Sunday, September 27, 2009

Meta-analysis: statins for primary prevention of cardiovascular disease in those with risk factors

A meta-analysis of controlled trial data found that in a range of people with cardiovascular risk factors but no overt cardiovascular disease, statin treatment reduced overall mortality and risk of major cardiovascular and cerebrovascular events. Absolute risk reductions, however, were fairly small.

Statin treatment for secondary prevention of cardiovascular disease is well accepted, however its use for primary prevention is still debated as this would have significant public health implications. In particular, the effects in women, older people, and those with diabetes are unclear. The authors of this meta-analysis aimed to use published clinical trial data to determine the effects of statins as primary prevention including assessment in the three subgroups mentioned. They carried out an extensive literature search to locate randomised controlled trials that compared statins with control in people with cardiovascular risk factors but no established cardiovascular disease. Eligibility criteria included follow-up for at least one year, mortality or cardiovascular disease as primary outcomes, at least 80% of participants without established cardiovascular disease or complete separate reporting of this subgroup. Primary end point of the meta-analysis was all cause mortality; secondary end points included composite major coronary events and composite major cerebrovascular events.

The initial search located 1,230 reports: 1,188 were excluded on the basis of title or abstract. Of the remaining 42 studies retrieved for full assessment, 32 were excluded to leave 10 eligible studies for analysis. These included 70,388 people, of whom 23,681 (34%) were women and 16,078 (23%) had diabetes. Specific data on the subgroups of interest were available for six of the studies, for one (ALLHAT) in the published paper and for five others from the original investigators. About 6% of the included participants were actually secondary prevention patients that could not be excluded from analysis, which was therefore also carried out excluding the three studies involved. Treatment allocation was evenly balanced (statin n=35,138, control n= 35,250); mean age was 63 years, and mean follow-up was 4.1 years.

Statin treatment was associated with a reduction in overall mortality: over the 4.1 years, this was 5.7% in the control group compared to 5.1% in the statin group for a relative risk reduction (RRR) of 12% (odds ratio 0.88; 95% CI, 0.81 to 0.96) and an absolute risk reduction (ARR) of 0.6% (NNT over 4.1 years 167).

There was also a reduction in major cardiovascular events: 5.4% in the control group vs. 4.1% in the statin group for a RRR of 30% (OR 0.70, 95% CI, 0.61 to 0.81) and an ARR of 1.3% (NNT over 4.1 years 77).
Similarly, cerebrovascular events were reduced: 2.3% vs. 1.9%, RRR 19% (OR 0.81; 95% CI, 0.71 to 0.93), ARR 0.4% (NNT over 4.1 years 250). There was no association between statin use and risk of cancer over the study period (OR, 0.97; 95% CI, 0.89 to 1.05).

Analysis by the defined subgroups showed no evidence of heterogeneity between the groups, and analysis without the studies that included secondary prevention patients did not change the results significantly.

The authors conclude that primary prevention with statins in patients with cardiovascular risk factors has similar relative effects on cardiovascular risk as secondary prevention, and that these are not significantly different in women, older people, and patients with diabetes. They comment that the absolute treatment benefit is low, however, and the current data do not allow identification of those patients who would most benefit. Correct identification of these people remains a challenge, however current risk scoring systems, as well as current data, indicate that older men (>65 years) with risk factors, or older women with diabetes and risk factors are the highest risk groups and would be most likely to benefit from long-term statin use.

BMJ 2009; 338: b2376 (link to abstract, fulltext freely available at time of posting)

No comments: