Browse Articles

Characterization of ENTPD1/CD39-mediated myeloid-T-cell crosstalk and its role in immunotherapy response in triple-negative breast cancer.

e14564 Background: Immunotherapy efficacy in triple-negative breast cancer (TNBC) is limited by mechanisms of immune suppression within the tumor microenvironment (TME) . Ectonucleoside triphosphate diphosphohydrolase 1 (ENTPD1, encoding CD39) has been implicated in immune evasion, yet its precise regulatory role in orchestrating intercellular crosstalk and driving resistance to PD-L1 blockade in TNBC remains unclear . Methods: We analyzed single-cell RNA sequencing (scRNA-seq) data from 167,166 immune cells from 27 tumor and metastatic samples across 14 TNBC patients treated with paclitaxel $\pm$ atezolizumab . Quantitative metrics—treatment contribution (TC), ENTPD1 dynamic index (EDI), and tumor response index (TRI)—were developed to capture dynamic immune responses . Findings were validated in an independent cohort of 112 TNBC patients via immunohistochemical (IHC) assessment of CD39 expression and survival analysis . Results: ENTPD1 was predominantly expressed in M2-like macrophages and a subset of regulatory T cells (CD4+CD39+ Tregs) . ENTPD1+ macrophages exhibited M2-like polarization and suppression of STING-type I interferon signaling . Cell-cell communication analysis identified a previously unrecognized lymphotoxin (LT)-centered axis (LTA-TNFRSF1B/14) linking CD4+CD39+ T cells with M2 macrophages, bridging adenosine metabolism with cytokine-driven immunoregulation . LT signaling was selectively enriched in non-responders and coincided with enhanced M2 macrophage interactions . In the validation cohort, high CD39 protein expression significantly correlated with reduced overall survival (OS) ($P < 0.001$) . Responders to PD-L1 blockade showed marked ENTPD1 downregulation post-treatment, whereas non-responders maintained high levels . Conclusions: ENTPD1/CD39 orchestrates an adenosine-rich immunosuppressive microenvironment that promotes M2 macrophage polarization and resistance to PD-L1 blockade via a novel LT-mediated crosstalk axis . CD39 serves as both a prognostic biomarker and a tractable immunotherapeutic target to overcome resistance in TNBC .

Corruption as an upstream determinant of global cancer outcomes.

e23031 Background: Corruption undermines health systems globally through resource diversion and weakened governance, yet its specific relationship with cancer outcomes has not been systematically quantified. This study evaluated the relationship between national corruption levels and cancer mortality-to-incidence ratios (MIR) globally, with a focus on the mediating factors through which corruption may influence cancer outcomes. Methods: We conducted a global pan-cancer ecological study of 173 countries using GLOBOCAN 2022 data. Corruption was quantified using the 2024 Corruption Perceptions Index (CPI). Univariable and multivariable linear regressions were performed. To evaluate mediation, we applied single-mediator causal mediation analysis using the Imai–Keele–Tingley framework with nonparametric bootstrapping, sequential attenuation modeling, and structural equation modeling. Results: CPI demonstrated a strong inverse association with cancer MIR in univariable analysis (β = −0.00567, p < .001; R² = 0.526). Causal mediation analyses estimated that approximately 94% of CPI's total association with MIR operated through downstream pathways, most prominently Human Development Index (74.1%) and GDP per capita (54.7%). In absolute terms, a 10-point increase in CPI was associated with a 0.059 reduction in MIR. Sequential attenuation modeling similarly demonstrated a 94.4% reduction in the CPI coefficient after adjustment for GDP, UHC, HDI, and health spending. Structural equation modeling estimated that HDI accounted for the largest proportion of the total indirect effect. Conclusions: National corruption levels are strongly associated with cancer outcomes globally, which is largely mediated through economic development and health system capacity. Anti-corruption efforts may therefore function as foundational enablers of effective cancer control by strengthening downstream institutional and economic conditions. Univariable analysis of national health system factors and cancer outcomes. Health System Measure N Beta (95% CI) P value R² Corruption Perceptions Index 173 -0.00567 (-0.0065 to -0.0049) <.001 0.526 GDP per capita (US$) 172 -5.10 × 10⁻⁶ (-5.8×10⁻⁶ to -4.5×10⁻⁶) <.001 0.605 Health spending (% of GDP) 172 -0.022 (-0.030 to -0.018) <.001 0.200 Physicians per 1000 173 -0.066 (-0.070 to -0.054) <.001 0.569 Nurses per 1000 173 -0.030 (-0.031 to -0.025) <.001 0.673 Surgical workforce per 1000 152 -0.0028 (-0.0031 to -0.0022) <.001 0.494 UHC Service Coverage Index 162 -0.0078 (-0.0083 to -0.0068) <.001 0.666 Human Development Index 172 -0.85 (-0.89 to -0.76) <.001 0.782 Gender Inequality Index 152 0.63 (0.57 to 0.70) <.001 0.686 Radiotherapy centers (count) 137 -0.00014 (-0.00024 to -0.00004) .005 0.056 Abbreviations: CI, confidence interval; GDP, gross domestic product; UHC, universal health coverage.

Long-term incidence of hypothyroidism and thyroiditis after CAR T-cell therapy versus standard chemotherapy in hematologic malignancies: A propensity-matched multi-institutional cohort study.

