1. Journal of the American College of Cardiology Vol. 47, No. 6, 2006
© 2006 by the American College of Cardiology Foundation ISSN 0735-1097/06/$32.00
Published by Elsevier Inc. doi:10.1016/j.jacc.2005.12.018
EDITORIAL COMMENT assessment of cardiovascular risk factors, a lack of medica-
tion compliance data, and problems with ascertainment of
Cardiovascular Risk case status (1,2). Despite these limitations, these studies
in Patients With Human (with one exception) have suggested that patients with HIV
infection are at increased cardiovascular risk, and that use of
Immunodeficiency Virus Infection protease inhibitors may be especially disadvantageous from
a cardiac standpoint. Although two prospective observa-
Incomplete Data* tional studies also showed increased risk of myocardial
James H. Stein, MD infarction in patients receiving HAART, they also suffered
Madison, Wisconsin from low cardiovascular event rates, and concerns have been
raised about problems with case ascertainment and
Since the advent of highly active antiretroviral therapy between-group differences in the duration of HIV infection
(HAART) in 1996, the death rate from acquired immune (2).
deficiency syndrome (AIDS) in the U.S. has decreased Because of these limitations, many investigators have
dramatically. Because of HAART, it is not uncommon to used surrogate imaging markers of atherosclerosis to evalu-
meet patients who have lived with human immunodefi- ate cardiovascular risk. With one exception, these studies
ciency virus (HIV) infection for well over two decades. have demonstrated that use of HAART is associated with
Shortly after its introduction, however, clinicians observed endothelial dysfunction, but each study was small and
an increasing prevalence of a lipodystrophic syndrome cross-sectional, with attendant biases and limitations. In
among patients taking HAART that included hyperlipid- these studies, atherogenic lipoproteins, markers of insulin
emia, impaired glucose metabolism, and central adiposity resistance, and in some instances markers of virological
(1,2). Although HAART-associated lipodystrophy shares control explained part of the differences in endothelial
similarities with the more common insulin resistance syn- function (2). These observations have been corroborated by
drome that predisposes patients to coronary artery disease studies demonstrating that protease inhibitors damage en-
dothelial cell mitochondrial deoxyribonucleic acid in cul-
See page 1117 ture, impair vasomotor function, and reduce endothelial
nitric oxide synthase expression by coronary artery rings
(3,4). Several studies also have shown an increased preva-
(“Metabolic Syndrome,” MetS), they clearly are different
lence of carotid plaque or intima-media thickness (CIMT)
disease processes. For example, an important component of
among patients receiving protease inhibitors, but they also
HAART-associated lipodystrophy is peripheral fat wasting
were small and cross-sectional.
(lipoatrophy), which is not seen in patients with coronary
The study by van Wijk et al. (5) adds to the growing
artery disease. It also is not clear if HAART-associated
literature suggesting that individuals receiving HAART
lipodystrophy carries the same cardiovascular risk as MetS.
have markers of vascular disease that are associated with
To complicate cardiovascular risk assessment further, un-
increased cardiovascular risk. In this study, individuals on
treated HIV infection also has been associated with dysli-
HAART had impaired flow-mediated vasodilation of the
poproteinemia and premature atherosclerosis (1). Thus, the
brachial artery, a marker of endothelial dysfunction. This
evaluation of cardiovascular risk in patients with HIV
agrees with previous literature, but is not conclusive because
hinges upon a complex interplay of direct and indirect
nitroglycerin-mediated vasodilation was impaired; making
vascular effects of HIV infection, antiretroviral therapy,
it unclear if the differences between the groups were due to
aging, and exposure to cardiovascular risk factors.
endothelial dysfunction or a non-endothelium-mediated
In the past five years, several cross-sectional and obser-
process.
vational studies have tried to provide clinical and patho-
A very interesting and important aspect of this study was
physiological insights into these interrelationships. Unfor-
classifying patients according to the presence of MetS, using
tunately, the observational studies that have been published
the National Cholesterol Education Adult Treatment Panel
or reported at meetings have had significant limitations,
III guidelines. Although this definition was not intended to
including low rates of adverse cardiovascular events, a short
be used for patients with lipodystrophy, HIV-infected
duration of exposure to HAART, and the usual limitations
patients with MetS had increased levels of apolipoprotein
associated with retrospective studies such as nonsystematic
B-100, C-reactive protein, insulin, and 2-h insulin com-
pared with HIV-infected patients without MetS and HIV-
*Editorials published in the Journal of the American College of Cardiology reflect the
views of the authors and do not necessarily represent the views of JACC or the negative controls. These findings suggest that HIV-infected
American College of Cardiology. patients with MetS have markers of insulin resistance and
From the Division of Cardiovascular Medicine, University of Wisconsin Medical increased cardiovascular risk (5). Indeed, the major strength
School, Madison, Wisconsin. Dr. Stein is the recipient of a research grant from
Bristol-Myers Squibb. He has served as a consultant to Abbott Labs and is on the of this report is the completeness of imaging and metabolic
Speaker’s Bureau for Merck (not related to HIV treatment). data. A study with this complete a dataset regarding

2. JACC Vol. 47, No. 6, 2006 Stein 1125
March 21, 2006:1124–5 Editorial Comment
HAART, metabolic risk factors, and cardiovascular imaging carefully matched triads based on HIV serostatus, antiret-
in patients with HIV and control subjects has not been roviral therapy, and cardiovascular risk factors. At baseline,
reported previously, so it is unique in that regard. there were no differences in CIMT between the groups of
The finding that CIMT was greater among HIV- HIV-infected patients (7). Three-year follow-up of changes
infected patients with MetS is consistent with the metabolic in CIMT, risk factors, and virological markers are expected
data provided by the authors. This finding would have been in early 2006. These studies will help clarify the major
more powerful if it had been corroborated by the aortic contributors to cardiovascular risk in patients with HIV
pulse wave velocity data; however, the sample size was infection.
