Development, Validation, and Regulatory Considerations for a Liquid Biopsy Test

Dana W.Y. Tsui, Gideon M. Blumenthal, Reena Philip, Jane Barrett, Clara Montagut, Kelli Bramlett et al. · 2020 · 15 citationsRead the paper

Genome-guided precision medicine has become an essential part of cancer management. However, tumor biopsies are not always accessible and often fail to capture complex genetic heterogeneity especially in the relapse setting. Liquid biopsy tests offer opportunities to explore the underlying evolving tumor in a safe, noninvasive way. Liquid biopsy tests can be broadly categorized in to 2 types: circulating tumor cells (CTC) and cell-free tumor nucleic acids, mostly circulating tumor DNA (ctDNA). Both CTC and ctDNA are important biomarkers to predict prognosis and to screen for somatic alterations that can inform cancer management. In this article, we will focus on ctDNA liquid biopsy tests. Early implementation of ctDNA tests started in cancer types with well characterized oncogene drivers such as KRAS mutation in colorectal cancers and EGFR mutations in nonsmall cell lung cancer (NSCLC). Since 2014, multiple large-scale clinical trials have demonstrated plasma-based EGFR testing predict similar outcomes as tumor-based testing in NSCLC, which led to FDA-approval of the first liquid biopsy test as a companion diagnostic for EGFR-targeted therapy in 2016. This marked the beginning of a new era of noninvasive genomic medicine. Despite this exciting development, concerns arose when researchers showed discrepancies between the results of different commercially available plasma DNA tests performed on the same samples, highlighting the importance of strict regulation and thorough analytical validation of individual liquid biopsy tests. In this article, we invited experts from diverse areas, including regulatory, pharmaceutical, and academic and commercial molecular diagnostic test providers, to discuss the different aspects of development, validation, and regulatory considerations for ctDNA liquid biopsy tests. Soma Ghosh: The US Food and Drug Administration has so far approved 2 liquid biopsy ctDNA-based companion diagnostic (CDx) tests, meaning they are essential for the safe and effective use of the corresponding therapeutic: the Roche's Cobas EGFR Mutation Test v.2 that detects EGFR mutations associated with patients with nonsmall cell lung cancer [see Summary of Safety and Effectiveness Data (SSED) for P150047 and P150044 for study designs and validation details] and the QIAGEN therascreen® PIK3CA RGQ PCR Kit to detect PIK3CA mutations associated with patients with advanced breast cancer (SSED P190004). We expect that the number of liquid biopsy test approvals will increase significantly in the coming years as a growing demand for these noninvasive testing modalities appear to revolutionize the management of cancer patients. The analytical and clinical validation studies in support of liquid biopsy tests depend on the intended use (IU) and Indications for Use of the test. For a CDx test, the clinical validity demonstrates that the CDx identifies the correct patient population for the therapeutic. For both the liquid biopsy CDx approvals (Roche's Cobas EGFR Mutation Test v2 and the QIAGEN therascreen PIK3CA RGQ PCR Kit), clinical evidence was presented from results of a retrospectively designed clinical bridging study using clinical trial plasma samples that were collected during patient enrollment for the pivotal clinical trial of the therapeutic. The clinical bridging studies demonstrated that the drug efficacy was preserved in patients identified using the plasma samples relative to the efficacy based on the tissue-based enrollment test. J. Carl Barrett: There are multiple uses of ctDNA in drug development including patient selection, monitoring of drug response, detection of minimal residual disease (MRD) and discovering resistance mechanisms to therapies. Of these, only patient selection involving a CDx is currently covered by regulatory practices, although guidance for the MRD and monitoring applications are provided. Regulatory standards for ctDNA-based CDx are consistent with other CDx assays. Only a few ctDNA assays are approved but this is expected to rapidly change in the near future. A major challenge for ctDNA assays is the limited sample volumes available for extensive analytical validation and for follow-on diagnostic development. Applying a follow-on diagnostics approach recently used for approval of a tissue-based next-generation sequencing CDx using appropriate commercial and contrived samples will enable increased access to high quality liquid biopsy tests for patients when limited clinical trial samples are available. Kelli Bramlett: As a kit developer, we are thinking a lot about the appropriate analytical validation studies for different indications such as tumor profiling or a companion diagnostic for our liquid biopsy technologies. These types of studies are particularly challenging for a liquid biopsy assay when getting sufficient quantity for repeat testing of clinical samples harboring a known targeted mutation is difficult if not impossible. We are working closely with the United States Food and Drug Administration (FDA), consortia, and colleagues to understand and define best practices for developing and commercializing regulated liquid biopsy tests. As members of the Blood Profiling Atlas in Cancer (BloodPAC) consortia, we have participated in drafting an FDA presubmission document outlining analytical validation tests for next-generation sequencing (NGS) based assays interrogating cell-free nucleic acid. The goal of this work is to work together with the regulatory agencies to think through the unique