11148 Background: Thyroid dysfunction is a common endocrine immune-related adverse event with modern cancer immunotherapies. Chimeric antigen receptor T-cell (CAR T) therapy is an adoptive cellular immunotherapy with expanding use in relapsed/refractory hematologic malignancies; its toxicity profile of cytokine release syndrome and immune effector cell–associated neurotoxicity syndrome are well described, whereas endocrine sequelae are comparatively less characterized. Large-scale population analysis regarding the effect of CAR T therapy on thyroid function remains limited. As CAR T survivorship increases and indications broaden, defining long-term thyroid-related outcomes is essential in optimizing care. Methods: We performed a propensity-matched cohort study using the TriNetX Research Network (Jan 1, 2018–Jan 24, 2026). Adult, non-pregnant patients with DLBCL, PMBCL, follicular lymphoma, mantle cell lymphoma, multiple myeloma, or ALL were included. Patients receiving CAR T in addition to chemotherapy were compared with chemotherapy alone. Exclusions included prior thyroid disease/malignancy, central endocrine disorders, radiation therapy, bone marrow transplant, and baseline exposure to thyroid-active/thyroid-disrupting therapies (e.g., amiodarone, immune checkpoint inhibitors, tyrosine kinase inhibitors, interferons, immunomodulatory agents like lenalidomide). Propensity score matching balanced age, sex, race, malignancy type, baseline comorbidities, steroid exposure, and antipsychotic exposure. The primary outcome was new-onset hypothyroidism or thyroiditis; effect estimates are reported as ORs with 95% CIs. Results: Before propensity-score matching, there were 921 patients in the CAR T group and 24,426 in the chemo-only group. Post-matching, there were 878 CAR T patients and 879 chemo-only patients. Hypothyroidism or thyroiditis occurred in 16 patients in the CAR T group (1.822%) and in 21 patients in the chemo-only group (2.389%). Absolute risk difference was −0.567% (95% CI -1.909%, 0.776%; p = 0.4081) and the OR was 0.758 (95% CI 0.393, 1.463). No statistically significant difference in incidence was observed. Mean follow-up for CAR T after chemo and chemo-only groups were 670 and 964 days, respectively. Conclusions: In this large, multi-institutional propensity-matched real-world analysis with long-term follow-up, CAR T therapy after chemotherapy was not associated with increased incidence of hypothyroidism/thyroiditis compared with chemotherapy alone. These findings provide reassuring thyroid safety data, supporting the expanding use of CAR T as indications broaden. Future studies should evaluate thyroid outcomes by CAR T construct/target and characterize timing and severity to inform monitoring strategies.

Patient experience with later line chemotherapy in four tumor types.

e24142 Background: Understanding patient experiences with prior therapies offers crucial context for assessing the quality of life benefits of antibody-drug conjugates (ADCs). This study aimed to qualitatively explore the perspectives of patients with advanced or metastatic endometrial cancer (EC), cervical cancer (CC), gastric cancer (GC), and epidermal growth factor receptor mutation (EGFRm) positive non-small cell lung cancer (NSCLC) who had progressed to second-line (2L) and later therapies, focusing on tolerability challenges and quality of life (QoL) impacts related to past and current treatments. Methods: This study employs qualitative data to capture treatment burden and impact from the patient perspective, emphasizing the value of patients’ own words and experiences. In-depth, semi-structured interviews were conducted with 25 patients (EC=8, CC=5, GC=5, NSCLC=7). Interviews explored patients’ experiences with treatment tolerability issues, impact on daily life, coping strategies, and motivations to continue their treatment. Data were analyzed using a modified content analysis framework to identify key concepts. Results: Most patients anticipated treatment-related side effects and proactively discussed these with their healthcare providers. Across tumor types, the most burdensome symptoms included gastrointestinal issues (nausea, diarrhea) and neurological effects (notably neuropathy and brain fog), resulting in significant, multi-domain QoL impacts. Commonly reported effects in the physical domain included persistent fatigue (particularly among NSCLC patients), reduced mobility due to neuropathy, and appetite loss from nausea. Emotional distress was prevalent, characterized by sadness, grief, and self-consciousness stemming from visible treatment effects (e.g., hair loss, IV marks). Reduction in social participation and feelings of isolation were also reported, such as staying home from important events or feeling the need to keep cancer-related struggles private. Many patients described disruptions to work and daily routines, with some troubled by job modifications or cessation. Despite these challenges, patients remained motivated to continue treatment by the desire to prolong survival and maintain connections with loved ones. Conclusions: This real-world, qualitative analysis highlights the extensive physical, emotional, and social impacts of 2L+ therapies on patients with advanced cancers. These insights are critical for informing ADC treatment integration as a patient-centered treatment option and facilitating scientific discourse to better address patient QoL needs.

Digital twin models for counterfactual prediction in NSCLC: A validation study using clinical trial data.

e20517 Background: Digital twin models (DTMs) developed on real world data can estimate a patient’s response to treatments they did not receive. These counterfactual (CF) predictions can help reduce a trial’s sample size without compromising its statistical power. Validation of DTM-based CF predictions remains limited. We evaluated CF outcome predictions in NSCLC by benchmarking DTM estimates against observed outcomes from IMpower131 and 132. 1 We then assessed potential trial sample size reductions enabled by DTMs. Methods: We trained machine learning (ML) models (penalized logistic regression [LR], XGBoost [XGB], MLP, GAT) to predict real-world overall survival (rwOS) for patients with stage IV NSCLC initiating 1L platinum chemotherapy (chemo) 2011-2016 from the Flatiron Health Research Database. Features were generated using structured and LLM-extracted clinical details (e.g. Charlson comorbidity index [CCI]; sites of metastases [SOM]). Models were internally validated on a held-out test set and externally validated (MAD 2 , rwOS and hazard ratio [HR] comparisons) using patient-level data from IMpower131 and 132. Predictions were made for stage IV patients in 1) the control arm to compare model predictions with observed trial outcomes (Table, Control Arm Validation) and 2) the experimental arm to simulate their outcomes as if they had instead received control-arm therapy (Table, CF Treatment Effect). We estimated trial sample size reduction based on the C-index from Cox Proportional Hazards models with model-based prognostic scores for covariate adjustment. Results: Key features across models included ECOG status, albumin, Brain+liver SOM, and CCI. LR and XGB performed best for IMpower131 and 132 respectively, with MAD 2 < 5% and predicted (pred) control arm rwOS similar to those observed (obs) in the trials 3 . CF predictions yielded HRs consistent with trial results (Table). Estimated sample size reduction was 9-15% and 15-21% for IMpower131 and 132 respectively. Conclusions: Validated DTMs demonstrate the feasibility of accurately predicting CF outcomes with advanced ML. This approach can enable efficient, well-powered trials with smaller sample sizes and supports a path toward broader validation and regulatory adoption. IMpower 131 IMpower 132 Control Arm Validation Pred vs Obs Median rwOS (months) 10 vs 12.6 3 15 vs. 13.1 3 MAD 2 3.9% 1.8% C-Index 0.61 0.67 HR (Pred:Obs; 1 is perfect prediction) 0.97 [0.86-1.00] 4 1.03 [1.00-1.06] 4 CF Treatment Effect Obs trial HR 0.86 (0.73-1.02) 3 0.82 (0.66-1.01) 3 DTM-derived HR 0.81 (0.79-0.84) 4 0.85 (0.83-0.86) 4 1 IMpower131 & IMpower132: atezolizumab plus chemo vs chemo alone in squamous and non-squamous NSCLC respectively. 2 Mean absolute difference between pred & obs OS curves. 3 Observed results differ slightly from the published trials as this analysis restricted to subset of stage IV patients. 4 95% CI based on 1000 bootstrapped samples of trial data.

