SEARCH · Search PubMed

Results for “External validity”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 235 records · Page 13Linked to original sources

PubMed2026

Predicting ACL injury risk in athletes: A systematic review of machine learning-based models.

BACKGROUND: Early ACL injury risk identification in athletes is essential. This systematic review examines machine learning (ML) models for predicting ACL injuries, evaluating their methodological quality, performance, and reliability. METHOD: A comprehensive electronic search was conducted across PubMed, Scopus, Web of Science, and IEEE Xplore databases, supplemented by Google Scholar for grey literature, covering articles published between January 1, 2015, and August 30, 2025. Eligible studies were appraised using the Prediction Model Study Risk of Bias Assessment Tool (PROBAST) for methodological quality and risk of bias, and the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) guidelines for quality of evidence. RESULTS: Ten studies were included. PROBAST showed eight studies had moderate risk of bias and two low risk. TRIPOD found only two studies met quality criteria. ML models included logistic regression (n = 5), support vector machines (n = 4), k-nearest neighbor (n = 3), decision trees (n = 3), random forests (n = 5), neural networks (n = 2), linear discriminant analysis (n = 1), and pre-trained CNNs (n = 1). AUC ranged from 0.63 to 0.98. Accuracy (reported in six studies) ranged from 26% to 95%; however, these values should be interpreted with caution due to the absence of confidence intervals, lack of class imbalance handling, and limited external validation across studies. Tree-based ensemble methods such as random forest achieved competitive accuracy (74-86%), while SVM, a non-ensemble classifier, reported accuracy ranging from 71% to 95%; however, the highest values were obtained in studies with notably small sample sizes (n = 12 to n = 39), raising concerns about overfitting and generalizability. CONCLUSION: Current ML algorithms show promise for identifying athletes at high ACL injury risk and detecting relevant risk factors. Although study quality was generally satisfactory, future research should prioritize external validation and model interpretability to support clinical translation.

Humans↗

PubMed2026

Identification of Drug-resistant Cell Subpopulations in Colorectal Cancer Through Single-cell Analysis and Exploration of Potential Therapeutic Strategies.

INTRODUCTION: The therapeutic efficacy of Colorectal Cancer (CRC) is often compromised by resistance to the standard chemotherapy agent oxaliplatin. METHODS: This study obtained single-cell RNA sequencing (scRNA-seq) data from the Gene Expression Omnibus (GEO) database. Differentially Expressed Genes (DEGs) between resistant and sensitive epithelial subpopulations were identified, followed by enrichment analysis. Pseudotemporal trajectory and cell-cell communication were analyzed using Monocle2 and CellChat, respectively. The candidate drug was predicted by Connectivity Map (cMAP) analysis. External validation included assessment of the EpC2 signature in an oxaliplatin-resistant cell line dataset (GSE76092), survival analysis using The Cancer Genome Atlas (TCGA) cohorts, and re-analysis of the GSE179784 dataset to assess the reproducibility of EpC2-like subpopulations and their DNA Damage Repair (DDR) scores. RESULTS: Cell subpopulations were divided into 10 clusters. Among them, epithelial cells comprised 5 subpopulations, with EPC2 identified as a potential oxaliplatin-resistant subset. DEGs were enriched in the TNF and IL-17 pathways. External validation confirmed the enrichment of EpC2 in resistant cell lines and its association with poor survival. Pseudotemporal trajectory revealed that epithelial cells underwent state transitions, forming two distinct branches. The resistant group exhibited enrichment in RNA splicing and NF-κB pathways. Cell-cell communication analysis revealed interactions involving MDK- NCL and PPIA-BSG. Dasatinib was predicted as a candidate drug. DISCUSSION: We identified an oxaliplatin-resistant subpopulation of Epithelial Cells (EpC2) in CRC, elucidated its multi-layered resistance mechanisms, and integrated multi- omics and cMAP database analyses to predict a potential intervention drug. CONCLUSION: This study provided potential therapeutic possibilities for oxaliplatin resistance, contributing to CRC treatment.

