Human Immunodeficiency Virus and Heart Failure in Low- and Middle-Income Countries

Abstract

Successful combination therapy for human immunodeficiency virus (HIV) has transformed this disease from a short-lived infection with high mortality to a chronic disease associated with increasing life expectancy. This is true for high- as well as low- and middle-income countries. As a result of this increased life expectancy, people living with HIV are now at risk of developing other chronic diseases associated with aging. Heart failure has been common among people living with HIV in the eras of pre- and post- availability of antiretroviral therapy; however, our current understanding of the pathogenesis and approaches to management have not been systematically addressed. HIV may cause heart failure through direct (e.g., viral replication, mitochondrial dysfunction, cardiac autoimmunity, autonomic dysfunction) and indirect (e.g., opportunistic infections, antiretroviral therapy, alcohol abuse, micronutrient deficiency, tobacco use) pathways. In low- and middle-income countries, 2 large observational studies have recently reported clinical characteristics and outcomes in these patients. HIV-associated heart failure remains a common cardiac diagnosis in people living with heart failure, yet a unifying set of diagnostic criteria is lacking. Treatment patterns for heart failure fall short of society guidelines. Although there may be promise in cardiac glycosides for treating heart failure in people living with HIV, clinical studies are needed to validate in vitro findings. Owing to the burden of HIV in low- and middle-income countries and the concurrent rise of traditional cardiovascular risk factors, strategic and concerted efforts in this area are likely to impact the care of people living with HIV around the globe.

People living with human immunodeficiency virus (PLHIV) around the globe and taking antiretroviral therapy (ART) now achieve a near-normal life expectancy (1,2). As a consequence, PLHIV increasingly experience the chronic diseases of aging. Of particular concern is the risk of coronary artery disease, myocardial infarction, and heart failure (HF) among PLHIV, as observed in North American, European, and Australian HIV cohorts (3–9). Sub-Saharan Africa (SSA), which accounts for 12% of the global population, is disproportionately affected by HIV, with 69% of all adults and 90% of all children living with HIV residing here (10). Cardiovascular manifestations of HIV infection in SSA have been reported for more than 25 years, but most of the reports are prior to 2004 when ART became widely available (11).

Prior to the advent of ART, studies supported a relationship between HIV and left ventricular (LV) dysfunction. Numerous terms were used to describe the syndrome, including HIV-associated cardiomyopathy, HIV-associated HF, and HIV-associated LV dysfunction. As initially described, HIV-associated HF was related to severe immune system compromise, had no specific therapy, and had a median survival of 101 days after diagnosis (12–15). Concurrent shifts in epidemiological patterns of HIV treatment and cardiovascular diseases produced a complex and evolving relationship between HF and HIV that merits a contemporary review. Most reports related to HF are focused on high-income countries (16) and are misaligned with the global predominance of HIV in low- and middle-income countries (LMICs) (Figure 1) (11,17–21). In this review, we summarize the epidemiology, approach to diagnosis, therapy, and prognosis of HIV-associated HF in LMICs in the ART era. Our review of the publications was accomplished by searching PubMed, EMBASE, and the Cochrane Central Register of Controlled Trials for studies published through 2014. We used the following terms as keywords and subject headings: human immunodeficiency virus, heart failure, cardiomyopathy, echocardiography, antiretroviral therapy, low- and middle-income countries, developing countries, and each of the countries in the World Bank low- and middle-income groups. We included studies that report HF characteristic of HIV-infected individuals in LMICs receiving ART. Studies were excluded if the abstract was not available in English. Data extraction (e.g., year of study, location, study description, and findings) was performed and confirmed by 2 investigators (F.A., G.S.B.). In the present discussion we use the terms “HIV-associated HF” or “HF in PLHIV” to describe the syndrome of HF, including diastolic dysfunction, and use the term “HIV-associated LV systolic dysfunction” (LVSD) only when discussing asymptomatic ventricular dysfunction or to describe heart muscle disease. Owing to the heterogeneity of definitions of HF in HIV, we defined HF according to the criteria proposed in each study. This represents the state of published reports. We ultimately propose a framework to invigorate research that will inform best-care practices for patients with HIV-associated HF in LMICs and the rest of the world.

