Originally published April 26.
By Turna Ray
Amid mounting evidence against the utility of its KIF6 Genotyping Assay, Celera is hoping to meet with the US Food and Drug Administration to discuss the regulatory status of the test after the agency deemed it "not approvable."
As part of its premarket approval submission for the KIF6 test, Celera conducted retrospective analysis of several large, prospective studies showing KIF6's association with cardiac events and statin benefit. However, in an April 7 letter, the FDA told the company that it needs to conduct a prospective, randomized trial proving that the test is clinically useful to doctors in identifying which patients at risk for coronary heart disease would benefit from taking statins.
The impact of the regulatory setback is further clouded by the timing of the FDA's decision, which comes amid Celera's pending acquisition by Quest Diagnostics.
The acquisition, which cleared antitrust review last week and is slated to close at the end of the month, might give Celera the financial backing to conduct the types of studies that the FDA is requesting.
Then again, the companies could decide to scrap the KIF6 development program altogether and advance new tests. After all, Quest's interest in Celera goes beyond the KIF6 test. The national lab services provider said last month when it announced its decision to acquire Celera that it is especially eager to harness the capabilities of the company's Berkeley HeartLab business, and is looking to add a number of tests to its pipeline in the area of infectious diseases, neurological disorders, organ transplantation, and genetic disorders (PGx Reporter 3/23/2011).
Complicating the issue is the fact that since Celera filed its PMA, new clinical data have emerged contradicting the evidence previously presented by the company regarding the utility of the test. Specifically, two peer-reviewed papers and several posters presented at the annual meeting of the American College of Cardiology earlier this month were unable to find a link between KIF6 and coronary artery disease risk or statin benefit.
Researchers who have independently tested the utility of Celera's KIF6 assay and found it wanting believe the body of evidence increasingly suggests there is no saving the test.
Themistocles Assimes of Stanford University, the lead author of a large meta-analysis on the KIF6 assay published in the Journal of the American College of Cardiology last year, believes that the available evidence doesn't weigh in the test's favor.
"Our meta-analysis of case-control studies combined with the recent publication of [additional studies] overwhelmingly supports the lack of association between the KIF6 assay and risk of coronary disease, as well as response to statin therapy among subjects of European ancestry," Assimes told PGx Reporter this week. "The size and power of these studies makes it almost impossible for this conclusion to change."
Celera wouldn't comment beyond a statement on its website regarding the FDA's negative regulatory action regarding the KIF6 test. The "not approvable" letter for the KIF6 Genotyping Assay doesn't technically stop Celera from marketing StatinCheck, its laboratory-developed KIF6 genotyping test. Similar to the KIF6 Genotyping Assay, StatinCheck also predicts coronary heart disease risk and statin therapy benefit.
Under the current dual regulatory pathway for molecular diagnostics, StatinCheck is being performed through Celera's CLIA-certified Berkeley HeartLab, and can be marketed without FDA approval. However, since the FDA has indicated that it intends to promulgate additional requirements for LDTs based on their risk level and intended use, Celera will likely have to iron out with the agency what it will need to do to keep marketing StatinCheck as an LDT.
Last year, Celera CEO Kathy Ordonez said that its Berkeley HeartLab subsidiary had performed nearly 200,000 KIF6 tests since the company launched StatinCheck in the US two years ago. Celera estimates that approximately 180 million people are carriers of the KIF6 variant worldwide. The company's 23 internal sales reps market the test in the US, and the Celera has inked partnerships with Abbott and other external labs to sell the test in Europe. The price to patients for one KIF6 test is around $100.
Celera filed a PMA for the KIF6 Genotyping Assay with the FDA in January seeking approval for the genotyping test as an in vitro diagnostic for gauging the risk of coronary heart disease and statin benefit.
At the time of the filing, the company claimed it had validated its test in 55,000 people involved in nine studies showing the link between the KIF6 Trp719Arg allele and coronary heart disease risk, and the association between KIF6 and statin benefit. The data, according to Celera, suggest that KIF6 carriers have between a 22 percent and 55 percent greater risk for cardiac events than non-carriers, and that people with this genetic polymorphism also benefit more from statin therapy.
[ pagebreak ]
But even before the company submitted its PMA, researchers were questioning the utility of the KIF6 test (PGx Reporter 10/13/2010).
