Cardiovascular Disease in Nonalcoholic Steatohepatitis: Screening and Management

Abstract

Purpose of Review

The global burden of non-alcoholic steatohepatitis (NASH) as a major cause of chronic liver disease continues to rise. Cardiovascular disease (CVD) is a leading cause of morbidity and mortality in this patient population. The current review summarizes recent advances in the understanding of CVD in NASH and strategies for screening and management.

Recent findings

Large genetic epidemiological studies support the intricate role of the metabolic syndrome in the pathophysiology of CVD risk in patients with NASH. Atherosclerotic CVD risk scores can predict elevated CV risk in NASH, but additional work is necessary to refine risk stratification and to guide optimal management. New antidiabetic agents may offer benefit in treating steatosis and reducing CV morbidity in NASH.

Summary

Achieving improved outcomes in patients with NASH requires that future efforts focus on optimizing methods for CVD screening and designing clinical trials with long-term cardiovascular endpoints in mind.

Introduction

Non-alcoholic fatty liver disease (NAFLD) is a chronic condition that is intricately associated with the metabolic syndrome. Mirroring the rising global burden of obesity, NAFLD is rapidly becoming one of the most common etiologies of chronic liver disease, affecting one out of every four individuals worldwide [1, 2]. Non-alcoholic steatohepatitis (NASH), a subtype of NAFLD characterized by inflammation and hepatocyte injury, can ultimately lead to cirrhosis and hepatocellular carcinoma and is now the second-most common indication for liver transplantation (LT) in the United States [3, 4]. However, cardiovascular disease (CVD), rather than liver decompensation, remains the leading cause of morbidity and mortality in patients with NAFLD and NASH [5]. While a significant proportion of the cardiovascular risk in NASH is likely attributable to shared metabolic risk factors, there is much interest in understanding the independent contribution of hepatic steatosis and associated inflammation and fibrosis. Extensive research documents a high prevalence of various sub-clinical and clinical manifestations of atherosclerotic CVD (ASCVD) in NAFLD, and yet there remains an ongoing need for evidence-based guidance on the optimal approach to CVD screening, prevention, and treatment in this high-risk population.

The purpose of this review is to provide an up-to-date summary focused on highlighting recent advances in the understanding of mechanisms, screening, prevention and management of CVD in patients with NAFLD.

Pathophysiologic mechanisms linking NAFLD and CVD

One of the principal pathologic underpinnings of CVD is the formation of the atherosclerotic plaque [6, 7]. Briefly, plaque forms when endothelial dysfunction allows oxidized low-density lipoprotein (LDL) to deposit within subendothelial macrophages. In turn, inflammatory cytokine release and vascular smooth muscle proliferation create a “vulnerable” plaque that is prone to rupture and occlusion, particularly in pro-thrombotic states. In individuals with NASH, numerous pro-atherogenic derangements in these processes are proposed to contribute to increased cardiovascular risk (Figure 1).

Figure 1 –

Pathophysiologic mechanisms linking hepatic steatosis to CVD

Dyslipidemia

Adipose tissue expansion and insulin resistance (IR), both precursors for the formation of hepatic steatosis, cause triglyceride (TG) accumulation within hepatocytes. Hepatic TG burden causes an atherogenic serum lipoprotein profile by increasing production of very-low-density lipoprotein (VLDL) and small-dense LDL (sdLDL) and decreasing high-density lipoprotein (HDL) levels [8–10].

Oxidative stress and systemic inflammation

“Lipotoxicity” within hepatocytes induces oxidative stress, upregulating hepatic inflammatory cascades. The result is increased serum levels of cytokines including interleukin-6 (IL-6), high sensitivity C-reactive protein (hs-CRP), IL-1β, and tumor necrosis factor alpha (TNF-α) and liver-specific proteins including Fetuin-A (FetA) and fibroblast growth-factor 21 (FGF21) [11–13]. Abundant research links these proteins to development of CVD, including stroke and coronary artery disease (CAD) [14–20].

Hemostatic and endothelial dysfunction

Hepatic fat accumulation alters hemostasis by increasing prothrombotic factors and plasminogen activator inhibitor type 1 (PAI-1) and exacerbates endothelial dysfunction through elevated serum levels of asymmetric dimethyl arginine (ADMA), a nitric oxide synthase (NOS) antagonist that is also a known biomarker of CVD [21–23].

