Detecting Lung Cancer From Simple Blood Draw

CHICAGO — Diagnosing lung cancer from a simple blood test?

Two years ago, this was a "pipe dream" but "today, we actually have data to show this is really feasible to find early-stage cancer in the blood," said lead investigator Geoffrey R. Oxnard, MD, associate professor of medicine at Dana-Farber Cancer Institute and Harvard Medical School in Boston, Massachusetts.

"It will take some time, but it's well on its way," he told a press briefing here at the American Society of Clinical Oncology (ASCO) 2018.

He presented an interim analysis of a study that compared the performance of three assays to measure cell-free DNA in over 1000 individuals, including over 150 patients with lung cancer.

Controlling for potentially confounding DNA shed by white blood cells, the study found that, regardless of the assay used, approximately 50% of early lung cancer cases could be detected, rising to 90% of advanced cases.

Commenting on the new results, ASCO expert David Graham, MD, from the Carolinas Medical Center in Charlotte, North Carolina, said that this is "a very exciting new methodology" and is "an important first step towards an easier way to detect lung cancer at earlier and hopefully more curable stages."

"If the promise of this report holds, we could easily see a day when a person could be screened for lung cancer possibly by simply going into their regular doctor's office for a blood draw," he added.

Such a test could up the game in lung cancer screening, which — despite the availability of low-dose CT (LDCT) for a few years now — remains very low compared with screening rates for other cancers. As reported by Medscape Medical News, data presented at this year's ASCO annual meeting showed that in 2016, fewer than 2% of the 7 million individuals eligible for lung cancer screening across the United States underwent LDCT screening.

One of Most Important Presentations at the Meeting

At the press briefing, Oxnard said that this first interim analysis of the study "demonstrates that comprehensive sequencing of plasma cell-free DNA can generate high-quality data across the entire genome and that it permits noninvasive cancer detection."

Noting that the assays "can detect lung cancer across stages, across histologies, across populations", he added: "These results support the promise of using cell-free DNA-based assays to develop an early cancer detection test with high specificity."

Oxnard said that ongoing studies, such as STRIVE and the current study, will allow for the "further optimization and focusing of this assay towards an eventual cancer diagnostic."

Hossein Borghaei, DO, chief of the Division of Thoracic Medical Oncology at Fox Chase Cancer Center, Philadelphia, Pennsylvania, told Medscape Medical News that a blood test for early lung cancer would be "welcome" from a clinical point of view.

While he felt that the study was "well conducted," he pointed out that the number of patients with lung cancer included was "a little bit low," thus underlining the need for confirmatory studies.

Moreover, Borghaei said that, when thinking of applying a cell-free DNA assay–based test to the wider population, "accessibility to the test and the cost of the test should be considered."

"If it's too expensive, that puts it out of reach for many patients. However, if it is truly proven to be effective in larger studies, it can move forward."

Other potential issues highlighted by Borghaei included the lack of sensitivity of the test for early-stage lung cancer, meaning that there is "still some work to do" to improve detection rates, and that the algorithms underpinning the calculations are "not trivial" and can be performed only in highly specialized centers.

One of the most important things to come out of the ASCO meeting in terms of screening and early detection. Dr Hossein Borghaei

Nevertheless, he believes that it is a "really positive, important step [and] is one of the most important things to come out of the ASCO meeting in terms of screening and early detection."

Study Details

In their study, Oxnard and colleagues investigated whether cell-free DNA–based tests, which have already been used to identify specific mutations for cancer genotyping, could be used more broadly for cancer detection by identifying a signature of mutations.

They established the Circulating Cell-free Genome Atlas Study (CCGA), which has so far enrolled 12,292 of a planned 15,000 individuals from 142 sites in the United States and Canada, of whom 70% have untreated cancer and 30% are noncancer controls.

From these, the team selected 2800 individuals, divided into a training set of 1785 individuals (127 with lung cancer) and a test set of 1015 participants (47 with lung cancer), leaving 12,200 for future validation studies.

They then selected 118 evaluable and analyzable patients with lung cancer and 561 noncancer controls from the training set and 46 patients with lung cancer and 362 noncancer controls from the test set.

The two groups were similar in terms of their baseline demographic characteristics, although, as expected, patients with lung cancer were more likely than noncancer controls to be male and to smoke.

The researchers then used three different types of cell-free DNA sequencing assays:

A targeted assay to detect noninherited somatic mutations, including single nucleotide variants and small insertions/deletions;
Whole-genome sequencing to detect somatic gene copy number changes; and
Whole-genome bisulfite sequencing of cell-free DNA to detect abnormal methylation patterns indicating epigenetic changes.

Crucially, the assays used white blood cell DNA filtering to reduce the false-positive rate down to just 2%, giving a specificity of 98%.

Oxnard explained, "The white blood cells are rich with mutations that could pollute the DNA and make you think that there's cancer present in the cell-free DNA [but] is in fact shed from the white blood cells."

"You screen out this interference from white blood cells and other biologic noise, and you're left with final features — mutations, copy number variations, and methylation signatures — which then go into the final assays being studied."

The results showed that regardless of the assay used, an average of 41% (29% to 54%) of early-stage lung cancers, defined as stage I to IIIA, were detected in the training set.

In the test set, 50% (29% to 71%) of early-stage lung cancers were detected.

For advanced lung cancers, defined as stage IIIB to IV, the performance of the assays was better, with an average of 89% (77% to 96%) of cases detected in the training set and 91% (71% to 99%) of cases detected in the test set.

Reality Check

Despite the enthusiasm for the idea of a liquid biopsy for early cancer detection, it may be some years yet before any such tests are available for use in clinical practice. A recent ASCO Special Article that reviewed the field concluded that, currently, "there is no evidence of clinical utility and little evidence of clinical validation of circulating tumor [ct] DNA assays in early stage cancer, treatment monitoring or residual disease detection."

While acknowledging that the "rapid pace of research" will necessitate a re-evaluation of this stance, it adds: "There is no evidence of clinical validity and clinical utility to suggest that ctDNA assays are useful for cancer screening, outside of a clinical trial."

This study was funded by GRAI, Inc. Geoffrey R. Oxnard disclosed consulting or advisory role with AstraZeneca, Inivata, Boehringer Ingelheim, Takeda, Genentech/Roche, Novartis, LOXO, Ignyta, DropWorks, and GRAIL; patents, royalties, other intellectual property with Chugai Pharma, Bio-Rad, Sysmex, and Guardant Health. Tara Maddala disclosed employment with GRAIL and stock and other ownership interests with GRAIL. Other authors list numerous potential conflicts of interest. Graham reports serving on a speakers' bureau with Biopep Solution.

American Society of Clinical Oncology (ASCO) 2018. Presented June 4, 2018. Abstract LBA8501

For more from Medscape Oncology, follow us on Twitter: @MedscapeOnc

Comments

Commenting is limited to medical professionals. To comment please Log-in.

Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.