Citations

Introduction: Pulmonary hypertension (PH) is a life-threatening, progressive disease with non-specific symptoms, often leading to delayed diagnosis. Early identification of World Health Organization Group 1 (Pulmonary Arterial Hypertension, PAH) and Group 4 (Chronic Thromboembolic Pulmonary Hypertension, CTEPH) is essential, as effective therapies can improve outcomes. Hypothesis: An electrocardiogram-based AI algorithm for PH detection (ECG-AI PH) may enable earlier diagnosis and reduce healthcare utilization, including hospitalizations and procedures. Methods: Retrospective analysis used a de-identified data platform of >7M clinical records from a multistate integrated health system. Adult precapillary PH patients (mPAP >20 mmHg, PVR >2 WU, PCWP ≤15 mmHg) were identified as PAH or CTEPH based on ICD codes, use of approved therapies, or surgical interventions (for CTEPH) between 2002 and 2024. ECG-AI PH was applied to ECGs within 30 days of diagnostic right heart catheterization, using a 5:1 randomly sampled PH-negative control cohort. Training set patients were excluded. Clinical event frequency was compared between two intervals: from first possible PH symptom (dyspnea, syncope, chest pain, fatigue, lower limb swelling) to diagnosis, and from symptom onset to first positive ECG-AI PH prediction. Results: A total of 1882 PAH and 359 CTEPH patients met inclusion criteria. Of these, 1340 PAH and 258 CTEPH patients had symptom codes prior to diagnosis. Both groups showed prolonged intervals from symptom onset to diagnosis, with multiple diagnostic procedures and hospitalizations (Figure). ECG-AI PH performance evaluation on the test set included 647 PAH and 152 CTEPH patients. ECG-AI PH achieved AUCs of 0.90 and 0.89 for PAH and CTEPH, sensitivities of 80.3% and 76.8%, and specificities of 83.4% and 82.4%. Among those tested, 576 PAH and 95 CTEPH patients had a positive ECG-AI PH prediction after symptom onset but before diagnosis. Compared to the current patient journey, the interval between initial symptoms and a positive output from ECG-AI PH was shorter and had fewer diagnostic tests/visits. Conclusion: ECG-AI PH demonstrated strong performance in detecting PAH and CTEPH. It may reduce diagnostic delays, support earlier PH-focused screening (e.g., echocardiograms evaluating the right heart), enable earlier intervention, and reduce pre-diagnosis healthcare burden, benefitting both patient outcomes and healthcare system efficiency.

Background: Immunotherapy-based regimens have expanded the treatment landscape for unresectable hepatocellular carcinoma (uHCC); however, real-world data are limited. Methods: This retrospective, observational study used data from electronic medical records from Mayo Clinic sites across the United States. Patients with uHCC who initiated a first-line (1L) systemic therapy between June 2020–October 2022 with ≥2 follow-up visits were included. Treatment patterns, overall survival (OS), and post-index gastrointestinal (GI) bleeding were assessed by GI bleeding risk defined by Child–Pugh Class B or C, pre-index GI bleeding, uncontrolled hypertension, or significant varices and band ligation. Results: Of 186 included patients, 68.8% had GI bleeding risk and 31.2% did not. Atezolizumab plus bevacizumab was the most common 1L systemic therapy in patients with or without GI bleeding risk (72.7% and 29.3%, respectively). Median OS (95% confidence interval) with atezolizumab plus bevacizumab was 12.8 (8.0–19.3) months and not reached in patients with and without GI bleeding risk, respectively. OS rates with atezolizumab plus bevacizumab in patients with or without GI bleeding risk, respectively, were 52.3% and 70.6% at 12 months, 41.6% and 57.8% at 18 months, and 34.6% and 51.3% at 24 months. Post-index GI bleeding with atezolizumab plus bevacizumab occurred in 19.4% and 5.9% of patients with and without GI bleeding risk, respectively. Conclusions: During this study period, atezolizumab + bevacizumab was the most common 1L therapy for patients with uHCC, regardless of GI bleeding risk. OS rates with atezolizumab + bevacizumab were lower in patients with versus without GI bleeding risk. Findings highlight the unmet need for guidance on characteristics-driven treatment decisions.

