Browse Articles

An early on-treatment "coordinated inflammatory niche" plasma signature as a predictor of response in the SYLT026 trial: Cadonilimab combined with FOLFIRINOX and bevacizumab as first-line treatment for microsatellite-stable/proficient mismatch repair (MSS/pMMR) metastatic colorectal cancer (mCRC).

3600 Background: The SYLT026 trial is a phase II study investigating the efficacy and safety of cadonilimab (an anti-CTLA-4/PD-1 bispecific agent) combined with FOLFIRINOX and bevacizumab as first-line treatment for microsatellite-stable/proficient mismatch repair (MSS/pMMR) metastatic colorectal cancer (mCRC). We previously reported the promising outcomes at ASCO GI 2026. Updated survival data are reported here. The determinants of effective immune responses in MSS/pMMR mCRC remain elusive; the "Coordinated Inflammatory Niche" plasma signature was analyzed to predict the immune checkpoint inhibitor (ICI) response. Methods: Plasma samples from eligible patients were collected before ICI treatment and profiled using three non-overlapping protein arrays comprising 120 human cytokines (QAH-CAA-2000). These cytokines were integrated into an Inflammatory Niche Index (INI) using machine learning (random forest algorithm). Its predictive value for durable PFS ( > 12 months) was evaluated and compared with single markers. Results: Between November 16, 2023, and July 26, 2024, 20 patients were enrolled. The ORR was 100% (all partial responses). As of the data cutoff (December 31, 2025), the median follow-up was 24.4 months (IQR 22.4–26.4), the median PFS (mPFS) was 17.2 months (IQR 13.6–20.7), and the OS maturity rate was 20%. Seventeen patients underwent cytokine array analysis. Cytokine profiling identified a 9-protein panel (IP-10, eotaxin, MCP-1, IGFBP-3, MCP-4, TARC, IL-17, CCL28, I-TAC) that was significantly upregulated in responders. Further random forest modeling revealed the importance ranking of each protein, with TARC and IP-10 as the two most significant predictors, consistent with their most pronounced differential expression in the heatmap. The INI demonstrated superior predictive accuracy (AUC = 0.944) compared with IP-10 alone (AUC = 0.71, p = 0.03). Patients with a high INI (n = 9) had promising outcomes: mPFS was not reached (12-month PFS rate 90%), compared with 6.5 months in the low INI group (n = 8) (HR = 0.16, p = 0.001). Gene set enrichment analysis linked a high INI to a baseline tumor microenvironment enriched for NK T cell-related signatures, suggesting a biological basis for the niche. Conclusions: CTLA-4/PD-1 immunochemotherapy induced deep and durable responses in MSS/pMMR mCRC. An early on-treatment plasma proteomic signature (INI) powerfully identifies patients destined for durable benefit, offering a transformative tool for ultra-early, biology-guided patient selection in MSS/pMMR mCRC. Clinical trial information: NCT05839470 .

Non-invasive characterization of intratumoral CD8+ T cells using standard-of-care (SOC) CT and ⁸⁹Zr-crefmirlimab berdoxam PET (CD8-PET) radiomic signature in solid tumors.

