What happens to cardiovascular system behind the undetectable level of HIV viremia?

A large body of evidences in scientific literature supports the epidemiological observation of an increase of morbidity and mortality related to clinical and sub-clinical cardiovascular diseases (CVD) in HIV patients in the era of cART. A significantly higher risk of acute myocardial infarction (AMI) was observed in HIV infected patients compared with uninfected in veterans aging cohort study virtual cohort (VACS-VC) that analyzed data collected in 82,459 veterans followed for an average of 5.9 years; unexpectedly the highest risk was found in patients in cART and with HIV-RNA less than 500 copies/ml [8]. Also Silverberg et al. [9] reported an higher risk of AMI among HIV+ patients with a low recent or nadir CD4 cells (<200) compared with HIV-subject in a cohort study population included 22,081 HIV positive and 230,069 HIV negative subjects.

In contrast Klein et al. [10] reported a decline of CVD in HIV patients as observed by the analysis of the data of Kaiser Permanente Southern California and Northern California health plans including 24,768 HIV positive patients and 257,600 controls. By these data the adjusted relative risk of AMI in HIV+ patients declined from 1.8 in 1996–1999 to 1.0 in 2010–2011 possibly due to the better control of the traditional CVD risk factors, the lower incidence of severe immunodeficiency, and the use of more lipid-friendly cART in the patients enrolled in this study.

Although the data reported by Klein et al. probably influenced by the specific characteristics of their cohort, epidemiological studies show a significant overall impact of HIV on CVD. Since the size of the problem (although decreased after the introduction of lipid free antiretroviral drugs) remains very important, the actual aim of HIV medical practitioners is to better understand the kind of cardiovascular (CV) prevention to be applied to HIV population.

Recent retrospective and prospective studies evaluating the epidemiology of CVD in HIV positive patients were reported in Table 1.

Many factors may contribute to the increase of incidence of CVD in HIV infected patients: the traditional determinants of disease (cigarette smoking, diabetes mellitus, dyslipidemia, obesity, systemic hypertension, sedentary lifestyle, stress and family history of coronary artery or vascular disease) appear to be more common than in HIV negative population [11‐13]. Moreover alterations of regulation of glucose homeostasis and lipid metabolism and metabolic syndrome are frequently related with cART and are higher in patients previous treated with older antiretroviral drugs [14].

Anyway the excess risk of CVD observed in course of HIV infection cannot be explained only by the factors mentioned above. Several studies have also observed that co-morbidities, like CVD, which are normally witnessed later on in life as a result of natural aging, were increasingly prominent among the HIV-infected population [15, 16]. These observations led to the hypothesis that the cART treated HIV-infected patients experienced a premature and accelerated aging probably driven by side effects of antiretroviral drugs, chronic inflammation and persistent immune activation (Fig. 1).

Undoubtedly a pivotal role in the development of accelerated aging of cardiovascular system and in the pathogenesis of CVD is attributed to the status of immune activation. In fact in HIV patients the levels of proinflammatory cytokines and biomarkers associated to endothelial dysfunction are always higher and they can accelerated the atherosclerosis process. For example monocyte activation caused by microbial translocation drives the releases of soluble CD14 and CD163: the first is associated to increased death and the second with the risk of coronary artery progression and atherosclerosis [17‐20]. High levels of these inflammatory markers and of factors of hypercoagulation (i.e. D-dimers, fibrinogen) are linked to systemic clotting and chronic inflammatory damage of vascular endothelium [21]. Also HIV itself is considered responsible of persistent immune activation and endothelial dysfunction in fact it can penetrate into endothelial cells by receptor of CD4-T cells or chemokine receptors pathway. Moreover some HIV viral proteins, such as Gp120 and tumor necrosis-alpha (TNF-α), induce important consequences on the vascular tone, on the adhesion and aggregation of the platelet by the decrease of nitric oxide (NO) levels in the endothelial cell [22‐24]. The residual viremia may be the source of harm linked directly to the presence of the virus in course of effective cART, but may also be linked to the maintenance of the process of chronic inflammation and persistent immune activation.

