NEW YORK – Liquid biopsy genomic profiling alone was as effective as tissue-based genomic testing in guiding treatment decisions in EGFR-mutated lung cancer patients in a recently published study out of MD Anderson Cancer Center.
The findings lend further support to oncologists' growing desire to use liquid biopsy as a frontline tool to quickly scan for targetable markers and, if that turns up empty, to use tissue-based next-generation sequencing to cast a wider net.
It's currently standard practice to identify targetable genetic alterations in tumor tissue samples from lung cancer biopsies, but there are challenges with this approach that make liquid biopsy testing a more attractive option for oncologists looking to evaluate patients quickly and get them on the right drug. Tissue-based testing is both time-consuming and invasive for patients with advanced cancer who need treatment quickly and may be in fragile health.
The turnaround time for getting results from tissue-based genetic testing can be as long as a month due to the need to schedule a biopsy and then process and send tissue samples for analysis, said John Heymach, chair of thoracic and head and neck medical oncology at MD Anderson Cancer Center.
"Liquid biopsies are clearly an advance in the field," Heymach said. "We knew we could detect mutations in EGFR, ALK, and others in the blood, but it's never been asked directly if you get the same clinical outcomes when you detect it by blood or by tumor and make treatment decisions from those results."
In a study published in JCO Precision Oncology in August, Heymach and MD Anderson researchers compared the outcomes of non-small cell lung cancer patients with EGFR mutations in their tumors identified using only liquid biopsy, with patients profiled using a tissue-based test. The researchers found that patients in both groups had similar outcomes on EGFR inhibitors.
The study findings support the use of liquid biopsy testing alone as a frontline tool for gauging EGFR mutations in NSCLC and may help patients receive personalized treatment more quickly, said Heymach, the senior author of the study.
The researchers retrospectively analyzed genomic profiling results and clinical outcomes data from 80 patients with EFGR-mutant advanced NSCLC, focusing specifically on the common EGFR exon 19 deletions and L858R mutations. Half of those patients were profiled using either the Guardant360 CDx ctDNA assay and the other half were tested on tissue-based NGS assays performed within MD Anderson's laboratory.
The study included ctDNA test results from more than 1,900 patients seen at MD Anderson between May 2015 and February 2018 and included in its Lung Cancer Moon Shot GEMINI database. Researchers focused their analysis on EGFR exon 19 deletions and L858R mutations as these were the most common targetable gene alterations in NSCLC patients. The analysis was further narrowed to a cohort of 80 patients based on the availability of complete treatment data.
While Guardant wasn't directly involved in this research beyond sequencing the samples, Heymach said, MD Anderson has an ongoing partnership with the company to research the utility of its liquid biopsy assay in NSCLC patients. The Guardant360 assay has been approved in the US as a companion diagnostic for osimertinib (AstraZeneca's Tagrisso) to identify treatment-eligible advanced NSCLC patients with EFGR exon 19 deletions, L858R, and the T790M resistance mutation.
MD Anderson's molecular diagnostic lab used two tissue-based tests to assess research participants' samples during the study period. Initially, they used a targeted NGS test to analyze hotspot regions in 50 genes. The lab later used an expanded NGS test, Thermo Fisher's OncoMine Comprehensive Assay V1, to detect a wider range of mutations.
Patients in both liquid biopsy and tissue testing groups received front-line US Food and Drug Administration-approved EGFR inhibitors, including erlotinib (Genetech's Tarceva), gefitinib (AstraZeneca's Iressa), afatinib (Boehringer Ingleheim's Gilotrif), or osimertinib (AstraZeneca's Tagrisso). There was no statistically significant difference in the median progression-free survival between patients who had tissue NGS and liquid biopsy testing, 379 days versus 353 days, respectively.
Heymach and his colleagues further assessed the concordance between ctDNA and tissue test results by looking at 217 matched results from the two types of tests within the GEMINI database. They found a high concordance between the two modalities. In a cohort of 109 patients from the GEMINI database with sensitizing EGFR mutations, including exon 19 deletion and exon 21 L858R, there was 98 percent concordance between ctDNA and tissue results.
The GEMINI database also included ctDNA results from NSCLC patients with mutations in BRAF, MET exon 14 skipping, and gene fusion groups ALK, RET, and ROS1, but there was not sufficient data from those patients to conduct a similar outcomes comparison.
