Citations

380 Background: KRAS amplifications (amp) occur in ~2% of solid tumors, representing a high unmet need for effective targeted therapies. BI3706674 binds to GDP-bound KRAS and locks KRAS protein in an inactive form. In preclinical models, BI3706674 has shown potent inhibition of KRAS-amp tumors. Methods: This is a Phase Ia/b trial (NCT06056024) to explore the safety, pharmacokinetics (PK), pharmacodynamics (PD) and efficacy of BI 3706674 in patients (pts) with unresectable or metastatic KRAS wild type amp or KRAS G12V mutated (mut) solid tumors. The trial aimed to define the maximum tolerated dose (MTD). Pts (ECOG 0 -1) received BI 3706674 daily per os at 50 - 1200 mg doses. Dose escalation (DE) was guided by a Bayesian logistic regression model. This report focuses on KRAS wt amp pts from Phase Ia (DE). Results: As of 31 July 2025, 46 pts had received BI 3706674. Pts (n=15/46) with KRAS wt amp tumors had upper gastrointestinal tract (uGI) cancer [n=9: gastric (GAC;n=5), oesophageal (EAC; n=2), gastroesophageal junction (GEJC; n=2) cancers], Cholangiocarcinoma (n=1), colorectal (CRC; n=4), and ovarian (OC; n=1) cancers. Pts were treated at 200 mg (n=2), 400 mg (n=1), 800 mg (n=5), 1200 mg (n=7) doses; males: n=11; ECOG 1: n=10; median age - 62 years (range 38–83); median number of prior treatment lines was 3 (range 1-6). AEs were reported in 15 pts (100%), including 12 pts (80%) with treatment-related AEs (TRAE). High-grade AEs (Grade 3-5) were reported in 10 pts (66.7%), including 4 pts (26.7%) with TRAEs. SAEs were reported in 7 pts (46.7%), including 2 pts (13.3%) with TRAEs. The most-frequent AEs were mostly low-grade (Grade 1-2) diarrhoea (10 pts, 66.7%), nausea (10 pts, 66.7%), and vomiting (9 pts, 60.0%), often reported as TRAE. Two dose limiting toxicities were observed in 1200 mg group: nausea (Grade 3) and vomiting (Grade 3). MTD was not reached. BI 3706674 PK parameters followed the predicted non-linear PK model; dose dependent systemic exposure reached plateau at 800 mg. 12/15 pts with KRAS wt amp tumors, including 10/12 pts in 800 – 1200 mg groups, were efficacy evaluable; Disease control rate (DCR) was 58.3% (7/12 pts): 3/ 12 pts with partial response (PR), 4/12 had stable disease (SD), 5/12 pts - disease progression (PD). In 800 – 1200 mg groups 1 - OC, 2 – CRC pts achieved SD. PK/PD/efficacy correlation data is pending. Conclusions: BI 3706674 showed a tolerable safety profile and clinical efficacy in pts with KRAS amp tumors. Study results provide clinical proof of concept. KRAS wt amp is a druggable target and KRAS-targeting treatment is active in KRAS wt amp uGI cancer pts. Clinical trial information: NCT06056024 . Anticancer activity of BI 3706674 in efficacy evaluable pts with KRAS wt amp tumors. All pts, 200 – 1200 mg, N=12 Pts with uGI cancers,800 – 1200mg, N=7 ORR, n (%) 3* (25%) 3* (43%) SD, n (%) 4 (33.3%) 1 (14%) DCR, n (%) 7 (58.3%) 4 (57%) PD, n (%) 5 (41.7%) 3 (43%) *Tumor shrinkage: -60.3%, -34% and 34.3%.

