Assessing the cardiovascular impact of HIV, abacavir, and new signals for lopinavir/r
Nathan Geffen, TAC and Simon Collins, HIV i-Base
Peter Reiss summarised the growing number of studies on the relationship between abacavir and cardiovascular disease (CVD).  Four of six studies show an increased risk, while two, based on clinical trial data, do not.
The Data Collection on Adverse events of Anti-HIV Drugs (D:A:D) study found an increased relative risk of 1.68 (95%CI 1.33-2.13) for myocardial infarction (MI) in subjects recently taking abacavir. The relative accumulated risk was much smaller (1.07 per year; 95%CI 1.01-1.44). Importantly, this years analysis from D:A:D looked at individual PI effects for the first time and reported that after adjusting for lipids, cumulative (but not recent) exposure to indinavir or lopinavir/ritonavir was associated with an annual increased relative rate of MI (RR, 95%CI 1.08 [1.02-1.14] and 1.09 [1.01-1.18], respectively).
Sensitivity analysis of boosted and unboosted use of indinavir and saquinavir in a limited numbers of patients in D:A:D did not identify ritonavir-boosting as a risk factor. There were no statistically significant associations between recent or cumulative use of tenofovir, ddC, AZT, d4T, or 3TC and MI risk or with cumulative exposure to nevirapine, efavirenz, nelfinavir or saquinavir.
The SMART study supported the D:A:D results on abacavir, with patients using abacavir having a significantly higher risk of heart disease in four measured categories, including MI (RR 4.3; 95%CI 1.4-13). 
In the STEAL study, a randomised trial comparing abacavir + 3TC versus tenofovir + FTC in 360 treatment experienced patients in Australia, David Cooper and colleagues reported eight CVD events in the abacavir group versus one in the tenofovir arm (HR: 7.7; 95%CI 0.02-0.98); p=0.046). However, the abacavir arm had significantly more current smokers at baseline (40% v. 29%). This is a small trial, but randomisation means any differences are unlikely to be due to confounding/channeling bias. 
Furthermore, a case-control study in the ANRS CO4 study, looking at the effect of specific antiretroviral drugs on MI risk among more than 11,500 patients in the French Hospital Database, showed recent abacavir (less than one year) to double the risk of a heart attack (OR=2.19, 95%CI: 1.19-4.02). This study also reported a significantly increased risk for lopinavir (OR = 1.38/year, 95%CI 1.10 to 1.74), and amprenavir/fos-amprenavir (OR = 1.55/year, 95%CI 1.20 to 1.99). 
Yet, GSK’s abacavir database which included nearly 15,000 patients, show no increased risk of MI (RR 0.86; 95%CI 0.4-1.86; p=0.71) or coronary artery disorders (RR 0.59; 95%CI 0.35-1.01; p=0.06). If anything their data shows a trend in favour of ABC for the latter. However, many commentators have pointed out that registrational trial databases are short duration in younger and generally healthier patient groups, and are not designed or powered to look for cardiovascular events. Additionally, many members of the control group may have received treatment with other drugs (e.g protease inhibitors) that may themselves increase the risk of MI. 
Constance Benson presented data of 3,200 patients randomised to their first ART regimen in one of five ACTG studies. Follow-up data was available for over 2,100 patients through the ALLRT long-term protocol. 
Follow-up was censored at the first of off-study, death, initiation of non-randomised abacavir or 6 months after the last visit or discontinuation of randomised [HAART]. Risk was estimated for multiple factors including abacavir exposure, gender, race, age, viral load, CD4 count, ddI use, smoking, hypertension, high cholesterol, hyperglycaemia and family history of CVD. An event was classified as MI if confirmed by two independent reviewers.
