Citations

Abstract Aims Left ventricular hypertrophy (LVH) is a strong predictor of cardiovascular disease. We previously compared supervised machine learning techniques to classify cardiac magnetic resonance (CMR)-derived LVH using electrocardiogram (ECG) and clinical variables in 37 534 UK Biobank participants, obtaining an area under the receiving operating curve (AUROC) of 0.85, but with limited specificity and requiring external validation. In this study, we develop a deep learning (DL) model to improve classification with external evaluation in the Study of Health in Pomerania (SHIP). Methods and results We analysed 12-lead ECGs of 48 835 participants from the UK Biobank imaging study. The dataset was split into a training set (70%), validation set (15%), and test set (15%) for performance evaluation. The model architecture was a fully convolutional network, for which the input was the participants’ median ECG and clinical variables and the predicted indexed left ventricular mass (iLVM) as the output. A subsequent logistic regression model was used to recalibrate iLVM predictions. In UK Biobank, 717 (1.5%) participants had CMR-derived LVH and the AUROC for the DL model was 0.97. The ECG components most predictive of LVH were the QRS complex and ventricular rate. The DL model outperformed our supervised algorithms, previous DL modelling efforts and clinical ECG benchmarks. There was modest generalizability of the DL model to 1423 participants in SHIP (AUROC 0.78), with differences in clinical profile, ECG acquisition, and CMR labelling as important factors. Conclusion Our findings support the feasibility of scalable DL-based screening tools for the prediction of LVH from the ECG, whilst highlighting the need for model development using larger datasets with greater diversity to ensure generalizability.

Abstract Objective To measure the relative levels of signal and noise in expert diagnosis of epilepsy. Methods Twenty multinational epileptologists independently reviewed 50 vignettes of adult and pediatric patients presenting with suspected seizure(s) on two separate occasions with a ≥30‐day washout period. Experts provided a diagnosis of epilepsy or non‐epilepsy based on clinical information and, if requested, routine EEG and neuroimaging data. Cases had an established clinical diagnosis of epilepsy or non‐epilepsy based on capture of habitual paroxysmal events on video‐EEG or long‐term clinical follow‐up. Experts' judgments were analyzed to decompose variability into different sources: signal (objective differences between cases), level noise (experts' bias toward over/under‐diagnosis), pattern noise (experts' idiosyncratic reactions to specific case features), and occasion noise (inconsistency across occasions). Results The probability of an expert making a different diagnosis for a given case on two different occasions was 16%. The probability of two different experts making a different diagnosis for the same case was 26%. Signal (case “difficulty”) accounted for 66–69% of total variation, with 31–34% attributable to noise. Level noise was the largest contributor in the absence of EEG/neuroimaging results (23%), while pattern noise dominated when test results were available (24%). Occasion noise contributed relatively little (1%) but was still sufficient to cause diagnostic reversals in 16–22% between occasions. Significance The degree of noise in expert diagnosis of epilepsy is substantial, stemming primarily from physicians' idiosyncratic interpretations of case features and variable dispositions toward over‐ or under‐diagnosis. Strategies to improve reliability are needed, including standardized data collection protocols and structured decision algorithms. For “difficult cases,” where expert reliability and accuracy are lowest, our findings support current clinical practice which favors early referral for video‐EEG monitoring over reliance on diagnostic anchoring. This diagnostic pathway may become more accessible with advances in EEG technology (e.g., wearable devices) and artificial intelligence.

