The arrival of effective antiretroviral therapy (ART) has meant that there are many people with human immunodeficiency virus (HIV) infection living to an advanced age. Therefore diseases more associated with the middle and older years will achieve increasing importance as comorbidities in patients with HIV infection. It has been estimated that, globally by 2030, the leading cause of death will be ischaemic heart disease with HIV/ acquired immune deficiency syndrome (AIDS) in third place (in low income countries these diagnoses will fill first and second places respectively). Therefore, these conditions represent a significant intersection of epidemics.
Cardiovascular disease (CVD) and its prevention are very important considerations in the overall management of patients with HIV, especially as both HIV infection and its treatment are implicated in the pathogenesis. As data from the START trial will now encourage earlier commencement of ART, patients are likely to be exposed to these drugs for longer periods of time and as a result may increasingly acquire cardiovascular complications. It can be confusing to read that ART is beneficial for cardiovascular health in the short term and then see data implicating ART as a threat to cardiovascular health over a longer period. Clearly ART is very good for overall health and wellbeing and CVD should not be considered as a contraindication to HIV treatment. However CVD may influence the choice of ART in certain high-risk individuals and all patients should be monitored for the more traditional CVD risk factors and undertake strategies to minimise their effect.
In the time before effective ART there were a number of cardiac and vascular conditions more commonly seen in patients with HIV than in those without the infection from similar backgrounds; these conditions included dilated cardiomyopathy, pulmonary hypertension and pericardial effusions. They are seen less frequently since the arrival of effective ART. Nonetheless these diagnoses are important to consider in patients with appropriate physical signs. Arrhythmias occur often in the in the setting of CVD and should be assessed and treated accordingly, additionally cardiac electrical dysfunction may be precipitated by ART agents and their interactions with other medications.
Cardiovascular disease resulting from HIV and its treatment
A diagnosis of HIV infection can influence the development of CVD in a number of ways (Table 1).
The virus precipitates a chronic inflammatory condition that damages the coronary artery endothelium producing a number of events that might accelerate vessel occlusion through atherosclerotic plaque formation or the creation of a prothrombotic state. There are many inflammatory and prothrombotic molecules implicated in this process, e.g. interleukin-6, C-reactive protein (CRP) and D-dimer and these have been directly linked to clinical risk of CVD. In addition, activation of monocytes and macrophages has been associated with increased arterial inflammation in individuals with HIV infection when compared to uninfected controls.
Control of HIV infection by ART is the most important HIV-related factor in prevention of myocardial infarction. The SMART trial has showed that continuous therapy (viral suppression arm) was superior to interrupted therapy (drug conservation arm) in regards to protection against death, major opportunistic infections and other morbidity including CVD. However, treatment for HIV takes its own toll on the cardiovascular system (Table 2).
Early data from the Data Collection on Adverse Events of Anti-HIV Drugs (D:A:D) Study revealed that, while the absolute rate of myocardial infarction was low, there was a significant increase in risk in the first few years of ART use. A number of cohort studies have shown a negative effect of ART on cardiovascular health by increasing the risk of cardiovascular events in subjects with HIV on ART compared with controls without HIV. In a Danish population cohort study including nearly 4000 patients with HIV, before initiating ART there was a small (though non-significant) increase in the relative risk (RR) of suffering an initial ischaemic event requiring hospitalisation (RR=1.39). However, after commencing ART, the adjusted RR rose to 2 which was statistically significant. Other cohort studies have shown similar effects. Islam et al. published a meta-analysis that ascribed a 52% increase in RR for a CVD-related event in patients with HIV on ART versus those that were treatment naïve. Increase CVD risk has been seen in treated over untreated patients with HIV infection as well, although in Currier’s 2003 study this risk was only significant in the 18-33-year-old age group. In addition asymptomatic disease has been found to be prevalent in HIV patients receiving ART, and its severity increases with the duration of use of that therapy.
It has emerged that some antiretroviral drugs are particularly associated with cardiac events. In the D:A:D study it was shown that the use of abacavir in the previous 6 months resulted in a RR of 1.90 for a myocardial infarct compared with comparator antiretroviral drugs. In the same study a smaller but still significant risk (RR=1.49) was noted for didanosine. There have been other studies to confirm this finding e,g. an Australian case-control study also found abacavir to be significantly associated with coronary heart disease, however others have not confirmed the correlation. A recent update of the D:A:D study data still found that current use of abacavir gave a 47% increased 5-year risk of cardiovascular disease. Other nucleoside reverse transcriptase inhibitors (NRTIs) e.g. zidovudine and older protease inhibitors (PIs) e.g. saquinavir may increase CVD risk via other mechanisms (see below) but newer agents seem not to have such a significant effect on cardiac health. Non-nucleoside reverse transcriptase inhibitors (NNRTIs) can precipitate dyslipidaemia though this is not a problem with nevirapine use. Integrase inhibitors and are often preferred for their low risk for adverse reactions including a better lipid and cardiovascular risk profile than most other classes of ART.