Severe CVD events were identified in 63 patients, of which 27 were MI. Significant increases in the risk of events were detected for hypertension (RR of 2.3 for severe CVD; 95%CI 1.3-4.1; p=0.007) and older age (RR of 2.0 per 10 years of age for MI; 95%CI 1.4-2.9; p<0.001. RR of 1.9 per 10 years of age for severe CV; 95%CI 1.5-2.4; p<0.001).
They found no association between either MI or severe CVD and recent abacavir use (RR of 1.2 for MI; 95%CI 0.5, 3.1; p=0.82. RR of 0.8 for severe CVD; 95%CI 0.4-1.5; p=0.5). Of note, however, male sex (a well known risk factor for MI) was not identified as a risk factor for MI in the study, emphasizing the lack of power.
Reiss recommended the following to deal with these complex results:
Although differences in study design, statistical power, endpoint definitions, and procedures to capture and validate endpoints may each contribute to these discrepant findings, additional possible explanations also need to be considered. Reviewing the characteristics of the various patient populations which were studied, one could for instance speculate whether the likelihood of identifying the CVD risk associated with abacavir may be greater in those who are first exposed after their HIV infection is already suppressed.
Data suggest a pathogenic mechanism (possibly of a proinflammatory nature) involving acute processes, such as plaque rupture or subsequent thrombosis, rather than a chronic one affecting atheroma formation.
For now, it seems prudent to withhold abacavir from patients with high underlying CVD risk if suitable alternative regimens are available. If not, patients absolute CVD risk in the presence of abacavir should be minimised by aggressive management of traditional CVD risk factors.
Potential abacavir mechanisms
Explaining the D:A:D and other finding are complicated by not having a clear mechanism of action for any effect. While this is common by definition for any unexpected reaction, especially in HIV care – most notably for fat accumulation – it is an area that many research groups are looking at.
The summary of these studies at CROI is similarly complex:
- GSK data from the HEAT study found no differences at 96 weeks in endothelial function markers (vascular cell adhesion molecule-1; sVCAM-1) or inflammation markers (IL-6 and hs CRP) between almost 500 patients randomised to either abacavir/3TC or tenofovir/FTC, each with lopinavir/r in a prospective, randomised study in treatment naive patients. 
- Frank Palella and the MACS cohort reported similar findings in over 300 matched pairs (194 women, 96 men). Abacavir use was not independently associated with elevated plasma levels of hsCRP, IL-6, and D-dimer. While changes in the levels of these markers were seen between the baseline and index visits (D-dimer and IL-6 decreases, hsCRP increases), they were comparable among persons who initiated ABC versus non-ABC containing HAART. Women had higher D-dimer and lower CRP levels than men. 
- However, Claudette Satchell and colleagues from University College Dublin conducted a prospective study to assess platelet function in 58 patient, 30 of whom were on abacavir-containing regimens (ABC group) and 28 who were on non-abacavir-containing ART (no ABC group).  They reported consistently higher platelet reactivity in the abacavir group when exposed to increasing levels of platelet agonists and that these differences remained significant when controlled for gender, age, ethnicity, mode of HIV acquisition, smoking history, diabetes, family history of CVD, systolic blood pressure, use of other classes of ART, use of aspirin and methadone and CD3+, CD4+, and CD8+ T cell count.
Impact of HIV
Several studies also provided evidence for the role of HIV in cardiovascular disease:
- Carl Grunfeld reported that in the Fat Redistribution and Metabolic Change in HIV infection (FRAM) study, even after adjustment for traditional CVD risk factors, HIV infection was independently associated with as severe an impact on atherosclerosis (measured by increased carotid intima media thickness (IMT)) as traditional CVD risk factors, such as smoking.  The presentation also suggested that the previous contradictory results looking at IMT in HIV infection may be explained by the two studies finding a link to HIV having measured both thickness in the common carotid and the internal and carotid bulb (a region associated with more vascular turbulence and impact of HDL and total cholesterol) the five studies finding no association having only measure the common carotid.