Figure 1. Worldwide Distribution of Studies of HF and Burden of HIV.

Country-level data for (A) the number of individuals included in clinical studies and registries for HF, and (B) the number of adults living with HIV Data from the World Health Organization Global Health Observatory Data Repository; U.S. Central Intelligence Agenct World Factbook, and the Centers for Disaese Control, Taiwan (ROC). (14,22-25,99-101). HF = heart failure; HIV = human immunodeficiency virus.

Causes of HIV-associated HF

There is no unifying definition of HIV-associated HF, and it is often a diagnosis of exclusion (22-25). This is due to the knowledge gaps regarding the causes of HF in PLHIV; however, there are a number of prevailing hypotheses (Figure 2) (21,26-29). Reports from high-, low-, and middle-income countries have shed light on the probable causes.

Figure 2. Pathways of HIV-Associated HF.

HIV may cause HF through direct and indirect pathways. Direct pathways involve HIV viral replication, an autoimmune milieu, and autonomic dysfunction. Viral replication also affects mitochondrial function, which is associated with cardiac myocyte dysfunction. Indirect pathways leading to HF include factors that are common among PLHIV such as opportunistic infections, antiretroviral therapy, alcohol (EtOH) abuse, micronutrient deficiency, and tobacco use. HF = heart failure; HIV = human immunodeficiency virus; PLHIV = people living with HIV.

HIV Viral Replication

One hypothesis implicates direct myocardial viral toxicity. HIV ribonucleic acid (RNA) concentrations ≥500 copies/ml are associated with a nearly 2.5-fold increased risk of developing HF compared with HIV-uninfected individuals (8). Targeted myocardial transgenic expression of HIV transactivator of transcription (Tat) protein activates endothelial cells, causing LV systolic dysfunction (LVSD), increased LV mass, and expression of natriuretic peptides in mice that may lead to hemodynamic compromise (30). How and whether HIV directly damages cardiac myocytes, which do not possess CD4 receptors, have been hotly debated (31-33). In vitro studies of human and rat cardiac myocytes show that HIV can enter cardiac myocytes through pathways independent of CCR5 or CXCR4 receptors (32,34).

Mitochondrial Injury

HIV may impair cardiac function through mitochondrial pathways. HIV infection initiates a mitochondrion-mediated cascade, releasing proteases that lead to cardiac myocyte damage and apoptosis (35). Tat protein also disrupts mitochondrial membrane permeability (36). The complete pathway between HIV infection and LVSD remains unclear. To date, a clear genetic predisposition does not appear to play a major role, as 1 specific mitochondrial deoxyribonucleic acid (DNA) polymorphism, mtDNA T16189C, is not more common in patients with HIV and HF than in HIV-infected controls without HF in South Africa (37).

Opportunistic Infections

In SSA, there have been 2 studies examining myocardial biopsy specimens among patients with HIV-associated LVSD. In a 1998 postmortem study of 16 patients with HIV/acquired immune deficiency syndrome (AIDS), myocarditis was attributed to HIV (50%) and infection by Toxoplasma gondii (19%), Cryptococcus neoformans (19%), and Mycobacterium avium intracellulare (13%) (38). An antemortem study of patients compared analyses of endomyocardial biopsy specimens among HIV-associated HF patients (n = 14), HIV-uninfected patients with idiopathic dilated cardiomyopathy (n = 8), and heart transplant recipients in South Africa (n = 11) (39). In contrast to the earlier study (38), infection with multiple cardiotropic viruses (average 2.5 viruses per individual) was common among those with HIV-associated HF. The most common viruses were Epstein-Barr virus (64%), herpes simplex virus (50%), and parvovirus B19 (14%). Whether viral suppression and restoration of immune function alter the role played by opportunistic infections in HIV-associated HF in LMICs is unknown.

Tobacco and Alcohol Use

Unhealthy behaviors have been implicated in the development of HF among PLHIV, specifically, heavy alcohol consumption and tobacco use (40). A systematic review and meta-analysis investigating LVSD in paucisymptomatic HIV-infected patients in the ART era found that active tobacco smoking was associated with a 1.57-fold higher odds of having HF and was second only to age as the univariate risk factor for HF (41). Only 1 of the included studies, however, was from an LMIC (42). A study from Rwanda has shown that the relationship among alcohol intake, cigarette smoking, and dilated cardiomyopathy in PLHIV is attenuated after adjusting for HIV stage, socioeconomic status, duration of HIV infection, and plasma selenium levels (43).

