Myocardial ischemia and left ventricular diastolic dysfunction in HIV infected patients and asymptomatic for coronary artery disease

Highly active antiretroviral therapy (HAART) allows chronicity of AIDS evolution, leading to association of other pathologies such as coronary artery disease (CAD). Myocardial ischemia (MI) and left ventricular diastolic dysfunction (LVDD) evaluation in HIV-infected patients may favor primary prevention of CAD. The study aimed to evaluate frequencies of MI and LVDD in the population living with the human immunodeficiency virus (PLHIV) and asymptomatic for CAD. We analyzed data from 110 HIV-infected patients who underwent clinical and laboratory evaluation, treadmill exercise stress test, and transthoracic echocardiogram, and compared it with 2,619 healthy individuals from the control group (non-HIV and non-CAD), selected from the database. HIV-infected patients presented lower average age (51.5 ± 7.7), systemic arterial hypertension (28.0%) and dyslipidemia frequencies (32.0%). On the other hand, their MI frequency was twice as high (14.7%); and diastolic dysfunction (DD) percentage was higher in ischemic patients (45.5%). In the HIV-infected group, MI frequency was 10.0%, while that of DD was 18.2%. MI was twice as frequent among HIV infected patients compared to uninfected, despite lower frequency of risk factors for CAD. Non-ischemic patients living with HIV had a frequency of DD more than twice compared to the control individuals.


Introduction
In 2020, the population living with the human immunodeficiency virus (PLHIV) was estimated at 37.7 million people worldwide, 73% of PLHIV received antiretroviral therapy (ART), and approximately 1,5000,000 were newly infected in 2020 (WHO, 2021). From 1997 onwards, the advent of highly effective or highly active antiretroviral therapy (HAART) brought a new perspective to the course of HIV infection as it provided control of viral load (VL) and consequent increase in life expectancy of these patients, transforming it into a chronic medical condition.
In parallel to longer longevity, other non-HIV related diseases start to increase, especially cardiovascular disease (CVD) (Thienemann, Sliwa & Rockstroh, 2013). Studies have shown a higher incidence of cardiovascular abnormalities in patients with HIV, including coronary artery disease (CAD), which is more prevalent and extensive in these individuals (Mangili et al., 2006;Post et al., 2014;Subramanian et al., 2012). As the incidence of opportunistic infections declines, prevalence of non-HIV/AIDS-related comorbidities, including CVD, continues to rise among HIV-infected patients compared to uninfected (Toribio et al., 2017).
Pathogenesis of this disorder encompasses complicated interactions between effects of chronic HIV infection, antiretroviral use, and patient's own factors, including genetic susceptibility.
Since the HIV-infected population has reached older ages, there has been an inherent increase in cardiovascular risk, over-added by effects of infection and its therapy. Knowledge and early management of this condition, including studying asymptomatic ones, are imperative.
The study aimed to evaluate frequencies of myocardial ischemia (MI) and left ventricular diastolic dysfunction (LVDD) in the PLHIV and asymptomatic for CAD. It also aimed to evaluate MI occurrence in this sample, besides to evaluate left ventricular (LV) systolic and diastolic functions in PLHIV on HAART and to compare the groups of HIV-infected and non-HIV-infected patients for the frequency of MI and diastolic dysfunction (DD).

Study design and population
An observational, cross-sectional, analytical study with a quantitative approach was conducted with prospectively collected data (Pereira et al., 2018). 110 PLHIV were selected consecutively and non-randomly, with non-established CVD, from the Infectious Disease Outpatient Clinic of the University Hospital of Federal University of Sergipe (HU-UFS) and the Outpatient Clinic Center of Aracaju (CEMAR).

Inclusion criteria
Patients were included independent of gender, with HIV infection diagnosed, 10 years old or older (due to the possibility of vertical transmission of HIV), asymptomatic for cardiovascular diseases, and who signed the Informed Consent Form (ICF). Patients under 10 years old, with established coronary artery disease, as well as those with clinical signs of opportunistic infections, with recent hospitalization, debilitated patients, and pregnant women were excluded.

