CardiovasCular CompliCations in the aCquired immunodefiCienCy syndrome

Rev Assoc Med Bras 2009; 55(5): 621–30 *Correspondência: Viale Anicio Gallo, 63 00174 Rome, Italy Tel/Fax:+39-6-71028.89 g.barbaro@tin.it Summary The introduction of highly active antiretroviral therapy (HAART) has significantly improved the clinical outcome of HIV disease, with increased survival rates. However, the introduction of HAART has generated a contrast in the cardiac manifestations of AIDS. In developed countries, we observed an approximate 30% reduction in the prevalence of HIV-associated cardiomyopathy, possibly related to a reduction of opportunistic infections and myocarditis. In developing countries, however, where the availability of HAART is limited and the pathogenic impact of nutritional factors is significant, we observed an increase of approximately 32% in the prevalence of HIV-associated cardiomyopathy and a related high mortality rate from congestive heart failure. Also, some HAART regimens in developed countries, especially those including protease inhibitors, have been shown to cause, in a high proportion of HIV-infected patients, a iatrogenic metabolic syndrome (HIV-lipodystrophy syndrome).This is associated with an increased risk of atherosclerosis-related cardiovascular events even in young HIVinfected people. A better understanding of the molecular mechanisms responsible for this syndrome will lead to the discovery of new drugs that will reduce cardiovascular risk in HIV-infected patients receiving HAART.


introduCtion
Studies published before the introduction of highly active antiretroviral therapy (HAART) have tracked the incidence and course of human immunodeficiency virus (HIV) infection in relation to both pediatric and adult cardiac illnesses 1 .These studies showed that subclinical echocardiographic abnormalities independently predict adverse outcomes and identify high risk groups to be targeted for early intervention and therapy.Understanding the effects of HAART on the cardiovascular system is only possible after first understanding the effects of HIV co-infections 2 .HAART is only available to a minority of HIV infected individuals worldwide and studies published before the introduction of HAART continue to be applicable globally 3 , especially in developing countries [4][5][6] .

Hiv-associated cardiomyopathy
Human immunodeficiency virus (HIV) disease is recognized as an important cause of dilated car diomyopathy, with an estimated annual incidence of 15.9/1,000 before the introduction of HAART 7 .The im portance of cardiac dysfunction is demonstrated by its effect on survival in the acquired immunodefi ciency syndrome (AIDS).Median survival to AIDS-related death is 101 days in patients with left ventricular dysfunction and 472 days in pa tients with a normal heart on echocardiography at a similar infection stage 7 .The unadjusted haz ard ratio for death in HIV-associated cardiomyopa thy compared to idiopathic cardiomyopathy is 4.0; the ratio adjusted after multivariate analysis is 5.86 7 .The introduction of HAART regimens, by preventing opportunistic infections and reduc ing the incidence of myocarditis, has reduced the prevalence of HIV-associated cardiomyopathy by about 30% in developed countries 8,9 .However, the median prevalence of HIV-associated cardiomyopathy is increasing in developing countries (about 32%), where the availability of HAART is scarce and the pathogenetic impact of nutritional factors is greater [4][5][6] .

myocarditis
Myocarditis continues to be the best-studied cause of di lated cardiomyopathy in HIV disease.HIV-associated myocarditis may be defined as «a process characterized by a lymphocytic infiltrate of the myocardium with necrosis and/or degenera tion of adjacent myocytes not typical of the isch emic damage associated with coronary artery dis ease in subjects infected by HIV with or without evidence of opportunistic infective agents» 10 .Myocarditis has been identified at autopsy in 40-52% of patients who died of AIDS before the introduction of HAART 11 .In the Gruppo Ital iano per lo Studio Cardiologico dei pazienti af fetti da AIDS (GISCA) autopsy series, histologi cal diagnosis of myocarditis was made in 30 of 82 patients (37%) with cardiac involvement 11 .Of 20 autopsy patients with dilated cardiomy opathy, 10 (83%) had active myocarditis at histo logical examination of myocardial tissue speci mens 11 .
Histological findings in HIV-infected pa tients with myocarditis do not substantially differ from those observed in HIV-uninfected patients.Lymphocytes, along with fewer macrophages, are distributed diffusely as single cells or in small clusters.Autopsy studies of AIDS patients who died of acute left ventricular dysfunction, almost in variably show a marked inflammatory infiltrate 11,12 .However, mild and focal mononuclear infil trates are frequently observed in the hearts of AIDS patients, irrespective of the presence of cardiac symptoms 11,12 .The intracardiac conduction sys tem is sometimes affected in AIDS patients, gen erally as a complication of myocarditis.Primary involvement of the conduction system is possible as a consequence of opportunistic infections, drug cardiotoxicity, or primary location of HIV-1 in the specific conduction tissue 11 .Histological ex amination of myocardium specimens at autopsy may show mononuclear infiltration of the intra cardiac conduction tissue, which is frequently as sociated with vasculitis and fragmentation of the bundles with lobulation and fibrosis.These find ings are generally associated with electrocardio graphic conduction abnormalities (e.g., left anterior hemiblock, left bundle branch block, and first-degree atrioventricular block) 11 .In the GISCA autopsy series, 5/12 patients (42%) with cardiomyopathy also had intracardiac conduction system alterations on histological examination.Two of them had lym phocytic infiltration of the conduction tissue as a complication of active myocarditis and the re maining three had fragmentation and fibroma tous degeneration of the left bundle 11 .