Lipids and HIV protease inhibitors
Early observations noted that patients with HIV without ART experienced reductions in total cholesterol and low-density lipoprotein (LDL) cholesterol. This reduction appeared beneficial except that a similar reduction of the more protective high-density lipoprotein (HDL) cholesterol was also measurable, and, additionally, as the infection progressed, serum triglyceride levels were noted to rise. Early treatment success with HIV PIs such as saquinavir, nelfinavir, lopinavir and ritonavir was tempered by the sometimes spectacular adverse effect on lipids, both cholesterol and triglycerides. Newer PIs such as atazanavir have been developed with a view to better lipid side-effect profiles. Ritonavir is now only recommended as the booster in combinations with other PIs but it should be noted that this small dose still has the potential to cause elevated lipids.
The treatment of hyperlipidaemia by HMG CoA reductase inhibitors (statins) has been hampered by the inhibition of the CYP3A4 enzyme by PIs. The metabolism of simvastatin is particularly affected and can result in severe rhabdomyolysis. Atorvastatin is only partially reliant on the enzyme so can be used cautiously at reduced doses perhaps starting at only 10 mg daily. Similar caution may be advisable for rosuvastatin. Pravastatin has been favoured as it does not interact with PIs but is not as effective as other statins in some patients. A newer agent, pitavastatin, has been approved for use and has few negative interactions with ART.  Sometimes multiple drugs directed at both cholesterol (e.g. ezetimibe) and triglycerides (fibrates, sometimes even nicotinic acid) control will be required. In some patients a multidisciplinary approach can be very helpful. Referral to specialised services including physiotherapy, dietetics and endocrinology may result in an acceptable plan for CVD risk reduction.
The metabolic syndrome is a complex multifactorial interplay of host, viral and treatment factors and is a very important consideration in the management of CVD. It is not fully understood but is certainly influenced by chronic inflammation and immune activation, which are manifestations of HIV-induced disease. Metabolic syndrome effects include adipose tissue dysfunction, disruption of insulin-glucose homeostasis and dyslipidaemia. A sometimes very noticeable manifestation is the lipoatrophy/lipodystrophy syndrome that can result in abnormal subcutaneous fat distribution and elevated serum lipid profiles. Thymidine analogue NRTI agents (zidovudine and stavudine) are particularly associated with lipoatrophy causing loss of subcutaneous fat from the cheeks, buttocks, arms and legs, which is thought to be due to mitochondrial toxicity. Lipodystrophy and lipohypertrophy (including the deposition of excess fat as a ‘buffalo hump’) may result from the use of several different agents, including PIs, some NNRTIs and several of the other NRTIs.
Approach to cardiovascular disease risks
The use of ART has been overwhelmingly beneficial for people living with HIV/AIDS. More and more people now have access to ART and The Joint United Nations Programme on HIV and AIDS (UNAIDS) global target for 2015 is for 15 million people to be taking ART. The START trial’s premature cessation will encourage earlier use of ART resulting in longer drug exposure for many people. This longer drug use will prolong life, but potentially increase the chance of cardiovascular morbidity. In patients with severe CVD risk factors some groups have recommended avoiding initial ART combinations that include certain agents. The British HIV Association guidelines uses the D:A:D study and French Hospital Database cohort data to suggest avoiding abacavir and some of the more lipid-elevating PIs such as lopinavir/ritonavir and fosamprenavir/ritonavir in patients with high CVD risk.
Given that treatment is beneficial but not without CVD risks, it is paramount to exercise a holistic management approach to address and minimise all CVD possible risks. Smoking in particular must be targeted. Patients with HIV infection are more likely to be tobacco consumers compared with their matched HIV-negative counterparts and they should be encouraged to stop – regardless of how many times they have tried before. Control of diabetes mellitus, hypertension, lipids and obesity will help. Regular checks and sometimes aggressive medical management can achieve significant reductions in CVD risk. Increased physical activity has an important part to play and should be encouraged.
It is important to note that co-infection with hepatitis C virus (HCV) can exacerbate CVD in people with HIV infection and control of this co-infection may affect CVD. Gender and family history are risk factors over which there can be little control, but their effects should give pause for thought. It has long been known that males have an increased absolute risk of ischaemic heart disease compared to age-matched females. However, in the setting of patients with HIV infection, women have a much greater relative risk of CVD when taking ART than their male counterparts. Women are often under-represented in trials and studies related to HIV and gender-specific questions should be addressed in future research.
