|Subject: Circulating Tumor DNA Testing for Cancer (Liquid Biopsy)|
|Document #: GENE.00049||Publish Date: 12/16/2020|
|Status: Reviewed||Last Review Date: 11/05/2020|
This document addresses cell-free circulating tumor DNA (ctDNA) testing, from a blood sample, as an alternative to tissue biopsy in the diagnosis of cancer, and for clinical response to targeted agents of cancer treatment. These tests are also known as liquid biopsies. Examples of these tests include, but are not limited to:
Note: This document does not address circulating tumor cell (CTC) testing. For more information on CTC tests, please see the following:
Note: This document does not address single-gene EGFR testing. For more information on EGFR tests, please see the following:
Note: When another document exists addressing a specific condition or genetic test, that document supersedes this one.
Note: For more information on related topics, please see the following:
Investigational and Not Medically Necessary:
The use of a circulating tumor DNA (ctDNA) test for the diagnosis or treatment of cancer is considered investigational and not medically necessary for all indications.
According to the American Society of Clinical Oncology (ASCO), the significance of ctDNA tests are determined by assessing analytical validity (the test can accurately and reliably detect a biomarker), clinical validity (the test can detect the presence or absence of cancer), and clinical utility (the test can improve the outcomes of individuals with cancer). Even though several biomarkers have been shown to be useful for targeting and treating cancer, it cannot be assumed ctDNA tests that look for these biomarkers are automatically valid. Each ctDNA test must demonstrate accuracy, as compared to a tissue biopsy, before it can be used to make clinical decisions. Furthermore, low shedding of tumor DNA in some individuals may result in false-negative testing findings, when ctDNA is ordered in the absence of a tissue biopsy.
In a single-center observational study, Thompson and colleagues (2016) examined the concordance between tissue biopsy samples and Guardant360 blood samples for individuals with non-small cell lung cancer (NSCLC). A total of 102 subjects with a diagnosis of NSCLC or suspected NSCLC were included in the study. Tissue samples (n=50) were processed using the Illumina TruSeq Amplicon 47 gene cancer panel (n=38) or the 20 gene Penn Precision Panel (n=12). For the 50 subjects who had both blood and tissue tests, the overall concordance was 60%. For EGFR mutations, the overall concordance was 79%. The authors concluded that ctDNA testing has potential for real-time molecular monitoring for individuals with advanced cancer. Several studies have also compared Guardant360 to tissue-based broad molecular profile tests with similar or mixed conclusions (Chae, 2016; Hahn, 2017; Pishvaian, 2016; Sandulache, 2017; Schwaederle, 2016; Schwaederle, 2017; Villaflor, 2016; Yang, 2017). These studies were limited by small sample sizes, qualitative methods, or imperfect comparators.
McCoach and colleagues (2018) performed a retrospective cohort study to determine the clinical utility of Guardant360 for detecting anaplastic lymphoma kinase (ALK) fusions in NSCLC during diagnosis or during treatment with ALK inhibitors. The researchers included 88 subjects with 96 plasma-detected ALK fusions from the Guardant360 de-identified database. Subjects were separated into 4 cohorts: cohort 1 (n=42) contained subjects with a newly discovered ALK fusion, cohort 2 (n=31) contained subjects with a known or presumed ALK fusion and whose cell-free DNA (cfDNA) was obtained at progression, cohort 3 (n=13) contained subjects without additional clinical information, and cohort 4 (n=6) contained subjects who had been treated with anti-EGFR targeted therapy and found to have an ALK fusion by cfDNA. In cohort 1, the Guardant360 test found an ALK fusion in 16 subjects who had been reported as tissue-negative or tissue insufficient. Of the 42 subjects in the cohort, 10 had tissue samples available (5 ALK-positive, 5 ALK-negative), 11 had insufficient samples, and 21 did not have ALK information available. For the 5 subjects who were identified by Guardant360 as ALK-positive despite negative tissue biopsies, 3 eventually responded to ALK inhibitor therapy while clinical data was not available for the other 2 subjects. For cohort 2, 16 samples contained 1-3 ALK resistant mutations. For 5 samples, an ALK kinase domain mutation was identified in cfDNA despite the ALK fusion not detected in cfDNA and the prior tissue sample showing an ALK fusion. For cohort 3, the clinical status was unknown and no resistance mutations or bypass pathways were identified. For cohort 4, 6 subjects were found to have ALK fusions. The authors concluded that cfDNA NGS testing is an “additional tool” for detecting alterations, resistance mutations, and bypass pathways. Limitations of the study included the retrospective design and lack of clinical data for some subjects. The authors noted that tissue evaluation was at the providers’ discretion and the testing method was not available for all subjects. Furthermore, no sensitivity or specificity information was provided.
