TL;DR
- ctDNA adoption in oncology trials has grown significantly over the past decade.
- Despite growing interest, ctDNA remains overwhelmingly a secondary or exploratory endpoint rather than a primary endpoint.
- Only four oncology trials in our dataset use ctDNA as a primary endpoint.
- Adoption is concentrated in solid tumors, particularly NSCLC, HER2-negative breast cancer, colorectal cancer, and CRPC.
- The recent slowdown in new ctDNA trials appears to reflect increasing emphasis on validation, standardization, and regulatory alignment rather than declining interest.
- Broader adoption will depend on standardized sampling, assay alignment, context-specific validation, and prospective correlation with accepted clinical outcomes.
Why ctDNA keeps returning to the boardroom agenda
Circulating tumor DNA (ctDNA) sits at the intersection of what oncology development teams want and what traditional endpoints often fail to deliver: earlier, more sensitive readouts than imaging, with the potential to shorten timelines and improve decision-making. In executive terms, ctDNA is attractive because it promises faster answers to expensive questions—whether a therapy is working, whether resistance is emerging, and whether a trial design is on track to support a credible label.
The appeal is also operational. If ctDNA can be measured repeatedly, it can support more adaptive trial strategies, reduce reliance on infrequent scans, and improve visibility into response dynamics between scheduled imaging timepoints. That matters most in settings where radiographic change lags biology, or where early clinical decisions determine whether an asset continues to receive investment.
The longer-term ambition is larger: ctDNA dynamics (reduction, clearance, or molecular progression) as a surrogate endpoint, particularly in early-stage disease where event-free survival and overall survival can take years. Yet the evidence package required to treat ctDNA as a decision-driving endpoint—standardized sampling, assay performance, consistent thresholds, and prospective confirmation—has not been established at scale.
This gap between promise and proof is exactly what the trial landscape reflects.
Distribution of ctDNA Trials by Trial Start Year
Source: Intelligencia AI oncology trial database (MRD endpoints excluded) · 111 total interventional trials
Trial adoption has accelerated meaningfully since 2019, with the 2023 peak reflecting broad exploratory integration across oncology. The 2025 pullback is best read as an evidence-engineering recalibration — not a loss of conviction — as sponsors refine endpoint definitions and assay standards to meet tightening regulatory expectations.
What our dataset shows: clear traction, almost no primary-endpoint use
Across oncology, our database identifies 111 interventional trials using ctDNA explicitly as an endpoint (excluding MRD endpoints by design). The trendline indicates meaningful uptake over time, but the positioning is consistent: ctDNA is overwhelmingly secondary or exploratory—used to enrich interpretation, support mechanistic narratives, or guide internal decisions rather than serve as a registrational anchor.
Only four oncology trials have used ctDNA as a primary endpoint in our dataset—an important reality check for anyone assuming ctDNA is already “the next PFS.” In practice, sponsors appear to be moving cautiously: integrating ctDNA broadly, but rarely placing it in the role that would carry the most regulatory weight.
This pattern echoes what we documented previously for MRD: substantial attention, limited primary-endpoint adoption.
The difference is that MRD’s validation path is often disease-specific and narrower, while ctDNA must earn credibility across heterogeneous solid tumor contexts.
ctDNA Endpoint Positioning Across Oncology Trials
Source: Intelligencia AI oncology trial database · 111 interventional trials
While ctDNA adoption has expanded rapidly across oncology development, only 4 of 111 identified trials use ctDNA as a primary endpoint. Sponsors are clearly integrating ctDNA into development programs, but overwhelmingly as a secondary or exploratory measure rather than a registrational anchor.
Where ctDNA is being used: concentrated in solid tumors, led by a few indications
ctDNA endpoint use in our dataset is almost exclusively concentrated in solid tumors, with the heaviest focus in:
- NSCLC
- HER2-negative breast cancer
- Colorectal cancer
- Castration-resistant prostate cancer
This concentration is strategically coherent. These indications combine high clinical need with a growing ecosystem of targeted therapies and biomarker-driven decision-making. ctDNA fits naturally where there is a desire to quantify molecular response earlier, detect emerging resistance sooner, and segment patients more precisely than imaging alone can support.