Humans↗

PubMed1993

A Pathfinder analysis of pedagogical knowledge structures: a follow-up investigation.

This study is an extension of an earlier investigation of undergraduate students' acquisition of key pedagogical concepts in a physical education teaching methodology course. In that study, Pathfinder, a method for eliciting associative memory networks, was used to describe and compare the pedagogical knowledge structures of students to that of the course instructor. After the course, students' pedagogical knowledge structures corresponded more closely with that of the instructor, and students who corresponded the closest performed better in the course. The results raised an interesting issue regarding the acquisition of knowledge in undergraduate students. Did students acquire a generalizable body of pedagogical knowledge applicable beyond the context of the teaching methodology course or a highly contextualized reflection of their course instructor's knowledge base? In the present study the external validity of the pedagogical knowledge base was examined by using Pathfinder to compare the knowledge structures of students from the initial investigation with knowledge structures of five experienced teacher educators from five different teacher education programs. The findings indicated that students' knowledge structures became significantly more correspondent with that of the experienced teachers' structures from the beginning to the end of the course. Also, students' correspondence with teacher educators' structures following instruction was found to be significantly correlated with academic and teaching performance. The findings point to the external validity of the domain of knowledge under study and the robustness of Pathfinder for capturing pedagogical knowledge.

Adult↗

PubMed1985

Attrition in prevention research.

Selective attrition can detract from the internal and external validity of longitudinal research. Four tests of selective attrition applicable to longitudinal prevention research were conducted on data bases from two recent studies. These tests assessed (1) differences between dropouts and stayers in terms of pretest indices of primary outcome variables (substance use), (2) differences in change scores for dropouts and stayers, (3) differences in rates of attrition among experimental conditions, and (4) differences in pretest indices for dropouts among conditions. Results of these analyses indicate that cigarette smokers, alcohol drinkers, and marijuana users are more likely to drop out than nonusers, limiting the external validity of both studies. For one project, differential rates of attrition among conditions suggested a possible attrition artifact which will interfere with interpretation of outcome results, possibly masking true program effectiveness. Recommendations for standardizing reports of attrition and for avoiding attrition through second efforts are made.

Adolescent↗

PubMed2001

Cheminformatic models to predict binding affinities to human serum albumin.

Models to predict binding affinities to human serum albumin (HSA) should be very useful in the pharmaceutical industry to speed up the design of new compounds, especially as far as pharmacokinetics is concerned. We have experimentally determined through high-performance affinity chromatography the binding affinities to HSA of 95 diverse drugs and druglike compounds. These data have allowed us the derivation of quantitative structure-activity relationship models to predict binding affinities to HSA of new compounds on the basis of their structure. Simple linear, one-variable models have been derived for specific families of compounds (r(2) > or = 0.80; q(2) > or = 0.62): beta-adrenergic antagonists, steroids, COX inhibitors, and tricyclic antidepressants. Also, global models have been derived to be applicable to the whole medicinal chemical space by using the full database of HSA binding constants described above. For this aim, a genetic algorithm has been used to exhaustively search and select for multivariate and nonlinear equations, starting from a large pool of molecular descriptors. The resulting models display good fits to the experimental data (r(2) > or = 0.78; LOF < or = 0.12). In addition, both internal (cross validation and randomization) and external validation tests have demonstrated that these models have good predictive power (q(2) > or = 0.73; PRESS/SSY < or = 0.23; r(2) > or = 0.82 for the external set). Statistical analysis of the equation populations indicates that hydrophobicity (as measured by the ClogP) is the most important variable determining the binding extent to HSA. In addition, structural factors (especially the topological (6)chi(ring) index and some Jurs descriptors) also frequently appear as descriptors in the best equations. Therefore, binding to HSA turns out to be determined by a combination of hydrophobic forces together with some modulating shape factors. This agrees with X-ray structures of HSA alone or bound to ligands, where the binding pockets of both sites I and II are composed mainly of hydrophobic residues.