Aggarwal and colleagues (2018) conducted a single-center, prospective study to assess mutation detection using Guardant360 for individuals with stage IV NSCLC. A total of 323 participants had Guardant360 plasma testing as part of clinical management. The primary outcomes were targetable alterations detected with plasma and tissue next-generation sequencing, the association between allele fractions of mutations detected in tissue and plasma, and the association of response rate with the plasma allele fractions of the targeted mutations. For 113 individuals, therapeutically targetable mutations were detected in EGFR, ALK, MET, BRCA1, ROS1, RET, ERBB2, or BRAF. Of 94 participants who had plasma testing alone, 31 had a targetable mutation detected and were considered to not need tissue biopsy. For the 229 participants who had concurrent plasma and tissue testing or were not able to have tissue testing, an additional 35 targetable mutations were detected. For those who received targeted therapy based on the plasma result, 36 out of 42 participants had complete/partial response or stable disease. Of the 128 subjects with concurrent plasma and tissue next generation sequencing results, 8 therapeutically relevant mutations were found in plasma only, 31 were detected in both plasma and tissue, and 16 were detected in tissue only, with an overall concordance of 81.3%. Therapeutically targetable mutation detection was highest for individuals with liver metastases (100% concordance with tissue [n=13]) compared with individuals with M1a disease (46.2% concordance). Based on the level of discordance found in the study, the authors note that “a tissue biopsy remains essential for initial cancer diagnosis”; however, in the setting of inadequate tissue DNA, “plasma NGS can be an adequate surrogate for molecular profiling.” The study was limited by a single-center design, potential user bias, and the consideration of plasma testing at a single point. The study was also enriched with individuals who underwent testing after progression to detect resistance mutations, which likely increased the frequency of individuals with EGFR T790M. The long-term outcomes of employing Guardant360 plasma testing in the clinical management of stage IV NSCLC versus, or in conjuction with standard tissue biopsy remains uncertain, as does the potential risk of false-negative results.
Leighl and colleagues (2019) reported on the multicenter, prospective NILE (Non-invasive versus Invasive Lung Evaluation) study, which aimed to assess the clinical utility of Guardant360 for the identification of eight genomic biomarkers (EGFR, ALK, ROS1, BRAF V600E, RET, MET, MET exon 14, ERBB2 [HER2]) in individuals with newly diagnosed metastatic NSCLC. A total of 307 individuals were enrolled with biopsy-confirmed, previously untreated non-squamous NSCLC (stage IIIB/IV) and tissue genotyping (genomic testing and PD-L1 expression analysis using next generation sequencing polymerase chain reaction “hotspot” testing, FISH and/or IHC, or Sanger sequencing). Participants submitted a pre-treatment blood sample for Guardant360 testing. A total of 282 individuals met all inclusion criteria and were included in the final analysis. Tissue genotyping for all eight biomarkers was completed in 51 individuals (18.1%) (the majority of individuals had sequential individual biomarker testing, and did not undergo physician-directed sequencing of all eight genomic biomarkers), and Guardant360 testing for all eight biomarkers was completed in 268 individuals. One of eight biomarkers was identified in tissue samples in 60 individuals compared to 77 individuals with Guardant360 (p<0.0001). For 60 individuals with tissue-positive results, one of the eight biomarkers was identified in tissue alone (n=12) but not with Guardant360, a false-negative rate of 20%. In regards to these 12 individuals, the researchers note: “the lack of full genomic assessment obtained by comprehensive cfDNA genomic profiling may have led to the patient being treated with a less efficacious therapy.” While the primary objective to demonstrate non-inferiority of Guardant360 compared to tissue-based genotyping was achieved, the study was limited in that only 18% of participants received comprehensive tissue genomic profiling. As with other research on the topic, a substantial number of false-negative results were obtained by cfDNA, which can lead to undertreatment.
