Cancer Medicine

Background Emerging evidence suggests irreversible electroporation (IRE) with standard-of-care (SOC) chemo-therapy may improve survival in patients with Stage 3 pancreatic ductal adenocarcinoma (PDAC) when compared to SOC alone. This study evaluates the overall survival (OS) and progression-free survival (PFS) of Stage 3 PDAC patients treated with SOC plus IRE with the NanoKnife System versus SOC alone. Methods This prospective, multicenter study included two cohorts from the DIRECT registry: an IRE cohort from sites offering IRE as part of clinical care, and a comparator SOC cohort of prospectively enrolled and contemporaneous retrospective patients. Enrollment spanned 08/05/2019 to 02/05/2023, with follow-up through at least 24 months, death, or loss to follow-up. Included were 137 patients (99 IRE; 38 SOC), aged [&ge;]18 years with Stage 3 PDAC and no progression after three months of SOC therapy. Results Median (interquartile range) time from diagnosis to enrollment was 8 (6-10) months for IRE and 4 (3-6) for SOC (p<0.0001). Median OS and PSF from enrollment were 18 (95% confidence interval [CI]: 15-24) months and 9 (95% CI: 7-12) months for IRE, and 10 (95% CI: 8-14) months and 6 (5-8) months for SOC, respectively (p<0.0001 and p=0.009). Adverse events occurred in 80% (79/99) of IRE patients and 95% (36/38) of SOC patients; 29% (29/99) of the IRE cohort experiencing an IRE-related adverse event. Conclusions IRE was associated with improved OS versus SOC alone and may be an effective consolidative treatment for Stage 3 PDAC after three months of induction chemotherapy.

BackgroundThe MDM2-p53 signaling pathway plays a central role in tumor suppression, and genetic variants that disrupt this pathway may influence breast cancer (BC) susceptibility. However, data from South Asian populations, particularly Bangladesh, remain limited. MethodsA case-control study was conducted in Bangladeshi women, including BC patients and healthy controls (HCs). Genotyping of MDM2 polymorphisms was performed using PCR-based methods. Circulating MDM2 and p53 protein levels were measured using enzyme-linked immunosorbent assays (ELISA). Associations between genotype, protein levels, BC status, and clinicopathological features were evaluated using appropriate statistical models. ResultsA strong and genotype-specific association was observed for MDM2 rs2279744. Women carrying the heterozygous TG genotype had a markedly increased risk of BC across additive, dominant, and over-dominant models, whereas the GG genotype showed a protective effect under the recessive model. In contrast, rs937282 did not show a significant association with BC risk. Circulating MDM2 levels were significantly elevated in patients compared with controls and varied by rs2279744 genotype, while circulating p53 levels showed an opposite trend. A strong inverse correlation was observed between serum MDM2 and p53 levels, supporting dysregulation of the MDM2-p53 feedback loop. Elevated MDM2 levels were also noted in HER2-positive and triple-positive BC subtypes. ConclusionTogether, these findings indicate that the MDM2 rs2279744 polymorphism contributes to BC susceptibility in a genotype-specific manner, likely through disruption of the MDM2-p53 regulatory balance. However, the absence of functional validation limits direct causal inference.

BackgroundFor pancreatic cancer, practical blood-based tests for early detection and postoperative surveillance remain elusive. We sought to develop a qPCR-measurable serum microRNA (miRNA) panel that robustly discriminates pancreatic cancer from non-cancer controls and other malignancies. MethodsWe profiled 255 serum miRNAs in batch 1 (n=72) and selected 27 candidates. Candidates were refined in batch 2 (n=552) and cross-batch evaluation was performed with batch 3 (n=391) to derive a miRNA model. Independent validation used batch 4 (n=616). Clinical relevance was assessed in an independent clinical cohort of resection patients with samples obtained preoperatively and at 1 and 12 months postoperatively. ResultsThe miRNA model trained on batches 2 and 3 achieved an area under the curve (AUC) of 0.91 and 0.83 for pancreatic cancer versus non-cancer controls and non-cancer plus other cancers, respectively, when independently validated in batch 4. Stage-wise AUCs in batch 4 were 0.91 (I), 0.94 (II), 0.86 (III) and 0.90 (IV). In the clinical batch, the score decreased postoperatively (preoperative vs month 1; p<0.01) and was higher in recurrence than non-recurrence (p<0.001). ConclusionsThe developed compact miRNA qPCR assay discriminated pancreatic cancer across independent assay batches and showed clinical relevance for postoperative surveillance. Clinical Trial RegistrationNot applicable.

