Trials should be clinical decision-driven not biomarker-driven, with appropriate biomarkers significantly adding to the prognostic and/or predictive information already obtained via validated methods. In addition, physicians need to have the appropriate tools to address this elevated risk.
Trials need to be performed in disease settings with tools and treatments that are known to be effective in order to appropriately assess the value of CTC/cfDNA to treatment efficacy and disease outcome.
When designing trials, investigations must consider the impact of treatment itself on the biomarker of interest. It is likely that the biomarker load will significantly decrease following treatment, making assessment of its true value difficult or even impossible.
Special consideration must be placed on designing, implementing and validating standard operating procedures (SOPs) for the collection and analysis of samples.
Appropriate selection of the timing of sample collection is critical, and should be based on the specific biology of each disease (e.g., baseline, throughout treatment, following treatment completion, and follow-up samples).
Must determine if characterization is necessary or if enumeration will suffice. If characterization is deemed necessary, one must then decide at what level the collected sample will need to be assessed (DNA, RNA, protein, functional assays) to properly answer the posed question(s).
Technology selection is important, especially with regards to previous trial data, and widespread feasibility based on overall cost must be considered.
Appropriate statistical evaluation of the number of patients required to answer posed questions. Typically, larger number of patients will be required than standard clinical trials due to the rare nature of CTCs/cfDNA and overall disease heterogeneity.
Most importantly, how results will be analyzed and interpreted, and if the obtained data can be compared head-to-head with previously performed or ongoing clinical trials.
