Stejskal, Goodarzi, Srovnal, Hajdúch, van ‘t Veer, & Magbanua (2023) — Circulating Tumor Nucleic Acids: Biology, Release Mechanisms, and Clinical Relevance

Bibliographic Reference

Stejskal, P., Goodarzi, H., Srovnal, J., Hajdúch, M., van ‘t Veer, L. J., & Magbanua, M. J. M. (2023). Circulating tumor nucleic acids: Biology, release mechanisms, and clinical relevance. Molecular Cancer, 22, 15. https://doi.org/10.1186/s12943-022-01710-w

Core Argument

This review consolidates current knowledge on the biology of circulating tumor nucleic acids (ctNAs) — both DNA (ctDNA) and RNA (ctRNA) — with a focus on the mechanisms by which they are released from tumor cells into circulation, their degradation and clearance, and how these processes shape their clinical utility as liquid biopsy biomarkers. The central thesis is that understanding the biology of ctNAs — not just their presence or absence — is a prerequisite for interpreting liquid biopsy data, distinguishing treatment-responsive from resistant cell populations, and moving ctNA assays into routine clinical practice.

The paper systematically addresses two major knowledge gaps. First, why some tumors shed high amounts of ctNAs while others have undetectable levels — linking shedding heterogeneity to differences in cell death rates (apoptosis vs. necrosis), active secretion via extracellular vesicles (EVs) and protein complexes, and tumor-intrinsic molecular features (e.g., cell-cycle gene expression, subclonal driver mutation status). Second, the contrast between ctDNA (well-studied, primarily genetic information, short half-life of minutes to 1-2 hours) and ctRNA (poorly understood, requires active secretion for stability, reflects dynamic transcriptional and post-transcriptional processes). The review argues that ctRNA provides complementary information to ctDNA — expression signatures that ctDNA alone cannot capture — and that a multi-marker approach combining both analytes may be necessary for comprehensive liquid biopsy.

The review is structured as a narrative synthesis organized around release mechanisms (passive: apoptosis, necrosis, CTC breakage, micronuclei/double minutes; active: exosomes, microvesicles, protein complexes including AGO2 and HDL), modulating factors (radiation, senescence, hypoxia, paracrine signaling, molecular determinants), properties of released molecules (fragment size, half-life, protection from degradation), and clinical applications (screening, MRD detection, tumor characterization, treatment monitoring, resistance detection). Throughout, the authors emphasize that the field lacks standardization — preanalytical, analytical, and biological — and that persisting technical challenges limit clinical implementation despite growing evidence of clinical validity.

Methods

This is a narrative review article. No systematic review methodology, formal literature search strategy, or meta-analytic framework is reported. The authors synthesize published findings from primary research articles and earlier reviews across the following domains: cell death biology (apoptosis, necrosis, necroptosis, ferroptosis, pyroptosis, parthanatos, oncosis), extracellular vesicle biology (exosomes, microvesicles, apoptotic bodies), nucleic acid biochemistry (fragment size analysis, nuclease degradation, protein complex protection), and clinical ctDNA/ctRNA assay development. The review covers approximately 212 references spanning foundational work on cfDNA (e.g., Snyder et al., 2016; Mouliere et al., 2018) through recent clinical studies (e.g., Abbosh et al., 2017; Powles et al., 2021). No original data are presented.

Key Findings

1. “Apoptotic cell-derived cell-free DNA possesses a ladder-like pattern profile” with a peak fragment size of 167 bp, corresponding to the length of DNA around one nucleosome (147 bp) plus linker DNA (20 bp). This characteristic results from caspase-activated DNase (CAD) and other nucleases fragmenting DNA at internucleosomal regions not protected by histones. Necrosis, by contrast, produces larger fragments (up to many kbp) that are digested by macrophages during phagocytosis, generating the residual ctDNA. These size differences are diagnostically useful: shorter fragments (< 145 bp) enrich for ctDNA, while longer fragments indicate a high necrotic rate.

2. “The estimated half-life of ctDNA in the circulation ranges from minutes to 1-2 hours”, making ctDNA analysis a real-time reflection of disease status. Clearance occurs through organ uptake (primarily liver and spleen, minimally kidneys), nuclease degradation, and association with protective complexes (EVs, protein complexes, nucleosomes) that modulate degradation rates. The short half-life means that ctDNA levels reflect current rather than historical tumor burden — a critical property for monitoring treatment response and detecting minimal residual disease.