Ethics
This study is approved by the Research and Ethics Committee from Federal University of Sergipe following the Brazil regulation for research with human subjects and the Declaration of Helsinki. Reference number: 2.244.171 Research, Society andDevelopment, v. 10, n. 11, e301101119756, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i11.19756 4

Statistical analysis
To estimate sample size, prevalence of atherosclerosis in HIV-infected patients, as documented in literature, was 48 to 63%. The 95% confidence interval and an acceptable difference of about 5% were also considered. This estimate used Findings in this sample were compared to meet one of specific objectives with a database of 2,619 non-HIV-infected individuals without prior CAD and asymptomatic from cardiovascular symptoms admitted to the Cardiology methods service of a Hospital at Aracaju-Sergipe, for elective transthoracic echocardiography (TTE), exercise test (ET) and laboratory tests.
Despite aforementioned losses, analyzes of different aspects were possible (sub analysis) and those variables based on TTE and ET exams were chosen, since these determine the investigative power of main outcome variables, LVDD and MI, respectively.

Clinical characteristics of HIV-infected patients
The sample of HIV-infected patients consisted of 110 individuals with a mean age of 45.3 ± 11.7 years old, minimum age of 19 years old and maximum of 73 years old. There was a higher frequency of males with 69.1%. Among traditional risk factors for CAD, lower frequencies of diabetes mellitus (DM) (6.4%), obesity (14.5%), family history of early CAD (FHECAD) (14.5%), Systemic Arterial Hypertension (SAH) (21.8%) and dyslipidemia (DLP) (27.3%), and higher frequencies of smoking (37.3%) and sedentary lifestyle (59.1%) were found. Half of patients are infected with HIV greater than 7.0 years, 25% greater than 11.3 years, and 25% had less than 3.0 years. Treatment time was over 6.0 years in half of sample, 25% of patients reached values over 9.5 years, and 25% less than 2.0 years (Table 1).
However, female sex, hypertension and sedentary lifestyle, whose odds ratios did not reach p less than 0.05, presented odds ratios and respective p values fulfilling criteria for entry into model to be adjusted (p-value less than or equal to 0.30 and odds ratio greater than or equal to 1.30). Likewise, it fulfilled the above criteria and included LVDD's occurrence. In adjusted models, after the first one, independent variables were excluded at a time, generating different logistic regression models, to establish for which association with MI was suggested. It was observed only age variable presented a significant relation with MI, with an odds ratio of 1.1 (Table 4).

HIV-infected patients with and without left ventricular diastolic dysfunction (LVDD)
Patients with HIV and DD had an average age (54.9 ± 9.5 years) higher than non-affected by this condition (43.2 ± 11.1 years) with p <0.0001, with difference between means of 11.7 ± 2.7 95% (CI 6.4 to 17.0 years); and greater participation of hypertensive (45%), diabetic (20%), obese (30%) and dyslipidemic (45%), but without significance for the last two cardiovascular risk factors (Table 5). Considering DD as dependent variable and age, male gender, SAH, DM, DLP, obesity, FHECAD and occurrence of MI as independent variables, association between independent variables and presence of DD in HIV-infected patients was analyzed by logistic regression, initially not adjusted (Table 6). These independent variables were listed either by fulfilling statistical criteria or by the researcher's choice. Unadjusted odds ratio showed age is associated with the presence of DD, as well as hypertension, DM, obesity and MI. On the other hand, DLP, whose odds ratio did not reach p less than 0.05, presented an odds ratio and its p-value fulfilling criteria for entry into the model to be adjusted (p-value less than or equal to 0.30 and odds ratio greater than or equal to 1.30). In adjusted models, after the first one, independent variables were excluded at a time, generating different logistic regression models, to establish an association with DD was suggested. There was a significant relationship of DD only with variables age and obesity, which presented odds ratios of 1.12 and 5.00, respectively (Table 7).