Background: Fibrodysplasia ossificans progressiva (FOP) is an ultra-rare, disabling, genetic disorder characterized by progressive heterotopic ossification (HO). FOP care pathways are complex, and there are many underrepresented patients globally. Objectives: This scoping review aimed to identify global FOP case reports and investigate real-world management practices. Evidence sources, eligibility criteria, and charting methods: International/local databases and congress proceedings were searched. Publications reporting patient-level data for ⩾1 patient with an FOP primary diagnosis, published January 2000–January 2024, were eligible. Publication details, patient demographics, disease presentation, and care pathway were extracted. Data were charted through categorization of themes and frequency analyses. Results: Of 6064 publications screened, 369 were eligible, reporting on 541 patients with FOP. Most publications (82.4% (304/369)) reported an individual case; the remaining reported 2–25 cases. Cases were from 57 countries, most commonly in the Asia/Pacific region (47.0% (254/541)). Most patients had reportedly seen a single specialist (64.8% (59/91)), and care from multidisciplinary teams was reported for only 5.5% (5/91). HO (84.7% (458/541)) and great toe malformations (66.7% (361/541)) were the most commonly reported signs, though mobility limitations (46.2% (96/208)) and growths/lesions (38.5% (80/208)) were most common before diagnosis. The predominant drug classes reported for FOP treatment were corticosteroids (52.6% (102/194)) and nonsteroidal anti-inflammatory drugs (39.7% (77/194)). Surgeries were reported in 35.1% (190/541) of patients. Where specified, the most common surgery types were malformation correction/bone resection (43.6% (78/179)) and biopsies (34.1% (61/179)). Conclusion: This scoping review synthesized valuable real-world insights from case reports of FOP, potentially supporting outreach initiatives in underserved regions and facilitating shared learnings on diagnosis and treatment. Although the inherent selection bias of case studies toward atypical cases limits definitive conclusions, some global trends in the clinical management practices were revealed.

Abstract Diffuse Large B Cell Lymphoma (DLBCL), the most common subtype of aggressive B-cell Non-Hodgkin Lymphoma (NHL), accounts for 30-40% of newly diagnosed NHL cases. Despite recent progress improving our understanding of molecular features of DLBCL, there is still a high relapse rate (~45%) among patients who undergo standard treatment (R-CHOP). Several groups have identified reproducible molecular clusters of DLBCL although, the biologic significance of many of the specific gene mutations within these clusters is poorly understood. DTX1 mutations are a feature of the BN2/C1 DLBCL classifiers and are among the mutations with unknown significance. BN2/C1 DLBCL are a mixture of ABC, GCB, and Unclassified cell of origin (COO) with a common signature being the presence of genetic alterations in genes known to regulate the NOTCH pathway; yet, these tumors lack transcriptional signatures of traditional NOTCH activation, suggesting a unique and undefined biology driven through non-canonical NOTCH pathway activation. DTX1 is an E3 ubiquitin ligase which directly binds the intracellular domain of the NOTCH receptor (NICD), targeting it for degradation, resulting in suppression of NOTCH pathway activity. We therefore hypothesize that DTX1 mutations lead to NOTCH pathway dysregulation and are a critical driver for a subset of BN2/C1 DLBCLs. Using whole exome sequencing (WES) data from 404 newly diagnosed DLBCL patients, we found 8% of patients with mutations in DTX1. Alterations in DTX1 were localized to the NICD binding domain, WWE1. In silico Alphafold modeling of DTX1 WWE1-interaction with NICD revealed that patient mutations clustered within predicted interacting surfaces with NICD. These findings lead to the hypothesis that mutations in the DTX1 WWE1 domain suppress its ability to bind to and degrade NICD, enhancing NOTCH pathway activation in DLBCL. To address this question, we first compared DTX1 protein expression by immunoblot in DTX1 WWE1 mutant (OCI-Ly3 and SUDHL2) and wildtype (WT) (OCI-Ly7, OCI-Ly19, SUDHL6, and U2932) DLBCL cell lines. We found higher expression of NICD1 (p=0.03) in OCI-Ly3 and NICD2 (p<0.0001) in SUDHL2 cell lines compared to WT. This increase was accompanied by an increase in NOTCH surface expression with higher NOTCH1 (p<0.0001) and NOTCH2 (p<0.0001) expression on OCI-Ly3 and SUDHL2 cells, respectively, compared to WT. To remove variability across cell lines we next used CRISPR to generate DTX1 knock out (KO) models using WT DTX1 DLBCL cell lines U2932 and OCI-Ly19. Analysis of NICD and NOTCH receptor protein expression confirmed our initial finding, showing an 8.5 and 4.8-fold increase of NICD1 and a 6 and 2.5-fold increase of NICD2 expression in the U2932 and OCI-Ly19 DTX1 KO cell lines, respectively, compared to empty vector (EV) controls. An increase in surface NOTCH1, and to a lesser extent NOTCH2, was seen in U2932 (p=0.01 and p=0.07, respectively) and OCI-Ly19 (p=0.001 and p=0.2, respectively) KO cells compared to EV controls. To identify pathways affected by DTX1 mutations, we next performed RNAseq on U2932 EV and KO cell lines. Of interest, we did not detect any significant changes in classical NOTCH driven genes (Hes1, Hey1, etc.), supporting prior studies on BN2/C1 tumors. However, we did identify potential non-canonical NOTCH-target genes that may contribute to lymphomagenesis. Gene expression in KO cells showed upregulation of genes involved in germinal center (GC) processes (AICDA, IRF4; p<0.001) and cellular growth and survival (IL6R, CCND2, STAT5B, BCL2; p<0.0001). Measurement of IL6R by flow cytometry and CyclinD2 by immunoblot further verified DTX1-mediated upregulation of IL6R (p<0.01) and CyclinD2 (p<0.01) for U2932 KO cells compared to EV control. We also observed downregulation of genes contributing to cell death (ex. BAX, CASP8, FAS; p<0.05) in KO cells compared to EV. Together, these novel findings support our hypothesis that DTX1 WWE1 mutations lead to dysregulation of the NOTCH pathway by driving overexpression of NICD and NOTCH, resulting in dysregulated expression of proteins involved in cell cycle, GC reaction, and apoptosis. We further provide new biologic insight into BN2/C1 DLBCL, ultimately highlighting a path toward novel therapeutic targets and the advancement of personalized medicine, as patients harboring these mutations may benefit from new therapies targeting inflammatory pathways, such as CELMoDs.

