FDA issues guidance on early-stage solid tumor drug development with ctDNA

Original link: https://kaopubear.top/blog/2022-05-16-fda-ctDNA-for-early-stage-solid-tumor-drug-development/

Earlier this month, the FDA issued a draft guidance for the use of ctDNA in early-stage solid tumor drug development .

For the relevant companies, the precautions in the opinion are indeed “instructive”. For relevant practitioners, it is equivalent to getting a reference answer to “ctDNA application in early solid tumor clinical research”.

The entire guidance reflects the FDA’s current view of the use of ctDNA as a biomarker for drug development and clinical trial design in early-stage clinical trials in solid tumors, and this section does not address the use of ctDNA for early-stage tumor detection screening or use in metastatic tumors .

The following is a non-complete translation of the meaning of the guidance.

background

Drug development in early-stage non-metastatic solid tumors often involves time-event endpoints in large trials and years of follow-up.

Some patients with early-stage solid tumors can be cured with only local therapy (eg, surgery, radiotherapy, or chemoradiation), some patients require (neo)adjuvant therapy, and some patients remain untreated despite surgery and/or systemic therapy. May develop metastatic disease.

ctDNA, as fragmented DNA generated by tumor drivers, is shed into the blood of patients, and the amount depends on tumor type, location, stage, tumor burden, and treatment response. As a biomarker, ctDNA has many potential clinical uses at an early stage that can aid and accelerate drug development.

In the early-stage tumor setting, ctDNA can be used to detect certain targeted alterations, enrich for high- and low-risk populations in screening assays, reflect patient response to treatment, or potentially serve as an early marker of efficacy.

Evidence supporting the clinical effectiveness or clinical utility of ctDNA varies across solid tumors, patient populations, and testing modalities. However, several small studies have shown that detection of molecular residual disease (MRD) in ctDNA after surgery or standard systemic therapy leads to poor prognosis, screening high-risk patients for recurrence.

Because ctDNA assessments vary between laboratories and the detection techniques used, sometimes results are inconsistent. The protocols developed by many clinical laboratories may affect the quantitative detection of ctDNA. Further standardization of detection methods will enable better use of ctDNA in regulatory settings.

Application direction

Sponsors should consult with the FDA if they plan to use ctDNA for patient screening or as an endpoint in an early-stage solid tumor clinical trial. The following are potential uses of ctDNA.

ctDNA for molecular alteration-based patient screening

In adjuvant therapy, patients typically receive curative local therapy followed by systemic therapy to prevent recurrence. In this case, sampling the patient’s plasma could detect ctDNA and potentially select patient populations with genetic or epigenetic alterations that could be specific drug targets under study.

  • ctDNA can be used as a clinical trial standard for patient selection.
  • ctDNA can also be used as a stratification factor if the trial enrolls both marker-positive and negative populations.
  • The sensitivity of ctDNA to detect all mutations of clinical interest contained within tumor tissue should be assessed. If no mutation is detected in the ctDNA, tumor testing may be required to confirm a negative result.

ctDNA MRD was used for patient enrichment.

Following surgery and/or (neo)adjuvant therapy, ctDNA can be used as a marker of MRD, thereby enriching in clinical trials patients at high risk and increased events of relapse or death.

  • ctDNA testing after surgery or (neo)adjuvant therapy can identify biomarker-positive populations.
  • Baseline ctDNA status can also be used as a stratification factor in studies that include both ctDNA-negative and positive patients.
  • Sequential tests can be used to test the intention-to-treat population (both ctDNA positive and negative), as well as the ctDNA positive group.
  • Design options can include an upgraded experimental design (high-risk patients with ctDNA-positive status) or a de-escalated experimental design based on a low-risk population with ctDNA-negative status, and clinical trials should be randomized.
  • The primary endpoint should be disease-free survival (DFS) if only adjuvant therapy is given, such as event-free survival (EFS) with neoadjuvant therapy (with or without adjuvant therapy), or overall survival (OS).
  • Due to limited events, an early interim analysis of the primary endpoint should not be performed. Later interim analyses can be considered, but these should be pre-specified before the start of the trial, adjusted for multiple testing and set up with a node with sufficient data maturity. For example, most patients would be expected to have completed treatment before any interim analysis was performed.

ctDNA as a measure of response

  • ctDNA can be used in early clinical trials to help uncover signals of drug activity, which may aid in drug development programs.
  • The FDA encourages the development of evidence on the usefulness of ctDNA response after neoadjuvant therapy, demonstrating the use of ctDNA beyond pathological complete response after neoadjuvant therapy.

ctDNA as an early endpoint in clinical trials

Although not currently validated for use, it is possible that changes in ctDNA response to drugs could be used as an early endpoint to support drug approval in early-stage tumors.