On October 7, an international team led by Assimes published a study in JACC looking at the association between the KIF6 Trp719Arg allele and the development of coronary artery disease in more than 17,000 CAD patients and more than 39,000 controls from 19 different studies conducted globally. Assimes et al. failed to find any association between the KIF6 variant and CAD.
Following the publication, Tom Quertermous, William G. Irwin Professor in Cardiovascular Medicine at Stanford and the study’s senior author, said that the meta-analysis "puts the nail in the coffin” for KIF6 as a marker for gauging risk of coronary disease. Given the size of the study, "if there was a significant association between this variant and coronary disease, we would have found it," Quertermous said.
The study was reviewed by cardiologists Eric Topol and Samir Damani of the Scripps Research Institute, who in an editorial accompanying the meta-analysis, suggested that Celera had commercialized its test as an LDT without appropriate validation that KIF6 carriers were at greater risk for coronary events and saw more benefit from statins than non-carriers.
At the time, Celera vigorously defended its test and countered that the deficiencies of the meta-analysis didn't paint an accurate picture of the KIF6 test's clinical utility.
The company's main quibble with the meta-analysis was that in the cohorts analyzed by Assimes et al., there was no reliable information available on the use of statins. As such, Celera held that if the majority of patients in these studies were already on statins, then that would bias the study's findings on KIF6's link to CAD. Additionally, the company noted that the JACC study only looked at patients with nonfatal cases of CAD, when previously published case-control studies suggested there was no association between KIF6 and nonfatal myocardial infarction and coronary artery disease.
Although the authors of the JACC paper cited the lack of knowledge about statin use as a weakness of their analysis, they noted that it was unlikely that there was much statin use among the patients in the 19 cohorts assessed, since most of these patients had early onset of coronary disease and fell below the age at which they would be prescribed statins.
Not swayed by Celera's arguments against his meta-analysis, Assimes doesn't believe that the company can save the test by narrowing its indication to a more specific subpopulation of cardiac patients.
"I see no compelling evidence that the KIF6 assay performs better when considering specific CAD outcomes (such as fatal vs. non fatal myocardial infarction, revascularization and/or angina) or any other subgroup," Assimes said via e-mail.
One remaining research option that hasn't been explored by Celera is looking at whether the KIF6 genotype yields different outcomes in non-European populations. "The KIF6 assay has not been extensively tested in non European groups but there is no compelling reason to believe that its effects would differ so markedly from the null effects in Europeans," Assimes said. "However, larger number of non-Europeans would have to be studied to support this hypothesis more definitely."
After receiving the "not approvable" letter from the FDA, Celera officials want an audience with the agency to put into context some additional negative clinical data on the KIF6 test that has emerged since the company submitted its PMA.
Celera acknowledged in a statement last week that clinical data that had emerged since its PMA submission may have swayed the FDA's view of the test. However, the company said it believes "that differences in study design and patients' clinical phenotypes could explain the contradictory results … and that research in progress on the biological function of KIF6 could add support for the association with risk for CHD and statin benefit."
[ pagebreak ]
The "contradictory" data include studies published in Circulation: Cardiovascular Genetics and the JACC and two posters presented at the annual meeting of the American College of Cardiology earlier this month, which "provided new information … that differed in several important respects" from data the company previously reported.
The CCG paper outlines the JUPITER (Justification for Use of Statins in Primary Prevention, An Intervention Trial Evaluating Rosuvastatin) study, in which researchers led by Paul Ridker of Brigham and Women's Hospital evaluated the effect of the KIF6 variant on outcomes among more than 8,700 Caucasian participants and in the trial as a whole. Ridker et al. found that there were no differences in percent LDL cholesterol or cardiovascular event rates between carriers and non-carriers of the allele.
The JUPITER trial enrolled men and women without prior cardiovascular disease or diabetes, who had normal LDL cholesterol. These participants were randomized to rosuvastatin (AstraZeneca's Crestor) or to placebo and followed until they had their first major vascular events and for all-cause mortality. "Genotype had no impact on rosuvastatin efficacy in further analyses that included all-cause mortality, in analyses conducted in the total trial cohort that adjusted for race, or in analyses using generalized models of inheritance rather than recessive models," the authors said. "Thus, at least for rosuvastatin, there appears to be no clinical utility to screening for KIF6 genotype as a method to determine vascular risk or to predict statin efficacy."