Nonviral myocarditis
The most common oppor tunistic infectious agent associated with myocar ditis in AIDS is Toxoplasma gondii, observed as often as 12% in one autopsy series of deaths from AIDS between 1987 and 1991 12 .There may be regional differences in the incidence of Toxoplasma gondii myocarditis, perhaps because the natural reservoir of organisms persists more easily in hu mid environments.Elevation of the myocardial frac tion of creatine kinase may commonly occur with myocardial toxoplasmosis.Toxoplasma gondii organisms can produce a gross pattern of patchy irregular white infiltrates in the myocardium similar to non-Hodgkin's lymphoma.Microscopically, the myocardium shows scattered mixed inflam matory cell infiltrates with polymorphonuclear leukocytes, macrophages, and lymphocytes.Toxoplasma gondii can produce fairly variable inflammation along with myocardial fiber necrosis.The three microscopic patterns of involvement by Toxoplasma gondii include acute diffuse myocarditis, focal myocardi tis, and presence of organisms without significant inflammation or necrosis 10,12 .In Toxoplasma myocarditis, true Toxoplasma gondii extracellular cysts, or pseudocysts within myocardial fibers, both of which contain the small 2-µ-sized bradyzoites, are often difficult to find, even if inflammation is exten sive.Immunohistochemical staining may reveal free tachyzoites, organisms that are found outside of cysts.Otherwise, with rou tine hematoxylin and eosin staining, it is difficult to distinguish these free tachyzoites from fragments of inflam matory cells or myocytes that have undergone necrosis within the areas of inflammation 10,12 .
Fungal opportunistic infections of the heart occur infrequently in HIV-infected patients.They are often incidental findings at autopsy, and car diac involvement is probably the result of widespread dissemination, as exemplified by Candida sp. and by the fungi Cryptoccocus neoformans, Coccidioides immitis, or Histoplasma capsulatum [10][11][12] .Fungal lesions are characterized gross ly by the appearance of multiple small rounded white plaques.They may have a hemorrhagic bor der, particularly lesions caused by Aspergillus that can be angioinvasive.Microscopically, fungal lesions have variable inflammatory infiltrates and necrosis and a specific diagnosis is made by iden tifying yeast forms or hyphae of specific organ isms, aided by standard histological stains such as Gomori methenamine silver or periodic acid Schiff [10][11][12] .The near absence of an inflam matory infiltrate accompanying fungal organisms is a manifestation of immune system failure with progression of AIDS to a late stage when oppor tunistic infections are more likely to be widely disseminated to organs such as the heart [10][11][12] .Patients living in endemic areas for Trypano soma cruzi may rarely develop a pronounced myocarditis 13,14 .Mycobacterium avium-complex infection can be widely disseminated and in volve the heart with microscopic lesions charac terized by clusters of large macrophages filled with numerous acid-fast rod-shaped organisms [10][11][12] .Pneumocystis carinii can involve the heart in cases with widespread dissemination of this or ganism 12 .Grossly, the epicardium and cut sur faces of the myocardium may have a sandpaper-like quality due to the presence of multiple pinpoint foci of calcification.Microscopically, this calcification is not accompanied by significant in flammatory cell infiltrates, but there may be de posits of amorphous granular pink exudate simi lar to that seen in alveoli with Pneumocystis cari nii pneumonia 10,12 .The cysts may be difficult to recognize, even with Gomori methenamine sil ver stain, and diagnosis is aided by immunohisto chemical staining 10,12 .

Viral myocarditis
Histology and immunohisto chemistry rarely detect the presence of viruses in the myocardium.However, in situ hybridiza tion or polymerase chain reaction studies reveal a high frequency of either cytomegalovirus or HIV-1 or both, in AIDS patients with lympho cytic myocarditis and severe left ventricular dys function 11,15 .These data support the hypothe sis that, at least in a subset of patients, HIV-1 has a pathogenetic action and possibly influences the clinical evolution toward dilated cardiomyopa thy.HIV-1 nucleic acid sequences were detected at autopsy by in situ DNA hybridization in 35% of the GIS CA patients with cardiac involvement; 86% of them had active myocarditis at histological ex amination 11 .Among patients with myocarditis, coinfection with coxsackievirus B3 was docu mented in 32%, with Epstein-Barr virus in 8%, and with cytomegalovirus in 4% 11 .In autopsy biopsy samples, myocytes with a positive hybrid ization signal were sparse, usually only one to four cells per section 11 .Although about 70% of patients with positive hybridization signals had active myocarditis at histological examination, most myocytes with a positive hybridization signal were not surrounded by inflammatory cells 11 .In the GISCA autopsy series, HIV-1 was document ed by in situ hybridization in 83% of patients with myocarditis 11 .
Coinfection with other viruses seems to have an important etiopathogenetic role.The GISCA autopsy records show that 83% of patients with myocarditis and 50% of those with dilated cardiomyopathy were coinfected with cardiotropic vi ruses (usually, coxsackievirus B3 and cytomega lovirus 11 ).Herskowitz et al. used in situ hybrid ization to detect myocardial cytomegalovirus infection in 48% of HIV-infected patients with myocarditis and left ventricular dysfunction who underwent endomyocardial biopsy 15 .Bowles et al. used polymerase chain reaction and found that 42% of HIV-infected patients with cardiomy opathy had cytomegalovirus or adenovirus in the myocardial tissue 16 .Some patients with adeno virus coinfection had congestive heart failure but not myocarditis, suggesting that the virus may be virulent without an associated inflammatory re sponse.