3. “Active secretion is an important release mechanism for ctNAs, especially for ctRNAs, which are rapidly degraded when passively released.” Active release occurs via extracellular vesicles (exosomes, 30-150 nm; microvesicles, 100-1000 nm; and apoptotic bodies) and macromolecular protein complexes (e.g., AGO2, NUCLEOPHOSMIN 1) and high-density lipoproteins (HDL). EV-associated ctNAs are protected from nuclease degradation, and their cargo (specific mRNAs, miRNAs, and other ncRNAs) is selectively sorted, reflecting the cell of origin and its physiological state. Exosomal ctDNA has been observed to represent the whole genome, while specific ncRNAs in exosomes show regulated sorting via RNA-binding proteins (AGO2, YB-1, hnRNPA2B1). Large EVs contain higher amounts of tumor-derived DNA than smaller EVs.

4. “CtDNA-based liquid biopsy can provide minimally invasive and real-time assessment of tumor heterogeneity and treatment response as it exploits ctDNA released from different tumor subclones that might not be examined by locally limited tissue biopsy.” Mutations found in ctDNA show concordance of up to 90% with matched solid tumors. ctDNA can capture heterogeneity across multiple tumor subclones simultaneously — a key advantage over single-site tissue biopsy. Studies have demonstrated the feasibility of tracking tumor evolution dynamics using ctDNA (e.g., Abbosh et al., 2017 in TRACERx lung cancer), and ctDNA fragmentomics (size profiling, nucleosome footprinting) and epigenetic features (methylation patterns) are emerging as additional informative layers beyond mutation detection.

5. “The proportion of ctDNA within the cfDNA varies, ranging from 0.003 to 95%”, with levels as low as up to 1% in early-stage tumors and potentially exceeding 10-40% in advanced disease. This enormous dynamic range — spanning four orders of magnitude — underpins both the promise and the challenge of ctDNA analysis. High inter-individual variability across patients with the same cancer type and across different cancer types renders ctDNA evaluation challenging. Brain, renal, prostate, and thyroid cancers show ctDNA detection rates under 50% even in advanced disease, while pancreatic, ovarian, colorectal, and breast cancers often exceed 75%. Factors beyond tumor size — particularly necrosis rate, proliferation index, and molecular subtype — better predict ctDNA levels than physical tumor dimensions alone.

Concepts Introduced or Used

• ctDNA (circulating tumor DNA) — Tumor-derived DNA fragments in circulation; carries genetic alterations (SNVs, CNAs, rearrangements) that provide specific markers for detection. Shorter than total cfDNA (< 145 bp). Released both passively (apoptosis, necrosis) and actively (EVs). Half-life of minutes to 1-2 hours enables real-time monitoring.

• ctRNA (circulating tumor RNA) — Tumor-derived RNA molecules in circulation, including mRNA, miRNA, and long ncRNAs. Requires active secretion via EVs or protein complexes for stability — passively released RNA is rapidly degraded. Reflects dynamic transcriptional and post-transcriptional processes not captured by ctDNA.

• cfDNA (cell-free DNA) — Total DNA fragments in circulation, predominantly derived from hematopoietic cell turnover via apoptosis. The ctDNA fraction is the tumor-derived subset. Background cfDNA from non-tumor cells creates the signal-to-noise challenge for ctDNA detection.

• Extracellular vesicles (EVs) — Heterogeneous lipid-bound particles (exosomes, microvesicles, apoptotic bodies) that carry and protect ctNAs from degradation. Exosomes (30-150 nm) are endosomally derived and contain selectively sorted cargo; microvesicles (100-1000 nm) bud from the plasma membrane and require ARF6 for formation. EVs are key mediators of intercellular communication in cancer, transferring oncogenic cargo, and their levels correlate with tumor invasiveness. See extracellular-vesicles.

• Liquid biopsy — Non-invasive sampling of blood or other body fluids for detection of tumor-derived material (ctDNA, ctRNA, CTCs). Enables serial monitoring impossible with tissue biopsy. See liquid-biopsy.

• Fragmentomics — Analysis of cfDNA fragment size profiles, end motifs, and nucleosome footprints to infer tissue of origin and detect tumor-derived fragments. Shorter fragment size enriches for ctDNA; tissue-specific nucleosome wrapping patterns can identify the source tissue.

• Clonal hematopoiesis (CH) — Age-related accumulation of somatic mutations in hematopoietic stem cells. A major source of false-positive ctDNA detection — CH-related mutations (e.g., in DNMT3A, TET2, ASXL1) can appear in cfDNA and be mistaken for tumor-derived alterations. See clonal-hematopoiesis.