Also Cytomegalovirus (CMV) infection appears to have a role in the premature aging of CV system: in a large Italian cohort CMV/HIV co-infection was associated with the risk of non-AIDS events including CV events independently of other prognostic factors. The potential role of CMV infection in CV disorders could depend by the complex interplay between viral and immunological activation, culminating in cyclical growth, damage and repair of endothelial cells. In fact from a pathogenic point of view, CMV can promote abnormal growth of endothelial cells and pathogenesis of atherosclerosis enhanced by proangiogenic factors including IL-6 and granulocyte macrophage colony stimulating factor [25, 26].

For this reason, the research’s interest is to understand the hidden inflammatory causes of CVD in HIV patients and the first objective is to control the levels of persistent immune activation. In fact, to date, immunological therapeutic strategies and virological interventions are investigated with the aim of controlling the levels of immune activation and then reduce the impact of CVD promoted by chronic inflammation. Table 2 shows the studies, registered on https://​www.​clinicaltrials.​gov and classified as “open”, and provides highlights for the current and future research topics in this area.

As previous described, an increase of morbidity and mortality related to clinical and sub-clinical CVD in HIV patients are reported. An important paradigm in the comprehension of this increased incidence is to understand the role of traditional risk factors. Evaluating the HIV cohorts, numerous data suggest that the prevalence of “traditional” risk factors is higher in HIV subjects if compared with general population. Triant et al. [27] observed greater incidence of hypertension (HIV: 21.2 vs. non-HIV: 15.9 %; p < 0.001), diabetes (HIV: 11.5 vs. non-HIV: 6.6 %; p < 0.0001), and dyslipidemia (HIV: 23.3 vs. non-HIV: 17.6 %; p < 0.0001) in HIV positive patients than in the controls. In particular, the serum triglyceride levels are higher in HIV population respect to controls [28]. Moreover, triglycerides work in a synergic way with both infection and cART: in fact, triglycerides levels increase when signs and symptoms of AIDS appear but also their serum levels are raised by protease inhibitors therapy [29].

However the high prevalence of “traditional” risk factors do not seem to be sufficient to explain the increased cardiovascular risk observed in the HIV positive population and the cardiovascular risk assessment based on conventional risk prediction models (i.e. Framingham system, PROCAM, SCORE) does not offer a good predictive value for the HIV-positive population [30]. In fact cardiovascular risk scores were built for general population, and were not validated in the HIV positive population. Probably the factors that most affect the failure in estimating the cardiovascular risk in HIV-positive patients are HIV infection itself and the premature aging of CV system driven by chronic inflammation and immune activation: in this sense the virus may be considered “the missing factor” essential for a correct evaluation of the overall risk [31].

Given the inadequacy of the risk estimation algorithms available, recently new models of risk prediction, specific for HIV-positive patients, were designed: for example Friis-Moller et al. [31] proposed the Data Collection on Adverse Effects of Anti-HIV Drugs (D:A:D) study equation in which age, sex, systolic blood pressure, serum cholesterol total and HDL-cholesterol level, diabetes, smoking status, family history of CVD, current use of abacavir, indinavir, or lopinavir; and the number of years on indinavir or lopinavir were the variables included. This cardiac risk model was created based upon data from the large cohort of HIV-infected patients who were followed longitudinally for cardiac events and performed better than the Framingham risk score among patients in that cohort. Moreover increased C-reactive protein (CRP) levels, uncontrolled HIV viral load at time of CV event and slower immunologic response were found to be associated with increased CVD risk in D:A:D: cohort [32]. Anyway a recent study of Thompson-Paul et al. [33] evaluating the D:A:D score and comparing with traditional algorithm for CV risk evaluation in HIV negative patients, highlighted that also the use of this algorithm was associated with the underestimation of CV risk. The authors concluded underlying that to better estimate CV risk in HIV-infected persons, additional risk factors, such as immunologic or virologic status may need to be considered.

The failure of traditional methods of CV risk assessment has made it necessary to evaluate new methods to better understand the risk of CVD in HIV population.