myocardial cytokine expression
Myocardial dendritic cells may play a role in the interaction between HIV-1 and the cardiac myo cyte and in the activation of cytotoxic cytokines 17 .Dendritic cells can initiate the primary immunologic response and present the antigen to T-lymphocytes.The interaction be tween dendritic cells and T-lymphocytes, particu larly CD8 cells, could promote a local elevation in the multifunctional cytokine tumor necrosis factor-(TNF) alpha, which can also be produced and secreted by infected macrophages 18 .TNF-alpha produces a negative ino tropic effect by altering intracellular calcium ho meostasis, possibly by inducing nitric oxide (NO) synthesis, which also reduces myocyte contractil ity.Kan et al. demonstrated that HIV gp120 enhances NO production by cardiac myocytes through p38 MAP kinase-mediated NK-kB activation 19 .Myocarditis and dilated cardiomyopathy are associated with markedly elevated cytokine pro duction, but the elevations may be highly local ized within the myocardium, making peripheral cytokine levels uninformative 20 .When myocar dial biopsies from patients with HIV-associated cardiomyopathy are compared to samples from patients with idiopathic dilated cardiomyopathy, the former stain more intensely for both TNFalpha and inducible nitric oxide synthase (iNOS).Stain ing is particularly intense in samples from patients with a myocardial viral infection and correlates with CD4 count, independent of antiretroviral treatment 20 .Staining is also more intense in samples from patients with HIV-associated car diomyopathy coinfected with coxsackievirus B3, cytomegalovirus, Epstein-Barr virus or adenovi rus 20 .Moreover, staining for iNOS is more in tense in samples from patients coinfected with HIV-1 and coxsackievirus B3 or cytomegalovirus than in samples from patients with idiopathic di lated cardiomyopathy and myocardial infection with coxackievirus B3 or who had adenovirus in fection alone 20 .In patients with HIV-associated dilated cardiomyopathy and more intense iNOS staining, the survival rate was significantly lower: those whose samples stained more than 1 optical density unit had a hazard ratio of mortality of 2.57 20 .Survival in HIV-infected patients with less intense staining was not significantly different from survival in pa tients with idiopathic dilated cardiomyopathy 20 .

autoimmunity as a contributor to HIV-associated cardiomyopathy
Cardiac-specific autoantibodies (anti-alpha-myosin autoantibodies) are more common in HIV-infected patients with dilated cardiomyopathy than in HIV-infected patients with healthy hearts.Currie et al. reported that HIV-infected patients were more likely to have specific cardiac autoantibod ies than were HIVuninfected controls 21 .Those with echocardiographic evidence of left ventricu lar dysfunction were particularly likely to have cardiac autoantibodies, supporting the theory that cardiac autoimmunity plays a role in the pathogenesis of HIV-associated cardiomyopathy and suggesting that cardiac autoantibodies could be used as markers of left ventricular dysfunction in HIV-infected patients with previously normal echocardiographic findings 21 .In addition, monthly intravenous immunoglobulin in HIV-infected pediatric patients minimizes left ventricu lar dysfunction, increases left ventricular wall thickness, and reduces peak left ventricular wall stress.This suggests that both impaired myocardial growth and left ventricular dysfunction may be immunologically mediated 22 .These effects may be the result of immunoglobulins inhibiting cardiac autoantibodies by competing for Fc re ceptors, or of immuno globulins dampening the secretion or effects of cytokines and cellular growth factors 22 .These findings suggest that immunomodulatory therapy might be helpful in HIV-infected patients with de clining left ventricular function, although further study of this possible therapy is needed.

Nutritional deficiencies as a factor in left ventricular dysfunction
Nutritional deficiencies are common in HIV in fection and may contribute to ventricular dys function independently of HAART.Malabsorp tion and diarrhea can both lead to trace element deficiencies which have been directly or indirectly associated with cardiomyopathy 23,24 .Selenium, as a component of glutathione peroxidase, is involved in the antioxidant response in cells and tissues.Selenium deficiency is associated with congestive cardiomyopathy and skeletal-muscle disorders.Furthermore, low levels of selenium or other macronutrients may be responsible for the cardiotoxic effects of coxsackievirus B3 and for the ability of these viruses to enhance the toxic effects of zidovudine on skeletal muscle.Both findings may be relevant to selenium-related ventricular dysfunction in HIV-infected patients 24,25 , especially in developing countries 6 .Selenium replacement may reverse cardiomyopathy and re store left ventricular function in seleniumdeficient patients 26 .HIV infection may also be as sociated with altered levels of vitamin B 12 , carni tine, growth hormone, and thyroid hormone, all of which have been associated with left ventricu lar dysfunction 23,24 .