Adrenergic beta-Antagonists↗

PubMed2026

ceRNA network of lncRNAs and mRNAs in OSF-to-OSCC progression: Diagnostic biomarkers and functional pathways.

BACKGROUND: Oral submucous fibrosis (OSF) is a chronic potentially malignant disorder that can progress to oral squamous cell carcinoma (OSCC). Although dysregulated non-coding RNAs have been implicated in oral carcinogenesis, the competing endogenous RNA (ceRNA)-mediated regulatory mechanisms underlying OSF-to-OSCC progression remain poorly understood. This study aimed to identify candidate regulatory molecules and construct a putative lncRNA-miRNA-mRNA network associated with malignant transformation. METHODS: Publicly available microarray datasets (GSE117973 and GSE125866) were analyzed to identify differentially expressed genes between OSF and OSCC. Differentially expressed transcripts were classified into mRNAs and lncRNAs based on public transcript annotations. Highly correlated lncRNA-mRNA pairs were identified using Pearson correlation analysis and integrated with multiMiR-supported miRNA-mRNA interactions obtained from public databases to construct a putative ceRNA regulatory network. Functional characterization focused on apoptosis, epithelial-mesenchymal transition (EMT), and immune checkpoint-related pathways. Receiver operating characteristic (ROC) analysis was performed to evaluate diagnostic performance, and selected biomarkers were externally validated using The Cancer Genome Atlas (TCGA) OSCC cohort. RESULTS: Integrated transcriptomic analysis identified several dysregulated mRNAs and lncRNAs associated with OSF-to-OSCC progression. Network analysis highlighted TBC1D3B, RREB1, TEAD3, SREBF1, TMEM41B, FOXK2, and KIAA1958 as prominent hub genes within the putative regulatory network. Functional analyses demonstrated significant associations with apoptosis-, EMT-, and immune checkpoint-related genes, suggesting potential involvement in multiple biological processes contributing to malignant transformation. Several hub genes exhibited strong diagnostic performance, with ROC analysis yielding AUC values ranging from 0.891 to 1.000, indicating excellent discrimination between OSF and OSCC samples. External validation using TCGA further supported the relevance of the identified biomarkers in OSCC. CONCLUSIONS: This study provides a comprehensive transcriptomic framework describing putative lncRNA-miRNA-mRNA regulatory interactions associated with OSF progression to OSCC. The identified hub genes and regulatory networks represent candidate biomarkers for early detection and provide a foundation for future mechanistic and experimental validation. As the proposed ceRNA interactions are computationally inferred, further biological validation is required before clinical application.

RNA, Long Noncoding↗

PubMed1991

An empirical subgrouping of Finnish learning-disabled children.

The internal and external validity of a subgrouping of 82 Finnish children with relatively mild learning disabilities and 84 Controls was explored. The sample was selected from a total population of 1607 second grade pupils. Eight neuropsychological measures from four function areas were selected as classification criteria in cluster analysis. Six consistent and clinically meaningful subgroups were derived. These subgroups were designated as follows: (1) Normal, (2) General Language, (3) Visuo-Motor, (4) General Deficiency, (5) Naming, and (6) Mixed. Most of the LDs were clustered in subgroups (2) through (6); and most of the Controls, in subgroup (1). Several internal and external validation procedures indicated at least moderate validity in the subgroups, with the exception of the Mixed subgroup. The five valid subgroups encompassed 82% of the LDs and 90% of the Controls, and moreover, resembled subgroups which previously have been found among English-speaking children. This suggests that language differences exert no significant effect on the types of the emerging subgroups.

Aging↗

PubMed1998

The problem of protocol driven costs in pharmacoeconomic analysis.