Taken together, individuals with NAFLD exhibit a state of dyslipidemia in a milieu of chronic inflammation, oxidative stress, hemostatic and fibrinolytic alterations, and endothelial dysfunction – an environment ripe for atherogenesis and CVD. However, as IR is a sine qua non in NAFLD, it remains difficult to disentangle the effects of IR and metabolic syndrome from the independent contribution of processes occurring in and around the liver when studying CVD risk. Recently, however, investigators have turned to evaluating known NAFLD susceptibility genes in an effort to help answer this perplexing question.

Genetic studies

Two of the most robust genetic polymorphisms associated with NAFLD development and progression are patatin-like phospholipase domain-containing protein 3 (PNPLA3) and transmembrane 6 super family 2 (TM6SF2). PNPLA3 I148M and TM6SF2 E167K are variants that interfere with hepatic TG metabolism [24, 25]. Both are consistently shown to predispose to steatosis and associate with increased severity of NAFLD at all stages [26].

In a meta-analysis of 48 genome-wide association studies, Simons et al. interestingly showed that both PNPLA3 and TM6SF2 conferred a small, but statistically significant, protective effect against CAD; this has previously been demonstrated for TM6SF2 alone [27, 28]. In addition, in a Mendelian randomization study of PNPLA3 in a Dutch cohort, despite a clear association between PNPLA3 and hepatic fat content, there was no association between the presence of the PNPLA3 variant and CAD (OR per allele 0.98, 95% CI 0.95–1.02) [29].

These genetic epidemiologic studies question the notion that hepatic steatosis directly and independently contributes to CV risk. Still, these studies are unable to completely untangle the connection between hepatic physiology and metabolic regulation. Indeed, a commonly posited theory for the above findings are that the cardioprotective effects of the PNPLA3 and TM6SF2 gene variants may be explained by a secondary effect of decreased circulating VLDL [27].

While the evidence thus far clearly supports a powerful supporting role of the liver in many pathophysiologic processes that lead to CVD, a direct and independent link between NAFLD and CVD, in particular CAD and atherosclerosis, remains difficult to isolate.

Subclinical and clinical CVD in patients with NAFLD

Subclinical disease

A variety of validated non-invasive measures of underlying ASCVD exist in clinical practice. Coronary artery calcium (CAC), a widely studied measure of subclinical atherosclerosis, has high sensitivity and specificity in detecting CAD [30]. Carotid intima-media thickness (CIMT) and brachial artery flow-mediated vasodilation (FMD) are also predictive of CV events, including myocardial infarction (MI) and stroke [31, 32]. In two recent meta-analyses, participants with NAFLD had significantly higher CAC scores, CIMT, and FMD than controls [33, 34].

Individuals with NAFLD also demonstrate cardiac structural changes associated with preclinical ventricular remodeling and dysfunction that may predispose to clinical heart failure (HF). Studies demonstrate increased left ventricular (LV) mass and left atrial (LA) volume, lower early diastolic relaxation velocity, higher LV filling pressures, and lower trans-mitral peak early to late ventricular filling ratios [35–41]. Many of these echocardiographic parameters are also shown to worsen over five years in persons with NAFLD [42]. Furthermore, the presence of underlying NASH or advanced fibrosis correlate with increased severity of the observed cardiac geometric changes and indices of diastolic dysfunction [37, 40].

Coronary artery disease

Individuals with NAFLD experience higher rates of clinical CAD and worse outcomes after coronary events. Patel et al. prospectively studied 228 patients undergoing coronary angiography as part of LT evaluation [43]. After adjusting for traditional CAD risk factors, individuals with NASH had significantly higher rates of severe CAD compared to those with hepatitis C or alcohol-related cirrhosis. Patients with NAFLD also have higher prevalence of coronary lesions requiring percutaneous coronary intervention, in-hospital mortality during an acute coronary syndrome episode, and 3-year mortality after acute ST-segment elevation MI [44–46].

Cerebrovascular disease

Whether NAFLD is associated with clinical cerebrovascular disease is less clear. A 2016 meta-analysis by Haddad et al. of six studies with 5,953 NAFLD patients demonstrated a 2-fold increase in the relative risk of ischemic stroke versus controls (RR 2.09, 95% CI 1.46–1.98) [47]. Furthermore, in a cohort of 306 patients with acute brainstem infarctions, NAFLD was associated with stroke severity, as determined by National Institutes of Health Stroke Scale scores >7 (HR 2.24, 95% CI 1.25–4.01), and with stroke progression [48].

