Q&A Report: Epigenomics at 360° ​in Liquid Biopsy​: Approaches and Case Studies for Biomarker Discovery and Drug Development​

Matteo Tosolini, PhD
Biopharma Business Development Manager
EMEA
Hologic Diagenode

Jessica Apulei, PhD
Scientific Liaison Manager
EMEA and Life Sciences
Hologic Diagenode

Plasma input requirements depend on the biomarker type. cfDNA methylation analyses typically require higher input (~ 2-5 ml plasma coming from respectively 5-10 ml whole blood). In contrast, optimized workflows for cfRNA and circulating nucleosome assays can operate with lower volumes (down to few hundred µl). This highlights the distinct input demand of DNA, RNA and chromatin-based analyses.

Pre-analytical variables are critical determinants of data quality. Key factors include blood collection tubes (e.g., EDTA vs. stabilization tubes), time to plasma separation, centrifugation protocols, and storage conditions. Delays or suboptimal handling can lead to cell lysis, background DNA contamination, and nucleic acid degradation. Standardization of these parameters is essential for reproducibility and accurate downstream analysis.

Early detection requires highly controlled pre-analytical workflows due to low biomarker abundance. Stabilization tubes are recommended to preserve analytes during transport, allowing shipment at room temperature within validated timeframes. In contrast, EDTA samples require rapid processing and cooled condition. Proper tube selection, timing, and storage are critical to maintain sensitivity and data quality.

The acceptable time depends on the collection tube used. For standard K2-EDTA tubes, plasma should be separated within 2-4 hours of blood draw to minimize cell lysis and background contamination. In contrast, nucleic acid stabilization tubes (e.g., Streck) allow extended stability, enabling processing within 24-72 hours at RT. Strict adherence to these timelines is essential to preserve cfDNA, cfRNA, and nucleosome integrity.

Plasma is strongly preferred over serum for liquid biopsy applications. During serum preparation, coagulation leads to cells lysis and release of genomic DNA, significantly increasing background and reducing assay sensitivity. Plasma, when properly processed, better preserves circulating biomarkers such as cfDNA, cfRNA, and nucleosomes, ensuring more reliable and reproducible results.

For DNA methylation analysis on plasma, we strongly recommend using Streck tubes. The ideal amount of cfDNA is 10–20 ng, which typically corresponds to approximately 4–5 mL of plasma. However, we can reliably work with as little as 5 ng, which is usually obtained from around 2 mL of plasma.

For cfRNA, on the other hand, we recommend K2 EDTA plasma. From 0.2–0.5 mL of platelet-poor or platelet-free plasma (PPP or PFP), we typically obtain between 200–1200 pg of cfRNA.

While cfDNA is known for its relative stability in circulation, histone modifications on circulating nucleosomes—such as acetylation—are generally more dynamic by nature. However, when nucleosomes are released into the bloodstream, they are protected by their association with DNA and histone proteins, which helps preserve key epigenetic marks to a meaningful extent.

Pre-analytical conditions (e.g., blood collection tubes, processing time, and storage) play a critical role in maintaining the integrity of these modifications. When samples are properly handled and stabilized, histone marks, including acetylation, can remain sufficiently preserved to enable reliable downstream analyses such as chromatin profiling and biomarker discovery.

We have previous experience working with CSF for DNA methylation studies. However, the main limitation is the sample input, both the volume (mL) that can be collected per patient and the corresponding DNA yield (ng).

While CSF is certainly the most proximal biofluid for neurodegenerative diseases, it is also possible to use other types of liquid biopsy, such as plasma, and computationally extract a neuro-specific signal from cfDNA fragments between those originating from different tissues across the body. We do not yet have a published case study specifically on neurodegenerative diseases, but this is not due to any limitation of liquid biopsy or epigenomic approaches in this area. On the contrary, we see strong potential for their application in this field, and we would be happy to explore how these approaches could support your biomarker strategy in neurodegeneration.

Sample choice depends on the biological context and assay. Plasma is the most established and robust source, supported by standardized workflows and broad applicability across indications. Urine and saliva provide non-invasive alternatives but often yield lower and more variable nucleic acid content, which may impact sensitivity. They are most relevant for tumor types in direct contact with these biofluids.

Both analyses can be performed from the same blood-draw but require separate processing workflows. Plasma is typically divided into aliquots to preserve material integrity, as nucleosome profiling and cfDNA analyses have distinct extraction and handling requirements.