While the calculator of cardiovascular risk that emerged from the Framingham cohort studies has not been formally validated in the setting of HIV it can be very useful to both monitor and educate patients, and encourage them to see where risk reduction might be achieved.
Pulmonary hypertension in the setting of HIV infection recently acquired its own specific category in the Updated Classification of Pulmonary Hypertension. It has been consistently found in 0.5% of people with HIV infection, a figure unchanged by the widespread use of effective ART and represents a rate about a thousand times greater than that of idiopathic pulmonary hypertension as reported in the general population. The pathogenesis is thought to be similar to that of CVD including both direct viral effects and a reaction to the inflammatory response that ensues, as well as aberrant cell responses, co-infection and comorbidities (e.g. HCV infection, systemic hypertension, diabetes) and the effects of ART (e.g. dyslipidaemia). Pulmonary hypertension is often best managed by a specialist and calcium channel blockers, sildenafil, diuretics and anticoagulation may all have some use). In one study of patients with HIV with pulmonary hypertension, long-term benefit was demonstrated with bosentan use.
Pericardial effusion and cardiomyopathy
Early in the HIV epidemic, high rates (11%) of pericardial effusion were noted in patients with HIV infection. With the advent of ART the condition is much less common and in one study of 802 patients (of which 85% were taking ART), only two patients (0.25%) were found to have echocardiographic evidence of a pericardial effusion and in neither case was the cardiac function compromised. However in Africa, pericardial effusions are likely to be the first manifestation of HIV/AIDS-related cardiac disease, often large enough to cause tamponade. Up to 70% can be due to Mycobacterium tuberculosis infection.
Cardiomyopathy in HIV is thought to be multifactorial, possibly resulting from
- direct infection with the virus (with or without myocarditis),
- release of inflammatory cytokines,
- opportunistic infections including cardiotropic viruses,
- nutritional disorders and
- drug toxicity.
Data from early in the HIV epidemic suggested that cardiomyopathy (defined as systolic dysfunction usually assessed by the ejection fraction) was much more common than at the present time. A meta-analysis of more than 2200 asymptomatic patients well-controlled on ART put the rate of systolic dysfunction at 8.3%, but diastolic dysfunction was present in 43.4%. Systolic dysfunction was associated with high CRP levels, smoking and previous myocardial infarction, while diastolic dysfunction was more common in older patients and those with hypertension.
Despite the reductions in both pericardial effusions and systolic dysfunction in HIV patients receiving ART, these conditions should still be considered in the setting of HIV infection when someone presents with dyspnoea, cardiomegaly and a raised jugular venous pressure, especially when CD4 T-lymphocyte (CD4) cell counts are low.
Optimising HIV viral load control should be achieved where possible with the addition of standard measures for the management of pericardial effusion and cardiomyopathy as appropriate. When end-stage cardiac failure results from any cause, consideration for cardiac transplantation should be entertained, as HIV infection is not a contraindication.
Malignant conditions of the heart are rare but have been seen in patients with HIV particularly in those with more advanced immunosuppression. Non-Hodgkin lymphoma and Kaposi sarcoma have both been reported to affect the heart and the effect of ART has been reported to be profound on the latter reducing the prevalence of Kaposi sarcoma from up to 28% (on autopsy studies) to being largely unseen.
HIV is not seen as a particular risk factor for infective endocarditis (IE), however it can occur and in a similar way to the population without HIV infection. It is more frequently seen in those people who inject drugs (though the advent of ART does not seem to have had a large impact on prevalence of IE in this particular population). However in a US nationwide study of IE, the prevalence of HIV infection among IE patients fell from 4.8% in 1998 to 1.5% in 2009 for reasons that are not entirely clear.
In a large US study of more than 30,000 veterans, 2.6% developed atrial fibrillation over a median time of 6.8 years and was more common (and independent of other risk factors) in those with lower CD4 cell counts and higher HIV viral loads. As in all cases of atrial fibrillation, treatment is usually required and standard therapy with anti-arrhythmic agents such as beta-blockers or digoxin and, where appropriate, anticoagulation is recommended. There have been mixed reports of PIs having a detrimental effect on electrical conduction in the heart; some found QT prolongation and others did not. While it is not necessary to perform an ECG on everyone commencing ART, it would be advisable to do so in those with CVD risk factors both before commencing treatment and a few months afterwards to see if there is a discernable effect. Care should be taken with prescribing those drugs that may also exacerbate QT prolongation and arrhythmias in addition to ART, including macrolides, fluoroquinolones, triazole and imidazole antifungal agents, pentamidine and methadone.