* Biomarker shows risk but itself does not cause disease
* "Fruitless" to develop drugs to lower protein levels
By Andrew Stern
CHICAGO, June 30 (Reuters) - Measuring a compound called C-reactive protein helps identify people with heart disease, but a study on Tuesday concluded there is no reason to think that elevated levels of the substance itself cause heart problems.
A pair of studies published in the Journal of the American Medical Association revealed limitations in the much-debated tests for biomarkers such as C-reactive protein, which is secreted by the liver and signals tissue inflammation and a risk for heart attack or stroke.
An international study involving 30,000 patients by researchers from Imperial College London and a dozen other universities and institutes in Europe and North America concluded that higher-than-normal levels of C-reactive protein were likely caused by tissue inflammation, and not a cause of the inflammation and resulting cardiovascular problems.
"This study puts the nail in the coffin ... in the question about (C-reactive protein): does it cause cardiovascular disease?" said Dr. James Lemos of the University of Texas Southwestern Medical Center, who helped write a commentary in the journal about the studies.
It clearly does not, he said.
Dr. Paul Elliott of Imperial College, who led the study, wrote that it would be "fruitless" to develop drugs aimed at lowering C-reactive protein to prevent heart disease.
Many heart attacks are caused by atherosclerosis, a narrowing of arteries that can be caused by a build-up of fatty deposits and inflammation.
UNEXPLAINED HEART ATTACK
For many years doctors have searched for tools to identify people at risk for sudden heart attack, but "you frequently encounter the unexplained heart attack," Lemos said in a telephone interview.
Traditionally, the risk factors used to identify risk are high blood pressure, diabetes, smoking and high cholesterol levels. Some doctors test their patients for biomarkers, but the usefulness of the results is debatable, Lemos said.
In the other study by researchers at Massachusetts General Hospital and Lund University in Malmo, Sweden, 5,000 Swedes were tested for six potential biomarkers for heart disease and then monitored for up to 15 years.
While five of the six biomarkers -- including C-reactive protein -- were useful in predicting future cardiovascular problems such as heart attacks, they were not helpful in classifying patients' risk or determining changes in the course of treatment.
Last year, the JUPITER study found patients who had higher-than-normal C-reactive protein levels but modest cholesterol levels had dramatically lower heart risks after taking the cholesterol-lowering statin drug Crestor, manufactured by AstraZeneca Plc AZN.L. That raised the question whether lowering C-reactive protein could curb heart disease.
Statins that include Crestor and Pfizer Inc's PFE.N Lipitor are the top-selling prescription drugs in the world.
"That is the goal of these tests -- to pick up these individuals (at risk for heart attack)," Lemos said. He said while the Swedish study was negative, others had mixed results.
"Both as a clinician and as an investigator, I still think at some point we will figure out how to use multiple biomarkers to predict risk. I don't think this is a failed idea," he said. (Editing by Maggie Fox and Xavier Briand)