Skip to main content
Premium Trial:

Request an Annual Quote

AACR: Posters Add to Clinical Data for Guardant360, Show First Evidence of Impact on Docs' Decisions


NEW YORK (GenomeWeb) – Investigators using Guardant Health's ctDNA sequencing test shared data from a variety of studies at this week's annual meeting of the American Association for Cancer Research.

Posters highlighted user experiences with the technology in ongoing studies and provided data on Guardant360's potential clinical utility in specific cancer subtypes like breast cancer. They also offered an encouraging hint at the test's real world impact via its influence on the attitudes and decision making of oncologists.

The company also presented its own poster updating a clinical validity analysis of its growing archive of commercial cases.

Guardant's test relies on proprietary technology it calls digital sequencing, which involves using unique barcodes to tag each single strand of a DNA fragment prior to enrichment and sequencing.

Guardant CEO Helmy Eltoukhy told GenomeWeb at the conference that the company believes it has overwhelmingly demonstrated Guardant360's analytical and clinical validity at this point, and is now firmly focused on establishing the test's clinical utility and impact on practice and patient outcomes.

"We are really beyond the clinical validation phase. We've done thousands of patients with matched tumor samples, and now it's about actually showing that [blood-based] genomically informed clinical decision making can inform clinical outcomes," Eltoukhy said.

The company's major presentation at AACR in that vein was a poster authored by researchers from MD Anderson, who reported on 105 metastatic colorectal cancer patients sequenced using Guardant's test as part of an ongoing clinical utility study.

According to the authors, plasma testing was successful for all 105 patients, while tissue sequencing was only possible in 90 percent.

Overall, 85 percent of the patients had at least one detectable genetic alteration, the group wrote. Among those with successful tissue sequencing results to compare, the researchers found that 38 percent had mutations detected only in the plasma that were not present in tissue. TP53 and SMAD4 mutations were the most common of these.

For many cases, the authors wrote, it appeared that these plasma-only mutations represent minor clones that may have evolved later in the course of patients' cancers. This hypothesis was supported by the fact that the team saw a much lower median allele frequency — about four percent — for these "discordant" mutations compared to those that were present in both blood and tissue.

According to Eltoukhy, this ability to uncover putative therapeutic targets that might not be detectable by tissue sequencing, especially later subclonal alterations associated with acquired resistance to standard therapies, is one of the main areas Guardant believes its 360 test can make an impact.

"We've had dramatic patient stories where a patient has had a re-biopsy done but nothing actionable was found, and then they do our test and we find an additional EGFR or BRAF mutation that is actionable. We've seen dramatic responses with significant tumor volume reduction and much longer progression-free survival," he said.

Beyond this research on mutation prevalence and the sensitivity of Guardant360, the MD Anderson group then went on to track the impact of test results on the clinical care of patients. According to the authors, about 87 percent of the time, doctors said that ctDNA testing improved the quality of care they provided their patient.

Overall, Guardant's ctDNA approach uncovered actionable alterations, as determined by the physician, in 37 percent of patients, the researchers reported. Of these, 56 percent have thus far enrolled or have plans to enroll in a clinical trial based on the results.

According to Eltoukhy, the MD Anderson study will continue to follow these patients through treatment to establish the impact of Guardant's test on their outcomes. 

Scott Kopetz, the study's first author, told GenomeWeb in an email that he and his colleagues expect it will be another six months to a year before they have their first meaningful outcome data.

Beyond the MD Anderson study, Guardant also presented a poster at the meeting highlighting its growing archive of data on commercially sequenced cases.

Eltoukhy told GenomeWeb that Guardant has processed close to 10,000 samples to date overall.

The company previously discussed about 1,500 of these in a presentation at Tri-Con earlier this year. AmirAli Talasaz, president and COO of Guardant, said in that talk that 77 percent of patients tested have had at least one alteration and 75 percent have had an actionable alteration when tested with Guardant360.

In its more recent analysis shared at AACR, Guardant researchers updated this data to more than 2,000 cases across a number of tumor types. About 1,300 were analyzed using a 54-gene version of the company's panel, while another 660 were tested using an updated 68-gene panel.

According to the authors, Guardant was able to successfully process and analyze 99.5 percent of the samples received. Use of the newer 68-gene panel apparently increased the mean number of alterations detected per patient — from 3.4 to 3.7.

The overall detection rate for SNVs held steady at 72 percent from the first to second panel, but the rate of copy number variations increased from 13 percent to 20 percent with the switch to the larger panel. In some specific cancer types, the test has proven itself to be even more sensitive, detecting at least one alteration in up to 85 percent of colorectal cancer patients, for example.

Interestingly, the company also broke out data in the poster on mutation frequencies and their putative link to either a tumor or germline origin.

In a study published last week in Science Translational Medicine by a team from Johns Hopkins, researchers concluded that analyzing only the tumor but not matched normal tissue can yield false-positive alterations that might misguide therapeutic decision-making.

In light of the debate over those results, Eltoukhy said it's worth noting that sequencing plasma has essentially a built-in control mechanism to distinguish both germline and tumor mutations. Since all circulating cell-free DNA in the blood is sequenced in Guardant's approach, it means the test also picks up germline variants, which can be distinguished by their relative concentration.

Germline mutations will be present at close to fifty percent or 100 percent of the dominant allele frequency depending on whether they are heterozygous or homozygous. Somatic mutations, meanwhile, are only present at much lower fractions, the company illustrated in its poster.

Other research

There were several other presentations at AACR from users of Guardant360. In one set of posters, researchers from Thomas Jefferson University Hospital reported on both the concordance of Guardant's ctDNA analysis with tissue-based sequencing and shared evidence of the platform's clinical utility in guiding care for patients with metastatic breast cancer.

The TJU team, led by Massimo Cristofanilli, reported that analysis of ctDNA in 33 metatstatic breast cancer patients showed both a high concordance with tissue sequencing results, and, much like in the MD Anderson study, also appeared to reveal novel subclonal mutations.

In a second poster, the same TJU group reported on ctDNA testing among 76 patients with advanced or metastatic breast cancer. According to the authors, ctDNA was detected in 84 percent of these patients, and of those patients with detected mutations, 52 percent had an actionable mutation that was incorporated into treatment planning.

In an email to GenomeWeb, Cristofanilli said that his group has been working with the Guardant test for more than 15 months and has tested approximately 130 patients overall.

"I think we have now an understanding of the powerful information provided by the longitudinal use of Guardant360, and we hope to summarize the data in a [future] original manuscript," he added.

The TJU team was not the only one at the meeting sharing data from use of Guardant360 for breast cancer. Another group from the University of Pennsylvania also presented a poster on the METAMORPH study, in which researchers are collecting and analyzing ctDNA using Guardant's platform alongside sequencing of metastatic biopsies using the Illumina TruSeq cancer panel.

In the poster, the group wrote that both tests have been performed for 28 patients. Among that group, 68 percent had at least one alteration in their metastatic sample, while 86 percent had at least one mutation in their ctDNA.

Overall, about half of the patients had the same alterations identified in both metDNA and cfDNA, and 80 alterations were identified in total, 23 of which were detected by both assays.