Among the inflammatory biomarkers of CV risk, several have been proposed for diagnostic use, including IL-6, D-dimer, MMP-9, and high-sensitivity CRP (hsCRP). In HIV seronegative patients, hsCRP has emerged as the most useful because it has been proven to add risk to the other factors described within the Framingham cohort. Anyway for the time being, the meaning of hsCRP cannot be relied upon in the same manner in patients with HIV infection, in fact the results of studies in this setting are conflicting [34, 35].

In the field of diagnostic imaging, carotid intima-media thickness (cIMT), brachial ankle pulse wave velocity and coronary artery calcium (CAC) score are innovative indicators for assessment of subclinical atherosclerosis and could give information about the development of CVD in patient without history of atherosclerotic diseases. These cardiovascular disease risk surrogate markers, are currently relegated to a secondary role, anyway their results should be considered complementary, and not alternative, to the information provided by CVD risk scores. In fact CVD risk assessment algorithms were not developed to predict coronary or carotid atherosclerosis, but rather cardiovascular events depending on the algorithm. For example baseline levels and change in levels of subclinical atherosclerosis, assessed by cIMT, could provide additional informations to determine how intensively to intervene on lifestyle and whether medication in correspondence with individual’s cardiovascular risk factors is indicated [36‐40].

Therefore in the last years the search for an optimal algorithm for estimating cardiovascular risk has been partly replaced by the search for exams increasingly able to estimate directly the presence of subclinical atherosclerotic lesions. The prevention is therefore increasingly moving towards the ability to diagnose the presence of silent atherosclerotic lesions able to progress and to determine a significant symptoms in asymptomatic subjects. On the basis of the above considerations the diagnostic applications of molecular imaging techniques becomes always more important in the detection of atherosclerotic plaques. The imaging by 18-fluoredeoxyglucose positron emission tomography (PET) and single photon emission computed tomography (SPECT) are considered important to identify activated macrophages infiltrating the arterial wall. Moreover both PET and SPECT provide information about the risk of plaque rupture and furthermore evaluate the vulnerability of itself. Recently Knudsen et al. conducted a prospective cross sectional study in which 56 HIV positive patients and 25 controls were scanned using 82Rb PET to obtain the stress flow (MFR). The results obtained were stratified in: low <1.5, borderline >1.5–2 or normal >2 and the Author did not observe difference between HIV positive and negative controls [41]. The data about the use of PET and SPECT in HIV population are not enough to conclude the useful of that in the real life. On the other hand the utility of imaging techniques in the detection of subclinical atherosclerotic plaques was reported in a cross-sectional study using dual-source CT (MDCT) coronary angiography conducted in HIV-infected subjects with low CV risk: MDCT showed an unexpected, age-associated high rate of significant coronary stenosis requiring coronary angiography in 29.1 % of asymptomatic HIV positive subjects enrolled. On the basis of these findings authors suggested that MDCT may be appropriate for CVD screening programs in HIV positive population [42].

Detection of subclinical atherosclerosis by noninvasive tests such as CAC score, cIMT, MDCT, and PET may improve risk prediction above that of established risk scoring models. Moreover interesting results are expected by promising innovative techniques, such as 3D quantification of carotid plaque by ultrasound (as in the Bioimage study), which are still waiting to be properly tested in HIV-positive population.

However, despite the interesting results obtained, more studies are need to investigate if diagnostic imaging exams can be used to guide patient screening, management and therapy in HIV setting.

As previously reported, HIV-infected patients receiving combination antiretroviral therapy may experience metabolic complications (i.e. dyslipidemia, impaired glucose metabolism and abnormal body fat distribution), potentially increasing their risk of cardiovascular diseases. In fact HIV infection is associated with a marked rise in the frequency of coronary heart disease and with accelerated coronary atherosclerosis and vasculopathy.

The progression from subclinical atherosclerosis to symptomatic disease is estimated to be higher and earlier in HIV patients than in general population. In fact HIV patients have a high burden of subclinical atherosclerosis including an increased amount of non-calcified coronary plaque compared with HIV-uninfected individuals independently by cardiovascular risk.

