CROI: DAD cohort finds increased risk of heart attack in people taking abacavir or ddI

This article is more than 16 years old. Click here for more recent articles on this topic

One of the more surprising reports at the Fifteenth Conference on Retroviruses and Opportunistic Infections in Boston on Monday was that the nucleoside analogue abacavir (Ziagen) appears to increase the risk of myocardial infarction (MI), or heart attack, by 90%, and didanosine (ddI, Videx) by 49%, No such increase was seen with zidovudine (AZT), stavudine (d4T, Zerit), and the additional risk due to abacavir and ddI largely disappeared after the drugs were discontinued. These findings were drawn from the large, multi-cohort DAD observational study.

However, although the relative risk increased in individuals taking abacavir or ddI, the absolute risk of heart attack remained low for individuals without a moderate to high cardiovascular risk. Older age, smoking, high cholesterol levels, diabetes, high blood pressure and a family history of cardiovascular disease are important risk factors.

The investigators noted that the predicted rate of heart attacks over ten years would be three per thousand person-years of follow-up among patients with low cardiovascular risk, less than ten per thousand person-years among people with moderate risk, and 32 per 1000 person-years among people with high cardiovascular risk.



Relating to the heart and blood vessels.


The result of a statistical test which tells us whether the results of a study are likely to be due to chance and would not be confirmed if the study was repeated. All p-values are between 0 and 1; the most reliable studies have p-values very close to 0. A p-value of 0.001 means that there is a 1 in 1000 probability that the results are due to chance and do not reflect a real difference. A p-value of 0.05 means there is a 1 in 20 probability that the results are due to chance. When a p-value is 0.05 or below, the result is considered to be ‘statistically significant’. Confidence intervals give similar information to p-values but are easier to interpret. 

myocardial infarction

Heart attack. Myocardial refers to the muscular tissue of the heart. An infarction is the obstruction of the blood supply to an organ or region of tissue.

person years

In a study “100 person years of follow-up” could mean that information was collected on 100 people for one year, or on 50 people for two years each, or on ten people over ten years. In practice, each person’s duration of follow-up is likely to be different.

body mass index (BMI)

Body mass index, or BMI, is a measure of body size. It combines a person's weight with their height. The BMI gives an idea of whether a person has the correct weight for their height. Below 18.5 is considered underweight; between 18.5 and 25 is normal; between 25 and 30 is overweight; and over 30 is obese. Many BMI calculators can be found on the internet.

In a statement issued by the steering committee of the study, investigators said that patients should discuss the study results with their doctor in order to assess their degree of risk, and should not stop abacavir treatment without prior discussion with their doctor.

For the study as a whole, the risk of heart attack was 1 in 64 over five years. Abacavir increased this risk by 90%, but the increase in risk was most concentrated among those with moderate to high cardiovascular risk.

Study background

The DAD study (Data Collection on Adverse Events of Anti-HIV Drugs) is a collaboration of eleven prospective cohorts from across Europe, Australia, and the US, totalling over 30,000 participants. The DAD researchers have previously reported data on cardiovascular risk and myocardial infarction in people on antiretroviral treatment (see, for instance, here). The goal of this new study was to investigate whether nucleoside analogues (NRTIs) – particularly the thymidine analogues zidovudine (AZT) and stavudine (d4T, Zerit), which have known associations with dyslipidaemia and insulin resistance – had an observable effect on rates of myocardial infarction.

For this analysis, 33,347 patients were followed from the time of enrolment, yielding 157,912 person-years of follow-up data. Over this time, 517 patients developed a myocardial infarction .

The impact of each of the five NRTIs (AZT, ddI, 3TC, d4T, and abacavir) on myocardial infarction risk was assessed by Poisson regression analysis. Since cardiovascular risk is influenced by many factors, risk ratios were adjusted for: age, sex, HIV risk factor, ethnicity, calendar year, cohort, smoking status, family history of cardiovascular disease, personal history of cardiovascular disease, body mass index (BMI), and exposure to other antiretroviral drugs (tenofovir, the NNRTIs, and the major PIs in use during the study).

