Warfarin and Aspirin for Primary Prevention

Source: The Medical Research Council's General Practice Research Framework. Lancet 1998;351:233-241.

Previous studies have suggested that warfarin and aspirin may prevent ischemic heart disease (IHD) events in higher risk individuals without overt coronary artery disease. Thus, the MRC General Practice Research Framework conducted the Thrombosis Prevention Trial, which randomized 5085 men between the ages of 45 and 69 years who were in the top 20-25% on an IHD risk score to four treatment groups: warfarin to an INR of 1.5 (W group); aspirin 75 mg/d (A group); both (WA); or placebo (P). The men were seen every three months for up to five years to adjust their warfarin dose and assess for events. The primary end point was total IHD events, coronary death, or non fatal myocardial infarction (MI). Secondary end points included stroke and other vascular events.

The mean INR of those on warfarin was 1.47 with a mean dose of 4.1 mg/d. There were 410 total IHD events-142 fatal and 268 non-fatal. Warfarin mainly reduced deaths by 17% (P = 0.04), but there were more strokes in those taking warfarin vs. those not (57 vs 49). Aspirin mainly reduced non fatal IHD events by 32% (P = 0.004). Hemorrhagic stroke was more common on aspirin, but the numbers were small (9 vs 1, P = 0.01). Reductions in total IHD events were similar for the warfarin and aspirin groups-21% (P = 0.02) and 20% (P = 0.04), respectively. WA compared to P showed a 34% decrease in all IHD events (P = 0.006). Again, hemorrhagic stroke was more common with WA vs. P (7 vs 0, P = 0.009). Also, ruptured aortic and dissecting aneurysms were more common in subjects on warfarin vs. those not (15 vs 3, P = 0.01). Major and minor bleeding episodes were more common on WA compared to the other treatment groups (P < 0.001), but there were few major episodes (WA = 12, W = 9, A = 8, P = 4) and most were upper gastrointestinal.

The authors conclude that these data support the concept that aspirin reduces non fatal IHD events and warfarin reduces fatal IHD events in men at high risk for IHD. The combination is more effective than either agent alone and significantly decreases all IHD events.

Comment by Michael H. Crawford, MD

This small but well-conducted trial flames the controversy over aspirin for primary prevention and forces us to reconsider the role of warfarin in low doses. In 1988-1989, the American Physicians Health Study and the British Male Doctors Study were published (N Engl J Med 1989;321:129-135; BMJ 1988;296:313-316, respectively). The American study was stopped early in 1988 because of a 40% reduction in first, non fatal MIs and a 65% decrease in fatal MIs. The British study did not show a reduction in first MI-hence the controversy. Both studies failed to show a reduction in total mortality, possibly because of an increase in fatal hemorrhagic strokes.

The Thrombosis Prevention Trial (TPT) results do not support the widespread use of aspirin for primary prevention for several reasons. In these high-risk men, aspirin decreased all IHD events by a modest 20% and had no effect on total mortality. The results in a lower risk group would probably be marginal, similar to the British Male Doctor's Study. Also, for every three IHD events prevented, there would be one disabling stroke. Finally, there are no data in women-at least half the primary prevention target population. However, in a high-risk male population, a stronger case for aspirin use can be made.

The theory behind warfarin for primary prevention comes partly from the observation in the Northwick Park Heart Study (QJ Med 1994;87:403-406), which reported that increased levels of factor VII were associated with higher IHD event rates. Since factor VII is vitamin K dependent, warfarin can reduce these levels. The dose of warfarin selected corresponded to an INR of 0.5 above normal, which would reduce factor VII from levels observed in high-risk men to those of low-risk men. This was accomplished with minimal inconvenience and relatively good safety. With every-third-month monitoring, major bleeding episodes were not statistically different from placebo, although non major bleeding occurred much more frequently.

The results of warfarin on the primary end point of all IHD events was similar to aspirin (21% reduction), mainly due to a 39% decrease in fatal events. The combination of low intensity warfarin and aspirin produced the greatest reduction in all IHD events vs. placebo (34%) but resulted in an increase in fatal strokes. For every 12 IHD events presented, there was one excess stroke, but they tended to be hemorrhagic and fatal. Also, warfarin use was associated with an excess of dissecting and ruptured aortas. Thus, low-intensity warfarin cannot be recommended for primary prevention of IHD because of adverse events-especially with aspirin-and the inconvenience. However, for those who need but cannot take aspirin, it is an alternative. Whether it is better than other alternative agents, such as ticlopidine and clopridogel, is unclear.

An interesting editorial accompanied this paper titled "Aspirin, the poor man's statin" by Verhengt (Lancet 1998;351:227-228). Verhengt points out that the risk reduction in non fatal MIs in the TPT with aspirin is similar to the reduction seen in the Western Scotland Statin Study, which also studied high-risk individuals (N Engl J Med 1995;333:1301-1307). (See Table.)


      Age Cholesterol %Risk Persons/event

    Study years mmol/L reductio prevented Relative Cost

    WSSS 55 7.0 31 263 13

    TPT 57 6.7 32 347 1

Thus, aspirin could be considered an inexpensive alternative to statin drugs for primary prevention in high-risk patients. Administration of both in more afluent high-risk subjects could be additive or synergistic, but we don't know. Warfarin and aspirin seem to have independent effects based upon the results of TPT, and it is likely that statins work by another mechanism as well.

The TPT study pushes the boundaries for prophylactic aspirin use and demonstrates that affecting clotting mechanisms not involving platelets by warfarin may also be beneficial. The current choice of warfarin is probably not the ideal agent, but future clotting inhibitors may be efficacious with a higher safety profile. TPT also demonstrates that preventing IHD events may need a multifactorial approach.