In contrast, in a population-based European cohort of over 120,000 patients with NAFLD followed for up to five years, there was no association between NAFLD and stroke [49]. Finally, a population-based study of over 30,000 individuals in the REGARDS (Reasons for Geographic and Racial Differences in Stroke) cohort found an inverse association between stroke risk and NAFLD (diagnosed by fatty liver index [FLI] > 60) in men [50]. Of note, a positive correlation was noted only in women with FLI above the 90th percentile.

There is, however, data to suggest that the presence of fibrosis in NAFLD may be associated with cerebrovascular disease. In the United States National Health and Nutrition Examination Survey (NHANES) 2005 to 2014, persons with NAFLD-fibrosis (defined using the FIB-4 index) experienced higher rates of stroke compared to those without fibrosis (OR 1.87, 95% CI 1.00–3.50) [51]. Baik et al. also described 395 patients with stroke or transient ischemia attack (TIA) undergoing transient elastography, noting that only the degree of fibrosis, not steatosis, predicted mortality [52].

Structural and functional cardiac disease

There is an increased incidence of cardiac valvular disease in NAFLD. Patients with type 2 diabetes (T2DM) and NAFLD have increased rates of aortic valve sclerosis and/or mitral annular calcification after adjustment for traditional CV risk factors and diabetes variables (adjusted OR 2.70, 95% CI 1.23–7.38) [53].

Despite knowledge of increased subclinical indices of ventricular dysfunction in patients with NAFLD, data on occurrence of HF events and its sequelae are sparse. In a study of 102 inpatients with HF undergoing liver ultrasound in Beijing, those with NAFLD had higher LV mass indices and LV fibrosis, but no significant difference in subsequent major adverse cardiac events (MACE) or readmission rates over one year [54]. In contrast, Valbusa et al. have reported data from Italy among patients admitted for decompensated HF, showing that NAFLD is associated with higher rates of 1-year all-cause (aHR 5.05, 95% CI 2.78–9.10) and cardiac rehospitalizations (aHR 8.05, 95% CI 3.77–15.8) and all-cause mortality (aHR 1.82, 95% CI 1.22–2.81) over a two-year period [55, 56].

Cardiac arrhythmias

The risk of atrial fibrillation (AF) has been studied extensively in NAFLD, with conflicting findings [57–60]. However, a 2019 meta-analysis incorporating nine cross-sectional and longitudinal studies of over 360,000 middle-aged and elderly individuals demonstrated a significant association between NAFLD and prevalent AF (OR 2.07, 95% CI 1.38–3.10), independent of other risk factors, and an increased risk of incident AF in those with NAFLD and T2DM [61].

Additional conduction abnormalities observed in patients with NAFLD include QTc prolongation, ventricular arrhythmias, and various forms of heart block [62–64].

Overall cardiovascular events and cardiovascular mortality

To date, the question of whether NAFLD can be considered an independent risk factor for the development of CV events or CVD-related mortality remains controversial. A number of recent cohort studies and meta-analyses have added to the cumulative trove of evidence and help shed additional light on the subject.

Regarding CV events, Zeb et al. studied 4,119 participants in the Multi-Ethnic Study of Atherosclerosis (MESA) over a median follow-up of 7.6 years and noted that NAFLD (diagnosed by computed tomography [CT]) was associated with incident non-fatal CAD events, defined as first occurrence of MI, resuscitated cardiac arrest, or angina with or without revascularization (HR 1.74, 95% CI 1.25–2.41) [65]. Additional studies of NAFLD patients with or without T2DM show a similar degree of increase in risk of CV events [66, 67].