Autonomic dysfunction as a factor in left ventricular dysfunction
There is evidence of an altered autonomic function in HIVinfected patients 27,28 .In a previous study, autonomic dysfunction in HIV-infected patients was assessed by a significant decrease of the corrected coefficients of electrocardiographic R-R interval variation (CVc) compared to healthy controls 29 .CVc correlated significantly with CD4 count and with values of hepatic, plasmatic, lymphocyte and erythrocyte concentrations of reduced glutathione as well as with erythrocyte malonyldialdehyde levels as the expression of increased lipoperoxidation 29 .This correlation was more evident in patients coinfected with hepatitis C virus (HCV) where the lipoperoxidation process and production of free radicals of oxygen are enhanced by the inflammatory process induced by HIV/HCV coinfection 29 .According to this study, the autonomic dysfunction in HIV-infected patients may be the expression of a reduced response to the oxidative stress related to HIV infection because of a systemic depletion of reduced glutathione associated with the state of immunodeficiency 29 .
Left ventricular asynergy may develop due to regional differences in the distribution of cardiac sympathetic nerve endings, even in the context of acute myocarditis.In fact, an alteration of catecholamine dynamics (or autonomic function) has been associated with a transient extensive akinesis of the apical and mid portions of the left ventricle with hypercontraction of the basal segment (takotsubo-like dysfunction) in an HIV-infected patient with cytomegalovirus myocarditis 30 .

Left ventricular dysfunction caused by drug cardiotoxicity
Studies of transgenic mice suggest that zidovudine is associated with diffuse destruction of the cardiac mi tochondrial ultrastructure and inhibition of mito chondrial DNA replication 31 .This mito chondrial dysfunction may result in lactic acidosis, which could also contribute to myocardial cell dys function.However, in a study of infants born to HIV-infected mothers followed from birth to age 5, perinatal exposure to zidovudine was not found to be associated with acute or chronic abnormalities in left ventricular structure or function 32 .Other nucleoside reverse transcriptase inhibitors, such as didanosine, zalcitabine or lamivudine, do not seem to promote or prevent dilated cardiomyopathy.In AIDS patients with Kaposi's sarcoma, reversible cardiac dysfunction was associated with prolonged, high-dose therapy with interferon alpha 33 .High-dose interferon alpha treatment is not associated with myo cardial dysfunction in other patient populations, so it has been proposed that it may have a synergistic effect with HIV-1 infection 33 .Doxorubicin (adria mycin), which is used to treat AIDS-associated Kaposi's sarcoma and non-Hodgkin's lymphoma, has a dose-related effect on dilated cardiomyopathy 34 , as does foscarnet sodium when used to treat cyto megalovirus esophagitis 35 .

Hiv infection opportunistic infections and vascular disease Endothelial dysfunction
Endothelial dysfunction and injury have been described in HIV infection 36 .Endothelial activation in HIV-1 infection may also be caused by cytokines (e.g., TNF-alpha) secreted in response to mononuclear or adventitial cell activation by the virus, or may be a direct effect of the secreted HIV-1-associated proteins gp 120 and tat (transactivator of viral replication) on the endothelium with possible induction of apoptosis process 37 .Opportunistic agents, such as cytomegalovirus, as well as human herpes virus-8 (a virus involved in the development of AIDS-associated Kaposi's sarcoma), frequently co-infect HIV-infected patients and may contribute to development of endothelial damage.In spite of all these observations, the clinical consequences of HIV-1 and opportunistic agents on endothelial function have not been elucidated yet.

Vasculitis
A wide range of inflammatory vascular diseases including polyarteritis nodosa, Henoch-Schonlein purpura and druginduced hypersensitivity vasculitis may develop in HIV-infected individuals.Kawasaki-like syndrome [38][39][40] and Takayasu's arteritis 41 have also been described.The course of vascular disease may be accelerated in HIV-infected patients because of atherogenesis stimulated by HIV-infected monocyte-macrophages, possibly via altered leukocyte adhesion or arteritis 39 .
The incidence of vasculitis (excluding adverse drug reactions) in HIV infection is estimated to be about 1% 39 .Some HIV-infected patients have a clinical presentation resembling systemic lupus erythematosus including vasculitis, arthralgias, myalgias, and autoimmune phenomena with a low titer positive antinuclear antibody, coagulopathy with lupus anticoagulant, hemolytic anemia, and thrombocytopenic purpura 39 .Hypergammaglobulinemia from polyclonal B-cell activation may be present, but often diminishes in the late stages of AIDS.Specific autoantibodies to double-stranded DNA, Sm antigen, RNP antigen, SSA, SSB and other histones may be found in the majority of HIV-infected persons, but their significance is unclear 39 .

HIV infection and coronary arteries
The association between viral infection (cytomegalovirus or HIV-1 itself) and coronary artery lesions is not clear.HIV-1 sequences have been detected by in situ hybridization in the coronary vessels of an HIV-infected patient who died from acute myocardial infarction 42 .Po tential mechanisms through which HIV-1 may damage cor onary arteries include activation of cytokines and cell-adhesion molecules and alteration of major histocompatibility complex class I molecules on the surface of smooth-muscle cells 42 .It is also possible that HIV-1-associated protein gp 120 may induce smooth-muscle cell apoptosis through a mitochondrion-controlled pathway by activation of inflammatory cytokines 37 .

periCardial effusion
The prevalence of pericardial effusion in asymptomatic HIV-infected patients has been estimated at 11% before the introduction of HAART 43 .According to some retrospective data, prevalence of pericardial effusion in HIV-infected patients is reduced by about 30%-35% after the introduction of HAART, with a trend similar to that observed for HIV-associated cardiomyopathy 8 .On the contrary, in developing countries, prevalence of pericardial effusion is increased by 35%-40%, mostly related to Mycobacteria infections 4,5 .HIV infection should be included in the differential diagnosis of unexplained pericardial effusion or tamponade.Pericardial effusion in HIV disease may be related to opportunistic infections (e.g., Mycobacterium tuberculosis, Mycobacterium avium-complex infection) or to malignancy (e.g., non-Hodgkin lymphoma), but most often a clear etiology is not found.The effusion may be part of a generalized serous effusive process, also involving pleural and peritoneal surfaces.This "capillary leak" syndrome is possibly related to enhanced cytokine expression (e.g., TNF-alpha) in the later stages of HIV disease and correlates with the immunodeficiency state of the patient 8,44 .

