As part of the meta-analysis published in the Journal of Liquid Biopsy, the authors performed individual-level (I-) and trial-level (T-) associations to assess associations between changes in ctDNA and long-term outcomes, the gold-standard approach for demonstrating biomarker surrogacy. I-associations showed robust associations between decreased ctDNA and improved outcomes across trials; however, T-associations were weaker. These results underscore the need for additional, prospective research to confirm the strength of trial-level associations.

“We performed the meta-analysis recognizing the data were not fit-for-purpose but knowing the findings would provide critical insights for prospectively designed meta-analyses,” said Hillary Andrews, Senior Director of Science Policy & Strategy at Friends and the ctMoniTR project manager. “Our findings raise important questions about how to handle crossover and homogeneity of RCTs for future T-association analyses.”

The landscape assessment published in JCO OA, synthesized data from 162 advanced solid tumor studies evaluating on-treatment reductions in ctDNA levels within the first 15 weeks of therapy, and associations with long-term outcomes. Of the 76 studies that assessed the association between ctDNA dynamics and OS, all but one reported a significant association between a reduction in ctDNA levels and improved OS. These findings were consistent across multiple cancer types, treatment modalities, ctDNA assays, and molecular response (MR) definitions.

“The landscape assessment shows that many clinical trials already incorporate ctDNA dynamics into their study designs and, importantly, that on-treatment ctDNA reductions consistently associate with improved overall survival for patients with advanced cancer,” said Elena Levi-D’Ancona, Science Policy Analyst at Friends and lead author for the landscape assessment. “To advance changes in ctDNA levels toward regulatory qualification as an early endpoint, the field requires a more standardized approach to data generation.”

These publications highlight the robustness of ctDNA dynamics as a predictive biomarker across solid tumors and treatment modalities and support the potential use of on-treatment ctDNA reductions as an early endpoint in clinical trials. In addition to continuing to build evidence to support ctDNA as an early endpoint, outstanding questions, including variability in sampling schedules, definitions of MR, and assay methods are the focus of the next steps of the ctMoniTR Project.

To learn more about the ctMoniTR Project, please visit: https://friendsofcancerresearch.org/ctdna/.