The Quest for a Liquid Assay for Cancer Detection—The Circulating Cell-free Genome Atlas (CCGA) Study

While targeted therapies and immune checkpoint inhibitors have grabbed the headlines at recent oncology conferences, it is important to remember that early diagnosis is key to successful cancer treatment. The liquid biopsy—a blood test that detects early stage cancer—has long remained the Holy Grail of cancer screening and diagnosis. Genetic testing for specific mutations in biopsy and cytology specimens is now an established part of the diagnostic workup for lung cancer.¹ It is not surprising, therefore, that researchers have been investigating the use of this technology using cell-free DNA, not for cancer genotyping, but for cancer detection. This requires a different approach using broad sequencing and looking for any signal of cancer.² The Circulating Cell-Free Genome Atlas (CCGA) study (NCT02889978) is the largest study ever initiated to develop a liquid biopsy for early cancer detection based on cell-free DNA. This prospective, observational, longitudinal study is taking place at over 140 different sites around the United States, and has currently enrolled 10,012 of a planned 15,000 participants, including people with a recent cancer diagnosis and also a control group of individuals with no known cancer.

Initial findings from the CCGA study were presented at the American Association for Cancer Research (AACR) Annual Meeting, which was held on April 14–18, 2018, in Chicago.³ The investigations involved blood samples from 1,627 participants (878 cases, 580 controls, and 169 technical assay controls) across 20 tumor types and all clinical stages. The cell-free DNA was isolated, then analyzed using three sequencing methods: a targeted assay to detect non-inherited (somatic) mutations such as nonsynonymous single nucleotide variants (SNVs) and indels (small insertions and/or deletions) in specific sections of the genome; whole-genome sequencing to assess differences in somatic gene copy number across the genome; and whole-genome bisulfite sequencing, to determine abnormal patterns of cell-free DNA methylation.

It is important to remember that DNA mutations can also be generated by processes other than cancer, and these could lead to a false-positive signal. One source of these mutations is clonal hematopoiesis, a common process during human aging that results in variations in the DNA of the white blood cell (WBC) population.⁴ Because a large proportion of cell-free DNA is derived from WBCs; investigators also sequenced paired WBCs during targeted and whole-genome sequencing, allowing them to subtract the signal generated from the WBCs, leaving only tumor-related DNA variants, which resulted in a cleaner signal and increased the specificity of the assay. The SNV/indel mutations in 66% and 78% of the participants with and without cancer, respectively, were found to be derived from WBCs and not from tumors.⁵

A consistent signal suggestive of cancer was observed across all assays in less than 1% of the participants who entered the study with no known cancer, representing potential undiagnosed cancers but also demonstrating high specificity. The average nonsynonymous SNVs/indels per megabase were 1, 2.4, and 6.5 in participants without cancer at study entry, with stage I–III cancer, and with stage IV cancer, respectively. In terms of whole-genome sequencing analysis, scores were 0, 0.3, and 2 in participants without cancer, with stage I–III cancer, and with stage IV cancer, respectively. Finally, methylation scores were 0, 1, and 3.9 in participants without cancer, with stage I–III cancer, and with stage IV cancer, respectively.³

“The early results from our prototype assays suggest that it is feasible to design a comprehensive blood-based test that allows us to detect cancer in asymptomatic individuals,” said lead investigator Alexander Aravanis, MD, PhD, vice president of research and development at GRAIL Inc. “Our initial data suggest that it will be possible to develop a screening test for multiple cancers with very high specificity, potentially greater than 99 percent.”⁵

