Cardiac Involvement in HIV-infected Patients
Cardiac abnormalities were appreciated early in the epidemic of acquired immunodeficiency syndrome (AIDS), even before the etiologic agent, human immunodeficiency virus (HIV) was isolated and characterized. In one small autopsy series, major cardiac pathology was noted in 24 percent of cases [1]. Symptomatic cardiac disease was described first in children but is also common in HIV-infected adults [2,3]. Patients with AIDS can have cardiac pathology related to opportunistic infections and tumors. The etiology and pathogenesis of myocarditis and cardiomyopathy in association with HIV infection is still the subject of intense speculation [4].
Patients with HIV infection can have a variety of cardiovascular manifestations. The most common are:
• Pericarditis
• Pulmonary vascular disease and pulmonary hypertension
• Valvular disease
• Myocarditis
• Cardiomyopathy
• An increased incidence of coronary artery disease.
Most of these patients have AIDS. Many of the studies of cardiovascular disease in HIV-infected patients were performed before the availability of highly active antiretroviral therapy (HAART); thus, the applicability of some of the following observations to current practice is uncertain.
PERICARDITIS – The most frequent clinical manifestation of cardiovascular disease in patients with AIDS is pericarditis. Echocardiographic studies have identified a pericardial effusion in approximately 20 percent (range 10 to 40 percent) of these patients, and the effusion is large in 4 percent [5-9]. Patients who develop a pericardial effusion generally have a poor prognosis (see below).
One report, for example, summarized results of serial echocardiograms in 195 HIV-infected subjects (59 asymptomatic, 62 with early symptoms, and 74 with AIDS) [9]. Pericardial effusions were essentially limited to the patients with AIDS: they were present in 5 percent at study entry and developed during follow-up at a rate of 11 percent per year. Almost all of the effusions were small and asymptomatic; only two of the 195 HIV-infected patients developed clinical pericarditis.
Among all patients who present with pericardial effusion, the proportion associated with HIV infection will vary markedly according to the type of hospital. Among hospitals in which HIV-infected patients are commonly seen, HIV infection has been found in 7 percent of all pericardial effusions [10] and 28 to 33 percent of clinically significant pericardial effusions [11,12].
The picture of clinical pericarditis is similar to that of pericarditis from etiologies other than HIV. Some patients are symptomatic with fever, pleuritic chest pain, and, in some cases, a pericardial friction rub. (See "Evaluation and management of acute pericarditis"). However, most patients present with an asymptomatic increase in the cardiac silhouette on chest x-ray. In the experience at San Francisco General Hospital, about one-third of the symptomatic patients present with or develop signs of pericardial tamponade. Similar findings have been noted by others [12].
In spite of cultures and microscopic examination of the pericardial fluid and tissue, the etiology of the pericarditis is not usually definable [10,12-14]. Mycobacterium tuberculosis, atypical mycobacteria, fungi, Staphylococcus aureus and Streptococcus pneumoniae have all been described, as well as more unusual pathogens such as Nocardia asteroides, Listeria monocytogenes, Rhodococcus equi and Chlamydia trachomatis. Neoplasms, particularly lymphoma and Kaposi's sarcoma, also can cause pericardial effusion.
The relative frequency of these disorders can be illustrated by the following observations:
• In a series of 14 hospitalized patients with pericardial effusion (10 of which were small), a probable etiology could only be established in four (two with endocarditis, and one each with lymphoma and myocardial infarction) [10].
• In a report of 29 patients with AIDS who underwent a pericardial window for pericardial effusion, culture and biopsy specimens established a diagnosis in only seven: (three with lymphoma, two with adenocarcinoma, and one each with tuberculosis and S. aureus) [13].
• In a review of 74 reported cases of cardiac tamponade in HIV-infected patients, the major causes were idiopathic (45 percent), mycobacteria (20 percent), bacteria (19 percent), lymphoma (7 percent), Kaposi's sarcoma (5 percent), viruses (3 percent), and fungi (1 percent) [12].
Tuberculous pericarditis may be difficult to diagnose in patients with AIDS. Cutaneous anergy is common in this setting (61 percent in one series) [15]; as a result, a negative tuberculin skin test does not exclude the diagnosis. Pericardial biopsy is more sensitive than smears and culture of pericardial fluid but may not be positive for caseating granulomata. (See "Tuberculous pericarditis").
