NEW YORK (GenomeWeb) – This month's annual meeting of the American Society of Clinical Oncology hosted numerous updates on the use of liquid biopsy tests to monitor cancer patients either after surgery or during treatment with certain drugs. Highlights included abstracts on bespoke or patient-specific platforms, as well as more generalized single-mutation and genome-wide assays.
Both the commercial companies that have staked claims in this arena and researchers pioneering new methods have argued that such techniques could significantly change the way oncologists are able to treat their patients: whether by catching relapses early, before they become symptomatic, or by giving a quicker and more precise picture of whether a particular treatment is working.
At the same time, leaders in the field have cautioned that the evidence is not yet in on the ultimate utility for these platforms — for example, whether catching a relapse earlier truly helps a patient live longer or better, or whether being able to track response to a drug more precisely would change significantly how a doctor treats a patient.
A joint review on circulating tumor DNA issued last March by ASCO and the College of American Pathologists concluded that widespread use of circulating tumor DNA (ctDNA) assays in most patients with advanced cancer —  especially for monitoring or surveillance — is still an area of ongoing research.
"The technologies are getting better and better and the field is still evolving. But if you still benchmark liquid biopsy to standard imaging techniques,  … I would say don't stop doing the imaging and just do liquid biopsies only [right now]," said Chi Van Dang, scientific director of the Ludwig Institute for Cancer Research.
However, he added, "the progress has been quite remarkable." Data is accumulating, for example, on how longitudinal monitoring of blood-borne DNA can provide a window into tumor heterogeneity and evolution that might be clinically actionable, at least in some cancers.
"Multiple studies have shown that [liquid biopsy] can [signal] resistance to treatment even before clinical signs manifest," said Luca Quagliata, global head of medical affairs for clinical NGS and oncology at Thermo Fisher Scientific.
Prior to the ASCO meeting, investigators working with Thermo products shared their own data at the European Association for Cancer Research Joint Conference on Liquid Biopsies in Bergamo, Italy, in May, comparing tissue sequencing results and liquid biopsies in 159 non-small cell lung cancer patients.
While concordance was extremely high when liquid biopsy was conducted immediately after the initial tissue biopsy, it dropped down to just 72 percent when plasma NGS analysis was performed at later stages — illustrating the "genetic drift" caused by the development of new mutations during treatment.
Importantly, investigators reported that they saw clinically relevant mutations that were not present in an initial tissue biopsy, including EGFR, ALK and BRAF mutations, emerge in later plasma samples.
At ASCO, new research hinted at how viewing this evolution in real time could impact clinical practice. Many of these results were from efforts to use liquid biopsies to provide a window into early treatment response or efficacy, something that can be hard to assess from clinical features, patient symptoms, or even imaging, especially in newer drug classes like immunotherapy.
In one poster at the conference, researchers from Mayo Clinic Florida and the University of Florida Health Cancer Center in Jacksonville reported on their implementation of Guardant Health's Guardant360 assay to assess patients at diagnosis and then again at progression.
In a cohort of 127 colorectal cancer patients who had this serial testing, the investigators saw mechanisms of resistance emerging at progression, including acquisition of KRAS, NRAS, and EGFR mutations, and emergence of HER2 and MET amplifications. When targeted or biologic drugs were then stopped, the group saw various subclonal, emergent mutations disappear from the patients' blood — something that did not happen for those on chemotherapy.
According to the authors, not only was the information informative of tumor evolution and resistance, it was of actual value for modifying treatment, and selecting clinical trials, at least for some patients.
The findings reflect those of other groups in this and other tumor types and provide a potential new paradigm for treatment in which molecularly targeted drugs might be started, stopped when resistance emerges, switched, or even started again after resistant subclones go quiet.
Investigators at the Princess Margaret Cancer Centre in Toronto, for example, reported on a study of patients with solid tumors receiving immune checkpoint blockade with pembrolizumab (Merck's Keytruda) in the Phase II INSPIRE trial who were tested using Natera's patient-specific assay methodology.