Regarding CV mortality, the picture is less clear. Table 1 summarizes four large meta-analyses published over the past five years. In 2016, Targher et al. evaluated 16 studies of 34,043 individuals (36.3% with NAFLD, defined by imaging or histology) and noted a higher risk of fatal and/or non-fatal CVD events (pooled OR 1.63, 95% CI 1.26–2.13) [5]. However, when the analysis was restricted to studies evaluating CV mortality alone, there was no significant association (pooled OR 1.31, 95% CI 0.87–1.97). In the same year, Wu et al. published a meta-analysis of 34 cross-sectional and cohort studies with 164,494 participants, showing an association between NAFLD (defined by imaging, histology, or liver enzymes) and both prevalent (pooled HR 1.81, 95% CI 1.23–2.66) and incident (pooled HR 1.37, 95% CI 1.10–1.72) CVD, but again there was no association with CVD-related mortality (pooled HR 1.10, 95% CI 0.86–1.41) [68]. In 2019, Liu et al. included 498,501 patients in 14 studies, of which seven included data on CVD-specific mortality. In patients with NAFLD (defined by imaging or histology), there was no significant association with the risk of death from CVD (pooled HR 1.13, 95% CI 0.92–1.38) [69]. These three meta-analyses collectively suggest against a significant association between NAFLD and CV mortality. A fourth meta-analysis, published by Haddad et al. in 2017 and including six studies with over 25,000 individuals, seems at first glance to offer contrary evidence [47]. In two out of the six studies in this meta-analysis with necessary outcome data, Haddad reported higher CV mortality in individuals with NAFLD (pooled RR 1.46, 95% CI 1.31–1.64). Importantly, Haddad et al. used unadjusted RRs in their analysis. However, neither of the original studies, which consisted of overlapping NHANES-III samples, were able to demonstrate increased CV mortality in NAFLD after adjust ment for major confounders [70, 71].

Table 1 –

Published meta-analyses as of 2016 investigating the association between NAFLD and cardiovascular events and/or mortality

Authors, year [Ref.]	Inclusion/exclusion criteriaa	NAFLD diagnosis	Number of studies (Total, reporting CV mortality)	Sample size (Total, with NAFLD)	Heterogeneity (I2 statistic)b	Pooled point estimate for CVD-related mortality, (95% CI)
Targher et al., 2016 [5]	Any cross-sectional, prospective or retrospective study Any language	Imaging or histology only	Total: 16 CV mortality: 7	Total participants: 34,043 No. with NAFLD: 12,381	90%	OR 1.31, (0.87–1.97)
Wu et al., 2016 [68]	Any cross-sectional, prospective, or retrospective study events or mortality Any language exclusion	Imaging, histology, or labs	Total: 34 CV mortality: 5	Total participants: 164.494 No. with NAFLD: not reported	64.9%	HR 1.10, (0.86–1.41)
Haddad et al., 2017 [47]	Prospective studies only Control group without NAFLD English language only	Imaging only	Total: 6 CV mortality: 2	Total participants: 25,837 No. with NAFLD: 5,953	0%	RR 1.46, (1.30–1.64)c
Liu et al., 2019 [69]	Excluded studies with patients undergoing LT or bariatric surgery	Imaging or histology	Total: 14 CV mortality: 7	Total participants: 498,501 No. with NAFLD: 95,111	57.5%	HR 1.13, (0.92–1.38)

^aThe following inclusion/exclusion criteria were common to all meta-analyses: adult (>18 years) participants only and included studies reported association between NAFLD and CV mortality,

^bI² statistic only for subset of studies with outcome of CV mortality

^cReported RRs are unadjusted. The original studies reported adjusted HRs and showed no statistical increase in CV mortality.

An important consideration is the role of NAFLD disease stage in cause-specific outcomes. In the aforementioned Targher et al. meta-analysis, more “severe” NAFLD was associated with an overall higher risk of fatal and/or non-fatal CVD events [5]. However, in the included studies, “severe” was defined in a variety of disparate ways, including elevated gamma-glutamyl transferase, high hepatic ¹⁸F-fluoro-2-deoxyglucose uptake on positron emission tomography, high NAFLD fibrosis score, or histology. Despite these limitations, the association between CV endpoints and NAFLD-fibrosis, rather than simple steatosis, is supported by other studies as well [72–74].

To summarize, the evidence to-date supports the notion that NAFLD, and in particular advanced fibrosis, is associated with increased risk of CV events independent of metabolic comorbidities. However, investigators have been unable to definitively demonstrate an increase in CV mortality in this patient population.

Screening

There is limited guidance on CVD screening and risk-stratification specific to the NAFLD/NASH population. The American Association for the Study of Liver Diseases (AASLD) and European Association for the Study of the Liver (EASL) recommend comprehensive work-up of CVD risk factors but offer no more detailed guidance [3, 75]. The American College of Cardiology (ACC) and American Heart Association (AHA) guidelines on CVD prevention include a major emphasis on calculation of 10-year ASCVD risk in the general popluation [76].

