icon- folder.gif   Conference Reports for NATAP  
  XVII International AIDS Conference
Mexico City
3-8 August 2008
Back grey_arrow_rt.gif
Abacavir and Heart Disease
  Does Abacavir Make Heart Disease More Likely? Two New Studies Disagree
Immediately below the textual report are both the GSK and SMART Studies' slidesets

XVII International AIDS Conference
August 3-8, 2008
Mexico City
Mark Mascolini
Does abacavir predispose people to cardiovascular disease? Yes, if you believe data from the prospective D:A:D cohort [1] and an analysis of patients enrolled in the SMART trial [2]. No, if you believe combined data from 54 clinical trials of abacavir [3].
In 2007 D:A:D investigators published findings tying protease inhibitors to a higher risk of myocardial infarction (MI) [4]. Turning their attention to nucleosides reverse transcriptase inhibitors (NRTIs), the D:A:D team hypothesized that the thymidine NRTIs zidovudine (AZT) and stavudine (d4T) may promote MIs. But those researchers surprised themselves--and everyone else--when their 33,347-person analysis determined that abacavir almost doubled the relative risk of MI compared with other NRTIs, while didanosine (ddI) made an MI about 50% more likely [1]. The higher MI risk with abacavir vanished when people stopped taking this drug.
Both GlaxoSmithKline (GSK), maker of abacavir, and Bristol-Myers Squibb, maker of ddI, scrutinized clinical trial databases and side effects reports on these two NRTIs and failed to confirm any link between their use and heart attacks. GSK presented its analysis at AIDS 2008, comparing 9639 people who took abacavir (including a few hundred children) and 5044 who did not in 54 clinical trials with at least 24 weeks of follow-up [2].
The adult abacavir trial populations were relatively young at medians of 36 years in people with no antiretroviral experience when they started abacavir and 40 years in treatment-experienced people starting abacavir. In contrast, median age of abacavir takers in D:A:D stood at 45 years. Almost all abacavir users in D:A:D, 92% were men, who have a higher risk of heart disease than women, while 86% of treatment-experienced enrollees in the trials and 78% of naive enrollees were men.
Ischemic heart disease rates were low throughout the abacavir trials. Neither MI rates nor rates of any ischemic coronary artery disorder differed between people who took abacavir in the trials and those who did not. In fact, a lower rate of coronary artery problems with abacavir approached statistical significance:
- Frequency of MI: 0.114% with abacavir vs 0.139% without abacavir
- Rate of MI per 1000 person-years: 2.04 with abacavir vs 2.36 without abacavir, P = 0.706
- Frequency of ischemic coronary artery disease or disorder: 0.249% with abacavir vs 0.416% without abacavir
- Rate of ischemic coronary artery disease or disorder per 1000 person-years: 3.45 with abacavir vs 5.82 without abacavir, P = 0.055
Looking at MI incidence only in 12 adult randomized trials, the GSK team recorded rates of 2.15 per 1000 person-years with abacavir versus 4.10 per 1000 person-years with nonabacavir regimens, a nonsignificant difference. Incidence of any ischemic coronary artery disease or disorder in these 12 adult trials was 4.30 per 1000 person-years with abacavir and 7.64 per 1000 person-years without abacavir, also a nonsignificant difference. Trawling through spontaneous reports of side effects in their own database and in FDA files, GSK fished out no hints that abacavir may heighten MI risk.
Jens Lundgren, a top D:A:D investigator who also helped run the SMART treatment-interruption trial, reported that abacavir use without ddI quadruped the MI risk in people randomized to continuous therapy in SMART [3]. Abacavir also raised the risk of cardiovascular disease (CVD) defined three ways. Lundgren produced evidence suggesting heightened vascular inflammation with abacavir as the mechanism behind the higher CVD risk.
SMART investigators split the 2752 people in the continuous-treatment group into three contingents: people taking abacavir without ddI, people taking ddI, and people taking neither of those NRTIs. Median age in these three groups was 44 or 45, older than in the abacavir trials, and 83% were men.
Lundgren figured rates for MIs and three heart-disease clusters: (1) major CVD (clinical and silent MI, stroke, surgery for coronary artery disease), (2) major CVD expanded (major CVD plus peripheral vascular disease, congestive heart failure, drug treatment for coronary artery disease, and unwitnessed deaths), and (3) minor CVD (congestive heart failure, peripheral vascular disease or coronary artery disease requiring drug treatment).
