Prevention of Heart Failure, Tachyarrhythmias and Sudden Cardiac Death in HIV

Abstract

Purpose of the review:

To summarize the state-of-the-art literature on the epidemiology, disease progression and mediators of heart failure (HF), tachyarrhythmias, and sudden cardiac death in people living with HIV (PLWH) to inform prevention strategies.

Recent findings:

Recent studies corroborate the role of HIV as a risk enhancer for HF and arrhythmias, which persists despite adjustment for cardiovascular risk factors and unhealthy behaviors. Immune activation and inflammation contribute to the risk. HF occurs more frequently at younger ages, and among women and ethnic minorities living with HIV, highlighting disparities. Prospective outcome studies remain sparse in PLWH limiting prevention approaches. However, subclinical cardiac and electrophysiologic remodeling and dysfunction detected by noninvasive testing are powerful disease surrogates that inform our mechanistic understanding of HIV-associated cardiovascular disease and offer opportunities for early diagnosis.

Summary:

Aggressive control of HIV viremia and cardiac risk factors and abstinence from unhealthy behaviors remain treatment pillars to prevent HF and arrhythmic complications. The excess risk among PLWH warrants heightened vigilance for HF and arrhythmic symptomatology and earlier testing as subclinical abnormalities are common. Future research needs include identifying novel therapeutic targets to prevent HF and arrhythmias and testing of interventions in diverse groups of PLWH.

Introduction

Access to anti-retroviral therapy (ART) has transformed HIV epidemiology. People living with HIV (PLWH) enjoy extended longevity but are vulnerable to non-AIDS-related chronic morbidities such as cardiovascular disease (CVD) that deter healthy aging. Numerous factors potentially contribute to prevalent and incident CVD among PLWH. Even with durable HIV viral suppression, immune dysregulation and heightened inflammation persist as do viral reservoirs and viral replication which are cardiotoxic. ART is associated with potential side effects including dyslipidemia, insulin resistance, lipodystrophy and pro-arrhythmogenicity. The burden of traditional cardiovascular risk factors increases as PLWH age. Behavioral factors such as smoking, substance use and unhealthy diet and exercise patterns, which are more common among PLWH, further heighten risk. Finally, disparities in health care access and utilization may also play roles.

Here, we aim to summarize the state-of-the-art knowledge in epidemiology, disease progression and prognosis and potential mediators of non-atherosclerotic CVD in PLWH, focusing on heart failure (HF), atrial and ventricular arrhythmia, and sudden cardiac death (SCD) outcomes. Such an understanding can inform prevention strategies, management and subsequent research directions. Prevention of atherosclerotic CVD has been reviewed recently and is not covered here [1*,2*].

EPIDEMIOLOGY OF HEART FAILURE IN PLWH (Table 1)

Table 1:

Prevalence, incidence and risk factors associated with HF, atrial arrhythmias and ventricular arrhythmias/sudden cardiac death in PLWH

	HF	Atrial arrhythmias	Ventricular arrhythmias / sudden cardiac death
Prevalence	6.5–7.2%	2–3%	--
Incidence	0.23–0.9 per 100 person-years to 2.1% per year	18.2 per 1000 person-years	232 per 100,000 person-years to 3.5–13% per year
Age	Younger age of onset	Older age	Younger
Sex	Women	---	Men
Race/ethnicity	Blacks, Asian-Pacific Islanders	Caucasian race	Caucasian race
Traditional cardiac risk factors	Increased burden	Increased burden	Increased burden
HIV viral load ≥500 copies/ml	HFrEF only	--	Increased risk
Nadir CD4 <200 cells/ml	Increased risk	Increased risk	Increased risk
Protease inhibitors	Increased risk	--	--
Substance use	Cocaine increases risk	--	Cocaine increases risk Opiates increase risk