Background. The European Society for Medical Oncology (ESMO) Scale for Clinical Actionability of molecular Targets (ESCAT) ranks genomic alterations by the evidence supporting the predictive value of the molecular target for response to targeted therapies. No openly available, systematically curated set of standard care biomarkers mapped to the ESCAT framework exists to support clinical decision-making or harmonize biomarker interpretation. Methods. We mapped all OncoKBTM Level 1 biomarkers to ESCAT tiers using evidence cited by OncoKBTM, excluding abstract-only data. Eight board-certified oncologists and hematologists independently assigned ESCAT tiers, with discrepancies resolved through structured consensus meetings. Recurring evidence scenarios that did not correspond to any existing ESCAT tier informed a set of a priori defined modifications, which were subsequently applied to biomarkers that could not be classified using native ESCAT criteria. Results. Of 188 OncoKBTM Level 1 biomarkers, 16 were excluded due to abstract-only evidence. Using native ESCAT criteria, 51% of the remaining biomarkers were classified as Tier 1, 3% Tier 2, 18% Tier 3, 6% Tier X and 22% could not be assigned to any tier. Applying the modified ESCAT criteria resolved all previously unclassifiable biomarkers and increased Tier 1 assignments to 73%. Inter-rater reliability (Krippendorffs alpha) was moderate (0.586) and 62% of classifications required consensus discussions. Comparison with ESCAT tiers reported in ESMO Clinical Practice Guidelines showed improved concordance when using the modified criteria. Conclusions. The native ESCAT criteria are highly stringent, resulting in many FDA-recognized, clinically validated biomarkers that are currently assigned level 1 by OncoKBTM not mapping to any existing tier. Our predefined modifications improved alignment with OncoKBTM Level 1 designations and with published ESMO clinical practice guidelines. The mapped set of standard care biomarkers are provided on the OncoKBTM website, offering a practical resource that harmonizes ESCAT tiers of evidence with a widely adopted levels of evidence schema.

Background Complex gastrointestinal (GI) oncologic surgeries carry substantial perioperative risk, and nationwide outcomes in low- and middle-income countries (LMICs) are underreported. This study aimed to evaluate national trends in surgical volume, in-hospital mortality, and intensive care unit (ICU) utilization for major GI cancer surgery in Brazils Unified Health System (SUS) over a 14-year period. Methods A population-based analysis was performed using national administrative databases to identify all adult patients undergoing colectomy, gastrectomy, pancreatic resection or esophagectomy for cancer in the SUS from 2010-2023. Annual rates were age-standardized according to the WHO standard population. Temporal trends were assessed using Poisson regression to estimate average annual percent change (AAPC) with 95% confidence intervals (CIs). Results A total of 179,337 hospital admissions were analyzed (median age 63 years; 48% female). Colectomies accounted for 72% of cases, followed by gastrectomies (19%), pancreatic resections (5%), and esophagectomies (3%). Although crude surgical volume increased, population-adjusted rates declined overall (AAPC -2.09%; 95% CI -2.58 to -1.59), mainly due to reductions in gastrectomies and esophagectomies. Median hospital stay decreased from 9 to 7 days (AAPC -1.93%; 95% CI -2.79 to -1.06). Overall in-hospital mortality declined from 8.1% to 5.7% (AAPC -2.88%; 95% CI -4.15 to -1.59). ICU utilization rose from 37% to 43% of admissions (AAPC +1.31%; 95% CI 0.91 to 1.71). Conclusion Over 14 years, in-hospital mortality and length of stay for major gastrointestinal cancer surgery declined within Brazils universal public health system. These temporal trends occurred alongside expansion of accredited oncology services and increased ICU utilization, although causal relationships cannot be established from administrative data. These findings should be interpreted as hypothesis-generating and highlight the need for more granular hospital-level data in LMIC settings.

Abstract Introduction: MET tyrosine kinase (TKI) therapy has improved outcomes in patients with non-small cell lung cancer (NSCLC) harboring MET alterations. However, primary and acquired resistance ultimately limits durability of response. This study evaluated the safety and efficacy of the MET inhibitor capmatinib with the MEK inhibitor trametinib in patients with metastatic MET-driven NSCLC who had progressed on prior treatment with at least one MET inhibitor. Methods: A multicenter phase I study evaluated capmatinib in combination with trametinib in patients with advanced stage NSCLC harboring activating MET alterations and prior exposure to at least one MET TKI. A 3+3 dose-escalation design was employed to assess safety and tolerability of the combination. Results: Three patients (n = 3) were enrolled in the study and completed a median of 3 cycles of therapy. Dose-limiting toxicities, including rash, edema, and nausea, necessitated dose reductions in the first two patients and initiation of the third patient at a lower dose level. Ultimately, all patients discontinued therapy due to treatment-related adverse events. The study was terminated early due to poor accrual and TRAEs. No radiographic objective responses were observed. Conclusions: In this phase I trial, capmatinib plus trametinib was associated with significant treatment-related adverse events and treatment was discontinued in all participants. Based on these findings, further investigation of this combination of MET and MEK inhibitors is not recommended.