Micronutrient Deficiency

Of the micronutrient deficiencies, selenium has been the most extensively studied, related to HIV-associated HF. Selenium's antioxidant properties protect against endothelial dysfunction, and its deficiency has been associated with cardiac dysfunction. Due to soil composition and agricultural practices in SSA, 28% of the population is at risk for selenium deficiency (44). Selenium deficiency is common in PLHIV with and without LVSD (43,45). Selenium supplementation improves cardiac function in PLHIV in small studies (46,47), but the debate as to whether HIV-associated LVSD responds to selenium supplementation is unresolved. Nonetheless, there is evidence of association between lower body mass index and LVSD in PLHIV, suggesting that overall nutrition may be a factor (48).

Antiretroviral Therapy Toxicity

Although the introduction of ART in high-income countries has been associated with a reduction in the incidence of HIV-associated HF of 30% to 50% (49,50), ART has also been implicated in causing LVSD (51). Zidovudine (AZT), a nucleoside reverse transcriptase inhibitor, has been associated with reversible, dose-dependent skeletal and cardiac myopathy attributed to mitochondrial damage (52,53). Studies pre- and post-widespread availability of combination ART link AZT to cardiomyopathy (33,54). There is also contemporary evidence linking AZT to diastolic dysfunction (55). These findings are especially important for LMICs, where AZT is still a first-line option for HIV treatment for adults, adolescents, and children.

Cardiac Autoimmunity

Circulating cardiac autoantibodies are detected more frequently in PLHIV with LVSD (43%) than in HIV-infected patients without LVSD (19%) or in HIV-uninfected controls (3%) (56). Autoantibody concentrations in PLHIV correlate with mortality, but biopsy-proven autoimmune myocarditis in PLHIV with LVSD responds well to steroid therapy (56,57). Together with findings of increased myocardial expression of human leukocyte antigen (HLA) class I antigen in PLHIV with HF (58), these observations support a role for cardiac autoimmunity in the pathogenesis of HIV-associated LVSD.

In the ART era, the causes of HIV-associated LVSD appear to depend on the degree of viral suppression. When viral suppression is adequate and immune function returns to normal, other cardiovascular risk factors or ART may be the main causative factor. On the other hand, when viral suppression has not been achieved and at worse stages of HIV, it is more likely that HIV, ART, opportunistic pathogens, autoimmunity, or nutritional deficiencies result in LVSD. As more data are generated from PLHIV worldwide, the causes and mediators of HIV-associated LVSD will be better understood.

Epidemiology of HIV-associated HF in LMICs in the pre-ART era

Studies of HIV-associated HF in the pre-ART era in LMICs were heterogeneous, with wide variation in patient population, patient selection, and definitions of HF outcomes, yielding a wide range in prevalence estimates. Observational studies from SSA indicated an HF prevalence of approximately 50% and an incidence of any cardiac abnormality of 55% over a 7-year period (12,13,38). LVSD was most common among symptomatic patients, whereas asymptomatic PLHIV more often had HF with preserved ejection fraction (EF) (59). HIV-associated HF occurred most often among young persons with CD4⁺ T-cell counts of <100 cells/mm³, lower socioeconomic status, longer duration of HIV infection, higher viral load, and advanced HIV stage (14,43). In-hospital mortality rates due to HIV-associated HF reached 15% (13).

Epidemiology of HIV-associated HF in LMICs in the ART era

Studies in the ART era have shown that HF is still a relevant form of HIV-associated cardiovascular disease (Table 1)(60-70). There have been 2 large prospective observational studies from SSA in the ART era that shed light on the contemporary epidemiology of HIV-associated HF: the HoS (Heart of Soweto) and THESUS-HF (the Sub-Saharan Africa Survey of Heart Failure) studies (7,22). In addition, a number of other studies also inform the clinical characteristics and outcomes of patients with HIV-associated HF in LMICs in the ART era.

Table 1. Studies of HIV and HF in LMICs in the ART Era.

