Skip to main content
Premium Trial:

Request an Annual Quote

Data Demonstrates Potential of Liquid Biopsy in Cancers of Unknown Primary


NEW YORK (GenomeWeb) – Researchers showed in a study last week that using blood-based genomic testing — specifically Guardant Health's Guardant360 test —should make it possible to uncover targetable genomic alterations in a significant number of patients with cancers of unknown primary origin.

Publishing their results in Cancer Research, the UC San Diego-led team studied 442 patients with CUP (carcinoma of unknown primary) and found mutations that could inform a specific treatment decision in more than half of them.

Though the results don't speak to how genomic testing — whether blood-based or tissue-based — affects patient outcomes and quality of life, they do provide similar evidence for efficacy of blood-based testing that has already been demonstrated for tissue sequencing.

Because liquid biopsy tests can be performed more easily and quickly than obtaining and testing tissue, they could offer an avenue for more CUP patients to be evaluated for targetable alterations, UCSD researcher Razelle Kurzrock, the study's senior author, said this week.

While the use of genomic testing, whether blood- or tissue-based, in CUP is not universal, Kurzrock said her institution has now integrated liquid biopsy in addition to tissue sequencing for these patients based on the results of their study with Guardant.

CUP, which presents as metastatic disease with no clear sign of where it originated, has long been a challenge to oncologists, predating the advent of precision or personalized drugs.

With therapies now available that can be effective against mutation-positive tumors regardless of where in the body they occur — most notably the recent US Food and Drug Administration approval of Merck's PD-1/PD-L1 inhibitor Keytruda (pembrolizumab) for patients with microstatellite instability-high (MSI-H) or mismatch repair deficient (dMMR) cancers — there may be even greater benefit to molecular testing of patients who present with CUP.

About two years ago, researchers from Foundation Medicine published a study in JAMA Oncology, in which they reported that their tissue sequencing was able to detect a potentially targetable alteration in 85 percent of a cohort of CUP patients.

Prior to that, most methods to help guide treatment for patients with CUP were focused on identifying an origin location in the body.

But Kurzrock said that in a new era of molecularly targeted drugs, knowing where a cancer originated is starting to be less important than knowing its genomic or other molecular features.

Over the last few years, blood-based cancer tests have emerged that offer the opportunity to search for these mutations without the need for a biopsy procedure.

According to Kurzrock, evidence comparing tissue-based and blood-based cancer sequencing very solidly demonstrates that the two approaches are complementary — each with limitations and advantages relative to the other.

Limitations of tissue testing include insensitivity to intratumor genomic heterogeneity, which may also be especially important in patients with CUP because they so often harbor numerous metastases, she and her coauthors wrote. There also is an inability to repeat biopsies over time in order to interrogate mutational processes that can occur alongside treatment.

For example, Kurzrock and her coauthors reported on the case of a 60-year-old woman who had five ctDNA analyses during her disease course. Her tumor initially demonstrated MYC N402N and JAK2 E621K alterations, but after treatment with cisplatin and gemcitabine, the MYC and JAK2 alterations disappeared from circulation and multiple new alterations emerged including APC, NF1, KIT, AR, and STK11 anomalies.

On the other hand, circulating tumor DNA doesn't always reflect the same mutations seen in tumors, so liquid biopsy may miss some relevant mutations, especially if patients are being treated with drugs that change or reduce the shed of DNA into the blood.

Although a number of liquid biopsy tests are now available, in the Cancer Research study, Kurzrock and her colleagues turned to Guardant Health's Guardant360 assay.

"I have been what you might call an early adopter of genomic testing, so I have a familiarity with new technologies as they come out. … We have done a lot of these, about 2,000 I think, and we had been favorably impressed with the [Guardant] test," Kurzrock said. "Competitors have now come out more recently, and we are also seeing great results with others now as well."

In the team's analysis, Guardant's testing identified some kind of circulating tumor DNA alteration in 80 percent of the 442 patients studied. Excluding variants of unknown significance (VUS), that number dropped to 66 percent, or 290 subjects.

According to the authors, TP53-associated genes were the most commonly altered, followed by genes involved in the MAPK pathway, PI3K signaling, and cell-cycle machinery.

More than 99 percent of the 290 patients with non-VUS findings had at least one targetable alteration, meaning a variant associated with either an FDA-approved drug or with an investigational clinical trial, the team reported.

Importantly, the study did not evaluate how many patients with genomic alterations actually received targeted treatment, nor did it analyze outcome measures like mortality or progression-free survival in patients with molecular findings.

But Kurzrock said that studies of this sort are ongoing. At UCSD, researchers are close to reporting on therapy matching rates and outcome measures for patients as part of the institution's I-PREDICT clinical trial.

"We are just curating the data now from that cohort, which includes some CUP cases," she said.  "We have just over 200 patients that have entered that study, and we haven't analyzed the results, but we are seeing about 45 percent of patients getting matched. That is more than other groups have reported, and also more than our group has reported previously."

The team's study also included a single case report of a CUP patient tested by Guardant who showed KRAS G12D and MLH1 R389W mutations in his ctDNA.

Physicians began treatment with a combination of the MEK inhibitor trametinib and the immunotherapy drug nivolumab. The patient achieved a partial response at 8 weeks and is still being treated, the authors reported.

"This patient has had a really beautiful response," Kurzrock said. "This doesn’t mean that everyone responds like that, but we have had a preliminary look at the data and it's very encouraging."