challenges of developing regulated liquid biopsy tests and agree on some clear recommendations that everyone can use as a starting point. Clara Montagut: The field of liquid biopsy is so new that regulatory requirements for clearance are unclear in the European Union (EU). Currently, the CE marking of all oncology tests including liquid biopsy tests follows the requirements of the EU in-vitro diagnostics (IVD) Directive (Directive 98/79/EC): The tests are classified as General IVDs and do not require the involvement of Notified Body (no CE certificate issued). When the manufacturer has the required documentation completed, the Competent Authority in the country of legal manufacturer is notified of CE marking. That means that the product can be placed on the EU, Norwegian and Swiss market. Importantly, the CE marking requirements shall change under the new IVD Regulation. The oncology tests will be classified as Class C products and will require review of technical documentation by a Notified Body. Only after the Notified Body issues the CE certificate, the product will be considered CE marked and will be allowed to be placed on the EU market. J. Carl Barrett: Tissue to plasma concordance can vary due to a number of factors including time of sample acquisition, tumor heterogeneity, clonal hematopoiesis of indeterminant potential and sequencing, and other technical errors. Nonetheless, tumor to plasma orthogonal assays are important. Orthogonal plasma tests to compare sensitivity and specificity of multiple ctDNA assays can help discern discordant calls and their origin. Standards are useful for establishing limits of detection but standardization of the use of standards is needed (and ongoing by the Foundation of National Institute of Health, FNIH). Standards may not establish false positives, however. Kelli Bramlett: There are many important considerations for an appropriate orthogonal assay. Most importantly is to identify a test covering genes and variants of interest that has been rigorously validated to an appropriate limit of detection (LOD). Historically, digital PCR (dPCR) methods have commonly been used to orthogonally validated clinically relevant mutations at limits of detection below 1% for a liquid biopsy application. It is now becoming possible to consider a broader NGS-based panel or assay as an orthogonal method for liquid biopsy assays. The key challenge is to make sure the NGS method is validated down to the required limit of detection to serve as a fitting orthogonal test. It is clear that performance of liquid biopsy controls or reference standards vary depending on how they were generated and manufactured and what type of technology is being utilized in the liquid biopsy assay. It is necessary to evaluate multiple types of control material and compare performance metrics to actual cell-free nucleic acid extracted from plasma to determine what control material best mimics actual human cell-free nucleic acid in a specific liquid biopsy test. It has been broadly observed that controls that perform well in dPCR methods generally work equally well for amplicon-based NGS assays, while different control materials may better mimic human cell-free nucleic acid in hybridization and/or capture-based NGS assays. Soma Ghosh: Preanalytical variables have to be adequately understood and appropriately standardized for sample integrity and success of liquid biopsy tests. Some key preanalytical factors that influence the quality and quantity of ctDNA for accurate and reliable results of liquid biopsy tests include patient characteristics (e.g., stage of tumor), specimen collection device (blood collection tube), the anticoagulant used, specimen volume, specimen stability (blood, plasma, extracted ctDNA), and specifics of specimen processing methodology such as extraction process, preparation, time between collection and testing, and other additional factors. Standardization of these variables and practices would be important in ensuring reliable and accurate results from liquid biopsy tests. J. Carl Barrett: Preanalytical variables are poorly understood and include both technical and biological variables. Collection tubes, time and method of processing, freezing and thawing, DNA extraction and library construction are some of the variables that have been examined. A comprehensive listing and reporting methodology is needed and the BloodPAC has prepared a guide to analytical variables. These variables may affect qualitative (is a variant present/absent) and quantitative measures of ctDNA (how much of the variant is detected). Biological variability in quantitative measures of ctDNA are significant and depending on the intended use may require multiple assessment of blood taken at different times. Additionally, qualitative assessments are less likely influenced by analytical variables than quantitative assessments. For example, collection tubes may not alter the outcome of the presence or absence of a marker. Clara Montagut: Preanalytic variables may affect analytic validity, clinical utility, and validity of ctDNA. Several challenges limit optimal collection of ctDNA, including the small proportion of ctDNA compared to total cell-free DNA, a potential contamination of sample by DNA released by blood cell lysis and the labile nature of the DNA molecule with implications for storage and transport. Recommendations include the use of a large gauge diameter needle for blood extraction, and the use of K2EDTA or cell stabilizer tubes (e.g., Streck cell-free DNA tubes) that should be stored at 4°C for the former and 10–30°C for the latter. Plasma isolation should be done preferably no later than 4–6 hours after blood drawn with K2EDTA or within 