Study, Year (Ref #)	Study Site (s)	Study Type	Sample Size (n)	HIV-Positive/Negative (n)	Key Findings
Jain et al. 2014 (60)	India	Cross-sectional descriptive study	100	100/0	Diastolic dysfunction (43%) and dilated cardiomyopathy (18%) were the most common abnormalities observed in HIV-infected patients.
Shaboodien et al., 2013 (39)	South Africa	Case-comparison study	33	14/19	Myocarditis was present in 44% of HIV-associated cardiomyopathy cases, 36% of heart transplant recipients, and 25% of participants with idiopathic dilated cardiomyopathy. Myocarditis was acute in 50% of HIV- and heart transplant-associated myocarditis and was chronic in all with idiopathic dilated cardiomyopathy.
Sliwa et al., 2013 (61)	Multiple in sub-Saharan Africa	International observational registry	1,006	65/500	Among patients with HF, HIV status predicts 180-day mortality.
Chillo et al., 2012 (62)	Tanzania	Cross-sectional descriptive study	102	102/0	10% of HIV-positive patients with cardiomyopathy (systolic dysfunction), 34% with hypertensive heart disease.
Damasceno et al., 2012 (22)	Multiple in sub-Saharan Africa	International observational registry	1,006	65/435	HIV-associated cardiomyopathy was rarely seen (2.6% [26/1000 cases]).
Ige et al., 2012 (63)	Nigeria	Cross-sectional descriptive study	300	150/150	LVSD is significantly more frequent in HIV-infected children and worsens with increasing age.
Okoromah et al., 2012 (64)	Nigeria	Cross-sectional observational study			In HIV-infected children, the prevalence rates of both dilated cardiomyopathy and LVSD were 33.7%.
Olusegun-Joseph et al., 2012 (65)	Nigeria	Cross-sectional descriptive study	150	100/50	Echocardiographic abnormalities were significantly more common in the cases than the controls (78% vs. 16%, respectively; p < 0.001), including systolic dysfunction (30% vs. 8%, respectively; p < 0.05) and diastolic dysfunction (32% vs. 8%, respectively; p < 0.01).
Pepeta et al., 2012 (66)	South Africa	Retrospective	34	34/0	There was improvement in LV function in patients with HIVAC treated with antiretroviral therapy.
Schwartz et al., 2012 (67)	Botswana	Cross-sectional descriptive study	179	106/73	HIV infection was strongly associated with pericarditis and cardiomyopathy; 18% of HIV+ patients had hypertensive heart disease. Prevalence of HIV among patients with cardiomegaly was higher than in the general population
Sliwa, et al., 2012 (7)	South Africa	Prospective clinical registry	53,28	518/4,810	HIV-related cardiac disease was a minor contributor to overall disease burden (4% prevalence). Cardiomyopathy was the most common HIV-related cardiac disease.
Tantchou Tchoumi et al., 2011 (68)	Cameroon	Prospective observational	462	7/455	7 of 462 patients (1.6%) of cases had HIVAC.
El Hattaoui et al., 2008 (42)	Morocco	Cross-sectional descriptive study	238	158/80	Cardiomyopathy was more common in HIV-infected patients with AIDS than in those with non-AIDS or HIV-uninfected patients. Cardiomyopathy was more common among HIV-infected patients with CD4 count below 100.
Twagirumukiza, et al., 2007 (43)	Rwanda	Prospective observational	416	416/0	Predictors of dilated cardiomyopathy were low socioeconomic status, duration of HIV infection, CD4 count, HIV viral load, advanced stage of HIV, and low plasma level of selenium.
Lubega et al., 2005 (69)	Uganda	Cross-sectional descriptive study	230	230/0	Heart abnormalities were common in children with symptomatic HIV disease and included sinus tachycardia (21%), LVSD (17%), and right ventricular dilation (14%).
Diogenes et al., 2003 (70)	Brazil	Case report	1	1/0	Perinatally-infected 1-year-old child with severe cardiomyopathy had full reversal of disease after 6 years of treatment and follow-up, with normal LV dimension and contractility.