Endocarditis
Prevalence of infective endocarditis did not vary in HIVinfected patients who use intravenous drugs after the introduction of HAART even in developed countries, being similar to that observed in HIV-uninfected intravenous drug addicts 7 .Estimates of infective endocarditis prevalence vary from 6.3% to 34% of HIV-infected patients who use intravenous drugs independently of HAART 11 .Among intravenous drug addicts, the tricuspid valve is most frequently affected and the most frequent agents are Staphylococcus aureus (>75% of cases), Streptococcus pneumoniae, Haemophilus influenzae, Candida albicans, Aspergillus fumigatus and Cryptococcus neoformans 11 .Avirulent bacteria, such as the HACEK group (Haemophilus species, actinomycetemcomitans,Cardiobacterium hominis, Eikenella corrodens and Kingella kingae), which are often part of the, endogenous flora, of the mouth, can cause endocarditis in HIV-infected patients 12 .Vegetations may form on the tricuspid or pulmonary valves with resultant pulmonary embolism and consequent septic pulmonary infarction which appear as multiple opacities on chest radiograms.Systemic emboli often involve coronary arteries, spleen, bowel, extremities and central nervous system.Cardiac rhythm alterations (i.e., atrioventricular block) may suggest the presence of an abscess near to the atrioventricular node.Peripheral pulses should be examined for signs of embolic occlusion or pulsating mass suggesting mycotic aneurysm.Mycotic aneurysms may occur in the intracranial arteries potentially leading to intracranial hemorrhage 45 .Patients with HIV infection generally have similar presentations and survival (85% vs. 93%) from infective endocarditis as those without HIV.However, patients with late-stage HIV disease have about 30% higher mortality with endocarditis than asymptomatic HIV-infected patients, which may be related to the degree of immunodeficiency 46 .Non bacterial thrombotic endocarditis, also known as marantic endocarditis, had a prevalence of 3%-5% in AIDS patients, mostly in patients with HIVwasting syndrome, before the introduction of HAART 11 .Friable endocardial vegetations, affecting predominantly the left-sided valves, consisting of platelets within a fibrin mesh with few inflammatory cells characterize it 11,12 .Marantic endocarditis is now more frequently observed in developing countries, with a high incidence (about 10%-15%) and mortality by systemic embolization 4,5 .

HIV-associated Pulmonary Hypertension
The incidence of HIV-associated pulmonary hypertension has been estimated in 1/200, much higher than 1/200,000 found in the general population and is increased after introduction of HAART 47 .The histopathology of HIV-associated pulmonary hypertension is similar to that of primary pulmonary hypertension.The most common alteration in HIV-associated pulmonary hypertension is the plexogenic pulmonary arteriopathy, while thrombotic pulmonary arteriopathy and pulmonary veno-occlusive disease are rarer histological findings 12 .This observation may suggest that similar etiopathogenetic mechanisms are at the basis of both HIV-associated pulmonary hypertension and primary pulmonary hypertension.A key pathogenetic role is played by pulmonary dendritic cells which are not sensitive to HAART and may hold HIV-1 on their surfaces for extended time periods 47 .Infection of these cells by HIV-1 causes a chronic release of cytotoxic cytokines (e.g., endothelin-1, interleukin-6, interleukin-1 beta and TNF-alpha) contributing to vascular plexogenic lesions and progressive tissue damage, independently of opportunistic infections, stage of HIV disease and HAART regimens 47 .Activation of alpha-1 receptors and genetic factors (increased frequency of HLA-DR6 and DR52) have also been hypothesized in the pathogenesis of HIV-associated pulmonary hypertension 47 .

CardiaC involvement in aids-assoCiated neoplasms
The prevalence of cardiac Kaposi's sarcoma in AIDS patients ranges from 12% to 28% in retrospective autopsy studies performed in the pre-HAART era 11 .Cardiac involvement with Kaposi's sarcoma usually occurs when widespread visceral organ involvement is present.Lesions typically measure less than 1 cm and may be pericardial or, less frequently, myocardial, and are only rarely associated with obstruction, dysfunction, morbidity, or mortality.Microscopically, there are atypical spindle cells lining slit-like vascular spaces 12 .Non-Hodgkin lymphoma involving the heart is not frequent in AIDS 11,12 .Most are high-grade B-cell (small non-cleaved) Burkitt-like lymphomas, with the rest classified as diffuse large B cell lymphomas (in the REAL classification).Lymphomatous lesions may appear grossly as either discreet localized or more diffuse nodular to polypoid masses 11, 48, 49 .Most involve the pericardium, with variable myocardial infiltration 11,48,49 .There is little or no accompanying inflammation and necrosis.Prognosis of patients with HIV-associated cardiac lymphoma is generally poor because of widespread organ involvement, although some patients treated with combination chemotherapy have experienced clinical remission 50 .The introduction of HAART led to a reduction by about 50% in the overall incidence of cardiac involvement by Kaposi's sarcoma and non-Hodgkin lymphomas.This may be attributable to the improved immunologic state of patients and prevention of opportunistic infections (human herpes virus-8 and Epstein-Barr virus) known to play an etiologic role in these neoplasms.On the contrary, an increased prevalence of cardiac involvement of AIDS-associated tumors may be observed in developing countries in relation to the scanty availability of HAART 4,7 .

