Citations

Abstract Background Lower extremity compression therapy has been proposed as an adjunct to mitigate diuretic resistance during acute decompensation of heart failure (ADHF). We evaluated the effect of lower leg compression with elastic bandages on diuretic efficiency in patients with diuretic-resistant ADHF. Methods Historical cohort study enrolling adult cardiac ICU admissions with ADHF who used elastic bandages (EB) at Mayo Clinic in Minnesota from 2007 to 2017. The diuretic efficiency before and after applying EB was the primary outcome. Results A total of 176 patients with heart failure who received elastic bandages were included. The in-hospital mortality rate, the median hospital length of stay, and 30-day hospital readmission were 9.7%, 11.4 days (IQR 7.1–19.0), and 11%, respectively. When EB was applied, the average daily loop diuretic dosage was decreased by a mean difference of 1.19 mg per 40 mg furosemide equivalent (95% CI: 0.22 to 2.2; p = 0.016). The average post-EB diuretic efficiency was 1,355 mL (IQR 1,119, 1,590) per 40-mg furosemide, which was significantly higher than before EB application, with a mean difference of 340 mL per 40-mg furosemide equivalent (95% CI: -554, -125; p = 0.002). A higher diuretic efficiency led to lower diuretic doses. Average daily urine output before and after EB application was not different (2,642 pre-EB vs. 2,792 post-EB mL/day; 95% CI: -455 to 154; p = 0.33). Conclusion EB use enhances diuretic efficacy during ADHF. Further studies are needed to determine the potential clinical benefits and risks of applying EB compression treatment in ADHF.

Systematic reviews of randomized controlled trials (RCTs) support evidence-based clinical advancement. Clinical trial reporting guidelines, including CONSORT, establish common data elements that support systematic synthesis of published trial outcomes. However, the rapid rise of digital health technologies presents new challenges for reporting guidelines, and substantial real-world reporting heterogeneity has been observed. Therefore, machine-assisted evaluation of reporting adherence offers a pathway for more agile information management to accommodate the diverse and disruptive nature of digital health technology-enabled RCTs. Here, we propose an accelerated retrieval-to-abstraction literature synthesis framework to conduct article screening, text and table data preprocessing, and full-text review using an unmodified pretrained large language model. We apply this methodology in two digital health technology-enabled RCT case studies evaluating adherence to technology and digital health equity guideline reporting items. Large language model performance was on par with inter-annotator agreement F1 scores (≥0.9) in both screening and information-extraction tasks. These findings support the scalability and rigor of accelerated evidence synthesis and demonstrate the importance of common reporting elements across digital health technology-enabled RCTs.

Abstract Glioblastoma (GBM) is a lethal, treatment-resistant brain cancer characterized by diffuse brain invasion, complex neuron-glioma signaling, and cellular, molecular, and spatial heterogeneity. Surgery is the current mainstay of treatment, aimed at maximizing tumor cytoreduction, decompressing the brain to improve neurological function, and providing material for analyses to inform prognostication and adjuvant treatment selection. To accurately model GBM surgical resection and tissue sampling in preclinical animal models, we developed the Maximal Safe Intracranial Surgery (MSIS) system. MSIS integrated a miniaturized neurosurgical resection device and enabled a safe, tunable, and reproducible approach for intracranial tumor resection, multi-regional biopsy, and longitudinal tissue sampling in murine models, with stereotactic, sterile, and high-viability tissue collection. MSIS was applied across five distinct orthotopic brain tumor models to reproducibly generate histologically and radiologically representative models of post-surgical GBM. Similar to humans, increasing the extent of resection correlated with improved animal survival, and recurrent growth patterns varied significantly by model type. Transcriptional profiles of matched primary vs. post-surgical recurrent tumors revealed enrichment of margin-associated transcriptional states consistent with neuron-glioma signaling. Overall, the MSIS system offers opportunities to model the standard-of-care, dissect the spatial and temporal dynamics of GBM, and pair therapeutic discovery with contextual tumor pathobiology.