On these basis, recently, an increasing number of HIV patients undergo invasive cardiovascular procedures as a result of acute cardiovascular events or severe atherosclerotic disease. Although percutaneous coronary intervention (PCI) frequently used to treat coronary artery disease in HIV, little is known regarding the outcomes and the effects of immune activation on the this procedure.

In literature an observational study on 50 HIV infected patients undergoing PCI found that this procedure is a safe treatment strategy of coronary revascularization in HIV positive patients without significant differences in terms of clinical restenosis from the control population [43].

On the other hand Martín-Reyes et al. [44] reported that in-hospital prognosis in patients with HIV, undergoing PCI, was worse than in control subjects: in fact HIV-positive patients presented a lower rate of PCI success (75 vs. 85 %) respect HIV-negative patients. Moreover Segev et al. [45] analyzing the long-term outcome of PCI in 12 HIV-infected patients, observed that 58.3 % of subjects suffered from severe clinical and/or angiographic restenosis requiring additional interventions or causing severe angina pectoris. They concluded that HIV-infected patients should be considered as high risk group and treated routinely with drug-eluting stents but they did not given any pathogenic interpretation of their observations. Although HIV-infected patients have a higher incidence of post-PCI ischemic events, restenosis, and stent thrombosis, Matetzky underlined that the intermediate-term mortality is low [46].

At this moment few data are available regarding the pathogenesis of restenosis after coronary revascularization in HIV positive patients. It is possible that the concomitant presence of traditional risk factors, assumption of antiretroviral therapy and his metabolic complications contribute to the restenosis events after PCI in HIV-positive population. However, although the presence of these well-known reasons of atherogenesis, the rapidity of progression of atherosclerosis and restenosis process seems linked principally to chronic inflammation and excess immune activation in HIV-infected patients [47‐49]. In fact chronic HIV infection is characterized by multifaceted systemic immune activation, including increased frequencies of activated T-cells and increased turnover of T-cells that correlate directly with disease progression [50]. Moreover a previous study have described other conditions (coronary allograft vasculopathy, native atherosclerosis) associated with an inappropriate inflammatory reactivity that may predispose to re-stenosis after PCI [51]. For these reasons therapies that reduce immune activation may be of benefit, particularly for such individuals [50, 52]; moreover a strictly monitoring of coronary atherosclerosis is required in management of HIV-infected patients with concomitant non-traditional risk factors for intense immune-activation.

In summary contrasting results are reported in the literature on the outcome of coronary revascularization; moreover the coronary revascularization with PCI and stents seems to be influenced by the excess of HIV related inflammation and immune activation [53]. Large clinical studies are needed to clarify aspects of the matter is not yet resolved.

As previously highlighted, HIV-infected patients experienced a premature and accelerated aging probably driven by chronic inflammation and persistent immune activation. For this reason currently immunological therapeutic strategies and virological interventions are investigated with the aim of controlling the levels of immune activation and then reducing the impact of CVD promoted by chronic inflammation.

An innovative immunological/pharmacological intervention, granted by experience in HIV negative populations, is based on use of statins. These class of drugs, widely used for primary prevention of atherosclerotic CVD, are known to decrease cholesterol level as well as inflammation in HIV seronegative patients. For example in the JUPITER trial rosuvastatin significantly reduced all-cause mortality and in particular the incidence of major cardiovascular events in healthy persons without hyperlipidemia but with elevated hsCRP levels, a well-known marker of inflammation [54]. Statins will emerge also as potential components of the therapeutic armamentarium for chronic inflammatory and autoimmune pathologies which share some pathogenetic aspects concerning chronic inflammation with HIV infection [55]. In fact statin therapy has been shown to reduce disease severity in patients with rheumatoid arthritis (RA) and to improve endothelium-dependent vasodilation in patients with RA and systemic lupus erythematosus (SLE) [56‐60]. In systemic inflammatory diseases (such as RA, SLE, inflammatory bowel diseases, psoriasis, spondyloarthritis and others) multiple factors, including tumor necrosis factor-α (TNF-α), circulating inflammatory cytokines, reactive oxygen species, autoantibodies, and oxidized low density lipoprotein (LDL) directly and indirectly activate endothelial cells, increased endothelial permeability, increased leukocyte adhesion and generation of a pro-thrombotic state. In this case statins could have a potential role in modifying initiation and amplification of immune inflammatory responses and their anti-inflammatory activity is linked to different mechanisms including (but not limited) the decrease in chemotaxis of monocytes and macrophages, the lipopolysaccharide (LPS) mediated release of TNF-α, and the activation of NO synthase [61‐64].

