About 25 to 30 percent of people prescribed statins dump them within a year. I flunked Lipitor after a few wretched months.
Statins are prescribed to lower cholesterol in people who show risk factors for cardiovascular disease or diabetes, or who already have them. Side effects can include muscle weakness, diabetes onset and, rarely, permanent muscle damage. These risks are higher in women, with age, and with certain heart and blood pressure drugs.
Frustrated with trial and error, I was ready to swap some DNA for some personalized insight. I sent a vial of spit to Boston Heart Diagnostics, a Framingham, Mass., company that tests for statin tolerance using SLCO1B1, a gene that was associated with statin myopathy in studies at Oxford and Duke universities. The company currently only offers this test as a blood test through health care providers but, because I'm a science writer, offered to guide me directly through the process.
Genetic tests are conducted using DNA fragments extracted from blood, cells or saliva, which are then copied and amplified to get enough material to test for the precise variant of interest. Then the results are analyzed and shared with the patient. Dr. Ernst Schaefer, the company's chief medical officer, would act as my health care provider to interpret the results.
The Boston Heart Diagnostics test identifies three genetic variants, called single nucleotide polymorphisms ,or SNPs, on the SLCO1B1 gene that are associated with how well one metabolizes statins.
According to the report I received the following week, I was among the 25 percent of people with one copy of the SLCO1B1 variant who metabolize statins like Lipitor or Zocor poorly. Meaning I am 4.5 times more likely to develop muscle myopathy than people without the gene. At least I wasn't among the 2 percent of the population who have both SLCO1B1 gene variants. Then I would be 17 times more likely to develop muscle weakness than the people without the gene.
I called Schaefer for more detail. "The variant is in the transporter that takes up statin in the liver, so people that have the variant have higher blood levels and the statin is more likely to go into muscle," he explained. "Our test is not diagnostic, not confirmatory, just indicates a higher risk of getting a problem on a statin. What is diagnostic is how the patient feels."
Schaefer recommended that instead of the high dose of Lipitor my doctor had prescribed, I consider a low dose of a different type of statin.
Boston Heart is initially targeting its test for people with genetic high cholesterol who are prescribed a PCSK9 inhibitor, a drug newly approved by the Food and Drug Administration. Injected monthly, PCSK9 inhibitors have fewer side effects than statins, but the first drug to hit the market costs over $14,000 a year, compared with $600 for a year's supply of atorvastatin, the generic name for Lipitor.
Boston Heart has so far tested over 250,000 patients for statin intolerance but can't yet comment on its prevalence in specific ethnic groups. While this test is not approved by the Food and Drug Administration, it is reimbursed by some insurance.
The test costs $100. My health plan still considers prescribing generic drugs for low-risk patients like me on a trial-and-error basis to be cheaper.
Insurers will start paying when genetic testing becomes the standard of care agreed upon by physician committees, said Dr. Julie Neidich, a geneticist and laboratory medical director at Pathway Genomics of San Diego, which also makes a genetic test panel that includes statin intolerance.
Could DNA tests like this be the vanguard of personalized medicine? A few months ago, President Obama announced a $215 million precision medicine initiative to collect DNA records from 1 million volunteers to develop medical treatments tailored to individual genetic makeup.
"I think we aren't there yet," said Neidich. "While humans have about 20,000 genes, only about 5,000 genes have been reported in the scientific literature to be associated with a disorder found in humans. We have no clue what the other ones do."
Four DNA-based drugs were just approved by the FDA, and in 2014, nine of the 41 FDA-approved medicines were personalized treatments.
With the exception of new genetically targeted drugs for chronic hepatitis C, where patients' response depends on the genotype of their infection, all these personalized drugs treat rare diseases and unusual forms of cancer, where disease is caused by one gene or a handful of genes.
Health insurance companies are watching these developments like hawks. Indeed, the research arm of Kaiser Permanente, my health care plan, has been accumulating a genetic data bank to study complex conditions like heart disease and diabetes.
So far over 200,000 Kaiser Permanente members have contributed DNA samples to be analyzed along with their health records. This matters, because small effects are easier to detect over a large population, particularly in ethnic groups not well-represented in prior clinical trials.
So what can cardiac researchers conclude so far from these genetic data? Dr. Ronald Krauss, a lipid specialist at Children's Hospital Oakland Research Institute and the University of California, San Francisco, has been studying cholesterol metabolism for over 40 years and has been collaborating with the Kaiser study.
Krauss is concerned about the unseen effects of statin use over time. "The focus on cholesterol and on relatively uncommon adverse effects of statins such as muscle damage and onset of diabetes could be the tip of the iceberg in terms of biological effects with clinical consequences for some individuals," Krauss said. "These could be either negative or positive. In the lab, we see widespread effects of statins at the cellular and molecular level. Are subtle things happening that affect health over the long term?"
But unlike my single gene test for muscle weakness, predicting a risk or treatment response for complex conditions like heart disease involves sifting through a dog's breakfast of genetic variants of varying impact.
"It is not like Huntington's disease, where a single identifiable genetic variant can give you the disease," said Krauss. "In complex conditions like heart disease, there are multiple genetic effects, but for the most part, each individual genetic trait has a modest effect."
But the complications don't stop there. Other factors that can affect a person's disease risk and response to treatment include gut bacteria, RNA, acquired DNA changes, lifestyle and environment. Parsing out the impacts of these variables will require massive computing power and gargantuan data storage capacity to deal with the flood of genetic information to be sequenced and interpreted.
"As we have learned more about the biology of statin response, it is clear that this involves major regulatory networks, and this is big stuff," said Krauss.
So, time for another cholesterol test. Apparently I'm prone to high cholesterol, no matter how many vegetables I ingest. At least I can experiment with Schaefer's statin recommendations until the bioinformaticians surfing the precision medicine data deluge figure out the big stuff.
Wendy Wolfson is a science writer in Orange County, Calif.