The increasing number of economic evaluations of healthcare interventions and of drug therapies in particular has been well documented. Surveys of the quality of studies have demonstrated that standards of conduct of such studies have not similarly increased. Concerns over the standards have led to increased calls that economic analyses be more closely linked to randomised controlled clinical trials (RCT). Seven potential threats to the external validity of results limit the generalisability of studies based on RCTs. One such threat is the existence of protocol driven costs. There are two main types of protocol driven costs. Protocol prescribed costs arise as a result of resource use mandated by the clinical trial design. Protocol derived costs occur when increased clinical investigations mandated by trial protocols lead to atypical disease management. Methods to control for protocol driven costs within pharmacoeconomic study designs are available. Modelling studies can be based on data within clinical trials combined with observational data representing more typical resource use. The adoption of pragmatic clinical trial designs provide greater external validity though reduced internal validity. Refinements to explanatory clinical trials can also lead to reduced protocol driven costs. The extent that current studies control for such costs is unclear due to the lack of transparency in the reporting of study methods. A review of published studies found little consideration of protocol driven costs although in several studies there was evidence of their existence. Future studies conducted alongside RCTs should explicitly address how the issue of protocol driven costs was handled within the study framework.

Economics, Pharmaceutical↗

PubMed2026

A comparative evaluation of multiple enlarged perivascular space segmentation tools.

BACKGROUND: Enlarged perivascular spaces (ePVS) are a marker of cerebral small vessel disease, potentially reflecting reduced waste clearance. Because manual quantification is unfeasible in large datasets, we developed and evaluated an automated tool. METHODS: Detection Of Regions of Enlarged perivascular Spaces (DORES), a 3D nnU-Net-based deep learning algorithm was developed for ePVS segmentation using T1-weighted and fluid-attenuated inversion recovery magnetic resonance imaging (MRI). DORES was developed in two stages: an initial model trained on 35 manually segmented scans and a final model on 1460 pseudo-labeled sessions from the Vanderbilt Memory and Aging Project (VMAP). A subset of VMAP participants with 3 T brain MRI underwent whole-brain manual ePVS tracing (n = 35, 73 ± 9 years, 51% male) and visual rating (n = 388, 71 ± 8 years, 54% male) by a neuroradiologist. DORES was evaluated and compared against three other segmentation tools using Dice and F1 scores, absolute volume and element differences, correlation, and agreement. External validation used an Alzheimer's Disease Neuroimaging Initiative 3 subset with manual tracings (ADNI3, n = 18, 73 ± 9 years, 67% female). RESULTS: DORES achieved Dice scores of 0.61 ± 0.16 (white matter) and 0.72 ± 0.08 (basal ganglia) in VMAP, with strong correlations and agreement for ePVS count and volume. Performances modestly declined in ADNI3 across algorithms. Scanner-stratified analyses showed stronger correlations for Philips versus Siemens images in the basal ganglia, indicating scanner-dependent differences in measurement consistency. CONCLUSIONS: DORES provides a multimodal nnU-Net-based pipeline for ePVS segmentation in older adults. The model demonstrates robust within-cohort performance and reasonable external validity, though scanner-related effects limit application across sites.

Humans↗

PubMed2026

Systemic Proteome Profiling to Differentiate Primary Glomerular Diseases.