BackgroundHomozygous deletion of the methylthioadenosine phosphorylase (MTAP) gene is a frequent genetic alteration in cancer. MTAP, which creates adenine from 5-methylthioadenosine (MTA), is constitutively expressed in all tissues throughout the body. Previously, we described a novel strategy to specifically target MTAP-deleted cancer cells by combining the antipurine prodrug 2-fluoroadenine (2FA) with MTA. In vitro, this combination efficiently killed MTAP- cancer cells, but in vivo the combination was much less effective in vivo. Here, we explored the role of xanthine oxidase (XO) in this process. Materials and MethodsVarious combinations of 2FA, MTA, and the xanthine oxidase inhibitor febuxostat (FX) were tested in various cancer cell lines grown in vitro and in mice. LC-MS/MS was used to examine the levels and ratio of intracellular 2-FA-containing nucleotides compared to adenine-containing nucleotides. Results and conclusionsThe treatment of cells with 2FA+MTA in vitro resulted in much higher 2FANP/ANP ratios than the same treatment in vivo. The addition of XO to culture media in vitro effectively abolished the killing by 2FA, and this effect was fully reversed by the addition of febuxostat (FX), a xanthine oxidase inhibitor. In vivo, the addition of FX to 2FA results in increased cell killing and toxicity and a 1000% increase in the amount of 2FA converted to 2-FA-monophosphate (2FAMP). Xenograft studies using MTAP- HT1080 and MiaPaCa-2 cell lines have shown that a 2FA/MTA/FX cocktail can cause tumor regression in vivo. These studies suggest that the combination of 2FA/MTA/FX should be explored as a treatment for MTAP- cancer.

Background: The All of Us Research Program represents a rich resource for cancer epidemiology research, with over 400,000 participants with whole genome sequences linked to electronic health records (EHR). Large cancer datasets often focus exclusively on cases without controls and neglect pre-diagnosis healthcare occurrences. Here, we perform a phenome-wide association study (PheWAS) of EHR data at least 1 year pre-diagnosis between cancer cases and matched controls, revealing co-occurring and mutually exclusive phenotypes. Methods: We identified 55,000+ cancer cases across 21 cancer types in All of Us version 8. To eliminate age-related confounding, we implemented a two-stage matching and censoring strategy: loose matching on demographics to establish index dates and cohort comparability, followed by right-censoring of EHR data (excluding 1 year pre-diagnosis/index), then 1:2 matching to address residual demographic imbalance. We tested associations between 23,193 cancer cases, 46,386 matched controls and approximately 1,600 clinical phenotypes using logistic regression adjusted for sex at birth, self-reported race, age at diagnosis/index date, and two censored EHR metrics: observation window and unique condition count, with Bonferroni correction for multiple testing. Results: Our analysis identified 232 significantly associated phenotypes, confirming established cancer risk factors including elevated prostate specific antigen (OR = 2.92, 95% CI: 2.65-3.23; p-value=1.8x10-101) and multinodular goiter (OR = 1.73, 95% CI: 1.56-1.91; p-value=6.7x10-27). Further investigation into the relationship between several phenotypes with seeming inverse effects is warranted. Conclusions: This PheWAS of EHR data at least 1 year pre-diagnosis leveraged the diversity of All of Us to examine how clinical phenotypes prior to cancer diagnosis vary across cancer types and racial groups. Our findings validate All of Us as a robust platform for cancer epidemiology research, confirming established risk factors at scale across diverse populations. This work provides methodological insights for EHR-based susceptibility analyses and demonstrates the value of agnostic phenome-wide approaches for generating hypotheses in precision medicine.

Purpose: Pathogenic or likely pathogenic (P/LP) variants are increasingly identified in genes more commonly associated with adult-onset cancer predisposition, but their prevalence and relevance to children who present with cancer remain unclear. Methods: We retrospectively analyzed 1,280 consecutive pediatric patients with cancer who underwent clinical germline sequencing, using a virtual panel, from 2021 to 2024. Genes with P/LP variants were categorized as aoCPG or pediatric-onset cancer predisposition genes (poCPG) according to cancer risk before age 18 years and pediatric surveillance recommendations. Variant relevance was adjudicated using tumor diagnosis/histopathology, immunohistochemistry, and tumor molecular features and classified as primary, secondary, or indeterminate. Results: Among 1,280 patients, 197 (15.4%) harbored 211 P/LP variants across 54 genes. Sixty-six variants (31.3%) occurred in aoCPG, 87 (41.2%) in poCPG, and 58 (27.5%) were heterozygous variants in autosomal recessive genes. Among adult-onset variants, 7 (10.6%) were primary, 54 (81.8%) secondary, and 5 (7.6%) indeterminate. Among pediatric-onset variants, 77 (88.5%) were primary and 10 (11.5%) secondary. Six patients (3 adult-onset variants; 3 pediatric-onset variants) received targeted therapy informed by germline/somatic sequencing results. Conclusion: In pediatric oncology, most variants in aoCPG are secondary rather than tumor-related findings. Tumor-informed interpretation, beyond variant classification, may improve reporting, counseling, and therapeutic decision-making

