The review, led by Chimbabantu Kaoma and colleagues at the Nuclear Medicine Research Infrastructure and the University of Pretoria in South Africa, synthesises evidence on circulating tumour DNA, or ctDNA, across the four major treatment classes in mCRPC: androgen receptor pathway inhibitors, taxane chemotherapy, PARP inhibitors, and PSMA-targeted radioligand therapy. The authors’ central message is nuanced. ctDNA is undeniably informative, but most of its proven value today is prognostic rather than predictive, meaning it tells clinicians how aggressive a patient’s disease is rather than which specific drug will work best.

The biological logic behind ctDNA is elegant. When tumour cells die, through apoptosis, necrosis, or therapy-induced destruction, they release fragments of their DNA into the bloodstream. This tumour-derived DNA mixes with cell-free DNA shed by normal cells, predominantly from white blood cells and their precursors. In healthy individuals, cell-free DNA consists mainly of short fragments around 40 to 200 base pairs. In patients with metastatic cancer, a variable proportion of this circulating DNA originates from the tumour itself, and that proportion, known as the tumour fraction, can range from less than 1 percent to more than 50 percent in advanced mCRPC. Because ctDNA has a half-life of roughly one to two hours, cleared rapidly by the liver, kidneys, nucleases, and macrophages, its levels reflect tumour dynamics almost in real time.

What makes ctDNA particularly powerful in mCRPC is that it captures the tumour’s evolutionary story. Prostate cancer under therapeutic pressure is a textbook example of Darwinian selection. Androgen receptor signalling remains the dominant oncogenic driver even after castration, and tumours accumulate gain-of-function mutations in the androgen receptor’s ligand-binding domain, such as T878A, H875Y, L702H, and F877L, that broaden ligand specificity or even convert drugs like enzalutamide from antagonists into agonists. The F877L mutation, for instance, can flip enzalutamide and apalutamide into activators of the very receptor they were designed to block, while L702H allows glucocorticoids given during treatment to activate the receptor. Serial ctDNA sampling can track these resistant subclones as they emerge, something a single archival tissue biopsy taken years earlier simply cannot do.

On the technical side, next-generation sequencing has transformed what is possible. Targeted hybrid-capture panels, amplicon-based deep sequencing, droplet digital PCR for hotspot mutations, and low-pass whole-genome sequencing for tumour fraction estimation can now detect variant allele frequencies below 0.1 percent under optimised conditions, thanks to molecular barcoding and error-suppression strategies. But the review is careful to distinguish analytical sensitivity from biological sensitivity. A technically exquisite assay still fails when the tumour sheds little DNA, as is common in bone-predominant disease, or when the relevant resistant clone sits below the assay’s limit of detection. A negative ctDNA result, the authors stress, never excludes active disease.

One of the most important confounders the review highlights is clonal haematopoiesis of indeterminate potential, or CHIP. As people age, some haematopoietic stem cells acquire somatic mutations in genes like TP53, ASXL1, DNMT3A, and TET2, and these non-malignant clones release mutated DNA into the plasma that can be mistaken for tumour-derived signal. Matched white blood cell sequencing is therefore essential to filter out this noise. Pre-analytical variability, differences in tube type, processing time, and storage conditions, plus a lack of harmonisation across platforms, further complicate interpretation and currently prevent universal thresholds from being applied in the clinic.

Despite these caveats, the prognostic evidence for ctDNA tumour fraction is remarkably consistent across every treatment class examined. In landmark trials of abiraterone and enzalutamide, ctDNA-positive patients had median overall survival reduced from 47.4 to 27.3 months and median radiographic progression-free survival reduced from 33.0 to 17.8 months, independent of PSA and other clinical variables. Each 10 percent increase in ctDNA aneuploidy fraction was associated with a proportional increase in the hazard of progression and death. Similar patterns emerged in the TALAPRO-2 trial of talazoparib plus enzalutamide, in the FIRSTANA and PROSELICA taxane trials, and in the PSMAfore and TheraP radioligand therapy studies, where baseline ctDNA fraction above 0.5 percent independently predicted inferior outcomes regardless of treatment assignment.

The one domain where ctDNA has achieved genuine predictive, treatment-defining status is in homologous recombination repair testing for PARP inhibitors. In the PROfound trial, ctDNA testing showed substantial concordance with tissue sequencing for BRCA1, BRCA2, and ATM alterations, and this evidence contributed directly to regulatory approval of olaparib and to the FDA clearance of FoundationOne Liquid CDx as a companion diagnostic. Beyond selection, ctDNA also monitors response in this setting: in TOPARP-A, a fall of at least 50 percent in cell-free DNA concentration at four to eight weeks was associated with improved survival, while rising ctDNA identified non-responders and revealed resistance mechanisms such as BRCA2 and PALB2 reversion mutations that restore the broken repair pathway.

For PSMA-targeted radioligand therapy with lutetium-177 PSMA-617, the review frames ctDNA as complementary to, not a replacement for, PSMA PET imaging. PSMA PET determines eligibility by confirming sufficient target expression across metastatic sites, while ctDNA refines molecular risk assessment and tracks clonal evolution. Exploratory analyses from TheraP suggest that ATM alterations may predict greater benefit from radioligand therapy over cabazitaxel, plausibly because impaired DNA repair enhances radiosensitivity, while AR amplification and alterations in TP53, RB1, and PTEN are enriched in poor responders, though these signals remain investigational. Early on-treatment ctDNA kinetics, measured at cycle 2 day 1 in PSMAfore, correlated more strongly with progression-free and overall survival than baseline levels and were largely independent of PSA, suggesting ctDNA can detect molecular response to radiation-induced DNA damage before anatomical changes become visible.

