Citations

Background: Hepatorenal syndrome–acute kidney injury (HRS-AKI) is associated with high mortality in cirrhosis. Recent FDA approval of terlipressin and updated clinical guidance have expanded management options, but real-world practice patterns remain unknown. Methods: We conducted a nationwide survey of U.S. hepatology providers regarding HRS-AKI management practices. The 33-question survey assessed provider demographics, vasoconstrictor selection, treatment timing, and adherence to guideline recommendations. Results: Among 162 respondents, most were hepatologists (94%) at academic centers (83%). Only 11% obtained a nephrology consultation at AKI detection, with 44% waiting until worsening renal function. Terlipressin was available at 80% of institutions but more commonly at academic centers (84% vs. 53%, p =0.003). Consistent with guideline recommendations, 77% of providers initiate vasoconstrictors only after completing a trial of volume expansion (if no renal improvement), particularly in academic centers (80% vs. 60% non-academic, p =0.03). Terlipressin (49%) and midodrine/octreotide (44%) were preferred first-line treatments, with providers at academic centers more likely to use midodrine/octreotide (46% vs. 30%, p =0.008). Mean arterial pressure was used by 62% of providers to guide vasoconstrictor dosing. Most providers (73%) discontinued vasoconstrictor treatment after 4 days if no improvement. Conclusions: Significant variations exist between guideline recommendations and real-world HRS-AKI management, especially first-line treatment choice, treatment monitoring, and nephrology consultation. These findings highlight opportunities to improve guideline implementation and identify areas where practice patterns might inform provider education and future guidance updates.

Abstract Objectives Patients referred for specialized care often arrive with outside medical records (OMRs) compiled into multi-report PDFs that include imaging, pathology, and clinical notes in unstructured formats. Reviewing these records is time consuming and mentally taxing, increasing the risk of delayed care, clinician frustration, and missed information affecting quality of care. This study aimed to automate the segmentation, classification, and date extraction of scanned OMRs, with a focus on records relevant to breast cancer care. Materials and Methods We used optical character recognition (OCR) to extract machine-readable text from 1303 scanned PDF documents from 116 distinct external institutions. Gemini 1.5, a large language model (LLM), was then used to segment multi-report files into individual documents, classify them into clinically meaningful categories such as mammograms and pathology reports, and extract study dates to build diagnostic timelines. Document categories were informed by clinical workflows in a breast cancer center. Results The system achieved an F1 score of 0.95 for segmentation, 0.96 for classification, and 0.90 for date extraction. In a pilot of 45 records reviewed by clinicians, only 2 classification errors and 1 date error were reported. Clinicians estimated that the tool reduced OMR review time by 40%, improved workflow efficiency, and increased satisfaction. Discussion Our findings demonstrate that combining OCR with LLMs can significantly enhance the processing of unstructured medical records, reducing manual burden and supporting timely clinical decision-making. Conclusion This study demonstrates the successful application of OCR and LLMs for organizing scanned OMRs within a specialty clinic. By automating a previously manual process, the approach supports scalable review of incoming outside records and has potential for adaptation to other clinical workflows. Future work will focus on evaluating the system across additional specialties and institutions.

Abstract Purpose: Knowledge of prognostic factors and long-term survival in patients with diffuse hemispheric glioma, H3 G34–mutant (DHG, H3 G34), remains limited in this tumor with a poor prognosis. Experimental Design: This retrospective, multi-institutional study investigated prognostic variables for patients with DHG, H3 G34, and their association with progression-free survival (PFS) and overall survival (OS). Uni- and multivariable Cox proportional hazard models were applied with multiple imputed datasets. Results: A total of 153 patients (142 G34R, 9 G34V, 2 via DNA methylation) were included. The median age at diagnosis was 17 years (range, 2–45). Initial gross/near total resection (GTR/NTR) was achieved in 43% of patients. Radiation was given in 91% (85% focal irradiation), and initial chemotherapy was given in 87% [70% temozolomide-based (TMZ), 25% TMZ/lomustine, 5% non-TMZ]. Median OS was 24 months [interquartile range (IQR), 22–28] with a median PFS of 14 months (IQR, 12–19). Twelve patients (8%) were found to be long-term survivors (≥5 years). Exploratory multivariable analysis showed that adjuvant radiotherapy [HR, 0.076; 95% confidence interval (CI), 0.033–0.17] and achieving GTR/NTR compared with < NTR (HR, 0.51; 95% CI, 0.33–0.78) were associated with improved PFS. Multivariable analysis showed improved OS with increasing age at diagnosis (HR, 0.70; 95% CI, 0.57–0.87), initial radiotherapy (HR, 0.38; 95% CI, 0.15–0.96), and initial GTR/NTR compared with < NTR (HR, 0.60; 95% CI, 0.37–0.97). Conclusions: This cohort highlights prognostic factors for patients with DHG, H3 G34, and describes relapse patterns and therapy approaches. Clinical trials and prospective registries are needed to improve outcomes.