Early and durable ART has decreased the frequency of HIV-associated dilated cardiomyopathies characterized by severe left ventricular systolic dysfunction, especially in developed countries. However, numerous studies report an excess burden of HF, due to both reduced (HFrEF) and preserved ejection fraction (HFpEF) [3–12]. In an analysis of a national database of electronic health care records, a HF diagnosis occurred in 7.2% of PLWH and 4.4% of controls (relative risk 1.66 [1.60–1.72], p<0.0001) [3]. In the Veteran Aging Cohort Study (VACS), a multicenter longitudinal study, veterans living with HIV had a 61% increased risk of incident HFrEF and a 21% increased risk of incident HFpEF compared with uninfected veterans after multivariable adjustment [4]. When physician-adjudicated HF was used instead of administrative coding in another study, the excess incident risk associated with HIV was even higher (relative risk 2.10, 95% confidence interval, 1.38–3.21 versus uninfected controls) [5]. Both prevalent and incident HFrEF and HFpEF are associated with lower nadir CD4+ T-cell count (CD4 count) <200 cells/mm³ [3,5,6], while HIV viral load ≥500 copies/mL is associated only with HFrEF [4]. Another analysis of national health insurance data reported lower incident HF risk with increasing ART duration [9]. Incident and prevalent HF risks among PLWH are higher at younger ages [3,4,6,7*,13*], among women [3,7*] [6], blacks [14], and Asian-Pacific Islanders [7*], and independent of coronary artery disease [3,4,7*]. PLWH with HF have an increased burden of cardiovascular risk factors and substance use compared to controls [14].

Once diagnosed with HF, PLWH have worse outcomes including higher rates of all-cause mortality and recurrent HF and all-cause hospital admission [14–16]. Outcomes are particularly worse among PLWH with low CD4 count or unsuppressed HIV viral load [14,15]. However, among HFrEF patients, for those PLWH with undetectable viral load and CD4 count ≥200 cells/mm³, outcome is similar versus uninfected patients [15]. Cocaine use and coronary artery disease are associated with increased risk for 30 day HF readmission [15] Use of protease inhibitors, compared to other ART, is associated with increased 30 day HF readmission rate [17]. Recipients of protease inhibitors have higher rates of hyperlipidemia, diabetes mellitus, and coronary artery disease and lower left ventricular ejection fraction (LVEF), further exacerbating adverse outcome [17]. Among PLWH hospitalized with HF, persistently reduced LVEF with lack of LVEF recovery at follow-up is more common than seen in other chronic inflammatory diseases [18].

SUBCLINICAL DISEASE MEASURES AND RISK OF HF IN PLWH (Table 2)

Table 2:

Subclinical manifestations of increased risk for HF, atrial arrhythmias and ventricular arrhythmias/sudden cardiac death in PLWH

HF	Atrial arrhythmias	Ventricular arrhythmias / sudden cardiac death
Echocardiography: Low LVEF Diastolic dysfunction Increased LV mass Reduced LV GLS CMR: Reduced LV strain Focal LGE, nonischemic pattern Increased T1/ECV Increased intramyocardial lipid levels	Echocardiography: Increased left atrial size CMR: Increased left atrial size	Echocardiography: Low LVEF CMR Macroscopic scar Interstitial fibrosis Electrocardiography Prolonged QT interval Increased QT variability Decreased heart rate variability Increased QT variability index LV hypertrophy Increased ventricular ectopy and non-sustained VT (among men with lower CD4, higher viral load)

There are few prospective longitudinal studies of and no interventional studies of HF prevention in PLWH. However, much can be learned from imaging-based investigations of subclinical structural heart disease associated with HIV. Changes in cardiac structure and function precede symptomatic HF (pre-HF Stage B) in the general population [19]. Echocardiography and cardiovascular magnetic resonance imaging (CMR) identify imaging features associated with increased risk for symptomatic HF. Reduced LVEF is much less frequent than in the pre-ART era with only a 2.4% prevalence reported in a recent large echocardiography study of men living with HIV enrolled in the Multicenter AIDS Cohort Study (MACS) which was statistically similar to the 1.9% prevalence among uninfected comparators [20*]. In uninfected individuals with preserved LVEF, LV global longitudinal strain (GLS) reflects contractility, detects subclinical impairment in deformation and is incrementally prognostic for HF outcomes [21]. In a retrospective cohort of PLWH, abnormal GLS was associated with lower CD4 count [22]. A Ugandan study of PLWH with high prevalence of hypertension also reported reduced GLS compared to uninfected controls [23]. CMR studies similarly note reduced LV GLS and circumferential strain among PLWH versus controls [24–27].

