NEW YORK (GenomeWeb) – The Illumina-sponsored Actionable Genome Consortium reported this week that cell-free DNA-based tumor genotyping for non-small cell lung cancer patients is concordant with tissue tumor-based genotyping most of the time and could be particularly useful in cases where acquiring a tissue biopsy is not feasible.
The group, which includes researchers from Memorial Sloan Kettering Cancer Center, Dan-Farber Cancer Center, MD Anderson Cancer Center, Grail, and Illumina, published its study this week in Annals of Oncology.
The researchers analyzed cell-free samples from 127 NSCLC patients, focusing on 37 lung-cancer-related genes, which they sequenced at very high depth — around 50,000-fold coverage. In addition, the patients all had tumor genotyping performed by their respective academic medical centers, according to standard of care — 64 of whom were tested by an NGS panel.
Tissue tumor sequencing identified driver mutations in 91 patients, while cfDNA testing was concordant in 68, for a sensitivity of 75 percent. For the 19 samples in which tumor sequencing did not find a driver mutation, cfDNA sequencing was 100 percent concordant. And, in 17 samples where tissue sequencing could not be performed due to insufficient amount of tissue sample, cfDNA testing identified driver mutations in four.
When the researchers evaluated the discordant samples — where tumor testing identified a driver mutation, but cfDNA sequencing did not — they noted a number of factors that may have played a role.
One key factor was the amount of cfDNA in the patient's blood. The lower the concentration, the less likely the cfDNA test would detect an alteration, despite the assay's ability to detect variants down to an allele frequency of 0.14 percent.
Another potential factor was the time interval between the tissue biopsy and blood biopsy. The researchers noted that in the 23 cases where the tissue biopsy-based assay detected driver mutations, but the cfDNA assay did not, plasma was collected for testing more than 200 days after a tissue biopsy was done, so the tumor may have evolved or treatment may have reduced the tumor burden. However, because of the small sample size, the researchers were not able to draw any definitive conclusions. The authors noted, however, that a "trend toward decreasing plasma driver detection with increasing time interval and intervening treatment between tissue and plasma collection was also observed in a previous study suggesting that tumor shedding of cfDNA may vary over time."
The study adds to a growing body of evidence suggesting that cfDNA sequencing can be used as a noninvasive way to genotype a patient's tumor, particularly in lung cancer patients, when obtaining a tissue biopsy can be risky.
In particular, Guardant Health has been working to make the case that blood-based cancer testing could be performed as a first-line genotyping test. Researcher plan to report results from the NILE (Noninvasive vs. Invasive Lung Evaluation) study in an abstract at the upcoming American Association for Cancer Research meeting.
According to the NILE study, Guardant's cfDNA test enabled researchers to identify more actionable mutations faster than tissue-based testing for the same NSCLC patients.
Guardant has said that establishing a "blood-first" paradigm will be key to securing US Food and Drug Administration approval and reimbursement for its test.
In Illumina's Actionable Genome Consortium study, the researchers concluded that cfDNA analysis is "not a replacement for histologic confirmation and immunohistochemistry from a diagnostic biopsy" and that the "data thus support a complementary role for plasma cfDNA NGS." However, they said that cfDNA could be used before tissue-based testing, given its "high specificity, noninvasive nature, and fast turnaround time." But, due to its "modest sensitivity" a negative cfDNA test should spur a tissue-based molecular test.