A new liquid biopsy technique, CancerSEEK, that evaluates tumor mutations and cancer-linked proteins can detect 70% of cancers in patients who have one of eight tumor types. The technique can pinpoint the location of a tumor in 68% of cases. However, its sensitivity drops off at earlier disease stages.

A new blood test for tumor-specific mutations and proteins may bring cancer screening with liquid biopsies closer to reality (Science 2018 Jan 18 [Epub ahead of print]). The procedure can identify 70% of patients who have any of eight common tumors, including five tumors for which no screening test is currently available.

Researchers have been trying to develop liquid biopsies to spot cancers early, but to date the techniques have had several shortcomings. For early-stage cancers, when detection is more likely to benefit patients, the blood levels of tumor DNA may be insufficient for tumor detection. Furthermore, the tests' specificity has remained unclear because researchers haven't tested them in large populations of healthy volunteers.

To overcome these limitations, Nickolas Papadopoulos, PhD, of Johns Hopkins School of Medicine in Baltimore, MD, and colleagues developed CancerSEEK, a blood test that combines sequencing with analysis of protein biomarkers. The test searches for tumor-specific mutations in 16 genes and, to improve detection, uses approaches such as barcoding the DNA fragments isolated from the blood.

Because many early-stage tumors release minuscule amounts of DNA, CancerSEEK then evaluates the levels of eight proteins produced in large quantities by cancer cells, including CA-125 and carcinoembryonic antigen. Papadopoulos and colleagues designed an algorithm that weighs the protein and DNA data to determine whether a patient is likely to have a tumor.

To evaluate their approach, the researchers applied CancerSEEK to blood samples from 1,005 patients with one of eight cancers—breast, colorectal, lung, esophagus, pancreas, ovarian, stomach, or liver—the last five of which don't have screening tests. The scientists also analyzed samples from 812 healthy subjects.

CancerSEEK identified 70% of the patients with cancer. It worked best for ovarian cancer, detecting tumors in 98% of the patients with the disease, and worst for breast cancer, where the detection rate was 33%. The false-positive rate for all cancers was less than 1%.

Because liquid biopsy techniques that rely only on genomic information usually cannot determine a tumor's location, the researchers asked whether CancerSEEK could do that. In the 626 patients who tested positive, CancerSEEK narrowed the tumor's location to either of two organs in 83% of patients and pinpointed the site in 68%.

Papadopoulos and his colleagues estimate that the test would cost about $500.

“I would say it's a step on the way” to a feasible liquid biopsy screening test, says Ian Cree, MBChB, PhD, of the World Health Organization's International Agency for Research on Cancer in Lyon, France. Cloud Paweletz, PhD, of Dana-Farber Cancer Institute in Boston, MA, agrees, saying the researchers “have done a really good job to maximize the information that they get at a reasonable price.”

Both scientists also agree that the approach requires refinement before it could be clinically useful. Cree notes that the sensitivity declined from 73% for stage II cancers to 43% for stage I. “We do need something that will do even better in low-stage disease,” he says.

The authors should strive for even higher specificity, says Paweletz, because the control group isn't likely to be representative of the patients who would undergo cancer screening. “We need to get the specificity higher to use it in a broader population.” –Mitch Leslie

For more news on cancer research, visit Cancer Discovery online at http://cancerdiscovery.aacrjournals.org/content/early/by/section.