• Apoptosis-induced proliferation (AIP) — Compensatory proliferation triggered by caspase-dependent release of mitogenic signals from apoptotic cells. A mechanism by which cell death can paradoxically stimulate tumor growth, potentially contributing to therapy resistance. EVs from apoptotic cells contain proliferative miRNAs, mRNAs, and long ncRNAs.

• Minimal residual disease (MRD) — Residual tumor cells below the detection limit of imaging after treatment. ctDNA analysis is emerging as a promising MRD detection modality, with serial ctDNA monitoring enabling earlier detection of impending relapse than conventional imaging. See minimal-residual-disease.

Entities Referenced

• MALAT1 — Metastasis-associated lung adenocarcinoma transcript 1; a long ncRNA detected in blood of NSCLC patients (96% specificity). Functions as a scaffold in gene and splicing regulation. See malat1.

• HOTAIR — HOX antisense intergenic RNA; a long ncRNA whose increased blood levels in colorectal cancer patients positively correlated with higher mortality.

• GIHCG — Gradually increased during hepatocarcinogenesis; a long ncRNA whose serum levels in renal cell carcinoma had 84.8% specificity and 80.7% sensitivity.

• H19, HOXA-AS2 — Long ncRNAs implicated in miRNA regulation; associated with proliferation, cell cycle progression, and cell migration in several cancer types.

• ARF6 (ADP-ribosylation factor 6) — GTPase involved in the formation and shedding of microvesicles and selective integration of their cargo. ARF6 expression associated with tumor invasiveness and detected on MVs from tumor cell lines.

• AGO2, YB-1, hnRNPA2B1 — RNA-binding proteins that mediate the selective sorting of miRNAs and other RNAs into exosomes. AGO2 also forms circulating protein complexes that protect miRNAs from degradation independent of EVs.

• CAD (caspase-activated DNase) — Nuclease that executes internucleosomal DNA fragmentation during apoptosis, producing the characteristic 167 bp ladder pattern of cfDNA.

• Double minutes (DMs) — Extrachromosomal circular DNA fragments containing amplified oncogenes. Can exit the nucleus by budding and be extruded from cells as micronuclei, contributing to ctDNA release. See double-minutes.

• FoundationOne Liquid CDx, Guardant360 CDx, Cobas EGFR Mutation Test v2 — FDA-approved ctDNA-based diagnostic tests for multiple cancer types. FoundationOne Liquid CDx and Guardant360 CDx use NGS-based panel sequencing; Cobas EGFR Mutation Test v2 uses RT-PCR for EGFR mutations in NSCLC.

• TRACERx — Prospective multi-region sequencing study referenced for demonstrating phylogenetic ctDNA analysis to track early-stage lung cancer evolution (Abbosh et al., 2017). See TRACERx.

Limitations (as stated by authors)

• “Mechanisms involved in ctNA release need to be better understood. There are considerable gaps in our knowledge regarding the presence, fluctuations, and characteristics of ctNAs and their potential roles in tumor resistance and evolution.” The exact contribution of different cell death types (apoptosis vs. necrosis vs. necroptosis, ferroptosis, pyroptosis, parthanatos, oncosis) to the ctDNA pool is unknown. Systematic investigations of active and passive ctNA release mechanisms have yet to be fully described.

• “The exact proportions of NAs released via different types of cell death are unknown” — apoptosis and necrosis are considered major contributors, but their relative contribution remains unquantified. The rates at which different cell death types occur and contribute to ctDNA shedding are difficult to estimate because cell death pathways are molecularly interconnected with significant crosstalk.

• Preanalytical and analytical standardization is lacking. The authors note that the ASCO/CAP joint review (Merker et al., 2018) “pointed to the lack of evidence of clinical validity and utility of the majority of ctDNA assays outside of a clinical trial.” They further state: “the lack of standardization of CTC detection methods, as well as the high false-negative rate of ctDNA assays, points to the need for further technological advancements.”

• “False positivity may arise from tumor heterogeneity but is more likely from clonal hematopoiesis and detection of somatic alterations in DNA released by normal blood cells.” The predominance of cfDNA over ctDNA, the release of cfDNA by hematopoietic cells, and the partial overlap of genes mutated in clonal hematopoiesis with tumor drivers “can significantly increase the risk of false-positive ctDNA detection and limit copy number alteration detection.”

• “CtRNA has been studied less than ctDNA, and the isolation of RNA subpopulations derived from EVs and lipoprotein complexes remains a current technical challenge.” The heterogeneity and overlapping sizes and densities of EVs (exosomes, microvesicles, apoptotic bodies, lipoproteins) hinder their selective separation and characterization, confounding profiling experiments.