KEY POINTS: Plasma proteome profiling identified distinct signatures across biopsy-proven primary glomerular disease subtypes. An elastic net model using 93 proteins classified primary glomerular disease subtypes and controls, with external validation. Integrating proteomics with machine learning yields biologically interpretable insights in primary glomerular diseases. BACKGROUND: Primary GN is a heterogeneous group of kidney disorders where understanding of their pathophysiology remains incomplete. Despite the diagnostic potential of high-throughput proteomics, constrained proteomic depth and a reliance on binary comparisons have left the feasibility of using systemic signatures to differentiate multiple GN subtypes largely unexplored. METHODS: To identify protein signatures that noninvasively differentiate major primary glomerular disease subtypes and provide mechanistic insights, we performed large-scale systemic proteome profiling of 5416 plasma proteins via Olink Explore HT in a discovery cohort ( n =147) and an external validation cohort ( n =85) of Korean participants (mean age, 41±13 years; 46% female). The study population included patients with four GN subtypes-focal segmental glomerulosclerosis, IgA nephropathy, minimal change disease, and membranous nephropathy-alongside healthy controls. We developed a machine learning (ML) model using logistic regression with elastic net regularization to classify disease groups based on proteomic profiles and evaluated its performance in the independent validation cohort. RESULTS: Plasma proteome profiles were distinct among disease subtypes, emerging as a significant source of data variation independent of conventional markers such as eGFR or proteinuria levels. The ML model performed robustly in both the discovery and validation cohorts, achieving an area under the receiver operating characteristic curve >0.8 for differentiating minimal change disease, membranous nephropathy, and IgA nephropathy. The model, even without clinical information, correctly identified 93% of minimal change disease cases (14 of 15) and 63% of IgA nephropathy cases (20 of 32), but its performance was limited for focal segmental glomerulosclerosis, with only 21% of cases (three of 14) correctly classified. Functional analysis of key proteins highlighted distinct biologic pathways, such as hemostasis in minimal change disease. CONCLUSIONS: We identified distinct systemic proteome signatures for primary glomerular diseases, where disease subtype served as a major determinant of proteomic variance alongside conventional clinical markers. ML models demonstrated robust discriminatory performance for minimal change disease, membranous nephropathy, and IgA nephropathy, underscoring the potential for proteome-based classification.

Humans↗

PubMed2000

Ethnic representation in a sample of the literature of applied psychology.

A number of authors have raised concerns over the external validity of psychological research. This study examined the extent to which empirical articles include human participants from diverse ethnic backgrounds. Articles published over a 5-year period in 14 selected journals representing 3 applied sub-disciplines of psychology were examined. Of the 2,536 articles coded, only 61% indicated the ethnicity of the participants. For those articles, the ethnic compositions approximated U.S. Census estimates, with the exception of an overrepresentation of African Americans and an underrepresentation of Hispanic Americans. The results imply that although the field is apparently adequately recruiting English speakers, representation of non-English speakers should be increased. To further enhance the external validity of psychological research, ethnicity of participants should be not only specified but also analyzed in relation to the results of a study.

Authorship↗

PubMed2000

[What is the evidence with regard to the effectiveness of orlistat?].

BACKGROUND: Obesity is worldwide one of today's most important medical and public health problems. Orlistat (Xenical) is a relatively new drug in the pharmacological treatment of obesity which partially blocks fat absorption. The following article analyses the available evidence of orlistat's effectiveness in the treatment of obese patients. METHODS: Three randomised controlled trials investigating the effect of orlistat in the treatment of obesity were identified by systematic Medline search. The internal and external validity of these studies was assessed using systematic criteria. RESULTS: All three studies consistently demonstrate a treatment benefit of orlistat compared to placebo. Patients treated with orlistat lost an average of 3.4 kg more than patients taking placebo while on a hypocaloric diet. Simultaneously, control of cardiovascular risk factors improved independently of the observed weight loss. Up to 40% of all patients experienced gastrointestinal side effects which were generally well tolerated. The studies prove that treatment with orlistat can result in a moderate weight reduction. However, the results of the studies cannot be easily generalised to obese patients in a primary care setting, due to limitations concerning the studies' internal and external validity. CONCLUSIONS: Based on these studies orlistat is an efficient pharmacological treatment for obesity in patients adhering to a hypocaloric diet. Studies demonstrating orlistat's effectiveness in a primary care setting are so far lacking. From a public health perspective there is a need for a randomised controlled trial showing not only orlistat's effectiveness on surrogate markers in a primary care setting but, ideally, a reduction in obesity-related mortality and morbidity.

