately powered to detect the smaller effect size since the number of subjects with revascularized CAD was relatively small. Biologically, the risk of poor outcomes following coronary thrombosis is linked to the robustness of collateral circulation.(26) Therefore, slowly developing coronary stenosis with an adequate collateral flow are less likely to lead to fatal events or sequelae following a coronary event. In contrast, rapidly developing stenosis due to accelerated atherosclerosis as has been demonstrated following LT with presumably inadequate collateral formation are more likely to lead to fatal outcomes or lasting sequelae.(27) Prior studies in LTR have demonstrated a close association between LT, atherogenic lipoproteins, and pro-inflammatory mediators.(9,10) Furthermore, a longitudinal increase in markers of endothelial dysfunction following LT has also been reported.(28) These factors are linked to atherosclerosis and give credence to the theory that rapid atherosclerosis following LT likely plays a more important role in post-LT outcomes than the presence of native CAD present at the time of LT.(9,10)

Dyslipidemia is a strong, independent, and more importantly, modifiable risk factor for CAD and CVD associated mortality.(29,30) In the present study, we define a time-dependent increase in the prevalence of dyslipidemia in LTR that varied with etiology of end-stage liver disease requiring LT. These findings reaffirmed prior published reports demonstrating the greatest prevalence of dyslipidemia among patients transplanted for NASH cirrhosis.(1) Dyslipidemia in patients after LT was largely characterized by a uniform increase in serum triglycerides, total cholesterol, and a concurrent decrease in HDL-C. Factors associated with the development of dyslipidemia included diabetes and NASH as etiology of chronic liver disease suggesting that these should potentially be targeted to reduce future risk. Mechanistic studies have previously demonstrated that cyclosporine leads to dyslipidemia by reducing the activity of hepatic cholesterol-7α-Hydroxylase, rate limiting enzyme in bile acid synthesis, and thereby by reducing incorporation of cholesterol into bile acids.(31) Furthermore, cyclosporine also increases hepatic lipoprotein production and reduces lipoprotein clearance.(32) Recent studies also implicate the association between cyclosporine and more atherogenic lipoproteins sub-particles such as small dense LDL-C, very low-density particle size, and concentration.(9,10) Thus, it may be prudent to use tacrolimus as the main immunosuppressant in LTR at risk for developing dyslipidemia, such as obese, diabetics and those transplanted for NASH cirrhosis, to mitigate the effects of immunosuppression on dyslipidemia following LT.

Dyslipidemia is a major modifiable risk factor for CAD, and the mainstay of therapy includes at least use of 3-hydroxy-3-methyl-glutaryl-coenzyme A reductase inhibitors or statins.(22) There is currently an abundance of literature demonstrating the efficacy of statin in primary and secondary prophylaxis for reducing myocardial infarction, stroke and CVD associated mortality.(29,33,34) Furthermore, statin therapy has demonstrated a mortality benefit even patients with chronic liver disease,(35) however, statins are often avoided in patients with chronic liver disease for fear of hepatotoxicity, even in patients with obstructive and severe CAD(16,36). In a recent study of patients with decompensated cirrhosis, statins were safely tolerated and did not impact survival or need for transplantation.(16) Our study expands the published literature by demonstrating that statin therapy is safe after LT and are not associated with hepatotoxicity. Furthermore, statins are under-utilized in LTR in whom there is an indication either due to dyslipidemia or presence of known CAD. This is concerning as the use of statin therapy after LT was associated with a reduction in overall mortality. While side effects can limit the use of statin therapy, in our study cohort only a small number of patients developed any reportable side-effects. Thus, statin therapy is safe, effective with limited adverse effects after LT and should be aggressively initiated and titrated to improve long-term survival following LT. Even though the results from the current study tend to a positive effect of aspirin use on survival, the confidence interval is wide and does not allow us to give any definite conclusion. Future studies with larger samples sizes will be needed to explore this complex interaction. Males and patients with HCV related liver disease were less likely to be on statins whiles patients with hypertension were more likely to be on aspirin therapy. This could from multiple interrelated factors, including awareness, social status, income levels as well as other coexisting medical problems.

There are several limitations in the current study worth mentioning. First, since all patients with obstructive CAD were revascularized, the current study is not able to provide the true natural history of obstructive CAD following LT; however, the study population is representative of patients receiving LT as those with obstructive CAD not amenable to revascularization are excluded from further LT consideration. Second, in the present study while the total sample size was large, the number of deaths attributable solely to CVD was relatively small to determine key relationship between pre- and post-LT clinical parameters and CVD associated mortality. The survival benefit from aspirin use may not have been demonstrated due to type 2 error. This limitation will be a major limitation of any single center study and requires a well-designed prospective natural history with a large sample size to address. Finally, due to the retrospective design of the current study, we are unable to explore what physician and patient-related factors may have influenced the decision not to initiate statins. Understanding this is of paramount importance if we are to improve the management of CVD following LT. Prior survey-based studies have demonstrated that only a small minority of primary care physicians feel comfortable taking care of LTR and the decision to not start statins in LTR therefore, may be reflective of this. However, well-planned studies that use mixed‐methods approaches are necessary to truly understand the underutilization of statins in LT recipients.

In summary, CAD at the time of LT does not impact overall survival following LT, suggesting the role of accelerated atherosclerosis post-LT. Statin therapy is safe, well-tolerated, and confer a mortality benefit in LTR but are underutilized.

Supplementary Material

Supp figS1-2

Supplementary Figure 1: Consort diagram: Result of Pre-LT CAD evaluation

Supplementary Figure 2 A: Changes in serum total cholesterol after liver transplantation stratified to etiology of liver disease.

Supplementary Figure 2 B: Changes in serum LDL-C after liver transplantation stratified to etiology of liver disease.

Supplementary Figure 2 C: Changes in serum HDL-C after liver transplantation stratified to etiology of liver disease.

Supplementary Figure 2 D: Changes in serum triglycerides after liver transplantation stratified to etiology of liver disease.

Abbreviations:

ALT

Alanine transaminase

AST

Aspartate transaminase

CAD

Coronary artery disease

CASS

Coronary Artery Surgery Study

CI

Confidence intervals

CVD

Cardiovascular disease

GVIF

Generalized variance inflation factor

HCV

Hepatitis C

HDL-C

High-density lipoprotein cholesterol

HR

Hazard ratio

IQR

Interquartile range

LDL-C

Low-density lipoprotein cholesterol

LOESS

Locally estimated scatterplot smoothing

LT

Liver transplantation

LTR

Liver transplant recipients

OR

Odds ratios

NASH

Nonalcoholic steatohepatitis

SD

standard deviation

VCU

Virginia Commonwealth University