COVID-19 mRNA vaccination and risk of immune-related adverse events in ICI-treated cancer patients.

2641 Background: Immune checkpoint inhibitors (ICIs) are now standard-of-care in cancer therapy; however, concerns persist regarding whether COVID-19 mRNA vaccination may enhance immune activation and exacerbate immune-related adverse events (irAEs). We evaluated the association between COVID-19 mRNA vaccination and irAE risk among patients initiating ICI therapy and assessed toxicity and survival outcomes. Methods: We conducted a retrospective cohort study using target trial emulation with 1:1 propensity score matching based on data from the TriNetX Research Database (January 2021–December 2025). Adult patients with metastatic cancer initiating their first ICI therapy were included. Patients with prior ICI therapy, active irAE within 30 days before ICI initiation or immunosuppressive conditions were excluded. Vaccinated patients who received COVID-19 mRNA vaccination within 100 days prior to ICI initiation were matched to unvaccinated controls. The primary outcome was composite irAE incidence over 36 months. Secondary outcomes included severe irAEs, organ-specific irAEs, all-cause mortality, pneumonia, and intensive care unit (ICU) admission. Cox proportional hazards models estimated hazard ratios (HRs) with 95% confidence intervals (CIs). E-values and quantitative bias analyses assessed robustness to unmeasured confounding. Results: The matched cohort included 7,218 patients (3,609 vaccinated; 3,609 unvaccinated). Vaccinated patients had a higher risk of overall irAEs (HR 1.22, 95% CI 1.17–1.28; E-value 1.74) and severe irAEs (HR 1.11, 95% CI 1.04–1.18; E-value 1.45). Organ-specific analyses demonstrated a significant increase in ocular irAEs (HR 1.48, 95% CI 1.17–1.87; E-value 2.33), with non-significant associations for dermatologic (HR 1.08, 95% CI 0.97–1.21) and rheumatologic irAEs (HR 1.12, 95% CI 0.96–1.32). No significant increases were observed for cardiac, endocrine, gastrointestinal, hematologic, or pulmonary irAEs. Despite higher irAE risk, vaccination was associated with lower all-cause mortality (HR 0.86, 95% CI 0.80–0.93) and fewer ICU admissions (HR 0.43, 95% CI 0.25–0.73). Subgroup analyses showed stronger overall irAE associations in females and patients aged ≥50 years. Quantitative bias analysis indicated moderate robustness, with a 74.8% probability that HR>1.0 persists despite plausible unmeasured confounding. All P<0.001. Conclusions: COVID-19 mRNA vaccination among patients receiving ICIs is associated with a modest increase in irAEs, particularly ocular toxicities, but may offer survival and critical illness benefits. These findings support continued vaccination in eligible ICI-treated patients, along with enhanced irAE monitoring, including ophthalmologic surveillance in higher-risk groups.

Use of surrogate endpoints in health technology assessment: A review of company submissions for cancer drugs in Singapore.

e23147 Background: Surrogate endpoints are increasingly used to accelerate clinical trials and regulatory approval of new cancer drugs. However, their use poses challenges for health technology assessment (HTA), which focuses on long-term clinical outcomes and cost-effectiveness. In Singapore, the Agency for Care Effectiveness (ACE) conducts HTA to inform funding decisions by the Ministry of Health Drug Advisory Committee. Since 2023, companies have provided evidence submissions to ACE for cancer drugs to be evaluated for funding via the company-led submission process. This review examines the extent to which surrogate endpoints are used, and how they are validated, in cancer drug submissions to ACE. Methods: We reviewed all cancer drug submissions to ACE between 2023 and 2025. Data were extracted on the use of surrogate endpoints, including their purpose, evidence base and methods used to validate the surrogate-final outcome relationship. Results: Of 31 submissions, 24 (77%) were for solid tumours and the remaining 7 (23%) for blood cancers. Surrogate endpoints were used as key evidence in 17 submissions (55%) to support clinical or cost-effectiveness analyses. In most of these submissions (82%), overall survival data were immature at the time of evaluation. Eight different surrogate endpoints were considered across the 17 submissions. Progression-free survival was the most commonly used (59%), followed by disease-free survival (18%) and event-free survival (12%). The strength of evidence supporting the validity of the surrogate relationship varied considerably. Five submissions (29%) provided no supporting evidence for the surrogate relationship. Ten submissions (59%) cited randomised controlled trials; however, these were often conducted in populations or settings not fully aligned with the submission context. Two submissions (12%) relied on assessments from overseas HTA agencies. Quantitative measures of surrogate validity were infrequently reported: correlation coefficients were provided in six submissions (35%), and surrogate threshold effects estimating the expected treatment effect on the final clinical outcome in only one submission (6%). Conclusions: Most cancer drug submissions for funding consideration in Singapore rely on surrogate endpoints, often without adequate validation. While surrogate endpoints may support accelerated regulatory approval, inadequate validation introduces significant uncertainty into HTA assessments of clinical benefit and cost-effectiveness. Strengthening evidence for surrogate validation in company submissions could support more robust HTA evaluations and help reduce uncertainty in decision-making.

