Citations

Abstract Background: The limited efficacy of oncolytic virotherapy (OV) in gastrointestinal malignancies has been attributed to an immunologically inert tumor microenvironment. While cisplatin is widely used as a cytotoxic agent, its immunomodulatory effects remain incompletely defined. We hypothesized that cisplatin may reprogram the tumor immune landscape to enable more effective OV-mediated anti-tumor responses. Methods: We evaluated the effects of cisplatin in combination with two distinct oncolytic platforms, vesicular stomatitis virus (VSV) and adenovirus (AdV), across both in vitro and in vivo colorectal cancer models. In vivo efficacy was assessed using MC38 subcutaneous and peritoneal metastatic models, with analysis of tumor growth, survival, and toxicity. Immune profiling was performed using single-cell RNA sequencing (scRNA-seq), flow cytometry, and bulk RNA sequencing to interrogate changes in CD8+ T cell states and functional programs. Guided by these preclinical findings, a clinical trial combining XELOX, PD-1 blockade, and OV therapy was initiated (ChiCTR2500107607). Results: Cisplatin markedly enhanced OV-mediated anti-tumor efficacy across models, resulting in sustained tumor control and improved survival in both subcutaneous and peritoneal disease settings. This effect was conserved across viral platforms, indicating a virus-agnostic interaction. Mechanistically, cisplatin induced a robust interferon (IFN)–dominant transcriptional program, characterized by upregulation of interferon-stimulated genes, including IRF7, IRF9, and ISG15. Single-cell analysis revealed that combination therapy reshaped the CD8+ T cell compartment, promoting the expansion of a population with transcriptional and functional features of central memory T cells (Tcm), alongside reduced exhaustion signatures. These changes were accompanied by enhanced cytokine signaling pathways associated with T cell maintenance and function, including IL-2, IL-7, and IL-15. In vitro, the combination of cisplatin and OV increased tumor cell death and interferon-γ production, supporting a model in which cytotoxic and immunostimulatory effects are mechanistically linked. A phase I clinical trial has been initiated. In gastric cancer, patients with retroperitoneal metastases are being treated with OV in combination with PD-1 blockade (ongoing). In colorectal cancer, the strategy is being evaluated in the neoadjuvant setting with an rectal preservation approach. Conclusions: Our findings identify a previously unrecognized role for cisplatin in priming an interferon-driven immune program that promotes Tcm differentiation and function, thereby enabling durable responses to oncolytic virotherapy. This work provides a mechanistic rationale for leveraging chemotherapy-induced immune reprogramming to enhance viro-immunotherapy and supports the clinical translation of this strategy in gastrointestinal malignancies. Citation Format: Joseph Brandon. Parker, Zhongyi Dong, Tao Liu, Qisen Guo, Haoyu Zhang, Bo Ni, Long Bai, Yiwen Chen, Jiasheng Zong, Xiang Xia, Jiaojiao Zheng, Hoang Tran Pham, Aya Elalfy, Demarcus Ingram, Ahmed Abdel-Hakeem, Muhammad Ali, Shannon Hartzell, Zizhen Zhang, Vivienne Yang. Cisplatin primes an interferon-driven CD8+ central memory T cell program to enable oncolytic virotherapy in gastrointestinal cancer malignancies [abstract]. In: Proceedings of AACR Drug Discovery and Development (AACR D3) Conference; 2026 Jul 21-24; Boston, MA. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(14_Suppl):Abstract nr B024.