Background Individuals with neurologic impairment, such as brain injury, are more likely to have autonomic dysfunction, including POTS, but often experience a lack of treatment options and access to quality care. Objective Examine the feasibility and outcomes of a group-based multimodal rehabilitation training program. Methods Nineteen females, ages 18–53 (M = 28.6) completed the outpatient training program. The intervention included eight 50-min weekly sessions using an interdisciplinary, multimodal approach, including education, movement, and mindful self-compassion. Participants completed demographic and medical history and assessment for baseline orthostatic intolerance. At pre-training, post-training, and follow-up, participants completed the Modified Fatigue Impact Scale (MFIS), Lower Extremity Functional Scale (LEFS), Rand 36-Item Health Survey 1.0 (RAND-36), Trail Making Test (A and B) (TMT-A, TMT-B), 10-Meter Walk Test (10MWT), and grip strength testing. Results At baseline, participants were significantly below norms in daily function (p < 0.001), grip strength (p ≤ 0.001) gait speed (p < 0.001), fatigue (p < 0.001), and quality of life (p < 0.001). Between pre-training and follow-up, participants showed significant improvement across function in daily life tasks (p < 0.01); grip strength in the dominant (p < 0.01) and non-dominant (p < 0.01) hands; gait speed (p < 0.05); levels of fatigue (p < 0.05); cognitive performance (p < 0.05); and quality of life (p < 0.05). Conclusion Individuals with POTS face functional challenges that can be significantly and objectively improved through multimodal rehabilitation. There is a need for provider education and further research to optimize care and quality of life for individuals with POTS.

