Molecular Residual Disease in Solid Tumors: Understanding ctDNA and NGS-Based Approaches

MGI

/

Science

/

Molecular Residual Disease in Solid Tumors: Understanding ctDNA and NGS-Based Approaches

Nov 20, 2026

Share article

Molecular Residual Disease
Molecular Residual Disease

Discover how ctDNA analysis and next-generation sequencing (NGS) enable molecular residual disease (MRD) detection in solid tumors, from tumor-informed and tumor-agnostic approaches to sensitive monitoring and relapse risk assessment.

From liquid biopsy to molecular residual disease (MRD)

Liquid biopsy is a minimally invasive approach that enables the analysis of biological material released by cells into body fluids, most commonly blood. In oncology, one of its main applications is the detection of circulating tumor DNA (ctDNA), which consists of small DNA fragments released by tumor cells into the bloodstream.

To understand ctDNA, it is important to distinguish it from cell-free DNA (cfDNA). cfDNA refers to all extracellular DNA fragments circulating in the bloodstream, originating from different tissues and cell types. In cancer patients, a proportion of this DNA originates from tumor cells and is referred to as ctDNA. The relative abundance of ctDNA can vary considerably depending on tumor type, disease stage, tumor burden and treatment response.

Following surgery, radiotherapy or systemic treatment, the amount of ctDNA in the bloodstream can decrease substantially. However, the absence of detectable disease on conventional imaging does not necessarily mean that all cancer cells have been eliminated. A small population of residual cancer cells may remain in the body and potentially lead to disease recurrence.

These residual cells can release small amounts of tumor-derived DNA into the bloodstream. Detecting this molecular signal after treatment is the principle behind Molecular Residual Disease (MRD) testing.

While liquid biopsy can also support tumor profiling, treatment selection, response monitoring and resistance detection, ctDNA-based MRD testing addresses a more specific question:

Is there molecular evidence of cancer remaining after treatment, even when conventional clinical assessments and imaging show no detectable disease?

In solid tumors, the term molecular residual disease is often preferred to minimal or measurable residual disease because the test detects a cancer-associated molecular signal rather than directly identifying or counting viable residual cancer cells. [1] [2]

What is MRD and why is it important?

Understanding residual disease after cancer treatment

Cancer treatment aims to eliminate tumor cells or reduce their number to a level where the disease can no longer be detected using conventional diagnostic methods.

For example, following surgery to remove a solid tumor, imaging may indicate that the tumor has been completely removed and that there is no visible evidence of disease. Nevertheless, microscopic cancer cell populations may remain locally or elsewhere in the body.

These remaining cells are not necessarily detectable through imaging because their number and distribution are below the detection capabilities of conventional techniques.

If these cells survive and continue to grow, they may eventually cause the cancer to return. This is known as disease recurrence or relapse.

MRD testing aims to identify molecular evidence of these residual cancer cells before recurrence becomes clinically or radiologically detectable.

How does ctDNA help detect residual disease?

Tumor cells release DNA fragments into the bloodstream through processes including cell death and turnover. These fragments can carry genetic alterations associated with the tumor, such as somatic mutations.

By analyzing plasma cfDNA, laboratories can search for these tumor-associated alterations and determine whether a cancer-related molecular signal is still detectable following treatment.

Two possible results can be obtained:

  • MRD-positive: Tumor-associated molecular signals are detected in the blood, suggesting a higher risk of disease recurrence.

  • MRD-negative: No tumor-associated signal is detected above the analytical detection threshold of the assay.

Importantly, an MRD-negative result does not prove that all cancer cells have been eliminated. The tumor may release very little ctDNA, the blood sample may contain too few informative molecules, or the assay may not cover the relevant genetic alterations.

Why is MRD clinically relevant?

MRD provides an additional molecular layer of information to conventional patient follow-up.

A 2024 meta-analysis including 80 prospective studies found that postoperative ctDNA positivity was associated with a pooled recurrence hazard ratio of 7.48 (95% CI 6.39–8.77). Longitudinal testing achieved a pooled sensitivity of 74% (95% CI 68–80%), compared with 50% (95% CI 46–55%) for landmark testing. [3]

These results combine different cancer types and assay designs and should not be interpreted as the expected performance of every MRD test. [3]

Key concept: MRD positivity is generally associated with an increased risk of relapse. However, MRD negativity means that no molecular signal was detected in the sample tested, not necessarily that no residual cancer cells remain.