Haart-associated lipodystrophy and metabolic syndrome
The introduction of HAART has significantly modified the course of HIV disease, with longer survival and improved quality of life in HIV-infected subjects.However, HAART regimens, especially those including protease inhibitors (PI) have shown to cause, in a high proportion of HIV-infected patients, somatic (lipodystrophy /lipoatrophy) and metabolic (dyslipidemia,insulinresistance) changes that in the general population are associated with an increased risk of cardiovascular disease, producing an intriguing clinical scenario.
Rev Assoc Med Bras 2009; 55(5): 621-30 HIV-associated lipodystrophy or lipoatrophy, unreported prior to introduction of HAART, was first described in 1998 51 .It is characterized by the presence of a dorsocervical fat pad (also known as buffalo hump), increased abdominal girth and breast size, lipoatrophy of subcutaneous fat of the face, buttocks and limbs, and prominence of veins on the limbs.The overall prevalence of at least one physical abnormality is thought to be about 50% in otherwise healthy HIV-infected patients receiving HAART, although reported rates range from 18% to 83% 52,53 .Differences in rates might be influenced by age, gender, the type and duration of antiretroviral therapy, and the lack of an objective and validated case definition 52,53 .As in genetic lipodystrophy syndromes, fat redistribution may precede the development of metabolic complications in HIV-infected patients receiving HAART.Among HIV-infected patients with lipodystrophy, increased total serum and low density lipoprotein cholesterol and triglyceride levels have been observed in about 70%, whereas insulin resistance (elevated C-peptide and insulin) and type 2 diabetes mellitus have been observed in 8% to 10% 52,53 .The severity of these metabolic abnormalities increases with growing severity of lipodystrophy, and are associated with a raised risk of cardiovascular events: approximately 1.4 cardiac events per 1000 years of therapy according to the Framingham score 53 .
The pathogenesis of HAART-associated lipodystrophy and metabolic syndrome is complex and a number of factors are involved, including direct effects of HAART on lipid metabolism, endothelial and adipocyte cell function, and mitochondrial dysfunction.

molecular mechanisms PI-associated lipodystrophy and metabolic alterations
PI targets the catalytic region of HIV-1 protease.This region is homologous with regions of two human proteins that regulate lipid metabolism: cytoplasmic retinoic-acid binding protein-1 (CRABP-1) and low density lipoprotein-receptor-related protein (LRP) 54 .It has been hypothesized, although without strong experimental support, that this homology may allow PI to interfere with these proteins, which may be the cause of metabolic and somatic alterations that develop in PI-treated patients 54 .The hypothesis is that PI inhibit CRABP-1-modified and cytochrome P450-3A-mediated synthesis of cis-9-retinoic acid and peroxisome proliferator-activated receptor type-gamma heterodimer.The inhibition increases the rate of apoptosis of adipocytes and reduces the rate at which pre-adipocytes differentiate into adipocytes, with the final effect of reducing triglyceride storage and increasing lipid release.PI-binding to LRP would impair hepatic chylomicron uptake and endothelial triglyceride clearance, resulting in hyperlipidemia and insulin resistance 54 .
Some data indicate that PI-associated dyslipidemia may be caused, at least in part, either by PI-mediated inhibition of proteasome activity and accumulation of the active portion of sterol regulatory element-binding protein (SREBP)-1c in liver cells and adipocytes 55 , or by apo-C III polymorphisms in HIV-infected patients 56,57 .Sequence homologies have been described between HIV-1 protease and human site-1 protease (S1P), which activates SREBP-1c and SREBP-2 pathways.A polymorphism in the S1P/SREBP-1c gene confers a difference in risk for development of an increase in total cholesterol with PI-therapy.This suggests presence of a genetic predisposition to hyperlipoproteinemia in PI-treated patients 58 .There is also evidence that PI directly inhibit the uptake of glucose in insulinsensitive tissues, such as fat and skeletal muscle, by selectively inhibiting the glucose transporter Glut4 59 .

TNF-alpha and lipodystrophy
The relationship between the degree of insulin resistance and levels of soluble type-2-TNF-alpha receptor suggests that an inflammatory stimulus may contribute to development of HIV-associated lipodystrophy 60 .TNF-alpha activates 11-betahydroxysteroid dehydrogenase type-1, which converts inactive cortisone into active cortisol.The activity of this enzyme is higher in visceral fat, compared to subcutaneous fat.Visceral fat is able to locally produce cortisol which could act inside adipocytes and increase lipid accumulation 61 .