The review’s conclusion is measured but forward-looking. ctDNA is best positioned today as a complementary tool for risk stratification and disease monitoring, with tumour fraction serving as a pan-therapy prognostic biomarker and genomic alterations providing therapy-specific predictive information only in validated contexts such as HRR testing. Emerging technologies, including fragmentomics, DNA methylation profiling capable of detecting neuroendocrine transformation, multi-omic liquid biopsy integrating circulating tumour cells and extracellular vesicles, and machine learning approaches to integrate complex datasets, promise to expand ctDNA’s clinical reach. But the authors are clear that routine implementation will require prospective biomarker-driven trials, assay standardisation, population-diverse validation, and a demonstration that acting on ctDNA results genuinely improves patient outcomes. Until then, the blood test that reads a tumour’s evolutionary diary remains a powerful adjunct rather than a standalone decision-maker, one whose full potential in precision oncology is still being written.

Subject of Research: Circulating tumour DNA as a liquid biopsy biomarker for monitoring and prognostication in metastatic castration-resistant prostate cancer.

Article Title: Circulating tumour DNA in mCRPC: bridging tumour biology and real-time treatment monitoring

Article References: Kaoma, C., Sathekge, M., Chalwe, J., Mokoala, K., & Kgatle, M. (2026). Circulating tumour DNA in mCRPC: bridging tumour biology and real-time treatment monitoring. Frontiers in Molecular Medicine, 6, Article 1852641. https://doi.org/10.3389/fmmed.2026.1852641

Image Credits: AI Generated

DOI: 10.3389/fmmed.2026.1852641

Keywords: circulating tumour DNA, liquid biopsy, metastatic castration-resistant prostate cancer, androgen receptor pathway inhibitors, PARP inhibitors, PSMA radioligand therapy, tumour fraction, clonal haematopoiesis, clonal evolution, precision oncology, biomarker validation, next-generation sequencing

Clonal haematopoiesis occurs when a haematopoietic stem cell acquires a somatic mutation that gives it a survival or proliferation advantage, allowing it to outcompete normal stem cells and produce an expanding fraction of the body’s blood cells. The phenomenon becomes increasingly common with age, and its presence in blood DNA can be detected through sequencing long before any disease develops. Researchers have previously tied clonal haematopoiesis to heightened risks of blood cancers, heart attacks and strokes, but its role in infectious disease has remained far less clear. The new study, led by Eleanor Fewings of the Illumina Artificial Intelligence Laboratory together with collaborators at the Broad Institute, Massachusetts General Hospital, Vanderbilt University Medical Center and Memorial Sloan Kettering Cancer Center, set out to answer that question at unprecedented scale.

The team assembled genetic and clinical data from three population cohorts totalling 590,537 individuals, applying a multi-parameter somatic variant calling and filtering pipeline designed to detect even low-frequency mutations in blood-derived DNA with high confidence. Across 43 genes implicated in clonal haematopoiesis, they identified 19,822 mutations in 40,560 people, corresponding to a prevalence of 6.87 percent. To guard against spurious findings, the researchers then validated their results in a combined analysis encompassing 889,215 individuals, one of the largest somatic genetics investigations ever conducted. The sheer scale of the resource allowed them to dissect risk not just for clonal haematopoiesis as a whole, but for individual mutated genes, specific recurrent mutations and expanding clones defined by variant allele fraction, a measure of the proportion of blood cells carrying the mutation.

The central result was unambiguous: people with clonal haematopoiesis were significantly more likely to develop serious bacterial infection, with a hazard ratio of 1.22 and a p-value of 4.66 × 10⁻²⁹, an association so statistically strong that chance can be essentially excluded. The effect was independent of haematological malignancy, meaning it was not simply a case of undiagnosed cancer causing infections, and it remained robust after adjustment for inflammatory comorbidities and genetic ancestry. Perhaps most strikingly, clonal haematopoiesis was also associated with a 1.84-fold increase in mortality within 28 days of infection, indicating that the mutations do not merely raise the likelihood of becoming infected but also compromise the body’s ability to survive the encounter.

The study’s most clinically consequential contribution, however, lies in its demonstration that risk is strongly dose-dependent and mutation-specific. Among individuals carrying recurrent mutations in five particular genes, SRSF2, DNMT3A, SF3B1, IDH2 and NRAS, a group numbering 2,143 people, the absolute risks were substantial. For those with expanded SRSF2 P95 clones, a specific recurrent mutation in a splicing factor gene, between 22 and 25 percent developed a serious bacterial infection within ten years. That figure rivals the predictive power of well-established clinical risk factors and suggests that a single detectable point mutation in a blood sample can flag a person as being at markedly elevated infectious risk years in advance.

The mechanistic logic behind these associations is rooted in what the mutated genes actually do. DNMT3A encodes a DNA methyltransferase essential for normal stem cell differentiation, while SRSF2 and SF3B1 are splicing factors that govern how RNA transcripts are processed, and IDH2 and NRAS participate in metabolic and signalling pathways respectively. When these genes are disrupted, the resulting blood cells can behave abnormally in ways that impair innate and adaptive immunity, including altered production of inflammatory cytokines and compromised function of neutrophils and macrophages, the cellular workhorses of bacterial defence. The new epidemiological findings align with this biology, showing that the immune consequences of clonal haematopoiesis are not confined to cancer predisposition but extend to everyday host defence against bacteria.