Zugazagoitia and colleagues (2019) evaluated the ability of the Guardant360 test to identify individuals with NSCLC in routine clinical practice who have tyrosine-kinase inhibitor (TKI) resistance. This was a prospective study that included 53 individuals with EGFR, ALK or ROS1-altered advanced stage NSCLC who experienced progression (clinical or radiological) on prior TKI therapy. The sample was divided into 3 subgroups; 1) EGFR-mutant NSCLC with resistance to first/second-generation EGFR TKI (cohort 1, n=31); 2) EGFR T790 + NSCLC with osimertinib resistance (cohort 2, n=15) and ALK/ROS1-rearranged NSCLC with resistance to crizotinib and/or next generation ALK/ROS1 TKI (cohort 3, n=7). Individuals with sufficient tumor DNA shedding such that plasma findings could be adequately interpreted were classified as “shedders”. In cohort 1, 20 individuals (65%) were classified as shedders and 9 (29%) were found to have EGFR T790 M mutations with Guardant360 testing. In 2 additional individuals, EGFR T790 M mutations were identified by another method; these 11 individuals received subsequent osimertinib therapy. In cohort 2, Guardant360 testing in 10 individuals were classified as shedders and, in 9 of these, at least 1 pathologic alteration in addition to the EGFR sensitizing and/or T790M mutation was detected. None of the individuals in cohort 2 received subsequent targeted therapies. In cohort 3, which included only 7 individuals, 4 individuals were shedders and were found to have actionable alterations. Two individuals in cohort 3 received subsequent treatment informed by Guardant360 testing. A substantial number of individuals in the study were not considered to be tumor DNA shedders and this study did not compare outcomes in individuals managed with and without the Guardant360 test.
Murray and colleagues (2017) investigated the analytical and clinical validity of the Colvera plasma test for the detection of methylated BCAT1 and IKZF1 in individuals with colorectal cancer (CRC). The researchers randomized 264 plasma samples and 120 buffer samples, divided the samples into 8 batches of 48, and processed the samples over 8 days using 2 equipment lines. Clinical validity was analyzed by using Colvera on 222 archived plasma samples (n=26 with known colorectal cancer) from individuals who were scheduled for colonoscopy as part of a previous trial (Pedersen, 2015). The researchers found that the limit of detection (LOD) was 12.6 pg/ml (95% confidence interval [CI], 8.6 to 23.9), the equivalent of 2 diploid genomes/ml of plasma. Colvera tested positive for 19/26 known cancer cases for an agreement of 73% (95% CI, 52% to 88%). For the 196 nonneoplastic subjects, Colvera had an agreement of 89% (95% CI, 84% to 93%). Total agreement was 87% (194/222; 95% CI, 82% to 91%). Limitations of the study included a small sample size.
In 2020, Musher and colleagues published a cross-sectional study evaluating the diagnostic accuracy of the Colvera test compared with carcinoembryonic antigen (CEA) for identifying recurrence of CRC. The study enrolled 537 adults who were undergoing surveillance after treatment for stage II or III CRC. Blood samples were collected at a single time point, within 6 months of surveillace radiological imaging, and evauated using the Colvera test and CEA. A total of 322 (60%) individuals were included in the final analysis; 20 were excluded because they did not meet eligibility criteria and 195 were excluded for insufficient information. Among the evaluable participants, CRC recurrence occurred in 27 (8.4%) of individuals. The sensitivity of the Colvera test for detecting CRC recurrence (63%) was significantly higher than CEA testing (48.1%), p=0.046. However, the specificity of CEA testing (96.3%) was significantly higher than Colvera testing (91.5%), p=0.012. A substantial proportion of study participants were excluded from the analysis.