AIDS = acquired immune deficiency syndrome; ART = antiretroviral therapy; HF = heart failure; HIVAC= ■ ■ ■; LMICs = low- and middle-income countries; LVSD = left ventricular systolic dysfunction.

Epidemiology and Clinical Characteristics

The HoS study is the largest and most comprehensive registry of de novo presentations of any cardiac disease in SSA (71). Of 5,328 cardiac presentations between 2006 and 2008, there were 518 PLHIV (7). Most PLHIV (54%) were taking ART at the time of enrollment, were younger, and had lower blood pressure, higher heart rate, and lower body mass index than the rest of the HoS cohort. Among PLHIV, HIV-associated HF was the most common diagnosis (38%). The average LVEF of those with HIV-associated HF was 46%.

There were more women than men with HIV-associated HF in the HoS study. Overall, the women in HoS were younger (∼6 years younger than men), and men were more likely to smoke tobacco. The viral load was significantly higher (110,000 vs. 19,000 RNA copies/ml), and CD4⁺ T-cell counts were significantly lower (180 vs. 211 cells/mm³) in patients with HIV-associated HF than in PLHIV without HF. Valve dysfunction, right HF, and hypertension were common comorbidities, and no PLHIV with HF were reported to have coronary artery disease.

The THESUS-HF study was a multinational prospective registry of patients with acute decompensated HF (ADHF) admitted to university hospitals in 9 African countries between 2007 and 2010 (22). Endemic causes of HF (i.e., HIV, rheumatic heart disease, pericarditis, and others) were compared with emerging causes such as hypertension and ischemic heart disease. Of 1,006 patients enrolled, 50% underwent testing for HIV, and 65 were found to be infected with HIV. HIV was the cause of HF in 2.6% of all cases. Most of these patients were from Mozambique, Cameroon, and Uganda, countries that also have high rates of HIV. Patients with endemic causes of HF (including HIV) were younger by 10 to 15 years and were less often smokers, hypertensive, or diabetic. According to echocardiographic results, LV systolic and diastolic dimensions were larger in patients with endemic causes; however, there were no significant differences in LVEF.

Most, but not all (72,73), observational studies from LMICs in the ART era are generally consistent with the above findings, showing prevalence of HIV-associated HF between 1% and 5% (68,74,75). PLHIV usually have HF diagnosed in the third decade of life and are more often women. Both systolic and diastolic dysfunctions are more common (∼30% prevalence) in PLHIV than in controls without HIV (65), and rarely, diastolic dysfunction is more prevalent (42,60). Ventricular dimensions and mass are large in PLHIV with HF in SSA, similar to findings from North America (65,76). Up to one-third of African HIV-infected children have LVSD (63,64). Cardiac dysfunction in HIV-exposed infants is identifiable shortly after birth in high-income countries (77).

Antiretroviral Therapy

Pediatric studies suggest a relationship between type of ART and LVSD. In Brazil, HIV-infected children treated with AZT were more likely to develop HF than children treated with combination ART (78). Conversely, combination ART, but not AZT monotherapy, improves LV function in children with HIV-associated HF (70). In a retrospective study from South Africa, 18 HIV-infected children with HF receiving ART were compared with 16 children with HIV-associated HF cared for before national rollout of ART (66). After 38 months of follow-up, the children treated with ART had improved LV fractional shortening compared with the nontreated controls. Four of the ART-treated children had worsening cardiac function that improved upon discontinuation of AZT, stavudine, or didanosine.

Outcomes

In SSA, patients with endemic causes of HF (including HIV) have slightly longer lengths of hospital stay for ADHF than those with emerging causes (9.8 vs. 8.7 days, respectively) (22). In-hospital mortality (28% vs. 14%, respectively), 60-day mortality (13% vs. 9.0%, respectively), and 180-day mortality (21% vs. 16%, respectively) are higher in the endemic causes group. HIV status is a significant predictor of death at 180 days for patients with ADHF (61).

Treatment

There are no specific international guidelines for the treatment of HIV-associated HF (24,25). There are no controlled trials from LMICs addressing the most effective treatment strategy. As such, data from the broader international studies guides management. Restoration of immune status and viral load suppression may ameliorate cardiac dysfunction as long as medications causing cardiac dysfunction are avoided (49,50). Clinicians should be aware of the potential drug-drug interactions (e.g., with calcium channel blockers, HMG-CoA reductase inhibitors, antiarrhythmic drugs, and ART) as well as cardiac side effects of certain ART. Online tools detailing significant drug interactions for ART and HF medication are available (79).

