Dr. Adham Jurdi discusses how serial ctDNA testing may support molecular residual disease monitoring, post-surgical risk assessment and more targeted oncology trial design.
After surgery for a solid tumor, one of the most consequential questions is whether microscopic cancer remains.
Computed tomography (CT) and magnetic resonance imaging (MRI) are essential for assessing radiographically visible disease, but they generally cannot detect very small amounts of residual cancer.
Circulating tumor DNA (ctDNA)-based molecular residual disease (MRD) testing is being studied as a way to address this gap by identifying tumor-derived DNA fragments in the blood.
Senior Medical Director,
GI Oncology and Innovations
Natera
Tumor-informed approaches begin by identifying variants in an individual patient’s tumor and then looking for the same molecular signature in serial blood samples. This can provide a personalized measure of whether residual disease is detectable after surgery, how it changes during treatment and whether molecular recurrence may be occurring before it becomes visible on imaging.
To explore the clinical and research implications of this approach, Xtalks spoke with Dr. Adham Jurdi, Senior Medical Director of GI Oncology and Innovations at Natera.
Moving Beyond Population-Level Recurrence Estimates
Post-surgical treatment decisions have traditionally been guided by tumor stage, pathology and other clinical risk factors. Although these measures remain essential, patients with apparently similar disease can follow very different clinical trajectories.
MRD testing introduces an individualized molecular measure into this assessment. In gastrointestinal cancers, particularly colorectal cancer, Dr. Jurdi said prospective evidence has shown that detectable ctDNA after surgery is a strong prognostic marker for recurrence.
“We’re moving away from group statistics towards understanding each patient’s own molecular journey and reality,” he said. “And that kind of precision is what we actually need to change outcomes in the clinic.”
This distinction is important because conventional staging estimates risk across groups of patients. MRD testing instead seeks evidence of residual cancer associated with the individual patient’s tumor.
“A single test is a snapshot. It’s a cross-section in time as to what’s going on in that patient at that moment.”
— Dr. Adham Jurdi
However, showing that MRD can predict recurrence risk is not the same as proving that it should guide treatment decisions. A biomarker may identify patients who are more likely to experience recurrence, but researchers must still show that changing treatment based on the result actually improves outcomes, such as disease-free or overall survival. Prospective interventional trials are therefore needed to determine how MRD results should be used in clinical practice.
A regulatory precedent now exists in muscle-invasive bladder cancer, where selecting patients for adjuvant treatment based on MRD status improved disease-free and overall survival. However, these findings apply to a specific cancer, treatment and testing approach. They do not show that MRD-guided treatment is clinically effective across gastrointestinal cancers or other solid tumors.
Why Serial ctDNA Testing Matters
A single blood draw provides information from one point in time. Serial testing can show whether ctDNA is clearing, remaining stable or increasing, potentially offering a more complete picture of disease dynamics.
“A single test is a snapshot. It’s a cross-section in time as to what’s going on in that patient at that moment,” Dr. Jurdi explained. “But serial testing is really what gives you the full story. Is the disease responding? Is it dormant? Is it coming back?”
For patients receiving adjuvant therapy, declining or clearing ctDNA may provide an early molecular indication that treatment is having an effect. Persistent or rising ctDNA may suggest that residual disease is not responding adequately, potentially creating an opportunity to investigate alternative strategies before radiographic recurrence.
Dr. Jurdi used the term “landmark MRD negativity” to describe ctDNA clearance at an early, predefined assessment after treatment began. Sustained clearance may have different implications from a single negative result, while a plateau or subsequent rise could indicate a different disease trajectory.
A negative result does not definitively exclude residual disease. Some tumors may release little ctDNA into the bloodstream, while very low concentrations may fall below an assay’s limit of detection. This is another reason serial testing may provide more information than a single negative blood draw. The FDA recommends that sponsors account for false-negative results, assay sensitivity and variation in tumor shedding when using ctDNA in early-stage cancer trials.