Abstract The advent of high-throughput sequencing has led to an exponential increase in genomic data, highlighting the need for efficient and accurate models to analyze and interpret this information. In this study, we introduce a novel, exomic foundational model that leverages a combination of the human reference genome and multispecies data to improve variant detection and interpretation. Our model utilizes a short-range transformer architecture and is trained on a large dataset of human exomic sequences derived from the Tapestry study. Through a series of ablation studies and scaling experiments, we demonstrate the effectiveness of our model in predicting next token accuracy and identifying clinically pathogenic variants. We also show that our model outperforms existing models in a range of downstream tasks, including variant effect prediction and disease state identification. In fact, our largest sequence transformer for range adaptive nucleotide decoding variant (1B parameters) surpassed previous benchmarks, demonstrating a mean accuracy of 0.880 [an 8.2% improvement over the original nucleotide transformer (NT) and a 7% improvement over NT-v2]. Furthermore, we construct a unique exomic ClinVar dataset to evaluate the model’s performance on pathogenicity and disease states. Our results highlight the potential of this model to improve our understanding of the human exome and its role in disease. The model and its applications have significant implications for genomics-based diagnosis and personalized medicine, including tailored therapeutic development.

Abstract Background Hospital discharge summaries are critical for ensuring continuity of care, but their quality often varies. Large language models (LLMs) have the potential to standardize and enhance the efficiency of this documentation process. Objectives To evaluate the quality of hospital discharge summaries created by an LLM‐based hospital course drafting tool created by Epic Systems compared with human‐written summaries. Methods Retrospective study at a single tertiary‐care institution in 2024. The cohort included 100 adult hospitalizations lasting >72 h across medical and surgical dismissing services. No interventions were performed. Summaries (LLM‐generated vs. human‐written) were independently reviewed using a standardized rubric covering nine domains (e.g., comprehensiveness, clarity, relevance). Scores were normalized and compared. Readability was assessed using Flesch Reading Ease. Results LLM‐generated summaries outperformed human‐written summaries across all criteria ( p < .05), with the greatest difference observed in comprehensiveness (LLM median 0.62 vs. human −0.23). Human‐written summaries from surgical services scored lower than those from medical services, but LLM performance was consistent across both. Human summaries had higher Flesch Reading Ease scores (33.11 vs. 26.2; p < .05), reflecting simpler language. Conclusions LLM‐generated summaries demonstrated superior quality, consistency, and clinical utility compared with human‐written summaries, highlighting their potential to improve documentation efficiency and standardization.

Abstract Background: Spatial Reasoning AI (SRAI) systems are emerging as a transformative class of clinical artificial intelligence, enabling real-time interpretation of complex anatomy in imaging, surgery, and chronic disease management. Unlike traditional AI approaches constrained to tabular or 2D inputs, SRAI models incorporate geometric priors, multimodal fusion, and 3D anatomical reasoning to support high-impact tasks such as volumetric segmentation, lesion localization, intraoperative guidance, and prognostic modeling. Objective: This scoping review aimed to (1) map the emerging landscape of SRAI in healthcare, (2) assess its clinical performance across diagnostic, prognostic, and workflow applications, and (3) identify opportunities and barriers for safe translation into practice. Methods: Following PRISMA 2020 guidelines, we systematically searched PubMed, Springer Nature, and IEEE Xplore (including CVPR, ICCV, ISBI, and EMBC) for original studies implementing spatially aware AI models with clinical validation or workflow relevance. Data extraction captured model architecture, input modality, anatomical scope, application domain, and performance outcomes. Results: We identified eligible studies spanning oncology, neurology, cardiology, surgery, and infectious diseases. SRAI systems consistently outperformed conventional methods in segmentation, lesion classification, and risk stratification. A paradigm shift is underway from task-specific CNNs to foundation-level models, including MedSAM and MedCLIP for zero-shot segmentation and retrieval, and NeRF-based pipelines for anatomically faithful 3D reconstruction. Randomized evidence shows that SRAI can improve workflow efficiency without compromising diagnostic accuracy. Conclusion: SRAI models demonstrate strong potential to advance diagnostic precision, prognostic accuracy, and workflow integration across multiple clinical domains. Their safe adoption will require standardized evaluation metrics, prospective multi-institutional validation, and interpretable outputs. As foundation models evolve to better encode spatial priors, SRAI is poised to become a cornerstone of workflow-integrated precision medicine.