COMFORT Survey: A comparison of the comfort level of hematologic and solid tumor oncology specialists with serious illness conversations.

12102 Background: Serious illness conversations (SICs) are essential in aligning medical care with the values and goals of patients with advanced cancer. It is unclear if hematologic and solid tumors specialists (i.e., physicians [MDs] and advanced practice providers [APPs]) differ in their comfort level with SICs, which may drive end-of-life care outcomes. We compared the comfort level with SICs between hematologic and solid tumor specialists and examined contributing factors. Methods: We surveyed 186 hematology and 186 solid tumor specialists at MD Anderson Cancer Center. The 64-question survey examined their comfort with SICs under 5 domains (Transitions of Care [TOC, primary outcome], Dying Process [EOL], Prognosis, Goals, Advance Care Planning [ACP]). Responses were rated from 0 (extremely uncomfortable) to 10 (extremely comfortable). The TOC domain included 4 questions (i.e., stopping cancer therapy / transfusions / artificial nutrition and recommending hospice). We compared groups using simple linear and multivariable models accounting for specialist demographics. Results: 245 specialists completed this survey, with a response rate of 66%. Hematologic and solid tumor specialists both reported high comfort with TOC (7.0 [2.4] vs 7.1 [2.2]), with no significant difference between groups (mean difference [95% CI]: 0.1 [-0.5, 0.7]; P=0.78). In univariate analysis, APPs were significantly less comfortable with TOC than MDs among both hematologic (2 [1.2, 2.8]; P<0.0001) and solid tumor specialists (1.4 [0.7, 2.2]; P=0.001). In multivariate analysis, APPs remained the only significant variable associated with less comfort with TOC in hematologic (1.5 [0.4, 2.7]) and solid tumor groups (1.2 [0.2, 2.1]). Similarly, we found no significant difference between hematologic and solid tumor specialists in comfort level with the other 4 SIC domains and observed significant differences between APPs and MDs (Table). Conclusions: Both hematologic and solid tumor specialists reported high levels of comfort with SIC; however, APPs were significantly less comfortable. Given that APPs are increasingly involved in SIC, our study underscores the need to provide more training and support for APPs. Self-reported comfort levels for 5 SIC domains. Mean [SD] MD vs. APP Mean differences [95% CI]; p-value Domains Solid MD Solid APP Heme MD Heme APP Heme Solid TOC 7.8 [1.7] 6.4 [2.5] 8.1 [1.9] 6.1 [2.4] 2 [1.2,2.8]; <.0001 1.4 [0.7,2.2]; 0.001 Prognosis 8.5 [1.4] 7.1 [2.1] 8.6 [1.8] 6.8 [1.8] 1.9 [1.2,2.5]; <.0001 1.4 [0.8,2]; <.0001 Goals 8.5 [1.2] 7.8 [1.7] 8.9 [1.4] 7.5 [1.9] 1.4 [0.8,2]; <.0001 0.7 [0.1,1.2]; 0.09 EOL 7.9 [1.9] 6.3 [2.6] 8 [2.3] 6.2 [2.8] 1.8 [0.9,2.7]; 0.0005 1.6 [0.8,2.4]; 0.0009 ACP 8.6 [1.2] 7.2 [2.5] 9.1 [1.1] 7.3 [2.4] 1.7 [1,2.5]; <.0001 1.4 [0.7,2.1]; 0.0006

Green synthesis of gallium-doped zinc oxide nanoparticles using piper betle extract for antibacterial applications

Dynamic Smart Membranes: Real‐Time Perception, Self‐Response, and AI‐Driven Optimization for High‐Safety Lithium‐Based Batteries

ABSTRACT Lithium‐based batteries are fundamental to modern energy storage systems, yet their safety remains a critical challenge due to risks such as thermal runaway, dendrite‐induced short circuits, and interfacial degradation. Conventional composite membranes, including composite separators and solid electrolytes, have been developed to improve thermal stability, mechanical strength, and ionic conductivity. However, these static materials lack the ability to dynamically respond to real‐time operational stresses such as local temperature spikes, mechanical deformation, or evolving electrochemical conditions, leading to persistent safety limitations. To overcome these issues, smart membranes have emerged as a transformative solution, integrating real‐time perception, self‐responsive mechanisms, and artificial intelligence (AI) to enhance battery safety proactively. This review systematically addresses the three primary safety issues of conventional composite membranes, including thermal instability, mechanical failure, and ion transport limitations, through detailing how smart membranes leverage embedded sensors for continuous perception, employ self‐protection and self‐healing functionalities to mitigate risks, and utilize AI for material optimization and failure prediction. Furthermore, we discuss the industrial viability of smart membranes, highlighting challenges related to cost, scalability, and integration, while outlining future directions including multifunctional coupling, wireless sensing, and advanced material designs. Smart membranes represent a transformative advance toward autonomous, safe, and durable next‐generation lithium batteries.

Beyond didactics: Simulation-based education in hematology/oncology training—A scoping review.

