Circulating tumor DNA (ctDNA) can be used as a biomarker in early-stage solid tumor clinical trials for purposes of patient selection, patient enrichment and as a measure of response. However, further data are needed to support use as an endpoint reasonably likely to predict long-term outcome, the US Food and Drug Administration said.
In a new draft guidance, the agency provides recommendations on uses of ctDNA as a biomarker in the early-stage setting, where the intent is curative. The guidance does not address the use of ctDNA for early cancer detection or screening, or in the metastatic setting.
The document appears to mark the FDA’s first formal, public recommendations on use of ctDNA in drug development for solid tumors. A December 2020 final guidance discusses the use of minimal residual disease in hematological malignancy drug development. (Also see “Oncology Endpoint Guidance: US FDA Reaffirms Support For EFS, MRD In Update ” – Pink Sheet, 19 Dec, 2018.)
ctDNA has been the focus of a collaborative effort, led by Friends of Cancer Research, to collect data from multiple clinical trials to determine if the blood biomarker can accurately predict whether a drug is working.
Patient Selection …
ctDNA is tumor-derived fragmented DNA shed into a patient’s bloodstream that is not associated with cells. The quantity of ctDNA can vary by individual and depends on various factors, including the tumor type, location, stage, tumor burden and response to therapy.
“The evidence to support the clinical validity or clinical utility of ctDNA varies across solid tumor malignancies, patient populations, and testing modalities,” the guidance states. “However, multiple small studies have suggested that residual ctDNA detecting molecular residual disease (MRD) after surgery or completion of standard systemic therapy confers a poor prognosis and selects a population at high risk of relapse.”
ctDNA can be used for patient selection for clinical trial eligibility based on molecular alterations.
In the adjuvant setting, patients typically receive curative local therapy followed by systemic therapy to prevent recurrence. “In this situation, sampling a patient’s plasma can allow for detection of ctDNA and for potential selection of a patient population harboring genetic or epigenetic alterations that could be targetable by a given drug under study,” the guidance states.
In addition to patient selection, ctDNA can be used as a stratification factor if a trial enrolls both marker-positive and marker-negative patients, such that endpoints may be tested in both the intent-to-treat population and biomarker-selected (ctDNA-positive) subgroup.
… Enrichment …
Another use case for ctDNA is as a marker of MRD after definitive surgery and/or (neo)adjuvant therapy to enrich a trial for patients with higher risk disease and increased events of disease recurrence or death, the guidance states.
“Design options could include an escalation design of adding an experimental therapy to standard of care compared to standard of care alone for patients with ctDNA positive status (higher-risk) or a de-escalation dosing based on ctDNA-negative status (lower-risk population),” the guidance states.
Such a trial should be randomized. The primary endpoint should be disease-free survival if only adjuvant therapy is given, event-free survival if neoadjuvant therapy is given, or overall survival, the FDA said.
There should not be early interim analyses of the primary endpoint due to limited events. Later interim analyses may be considered but should be prespecified near the start of the trial and set for a time when data will be sufficiently mature. “For example, it would be expected that most patients should have completed treatment prior to any interim analyses being conducted,” the guidance states.
… And As A Measure Of Response …
The FDA also sees a potential role for ctDNA in early-phase trials to aid in signal finding for drug activity, which could help sponsors in their drug development plans.
“FDA encourages sponsors to develop evidence regarding the usefulness of ctDNA response in addition to or supporting pathologic complete response information after neoadjuvant therapy,” the guidance states.
Pathological complete response, which is the absence of residual, invasive cancer in tissue samples removed during surgery or biopsy after treatment, is an accepted surrogate endpoint to support accelerated approval in neoadjuvant treatment of high-risk, early stage breast cancer. However, pCR has proved more problematic in other disease settings, such as muscle-invasive bladder cancer. (Also see “Bladder Cancer: More Questions Than Answers With pCR Endpoint” – Pink Sheet, 19 Aug, 2019.)
… But Not Yet As An Early Endpoint
Although changes in ctDNA following drug treatment may have potential use as an early endpoint to support product approval in early-stage solid tumors, this use has not yet been validated, and further data are required to establish the biomarker as reasonably likely to predict benefit with regard to disease-free survival, event-free survival or overall survival, the guidance states.
If ctDNA data are collected in a trial before and after drug treatment, long-term outcomes data also should be collected to characterize the association between ctDNA clearance and outcome, the FDA said.
The agency also explains what it would need to see in terms of a meta-analysis aimed at validating ctDNA for use as an early endpoint.