PurposeGlasdegib is a Sonic Hedgehog (SHH) pathway inhibitor used for treating newly diagnosed acute myeloid leukemia in elders or patients unfit for intensive chemotherapy. This study sought to demonstrate growth inhibition and increased apoptosis of B-cell acute lymphoblastic leukemia (B-ALL) in vitro under glasdegib, alone and combined with inotuzumab, using a novel co-culture system and validated chemosensitivity testing model to determine whether glasdegib with and without inotuzumab may represent a promising treatment strategy in B-ALL. MethodsSeven blood and marrow samples from B-ALL patients were co-cultured with HS-5 stromal cells in a co-culturing system designed to mimic the tumor microenvironment to maintain B-ALL cell viability for chemosensitivity testing under glasdegib and inotuzumab. ResultsCo-culturing improved B-ALL viability from four to nine days. Dosage-dependent responses to glasdegib were consistent among B-ALL samples on day four based on culture viability, and varied based on expressions of SSH genes GLI1, GLI3, SMO, and PTCH1. Combination with inotuzumab had varied effects on treatment response. ConclusionCo-culturing B-ALL cells with HS-5 stromal cells improves B-ALL growth and viability. Glasdegib with and without inotuzumab treatments impact the viability of co-cultured B-ALL cells by day four. SHH gene expressions suggest different B-ALL patients may be sensitive or resistant to glasdegib and inotuzumab.

Background: Self-reported confidence in health information seeking does not reliably predict accurate health knowledge, yet the population-level distribution of this discordance and its demographic predictors have received limited direct study. This study aimed to identify and characterize a Confident-Incorrect phenotype among U.S. adults: individuals with high perceived health information competence who simultaneously hold inaccurate or fatalistic beliefs about cancer. Methods: Cross-sectional analysis of HINTS 7 (N = 7,278). A Confidence Index (3-item digital literacy composite (Cronbach's = 0.674) and an Evidence-Consistent Knowledge Score (factual cancer knowledge minus a cancer fatalism composite; fatalism subscale = 0.563) were computed and combined into a discordance framework. Median-split classification produced four phenotypes. Gaussian Mixture Model clustering with four components provided moderate independent validation (inter-method agreement = 65.2%). Survey-weighted multinomial logistic regression (n = 5,771; McFadden pseudo-R2 = 0.129) examined phenotype predictors. Results: An estimated 20.3% of U.S. adults were classified as Confident-Incorrect. They reported confidence levels similar to Well-Informed adults (z = 0.72 vs. 0.82) but scored 2.8-fold lower on objective cancer knowledge (0.74 vs. 2.06 out of 4) and exhibited the highest cancer fatalism of any phenotype (3.17 vs. 1.65 out of 4). Only 14.3% correctly identified alcohol as a cancer risk factor (vs. 58.8% of Well-Informed adults). Cancer screening rates did not differ meaningfully across phenotypes. Lower education (OR = 0.754), Hispanic ethnicity (OR = 1.788), non-Hispanic Black race (OR = 1.893), higher social media use (OR = 1.097), and lower trust in scientists (OR = 0.749) independently predicted Confident-Incorrect membership. Conclusions: An estimated one in five U.S. adults is overconfident in health information competence while holding substantially inaccurate beliefs about cancer prevention. Cancer screening rates did not follow the expected gradient across phenotypes, a null finding that cautions against inferring immediate behavioral impact from observed belief gaps. Interventions targeting specific factual errors and cancer fatalism are more likely to reach this group than general health literacy programs.