Importantly, the associations held across diverse populations and were largely independent of the specific bacterial strain involved, suggesting a broad, generalised weakening of antibacterial immunity rather than vulnerability to any single pathogen. This universality strengthens the case that clonal haematopoiesis acts as a systemic modifier of immune competence. It also helps explain a long-standing clinical observation: older adults vary enormously in their susceptibility to sepsis and pneumonia, and much of that variation has remained unexplained by conventional risk factors such as diabetes, chronic kidney disease or immunosuppressive medication. Somatic mutations in the blood may account for a meaningful share of that unexplained risk.

The study was made possible by an extraordinary confluence of large-scale biobank resources, including the UK Biobank, Genomics England’s National Genomic Research Library and the BioVU biobank at Vanderbilt University Medical Center, with sequencing of 250,000 whole genomes funded by the Alliance for Genomic Discovery, a consortium of NashBio, Illumina and several pharmaceutical partners. This infrastructure reflects a broader shift in biomedical research toward population-scale sequencing, in which millions of routinely collected blood samples become windows into the somatic evolution occurring inside the body. What was once visible only through the lens of leukaemia genomes can now be quantified across entire populations, revealing subtle but consequential effects on common diseases.

The implications for clinical practice are potentially far-reaching. If clonal haematopoiesis screening were incorporated into routine care, individuals with high-risk clones, particularly expanded SRSF2 mutations, could be flagged for closer monitoring, earlier vaccination strategies, aggressive management of modifiable infection risks or a lower threshold for empirical antibiotic treatment when symptoms appear. The findings also raise the prospect of therapies that target the mutated clones themselves or their inflammatory downstream effects, an area of intense pharmaceutical interest given the established links between clonal haematopoiesis and cardiovascular disease. Several companies are already developing agents aimed at damping the inflammation driven by these clones, and infection risk may now join cardiovascular events and malignancy as an outcome worth tracking in such trials.

Caveats remain. The study is observational, and while the authors adjusted extensively for malignancy, comorbidities and ancestry, residual confounding cannot be fully excluded, and the precise molecular mechanisms connecting each mutation to impaired bacterial clearance will require experimental work in cellular and animal models. The published version is also an accepted manuscript shared early for rapid access and is subject to further editorial refinement. Nevertheless, the consistency of the dose-dependent signal across hundreds of thousands of people, multiple cohorts and diverse ancestries makes a compelling case that clonal haematopoiesis deserves recognition as a genuine risk factor for serious bacterial infection. For the millions of people who carry these silent mutations, the study transforms a hidden aspect of biological ageing into an actionable piece of medical information, and it signals that the era of somatic mutation screening for non-cancer diseases may be arriving faster than anyone anticipated.

Subject of Research: Association of clonal haematopoiesis mutations with risk of serious bacterial infection

Article Title: Clonal haematopoiesis subtypes confer dose-dependent risk of serious bacterial infection

Article References: Fewings, E., Yu, Z., Gutierrez, J. C., Truong, B., Fiziev, P. P., McRae, J., Schwartzentruber, J., Bick, A. G., Natarajan, P., Lareau, C. A., Ulirsch, J. C., & Farh, K. K.-H. (2026). Clonal haematopoiesis subtypes confer dose-dependent risk of serious bacterial infection. Nature Communications. https://doi.org/10.1038/s41467-026-77934-x

Image Credits: AI Generated

DOI: 10.1038/s41467-026-77934-x

Keywords: clonal haematopoiesis, somatic mutations, serious bacterial infection, SRSF2, DNMT3A, SF3B1, IDH2, NRAS, haematopoietic stem cells, sepsis, population genomics, immune ageing

The research team, led by Adeena Musheer and colleagues working across institutions in Pakistan and the United States, conducted the review according to PRISMA 2020 guidelines, the current international standard for systematic evidence synthesis. They searched MEDLINE, Embase, the Cochrane Central Register of Controlled Trials, and grey literature sources from database inception through December 2025. To be included, studies had to report myeloid neoplasms diagnosed after CAR-T therapy or ASCT and provide extractable data on at least one of three parameters: incidence, latency, or molecular characteristics. Because the included studies varied enormously in design, patient populations, and reporting conventions, the authors synthesized their findings descriptively rather than attempting a meta-analysis, a decision that reflects an honest assessment of the state of the evidence. Nine studies comprising 2,933 patients ultimately met the inclusion criteria, and within those cohorts a total of 195 therapy-related myeloid neoplasm cases were reported.

One of the most striking findings concerns latency, the interval between the initial therapy and the diagnosis of the secondary cancer. The review found that median latency varied widely across studies, but the pattern differed meaningfully between the two treatment modalities. Following CAR-T therapy, reported median latency ranged from just 3 to 78 months, a window that in some cohorts was remarkably short. In contrast, ASCT cohorts reported longer intervals, spanning from 2.4 to as much as 262.8 months, meaning that some secondary leukemias appeared more than two decades after transplantation. This difference in timing is biologically intriguing. CAR-T therapy itself is not a classical DNA-damaging agent in the way that alkylating chemotherapy or total body irradiation are, so the shorter latency observed after CAR-T likely reflects the fact that these patients arrive at the therapy with bone marrows already heavily scarred by prior cytotoxic exposure, with pre-existing mutant clones poised for outgrowth once the selective pressures of treatment remove competing normal hematopoiesis.

That mechanistic picture is supported by the molecular data the review assembled. Half of the included studies assessed clonal haematopoiesis, the age-related phenomenon in which blood cells descended from a single mutated stem cell progressively dominate the marrow. Clonal haematopoiesis is increasingly understood as the soil in which therapy-related myeloid neoplasms grow: pre-leukemic clones carrying mutations in specific genes tolerate chemotherapy better than normal stem cells, expand to fill the void left by treatment, and in some patients acquire additional hits that drive full-blown malignancy. In the studies reviewed, the prevalence of clonal haematopoiesis ranged from 20 percent to over 80 percent, and crucially, higher rates were observed among patients who went on to develop therapy-related myeloid neoplasms. Recurrent mutations in three genes in particular, DNMT3A, TET2, and TP53, were consistently identified across both CAR-T and ASCT cohorts, underscoring that the two treatments share common biological underpinnings for this complication.