In a prospective-retrospective cohort study, Grasselli and colleagues (2017) explored the concordance between RAS in tumor tissue and ctDNA for metastatic colorectal cancer to establish eligibility for anti-EGFR therapy. Blood plasma samples and tissue samples were obtained from 146 individuals with a diagnosis of colorectal cancer. RAS status was determined with plasma and tissue samples using BEAMing and real-time PCR as standard of care (SoC) technique. The median time from tissue specimen to ctDNA collection was 1.2 months in therapy-naïve individuals and 20.2 months in previously exposed individuals. The ctDNA BEAMing RAS agreement with SoC was 89.7% compared with 90.0% agreement with SoC for BEAMing in tissue. A total of 15 individuals had discordant tissue-plasma results; there were 9 individuals with low frequency RAS mutations not detected in tissue and 6 individuals with RAS mutations not detected in plasma. Prediction of treatment benefit for individuals receiving anti-EGFR plus irinotecan was equivalent between the two groups. The authors concluded that “ctDNA analysis in plasma can detect RAS mutations to an equivalent level as SoC techniques in tissue.” They noted the study was limited by a small cohort size.
In 2018, Garcia-Foncillas and colleagues published a prospective, multicenter study that compared OncoBEAM to tissue analysis for metastatic colorectal cancer. A total of 236 individuals underwent tissue biopsy and blood sample collection at 10 centers between November 2015 and October 2016. RAS mutations were detected in 55.5% of tissue samples and 51.3% of plasma samples. The researchers found that the overall percentage agreement between plasma-based and tissue-based RAS mutation testing was 89%. Performing a re-analysis with BEAMing of tissue from discordant cases, they found 2 false negative and 5 false positive tumor tissue RAS cases, making the final concordance 92%. They concluded that ctDNA analysis by OncoBEAM is comparable to SoC tissue testing techniques for RAS in individuals with colorectal cancer.
Other ctDNA Tests
CancerSEEK (Johns Hopkins Kimmel Cancer Center, Baltimore, MD) is a liquid biopsy test that is in research and development and not commercially available at this time. Cohen and colleagues (2018) developed CancerSEEK, a genetic alteration and protein biomarker assay, to detect early cancers and reveal the origin of the cancer. The researchers used CancerSEEK to evaluate 1005 subjects already diagnosed with stage I-III cancers of the ovary, liver, stomach, pancreas, esophagus, colorectum, lung, or breast. A control cohort consisted of 812 subjects with no known history of cancer. The researchers found that CancerSEEK had a median sensitivity of 70% and a specificity greater than 99%. In the healthy cohort, 7 subjects tested false-positive with CancerSEEK. The researchers compared CancerSEEK to tissue samples for 153 subjects and found that the mutation in the plasma sample was identical to the mutation in the tumor for 138 subjects (90%). The researchers used supervised machine learning to examine CancerSEEK’s ability to find the origin of cancer. The test was able to localize the origin of the cancer to two anatomic sites in a median of 83% of subjects and to a single organ in a median of 63% of subjects. While the results for CancerSEEK are promising, the researchers state that “to actually establish the clinical utility of CancerSEEK and to demonstrate that it can save lives, prospective studies of all incident cancer types in a large population will be required.”
There is a paucity of published peer-reviewed literature, especially large-scale, high-quality prospective randomized trials, which determine the validity and utility of these tests compared to traditional pathologic examination. In some cases, studies on plasma-based testing are difficult to compare to each other given that different tissue tests are used for comparison. In addition, there is insufficient evidence that these tests improve health outcomes.