ImportanceReceipt of anthracycline-based chemotherapy and statin therapy have been found to be associated with deterioration of cognitive function in patients with cancer.ObjectiveTo assess the association of statins with attention, verbal fluency, and executive function in patients receiving doxorubicin treatment for cancer.Design, Setting, and ParticipantsThis secondary analysis of the Preventing Anthracycline Cardiovascular Toxicity With Statins randomized clinical trial, conducted across 31 US community and academic sites from February 5, 2014, through September 24, 2020, involved participants with stage I to IV lymphoma or stage I to III breast cancer undergoing doxorubicin treatment. The data analysis was conducted between August 1, 2024, and April 4, 2025.InterventionsParticipants were randomized (1:1) to receive a daily 40-mg dose of atorvastatin or placebo before initiating doxorubicin treatment and then continued for 24 months.Main Outcomes and MeasuresAttention, executive function, and verbal fluency were assessed using the Trail Making Test part A (TMT-A) and part B (TMT-B) and the Controlled Oral Word Association test, respectively. Time to complete the assessments were modeled using a linear mixed model, while errors (counts) were modeled using a generalized linear mixed model assuming a Poisson distribution.ResultsA total of 238 participants (mean [SD] age, 49 [12] years; 217 female [91.2%]) were randomized to the statin (n = 118) or placebo (n = 120) groups. Values for TMT-A at 6 and 24 months were similar between the statin group (mean, 32.5 seconds [95% CI, 29.4-35.7 seconds] and 29.8 seconds [95% CI, 26.4-33.2 seconds], respectively) and the placebo group (mean, 28.4 seconds [95% CI, 25.2-31.5 seconds] and 27.8 seconds [95% CI, 24.5-31.0 seconds], respectively). From before treatment to 24 months after treatment, statin recipients had a significant mean improvement of 10.2 seconds for TMT-B (95% CI, 1.9-18.5 seconds), whereas placebo recipients had a nonsignificant improvement of only 0.2 seconds for TMT-B (95% CI, −8.5 to 8.1 seconds). Analysis of the time-by-treatment interaction showed no significant difference between groups. For Controlled Oral Word Association scores in the same period, both the placebo and statin groups showed a mean improvement of 3.62 points (95% CI, 1.71-5.54 points) and 4.74 points (95% CI, 2.69-6.79 points), respectively, with no difference between groups.Conclusions and RelevanceThis secondary analysis of a randomized clinical trial suggests that the receipt of statins during doxorubicin therapy for lymphoma or breast cancer was not associated with worsening cognitive functions over 24 months. Within-group analyses suggested that statins may have contributed to better executive function on the TMT-B.Trial RegistrationClinicalTrials.gov Identifier: NCT01988571

Background: Early detection and prediction of arterial hypertension (HTN) are vital to reduce cardiovascular morbidity and mortality. AI has enhanced ECG diagnostic capabilities, enabling detection of latent cardiovascular abnormalities. Aims: This systematic review evaluates current evidence on AI-enhanced ECG for detection and prediction of HTN in in both clinical and subclinical populations. Methods: A search was conducted on April 28, 2025, in PubMed, Scopus, and Web of Science. Gray literature and study reference lists, plus the first 10 pages of Google Scholar, were screened. The review followed PRISMA 2020 guidelines. Bias risk was assessed using PROBAST (for prediction models) and QUADAS-2 (for diagnostic studies). Two reviewers independently performed study selection and data extraction after pilot training. Results: A total of 26 studies were included: 16 focused on detection and 10 on prediction of HTN using AI-enhanced ECG. Overall, 96% reported successful outcomes, underscoring the strong diagnostic and prognostic potential of this approach. Among detection studies, 94% successfully identified hypertensive individuals, classified blood pressure (BP) levels, or evaluated cardiovascular risk. Most were retrospective in design (75%), with internal validation in 44% and external validation in 19%. Sample sizes ranged from 100 to over 120,000, including healthy and cardiovascular patients. Commonly used algorithms included deep Learning, random forest, and ensemble methods, often using single-lead ECG data, highlighting the feasibility of wearable technologies. QUADAS-2 identified 56% of studies had low risk of bias. All prediction studies achieved accurate estimation of systolic/diastolic BP or predicted future HTN onset. Designs were 40% experimental and 50% retrospective. All studies reported internal validation, and 50% included external validation. AI models such as CNNs, ResNet-LSTM, and U-Net were used across clinical and community-based populations, with follow-up periods of up to 6.8 years. Populations were predominantly normotensive or mixed. PROBAST assessments showed 60% low and 40% moderate risk of bias. Conclusion: AI-enhanced ECG is a promising, non-invasive tool for early detection and prediction of HTN. However, external validation and clinical trials are needed to support integration and generalizability.