Several different statistical models were used to estimate the relative impact (compared to no use) of: recent use (current, or within the past six months), past use (more than six months previously), and cumulative use (per year) of each NRTI. Abacavir and ddI were the only two NRTIs to consistently show a significant impact on myocardial infarction risk. However, whether that impact was seen with recent, past, or cumulative use varied according to which model was used. In the most complete model, which included all three time frames, only recent (not past or cumulative) use of abacavir and ddI was found to increase MI risk – by 94% and 53%, respectively. Complete results were as follows:


Relative rate of MI due to:

Cumulative (per year)

Recent use (within past 6 mo)

Past use (more than 6 mo prev)

AZT (zidovudine)

1.04 [95% confidence interval (CI), 0.99-1.09]


1.22 [0.82-1.81]


1.29 [0.89-1.85]


ddI (didanosine)

1.00 [0.93-1.07]


1.53 [1.10-2.13]


1.08 [0.84-1.39]


d4T (stavudine)

1.02 [0.95-1.09]


1.22 [0.84-1.77]


1.24 [0.93-1.66]


3TC (lamivudine)

0.99 [0.93-1.06]


1.69 [1.02-2.80]


1.45 [0.88-2.40]



1.00 [0.92-1.08]


1.94 [1.48-2.55]


1.29 [0.94-1.77]


Lamivudine (3TC) occupies a somewhat ambiguous position in the DAD report. While abacavir and ddI had statistically significant effects in all three of the analytical models (the other two of which are not shown here), 3TC shows a significant effect (a 69% myocardial infarction risk increase due to recent use) only in the third and final model (summarised above). The poster presented at CROI states that, contrary to the original hypothesis, "neither cumulative nor recent use of the two thymidine analogues or lamivudine [3TC] was associated with risk of myocardial infarction."

The risks of myocardial infarction associated with recent abacavir and ddI use remained after adjustment for HIV viral load, CD4 cell count, dyslipidaemia and other metabolic factors. The excess risks attributable to these two drugs was, however, most pronounced in patients with high underlying cardiovascular risk. There is some question as to whether the results might be biased by this fact – for instance, patients at higher cardiovascular risk might be more likely to be placed on abacavir in the first place due to a perception of abacavir having a safer cardiovascular profile.

The researchers acknowledge this possibility, but argue that (a) adjusting for known cardiovascular risk factors had little effect on the outcomes, and (b) myocardial infarction risk was seen to decrease after abacavir cessation – which would be implausible if the risk was due to persistent underlying factors. Hence, they believe, "preferential use of abacavir and didanosine in patients with an a priori elevated CV risk appears not to explain the findings."

The researchers conclude that, "although it is impossible to rule out bias as an explanation, if these associations are [in fact] causal, the unknown biological mechanism(s) appear reversible upon cessation of these drugs."

The full poster, as presented at CROI, is available at the Copenhagen HIV Programme site at, as is a position statement by the DAD steering committee explaining the results for lay readers and placing them in broader context. Notably, while the committee unequivocally states that these results "show that using abacavir or ddI does increase the chance of a heart attack," there are many other controllable risk factors of comparable or greater impact. E.g., "a current smoker who is taking abacavir would reduce their risk of a heart attack slightly more by stopping smoking than by switching […] to an alternate drug." They "recommend that patients receiving abacavir or ddI should consult their doctor [to] discuss whether a modification of […] their regimen is appropriate."


Sabin C et al. Do thymidine analogues, abacavir, didanosine and lamivudine contribute to the risk of myocardial infarction? The D:A:D study. Fifteenth Conference on Retroviruses and Opportunistic Infections, Boston. Abstract 957c, 2008.

Update: Following the conference presentation, this study was published in a peer-reviewed journal:

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: a multi-cohort collaboration. The Lancet, 371: 1417-1426, April 2008.