Nu.Q® nucleosome immunoprecipitation demonstrates high and reproducible enrichment efficiency, with reported depletion of circulating nucleosomes from plasma often exceeding ~80% (Van den Ackerveken, et al. 2021). The workflow enables quantitative and proportional recovery relative to input levels, supporting robust downstream analyses such as histone PTM profiling and biomarker discovery.

Tumor stage can influence detectability. Early-stage cancers typically release lower levels of circulating biomarkers (such as cfDNA or cfRNA), which can make detection more challenging. In contrast, later-stage tumors generally shed higher amounts of these biomarkers, resulting in stronger and more easily detectable signals. Advanced, high-sensitivity approaches, such as those implemented at Hologic Diagenode via the epigenomic layer, are specifically designed to improve detection even in early-stage diseases, where signals are more limited.

Comprehensive cfRNA profiling beyond miRNA is still relatively new, and many aspects remain to be fully explored. While we do not yet have internal data specifically addressing this point, it is plausible that increasing tumor burden could lead to a stronger cfRNA signal due to a higher abundance of tumor-derived RNA fragments circulating in plasma (somewhat analogous to what is observed with cfDNA, where the ctDNA fraction often increases with tumor progression).

At Hologic Diagenode, we support several biomarker discovery programs in liquid biopsy, to help identify clinically relevant signatures in oncology and other disease areas. We typically begin with pilot studies designed to address a specific biological or clinical hypothesis and generate the data needed to guide larger discoveries or validation programs.

Liquid biopsy enables sensitive detection of MRD through analysis of ctDNA, cfRNA, or other tumor-derived biomarkers. Because MRD signals are typically present at low levels, assays must achieve high sensitivity and specificity.

Reliable detection depends on optimized pre-analytical workflows, sufficient plasma input, and longitudinal sampling. Clinical utility requires robust assay performance and careful interpretation of low-frequency signals.

Liquid biopsy is suited for longitudinal monitoring because it allows repeated, minimally invasive sampling to track disease dynamics, treatment response, or relapse. Key considerations include standardizing pre-analytical handling (tube type, processing time, storage) and maintaining consistent assay protocols across time points to minimize technical variability. Sufficient plasma volume and optimized sensitivity are critical to detect subtle changes, especially in minimal residual disease settings. Additionally, interpreting longitudinal data requires understanding baseline fluctuations and establishing meaningful thresholds for clinical decision-making.

It is challenging to define a threshold as the likelihood of success often depends on several factors, such as the clinical context (e.g., colorectal vs. breast cancer), the intended application of the assay (treatment monitoring vs. early detection), and the molecular layer that may be most informative for the project (chromatin profiling, DNA methylation, or RNA-based approaches).

At Hologic Diagenode, we have been supporting several biomarker discovery programs using our multi-omic epigenomic approaches in liquid biopsy. Would you like to know more about it? Please visit our service page.

Integrating multiple biomarker types into a single liquid biopsy provides a more comprehensive view of the disease state. While cfDNA can reveal mutations and methylation patterns, cfRNA provides information on gene expression, and nucleosome profiles can reflect chromatin organization and epigenetic changes. Combining these layers of information can improve sensitivity and specificity, especially in early detection or minimal residual disease monitoring, by capturing signals that might be missed when relying on a single biomarker type. This multi-omics approach also allows better insight into tumor heterogeneity, disease progression, and treatment response, ultimately supporting more personalized clinical decisions.

Nucleosome-based assays (e.g., Nu.Q® Discover) offer a robust, scalable, and cost-effective solution for profiling chromatin features in liquid biopsy samples. They require low input material, and provide fast turnaround times, making them well-suited for clinical and high-throughput applications.

However, these methods provide a more global view of chromatin organization and lack locus-specific resolution, limiting their ability to precisely map histone modifications or discover novel regulatory regions.

In contrast, sequencing-based approaches such as cfChIP-seq deliver high-resolution, genome-wide information on histone modifications, enabling deeper biological insights. This comes at the cost of greater technical complexity, higher sample input requirements, longer turnaround times, and increased variability. In summary, nucleosome-based assays are optimized for robustness and clinical scalability, while cfChIP-seq is better suited for detailed epigenomic research and discovery.

Hologic Diagenode’s technologies are designed from scalability from exploratory studies to large clinical trials. Our workflows support low-input samples, high reproducibility, and standardized processing across cohorts. This enables generation of robust, clinically relevant datasets suitable for translational research, patient stratification, and regulatory applications, with flexible service models covering sample processing through data analysis.