BackgroundBreast cancer stem cells (BCSCs) contribute significantly to breast cancer (BC) mortality among women globally. It underpins tumor heterogeneity in BC by driving variations in stemness potential and altering immune microenvironment. However, how BCSCs, subpopulations of breast cancer cells from distinct molecular subtypes differentially modulate CD8 T-cell exhaustion and immune dysfunction remain unclear. MethodsWe conducted our study from patients with BC of four subtypes. MACS sorted (Lin-CD44+CD24-) BCSCs were prepared for mammosphere formation assay from mastectomies samples. Flow-cytometry was used to analyze breast cancer stem cells (BCSCs). Immunofluorescence, immunohistochemistry, Real Time and Reverse Transcriptase PCR array, Chromatin-immunoprecipitation assay, Transwell, ELISA, Western blotting, Cloning, Transfection, Knockdown, chromatin immunoprecipitation approaches were used to investigate the underlying mechanisms. ResultsHere, we report that BCSCs actively participate in tumor progression by modulating effector CD8 T-cells. Triple-negative breast cancer (TNBC), being the subtype with the most adverse outcomes, sustains the enrichment of stem cell regulating transcription factors like NANOG, OCT4 and SOX2 compared to HER2, Luminal B, and Luminal A subtypes. Tumor from TNBC patients exhibited an exhausted phenotype within CD8 T-cell infiltrates with PD1high TIM3high LAG3 high IFN{gamma}low signature. BCSCs induced increased proportion of exhausted CD8 T-cells, predominantly in the TNBC subtype. Cell-surface Notch1 expression was upregulated in BCSCs across all molecular BC subtypes, with the highest elevation observed in TNBC. Knockdown or inhibition of Notch1 downregulated stemness-associated genes and diminished CSC-mediated induction of CD8 T-cell exhaustion. Cumulatively, these findings suggest that assessment of high Notch1 and Nanog frequency within BCSCs can guide Notch1-targeted therapies and may formulate for new combinatorial treatment strategies to improve patient outcomes. Additionally, therapeutic targeting of BCSC-intrinsic NOTCH1-NANOG/NOTCH1-PD-L1 axis could represent an effective strategy to reduce stemness programs and alter BCSC-driven CD8 T-cell exhaustion, majorly in aggressive subtypes such as TNBC. ConclusionsBCSCs aggressiveness is perpetuated through Notch1-mediated axis. Targeting Notch1 would reduce stemness (majorly NANOG), survival, as well as prevent CD8 T-cell exhaustion (upregulating PD-1, TIM3, LAG3), thereby weakening tumor progression.

Importance Prognostic tools beyond staging are needed to guide treatment and counseling in head and neck squamous cell carcinoma (HNSCC). Objective To develop and externally validate a machine learning model predicting survival in advanced HNSCC using routinely collected clinical and biomarker data. Design, Setting, and Participants Retrospective, multi-institutional cohort study including 2,385 patients with stage III-IV HNSCC diagnosed from 2012-2022 in the University of California Health Data Warehouse (UCHDW). Patients were randomly split into training (n = 1,908) and test (n = 477) sets. Partial external validation used 7,749 patients from the Surveillance, Epidemiology, and End Results (SEER) registry (2010-2020). Exposures Demographic, tumor, treatment, comorbidity, and biomarker variables recorded at or before diagnosis. Main Outcomes and Measures The primary outcome was all-cause mortality within 70 months. Cox proportional hazards models included all predictors. Discrimination was assessed with Harrell's concordance index (C-index), calibration with predicted vs observed survival, and stratification with Kaplan-Meier curves. A Random Survival Forest (RSF) was trained for benchmarking and interpretability using Shapley Additive exPlanations (SHAP). Results Among 2,385 patients in UCHDW (median age, 63 years; 29.0% mortality), the Cox model achieved a C-index of 0.735 in the internal test set. Risk quartiles showed clear separation on Kaplan-Meier curves (log-rank p < 0.0001). In the SEER cohort (n = 7,749), where only demographic, staging, subsite, and treatment variables were available, the reduced Cox model achieved a C-index of 0.688, with calibration showing modest underestimation of survival in high-risk groups. Age, T stage, Charlson Comorbidity Index, neutrophil-to-lymphocyte ratio, and platelet count were among the strongest predictors, while surgery was associated with improved survival. The RSF achieved a C-index of 0.758 internally, with SHAP highlighting nonlinear effects of albumin, BMI, and inflammatory markers. Conclusions and Relevance A machine learning model using routine clinical and biomarker data demonstrated good prognostic performance in advanced HNSCC, with partial external validation. Such approaches may support individualized survival estimates, risk stratification, and treatment discussions, but broader validation is required before clinical adoption.

