(UroToday.com) The 2025 American Society of Clinical Oncology Genitourinary (ASCO GU) cancers symposium held in San Francisco, CA between February 13th and 15th 2025, was host to the Panning for Gold: The Role of ctDNA as a Biomarker for Bladder Cancer Session. Dr. Alexander Wyatt discussed the evolution of ctDNA detection with next generation technology.
Dr. Wyatt began by discussing the properties of circulating tumor DNA (ctDNA). In individuals with cancer, a fraction of cell-free DNA (cfDNA) is tumor-derived, known as ctDNA. CtDNA is slightly smaller than cfDNA and can be distinguished by features absent in normal blood cfDNA and germline DNA, such as mutations, structural rearrangements, and lineage-specific methylation marks. CtDNA analysis also utilizes lineage-specific nucleosome patterns (fragmentomics). The percentage of ctDNA correlates with the volume of active and proliferative cancer and is rapidly suppressed by effective anticancer therapy, though its levels can be influenced by cancer type and subtype.
New technology is advancing the long-standing concept of the "liquid biopsy," a promise first recognized over a century ago. The earliest documented evidence of solid tumor material in peripheral blood was recorded in Melbourne, Australia, in 1869.
Dr. Wyatt focused on cell-free DNA (cfDNA) as a well-established assay, emphasizing its role as a practical and minimally invasive biomarker. cfDNA in blood plasma is an established analyte that does not require specialized purification protocols. It is distinct from circulating tumor DNA (ctDNA) and circulating tumor cells (CTCs). Interestingly, the amount of ctDNA is highly influenced by cancer type and volume. For example, neuroendocrine tumors and small cell variants tend to have higher ctDNA levels in the blood compared to prostate and urothelial cancers.
Apoptosing cells shed cfDNA into the bloodstream, with most normal cfDNA originating from the blood lineage. The median fragment length of cfDNA is 167 base pairs, matching the size of a nucleosome unit. An average person has approximately 5–10 ng of cfDNA per mL of plasma, equating to around 1,500 diploid genome copies, which is crucial for cfDNA assays. Proper collection protocols are essential to control pre-analytical variables and prevent additional white blood cell lysis. In cancer patients, a fraction of cfDNA is tumor-derived (ctDNA), which tends to be slightly smaller than cfDNA. ctDNA is identified by unique features absent in blood cfDNA and germline DNA, including DNA mutations, structural rearrangements, lineage-specific methylation marks, and nucleosome patterns (fragmentomics).
The percentage of ctDNA in the bloodstream is directly related to the volume of active and proliferative cancer and is rapidly suppressed by effective anti-cancer therapy. The extent of ctDNA shedding varies significantly by cancer type. Prostate and bladder cancers tend to shed more ctDNA due to their higher cell turnover rates, frequent genomic instability, and direct exposure to the bloodstream via the rich vasculature of the genitourinary tract. In contrast, kidney and testicular cancers shed relatively little ctDNA, likely due to lower baseline cell turnover, different tumor microenvironments, and less direct interaction with circulating blood.
In patients with advanced cancer, plasma cell-free DNA (cfDNA) is not only contributed by the tumor but also by various non-tumor sources, such as clonal hematopoiesis (CH). CH, a condition in which hematopoietic stem cells acquire somatic mutations leading to the expansion of clonal populations of blood cells, can lead to the presence of CH-related variants in plasma cfDNA. These variants are pervasive in the elderly cancer population and can complicate the interpretation of cfDNA in cancer diagnostics, including those with urothelial cancer.
To minimize the influence of CH variants, it's recommended to sequence white blood cell DNA from the patient to subtract the non-tumor-related mutations associated with CH. This helps ensure that the tumor-derived ctDNA is accurately identified and analyz
ctDNA has potential utility across the spectrum of disease for cancer management by:
- Improving cancer screening: Using ctDNA detection to screen high-risk populations, such as those with hereditary cancer syndromes.
- Detecting minimal residual disease (MRD): Offering greater sensitivity than current clinical tools.
- Monitoring for response to therapy: Measuring ctDNA levels to track treatment response.
- Estimating cancer aggression: The amount of ctDNA correlates with tumor aggressiveness.
- Predicting treatment efficacy: Identifying alterations like ERBB2, FGFR3, and subclonal shifts to guide therapy and prognosis.
- Identifying resistance mechanisms: Detecting genetic changes that inform resistance.
- Characterizing tumor biology and evolution: Gaining insights into how tumors evolve over time.
Dr. Wyatt used the analogy of a fire to explain ctDNA testing, emphasizing that one test does not fit all situations. He compared the assay to a smoke detector, which simply needs to alert you to the presence of a fire (detection). However, for a more detailed understanding, we want to know the characteristics of the fire, what’s sustaining it, and what started it (characterization). This same approach is needed for ctDNA in every cancer type. He highlighted three variables that we should be thinking about when ordering tests for ctDNA.
To detect, for example, minimal residual disease (MRD), there are tumor-informed and tumor-naïve ctDNA assays.
Tumor-informed
- Uses information about the individual cancer from tissue
- Pros: Specific and established
- Cons: Time and cost
Tumor-naïve
- Relies on de novo detection of ctDNA features (e.g., mutations)
- Pros: Quick; can also screen
- Cons: Lower specificity
To improve sensitivity in ctDNA detection, the newest tumor-informed assays use more ctDNA features, such as additional mutations from tissue whole genome sequencing and ctDNA fragment features. In contrast, tumor-naïve assays focus on epigenomic features, like methylation marks, fragment features, and scanning the entire genome, to enhance sensitivity. Dr. Wyatt expressed optimism about the next generation of tumor-naïve tests, as these advancements offer greater input of ctDNA, which could significantly improve detection capabilities.