2614 Background: Noninvasive estimation of intratumoral CD8+ T-cell density using radiomics may enhance comprehensive immune profiling and support immuno-oncology development and treatment decisions. We evaluated the feasibility of using a multimodality radiomic approach incorporating SOC CT and CD8-PET/CT to characterize CD8+ T-cell density in solid tumor lesions. Methods: 71 soft-tissue lesions from 52 patients with solid tumors enrolled in a Phase II iCorrelate trial (NCT03802123) were retrospectively analyzed. Patients received a pre-treatment CD8-PET/CT scan (ImaginAb, Inc) before initiating SOC immunotherapy (IOT) and a second CD8-PET/CT scan on-treatment 4-8 weeks later, with selection of biopsied tumor lesions of known CD8+ T-cell density for radiomic analysis. Radiomic features were extracted from volumetric segmentations of these tumor lesions using 3D Slicer on Diagnostic contrast CT (DCT, n=27), CT attenuation correction (CTAC; n=71), and CD8 PET (n=71) images. Features were integrated with acquisition parameters, lesion location, and adenopathy status. Feature selection used maximum relevance–minimum redundancy and variance inflation factor methods. Elastic-Net classifiers were trained to predict binarized (median split at the 328 cells/mm 2 level) CD8+ T-cell density, with hyperparameter optimization via grid search and 5 fold Cross-Validation (CV). In each iteration of the CV, 1 fold (20%) was left out and models were trained on the remaining 4 folds (80%). Model performance was assessed using AUC and F1 score. Results: In the test set, Multimodality models (CTAC+DCT+PET) outperformed single-modality (AUC up to 0.94 vs. 0.80), with improved AUC and F1 scores and robust performance across endpoints. Single-modality models achieved AUCs of 0.76 (F1 0.55) for DCT, 0.80 (F1 0.72) for CTAC, and 0.80 (F1 0.70) for PET. The combined CTAC+DCT+PET model demonstrated superior performance with a test AUC of 0.85 (F1 0.79) and training AUC of 0.94 (F1 0.89). Application of previously reported DCT radiomic weights (Sun et al.) to this cohort yielded a lower AUC of 0.68, compared with the multimodality model. Conclusions: This study demonstrates robust radiomics-based characterization of intratumoral CD8+ T-cell density in patients with solid tumors is feasible using non-invasive multi-modality approach utilizing CD8-PET and SOC CT scans. These findings may lead to further clinical adoption and non-invasive monitoring of the tumor immune microenvironments with CD8-PET and SOC CT scans using radiomics, informing IOT decision-making, and enhancing patient stratification in both clinical development and routine oncology practice. Validation in larger, multicenter cohorts is still required to confirm generalizability and robustness of these radiomic signatures. Clinical trial information: NCT03802123 .

Reproductive and pregnancy-related outcomes among adolescent and young adult female cancer survivors (15–39 years): A nationwide population-based study in South Korea.

12104 Background: As survival rates among adolescent and young adult (AYA) female cancer patients improve, increasing numbers of survivors pursue pregnancy and childbirth. However, population-level data on reproductive outcomes and pregnancy-related complications among female cancer survivors remain limited. Methods: This study collected de-identified data from the South Korean National Health Insurance Service Database. Among 95,264 women aged 15–39 years newly diagnosed with cancer between 2007 and 2010, we identified 58,107 eligible survivors. This cohort was age-matched 1:10 with 719,213 women without cancer. Reproductive outcomes were assessed in the full cohort, and pregnancy-related complications were evaluated among women with documented pregnancies. Odds ratios (ORs) and 95% confidence intervals (CIs) were estimated using logistic regression. Results: Compared with matched controls, female cancer survivors had a higher risk of infertility (OR 1.25; 95% CI 1.21–1.28) and lower odds of pregnancy (OR 0.94; 95% CI 0.93–0.96) and childbirth (OR 0.81; 95% CI 0.78–0.84). Among women who became pregnant, cancer survivors experienced higher risks of ectopic pregnancy (OR 1.20; 95% CI 1.11–1.29), miscarriage or abortion (OR 1.23; 95% CI 1.14–1.32), intrauterine infection (OR 1.26; 95% CI 1.01–1.57), and preterm labor (OR 1.21; 95% CI 1.14–1.29). Conclusions: AYA female cancer survivors face increased risks of infertility and adverse pregnancy-related outcomes. These findings highlight the need for specialized reproductive counselling and high-risk obstetric care for female cancer survivors. Reproductive outcomes in the full matched cohort and pregnancy-related outcomes among the pregnancy cohort. Outcomes, No. (%) Survivor (N=58,107) Control (N=719,213) OR 95% CI Reproductive Outcomes Infertility 5,439 (9.4) 55,111 (7.7) 1.245 1.209 1.282 Pregnancy 16,974 (29.2) 218,901 (30.4) 0.944 0.926 -0.961 Childbirth 10,912 (18.8) 98,229 (20.8) 0.810 0.780 -0.842 Pregnancy-related Outcomes Ectopic pregnancy 783 (4.6%) 8,495 (3.9%) 1.198 1.111 -1.291 Miscarriage / abortion 866 (5.1%) 9,182 (4.2%) 1.228 1.143 -1.319 Preeclampsia / eclampsia 507 (3.0%) 6,488 (3.0%) 1.008 0.920 -1.105 Gestational diabetes 3,139 (18.5%) 40,317 (18.4%) 1.055 0.965 -1.046 Intrauterine infection 85 (0.5%) 872 (0.4%) 1.259 1.007 -1.574 Premature Rupture of Membranes / amniotic fluid disorders 2,131 (12.6%) 28,002 (12.8%) 0.978 0.933 -1.025 Placental disorders 238 (1.4%) 2,776 (1.3%) 1.107 0.969 -1.265 Preterm labor 1,168(6.9%) 12,601 (5.8%) 1.210 1.137 -1.287 CI, confidence interval; No., number; OR, odds ratio.

