Safeguarding genomic integrity in pluripotent stem-cell therapies.

# Safeguarding genomic integrity in pluripotent stem-cell therapies *The Lancet 2026* Comment Safeguarding genomic integrity in pluripotent stem-cell therapies Published Online Human pluripotent stem-cells (HPSCs)—including transplanted cells. Genomic integrity refers to the February 27, 2026 human embryonic stem-cells and human induced stability of a cell’s DNA—the absence of mutations, https://doi.org/10.1016/ S0140-6736(26)00095-4 pluripotent stem-cells—are reshaping the landscape of structural variants, or copy-number changes that regenerative medicine.1,2 In Parkinson’s disease, where can increase tumourigenic potential or impair selective loss of midbrain dopaminergic neurons drives differentiation and graft function.9 As enthusiasm for major motor symptoms, generations of investigators HPSC-based cell therapy accelerates and programmes have evaluated diverse dopamine-producing donor expand into larger and more geographically tissues, including fetal ventral mesencephalon. Although distributed trials across Parkinson’s disease and other fetal grafts established clinical proof-of-concept for cell degenerative disorders,10 genomic integrity should replacement therapy, they are no longer pursued because shift from a peripheral technical issue to a primary of ethical, logistical, and practical constraints.3 determinant of translational safety; it underpins Early feasibility trials of HPSC-based cell replacement patient protection, therapeutic consistency, and public therapy now show survival of transplanted trust. dopaminergic neurons in the human brain, together This challenge is inherent to HPSC biology.1,2 As with PET-based evidence of restored dopaminergic HPSCs proliferate indefinitely and can differentiate activity.4–7 These findings, accompanied by modest into any lineage, they are uniquely prone to acquiring symptomatic improvement and the absence of tumour somatic variants at multiple stages. Variants could formation or graft-induced dyskinesia, provide the arise from pre-existing donor mosaicism (ie, somatic strongest indication to date that HPSC-derived neural variants already present in a subset of donor cells grafts can function in vivo. Although not powered to before reprogramming), especially in cells derived evaluate efficacy, these trials constitute a meaningful from older individuals. Additional variants accumulate proof-of-concept and mark a pivotal moment for during reprogramming under oxidative and replicative regenerative neurology.8 stress, and continue to emerge during expansion and Yet this progress exposes a crucial and insufficiently differentiation, where clonal selection favours growth- addressed vulnerability: the genomic integrity of advantage populations. Recurrent chromosomal gains11 Panel: Genomic-integrity standards for HPSC-derived clinical products Pluripotent stem-cells can acquire genetic alterations through position is sequenced ~50 times]) whole-genome donor mosaicism, reprogramming stress, and clonal selection, sequencing to detect de novo somatic variants—including underscoring the need for systematic genomic safeguards. single-nucleotide variants, structural alterations, and A shared genomic-integrity framework is required to support copy-number changes; lower-depth (≤30× mean coverage) safe and reproducible translation of HPSC-based therapies. methods can miss clinically relevant alterations • Predefined exclusion criteria for high-risk variants: loss-of- • Independent confirmation of low-frequency findings: function variants in core tumour-suppressor genes with well subclonal, low-allele-fraction somatic variants should be established mechanisms should preclude progression to validated with an orthogonal method to avoid clinical manufacturing; this approach establishes misclassification and inadvertent expansion of unsafe measurable, consistent, and patient-centred safety clones thresholds • Transparent reporting to support comparability and • Genomic quality control during manufacturing: new regulation: explicit disclosure of sequencing methods, variants can emerge at multiple production stages; clinical somatic variant-calling pipelines, depth and coverage programmes should therefore incorporate predefined metrics, and evaluated gene sets strengthens interpretation genomic testing points during production and before and facilitates comparison across trials clinical release HPSC=human pluripotent stem-cell. • High-depth baseline whole-genome assessment of cell banks: use high-depth (50× mean coverage [ie, each 1492 Comment and tumour-suppressor gene mutations12 have been orthogonal methods, and implement genomic checks documented in multiple HPSC lines. The small founding at predefined production stages to detect emerging population and extensive passaging required to variants before release. establish HPSC lines make such changes not exceptional Experience from the gene therapy