The identity of those three genes tells a coherent story. DNMT3A and TET2 are the two most commonly mutated genes in age-related clonal haematopoiesis in the general population, and both encode enzymes involved in DNA methylation regulation, a key epigenetic control mechanism in stem cells. Mutant TET2 and DNMT3A clones are known to be chemotherapy-resistant and to expand preferentially after cytotoxic treatment. TP53, by contrast, is the guardian of the genome, and mutations in it confer both resistance to certain therapies and a propensity toward genomic instability and aggressive leukemic transformation. The consistent finding of TP53 mutations in therapy-related myeloid neoplasms after both CAR-T and ASCT suggests that, in a subset of patients, the seeds of the secondary cancer are present before the transplant or CAR-T infusion even occurs, having been selected and amplified by earlier chemotherapy cycles.

The survival data in the review are sobering and represent perhaps its most clinically consequential message. Once a therapy-related myeloid neoplasm developed, outcomes were poor in both settings, but the prognosis was markedly worse after CAR-T therapy. Patients in CAR-T cohorts had a median overall survival of only 4.5 months following diagnosis of the secondary neoplasm, while ASCT cohorts fared better with a median overall survival of 18.4 months. Several factors could contribute to this disparity. Patients receiving CAR-T in the included studies may have had more heavily pretreated disease, more advanced clonal evolution by the time of the CAR-T infusion, or less reserve to tolerate intensive induction chemotherapy for the secondary leukemia. The authors themselves caution that these observations should be interpreted in the context of heterogeneous study designs and differing degrees of prior treatment exposure, so the comparison between modalities is not a head-to-head trial result but a signal that warrants prospective investigation.

The methodological limitations of the underlying evidence base are significant and the review does not shy away from them. Risk of bias was judged to be moderate to high across the included studies, primarily because of retrospective designs, confounding, and heterogeneity in how outcomes were reported. Retrospective cohort studies are vulnerable to incomplete follow-up, ascertainment bias in detecting secondary cancers, and the impossibility of fully adjusting for the many differences between patients who receive CAR-T versus those who proceed to ASCT. Because pooled incidence estimates could not be calculated, clinicians reading this review will not find a single number they can quote to a patient about their absolute risk. What they will find instead is a consistent qualitative picture: therapy-related myeloid neoplasms are a real and clinically significant complication in both settings, they occur with distinct latency patterns, and they are driven by shared clonal biology.

The implications for clinical practice are already taking shape. As CAR-T therapy expands from relapsed and refractory lymphomas and myelomas into earlier lines of treatment, the population of long-term survivors exposed to this modality is growing rapidly, and the review’s findings argue strongly for structured long-term surveillance of these patients. The authors highlight the need to incorporate molecular risk stratification strategies into the care of heavily pretreated patients. In practical terms, this could mean screening for clonal haematopoiesis before or after CAR-T infusion or transplant, with the presence of high-variant-allele-fraction TP53 or other high-risk mutations flagging patients who merit closer blood count monitoring and earlier referral for evaluation. Such an approach mirrors what is already being explored in solid tumor oncology, where clonal haematopoiesis testing is increasingly used to interpret circulating tumor DNA and to anticipate chemotherapy-associated complications.

For the field at large, this systematic review arrives at a pivotal moment. CAR-T cell therapy is one of the most celebrated advances in cancer medicine of the past decade, producing durable remissions in patients whose disease had exhausted every conventional option, and autologous transplantation remains a cornerstone of treatment for myeloma and lymphoma. The finding that 195 secondary myeloid cancers occurred among roughly 2,900 patients across nine studies does not diminish the value of these therapies, but it does define the frontier of the next challenge: preserving their remarkable benefits while identifying, before treatment, the minority of patients whose marrow clones place them at highest risk of a second malignancy. The review’s central contribution is to consolidate scattered single-center reports into a coherent evidence map, one that shows shorter latency and poorer survival after CAR-T, longer and more variable latency after ASCT, and a common molecular signature rooted in DNMT3A, TET2, and TP53. Converting that map into prospective screening programs, standardized reporting, and ultimately preventive strategies is the task the authors have set for the field, and it is one that will only grow more urgent as these therapies reach ever larger numbers of patients.

Subject of Research: Therapy-related myeloid neoplasms following CAR-T cell therapy and autologous stem cell transplantation

Article Title: Therapy-related myeloid neoplasms after CAR-T therapy and autologous stem cell transplantation: a systematic review of reported incidence, latency, and clonal haematopoiesis

Article References: Musheer, A., Mannan, M. S., Khan, M. W., Bhatti, A., Hafeez, A. S., Junaid, A., & Muhammad, S. (2026). Therapy-related myeloid neoplasms after CAR-T therapy and autologous stem cell transplantation: a systematic review of reported incidence, latency, and clonal haematopoiesis. Annals of Hematology. https://doi.org/10.1007/s00277-026-07305-5

Image Credits: AI Generated

DOI: 10.1007/s00277-026-07305-5

Keywords: therapy-related myeloid neoplasms, CAR-T therapy, autologous stem cell transplantation, clonal haematopoiesis, acute myeloid leukemia, myelodysplastic syndrome, DNMT3A, TET2, TP53, latency, systematic review, hematology

Clonal haematopoiesis, or CH, arises when somatic mutations in haematopoietic stem cells allow them to outcompete their neighbours and produce expanded populations of mutant blood cells detectable in the peripheral circulation. The condition is remarkably common, present in an estimated 10 to 20 percent of people over the age of 65, and by mid-century the global population over 65 is projected to double, placing hundreds of millions of people at risk of CH-associated complications. Mutations in DNMT3A and TET2 together account for roughly 70 percent of all CH variants, and both have been linked to heightened inflammatory responses in myeloid cells, as well as increased risk of haematologic malignancies and cardiovascular disease. Yet the cell-intrinsic mechanisms connecting these mutations to inflammatory signalling have remained incompletely understood, and clinicians have long struggled to explain why individuals carrying the same mutation can show such different clinical outcomes.