Abstract BRCA2 mutations contribute to the pathogenesis and treatment sensitivity of a subset of ovarian, breast, prostate, and pancreatic cancers. When these cancers become therapy resistant, secondary mutations that restore the BRCA2 open reading frame are found in half the cases, but other causes of resistance remain incompletely understood. Here, we identified translational readthrough of a premature termination codon (PTC) as a cause of resistance to poly(ADP-ribose) polymerase inhibitors (PARPis) and cisplatin in cells derived from the BRCA2-mutated ovarian cancer line PEO1 by PARPi selection. Despite persistence of the signature 4965C > G (p.Y1655X) BRCA2 mutation, low-level expression of full-length BRCA2 protein was detectable in these cells by immunoblotting and tandem mass spectrometry. Either BRCA2 knockdown or gene interruption 5′ or 3′ to the PTC restored treatment sensitivity, implicating BRCA2 in the resistance. Reporter assays demonstrated UAG-selective readthrough in the resistant clones but not parental cells. Moreover, custom searching of global proteomic data indicated readthrough of stop codons, particularly UAGs, in additional proteins in the resistant clones. Finally, multi-omic analysis identified multiple changes in the nonsense-mediated decay and termination machineries that favor readthrough. Accordingly, the present results identify PTC readthrough as a potential mechanism of drug resistance in cells with BRCA2 nonsense mutations.

Background Sinonasal pathology, such as chronic inflammatory disease or neoplasm, can cause secondary nasolacrimal duct obstruction (sNLDO). There is a paucity of data on dacryocystorhinostomy (DCR) techniques and outcomes for NLDO secondary to known sinonasal pathologies. Objective To describe case characteristics, management, and surgical outcomes of sNLDO at a tertiary academic center. Methods Charts of DCR cases with at least 3 months follow-up done at a single institution over a 5-year period (2018-2022) were reviewed. Patient and surgical characteristics were recorded and compared between secondary and primary NLDO groups. The primary outcome was early functional success of DCR, assessed by improvement or resolution of epiphora at 3 months. Results A total of 236 cases were included. Of those, 79 (33.5%) were classified as secondary NLDO associated with sinonasal pathology. Reasons for sNLDO were sinonasal cancer (46.8%), chronic inflammatory disease (24.1%), trauma (16.5%), or prior head and neck radiation (31.6%). When comparing primary acquired NLDO to sNLDO cases, DCR for sNLDO was more likely to be performed endoscopically (68.4% vs 31.8%, P < .01), by a multidisciplinary team of rhinologist and oculoplastic surgeon (70.9% vs 8.3%, P < .01), and required bilateral surgery (36.7% vs 5.7%, P < .01). The overall functional success rate was 79.7% for sNLDO compared to 87.9% for primary acquired NLDO ( P < .01). Revision surgery for sNLDO, history of sinonasal malignancy, and radiation were associated with lower functional success rates (66.7%, 70.3%, and 68.0%, respectively). Canalicular pathology was more common in sNLDO cases (29.0% vs 8.3%, P < .01). Conclusions Secondary NLDO was more commonly managed with an endoscopic approach by a multidisciplinary team. Revisions for sNLDO, as well as cases with a history of sinonasal malignancy and radiation, had a markedly lower success rate, and sNLDO was frequently associated with canalicular obstruction. These findings may facilitate surgical planning and patient counseling.

Importance Advances in Alzheimer disease (AD) have shifted research focus to earlier disease stages, necessitating more scalable approaches to identify cognitively unimpaired individuals with amyloid β (Aβ) pathology. Objective To assess the utility of plasma phosphorylated tau 217 (p-tau217) for classifying Aβ status in cognitively unimpaired individuals, both as a stand-alone test and in a 2-step approach where positive plasma results were confirmed using a second modality (Aβ positron emission tomography [PET] or cerebrospinal fluid [CSF]). Design, Setting, and Participants This cross-sectional cohort study used data collected between June 2009 and March 2024. We included 2916 cognitively unimpaired participants from 12 international independent observational cohorts in the US, Europe, Australia, and Canada with available plasma p-tau217 levels and CSF or PET Aβ biomarkers. Performance comparisons between mass spectrometry and immunoassay-based p-tau217 measurements were also performed (n = 964). Exposures Plasma p-tau217 levels measured by immunoassay. Main Outcome and Measures Aβ status, determined by CSF or Aβ PET biomarkers. Results Participants had a mean (SD) age of 66.9 (9.9) years; 971 (33.3%) were Aβ positive by either CSF or PET, 1667 (57.2%) were women, and 1108 (38.1%) carried at least 1 APOE ε4 allele. As a stand-alone test, plasma p-tau217 achieved a positive predictive value (PPV) of 79% (95% CI, 74-84) and an overall accuracy of 81% (95% CI, 80-82). In a 2-step workflow, the PPV and accuracy significantly increased to 91% (95% CI, 86-95). While this approach required screening of 677 individuals with plasma p-tau217 to identify 100 Aβ-positive individuals, compared to 536 participants when using PET alone, it reduced the need for PET testing to 124. Immunoassays demonstrated comparable PPVs to mass spectrometry (80% [95% CI, 74-86] vs 85% [95% CI, 81-90]; P = .12) but significantly lower overall accuracy (82% [95% CI, 79-84]% vs 88 [95% CI, 86-90]; P &amp;lt; .001) and true Aβ-positive detection rate (49% [95% CI, 43-55] vs 69% [95% CI, 64-75]; P &amp;lt; .001). Conclusions and Relevance The findings highlight the potential of plasma p-tau217 as a stand-alone test—or when used in a sequential 2-step approach alongside PET or CSF testing—as a cost-effective, scalable, and minimally burdensome strategy for identifying preclinical AD. Tailored screening workflows that incorporate p-tau217 can improve efficiency in participant selection for preclinical AD trials and, in the future, help guide access to disease-modifying treatments.