How did we arrive at ctDNA-based MRD testing?

From conventional follow-up to molecular monitoring

Traditionally, cancer patients are monitored after treatment through a combination of clinical examinations, medical imaging and, in selected cancers, circulating protein biomarkers.

Although these methods remain essential, they have limitations when residual disease is present at a very early stage.

For example, imaging techniques detect anatomical changes. If the remaining cancer cell population is extremely small, it may not produce a detectable lesion.

Liquid biopsy introduced a different possibility: instead of waiting for a tumor to become visible, laboratories can investigate whether tumor-derived DNA is circulating in the bloodstream.

However, detecting these rare DNA fragments requires highly sensitive analytical methods.

The role of next-generation sequencing

The development of next-generation sequencing (NGS), together with molecular barcoding and advanced bioinformatics, has made it possible to identify very low-frequency genetic alterations within a large background of normal cfDNA.

Several technological developments have contributed to this progress:

  • High sequencing depth: Generates more sequencing observations to improve the opportunity to detect rare variants.

  • Unique molecular identifiers (UMIs): Molecular barcodes attached to DNA fragments help distinguish original DNA molecules from sequencing duplicates.

  • Error correction: Helps reduce the impact of sequencing errors that could otherwise be mistaken for genuine low-frequency mutations.

  • Bioinformatics analysis: Identifies, filters and interprets cancer-associated signals within complex sequencing data.

This evolution has progressively enabled more sensitive molecular monitoring after cancer treatment.

Evolution of cancer monitoring

Clinical follow-up and imaging → Molecular biomarkers → Liquid biopsy → ctDNA analysis → Highly sensitive MRD assessment.

Another important development is the move from single post-treatment measurements to longitudinal monitoring. Because ctDNA concentrations can fluctuate over time, repeated blood sampling increases the opportunity to detect molecular signals that may be absent or below the detection threshold in an earlier sample. [2] [3]


How does the MRD workflow work?

The two main approaches to ctDNA-based MRD testing are tumor-informed and tumor-agnostic. In this article, tumor-agnostic refers to plasma-based testing that does not require prior molecular characterization of the patient’s tumor tissue.

The fundamental difference is whether the patient's tumor must first be genetically characterized to identify the specific alterations that will subsequently be monitored in blood.

Tumor-informed approach: Tracking a patient's tumor fingerprint

The tumor-informed strategy starts by identifying the genetic characteristics of the patient's original tumor.

Step 1: Tumor tissue characterization

A tissue sample obtained from the original tumor is analyzed using a sequencing approach such as a targeted panel, comprehensive panel, whole-exome sequencing (WES) or whole-genome sequencing (WGS).

The objective is to identify somatic alterations, meaning genetic changes acquired by tumor cells that are not inherited.

Matched white-blood-cell DNA, commonly obtained from the buffy coat, may also be analyzed to distinguish tumor-specific mutations from germline variants and alterations associated with clonal hematopoiesis.

Step 2: Selection of tumor-specific alterations

Once the tumor's genetic profile has been established, a selection of informative alterations is identified.

These alterations form a patient-specific molecular fingerprint that can be used to search for traces of the original tumor in subsequent blood samples.

Step 3: Plasma monitoring

Blood is collected and processed to obtain plasma, from which cfDNA is extracted.

The selected alterations are then monitored using a highly sensitive method. In NGS-based workflows, sequencing is used to search for these targets, while digital PCR (dPCR) may also be suitable when only one or a few known variants are monitored.

If these alterations are detected, this provides molecular evidence that tumor-associated DNA is still present in the bloodstream.

Step 4: Longitudinal monitoring

The process can be repeated at different time points to monitor changes in the molecular signal over time.

The advantage is that the analysis focuses on alterations already known to be present in the patient's tumor.

Personalization does not necessarily require a completely unique laboratory panel for every patient. Laboratories may use patient-specific assays or a common panel combined with patient-specific bioinformatics analysis. [2]

In simple terms: identify the tumor's genetic fingerprint once, then repeatedly search for that fingerprint in the patient's blood.

Tumor-agnostic approach: Searching for cancer-associated signals without prior tumor profiling

Unlike tumor-informed testing, the tumor-agnostic approach does not require prior sequencing of the original tumor.

The assay analyzes plasma cfDNA directly using a predefined strategy designed to identify cancer-associated signals across different patients.