Trends and burden of retinoblastoma incidence, prevalence, and mortality in the United States, 1990–2023: A retrospective analysis with machine learning projections to 2050.

e22609 Background: Retinoblastoma, a rare intraocular malignancy primarily affecting children, has seen declining age-standardized burden in high-income settings due to improved early detection, treatment, and survival. This study examined trends in age-standardized DALYs, deaths, incidence, and prevalence among US adults (≥20 years) from 1990 to 2023 using IHME GBD 2023 data, calculated estimated annual percentage changes (EAPC), and projected future rates to 2050 via ARIMA forecasting. Methods: Age-standardized rates (ASRs) per 100,000 population were extracted from the Global Burden of Disease 2023 database for the United States, stratified by sex (Both, Female, Male). Historical trends (1990–2023) were quantified using EAPC from linear regression on log-transformed ASRs. Future projections (2024–2050) utilized ARIMA models fitted to historical time series, producing point estimates and 95% prediction intervals (PI). Results: From 1990 to 2023, age-standardized DALYs (Both sexes) declined markedly from 0.823 (95% UI: 0.665–1.012) to 0.391 (95% UI: 0.263–0.571), yielding an EAPC of -2.35% (95% CI: -3.08 to -1.63); females exhibited a steeper decline (EAPC -2.91%, 95% CI: -3.77 to -2.04) than males (EAPC -1.92%, 95% CI: -2.55 to -1.27). Age-standardized death rates fell from ~0.0088 to ~0.0041 (EAPC -2.47%, 95% CI: -3.22 to -1.71), with stronger reductions in females (EAPC -3.02%) than males (EAPC -2.03%). Incidence decreased from 0.106 to 0.078 (EAPC -1.19%, 95% CI: -1.74 to -0.64), and prevalence from 0.972 to 0.683 (EAPC -1.23%, 95% CI: -1.75 to -0.70), reflecting improved survival and reduced new cases in adults. ARIMA forecasts indicate continued substantial declines in age-standardized DALYs (Both: from 0.372 in 2024 toward near-zero values by 2050), deaths (approaching zero by mid-century), incidence (stabilizing near 0.074 then trending downward), and prevalence (gradual erosion to low levels by 2050), driven by historical momentum and near-elimination of adult-onset cases. Conclusions: Age-standardized burden of retinoblastoma in US adults declined sharply from 1990–2023 across all metrics, with the steepest reductions in DALYs and mortality, particularly among females. Projections to 2050 suggest near-eradication of age-standardized impact in adults, underscoring the success of pediatric retinoblastoma management and the rarity of adult presentations in high-resource settings. Measure Sex EAPC Lower 95%CI Upper 95%CI DALYs Both -2.35 -3.08 -1.63 DALYs Female -2.91 -3.77 -2.04 DALYs Male -1.92 -2.55 -1.27 Deaths Both -2.47 -3.22 -1.71 Deaths Female -3.02 -3.9 -2.12 Deaths Male -2.03 -2.7 -1.36 Incidence Both -1.19 -1.74 -0.64 Incidence Female -1.75 -2.44 -1.05 Incidence Male -0.76 -1.24 -0.27 Prevalence Both -1.23 -1.75 -0.7 Prevalence Female -1.77 -2.44 -1.09 Prevalence Male -0.81 -1.26 -0.36

Efficacy and safety of engineered TCR-T cell therapy (afami-cel, lete-cel) in pediatric synovial sarcoma.