Anti-Obesity Agents↗

PubMed2026

Resolution-dependent self-supervised transfer in chest radiograph classification.

BACKGROUND: Self-supervised learning (SSL) has improved visual representation learning, but its value in chest radiography remains uncertain. DINOv3 extends earlier SSL models through Gram-anchored self-distillation and explicit high-resolution adaptation. Whether these changes improve transfer learning for chest radiograph classification has not been established. METHODS: We benchmarked DINOv3 against DINOv2 and supervised ImageNet initialization across seven chest radiograph datasets comprising 816,183 radiographs from pediatric and adult cohorts. ViT-B/16 and ConvNeXt-B were evaluated under full fine-tuning at 224 × 224 and 512 × 512 pixels, with targeted 1024 × 1024 experiments on three cohorts. Additional analyses examined parameter-efficient adaptation, synthetic label corruption, external validation, frozen 7B features, and computational efficiency. The primary outcome was the mean area under the receiver operating characteristic curve across labels. RESULTS: In adult cohorts, DINOv3 did not consistently outperform DINOv2 at 224 × 224 pixels, but became the strongest initialization at 512 × 512 pixels, especially with ConvNeXt-B. Gains were greatest for small focal and boundary-dependent abnormalities, whereas large-structure findings changed little. The pediatric cohort showed no significant benefit from DINOv3, higher resolution, or backbone choice. Scaling to 1024 × 1024 rarely improved performance and markedly increased computational cost. ConvNeXt-B remained superior to ViT-B/16 under both full and parameter-efficient adaptation. External validation preserved the 512 × 512 DINOv3 advantage, whereas synthetic label corruption showed that this benefit should not be interpreted simply as superior noise robustness. Frozen DINOv3-7B features underperformed relative to fully adapted 86 to 89M-parameter backbones. CONCLUSIONS: For adult chest radiograph classification, DINOv3 provides its most reliable benefit at 512 × 512 pixels, particularly with ConvNeXt-B. Fully adapted mid-sized models at 512 × 512 pixels provided the best performance-cost trade-off in our benchmark.

Journal Article↗

PubMed1997

Analysis of alcohol use clusters among subcritically injured emergency department patients.

OBJECTIVES: 1) To cluster patients according to self-reported drinking patterns using cluster analysis; 2) to externally validate clustered groups on variables related to drinking but not used in the cluster analysis; and 3) to use the clustered patients' responses to alcohol consumption questions to develop a brief screening tool emergency physicians can use to identify patients in need of referral or intervention related to potentially hazardous alcohol consumption. METHODS: A self-report battery was administered to 95 subcritically injured patients. Patients also were saliva alcohol-tested upon arrival to the ED. Using the patients' self-reported quantity, frequency of alcohol consumption, and frequency of having > or = 6 drinks on a drinking occasion, patients were categorized into 3 groups using cluster analysis. The 3 clusters were externally validated using injury-related variables, alcohol-related consequences, and the patients' reported readiness to change drinking. A screening tool was developed using cutoff values reported by the patients' answers to drinking pattern questions. RESULTS: Fifty-nine patients were alcohol-negative, and 36 tested alcohol-positive (i.e., > 4 mmol/L [> 20 mg/dL]) or had elevated scores on an alcohol problem screening instrument. Three distinct drinking pattern clusters were found. Clusters were validated using discriminant function analysis and multivariate analyses of variance to confirm cluster classifications. Steady and high-intensity drinkers reported more alcohol-related negative consequences, and high-intensity drinkers indicated they would consider changing their drinking. The screening tool correctly classified 97% of the patient sample into their respective clusters. CONCLUSIONS: Using the drinking pattern questions in the clustering procedure was effective for grouping injured patients into clusters that could be differentiated on other drinking-related variables. The resulting screening tool can be used in the ED setting to screen patients for further assessment and intervention. The readiness-to-change results support the assertion that the injury event provides a "teachable moment" for subcritically injured patients whose injury may be related to their alcohol consumption.