Abstract Background Regorafenib and anti-EGFR monoclonal antibodies are standard therapies for RAS/BRAF wild-type (WT) metastatic colorectal cancer (mCRC) in the refractory setting, but the optimal sequencing remains unknown. Herein, we report results from REVERCE II, a randomized phase II trial evaluating the efficacy and safety of sequencing these therapies. Methods Patients with RAS WT mCRC previously treated with fluoropyrimidine, oxaliplatin, and irinotecan were randomized to receive (R-E) regorafenib followed by anti-EGFR therapy ± irinotecan or (E-R) anti-EGFR therapy ± irinotecan, followed by regorafenib sequentially. The primary endpoint was overall survival (OS). Key secondary endpoints included progression-free survival (PFS) with initial (PFS1) and second treatment (PFS2), sequential treatment PFS (stPFS), objective response rate (ORR), and safety. Exploratory endpoints included serial biomarker analyses. Unfortunately, the trial was closed prematurely due to slow enrollment and lack of funding. Results Twenty-two patients were enrolled (R-E: n = 12; E-R: n = 10) with sequential treatment initiated in 33% (4/12) and 80% (8/10) of patients in R-E and E-R, respectively. Median OS for R-E and E-R was 21.9 and 13.3 months, respectively (one-sided P = .114). PFS1 was 2.5 (R-E) vs 5.6 months (E-R) (one-sided P = .998), while PFS2 was 3.7 vs 2.0 months (one-sided P = .086). stPFS was similar (9.3 vs 9.6 months; one-sided P = .358). ORR for initial treatment was 0% in R-E and 20% (2/8) in E-R. Sequential treatment ORR was 25% (n = 1/4) in R-E and 0% in E-R. No significant differences were observed adverse effects between arms. Conclusion While the study was closed prematurely prior to completion, the sequence of regorafenib before to anti-EGFR therapy suggested a possible survival benefit. Further studies may help better determine the optimal treatment sequencing strategy in refractory mCRC.

Background: Large language models (LLMs) are increasingly used in patient-facing clinical applications, creating a need for scalable methods to evaluate the accuracy, safety, and appropriateness of their responses. Although physician review remains the reference standard, it is resource-intensive and difficult to scale. Objective: To develop and clinically validate an LLM-as-a-judge framework for evaluating responses to patients’ periprocedural questions. Methods: We retrospectively evaluated 1285 patient–AIVA interactions from 112 patients across two Mayo Clinic sites and multiple surgical specialties. Physician reviewers classified interactions using a predefined true-positive (TP), false-negative (FN), true-negative (TN), and false-positive (FP) framework. The finalized LLM judge independently evaluated the same interactions. Agreement was assessed using four-class and category-specific agreement, Cohen’s kappa, and a secondary binary analysis of response correctness. Results: Overall, four-class agreement was 93.1% (95% CI, 91.7–94.5%), with Cohen’s κ = 0.852 (95% CI, 0.823–0.882). Category-specific agreement was 94.1% for TP, 90.7% for FN, 95.5% for TN, and 84.9% for FP classifications. In the binary correctness analysis, accuracy was 93.4% (95% CI, 92.0–94.7%), sensitivity 94.6%, specificity 90.2%, precision 96.2%, and F1-score 0.954. McNemar’s test demonstrated no significant asymmetry between paired classifications (p = 0.11). Conclusions: A clinically grounded LLM judge closely approximated physician evaluation of patient-facing AI responses. These findings support its potential as a scalable assistive monitoring tool while preserving physician oversight for uncertain or safety-sensitive interactions.