11569 Background: Synovial sarcoma (SyS) represents ~ 5% of adult soft tissue sarcomas, however, SyS is the most common non-rhabdomyosarcoma soft tissue sarcoma in the pediatric population. Both lete-cel, an engineered TCR-T cell therapy targeting NY-ESO-1, and afami-cel, an engineered TCR-T cell therapy targeting MAGE-A4, have shown efficacy and an acceptable safety profile in adult patients with SyS. We report here the largest cohort of pediatric SyS patients treated with either afami-cel or lete-cel. Methods: Pediatric SyS patients (pts) were treated on four phase II trials: lete-cel SyS Pilot study (NCT01343043), lete-cel IGNYTE-ESO trial (NCT03967223), afami-cel SPEARHEAD-1 (NCT04044768) and SPEARHEAD-3 (NCT05642455). Initial data from SPEARHEAD-3 will be available and added at the time of presentation. Eligibility for this analysis: advanced (unresectable/metastatic) SyS, human leukocyte antigen (HLA)-A*02, NY-ESO-1 or MAGE-A4 expression, baseline ECOG 0–1 and received any dose of lete-cel or afami-cel as a single infusion. Key efficacy endpoints: independent review-assessed overall response rate (ORR) per RECIST v1.1, best overall response, duration of response (DoR). Key safety endpoints: adverse events (AEs), serious AEs, AEs of special interest (ex. Cytokine release syndrome (CRS)). Data is presented separately for lete-cel and afami-cel. Results: Seven pediatric pts (age 10 to 17) were treated with lete-cel for advanced disease. The median lete-cel dose was 4.9 x 10 9 transduced cells. The ORR was 43% (3/7 pts, 2 PR, 1 CR). The duration of response was 2.83, 7.39, and 7.39 months respectively; one response was ongoing at the time of this analysis. Pediatric pharmacokinetics and exposure-response data were in agreement with data from adult pts. Seven pediatric pts (age 13 to 16) were treated with afami-cel for advanced disease. The median afami-cel dose was 8.2 x 10 9 transduced cells. The ORR was 28.6% (2/7 pts, 2 PR). The duration of response was 4.7 and 17.4 months respectively; one response was ongoing at the time of this analysis. Similar pharmacokinetics were observed in pediatric pts compared to adult pts. CRS was common occurring in >50% of pts for both lete-cel and afami-cel; events were grade 1 or 2, except one patient with lete-cel experiencing grade 3 CRS. Cytopenias after lymphodepletion and T cell infusion were the most common grade 3 or 4 adverse events (>30% of pts) for both therapies. Conclusions: Engineered TCR-T cell therapy targeting NY-ESO-1 (lete-cel) or MAGE-A4 (afami-cel) has shown responses and durable responses in the pediatric SyS population. The safety profile in this age group is consistent with that observed in the adult population. TCR-T represents a potential new treatment paradigm for pediatric SyS patients. Further studies should help to determine the optimal timing for this cellular therapy. Additional Information: Coauthor Brian Van Tine, MD, died Nov 2025. Clinical trial information: NCT01343043 , NCT03967223 , NCT04044768 , NCT05642455 .

Inflammatory markers as predictors of pathological complete response in patients with breast cancer after neoadjuvant chemotherapy.