In the characterization setting, ctDNA is typically applied to advanced cancers to identify targetable or prognostic gene alterations, such as FGFR3, ERBB2, or MSI status. It can also help assess inter-cell phenotypes, such as the cell lineage-of-origin (e.g., neuroendocrine). Additionally, ctDNA characterization can reveal transcriptome pathway activity and is increasingly being used in new research. It also provides insights into resistance mechanisms and new tumor biology, offering a view of cancer evolution across serial blood collections.
When looking for genomic alterations with ctDNA, deep targeted sequencing across large (tumor-naïve) gene panels is used, though it is less sensitive than detection tests, especially since ctDNA% is typically higher in advanced disease.
There are three key variables to consider when using ctDNA for genotyping:
- ctDNA fraction: A low ctDNA% can lead to inconclusive results, making it important to ensure an adequate amount of ctDNA is present.
- The second variable to consider is alterations beyond mutations. Tests may not report copy number and structural variants, so it is essential to be aware of the limitations of the test and the potential for false negatives. Deletions and copy number alterations can be important but may be missed if not specifically tested for.
- Another important factor is clonal hematopoiesis (CH), which can be a confounder in plasma-only cfDNA tests. This can be difficult to resolve without matched white blood cell sequencing. CH-variants can also fall in bladder cancer-relevant genes, such as TP53 and ATM, so it's crucial to consider these when interpreting ctDNA results.
Dr. Wyatt presented unpublished data from his group, including a study of 300 patients with metastatic renal cell carcinoma. The study involved synchronous plasma ctDNA and clonal hematopoiesis profiling, accounting for TP53. Around 50% of the variants fell into both the ctDNA and clonal hematopoiesis compartments, highlighting how easily the origin of these variants can be confused. This underscores the importance of carefully distinguishing between tumor-derived and clonal hematopoiesis-derived variants when analyzing ctDNA.
Several emerging approaches for indirectly measuring tumor transcriptional activity through ctDNA have been developed, as ctDNA carries the epigenomic modifications and features of the cancer cell. Three notable approaches to be aware of include:
- DNA methylation marks on ctDNA, which reflect the methylation features of the cancer it originated from.
- Fragment information of ctDNA, which helps understand the positioning of nucleosomes in the cell and can indicate whether a specific gene is highly expressed.
- ChIP-seq directly for plasma to identify histone modifications of the chromatin, providing insights into the transcriptional activity of the tumor.
All these approaches seem to indicate CDNA cell-of-origin and identify small cell cancers.
There have also been interesting signals from non-cancer cell-free DNA that could open up opportunities to leverage non-cancer signals. For example, plasma ChIP-seq was recently used to identify a high GI tract signal (H3K4me2) in plasma cfChIP-seq data from a patient with metastatic castration-resistant prostate cancer (mCRPC). Notably, the patient had rectal invasion from his prostate cancer. This observation led to the conclusion that epigenomic alterations can induce normal tissue destruction. In the future it is possible to understand the site of metastasis by looking at ctDNA.
In the context of ctDNA there are several challenges to consider. The technology is still in its early stages and is immature compared to genomic tests. While it shows promise as a biology probe, its clinical utility remains to be determined. There are no head-to-head comparisons between platforms, and the sensitivity and lower limits of detection are unknown and will likely vary by factors such as phenotype, background signal in blood, the number of 'targets' interrogated, and the ctDNA% (with lower ctDNA% leading to reduced signal). Additionally, unpredictable patient-specific biology and non-tumor cell confounders complicate the interpretation. A key question remains: why do some cancers shed ctDNA more effectively than others?
Urine tumor DNA (utDNA) shows clear promise for improving the diagnostics of urothelial cancer (UC). It contains a mix of both ctDNA and cellular DNA, including urothelial cells, which is relevant for UC detection. However, lessons from plasma technology are also important for urine DNA analysis. The specificity of somatic mutations in urine DNA remains less clear, due to the challenge of field cancerization. There is still an unmet need in this area, as there is no consensus on how urine should be collected and processed, what preservatives and volumes are required, or the optimal methods for DNA extraction. Unlike blood, the standardization of urine-based testing for UC remains a work in progress.
Dr. Wyatt concluded his presentation with the following key take-home messages:
- One size does not fit all when selecting a ctDNA test.
- ctDNA tests have an emerging impact across oncology.
- Test design varies based on the clinical goal: DETECT or CHARACTERIZE cancer.
- Sensitivity and specificity remain sub-optimal in certain contexts.
- New epigenomic technologies are enhancing the sensitivity of detection tests and enabling ctDNA lineage phenotyping.
- Choose your tests wisely and be aware of the limitations of the technology.
Presented by: Alexander William Wyatt, PhD, Vancouver Prostate Centre, University of British Columbia. Vancouver, BC, Canada.
Written by: Julian Chavarriaga, MD – Urologic Oncologist at Cancer Treatment and Research Center (CTIC) via Society of Urologic Oncology (SUO) Fellow at The University of Toronto. @chavarriagaj on Twitter during the 2025 Genitourinary (GU) American Society of Clinical Oncology (ASCO) Annual Meeting, San Francisco, CA, Thurs, Feb 13 – Sat, Feb 15, 2025.