The Framingham Risk Score and the ASCVD risk score are two such commonly used tools. The severity of hepatic steatosis correlates with estimated CVD risk using both scores [77]. Furthermore, in NHANES-III from 1988 to 1994, participants with NAFLD and a high-risk ASCVD score (≥7.5%) had a two-fold increase in CV mortality (aHR 2.02, 95% CI 1.12–3.65) compared to NAFLD participants with scores <7.5% [78]. Notably, the cardiac-specific mortality risk was reduced by approximately 45% if the 10-year ASCVD score was lowered to below 7.5%. Interestingly, the presence of advanced fibrosis (as determined by high NFS scores) did not add to CV risk determination in this study.

The available evidence suggests that established risk scores can estimate risk of ASCVD risk in patients with NAFLD; however, many unanswered questions remain. In younger patients with NAFLD (e.g., age < 40 years), in whom available risk scores are not validated, how can ASCVD risk be estimated? In these patients, should metabolic assessment occur more frequently than every 4 to 6 years as currently recommended by ACC/AHA guidelines? Is there a role for routine assessments of subclinical ASCVD, such as the use of CAC scoring or CRP to re-classify risk? Can newly developed HF risk scores predict HF events among persons with NAFLD [79]? Finally, does the presence of steatohepatitis or fibrosis alter CV risk such that they should be incorporated into CV risk scores in persons with NAFLD? These and other questions represent important future directions for research in the field.

Management

There is significant overlap in many aspects of management of NAFLD and CVD. Below, and in Table 2, we summarize the important components of management for each condition with a focus on strategies to optimize CV morbidity and mortality.

Table 2 -.

Summary of management of metabolic risk factors and cardiovascular disease in patients with NAFLD

Intervention	Guidance for CVD	Guidance for NASH
Dietary modification	Emphasis on vegetables, fruits, legumes, nuts, whole grains, and fish Reduced intake of cholesterol and sodium Limit intake of processed meats, refined carbohydrates, and sweetened beverages Avoidance of trans fats Consider plant-based or Mediterranean diet	Calorie restriction Small studies suggest possible benefit of Mediterranean diet
Exercise	≥ 150 minutes moderate-intensity or ≥ 75 minutes vigorous-intensity exercise per week
Alcohol	Moderate alcohol use (≤ 2 drinks daily in men, ≤ 1 drink daily in women) lowers CV morbidity and mortality	Moderate use may worsen steatosis, NASH, and fibrosis with prospective study demonstrating no CV benefit
Aspirin	Proven benefit for secondary > primary CVD prevention	Prospective study suggesting decrease in fibrosis progression Limited data on bleeding risk in advanced liver disease
Statins	Treat based on lipoprotein profile, presence of T2DM, and calculated 10-year ASCVD risk Proven benefit in primary and secondary prevention	Safe to prescribe in NAFLD May improve steatosis, liver chemistries, and NASH
Metformin	Possible improvement in late CV outcomes	No direct benefit in NASH histology
Pioglitazone	Reduces risk of MACE in T2DM Avoid if history of HF	Improves NASH histology
GLP-1 agonists	Reduces CV risk in T2DM	Increases NASH resolution (possibly through weight loss)
SGLT-2 inhibitors	Reduced CV risk in T2DM	Improves hepatic fat content
Obeticholic acid	Increases LDL, decreases HDL	Improves fibrosis without worsening NASH

ASCVD: atherosclerotic cardiovascular disease; GLP-1: glucagon-like peptide 1; HDL: high-density lipoprotein; HF: heart failure; LDL: low-density lipoprotein; MACE: major adverse cardiovascular events; NAFLD: nonalcoholic fatty liver disease; NASH: non-alcoholic steatohepatitis; SGLT-2: sodium-glucose cotransporter 2; T2DM: type 2 diabetes mellitus.

Lifestyle interventions

Lifestyle modification, including dietary intervention and exercise, is universally recommended as first-line therapy in both NAFLD and CVD [3, 75, 76]. Specific ACC/AHA guidelines endorse Mediterranean or plant-based diets and recommend specific minimum durations of moderate or vigorous intensity aerobic exercise per week [76]. In NAFLD, the well-known benefits of weight loss were recently re-demonstrated by Vilar-Gomez et al., showing that loss of 5% or greater of body weight improved all histological features of NAFLD [80]. No specific diets have proven superior in NAFLD; however, various small-scale studies of the Mediterranean diet suggest a modest decrease in steatosis in a proportion of patients [81].

