By Beth Daley
The area of pharmacogenetics, or how genes affect a person’s response to drugs, is a fast-growing commercial segment of genetics. The basic science is sound and decades old. It involves identifying how a patient responds to medicine, and so helps physicians avoid bad reactions and figure out what dose is best for each individual. Today, clinical laboratories are selling hundreds of these tests to patients and doctors without substantive trials, independent validation or solid proof that they actually are accurate or even useful to patients. Among the fast-growing areas for these tests are psychiatry and in assessing how patients respond to opioids.
When writing about how these tests work, skepticism is in order. These tests are highly complex and their algorithms are proprietary. For example, if a lab develops a test to help a doctor decide if a patient with ADHD should take a certain drug and at what dose, it is not always clear to me which genes these tests are targeting. In 2005, the FDA approved its first pharmacogenetic test and since then it has approved dozens of other tests and included warnings about such tests. But the fact remains most of these tests are not FDA vetted or approved.
As I reported last fall, there is an issue of conflict of interest because patients often cannot determine if their doctors stand to gain financially by recommending personalized medicine tests. In 2014, the federal government established a database to disclose financial relationships between physicians and the health care industry, but it excludes most genetic tests.
Most pharmacogenetic tests often fall under the category of laboratory-developed tests (LDTs) which the FDA does not regulate under a decades-old regulation exemption. This year, the FDA has pledged to start regulating LDTs, a process that will take up to a decade because there are an estimated 15,000 LDTs available in the United States. However it is still unclear if those rules will actually go into place. It is unknow how many are pharmacogenetics tests. The end result is there is little independent or meaningful review of these tests and so it is difficult to know if they do what the labs say they do.
For this reason, journalists writing about these tests should ask for any peer-reviewed studies that researchers have done on these tests. Unfortunately, however, such studies may be biased because many are funded by the companies that develop these tests. I found one pharmacogenetic test for depression that a Medicare contractor approved, but a review of the studies the company submitted revealed that all were funded by the company that developed the test. Plus, the tests were not as robust as the company claimed. One supposedly double-blind randomized trial was not double-blind or randomized in the way most drug trials are conducted.
In addition to asking for peer-reviewed studies, journalists should review the arrangement the lab company developing these tests is getting them to patients. Sometimes, these companies hold trials in doctors’ offices and so the trials are not listed on ClinicalTrials.gov, a registry and database of results of publicly and privately supported clinical studies on human participants. If the lab’s so-called study is not listed on clinicaltrials.gov, it may be just a way to get patients to take the tests so that the lab can bill the patient’s health insurance company. If this is the case, the patient could be stuck with a large bill for a test the insurer rejects for payment.
One problem when reporting on pharmacogenetics tests is whether there is any harm to patients from taking these tests. Showing harm is difficult because these tests often guide physicians and patients to avoid FDA-approved drugs or they are used to guide a physician the proper dosage for a patient. Therefore, saying a test result is harmful can be hard to prove for several reasons. First, we don’t know if the test is valid and, second, it guides patients to drugs a doctor can prescribe. For me, determining patient harm was one of the most challenging issues when writing about these tests.
While a test may or may not cause harm, these tests can cost thousands of dollars and so what patients and insurers pay can be harmful.
Looking back over what I’ve learned in covering pharmacogenetics is that the issue is a big, relatively new, and the technology is changing quickly. Therefore, I have relied heavily on non-partisan sources such as the Hastings Center in New York, academics, and experts in the specialty involved, such as pharmacogenetics for cancer drugs or psychiatry.
Also, I found that I wasted a lot of time just learning the basics of the field and reading scientific papers but uncertain about whether my research was taking in the right direction. Now, I go directly to experts to help me evaluate research papers and to comment on them. Also I got too deep in the weeds trying to understand the genetics and how patients respond to drugs. I didn’t need that level of detail to write the stories I’ve done.
All of these lessons led me to conclude that as a journalist, I need to trust my gut instincts. When a company website promises great results quickly without good independent published research in the literature, I’ve learned to ask if this company is making a major leap forward in science that no one else is making or is this company marketing a test that is as yet unproven?
Here are links to the award-winning work on Unregulated Tests:
- Oversold and misunderstood: Prenatal screening tests prompt abortions
- Lyme disease
- Physician society urges better understanding of new prenatal tests
Beth Daley (@BethBDaley) is an investigative reporter and trainer for the New England Center for Investigative Reporting. In 2014 she won a first-place Award for Excellence in Health Care Journalism in the Investigative Reporting (small) category for her work on Unregulated Tests.