The research team hypothesised that retrotransposable elements, or RTEs, might provide the missing mechanistic link. These genomic parasites, including long interspersed nuclear elements (LINEs), short interspersed nuclear elements (SINEs), and long terminal repeat (LTR) elements also known as human endogenous retroviruses, are normally silenced by DNA methylation and repressive histone modifications. When that silencing breaks down, RTEs can produce RNA and even DNA copies that the cell mistakes for viral invaders, activating innate immune sensors such as cGAS and triggering type-I interferon responses. RTE derepression is increasingly recognised as a hallmark of ageing and has been implicated as a causative factor in cellular senescence, making it a natural suspect in a condition defined by age, mutation, and inflammation.

To test the hypothesis, the researchers analysed peripheral blood mononuclear cell samples from 56 individuals with clonal haematopoiesis and 12 non-CH controls, drawn from a larger population study of 738 otherwise healthy adults over 50 recruited through elective hip replacement surgery at King’s College Hospital in London and The Robert Jones and Agnes Hunt Orthopaedic Hospital in Oswestry. Genomic DNA was sequenced with a customised myeloid panel achieving deep coverage through error-corrected duplex consensus reads, while RNA sequencing without poly(A) selection captured both gene and repetitive element expression, averaging 65.1 million paired-end reads per sample. Participants were stratified into five groups: DNMT3A-mutant high variant allele frequency (VAF above 10 percent, n = 10), DNMT3A-mutant low VAF (n = 30), DNMT3A/TET2 double-mutant (n = 9), TET2-mutant (n = 7), and non-CH controls (n = 12).

The results were unambiguous for the largest expanded clones. High-VAF DNMT3A-mutant clones exhibited widespread derepression of retrotransposable elements, dominated by the LINE:L1 and LTR:ERV1/ERVL families, which are typically repressed by DNA methylation. This pattern fits neatly with prior work showing that DNMT3A loss disrupts de novo methylation and permits repetitive element reactivation in haematopoietic and embryonic stem cells. Interestingly, the high-VAF DNMT3A group was not uniform: the team identified two distinct subgroups, one with markedly elevated RTE expression and another resembling controls, revealing substantial heterogeneity within the same mutation class. The position of the mutation within the protein appeared to matter, with variants affecting the PWWP domain exclusively associated with the RTE-high phenotype, and within the methyltransferase domain most high-VAF mutations showing elevated RTE expression except the recurrent R882H hotspot, which clustered in the RTE-low group. The authors caution that with only ten high-VAF cases these observations require validation in larger cohorts, but they suggest that different DNMT3A mutations may not be biologically equivalent.

Functionally, the RTE-high DNMT3A clones looked dramatically different from their RTE-low counterparts. Gene set enrichment analysis showed that high-VAF DNMT3A clones with elevated RTE expression were enriched for inflammatory signalling pathways, including TNF-alpha/NF-kappaB signalling and broad inflammatory response programmes. At the effector level, these clones expressed significantly higher levels of TNF and the alarmin S100A9, and showed upregulation of multiple interferon regulatory genes, including ISG15, IFI44, IFIT1, IFIT2, and IFIT3, a signature consistent with RTE-derived nucleic acids activating cytosolic sensors and driving type-I interferon responses. The researchers also observed enrichment of pathways involving the Human Silencing Hub (HUSH) and Sirtuin 1 (SIRT1), which they interpret as possible compensatory attempts by alternative epigenetic regulators, including HUSH, Polycomb Repressive Complex 2, and SIRT1, to preserve transcriptional repression in the face of failing DNMT3A-mediated methylation, a coordinated but ultimately failing cellular response rather than random transcriptional noise.

TET2-mutant clones told a completely different story. Far from derepressing retroelements, TET2-mutant haematopoiesis showed a trend towards reduced RTE expression relative to controls, and no significant associations between RTE activity and any of four curated inflammatory and ageing signatures: interferon regulatory genes, TNF-NF-kappaB, senescence, and the senescence-associated secretory phenotype (SASP). Instead, pathway analysis revealed that TET2-mutant clones were enriched for oxidative phosphorylation, reactive oxygen species signalling, and mechanistic target of rapamycin complex 1 (mTORC1) signalling, pointing to a metabolically driven, redox-associated inflammatory phenotype rather than a retroelement-driven one. This aligns with recent publications implicating metabolic and inflammasome pathways, rather than transposable element activation, in TET2-mutant inflammation. In regression analyses, RTE family expression was significantly and positively associated with all four signature scores in DNMT3A high-VAF cases, showed more selective associations with interferon and senescence programmes in low-VAF DNMT3A cases, and showed no significant associations in TET2-mutant cases.