Abstract Disclosure: S.R. Chacko: None. C.A. Villavicencio Torres: None. K. Yu: None. S.G. Waguespack: None. L. Lu: None. Z. Belaya: None. T. Bandgar: None. A. Sada: None. O. Ragnarsson: None. B. Altieri: None. A. Newman: None. P. Vibhatavata: None. M. Tóth: None. E.V. Varlamov: None. C. Badiu: None. A. Paissan: None. M. St-Jean: None. H. Falhammar: None. N. Imamudeen: None. L. Haberbosch: None. M. Bobrowicz: None. A. Tabarin: None. R. Shah: None. M.R. Gadelha: None. A. Peersen: None. W.F. Young: None. J. Kaplan: None. J. Varghese: None. M.A. Habra: None. O. Golounina: None. G.A. Melnichenko: None. A.R. Lila: None. C. Yamichannaiah: None. I. Bancos: None. Context: Cushing syndrome (CS) due to ectopic ACTH secretion (EAS) is a rare condition, associated with severe hypercortisolism and consequent morbidity and mortality. Its varied and often emergent presentation make diagnosis and treatment challenging. Objectives: 1.Characterize clinical and biochemical presentation of EAS. 2.Describe the scope of hypercortisolism-related complications, overall and CS-specific mortality. 3.Identify factors associated with CS-specific mortality. Methods: We conducted a multicenter retrospective cohort study in patients with EAS at 37 centers. Outcomes were clinical/biochemical severity scores, complications, overall and CS-specific mortality. Results: In 1053 patients [median age at diagnosis 52 years (IQR 37-64), 57% women], metastatic disease was present at diagnosis in 585 (56%). The most common sources of EAS were bronchial neuroendocrine neoplasm (NEN) (304, 29%), pancreatic NEN (114, 11%) and small cell lung carcinoma (114, 11%), with the source of EAS being occult in 192 (18%). Clinical and biochemical severity scores were severe in 475 (45%) and 873 patients (92%) respectively. Hypokalemia was present in 819 patients (78%) in the year prior to diagnosis. Of 342 patients (32%) who had IPSS, 301 (88%) suggested an ectopic source, 7 (2%) a pituitary source, and 23 (7%) were non-diagnostic. Of 595 patients who had both functional and cross-sectional imaging, discordant findings were seen in 122 (21%), concordant lesion was seen on both in 326 (55%), and no lesion in 147 (25%). Of those with discordant results, cross-sectional imaging was accurate in 66 (54%) and functional imaging in 56 (46%).Management included medical therapy alone in 237 (23%), surgical excision of the source of EAS in 158 (15%), bilateral adrenalectomy in 77 (7%) and multimodal therapies in 498 patients (47%). Complications within 5 years after diagnosis included hospitalization related to CS or its therapy in 351 (33%), infection in 337 (32%), thromboembolic event in 102 (10%), cardiovascular event in 83 (8%) and cerebrovascular event in 27 (3%). During a median follow-up of 35 months (IQR 20-122) from CS diagnosis, death occurred in 462 patients (44%), with the cause of death related to CS in 149 (14%). CS-specific mortality was higher in metastatic compared with non-metastatic disease (19% vs. 7%, P&lt;0.001). On multivariable analysis of age, sex, clinical and biochemical severity, older age (HR 1.03, 95% CI: 1.02-1.04) and lower clinical severity score (HR 0.97, 95% CI: 0.94-0.99), were associated with higher CS-specific mortality. Conclusion: Our interim analysis, the largest worldwide, found that most patients with EAS had severe hypercortisolism and hypokalemia. Metastatic disease was present in 56%, and the source of EAS was occult in 18%. CS-related death occurred in 14%, emphasizing the need for urgent management. Presentation: Monday, July 14, 2025

Abstract Aims Retatrutide, a novel synthetic molecule, is a triple agonist activating the glucose‐dependent insulinotropic polypeptide, glucagon‐like peptide‐1 and glucagon receptors. The TRIUMPH clinical development program evaluates its safety and efficacy concurrently for the treatment of obesity and two related complications—obstructive sleep apnea (OSA) and knee osteoarthritis (OA). A novel basket trial design simultaneously evaluates retatrutide treatment across these multiple adiposity‐related disease states. Materials and methods TRIUMPH consists of four Phase 3, multicenter, randomized, double‐blind studies assessing weekly subcutaneous retatrutide compared to placebo, in conjunction with healthy diet and physical activity in over 5800 participants. The four trials consist of two weight management basket trials (TRIUMPH‐1 and TRIUMPH‐2) with OSA and/or OA protocols nested within the weight management trial; one weight management trial in a population with CVD (TRIUMPH‐3); and one stand‐alone OA trial (TRIUMPH‐4). The primary endpoint for weight management is percent change in body weight, for OSA is change in Apnea‐Hypopnea Index and for knee OA includes change in the Western Ontario and McMaster Universities Osteoarthritis Index pain subscale score. The basket trial permits independent analysis of weight management, OSA and OA studies with type I error rate controlled at α = 0.05, split between the overarching weight management and each basket trial. Conclusions By recruiting participants with shared disease exposures, the TRIUMPH program will assess the safety and efficacy of retatrutide for the treatment of adults with obesity and two of its common complications—OSA and OA.