Inflammation and Fibrosis in HIV: Getting to the Heart of the Matter

The treatment and disease course of HIV infection have changed dramatically since the discovery of the HIV in 1983. Acquired immunodeficiency syndrome (AIDS), once the axiomatic cause of early death in HIV infected individuals, has now become an uncommon complication among individuals treated with antiretroviral therapy (ART). However, with the the success of ART comes new challenges for the HIV-infected, namely, a substantial burden of cardiovascular disease. HIV infection is associated with high rates of cardiovascular disease (CVD) complications, including acute myocardial infarction¹, sudden cardiac death², and heart failure³.

The mechanisms leading to CVD in HIV remain incompletely understood. HIV-infected individuals often have a greater burden of traditional cardiac risk factors, though such traditional risk factors alone do not nearly account for the observed increased risk. A wealth of research suggests that, paradoxically, chronic up-regulation of inflammatory activity, which is present even among anti-retrovirally treated and suppressed individuals, may play an important role in predicting mortality, cardiovascular disease as well as other non-AIDS conditions^4,5,6. While the hallmark of untreated HIV infection is an immunodeficient state, treated, virally suppressed HIV disease is associated with immune activation^7-9. Numerous potential causes are thought to contribute to this immune activation, including: toxicity from ART, low-level viral replication, and disease-mediated shifts in inflammatory cell subsets, among other mechanisms. Delineation of the precise mechanisms underlying immune activation in HIV, as the driver of cardiovascular complications, is the focus of an increasing number of studies.

In line with the inflammatory hypothesis of CVD complications in HIV, several studies have clearly established the predictive value of inflammatory and coagulation biomarkers for CV events and mortality in HIV.^{10, 11} Recent imaging studies have added to our understanding of the relationship between immune activation and atherosclerotic inflammation in HIV. In studies that employed FDG-PET/CT, individuals with HIV have been shown to have higher levels of atherosclerotic inflammation compared to non-infected controls^{12, 13}. One such study showed that the atherosclerotic inflammation correlates with markers of monocyte activation. Subsequent studies showed an association between arterial inflammation and structural measures of coronary atherosclerosis, including the presence of high-risk morphological features ¹⁴. Further, among HIV-infected individuals, immune activation and chronic inflammation persist and are thought to lead to collagen deposition and fibrosis of lymphoid tissue;¹⁵ as fibrosis has been reported in other organs such as the liver¹⁶ as well as the myocardium. Myocardial fibrosis, which has been shown to be a consequence of several chronic inflammatory conditions as well as traditional cardiovascular risk factors¹⁷, is characterized by a accumulation of collagen and has been identified as a contributor to diastolic dysfunction, heart failure, and sudden cardiac death ¹⁸. Indeed, ART-treated HIV as associated with a greatly increased incidence of myocardial fibrosis¹⁹ as well as both systolic and diastolic LV dysfunction²⁰. However, the relationship between myocardial inflammation and fibrosis in HIV has been less well studied.

Cardiac magnetic resonance (CMR), can be used to measure many of the features needed to develop an improved understanding of myocardial fibrosis and inflammation in HIV. LV function and fibrosis are well-characterized by CMR. While CMR cannot directly measure myocardial inflammation, it can be used to build a case for its presence through circumstantial evidence. CMR can image several of the hallmark components of inflammation: hyperemia (EGE: early gadolinium enhancement), edema (T2 weighted imaging), and scar/necrosis (LGE: late gadolinium enhancement). Prior studies suggest that the diagnostic accuracy of acute myocarditis is improved when two out of the three criteria are present, also known as the “Lake Louise criteria”²¹. Accordingly, the two studies published in this issue of Circulation: CV imaging, which used CMR to study HIV-infected individuals, may help shed additional light on the relationship between HIV, inflammation, fibrosis, and cardiovascular disease.