Why ctDNA is “solid-first”
ctDNA endpoint validation is constrained by shedding variability and tumor biology. Detectability and kinetics vary meaningfully across tumor types, disease stages, metastatic patterns, and treatment contexts. That variability complicates universal thresholds and limits cross-trial pooling—one reason ctDNA remains concentrated in tumor types where shedding is common and assays are operationally viable.
In liquid cancers, the field already has a more mature biomarker framework—MRD—with established sampling conventions and sensitivity standards. That makes “pure ctDNA endpoints” less central there, even though the underlying technologies overlap. Our earlier MRD drop captured how even MRD—despite stronger precedent—has still seen limited primary-endpoint adoption.
The 2025 dip: why program starts may have slowed
The visible decline in 2025 program starts likely reflects a mix of factors that often appear when a field moves from enthusiasm to standard-setting:
- A recalibration toward standardization
Across sponsors and stakeholders, the ctDNA discussion has increasingly shifted from “ctDNA is promising” to “what exact definitions, thresholds, timing, and assay rules are required to make it credible as an endpoint?” When that transition occurs, trial initiation can temporarily slow as protocols are redesigned and endpoint language becomes more conservative.
- Regulatory messaging has moved from general support to tighter expectations
In late 2024, FDA issued guidance on ctDNA use as a biomarker in early-stage solid tumor drug development, reinforcing the legitimacy of ctDNA—but also making clear that analytical validity, context of use, and evidence generation must be explicit.
Guidance often triggers an industry “alignment phase,” where sponsors refine trial designs to match evolving expectations.
- Broader scrutiny of surrogate endpoints
ctDNA’s promise is playing out in a climate where surrogate endpoints are under greater scrutiny—especially when positioned as time-saving substitutes for long-term outcomes. That scrutiny does not negate ctDNA’s potential; it raises the burden of proof and increases caution in how endpoints are framed.
Net: the 2025 dip reads less like declining interest and more like an evidence-engineering phase—a shift from exploratory adoption to disciplined validation-building.
What it will take to validate ctDNA endpoints in practice
Correlation alone is insufficient; ctDNA needs prospective, standardized evidence packages before it can carry decision weight. The requirements are becoming clearer:
- Standardized timing and sampling protocols (when blood is collected, how it is processed), enabling comparability across trials and sponsors.
- Assay performance clarity and threshold agreement, including consistent definitions for “molecular response” and “clearance.” Differences across assays materially limit pooling and meta-analysis.
- Context-specific validation, because ctDNA behavior varies by tumor type, stage, and treatment setting—one-size-fits-all criteria are unlikely to hold.
- Prospective embedding in trials that can credibly link ctDNA changes to accepted outcomes (EFS, PFS, OS), particularly in early-stage disease where time-to-insight value is highest.
This is also why ctDNA differs from MRD. MRD’s path to credibility has been narrower and more disease-specific, while ctDNA must earn evidence across broader tumor biology and clinical contexts.
How we support decision-makers evaluating ctDNA endpoint strategy
As with our prior MRD analysis, the ctDNA story is defined by a gap between headline interest and true adoption as a primary, decision-driving endpoint.
We help teams navigate that gap with evidence rather than optimism:
- Clinical teams: we benchmark how ctDNA endpoints are being used (primary vs secondary/exploratory), by indication, phase, sponsor, and endpoint framing—so protocol choices align with real precedent and realistic regulatory posture.
- Portfolio leaders: we quantify traction versus validation risk—supporting decisions on where ctDNA adds value as an exploratory signal versus where it introduces avoidable uncertainty.
- BD and strategy: we contextualize assets making ctDNA claims by comparing design patterns and endpoint positioning against the landscape—so differentiation is assessed against what the market is truly doing.
Bottom line
ctDNA endpoint use is rising, but validation has not arrived; the winners will be sponsors who treat ctDNA as an evidence-building program—not just a measurement layer.
For more tailored, data-rich insights, let’s talk.