Step 1: Blood collection and cfDNA extraction

A blood sample is collected and processed to isolate plasma. cfDNA is then extracted for molecular analysis.

Step 2: Direct molecular analysis

The extracted cfDNA is analyzed using a predefined assay that may investigate several types of cancer-associated signals, including:

  • Recurrent somatic mutations.

  • DNA methylation patterns.

  • Fragmentation characteristics of cfDNA.

  • Copy-number alterations.

  • Combinations of genomic and epigenomic features.

Unlike a tumor-informed approach, the assay does not need to know the original tumor's specific genetic fingerprint before testing.

Step 3: Bioinformatics analysis

Sequencing data are processed to identify patterns that may indicate the presence of tumor-derived DNA.

Matched white-blood-cell DNA may also be used to filter alterations associated with clonal hematopoiesis, even though original tumor tissue is not required. [1] [2] [4]

In simple terms: instead of looking for a previously identified tumor fingerprint, the assay searches directly for predefined cancer-associated molecular patterns.


Tumor-informed vs tumor-agnostic: What are the differences?

Both approaches aim to detect molecular evidence of residual disease, but their underlying strategies have important implications for sensitivity, workflow complexity, scalability and cost.

Parameter

Tumor-informed

Tumor-agnostic

Basic principle

Tracks known tumor-specific alterations

Searches for predefined cancer-associated signals

Starting material

Tumor tissue, often with normal DNA

Blood/plasma

Follow-up sample

Plasma cfDNA

Plasma cfDNA

Target selection

Patient-specific alterations

Predefined genomic, epigenomic or fragment-based signals

Analytical sensitivity

Often advantageous at very low ctDNA levels because the expected signal is known

Depends strongly on assay design and signal integration

Time to first assay

Longer due to tumor characterization

Faster because prior tumor profiling is not required

Scalability

More complex, patient-specific analysis

More standardized and easier to batch

Cost structure

Higher initial setup cost; potentially focused follow-up sequencing

Lower personalization cost; potentially greater sequencing requirements per follow-up

Tumor evolution

May miss new clones not represented by selected targets

May identify new signals within assay coverage, but this is not guaranteed

Which approach offers better sensitivity?

Knowing the tumor's genetic fingerprint can provide an analytical advantage, particularly when ctDNA concentrations are extremely low.

Because the assay already knows which alterations to look for, sequencing and analysis can be concentrated on a selected group of informative targets.

However, sensitivity is not determined by the approach alone.

In a breast cancer cohort, Parsons et al. found greater sensitivity at very low ctDNA levels with tumor-informed assays tracking multiple patient-specific variants than with the tumor-naïve copy-number approaches evaluated. However, assay performance depends on its design, so these findings should not be generalized to all tumor-informed or tumor-agnostic assays. [4]


Why use NGS for MRD?

The main advantage of NGS is its ability to analyze multiple molecular targets simultaneously.

When ctDNA represents only a tiny fraction of total cfDNA, relying on the detection of a single mutation may be insufficient.

By analyzing multiple tumor-associated alterations in parallel, NGS can combine information from several independent molecular signals to improve detection confidence.

What makes NGS particularly relevant?

  • Multiplexing: Multiple mutations and genomic regions can be investigated simultaneously.

  • High analytical depth: Supports the detection of low-frequency variants when sufficient informative molecules are available.

  • Error correction: Molecular barcodes and bioinformatics help distinguish genuine low-frequency alterations from sequencing noise.

  • Flexibility: NGS can support both broad tumor characterization and focused longitudinal monitoring.

  • Multiple molecular features: Depending on the assay, analysis can include mutations, copy-number changes, structural variants, methylation or fragmentation information.

Is NGS always the best option?

Not necessarily.

If a laboratory only needs to monitor one or a few known genetic alterations, dPCR may offer a faster and less expensive alternative.

NGS becomes particularly valuable when several molecular targets need to be combined, when the tumor fingerprint must first be identified, or when broader genomic information is required. [2] [4]

The choice of technology should therefore be guided by the biological question and analytical requirements, rather than by sequencing technology alone.


Which NGS approach can be used for MRD?

One important distinction is that tumor characterization and MRD monitoring do not necessarily require the same sequencing strategy.

The first objective is to identify informative tumor-associated alterations. The second is to detect these alterations at very low concentrations in plasma.

This means that sequencing breadth and sequencing depth have different roles.

