Imagine a simple blood test that could flag most kinds of cancers at the earliest, most curable stage. For decades that idea—the “liquid biopsy”—has been a holy grail of oncology. Liquid biopsies could, in theory, detect a tumor well before it could be found by touch, symptoms or imaging. Blood tests could obviate the need for surgeons to cut tissue samples from suspicious lumps and lesions and make it possible to reveal cancer lurking in places needles and scalpels cannot safely reach. They could also determine what type of cancer is taking root and what treatment might work best to squash it.

The grail is not yet in hand, because it is hard to find definitive cancer signals in a tube of blood, but progress in recent years has been impressive. Last year the journal Science published the first big prospective study of a liquid biopsy for DNA and proteins from multiple types of cancers, conducted in 10,000 healthy older women. Though far from perfect, the blood test called CancerSEEK, developed at Johns Hopkins University and licensed by diagnostics company Thrive, found 26 malignancies that had not been discovered with conventional screenings such as mammography and colonoscopy. An even larger study is getting underway in London with 25,000 adults who have a history of smoking, using a blood test from a company audaciously named Grail.

No multicancer blood test is close to being approved, but the U.S. Food and Drug Administration has signaled its enthusiasm by designating CancerSEEK as a “breakthrough device,” due to its lifesaving potential. That status was also achieved this year by some more narrowly targeted blood tests, including one (from Bluestar Genomics) aimed at picking up pancreatic cancer in high-risk individuals and others that look for glimmers of recurrence in patients already treated for cancer. The breakthrough listing “accelerates the review process,” explains Nickolas Papadopoulos, one of the Johns Hopkins scientists who developed CancerSEEK.

Liquid biopsies can rely on a variety of biomarkers in addition to tumor DNA and proteins, such as free-floating cancer cells themselves. “It's believed that there's a lot of turnover within a tumor,” explains cancer biologist Ana Robles of the National Cancer Institute, “and as cells are dying, fragments are released into blood.”

What makes the search difficult, Robles explains, is that “if you have an early-stage cancer or certain types of cancer, there might not be a lot of DNA being shed,” and tests might miss it. The ideal blood test will be both very specific (uncovering a mutation or other signal that can only be cancer) and very sensitive so that even tiny tumors can be found. To tackle this challenge, CancerSEEK looks for cancer-specific mutations on 16 genes and for eight proteins that are linked to cancer and for which there are highly sensitive tests. In a study that used an updated version of the test, it detected more than 95 percent of ovary and liver tumors and about 70 percent of cancers of the stomach, pancreas and esophagus but only 33 percent of breast tumors and just 43 percent of stage 1 cancers. It will take further giant, costly studies for pan-cancer liquid biopsies to prove their efficacy for early detection.

And simple detection is not the only goal. An ideal liquid biopsy will also determine the likely location of the cancer so that it can be treated. “Mutations are often shared among different kinds of cancer, so if you find them in blood you don't know if that mutation is coming from a pancreatic cancer or lung cancer,” says Anirban Maitra, a pancreatic cancer scientist at the M.D. Anderson Cancer Center in Houston. To solve that problem, some newer liquid biopsies look for changes in gene expression—whether they are turned on or off—as opposed to changes in the genes themselves. Such changes, Maitra notes, are “more organ-specific.”

On the nearer horizon are liquid biopsies to help people already diagnosed with cancer. Last year the FDA approved the first two such tests, which scan for tumor DNA so doctors can select mutation-targeted drugs. Scientists are working on blood tests to detect the first signs of cancer recurrence in patients who have completed treatment. This work on “minimal residual disease” is moving fast, Papadopoulos says. “The question is: Does it save lives?”

That is the question companies such as Thrive and Grail must answer for their broadly ambitious screening tests. A high rate of false positives or negatives or a tendency to detect cancers that are slow-growing and trivial will not be useful. “These companies have to prove that they can detect early cancer and, more important, that the early detection can have an impact on cancer survival,” Maitra observes. “That is the holy grail of the holy grail.”