Background Pancreatic Ductal Adenocarcinoma will be the 2nd-most common cause of cancer mortality by 2030. It is associated with rapid deterioration, severe symptoms, and significant quality-of-life concerns. Using input from patients, family caregivers (FCGs), and provider stakeholders, we designed an intervention, PAL-CHW-PDAC, delivered by a community health worker that involves proactive symptom monitoring and management, care navigation, and disease education. Methods We conducted a pilot randomized controlled trial of 60 patients with newly diagnosed PDAC (within 2 weeks of diagnosis) and their caregivers at Loma Linda University Health from 09/2025 to 05/2026. Patients were randomized 1:1 to receive the PAL-CHW-PDAC intervention (6 CHW visits over 3 months) or an attention control. The control comparator involved receiving standard handouts and videos on pancreatic cancer, along with check-in visits with research staff. The primary outcome was symptom burden, defined using the NCCN/FACT Hepatobiliary Symptom Index. Secondary outcomes included quality of life (QoL) measured by the FACT-Hep and psychological distress (measured by the NCCN-Distress Thermometer). Caregiver outcomes included burden, preparedness, quality of life, and psychological distress. Results: 60 out of 74 eligible (81%) were enrolled. The median age was 71, 60% of patients were Hispanic. 68% of patients presented with metastatic PDAC, 23% with borderline resectable disease and 9% with resectable PDAC. There was a trend towards improved symptom burden at 12 weeks (mean increase of 5.3 points vs. decrease of 3.2 points; p=0.093) with the intervention compared to the attention control. The intervention group also had improved psychological distress at 12 weeks (3.31 vs. 5.95, p=0.01), caregiver psychological distress (3.26 vs. 6.86, p<0.001) and caregiver preparedness (2.92 vs. 2.11) at 12 weeks. Telehealth utilization for symptom-focused visits improved with the intervention (82%) compared to the control. (14%, p=0.01) Hospice utilization also improved with the intervention (41% vs 7%, p-0.12). Conclusions: A pilot RCT of the PAL-CHW-PDAC intervention demonstrated preliminary efficacy with a trend towards improved symptom burden, psychological distress, and caregiver psychological distress and preparedness. A larger definitive clinical trial is needed to understand the impact of this promising intervention. ClinicalTrials.gov number, NCT07591571

Importance: While gut dysbiosis is known to impair response to immune checkpoint inhibitors (ICIs), the relative clinical impact of antibiotic timing (pre- vs. post-ICI initiation) remains unclear. Objective: To evaluate whether antibiotic timing differentially influences overall survival (OS) in a large, multi-institutional pan-cancer cohort. Design, Setting, and Participants: This retrospective cohort study utilized deidentified electronic health record data from six academic medical centers within the University of California Health system. We included 21,108 adults with any malignancy who received PD-1, PD-L1, or CTLA-4 inhibitors between January 2014 and December 2024. Exposures: Antibiotic exposure windows were categorized as pre-only (-60 to -1 days), post-only (+1 to +60 days), both windows, or none. Main Outcomes and Measures: The primary outcome was overall survival (OS) calculated from the first ICI dose. Multivariable Cox proportional hazards models adjusted for demographics, tumor type, line of therapy, and baseline health indicators (albumin, NLR, and recent hospitalization). Results: Among 21,108 patients, 17.3% had pre-only exposure, 13.3% had post-only exposure, and 60.6% had no exposure. In the multivariable model, post-only exposure (HR, 1.27; 95% CI, 1.20-1.35) and combined pre- and post- exposure (HR, 1.31; 95% CI, 1.23-1.40) were significantly associated with higher mortality. Pre-only exposure was not significantly associated with OS (HR, 1.04; 95% CI, 0.99-1.10). Subgroup analyses by tumor type showed consistent trends across major malignancies, including head and neck (Post HR, 1.46) and renal cell carcinoma (Post HR, 1.26). Conclusions and Relevance: In contrast to some smaller studies, this large-scale analysis indicates that antibiotic exposure after ICI initiation carries a greater risk than exposure prior to treatment. These findings highlight the need for rigorous antibiotic stewardship strategies specifically during the early phases of immunotherapy treatment.

Background: Preparing tumor board patient summaries is time intensive. Large-language-model based systems may automate summarization but require real-world evaluation prior to clinical use. We performed an exploratory retrospective evaluation of the Microsoft Healthcare Agent Orchestrator (HAO), deployed in a Mayo Clinic controlled staged environment, to generate tumor board-style patient summaries from retrospective Electronic Health Record (EHR) notes. Methods: HAO generated summaries for breast, hepatobiliary, and neuro-oncology tumor board cases using up to the most recent 1,000 clinical notes. Clinician reviewers evaluated outputs via REDCap surveys across perceived factuality, completeness, clarity/conciseness, temporal cohesion, comparative performance, safety, and clinical utility (0-4 Likert scale). Reviewers were permitted to query the HAO chat interface to address missing details. Automated factuality was assessed using TBFact (bidirectional entailment), reporting precision and recall against available reference summaries. Results: Among 57 survey responses from 5 different physicians, mean scores exceeded 2.8 across domains, with medians of 3 for most axes. In an exploratory comparison, oncology fellows required less time to review HAO-generated summaries than to manually generate patient summaries (mean difference 13.57 minutes per patient, p<0.001), although this difference may be influenced by prior familiarity with the same cases; 96% of survey responses indicated that HAO would save time. TBFact evaluations showed higher recall than precision across domains, consistent with broad capture of reference content alongside additional content that was not present in gold-standard summaries. Attribution was viewed favorably but showed issues with primary-source specificity and link reliability. Conclusions: In a controlled Mayo environment, HAO demonstrated moderate performance and was associated with reduced review time for tumor board preparation. These findings are promising but preliminary and do not establish clinical safety, noninferiority to manual review, or readiness for routine clinical use. Limitations, including verbosity, specialty-specific content gaps, and inconsistent attribution, highlight the need for iterative refinement and further evaluation.