The findings held up under independent scrutiny. Single-cell TARGET-seq data from a separate study, covering five DNMT3A-mutant and three TET2-mutant samples, showed that DNMT3A-mutant cases contained more upregulated RTE subfamilies across individual haematopoietic lineages, with the erythroid and megakaryocyte compartments showing the strongest enrichment and ERV1, ERVL, ERVL-MaLR, and L1 families contributing most. An external validation cohort of 92 healthy middle-aged and older adults from a publicly available RNA-seq dataset, in which RNAVAR mutation calling identified seven DNMT3A and six TET2 high-VAF cases, reproduced the core pattern: retroelement expression correlated significantly with interferon, TNF-NF-kappaB, and senescence signatures in DNMT3A high-VAF clones but not in TET2-mutant cases. Double-mutant clones carrying both mutations occupied an intermediate position, with RTE levels higher than TET2-mutants but lower than DNMT3A high-VAF clones, and attenuated TNF-NF-kappaB and interferon programmes, hinting at non-redundant and possibly antagonistic roles for the two epigenetic regulators.

The study has limitations the authors acknowledge candidly. Bulk RNA sequencing of peripheral blood mononuclear cells can be influenced by differences in cellular composition, and although the TARGET-seq analyses supported cell-intrinsic differences in RTE regulation, the small single-cell cohort cannot fully exclude compositional effects. The number of high-VAF cases remains modest, and, critically, the data demonstrate an association between RTE derepression and inflammatory pathway activation rather than a proven causal relationship, since both could arise as parallel consequences of DNMT3A-associated epigenetic dysregulation. Recurrent co-occurring CHIP mutations were found in only 10 percent of DNMT3A-mutant cases and showed no enrichment in the RTE-high or high-VAF groups, making co-mutations an unlikely explanation for the observed heterogeneity.

Even so, the implications are considerable. By showing that inflammation in clonal haematopoiesis is mutation-specific, with retrotransposon reactivation powering the inflammatory programme of DNMT3A-mutant clones while metabolic and redox dysregulation drive TET2-mutant inflammation, the study lays the groundwork for mutation-specific biomarkers and therapies. If future functional and preclinical studies confirm that modulating RTE activity can dampen the inflammatory phenotype of DNMT3A-mutant clonal haematopoiesis, interventions could one day be tailored to the specific molecular mechanism behind each patient’s mutated clone, replacing a one-size-fits-all approach with precision medicine for age-related blood disorders and, potentially, the cardiovascular and neurodegenerative diseases that travel with them.

Subject of Research: Mutation-specific retrotransposable element derepression and inflammatory signalling in DNMT3A- and TET2-mutant clonal haematopoiesis

Article Title: Retrotransposable element derepression distinguishes DNMT3A-mutant from TET2-mutant clonal haematopoiesis

Article References: Hasan, M., Durandeau, S., Thompson, I. R., Roussotte, H., Tsai, Y.-T., Seymen, N., Gerlevik, S., Bianchini, N., Alishah, K., Albuquerque, M. M., Lewis, J., Bonganay, L., Jakobsen, N. A., Zeisig, B., Irshad, S., Vyas, P., So, E. C. W., Iacoangeli, A., Zheng, X., … Karimi, M. M. (2026). Retrotransposable element derepression distinguishes DNMT3A-mutant from TET2-mutant clonal haematopoiesis. GeroScience. https://doi.org/10.1007/s11357-026-02510-6

Image Credits: AI Generated

DOI: 10.1007/s11357-026-02510-6

Keywords: clonal haematopoiesis, DNMT3A, TET2, retrotransposable elements, LINE-1, endogenous retroviruses, inflammation, ageing, DNA methylation, interferon signalling, senescence, haematopoietic stem cells

Clonal haematopoiesis occurs when a blood stem cell acquires a somatic mutation and its descendants expand to form a detectable clone, often without any symptoms. Population sequencing shows it in as many as one in five adults over 65, yet transformation to an overt myeloid neoplasm such as acute myeloid leukaemia or myelodysplastic syndrome happens at only 0.5 to 1.0 percent per year. The clinical risk landscape is starkly uneven. According to the clonal haematopoiesis risk score, or CHRS, only about one percent of carriers detected in population-scale sequencing fall into the high-risk category, facing a 10-year blood cancer incidence of 52 percent, while roughly 90 percent are low risk with a 10-year incidence below one percent. Sequencing everyone, therefore, would generate enormous volumes of clinically meaningless results.

The research team, led by investigators drawing on the UK Biobank, hypothesised that the molecular behaviour of dangerous clones leaves fingerprints in ordinary blood measurements. Mutant stem cells alter the production, size, and morphology of red cells, white cells, and platelets, and those changes are captured in the two dozen parameters reported with every complete blood count. To test the idea, they assembled an enormous training dataset of 369,260 UK Biobank participants aged 40 to 69 who had both whole-exome sequencing and blood counts drawn on the same day, then held back 92,316 participants for internal testing.

Defining the target was itself a technical exercise. The team computed a molecular clonal haematopoiesis score for each person, summing points for high-risk mutations in genes such as TP53, JAK2, SRSF2, and RUNX1, for a solitary DNMT3A mutation, for multiple mutations, and for a variant allele fraction of 20 percent or more. A score above 4 captured nearly everyone who later developed myeloid neoplasia or who was high or intermediate risk by CHRS. Participants were then labelled positive if they scored above the threshold and developed blood cancer, negative if they scored at or below it, and equivocal if they scored high but remained cancer-free during follow-up.

The model itself is a balanced random forest, a tree-based classifier chosen after comparing nine architectures on a resampled training set of 3,000 individuals, 1,000 per class. Synthetic minority over-sampling and random under-sampling corrected the extreme class imbalance, since true positives made up less than one tenth of one percent of the cohort. Feature selection was aggressive: highly correlated parameters such as white cell count and haematocrit were removed, the Boruta algorithm discarded sex and several less informative measures, and features inconsistently annotated across datasets were dropped. The final inputs are strikingly mundane: age plus absolute counts of red cells, lymphocytes, monocytes, neutrophils, eosinophils, reticulocytes, and platelets, along with haemoglobin, red cell distribution width, platelet crit, mean corpuscular volume, and giant platelets.