e12653 Background: Breast cancer is the most common cancer and the leading cause of cancer-related death in women. Pathological complete response (pCR) is associated with better overall survival and progression-free survival. Inflammation plays an important role in the occurrence and development of tumors, and in previous studies there is information that inflammatory markers such as the neutrophil-lymphocyte ratio (NLR), the lymphocyte-monocyte ratio (LMR), the platelet-lymphocyte ratio (PLR), the absolute lymphocyte count (ALC), the systemic inflammatory response index (SIRI), the systemic immune inflammation index (SII) and even the body mass index (BMI) can predict response in different types of cancer. The present study evaluated the expression of these inflammatory markers in breast cancer patients who received neoadjuvant chemotherapy to determine their potential as predictive biomarkers of pathological complete response. Methods: We retrospectively analyzed the clinicopathological parameters and pretreatment peripheral blood characteristics of breast cancer patients who received neoadjuvant chemotherapy between January 2018 and December 2023 and correlated these with the pathological findings after surgery. Data was collected from patients’ electronic health records. Patients with conditions that could act as confounding factors in the analysis of inflammatory markers, such as prior inflammatory diseases and pregnancy were excluded. Receiver Operating Characteristic (ROC) curves were used to determine the cutoff of NLR, LMR, PLR, ALC, SII, SIRI and BMI and the value with the highest Youden index was defined as the optimal cutoff. The chi-square test was used to explore the relationship between inflammatory markers with pCR and other clinicopathological parameters. Multivariate analyzes were performed using the logistic regression model. Results: A total of 263 patients were included during the period. The median age was 52 years; most patients were female (99.62%), with only one male patient (0.38%). Most tumors were hormone receptor–positive (67.79%) and HER2-negative (76.4%). Among the 267 tumors analyzed, 86 (32.21%) achieved pCR and this was significantly related to a higher ALC. Compared to patients with ALC ≤ 2985/mm 3 , patients with ALC > 2985/mm 3 had a 2.5 times greater chance of achieving pCR (odds ratio = 2.578; 95% CI: 1.146-5.798, p = 0.022). The other inflammatory markers were not significantly associated with pCR. Multivariate analysis indicated that a higher ALC, Ki67 > 20% and the HER-2 status were independent prognostic factors for pCR in breast cancer patients. Conclusions: Pretreatment ALC is predictive of pCR in patients with breast cancer receiving neoadjuvant chemotherapy and should be taken into consideration when discussing the treatment plan, as the index may aid in the formulation of personalized treatment strategies.

Study protocol for OtoSurg 1: A prospective evaluation of worldwide tonsillectomy indications, techniques, and outcomes

Background Tonsillectomy is the most common pediatric surgery performed globally, yet little is known about how tonsillectomy outcomes vary worldwide. Preliminary Data Although generally considered a safe surgery, tonsillectomy carries significant risks. Approximately 3% of US children require hospital readmission within 30 days due to life-threatening complications. While mortality rates in high-income countries (HICs) are relatively low (~0.0005%), mortality rates in low- and middle-income countries (LMICs) are not well characterized, with some reports suggesting rates many times higher (~3%). Research gap Despite the frequency and significance of tonsillectomy, it remains unclear how surgical indications, techniques, and settings may impact outcomes worldwide. In addition, there is currently no mechanism to perform a prospective investigation of global tonsillectomy outcomes. Hypotheses and specific aims We hypothesize that the morbidity and mortality of routine tonsil surgery vary significantly according to geographic region and healthcare setting, as well as surgical indications and techniques. We aim 1) to quantify global variation in 30-day major postoperative complications and mortality following pediatric tonsillectomy across World Health Organization (WHO) regions and 2) to characterize global variation in pediatric tonsillectomy surgical indications and operative techniques across WHO regions. Design OtoSurg 1 will be an international, multi-site, prospective cohort study. The central IRB is approved and hosted by Emory University (IRB ID: STUDY00009113). All healthcare facilities globally that perform pediatric tonsillectomy will be invited to participate. Participating investigators will be trained to collect data on patients undergoing primary tonsillectomy over a 90-day period at their respective institutions. Site-based teams will record demographic data, surgical indication(s), surgical technique(s), postoperative complications, and mortality directly into a global de-identified tonsillectomy outcomes data registry. We will leverage existing relationships and professional networks to recruit participants from 30 sites in each of the six WHO regions for a target of 180 partner sites. Potential impact This study will highlight potential opportunities for intervention to standardize and improve outcomes for the world’s most common pediatric surgery, focusing on LMICs. Beyond its primary objectives, this project will also seek 1) to recruit and train a collaborative research network of otolaryngology surgeons from around the world and 2) to build the digital infrastructure necessary to support this network sustainably. This collaborative research network will serve as the foundation for future large-scale research initiatives and for the development and implementation of novel, data-driven interventions.

