The results from two widely used genetic sequencing tests can differ "markedly" in the same patient, according to investigators who compared the FoundationOne (F1; Foundation Medicine) and Guardant360 (G360; Guardant Health) tests.
In other words, the genetic characteristics of a single patient's cancer could have one profile with one test and a very different one with the other test.
The test results are used to guide cancer treatment choices in patients who have advanced cancer and in whom standard therapies have failed.
"Since both the F1 and the G360 tests are performed in thousands of patients with cancer each year, these findings are clinically relevant," write the study investigators, led by C. Anthony Blau, MD, from the University of Washington in Seattle.
Their research letter, with data on just nine patients with advanced cancer, was published online December 15 in JAMA Oncology.
The new study reveals "widely discordant results," agreed Vinay Prasad, MD, from the Oregon Health and Sciences University in Portland.
Dr Prasad, who was asked for comment, explains that despite having a small number of patients, the analysis has a big impact.
"What we are talking about here is precision oncology — the use of sequencing in patients with a range of cancers who have exhausted other options in order to pair them with therapies," he told Medscape Medical News. "A very basic question about this approach has been: Would two different tests give the same result?"
"This [a high level of concordance] is a prerequisite to the tests being of any clinical use," he continued.
"This paper shows that the majority of the time, the tests give different answers. That is the equivalent of dropping a ton of bricks on this approach," Dr Prasad.
Dr Blau was also critical.
"Our paper shows that these tests do not provide a 'plug and play' solution for arriving at effective therapies," he told Medscape Medical News.
The results from two widely used genetic sequencing tests can differ.
He was referring to the fact that drugs are recommended according to the test results, but that, in the study, the recommendations were mostly not in unison.
However, Dr Blau also believes that the tests are valuable. "Thousands of patients have benefitted from the information that these tests provide," he said. In their research letter, the study authors say the tests need to be compared in larger numbers of patients in order "to improve concordance and clinical utility."
Study Only Compared Alterations Identifiable by Both Tests
The F1 test sequences clinical tumor samples to characterize 315 cancer-associated genes and 28 genes involved in rearrangements. The G360 test uses circulating DNA from blood to sequence 70 genes.
In the study, all 9 patients (5 with breast cancer and 4 others with pancreatic, lung, salivary gland, and thymic cancers, respectively) had both tests.
Importantly, the researchers limited comparisons to alterations identifiable by both F1 and G360 testing.
The genomic alterations reported by each test were compared for each patient. And the results were also assessed to compare the recommended drugs.
One patient had no genetic alteration identified by using either platform. The other 8 patients had 45 alterations, but only 10 (22%) were concordant between the platforms.
In other words, the majority of the test results were discordant.
Furthermore, for 2 of these 8 patients, there was zero concordance among the described alterations.
For the 8 patients, a total of 36 drugs were recommended. However, the authors say that only 9 drugs (25%) were recommended for the same patients by both platforms, and in 5 patients, there was no overlap between the drugs recommended by the F1 test and those recommended by the G360 test.
So, in the majority of the patients, there was zero concordance in terms of drug recommendations.
Dr Prasad believes that these genetic tools should ultimately be required to prove that they improve outcomes — namely, overall survival, because that is the standard for other cancer practices. "These tests have no credible evidence they make people better off," he said.
"They certainly drive up costs, with prices of several thousand dollars," Dr Prasad added.
Many critics have called for clinical trials to determine whether the mushrooming industry of cancer genetics testing in the context of treatment decision making actually improves outcomes, as previously reported by Medscape Medical News.
Dr Blau said "realistic goals" are needed in assessing the clinical value of information provided by tests such as F1 and G360. "These tests are typically used in the setting of advanced cancer, where cures are few and far between. Even under the best of circumstances, treatments based on the results of these tests are almost never curative," he commented .
Dr Blau and Dr Prasad have disclosed no relevant financial relationships.
JAMA Oncol. Published online December 15, 2016. Abstract