Objective Respiratory syncytial virus (RSV) is an important viral pathogen in children, older adults and adults with certain high-risk conditions. In our prospective community-based cohort study of adults 50 years and older before the COVID-19 pandemic, the incidence of RSV acute respiratory infection (ARI) was 48.6 cases/1000 person-years (PY), which decreased to near zero during the COVID-19 pandemic. Our objective was to determine the incidence of RSV-ARI during the 2 years following the initial phase of the pandemic (2021–2022 and 2022–2023). Design This is a community-based prospective cohort study. Setting Adults living in southeast Minnesota from October 2021 to September 2022 (year 3) and October 2022 to September 2023 (year 4). Participants Adults≥50 years (n=2500). Primary and secondary outcomes We calculated incidence and attack rates for RSV-ARI as primary outcome and reported hospitalisations, pneumonia and death following ARI as secondary outcome. Results There were 2500 participants in the study with a mean age of 68.2 years (SD 9.3) at the start of year 3. Participants were predominantly female (60%), non-Hispanic white (96%) and residing in urban areas (76%). The incidence rate of RSV-ARI was 6.12/1000 PY (95% CI 3.42 to 10.09) in 2021–2022 and 16.40/1000 PY (95% CI 11.66 to 22.42) in 2022–2023. We noted higher attack rates of RSV-ARI during the winter months. There were no hospitalisations, pneumonia or deaths within 30 days of RSV-ARI. Conclusions Compared with the period before the COVID-19 pandemic, RSV-ARI incidence was lower in both 2021–2022 and 2022–2023 periods. In 2022–2023, the incidence increased 2.7-fold compared with the 2021–2022 period. The observed incidence rates, particularly the significant increase in the most recent season, underscore the continued public health relevance of RSV-ARI and support the rationale for ongoing RSV vaccination efforts to mitigate its overall burden in the population.

Abstract Motor speech disorders (MSDs) are early markers of neurological disease, but expert perceptual analysis is rarely available outside specialized centers. Automated speech analysis offers a scalable alternative, yet prior studies have not systematically compared modeling approaches or assessed clinically relevant metrics in independent datasets. This study compared static acoustic features, articulatory informed Phonet features, and self-supervised pretrained models for binary and multi label MSD classification. We trained and evaluated models on 583 speech samples using speaker level splits. Baseline models included logistic regression and Gated Recurrent Units (GRUs) trained on eGeMAPS and MFCCs. We extracted three types of Phonet derived features and evaluated pretrained HuBERT and SSAST models in frozen, partially fine-tuned, and fully fine-tuned configurations. Binary classification distinguished MSDs from controls, while multi label classification identified six MSD subtypes. Models were assessed using validation AUC, and cut points were tested on two independent datasets. Pretrained and Phonet based models substantially outperformed static acoustic features. In binary classification, HuBERT achieved the highest AUC (0.95), while compact Phonet derived GRUs achieved comparable performance (up to 0.94). These models generalized well to independent datasets, maintaining high sensitivity (0.94) and specificity (0.97). In multi label classification, Phonet models achieved the highest macro average AUC (0.86), but threshold-based subtype performance declined on unseen data. Automated MSD detection is feasible and clinically promising. Binary classification generalized well, whereas multi label classification showed limited threshold stability across datasets.

Background The current classification of pulmonary hypertension (PH), based largely on expert opinion, has limitations in prognostication and guiding therapies. We hypothesize that novel PH clusters that predict survival will reveal mechanistic phenotypes associated with biomarkers of vascular health across all PH groups. Methods We first identify novel PH clinical clusters by performing unsupervised clustering analysis on the CC‐PH (Cleveland Clinic PH) registry (N=1529). We develop classification models to predict the new PH clusters and then apply them to the multicenter PVDOMICS (Pulmonary Vascular Diseases Phenomics) cohort (N=853) for validation. We compare transplantation‐free survival across the new PH clusters. We quantify metabolites of the arginine‐nitric oxide pathway and D‐dimer levels and calculate global arginine bioavailability (arginine/[ornithine+citrulline]) to assess endothelial function and activation in the new clusters and link these biomarkers to clinical outcomes. Results Clustering analysis identify 3 clear clusters in CC ‐ PH that are validated in PVDOMICS and outperform conventional classifications in predicting transplantation‐free survival. The phenotype associated with the worst survival is characterized by reduced lung diffusion capacity, decreased arginine bioavailability and nitrate levels, and elevated D‐dimer levels, consistent with loss of pulmonary microcirculation and endothelial dysfunction. Conclusions We identify new informative PH phenotypes associated with mortality and defined by biomarkers of endothelial function and activation. Loss of endothelial health and pronounced pulmonary vascular rarefication contribute more substantially to mortality across the spectrum of PH than right heart function. REGISTRATION: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT02980887.