- A second analysis from FRAM following over 900 HIV-positive patients and almost 300 HIV-negative controls reported a mortality risk that was 3 times higher among the HIV-positive group even after adjustment for demographic and traditional CVD risk factors. 
- Priscilla Hsue and colleagues from San Francisco General Hospital, whose group has also reported significantly increase carotid IMT associated with HIV and treatment, reported a new study showing higher levels of coronary artery calcium (CAC) in almost 250 HIV-positive patients compared to 45 HIV-negative matched controls. After adjusting for age, gender, race, smoking, hypertension, and diabetes mellitus, the HIV-infected subjects had a significantly higher prevalence of detectable CAC (OR = 2.7, 95%CI 1.06 to 6.7, p = 0.037). In the adjusted analysis, only age, gender, and race were significant with no impact seen for duration of HAART, PI-use and time with undetectable viral load. 
While consensus on the use of abacavir in people at high cardiovascular risk appears to be following Peter Reiss’s summary statement. The data on individual PI effects are new, linked to lopinavir and not apparently related to the boosting effect of ritonavir. Lower use of both indinavir and ddI should make those data largely of historical interest.
The focus on the potential mechanisms is likely to continue, but drug-related side effects are often observed and yet poorly understood. Indeed the exact mechanisms for the activity and toxicities of many drugs is frequently poorly understood.
Reducing modifiable cardiovascular risks is clearly an important goal on an individual patient level given the accumulating evidence linking untreated HIV infection to heart disease, but this can be a challenge in practice.
Unless stated otherwise, all references are to the Programme and Abstracts of the 16th Conference on Retroviruses and Opportunistic Infections. 8-11 February 2009, Montreal. Oral abstracts are available as a web cast.
- Reiss P. Abacavir and cardiovascular risk. 16th CROI. Oral abstract 152. http://www.retroconference.org/2009/Abstracts/36788.htm
- Lundgren J. et al. Risk of myocardial infarction with exposure to specific ARV from the PI, NNRTI, and NRTI drug classes: the D:A:D study. 16th CROI, 2009. Oral abstract 44LB.
- Cooper DA et al. Simplification with fixed-dose tenofovir/emtricitabine or abacavir/lamivudine in adults with suppressed HIV replication: the STEAL study, a randomised, open-label, 96-week, non-inferiority trial. 16th CROI, 2009. Poster abstract 576.
- Lang S et al. Impact of specific NRTI and PI exposure on the risk of myocardial infarction: a case-control study nested within FHDH ANRS CO4. 16th CROI 2009. Oral abstract 43LB.
- Cutrell A et al. Abacavir and the potential risk of myocardial infarction. The Lancet. 2008 Apr 26;371:1413.
- Benson C et al. No association of abacavir use with risk of myocardial infarction or severe cardiovascular disease events: results from ACTG A5001. 16th CROI 2009. Poster abstract 721.
- McComsey G et al. Similar reductions in markers of inflammation and endothelial activation after initiation of abacavir/lamivudine or tenofovir/emtricitabine: the HEAT study. 16th CROI 2009. Poster abstract 732.
- Palella F et al. Inflammatory markers among abacavir and non-abacavir recipients in the Womens Interagency HIV Study and the Multicenter AIDS Cohort Study. 16th CROI 2009. Oral late-breaker abstract 150LB.
- Satchell C et al. Platelet hyper-reactivity in HIV-1-infected patients on abacavir-containing ART. 16th CROI 2009. Oral late-breaker abstract 151LB.
- Grunfeld C et al. HIV infection Is an independent risk factor for atherosclerosis similar in magnitude to traditional cardiovascular disease risk factors. 16th CROI 2009. Oral abstract 146
- Modrich L et al. Factors associated with mortality in the study of fat redistribution and metabolic change in HIV infection. 16th CROI 2009. Poster abstract 706.
- Hsue P et al. HIV infection is independently associated with detectable coronary artery calcium. 16th CROI 2009. Poster abstract 724.