What Does Guardant360 Mean?

Guardant360 CDx is a blood test that reads tumor DNA. FDA describes it as a qualitative next generation sequencing in vitro diagnostic device. In plain terms: a machine reads the DNA letters in your blood plasma and reports which cancer-related changes it finds.

The "CDx" matters. It marks the version FDA reviewed and approved as a companion diagnostic, meaning a test cleared to decide who gets a specific drug. NCI notes the approval date as August 7, 2020, for solid tumors, not blood cancers.

Tumors shed DNA into the bloodstream. Those fragments are called circulating tumor DNA, or ctDNA. They float among a much larger pool of ordinary cell-free DNA shed by healthy cells.

The FDA summary lists the collection details. Blood goes into Streck Cell-Free DNA Blood Collection Tubes. The minimum is 5 mL of whole blood. From that, 5 to 30 nanograms of cell-free DNA are used for the analysis.

No needle goes near the tumor. That is the appeal, and also the weakness.

The FDA summary describes the reviewed content precisely:

• 55 genes examined for single nucleotide variants, meaning single-letter changes, and for insertions and deletions
• copy number amplifications, meaning extra gene copies, in 2 genes: ERBB2 and MET
• fusions, meaning two genes joined abnormally, in 4 genes: ALK, NTRK1, RET, and ROS1

NCI's summary describes the test as covering more than 60 genes. Descriptions differ because laboratories offer versions beyond the FDA-approved one. Read the gene list printed on your own report rather than any general description.

This is the most misread part of a Guardant360 report. A companion diagnostic claim is drug-specific and alteration-specific. FDA approved a short list, one at a time.

The original claim covered non-small cell lung cancer and osimertinib, sold as Tagrisso. The qualifying alterations were EGFR exon 19 deletions, L858R, and T790M.

A 2023 supplement added breast cancer. It approved detection of ESR1 missense mutations between codons 310 and 547 to identify patients for elacestrant, sold as Orserdu.

A later supplement added a second breast cancer claim, for imlunestrant, sold as Inluriyo, in ER-positive, HER2-negative advanced disease. The qualifying ESR1 changes named are E380, a V422 deletion, S463, L469, L536, Y537, and D538. That claim rests on the EMBER-3 trial. Among patients with an ESR1 mutation, median progression-free survival was 5.49 months with imlunestrant against 3.84 months with the comparison endocrine therapy, a hazard ratio of 0.635.

Everything else the report lists is information. It is not an FDA-backed instruction to use a particular drug.

Sensitivity here is not a slogan. It is a measured fraction.

FDA reports the limit of detection by how much DNA went in. With 5 nanograms of input, the limits for single-letter changes and small insertions or deletions ran from about 0.8% to 3.0% variant allele fraction. Variant allele fraction is the share of DNA fragments carrying the change. With 30 nanograms of input, those limits improved to about 0.2% to 0.8%.

Fusions follow the same pattern: about 0.9% to 1.9% at 5 nanograms of input, and 0.1% to 0.2% at 30 nanograms.

The practical meaning: a tumor that sheds little DNA can be genuinely present and still invisible in the tube.

FDA states it plainly. A negative plasma result does not assure that the tumor is negative for genomic findings. FDA's guidance is that patients with a negative plasma result should have tissue testing to confirm.

An oncologist interviewed for NCI's Cancer Currents blog put the same point in blunter language. Whether a blood draw catches enough tumor DNA is partly a matter of luck. When a blood-based profiling test finds nothing, he said he would treat the result as inconclusive rather than negative.

If your report shows no actionable alterations, the next question is not "what drug now." It is: can we get tissue, and should we?

FDA's supplement summary notes that findings in cell-free DNA can come from three sources. One is ctDNA from the tumor. One is germline alterations, meaning changes you were born with. One is non-tumor somatic changes, especially clonal hematopoiesis of indeterminate potential, or CHIP.

CHIP is an age-related expansion of blood cells carrying a mutation. It is common, and it is not cancer. Certain genes, particularly ones involved in blood cell regulation, throw CHIP signals often.

So a mutation on the report can be real, reproducible, and still irrelevant to your tumor. If a finding is surprising for your cancer type, ask whether CHIP was considered.

The oncologists interviewed for that same NCI blog post framed it this way. A liquid biopsy helps when a tumor is hard or unsafe to reach, or when a person cannot have a core biopsy. If the tumor can be reached easily and safely, a core biopsy is still preferred.

Two practical points follow. First, blood tests shine when a biopsy is risky, when the tumor is hard to reach, or when the previous biopsy is exhausted. Second, blood tests can be repeated, which makes them useful for detecting resistance changes that appear during treatment.

Tests can also fail outright. In the FDA-reviewed concordance study, samples were collected from 386 donors, and 11 of them failed testing on Guardant360 CDx because of an instrument failure during a power outage.

• Which specific alterations were reported, and at what variant allele fraction?
• Is any of them an approved companion diagnostic match, or only informational?
• Was the amount of cell-free DNA in my sample adequate?
• Could any finding be from clonal hematopoiesis instead of the tumor?
• If nothing was found, is tissue testing possible now?
• Was this the FDA-approved CDx version of the test, or a broader laboratory version?

• https://www.accessdata.fda.gov/cdrh_docs/pdf20/P200010B.pdf
• https://www.accessdata.fda.gov/cdrh_docs/pdf20/P200010S021B.pdf
• https://www.fda.gov/medical-devices/recently-approved-devices/guardant360-cdx-p200010s010
• https://www.cancer.gov/news-events/cancer-currents-blog/2020/fda-guardant-360-foundation-one-cancer-liquid-biopsy