In a joint-review analysis on circulating tumor DNA (Merker, 2018), ASCO and the College of American Pathologists (CAP) stated:
The National Comprehensive Cancer Network (NCCN) guidelines (V8.2020) on non-small cell lung cancer stated:
A joint guideline (Lindeman, 2018), Updated Molecular Testing Guideline for the Selection of Lung Cancer Patients for Treatment with Targeted Tyrosine Kinase Inhibitors, from the CAP, International Association for the Study of Lung Cancer (IASLC), and the Association for Molecular Pathology (AMP) stated the following:
According to the American Cancer Society (ACS, 2019), there are approximately 1.8 million new cancer diagnoses and 600,000 cancer-related deaths. Cancer develops from genetic alterations in DNA that affect the way cells grow and divide. A tissue biopsy is the gold standard for detecting DNA alterations that can be used to identify cancer, determine treatment options, or evaluate responsiveness to treatment. Tissue biopsies have several disadvantages: the biopsy procedure may be painful, such as the insertion of a long needle or a surgical procedure; the retrieved tissue may be too small for analysis; or an individual may not be able to physically tolerate the procedure. In addition, because tissue biopsies only represent cellular samples from parts of a tumor, important diagnostic data could be missed (Weber, 2014).
Liquid biopsy is proposed as a less-invasive method for cancer identification, surveillance, and treatment guidance. The National Cancer Institute (NIH) defines liquid biopsy as “a test done on a sample of blood to look for cancer cells from a tumor that are circulating in the blood or for pieces of DNA from tumor cells that are in the blood.” ctDNA tests detect small fragments of mutated DNA that are released from tumors into blood, presumably by apoptosis and/or necrosis. Some ctDNA liquid biopsy tests are targeted for specific gene mutations. For example, in the instance of non-small cell lung cancer, a targeted liquid biopsy may be used to identify the presence of the EGFR mutation and determine if individuals may benefit from kinase inhibitor medication. Other liquid biopsy tests analyze multiple biomarkers and are purported to detect various cancers or treatments (Perakis, 2017).
There are several limitations of liquid biopsies. In regard to cancer management, many cancers do not have specific DNA mutations that can be identified and, when present, can be different in individuals with the same cancer. The DNA found in the fluid sample may not fully represent the tumor and mislead treatment decisions. The mutations found may not be “driver” mutations and may not provide useful information about the cancer. In regard to cancer detection, liquid biopsies can test positive for cancer when no cancer is present (false-positive) or test negative when cancer is present (false-negative). Because cancer cells release more mutated DNA fragments in later cancer stages, the test may not identify early cancer. Likewise, a liquid biopsy can detect cancerous cells that may never actually cause harm, leading to overtreatment (NIH, 2018). While liquid biopsies are promising, a great deal of research is still needed to determine if these tests improve outcomes for individuals with cancer.
Liquid biopsies are regulated by the Clinical Laboratory Improvement Amendments (CLIA) program, which oversees and certifies the laboratories conducting FDA-approved and non-FDA approved tests. In vitro diagnostic liquid biopsies, tests that are manufactured and then commercially sold to multiple labs, are also regulated by the FDA and must meet premarket review requirements. Liquid biopsies that are considered laboratory determined tests (LDTs), tests manufactured and performed in the same CLIA laboratory, have not thought to be subject to FDA regulations. However, in 1976, the FDA was given authority to regulate all in vitro diagnostics as devices. Because LDTs were not complex during that time, the FDA did not enforce premarket review and other requirements. As LDTs have grown more complex, the FDA has taken a renewed interest in overseeing LDTs to ensure public health and safety. On January 13, 2017, the FDA issued a discussion paper on LDTs but has not released an enforceable position at this time (FDA, 2018).
In 2012, the Institute of Medicine released recommendations on genomic-based test development and evaluation. They stated that genomic tests should be designed and properly validated in a CLIA-certified lab, and intended use of the tests, including LDTs, should be discussed with the FDA before validation studies begin. In addition, the authors stated the following:
Of note, FDA review of a biomarker test has been focused principally on analytical and clinical/biological validity, but not on demonstration of clinical utility…Therefore, FDA approval or clearance does not necessarily imply that the test improves clinical outcomes or should be used for patient management. LDTs performed in CLIA-certified laboratories also do not require evidence of clinical utility; only analytical and clinical validity of the test must be demonstrated prior to clinical use.
On March 16, 2018, the Centers for Medicare and Medicaid Services (CMS) approved NGS-based in vitro companion diagnostic laboratory tests for national coverage after an FDA-CMS parallel review. In the decision memo they state that “at this time, liquid-based multi-gene sequencing panel tests are left to contractor discretion if certain patient criteria are met.”