field reinforces but expected.13 the need for rigorous safety oversight and clear These variants have biological and clinical risk disclosure. One example is Jesse Gelsinger, who consequences.13–15 High-risk somatic variants can died of a massive immune response during a safety increase tumourigenicity by inactivating tumour- trial of gene therapy for ornithine transcarbamylase suppressor genes or activating oncogenes, and can deficiency in 1999; this trial revealed how also impair neuronal differentiation, resilience, and unanticipated biological variables can cause graft function.16 These perturbations threaten the two devastating outcomes and stall an entire therapeutic outcomes that matter most in cell-based therapy: safety domain.22 Only after harmonised vector design, and durable clinical benefit.17 They can increase the risk genomic analysis, and coordinated safety oversight of graft-derived tumours and compromise neuronal were established did the field regain its momentum. differentiation, survival, and sustained dopaminergic HPSC-based cell replacement therapy now stands at a function. For a therapeutic product intended for juncture comparable to that of the gene therapy field: intracerebral transplantation, the stakes are uniquely early feasibility has been shown, but broad clinical high; even rare tumourigenic clones or subtle translation will require shared definitions, transparent functional impairments could have irreversible clinical reporting, and consistent quality control. consequences.18,19 The path forward demands a decisive shift from Despite these risks, systematic genomic integrity narrative reassurance to actionable transparency. surveillance is largely absent from recent clinical Clinical manuscripts should go beyond summary reports.4–6 Manuscripts rarely specify how somatic statements and report concrete genomic integrity variants were assessed, what sequencing depth practices, specifying testing timepoints, sequencing was used, which genes were evaluated, or what depth (eg, high-coverage whole-genome sequencing), exclusion thresholds were applied.20 These omissions and the rationale for excluding defined variant risk implying uniform safety and deprive clinicians, classes. Regulators should establish minimum regulators, and patients of essential information. As requirements for surveillance of high-risk genes and more centres initiate HPSC-based programmes and mandate orthogonal validation of low-frequency multisite trials expand, divergence in genomic quality variants. Manufacturing centres should incorporate control practices is likely to widen. Such variability genomic quality control not only at release but also at introduces inequities in patient safety, undermines intermediate production steps, applying prespecified interpretability across trials, and poses a systemic risk to action limits when culture-acquired alterations arise a dynamic therapeutic modality. in core tumour-suppressor genes. Investigators A minimum genomic integrity standard is should adopt shared definitions and concise reporting therefore essential (panel). Recent best practices checklists that enable comparability across trials. These from the International Society for Stem Cell Research aligned measures will curb preventable divergence point towards explicit standards for genomic and protect patients from uneven risk. Without this characterisation, recommending high-depth whole- discipline, the field risks remaining confined to early genome sequencing of pluripotent cell banks and feasibility, unable to translate potential into reliable genome-edited or highly expanded clones.21 A clinical practice. concise, clinically actionable framework is outlined Feasibility has shown that HPSC-based cell in the panel and should guide all clinical-grade replacement therapy can work in humans. Whether it manufacturing. In brief, programmes should apply will do so safely, consistently, and equitably will depend high-depth whole-genome sequencing, use explicit on the decisions made now. The window for establishing exclusion rules for high-risk tumour-suppressor shared genomic integrity standards is narrow. Before gene variants, confirm low-frequency findings with programmes expand further, the field should converge Comment on a common genomic integrity framework and 6 Sawamoto N, Doi D, Nakanishi E, et al. Phase I/II trial of iPS-cell-derived embrace a culture of genomic stewardship. Doing so will dopaminergic cells for Parkinson’s disease. Nature 2025; 641: 971–77. 