Performance in the UK Biobank test set was strong where it mattered most. The area under the receiver operating characteristic curve for the positive class reached 0.89, and sensitivity for high-risk clonal haematopoiesis was 0.73, with the three missed cases all under 60 with entirely normal blood counts. More strikingly, positive classification captured 100 percent of the 42 individuals deemed high risk by CHRS, and the overall negative predictive value was 0.993, meaning a negative result reliably rules out clinically significant clonal haematopoiesis. The model was intentionally permissive of false positives, accepting that sequencing would resolve them, and achieved one detected high-risk case for every 247 people flagged, compared with one per 2,500 under a screen-everyone strategy.

External validation demonstrated genuine portability. In a clinical cohort of 2,233 patients evaluated in Dana-Farber haematology clinics, where cytopenias were far more common, PRISM correctly classified 16 of 18 positive-label individuals, an AUC of 0.74 and sensitivity of 0.89, with a negative predictive value of 0.968. In the All of Us Research Program cohort of 14,674 Americans, the positive-class AUC was 0.87 with perfect sensitivity across ten cases and a negative predictive value of 0.998. Across all three cohorts, every CHRS-defined high-risk individual received a positive PRISM label, and intermediate-risk carriers almost never received a negative one. Benchmarking against the clonal cytopenia risk score showed that 18 of 19 high-risk clonal cytopenia cases were also flagged as positive.

The downstream consequences were equally telling. Among UK Biobank participants with clonal haematopoiesis who later developed myeloid neoplasia, 63 percent had been labelled positive and 25 percent equivocal, and the misclassified negatives were overwhelmingly low risk by post-sequencing CHRS. Ten-year cumulative incidence of myeloid neoplasia was 1.44 percent in the positive class versus 0.11 percent in the negative class, a hazard ratio of 8.81 that outperformed age, cytopenia, mean corpuscular volume, and red cell distribution width in multivariable models. Because clonal haematopoiesis also drives cardiovascular disease, the team checked whether negative labels would exclude people at cardiac risk; reassuringly, 10-year cardiovascular risk was lowest, at 4.6 percent, in the PRISM-negative group, compared with 15.7 percent in the positive class.

The authors are careful about scope. PRISM is a pre-sequencing triage tool, not a diagnostic test, and its equivocal class, which includes people with abnormal counts but no dangerous clones, cannot be resolved from a single blood draw; serial sampling and clinical correlation remain essential. The UK Biobank’s predominantly European ancestry limits generalisability, and specialised settings such as chemotherapy exposure or inherited bone marrow failure would need context-specific recalibration. Still, if prospective validation succeeds, the implications are considerable: haematology clinics could reassure worried patients with a negative label, transplant programmes could screen donors cheaply before sequencing, and health systems could cut unnecessary molecular testing by 14 to 55 percent while sparing low-risk individuals the psychological and financial burden of overdiagnosis. A model built from the humble blood count may become the gatekeeper for the genomic era of blood cancer prevention.

Subject of Research: A machine learning model using complete blood count parameters to prioritise sequencing for high-risk clonal haematopoiesis

Article Title: Derivation of a blood count-based machine learning model to prioritise sequencing for high-risk clonal haematopoiesis: a retrospective cohort study

Article References: Nandi, R., Zhao, K., Pershad, Y., Khondoker, N. N., Vlasschaert, C., Uddin, M. M., Natarajan, P., DeAngelo, D. J., Luskin, M. R., Lindsley, R. C., Uno, H., Shimony, S., Bick, A. G., & Weeks, L. D. (2026). Derivation of a blood count-based machine learning model to prioritise sequencing for high-risk clonal haematopoiesis: a retrospective cohort study. eClinicalMedicine, 100, Article 104238. https://doi.org/10.1016/j.eclinm.2026.104238

Image Credits: AI Generated

DOI: 10.1016/j.eclinm.2026.104238

Keywords: clonal haematopoiesis, machine learning, complete blood count, myeloid neoplasia, leukaemia risk, CHRS, UK Biobank, risk prediction, sequencing, premalignant conditions, cardiovascular disease, precision medicine

The conceptual foundation dates back to 1976, when Peter Nowell proposed that tumour cell populations evolve through acquired genetic lability, with stepwise selection of variant sublines driving progression. Half a century of bulk sequencing vindicated the model, revealing branched evolution, intratumoural heterogeneity and the selective sweeps that follow treatment. Yet bulk measurements average across millions of cells, obscuring which mutation resides in which cell and, crucially, what that mutation actually does to the cell carrying it. The review’s authors argue that closing this genotype-to-phenotype gap is now the central task of cancer evolutionary biology.

One striking motivation comes from healthy tissue. Landmark studies of normal skin, oesophagus, colon, endometrium, bladder and bronchial epithelium have shown that somatic mutations in canonical cancer driver genes are under pervasive positive selection in tissue that looks entirely normal under the microscope. In sun-exposed skin, roughly a quarter of cells carry cancer-associated mutations, and mutation burdens in some cells rival those of tumours. Clonal haematopoiesis, the age-related expansion of blood cells carrying mutations in genes such as DNMT3A, TET2 and JAK2, likewise demonstrates that genetic mosaicism is a feature of ordinary physiology and ageing, seeding the pre-malignant landscape from which frank cancers emerge.