There have been no trials to assess whether guideline-directed medical therapy for HF is also suitable for PLHIV. Nonetheless, use of angiotensin-converting enzyme inhibitors (26%), β-adrenergic blockers (29%), and spironolactone (35%) is low (7,22). The use of digoxin in HIV-associated HF may have particular relevance because of its inhibitory effects on HIV viral replication as demonstrated in vitro (80). Trials are needed to further elucidate whether this translates into clinical benefit in humans with HIV-associated HF.

Diastolic Dysfunction in HIV+ Patients

Among PLHIV with HF, LVSD is currently less common in high-income countries. Diastolic dysfunction, however, is present in up to 64% of asymptomatic PLHIV taking ART in high-income countries (81,82) and appears to be independent of traditional risk factors including age and hypertension (76). Early stage diastolic dysfunction is often the only echocardiographic abnormality found in asymptomatic PLHIV (83). Asymptomatic PLHIV have higher LV mass than HIV-uninfected subjects, and higher LV mass is inversely proportional to nadir CD4⁺ T-cell count (76). The mechanisms of diastolic dysfunction in PLHIV are as yet unknown but may involve direct myocardial effects of HIV (84). Data from LMICs in this field are sorely needed.

Impact of HIV Infection on Cardiac Deformation

Early detection of subclinical myocardial dysfunction, assessed by 2-dimensional strain and strain rate using speckle tracking, may maximize treatment benefits by identifying individuals who would benefit from preventive strategies (85). Asymptomatic, otherwise healthy PLHIV have lower strain and strain rates than HIV-uninfected people (86). HIV infection appears to affect mainly longitudinal systolic cardiac strain with preserved circumferential myocardial deformation (Figure 3) (87). Magnetic resonance imaging shows that these subclinical changes may relate to myocardial fibrosis and steatosis (84). Identification of early markers of myocardial dysfunction may identify PLHIV at high risk for cardiac dysfunction and thus allow early initiation of life-saving therapy. Such data are not yet available from LMICs.

Figure 3. Myocardial Deformation Tracings in HIV-Infected Individuals With Normal LVEF.

(A) Normal global longitudinal strain (GLS) pattern in asymptomatic PLHIV (average GLS = −19%). (B) Abnormal GLS pattern in asymptomatic PLHIV (average GLS = −6%). (Illustrations courtesy of Duke University Medical Center Cardiac Diagnostic Unit.) HIV = human immunodeficiency virus; LVEF = left ventricular ejection fraction; PLHIV = people living with HIV.

Other Risk Factors for HIV-associated HF in LMICs

Data are mixed regarding the prevalence of traditional cardiovascular risk factors in PLHIV in LMICs. A meta-analysis examining the association between HIV and cardiovascular risk factors in SSA showed that the burden of most factors was generally low (88). PLHIV had lower body mass indexes (standard mean differences [SMD]: 0.3 kg/m²), lower high-density lipoprotein (HDL; SMD: −0.6 mg/dl) and higher triglyceride levels (SMD 0.3 mg/dl) than their HIV-uninfected counterparts. ART use was associated with higher HDL (SMD 0.4 mg/dl), low-density lipoprotein (LDL; SMD: 0.4 mg/dl), and lower hemoglobin A_1C levels (SMD: −0.3%). Smoking, however, is common in HIV-infected adults (22%) and adolescents (13% to 17%) in SSA (89).

High blood pressure is 1 of the most common causes of HF in SSA (22,23,73,75). Among PLHIV in SSA, the overall rate of hypertension is 8% to 19% (90). HIV infection is also associated with ambulatory non-dipping status, which suggests an underlying dysregulation of the cardiovascular system (91). Higher blood pressure increases mortality in HIV-infected men with well-controlled HIV disease nearly 3-fold than in those with more advanced disease (92).