These promising findings were quickly followed by another presentation at the 2018 American Society of Clinical Oncology (ASCO) Annual Meeting, which was held on June 1–5 in Chicago. This sub-study investigated the ability of the three assays to detect cancer in 127 people with stage I–IV lung cancer. Signals suggesting lung cancer were detected across all assays, and increased according to cancer stage. The targeted assay detected 51% of early-stage (stage I–IIIA) lung cancers and 89% of late-stage (stage IIIB–IV) lung cancers. Whole-genome sequencing detected 38% of early-stage cancers and 87% of late-stage cancers. Whole-genome bisulfite sequencing had similar efficacy, detecting 41% of early stage lung cancers and 89% of late-stage cancers. Stratifying participants by histological subtypes found similar sensitivities across all assays for adenocarcinoma, squamous cell and small cell lung cancer. The rate of false positives was low, with five (<1%) of those who had no cancer at study entry showing positive signals across all three assays. Of these, one was subsequently diagnosed with stage III ovarian cancer, and one with stage II endometrial cancer.⁶

Lead investigator Geoffrey Oxnard, MD, Associate Professor of Medicine at Dana-Farber Cancer Institute and Harvard Medical School, said: “Lung cancer is the single greatest cause of cancer death globally and is typically diagnosed at late stages when, despite recent treatment breakthroughs, survival is often measured in months. These initial data from CCGA are exciting because they suggest it is possible to detect lung cancer through DNA signals in the blood at earlier stages when survival rates may be higher.”⁷

While these high detection levels were seen in late-stage cancer, the levels of early stage cancer detection were less impressive and are not likely to become practice-changing at this stage. The next step of the CFCA study will be verification of these results in an independent group of approximately 1,000 participants as part of the same sub-study. Improved detection rates from larger populations will be necessary for this test to be approved. Other systems, such as Cancer SEEK, which is being developed at John Hopkins University, have also shown early promise.⁸ It is also important to remember that finding a tumor at an early stage may not be enough; not every cancer can be treated and not all need treating—many prostate cancers are inconsequential and only found during autopsies. We don’t know whether a liquid biopsy can give other useful information, for example how aggressive a cancer is. Nevertheless, the potential for a liquid biopsy in early-stage cancer detection is enormous, and these studies represent an important first step.

References

1. Sholl LM, Aisner DL, Varella-Garcia M, et al., Multi-institutional oncogenic driver mutation analysis in lung adenocarcinoma: The Lung Cancer Mutation Consortium experience. J Thorac Oncol. 2015;10:768–77.
2. OncLive. Dr. Oxnard Discusses the Circulating Cell-Free Genome Atlas Study. 2018. Available at: www.onclive.com/conference-coverage/asco-2018/dr-oxnard-discusses-the-circulating-cellfree-genome-atlas-study (accessed June 21, 2018).
3. Aravanis AA, Oxnard GR, Maddala T, et al. LB-343 – Development of plasma cell-free DNA (cfDNA) assays for early cancer detection: first insights from the Circulating Cell-Free Genome Atlas Study (CCGA). Presented at the American Association for Cancer Research (AACR) Annual Meeting, April 14–18, 2018, Chicago, IL, US.
4. Steensma DP, Bejar R, Jaiswal S, et al. Clonal hematopoiesis of indeterminate potential and its distinction from myelodysplastic syndromes. Blood. 2015;126:9–16.
5. AACR. Press Release: Prototype Assays Suggest Highly Specific Blood Test to Screen for Cancer is Feasible. Available at: www.aacr.org/Newsroom/Pages/News-Release-Detail.aspx?ItemID=1187#.Wy4MSqdKiUk (accessed June 21, 2018).
6. Oxnard GR, Maddake T, Hubbell E, et al. Genome-wide sequencing for early stage lung cancer detection from plasma cell-free DNA (cfDNA): The Circulating Cancer Genome Atlas (CCGA) study. J Clin Oncol. 36:18(suppl):LBA8501.
7. BusinessWire. Grail, Press release: GRAIL Announces Data on Detection of Early-Stage Lung Cancers. www.businesswire.com/news/home/20180602005048/en/GRAIL-Announces-Data-Detection-Early-Stage-Lung-Cancers (accessed June 21, 2018).
8. Cohen JD, Li L, Wang Y, et al. Detection and localization of surgically resectable cancers with a multi-analyte blood test. Science. 2018;359:926–30.