Key takeaways:
- Nearly all studies show ctDNA reductions have an association with improved survival across solid cancer types.
- Standardization of ctDNA in clinical trials could provide more details on benefits.
A landscape assessment of more than 150 studies revealed reductions in circulating tumor DNA are associated with improved OS across solid cancer types, despite substantial heterogeneity in assays and collection and sampling timepoints.
“There’s a remarkably consistent signal across a diverse body of research,” Elena Levi-D’Ancona, PhD, science policy analyst at Friends of Cancer Research, told Healio. “There’s a strong signal here.”
‘Lack of harmonization’
Healio has reported on multiple studies in the past year detailing potential benefits of circulating tumor DNA (ctDNA) in several malignancies, including:
- melanoma, where ctDNA after immunotherapy could identify patients who have the highest risk for recurrence or death;
- non-small cell lung cancer, where ctDNA may help identify individuals with greater risk for worse survival outcomes; and
- muscle-invasive urothelial cancer, where ctDNA could help identify patients who would benefit from adjuvant immunotherapy.
Clinicians and researchers use ctDNA for a variety of reasons, including clinical trial participation, deciding if patients should continue or stop therapies, and treatment efficacy and response, according to study background.
Additionally, ctDNA has been evaluated as a surrogate endpoint for OS, but there is “substantial variation” in how it is used, Levi-D’Ancona said.
“Trials may be using different assays for measuring ctDNA, different assessment time points, or definitions of molecular response,” she explained. “This heterogeneity and lack of harmonization makes it more difficult to compare findings across studies, aggregate evidence, and understand which approaches may be most appropriate if ctDNA is ultimately going to support regulatory decision-making.”
Levi-D’Ancona and colleagues conducted a systematic landscape assessment to analyze ctDNA dynamics within 6 months of patients starting treatment for solid tumors.
The association of ctDNA with OS or PFS served as the primary endpoint.
‘Striking’ consistency
Overall, researchers included 162 studies (52.5% published between 2023 and 2025) in the assessment.
“I was really surprised simply by how much research is already being done in this space,” Levi-D’Ancona said. “We started this landscape assessment close to a year ago, and I don’t think we expected to find 162 studies evaluating decreases in ctDNA and their relationship to survival outcomes. I kept coming back to my mentor, saying, ‘I found more studies!’ That was really encouraging to see. We were expecting to see 30 to 50, maybe.”
The studies had considerable heterogeneity.
Research included ctDNA evaluations in more than 10 different tumor types, the most common being lung cancer (50%), colorectal cancer (10.5%) and melanoma (7.4%).
Studies have been conducted in all phases of trials, as well as observational, retrospective and prospective investigations.
Multiple assays were used, the most common being next-generation sequencing (67.9%). Collection after baseline occurred at various time points, the most common being 6 weeks after treatment (22.5%), 4 weeks (19.2%), 8 weeks (15.4%) and 3 weeks (14.8%).
Additionally, ctDNA sampling varied substantially, with the most frequent time points being cycle 3 day 1 (48.5%), cycle 2 day 1 (31.3%) and cycle 4 day 1 (8.2%).
Yet, of 76 studies that reported hazard ratios, 75 showed ctDNA reductions had associations with improved OS, and most were significant.
“I found that consistency particularly striking, especially given how heterogeneous these studies were in terms of cancer types, treatments, assays, time points and definitions of molecular response,” Levi-D’Ancona said. “Across all these differences, there was only one study that did not show this positive association. The results were really encouraging.”
Researchers acknowledged study limitations, including potentially missing certain studies.
Levi-D’Ancona emphasized more prospective evaluations are needed, and researchers should “consider incorporating ctDNA into trial planning.”
“That includes carefully thinking about when blood samples should be collected, rather than just relying on whatever samples or timepoints happen to be available retrospectively,” she said.
Levi-D’Ancona also noted the importance of standardization.
“We need to better understand whether differences in assays, timing, definitions of molecular response, and other study characteristics affect its relationship with survival,” she said. “If we can be more intentional about prospective data collection, we can build a stronger and more consistent body of evidence that can be used for a larger meta-analysis.
“We know that ctDNA is associated with survival. Now the question is, how can we best measure and use ctDNA so that it can reliably inform clinical trials and ultimately benefit patients?”