Mitochondrial dysfunction and lipodystrophy
There is evidence of nucleoside-induced mitochondrial dysfunction in HIV-infected patients treated with nucleosidecontaining HAART, because lipodystrophy with peripheral fat wasting is associated with a decrease in subcutaneous adipose tissue mitochondrial DNA content 62 .Disrupted pools of nucleotide precursors and inhibition of DNA pol-gamma by specific nucleoside reverse transcriptase inhibitors are mechanistically important in mitochondrial toxicity 63 .This effect has been especially described with use of stavudine and was correlated with length of exposure to this drug 64 .HAART regimens with didanosine plus stavudine are more likely to produce a greater increase in serum lactate and lipodystrophy than therapies based on zidovudine plus lamivudine within the first year of therapy 62 .Substitution of stavudine by abacavir or zidovudine improves mitochondrial indices and fat apoptosis in the setting of lipoatrophy 65 .

adipocytokines and lipodystrophy
Adipocytes secrete a range of adipocytokines which control insulin sensitivity 66 .There is evidence that an adipocytokine, adiponectin, a protein product of the apM1 gene, which is expressed exclusively in adipocytes, plays a role in development of HIV-associated lipodystrophy as well as in congenital and acquired lipodystrophies in non-HIV infected subjects 67 .In vitro and animal studies and cross-sectional studies in humans have shown that adiponectin is inversely correlated with features of HAART-associated metabolic syndrome.This syndrome has recently been linked to a quantitative trait locus on chromosome 3q27, the location of the apM1 gene 66,68,69 .These studies have shown that both adiponectin levels and the adiponectin-to-leptin ratio are positively correlated with features of HAART-associated metabolic syndrome 66,68,69 .According to these studies, this ratio could be used to predict insulin sensitivity and potential cardiovascular risk in HIV-infected patients receiving HAART.

HaarT and cardiovascular disease HAART-associated endothelial dysfunction
In vitro data reported by Fiala et al. suggest that some HAART regimens, such as those including zidovudine, some non-nucleoside reverse transcriptase inhibitors (e.g.efavirenz) and PI disrupt endothelial cell junctions and cytoskeleton action of the endothelial cells leading to endothelial dysfunction 70 .These findings are in agreement with those previously reported in vivo by Stein et al. 71 and in vitro by Zhong et al. 72 .In particular, Zhong et al. demonstrated that ritonavir at concentrations near clinical plasma levels is able to directly cause endothelial mitochondrial DNA damage and cell death mainly through a necrosis pathway but not through apoptosis 72 .Chai et al. investigated and compared the effects of PI on isolated porcine arteries and observed a reduced endothelial NO synthase expression and increased levels of superoxide anion as expression of increased endothelial oxidative stress 73 .The HIV-1 entry inhibitor TAK-799 is an antagonist for the chemokine receptors CCR5 and CXCR3, which are expressed on leucocytes, especially T-helper-1 cells, and these receptors, may be involved in recruitment of these cells to atherosclerotic vascular lesions 74 .In low-density lipoprotein receptordeficient mice treated with TAK-799, the number of T cells in the atherosclerotic plaque was reduced by 95%, concurrently with a 98% reduction in the relative interferon-gamma area 74 .According to this study, TAK-779 not only suppresses HIV-1 entry via blockade of CCR5 but also attenuates atherosclerotic lesion formation by blocking influx of T-helper-1 in the atherosclerotic plaque 74 .Since TAK-799 impairs atherogenesis, treatment with TAK-799 could be beneficial for young HIV-infected patients facing lifelong HAART regimens.

HaarT-associated vasculitis
Drug-induced hypersensitivity vasculitis is common in HIVinfected patients receiving HAART 39 .Vasculitis associated with drug reactions typically involves small vessels and has a lymphocytic or leukocytoclastic histopathology.The pathologic mechanisms include T-cell recognition of haptenated proteins or deposition of immune complexes in blood-vessel walls 39 .Medical practitioners need to be especially aware of abacavir hypersensitivity reactions because of the potential for fatal outcomes.Hypersensitivity reactions of this type should always be considered as a possible etiology for a vasculitic syndrome in an HIV-infected patient 39 .

HAART-associated coagulation disorders
HIV-infected patients receiving HAART, especially those with fat redistribution and insulin resistance, might develop coagulation abnormalities, including increased levels of fibrinogen, D-dimer, plasminogen activator inhibitor-1, and tissue-type plasminogen activator antigen, or deficiency of protein S 75,76 .For instance, protein S deficiency has been reported in up to 73% of HIV-infected men 75,76 .These abnormalities have been associated with thromboses involving veins and arteries and seem to be related to HAART regimens that include PI 77 .Thrombocytosis has been reported in 9% of patients receiving HAART, with cardiovascular complications in up to 25% of cases 78 .

HaarT-associated systemic arterial hypertension
Prevalence of arterial hypertension in HIV disease has been estimated to be about 20-25% prior to introduction of HAART 79 .HIV-associated endothelial dysfunction and injury, autoimmune reaction to viral infection (vasculitis), and renal disease have been hypothesized to play a role in the pathogenesis of HIV-associated hypertension.Endothelial dysfunction has been described in association with visceral fat accumulation and related metabolic alterations (e.g., insulin resistance) and activation of proinflammatory cytokines (TNF-alpha, IL-6), as well as with a direct action of drugs included in HAART regimens.HIV-associated renal impairment can present as acute or chronic kidney disease; it can be caused directly or indirectly by HIV-1 and/or by drug-related effects that are directly nephrotoxic or lead to changes in renal function by inducing metabolic vasculopathy and renal damage 80 .Chronic renal disease can be caused by multiple pathophysiologic mechanisms, leading to HIV-associated nephropathy, a form of collapsing focal glomerulosclerosis, thrombotic microangiopathy, and various forms of immune complex glomerulonephritis 80 .Arterial hypertension, even in agreement with the Adult Treatment Panel-III guidelines 81 , is currently considered part of HAART-associated metabolic syndrome 82 .It appears to be related to PI-induced lipodystrophy 83 and metabolic disorders, especially to elevated fasting triglycerides and insulin resistance 82,84 .