A number of studies have identified abnormal imaging metrics associated with diastolic impairment among PLWH [10,20,23,28,29]. These include increased LV mass/hypertrophy and diastolic dysfunction using various criteria including the 2009 [30] and 2016 American Society of Echocardiography guidelines [31] and those proposed by the Characterization of Heart Function on Antiretroviral Therapy (CHART) study [28]. Prevalent diastolic dysfunction ranges from 22–37% among cohorts of middle-aged PLWH [10,20,29]. However, among some studies of PLWH in African nations, diastolic dysfunction is much less prevalent (1–7%) [32,33] and similar by HIV serostatus [32,33], perhaps due to younger participants, few traditional cardiac risk factors and yet unidentified differences in environmental factors. Nonetheless, LV mass was higher among these younger African PLWH compared to uninfected controls which may predispose to future diastolic dysfunction [32,33]. Risk factors for LV hypertrophy include CD4 count ≤200 cells/mm³, lack of viral suppression [29], and later initiation of ART [33] while lower CD4 count [29] and more severe HIV disease [33] correlate with diastolic dysfunction.

Biomarker studies provide additional insight into potential mechanisms of cardiac dysfunction in HIV. An earlier study correlated levels of suppression of tumorgenicity-2 (ST2) with diastolic dysfunction [34]. More recently, diastolic dysfunction was associated with higher levels of N-terminal pro-B-type natriuretic peptide (NT-pro-BNP), troponin-I, and carboxyl-terminal telopeptide of collagen type I with borderline correlation with interleukin-6 [28] suggesting roles for cardiac stress, injury, collagen turnover and inflammation. Among a group of HIV seropositive and seronegative Ugandans, NT-proBNP, growth differentiation factor-15 (GDF-15), ST2, and cystatin C were associated with LV mass while NT-proBNP and GDF-15 were associated with diastolic dysfunction [35*]. The relationship between the serum biomarkers and LV mass was stronger for men than women, highlighting sex as well as potential geographic and/or racial/ethnic differences [35*].

CMR also provides insight into HIV-related myocardial involvement by detecting focal, macroscopic scar via late gadolinium enhancement (LGE), diffuse fibrosis via T1 mapping techniques, and myocardial steatosis by spectroscopy. Several studies have reported increased prevalences of focal LGE among PLWH [24–26,36] often in a nonischemic pattern involving the midwall and subepicardium [24,25,37*]. PLWH have increased extracellular volume fraction (ECV) or native T1 times reflecting diffuse fibrosis compared to uninfected referents [25–27,37*,38**]. Such patterns are adverse prognostic markers in HIV seronegative cohorts [39]. Elevated native T1 times among PLWH were recently associated with adverse cardiovascular outcomes including HF [38**]. Studies have also reported higher intramyocardial lipid levels in PLWH [25,36,40] which, in combination with higher amounts of fibrosis, are associated with strain abnormalities [36]. While dysfunction of immune and metabolic fat pathways likely exacerbate cardiac stress and injury and increased collagen turnover, our understanding of the specific pathways by which HIV increases burden of myocardial scar, fibrosis, and fat remains incomplete. A recent report correlating levels of transmethylamine-N-oxide (TMAO), a dietary gut metabolite, with ECV among PLWH suggests a possible role for gut permeability and microbial translocation contributing to immune activation that warrants further investigation [41**].