Understanding the available sequencing approaches

Targeted or specific panels

These panels focus on selected genes or genomic regions.

They are particularly useful when the relevant tumor-associated alterations are already known and need to be monitored at high sequencing depth.

Their main limitation is that alterations outside the selected regions cannot be detected.

Comprehensive gene panels

These panels cover hundreds of genes or larger predefined genomic regions.

They provide broader opportunities to identify informative variants and can support tumor characterization or standardized plasma testing.

However, their coverage remains limited to the genomic regions included in the panel.

Whole-exome sequencing (WES)

WES investigates protein-coding regions of the genome, known as exons.

It provides broader discovery opportunities than targeted panels and can help identify patient-specific somatic mutations for subsequent monitoring.

Its main limitation is that it does not cover the entire genome.

Whole-genome sequencing (WGS)

WGS investigates both coding and non-coding regions across the genome and can also provide information about structural genomic alterations.

It offers the broadest discovery space and the potential to identify large numbers of patient-specific variants.

However, it generates substantial sequencing data and requires significant bioinformatics resources.

Breadth vs depth: Understanding the balance

A common misconception is that broader sequencing always means better MRD detection.

In reality, sequencing breadth and depth must be considered together.

For a fixed sequencing read budget, a targeted assay can concentrate more reads on selected genomic regions, allowing greater depth at those positions.

A broader approach such as WES or WGS distributes sequencing resources across a much larger genomic space.

However, broad discovery may identify hundreds or thousands of informative tumor-specific variants that can subsequently be combined during focused monitoring.

Therefore, the most appropriate strategy depends on the number of informative targets, the available cfDNA input, the required analytical sensitivity and the sequencing resources available.

Key concept: Broad sequencing helps identify the tumor fingerprint; focused, deep sequencing helps monitor that fingerprint over time.


How does MRD testing benefit patients?

The potential value of MRD testing lies in its ability to provide molecular information that may not yet be visible through conventional diagnostic methods.

By identifying cancer-associated DNA signals in blood, MRD testing can contribute to a more detailed understanding of a patient's disease status following treatment.

Potential clinical applications

1. Relapse risk assessment

An MRD-positive result following treatment can indicate an increased risk of disease recurrence, helping identify patients who may require closer clinical follow-up.

2. Earlier molecular detection

In some clinical settings, ctDNA can become detectable before disease recurrence is visible through conventional imaging.

3. Longitudinal disease monitoring

Repeated blood sampling allows clinicians and researchers to follow changes in tumor-associated molecular signals over time.

4. Treatment strategy evaluation

MRD results may potentially support decisions about escalating, de-escalating or adjusting the duration of adjuvant treatment, when supported by clinical evidence for the specific cancer type and treatment setting.

5. Clinical research

MRD testing provides a valuable tool for investigating treatment response and evaluating new therapeutic strategies in clinical studies.

Clinical validity vs clinical utility: An important distinction

Although ctDNA-based MRD testing has demonstrated strong prognostic potential, it is important to distinguish two concepts.

  • Clinical validity: Does the MRD result reliably indicate a patient's risk of recurrence?

  • Clinical utility: Does acting on that result actually improve patient outcomes?

MRD has demonstrated prognostic and clinical validity in several settings, but its routine use to guide treatment or replace standard surveillance is not generally recommended unless an evidence-based clinical action is available. [9]

An MRD-positive result may identify a patient with a higher risk of relapse. However, this does not automatically mean that changing their treatment based on that result will improve survival or quality of life.

Evidence supporting MRD-guided treatment decisions remains dependent on the cancer type, treatment setting and specific clinical intervention.

For this reason, MRD testing should currently be considered complementary to established follow-up methods, including imaging, pathology, clinical assessment and validated biomarkers. [1]

 

Key take-home messages

  • MRD is a specific application of liquid biopsy: ctDNA analysis enables the detection of cancer-associated molecular signals after treatment.

  • Residual disease can exist without visible tumors: Molecular testing may identify signals that remain undetectable through conventional imaging.

  • Two main strategies are available: Tumor-informed testing tracks previously identified tumor-specific alterations, while tumor-agnostic testing searches for predefined cancer-associated signals without requiring prior tumor profiling.

  • NGS supports highly sensitive detection: Multiplexing, sequencing depth and error correction help identify rare molecular signals.

  • Different sequencing approaches serve different purposes: Targeted panels, comprehensive panels, WES and WGS offer different balances of breadth, depth and data requirements.