“An appropriate meta-analysis to validate ctDNA at a trial level association should include only randomized trials,” the guidance states. “Sponsors should discuss and provide details of any proposed meta-analysis plan to validate use of ctDNA in a particular context of use with the FDA.”
This meta-analysis plan should include details about trial design, inclusion/exclusion criteria, ctDNA assessment methods and disease setting. It also should include an adequate number of randomized trials with sufficient follow-up time, and those trials should include patients who are representative of the population in which the endpoint will ultimately be used.
The trials should include long-term clinical endpoints that have been clearly and consistently defined across studies, and sponsors should explore the effects of missing data on study results.
Analysis based on individual patient-level data should allow for an assessment of individual-level association, and criteria should be prespecified for concluding association based on both trial-level and individual-level association measures.
The guidance recommends that sponsors consult the agency if they plan to incorporate ctDNA for patient selection or as an endpoint in early-stage solid tumor trials.
ctMoniTR Project
The agency’s view that ctDNA as an early efficacy endpoint is not yet ready for prime time aligns with the conclusions of a multistakeholder work group convened by Friends of Cancer Research in 2021. The group said that validating ctDNA as a predictor of longer-term outcomes would require large amounts of data from multiple retrospective and prospective trials, along with harmonization across data elements, assays and blood collection procedures. (Also see “Cancer Drug Endpoints: ctDNA Promising But Presents Many Challenges, Experts Say” – Pink Sheet, 1 Dec, 2021.)
FOCR is spearheading a research project, ctMoniTR, aimed at answering the question of whether changes in the biomarker reflect a response to treatment.
In step one of the project, clinical and academic leaders, government, industry and advocates aligned on a methodology to combine data from multiple prior clinical trials in lung cancer investigating immune checkpoint inhibitors as an initial case study. Results from this first part of the project showed that early changes in ctDNA following PD-1/L1 treatment in advanced lung cancer were a strong indicator of long-term outcomes across multiple clinical trials, FOCR said.
Step two of the project is aimed at collecting data on the use of ctDNA to monitor treatment response in more than 25 studies representing over 3,000 cancer patients, 16 additional treatments, and several cancer types. Preliminary data are expected in 2022.
A key distinction between the ctMoniTR project and the FDA guidance is that the former is focused on trials involving advanced disease, while the guidance addresses the biomarker’s use only in early-stage disease.
FOCR present and CEO Jeff Allen said ctMoniTR has been focused on the advanced disease setting because that’s where the data were available in terms of the biomarker’s incorporation into cancer drug studies. Nevertheless, results from this project could provide foundational evidence for use of ctDNA that could be applied to the earlier disease setting, where use of the biomarker is likely to be most beneficial, Allen said in an interview with the Pink Sheet.
FOCR also is leading a project looking at ctDNA in the early-disease space, trying to create a roadmap of the data needed to establish surrogacy. He noted that earlier-stage disease trials are now embedding ctDNA monitoring into their protocols.
Allen said that although the draft guidance did not contain anything that was particularly surprising, it is helpful to the drug development community for the FDA to spell out its view of the use cases and its data expectations, as well as identify the gaps in existing data.
He noted the document is not very definitive with regard to use of ctDNA changes as a surrogate endpoint, but “that’s reflective of the experience. There hasn’t been a lot of data that allows FDA to be definitive.” The hope is that data from the CTMoniTR project will help fill in some of those gaps, he said.
The Importance Of Timing
The draft guidance also discusses assay considerations, including types of MRD panels, sampling considerations and analytical validation.
The document notes that the ctDNA assessment can vary among labs and technologies used to detect the biomarker, which can result in discordant results. In addition, many clinical labs develop their own protocols that can impact ctDNA measurements and detection. “Further standardization of assays will allow for better use of ctDNA in a regulatory setting and will allow for analyses across studies to validate the use of ctDNA,” the guidance states.
In terms of sampling considerations, the timing of ctDNA testing should be discussed with the FDA and supported by performance characteristics of the test, disease characteristics, and tumor biology.
A set timepoint should be chosen for enrollment into the study and prespecified, and the timing of ctDNA collection should be the same across study arms. “A baseline pre-treatment sample should be collected to allow for consideration of the impact of variation in tumor shedding rates on assay performance,” the FDA said. “In addition, this sample will allow for interpretation of the post-treatment sample for study enrollment.”
If a sponsor seeks to use multiple ctDNA timepoints to determine patient eligibility, this should be supported by scientific data and rationale, and sensitivity analyses based on different time windows could be explored but should be predetermined and discussed in advance.
