icon-folder.gif   Conference Reports for NATAP  
  10th Conference on Retroviruses and Opportunistic Infections
Boston, Mass, Feb 10-14, 2003
Back grey_arrow_rt.gif
Heart Disease in HIV: Retrovirus Report
  There were 4 studies reported back to back in an oral session at the Retrovirus Conference related to heart disease, HIV, and HAART. These were the 4 main studies at the Conference examining the risks for premature heart disease in HIV. In brief, the data presented from these studies raise concern about a premature risk for heart disease in HIV-infected individuals on HAART, but results are a bit mixed. Three studies reported find a risk, D:A:D, study the Hsue study, and the Hopkins study. The ACTG study does not find an increased risk but followed patients on HAART for just 4 years. Both the ACTG and Hsue studies examine carotid IMT (artery thickness) and find different results. Importantly, the risk for premature heart disease on average takes years to develop and the risks on average for people may take years before cardiovascular disease events occur. The risks may be increased for people with serious traditional risk factors and you should talk to your doctor about this. The D:A:D study ("Data Collection of Adverse Events of Anti-HIV Drugs") used a new acronym called CART (combination ART). This study started in 1999 and aims to assess whether exposure to CART is associated with an increased risk of myocardial infarction (MI). The objective is to assess the risk of MI in a population of HIV-infected individuals enrolled in this study and to examine the association of CART with MI (NNRTI or PI in combination with NRTIs); and to identify factors associated with increased risk. 23,468 patients from the USA, Europe, and Australia are participating in this study. The average age is 39; 24% are female; 45% MSM; 20% IVDU; 26% had prior AIDS; average CD4 count 418; 55% had undetectable HIV viral load.
--20% of patients had a family history of cardiovascular disease (CHD)
--a low percentage had previous CHD
--60% smoked ever
--about 5% had hypertension
--a low percentage had diabetes
--about 22% had >240 mg/dL (6.2 mmol/l) total cholesterol
--about 25% had HDL (good cholesterol) <35 mg/dl
--about 35% had triglycerides >204 mg/dl
So a good number of patients had risk factors for heart disease: smoking, elevated cholesterol, low HDL, high TG.
The definition of MI is as defined in the WHO MONICA study. The diagnostic algorithm includes information on: cardiac pain; cardiac enzymes; troponine; ECG changes; autopsy findings. Fatal and non-fatal (survival 28 days). Definite, possible, or unclassifiable.
126 patients experienced an MI. Incidence overall was 3,5 per 1000 person years (PY). 36 cases (28%) were fatal. 90% were in men. Average age was 48 (25-82). 55% were definitive, 29% possible, and 16% unclassifiable.
With <1 year on CART there were 9 MIs. With 1-2 years on CART there were 14. With 2-3 years on CART there were 22. With 3-4 years on CART there were 31. And with >4 years on CART there were 47 Mis.
CART was associated with a 26% increased risk of MI per year of exposure. Traditional risk factors associated with risk of MI were: age, gender, smoking, previous CHD. Other risk factors were total cholesterol, diabetes, and, and lipodystrophy. The authors concluded by emphasizing that the absolute risk of MI overall remains low (3.5 per 1000 PY) and should be balanced against the known beneficial effects of CART. Individuals already at high risk of CVD and on CART need special attention with regards to reducing modifiable risk factors. Further follow-up of these patients is required to monitor incidence trend over time, identify factors explaining association, and distinguish association with ART drug classes. The D:A:D study will continue at least until 2005.
Judith Currier reported for the ACTG on a study evaluating carotid IMT (intima thickness of the carotid artery) as measure of risk for coronary atherosclerosis and clinical cardiovascular disease. Currier reported that measurement of carotid IMT assessed with a non-invasive B-mode ultrasound technique, correlates with coronary artery atherosclerosis events and clinical cardiovascular events. Carotid IMT has been shown in 5 studies to be predictive od clinical cardiovascular events in both asymptomatic and symptomatic individuals. Prior studies of carotid IMT in patients with HIV-infection have produced conflicting results. In one study, PI-treated men had increased IMT compared to ARV-naive HIV-infected men and uninfected controls (Seminari E, et al. Artherosclerosis 2002; 162:433-438). It is critical that studies of artherosclerosis in HIV-infection control for confounders that increase cardiovascular risk: age, blood pressure, sex, family history, and lipids).
This is a prospective, longitudinal, matched cohrt study. Subclinical artherosclerosis is assessed by the measurement of carotid IMT. All subjects are to be evaluated at baseline and at weeks 24, 48, 72, and 96. There are 45 triads. In each triad is 3 patients: 1 with HIV and PI treatment for on average 4 years; a second patient with HIV but no PI; and a third uninfected person. The data presented here is just the baseline evaluation. Study participants were matched by cardiovascular risk factors: age, race, sex, blood pressure, smoking, menopausal status.