Treatment – The evaluation and therapy of pericardial effusion in HIV-infected patients varies with the clinical setting. An asymptomatic, small pericardial effusion without tamponade often requires no further testing except for follow-up [8,16]. If, on the other hand, the patient is symptomatic and the effusion large, even without tamponade, pericardiocentesis is indicated to obtain cultures, cytology and chemistries.
If tamponade occurs, which, as noted above, occurs in as many as 33 to 40 percent of patients with clinical pericarditis [12], immediate drainage is necessary. A catheter is often left in the pericardial space to drain by underwater-seal suction for the next 48 hours. If the effusion returns, subxyphoid pericardiotomy with creation of a pericardial window can be performed [13]. An alternative with which there is much less experience is balloon pericardiotomy to rupture the pericardium into the left pleural cavity [17]. (See "Pericardial compressive syndromes").
It is generally recommended that patients with a large pericardial effusion in whom no etiology is established be treated empirically for M. tuberculosis. Patients with AIDS have responded successfully to conventional antituberculous therapy for tuberculous pericarditis [18,19]. Among all patients with tuberculous pericarditis, the addition of a one month course of prednisone followed by a slow taper is associated with more rapid recovery, lower rates of mortality, and a decrease in surgical intervention. (See "Tuberculous pericarditis"). It is not known if a similar benefit applies to patients with advanced HIV infection in whom such a regimen may increase the opportunistic infections [20].
Other causes of pericarditis in patients with AIDS should also be treated when identified. Bacterial or fungal infection should be treated with appropriate drugs. Pericarditis due to lymphoma has been treated with radiation therapy and chemotherapy [21,22]. However, responses are generally transient and combination chemotherapy is associated with a significant risk of early death due to opportunistic infection [21].
Prognosis – The development of pericardial effusion in a patient with AIDS is a bad prognostic sign, even if asymptomatic [9,16]. In the series of 195 patients described above, mortality in the patients with pericardial effusion was significantly higher than in the patients with AIDS who did not have an effusion (64 versus 7 percent at six months) [9]. The shortened survival remained significant (relative risk 2.2) after adjustment for lead time bias and risk factors. The pericardial effusions rarely directly contribute to mortality but rather serve as a marker of advanced HIV infection.
MYOCARDIAL DISEASE – There are three major forms of myocardial disease in patients with AIDS:
• Focal myocarditis, usually seen incidentally at autopsy
• Echocardiographic evidence of decreased left ventricular function
• Clinical cardiomyopathy.
Focal myocarditis – The prevalence of focal myocarditis depends upon how it is determined. Autopsy studies have found myocarditis in 9 to 52 percent of HIV-infected patients (average 33 percent) [14,23]. Nearly all had focal collections of mononuclear cells with or without evidence of myocardial necrosis; ventricular dilatation was occasionally seen [23]. Most of these patients had had no cardiovascular signs or symptoms; however, congestive heart failure can occur.
Among patients with myocarditis, opportunistic bacterial, fungal, and protozoan pathogens can be identified in 10 to 15 percent of cases. In one autopsy series, 12 percent of 182 patients had cardiac toxoplasmosis [24]. (See "Etiology and pathogenesis of myocarditis"). However, symptomatic disease appears to be rare. A French survey estimated the prevalence of clinical extracerebral toxoplasmosis at 1.5 to 2 percent; only 3 percent of these patients had cardiac disease [25].
Cytomegalovirus is another common opportunistic infection in patients with late stage AIDS that can cause myocarditis in selected patients [26]. Coxsackievirus infection is another possible etiology [27].