Hot carrier transfer in Bi2Se3/Te heterostructure promoted by substrate-introduced effective field

Van der Waals heterostructures with negligible lattice mismatch and flexible fabrication provide a platform to modulate the photoactive properties and extend the applications of two-dimensional materials. In addition to the band alignment, the substrate-introduced effective electric field is also indispensable for manipulating the carrier dynamics at the heterointerface. Here, we have proposed a heterostructure composed of tellurium (Te) and bismuth selenide (Bi2Se3) nanofilms to regulate the influence of the substrate field via the stacking order. Our experimental results demonstrate that the terahertz transient photoresponses of the two types of heterojunctions exhibit distinct modulation depths and carrier lifetimes. Notably, unlike the Te/Bi2Se3 configuration, where the Bi2Se3 interlayer serves as a spacer to screen the substrate field, the Bi2Se3/Te heterostructure possesses faster relaxation due to hot carrier transfer promoted by the substrate-introduced effective field. This work provides in-depth physical insights into substrate engineering, which is crucial for further exploration of terahertz optoelectronic devices.

Photodegradation of microplastics using graphitic carbon nitride (g-C3N4) based nanocomposite materials: A recent development on environmental remediation and bibliometric study

Compact‐Type Quasi‐2D Perovskites MAPbBr ₃ @FABr: Reduced Interlayer Distances Enable Ultralow Threshold Lasing and High Electron Mobility

ABSTRACT Quasi‐2D perovskites have garnered significant attention as promising materials for laser applications due to their superior optical gain and environmental stability. However, the thick insulating spacer layers between semiconducting perovskite slabs impede efficient charge transport, constraining their overall performance. Herein, we introduce a new family of compact‐type quasi‐2D perovskites with substantially reduced interlayer spacing. This design promotes rapid energy funneling, yielding an amplified spontaneous emission (ASE) threshold of 1.82 µJ cm −2 , the lowest reported for green‐emitting quasi‐2D perovskites, and an optical gain coefficient of 413.77 cm −1 . Furthermore, the short interlayer distance, coupled with precisely aligned semiconducting layers, yields electron mobilities exceeding those of traditional counterparts by more than one order of magnitude. By varying the halogen composition from bromine to chlorine, we produced mixed‐halide variants with tunable, low‐threshold ASE across a broad spectral range. We also fabricated high‐quality MAPbBr 3 @FABr micro‐ring laser arrays, exhibiting whispering‐gallery‐mode lasing with a quality factor of ∼1802, thresholds of ∼1.89 µJ cm −2 and good operational stability. This work advances the design of quasi‐2D perovskites, providing a pathway for development of practical perovskite lasers with enhanced performance and broader applications.

Abdominal aortic calcification is associated with hip fracture in the elderly

The chromatin reader ZMYND8 recruits the NuRD component GATAD2A through its MYND domain to regulate MAPT213 long noncoding RNA transcription

Integrated AI-informed morphologic and spatial transcriptomics features from H&E slides to enable risk stratification of primary melanoma patients.

9559 Background: Tumor proliferative activity and the tumor-stroma microenvironment are key determinants of melanoma aggressiveness and prognosis, yet their assessment from routine histology remains limited and subjective. Mitotic activity and stromal composition are histologic markers of tumor aggressiveness associated with overall survival (OS). While spatial transcriptomics (ST) can capture tumor biology associated with prognosis, its cost and limited availability restrict clinical use. Recent advances in computational pathology enable inference of spatial gene expression directly from routine hematoxylin and eosin (H&E) whole slide images (WSIs). We evaluated whether integrating AI-informed histologic biomarkers with ST features derived from routine H&E slides improves OS prediction in primary melanoma. Methods: Formalin-fixed paraffin-embedded H&E WSIs from TCGA-SKCM primary melanoma cases with available survival data were analyzed (N=291). Automated mitotic index (MI) in tumor region was quantified using a Detectron2-based mitosis detection model, and tumor-stroma ratio (TSR) was computed from automated tissue segmentation. Spatial gene expression for 18 melanoma relevant genes was inferred from WSIs using a pretrained DeepSpot model, and quantified for spatial heterogeneity and clustering. Three Cox regression models were created: i) M Path+Clinical (MI, TSR, age), ii) M AI-ST+Clinical (ST features, age), and iii) M I ntegrated (MI, TSR, ST features and age). Feature selection was performed on training cohort (N=174), and model performance for 10-year OS prediction was evaluated on holdout test cohort (N=117) using hazard ratio (HR) with 95% confidence interval, C-index, and log-rank tests. Independent contributions of various features were assessed by multivariable analysis. Results: In hold-out test cohort, both M Path+Clinical model (C-index = 0.639, p = 0.001, HR = 2.71 [1.49-4.94]), and M AI-ST+Clinical model (C-index = 0.603, p = 0.002, HR = 2.51 [1.37-4.59]) were prognostic for 10 year OS. AI-ST signature consisted of spatial heterogeneity and clustering metrics of ARG1, TFAP2A, and KRT6B genes, along with global gene expression metrics. The integrated model M I ntegrated resulted in the best OS stratification (C-index = 0.644, p = 0.007, HR = 2.24 [1.23-4.07]). Multivariable analysis demonstrated independent association of MI, ST features and age with OS. Conclusions: AI-derived ST features contributed additional prognostic value beyond histologic and clinical features, yielding the strongest OS stratification when integrated into a unified model. By combining established histologic biomarkers with virtual spatial omics, this approach improves survival prediction without additional tissue or molecular assays, supporting its potential clinical utility for risk adapted management of melanoma patients.