In 70 patients tested at baseline and then again at the start of cycle 3 in the trial, the group saw that a change from a positive to a negative test between those two timepoints was strongly correlated with a variety of outcome measures.
The meeting also hosted data featuring approaches that don't require development of a patient-specific assay to track ctDNA. Memorial Sloan Kettering Cancer Center researchers reported on an analysis of patients in three trials of the immunotherapies durvalumab and tremelimumab who were tested with Guardant Health's Guardant360, for example, concluding that patients showing greater reduction of variant allele frequency had both improved progressionfree and overall survival.
Currently, monitoring whether a drug is working or not presents more immediate usefulness for pharmaceutical trials than it does for the practice of oncology. But some data is also beginning to supply evidence for impact in the clinic.
Investigators reporting on the COMET study, which assessed the prognostic value of early ctDNA and circulating tumor cell changes in metastatic breast cancer patients treated with first-line chemotherapy- found that ctDNA changes were predictive of favorable outcome, though the clinical utility of this "remains to be demonstrated."
Researchers from New York University School of Medicine and the Perlmutter Cancer Center shared results from another study using Bio-Rad's droplet digital PCR platform to track blood-borne BRAF mutations in advanced-stage melanoma patients treated with the BRAF inhibitors dabrafenib and trametinib.
The team could detect BRAF alterations in the blood of 93 percent of a cohort of about 350 patients before they began treatment. Looking at the subset of patients who responded best to the treatment, about 40 percent no longer had detectable mutations in their blood after one month of therapy. In contrast, 60 percent of the patients with poor responses had lingering BRAF-mutated ctDNA.
Based on these results, the study authors said in a statement, tracking BRAF ctDNA might be an improvement over the tools clinicians have now for assessing whether a patient is responding to treatment, as they are beating enzyme and protein biomarkers on accuracy, and CT scans in terms of timeliness.
In a study of lung cancer patients, researchers from the British Columbia Cancer Agency, also using a ddPCR approach, found that cell-free DNA levels, rather than mutation status, may be a better way of deciding whether to switch patients with evidence of resistance to EGFR TKI treatment to chemotherapy than looking for presence of the T790M resistance mutation.
Another area where groups see the possibility of real clinical utility for blood-based DNA assays is testing patients who have had tumors surgically removed, either to identify those in whom surgery may have failed to clear them of disease, or to find inklings of a later recurrence.Â
Breast cancer and colorectal cancers, for example, can be diagnosed at early stages, offering patients the chance for a putatively curative surgery. Some of these patients will successfully go decades if not the rest of their lives without their disease recurring, while others will not. And without tests or tools to be able to distinguish these two groups, clinicians are stuck with watch-and-wait strategies.
Multiple studies have now shown that circulating DNA mutations appear in the blood significantly earlier than when a recurrence shows up in imaging or starts to cause clinical signs or symptoms.
At ASCO, Massachusetts General Hospital researchers working with Guardant Health shared data from 72 colorectal cancer patients treated with surgery with or without adjuvant therapy who were tested with the "Lunar" assay Guardant has developed to profile both genomic alterations and epigenetic signatures, and were classified as either ctDNA positive or ctDNA negative.
Overall, the group calculated that the testing had a positive predictive value of 93 percent and a negative predictive value of 80 percent in anticipating future recurrences, though this varied for different subsets of the cohort, based on stage or other features.
Guardant said earlier this year that it is planning a 10,000-individual prospective clinical trial to test the ability of Lunar to detect colorectal cancer in asymptomatic individuals.
Thermo Fisher's Quagliata argued that the commercial landscape may have to shift to support liquid biopsy monitoring, if efforts succeed to prove its clinical utility. While single-timepoint tests for diagnosis and mutational profiling may make sense to send out to a central facility, labs, especially in markets outside the US, will want tools to perform repeat monitoring tests in house, he argued.
Nevertheless, central lab companies like Guardant Health have stated their intentions to begin providing clinical liquid biopsy monitoring, some even within the next year, indicating that they anticipate a market for centralized repeat tests.