October 2, 2018 by Dale Yuzuki
The first cell-free Pillar Biosciences publication answers several questions about reported discordance between genotyping matched tumor tissue and circulating tumor DNA collected and processed under different conditions.
Last year two studies were published comparing available commercial tissue genotyping and circulating tumor DNA tests, including Foundation Medicine’s FoundationOne CDx™, Guardant Health’s Guardant360®, and Personal Genome Diagnostics’ PlasmaSELECT™ tests. What they found was disturbing.
Background on circulating tumor DNA tests
With Foundation Medicine being acquired by Roche for $2.4B in June of 2018 and Guardant Health going public in early October in an IPO to raise $200M to say there is a lot of activity in the liquid biopsy test market is an understatement. Other companies such as Personal Genome Diagnostics received an FDA Breakthrough Device Designation in July 2018, and Inivata recently received draft Medicare Local Coverage Determination (LCD) from Palmetto GBA, a Medicare administrative contractor.
Should you dig into the scientific literature, the validation studies of each company’s standalone test is impressive. For example, Guardant Health’s 2015 PLOS One study states ‘Analytic sensitivity down to 0.1% mutant allele fraction is demonstrated’ and ‘Near-perfect analytic specificity (>99.9999%) enables complete coverage of many genes’ in its abstract should leave an impression. The cell-free assay validation for Foundation Medicine’s test FoundationOne Liquid was recently published, in addition to Inivata’s earlier this year. Each of these studies show excellent accuracy, sensitivity and specificity when they stand alone.
Concordance studies between technologies are troubling
However, one study titled “Comparison of 2 Commercially Available Next-Generation Sequencing Platforms in Oncology” compared the FoundationOne CDx tumor tissue test against the Guardant360 cell-free circulating DNA blood test from 9 individuals. Although the sample size was small, only 22% concordance was observed, and only 25% of the drug therapies recommended were identical.
A second study was similar, comparing tissue mutations with the FoundationOne CDx test to cell-free DNA mutations from the Guardant360 test in a study in 2016 called “Concordance between genomic alterations assessed by next-generation sequencing in tumor tissue or circulating cell-free DNA”. The researchers found that with patient samples with both tissue and plasma “over 50% of mutations detected in either technique were not detected using the other biopsy technique, indicating a potential complementary role of each assay.” At the gene-level where any mutation in the same specific gene would count as concordant between a tissue mutation and one from circulating tumor DNA, this ‘overall concordance’ was measured at 91.9% – 93.9%, but drops to 11.8% – 17.1% considering specific alterations.
Another study, “Patient-Paired Sample Congruence Between 2 Commercial Liquid Biopsy Tests” compared the Guardant360 test to the PlasmaSELECT test from Personal Genome Diagnostics, with 40 patient samples. Only 7.5% of the samples had complete congruence, 15% had partial congruence, and 40% had no congruence. The authors conclude “These data cannot determine which test is more accurate but suggest that reported gene alterations will not be the same across different platforms, raising the specter that patients could potentially receive different treatments depending on the cfDNA platform. Insufficient genetic profiling congruence could jeopardize the clinical benefit of personalized medicine.”
A new publication shows Pillar Biosciences solid tumor and cell-free DNA assay sensitivity
A recent publication “Heterogeneous mutation pattern in tumor tissue and circulating tumor DNA warrants parallel NGS panel testing” (Mol. Cancer 17:131, http://doi.org/10.1186/s12943-018-0875-0) examines 56 non-small cell lung carcinoma (NSCLC) patient samples, looked at tissue handled in several different ways (fresh-frozen and formalin-fixed paraffin embedded), and plasma handed with different processing times post-collection.
Using the Pillar Biosciences’ SLIMamp-based Lung and Colon Cancer Panel for the tissue samples, and an early-access version of Pillar Biosciences’ Lung cfDNA Panel, error rates for the tissue panel was limited to 0.5% minor allele frequency (see Figure 1 of the reference) and the error rates for the cell-free panel was less than 0.2% minor allele frequency.
Room for improved concordance
One major area for improvement of the concordance rate has to do with sample handling from point of collection. There are several publications that make this important point.
For example, here is publication comparing K2EDTA to proprietary Streck Cell Free DNA Blood Collection Tubes, and here another publication showing the importance of spinning the plasma twice for optimum yield of cell-free DNA. A collection tube called CellSave (co-developed with the CellSearch CTC technology now owned by Merini Biosystems) was compared to K2EDTA in this publication.
In this Pillar Biosciences study there were institutional restrictions to alternative blood collection tubes, as well as how quickly the samples (after collection in K2EDTA tubes) could be processed. It is expected that the concordance rate would improve with alternative collection approaches and/or faster processing.
Suffice it to say here that the discordance rate was consistent with the existing literature referenced above. What was the effect of longer times from collection of blood to purification of cell-free DNA? What was the difference in mutation detection rate comparing fresh-frozen samples to FFPE-processed ones? What did orthogonal testing of high-frequency discordance by digital PCR reveal?
And If you are looking for a sensitive, accurate and streamlined enrichment for your clinical laboratory, contact us for a trial kit to evaluate.