Evidence Building for Use of Liquid Biopsy to Drive Treatment Decisions, Monitor Disease Activity

Emerging evidence shows that liquid biopsy is a viable option for detecting and monitoring genomic mutations to inform treatment decisions and to assess disease activity in real-world clinical settings.

Among the technology’s many hopes is that it can inform genetic inquiries when tissue biopsies are not safely feasible, that it has the sensitivity to detect changes in resistance and disease burden before imaging modalities, that it can overcome the sampling challenge of tumor heterogeneity, and that eventually, the technology will enable pan-cancer screening for earlier disease detection.

Despite the promise that liquid biopsy can provide valuable clinical information in a minimally invasive manner, experts caution that the evidence base remains incomplete to warrant widespread clinical adoption of blood-based applications of next-generation sequencing (NGS) in cancer care.

The FDA has approved the CellSearch system (Menarini-Silicon Biosystems) for clinical use in testing for circulating tumor cells in cancer patients, the ability to detect circulating tumor cells, particularly in low-volume samples, remains a concern. Some experts are more encouraged about the prospect for adoption of sequencing technologies that can analyze circulating tumor DNA (ctDNA), even though estimates are that tumor-derived cfDNA accounts for less than 1 percent of total circulating DNA.

There is early evidence supporting the potential for liquid biopsy modalities to inform treatment of lung, melanoma, breast, ovarian, cervical, and bladder cancers.  But currently, availability of tests outpaces the data supporting their use. DTET examined some recent studies providing valuable evidence supporting early clinical adoption.

Driving Targeted Therapy in Lung Cancer
Plasma NGS genotyping is feasible, rapid, and useful in the real-world clinical practice setting for patients with advanced non-small cell lung cancer, according to a prospective study published Nov. 28, 2018 in the Journal of the National Cancer Institute. A variety of oncogenic drivers can be identified from plasma ctDNA that can drive treatment decisions (e.g., targeted treatment) more quickly than tissue-based sequencing.

“We have shown that plasma-based ctDNA NGS assays allow for rapid and noninvasive genotyping that could immediately guide precision therapy, providing an important supplement to tissue NGS, as well as an important alternative when tissue biopsy is not feasible,” write the study authors.

Researchers in the United States and Australia performed NGS targeting 21 genes in ctDNA from plasma samples of 210 patients with advanced non-small cell lung cancer. A subset of 106 patients had concurrent tissue NGS testing using a 468-gene panel. Most patients (81.4 percent) had previous conventional molecular testing for EGFR, ALK, and ROS1 mutations. The laboratory-developed test, ResBio ctDx-Lung assay (Resolution Bioscience), was used to evaluate plasma samples and the hybridization capture-based MSK-IMPACT assay (Memorial Sloan Kettering) was used for tissue-based genotyping.

Using ctDNA, the researchers detected somatic mutations in just under two-thirds of patients (135 of 210). Plasma NGS genotyping identified actionable driver mutations that led to a match with targeted therapy and clinical response in 46 of the 135 patients with identified mutations (EGFR, ALK, MET, and BRAF).  As might be expected, ctDNA detection of mutations was significantly lower in patients who were on systemic therapy at the time of plasma collection versus those who were not (42.9 percent versus 75.0 percent).

The median turnaround time for plasma NGS results was significantly shorter than for tissue NGS (9 versus 20 days). In 60 of the 105 patients who had concurrent plasma and tissue NGS, 56.6 percent of patients had at least one identical genomic alteration identified in both tissue and plasma. Among patients who tested plasma NGS positive, 89.6 percent were also concordant on tissue NGS. For patients who tested tissue NGS positive, 60.6 percent were also concordant for plasma. More specifically, for patients who tested plasma NGS positive for oncogenic drivers, tissue NGS concordance was 96.1 percent.

“A driver alteration identified by plasma NGS can immediately direct clinical care,” write the authors led by Joshua Sabari, M.D., from Memorial Sloan Kettering Cancer in New York. “However, a negative result requires further investigation. Based on our findings, plasma NGS genotyping is best performed at initial diagnosis in conjunction with tissue biopsy and at the time of clinical or radiologic progression, as the yield of ctDNA might be highest at those times based on its correlation with tumor burden.”

