University Children's Hospital Tübingen

The central importance of neutrophils for human health is clearly demonstrated by the clinical presentation of patients with severe congenital neutropenia (CN) and cyclic neutropenia (CyN). The majority of patients with CN with neutropenia as their only symptom do not survive the first year of life unless they are treated with recombinant human cytokine granulocyte colony-stimulating factor (rhG-CSF) or undergo allogeneic bone marrow transplantation. Understanding the processes of defective granulopoiesis downstream of CN-associated gene mutations sheds light on the mechanisms that regulate granulopoiesis and granulocyte functions. Furthermore, given that CN is a preleukemia syndrome, studying CN can help to define the role of defective granulopoiesis and deregulated neutrophil functions in leukemogenesis. The development of advanced experimental models of CN and CyN overcomes the limitations of insufficient patient material for research, leading to breakthroughs in the understanding of their pathogenesis and the development of new potential therapies.

Inflammation is a key driver of hematopoietic dysfunction in myeloid malignancies, but its role in the context of hypomethylating therapy remains incompletely understood. Although 5-Azacytidine is used posttransplant in high-risk myelodysplastic syndrome (MDS), only 50% of patients show a clinical response. We provide evidence that inherent inflammatory properties of healthy donor CD34+ stem cells exist that are likely to contribute to the "response" seen in MDS patients. These are linked to epigenetic priming of the myeloid niche, resulting in an S100A8/A9-driven inflammatory program that promotes functionality of immature NK cells. Using in vitro differentiation systems, multi-omic profiling, and an S100A9-/- mouse model, we find that 5-AzaC modulates inflammatory transcriptional programs through epigenetic rewiring of upstream regulatory elements. Loss of S100A9 disrupts myeloid differentiation, impairs NK cell maturation, and alters key developmental regulators including CEBPB, JUN, and NFIL3. In vivo, 5-AzaC restores these defects and primes NK cells in a time- and context-dependent manner. Re-analysis of the published Australian MDS/CMML cohort shows that "responders" display increased S100A8/A9 expression together with enhanced IFN-gamma, IL6-JAK-STAT3, and TNF signaling. These findings suggest that inflammatory myeloid programs may serve as predictive biomarkers and therapeutic targets to enhance NK cell-mediated graft-versus-leukemia activity posttransplant.

Dementia is a neurodegenerative disease that affects more than 45 million patients worldwide. Treatment options for dementia include lecanemab for mild cognitive impairment, acetylcholinesterase inhibitors, which are mainly used for mild to moderate dementia, as well as N-methyl-D-aspartate receptor antagonists for patients with severe dementia or certain antipsychotics. Pharmaceutical approaches are limited by potential side effects. To alleviate symptoms or at least slightly improve cognitive functions, additional herbal medicines (HM) such as Ginkgo biloba and non-pharmacological approaches, e.g., behavioral, exercise, music, and reminiscence therapy, yoga, tai chi, or acupuncture, have been used with heterogeneous results. Helleborus niger, a plant containing multiple bioactive compounds such as ecdysteroids and bufadienolide, is used clinically for its immunomodulatory, anti-inflammatory, analgesic, and antioxidant properties. In a German hospital specializing in geriatric patients, an association between the subcutaneous application of Helleborus niger 12x and changes in dementia was observed. Retrospective study on the effect of subcutaneous application of Helleborus niger 12x for an average of 3 weeks in patients with dementia. Eligible patients were aged ≥ 60 years and treated at the Humboldt Clinic in Bad Steben, Bavaria, Germany, with two different treatment regimens: standard care with conventional medication alone (control group, CG) vs. standard care with conventional medication plus Helleborus niger 12x preparations (Helleborus group, HG). The primary outcome was the between-group difference in pre-post (T0 vs. T1) changes in the Mini-Mental State Examination (MMSE). Secondary outcomes were pre-post changes in the Dementia Detection test (DemTect), the Shulman clock-drawing test (CDT) and the Geriatric Depression Scale (GDS). For the primary outcome, there was a statistically significant difference in MMSE between the groups (p < 0.001; d = 1.5) at T1 with significantly higher values in HG compared to CG. Regarding the secondary outcomes, significantly improved values were seen in HG compared to CG for DemTect, CDT and GDS with medium to high effect sizes. The study gives preliminary evidence that Helleborus niger 12x applied subcutaneously might improve cognitive function in patients with dementia at least during the three-week administration. Further randomized, blinded studies with longer follow-up are needed to confirm preliminary results. Identifier, DRKS00033972.