This type of longitudinal assessment could be particularly relevant in the adjuvant setting, where there is usually no measurable tumor to monitor. Patients may receive several months of therapy without knowing whether microscopic disease has been eliminated, and the effectiveness of treatment may not become apparent until recurrence occurs.
Treatment Escalation and De-escalation Require Clinical Evidence
MRD testing could eventually help identify patients who may benefit from more intensive or prolonged treatment, as well as those who could potentially avoid unnecessary therapy.
A positive result after surgery may indicate that a patient remains at substantial risk of recurrence despite having no visible disease on imaging. Conversely, an MRD-negative result could support treatment de-escalation in clinical trials or specific settings where evidence is available, potentially reducing unnecessary chemotherapy exposure. However, a negative result should not serve as a standalone reason to withhold treatment across tumor types.
The critical question is whether altering treatment according to MRD status improves outcomes within each clinical setting.
“MRD should inform clinical decisions, not replace them,” Dr. Jurdi emphasized.
That distinction is especially important for treatment de-escalation. Avoiding unnecessary chemotherapy could reduce neuropathy, fatigue, gastrointestinal effects and other short- and long-term toxicities.
Randomized evidence from the broader ctDNA field has shown that a ctDNA-guided strategy can reduce adjuvant chemotherapy use in selected patients with stage II colon cancer without compromising recurrence-free survival. Additional trials are evaluating whether similar approaches can guide treatment across other patient populations and testing platforms.
For treatment escalation, studies must similarly determine whether intensifying or changing therapy for MRD-positive patients can eliminate residual disease and prevent clinical recurrence, rather than simply identifying patients with a poorer prognosis.
Get industry leading pharma and biotech news, events and expert insights delivered to your inbox.
Pan-Cancer Evidence and Tumor-Specific Actionability
Consistent performance across several tumor types could strengthen confidence that ctDNA-based MRD testing reflects a broadly relevant feature of cancer biology and that the testing approach is reliable.
Dr. Jurdi pointed to findings across colorectal, gastric, biliary, breast, lung, melanoma and bladder cancers. Consistent associations between MRD status and outcomes across geographic regions and clinical settings may indicate that ctDNA detection reflects a broadly relevant feature of cancer biology rather than a phenomenon limited to one disease.
However, pan-cancer prognostic consistency does not necessarily mean that the same clinical action should follow a positive or negative result in every tumor type.
Treatment options, patterns of recurrence, tumor shedding, surveillance practices and the maturity of the evidence differ between cancers. Integration into clinical care will therefore require both broad analytical confidence and disease-specific evidence showing how results should influence management.
For clinicians, regulators and payers, this means evaluating more than whether an assay can identify recurrence risk. Evidence must also address how testing changes care, whether those changes improve outcomes and which testing intervals are appropriate for each clinical setting.
MRD Testing in Oncology Trial Design
Beyond routine care, MRD may have significant implications for oncology clinical trials.
Traditional adjuvant studies often require years of follow-up before enough recurrence events occur to evaluate disease-free survival. Overall survival can take even longer to mature. This creates challenges for sponsors attempting to evaluate new therapies while standards of care and development pipelines continue to evolve.
One potential strategy is to enroll or enrich trials with MRD-positive patients who have molecular evidence of residual disease. These participants may be more likely to experience recurrence and may also have a greater opportunity to benefit from an intervention intended to eliminate microscopic cancer.
By contrast, enrolling a large proportion of MRD-negative patients with relatively low recurrence risk could reduce the observed event rate and make it more difficult to detect a treatment effect.
MRD status could also be used to stratify participants, define biomarker-selected cohorts or evaluate molecular clearance during treatment. According to Dr. Jurdi, these approaches could support more targeted enrollment and provide earlier indications of biological activity.
Another possibility is using ctDNA clearance as a surrogate endpoint. Dr. Jurdi estimated that some conventional oncology trials may take five to eight years to produce disease-free survival results, whereas a validated molecular response endpoint could potentially support readouts within 12 to 18 months.