Abstract Aims Calcium release channel deficiency syndrome (CRCDS) results from loss-of-function (LOF) variants in the RYR2-encoded type 2 ryanodine receptor (RyR2), predisposing patients to sudden cardiac arrest/death (SCA/SCD) without abnormalities on a stress electrocardiogram (ECG). Undetected CRCDS may underlie idiopathic ventricular fibrillation (IVF) and sudden unexplained death in the young (SUDY). We aimed to determine the prevalence of potential CRCDS-causative RYR2 variants in IVF and SUDY. Methods and results We reviewed clinical evaluation and RYR2 genetic analysis of 169 IVF patients and 279 SUDY victims. Only ultra-rare (<0.005% in gnomAD) nonsynonymous RYR2 variants were considered potentially pathogenic. Among IVF patients, 6/169 (3%) overall—and 6/67 (9%) with exertion-related SCA—harboured an RYR2 variant and represent potential CRCDS cases. All exhibited normal resting and stress ECGs. Genetic analysis revealed six distinct RYR2 variants, two previously characterized as LOF. In SUDY, 31/279 victims (11%) had a RYR2 variant (30 unique variants), predominantly observed in exertion-related SCD 20/83 (24%) vs. rest-related 11/196 (6%). Of the 14 SUDY victims with functionally characterized RYR2 variants, five (2% of total cohort) had a LOF variant; among the 56 exertion-related SUDY cases, four (7%) had a LOF variant. Conclusion CRCDS may account for 3% of IVF overall and 9% of exertion-related SCA in IVF. Ultra-rare RYR2 variants may underlie up to 11% of SUDY, with 65% of RYR2-positive cases occurring during exertion. LOF-RYR2 variants may contribute to ≥7% of exercise-associated SUDY. Accurate identification of the underlying ryanodinopathy is essential for clinical management of affected patients.

Background: A summary of pharmacist-based practices and preferences regarding the prevention and management of high-dose methotrexate (HDMTX) toxicity in patients with lymphoma does not exist. The objective of the study was to describe current practices surrounding management of HDMTX therapy. This was done through a self-administered, web-based, cross-sectional survey of hematology/oncology pharmacists to ascertain practices and individual attitudes about HDMTX therapy management. Responses were summarized with descriptive statistics. Response comparisons were made using Chi-square or Fisher’s exact test for categorical data, and Kruskal-Wallis test for variables on the Likert scale. A total of 175 pharmacists provided 116 eligible surveys [68 (59%) complete responses and 48 (41%) partial responses] for analysis. The Cockcroft-Gault estimated creatinine clearance formula was the most preferred method (61%) to estimate kidney function and determine HDMTX dosing. Respondents would proceed with HDMTX therapy until the predicted risk of AKI, or AKI stage 2 or 3, exceeded 50% and 20%, respectively. Preferred treatment modifications for a high predicted AKI risk after HDMTX exposure included additional prehydration, HDMTX dose reduction, and increased kidney function monitoring, although levels and degrees of agreement varied. Respondents indicated that serum cystatin C-based GFR estimates lacked evidence, accuracy, and practicality in HDMTX therapy management. There was limited familiarity with cell cycle arrest biomarker use during HDMTX therapy. Glucarpidase was considered time-sensitive and highly effective; however, acquisition cost represented a barrier to use, and the optimal dose was felt to be unknown. In conclusion, there is significant heterogeneity in supportive practices and beliefs about the optimal management for patients with lymphoma receiving HDMTX therapy.