• ctDNA levels do not consistently correlate with tumor burden. The authors note: “data also show inter-individual variability among patients with the same cancer and across different cancer types rendering ctDNA evaluation challenging,” and “ctDNA detection was under 50% in primary brain, renal, prostate, and thyroid cancers” even when >75% of other advanced cancers had detectable ctDNA.

• Temporal variability in ctDNA shedding is acknowledged: “the impact of the timing of specimen collection on ctDNA analysis must be considered since ctDNA shedding fluctuates over time.”

Relevance to Clonal Evolution

This review provides the biological foundation for understanding why ctDNA works as a clonal evolution measurement modality, which is directly relevant to the wiki’s empirical test design for the compression-entrenchment hypothesis (docs/superpowers/specs/2026-07-05-ith-outcome-test-design.md). Several connections are critical:

ctDNA captures multi-clonal information that tissue biopsy misses. Because ctDNA is shed from all tumor subclones — not just the one sampled by a needle biopsy — it captures a systemic view of intratumor heterogeneity. The authors note that ctDNA exploits “ctDNA released from different tumor subclones that might not be examined by locally limited tissue biopsy,” and that ctDNA “may reflect systemic disease.” This is the same rationale underlying the compression-entrenchment test design’s use of single-biopsy ITH as a proxy for whole-tumor diversity: ctDNA avoids the spatial sampling problem entirely, though it introduces its own biases (differential shedding, cfDNA background).

Sample type considerations for the ITH test design. The test design lists “Sample type: FFPE vs. fresh vs. cfDNA” as a required confounder variable. This review clarifies why cfDNA is a fundamentally different measurement modality from tissue-based DNA: (i) ctDNA is generally shorter (<145 bp) than tissue-derived DNA, biasing detection toward shorter fragments and altering VAF distributions; (ii) ctDNA has a half-life of minutes to 1-2 hours, providing a real-time snapshot rather than the historical record captured by FFPE; (iii) ctDNA detection sensitivity depends on tumor-specific shedding rates, which vary by cancer type (under 50% in brain, renal, prostate, thyroid; >75% in many others) and by molecular subtype (higher in triple-negative breast cancer than other breast subtypes); (iv) plasma is preferred over serum for ctDNA analysis because clotting-induced cell lysis in serum increases high-molecular-weight cfDNA background.

ctDNA fragmentomics as an ITH surrogate. The review describes how cfDNA fragment size profiling can distinguish tumor-derived from non-tumor-derived DNA, and how tissue-specific nucleosome wrapping patterns can identify the tissue of origin. The size profile of cfDNA — shorter fragments in cancer patients, correlation between fragment shortening and poor survival — provides an orthogonal dimension to mutation-based ITH measurement. This is relevant to the fragmentomic analysis mentioned in the test design’s future perspectives.

Tumor evolution tracking via ctDNA. The review cites Abbosh et al. (2017) demonstrating phylogenetic ctDNA analysis in early-stage lung cancer (TRACERx) and describes how ctDNA can track clonal evolution dynamics in real time — detecting emergent resistance mutations, monitoring MRD, and distinguishing treatment-responsive from resistant cell populations. The short half-life means that ctDNA levels reflect current tumor status, not the cumulative evolutionary history captured by tissue. This temporal resolution is both a strength (real-time monitoring) and a limitation (single time points may miss clonal dynamics that unfold between sampling intervals).

False positives from clonal hematopoiesis. The review identifies clonal hematopoiesis as a major confounder in ctDNA analysis — mutations in CH-related genes (DNMT3A, TET2, ASXL1) can be misattributed to tumor-derived ctDNA. This confound is directly relevant to any ITH measurement from cfDNA: without matched normal (blood) sequencing, CH variants inflate apparent tumor mutation burden and alter the VAF distribution used for clonality inference. The test design correctly requires “matched normal (blood) filtering status” as a confounder variable.

ctRNA as a complementary evolutionary signal. While ctDNA captures genetic alterations, ctRNA reflects dynamic cellular processes — gene expression programs, signaling pathway activation, phenotypic state — that change during clonal evolution without sequence change. The review positions ctRNA as providing “information gaps from ctDNA analysis alone” by reflecting “the expression signature of tumor cells” and “the tumor microenvironment and evolution.” This connects to the dual-regime-evolution framework: ctDNA captures the Darwinian (genetic) regime, while ctRNA captures the non-Darwinian (transcriptional/epigenetic) regime of cancer evolution.