Mapping this diversity, however, is only the first step. The review emphasises that understanding somatic clonal evolution requires defining the phenotypes that give mutated clones a fitness advantage, whether those phenotypes involve altered differentiation, survival, proliferation or interaction with the microenvironment. This is where genotype-aware single-cell multi-omics enters. Methods such as G&T-seq and its descendants physically split or barcode the genome and transcriptome of an individual cell, while genotyping-of-transcriptomes approaches recover expressed mutations directly from single-cell RNA-sequencing data. Targeted strategies enrich for known mutant loci, and chromatin-focused assays now co-capture mutations alongside single-cell accessibility profiles.

The biological payoffs have been substantial. In clonal haematopoiesis, single-cell multi-omics has shown that the effects of a mutation are often cell-state specific. DNMT3A R882 mutations, for example, were found to perturb early progenitor states through selective hypomethylation, a phenotype invisible to bulk assays. Splicing aberrations in haematopoietic clonal outgrowths display distinct cell-type-specific impacts, and maps linking genotypes to chromatin accessibility profiles reveal how individual mutations reshape regulatory landscapes in a lineage-dependent manner. In myeloproliferative neoplasms, clonally resolved analyses have traced how JAK2 and CALR mutations propagate through differentiation hierarchies, while work in acute myeloid leukaemia has connected RAS-mutant leukaemia stem cells to clinical resistance against the BCL-2 inhibitor venetoclax.

Beyond single time points, the review highlights the power of coupling phylogenetic reconstruction with phenotypic measurement. Endogenous marks such as somatic point mutations, copy-number alterations, mitochondrial DNA mutations, microsatellite shifts and stochastic epimutations each leave heritable traces that allow researchers to infer the ancestral relationships among single cells. Mitochondrial mutations in particular have enabled lineage tracing directly in human samples, and somatic epimutations have recently been used to chart the dynamics of blood ageing. Reconstructed single-cell phylogenies can then be time-calibrated, converting a branching diagram into a chronogram that estimates when a clone originated within a patient’s lifespan.

Such temporal mapping demands careful statistical treatment. The authors describe phylogenetic frameworks that quantify heritability and plasticity of cell states, decoupling genetic inheritance from non-genetic, environmentally driven transitions. Molecular clock models, whether strict or relaxed, permit inference of mutation rates and timing of clonal expansions, and phylodynamic approaches borrowed from pathogen genetics now illuminate how tumour population sizes fluctuate over the course of disease. Applied to colorectal cancer, these tools have revisited the Big Bang model of tumour growth, in which most subclonal diversity is generated in an early expansion rather than through later selective sweeps, and have documented phenotypic plasticity under genetic control during malignant progression.

Spatial context adds a further dimension. Multiclonal invasion patterns in breast tumours, spatially resolved copy-number maps in benign and malignant tissue, and spatial genomics of cancer clones all demonstrate that evolutionary dynamics are constrained by tumour architecture. Mechanical confinement has been shown to govern phenotypic plasticity in melanoma, and harsh microenvironments select for glycolytic phenotypes in early breast cancer. Integrating spatially resolved or lineage-resolved phenotypes with genotype maps is therefore revealing how selection operates not just on mutations but on the cell states and niches in which those mutations find themselves.

Translationally, the authors argue that genotype-to-phenotype mapping can expose therapeutic vulnerabilities for the precision elimination of disease-propagating mutant cells. Because mutant phenotypes are often confined to specific cell states or lineages, vulnerabilities may exist that spare wild-type tissue. Single-cell analyses have identified drug-tolerant persister states, non-genetic determinants of clonal fitness, and epigenetically inherited plasticity that drives drug resistance through one-to-many genotype-to-phenotype relationships. In glioblastoma, recurring cellular states whose abundance is modulated by genetic aberrations suggest combination strategies; in IDH-mutant oligodendroglioma, mutant IDH inhibitors induce lineage differentiation detectable at single-cell resolution. Evolutionary steering, in which treatment is designed to guide tumours toward collateral sensitivities, becomes more tractable when clonal phenotypes can be read directly in patients.

The review closes with a sober assessment of remaining challenges. Whole-genome amplification artefacts, allelic dropout and tissue dissociation biases still limit sensitivity and fidelity, and artifacts in mitochondrial DNA analyses can misinform phylogenetic inference if uncorrected. Computational methods continue to mature, from probabilistic single-cell phylogeny inference that accounts for sequencing error to models that relax the infinite-sites assumption when back mutations and parallel evolution occur. Yet the trajectory is clear: by pairing genotype with phenotype in the same cell and embedding those pairs within time-calibrated phylogenies, researchers can now define, directly in primary human samples, the mechanisms underlying clonal expansion in both healthy and malignant tissues. For a disease that evolves to evade every therapy thrown at it, that may prove the most consequential lens oncology has yet acquired.

Subject of Research: Single-cell multi-omics mapping of genotype and phenotype co-evolution in clonal evolution of healthy and cancerous tissues

Article Title: A single-cell lens into the co-evolution of genotypes and phenotypes in cancer

Article References: Izzo, F., Prieto, T., Potenski, C., & Landau, D. A. (2026). A single-cell lens into the co-evolution of genotypes and phenotypes in cancer. Nature Reviews Cancer. https://doi.org/10.1038/s41568-026-00970-8

Image Credits: AI Generated

DOI: 10.1038/s41568-026-00970-8

Keywords: single-cell multi-omics, clonal evolution, genotype-phenotype mapping, tumour heterogeneity, clonal haematopoiesis, single-cell phylogenetics, lineage tracing, somatic mosaicism, phenotypic plasticity, cancer genomics, epigenetics, therapeutic vulnerabilities