The proposed 12- to 18-month timeline represents a forward-looking estimate rather than an established regulatory development pathway. The FDA has stated that changes in ctDNA levels or ctDNA clearance are not currently validated as early endpoints for drug approval in early-stage solid tumors. Prospective randomized studies must demonstrate that molecular response reliably predicts longer-term outcomes such as disease-free, event-free or overall survival.
If studies confirm that ctDNA changes reliably predict longer-term clinical outcomes in a specific setting, these endpoints could help sponsors stop ineffective approaches earlier, prioritize more promising therapies and design smaller, more targeted trials.
“You move from having a single window into a tumor to building a comprehensive, dynamic portrait of disease.”
— Dr. Adham Jurdi
Interpreting MRD When Imaging Is Negative
One of the most challenging situations occurs when ctDNA is detected but imaging does not yet show visible disease.
Dr. Jurdi described this as a potential lead-time advantage, because molecular recurrence may be detectable months before a lesion becomes large enough to appear on a scan. A positive result should therefore not be dismissed simply because imaging remains negative.
However, no test is perfect. Low ctDNA levels, technical limitations and differences between assays mean that a positive result should be interpreted alongside the timing of the blood draw, results from repeat testing, imaging findings and the patient’s overall clinical and pathological profile.
The appropriate response may include repeating the blood test, increasing the frequency of surveillance, using a different imaging method, discussing the case with a multidisciplinary tumor board or considering enrollment in a clinical trial.
A positive result does not necessarily mean that treatment should begin immediately. Instead, clinicians must determine whether the finding creates an opportunity for closer monitoring or earlier intervention before the disease becomes visible on imaging.
In some cancers, detecting recurrence earlier could increase the chance that it can still be surgically removed or treated with the goal of achieving a cure. An important area of ongoing research is whether this earlier molecular detection ultimately leads to better patient outcomes.
Integrating MRD into Oncology Workflows
Dr. Jurdi identified access and reimbursement, result interpretation and workflow integration as three major barriers to wider MRD adoption.
Clinicians need clear frameworks explaining what positive, negative, persistent and changing results mean within specific disease and treatment settings. They also need evidence-based guidance on what action, if any, should follow each result.
Education and multidisciplinary discussion will be important as MRD testing moves into oncology practice. Incorporation into clinical guidelines could provide more consistent recommendations, but guideline development will depend on the strength and maturity of prospective clinical utility data.
Operationally, testing must also fit into existing care pathways. Potential collection points include the post-surgical visit, the beginning and end of adjuvant therapy and scheduled surveillance appointments. Results must be available at a time when they can meaningfully inform discussions between clinicians and patients.
Without this integration, testing risks becoming an additional data point without a defined role. With appropriate timing and interpretation, serial ctDNA measurements could become part of a longitudinal monitoring strategy.
Building a More Complete View of Disease
Although ctDNA can provide a sensitive molecular signal, it represents only one layer of tumor biology.
Dr. Jurdi sees the next phase of precision oncology involving the integration of ctDNA with proteomics, epigenomics, transcriptomics, digital pathology and artificial intelligence. Combining these data types could help researchers move beyond determining whether residual disease is present to understanding how that disease is likely to behave.
AI may be particularly useful for identifying patterns across thousands of molecular and pathological features that would be difficult to detect using conventional analysis.
“You move from having a single window into a tumor to building a comprehensive, dynamic portrait of disease,” Dr. Jurdi said.
These models could potentially identify emerging treatment resistance, improve estimates of recurrence risk and support more personalized treatment decisions. However, they will require rigorous validation, transparent and interpretable algorithms, and evidence that combining these data leads to better clinical decisions.
For the oncology field, the central challenge is moving from detecting residual disease to knowing how to act on that information. MRD testing can help identify recurrence risk and track changes in disease over time. Its broader clinical impact will depend on prospective research establishing when testing should be performed, how clinicians should respond to the results and whether MRD-guided strategies improve patient outcomes.
Join or login to leave a comment
JOIN LOGIN