Children with rhabdomyosarcoma whose tumours lack the genetic marker traditionally associated with the most dangerous form of the disease may nevertheless harbour cancer cells with the same aggressive biology, according to a new study. The findings suggest that current classification systems can overlook children whose disease is likely to relapse or resist treatment, and could eventually help doctors identify those patients at diagnosis. Researchers say the work also points towards a more precise way of understanding childhood cancer: rather than focusing only on a single genetic alteration, clinicians may need to recognise a broader, shared cellular state that allows tumours to grow and spread aggressively.
Rhabdomyosarcoma, or RMS, is a cancer that develops in cells destined to become skeletal muscle and is among the most common soft-tissue cancers affecting children under 15. Around 55 children are diagnosed with the disease in the United Kingdom each year. Unlike many cancers in adults, which are frequently associated with the accumulation of mutations over decades or with environmental exposures, childhood cancers often arise from genetic and developmental errors occurring as tissues form. RMS is commonly divided into fusion-positive and fusion-negative disease, based on whether tumour cells contain a rearrangement joining two genes that are normally separate. The best-known alteration involves PAX3 or PAX7 fused with FOXO1, producing an abnormal transcription factor that can disrupt normal cell development and promote tumour formation.
The presence of a PAX-FOXO1 fusion is strongly associated with an unfavourable outcome. Children whose tumours carry the fusion generally have lower survival rates than those whose tumours do not, even when treated with intensive chemotherapy, surgery and radiotherapy. This distinction has become an important part of risk assessment and treatment planning. Yet the classification is not absolute. Some children with fusion-negative RMS develop rapidly progressive, treatment-resistant disease, while others with apparently similar tumours respond more successfully. Until now, the biological explanation for this difference has remained incomplete, raising concerns that conventional testing may be concealing a high-risk population within the group considered to have lower-risk disease.
In the study, published in Cancer Research on 19 August 2026, scientists from the Wellcome Sanger Institute, the University of Cambridge, Great Ormond Street Hospital and University College London analysed RMS tumours using genomic methods capable of resolving differences between individual cancer cells. The research cohort included four children with fusion-negative lethal disease, four with fusion-positive disease and eight children with fusion-negative tumours who remained alive. This design enabled the team to compare the molecular characteristics of aggressive and less aggressive tumours across the two major genetic categories, while examining whether the behaviour of individual cells matched the broader diagnosis assigned to each tumour.
A central technique was single-cell RNA sequencing, which measures the RNA molecules present in individual cells. RNA provides a snapshot of which genes are active, meaning that researchers can use it to determine whether a cell is behaving like a developing muscle cell, a proliferating cancer cell or another specialised population within the tumour. Bulk sequencing, in which genetic material from thousands or millions of cells is mixed together, can obscure rare but important groups. A small population of highly aggressive cells may produce only a faint signal in an averaged molecular profile. By analysing cells separately, the researchers were able to identify these hidden populations and examine the gene networks associated with their growth, developmental state and resistance to normal controls.
The results revealed that aggressive fusion-negative tumours contained cells with gene-expression programmes closely resembling those found in fusion-positive high-risk RMS. In other words, the cells appeared to have reached a similar biological destination through different genetic routes. The absence of the PAX-FOXO1 fusion did not guarantee that the tumour lacked the cellular features linked to aggressive behaviour. Instead, some fusion-negative cancers showed a convergent cell state, marked by coordinated activity across multiple pathways that influence cell identity, proliferation and the ability to maintain an immature, tumour-promoting phenotype. This convergence may explain why tumours carrying different initiating alterations can ultimately behave in similar ways in children.
The investigators also identified rare genetic changes in aggressive fusion-negative tumours that affect cellular pathways overlapping with those disrupted in fusion-positive disease. These alterations may not create the same fusion protein, but they can disturb related biological systems and push developing muscle cells towards a comparable malignant state. The finding supports a model in which RMS is not driven by one universal mutation. Instead, several independent genetic events may interfere with developmental programmes that normally guide muscle formation, producing a shared state of cellular immaturity and uncontrolled growth. Such a state could be more informative for predicting clinical behaviour than the presence or absence of a single risk marker alone.