Compared with use of NRTIs other than abacavir or ddI, current abacavir without ddI yielded an excess risk for MI alone in all three CVD categories:
- MI alone (n = 19): 4.3 times higher with abacavir
- Major CVD (n = 70): 1.8 times higher with abacavir
- Major CVD expanded (n = 112): 1.9 times higher with abacavir
- Minor CVD (n = 58): 2.7 times higher with abacavir
The D:A:D analysis did not determine the potential impact of tenofovir therapy on MI risk because not enough people were taking tenofovir at the time of the analysis [1]. When the SMART team used tenofovir takers as the comparison group (instead of all non-abacavir/ddI takers), current abacavir still yielded an excess risk of CVD.
The abacavir-MI link in D:A:D engendered much head scratching about why abacavir may predispose people to a higher heart attack risk. In their analysis of preclinical and clinical data, GSK investigators turned up no clues to a mechanism that may explain why abacavir would imperil the heart [2]. The SMART team did, finding higher rates of two markers of vascular inflammation in people taking abacavir than in those not taking this antiretroviral [3]: Levels of high-sensitivity C-reactive protein (hsCRP) were 27% higher (P = 0.02) and levels of interleukin 6 (IL-6) were 16% higher (P = 0.02) in the abacavir group.
Current ddI use in SMART did not heighten the risk of CVD or correlate with higher levels of inflammation or coagulation markers.
Lundgren and colleagues proposed that abacavir "causes an increased propensity for subclinical atherosclerosis to cause CVD." That increased propensity may be caused by inflammation-promoting properties of abacavir reflected in higher levels of hsCRP and IL-6. The bottom line, SMART investigators proposed, is that abacavir's apparently adverse impact on the heart appears to be clinically relevant only in people who already run a high risk of CVD.
Lundgren offered a laundry list of limitations to the SMART analysis: (1) Physicians may have preferred abacavir for people with apparent cardiovascular risk, but risk factors were about even between abacavir and nonabacavir groups in SMART, and statisticians adjusted the analyses for risk factors. (2) People taking abacavir may have had elevated hsCRP and IL-6 for reasons other than that drug. (3) Statistical power for some endpoints was modest. (4) SMART and D:A:D populations partially overlapped.
None of these studies has the final word on whether abacavir--combined with other risk factors in people with HIV--makes heart attacks or other types of heart disease more likely. Randomization in some of the trials GSK examined resulted in equivalent distribution of HIV traits and heart risk factors in the abacavir and nonabacavir groups [2]. D:A:D cohort members are not randomized to one treatment or another, so biases may creep in and resist attempts to control them away.
At the International AIDS Conference D:A:D investigator Andrew Carr observed that passive reporting of events like MI by trial clinicians is poor, so the GSK trial findings may underestimate true MI incidence. But there is no reason to assume underestimation would affect abacavir arms more or less than nonabacavir arms. Another attendee suggested that the trial analysis could be confounded by better virologic suppression rates in abacavir-containing arms, because other research links uncontrolled viremia to heart disease.
Compared with the abacavir trials, the D:A:D analysis had a hefty advantage in size and follow-up: The trials included 9639 people taking abacavir through 7845 person-years of follow-up and 5044 not taking abacavir through 4653 person-years; D:A:D included 33,347 people through 157,912 person-years of follow-up. Perhaps it takes that many people and that much follow-up to turn up an inflated heart disease risk in people taking antiretrovirals.
1. D:A:D Study Group. Use of nucleoside reverse transcriptase inhibitors and risk of myocardial infarction in HIV-infected patients enrolled in the D:A:D study. Lancet. 2008;371:1417-1426.
2. Cutrell A, Hernandez J, Yeo J, et al. Is abacavir-containing combination antiretroviral therapy associated with myocardial infarction? No association identified in pooled summary of 54 clinical trials. XVII International AIDS Conference. August 3-8, 2008. Mexico City. Abstract WEAB0106.
3. Lundgren J, Neuhaus J, Babiker A, et al. Use of nucleoside reverse transcriptase inhibitors and risk of myocardial infarction in HIV-infected patients enrolled in the SMART study. XVII International AIDS Conference. August 3-8, 2008. Mexico City. Abstract THAB0305.