The role of alcohol in NAFLD is hotly debated. Prior retrospective studies indicated a possible protective benefit of moderate (non-heavy) alcohol use in NAFLD, which is of interest in particular because of the possible CV benefit of moderate alcohol use [82]. The AASLD currently does not make a recommendation on non-heavy alcohol consumption in NAFLD due to insufficient data [3]. However, recent studies suggest that even light alcohol consumption may be detrimental in NAFLD. In a cohort of over 58,000 Korean adults with NAFLD, light (up to 9.9g/day) or moderate (women: <20g/day, men: <30g/day) drinking worsened fibrosis scores [83]. In a Mendelian randomization study by Sookoian et al., carriers of a polymorphism in the alcohol dehydrogenase gene that is robustly associated with decreased alcohol consumption served as a control group with reliably low levels of long-term alcohol exposure [84]. Non-carrier “moderate” drinkers (mean alcohol use 8.2 +/− 21g/day), in comparison to carriers (mean 2.3 +/ 5.3g/day), demonstrated higher degrees of steatosis, lobular inflammation, and NAFLD activity scores on biopsy. More recently, in persons with longitudinally assessed biopsy-proven NASH from the NASH Clinical Research Network (NASH-CRN), modest alcohol use was actually associated with less improvement in markers of NASH severity and lower odds of NASH resolution [85]. Finally, among black and white middle-age adults with CT-defined NAFLD in the population-based Coronary Artery Risk Development in Young Adults (CARDIA) study, prospectively-assessed moderate alcohol use was not associated with established risk factors for CVD or subclinical markers of CVD [86]. In sum, we believe that the evidence to-date suggests a possible harm of moderate alcohol consumption in NAFLD and no apparent CV benefit in this subpopulation.

Aspirin

While the benefits of aspirin for primary prevention of CVD are questioned, its use is well-established for secondary CV prevention [76]. Significantly less data exists in patients with NAFLD. Prior studies suggest a lower prevalence of fibrosis (determined by fibrosis scores and transient elastography) in NAFLD patients on aspirin therapy, however the cross-sectional nature of these studies prevents any causal inferences [87, 88]. A recent prospective study conducted by Simon et al. followed 361 adults with biopsy-proven NAFLD over a median of 7.4 years. Daily aspirin use was associated with a significantly lower incidence of worsening fibrosis indices (aHR 0.63, 95% CI 0.43–0.85) [89].

Future research should aim to elucidate the degree of CV risk reduction that patients with NAFLD experience with aspirin. Further, in more advanced stages of liver disease, the possible benefit of aspirin due to more prevalent clinical CVD needs to be balanced with the concerns of higher bleeding risks.

Statins

ACC/AHA guidelines recommend that decisions for statin therapy be determined by lipoprotein profile, presence of T2DM, and 10-year ASCVD risk or presence of clinical CVD [90]. The liver disease community does not make specific recommendations other than to advise that NAFLD does not increase the risk of liver injury from statins and that practitioners should follow current lipid guidelines [3, 75]. However, the story is more complex.

First and foremost, despite the clear prevalence of CVD in NAFLD and substantial evidence to suggest against statin hepatotoxicity in these patients, statins are under-prescribed in patients with fatty liver disease [91, 92]. Furthermore, post-hoc analyses of multiple large randomized controlled trials (RCTs) of the effect of statins on CV morbidity and mortality demonstrate that in the subgroups of patients with abnormal liver chemistries and NAFLD, statin therapy also improves liver chemistries, steatosis, and the risk of CV events [93–95].

More recently, smaller studies using histology also suggest the possibility of liver-related benefit in NAFLD. Kargiotis et al. prospectively studied 20 individuals with biopsy-proven NASH and dyslipidemia treated with rosuvastatin (10mg/day) and showed complete resolution of NASH in 95% after 1 year of therapy [96]. One criticism of this paper was the lack of reporting of other confounding parameters such as weight changes. In another cross-sectional study of patients undergoing liver biopsy for suspected NASH, Dongionvanni et al. noted that the subgroup of patients on a statin at the time of biopsy were significantly less likely to show steatosis, NASH, or advanced fibrosis [97]. Acknowledging the limitations of these two studies, at a minimum they suggest a need for ongoing research into the potential benefits of statins in NAFLD.