e21010 Background: Clinical oncology exposure in residency is often varied and limited, contributing to difficulty in transitioning to fellowship. Simulation-based education (SBE) is one of the established approaches in graduate medical education (GME) to develop clinical reasoning, procedural skills, teamwork, and high-stakes communication. We present a scoping review of the applications of SBE among medical professionals in the field of hematology/oncology, aiming to map the existing literature on effectiveness and implementation. Methods: PubMed and Cochrane Library were searched for original articles in English, published from 2016 to January 2026. Inclusion criteria included simulation education involving medical professionals (residents, fellows, physicians) with hematology and/or oncology-related content. We excluded studies focused primarily on surgical techniques or procedural skills; anesthesiology or obstetrics/gynecology; point-of-care ultrasound; or non-SBE interventions. Two researchers adhered to the PRISMA-ScR checklist and independently screened titles and abstracts; potentially eligible articles underwent full-text review. Data were charted from included studies. Results: The search yielded 483 articles; 40 underwent full-text review, and 19 were selected for final analysis. Studies were conducted mainly in the United States (15/19, 79%), with additional articles from Canada, Switzerland, Germany, and China. Simulation methods included virtual simulation, natural-language processing based patient avatars, role-play, and simulated patients and computer-based case simulation. Learners were mostly Fellows in Hematology and Oncology, but studies also included medical students, and interprofessional practitioners. Simulation-based education was applied across diverse competency domains, mostly in communication or palliative care-focused training. Other areas include clinical management, systems-based practice (root cause analysis, resource stewardship), diagnostics and disease specific learning with case-based scenarios, and therapy and procedural skills (chemotherapy prescribing, bone marrow biopsy). Intervention duration ranged from brief 30-minute sessions to multi-day workshops. Conclusions: Simulation-based education in hematology/oncology can be applied widely to different clinical settings, but is concentrated in communication-based training. Further studies could utilize virtual simulation, and expand simulation to include clinical and systems based scenarios to improve education in the field, which may help reduce resource barriers and promote more equitable access to high-quality training.

Neoadjuvant therapy vs. standard of care in oral cavity squamous cell carcinoma: A meta-analysis.

e18069 Background: Historically, locally advanced, resectable oral squamous cell carcinoma was treated with upfront surgery followed by adjuvant therapy. Advances in systemic treatment, particularly immunotherapy, have led to paradigm shifts with current NCCN guidelines now supporting the use of neoadjuvant immunotherapy for selected patients with PD-L1–expressing, locally advanced disease. While multiple individual studies and meta-analyses have compared neoadjuvant therapy to surgery alone; no prior meta-analyses have directly compared neoadjuvant and adjuvant approaches in this population. Methods: Pubmed, Embase, and Scopus were searched for articles written since 2015 comparing disease-free survival and overall survival in patients with locally advanced oral squamous cell carcinoma receiving neoadjuvant therapy (chemotherapy +/-immunotherapy, immunotherapy alone) vs. standard of care. Retrospective cohort studies and randomized control trials (RCT) were analyzed separately with combined hazard ratios. Random effects were used when I² > 50% and fixed-effects models were used when I² ≤ 50%. Results: Out of 561 articles, five met criteria for inclusion in the systematic review; 4 had enough data to be included in the meta-analysis. Two of these were retrospective cohort studies (n=221) and two were RCTs(n=996). In the retrospective cohort studies, neoadjuvant therapy was associated with increased overall survival but not disease-free survival. Heterogeneity between those studies was 0%. RCTs showed no difference in overall survival between neoadjuvant and SOC but had a high level of heterogeneity (81%). Conclusions: Neoadjuvant therapy in OCC consistently trended toward increased progression free-survival and overall survival, with retrospective studies retaining significance for overall survival with a low degree of heterogeneity. Prior meta-analyses limited to neoadjuvant chemotherapy have not demonstrated clear survival benefit. By incorporating broader neoadjuvant strategies, including immune checkpoint inhibition, this analysis identifies a consistent directional signal favoring neoadjuvant approaches. Direct comparisons between neoadjuvant pembrolizumab and adjuvant systemic therapy remain absent, underscoring the need for prospective trials. Disease-free and overall survival in patients with oral cavity squamous cell carcinoma receiving neoadjuvant therapy vs standard of care. Study Type Outcome Hazard Ratio 95% Confidence Interval p-value I² Retrospective Cohort Studies Disease-Free Survival 0.45 0.18-1.12 .087 79.71% Retrospective Cohort Studies Overall Survival 0.26 0.16-0.45 <.001 0% Randomized Control Trials Overall Survival 0.53 0.18-1.53 .24 84.32%

EGFR tyrosine kinase inhibitor plus angiogenesis inhibitor combination therapy in treatment-naïve <i>EGFR</i> -mutant advanced non–small cell lung cancer: A systematic review and meta-analysis.

e20754 Background: The benefit of adding angiogenesis inhibitors to EGFR-TKIs in treatment naive EGFR mutant NSCLC remains controversial. We conducted an updated meta analysis to evaluate efficacy and safety. Methods: We searched PubMed, Embase, Cochrane Library, and oncology conferences through January 2026 for RCTs comparing EGFR TKI plus angiogenesis inhibitor versus EGFR TKI alone in treatment-naïve EGFR mutant advanced NSCLC. Primary endpoint was PFS. Random effects meta analyses with Hartung Knapp adjustment were performed. Risk of bias was assessed using RoB 2.0; certainty using GRADE. Results: Thirteen RCTs enrolling 2,580 patients were included. Combination therapy significantly improved PFS (HR 0.65, 95% CI 0.60-0.71; p &lt; 0.0001; I² = 0%). Median PFS ranged from 13.7-24.8 months versus 8.2-20.2 months in controls. PFS benefit was consistent across subgroups: exon 19 deletion (HR 0.59), L858R (HR 0.64), brain metastases present (HR 0.59) or absent (HR 0.63), and across angiogenesis inhibitor classes. OS was not improved (HR 0.97, 95% CI 0.87-1.07). Combination therapy increased toxicity: grade ≥3 AEs (RR 1.53), hypertension (RR 6.30), and proteinuria (RR 10.56). GRADE certainty was high for PFS benefit and absence of OS benefit. Conclusions: EGFR TKI plus angiogenesis inhibitor significantly prolongs PFS but does not improve OS and increases toxicity. These findings provide high certainty evidence to guide shared decision making in the first line setting. Key subgroup analyses for PFS. Subgroup HR 95% CI I² Overall (13 studies) 0.65 0.60–0.71 0% Exon 19 deletion 0.59 0.49–0.71 0% L858R mutation 0.64 0.52–0.79 0% Brain mets present 0.59 0.46–0.74 0% Brain mets absent 0.63 0.53–0.76 0% Phase III trials 0.64 0.59–0.70 0%

Frailty among hospitalized patients with pancreatic cancer: A National Inpatient analysis.