Barriers to integrating obesity management into cancer survivorship care: Results from a single-institution survey.

e13783 Background: Cancer survivors with obesity are at increased risk of cancer recurrence, cancer-related mortality, and secondary cancer development. Evidence-based obesity treatments such as glucagon-like peptide-1 receptor agonists (GLP-1RAs) and bariatric surgery reduce these risks yet remain profoundly underutilized in survivorship care. We surveyed oncology clinicians to characterize barriers to adoption and identify potential implementation solutions. Methods: We conducted a cross-sectional, anonymous web-based survey of oncology clinicians at a single academic institution, purposively sampling across specialties to ensure broad representation, including surgical oncologists, medical oncologists, and advanced practice practitioners. Likert-scale items assessed clinician knowledge, comfort with counseling and referral, and perceived patient awareness, while short-answer items elicited barriers to integrating GLP-1RAs and bariatric surgery in survivorship care. Quantitative responses were summarized using descriptive statistics, and qualitative responses were reviewed to identify common themes. Results: Twenty-eight clinicians completed the survey (76% response rate). Respondents reported low utilization of these therapies, with 84% reporting they never or only sometimes discussed bariatric surgery or GLP-1RAs with patients, and 63% reporting they had never referred a patient for either therapy. Respondents also described substantial knowledge and confidence gaps: 79% reported low understanding of the oncologic benefits of these treatments, 84% reported low confidence discussing them in practice, and 76% perceived that cancer survivors with obesity were not at all or only slightly aware of their benefits. Despite these gaps, 79% of respondents agreed that bariatric surgery and GLP-1RAs should be a standard component of survivorship care. Reported barriers included unclear referral pathways (71%), limited knowledge of eligibility criteria (54%), and lack of support or resources (46%). Across candidate implementation strategies, 40-60% of respondents rated clarified referral pathways, decision support, and enhanced education/training as “highly helpful.” Conclusions: Oncology clinicians endorsed GLP-1RAs and bariatric surgery as an important component of survivorship care, but reported low utilization driven by knowledge gaps and system-level barriers. These findings highlight the need for multi-level, multi-component implementation strategies to support the integration of evidence-based obesity treatment into cancer survivorship care.

Prospective evaluation of AI-assisted documentation and summarization in virtual multidisciplinary cancer conferences.

