Originally published March 27.
By Turna Ray
CHICAGO – Researchers from the Medco Research Institute and Quest Diagnostics subsidiary Celera this week reported data from an observational study that found that patients who are taking statins are more inclined to keep taking the medication if they know about their genetic risk for cardiac conditions.
At the American College of Cardiology's annual meeting this week, Medco Research Institute's Scott Charland and others presented data from the Additional KIF6 Risk Offers Better Adherence to Statins, or AKROBATS, study. The prospective, non-randomized, observational trial, involving more than 1,200 patients whose prescriptions Medco manages, demonstrated that during a six-month period, study participants who had been genetically tested for the KIF6 variant were significantly more "adherent and persistent" in filling their prescriptions for statin meds compared to those who didn't receive genetic testing.
Medco conducted the genetic adherence study in spite of the fact that data in the published literature have yet to definitively establish that the KIF6 Trp719Arg allele is associated with increased risk of cardiovascular events. Furthermore, the retrospective pharmacogenetic analysis that Celera filed as part of its premarket approval application for the KIF6 test elicited a "not approvable" letter from the US Food and Drug Administration last year.
However, Eric Stanek, VP of the Medco Research Institute and a senior study author, maintains that the totality of evidence on the marker suggests there is a genotype-phenotype interaction. Data from AKROBATS suggest that "patients may have more personal utility for knowledge of their KIF6 carrier status, which is then translated into behavioral modification," the study authors concluded in a poster presented at ACC.
Without factoring in genetics, studies have shown that when patients take statins as prescribed by their doctors, they experience positive health benefits and contribute to lower healthcare costs. A Medco study published in the American Journal of Cardiology in June reported that patients who took statins as prescribed more than 90 percent of the time reduced total healthcare costs by $944 per patient over 18 months and significantly lowered their risk of hospitalization due to heart complications.
Through its personalized medicine programs, Medco is conducting observational studies in order to gather the clinical evidence to drive adoption of genetic tests in medical practice when they can reduce adverse reactions, improve drug dosing, and encourage treatment compliance, such as with KIF6 testing. These interventions, Medco is hoping, will not only improve patient outcomes, but also increase cost savings.
Celera and Medco announced two years ago they would collaborate on the AKROBATS trial (PGx Reporter 9/23/2009). At the time, studies suggested that patients who harbor the KIF6 Trp719Arg allele may be at heightened risk for coronary heart disease and predisposed to having a more robust response to statins than those without the mutation. Based on this data, Celera began marketing the KIF6 Genotyping Assay as a lab-developed test through its Berkeley HeartLab business. Quest acquired Celera last year.
Patients were enrolled in AKROBATS if they were 18 years or older and newly prescribed statins during an eight-month period in 2010. Medco contacted more than 1,500 customers and enrolled 647 people in the genetic testing arm. KIF6 results were mailed to participants within 30 days of starting statin therapy. In the comparator arm, the researchers enrolled the same number of patients who weren't tested for the KIF6 variant but were similar to those in the intervention arm in terms of statin prescription delivery channel, age, gender, and chronic medications.
Analyzing participants' prescription claims records, the researchers used a metric called the proportion of days covered, or PDC, to determine that participants who were genetically tested had prescribed statins on hand for an average of 77 percent of the days during the six-month period, compared to 68 percent of the time in the control arm.
Stanek told PGx Reporter that patients who had their meds on hand 80 percent of the time were categorized as being adherent. Based on this criterion, 63 percent of those tested were adherent compared to 45 percent of controls.
The researchers also found that 69 percent of genetically tested patients were "persistent" with statin treatment – meaning that they continuously refilled their prescriptions without gaps – compared to 53 percent of individuals in the control group.
The study's authors note, however, that measuring PDC may overestimate actual adherence to statins, since it assumes that individuals readily filling their prescriptions were also taking the medications. Stanek acknowledged that the study did not assess whether participants were actually taking their statins, but noted that PDC is a well-validated measure to track treatment compliance.