In-context learning (ICL) offers a promising paradigm for universal medical image analysis, enabling models to perform diverse image processing tasks without retraining. However, current ICL models for medical imaging remain limited in two critical aspects: they cannot simultaneously achieve high-fidelity predictions and global anatomical understanding, and there is no unified model trained across diverse medical imaging tasks (e.g., segmentation and enhancement) and anatomical regions. As a result, the full potential of ICL in medical imaging remains underexplored. Thus, we present Medverse, a universal ICL model for 3D medical imaging, trained on 22 datasets covering diverse tasks in universal image segmentation, transformation, and enhancement across multiple organs, imaging modalities, and clinical centers. Medverse employs a next-scale autoregressive in-context learning framework that progressively refines predictions from coarse to fine, generating consistent, full-resolution volumetric outputs and enabling multi-scale anatomical awareness. We further propose a blockwise cross-attention module that facilitates long-range interactions between context and target inputs while preserving computational efficiency through spatial sparsity. Medverse is extensively evaluated on a broad collection of held-out datasets covering previously unseen clinical centers, organs, species, and imaging modalities. Results demonstrate that Medverse substantially outperforms existing ICL baselines and establishes a novel paradigm for in-context learning.

Uncertainty is an inherent part of geoscience research and arises at multiple stages of the scientific process, from data collection and modelling to analysis and interpretation. In recent years, growing attention has been devoted to uncertainty quantification and assessment, alongside increasing recognition of the importance of uncertainty communication. These aspects are closely linked, as robust characterization of uncertainty provides an essential basis for transparent communication within the scientific community and beyond it.Communicating uncertainty not only plays a key role in improving the understanding of how scientific knowledge is produced, but can also help to foster trust by increasing transparency and contextualizing results. Nevertheless, reluctance to explicitly assess and communicate uncertainty persists, particularly when addressing non-expert audiences. This challenge is especially relevant in the context of natural hazard risk assessment and management: Here, adequate communication of uncertainties can add particularly valuable information for decision-making, risk governance, and a better understanding of the risks at hand among public audiences.This contribution presents an exploratory, database-driven overview of the scientific literature on uncertainty communication in geoscience, with a particular focus on natural hazards. Using structured queries in the Web of Science database, we examine publication trends over time, disciplinary distributions, thematic emphases, and possible blind spots. Keyword combinations range from general terms such as “uncertainty communication” and “multi-hazard communication” to more specific queries combining uncertainty, communication, and individual natural hazards (e.g., floods, earthquakes, droughts).Preliminary results indicate that uncertainty communication spans a broad range of scientific categories, while the level of attention varies substantially across hazard types, with flood-related studies being more prominent than others. Initial findings also suggest that multi-hazard uncertainty communication remains comparatively underrepresented, despite the increasing emphasis on multi-hazard and multi-risk assessments in recent research and policy frameworks. The growing volume of publications further highlights the need for systematic approaches to literature mapping, as well as the potential role of data-driven and AI-assisted tools in supporting such analyses.This research was partially funded by the European Civil Protection and Humanitarian Aid Operations (ECHO) of the European Commission (EC) through the VOLCAN project (ref. 101193100) and by the 2024 Research Prize of the Dr. K. H. Eberle Foundation to Mohadjer, Pelzer and Dietrich.