HaarT-associated coronary artery disease
HIV-infected patients receiving HAART with pre existing additional risk factors (e.g.hypertension, diabetes or increased plasma homocysteine levels) might be at raised risk of developing coronary artery disease because of accelerated atherosclerosis.Conflicting data exist, however, on the relationship between HAART and incidence of acute coronary syndromes, such as unstable angina or myocardial infarction, among HIV-infected patients receiving PI-containing HAART [85][86][87][88][89] .Differences in the study design, selection of patients and statistical analyses might explain this disparity.However, longer exposure to HAART and/ or PI seem to increase the risk of myocardial infarction.Results of the Data Collection on Adverse Events of Anti-HIV Drugs study showed that HAART therapy is associated with a 26% relative risk increase in the rate of myocardial infarction per year of HAART exposure 90 .

HaarT-associated peripheral vascular disease
The issue of surrogate markers of subclinical atherosclerosis has also been addressed.A study was performed on a cohort of 168 HIV-infected patients to measure the intima-media thickness (IMT) and assess indirectly the cardiovascular risk.In this population a high prevalence of atherosclerotic plaques within the femoral or carotid arteries was observed, but their presence was not associated with use of PI 91 .Similar results were also reported by Hsue et al. 92 and by Currier et al. 93 in case-control studies suggesting that traditional risk factors may contribute to atherosclerosis in HIV-infected patients independently of PI exposure.Alonso-Villaverde et al. reported that HIV-infected patients with subclinical atherosclerosis have higher circulating levels of monocyte chemoattractant protein-1 (MCP-1), especially of the allele MCP-1-2518G, compared to patients without atherosclerotic lesions, independently of HAART regimen 94 .Different results were reported by Maggi et al. who observed a higher than expected prevalence of premature carotid lesions in PI-treated patients compared to PI-naive patients 95 .Similar results have been reported in a study by Jerico et al. in 68 HIV-infected patients 96 .These authors conclude, stating that HAART should be considered a strong, independent predictor for the development of subclinical atherosclerosis in HIV-infected patients, regardless of known major cardiovascular risk factors and atherogenic metabolic abnormalities induced by this therapy 96 .The impact of individual measures to reduce cardiovascular risk and progression of atherosclerosis has been addressed by Thiebaut et al. 97 .According to these authors, the increased use of lipid-lowering agents, of PI-free HAART regimes and reduction of smoking may decrease the IMT in HIV-infected patients over time 97 .In spite of these different results, markers of subclinical atherosclerosis should be carefully assessed in HIV-infected patients receiving HAART, especially in those with lipodystrophy.

New markers defining cardiovascular risk in patients receiving HaarT
New insights in defining the cardiometabolic risk in patients with HAART-associated metabolic syndrome have been provided recently by the echocardiographic measurement of the epicardial adipose tissue.Epicardial adipose tissue is the true visceral fat of the heart and is significantly correlated with abdominal visceral fat measured by MRI 98 .Given its potential as an easy and reliable marker of visceral fat, epicardial fat has been recently evaluated in patients with HIV-lipodystrophy syndrome 98 .In these patients, echocardiographic epicardial fat correlated with intra-abdominal visceral fat, carotid IMT, and clinical parameters of the metabolic syndrome (especially waist circumference, blood pressure, fasting glucose, and insulin and markers of fatty liver) [98][99][100] .Taken together, these findings suggest that echocardiographic assessment of epicardial fat may have the potential to be a simple and reliable marker of visceral adiposity and increased cardiovascular risk in patients with HIV-lipodystrophy syndrome 98,99 .

ConClusion
Cardiac and pulmonary complications of HIV disease are generally late manifestations and may be related to prolonged effects of immunosuppression and a complex interplay of mediator effects from opportunistic infections, viral infections, autoimmune response to viral infection, drug-related cardiotoxicity, nutritional deficiencies, and prolonged immunosuppression.In developed countries HAART has significantly reduced prevalence of HIV-associated cardiomyopathy which heavily influenced the prognosis of HIV-infected patients in the pre-HAART period.However, HAART-associated metabolic syndrome is an increasingly recognized clinical entity.The atherogenic effects of PI-including HAART may synergistically promote acceleration of coronary and cerebrovascular disease and increase the risk of death from myocardial infarction and stroke even in young HIV-infected people.A better understanding of the molecular mechanisms responsible for this syndrome will lead to discovery of new drugs that will reduce the cardiovascular risk in HIV-infected patients receiving HAART.
* This article is a revised and updated version of the chapter: Barbaro G. "Pathogenesis of cardiovascular complications in the acquired immunodeficiency syndrome".In: Watson RR, Larson DF, editors.Immune dysfunction and immunotherapy in heart disease.Blackwell Futura; 2007:135-51.