German scientists have found a new genetic basis for why some patients respond better to the anti-clotting medicine Plavix, which could be used to test people before they get the world's second-biggest drug.

In a study in the Nature Medicine journal on Sunday, researchers from the University of Cologne hospital found that an enzyme known as paraoxonase-1, or PON1, is crucial for the activation of Plavix, made by Sanofi-Aventis and Bristol Myers Squibb, and known generically as clopidogrel.

The finding will fuel a debate about whether genetic or blood testing should be introduced for patients who need anti-clotting drugs in the light of several studies suggesting they are more likely to work in certain gene types.

Plavix had sales last year of more than $9.5 billion. The medicine is already off patent in parts of Europe and will lose U.S. patent protection in 2012.

However, experts say Plavix's effectiveness is hampered by differences in patients' metabolic processing.

This variability has been linked to genetic factors, but scientists do not agree about which specific genes may be responsible for differing response rates.

The issue hit the headlines in March, when the U.S. Food and Drug Administration said Plavix's label should carry a warning that some patients have a poor response to it.

Previous studies by U.S. researchers have found that variations of a gene known as CYP2C19 played a major role in patients' response to the drug. But a study published in August and sponsored by Sanofi and BMS contradicted the idea that people with a certain genetic make-up did not benefit from Plavix.

"There is a large dispute about the genetic factors," Dirk Taubert, who led the German study, said in a telephone interview. "CYP2C19 is currently the favored factor, but in our study we found no involvement of that factor."

"We have found a new genetic factor that we think is the crucial factor responsible for the variability in responses."

Taubert's team tested the relevance of genetic variants of PON1 in a group of patients with coronary artery disease who had a stent implanted to prop open clogged arteries, and then received Plavix to prevent blood clots, or thrombosis.

Their results showed that patients with one version, called PON1 QQ192, showed a higher risk of stent thrombosis than patients with another type, known as RR192. Taubert said the QQ192 patients also had activity levels of PON1 in their blood, as well as lower concentrations of active Plavix.

Taubert said the study found that PON1 differences accounted for around 75 percent of the response variability in this group of patients, who were of European origin, suggesting it would be a good factor to test for if doctors wanted to predict which patients might not respond well.

"It could be possible to make a genetic test or to test for blood levels of the enzyme activity, then it would be possible to predict which patients may have a poor response and therefore a higher risk of stent thrombosis," he said.