Alcohol Drinking↗

PubMed2026

Deep learning-based cross-attention fusion of multimodal MRI for survival prediction and risk stratification in IDH-wildtype glioblastoma: a multicenter study.

BACKGROUND: Glioblastoma (GBM) exhibits profound molecular and spatial heterogeneity, complicating prognostic evaluations. While multiparametric MRI provides crucial multidimensional biological information, conventional end-to-end deep learning integration strategies, such as early or late fusion, often fail to capture complex nonlinear cross-modal interactions. We aimed to systematically evaluate a cross-attention fusion (CAF) architecture for GBM survival prediction and quantify its incremental prognostic value relative to existing clinical tools. METHODS: In this multicenter retrospective study, 386 adults with IDH-wildtype, WHO grade 4 GBM were assembled from an institutional cohort (n = 226), the Chinese Glioma Genome Atlas (CGGA, n = 62), and The Cancer Genome Atlas (TCGA, n = 98). Using a unified 3D ResNet-18 backbone, we compared single-modality models, early fusion, late fusion, and CAF on preoperative T1-weighted, contrast-enhanced T1-weighted (T1CE), and T2-weighted MRI, and integrated the resulting deep learning risk score with routine clinical variables through multivariable Cox regression. Performance was assessed using Harrell's C-index, time-dependent AUC, and decision curve analysis. RESULTS: CAF showed numerically higher, more consistent C-index trends than early fusion, late fusion, and single-modality models (pooled C-index 0.629, 95% CI 0.594-0.664), although pairwise differences in time-dependent AUC were not statistically significant. Integrating clinical variables raised the pooled C-index to 0.691 (95% CI 0.660-0.721) in the treatment-era model, with comparable performance across the three cohorts (Local 0.688; CGGA 0.716; TCGA 0.689); a pre-treatment configuration excluding adjuvant therapy yielded a pooled C-index of 0.642. Under leave-one-cohort-out external validation, the combined model retained significant risk stratification in all held-out cohorts (C-index 0.63-0.71; all log-rank P < 0.01), albeit with attenuated discrimination. The deep learning risk score remained independent after multivariable adjustment (HR 1.41 per SD, 95% CI 1.26-1.57; P < 0.001). Kaplan-Meier analysis confirmed significant high- versus low-risk separation in all cohorts, and decision curve analysis showed greater net benefit than clinical-only and deep-learning-only models. CONCLUSION: The CAF-derived risk score offers prognostic information complementary to routine clinical variables, representing a promising noninvasive tool for individualized risk stratification when molecular profiling is incomplete or unavailable; these findings warrant prospective external validation before clinical use.

cross-attention fusion↗

PubMed2003

The incorporation of potential confounding variables in Markov models.

OBJECTIVE: To improve the quality of the methods used in Markov modelling studies by increasing the external validity by means of the incorporation of confounding variables. STUDY DESIGN: The concepts were illustrated using a hypothetical Markov model for Parkinson's disease. METHODS: The methodology consisted of incorporation of an extra explanatory variable in the Markov health states by means of health state-specific relationships between this explanatory variable and costs as well as time-dependent values of the extra explanatory variable. In addition, we determined the relevance of the incorporation of an extra explanatory variable by means of various sensitivity analyses. RESULTS: The results showed that the outcomes of a health economic model may be severely biased, when a confounding effect of an extra explanatory variable is not taken into account. Hence the external validity of Markov models may be limited, and consequently the results of the model are not an accurate reflection of reality. CONCLUSION: This study proves the need for the incorporation of all relevant explanatory variables in a health economic model.

Antiparkinson Agents↗

PubMed2026

Integrated Genomic and Proteomic Analysis Reveals T-B Lymphocyte Signatures in the MYCN Driven "Immune Desert" of Specific Neuroblastoma Subtypes.