Baseline clinical characteristics of people living with HIV
Laboratory data are shown in table 9. HIV infection is suppressed in most of patients evaluated. This was expressed by undetectable VL (< 40copy/ml) and lymphocytes CD4 + count above 200 cells/mm³. Furthermore, regarding metabolic profile of PLHIV group, it is evident significant majority of this sample shows a reduction of blood glucose, total cholesterol, LDL-cholesterol and triglyceride levels (by 60.0%), and an increase in HDL-cholesterol level (in 73.2%).

Baseline clinical characteristics of non-HIV-infected patients
Sample of seronegative HIV patients consisted of 2,619 asymptomatic individuals with an average age of 58.8 ± 9.8 years, a minimum of 41 and a maximum of 91 years. There was a discreet higher frequency of females with 51.1%. Among the traditional risk factors for CAD, we found higher frequencies of hypertension (50.3%) and of DLP (48.3%) in practically half of the individuals; and lower frequency for smoking (4.4%) (Table 10). Age expressed as mean and standard deviation; risk factors expressed as number of patients and percentage in parentheses. Source: Authors.
In another words, we observed that the frequencies of MI and DD are equally high in the group of patients with HIV, even at a lower average age and a lower frequency of risk factors for CAD, when compared to the other uninfected individuals.
While logistic regression analysis, adjusted for factors associated with MI, shows a strong association only with age. In addition, regarding the dependent variable DD, it is suggested that age and obesity are the only associated factors, with caveats the sample size.