Abstract Objectives Accurate and timely prediction of physiological deterioration in hospitalized patients is essential, but existing deep learning models, which rely solely on structured electronic health record (EHR) data, generate high false-positive rates by missing crucial contextual information in clinical notes. We present DETERIO-LLM, a hybrid prediction model that selectively integrates large language models (LLMs) with a deep learning model (DETERIO) to reclassify borderline-risk alerts using narrative clinical notes. Materials and Methods This retrospective study used a cohort of 1000 inpatients (4.6% deterioration prevalence). DETERIO-LLM selectively analyzed narrative notes for alerts in the uncertainty range (scores 2.5-3.5). Performance was evaluated using sensitivity, positive predictive value (PPV), and F1 score, benchmarked against comparators including the best-established ML model, eCART. Results DETERIO-LLM significantly improved predictive precision over all comparators. At its optimal threshold, the model achieved a PPV of 30.6% and an F1 score of 37.3%, outperforming eCART. This selective integration reduced false-positive alerts by 46.5% within the uncertainty range while maintaining comparable sensitivity. Discussion This study demonstrates a practical, interpretable method for integrating LLM-derived contextual knowledge into existing structured EHR prediction workflows. By focusing LLM analysis on high-uncertainty alerts, DETERIO-LLM substantially improves clinical utility through a marked reduction in false-positive cases, which is critical for mitigating alert fatigue. Conclusion Selective LLM integration with structured EHR models offers a viable, interpretable path to improving early warning systems in clinical care by leveraging narrative context to reduce alert fatigue while enhancing predictive performance.

Abstract This study examined the extent to which goodness of pronunciation (GoP) scores and phonological posterior probabilities capture perceptual ratings of speech severity in individuals with motor speech disorders (MSD). Speech recordings of the word catastrophe were obtained from 489 participants, including 333 neurologically typical controls and 156 individuals with MSD. GoP scores were derived using traditional acoustic features and self-supervised speech representations, including WavLM and XLS-R, across multiple modeling approaches, while phonological posterior probabilities were extracted using Phonet. Model performance was evaluated using Kendall’s rank correlations, regression, and receiver operating characteristic analyses against speech-language pathologists’ perceptual ratings of sound distortion and intelligibility. Both GoP and phonological posterior probabilities were significantly associated with perceptual ratings. Self-supervised speech representations substantially outperformed traditional acoustic features, with WavLM-based GoP using k-nearest neighbors achieving the strongest performance. Across correlation, regression, and classification analyses, GoP consistently outperformed phonological posterior probabilities for both sound distortion and intelligibility. Age and gender had minimal influence on model-derived measures or their relationships with perceptual ratings. These findings demonstrate the value of self-supervised GoP as an objective measure of speech impairment while highlighting the complementary role of phonological posterior probabilities in characterizing articulatory aspects of motor speech disorders.

Most genetic variants associated with complex traits are hypothesized to regulate gene expression. To understand the genetics underlying gene expression variability, we characterized 14,324 RNA-sequencing samples from the Trans-Omics for Precision Medicine program and performed expression and splicing quantitative trait locus (e/sQTL) analyses in six tissues and cell types, including whole blood ( n = 6454) and lung ( n = 1291). We detected tens of thousands of secondary cis-e/sQTLs, showing that secondary cis-e/sQTL discovery remains unsaturated. We fine-mapped UK Biobank–derived genome-wide association study (GWAS) signals from 164 traits and identified e/sQTL colocalizations for 10,611 GWAS signals, including 7096 that colocalize with secondary e/sQTLs. Our results suggest that even larger e/sQTL analyses will uncover additional secondary e/sQTLs, further benefiting GWAS interpretation.