In August, 2020, the FDA approved the Guardant360 test as a companion diagnostic for individuals with NSCLC considering treatment with osimertinib. The approved indication involves testing for EGFR exon 19 deletions, L858R, and T790M. Also in August, 2020, the FDA approved the FoundationOne Liquid Cdx test as a companion diagnostic for individuals with NSCLC considering treatment with osimertinib, gefitinib or erlotinib (EGRF exon 19 deletions and EGFR exoin 21 L858R alteration) and for individuals with prostate cancer considering treatment with rucaparib (BRCA1 and BRCA2 alterations).
Cell-free DNA (cfDNA): DNA that is circulating freely in body fluids, such as blood plasma, and is released from all types of cells.
Circulating tumor DNA (ctDNA): Fragments of DNA that are released from a tumor and migrate into bodily fluids, such as blood plasma.
The following codes for treatments and procedures applicable to this document are included below for informational purposes. Inclusion or exclusion of a procedure, diagnosis or device code(s) does not constitute or imply member coverage or provider reimbursement policy. Please refer to the member's contract benefits in effect at the time of service to determine coverage or non-coverage of these services as it applies to an individual member.
When services are Investigational and Not Medically Necessary:
For the following procedure codes or when the code describes a procedure indicated in the Position Statement section as investigational and not medically necessary.
Unlisted molecular pathology procedure [when specified as a liquid biopsy using plasma specimen]
Oncology (non-small cell lung cancer), cell-free DNA, targeted sequence analysis of 23 genes (single nucleotide variations, insertions and deletions, fusions without prior knowledge of partner/breakpoint, copy number variations), with report of significant mutation(s)
BCAT1 (Branched chain amino acid transaminase 1) or IKZF1 (IKAROS family zinc finger 1) (eg, colorectal cancer) promoter methylation analysis
Targeted genomic sequence analysis panel, solid organ neoplasm, cell-free DNA, analysis of 311 or more genes, interrogation for sequence variants, including substitutions, insertions, deletions, select rearrangements, and copy number variations
Peer Reviewed Publications:
Government Agency, Medical Society, and Other Authoritative Publications:
|Websites for Additional Information|
Cell-Free DNA (cfDNA)
Circulating Tumor DNA (ctDNA)
FoundationOne Liquid CDx
OncoBEAM Colorectal Panel
OncoBEAM Lung2 Panel
OncoBEAM Melanoma Panel
The use of specific product names is illustrative only. It is not intended to be a recommendation of one product over another, and is not intended to represent a complete listing of all products available.
Medical Policy & Technology Assessment Committee (MPTAC) review. Rationale, Background/Overview and References sections updated. Updated Coding section with 01/01/2021 CPT changes; added 0229U, 0239U.
Updated Coding section with 07/01/2020 CPT changes; added 0179U.
Updated related topics in Description/Scope section.
MPTAC review. Description/Scope, Rationale, Background, References, Websites and Index sections updated.
MPTAC review. Description/Scope, Rationale, Background, References and Websites sections updated.
Hematology/Oncology Subcommittee review. Initial document development.
Applicable to Commercial HMO members in California: When a medical policy states a procedure or treatment is investigational, PMGs should not approve or deny the request. Instead, please fax the request to Anthem Blue Cross Grievance and Appeals at fax # 818-234-2767 or 818-234-3824. For questions, call G&A at 1-800-365-0609 and ask to speak with the Investigational Review Nurse.
Federal and State law, as well as contract language, including definitions and specific contract provisions/exclusions, take precedence over Medical Policy and must be considered first in determining eligibility for coverage. The member’s contract benefits in effect on the date that services are rendered must be used. Medical Policy, which addresses medical efficacy, should be considered before utilizing medical opinion in adjudication. Medical technology is constantly evolving, and we reserve the right to review and update Medical Policy periodically.
No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, photocopying, or otherwise, without permission from the health plan.
© CPT Only – American Medical Association