The efficacy and safety of different statins for HIV-infected individuals in the primary prevention setting remain to be established. Also the effect of statins use on overall mortality in HIVpositive patients remains controversial [65, 66].

A recent meta-analysis, including 18 clinical studies with 736 HIV-positive patients under cART treated with statins in the primary prevention, showed that this therapy significantly lowers plasma total cholesterol (TC) and low density lipoprotein (LDL) levels. In particular rosuvastatin 10 mg and atorvastatin 10 mg provided the largest reduction in TC levels while atorvastatin 80 mg and simvastatin 20 mg had the largest reduction in LDL [67].

Moreover atorvastatin and rosuvastatin appear to reduce non-calcified coronary plaque volume and slow progression of carotid intima-media thickness in HIV positive patients under cART [68].

Conversely the effects of statins on inflammation and are less conclusive at the moment: a recent study, evaluating the effect of rosuvastatin on markers of immune activation in treatment-naive HIV-patients, showed that this molecule had a small but significant positive effect on CD4/CD8 T cell ratio, but not influenced other activation markers (neopterin, soluble Toll-like receptor (TLR) 2, sTLR4, interleukin (IL)-6, IL-1Ra, IL-18, D-dimer, hsCRP) [69]. Interesting the protective effect of statins on CV system is related not only to their anti-inflammatory activity but also to a control of oxidative stress: in fact oxidative stress plays a significant role in atherosclerosis development and is also elevated in HIV+ patients due to excessive production of free radicals that play a major role in oxidative modification in LDL particle. Serum oxidized LDL levels may contribute to monocyte activation, are associated with clinical manifestations of atherosclerosis, decreases in response to statin therapy and relates independently to reductions in coronary plaque in patients with HIV. Statins may reduce oxidative stress and subsequently subclinical vascular disease in HIV but further study in the context of HIV disease is warranted [70‐72].

Considered the importance of the topic and the still existing doubts, the US National Institutes of Health (NIH) recently launched the REPRIEVE trial to definitively assess the efficacy of statins as a primary prevention strategy for CVD in this at-risk population. This clinical study launched in 2015 is the largest ongoing trial of HIV-related CVD to date (see Table 2) [73].

Some concerns are the potential side effects described for these class of drugs. i.e. statin use was found associated with a modestly increased risk of incident diabetes mellitus in HIV Outpatient Study (HOPS) participants [74]. Also severe rhabdomyolysis, hepatotoxicity and acute renal failure were described in HIV positive population [75‐77]. Concerns are also the potential numerous drug–drug interactions with cART dependent on statins’ pharmacokinetic profile: i.e. simvastatin and lovastatin, metabolized through cytochrome P450 (CYP) 3A, have the significant potency for drug–drug interaction with potent CYP3A inhibitors such as ritonavir- or cobicistat-boosted HIV-protease inhibitors (PIs). Less potent drug–drug interactions have been reported for atorvastatin although it is also a CYP3A substrate. Non-CYP3A-dependent statin concentrations are also affected when co-administered with PIs, mainly through interaction with OATP1B1 [78, 79].

Interesting pitavastatin, a newer statin that does not have substantial interactions with antiretroviral drugs, will be tested in the REPRIEVE trial. At the present time, however, many aspects of potential drug interactions remain to be clarified. Despite all the warnings described, the overall use of statins in HIV population is associated with low rates of adverse events and the mean discontinuation rate reported in meta-analysis and attributable to these problems was 0.12 per 100 person-years [95 % CI (0.05, 0.20)] [67].

