NEW YORK (GenomeWeb) – A report last week authored jointly by researchers from the American Society for Clinical Oncology and the College of American Pathologists has put a damper on what some view as an overenthusiasm around liquid biopsy tests.
Though some platforms and some distinct uses of circulating tumor DNA have good evidence supporting both their clinical validity and utility, other areas remain unproven, including the use of broad sequencing tests to guide better drug choices, or the use of longitudinal liquid biopsy testing to monitor patient response to treatment or predict recurrence, the authors wrote.
The ASCO/CAP report, which is based on a literature review, doesn't directly discuss the growing marketing of commercial liquid biopsy tests, some of which are being sold for the same indications that the panel argues are not ready for prime time.
But Daniel Hayes, former ASCO president and an author on the report, said that the impetus for the review included both the fact that companies are advertising and promoting tests in these areas, and that the physician community may not have a robust understanding of whether or not these assays have proven beneficial for patients.
"Some of my colleagues at the University of Michigan won't have lunch with me anymore because of this … but in my opinion these tests should not be out on the market without proving that they improve care," Hayes said.
According to Hayes, reactions to the report have been a bit extreme, with some calling the conclusions "damning" for liquid biopsy testing. But he said that this "sky is falling" reading misses the parts of the report that are focused not on identifying problems, but on laying out clearly what types of research can be done and what kinds of evidence should be collected to fill these gaps.
"This is about stepping back and taking stock," or defining the questions that the clinical community should be asking about these tests and the kind of data that labs and companies should be preparing themselves to produce, Hayes said.
Rather than throwing a "wet blanket" on liquid biopsy, "if anything, I think having this kind of road map can help lead to doing the right thing faster," he added.
Hayes and his colleagues' analysis speaks to three important areas where enthusiasm for liquid biopsy is high. The first is the use of blood-based next-gen sequencing tests to probe patients for a broad panel of mutations in different genes to guide personalized therapy.
The issue here, according to Hayes, is that the question of the utility of broad NGS liquid biopsy tests depends on the question of the utility of NGS in general — something that the field has not yet definitively proven in tissue, let alone in blood.
"If we are not even sure this is the right thing to do using tissue, we should be careful about saying that it's the right thing to do in blood," he said.
The only randomized controlled study investigating the benefits of NGS-guided cancer treatment, SHIVA, showed that patients matched to a molecularly targeted therapy did not experience improved median progression-free survival compared to the control group, but had more serious adverse events. While experts have criticized SHIVA as flawed, the fact remains that a study correcting the conclusions of that trial has yet to emerge.
"Academic labs wanting to get this stuff out to patients are well-meaning and I understand that," Hayes added. "In [ASCO's TAPUR trial] I have had patients go on drugs I never would have given them otherwise with amazing responses … But it needs to be thought through a little more carefully than it is."
The report also provides some insight into the question of proving clinical utility — whether for tissue testing, or for liquid biopsy NGS. It can be easier, for example, the authors wrote, to examine utility retrospectively, but with a recognition that this is an imperfect measure and needs to be followed up with prospective confirmation studies.
A complicating issue for liquid biopsy NGS tests is that the field still lacks definitive data that defines exactly how blood-based testing compares to tissue, or how different assays compare to one another.
For single-gene tests, there is rationale for inferring clinical utility based on defining how the test compares to an equivalent test performed on tissue that has an established clinical function.
"If there is an established paradigm in tissue and your assay has a high analytical validity and if it gives you an answer that is equivalent to what you'd get in tissue and the utility in tissue is established, then you can do it," Hayes explained.
But even if clinical utility for tissue-based NGS becomes well established, it could be hard for broad liquid biopsy tests to take such shortcuts.
A second prominent area that the ASCO/CAP analysis addresses is the use of liquid biopsy assays for longitudinal or repeat testing to monitor whether patients are responding to a drug, to try to predict whether they will recur after a particular treatment, or to pick up early signs of recurrence.
Despite growing enthusiasm and encouraging small studies, the report concludes that based on a review of available evidence, the field has not yet proven clinical validity let alone clinical utility in this vein.
"Published studies are mostly retrospective, and few rigorous comparisons to established response metrics have been performed. In addition, no studies convincingly demonstrate improved patient outcomes or any cost savings when compared with standard-of-care monitoring approaches. There is no evidence supporting changing treatment at the time of ctDNA progression before clinical progression,” authors wrote.
Among the challenges for labs and commercial firms in filling the existing gaps is the fact that validating a test that quantifies tumor burden is more technically challenging than validating one that merely dichotomizes patients as variant detected or not detected.
For example, Hayes and colleagues wrote, "the best unit for quantifying DNA burden is not [even] established; most current approaches measure either the somatic variant allele fraction or detected somatic variant events per unit of plasma … The best approach to quantitation is currently unclear and will likely evolve in concert with what is needed for clinical utility and patient management."
In predicting relapse based on detecting minimal residual disease, the authors wrote that assays have yet to establish basic measures like the false-negative and false-positive rate, and the field is far from concluding that changing treatment based on the detection of lingering ctDNA might actually improve patient outcomes.
This doesn't mean that answering these questions is impossible, Hayes stressed. While the report's conclusions repudiate clinical adoption of liquid biopsy for monitoring right now, he said that the group tried to also express that excitement about this area is well-justified, considering proof of concept has been established in a growing number of tumor types.
The issue, Hayes said, is that these results, however exciting, don't rise to the level of definitive validity and utility evidence, yet this has not prevented companies from launching tests for commercial clinical use.
Several NGS panels are available commercially, despite the lack of clinical utility evidence highlighted by the report. And in monitoring and minimal residual disease testing, companies have also promoted clinical products.
Cynvenio Biosystems, for example, announced a CTC- and protein-based monitoring service last year intended to track disease status or surveil for early signs of recurrence in breast cancer patients.
Other early movers in this area include CellMax Life, which offers liquid biopsy tests for both early detection of new colorectal cancers and monitoring for resistance or recurrence in treated patients. Pathway Genomics was also at one point offering a blood-based mutation test called CancerIntercept Monitor both for early detection and recurrence monitoring. However, that test is no longer listed on the company's website.
Natera, which has been working to expand from non-invasive prenatal testing into the liquid biopsy space, is also planning on marketing tests specifically for ctDNA monitoring but has taken its time relative to some of these other firms.
The company is planning to launch a clinical version of its Signatera testing approach — which involves using whole-exome sequencing data from a patient's tumor tissue to define a small panel of mutations to follow in the blood — at the end of this year.
The company said it plans to share some of the first data on the analytical validity of the approach at the upcoming annual meeting of the American Association for Cancer Research. "We have narrowed down to indications and areas where we think ctDNA can make a real difference — where the tumor is going to shed enough for this to work as a biomarker, and where we can generate data relatively quickly," said Solomon Moshkevich, Natera's senior vice president of product and strategy, .
"Our plan is, as long as the retrospective data is good enough, then we can move forward with interventional clinical studies once our CLIA test launches," Moshkevich added.
According to Hayes, the intent of the report is not to detract from the promise of such tests, but to highlight the specific areas where proof is still lacking and help to define the types of studies that will need to be performed, and the data that clinicians should know to expect.
Moshkevich noted the final line of the report abstract — which reads "given the rapid pace of research, re-evaluation of the literature will shortly be required, along with the development of tools and guidance for clinical practice."
"We agree with the need for the data this report calls for, and we welcome the forward stance here. And from that [conclusion] what I really take away is that these guys want to stay ahead of the technology and are committed to keep pace with new evidence as it is published and to incorporate new data and information into clinical practice, as rapidly as possible" he said.