2–7 days if using cell stabilizer tubes, by sequential centrifugation at low and high speed (800–1600 g) at 4°C. Blood cells obtained from the buffy coat should be stored and may be used as a source for assessment of germline and/or clonal hematopoiesis of indeterminate potential (CHIP) mutations. Aliquoting plasma into single use small volume tubes (1–2 mL) avoids multiple freeze-thaw cycles of plasma samples. Kelli Bramlett: Work is ongoing to evaluate preanalytical variables on the quality of liquid biopsy test results and generate clear recommendations for key variables. Areas of known impact include the type of blood collection tube used, time from blood draw to initial processing step/s to separate plasma from whole blood. Another key factor is what type of cell-free DNA, cell-free RNA, or cell-free total nucleic acid extraction method is used and required for a particular liquid biopsy assay. A recent review from the American Society of Clinical Oncology and the College of American Pathologists (CAP) has offered some clear recommendations around many of these parameters including: (a) EDTA tubes to be processed within 6 hours of blood draw; (b) cell-free stabilizing blood tube should be processed within 48 hours of blood draw; (c) sequential 2-spin method is recommended with initial low speed spin followed by higher speed spin; and (d) minimizing freeze-thaw of stored extracted cell-free DNA by aliquoting prior to storage. Additional areas that are currently being investigated include factors such as time of day for the blood draw and method of blood draw, method of extraction, and how that is closely influenced by testing technology. The liquid biopsy field is still in the initial stages of understanding the most relevant preanalytical variables and being able to provide broad guidance on best practices. J. Carl Barrett: The reporting of results should indicate whether the alteration is likely germline or somatic depending on the allele fraction. Clonal hematopoietic (CH) alterations should be noted if predicted from known CH alterations or ideally from sequencing normal blood cells. Variant classification based on established standards such as the American College of Medical Genetics and Genomics Standards and guidelines for the interpretation of sequence variants (i.e., known or likely deleterious variants of the gene) should be provided in the report. The allele fraction of the observed variant and the most abundant variant should be reported. Calculations of clonal or subclonal variants would be very beneficial if possible. The number of unique reads of the variant in the assay is also important to report. Marc Ladanyi: Using white blood cell (“buffy coat”) DNA as a matched normal for ctDNA, as we do at Memorial Sloan Kettering (MSK) in the context of our MSK-ACCESS cell-free DNA panel assay, can be invaluable in clarifying private germline variants (still not uncommon in spite of expanding germline databases) and CH mutations among the variants called in the cell-free DNA. Kelli Bramlett: The need to rule out CH alterations and germline variants on liquid biopsy variant reporting may depending on the technology and test method For a very specific targeted test, these types of may not be There are methods that can be used to the impact of these types of variant on a liquid biopsy sample and and specific variant include of parameters to germline variants and of a germline control may be during initial development and testing to understand the potential impact of clonal hematopoiesis on results of a specific liquid biopsy test The germline control can be used to inform of for Clara Montagut: ctDNA results on the of germline mutations in a liquid biopsy test is a which considerations as well as clinical implications for the patients and the between the test and the to the is key and will help the to the to patients. Soma Ghosh: results may be associated with low of tumor DNA. In if a sample with fraction or DNA than the a the may not be reliable and may in a false test if a liquid biopsy test has a low when compared to a biopsy test, a from a liquid biopsy test may not be reliable and can testing using a patient sample by a biopsy test. to this both FDA approved liquid biopsy CDx tests, Roche's Cobas EGFR Mutation Test v.2 and the QIAGEN therascreen PIK3CA RGQ PCR have a testing by a tissue-based test when the test is Marc Ladanyi: define a plasma as a are prior tumor or plasma results available that document mutations that should be by the cell-free DNA panel assay a In that results are when the testing is performed prior to initial biopsy or Kelli Bramlett: results is always a of whether appropriate was used and what or of tumor cell-free nucleic acid was in the cell-free nucleic acid A is that tumor cell-free nucleic acid may only of the total cell-free nucleic acid that is used as into a liquid biopsy test for of somatic Some NGS-based liquid biopsy assays include in the variant sufficient material was was used to evaluate variants at a specific allele This may help evaluate a As liquid biopsy methods become established with appropriate clinical validation, an additional sample from the same individual may be used to cell-free somatic variant calls in both and Clara Montagut: results (i.e., the test is not specific mutations in the plasma specimen of the patient although the tumor specific may in clinical management and are of for the use of ctDNA in clinical and interpretation are needed to understand the of false results can be due to technical and biological including ctDNA from the tumor and of the specific Additional of the and DNA such as the presence of DNA extracted from the plasma specimen or the presence of other mutations