The possibility that residual immune activation (and the consequent inflammation) is related to persistent virus replication has suggested that more potent anti-retroviral regimen (i.e., ART “intensification”) may be more effective in reducing the HIV-associated immune activation. In one study intensification was carried out for 10 weeks with cART regimens (atazanavir/ritonavir, lopinavir/ritonavir, efavirenz) consisting of drugs that were not used previously in the enrolled patients. Since the median levels of viremia were not significantly different between pre-intensification and post-intensification period, the authors concluded that the levels of residual virus replication, and thus immune activation, may depend on the size of stable reservoir compartments that are established prior to the initiation of cART [80]. Another study of Llibreet et al. [81] assigned randomly 69 patients with undetectable levels of viraemia for more than 1 year to treatment groups in which 45 intensify therapy with raltegravir and 24 continue the same therapy for 48 weeks. The author observed after 24 weeks in the group of intensification therapy a reduction of the size of the reservoirs and of the levels of immune activation (measured as fraction of CD8+ T-cells expressing CD38+ and/or HLA-DR). Further studies on the effects of cART intensification on the residual immune activation will be needed to ascertain whether a complete recovery of the “ideal” pre-HIV infection immune system function can be achieved through virological interventions alone. In absence of conclusive results in this research field, the alternative possibilities, i.e. specific immune-based interventions, will be necessary to try to protect HIV infected individuals by damages of immune activation. In this sense we know that the most important source of the status of generalized inflammation is attributed to the breakdown in the integrity of the gut mucosa with the increase of microbial translocation and the transition of the microbial products in the systemic circulation. For this reason different trials were conducted to explore the possibility of intervention on this source of chronic inflammation. In a recent study in order to enhance GI tract immunity, we recruited and treated 20 HIV-infected humans with cART supplemented with probiotics and followed inflammation and immunological parameters. We observe that cART did not normalize the levels of immune activation in HIV positive patients, anyway inflammation and markers of microbial translocation were significantly reduced with probiotic supplementation. Patients enrolled showed a clear and statistically significant reduction in the levels of immune activation on CD4 T-lymphocytes, for both markers CD38 and HLA-DR and their simultaneous expression, LBP and hsCRP plasma levels after probiotic diet supplementation. In conclusion this study, like other recent reports [82‐85], evidenced that supplementing cART with probiotics in HIV-infected individuals may improve GI tract immunity and thereby mitigate inflammatory sequelae, ultimately improving prognosis [86].

Another potentially interesting therapeutic concept for the reduction of high levels of inflammation and to control CVD is to use non pharmacological tools of intervention such as physical activity. Several published studies indicate that physical activity could represent a beneficial non-pharmacological intervention to reduce chronic inflammation. Currently available data are limited, nonetheless increasing evidence suggests that the introduction of regular physical exercise in the clinical management of HIV-infected individuals may have a significant therapeutic impact in reducing inflammation and CVD. For example Longo et al. [87] conducted a longitudinal study on 50 ART-treated HIV-infected patients with sedentary life style to evaluate the effect of 12 weeks of moderate intensity exercise on parameters of immune activation and metabolic profile. At the enrollment and at 12 weeks all participants were assessed by a 6-min walking test and 1-RM, DEXA, metabolic and immune-virologic parameters, and markers of immune activation (IL-6, d-Dimer, sCD14 and IL-18). The authors found a significant improvement of both fitness and immune activation at the week-12 time point.

On the basis of these evidences, several authors proposed that regular physical exercise should be further studied as a potential anti-inflammatory, non-pharmacological approach to be used to treat HIV residual disease and non-AIDS-defining illnesses in cART-treated HIV-infected individuals [88‐94].

Likewise, lifestyle interventions are another non pharmacological tool reputed able to reduce cardiovascular risk in general population: nevertheless a recent randomized pilot study with a follow-up 36 months evidenced that a multidisciplinary lifestyle intervention resulted in a slight improvement in some cardiovascular risk factors and did not prevent cIMT progression in HIV-infected patients with Framingham scores >10 % [95]. Further studies are required to confirm the importance of all these data.