Abstract Background Evaluation of middle ear prostheses requires high-resolution imaging to assess position and structural integrity. Purpose To assess the diagnostic performance of photon-counting-detector (PCD)-CT for middle ear prosthesis imaging in cadaveric specimens. Materials and Methods Nine different middle ear prostheses were implanted in five fresh-frozen whole-head specimens, both in correct anatomic position and displaced to mimic malposition. Specimens were scanned on PCD-CT (NAEOTOM Alpha, Siemens) and energy-integrating-detector (EID)-CT (SOMATOM Force, Siemens), reconstructed at 0.2mm and 0.4mm, respectively. Eighteen cases (9 prostheses, 2 positions) per scanner were evaluated by two neuroradiologists (36 readings), blinded to scanner and position, who scored prosthesis position, reader confidence, and overall image quality. Agreement with the reference standard and between readers was compared. Results PCD-CT showed higher diagnostic confidence compared with EID-CT, while accuracy and inter-reader agreement did not demonstrate significant differences between scanners. Correct readings, PCD-CT versus EID-CT: lateral contact, 31/36 (86.1%; 95% CI: 58.3%-94.4%) vs. 29/36 (80.6%; 55.6%-91.7%), p = 0.500; medial contact, 30/36 (83.3%; 55.6%-94.4%) vs. 26/36 (72.2%; 50.0%-83.3%), p = 0.250. PCD-CT identified all cases of complete contact loss (score 3), while EID-CT missed one reading per contact. Inter-reader agreement may be higher with PCD-CT vs EID-CT (lateral: κw: 0.804 [0.578-0.941] vs. 0.571 [0.302-0.778], p = 0.250; medial: 0.684 [0.191-0.934] vs. 0.481 [0.024-0.835], p = 0.250). Reader confidence was significantly higher with PCD-CT; mean paired differences for Readers 1 and 2 were 9.4 (1.8-17.1; Holm-adjusted p = 0.019) and 23.1 points (13.4-32.7; p < 0.001) for lateral contact, and 16.4 (10.0-22.8; p < 0.001) and 17.8 points (6.8-28.8; p = 0.007) for medial contact. Readers preferred PCD-CT image quality in 16/18 (88.9%; 65.3%-98.6%) and 17/18 cases (94.4%; 72.7%-99.9%), respectively. Conclusion PCD-CT was associated with significantly higher reader confidence and preferred image quality, without significant differences in diagnostic accuracy or inter-reader agreement. These findings suggest PCD-CT is a promising modality for postoperative middle ear evaluations that warrants further investigations.

Abstract Background Optimal treatment following first-line immunotherapy for advanced hepatocellular carcinoma (HCC) remains uncertain. This study evaluated futibatinib, a pan-FGFR inhibitor, combined with pembrolizumab in previously treated patients with FGF19-expressing HCC. Methods Eligible patients were ≥18 years old with advanced HCC, prior systemic therapy, Child-Pugh A-B7 liver function, ECOG 0–2, and FGF19 overexpression. Patients received futibatinib 20 mg orally daily plus pembrolizumab 200 mg intravenously every 3 weeks. The primary endpoint was progression-free survival at 6 months (PFS6). Secondary endpoints included overall survival (OS), objective response rate (ORR), quality of life, and safety. Results Between August 2021 and March 2024, 14 patients enrolled and 13 were evaluable. Median age was 72 years (range, 60–87), 53.8% were male, 92.3% had Child-Pugh A liver function, and all had received prior immunotherapy. Median progression-free survival was 15 weeks (95% CI, 9.0–NE), with a PFS6 rate of 23.1% (95% CI, 8.6–62.3%). ORR was 0% (95% CI, 0–24.7%), while stable disease was observed in 53.9% of patients. Median OS was 55.7 weeks (95% CI, 47.3–NE). Grade ≥3 adverse events occurred in 69.2% of patients, and grade ≥4 adverse events in 15.4%. One grade 5 event (acute-on-chronic kidney injury) was deemed unrelated to treatment. Common adverse events included hyperphosphatemia, fatigue, diarrhea, elevated transaminases, dry eye, nail changes, rash, and oral mucositis. Conclusions Futibatinib plus pembrolizumab demonstrated an acceptable safety profile but did not meet efficacy thresholds supporting further study in FGF19-expressing HCC. Effective post-immunotherapy treatment strategies remain an unmet need.