11183 Background: Frailty is increasingly recognized as a key determinant of outcomes in older adults with cancer; however, its prevalence and inpatient correlates among hospitalizations for pancreatic cancer remain incompletely characterized. We examined the national inpatient burden of frailty among adults hospitalized with pancreatic cancer and compared demographics, comorbidities, and healthcare utilization between frail and non-frail patients. Methods: We analyzed National Inpatient Sample data (2016–2020) to identify adult pancreatic cancer hospitalizations using ICD-10 codes. Frailty was defined by ICD-10 R54. Survey-weighted analyses accounted for complex sampling, with statistical significance set at p &lt; 0.05. Results: Among an estimated 775,510 weighted adult hospitalizations with pancreatic cancer, frailty was identified in 0.39%. Frail patients were significantly older (74.96 vs 68.32 years, p &lt; 0.001) and more often female (56.4% vs 47.7%, p &lt; 0.001). Frailty was more prevalent among patients from lower-income neighborhoods and varied by hospital region (both p &lt; 0.01). Frail patients experienced longer hospital stays (7.16 vs 6.16 days, p &lt; 0.001), with similar total hospitalization charges ($77,978 vs $75,477). Frailty was associated with higher burdens of congestive heart failure (15.9% vs 10.3%) and sepsis (22.8% vs 17.0%) (both p &lt; 0.001). Frail patients demonstrated marked nutritional and functional vulnerability, including underweight status (24.6% vs 9.1%), anemia (52.5% vs 44.0%), malnutrition (46.2% vs 28.8%), chronic pain (24.4% vs 19.5%), and falls or mobility impairment (8.5% vs 3.6%) (all p &lt; 0.01). In multivariable analysis, advanced age, female sex, underweight status, malnutrition, anemia, chronic pain, sepsis, congestive heart failure, functional impairment, and lower socioeconomic status remained independently associated with frailty. Conclusions: Frailty was uncommon but clinically meaningful among adults hospitalized with pancreatic cancer and was associated with advanced age, nutritional and functional vulnerability, acute illness severity, and socioeconomic disadvantage. These findings underscore the need for improved inpatient frailty recognition and targeted supportive care strategies. Predictor of frailty in pancreatic cancer. Predictors Frail (%) Non-frail (%) p-value Congestive heart failure 15.9 10.3 &lt;0.001 Sepsis 22.8 17.0 &lt;0.001 Underweight 24.6 9.1 &lt;0.001 Anemia 52.5 44.0 &lt;0.001 Malnutrition 46.2 28.8 &lt;0.001 Chronic pain 24.4 19.5 &lt;0.001 Falls/mobility impairment 8.5 3.6 &lt;0.001 Obesity 4.4 8.6 0.0002 Smoking 9.3 12.2 0.034 Chronic kidney disease 16.2 12.8 0.012

Genetic and epigenetic modulation of CAR T-cell responses: A systematic review exploring the influence of host genetic variations and epigenetic modifications on CAR T-cell therapy outcomes.

e15197 Background: Chimeric antigen receptor (CAR) T-cell therapy demonstrates variable efficacy in hematological malignancies (complete response rates: 50-90%), partially attributable to host T-cell characteristics present before manufacturing. This review examines T-cell intrinsic genetic variants and epigenetic states that influence CAR-T expansion, persistence, and anti-tumor function. Methods: This systematic review followed PROSPERO-registered protocol (CRD42024548616).PubMed, Embase, and Scopus were systematically searched through April 2025, identifying 25 eligible studies wherein CAR-T targets included CD19 and GD2. Epigenetic interventions covered DNA methyltransferase inhibitors and histone deacetylase inhibitors; genetic profiling included whole-exome sequencing or targeted mutation panels. Quality was assessed using Cochrane Risk of Bias 2, Newcastle-Ottawa Scale, and ROBINS-I tools. GRADE framework was used to evaluate evidence certainty. Results: Evidence quality ranged from low to very low (GRADE assessment). Pre-infusion DNA methylation at 54 CpG sites associated with CD19 CAR-T outcomes in one chronic lymphocytic leukemia cohort (hazard ratio for event-free survival 0.12, 95% CI 0.03-0.51, n = 43), with partial validation in B-cell acute lymphoblastic leukemia/non-Hodgkin lymphoma (n = 24) but contradictory null findings in diffuse large B-cell lymphoma (n = 2 studies). Decitabine treatment during manufacturing reprogrammed T cells from Th2 toward Th1 phenotype (7 experimental studies; standardized mean difference 1.24, 95% CI 0.78-1.70) but showed no survival benefit in one phase 1/2 clinical trial (n = 32). Cytotoxicity of CAR-T cells in vitro was enhanced by histone deacetylase inhibitors (6/6 preclinical studies),clinical outcomes didn't improve and failed during post-infusion (tucidinostat trial, n = 32). TET2-mutant CAR-T products demonstrated enhanced complete response rates in chronic lymphocytic leukemia (54% vs. 12%, p &lt; 0.01, n = 43) but not in diffuse large B-cell lymphoma (2 null studies). In solid tumor models (3 preclinical studies), tumor-associated macrophage M2 polarization and myeloid-derived suppressor cell infiltration correlated with reduced CAR-T expansion (correlation coefficients -0.62 to -0.74), though causality was not established. Conclusions: T-cell intrinsic genetic and epigenetic factors show associations such as very low certainty evidence, (GRADE ⊕⊝⊝⊝). However, effect sizes are modest, findings are inconsistent across disease, and clinical translation faces manufacturing timeline constraints. Pre-specified methylation panels, standardized profiling platforms, and manufacturing-integrated epigenetic interventions are required before these biomarkers can guide better therapeutic optimization.

