Increased risk of myocardial infarction associated with abacavir and ddI

Simon Collins, HIV i-Base

A poster presented by Caroline Sabin from the Royal Free Hospital looked at whether nucleoside analogues were associated with cardiovascular disease (CVD) seen in the large international D:A:D cohort study. [1] D:A:D includes over 33,000 patients from 11 prospective cohorts, in which 517 myocardial infarctions (MI) occurred over approximately seven years of follow up (157,912 patient years).

This group first reported that ARV treatment as a whole was associated with an increased risk of CVD. Over subsequent years, the study gained more power to look at drug-class effects and last year reported a 110% increased risk per year related to PI use, after adjusting for blood lipids and other metabolic parameters and demogrpahics. No increased risk was seen with use of NNRTIs.

This study is now sufficiently powered to look at the impact of some individual ARVs, and in this adjusted multivariate analysis found a 90% increased risk from current or recent use (within six months) of abacavir (RR 1.90, 95%CI 1.47-2.45, p<0.01) and a 49% increased risk with ddI (RR 1.49, 1.14-1.95, p<0.01).

The absolute risk was highest in patients with highest underlying risk factors for CVD. When including the predicted 10-year Frammingham risk in the regression model, the MI rate increased by 95% (1.95 [1.50, 2.55]; p=0.0001) in those with a moderate 10-year CHD risk and by 142% (2.42 [1.74, 3.36]; p=0.0001) in those with a high 10-year CHD risk, when compared to those with a low 10-year CHD risk.

This effect remained after adjustment for viral load, CD4, blood lipids and other metabolic factorsÂ’. No link was found with duration of use, or past use of either drug, leading the researchers to conclude that if this effect is causal, the unknown biological mechanism (s) appears reversible after stopping these drugs.

No increased risk was seen with the use of ATZ, d4T or 3TC. Exposure to tenofovir or FTC in the cohort was insufficient to include these drugs in the analysis.


This is likely to be one of the most discussed studies from the meeting and it is unclear why it didnÂ’t receive an oral session.

Even though the mechanism is unknown, reporting a ARV risk that is only slightly short of the lower estimated risk for a current smoker, in a trial as well powered as D:A:D, is likely to impact prescription practice for patients with high underlying Framingham risk.

For patients currently stable on treatment, with low cardiovascular risk, or who have limited treatment choices, the benefit of abacavir may still outweigh any low increase in absolute risk. Recent analyses from the SMART study suggests that viral suppression <50 copies/mL reduces risk while higher levels of viraemia increase risk.

These decisions clearly should be assessed on an individual case basis.

Simon Collins is the community representative on the D:A:D Steering Committee.


Sabin C, Worm S, Weber R et al. Do thymidine analogues, abacavir, didanosine and lamivudine contribute to the risk of myocardial infarction? The D:A:D study. 15th CROI, 3-6 Ferbruary 2008, Boston. Abstract 957c.

The poster is available in PDF format from:

Links to other websites are current at date of posting but not maintained.