Background: Pneumocystis jirovecii pneumonia (PJP) is life-threatening for immunocompromised patients. No consensus exists on PJP prophylaxis for immunosuppressed patients without HIV, transplant, or cancer. Methods: We retrospectively reviewed the electronic health records of adult immunosuppressed patients with PJP diagnosed between 1990 and 2020 at Mayo Clinic. Patients with HIV, solid organ transplants, or cancer were excluded. Demographic data, treatments, and outcomes were manually abstracted. Results: The most common indications for immunosuppression were rheumatoid arthritis (19.7%), vasculitis (18.1%), and interstitial lung disease (ILD) not related to connective tissue disease (17.6%). Despite having high risk of PJP, 86.0% of patients did not receive PJP prophylaxis. Corticosteroids were the most common immunosuppressive agent used (84.5%), with 64.4% of patients receiving high-dose treatment. Nonbiologic disease-modifying antirheumatic drugs were used for 49.7%, including methotrexate (51.0%), azathioprine (22.9%), and hydroxychloroquine (11.5%). Biologics were prescribed for 25.4%, primarily rituximab (59.2%) and infliximab (22.4%). Hospitalization occurred for 76.7% of patients; 70.3% required intensive care unit (ICU) admission, and 46.6% received mechanical ventilation. The in-hospital mortality rate was 30.4% overall and 53.6% for patients on ventilation. Predictors of death included ILD [odds ratio (OR), 4.61; 95% CI, 1.75-13.00], ICU admission (OR, 3.60; 95% CI, 1.19-11.08), and ventilator use (OR, 3.46; 95% CI, 1.30-9.79). Biologic use was associated with lower odds of death (OR, 0.34; 95% CI, 0.11-0.89). Conclusions: Most patients in our cohort did not receive PJP prophylaxis, and outcomes were poor with high mortality rates. Standardized risk stratification and prophylaxis protocols are needed to improve outcomes.

Abstract Background: Spatial Reasoning AI (SRAI) systems are reshaping clinical practice by enabling real-time interpretation of complex anatomical structures in medical imaging, surgery, and chronic disease management (Lin et al., 2023, Qiao et al., 2021; Song et al., 2024) . Unlike traditional AI approaches that rely on tabular or 2D data, SRAI models incorporate geometric priors, multimodal fusion, and 3D anatomical reasoning to support high-impact tasks such as volumetric segmentation, disease localization, and intraoperative guidance (Milletari et al., 2016; Hatamizadeh et al., 2022; Zhang et al., 2022) . Objective: This scoping review aimed to (1) characterize the emerging landscape of spatial reasoning models in healthcare, (2) evaluate their clinical performance across diagnostic and prognostic applications, and (3) identify architectural trends that underpin successful translation into practice. Methods: Following PRISMA guidelines, we systematically screened PubMed to identify studies that implemented spatially aware AI models with real-time or clinically validated outcomes (Page et al., 2021) . Data extraction captured model architecture, modality, anatomical scope, clinical application, and performance metrics. Results: We identified 124 eligible studies across the clinical domains of oncology, neurology, cardiology, orthopedic surgery, and infectious diseases (Nie et al., 2019; Hao et al., 2020; Tasken et al., 2023) . SRAI systems consistently outperformed conventional methods in segmentation, lesion classification, and risk stratification. (Lin et al., 2023; Qiao et al., 2021; Mahootiha et al., 2024) . A paradigm shift from task-specific CNNs to foundation models that reason across spatial structures has emerged, beginning with adaptations of generalist models such as MedSAM and MedCLIP for zero-shot segmentation and cross-modal retrieval, and NeRF-based reconstruction for anatomically faithful 3D modeling (Ma et al., 2024; Mildenhall et al., 2020; Radford et al., 2021). Beyond technical performance, randomized evidence shows that SRAI systems can improve workflow efficiency without compromising outcomes (Lang et al., 2023) . Conclusion: SRAI models hold promise for enhancing diagnostic accuracy, prognostic precision, and clinical efficiency across diverse applications. Their successful integration will depend on standardized evaluation metrics, robust validation, and interpretable outputs. As foundation models evolve to better encode spatial priors, SRAI is poised to become a cornerstone of personalized and procedural medicine.