Mortality trends and demographic disparities in breast cancer deaths with co-existing diabetes mellitus among U.S. adults ≥45 years, 1999–2020.

e13135 Background: Breast cancer stands out as a considerable cause of mortality attributed to cancer, and diabetes mellitus is a prevalent health issue that can lead to detrimental consequences. Existing literature provides an insufficient overview of the patterns and differences in mortality rates associated with breast cancer, in the presence of diabetes mellitus, within the older demographic. Methods: Data regarding mortality from the CDC WONDER spanning 1999 to 2020 were scrutinized, focusing on the relevant cohort of United States (US) individuals aged 45 and above. Fatalities were identified using malignant neoplasm of the breast as the underlying cause of death UCD (ICD-10 code C50), coupled with diabetes mellitus as a contributory cause of death (ICD-10 codes E10 to E14). Age-adjusted mortality rates (AAMR) per 100,000 were determined utilizing the 2000 United States Standard Population. All evaluations were segmented by gender, age ranges, state, and classification of urban or rural. Crude rated were determined for ten-year age group stratification. Results: Between 1999 and 2020, a total of 65419 death rates from breast cancer and diabetes-related conditions were reported in US adults aged 45 and older. Between 1990 and 2018 the total deaths decreased. Still, in 2020, there was a significant increase, bringing the average yearly death rate to 3.0 per 100,000 people. The total death toll among men (1999-2020) was 1,008, with a steady average yearly mortality rate of 0.1 per 100,000 people in both 1999 and 2020. Among women total death toll (1999-2020) was 64411, AAMRs rose slightly from 1999 to 2004 (APC 1.27%), declined from 2004 to 2018 (APC −1.59%), then increased sharply from 2018 to 2020 (APC 13.97%). The basic death rates rose significantly with increasing age (per 100,000): 0.3 for ages 45–54, 1.1 for ages 55–64, 3.1 for ages 65–74, 7.2 for ages 75–84, and 14.0 for those 85 and older. The highest average yearly mortality rates per state were observed in the District of Columbia (4.3), Ohio (4.0), Nebraska (3.8), Mississippi (3.7), Oklahoma (3.7), and West Virginia (3.6). Conclusions: Mortality trends among U. S. Adults, due to breast cancer deaths with concurrent diabetes among adults aged 45 years and older, has decreased in the last two decades but has risen markedly in recent years. Deaths were more common among those above 60 years of age, women, nonmetropolitan inhabitants, in addition to people living in particular states. The recent increase in mortality rate and disparities among age groups, gender and states emphasize the necessity of all-encompassing strategies for handling breast cancer and diabetes.

SMS message sentiment as a predictive indicator of quality of life and program adherence in oncology nutrition.

1627 Background: A text-message based artificial intelligence nutrition platform called “Ina” was previously developed and shown to be equivalent to human dietitians in providing nutritional recommendations to patients with cancer. Once registered on the platform, patient users exchange conversational texts with and receive tailored dietary recommendations and recipes from Ina. We investigated whether the emotional sentiment of SMS messages can act as a proactive predictor of program-related Quality of Life (QoL) and program adherence. Methods: QoL was captured via a 1-5 Likert scale (“Has using Ina improved your quality of life?”) survey, and program adherence was captured by a binary response to a question (“Have you used any of Ina’s suggested tips?”). We extracted SMS messages within a 60-day window prior to each survey and question. Sentiment was measured using the Jockers-Rinker lexicon via the validated sentimentr package; scores were calculated as a weighted average ranging from -1 (negative sentiment) to +1 (positive sentiment). Linear mixed-effects models (LMM) were used to quantify the relationship between pre-survey message sentiment and program QoL/adherence outcomes, adjusting for total number of surveys completed and user’s unique baseline sentiment. Results: Between October 2019 and January 2026, 14,700 unique text messages were analyzed from 540 unique patient users of the platform (mean age: 51± 20 years; 91% female). Cancer types included breast (40%), ovarian (24%), bladder (16%), lung (13%), and colon (7%); metastases were present among 34%. For every one-point improvement in digital sentiment, users experienced a 14% improvement in subsequent program-related QoL (𝛽= 0.72, p &lt; 0.001). Furthermore, there was a trend toward an association between higher positive digital sentiment and subsequent reported adherence to nutritional recommendations (𝛽 = 0.12, p = 0.097). Conclusions: Sentiment analysis of unstructured SMS messages successfully identified subtle shifts in digital tone that precede structured assessments. This 'leading indicator' of sentiment could enable proactive supportive care and provider escalation for timely triage and intervention. Given rapid symptomatic deterioration in oncology that often leads to non-adherence, text-based sentiment analysis could enable early detection of declining engagement to support improved quality of life and long-term outcomes.

Clinicopathologic characteristics and cancer-specific survival in adolescents and young adults with endometrial cancer: A population-based SEER analysis.

e17615 Background: Endometrial cancer is the most common gynecologic malignancy in the United States, with an increasing incidence among younger women. Adolescents and young adults (AYAs) represent a distinct population with differing tumor biology, treatment patterns, and survivorship considerations compared with older adults. We evaluated age-based differences in presentation, treatment utilization, and CSS among patients with endometrial cancer. Methods: We conducted a retrospective population-based study using Surveillance, Epidemiology, and End Results (SEER) Research data (17 registries; diagnoses 2000–2022; November 2024 submission). Women with primary endometrial cancer were identified using ICD-O-3 site codes. AYAs were defined as age &lt;40 years and compared with older adults (≥40 years). Variables included stage, histology, grade, race, year of diagnosis, and receipt of surgery. Descriptive comparisons used chi-square testing. CSS was estimated using the Kaplan–Meier method and compared using the log-rank test. Multivariable Cox regression identified independent predictors of CSS. Analyses were performed using Stata/SE 19.5. Results: Among 279,829 women with endometrial cancer, 10,711 (3.8%) were AYAs. Median age was 35 years (IQR 32–38) in AYAs and 63 years (IQR 55–71) in older adults. AYAs more frequently presented with localized disease (52.3% vs 48.4%) and less often with regional (10.1% vs 14.6%) or distant disease (4.6% vs 6.4%). Endometrioid histology predominated in both groups but was more common in AYAs (68.3% vs 63.2%), whereas non-endometrioid tumors were markedly less frequent (2.1% vs 14.0%). High-grade disease occurred less often in AYAs (7.6% vs 17.1%). AYAs had lower proportions of non-Hispanic White patients (71.2% vs 80.0%) and higher representation of Asian/Pacific Islander patients (15.1% vs 8.6%). Surgery was performed less frequently in AYAs (82.5% vs 90.3%). Median CSS was not reached among AYAs, while older adults experienced significantly inferior CSS (p&lt;0.001). On multivariable analysis, AYA status was independently associated with improved CSS (HR 0.39, 95% CI 0.36–0.43). Advanced stage, non-endometrioid histology, high-grade disease, Black race, and lack of surgery were independently associated with worse CSS (all p&lt;0.001). Conclusions: In this extensive population-based analysis, adolescents and young adults with endometrial cancer demonstrated significantly more favorable cancer-specific survival compared with older adults. Across all ages, advanced stage, non-endometrioid histology, high-grade disease, Black race, and lack of surgery were independently associated with worse survival. These findings underscore important age- and race-based disparities and highlight the need for tailored management and survivorship strategies in younger patients.