OBJECTIVE To examine kidney disease outcomes associated with dipeptidyl peptidase 4 inhibitors (DPP-4i). RESEARCH DESIGN AND METHODS In this target trial emulation, claims data from enrollees in commercial, Medicare Advantage, and traditional Medicare health plans were used to identify adults ≥21 years old with type 2 diabetes, with moderate cardiovascular risk, and without stage 3 or worse chronic kidney disease (CKD) initiating alogliptin, linagliptin, saxagliptin, or sitagliptin between 2014 and 2021. The primary outcome was a composite of incident CKD stages 3–5, kidney failure, or kidney replacement therapy. Random treatment assignment was emulated using propensity scores, estimated using the super learner ensemble method and incorporated as inverse probability of treatment weights into proportional hazards models. RESULTS The weighted study cohort included 3,026 individuals starting alogliptin, 38,213 starting linagliptin, 14,934 starting saxagliptin, and 115,091 starting sitagliptin. One year after treatment initiation, the primary outcome occurred in 1.4%, 2.7%, 2.2%, and 2.4% of patients initiating alogliptin, linagliptin, saxagliptin, and sitagliptin, respectively, increasing to 3.2%, 6.2%, 5.0%, and 5.4%, respectively, at 3 years. Alogliptin was associated with lower risk of stage 3 or worse CKD than linagliptin (hazard ratio [HR] 0.50; 95% CI 0.37–0.69), saxagliptin (HR 0.63; 95% CI 0.45–0.86), and sitagliptin (HR 0.58; 95% CI 0.42–0.79), while saxagliptin was associated with lower risk than linagliptin (HR 0.80; 95% CI 0.73–0.87). Conversely, linagliptin was associated with higher risk than sitagliptin (HR 1.15, 95% CI 1.10–1.21). CONCLUSIONS Alogliptin was associated with lower risk of moderate-to-severe kidney disease than other DPP-4i, while linagliptin was associated with higher risk. Limitations include lack of laboratory data and potential unmeasured confounding.

Objectives To understand the trade-offs between different statistical modeling approaches, using real world data with small sub-populations, with rare exposures and increasingly rare outcomes. In particular, to compare adjusted regression, inverse probability weighting, and matching. Methods Data for these analyses came from the RADAR (N=1,134) and combined CNICS/JHHCC (N=14,434) cohorts. We estimated prevalence ratios (PRs) for self-reported use of specific substances comparing subpopulations (SP), SP-1 vs SP-3 and SP-2 vs SP-3, where SP-1 was the largest proportion (92% in RADAR, 81% in CNICS/JHHCC), SP2 was moderate proportion (18% in CNICS/JHHCC) and SP-3 was the smallest proportion (8% in RADAR, 1% in CNICS/JHHCC) of the population. We calculated PRs using 1) unadjusted relative risk regression (RR); and adjusted estimates controlling for age, race/ethnicity, study site, and year of interview using: 2) standard adjustment in RR; 3) stabilized inverse probability of treatment weighting (IPTW); and 4) matching with up to 3 matches from SP-1 or SP-2 per SP-3 participant. Results For most substances, all methods yielded consistent estimates. There were large weights in some of the IPTW analyses and in three cases these resulted in substantially divergent estimates. For the comparison between SP-1 and SP-3, the estimate for smoking was 1.3-fold greater in the matched analysis (PR=1.33, 95% CI: 1.02-1.75) than in IPTW (PR=1.03, 95% CI: 0.78-1.37 ATE and PR=1.05, 95% CI: 0.88-1.26 ATT). Even more extreme divergence in estimates was observed for differences between SP-2 and SP-3 with respect to methamphetamine/amphetamine, (IPTW-ATE: PR=1.51, 95% CI: 0.79-2.89; IPTW-ATT: PR=2.36, 95% CI: 1.36-4.12); vs Matching: PR=2.71, 95%CI: 1.47-4.99) and cocaine (IPTW-ATE: PR=1.15, 95% CI: 0.55-2.38; IPTW-ATT: PR=1.81, 95% CI: 1.03-3.18); vs Matching: PR=1.38, 95%CI: 0.76-2.53). Conclusion The combination of a rare exposure and a rare outcome can produce challenges for commonly used confounding adjustment strategies, and it is often best to compare different modeling approaches to gain greater insight.