This article reports the results of the second iteration of the autoPET challenge on automated lesion segmentation in whole-body PET/CT, held in conjunction with the 26th International Conference on Medical Image Computing and Computer Assisted Intervention in 2023. In contrast to the first autoPET challenge, which served as a proof of concept, this study investigates whether machine learning-based segmentation models trained on data from a single source can maintain performance across clinically relevant variations in PET/CT data, reflecting the demands of real-world deployment. Methods: A comprehensive biomedical segmentation challenge on PET/CT domain generalization was designed and conducted. Participants were tasked to train machine learning models on annotated whole-body 18F-FDG data (n = 1,014). These models were then evaluated on a test set of 200 samples from 5 clinically relevant domains, including variations in institutions, pathologies, and populations and a different tracer. Performance was measured in terms of average dice similarity coefficient, average false-positive volume, and average false-negative volume. The best-performing teams were awarded in 3 categories. Furthermore, a detailed analysis was conducted after the challenge, examining results across domains and unique instances, along with a ranking analysis. Results: Generalization from a single-source domain remains a significant challenge. Seventeen international teams successfully participated in the challenge. The best-performing team reached an average dice similarity coefficient of 0.5038, a mean false-positive volume of 87.8388 mL, and a mean false-negative volume of 8.4154 mL on the test set. nnU-Net was the most commonly used framework, with most participants using a 3-dimensional U-Net. Despite competitive in-domain results, out-of-domain performance deteriorated substantially, particularly on pediatric and prostate-specific membrane antigen data. Detailed error analysis revealed frequent false-positives due to physiologic uptake and decreased sensitivity in detecting small or low-uptake lesions. A majority-vote ensemble offered minimal performance gains, whereas an oracle ensemble indicates hypothetical gains. Ranking analysis showed no single team consistently outperformed all others across ranking schemes. Conclusion: The second autoPET challenge provides a comprehensive evaluation of the current state of automated PET/CT tumor segmentation, highlighting both progress and persistent challenges of single-source domain generalization and the need for diverse public datasets to enhance algorithm robustness.

NUT carcinoma (NC) is a rare, yet aggressive disease (median survival of 6.5 months) of the young defined by a translocation of the nuclear protein in testis gene 1 (NUTM1). Neither surgery nor radiochemotherapy or targeted therapies provide effective disease control, thereby creating a need for innovative therapeutic strategies. Recently, the recombinant oncolytic virus (OV) VSV-GP entered phase I clinical testing (NCT05155332). We (1) visualized VSV-GPs' replication cycle in transmission electron microscopy; (2) analyzed VSV-GP's oncolytic efficacy and replication kinetics using viability assays, xCELLigence analyses, RT-qPCR, TCID50 assays, and FACS analysis; as well as (3) explored potential resistance mechanisms in seven human NC cell lines. Upon infection with VSV-GP, we noticed profound oncolysis, apoptosis, and high replication kinetics in five out of seven NC cell lines. Investigations of molecular determinants of resistance showed that VSV-GPs' entry receptor α-dystroglycan was equally expressed by VSV-GP-permissive and -resistant cell lines. While synthesis of intrinsic interferon (IFN)-β was found to be suppressed in VSV-GP-infected NC cell lines, intracellular IFN-β signaling was functional. Since VSV-GP is a promising compound for virotherapy of NC patients, exploration of virus-induced immunogenicity and identification of resistance mechanisms could pave the way for additional combination treatment regimes in NC.