One challenge in using liquid biopsy tests alone, however, is their sensitivity. The researchers noted there has been concern among oncologists that liquid biopsy testing would detect tiny amounts of a mutation in ctDNA that may not represent the tumor DNA. The Guardant360 test used in this study has a sensitivity of 0.1 percent.
To test whether this affected treatment outcomes, the researchers split the liquid biopsy group into two subsets: patients with low levels of mutated EGFR in ctDNA with a variant allele frequency (VAF) between 0.01 percent and 0.99 percent, and those with high levels of EGFR mutations in ctDNA with a VAF of 1 percent or greater. They hoped to determine if patients in the low-ctDNA group benefitted from EFGR inhibitors like patients in the high-ctDNA group.
EGFR inhibitors reduced tumors in patients in both groups. The high-ctDNA group, on average, saw tumors shrink by nearly half and the low-ctDNA group saw a mean tumor reduction of 26 percent. The low-ctDNA group, however, fared better in terms of median progression-free survival compared to the high-ctDNA group, 424 days versus 293 days, respectively. This suggested that the low frequency of EGFR mutations in ctDNA did not affect treatment outcomes.
While the present study focused only on patients with EGFR mutations, oncologists should test NSCLC patients for a number of other biomarkers to give them the best chance of matching to a targeted drug. While liquid biopsy is a promising new tool to do that, its performance isn't as well established for all the potential targetable biomarkers in NSCLC and there's a small chance liquid biopsy profiling could miss actionable mutations. As such, there continues to be a need for tissue-based NGS profiling despite its challenges, Heymach noted.
Moreover, liquid biopsy is still relatively new, and doctors have become comfortable ordering tissue-based genomic profiling for NSCLC patients using tumor samples from patient biopsies used to diagnose their lung cancer subtype.
Although the method of obtaining a tumor tissue sample has become less invasive with the use of fine needle aspirate to avoid further damage to the patients' lungs, those procedures yield smaller samples. "Tissue specimens are getting smaller as we do less invasive biopsies, but there's also more molecular tests to do," Heymach said. "So, we have more and more demands on tissue, but less and less tissue to do it with."
That can lead to patients seeing even longer delays for genomic test results and more biopsy procedures if there is insufficient tissue to conduct all the tests needed, Heymach added. In his experience, about a quarter of biopsies don't get sufficient tumor tissue for genomic analysis.
Heymach hopes his group's study will assure oncologists that they can safely use a liquid biopsy test to assess EGFR mutations in NSCLC patients, and that if an EGFR mutation is identified by a liquid biopsy assay, even at low frequencies, then their patients will likely benefit from a targeted therapy. He is optimistic that liquid biopsy will continue to demonstrate its utility as a first-line tool for guiding cancer therapy decisions in other lung cancer indications and tumor types. In the meantime, he sees oncologists reaching for both types of tests to maximize precision oncology treatment opportunities for patients.
Specifically, he proposed that clinicians could order a liquid biopsy assay upon diagnosis to quickly determine if a patient can be matched to an EGFR-targeted therapy, and if that test is inconclusive, then a tissue-based test should be ordered.
The study comes at a time when the oncology community is increasingly willing to use both liquid biopsy and tissue NGS-based companion tests to identify targetable tumor mutations. The FDA certainly has been approving both modalities in a number of NSCLC indications to expand testing options for patients. For example, to assess the common EGFR mutations in lung cancer, oncologists now have access to FDA-approved liquid biopsy and tissue-based NGS companion tests for several EGFR inhibitors.
The FDA is also expanding the CDx indications of both types of tests into other biomarker-defined NSCLC settings. For example, in July, the FDA approved Foundation Medicine's FoundationOne Liquid CDx to identify NSCLC patients whose tumors have mutations resulting in MET exon 14 skipping for treatment with capmatinib (Novartis' Tabrecta). Earlier, the agency had approved the tissue-based version of Foundation's test as a companion test for the same drug.
"It makes sense to consider doing both a liquid biopsy and tumor biopsy to make sure you don't miss any of the alterations, because both methods can miss something occasionally," Heymach said, noting that missing a targetable mutation can really impact patients' treatment options and outcomes. "If you find an EGFR mutation, it can add years of life. If you miss that mutation, their options are usually just chemotherapy and immunotherapy, which often doesn't work very well for these patients."