People were excluded from this study if they had: HIV RNA >10,000; family history of MI in first degree relative (male <55, female >65); diabetes (fasting glucose >126 or use of oral anti-diabetic drugs); uncontrolled hypertension; renal disease (Cr >1.5 mg/dL); AST or ALT >2.5 ULN; hypothyroidism; use of systemic steroids.
The study objectives are to compare differences in baseline IMT of the carotid artery between HIV-infected individuals on PI therapy and subjects not receiving PI therapy; to compare differences in baseline IMT between HIV-infected and HIV-uninfected subjects; and to examine predictors of baseline IMT.
There was no difference between the 134 subjects in 45 triads (PI vs no PI vs HIV-) regarding CD4 count, viral load (%<50 copies/ml), and smoking. All subjects had normal blood pressure. Triads were matched by race/ethnicity: 76% white, 3% black, 16% hispanic, 4% asian/pacific islander.
Fasting glucose was the same (84-87) across all 3 groups); total cholesterol and triglycerides were higher in the PI group.
The average IMT was not significantly different between the PI, no PI, and HIV negative subjects. PI-treated patients (median duration of PI use of 4 years) did not have increased values for carotid IMT when compared to well matched HIV-infected patients not on PI therapy or to HIV-uninfected controls in this baseline analysis. Factors considered but not found to be significant: fasting glucose, total cholesterol, TG, LDL cholesterol, smoking, CD4, HIV RNA, duration of PI use, race/ethnicity, waist to hip ratio. Using a multi-variate analysis low HDL cholesterol when TG were >300, older age, and increased body mass index were significant factors in predicting increased carotid IMT and thus risk for heart disease. Every 10-unit decrease in HDL cholesterol was a borderline risk. PI-treated patients were more likely to have increased total cholesterol, TG, and low HDL compared to the other 2 groups, but the study did not find an association between PI exposure per se and increased carotid at this baseline analysis. Study subjects will continue to be followed beyond the 4 years they’ve been followed so far and perhaps over time we will see different results.
Gregory Lucas (Johns Hopkins University School of Medicine) reported on their study called "Increasing Risk of Cardiovascular Disease in HIV-infected Persons in Care". This study was designed to assess the risk of CVD in a large urban HIV cohort. The Johns Hopkins HIV Cohort is an observational clinic-based cohort that collects longitudinal demographic, clinical and therapeutic data on HIV-infected patients who receive primary care in Johns Hopkins HIV Clinics. The clinic has been enrolling patients since 1990.
Cases of CVD were defined as coronary heart disease (a new myocardial infarction or unstable angina) and cardiovascular disease (stroke) ( a new ischemic stroke or TIA; hemorrhagic events were excluded. Controls included patients with no clinical evidence of CVD. 4-5 controls per case were randomly selected. It is generally considered that retrospective case-control studies may not be as reliable as prospectively conducted studies. This analysis included both a prospective assessment of the incidence of CVD, and used a nested case-control design to assess risk factors for CVD.
2671 patients were followed after January 1996 contributing 7330 person-years of longitudinal follow-up time through October 1, 2003. CVD events: 43 coronary heart disease events; 37 strokes; 2 patients had both events.
Incidence of CVD:
--5.9 events per 1000 person-years for coronary disease
--5.0 events per 1000 person-years for stroke
--10.6 events per 1000 person-years for CVD overall
In comparison, the event rates in the National Health and Nutrition Examination Survey Epidemiologic Follow-Up Study (NHANES; persons without HIV) were:
--2 to 2.5 events per 1000 person-years for coronary disease
--3 to 3.5 events per 1000 person-years for stroke
Prior to 1/1/96 there were 11 events during 4711 person-years of follow-up time from 1/1/90 through 12/31/95, and 2.33 events per 1000 person-years for CVD.
In a multi-variate analysis Lucas said total cholesterol (relative risk=1.4), diabetes (RR=1.41), and hypertension (RR=3.18) were strongly associated with increased risk for CVD—Lucas said "all classical risk factors for CVD". Age (RR=1.81), d4T use (RR=2.51) and 3TC use (RR=1.75) were also associated with increased risk.
I did not see and data presented on smoking and if there were differences between the cases and controls regarding smoking.
Lucas found that the rate for CVD is 5-fold higher than the risk before 1996. The incidence of both coronary heart disease and stroke appear to be higher than expected from a US population of similar age, race, and sex, and higher than rates in the Hopkins cohort prior to 1996. The use of protease inhibitors was weakly associated with CVD, there did not appear to be ant specific protease inhibitor associated with CVD.