AIMS: This study aims to systematically dissect how MYCN amplification shapes the immunosuppressive tumor microenvironment (TME) in high-risk neuroblastoma, elucidating key mechanisms underlying immune evasion. METHODS: We performed an integrated multi-omics analysis of bulk RNA-seq (n = 721), single-cell RNA-seq (n = 9), proteomic data (n = 49) and spatial transcriptomics (Visium, with external validation in melanoma). Analyses included unsupervised clustering, cell-cell communication inference, transcriptional regulatory network reconstruction, and spatial proximity assessment to map the immune landscape. RESULTS: A distinct molecular subtype (Class C), defined by MYCN amplification and poor prognosis, exhibited a comprehensive "immune desert" phenotype characterized by low immune scores and minimal leukocyte infiltration. Single-cell analysis confirmed significant depletion of T and B lymphocytes within the Class C TME. Dysregulated transcriptional networks were identified, including upregulation of REL and EOMES in T cells-with EOMES potentially driving exhaustion via regulation of Transient Receptor Potential (TRP) genes, and REL inhibition enhancing cytotoxic function in vitro. A unique immunosuppressive B-cell subset (B7) engaged in enhanced crosstalk with exhausted T cells and harbored a MYC-centered network linked to cell cycle dysregulation and poor survival. Spatial transcriptomics revealed significant proximity between B7-active regions and Treg/exhaustion-enriched areas, externally validated in melanoma. Proteomic data validated elevated REL expression in MYCN-amplified tumors. CONCLUSION: This work delineates the immunosuppressive architecture of MYCN-driven neuroblastoma, revealing novel regulatory nodes within specific lymphocyte compartments. Integrating single-cell, spatial, and proteomic evidence, we propose REL inhibition as a therapeutic candidate, the EOMES/TRP axis as a bioinformatically supported hypothesis, and the B7/MYC hub as a hypothesis supported by transcriptomic and spatial evidence.

Humans↗

PubMed2026

Integrating genetic predictors into subsequent breast cancer risk prediction in survivors of childhood cancer.

PURPOSE: Female survivors of childhood cancer are at high risk for developing breast cancer. The contributions of most general population primary breast cancer genetic predictors to this risk have not been explored. METHODS: Analyses included females who survived ≥5 years after their childhood cancer diagnosis with available array (N = 2096, subsequent breast cancer [SBC]=218) or whole-genome sequencing (WGS; N = 3292, SBC=101) data from the Childhood Cancer Survivor Study and St. Jude Lifetime Cohort. We computed 99 externally-validated primary breast cancer polygenic risk scores (PRS). Using deep-coverage WGS, ClinVar-annotated pathogenic/likely pathogenic (P/LP) variants in breast cancer susceptibility genes were identified. Cox proportional hazards models assessed associations with SBC risk, adjusting for treatments and genetic ancestry. RESULTS: Among 5388 female survivors (genetic ancestry, European: N = 4,752; African: N = 444; East Asian: N = 192), 319 developed SBC. Most (90.9%) PRSs were nominally associated with SBC risk (P < 0.05), but effect sizes varied substantially. PRSs with superior discriminatory ability had greater genome-wide coverage (e.g., 6.4 million-variant PRS, HR per SD = 1.71, 95% CI = 1.43 to 2.05; P = 4.2x10-9) and 7.7-fold higher odds (P = 7.0x10-4) of including variants in multiple DNA damage repair pathways compared with PRSs with weaker risk associations. Among survivors with WGS, 1.6% carried P/LP variants in clinical testing panel genes, which was associated with a 7.4-fold greater risk (95% CI = 3.16 to 17.19). Including genetic factors improved SBC risk prediction by age 40 (P < 0.001) compared to treatment exposures alone. CONCLUSIONS: Externally-validated primary breast cancer genetic susceptibility predictors are relevant for SBC risk prediction and should be prioritized for risk stratification in survivors.

Journal Article↗