Role of direct HIV infection – The cause of the 80 to 85 percent of cases of apparently idiopathic myocarditis is uncertain. HIV itself may be responsible but other cardiotropic viruses, such as cytomegalovirus, coxsackievirus, or Epstein-Barr virus, also may be important. It is not clear if HIV-associated disease is due to direct myocardial infection, an autoimmune process induced by HIV or other cardiotropic viruses, or coexisting opportunistic infection. Since the myocardial cell lacks CD4 receptors, HIV is denied the usual mode of entrance into the cell. However, the organism could enter the cell if the cell were injured by another virus. In in vitro studies, for example, Epstein-Barr virus permitted the entrance and replication of HIV into CD4 receptor-negative myocardial cells [28]. In addition, HIV has been localized in or near the myocardial cell by immunocytochemical techniques [29,30] and in situ DNA hybridization [31,32]. However, HIV sequences also can be identified in HIV-infected patients without apparent cardiac disease.
Perhaps the best evidence for a role for HIV infection comes from a study of 952 asymptomatic HIV positive patients who were followed for 60 months [33]. A dilated cardiomyopathy was diagnosed on echocardiography in 8 percent, most of whom had evidence of myocarditis on biopsy; the mean annual incidence was 1.6 percent. Among the patients with dilated cardiomyopathy, 76 percent had HIV nucleic acid sequences present in myocytes, while 26 percent also had evidence of coxsackievirus, cytomegalovirus, or Epstein-Barr virus infection.
Subclinical and clinical dilated cardiomyopathy – Echocardiographic studies have identified cardiac abnormalities in 10 to 40 percent of patients with AIDS, with asymptomatic pericardial effusion and left ventricular dysfunction being the most common findings [5-7,16,34]. Clinically important cardiomyopathy is less common, but has been reported in up to 10 percent of patients [35-39]. These functional abnormalities can be transient [16,40-42].
The variability in the prevalence of cardiomyopathy in different studies may depend in part upon whether the patient is seen in a primary-care center or a tertiary referral center. This variance also depends upon whether or not the patient is an injection drug user (IDU). The range of findings can be illustrated by the following observations:
• In one series in which serial echocardiography was performed, global left ventricular hypokinesis was found in 10 of 69 (14 percent) asymptomatic HIV-infected patients [39]. Of the 59 patients with initially normal echoes, eleven developed hypokinesis over mean follow-up of 11 months (1.5 percent per month). It was estimated that the crude rate for overt cardiomyopathy was 2.1 percent in a clinic population of almost 2000 patients who were actively followed.
• A much lower incidence of echocardiographic dilated cardiomyopathy was noted in the series of 952 asymptomatic patients described above (1.6 percent per year) [33] and in another study of 136 HIV-infected patients without clinical, electrocardiographic or echocardiographic evidence of cardiovascular dysfunction on admission who were prospectively studied with serial echocardiograms; 93 of the patients had AIDS [36]. During a mean follow-up period of 415 days, seven patients, all in the AIDS subgroup, developed clinical and echocardiographic findings of acute global left ventricular dysfunction; six of these seven patients died of congestive heart failure. Necropsy findings in five of these patients revealed acute lymphocytic myocarditis in three, cryptococcal myocarditis in one, and interstitial edema and fibrosis in one.
Summarizing the longitudinal studies, clinical dilated cardiomyopathy is seen in approximately 1 to 3 percent of patients with AIDS.
Possible etiology – Some patients with asymptomatic or overt left ventricular dysfunction have a known etiology such as toxoplasmosis, cryptococcosis, cocaine use, toxo-serology hypertrophic cardiomyopathy, alcoholic heart disease, or drug toxicity. The diagnostic evaluation should include Toxoplasma serology cryptococcus neoformins serum antigen assay and review of drugs. In addition, as noted in the section on focal myocarditis, HIV infection itself or infection with coxsackievirus, cytomegalovirus, and Epstein-Barr virus may be responsible in selected patients [33].
A variety of other factors may contribute to the development of myocardial dysfunction. These include:
• Postviral autoimmunity [43]
• Cachexia [44-46]
• Selenium deficiency [47,48]
• Cardiotoxicity from therapeutic drugs such as pentamidine [49] and possibly zidovudine [50]
• Cytokines released by HIV-infected macrophages such as tumor necrosis factor, interleukin (IL)-1, IL-2, and alpha interferon [51,52]. These cytokines increase the production of inducible nitric oxide synthase and nitric oxide; high levels of nitric oxide are cytotoxic to myocardial cells [53]. (See "Etiology and pathogenesis of myocarditis" and see "Other hormones; cytokines; and chemokines in congestive heart failure").