Future investigation must also focus on whether indications for statin therapy in NAFLD should extend beyond current ACC/AHA guidelines. Is a greater degree of CV risk reduction achievable by prescribing statins to a wider cohort of patients with NAFLD? Should NAFLD be considered amongst the “risk-enhancing” factors described by the ACC/AHA to help guide treatment decisions in intermediate risk patients?

Until more data is available, clinicians should at least feel comfortable initiating statin therapy in those individuals with clear indications based on the lack of hepatotoxicity and the known CV benefits in the general population.

Diabetic medications

The agents used to treat T2DM in NAFLD can be viewed from two perspectives: those shown to affect CVD risk and those with potential efficacy in treating NAFLD.

Metformin is a well-established first-line therapy for newly-diagnosed patients with T2DM [98]. It also continues to be recommended by the ACC/AHA, citing previous studies showing improvement in late CV outcomes, however it has not been associated with improved liver histology in NAFLD [99, 3].

Pioglitazone improves liver histology in patients with NASH, regardless of the presence of T2DM [100]. As a result, the AASLD and EASL both endorse consideration of pioglitazone in persons with NASH and T2DM after a careful risk assessment given the known risk of weight gain and HF [3, 75]. In a recent meta-analysis, Liao et al. reviewed 9 studies of CV outcomes with pre-diabetes or T2DM, finding pioglitazone to be associated with a significantly lower risk of MACE but increased weight gain and HF [101].

There is increasing evidence that glucagon-like peptide 1 (GLP-1) agonists and sodium-glucose co-transporter 2 (SGLT2) inhibitors are effective antidiabetic agents in reducing CV risk [102–104]. The American Diabetes Association now recommends use of these agents independent of hemoglobin A1C levels in persons with high ASCVD risk or established CVD [105]. Evolving research also suggests a potential role in treating NASH. In 2016, Armstrong et al. published the Liraglutide Efficacy and Action in NASH (LEAN) study, a multicenter RCT of 52 patients with biopsy-proven NASH over 48 weeks. The GLP-1 agonist liraglutide was superior to placebo for NASH resolution (RR 4.3, 95% CI 1.0–17.7) with less progression of fibrosis (RR 0.2, 95% CI 0.1–1.0) [106]. Multiple RCTs comparing SGLT2 inhibitors to placebo in patients with T2DM and NAFLD show significant improvements in hepatic fat content as measured by controlled attenuation parameter (CAP) or magnetic resonance imaging proton density fat fraction (MRI-PDFF) [107–110]. While these have been relatively small (approximate sample sizes 50–80 patients) with short-term follow-up of 12–24 weeks, the data is nonetheless promising, and clinical trials with both GLP-1 agonists (clinicaltrials.gov ID: NCT03648554) and SGLT2 inhibitors are ongoing (clinicaltrials.gov ID: NCT03723252).

Obeticholic acid

Obeticholic acid (OCA) is a farnesoid X receptor (FXR) agonist that has gained recent press due to the expected upcoming Food and Drug Administration (FDA) approval for treatment of NASH. In a multicenter, double-blinded, phase III RCT by Younossi et al., compared to placebo, OCA significantly improved fibrosis without worsening NASH (OCA 10mg: RR 1.5, 95% CI 1.0–2.2; OCA 25mg: RR 1.9, 95% CI 1.4–2.8) [111]. However, a notable side effect of OCA is increased LDL and decreased HDL. Future studies will help to understand the overall impact of OCA on CVD incidence and progression in patients with NASH.

Conclusion

With the rising global burden of NAFLD, in which cardiovascular disease persists as the major contributor to morbidity and mortality, it is imperative that clinicians understand the metabolic risk factors and subsequent manifestations of CVD in this population. Lifestyle modification and weight loss remain cornerstones of treatment in NAFLD and NASH and in primary and secondary prevention of CVD. Significant work has been done thus far to elucidate underlying mechanisms of disease and to study management paradigms. Going forward, it will be necessary to develop and validate more precise methods for screening for CVD in NAFLD. Ultimately, improving overall morbidity and mortality in NAFLD will require that clinical trials of new medications be designed with cardiovascular endpoints in mind.