7 Schweitzer JS, Song B, Herrington TM, et al. Personalized iPSC-derived mitigate risk, prevent fragmentation, and safeguard the dopamine progenitor cells for Parkinson’s disease. N Engl J Med 2020; 382: 1926–32. future of regenerative medicine as it enters a decisive 8 Cha Y, Leblanc P, Kim KS. A new era in regenerative medicine: cell decade. replacement therapy for Parkinson’s disease is on the horizon. Cell Stem Cell 2025; 32: 864–66. K-SK is a co-founder of NurrOn Pharmaceuticals, holds equity in the company, 9 Halliwell J, Barbaric I, Andrews PW. Acquired genetic changes in human and receives consulting fees from NurrOn Pharmaceuticals, in accordance with pluripotent stem cells: origins and consequences. Nat Rev Mol Cell Biol 2020; institutional conflict-of-interest policies. K-SK receives royalties paid to McLean 21: 715–28. Hospital from Fujifilm Cellular Dynamics for a licensed patent related to 10 Kirkeby A, Main H, Carpenter M. Pluripotent stem-cell-derived therapies in methods for purifying midbrain dopaminergic neural progenitor cells clinical trial: a 2025 update. Cell Stem Cell 2025; 32: 329–31. (US# 9,750,768). K-SK is an inventor on issued and pending patents related to 11 Krivec N, Ghosh MS, Spits C. Gains of 20q11.21 in human pluripotent stem pluripotent stem-cell technologies and Parkinson’s disease cell therapy cells: insights from cancer research. Stem Cell Reports 2024; 19: 11–27. (US#9,657,273, US 17/239,059, and PCT/US2024/028596). YC declares a patent 12 Merkle FT, Ghosh S, Kamitaki N, et al. Human pluripotent stem cells issued for synergistic genome-non-integrating reprogramming by microRNAs recurrently acquire and expand dominant negative P53 mutations. Nature and transcription factors (US 17/239,059). All other authors declare no 2017; 545: 229–33. competing interests. During the preparation of this manuscript the authors used 13 Trounson A. Potential pitfall of pluripotent stem cells. N Engl J Med 2017; ChatGPT (OpenAI) to assist with grammar and language editing. The authors 377: 490–91. reviewed and edited all content generated by this tool and take full 14 Derks LLM, van Boxtel R. Stem cell mutations, associated cancer risk, and responsibility for the integrity, accuracy, and originality of the work. consequences for regenerative medicine. Cell Stem Cell 2023; 30: 1421–33. 15 Lezmi E, Jung J, Benvenisty N. High prevalence of acquired cancer-related Sek Won Kong, In-Hee Lee, Young Cha, Bob Carter, mutations in 146 human pluripotent stem cell lines and their *Kwang-Soo Kim differentiated derivatives. Nat Biotechnol 2024; 42: 1667–71. 16 Andrews PW, Barbaric I, Benvenisty N, et al. The consequences of recurrent kskim@mclean.harvard.edu genetic and epigenetic variants in human pluripotent stem cells. Department of Pediatrics, Harvard Medical School, Boston, MA, USA (SWK); Cell Stem Cell 2022; 29: 1624–36. Computational Health Informatics Program, Boston Children’s Hospital, Harvard 17 Merkle FT, Ghosh S, Genovese G, et al. Whole-genome analysis of human Medical School, Boston, MA, USA (SWK, I-HL); Health Sciences, University of embryonic stem cells enables rational line selection based on genetic Utah, Salt Lake City, UT, USA (BC); Molecular Neurobiology Laboratory, McLean variation. Cell Stem Cell 2022; 29: 472–86.e7. Hospital and Department of Psychiatry, Harvard Medical School, Belmont, 18 Andrews PW, Ben-David U, Benvenisty N, et al. Assessing the safety of MA 02478, USA (YC, K-SK); Program in Neuroscience and Harvard Stem Cell human pluripotent stem cells and their derivatives for clinical applications. Institute, Harvard Medical School, Belmont, MA, USA (YC, K-SK) Stem Cell Reports 2017; 9: 1–4. 19 Rossi A, Lickfett S, Martins S, Prigione A. A call for consensus guidelines on 1 Yamanaka S. Pluripotent stem cell-based cell therapy—promise and monitoring the integrity of nuclear and mitochondrial genomes in human challenges. Cell Stem Cell 2020; 27: 523–31. pluripotent stem cells. Stem Cell Reports 2022; 17: 707–10. 2 Daley GQ, Scadden DT. Prospects for stem cell-based therapy. Cell 2008; 20 Horak P, Griffith M, Danos AM, et al. Standards for the classification of 132: 544–48. pathogenicity of somatic variants in cancer (oncogenicity): joint 3 Park TY, Jeon J, Cha Y, Kim KS. Past, present, and future of cell replacement recommendations of Clinical Genome Resource (ClinGen), Cancer therapy for Parkinson’s disease: a novel emphasis on host immune Genomics Consortium (CGC), and Variant Interpretation for Cancer responses. Cell Res 2024; 34: 479–92. Consortium (VICC). Genet Med 2022; 24: 986–98. 4 Tabar V, Sarva H, Lozano AM, et al. Phase I trial of hES cell-derived 21 International Society for Stem Cell Research. Best practices for the dopaminergic neurons for Parkinson’s disease. Nature 2025; development of pluripotent stem cell-derived therapies. https://www.isscr. 641: 978–83. org/translational-best-practices (accessed Jan 27, 2026). 5 Chang JW, Na HK, Chang KW, et al. Phase 1/2a clinical trial of hESC-derived 22 Collins FS, Gottlieb S. The next phase of human gene-therapy oversight. dopamine progenitors in Parkinson’s disease. Cell 2025; 188: 7036–48.e11. N Engl J Med 2018; 379: 1393–95. 1494 --- [PDF原文](https://sci-net.xyz/storage/7932541/1f82d064715b5fed39bdda957a881fcee2e91f403daa99b0b1749c9638e10c0f/Safeguarding-genomic-integrity-in-pluripotent-stem-cell-therapies.pdf) DOI: 10.1016/S0140-6736(26)00095-4