NEW YORK (GenomeWeb) – Researchers from several Boston-area institutions have developed a method to quantify circulating tumor DNA in a blood sample, allowing them to quickly identify which patients have high enough levels to allow for accurate genome-wide sequencing without the need to take a tissue biopsy.
In a study this week in Nature Communications, investigators from the Broad Institute, Dana Farber Cancer Institute, the Koch Institute for Integrative Cancer Research, and Massachusetts General Hospital described their evaluation of the tool — called ichorCNA — in 1,439 blood samples from 520 metastatic breast or prostate cancer patients.
The approach uses low-coverage whole-genome sequencing to both detect copy number alterations and to quantify tumor fraction. According to the paper, the authors have filed a patent application on their method, but their use of it currently is centered on applying it to research studies.
"Using copy number [changes] to distinguish tumor from normal is not new, but using it to quantify tumor fraction has been challenging," the study's first author Viktor Adalsteinsson said this week.
Adalsteinsson, who leads the Blood Biopsy Team at the Broad, a multi-institutional collaboration to profile cancer genomes from blood, has been in charge of a variety of efforts that use liquid biopsy to identify mechanisms of response and resistance to therapy and explore the routine genomic monitoring of patients with cancer.
"Current methods typically require 10X to 30X coverage to count allelic copy number — that is, coverage of germline heterozygous SNP sites," he added. "But our method [developed by Broad Institute Postdoc Gavin Ha] was calibrated to simultaneously detect copy number [alterations] and quantify tumor fraction from 100-fold less sequencing; about 0.1X coverage."
Interestingly, when the researchers used the method to investigate patient blood samples, they found that 43 percent of the breast cancer and 49 percent of the prostate cancer patients had enough ctDNA for whole-exome sequencing in at least one of their samples, based on the ichorCNA analysis.
And when the team performed WES on blood samples from a smaller group of subjects, they found that they could recapitulate about 90 percent of the genetic features seen in the same patients' matched tissue samples.
Much excitement in the liquid biopsy space is around pushing sequencing technology toward the detection of rare and minute fragments of tumor DNA present in people who don't yet know that they have a cancer.
But investigators also see a great benefit in being able to use noninvasive methods to study tumor genomes of patients with advanced cancers, especially in the research space , but also increasingly in clinical practice.
Commercial tests — like those from Guardant Health, Foundation Medicine, and Personal Genome Diagnostics — and others have relied on targeted approaches that interrogate preselected regions of the genome that are known to be frequently mutated in various cancers. But mutations of interest do arise outside of these areas and could potentially be of clinical importance as science advances.
Whole-exome and whole-genome sequencing of tissue samples is increasingly used in the study of the genomic drivers of cancer inception, progression, response, and resistance to treatment. According to Adalsteinsson, the ability to do the same without having to obtain tissue samples could open up new avenues for research, and could also accelerate existing efforts by broadening the number of patients who can safely and easily participate.
But not all patients have enough tumor DNA in their blood to allow for genome-wide sequencing , hence the team's development and implementation of ichorCNA.
When the team used the approach in its study, close to half of the advanced cancer patients analyzed did have enough tumor DNA for exome sequencing, although in some cases this required multiple samples. Meanwhile, the concordance between blood and tissue was high.
The researchers were able to compare blood-based and tumor whole-exome sequencing in 41 patients The data matched closely across a number of genetic features, such as clonal somatic mutations (an 88 percent match) and copy number alterations (an 80 percent match), the authors wrote.
Adalsteinsson said that this bodes well for the use of this methodology to study larger groups of advanced cancer patients, something he and his colleague are already doing.
For example, investigators are using theichorCNA and subsequent whole-exome sequencing approach in the Broad Institute's "Genomics Platform" to comprehensively map metastatic and drug-resistant tumors from blood samples, he said.
Researchers are also using the method in direct-to-patient research efforts underway at the Broad, including the Metastatic Breast Cancer Project. In these efforts, the organization involved are using patient advocacy groups, the internet, and social media to connect directly with patients, who can contribute samples remotely. In addition to the MBCP, the team plans to integrate blood biopsies in similar efforts for angiosarcoma and metastatic prostate cancer.
In the meantime, the clinical sphere is still grappling with the utility of methods like whole-exome or whole-genome sequencing in routine clinical care of cancer patients. Adalsteinsson said that as the costs of sequencing continue to go down, genome-wide analyses in blood, as well as in tissue, may become a more standard part of clinical practice, but likely as an adjunct to, not a replacement of, existing targeted technologies.