Treatment patterns and outcomes in invasive lobular versus ductal breast carcinoma: A real-world analysis.

e12665 Background: Invasive lobular carcinoma (ILC) and invasive ductal carcinoma (IDC) differ in tumor biology; however, current clinical guidelines frequently recommend similar treatment strategies. Real-world comparative data across the full spectrum of disease stages remain limited. We compared clinicopathologic characteristics, treatment patterns, and outcomes between ILC and IDC in a real-world cohort. Methods: We conducted a retrospective cohort study of adults with invasive breast carcinoma treated at a tertiary cancer center in Colombia. Tumors were classified as ILC or IDC, and patients across all clinical stages were included. Clinicopathologic features, treatment patterns, and survival outcomes were compared between groups. Results: A total of 209 patients were included, of whom 31 (14.8%) had ILC and 178 (85.2%) had IDC. Median age was 56.3 years (IQR 46.3–66.0) in the ILC group and 58.0 years (IQR 51.8–67.6) in the IDC group. ILC tumors demonstrated lower proliferative activity, with a significantly lower median Ki-67 (10.0% [IQR 5.0–20.0] vs 27.5% [IQR 15.0–50.0], p &lt; 0.001). Neoadjuvant therapy was less frequently administered in ILC compared with IDC (6.5% vs 32.6%, p = 0.007), while upfront surgery was more common (90.3% vs 62.4%, p = 0.009). Use of adjuvant therapy was similar between groups (87.1% in both). After a median follow-up of 55.6 months, overall survival (OS) and progression-free survival (PFS) were comparable between histologic subtypes. Median OS was 55.9 months for ILC and 55.1 months for IDC, and median PFS was 49.8 and 51.9 months, respectively (p &gt; 0.9). These survival estimates should be interpreted in the context of a limited number of events. Conclusions: In this real-world cohort, ILC exhibited distinct biologic features and treatment patterns compared with IDC, including lower proliferative activity and less frequent use of neoadjuvant therapy, without observed differences in survival outcomes. These findings underscore the importance of considering histologic subtype in therapeutic decision-making and trial design. Baseline clinicopathologic characteristics by histologic subtype. Variable Overall cohort (n=209) Invasive ductal carcinoma (n=178) Invasive lobular carcinoma (n=31) Age in years, median (IQR) 58.2 (52-68) 56,3 (46-66) 58,0 (51,8-67,6) ER positive, n (%) 167 (79.9%) 137 (77.0%) 30 (96.8%) HER2 3+, n (%) 31 (14.8%) 31 (17.4%) 0 (0.0%) Ki-67, median (IQR) 25 (10- 50) 27.5 (15-50) 10 (5-20) Histologic grade 3, n (%) 57 (27.3%) 54 (30.3%) 3 (9.7%) Classic LCIS, n (%) 9 (4.3%) 1 (0.6%) 8 (25.8%) AJCC stage (I–II vs III–IV), n (%) 145 (69.4%) vs 44 (21.1%) 126 (70.8%) vs 39 (21.9%) 19 (61.3%) vs 5 (16.1%) OS, months (IQR) 55.6 (38.8-77.4) 55.1 (40.5-76.1) 55.9 (9.0-88.5) PFS, months (IQR) 51.3 (34.0-75.6) 51.9 (35.3-74.9) 49.8 (9.0-82.3)

Residual cancer burden after neoadjuvant therapy in lung cancer: A single-institution cohort.

e20050 Background: Residual cancer burden (RCB) after neoadjuvant therapy may predict recurrence and survival, but real-world data in lung cancer are limited. We evaluated RCB, clinical characteristics, and outcomes in a single-institution cohort. Methods: We retrospectively analyzed 53 patients (RCB classes pCR,1,2,3) treated with neoadjuvant therapy. Baseline demographics (age, sex, BMI, tobacco history), recurrence, and overall survival were compared across RCB classes. Cox proportional hazards models assessed the association of RCB with mortality, adjusting for age, sex, BMI, and tobacco history. Results: Baseline characteristics were similar across RCB classes. Recurrence rates increased with RCB class (pCR: 14.3%, 1: 20%, 2: 10%, 3: 25%). Mortality also trended higher with increasing RCB (pCR: 7.1%, 1: 20%, 2: 13.3%, 3: 50%). In fully adjusted Cox models, RCB 3 vs pCR showed a non-significant trend toward higher mortality (HR 4.98, 95% CI 0.49–50.7, p = 0.175). Age was independently associated with mortality (HR 1.10 per year, 95% CI 1.01–1.19, p = 0.025). Conclusions: Higher RCB may be associated with increased mortality in lung cancer, though statistical significance was not reached, likely due to small sample size. Age was an independent predictor of mortality. These findings suggest that RCB could inform post-neoadjuvant risk stratification, but larger studies are needed to confirm its prognostic value.