In the study by Luetkens et al, 28 HIV-infected individuals with chronic HIV on antiretroviral medication with a HIV RNA < 200 copies/ml underwent CMR and were compared healthy controls. Authors found that HIV-infected individuals had lower ejection fraction and lower strain values. Of note, myocardial inflammation (native T1 relaxation times: 1128 msec vs. 1086msec, p=0.009; relative T2 signal intensity ratio:1.6 vs. 1.4, p=0.046, and early gadolinium enhancement ratio: 3.1 vs. 2.1) and myocardial fibrosis were elevated in among the HIV-infected individuals (82% vs. 27%, p<0.001) as compared to controls. The patchy pattern of myocardial fibrosis²² in HIV has been previously reported by several groups, {Holloway, 2013 #163} ¹⁹, who observe a pattern of fibrosis that is similar to that seen with other causes of myocarditis. Authors note that while the prevalence of fibrosis is higher in HIV, the difference in scar burden was low in both studies, namely 2-3%. Though the study is small and descriptive in nature, it is the first to report myocardial inflammation using CMR tissue characterization in HIV. While HIV has been described as a cause of myocarditis²³, detection of HIV in cardiac myocytes remains controversial.²⁴ it is likely that the cardiac fibrosis and myocardial inflammation reported in the study by Luetkens result from chronic inflammation (as opposed to a direct effect of HIV on the heart).

The study by Ntusi et al provides additional evidence that treated HIV is associated with chronic myocardial inflammatory changes, including chronic subclinical myocardial edema and a high incidence of pericardial effusions. The investigators extend their prior study²² and report that compared to controls, HIV-infected individuals had lower (though still normal) LVEF, higher myocardial mass, and lower peak diastolic strain rate, findings that are in-line with prior observations. Furthermore, the investigators make the novel observations that pericardial effusions and myocardial fibrosis are three to four times more common in the setting of HIV. No differences were seen between ART-treated vs. untreated individuals in CMR characteristics with the exception of higher LVEF (68% in ART treated vs. 64% in ART naïve, p=0.02). The study identified a modest negative correlation between LVEF and Cd4 count nadir (R=-0.23, p=0.03).On the other hand, no associations were observed between imaging findings and: HIV medication, HIV RNA levels, CD4 nadir or CD4 count.

The findings from both papers provide additional information to the field of HIV-associated cardiovascular disease by applying CMR imaging parameters to describe subtle differences in myocardial tissue which include myocardial edema, inflammation, fibrosis and pericardial effusions. Authors suggest that chronic inflammation in the setting of treated HIV likely underlies the findings which include structural and functional differences and high levels of fibrosis in the setting of treated HIV. As patients from both studies were relatively young and asymptomatic, it is remarkable that subtle differences in myocardial inflammation and edema compared to uninfected individuals were found. Both studies also confirm the previous reports of high rates of myocardial fibrosis in HIV. Similar findings have been reported in other chronic inflammatory conditions such as rheumatoid arthritis.^{25, 26}

Along with the CMR imaging findings come unanswered questions. First, authors note that the myocardial volumes, function and tissue characteristics overall are within the normal range among individuals with HIV, making the clinical impact of these findings unclear. Second, aside from hsCRP, no biomarkers or tissue samples which might shed some light on underlying mechanism of the CMR findings were performed. Similarly, aside from HIV infection in general, no definitive correlations between indices of HIV infection, chronic inflammation or antiretroviral therapy were uncovered in these studies. The two studies highlight the advantages of CMR in general over echocardiography with respect to reproducible measurements of ventricular volume and function²⁷, smaller sample size²⁸, and assessment of fibrosis, injury, and inflammation. Of note, markers of cardiac stress such as ST2 and cardiac inflammation, GDF-15 are associated with cardiac dysfunction and all-cause mortality in HIV²⁹.

Taken together these new reports extend prior observations and suggest that treated and suppressed HIV infection is associated with myocardial inflammation and fibrosis as well as LV dysfunction. These findings may represent important culprits responsible for the observed excess incidence of CV events seen in HIV infection– a hypothesis which will require larger studies, detailed biomarker analysis, and proof of concept clinical trials to definitively establish.