Few studies have examined the time course of imaging characteristics among PLWH and are restricted to echocardiography. A retrospective analysis of clinically-acquired echocardiograms from PLWH with LVEF>40% observed higher LV mass and lower LV end-diastolic volume over 5 years independent of CD4 count or HIV viral suppression and no change in LV systolic volume, LV GLS or E/e’ ratio [42*]. A smaller cohort study of women with and at risk for HIV enrolled in the Women’s Interagency HIV Study (WIHS) performed repeat echocardiograms 12 years following baseline [43*]. Incident LV dysfunction occurred in 13.8%, most of which (79.2%) comprised diastolic dysfunction and neither rate of new systolic nor diastolic dysfunction differed by HIV serostatus [43*]. However, coinfection with hepatitis C correlated with a three-fold increased risk for incident LV dysfunction in women living with HIV [43*]. Larger validation studies are needed.

EPIDEMIOLOGY OF ATRIAL ARRHYTHMIAS IN PLWH (Table 1)

The literature is mixed regarding the association between HIV and atrial arrhythmias [44–47] due to differences in rhythm validation (hospital diagnosis coding versus physician adjudication) and cohort selection. Prevalent atrial arrhythmia is estimated at 2–3% [45,46] with an incidence rate for PLWH of 18.2 (versus 8.9 per 1,000 person-years for controls) [47]. Most studies report associations between older age, white race, diabetes, hypertension, coronary artery disease, HF, and/or chronic obstructive pulmonary disease and incident atrial arrhythmias among PLWH, common to those in the general population [45,48,49]. Additionally, nadir CD4 <200 cells/mm³ [45], lack of ART or the prescription of multiple ART regimens [48], perhaps a marker of treatment failure and/or duration of HIV, also correlate with increased risk for incident arrhythmia.

SUBCLINICAL DISEASE MEASURES AND RISK OF ATRIAL ARRHYTHMIAS IN PLWH (Table 2)

Abnormal atrial structure and function predispose to subsequent atrial fibrillation in people without HIV [50]. While subclinical changes in atrial size and function are described among PLWH [20*,37*], the progression to clinical arrhythmias and differences in trajectory or risk predictors from that of uninfected individuals are unknown.

EPIDEMIOLOGY OF VENTRICULAR ARRHYTHMIAS AND SUDDEN CARDIAC DEATH IN PLWH (Table 1)

Published studies suggest an increased risk for SCD among PLWH, estimated at 232 per 100,000 person-years to 3.5–13% per year, depending on the definition of SCD used and the cohort – e.g. public HIV clinic [51], longitudinal cohort study [52**], administrative hospital database [53*] versus hospitalized HF patients [54]. Low CD4 counts and higher HIV viral loads correlate with increased SCD risk [52,54]. Rates of SCD among PLWH with CD4 count ≥200 cell/mm³ and undetectable viral load are similar to those of controls [54]. Other risk factors for SCD in PLWH generally mirror those of uninfected referents, including coronary artery disease, low LVEF, worse functional class, as well as cocaine use, lower prescription of beta blockers, and longer QRS and QT interval durations [52,54]. In a study of PLWH hospitalized for HF with low rates of viral suppression (51%) and clinically-indicated implantable cardioverter defibrillators, appropriate and inappropriate discharges occurred more frequently compared to uninfected comparators and were associated with greater risk for HF admission and cardiovascular mortality [55]. Among PLWH, cocaine use and coronary artery disease as well as longer QRS duration and worse functional class were associated with increased odds of a device discharge [55].