Abstract Erdheim-Chester disease (ECD) is a rare systemic histiocytic neoplasm, with cardiac morbidities, including cardiovascular (CV) risk factors and cardiac conditions, playing a significant yet poorly understood role in the disease burden. This study evaluated the prevalence, characteristics, and prognosis of ECD in patients with (ECD-C) and without cardiac involvement (ECD-noC) and compared the burden of cardiac morbidities with matched controls. Patients diagnosed with ECD between 1990 and 2021 at a tertiary center were included, with cardiac involvement centrally assessed using radiographic studies. Cardiac morbidities were compared with a control group without ECD, matched for age, sex, body mass index, and smoking history. Among 104 patients with ECD, 39 (37%) had cardiac involvement. Patients with ECD-C had higher rates of hypertension (67% vs 46%), hyperlipidemia (67% vs 40%), heart failure (36% vs 8%), and pericardial effusion (28% vs 2%) than those with ECD-noC. Compared with the matched non-ECD cohort, patients with ECD had higher prevalence of coronary artery disease (20% vs 7%), heart failure (18% vs 4%), and antihypertensive drug use (55% vs 40%). Notably, patients with ECD-C had inferior progression-free survival (PFS) from frontline therapy compared with patients with ECD-noC (5-year PFS, 28.3% vs 70.5%). These findings highlight the burden of CV risk factors and cardiac conditions in ECD, even without a clinical diagnosis of ECD-C. Importantly, this cardiac morbidity burden is substantial for patients with ECD-C compared with ECD-noC. Our findings highlight the need for comprehensive cardiac risk assessment and management strategies to improve patient outcomes.

Background Birt–Hogg–Dubé syndrome (BHD) is a rare autosomal dominant disease characterised by the development of lung cysts, skin fibrofolliculomas and an increased risk for renal cancer. The progression of cystic lung disease with ageing in BHD has not been well characterised. Objectives To assess the severity of lung involvement in patients with BHD aged 50 years or older. Methods A retrospective computer-assisted search was performed to identify patients with BHD seen at Mayo Clinic in Rochester, Minnesota and the First Affiliated Hospital of University of Science and Technology of China from January 2000 to May 2023. Demographic, clinical, pulmonary function and radiological parameters were extracted and analysed. Results Among 82 patients with BHD aged 50 years or older, 49 (59%) were female and the median age was 57 years (range, 50–79) at baseline. Pneumothorax occurred in 46 patients (56%). Lung cysts were observed on CT in 77 patients (94%) and were bilateral in nearly all cases. The number of cysts was≥10 for most patients (84%) and the median size of the largest cyst was 31 mm (range, 4–160 mm). Median (range) % predicted values of forced vital capacity, forced expiratory volume in 1 s and diffusion capacity of the lung for carbon monoxide were 98 (62–147), 92 (56–145) and 85 (61–133), respectively, for those with data available. During follow-up (median, 41 months; range, 3–259), 35% of patients manifested worsening in pulmonary function or cystic lung disease. Conclusion Lung involvement in older patients with BHD manifests varying severity of cystic lung disease with relatively mild functional impairment, if any, and indolent progression.

Abstract Background Despite advances in diagnosis and management, the interstitial lung disease (ILD) lexicon is plagued by ambiguous and inconsistent terminology that complicates communication and impedes knowledge generation. The objective of this Fleischner Society Consensus Statement was to produce standardized terminology for ILD multidisciplinary diagnoses and major phenotypes. Methods Interviews with 10 experts were used to identify ILD clinical diagnoses and major phenotypes. The preferred terms for each entity and potential alternatives were identified, alongside the rationale for the preferred term. Entities with more than one potential term were the subject of an online modified Delphi survey posed to the 29 committee members, aiming to achieve consensus. Committee members rated their agreement with the initially preferred term—5 (strongly agree), 4 (agree), 3 (neutral/unsure), 2 (disagree), and 1 (strongly disagree)—with the option to provide additional comments. Median score ≥4 and interquartile range ≤1 were considered consensus agreement. Terms not reaching agreement were discussed by video conference, followed by an additional survey that incorporated feedback. Results From the 60 initial terms, there were two root terms that required upfront consensus before survey initiation (ILD and interstitial pneumonia) and another eight terms that had no alternative suggested by the committee or in the literature. Agreement was met by 47/50 terms (94%) in Round 1 of the survey. The three terms (6%) that did not reach agreement met agreement in Round 2. Conclusions This document provides standardized recommended terms for ILD multidisciplinary diagnoses and major phenotypes that will facilitate communication among clinicians, researchers, patients, and other stakeholders.