Dynamic Patient and Situation Simulation in the Context of Mass Casualty Incidents -Insights from the Research Projects D2PuLs and D2PuLs PRO builds directly on this approach by presenting a modular digital training environment for mass-casualty incidents. Dynamic patient models, which react physiologically to interventions, are linked with situation and manikin simulators so that complete rescue chains-from prehospital triage to in-hospital trauma carecan be trained. The study demonstrates that such digital simulations can supplement resource-intensive live exercises and generate detailed data for debriefing and quality improvement.The systematic review Extended Reality Technology for Emergency Medical Service Training: Systematic Review adds an evidence-synthesis perspective on immersive training tools. The review reports that virtual, augmented, and mixed-reality applications can improve engagement, perceived realism, and skill acquisition in emergency medical services, especially for rare or high-risk scenarios. At the same time, it highlights methodological limitations of existing studies and practical barriers such as ergonomic issues, user fatigue, and technical instability. Taken together, these three articles support the role of simulation and extended reality as important elements of disaster preparedness, while emphasising the need for structured curricular integration and robust evaluation.A second set of contributions focuses on artificial intelligence (AI) and data-driven decision support. Applications of Artificial Intelligence-Guided Clinical Decision Support in Disaster Medicine: An International Delphi Study explores, using a Delphi process, which AI functions are prioritised by disaster medicine experts. The study finds broad agreement that AI is particularly useful for improving situational awareness and logistics-such as estimating affected populations, supporting hazard vulnerability analysis, allocating scarce resources, and optimising patient distribution-whereas there is significantly more caution regarding AI involvement in ethically sensitive triage or end-of-life decisions. These results provide important guidance for the design of AI systems that align with practitioner expectations.Anomaly detection and early risk identification in digital disaster response-based on deep learning in public health proposes a deep-learning framework to identify aberrant patterns in health-related data streams. By combining temporal and contextual modelling, the approach improves detection performance and reduces false alarms in comparison with traditional methods, illustrating how AI can support early recognition of emerging threats in digital operations centres. Filling the gap: Artificial Intelligence-driven One Health integration to strengthen pandemic preparedness in resource-limited settings extends the focus to One Health. This article discusses AI-enabled architectures that integrate human, animal, and environmental data for improved surveillance in low-and middle-income countries, and emphasises governance, capacity building, and offline-capable, context-sensitive tools as prerequisites for effective deployment. Finally, The Information Challenge in Public Health Crises: A Study on the reliability and readability of information provided by large language model for thunderstorm asthma examines large language models as direct information sources in crises. By assessing reliability and readability of responses to questions on thunderstorm asthma, the study shows that, although medically relevant content can often be generated, readability frequently exceeds recommended levels for lay audiences and quality differs between systems. The authors conclude that current models should not be used as unsupervised tools for public crisis communication and that professional review and adaptation to target groups are essential.Overall, the eleven articles in this Research Topic demonstrate that digital innovations already contribute to multiple core domains of disaster medicine: competence development through simulation and extended reality, AI-supported situational awareness and One Health surveillance, new service-delivery models via telemedicine and drones, and enhanced information processing using sentiment analysis, named entity recognition, and large language models. At the same time, they consistently show that technological sophistication alone is insufficient. Acceptance by professionals, integration into organisational structures, rigorous evaluation in real-world environments, ethical and legal governance, and a strong focus on equity-particularly in resource-limited settings-are decisive for whether digital tools will truly bridge gaps and improve outcomes in disasters.

Despite considerable advancements in recent decades, mortality and complications following liver resection remain high. The volume-outcome relationship has been the subject of extensive research and offers relevant potential for improvement of surgical outcomes. This review aims to examine the impact of hospital and surgeon volume on patient-relevant outcomes in liver resections and synthesize the available evidence. A rapid systematic literature review was conducted, searching CENTRAL, Embase, PubMed, and study registries for articles published from 2000 to 2023. Eligible studies investigated the association between hospital or surgeon volume and patient-relevant outcomes in anatomical and non-anatomical liver resections. Study quality was assessed using the ISPOR and ROBINS-E checklists and reported alongside the results. The review protocol registered with PROSPERO (CRD42023398566). The search yielded 3287 records, of which 38 publications met the inclusion criteria. All included studies were retrospective observational studies. A higher surgical volume was associated with improved patient-relevant outcomes, such as reduced mortality following both anatomical and non-anatomical liver resections and lower rates of postoperative complications. However, the results indicate that the impact of hospital or surgeon volume is limited and likely depends on the respective outcome parameter. A considerable gap remains with respect to long-term outcomes and quality of life, and studies investigating surgeon volume are scarce. The findings provide evidence supporting a positive association between higher hospital volume and improved patient-relevant outcomes in liver resection. However, surgeon volume remains underexplored and the evidence from subgroups indicates that the impact of hospital or surgeon volume likely depends on study quality, procedure type, volume thresholds, and respective outcome parameters. Patient care could benefit from further research on long-term outcomes as well as quality of life, for which the current evidence is scarce.