Monitoring Melanoma Activity
Assessing ctDNA can provide evidence of melanoma activity before it is detectable on imaging, according to a study published in the October issue of Molecular Oncology. The authors say this improves the current standard of care, particularly in melanoma patients, in whom the disease can metastasize to unusual sites.

The researchers prospectively enrolled patients into three cohorts. The first group had 60 patients with radiographically measurable metastatic melanoma. The second group of 29 patients had surgically removed high-risk (stage IIB-IV) melanoma whose tumor tissue revealed any of the seven common mutations. The third group included 30 patients who were receiving or had received therapy and had any of the seven common mutations. All patients’ plasma samples were evaluated with a polymerase chain reaction-based BEAMing (beads, emulsions, amplification, magnetics) assay (Sysmex Inostics) to assess mutational status (seven BRAF and NRAS somatic mutations), total tumor burden, and when appropriate, inform targeted therapy selection.

The researchers identified 260 plasma mutations with BEAMing across all three cohorts. In 60 patients who underwent both plasma and tissue testing, tumor tissue testing revealed one of the seven mutations of interest in nearly two-thirds of patients (38 of 60). In 33 of those 38 patients (86.8 percent), mutations identified in ctDNA exactly matched the mutations found in tissue samples, yielding a sensitivity and specificity of the ctDNA assay of 86.8 percent and 100 percent, respectively. As might have been expected, higher tumor burden and visceral metastases were associated with detectable ctDNA.

In the 29 patients with surgical removal of the tumor, five had recurrent melanoma during the study, which was detectable by ctDNA in two cases. Among the 30 patients receiving treatment, 17 responded to treatment. Among the 13 nonresponders, for four of those 13, CT scan and ctDNA results detected disease activity simultaneously. In four other cases, ctDNA results predicted disease progression that was later confirmed by imaging.

Interestingly the sensitivity of the BEAMing assay appears to be impacted both by disease burden and by location of the tumor, the authors say. ctDNA levels were not detectable in patients with only locoregionally recurrent disease in skin and lymph nodes. Further, metastatic disease confined exclusively to the lungs and/or brain often failed to produce detectable amounts of ctDNA.

“Our study results demonstrate that incorporating ctDNA assessments into real-world melanoma patient management can influence patient care decisions, alter radiographic interpretations, and impact clinical outcomes,” write the authors led by Steven Rowe, M.D., Ph.D., from Johns Hopkins University in Baltimore, Md.

The Future of Liquid Biopsy
While the emerging evidence supporting the clinical use of liquid biopsy is promising, many unanswered questions remain. For instance, it is not yet clear how frequently plasma-based assessments of targetable somatic mutations are needed to improve patient outcomes, particularly when the mutation is not identifiable in tumor-based molecular testing and tumor growth is not apparent on imaging studies.

“Many technical hurdles have been resolved thanks to newly developed techniques and NGS analyses, allowing a broad application of liquid biopsy in a wide range of settings. … Still application is far from reality but ongoing research is leading the way to a new era in oncology,” writes Raffaele Palmirotta, from University of Bari Aldo Moro in Italy, in a review published Aug. 29, 2018 in Therapeutic Advances in Medical Oncology. “Large-scale and multicenter trials are also ongoing to confirm all the potentialities that are now being studied in order to fully define the exact settings and conditions for the application of liquid biopsy and confirm the comparison of performance with current solid biopsy methods.”

Experts also say that financial data is lacking, including cost benefit analysis compared to conventional biopsies and imaging, evidence of improved outcomes, and the potential downstream costs (or savings) resulting from plasma-based NGS.

“For successful integration of plasma NGS into clinical practice, universal guidelines, both from an informatics and a technical standpoint, are essential and need to span multiple companies and institutions,” writes Sabari and colleagues. “Not all assays are the same.”

Takeaway: Evidence is mounting supporting plasma-based NGS for detection of genomic mutations in cancer patients. While the potential exists to apply the technology to screening, diagnosis, prognosis, and monitoring of disease, experts believe use of the technology to inform treatment decisions and to monitor disease activity are the applications closest to becoming a clinical reality.