NEW YORK (GenomeWeb) - With personalized medicine and clinical cancer genomics moving further into the medical and social mainstream, patients are embracing the hope of new diagnostics and therapies.
But along with this hope comes the risk of hype, according to the authors of a new study published today in the Journal of National Cancer Institute, especially as more and more of these services are marketed or promoted online where patients encounter them directly, outside of an interaction with their physician.
According to the results of the study, while many genomic or other personalized medicine tests for cancer promoted online have well-defined clinical utility, many also do not.
"I think one of the things that happens is that there is a lot of hope on the patients' side and a lot of interest in new genomic technologies and targeted therapies," Stacy Gray, one of the study's lead investigators told GenomeWeb this week.
Gray and her co-lead investigator, Katherine Janeway, are both clinicians at Dana-Farber Cancer Institute. "Patients often come to the visit with lots and lots of printouts of things they've seen, whether its clinical trials or genomic tests they might want. So we wanted to really take a careful and critical look at the internet landscape to see what our patients are seeing when they go online," Gray said.
Several studies have examined the claims made in the internet genetic testing space and have found real instances of inaccurate representation of the science, Gray added. "But this had never been applied specifically to a cancer testing context and since it's an area where we are using genomic technology with increasing frequency, it looked like a particularly interesting case to study some of those question in a more granular way," she said.
To better understand the landscape of cancer personalized medicine tests that are promoted online and how often they do or do not meet the threshold for clinical utility, Gray, Janeway and co-authors set out to systematically catalog all the websites they could find marketing or promoting personalized cancer medicine and to determine whether the tests they identified met the current standard of care or not.
Overall, using a variety of internet search approaches, as well as a literature review, the group distinguished 55 websites, sponsored by commercial companies, academic centers, research institutions, or other entities, which marketed tests either including somatic or germline DNA analysis and interpretation, or other personalized care services over the internet.
Of these 55 sites, 32 specifically offered somatic testing or analysis, and of those, about half included information about specific tests or genes included.
The researchers then enlisted a panel of experts who were asked to determine whether the tests offered by the sites should be considered "standard" or "nonstandard" according to the definitions made by the US Center for Disease Control's Evaluation of Genomic Applications in Practice and Prevention working group. If 90 percent of the 14-member expert panel agreed that a test was "nonstandard," the study authors marked it so. Otherwise they did not.
"Unlike some of the other reports where they looked at claims and whether they were supported by the science ... We looked more at whether the tests that are being sold or promoted over the internet right now have evidence of clinical utility," Gray explained.
"We asked [this group of experts] to evaluate each test as to whether, at that point in time, it would be considered standard of care," she said.
Among the online sites that did describe details about what genes they tested for, or other aspects of their technology, the panel found overall that the vast majority of sites marketing personalized cancer treatment or testing promoted at least one test that did not currently have clinical utility according to 90 percent of the study's expert panel.
"We saw that it really runs the whole spectrum, from sites marketing tests that clearly have clinical utility, that are guideline-endorsed, and that really should be used in many patient populations, to things that we really don't think are very helpful at all," Gray said.
For example, many sites offered testing for KRAS and EGFR mutations, which the panel widely agreed have clear clinical utility in current practice. On the other end of the spectrum, Gray highlighted whole-exome sequencing and cell-based chemotherapy sensitivity testing as two prime examples of tests promoted online that don't have definitive clinical utility.
According to Gray, WES is an extremely promising technology, "but right now the use of whole-exome sequencing to identify targets for treatment for cancer has not been proven," she said.
Where chemotherapy sensitivity testing is concerned, "there have been actually two statements from ASCO saying that those tests do not have clinical value," Gray said, "so not only are they not guideline-endorsed, but there are actually guidelines against them, yet they are being marketed as a way to personalize cancer therapy."
Of course, Gray said, the goalpost of clinical utility is constantly shifting, and arguments can be made that tests that haven't yet been successfully adopted into the forefront of clinical practice could still prove themselves to have important impacts on patient outcomes in the near future.
But interestingly, Gray said, when the researchers compared how many tests were "nonstandard" according to 90 percent of their expert panel with how many were considered "nonstandard" by only 50 percent of the panel, they found that the number was almost identical. This suggests that, at least among this group of experts, there were not a lot of assays for which there was significant disagreement on their current clinical utility.
"It was pretty much the same no matter how we categorized it, meaning that a lot of the things that were out there just really had no evidence that they would help improve outcomes," Gray said.
According to Gray, having completed their profile of online marketing and promotion, the researchers are now working on finishing a similar study cataloguing the portrayal of genomic testing and personalized medicine in the media.