4. D:A:D Study Group. Class of antiretroviral drugs and the risk of myocardial infarction. N Engl J Med. 2007;356:1723-1735.

Abacavir and Heart Disease

Intl AIDS Conference
Aug 7, 2008 Mexico City
Jens Lundgren from the DAD Study group reported results this afternoon thursday from an analysis of the SMART study and found results similar with what he found in the D:A:D study, that abacavir was associated with an increased risk of myocardial infarction (MI) and said the drug may cause vascular inflammation that may precipitate a CVD event. I conducted a brief interview with Dr Lundgren immediately after his presentation today. He said we should separate the management of patients from science here. "We are not saying to take patients off abacavir now.....we are cautious to say not that abacavir is causing myocardial infarctions but it may". We need more information regarding the science, more data regarding a mechanism of action. Ongoing now are several studies looking at cardiovascular disease (CVD) biomarkers to see if abacavir increases the biomarkers. Dr Lundgren said that outside of HIV among the general population increases in CVD biomarkers such as CrP and IL-6 are associated with increased risk for CVD. Of note, he said the DAD study findings he reported "were driven by the high-risk patients". Dr Lundgren expressed concern only about patients at high risk. That is patients at high risk for developing CVD drove the finding for increased risk for myocardial infarction. Patients with low-risk for CVD disease did not have increased risk for MIs. Regarding a potential explanation, he thinks perhaps there is an inflammatory reaction to abacavir so if a patient already has plaque in their artery this inflammation might make it worse. Of course, smoking cigarettes is the worst offender in terms of causing inflammation. Dr Lundgren was clear that patients should stop smoking, improve their diet, and perform exercise. Dr Lundgren said "the controversy is over-heated artificially". We need more information and study data such as the biomarker-abacavir data to say more. Patients should request their doctor or care provider to evaluate their CVD risk by using the Framingham test, which evaluates the risk for developing heart disease over the next 10 years. He said, if a patient is at high-risk for CVD from a Framingham evaluation that means they have a 20% chance of CVD over the next 10 years, and based on his findings a patient's risk would increase to 38%.
GSK also reported in an oral presentation yesterday wednesday there analysis of this question by going back and putting together all the data they had from abacavir studies. One important difference mentioned by Dr Lundgren between the DAD study and the GSK analysis is the age of patients. The average age of patients in the GSK studies was 36-40 yrs and the average age Dr Lundgren said was 45 in the DAD study. Age is an important factor in CVD. The older one is the more likely they are to accumulate CVD risk factors. Also, a cohort study like DAD looks at ordinary clinic patients while randomized controlled prospective studies like previously conducted GSK abacavir studies may not enroll sicker patients. In Dr Lundgren's slides, which are in the attached document he lists the limitations of the SMART analysis.
So, I think both studies have limitations. The SMART finding bolsters the case from the DAD study. There will be more data reported at CROI in 2009 and I'm sure over the next year more data will emerge on this issue. researchers will be trying to answer this question. In the meantime here is a paragraph I emailed 2 days ago that still works:
---- I did an informal survey here speaking to many thought leaders and important HIV researchers regarding the abacavir question. The feeling among most of them was that the GSK analysis of 54 abacavir studies from their database was less convincing than the the DAD study analysis by Jens Lundgren. The DAD group said in their published report that their data finds an association between abacavir and MIs, but not a casual relationship, but most who I surveyed feel these results are good enough to impact clinical decisions, they feel the study results created a reasonable doubt and if a patient has high CVD risk and their is another option they will consider it. However, several I spoke with today are not adequately impressed with DAD results to affect how they feel about abacavir. Most who I spoke with however agree and so does the DAD group say that their study is a cohort, although prospective, and as Jens Lundgren and GSK I recall said today at a session both the GSK and the DAD analyses are subject to limitations. Still, thursday Jens Lundgren will present his findings which I expect will find similar results from the DAD analysis, which it did. We still have to wait for the DAD analysis to see how tenofovir performed, they found abacavir associated with MIs but did not yet complete the tenofovir analysis. Unfortunately, this unanswered question leaves patients in limbo. I think this question is so highly charged that researchers will be trying hard over the near future to try to answer this question, a lot of research will be generated at this question. See the attached report which contains the slides from both the GSK & Lundgren presentations. Jules Levin