BACKGROUND: The implementation of data science concepts, skills, and tools in critical care research and practice faces multiple, complex barriers. METHODS: We developed an implementation science-based framework and method for measuring the adoption, implementation, and sustainment of data science concepts, skills, and tools in critical care—the Society of Critical Care Medicine (SCCM) Discovery Data Science Campaign (DSC) Implementation Research Logic Model (IRLM). Our IRLM specifies constructs for: 1) key determinants (i.e., barriers and facilitators) influencing the implementation of data science concepts, skills, and tools in critical care; 2) implementation strategies deployed by the SCCM Discovery DSC to address these determinants; 3) theorized mechanisms of action by which these strategies affect outcomes; and 4) upstream and downstream implementation outcomes influenced by implementation strategies. RESULTS AND CONCLUSIONS: We believe that our model can facilitate more rigorous measurement of theoretically grounded, empirically assessable factors driving implementation of data science concepts, skills, and tools in critical care.

Abstract Background Integrating advanced artificial intelligence (AI) into clinical decision-support often requires the sharing of sensitive patient data with external services, raising privacy concerns. Homomorphic encryption (HE) allows computing directly on encrypted data, without revealing the underlying patient information. Objectives To develop a large language model (LLM)-assisted diagnosis framework while preserving patient privacy in the clinical text analysis, by leveraging HE and using rare disease (RD) diagnosis as a representative application. To demonstrate HE does not hinder the system performance. Materials and Methods Texts from patient histories and a RD knowledge base were embedded by LLMs into vectors, then encrypted using HE to obscure private information while retaining the semantic nuances. Diagnostic recommendations were generated by computing and ranking the similarities between the patient history and RD vectors in the encrypted space. The system was evaluated using 50 synthetic case reports (5 RDs, each with 10 reports). Results Applying HE did protect private information from reverse-embedding attacks. HE imposed little disruption to the diagnostic accuracy, with normalized discounted cumulative gains (nDCG) of 0.6108 ± 0.3412 (encrypted) versus 0.6083 ± 0.3415 (unencrypted). The accuracy and computational performance were tunable and consistent, as demonstrated across five different LLMs. Discussion Our privacy-preserving framework opens tremendous opportunities toward hosting and serving powerful AI solutions across institution boundaries, which would remove the need for local deidentification and incentivize users to access secure external decision-support services. Conclusions Integrating HE with LLM retrieval can promote the dissemination of nonredundant, high-capacity AI services by preserving both privacy and accuracy.