  • Broader sequencing is not automatically better: The optimal strategy depends on assay design, informative targets, cfDNA input and analytical sensitivity.

  • Assay selection depends on workflow requirements: The number of targets, cfDNA input, analytical sensitivity, throughput and turnaround time should guide the choice of assay.

  • MRD results provide valuable prognostic information: However, their use to guide treatment decisions requires clinical evidence specific to the cancer and treatment setting.

Ultimately, successful MRD testing is not about sequencing as much DNA as possible. It is about detecting an extremely small cancer-associated molecular signal with sufficient sensitivity and specificity to make the result meaningful.


References

  1. Pascual J, et al. ESMO recommendations on the use of circulating tumour DNA assays for patients with cancer. Annals of Oncology. 2022;33:750–768. https://www.annalsofoncology.org/article/S0923-7534(22)01721-5/fulltext

  2. Moding EJ, Nabet BY, Alizadeh AA, Diehn M. Detecting Liquid Remnants of Solid Tumors: Circulating Tumor DNA Minimal Residual Disease. Cancer Discovery. 2021;11:2968–2986. https://aacrjournals.org/cancerdiscovery/article/11/12/2968/674721/Detecting-Liquid-Remnants-of-Solid-Tumors

3.      Zheng J, Qin C, Wang Q, et al. Circulating tumour DNA-based molecular residual disease detection in resectable cancers: a systematic review and meta-analysis. eBioMedicine. 2024;103:105109. https://pmc.ncbi.nlm.nih.gov/articles/PMC11021841/

  1. Parsons HA, et al. Comparison of tumor-informed and tumor-naïve sequencing assays for ctDNA detection in breast cancer. EMBO Molecular Medicine. 2023;15. https://pmc.ncbi.nlm.nih.gov/articles/PMC10245040/

  2. Japan Society of Clinical Oncology. Position paper on appropriate clinical use of molecular residual disease testing. International Journal of Clinical Oncology. 2025. https://pubmed.ncbi.nlm.nih.gov/39920551/

  3. Semenkovich NP, et al. Genomic approaches to cancer and minimal residual disease detection using circulating tumor DNA. Journal for ImmunoTherapy of Cancer. 2023;11. https://jitc.bmj.com/content/11/6/e006284

  4. Larribère L, Martens UM. Advantages and Challenges of Using ctDNA NGS to Assess the Presence of Minimal Residual Disease in Solid Tumors. Cancers. 2021;13:5698. https://www.mdpi.com/2072-6694/13/22/5698

  5. Chen H, Zhou Q. Detecting liquid remnants of solid tumors treated with curative intent: circulating tumor DNA as a biomarker of minimal residual disease. Oncology Reports. 2023;49:106. https://pmc.ncbi.nlm.nih.gov/articles/PMC10152452/

  6. Lockwood CM, Messersmith HJ, Kim AS, et al. Circulating Tumor DNA Testing in Solid Tumors and Lymphoma: ASCO Guideline. JCO Oncology Practice. 2026;22(9):1468–1489.
    https://ascopubs.org/doi/10.1200/OP-26-00311

ctDNA

liquid biopsy

MRD

next generation sequencing

precision medicine

Share this article :

Share

Join our newsletter to stay up to date on features and releases.

I have read and understood MGI’s Privacy Policy, and I consent to the collection and processing of my personal data for handling, responding to my contact, receiving your newsletter as well as promotion and marketing activities.

*For Research Use Only

Not for use in diagnostic procedures (except as specifically noted).

Copyright © 2026 MGI tech GmbH, Ltd. All Rights Reserved.

Join our newsletter to stay up to date on features and releases.

I have read and understood MGI’s Privacy Policy, and I consent to the collection and processing of my personal data for handling, responding to my contact, receiving your newsletter as well as promotion and marketing activities.

*For Research Use Only

Not for use in diagnostic procedures (except as specifically noted).

Copyright © 2026 MGI tech GmbH, Ltd. All Rights Reserved.

Join our newsletter to stay up to date on features and releases.

I have read and understood MGI’s Privacy Policy, and I consent to the collection and processing of my personal data for handling, responding to my contact, receiving your newsletter as well as promotion and marketing activities.

*For Research Use Only

Not for use in diagnostic procedures (except as specifically noted).

Copyright © 2026 MGI tech GmbH, Ltd. All Rights Reserved.