A recent study compared results of autopsy and toxicology and histology testing in out-of-hospital arrest victims with and without HIV [56**]. Occult drug overdose was the leading cause of sudden death among PLWH, exceeding that in referents (34% vs 13%). Arrhythmic SCD comprised 47% of all sudden deaths with a trend toward higher incidence rate among PLWH versus controls (25.0 versus 13.3 arrhythmic deaths per 100,000 person-years; incidence rate ratio 1.87, 95% confidence interval 0.93–3.78). PLWH with arrhythmic SCD were predominantly men (94%), Caucasian (77%), and ~9 years younger than HIV seronegative cardiac arrest victims. The prevalences of cardiovascular risk factors and disease, including myocardial infarction and HF, were similar by HIV serostatus. The burden of total myocardial fibrosis, and specifically interstitial (but not replacement) fibrosis, was higher in PLWH. These findings highlight the lower rates of arrhythmic SCD when using an autopsy-based definition; the frequency of occult drug overdose as a cause of presumed cardiac arrest among PLWH; and the association between myocardial fibrosis burden and ventricular arrhythmic outcomes not unique to HIV. The study also confirms the increased prevalence of non-ischemic scar among PLWH reported in the CMR studies discussed in the HF section above.

SUBCLINICAL DISEASE MEASURES AND RISK OF VENTRICULAR ARRHYTHMIAS AND SUDDEN CARDIAC DEATH IN PLWH (Table 2)

Abnormal myocardial structure and function, including reduced LVEF and myocardial scar, not only correlate with increased risk for HF, as reviewed above, but also predispose to ventricular arrhythmias and SCD [57]. Additional factors that increase arrhythmic risk include electrophysiologic and autonomic dysfunction. PLWH may be susceptible to prolonged ventricular repolarization, as measured by the electrocardiographic QT interval, due to direct HIV viral effects on cardiac ion channels [58], side effects of ART [59], drug-drug interactions between ART and other medications known to increase QT interval duration [60], and substances such as opioids [61]. Prevalent QT prolongation ranges from 3–23% [62–66*]. Several studies report longer absolute corrected QT and QT subcomponent intervals among PLWH versus controls, both in developed and low income countries [64,67–69*]. Risk factors for prolonged QT in PLWH include male sex [66*], Asian race [66*], high HIV viral load [63,65,66*], lower CD4 count [63,65], ART use [69*], and increased inflammatory biomarker levels [64,68]. The data correlating specific ART classes and QT prolongation are conflicting but support increased risk with older generation protease inhibitors, efavirenz, and rilpivirine [65,70]. Few studies have examined longitudinal changes in QT intervals. One recent study reported a 1.6% incidence of new QT prolongation within 2 years among virally suppressed PLWH [71*]. Relatively high prevalences of early repolarization abnormalities (12.4%) and electrocardiographic LV hypertrophy (8.3%) were also recently reported in the multinational Randomized Trial to Prevent Vascular Events in HIV (REPRIEVE) Study [66*]. While studies in the general population suggest the association between electrocardiographic early repolarization [72] and LV hypertrophy with SCD [73], long term clinical follow-up is required to determine clinical significance in PLWH.

Complementary to static measurements, QT variability (QTV) quantifies spontaneous, temporal fluctuation in the QT interval, reflects electrical instability and correlates with increased risk for SCD in HIV seronegative cohorts [74]. Heart rate variability (HRV), a measure of autonomic function, is also an informative marker of arrhythmic risk in the general population. HIV may result in sympathetic dominance that can potentially alter both HRV and repolarization dynamics, predisposing to ventricular arrhythmias [75,76]. This hypothesis was tested in a recent study of MACS participants which used an ambulatory electrocardiography patch monitor to examine differences in QTV index (QTVI) by HIV serostatus and is defined as the ratio between QTV and HRV [77]. A higher QTVI (prognostically worse) was seen among men living with HIV compared to uninfected referents, even after adjusting for traditional cardiovascular risk factors. The association between higher QTVI and HIV seropositivity was stronger among men with HIV and detectable viral loads. In examining the QTV and HRV subcomponents of QTVI, HIV seropositivity was associated with lower (prognostically worse) HRV and higher (prognostically worse) QTV, supporting HIV-associated autonomic dysfunction and repolarization abnormalities as mechanisms for proarrhythmia. Higher inflammatory biomarker levels correlated with higher QTVI and partly attenuated the HIV association, supporting a role for inflammation-mediated changes in electrophysiology and autonomic function.