or in the plasma specimen may increase the that a patient has a type tumor when no mutation is in the ctDNA. When plasma is sequencing of the may help false and However, this is associated with additional and associated with sequencing of the tumor of when of in a cancer sequencing of an initial biopsy that was performed in a tumor may not capture of associated of mutations and may not capture an of the molecular J. Carl Barrett: allele calls can be due to low of to assay sensitivity and to subclonal in a As variants should be identified and if based on normal blood The clinical of low clonal variants is variants should be considered in the context of the cancer that are not effective on subclonal variants should be with as they may for the tumor cells or of a the subclonal variant is a known mutation to the variants should be time and Clinical trials based on low allele are Marc Ladanyi: Variant need to be rigorously validated prior to clinical implementation of the assay. variant and reporting likely for some of the of concordance between different commercial assays. Kelli Bramlett: Clinical validity has been for resistance for targeted in lung such as EGFR when at allele Clinical validity of resistance may be to establish than other targeted Additional efficacy of testing in studies cancer using liquid biopsy methods as compared to on the cancer type and detection of cancer from to liquid biopsy tests than detection testing is on these types of in studies and establishing efficacy in clinical trial Soma Ghosh: patients based on and liquid biopsy modalities will likely be needed to understand if liquid biopsy tests are to guide patient in cancer will be important for the clinical including the between the and plasma-based testing modalities for a specific to be and the efficacy based on the plasma test to be clinically significant and to that of the test. J. Carl Barrett: trials will likely include patients on ctDNA as the to population approval of tests. The of trials with known therapy with from CDx to the use of plasma tests for would from FDA studies establishing the and challenges of ctDNA to identify patients that from approved will to better in testing patients with ctDNA first due to the and the biological of multiple tumor at Marc Ladanyi: a liquid biopsy the necessary to guide first by EGFR mutation in a patient with lung and advanced stage disease by as a test. It also important opportunities for targeted such as Kelli Bramlett: Liquid biopsy testing for oncology variants in stage is the most and best understood use indicate the use of liquid biopsy tests for initial in when no tumor sample is available. The FDA has approved the use of liquid biopsy testing to detect EGFR mutations in with the test. The use of liquid biopsy as test will increase as additional become available clinical validity in additional tumor types and intended The type and stage of cancer will influence the use of liquid biopsy tests as have a of cell-free DNA and detection that vary with tumor type and cancer stage or tumor Soma Ghosh: noninvasive liquid biopsy assays have a significant in cancer compared to biopsy tests, are some challenges that need to be to support such as cancer detection and and false results to limit the to detect alterations and false can be by establishing assay that are not very low as compared with the In studies have that clonal hematopoiesis of indeterminate or may to the detection of mutations in and in a high false in liquid biopsy tests. in to reliable and accurate ctDNA-based liquid biopsy test should consider sequencing white blood followed by and with to rule out false due to accurate and reliable liquid biopsy tests and in the biopsy are developing reference materials and test validation these liquid biopsy tests will be used for clinical applications such as monitoring of residual and to in to being used for and patient management. Marc Ladanyi: Liquid biopsy testing will be an especially important for to This testing should not only evaluate detection of provided by but also should require reporting to of all called variants so that that are variants can some that would help their J. Carl Barrett: This is rapidly developing and can be in drug development and for patient only patient selection, but also monitoring of to of tumor and minimal residual disease after are new areas of drug development that ctDNA can the of number and from plasma ctDNA. in this will this and the use of ctDNA for patients with these genomic (e.g., and and Clara Montagut: the interpretation of liquid biopsy tests is will be to clinical reporting and effective between test and to help the results in the specific context of the disease and the relevant into clinical Kelli Bramlett: The potential of liquid biopsy tests to in cancer and guide is recent in technology methods to the impact of false on liquid biopsy and as well as clinical validity and tests, we can liquid biopsies will become a method available to when for oncology patients. The and of significant clinical will this of samples to the required testing is a significant as we work to establish broader clinical efficacy of liquid biopsy As liquid biopsy testing is established as a key to with higher associated and we will need to focus on from to the potential impact of liquid biopsy testing on the in the United oncogene factor human factor they have to the of this and have the (a) significant to the and of or and interpretation of (b) drafting or the for (c) approval of the and (d) to be for all aspects of the ensuring that to the or integrity of part of the are appropriately investigated and all the and/or potential of Cancer for cancer US

1 idea Seedlabs derived from this research