Background: Accurate identification of patients with metabolic dysfunction–associated steatohepatitis (MASH) at risk of disease progression (at-risk MASH) is critical for guiding resmetirom and semaglutide therapy. Inconsistent definitions of at-risk MASH complicate clinical risk classification and comparisons across studies. This study aimed to standardize the definition of at-risk MASH and propose optimized MRE-based liver stiffness (LS-MRE) cutoffs to improve noninvasive risk stratification and better guide pharmacotherapy. Methods: In this prospective study, 413 patients with suspected or diagnosed MASLD from two medical centers underwent liver biopsy and MRE. Results: A simplified definition of at-risk MASH (MASH with fibrosis stage≥2) identified a broader group of patients requiring clinical attention than a stricter definition requiring NAFLD Activity Score≥4 with ≥1 point in steatosis, inflammation, and ballooning (n=122 vs. 77). LS-MRE alone showed the highest diagnostic accuracy for identifying at-risk MASH using the simplified definition (AUC 0.91 [0.87, 0.95]) and cirrhosis (AUC 0.93 [0.88, 0.99]), compared with FIB-4 and individual laboratory parameters (AST, ALT, and platelet count) ( p <0.01 for all, paired DeLong tests). Optimized LS-MRE cutoffs (2.8–6.4 kPa) captured 25 additional biopsy-proven pharmacotherapy candidates (28% of all eligible patients) compared to AASLD guidelines–recommended cutoffs (3.1–4.4 kPa), improving sensitivity (80% vs. 50%) with acceptable specificity (73% vs. 88%). Conclusions: A simplified definition of at-risk MASH enables broader identification of treatment candidates. LS-MRE demonstrates excellent diagnostic performance and improves noninvasive risk stratification when paired with optimized cutoffs. These findings support the use of MRE to guide resmetirom and semaglutide eligibility and improve access to emerging therapies.

Background: Societal guidance recommends sequential noninvasive testing for metabolic dysfunction–associated steatotic liver disease (MASLD), using Fibrosis-4 (FIB-4) ≥1.3 as a gatekeeper before second-line assessment with the Enhanced Liver Fibrosis (ELF) score (<7.7 as a rule-out). The impact of this pathway design on the detection of fibrotic MASH remains uncertain. We aimed to evaluate the performance of guideline-endorsed sequential testing and to determine optimized ELF thresholds within alternative strategies. Methods: In this prospective study, patients with or at risk for MASLD underwent FIB-4, ELF testing, magnetic resonance elastography (MRE), and liver biopsy within 1 month. Fibrotic MASLD was defined as fibrosis stage ≥2 on histology or liver stiffness ≥3.1 kPa on MRE. Diagnostic performance of ELF was assessed, and sequential (ELF applied after FIB-4 ≥1.3) and concurrent (either test positive) testing strategies were compared. Results: Among 186 participants (median age 51 years; 69% women), fibrotic MASLD was present in 71 (38%). ELF demonstrated AUROC 0.77 (95% CI 0.70–0.84); the cohort-derived cutoff was 9.4 (80% sensitivity, 65% specificity). Lower cutoffs (8.7 and 7.7) increased sensitivity (≥90% and 100%) but low specificity (34% and 4%). FIB-4 ≥1.3 showed limited sensitivity (57%), resulting in 43% of fibrotic MASH cases being excluded from second-line testing. Sequential application of ELF, therefore, failed to overcome this sensitivity ceiling. A concurrent strategy using FIB-4 ≥1.3 or ELF ≥9.4 improved sensitivity to 87% with 64% specificity. Conclusions: Sequential FIB-4–based testing substantially limits the detection of fibrotic MASLD due to gatekeeping sensitivity constraints. Concurrent first-line testing with FIB-4 and ELF improves case identification, underscoring the importance of pathway design in optimizing noninvasive fibrosis assessment.