Discussion
The present study reveals HIV-infected patients had a lower average age (51.5 ± 7.7 vs 58.8 ± 9.8 years; p = 0.003) and lower frequencies of classic risk factors for CAD, especially hypertension (28.0% vs 50.3%; p <0.0001) and DLP (32% vs 48.3; p = 0.007). However, higher frequency was found for MI (14.7% vs 7.0%; p = 0.021) in relation to non-HIV-infected individuals. These findings are present even if HIV infection is under reasonable control (CD4 + count above 200 cells/mm³ and undetectable VL in most patients, besides use of HAART regimen). There are significant differences between the PLHIV group and the non-HIV-infected group. It suggests there is something more in HIV patients, which raises potential risk for coronary events, in contrast to the non-HIV group. However, it was observed a favorable adjusted odds ratio only for age.
Literature really points to the fact such behavior is due to HIV infection and its treatment -ART use (Koenig, 2017;Lang et al., 2015;Vilela et al., 2011)-, with inflammation and immunological activation widely proven as HIV infection part, contributing to CAD emergence (Boettiger et al., 2020;Freiberg et al., 2013;Vachiat et al., 2017). In fact, a large cohort found a strong association of HIV-positive individuals along increased risk of MI, regardless of traditional risk factors (Katoto et al., 2018;Silverberg et al., 2014).
Higher prevalence of smoking, DM, SAH and DLP in HIV patients is known to contribute to inflammation, as are coinfections that, in addition to this immune activation leading to inflammation, account for coagulation disorders (Freiberg et al., 2013). In this study, a high frequency of smoking (39.7% vs 4.4%; p <0.0001) among individuals living with HIV was found. Challenge for current medicine is to promote cardiovascular risk prevention in these patients, especially to reinforce the need for smoking cessation, taking into account genesis linked to multiple factors (Boettiger et al., 2020;Lang et al., 2015).
Evidence of satisfactory VL control (without treatment interruption), in parallel with elevation in blood glucose, total cholesterol, LDL-cholesterol and triglyceride levels, besides HDL-cholesterol level reduce, confirm the premise that HAART controls HIV infection at the expense of a worsening metabolic profile. However, the opposite was found in this study, with low blood glucose, total cholesterol, LDL-cholesterol and triglyceride levels, and high HDL-cholesterol level. In this respect, a question arises -studies such as SMART (Strategies for Management of Antiretroviral Therapy) show a large increase (in the range of 70%) in the risk of CVD in ART interruption, suggesting the need for continuous treatment to prevent HIV-associated inflammation and to reduce cardiovascular risk (Chihana et al., 2012;Longo-Mbenza et al., 1998). This fact also explains the reduced frequency of LV Systolic Dysfunction (only 2.7%) in our study for the group of patients with HIV, compared to prevalence of 35% in pre-HAART era (SMART, 2006).
Literature establishes HIV infection as a chronic condition in areas of wide HAART coverage (Baldassarre et al., 2007), allowing its association with an increase in non-immunodeficiency complications, notably CAD. High prevalence of traditional risk factors increases overall risk for CVD. Some studies estimate around 1.5 to 2.0 times general risk for this population (Triant et al., 2007;WHO, 2018).
Among HIV-infected patients, those with DD had a higher average age (54.9 ± 9.5 vs 43.2 ± 11.1 years; p <0.0001) and higher frequencies of risk factors for CAD, notably SAH (45%), DLP (45%), DM (20%) and obesity (30%) compared to those without DD, but with adjusted odds ratios favorable only for age and obesity. This fact suggests a probable participation also of metabolic syndrome in DD genesis, and not only MI process (Tavares et al., 2012), which in our study was present in a quarter of patients with DD.
Meanwhile, the subgroup of patients with HIV and without MI had a significant frequency of DD (15.2%) compared to low prevalence (1 to 7%) for DD in the general population (Almeida et al., 2018;Nagueh et al., 2016).
Long period of infection and HAART use seen in this population of HIV-infected patients reveals extensive exposure not only for the virus, but also to HAART regimens, which has been implicated in a more aggressive profile for development of atherosclerosis. Ninety-seven percent of patients in this study were using HAART, with 45.5% using Protease Inhibitors (PI), the class that most interfered with patients' metabolic profile (Gilbert, Fitch & Grinspoon, 2015). Prevalence of subclinical atherosclerosis in contrast to seronegative people, were not uniform, apart from there's an enormous absence of information about developing countries' situation. Previous studies imply HAART can cause DLP, SAH, endothelial dysfunction, particularly PI drugs (Ingle et al., 2014).
Regions of the globe with greatest poverty are still most affected, demanding a confrontation permeates production of scientific information (Senoner et al., 2019;Sinha & Feinstein, 2019). The evaluation of HIV infection in patients from northeast region of Brazil is essential to carry out public policy actions with elaboration, for example, of clinical protocols providing advanced preventive medical care for this particular population.

Study limitations
Patients' inclusion and, consequently, the sample size were limited due to spontaneous demand in outpatient clinicsboth due to availability and impossibility caused by health status. Socioeconomic condition was an impediment to displacement of many of these patients. Most of them are from Sergipe countryside; and even though they were Aracaju metropolitan area's residents, many patients were economically vulnerable. The fear of disclosing his identity was also a barrier to recruitment. In addition, it was seen a great resistance in successive stages, from questionnaires to exams. In order to contour these limitations, a satisfactory reception of patients was offer, including explanation about the ICF, confidentiality maintenance and benefits offered to them, especially to CVD prevention. Another limitation was several confounders' presence (such as DLP, which has ART itself as one of possible causes), which it will bring difficulties in defining atherosclerosis genesis in the PLHIV. Anyway, such limitations have been pointed out in other studies in same research area, complex interaction of factors in aforementioned genesis being a common argument, which still needs further clarification.

Conclusion
Myocardial ischemia occurred in HIV-infected patients with acceptable infection control and lower frequencies of the main risk factors for CAD, approximately twice as high as in non-HIV-infected individuals. Non-ischemic patients with HIV had a significant frequency of LVDD (more than twice) compared to healthy control individuals.
Moreover, additional and prospective studies that can follow PLHIV for a longer time are suggested to longitudinally verify the frequency of cardiovascular disorders, in addition to studies that can refine the presence of atherosclerosis in HIVinfected patients, without the presence of confounding events.