Neal Lindeman (left) and Lynette Sholl (right) are leading the effort to bring liquid biopsies out of research labs and into clinical settings.

Surgical biopsies have been the gold standard for cancer diagnosis and for good reason—getting direct access to tumor tissue leaves little doubt that the procedure will yield the information pathologists and oncologists need in order to determine an appropriate treatment plan.

But surgery is still surgery, carrying its usual risks, and some patients with advanced or relapsed cancer may not be well enough to undergo an invasive procedure. An open lung biopsy, which involves opening the chest while the patient is under general anesthesia, can result in an infection or air leak in the chest. One potential risk of a needle biopsy—a less-invasive method, in which a needle is inserted into the chest under light sedation or local anesthesia—includes a partial lung collapse.

That’s why a new kind of diagnostic tool that uses genomic information derived from a simple blood test—a process known as a liquid biopsy—is generating excitement among BWH pathologists who are piloting the method in clinical labs as part of a joint effort with Dana-Farber Cancer Institute.

The pilot is limited to a narrow pool of lung cancer patients—specifically, those with EGFR-positive adenocarcinoma of the lung who have relapsed. Patients with EGFR mutations account for 15 to 20 percent of lung cancer cases, according to Neal Lindeman, MD, director of Molecular Diagnostics at BWH. Patients who test positive for certain mutations in the EGFR gene are prescribed and often respond well to targeted therapies such as erlotinib, but may relapse if their tumor acquires a second mutation, known as T790M, which confers resistance to these therapies. The liquid biopsy test being conducted by Lindeman and colleagues can detect the presence of this second mutation.

Liquid biopsies have moved out of research labs and directly into the clinical labs at BWH starting in April. In other words, any patient who meets the diagnostic criteria can have a liquid biopsy done—no enrollment in a clinical trial is required. BWH researchers say that, as far as they know, this is the only such clinical diagnostic program happening in a nonprofit health care setting.

The Food and Drug Administration approved the first T790M inhibitor, osimertinib, this past November, which makes the ability to accurately and rapidly detect T790M mutations vital for getting the right treatment to the right set of patients at the right time.

The test is extremely accurate at making positive identifications, according to Lindeman, who directs the lab. In a joint study from BWH and Dana-Farber published earlier this year in JAMA Oncology, liquid biopsies were able to identify every patient who had the primary EGFR mutation. While the test does have a somewhat lower sensitivity, returning a false negative 30 percent of the time—which would likely trigger the need for a traditional biopsy—experts say the results stands out for the apparent absence of false positives. What’s more, a positive allows the oncologist to immediately prescribe a targeted therapy.

Being able to identify genetic material from cancer DNA circulating anywhere in the blood—not just in the tumor—with such precision is a significant advancement, albeit a somewhat unexpected one, Lindeman said.

“The thing that really surprised me is that it works so well. There are an awful lot of cells in the body, and the number of cancer cells is relatively low,” he said. “The fact that you can find this needle in a haystack is very powerful.”

It’s given Geoffrey Oxnard, MD, a thoracic oncologist at Dana-Farber working with BWH Pathology on the pilot, enough confidence to forego surgical biopsies for qualified patients.

“We trust the liquid biopsy so much that we can cancel the tumor testing for these patients,” Oxnard said.

Pathologists and oncologists are leading the liquid biopsy pilot with this particular type of lung cancer because it requires testing for only a small number of known genetic anomalies, Lindeman explained.

“There is a series of very well-known, single and specific mutations that are amenable to this kind of testing,” said Lindeman, who was part of the group that discovered the EGFR mutation and pioneered testing for it in 2004. “Right now, liquid biopsies work best if you know exactly what you’re looking for.”

Biopsy Results in Days, Not Weeks

Patients who test positive for the T790M mutation could forego a surgical biopsy; the information also enables an oncologist to quickly identify the best treatment, a T790M inhibitor, based on a positive result.

Quickly is the keyword here, explained Lynette Sholl, a molecular and surgical pathologist at BWH who specializes in lung cancer and tests the liquid biopsy specimens. Traditional surgical biopsies often take weeks to process. Liquid biopsies, on the other hand, can be processed in a matter of days.

“We’ve learned that the turnaround time for the blood plasma testing has the potential to be very short,” Sholl said. “We’ve reported our results in as little as a day, whereas usually your typical biopsy could take up to two to three weeks—from the time of the sample acquisition to the reporting of molecular testing of the tumor biopsy. Our goal is to keep the turnaround time for liquid biopsies under four days.”

Dana-Farber’s Oxnard, who identifies qualified cancer patients for the liquid biopsies that BWH’s pathologists test, says the time it takes to process biopsy results can be a matter of life and death for patients with this type of cancer.

“I’ve had patients die waiting for resistance biopsies,” he said. “We want to lower the bar for getting patients onto these smart, targeted therapies.”

Since launching the liquid biopsy pilot in April, BWH pathologists have processed 11 samples from clinical labs. However, the test’s origins are in the Translational Research Laboratory at the Belfer Center for Applied Cancer Science at Dana-Farber, which initiated the research and development for plasma-based lung cancer genotyping. The clinical trial that followed involved 180 patients with pulmonary adenocarcinoma, 120 of whom were newly diagnosed and 60 of whom had developed a resistance to EGFR inhibitors. The trial was validated at Dana-Farber’s Lowe Center for Thoracic Oncology.

At that point, Dana-Farber’s team got in touch with BWH Pathology, Sholl said. Their research demonstrated the feasibility and proper ways to handle liquid biopsies in a clinical setting, which spurred Sholl to lead the validation effort for bringing liquid biopsies into BWH’s clinical labs.

“In order to show that plasma testing for lung cancer mutations is equivalent, you need to have specimens of both types of biopsies, surgical and liquid,” Sholl said. After confirming that the science was sound, Sholl made a few workflow changes in the clinical lab to accommodate the assay, such as establishing a process for getting specimens to the lab quickly, before the DNA began degrading.

What’s Next for Liquid Biopsies?

Some unanswered questions still remain. Pathologists have been unable to predict how much free-circulating tumor DNA a patient has in his or her blood, Sholl said. Some have more, others have less. There appears to be a correlation with how advanced the tumor is—including whether it’s metastasized—but it remains an area of active investigation.

As a result, liquid biopsies are a qualitative test for the time being, providing a yes or no answer to whether a T790M mutation is present. Lindeman and Oxnard say they hope to see liquid biopsies become a quantitative test that can indicate the levels of cancer DNA in the blood, allowing them to be used for monitoring the progress of cancer in a patient or the likelihood of a relapse.

Another area of active investigation is broadening the panel of mutations the test can profile, which would open up liquid biopsies to a larger number of lung cancer patients, Sholl said. The current pilot for EGFR tests a handful of genetic mutations, which has made it easier to validate and roll out. Other genes that the research team hopes to profile—such as KRAS, BRAF and ERBB2—present a greater challenge, as they can be affected by dozens of different cancer-related mutations and are not yet associated with approved, targeted therapies in lung cancer.

Oxnard envisions liquid biopsies as a compelling complement to imaging technologies, such as CT scans, which he says don’t always give a clear picture of what’s going on with a tumor.

“There’s a remarkable amount of uncertainty in oncology, and this is a tool that may be able to provide greater insight, specifically because it’s measuring the key mutations in this cancer,” he said.