Tuberculosis infection is a common antecedent among PLHIV with HF. Two-thirds of patients with moderate-to-large tuberculous pericardial effusions in Africa have HIV infection (93). Patients tend to be young and have low CD4⁺ T-cell counts, and development of tuberculous pericarditis is usually regarded as progression to end-stage HIV disease (94).

In contrast to high-income countries, where acute coronary syndromes are more common in PLHIV than in those infected with HIV (3,4), coronary artery disease and ischemic cardiomyopathy are rarely reported in PLHIV in LMICs (7). When coronary artery disease is identified in PLHIV, patients tend to be younger and are more commonly smokers, and coronary thrombosis predominates over coronary atherosclerosis (95).

Research Directions for PLHIV with HF

There are many opportunities for discovery in the field of HIV-associated HF in LMICs. From an epidemiological perspective, it remains unclear whether the prevalence of HF is higher among PLHIV than with HIV-uninfected patients and whether there are true sex-based differences. Most data for HIV-associated HF in LMICs emanate from a small number of countries in SSA with high HIV burden. However, there may also be important geographic differences among countries in Africa and Latin America and the Caribbean, East Asia, and the Pacific, Eastern Europe, Central Asia, South Asia, and the Middle East that have yet to be described.

Establishing diagnostic criteria for HIV-associated HF should be a high priority. Diagnostic criteria should consider the degree of immunosuppression and stage of HIV, because causes of HIV-associated HF likely vary depending on HIV clinical stage.

There is also a need to clarify the appropriate therapy for HIV-associated LVSD. It is unclear whether guideline-directed medical therapies for HF, as described by society guidelines, have similar efficacy in ART-treated PLHIV and/or whether specific therapies (e.g., digoxin) would provide additional benefit for HIV-associated HF. The tachycardia and low blood pressure that are common in HIV-associated HF suggest a potential role for medications that affect sinoatrial node rate without lowering blood pressure (96).

The best approach to integrating care for HF in HIV treatment programs needs to be systematically evaluated. Many HIV treatment programs in LMICs have benefitted from strengthening of the overall health care system. These investments should be leveraged and expanded to address other comorbid chronic diseases.

It is unknown whether identifying LV dysfunction in an asymptomatic PLHIV warrants further action. Evaluating for contractile reserve with stress echocardiography, diastolic dysfunction, and abnormal myocardial deformation may identify higher-risk patients (97,98), but it is unknown whether a strategy incorporating such techniques alters clinical decision making for patients with HIV-associated HF. Last, the long-term impact of HIV on cardiac structure and function needs to be described in the ART era in order to guide care for children, as the natural history of cardiac function in perinatally HIV-infected children is unknown.

Conclusions

In the ART era, HF remains an important contributor to the cardiovascular disease burden in PLHIV in LMICs. Although the causative factors related to HIV-associated HF have shifted over time, HIV-infected patients in LMICs are at risk of developing HF even as access to effective ART expands. Outcomes for these patients are grim, and HIV infection increases in-hospital and post-discharge HF mortality. Ischemic HF is reportedly rare but may become more prevalent as cardiovascular disease epidemiology shifts. A number of causative factors appears to be operating simultaneously. Both adults and children are affected, yet early institution of ART may ameliorate the prognosis. Further insights into the disease may be gained by leveraging efforts from the numerous cardiovascular disease registries and HIV clinical care programs in LMICs. Finally, novel imaging techniques and parameters of cardiac function may reveal early manifestations of cardiac disease in PLHIV that could pave the way for early screening, institution of therapy, and prevention. The magnitude of HIV in LMICs should prompt the scientific community and funders of research to fill the knowledge gaps and improve our ability to manage and prevent HF in PLHIV.

Central Illustration. Current Knowledge, Knowledge Gaps, and Future Research.

Directions for HIV-associated heart failure in low- and middle-income countries. HIV = human immunodeficiency virus.

Abbreviations and Acronyms

ADHF

acute decompensated heart failure

AIDS

acquired immunodeficiency syndrome

ART

antiretroviral therapy

AZT

zidovudine

DNA

deoxyribonucleic acid

EF

ejection fraction

HF

heart failure

LMICs

low- and middle-income countries

LVSD

left ventricular systolic dysfunction

PLHIV

people living with human immunodeficiency virus

SSA

sub-Saharan Africa

Tat

transactivator of transcription