Lucas offered these caveats regarding his study. These results were found in his urban clinic. The associations found in this study for CVD were more strong for more classical risk factors than between anti-retroviral drug use. Selection biases may occur because this study is not randomized and certain therapies may have been selected for certain patients based on patient or family background. We would need comprehensive data on known CVD risk factors such as cigarette smoking, lipoproiein levels which would be important to include when assessing the effect of antiretroviral drugs. Lucas concluded that heart disease and stroke appear to be occurring at an increased rate than would be expected in his Johns Hopkins Cohort.
P Hsue (University of California, San Francisco General Hospital) reported her study "Increased Atherosclerosis Progression in Patients with HIV: The Rols of Traditional and Immunologic Factors". The objectives of this study were to measure carotid IMT in a cohort of adult patients with HIV, to identify clinical and laboratory predictors of carotid artery IMT in adults with HIV, and to measure the rate of IMT progression over 1 year in a subset of patients. This is an ongoing clinic-based cohort study evaluating long-term outcomes of treatment (SCOPE Study). Subjects in SCOPE were not selected based on metabolic or cardiovascular risk factors. In the Currier study individuals with certain risk factors (family history, diabetes, etc) were eliminated from the study I guess in an attempt to evaluate the effect of HAART and HIV on IMT compared to HIV-negative individuals. The following are measured annually in this longitudinal ongoing cohort: fasting LDL-C, HDL-C, TG, hs C-reactive protein, lp(a), fibrogen, anthropometry (waist to hip ratio), carotid artery IMT.
In this study Hsue I think used the average of 12 measurements of IMT of 6 preselected sites of the artery while in the Currier study they are looking at one specific site in the artery. In my discussions at the conference I was not able to discern if one procedure was better than the other. I heard mixed opinions, one researcher felt the 12 segment approach might be better and others felt there would be no difference.
There were 106 subjects in this study; average age was 44; 83% men; duration oh HIV-infection 11 yrs; 82% had PI exposure with average 4 yrs of duration; 58% had cigarette smoking; 25% hypertension; 7% diabetes. Average CD4 count was 354; CD4 nadir 100; HDL 40 mg/dl; LDL 104 mg/dl; TG 225 mg/dl; Hs CRP 1.75 mg/dl; LP(a) 13 mg/dl. The mean baseline IMT was significantly higher than expected in these patients than those in the large ARIC study of HIV negative males average age 45, which used a similar measurement technique (0.9 +/- 0.3 mm vs 0.61 mm).
Predictors of baseline IMT in multi-variate analysis: age (p<0.001), LDL cholesterol (p=0.018), hypertension (p=0.008), and nadir CD4 count <200 (p=0.003). Non-predictors were: duration oh HIV-infection, HIV RNA level, current CD4 count, HDL-C, TG, C-reactive protein, fibrogen, Lp(a).
The average age was 44 years; 21 were men; median duration of HIV-infection was 11 years; 20/22 on a PI-based regimen (median duration of PI therapy 4 years; CAD risk factors: tobacco use (8); hypertension (9); diabetes (1).
The median rate of progression was 0.1 mm/year which is higher than rates seen in other previously conducted studies Hsue presented.
Hsue concluded that HIV patients have greater IMT compared to historical controls. Both traditional risk factors (age, hypertension, and LDL-C) and advanced immunodeficiency (nadir CD4 <200) were independently associated with IMT progression in HIV+ individuals. Hsue said this latter finding may suggest that chronic inflammation is associated with coronary artery disease in HIV+ individuals. In a very small subset of patients in a short follow-up period of one year IMT progression was accelerated compared to historical controls. This subset of patients were mostly PI-treated and had risk factors such as hypertension, cigarette smoking, and coronary artery disease. Artherosclerosis, as measured by IMT and progression, may represent an important source of morbidity in HIV-infected individuals. Clinicians should be aware of this and work to control all correctable cardiovascular risk factors.
Since Currier excluded patients with certain risk factors such as diabetes and family history of CVD, I think you might say this study takes a look at individuals with HIV and their inherent risk factors. Perhaps, this study suggests that people with HIV and people on HAART are more prone to heart disease and IMT progression. One criticism of this study is that historical controls of study subjects were used for comparison. Today, more individuals appear to have bad diets, are obese, and may have greater risk for heart disease. So by comparing current data from HIV+ individuals to data from Hiv-negative individualsfrom previous years you may not see the same results if comparing to current data from HIV-negative individuals.
For additional details from these studies and perspectives from NATAP writers you can see these links other NATAP Retrovirus Conference Reports:
Cardiovacular Risk in HIV, written by Carl Fichtenbaum, MD
Fat and Bone: What D:A:D Says and Other Strangeness at CROI
Written by David Wohl, MD
HAART to HEART: cardiovascular risk in HIV, written by Judith Aberg, MD