In an effort to help physicians give safer doses of cholesterol-lowering statin drugs, researchers from the University of Western Ontario have created an algorithm incorporating both genomic and clinical variables to predict which patients are more likely to have statins linger in their blood where they can cause uncomfortable, as well as rare and serious, side effects.
The team published a study describing the discovery this month in the journal Circulation: Cardiovascular Genetics. Central to the new predictor is the identification of a group of SNPs in genes that control transport of drugs into the liver, which the researchers found were significantly associated with higher statin blood-levels in a cohort of 299 patients treated with one of two statins, atorvastatin (Pfizer's Lipitor) or rosuvastatin (AstraZeneca's Crestor).
Richard Kim, the study's senior author and a researcher at the Lawson Health Research Institute, told PGx Reporter this week that though the group plans to continue to study the predictor to determine how much of an impact it would actually have on patient outcomes, they are planning to incorporate the pharmacogenomic information into clinical practice in the meantime.
"We do think that having these types of biomarkers is helpful today," he said. "What is not easy to quantify yet is how much better it is than standard of care."
The use of statins, which lower cholesterol levels in the blood, increased by 17 percent from 2007 to 2012, according to IMS Health statistics. Last year, while doctors wrote 214 million monthly prescriptions for statins, the US Food and Drug Administration warned healthcare providers and patients about the rare risk of liver injury, as well as the potential for developing type 2 diabetes and muscle damage among statin users. The agency has also been tracking reports of memory loss and confusion linked to statins.
However, given the rarity of these statin-related adverse events, doctors need tools to help them predict which patients are at risk. According to Kim, the team is currently seeking grant funding to support a prospective trial comparing standard dosing with PGx-guided dosing using the group's algorithm.
In the team's recent study, Kim and his collaborators looked at the presence of polymorphisms in several genes related to drug transport. The group prospectively followed nearly 300 patients taking atorvastatin or rosuvastatin at various doses. In order to identify polymorphisms linked to statin-induced muscle side effects, the researchers compared patients who took the same drug doses but had different drug concentration levels in their blood.
Kim explained that based on his group's ongoing research, it has become clear that muscle pain and weakness side effects — as well as the rare, more serious muscle damage, rhabdomyolysis — are linked to the impairment of the transport proteins that normally bring statins into the liver, leading to an overconcentration of the drug in the blood that can then affect muscle tissue.
In the study, the researchers found that several SNPs in these transport genes accounted for a significant amount of statin blood-level variability in the 300-patient cohort — more than 80 percent of variation in blood levels for patients on rosuvastatin and 40 percent of variability for patients on atorvastatin.
While age was also a significant factor in the variability for both drugs, gender, ethnicity, and BMI were not, the authors reported.
Based on the results, the group developed a predictive algorithm that recommends maximum doses for either of the two statins based on a patient's age and genotype.
Because about 10 percent of people on statin therapy are expected to experience muscle side effects, the predictor recommends doses that should result in plasma concentrations lower than the 90th percentile, the authors wrote.
Although Kim said his team is already planning to implement the algorithm in clinical practice, it will take a large prospective trial comparing the PGx-guided methodology to standard dosing strategies to determine whether the algorithm has a significant impact on patients' health and outcomes.
"There are many choices of statins, as well as doses at this point … so using pharmacogenomic [information] here would be, I think, better than standard care at this point," Kim said. "What's needed in terms of future study is asking the question – does this actually improve care, or reduce the number of severe muscle damage cases, or hard outcomes like death?"
According to Kim, his group is working on planning and gaining funding for such future research.
In the meantime, the study did offer some suggestion of the potential benefit of the PGx strategy. By retrospectively genotyping a second, nearly 600-patient cohort, the researchers found that almost half of patients receiving the highest statin doses using standard clinical criteria would be on a dose that exceeds the maximum recommended by their new algorithm.
Of 16 patients in the retrospective cohort taking 80mg of atorvastatin, for example, nine exceeded the maximum recommended dose according to their genotype, and only seven were still on such a high dose one year later. Conversely, the study authors highlighted that another seven subjects taking the same 80mg dose who were not predicted to be on an excessive dose based on their genotype remained at the same dose one year later.
For patients on lower doses in the retrospective cohort, none were predicted by the group's algorithm to be at an excessive dose based on their genotype, suggesting that the predictor may be most useful for patients being treated with or considering the highest doses of the drugs.
Kim said another important future research goal will be to study whether there is significant variation between different ethnic populations. The team's initial study was of a predominantly Caucasian cohort.