Abstract Background The isocitrate dehydrogenase (IDH) mutation status is a key molecular marker in diagnosing and treating brain tumors. Currently, it is determined via invasive tissue biopsy. Recent advances in deep learning (DL) have offered promising non-invasive alternatives for determining IDH status. However, their clinical translation is hindered by a significant gap between DL predictions and their clinical applicability. The limited transparency of many DL-networks and inadequate evaluation metrics hinders trust and adoption, as clinicians require clear and validated insights for determining IDH status. These challenges highlight the need for robust validation and measures of predictive reliability to make DL-predictions clinically actionable. Methods We developed a unique approach for non-invasive prediction of IDH status using MRI. We combine a voxel-wise-segmentation network(MC-net) with Bayesian logistic regression (BLR) to provide an IDH status and estimate confidence scores. We utilized a comprehensive dataset of 2,481 glioma cases from eight institutions. Results Our framework(MC-net + BLR) demonstrated robust performance achieving 96.4% and 95.1% classification accuracies on diverse databases, with an AUC of 0.98. The BLR was implemented exclusively on held-out test data, ensuring that the derived confidence scores are independent of the training or validation phases. The derived confidence scores showed a low Brier score of 0.0125, highlighting its superior calibration and uncertainty quantification. Conclusion The developed framework provides an IDH status and a confidence score, offering clinicians an additional layer of assurance in prediction reliability. It bridges the gap between high-performing DL models and their clinical applicability by addressing the challenges in prediction reliability. Our framework is a significant advancement in non-invasive determination of IDH-status and confidence-informed therapeutic decision-making in neuro-oncology.

Purpose: To evaluate the role of women physicians in thyroid eye disease (TED) research and clinical care and describe relevance for academic medicine and patient access. Methods: A cross-sectional study of TED-related publications indexed on PubMed in 2000, 2010, and 2020, and of completed or currently active clinical trials was conducted. Author, journal, and principal investigator (PI) characteristics were noted. Author/PI gender was determined from institutional profile or subsequently via a previously described Google-based algorithm. American Society of Ophthalmic Plastic and Reconstructive Surgery (ASOPRS), American Board of Ophthalmology (ABO), and American Thyroid Association (ATA) information was obtained via direct communication. Trends in participation of men and women were evaluated. Results: A total of 444 manuscripts (61 from 2000; 117 from 2010; 266 from 2020) were included. The percentage of women as first and senior authors increased from 2000 to 2020 (p < 0.0001 and p = 0.0003, resp.). First author gender distribution was near equal in 2020 (52.6% men, 47.4% women [p = 0.39]), whereas senior authors remained mostly men (68.0% men, 32.0% women [p < 0.0001]). As of 2020, the gender imbalance of senior authors was greater on multidisciplinary author teams (74.8% men, 25.2% women) compared with single specialty author teams (62.2% men, 37.8% women). Of the 98 TED-related clinical trials included, 34 had women PIs (35%). The percentage of women as members in ASOPRS, ABO, and ATA all increased significantly from 2000 to 2020, but remained disproportionately men in 2020 (all p < 0.0001). Conclusion: Gender-based discrepancies persist in TED research, especially among senior authors, who determine research direction, and on multidisciplinary teams, which reflect the preferred model of care. Further efforts are required to determine whether these findings affect patients with TED.