Abstract BACKGROUND: DNA methyltransferase inhibition offers palliative benefit but does not alter disease biology. in chronic myelomonocytic leukemia (CMML, Merlevede J et al. Nat Comm 2016). This was conclusively demonstrated in the phase 3 DACOTA study, where decitabine did not improve event-free survival in comparison to hydroxyurea (median 12.1 versus 10.3 months, P=0.27) in advanced myeloproliferative CMML (MP-CMML). Hence, there is in urgent need to develop disease modifying therapies for this dismal condition. We and others have demonstrated a role for RAS mutations in MP-CMML through the activation of the KMT2A-PLK1 axis. Targeting of this axis using onvansertib, an inhibitor of PLK1, has demonstrated its pre-clinical efficacy in MP-CMML patient-derived organoid and xenograft models. (Carr RM et al. Nat Comm 2021). METHODS: We conducted a single center phase 1 study to assess the safety and efficacy of onvansertib in relapsed/refractory MP-CMML (NCT05549661). We used a Bayesian Optimal Interval (BOIN) study design to determine the maximum tolerated dose (MTD). The three dose cohorts included 6, 9 and 12 mg/m2 given orally from days 1 to 21 of each 28-day cycle. Key inclusion criteria included a WHO-defined diagnosis of MP-CMML with a white blood cell (WBC) count > 13,000/mm3 and adequate organ function (estimated GFR > 50 mL/min/m2). Response was evaluated per the MDS/MPN International Working Group Criteria (Savona et al. Blood 2015). Targeting efficacy was assessed by measuring phosphorylation of TCTP (a PLK1 substrate), specifically pTCTP/TCTP ratios by immunoblotting in peripheral blood mononuclear cell (PBMC) samples pre and post cycle 1 (Days 1, 7, 21) and end of cycle 3. Pharmacokinetic studies and translational correlative studies were conducted per protocol. RESULTS: Between April 2023 and July 2025, a total of 9 patients were enrolled. The median age at enrollment was 68 years (range: 60-83); all caucasian with 4 (44%) males. All patients had a diagnosis of CMML-1 with a median bone marrow (BM) blast percentage of 4 (0-9), median hemoglobin 10 (7.1-13.2) g/dL, median WBC count of 11.1 (3.1-72.7) x 109/L (4 patients had a WBC count < 13 x 109/L at enrollment due to hydroxyurea use but were confirmed to be MP-CMML at diagnosis), absolute monocyte count 2.1 (1.9-19.6) x 109/L, and platelet count of 36 (20-494) x 109/L. TET2 (89%), SRSF2 (78%), ASXL1 (56%), NRAS, KRAS and CBL (22% each) were the most frequent molecular abnormalities. Previous therapies included hydroxyurea (n=8) and decitabine (n=1). Four patients were enrolled at the 6mg/m2dose level, 2 at the 9 mg/m2 dose level and we currently have one patient at the 12 mg/m2 dose level. While there were 2 planned patients at the 6 mg/m2 dose level, due to a case of pericardial effusion that was possibly drug related, an additional two patients were enrolled at the same dose level, without any dose limiting toxicity (DLT). Targeting efficacy was confirmed by pTCTP immunoblotting in PBMC samples from 4 patients (6 and 9 mg/m2 dose levels) showing suppressed pTCTP/TCTP ratios. No confirmed DLT has been identified thus far, and no grade 3 or 4 toxicities attributable to the study drug have been reported. Possible or unrelated grade 3 adverse effects included infections (sepsis/cellulitis, n=2, 22%), acute kidney injury (n=1, 11%), pericardial effusion (n=1, 11%), and atrial fibrillation with rapid ventricular response (n=1, 11%). Median duration on therapy was 124 (range: 15-654) days with most common cause of study discontinuation being disease progression (33%). Response was assessed as per the MDS/MPN IWG criteria in patients who received at least 1 cycle (n=8) with 3 (38%) patients meeting criteria for either hematological or optimal marrow response. Specifically, there was one (14%) patient with >100% platelet response, one (14%) with normalization of WBC count (baseline WBC count >50 x 109/L), one (14%) optimal marrow response with BM blast% reduction from 8 to 2, facilitating candidacy for an allogeneic hematopoietic stem cell transplantation, and three patients (38%) remained as stable disease. CONCLUSION: In MP-CMML, PLK1 inhibition with onvansertib is relatively well tolerated and shows preliminary efficacy in approximately 40% of patients, with one patient having an optimal marrow response at the 9 mg/m2 dose. Future directions include a dose expansion phase and studying upfront combination with DNA methyltransferase inhibitors.

ABSTRACT Alzheimer's disease (AD), a leading cause of dementia in the elderly, is traditionally characterized by neurodegeneration driven by amyloid‐beta plaques and tau tangles. However, emerging evidence reveals that AD's impact extends beyond the brain, significantly affecting skeletal health. This review integrates clinical and transgenic mouse model data to elucidate the mechanistic interplay between AD pathology and bone metabolism. AD patients exhibit increased risk for hip fractures and low bone mineral density (BMD), independent of cognitive impairment severity. We found altered calcium and alkaline phosphate levels in the blood of patients with mild cognitive impairment and AD, as assessed from the Alzheimer's Disease Neuroimaging Initiative data. Convergent risk factors—age, sex, APOE4 genotype, smoking, and vitamin D deficiency—contribute to both neurodegeneration and bone fragility. Key molecular pathways, such as Wnt/β‐catenin signaling and TREM2‐mediated osteoclast regulation, underscore shared mechanisms driving disease pathology in both systems. Mouse models of AD consistently demonstrate disrupted bone remodeling, impaired osteoblast function, and heightened osteoclast activity. Therapeutic strategies targeting overlapping pathways, including lithium, anti‐FSH antibodies, and NF‐κB inhibitors, show promise in mitigating both cognitive decline and bone loss. Collectively, these insights advocate for a more integrated view of AD that includes skeletal comorbidities, potentially guiding the development of dual‐purpose interventions.