Thursday March 22 2012
Claims of a new 'heart attack prediction test' are innacurate
“Heart attacks could be predicted in advance,” The Daily Telegraph has reported, while the Daily Mail assures readers that a “simple and inexpensive test that can spot a heart attack waiting to happen could be available next year”.
However, these claims aren’t quite as impressive as they first appear. The reports are based on a small study looking at how levels of particular cells in the blood changed after a heart attack, not before one had occurred. Researchers tested levels of “circulating endothelial cells” (CECs) in 50 patients who had a heart attack and 44 healthy volunteers. Endothelial cells are normally found lining blood vessels but can leak into the bloodstream when arteries are damaged, and this can lead to a heart attack.
Researchers found that CEC levels rose after a heart attack, leading them to conclude that CECs may be a useful predictive tool to detect heart attacks before they occur. The flaw in this logic is that, without finding evidence that CEC levels rise before a heart attack, we cannot tell whether they signal an impending heart attack or merely rise as the result of one. The most practical way to test if CEC levels really can be a heart attack predictor is to test a group of subjects regularly over time and assess whether CEC levels do change prior to a heart attack.
For now at least, don’t rely on CECs to predict a heart attack, or on the media to predict a proven new heart attack test.
Where did the story come from?
The study was carried out by researchers from Scripps Translational Science Institute in the USA and was funded by the Clinical and Translational Science Awards and the American Reinvestment and Recovery Act. The study was published in the peer-reviewed Science Translational Medicine Journal.
The Telegraph’s coverage was generally well-balanced and included comments from other experts in the field highlighting the limitations of the study, which were that the blood samples were actually taken after the heart attack occurred: pouring water on other newspapers’ claims that scientists have found a way to predict a heart attack. For example, the Daily Mail’s claim that “a simple and inexpensive test that can spot a heart attack waiting to happen” seems overstated given the criticism the study has received from experts. This makes the Mail’s claim that the test “could be available next year” seem rather unlikely.
What kind of research was this?
This was a human study examining how the numbers and characteristics of types of cells called circulating endothelial cells (CECs) differed in patients who had had a heart attack and healthy volunteers. Endothelial cells are usually present in the lining of blood vessels such as arteries but can slough off into general circulation when the blood vessels become damaged.
Heart attacks, also known as acute myocardial infarction, are usually caused by a blockage in one or more of the coronary arteries that supply the heart muscle (myocardium). This interruption of blood supply starves the heart muscle of oxygen and causes muscle tissue to die (infarction). This is usually caused by an atherosclerotic plaque: a mass of cholesterol, other fatty substances and cellular matter that has built up on the wall of the coronary artery. These unstable atherosclerotic plaques can rupture or tear, causing a blood clot (thrombus) to form which can block the artery and cause a heart attack. Heart attacks are called by different names depending on findings on ECG (electrocardiogram) scans that suggest whether the coronary artery has been completely blocked or only partially blocked. The patients in this study had all suffered from a full blockage, known as an ST elevation MI.
The researchers stated that it is still very difficult to predict an acute heart attack with full blockage of an artery, even given recent advances in the diagnosis and treatment of heart diseases. They said that a clinical measurement that can predict an impending heart attack is still “desperately needed”.
This research aimed to determine whether CECs isolated from heart attack patients might show differences in number and characteristics that may be useful to predict heart attack events due to plaque rupture.
What did the research involve?
The researchers analysed the volume and cellular characteristics of CECs in 50 patients who had an ST elevation MI (an acute heart attack caused by a complete blockage of a coronary artery due to arterial plaque rupture). The CECs of the patients who had experienced a heart attack were compared with 44 healthy controls.
Heart attack patients were recruited from the emergency room at four regional hospitals in San Diego from Jan 2010 to Feb 2011, where they had blood samples taken to analyse their CECs. Blood samples were taken before their heart attacks were confirmed using an angiography (an imaging technique to look inside the coronary arteries) or before any non-emergency treatment was given.
Healthy controls were recruited from the blood donor programme at the Scripps Research Institute and were aged between 18 and 35. They were asked to report details of any chronic diseases they had, and only those reporting none were included.
Researchers used what they described as a “clinically feasible” microscope technique to detect and isolate CECs from the blood of study participants. This means that doctors should be able to use the technique in a medical setting such as a hospital rather than requiring specialist equipment unavailable to most doctors. The number of CECs, their size and shape and the genes they expressed were analysed from the blood sample and isolated CECs.
The analysis compared the number of CECs in the heart attack patients with the healthy controls. They did further analysis to see if any differences in CECs were related to age or other markers of heart attack.
What were the basic results?
The average age of the heart attack patients was 58.5 years and 82% were male. By contrast, the healthy volunteers had an average age of 30 and only 51% were male.
The study found that numbers of CECs were significantly elevated in patients who had a heart attack compared with controls: CEC counts were 19 cells per millilitre of blood in heart attack patients, compared to just 4 cells per millilitre in the healthy controls.
This raised CEC count did not correlate with other recognised measures of heart muscle death, such as test results relating to creatine kinase-myocardial band and troponin levels. Blood levels of these proteins will normally be raised in a person who has had a heart attack.
The researchers also noted that heart attack patients had CECs with different cellular characteristics from controls. Compared to healthy controls, the heart attack patients had CECs that were significantly larger, misshapen and had larger nuclear areas (the areas within cells containing the cellular DNA).
Cellular differences in heart attack patients were also observed when comparing them with patients with pre-existing peripheral vascular disease, a condition that leads to the narrowing and hardening of blood vessels in the legs and feet. It can cause blood clots and other circulatory problems.
How did the researchers interpret the results?
The researchers concluded that CEC counts “may serve as a promising clinical measure” for the prediction of heart attacks caused by atherosclerotic plaque ruptures.
This small study shows that CEC counts were significantly higher in 50 people who had a heart attack, compared with 44 health controls, which the authors claim could be useful in developing a predictive tool to detect imminent heart attacks.
The major flaw with this study is that they have collected blood samples after the patient had already been admitted to an emergency hospital ward for a heart attack, rather than before. Therefore, it is difficult to know whether the differences in the number and characteristics of the CEC observed in this study were present before the heart attack occurred, or were caused by the heart attack itself. Further research would be needed to establish whether there could be any meaningful, accurate and practical way to test their levels and predict a heart attack. For example, even if CEC characteristics do change before a heart attack, the length of time between the change and the attack needs to be established. If it’s just a few days it seems unlikely that at-risk patients will want to take precautionary weekly blood tests or that hospitals will have the resources to continually test people’s blood just in case they might have a particular type of heart attack.
To assess this further there would probably need to be a prospective study that follows a large group of people over a long period of time (for example, those identified to be at risk of cardiovascular disease), taking regular blood sample measurements and examining whether high levels of CEC are present before a heart attack occurs. Only when this type of study has reported its results will it be clearer whether elevated CECs could be useful in predicting heart attacks.
This current study by itself only shows the potential of CEC counts to be useful in the future; it does not prove that they are useful now or that they can actually predict a heart attack.
Analysis by Bazian