relatively small. Also, certain markers of MetS in HIV- Conclusions. In regard to cardiovascular risk in patients
infected patients did predict increased pulse wave velocity, with HIV, we have incomplete data. Compared with the
such as blood pressure, 2-h glucose, and fasting insulin, high death rate from AIDS in patients with inadequate viral
suggesting that insulin resistance may be involved in the suppression, cardiovascular event rates are low and control
pathophysiology of vascular dysfunction in patients on of viremia, regardless of the treatment strategy, is more
HAART. It is interesting that markers of inflammation and important for long-term survival than any increase in
viral load were more associated with CIMT than flow- cardiovascular risk that may be related to metabolic changes
mediated vasodilation or pulse wave velocity, but these associated with HAART. Uncontrolled viremia may be
associations are somewhat difficult to interpret in the more of a cardiovascular risk than controlled infection that
absence of information about the specific antiretroviral results in hyperlipidemia and insulin resistance. The overall
agents used by the subjects, because agents within classes message is that obtaining and maintaining virological con-
have different effects on insulin-glucose metabolism and trol is the overriding concern in patients with HIV infec-
lipoproteins (1). Although the imaging data in this study do tion. Metabolic and vascular effects secondary to HAART are
not conclusively demonstrate that MetS, as traditionally secondary considerations, but are worthy of rigorous investiga-
defined, is useful for identifying increased cardiovascular tion, as suggested by the paper by van Wijk et al. (5).
risk in HIV-infected patients, the metabolic abnormalities
observed in MetS patients suggest that insulin resistance Reprint requests and correspondence: Dr. James H. Stein,
may be involved and that further studies are needed (5). University of Wisconsin Medical School, Cardiology, G7/341
Future directions. To better understand cardiovascular CSC, MC 3248, 600 Highland Avenue, Madison, Wisconsin
53792. E-mail: jhs@medicine.wisc.edu.
risk in patients on HAART, larger, prospective studies are
needed. The biases inherent in observational and cross-
REFERENCES
sectional studies are well known, and their results can be
misleading. In designing prospective studies, longitudinal 1. Dubé MP, Stein JH, Aberg JA, et al., for the Adult ACTG Cardio-
data regarding changes in structural and functional markers vascular Disease Subcommittee. Guidelines for the evaluation and
management of dyslipidemia in HIV-infected adults receiving antiret-
of vascular disease, risk factors, antiretroviral therapy, and roviral therapy: recommendations of the HIV Medical Association of
virological control must be considered, as well as the effect the Infectious Disease Society of America and the Adult AIDS Clinical
of aging itself. Trials Group. Clin Infect Dis 2003;37:613–27.
2. Stein JH. Managing cardiovascular risk in patients with HIV infection.
Two well-designed studies sponsored by the AIDS Clin- J Acquir Immune Defic Syndr 2005;38:115–23.
ical Trial Group (ACTG) recently have been completed. 3. Fu W, Chai H, Yao Q, Chen C. Effects of HIV protease inhibitor
ACTG 5152s is a prospective study of 82 HIV-positive, ritonavir on vasomotor function and endothelial nitric oxide synthase
expression. J Acquir Immune Defic Syndr 2005;39:152– 8.
treatment-naïve patients randomized to receive one of three 4. Zhong DS, Lu XH, Conklin BS, et al. HIV protease inhibitor ritonavir
HAART regimens that would be predicted to have very induces cytotoxicity of human endothelial cells. Arterioscler Thromb
different effects on lipids and insulin-glucose metabolism. Vasc Biol 2002;22:1560 – 6.
5. van Wijk JPH, de Koning EJP, Castro Cabezas M, et al. Functional and
As expected, flow-mediated dilation was impaired before structural markers of atherosclerosis in human immunodeficiency virus-
starting treatment, but improved significantly and to a infected patients. J Am Coll Cardiol 2006;47:1117–23.
similar degree in each arm after as few as four weeks of 6. Stein JH, Cotter BR, Parker RA, et al. Antiretroviral therapy improves
endothelial function in individuals with human immunodeficiency virus
HAART (6). Nitroglycerin-mediated vasodilation did not infection: a prospective, randomized multicenter trial (Adult AIDS
change, indicating that use of HAART rapidly improved Clinical Trials Group Study A5152s) (abstr). Circulation 2005;112:
endothelial function and that cardiovascular risk may de- II237.
7. Currier JS, Kendall MA, Zackin R, et al. Carotid artery intima-media
crease as HIV infection comes under control. ACTG 5078 thickness and HIV infection: traditional risk factors overshadow impact
is a prospective cohort study of 134 subjects recruited into of protease inhibitor exposure. AIDS 2005;19:927–33.