  • Further data are needed to support the use of ctDNA as an endpoint with potential to predict long-term outcomes (DFS/EFS/OS).
  • Trials that collect ctDNA data before and after drug treatment should also collect long-term outcome data to determine the association between ctDNA negativity and outcome.
  • Various statistical criteria for validation endpoints have been proposed, and meta-analysis methods are often used. Meta-analyses to validate ctDNA at the trial-association level should include only randomized trials. Sponsors should discuss and provide FDA with details of a meta-analysis program to validate the use of ctDNA in specific settings.
    • The plan should include details of trial design, patient inclusion and exclusion criteria, and ctDNA assessment methods.
    • Clinical trials should include a representative population of patients for which the experimental endpoint will be used.
    • A sufficient number of randomized trials with adequate follow-up should be included.
    • Analyses based on individual patient-level data should allow for accurate assessment of individual-level associations.
    • Test-level and individual-level association criteria should be provided, including pre-determined time points for ctDNA assessment.
    • Long-term clinical endpoints such as EFS, DFS, and OS should be included, with clear and consistent definitions across studies.
    • Sponsors should explore the impact of missing data on trial results.

Precautions

Panel type detected by MRD

MRD can utilize tumor-informed methods, tumor-naive methods, or fewer candidate gene panels, each with its own advantages and limitations.

  • Tumor-informed panels select a set of mutations for detection by sequencing the tumor.
    • Limitations of this approach include lag time between tumor detection and ctDNA panel construction, and sensitivity and specificity may depend on clinical time points, product assay sensitivity, and the number of mutations detected.
  • Tumor-naive or tumor-agnostic panels refer to those panels that have not sequenced the primary tumor.
    • Limitations include that some tumor markers are not covered by the ctDNA panel, and additional characterization of the cohort is required to understand what proportion of patients can be followed with this technique.
    • It is possible that whole-genome sequencing (WGS), which can use other biomarkers in addition to mutations, epigenetic changes (such as methylation), or fragmentation analysis of ctDNA, to capture tumor-driven ctDNA signaling.

Sampling Considerations

Several sampling factors related to clinical trial design and intended patient populations should be considered.

  • ctDNA shedding is influenced by tumor histology, stage grade, and size, so timing of ctDNA testing should be discussed with FDA and should be based on the performance characteristics of the test, disease characteristics, and tumor biology.
  • A fixed time point should be pre-designated to join the study.
  • If sponsors wish to use multiple ctDNA time points to determine inclusion criteria (eg to assess whether interventions in early detection of relapse would affect outcomes) there should be scientific data/justification to support this. Sensitivity analyses based on different time points can be explored (but should be identified and discussed in advance).
  • The time of ctDNA detection should be the same for each group.
  • Pre-treatment baseline samples should be collected to account for the impact of changes in tumor shedding rates on assay performance. In addition, this sample will allow interpretation of the post-treatment sample to facilitate study enrollment.
  • All sites in the study should follow standardized sample collection, storage, handling, and handling protocols.

Analytical Verification Considerations for Market Application Detection

Validation studies should use specified technical protocols for the sensitivity, specificity, accuracy, precision, and other relevant performance characteristics of the assay, which may involve specimen collection, processing, and storage.

  • MRD test validation should include the entire test system, from sample collection (eg, collection of blood with a specific test collection tube that will be used for eventual marketing) to test result output, including thresholds for determining positive versus negative patients.
  • Assay cutoff points should be determined to optimize assay sensitivity and specificity for clinical use. MRD positivity should be accurately and reproducibly detected when analyzing performance.
  • Tests should have high sensitivity and negative predictive value (NPV) to support treatment de-escalation; high specificity and positive predictive value (PPV) to support treatment escalation.
  • The method of validation of the MRD test will depend on the type of MRD test.
  • It is recommended to use samples from clinical trials (clinical specimens) for critical assay validation studies, such as confirming the limit of detection (LOD), assay precision, and analytical precision. In some analytical validation studies, clinical samples can be supplemented with artificial samples due to the large number of samples required. When artificial samples are used in validation studies, functional equivalence between artificial and clinical samples should be demonstrated, and justification should be provided if artificial samples are used to replace or supplement clinical samples in some studies.
  • For immobilized panels, cell lines carrying specific mutations can be used as artificial samples. For personalized panels, cell lines that represent the distribution of mutation numbers and types should be developed based on data from earlier clinical studies.
  • Samples within the detection range should be used to demonstrate detection accuracy.
  • An appropriate set of reference materials should be developed to allow comparability across multiple MRD tests.

This article is reprinted from: https://kaopubear.top/blog/2022-05-16-fda-ctDNA-for-early-stage-solid-tumor-drug-development/
This site is for inclusion only, and the copyright belongs to the original author.

Leave a Comment