Other subclinical markers of ventricular arrhythmic risk studied in PLWH include ventricular ectopy (VE) and non-sustained ventricular tachycardia (NSVT), using administrative coding [78] and ambulatory electrocardiographic patch monitoring [79]. Burden of VE/NSVT appears similar by HIV serostatus [78,79] though among men with HIV, higher viral load [78], lower CD4 count [78], and higher QTVI correlate with more frequent VE/NSVT [77].

The Figure shows a conceptual diagram of the risk factors, injury mechanisms and interplay between adverse cardiac and electrophysiologic remodeling that contribute to ventricular arrhythmias/SCD.

Figure 1.

Conceptual diagram of mechanisms of ventricular arrhythmia/SCD in PLWH. The risk factors, injury mechanisms and interplay between adverse cardiac and electrophysiologic remodeling that contribute to ventricular arrhythmias/SCD are shown in this Figure (created with Biorender.com).

Conclusions

The recent literature corroborates HIV infection as a risk enhancer for HF and arrhythmias. While PLWH often have an increased burden of traditional cardiovascular risk factors and unhealthy behaviors, increased risk for these outcomes persists despite multivariable adjustment, reinforcing other HIV-specific mechanisms. HF endpoints in particular occur at younger ages, and with higher frequency among women and ethnic minorities living with HIV, highlighting disparities. The evidence supports a role for immune activation as a mediator, though exact pathways remain unclear, limiting intervention options. The associated metabolic sequelae of ART regimens also contribute but management of the individual patient is restricted to knowledge of general class effects.

There remains a lack of HIV-specific strategies and guidelines directing prevention, diagnostic and treatment approaches stemming from limitations of current outcomes studies. Few include independently adjudicated endpoints or individual HF, tachyarrhythmic and SCD events. Current estimates of the excess risk attributable to HIV suffer from imperfect comparison groups. Many observational studies include HIV seronegative referents matched only on a few demographic parameters but otherwise lack the risk factors known to be enriched in PLWH. Few include large numbers of women and diverse races/ethnicities. However, subclinical cardiac and electrophysiologic remodeling and dysfunction detected by noninvasive testing remain powerful surrogates of disease that inform our mechanistic understanding of HIV-related CVD, help identify potential preventive and therapeutic targets and offer opportunities for early diagnosis.

Durable and aggressive control of HIV viremia and cardiac risk factors as well as abstinence from unhealthy substances and behaviors remain primary treatment goals to prevent cardiovascular complications. The enhanced risk among PLWH warrants heightened vigilance for potential HF and arrhythmic symptomatology and merits earlier testing and diagnosis. Future research needs include the identification of therapeutic targets specifically for prevention of HF and arrhythmias and testing of interventions in diverse groups of PLWH.

Key points.

• Enrichment for traditional cardiac risk factors as well as unhealthy behaviors and substance use contributes to the excess risk associated with HIV for HF and arrhythmic outcomes, but does not fully explain the relationships.

• Higher HIV viral load, lower CD4 count, and younger age are associated with more frequent HF and ventricular arrhythmia endpoints among PLWH while women and ethnic minorities with HIV are at higher risk for HF endpoints.

• Among PLWH, the subclinical phenotype associated with increased risk for HF includes preserved LV ejection fraction but reduced global longitudinal strain, increased LV mass, larger left atrial size, diastolic dysfunction, and larger amounts of myocardial interstitial fibrosis.

• Electrocardiographic abnormalities are common among PLWH and prolonged QT duration, lower heart rate variability, higher QT variability and QT variability index support increased risk for ventricular arrhythmias and sudden cardiac death.

• Inflammation and immune activation mediate subclinical changes in cardiac structural and functional and electrophysiologic properties in PLWH compared to HIV seronegative comparators.