Importance BRCA genetic testing is critical for cancer risk assessment, treatment and personalization, yet substantial underutilization persists. Socioeconomic and clinical factors may strongly influence testing uptake; therefore, identifying the potential drivers to BRCA testing and treatment is essential for addressing gaps in access, increasing retention into care, and improving cancer outcomes. Objective To quantify the putatively causal effects of SDoH on BRCA genetic testing among individuals with breast, ovarian, pancreatic, and prostate cancers and to develop a predictive model to identify patients at risk for underuse of testing. Design, Setting, and Participants This observational case-control study used data from a large multistate clinical research data network covering southern US (2012-2023). The network contained records of more than 26 million individuals and was linked with ZIP code-level SDoH variables derived from national socioeconomic datasets. Adults diagnosed with breast, ovarian, pancreatic, or prostate cancer were eligible for cases (received BRCA testing) or controls (did not receive BRCA testing, matched by cancer diagnosis). Exposure SDoH categories, including economic conditions, education, healthcare access, neighborhood conditions, and social connectedness. Main Outcomes and Measures The primary outcome was receipt of BRCA genetic testing after cancer diagnosis. Results Among 3,279 people diagnosed with cancer, 748 received BRCA testing and 2,531 served as controls. Study population mean [SD] age was 66.8 [15.7] years; 1,758 were women [53.6%], 2,238 [69.6%] were White and 616 [18.8%] were Black or African American. Breast (1,420 [42.8%]) and prostate (1,342 [40.9%]) cancers were the most common diagnoses, followed by pancreatic (242 [7.4%]), ovarian (238 [7.2%]), and multiple cancers (55 [1.7%]). Upon adjusting for potential confounding, higher educational attainment (odds ratio [OR], 1.19), public-sector employment (OR, 1.42), neighborhood safety (OR, 1.28), and social participation (OR, 1.72) showed an increased likelihood of undergoing BRCA testing, whereas economic instability, including housing cost burden and reliance on public insurance, had an effect of reduced testing. A random forest classifier demonstrated good discriminative performance (AUROC, 0.776) to predict cancer patients who were likely to take BRCA testing, where nativity, language, and residential stability ranked among the most influential social determinants according to SHapley Additive exPlanations (SHAP) analysis. Conclusions and Relevance In this observational case-control study, SDoHs were strongly associated with receipt of BRCA genetic testing among people with cancer. These findings suggest that disparities in genetic testing may reflect structural and social barriers rather than differences in clinical eligibility alone. Efforts to improve equitable access to genetic testing may benefit from integrating social-context information into clinical workflows and targeting outreach or navigation strategies toward socially disadvantaged populations.

