Date: March 5, 2018
So called “liquid biopsies” — which can detect circulating tumor DNA (ctDNA) in blood samples — are not yet ready for prime time in the diagnosis or management of early-stage or advanced solid tumors, a new expert review concludes.
These assays are also not useful, outside of clinical trials, for monitoring patients for minimal residual disease following definitive treatment of cancer, nor for cancer screening, the expert review concludes.
The review was prepared jointly by the American Society of Clinical Oncology (ASCO) and the College of American Pathologists (CAP) and was published online March 5 in the Journal of Clinical Oncology.
“This is an area of great interest to both pathologists and oncologists, [and] it’s also an area where we see a lot of commercial advertisement and a lot of enthusiasm from the public,” Jason Merker, MD, PhD, cochair of the expert panel, who was representing the CAP, said in a statement.
“We thought it was a good time to look at the literature and take an evidence-based approach to various uses for ctDNA assays,” he added.
“Like all new things in medicine, the use of ctDNA assays in routine cancer care requires evidence of clinical utility. At present, there is insufficient evidence of clinical validity and utility for the majority of ctDNA assays in advanced cancer, including those that interrogate a panel of genes,” said Daniel F. Hayes, MD, coauthor of the review, who was representing ASCO.
“What is promising is that this area of research is rapidly evolving, so there should be enough evidence soon to formulate evidence-based guidance for a variety of clinical scenarios,” Hayes suggested.
Review Based on Literature Review
For the review, panel members identified 77 relevant articles in the literature. They limited their analysis to variants in ctDNA for solid tumors and to sequence or copy number variants in DNA.
They assessed overall evidence of the ability of a test to reliably detect the variant or variants of interest (analytic validity); whether a test accurately detects the presence or absence of a pathologic state or predicts outcomes for patients (clinical validity); and whether the use of the test improves patient outcomes compared with not using it (clinical utility).
The authors focused largely on the use of ctDNA assays in the setting of metastatic cancer, because that is the area for which there is most evidence. Much less research supports the use of liquid biopsy in other settings.
“Fundamentally, there are two paradigms to demonstrate clinical utility and the adoption of ctDNA as a clinically useful test,” the panel members write.
The most reliable strategy is to conduct a prospective clinical trial designed to evaluate how well the test performs as a stand-alone diagnostic test. However, no such trial has yet been carried out, they write.
The second approach is to assess whether the ctDNA test in question delivers the same information that physicians would seek through tissue genomic evaluation.
“[D]emonstrating that a ctDNA assay has high agreement with tumor tissue genotyping may provide sufficient evidence of utility for ctDNA assays in driving patient treatment decisions,” the panel members explain.
They go on to document how far short ctDNA assays fall with respect to meeting this high level of agreement.
First, only one polymerase chain reactin–based ctDNA test has been approved by the US Food and Drug Administration and the European Medicines Agency. That test is the COBAS assay for the detection of EGFR genetic variants in non–small cell lung cancer (NSCLC).
Another assay is available in Europe for the detection of the KRAS mutation in colorectal cancer (CRC).
“These assays have demonstrated clinical validity but the clinical utility in this setting is based on retrospective analyses,” the panel members point out.
Even in light of these approvals, the panel members note that physicians should continue to rely on tissue sample analysis if nothing is detected on ctDNA testing.
They also point out that ctDNA levels may drop while a tumor is still responding to treatment, and as a result, the sensitivity of the test may be compromised.
As for tumor types other than NSCLC and CRC, there is limited evidence to support the clinical validity of ctDNA analysis, the panel members conclude.
The clinical utility of assays developed for the detection of other potentially targetable variants, such as BRAF in melanoma, is not yet established, they add.
Further confounding the diagnostic potential of ctDNA testing is the fact that advanced cancers may be “genetically heterogeneous.”
Although ctDNA tests may be able to pick up subclonal variants in cancer, these variants may not predict how well patients will respond to treatments that theoretically target the variant.
“[S]ubclonality may undermine the clinical utility of ctDNA assays,” the panel members state.
Monitoring Response to Therapy
Ideally, liquid biopsies could help physicians monitor patients’ response to treatment, as some tumor-associated proteins now enable them to do.
This, too, remains a challenge, because quantifying changes in ctDNA over time is not as simple as determining whether or not a variant is present. Nor has the best unit by which to measure DNA burden been established.
“Correlations between changes in ctDNA levels and tumor responses or outcomes have been demonstrated in small proof-of-principle studies in a variety of cancer types,” the panel members acknowledge.
“However, currently there is a lack of rigorous evidence on clinical validity, let alone clinical utility, because few large, prospective validation studies have been performed on ctDNA-based monitoring,” they conclude.
Use to Monitor Residual Disease
Researchers expressed hope that after curative treatment of a solid tumor, ctDNA assays could be used to monitor patients for minimal residual disease, much as is done in hematogic malignancies using assays to detect leukemic cells in blood following completion of chemotherapy.
However, there is not enough evidence to support the ability of ctDNA tests to detect low levels of minimal residual disease in a manner similar to assays used in the management of diseases such as leukemia, the panel members conclude.
Moreover, “[t]he false-negative rate of ctDNA analysis in…patients who relapse without ctDNA being detected and the false-positive rate [in] patients who do not relapse despite the ctDNA assay being positive have not been established sufficiently for any assay,” they caution.
There is also no evidence that treatment based on detection of ctDNA improves patient outcomes, a major metric of clinical validity.
Screening for Cancer
In an ideal world, ctDNA tests could be used in the early detection of cancer in patients who have no signs of disease.
However, the feasibility of using ctDNA tests to screen asymptomatic individuals has not been demonstrated.
“[D]iagnosing the presence of cancer in a patient without cancer, and determining tissue of origin, have not been established,” the authors point out.
There is also a risk that such tests might be positive for cancers in cases in which none exists and thus lead to overdiagnosis. Currently, overdiagnosis is a major problem with, for example, mammography in breast cancer and prostate-specific antigen screening for prostate cancer, the panel members observe.
Dr Merker has served as a consultant to or in an advisory role for Bio-Rad Laboratories, Rainbow Genomics, and Genoux and has received patents or royalties or has other intellectual property in the measurement and monitoring of cell clonality. Dr Hayes owns stock or has other ownership interests in OncImmune and InBiomotion and has served as a consultant or as an advisor to Cepheid. He has also received research funding, mostly for his institution, from AstraZeneca, Puma Biotechnology, Pfizer, Eli Lilly, Merrimack Pharmaceuticals, Parexel, and Menarini Silicon Biosystems (aka Veridex/Johnson & Johnson). He also holds a number of patents and receives royalties for some of them.