Statins are a ”mixed blessing” that can cut the risk of stroke, but trigger bleeding in the brain warns the Daily Mail today. The newspaper goes onto say that a study found that “statins can significantly cut the risk of stroke”, but “this benefit was partially undermined by a slight increase in the risk of suffering a haemorrhagic stroke”.
The news story is based on an online publication where researchers have taken a second look at findings from a large international study. The people enrolled in this study had all already had one stroke and were given the maximum dose (80mg) of a powerful statin, atorvastatin (Lipitor). Overall, there was a 16% drop in total stroke with atorvastatin and a significant fall in the number of coronary heart events. However, there was a small increase in the less common type of stroke – haemorrhagic stroke.
Where did the story come from?
Dr Larry Goldstein and colleagues form the Duke University Medical Centre, Durham, America conducted this study. The original Stroke Prevention by Aggressive Reduction in Cholesterol Levels (SPARCL) trial was funded by Pfizer, the drug company that developed and markets the branded version of atorvastatin. All the main authors disclose honoraria or grants received from the drug company. Employees of Pfizer were also involved in the interpretation of the data. The study was published ahead of print in the peer-reviewed medical journal Neurology .
What kind of scientific study was this?
This was a secondary analysis of the findings from a randomised controlled trial. The authors looked again at the SPARCL trial and used statistical modelling, to see how different factors, such as age, sex and type of stroke, influence the length of time before a second stroke.
The original SPARCL study looked at 4,731 patients who had already suffered a stroke in the previous six months and who had high cholesterol levels. Most of the participants had suffered an ischaemic stroke or transient ischaemic attacks, where the blood supply to the brain is permanently or temporarily blocked a clot. Only 2% of participants had suffered a haemorrhagic, or bleeding type of stroke. In the trial, people who were already taking lipid lowering medication stopped taking it, and then were randomised to either 80mg of atorvastain (brand name Lipitor) or the placebo (inactive) tablet. Overall, the SPARCL trial reported that atorvastain was associated with a 16% drop in total stroke and significant reductions in coronary heart events.
In the secondary analysis, the researchers used a modelling process which took all the background information collected about the participants in the study and related it to the time until participants suffered a second stroke (for those that did). The researchers presented the results in un-adjusted and adjusted graphs and tables. Adjusting the data removed the effect of other factors known to have an effect on stroke. For example, when estimating the increase in risk for men compared with women, the effect of age, blood pressure and atorvastain treatment was removed form the equation. When estimating the increase in risk for those taking atorvastatin the effect of age, sex and blood pressure were statistically removed.
What were the results of the study?
Over five years of treatment with atorvastatin, there was a 21% reduction in ischaemic stroke. Overall, 88 (1.9%) of the 4,731 people enrolled in the trial had a haemorrhagic stroke. Atorvastatin treatment increased the chance of developing a haemorrhagic stroke by 69%. Of those on atorvastatin, 2.3% experienced a haemorrhagic stroke during the study compared with 1.4% of those taking placebo tablets.
Men were 77% more likely to have a haemorrhagic stroke as their second stroke than women. People who were older were also at increased risk: for every 10-year increase in age, there was a 37% increase in risk of haemorrhagic stroke as a second stroke. The risk of a second hemorrhagic stroke was almost six times higher in those who had already suffered a first hemorrhagic stroke when they entered the study.
What interpretations did the researchers draw from these results?
The researchers conclude that hemorrhagic stroke was more common in those treated with atorvastatin than those treated with placebo.
These findings confirm what previous studies found about risk factors for haemorrhagic stroke: that it is more common in people who have already had a similar type of stroke before, more common in men, in people with high blood pressure, and with increasing age. The risk for haemorrhagic stroke was not linked to the level of cholesterol found in the participants at entry to the study or before the second stroke.
What does the NHS Knowledge Service make of this study?
This study used an exploratory statistical model which the researchers emphasise is useful in generating theories, but cannot prove a cause and effect between atorvastatin and a second haemorrhagic stroke. The study has strengths, as it was large and conducted over a long period so it amassed findings on the less common types of stroke. Only a small proportion (2%) of people eventually suffered a haemorrhagic stroke over the study period.
The dose of atorvastatin used in this study is high when compared with the typical doses used as a preventive measure in patients who have not had a stroke or heart attack, and are at lower risk of these events. This means that the results may not apply to most patients who take the drug.
When making choices about medication and therapy, the authors suggest that clinicians and patients should balance the increase in risk of haemorrhagic stroke against the overall benefit of these drugs in reducing overall strokes, as well as heart disease. This call for an individualised approach to treatment, based on evidence, is also repeated by most newspapers.
Sir Muir Gray adds...
Stroke is a dreadful disease; and dreadful decisions are sometimes needed because the prevention and treatment of the disease carries, like almost all treatments, the chance of risk as well as benefit. However, the specialists who deal with stroke have thought more than any other group about the best way to individualise the evidence, and relate it to the needs and values of each patient.