BackgroundObesity significantly increases the risk of prognosis and clinical outcomes in pancreatic ductal adenocarcinoma (PDAC). While research on the interactions between obesity and the tumor microenvironment (TME) is mostly confined to a few interactions at a time, leaving a gap in the comprehensive understanding of obesity-driven PDAC. We set out to develop a cell-type-resolved model of obesity-driven PDAC using bulk transcriptomic data to investigate TME changes. MethodsWe conducted an integrated transcriptomic analysis of PDAC patients from the CPTAC-3 cohort (n=140) stratified by BMI. A custom immune and stromal functional gene signature database covering 65 cell types was constructed, followed by LLM-assisted review, overlap control, and validation. BayesPrism deconvolution using matched single-cell references was used to derive expression profiles for each cell type. Stabl, a machine-learning algorithm, was used to identify BMI-associated signatures. Bayesian hierarchical modeling, using both continuous and categorical BMI change, was applied to estimate effect sizes and assess the statistical credibility of the signature changes using the 95% Highest Density Interval (HDI) excluding zero. Virtual multiplex immunofluorescence was generated from whole-slide H&E images using gigaTIME to assess the spatial manifestation of BMI-associated TME changes in tissue ResultsBulk pathway analysis showed that ECM homeostasis and primary immunodeficiency pathways deteriorated with increasing BMI. However, Bayesian modeling revealed cell-type-specific, non-linear dynamics. Stromal populations in overweight (OW) individuals were altered, with changes in ECM synthesis and inflammatory signaling that stabilized rather than intensified during obesity. Immune compartments also showed diverse trajectories: CD4+ T cells remained functional in OW but collapsed in obesity; CD8+ T cells progressed linearly from activation to chronic exhaustion. NK cells exhibited non-monotonic behavior, and monocyte and B cell lineages became impaired prior to clinical obesity. Cell-cell interaction analysis showed a shift from a T cell and dendritic cell-centric adaptive interactome in normal weight patients to a neutrophil-dominated inflammatory network in OW. Spatial analysis showed stromal-trapped CD8+ T cells were compressed closer to the tumor boundary with rising BMI. ConclusionsOverweight status represents a critical tipping point in tumor microenvironmental reprogramming, challenging linear models of obesity-associated immune modulation and suggesting that early metabolic interventions may prevent PDAC functional deterioration. Model is available at https://obese-pdac-model.streamlit.app/ O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=138 SRC="FIGDIR/small/721695v1_ufig1.gif" ALT="Figure 1"> View larger version (36K): org.highwire.dtl.DTLVardef@b1c8cdorg.highwire.dtl.DTLVardef@1f61b7forg.highwire.dtl.DTLVardef@876c60org.highwire.dtl.DTLVardef@dc32b2_HPS_FORMAT_FIGEXP M_FIG C_FIG

Introduction Coding tumor diagnoses from free-text clinical documentation currently requires substantial manual effort. Promising approaches for automating this process include large language mod-els (LLMs), embedding models, and retrieval-augmented generation (RAG). While previous studies often focus on a single method, we directly compare these approaches on a real-world dataset of tumor diagnosis descriptions to assess their strengths and limitations. Methods We evaluated nine different embedding models using similarity search and embedding-based classification, as well as LLM-based coding, with and without RAG, on a real-world dataset of 2,024 unique German tumor diagnosis descriptions labeled with ICD-10 and ICD-O topography codes. The retrieval knowledge base was constructed exclusively from stand-ardized Alpha-ID, ICD-10-GM, and ICD-O-3 classifications. Performance was assessed for exact (full-code) and partial (three-character) code prediction. For RAG, we evaluated base and fine-tuned versions of Llama 3.1 8B and Llama 3.3 70B. Results Qwen3-Embedding-8B, the largest embedding model, yielded the best results. It achieved 47.8% exact-match and 72.1% partial-match accuracy for ICD-10 coding with classification, and 42.7% exact-match and 73.5% partial-match accuracy for ICD-O coding with similarity search. The other embedding models, including medically specialized ones, showed varied but lower performance. RAG improved base LLM perfor-mance and outperformed embedding-based approaches on partial-match accura-cy (80.6% partial-match accuracy for ICD-10 and 75.0% for ICD-O with Llama 3.3 70B), but not on exact-match accuracy. Conclusion A direct comparison with embedding-based approaches is essential to determine whether the additional effort of RAG is justified. The strong variation in performance also highlights the importance of model selection. Further advances in embedding-based methods, potential-ly supported by larger and more diverse training data, may offer a promising direction for future work.

BackgroundATR activation following DNA damage from cancer treatments such as radiation can mitigate anticancer efficacy, making ATR inhibitors (ATRi) an attractive therapeutic. In vivo and in vitro studies have shown enhanced tumor cell radiosensitivity with the ATRi ceralasertib, elimusertib, and berzosertib, however, the potentiating effect of ATRi on ionizing radiation (IR) through immune-based mechanisms has only been studied with ceralasertib. MethodsWe aimed to determine if antitumor immune responses observed with ceralasertib in combination with IR extend to the other ATRi class members in the preclinical CT26 mouse model. We also examined the relationship between exposure and immune stimulation, efficacy and survival outcomes of each ATRi when combined with IR. ResultsCeralasertib and elimusertib, not berzosertib, synergized with IR in a dose and schedule-dependent manner to modify tumor antigen-specific CD8+ T cell populations in the draining lymph node. Transient ATRi therapy, combined with IR, enhances antitumor efficacy, promoted tumor shrinkage, and increased survival. ATRi elicited differential inflammatory gene induction and dose-dependent unique cytotoxicity profiles in vitro. ConclusionThe immune mediated antitumor effect of ATRi combined with radiation is dose and schedule dependent, and while likely a class effect, may differ between ATRi compounds.