Previous studies cited by Celera had looked at KIF6's impact on patients' response to atorvastatin (Pfizer's Lipitor) or pravastatin (Bristol-Myers Squibb's Pravachol). Since the same effect of the allele was not observed in JUPITER with rosuvastatin, Celera noted that the "the differential benefit in KIF6 carriers may not apply to all statins as a class."
On March 30, JACC published data from the Heart Protection Study, a five-year trial that enrolled 18,348 patients with prior cardiovascular disease, noncoronary vascular disease, diabetes, or hypertension and compared their outcomes when they received simvastatin (Merck's Zocor) or placebo. In HPS, researchers led by Jemma Hopewell of the University of Oxford found that the KIF6 genotype was not significantly associated with cardiac events in placebo-treated patients. Furthermore, in the simvastatin arm, patients had a 42 percent reduction in LDL cholesterol, regardless of KIF6 carrier status.
"Statin therapy significantly reduces the incidence of coronary and other major vascular events to a similar extent, irrespective of KIF6 genotype," Hopewell et al. concluded. "Consequently, the use of KIF6 genotyping to guide statin therapy is not warranted."
However, Celera plans to point out to the FDA that in this study, crossing over placebo-treated patients to receive a statin other than simvastatin could have impacted the findings. "All patients in the HPS placebo group received simvastatin therapy during a four-to-six week run-in period, and by the end of the study 33 percent of those in the placebo group were using a non-study statin, factors which could have confounded the results," Celera said in a statement.
Separately, posters presented at the annual meeting of the American College of Cardiology covered KIF6-related outcomes in two studies: the Treating to New Targets trial and the IDEAL trial.
TNT was a dosing study involving atorvastatin in patients with stable coronary heart disease. In this study, although carriers of two copies of the KIF6 719Arg variant saw a 56 percent relative risk reduction in coronary heart disease, carriers of one copy of the KIF6 variant did not experience a significant event reduction. Celera said that this was "possibly because the treatments compared in TNT were with the same statin [atovastatin] and differed only in dose, a difference that may be less significant than differences between statins."
Additionally, Celera claims that data from other studies suggest that patients with more serious heart conditions might see a more pronounced KIF6-related statin benefit. "The previous PROVE-IT trial had enrolled acute coronary syndrome patients who may benefit more from a KIF6-related pleiotropic effect of high-dose statin therapy than would the stable CHD patients in TNT," Celera said in a statement. The company added that in this study, because there was no placebo arm, the link between KIF6 and coronary heart disease also couldn't be investigated.
Lastly, the poster also discussed IDEAL, which compared high-dose atorvastatin versus standard-dose simvastatin in patients with stable coronary heart disease. In the original IDEAL trial, patients treated with high-dose atovastatin didn't have significantly fewer major coronary events than those receiving regular dose simvastatin. A genetic assessment of IDEAL study patients showed that coronary event rates didn't change based on KIF6 carrier status.
But, here again, Celera argued that "since there was no placebo arm in IDEAL, an association of KIF6 genotypes with risk for CHD could not be assessed."
In the meantime, Celera is working with pharmacy-benefit manager Medco to build the clinical utility evidence in support of KIF6 testing. Last year, Medco and Celera launched a prospective, randomized study to evaluate whether patients' knowledge of their KIF6 gene variant status increases their adherence to statin therapy compared to those not offered the test (PGx Reporter 09/23/09). Positive findings from this Medco study would likely lead to the inclusion of the KIF6 test in Medco's personalized medicine program.
The "not approvable" letter from the FDA and the growing evidence questioning the utility of the KIF6 test do not appear to have changed Medco's plans for continuing the study. A Medco spokesperson told PGx Reporter that enrollment in the prospective study to evaluate statin adherence is complete and that Medco expects trial results in the third quarter.
Assimes believes that there is little reason for Medco to continue this study given the growing evidence showing that patients with the KIF6 variant don't see greater impact from statins. "This study should be stopped given the new evidence provided by our study, as well as the HPS and JUPITER studies," Assimes said.
"It is not ethical to motivate or not motivate individuals to adhere to statins (or any other therapy for cardiovascular disease) based on information about risk that is not accurate or precise."
Have topics you'd like to see covered in Pharmacogenomics Reporter? Contact the editor at tray [at] genomeweb [.] com.