By Kirell Lakhman

Berkeley HeartLab's KIF6 biomarker test may not be as good as previously thought at identifying patients for whom statins might help reduce the risk for heart disease, according to a massive new study.

The meta-analysis, published in the current Journal of the American College of Cardiology, found that the KIF6 Trp719Arg polymorphism "was not associated with the risk" of clinical coronary artery disease, as BHL parent company Celera has been claiming since it began offering the test in 2008.

In fact, the results of the study were so profound that an accompanying editorial said the assay "now seems to be useless."

And because of the sheer size of the study — its 69 authors replicated 19 case-control trials comprising 17,000 cases and 39,369 controls — its results could stymie efforts currently underway by Celera to more broadly offer the assay in an attempt to revive slumping orders.

The $100 test has been performed more than 160,000 times in the US since it became available in 2008. In January the University of California, San Francisco, became the first non-Celera lab in the US to be granted a license to develop and perform a KIF6 LDT, while only three other non-Celera facilities currently offer it on a send-out basis.

If your lab performs molecular assays for cardiologists and the JACC study stirs the whiff of déjà vu, you're not suffering from phantosmia. In late August I wrote that a widely reported study of the widely prescribed anticoagulant Plavix claimed that reports of adverse events tied to how the CYP2C19 gene affects the drug's metabolism — a notion broadly taken as given — are virtually non-existent.

A couple of weeks later two more papers appeared claiming that a pair of next-generation anticoagulants do not need genetic testing to determine dosing; that two 2C19 alleles, *2 and *3, do not affect their efficacy in the way that affects patients taking Plavix; and that the drugs may even be superior to it.

This small cluster of large, influential papers will spur more than their share of debates and diatribes, and might even retard clinical molecular testing, at least for the assays they covered.

At the very least, the JACC study has prompted two respected geneticists to caution "the potential pitfalls present in prematurely adopting a genomic test without sufficient evidence."

As for that paper, Celera, in a highly critical reaction, said the study "did not investigate the association between KIF6 and statin benefit, and the results are subject to fundamental biases that limit any comparison to the previous prospective studies that found an association between KIF6 and coronary heart disease."

Moreover, the company said 70 percent of the CAD samples used in the JACC paper, as well as several of its co-authors, "are employed by a commercial firm that sells a competing genetic test for CHD risk.”

Celera was presumably referring to Decode Genetics, which sells a molecular test, dubbed deCODE MI, that is designed to detect increased risk for myocardial infarction by interrogating SNPs in the CDKN2A/2B gene region on chromosome 9. [CAD and CHD are used interchangeably in the JACC paper and by Celera.] In fact, one of the 19 studies reviewed by the JACC authors — conducted by Decode — showed the allele posed a 7-percent decreased risk of developing CHD.

Not 'Plausible'

In their paper, the JACC authors, claiming that recent studies, including many by Celera, "suggest" that carriers of the KIF6 719Arg allele are "at increased risk of clinical CAD compared with noncarriers, "sought to replicate" the association between the polymorphism and CAD.

To that end they genotyped the allele in 19 case-control studies comprising 17,000 cases of nonfatal CAD and 39,369 controls either as part of a genome-wide association study or in a "formal attempt to replicate the initial positive reports." The cases and controls comprised individuals of European descent and a "modest number" of South Asians, African Americans, Hispanics, East Asians, and admixed cases.

According to the results, "none of the 19 studies demonstrated an increased risk of CAD in carriers of the 719Arg allele compared with noncarriers."

Moreover, regression analyses and fixed-effects meta-analyses "ruled out with [a] high-degree of confidence" an increase of 2 percent or more in the risk of CAD among European 719Arg carriers, and "observed no increase" in the risk of CAD among 719Arg carriers in the subset of Europeans whose disease manifests at age 50 years for men and 60 years and younger for women compared with similarly aged controls and all non-European subgroups.

"This is such a big study [that] if there was a significant association between this variant and coronary disease, we would have found it," senior author Tom Quertermous, a cardiologist at Stanford, said in a statement.

To be fair, the authors conceded that the study may have been flawed by selection bias since they "looked at only non-fatal cases of coronary artery disease." If the allele increases the risk of fatal CAD, the data "could be skewed, although that's "not likely."

In addition, the researchers were unable to control for "all traditional risk factors" of CAD and failed to determine if statin use "modified the relationship because information on drug use wasn't available."

In an accompanying editorial, Eric Topol, director of the Scripps Translational Science Institute in La Jolla, Calif., and Samir Damani, also of Scripps, called the JACC study "well-conducted and elegant" and stressed that "most importantly, no plausible biological mechanism of KIF6-mediated statin response is currently available."

"Proponents of KIF6 testing will almost certainly continue to argue that a high prevalence of statin use in the CAD cases may have abrogated the KIF6 risk of CAD, and, therefore, no significant effect for KIF6 was observed," they wrote. "However, that argument holds little merit."

Topol and Damani concluded that "the KIF6 story should serve as a valuable reminder of the potential pitfalls present in prematurely adopting a genomic test without sufficient evidence."

Separately, Topol said the findings suggest that "we desperately need better regulation for genetic testing." The assay, which is CE Marked in the EU and co-marketed with Abbott, is performed as an LDT in the US, though Celera has said it plans to file a pre-market approval application for it by the end of the year.

'Mischaracterization'

For its part, Celera said the JACC study "does not refute prior research on KIF6's association with coronary heart disease and statin benefit," and argued that it "is not a replication study of the published prospective studies of KIF6 Trp719Arg, which involved rigorous prospective and randomized controlled trial designs."

“We believe this case-control publication mischaracterizes the prior peer-reviewed papers on KIF6’s association with CHD," Celera said in a statement. "To ascribe more credence to results from case-control studies, with their long history of known selection biases than to results from prospective cohort studies and prospective, placebo-controlled, randomized clinical trials, runs counter to the accepted hierarchy of medical evidence."

The company also said Topol and Damani's editorial "contains numerous inaccuracies, misstatements, and omissions," and "wrongly asserts that the prior KIF6 results 'stems primarily from a few observational studies.'"

Celera concludes that the editorial "inaccurately states that the reports of CHD event reduction from statin therapy in KIF6 carriers are "'largely unsubstantiated.'"

Got a tip? Click my byline above to send it along or submit it as a comment below. Follow my headlines on Twitter or subscribe to my RSS feed.