The pharmacy benefits manager also accounted for the fact that study participants — who were contacted about the study, told about genetic testing, and agreed to participate — would be more prone to take their statins simply because they had committed to being in the trial. "We controlled for non-study related physician and pharmacist contacts," Charland said.
These controls were intended to prevent the type of criticism that Medco received
at the 2010 ACC annual meeting, when it presented data from a observational study investigating whether PGx testing to administer the anticoagulant warfarin improved patient outcomes over standard monitoring methods. In that investigation, the researchers did not randomize the study population or blind physicians to which patients were genetically tested. The lack of blinding was viewed by many in the research community as a study design flaw, since the benefit of genetic testing may have been positively bolstered by the so-called Hawthorne effect — a scenario in which study subjects improve behavior simply in response to being studied and not as a result of the intervention (PGx Reporter 3/17/2010).
Researchers informed patients who declined to participate in AKROBATS that Medco still could track whether they were filling their statin prescriptions. This knowledge appeared to improve the adherence in non-participants, but not as much as participants who had information about their KIF6 status.
"As part of the presentation of the study opportunity to those who went on to decline, the concept that Medco has the routine ability to monitor adherence to prescription medication was introduced," Stanek explained. "This could have had the effect of improving their adherence, and since the KIF6 testing group still did significantly better than they did in this control group, it helps reinforce the validity of our findings."
Strength of the Association
While Medco has been conducting AKROBATS, published studies about the association between the KIF6 719Arg variant and cardiac events have come to divergent conclusions. As such, industry observers have questioned whether the KIF6 adherence study should be conducted since the genotype-phenotype association hasn't been validated.
In October 2010, a large meta-analysis conducted by Themistocles Assimes of Stanford University and colleagues failed to find any association between the KIF6 719Arg allele and coronary artery disease. After investigating data from 17,000 CAD patients and more than 39,000 controls from 19 different studies conducted globally, the researchers said that if there had been a significant association between the KIF6 variant and coronary disease, they would have found it (PGx Reporter 10/13/2010).
Then, Celera received a "not approvable" letter last year from the FDA for its KIF6 Genotyping Assay. As part of its premarket approval submission for the KIF6 test, the company conducted retrospective analysis of several large, prospective studies showing KIF6's association with cardiac events and statin benefit. However, the FDA informed the company that it needed 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 (PGx Reporter 4/27/2011).
After the Assimes et al. meta-analysis was published, Medco discussed it with the institutional review board overseeing the AKROBATS trial. "We took this data to the IRB and asked if we should tell patients about this data," and the IRB advised that the study should continue under the original protocol, Stanek said.
However, Stanek cited a more recent study to note that subsequent investigations are beginning to gather insight into how KIF6 affects cardiac events.
Brian Ference of Wayne State University and others conducted several analyses to study the association between the KIF6 allele and the risk of CVD, including a meta-analysis of 144,000 patients and a meta-regression analysis of 88,000 participants. Although the meta-analysis of 144,000 patients showed no direct association between the KIF6 719Arg allele and relative risk of cardiovascular disease, the meta-regression analysis suggested that the variant lowers CVD risk by lowering LDL levels.
"The KIF6 719Arg allele increases vulnerability to LDL cholesterol and thereby influences the expected clinical benefit of therapies that reduce LDL cholesterol," the study authors wrote in PLoS One last year.
Stanek said that after considering the totality of the published data, he remains confident that messages to AKROBATS participants about the KIF6 marker wouldn't have changed during the study.
A Quest spokesperson told PGx Reporter that the company is continuing to discuss the regulatory status of the KIF6 test with the FDA but could not provide further details. However, the assay continues to be marketed as an LDT.
AKROBATS "did not provide insights into the clinical value of KIF6 testing, and we do not therefore expect the study alone will bolster adoption of KIF6 testing," the spokesperson added. "However, we believe the study's findings are significant for they demonstrate that patients who are engaged in their own healthcare make better patients. This finding has implications for clinical management of patients across a broad range of diseases, testing services and treatments."
Have topics you'd like to see covered in Pharmacogenomics Reporter
em>? Contact the editor at tray [at]
genomeweb [.] com.