PULMONARY HYPERTENSION – Severe pulmonary hypertension, with and without cor pulmonale and right heart failure, have been described in patients with AIDS [54-58]. In one series of HIV-infected patients, pulmonary hypertension was present in 0.5 percent overall and 8 percent of those with cardiopulmonary complaints [55]. The etiology of this problem is not well understood. Many, but not all, of these patients have had multiple pulmonary infections, often with Pneumocystis carinii; and many were IDUs. As a result, interstitial disease and vascular bed destruction by multiple pulmonary infections were thought to be important.
However, some patients have never had a pulmonary infection, and the histology of the lung and the clinical pattern are similar to that seen in primary pulmonary hypertension [54-59]. (See "Pathophysiology and clinical aspects of primary pulmonary hypertension"). In a review of 88 reported cases, for example, lung tissue was evaluated microscopically in 33: 28 were of the plexogenic variant of primary pulmonary hypertension; three had thrombotic pulmonary arteriopathy (one due to recurrent thromboembolism, and two to in situ thrombosis); and two had pulmonary venoocclusive disease [54]. The mechanism by which primary pulmonary hypertension might occur in HIV-infected patients is not known.
Affected patients may present with several months of easy fatiguability and shortness of breath, or with frank right ventricular failure. The diagnosis is made by Doppler echocardiography which shows an enlarged and hypertrophied right ventricle and a possibly enlarged right atrium in the presence of a normal left ventricle (show echocardiogram 1 and show echocardiogram 2). The pulmonary artery systolic pressure can be estimated from the tricuspid regurgitation jet. (See "Pathophysiology and clinical aspects of primary pulmonary hypertension").
The HIV-infected patient with pulmonary hypertension has a poor prognosis (median survival 1.3 versus 2.6 years in those without pulmonary hypertension in one report) [59]. Antiretroviral therapy may be of some benefit [59].
CARDIAC TUMORS – One of the findings suggesting that HIV infection was a new, unusual disease was the appearance of Kaposi's sarcoma in young men, a disease seen previously almost exclusively in old men. (See "AIDS-related Kaposi's sarcoma: Epidemiology and pathogenesis"). Among patients with AIDS, Kaposi's sarcoma can involve the myocardium or the pericardium, and can cause pericardial effusion and, in some cases, tamponade [12,60-62].
Non-Hodgkin's lymphoma is a malignancy often seen in the patient with AIDS. The tumor is usually widespread, but can present as a primary cardiac lymphoma [63]. When lymphoma involves the heart, it is usually diffusely infiltrative, but can form nodules and even intracavitary masses [14,64-67]. Cardiac lymphoma can cause congestive heart failure, atrial and ventricular arrhythmias, and heart block. With intracavitary growth, the masses cause mechanical obstruction to the blood flow across the valves. In such cases, surgical resection is indicated. Chemotherapy and radiation therapy have produced variable results [64].
VALVULAR DISEASE – Three main types of valvular disease have been associated with HIV infection:
• Nonbacterial thrombotic endocarditis
• Infective endocarditis
• Mitral valve prolapse.
Nonbacterial thrombotic endocarditis – Nonbacterial thrombotic (marantic) endocarditis consists of sterile vegetations which can occur on any of the valves. This is a common finding at autopsy, especially in patients with wasting diseases. It may present as systemic embolization, but is otherwise clinically silent; no valvular disturbance occurs.
Infective endocarditis – Infective endocarditis in HIV-infected patients occurs almost exclusively in IDUs [7,68]. Among other HIV-infected patients, infective endocarditis occurs with about the same incidence as in the non-HIV population. The clinical manifestations, diagnosis, and therapy are similar to that of infective endocarditis in the general population. (See "Diagnostic approach to infective endocarditis" and see "Treatment of infective endocarditis").