To understand how these aggressive cells were arranged inside tumours, the researchers used spatial transcriptomics. This approach preserves information about the location of cells within tissue while measuring gene activity, allowing scientists to connect molecular identity with physical organisation. Rather than viewing a tumour as a uniform mass, spatial analysis treats it as an ecosystem made up of distinct cell populations occupying different niches. The study showed how aggressive populations were organised within the cancer tissue, offering clues about how neighbouring cells and the tumour environment may support disease progression. Mapping these patterns could eventually help researchers determine whether particular regions of a tumour are especially likely to survive treatment or seed a relapse.
The discovery may have consequences for diagnosis and therapy, although it will require validation in much larger groups of patients before it can change clinical practice. Researchers are expanding the work to hundreds of additional RMS samples in an effort to find more of the rare genetic pathways that can generate the convergent aggressive state. If reliable molecular markers can be developed, tumour testing could be extended beyond fusion status to include the activity of high-risk cell programmes. Children whose disease currently appears non-high-risk might then be identified for closer monitoring or treatment adjustments, while those unlikely to benefit from additional chemotherapy could potentially avoid some of its long-term toxicity. Any change in treatment would need to be guided by clinical trials, because increasing therapy intensity can itself cause serious harm.
The researchers also hope that the shared markers found on aggressive RMS cells could provide targets for precision treatments, including immunotherapies such as CAR-T cell therapy. CAR-T treatment involves collecting a patient’s T cells and genetically engineering them to recognise a molecule on cancer cells before returning them to the body. For this strategy to work safely, the target must be present on tumour cells but absent, or present at much lower levels, in essential healthy tissues. The newly identified cell-state markers could help guide the search for such targets, although the study does not demonstrate that a CAR-T therapy for RMS is ready for patients. The immediate significance of the research is the biological insight: aggressive childhood cancers that look different genetically may share a vulnerable molecular identity, creating a possible foundation for earlier detection and more personalised treatment.
Subject of Research:
Aggressive and high-risk childhood rhabdomyosarcoma, including fusion-negative tumours with high-risk cellular features.
Article Title:
High-Risk Rhabdomyosarcomas Feature a Convergent Cell State
News Publication Date:
19 August 2026
Web References:
Wellcome Sanger Institute: https://www.sanger.ac.uk/
Wellcome: https://wellcome.org/
Cancer Research UK: https://www.cancerresearchuk.org/
Alice’s Arc: https://www.alicesarc.co.uk/
References:
Whitfield, H. J. et al. (2026), “High-Risk Rhabdomyosarcomas Feature a Convergent Cell State,” Cancer Research. DOI: 10.1158/0008-5472.CAN-25-4403.
Skapek, S. X. et al. (2013), “PAX-FOXO1 fusion status drives unfavourable outcome for children with rhabdomyosarcoma: a Children’s Oncology Group report,” Pediatric Blood & Cancer. DOI: 10.1002/pbc.24532.
Children with Cancer UK, “Rhabdomyosarcoma Overview.”
Keywords:
Rhabdomyosarcoma, childhood cancer, cancer genetics, PAX-FOXO1, fusion-negative RMS, single-cell RNA sequencing, spatial transcriptomics, cancer biology, precision medicine, immunotherapy, CAR-T cell therapy, oncology
Tags: aggressive childhood cancer driverscancer relapse prediction in childrencellular states in pediatric tumorschildhood cancer geneticsgenetic and developmental errors in childhood cancergenetic markers in pediatric cancerlimitations of traditional cancer classificationmolecular mechanisms of childhood rhabdomyosarcomapersonalized diagnosis of childhood cancersrhabdomyosarcoma tumor biologytumor classification in childhood cancerstumor growth and spread in pediatric patients