Abstract Background: Current data support that metastatic hormone receptor-positive (HR+) and HER2-positive (HER2+) breast cancer offers distinct treatment options targeting both the hormone receptor and the HER2-pathway. This might lead to increased treatment efficacy with the potential of omitting chemotherapy in these patients. The aim of the randomized multicenter phase III DETECT V trial was to analyze the efficacy of chemotherapy free and CDK4/6 inhibitor containing treatment strategies compared to current standard treatment. Methods: Between 09/2015 and 11/2022, 271 patients with HER2+ and HR+ MBC in the 1st-3rd line setting were randomized 1:1 to receive trastuzumab and pertuzumab combined with either endocrine therapy or chemotherapy followed by maintenance therapy with dual HER2 targeted and endocrine therapy. Chemotherapy and the endocrine agents could be chosen from a variety of available regimens according to physicians’ choice. In January 2019, the protocol was amended with the addition of the CDK4/6 inhibitor ribociclib to both treatment arms. The primary objective of DETECT V was to compare tolerability between the treatment arms, and secondary objectives comprised the comparison of overall survival (OS), progression-free survival (PFS) and safety, as well as the evaluation of the effect of adding ribociclib to both treatment arms. Here we report results of the final efficacy analyses as based on the modified ITT set of 262 patients (9 patients did not receive any study treatment and were excluded). Of these, 120 patients were recruited before the addition of ribociclib (59 and 61 patients in the chemotherapy-free and chemotherapy-containing treatment arm, respectively), and 142 patients were recruited after the addition of ribociclib (73 and 69 patients in the chemotherapy-free and chemotherapy-containing treatment arm, respectively). Results: Median patient age was 60 years, 202 (77.1%) of patients were in the first line setting and 137 (52.3%) patients had a metastasis-free interval (from primary diagnosis) exceeding 12 months. The overall comparison between all patients receiving chemotherapy-free and chemotherapy-containing treatment showed no significant difference with regard to OS and PFS (median OS not reached vs 46.1 months, HR 0.98, 95% CI 0.60 - 1.59, p = 0.93; median PFS 19.5 vs 23.0 months, HR 1.15, 95% CI 0.82 - 1.62, p = 0.42). However, both OS and PFS were significantly improved by the (non-randomized) addition of ribociclib (OS: median OS not reached vs 46.1 months, HR 0.43, 95% CI 0.26 - 0.72, p = 0.001; PFS: median PFS 29.7 vs 15.6 months, HR 0.48, 95% CI 0.34 - 0.67, p &amp;lt; 0.001). A separate analysis for the two treatment arms showed that the addition of ribociclib to chemotherapy-containing treatment significantly improved both OS (median OS not reached vs 38.7 months, HR 0.30, 95% CI 0.14 - 0.64, p = 0.002) and PFS (median PFS 33.1 vs 15.4 months, HR 0.35, 95% CI 0.21 - 0.59, p &amp;lt; 0.001), while the addition of ribociclib to chemotherapy-free treatment significantly improved PFS (median PFS 27.2 vs 15.6 months, HR 0.61, 95% CI 0.38 - 0.98, p = 0.040) but not OS (median OS not reached in both groups, HR 0.60, 95% CI 0.30 - 1.23, p = 0.163). An exploratory cross-cohort comparison revealed significantly improved outcome for patients receiving chemotherapy-free treatment plus ribociclib vs chemotherapy-containing treatment without ribociclib (OS: median OS not reached vs 38.7 months, HR 0.48, 95% CI 0.24 - 0.94, p = 0.033; PFS: median PFS 27.2 vs 15.4 months, HR 0.53, 95% CI 0.33 - 0.85, p = 0.008). Conclusion: Our results suggest that chemotherapy-free treatment for patients with HER2+ and HR+ MBC may be a valuable and effective treatment alternative, and may improve survival with the addition of ribociclib. Citation Format: W. Janni, T. Fehm, V. Müller, J. Blohmer, A. De Gregorio, T. Decker, A. Hartkopf, N. Ditsch, M. Schmidt, P. Wimberger, T. Engler, M. Banys-Paluchowski, P. A. Fasching, B. Rack, A. Schneeweiss, K. Pantel, T. W. Friedl, F. Schochter, J. Huober. Efficacy Analysis of the Randomized Phase III DETECT V trial: Treatment De-escalation by Omission of Chemotherapy and the Effect of Adding Ribociclib in HER2-positive and Hormone-receptor Positive Metastatic Breast Cancer [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr RF4-02.