METABOLIC ABNORMALITIES WITH HIV PROTEASE INHIBITORS – The protease inhibitors that are part of most HAART regimens are associated with several problems that can increase cardiovascular risk including diabetes mellitus, hypertriglyceridemia, and abnormal fat distribution. The following hypothesis has been proposed to explain this constellation of findings [69]. The catalytic region of HIV-1 protease, to which the drugs bind, has 60 percent homology to regions of two proteins that regulate lipid metabolism: cytoplasmic retinoic acid-binding protein type I (CRAB P1) and low-density-lipoprotein receptor-related protein (LRP). Protease inhibitor binding to LRP impairs hepatic chylomicron uptake and triglyceride clearance by the LRP-lipoprotein lipase complex. The ensuing hypertriglyceridemia contributes to central fat deposition, insulin resistance, and, in susceptible subjects type 2 diabetes.
Diabetes mellitus – New-onset hyperglycemia or worsening of preexisting diabetes has been described with an incidence ranging from less than 1 percent to 6 percent [70-72]. This complication is associated with all of the protease inhibitors, and can occasionally result in ketoacidosis. Affected patients have two characteristics similar to type 2 diabetes – insulin resistance and a positive response to sulfonylureas [69]. The protease inhibitor should not be stopped since the value of these drugs outweighs the risks of hyperglycemia.
Lipid disorders – Although hypertriglyceridemia can occur in the absence of protease inhibitors, extreme elevations in serum triglycerides to above 1000 mg/dL have been seen in some patients treated with these drugs [73,74]. Modest elevations in serum cholesterol also can occur [72]. In a study of ritonavir plus saquinavir, for example, increases in serum lipids were seen within two weeks of starting therapy: 11 percent developed serum triglyceride concentrations above 1500 mg/dL. There were no cases of pancreatitis.
It has been suggested that protease inhibitor therapy may be associated with premature coronary disease which may be due at least in part to the lipid abnormalities [75]. Risk assessment must consider other risks for cardiac disease such as smoking, obesity, family history, sedentary lifestyle, hypertension, etc. The hypertriglyceridemia can be treated with gemfibrozil and a statin can be given for concurrent hypercholesterolemia. There are, however, potential drug interactions with this approach: the statin with gemfibrozil (which may be minimized by the administration of pravastatin) and an interaction of protease inhibitors with some HMG-CoA reductase inhibitors due to metabolism by CYP3A4 (show table 1). (See "Drug-induced myopathies").
Abnormal fat distribution – An abnormal redistribution of fat from the periphery to the thorax and abdomen has been described. Various terms such as peripheral lipodystrophy, "protease paunch" and "buffalo hump" have been employed to describe abnormal accumulations of fat in the posterior neck, upper back, breasts and abdomen associated with protease-inhibitor drugs [72,76]. This complication seems to be worse in patients on ritonavir-saquinavir combinations than with indinavir. The abnormal distribution of fat does not respond to limitation of caloric intake and exercise.
CLINICAL APPROACH – The history and physical examination are the best screens for cardiovascular involvement in HIV-infected patients, looking for signs and symptoms of possible cardiovascular illness. There appears to be no advantage in finding subclinical disease; thus, routine echocardiography is not indicated.
The clinical presentation of cardiomyopathy is similar in HIV-infected patients as in those who are not infected. The initial symptoms are nonspecific: as an example, dyspnea may be due to pulmonary involvement or cardiac disease. Echocardiography can help differentiate between these possibilities.
More specific signs of cardiac involvement are an S3 gallop, a pericardial friction rub, pulmonary edema or a pathologic murmur. Right-sided congestive heart failure can be due to primary pulmonary hypertension. When these findings are present, transthoracic echocardiogram can demonstrate pericardial effusion, valvular insufficiency, and regional hypokinesis affecting the left ventricle. Invasive catheterization is not necessary, unless coronary artery disease has to be excluded.
Myocardial biopsy is recommended by some physicians, but its utility is controversial. The finding of a treatable etiology of myocarditis is rare. At present, there is no evidence that steroids or immunosuppressive agents are more effective than conventional management in patients with myocarditis. (See "Natural history and therapy of myocarditis").
The treatment of the AIDS patient with congestive heart failure is similar to that of other patients, consisting of angiotensin converting enzyme inhibitors, digoxin, and diuretics for fluid control. In addition, drugs that have been implicated in cardiomyopathy should be discontinued and a possibly causal infection (such as toxoplasmosis) should be treated if present.
References : UpToDate Vol 8 No1
