Liquid Biopsy for Cancer Monitoring: How ctDNA Is Transforming Recurrence Detection

What Is a Liquid Biopsy and How Does ctDNA Work?

When cancer cells divide, die, or are destroyed by the immune system, they release fragments of their DNA into the bloodstream. This circulating tumor DNA (ctDNA) carries the same genetic mutations and alterations present in the tumor itself. While ctDNA typically constitutes less than 1% of total cell-free DNA (cfDNA) in the blood, advances in next-generation sequencing (NGS) and polymerase chain reaction (PCR) technologies have made it possible to detect and analyze these trace amounts with high sensitivity and specificity.

Two main approaches are used in ctDNA analysis. Tumor-informed assays, such as Natera's Signatera test, first sequence the patient's tumor tissue to identify 16 or more patient-specific somatic mutations (single nucleotide variants, or SNVs), then design a personalized multiplex PCR panel to detect those exact mutations in blood samples. This approach achieves very high sensitivity (detection limits below 0.01% variant allele frequency) because it looks for known, patient-specific alterations. Tumor-agnostic (or tumor-naive) assays, such as Guardant Health's Guardant Reveal, analyze blood samples without prior tumor sequencing by screening for a broad panel of cancer-associated mutations, methylation patterns, or fragmentation patterns. These are less sensitive than tumor-informed approaches but offer the advantage of not requiring tumor tissue.

The clinical applications of ctDNA liquid biopsy span the cancer care continuum: detecting molecular residual disease (MRD) after curative-intent surgery, monitoring for recurrence during surveillance, assessing treatment response in real time, identifying resistance mutations that emerge during therapy, and potentially screening for early-stage cancer (though this application is less mature). The non-invasive nature of blood draws allows serial monitoring over time, providing a dynamic view of cancer biology that static imaging cannot offer.

How Is ctDNA Used to Monitor Cancer Recurrence?

The most impactful current application of ctDNA is detecting molecular residual disease (MRD) after surgery with curative intent. MRD refers to cancer cells or tumor DNA remaining after treatment that are below the detection threshold of conventional imaging (CT, MRI, PET scans, which typically require tumors of 5–10mm to be visible). Studies have consistently shown that patients who are ctDNA-positive after surgery (indicating MRD) have a dramatically higher risk of recurrence compared to ctDNA-negative patients. In stage II–III colorectal cancer, the DYNAMIC study (published in NEJM, 2022) found that ctDNA-positive patients had a recurrence rate of approximately 79% without adjuvant chemotherapy.

Serial ctDNA monitoring during post-surgical surveillance can detect recurrence months before it becomes visible on imaging. In colorectal cancer, the GALAXY study (part of the CIRCULATE-Japan initiative) demonstrated that ctDNA positivity at 4 weeks post-surgery was the strongest prognostic factor for recurrence, outperforming traditional staging, and that ctDNA became positive a median of 5.5 months before radiological recurrence. In breast cancer, the c-TRAK TN trial showed ctDNA detected relapse a median of 10.7 months earlier than clinical or imaging detection in triple-negative breast cancer patients.

This lead time creates a therapeutic window during which recurrence can potentially be treated while tumor burden is minimal and more likely to respond to therapy. Early detection of recurrence through ctDNA may allow for curative re-intervention (repeat surgery or targeted radiation) rather than palliative systemic therapy, potentially improving long-term outcomes. However, it is important to note that while ctDNA monitoring reliably identifies patients at high risk of recurrence, randomized trials are still ongoing to definitively prove that acting on ctDNA results (rather than waiting for imaging evidence) improves overall survival.

How Does ctDNA Guide Adjuvant Therapy Decisions?

One of the most promising applications of ctDNA is guiding adjuvant therapy decisions—the chemotherapy given after surgery to eliminate residual microscopic disease. Currently, adjuvant chemotherapy decisions are based primarily on pathological staging (tumor size, lymph node involvement, grade), but this approach is imprecise: many patients with early-stage disease receive unnecessary chemotherapy, while some with presumably favorable features still recur. ctDNA-guided therapy aims to match treatment intensity to actual residual disease risk.

The DYNAMIC trial (Circulating Tumour DNA Analysis Informing Adjuvant Chemotherapy in Stage II Colon Cancer, NEJM 2022) was a landmark randomized trial that used ctDNA results to guide adjuvant therapy in 455 patients with stage II colon cancer. In the ctDNA-guided group, patients who were ctDNA-positive received adjuvant chemotherapy (fluoropyrimidine with or without oxaliplatin), while ctDNA-negative patients were observed without chemotherapy. The ctDNA-guided approach reduced chemotherapy use by 49% (from 28% to 15% of patients receiving treatment) without compromising 2-year recurrence-free survival (93.5% in ctDNA-guided vs. 92.4% in standard management). This demonstrated that ctDNA could safely spare nearly half of stage II patients from unnecessary chemotherapy.

In breast cancer, the monarchE trial biomarker analysis and the PENELOPE-B trial have explored ctDNA for risk stratification in hormone receptor-positive, HER2-negative early breast cancer. Ongoing trials including CIRCULATE-Japan (colorectal), DYNAMIC-III (stage III colon cancer), and ZEST (breast cancer, evaluating the CDK4/6 inhibitor ribociclib in ctDNA-positive patients after standard adjuvant therapy) are expected to provide definitive evidence on whether ctDNA-guided escalation and de-escalation of adjuvant therapy improves overall survival. The paradigm shift from anatomy-based staging to molecular-based risk stratification represents one of the most significant advances in precision oncology.

What Are the Current Limitations and Future Directions?

Despite rapid progress, ctDNA liquid biopsy faces several important limitations. Sensitivity varies significantly by tumor type: cancers with high ctDNA shedding (colorectal, lung, pancreatic) are more readily detected than low-shedding tumors (renal cell carcinoma, brain tumors, and some early-stage cancers). In stage I cancer of any type, ctDNA detection sensitivity may be as low as 30–50%, compared to 80–90% in stage III–IV disease. False-negative results can provide false reassurance, and false-positive results (from clonal hematopoiesis of indeterminate potential, or CHIP—age-related mutations in blood cells) can cause unnecessary anxiety and overtreatment.

The most critical unanswered question is whether ctDNA-guided treatment decisions improve overall survival compared to standard care. While observational data strongly suggest that ctDNA status is prognostic (it predicts who will recur), proving that it is predictive (that acting on the information changes outcomes) requires randomized clinical trials with survival endpoints. Multiple such trials are underway, including CIRCULATE (colorectal, multiple countries), DYNAMIC-III (stage III colon, Australia), and COBRA (stage IIA colon, US), with results expected between 2025 and 2028.

Looking forward, multi-cancer early detection (MCED) tests represent the ambitious next step in liquid biopsy technology. Grail's Galleri test, which analyzes cell-free DNA methylation patterns to screen for over 50 cancer types in asymptomatic individuals, is currently being evaluated in the NHS-Galleri trial involving 140,000 participants in the United Kingdom. Other MCED platforms under development include CancerSEEK (Exact Sciences), DETECT-A, and various academic research programs. While MCED tests show promise for detecting cancers that lack established screening methods (pancreatic, ovarian, liver), questions about sensitivity, specificity, overdiagnosis, and cost-effectiveness must be resolved before widespread implementation. The FDA has signaled interest in creating a regulatory pathway for MCED tests, and a premarket review framework is under development.