1614 Background: Multidisciplinary cancer conferences (MCCs) are central to high-quality oncologic decision-making but are frequently constrained by time-intensive and variable documentation processes. Artificial intelligence (AI) can enhance multidisciplinary collaboration by improving the accuracy, timeliness, and consistency of documentation. However, prospective evaluations of AI tools embedded within MCC workflows remain limited. Methods: We conducted a prospective pilot study at the Cancer Centre of Southeastern Ontario, a regional academic cancer center in Ontario, Canada. In October 2025, a hospital-approved AI system, Microsoft Copilot, was integrated into all virtual lung and gastrointestinal (GI) MCCs conducted via Microsoft Teams. All conferences were recorded and transcribed. Structured prompts for AI-generated summaries were co-developed with MCC stakeholders and refined through iterative Plan-Do-Study-Act (PDSA) cycles. AI-generated summaries were validated against clinician-prepared electronic health record (EHR) documentation. Two independent reviewers assessed case identification accuracy and concordance of key clinical content. Following validation, AI summaries were distributed to MCC participants after case discussions to support EHR documentation, with mandatory clinician verification prior to finalization. Results: A total of 125 patient cases from lung and GI MCCs were analyzed, of which 108 had corresponding clinician-prepared EHR documentation available for comparison. Inter-rater agreement between reviewers assessing concordance of AI-generated and clinician-prepared summaries demonstrated high agreement (Cohen’s κ = 0.87). Case identification accuracy was high, with agreement rates of 97% for case number, 93% for medical record number, 96% for patient initials, 98% for presenter, 99% for diagnosis, and 99% for cancer stage; most identification discrepancies were minor and improved over time. Clinical content concordance was high, including discussion summaries at 99%, radiology at 99%, pathology at 98%, and treatment plans and next steps at 96%. Major discrepancies involving key clinical decisions were identified in 0.9% of cases. All AI-generated summaries were distributed within 24 hours following MCC completion, with a median clerical preparation time of 14 minutes per case. Conclusions: In this prospective, real-world evaluation, AI-assisted summaries integrated into virtual MCC workflows demonstrated high accuracy, strong safety performance, and meaningful efficiency gains. With mandatory clinician oversight, AI-supported summarization was feasible and timely, supporting its role as a scalable model for MCC documentation across oncology programs. Further multicenter studies are warranted to evaluate generalizability, clinician workload impact, and downstream effects on care delivery.

Rare germline heterozygous pathogenic BLM variants and correlations with co-occurring pathogenic variants and severe treatment-related adverse events in breast cancer.

e12603 Background: The BLM gene encodes a RecQ family DNA helicase. Biallelic pathogenic mutations in BLM cause Bloom syndrome, and carriers are at an increased risk of various malignancies, including breast and colorectal cancer. However, data on the clinical features and prognosis of breast cancer patients harboring germline pathogenic BLM variants remain limited. Methods: From 13,269 breast disease patients who underwent germline genetic testing, we identified 12 cases with germline pathogenic BLM variants (BLM group) and selected 24 matched controls without such variants. We collected data on clinicopathologic features, co-occurring pathogenic variants, grade ≥3 adverse events (AEs, per CTCAE v5.0), and overall survival (OS). Statistical analyses were performed using Fisher's exact test and log-rank test. Results: In the BLM group, 30.0% (3/10) of patients carried concurrent pathogenic variants (BRCA1, BRCA2, or MLH3), a finding not observed in the control group (0/24), with a statistically significant difference (P = 0.020). Furthermore, the incidence of grade ≥3 treatment-related AEs was higher in the BLM group (70.0%, 7/10) compared to the control group (29.2%, 7/24), approaching statistical significance (P = 0.054). No statistically significant difference in median OS was observed between the two groups (log-rank P = 0.134). Conclusions: Despite the absence of biallelic BLM variants in this cohort, germline heterozygous pathogenic BLM variants are associated with a higher burden of co-occurring pathogenic variants and a clinically relevant trend toward elevated grade ≥3 treatment-related AEs, supporting the need for tailored clinical monitoring in this patient subset. Larger prospective studies are warranted to validate these findings and establish optimal clinical monitoring strategies for breast cancer patients with germline pathogenic BLM variants.

Tele-nutritional collaborative care in patients with stage IV gastrointestinal cancer receiving chemotherapy: A randomized clinical trial.

12043 Background: Cancer cachexia is characterized by systemic inflammation and progressive muscle loss, fundamentally compromising quality of life (QOL) and overall survival. While early nutritional intervention is imperative, patients in resource-constrained regions like Western China often lack timely access. We conducted the tele-nutrition collaborative care (T-NICE) trail to evaluate the efficacy of a hybrid remote care delivery model for patients with stage IV gastrointestinal cancer undergoing chemotherapy. Methods: This randomized clinical trial was conducted at West China Hospital. Patients with histologically confirmed stage IV gastrointestinal cancer (esophageal, gastric, or colorectal) were randomly assigned (1:1) to the T-NICE intervention versus Standard Nutrition Care (SNC). The T-NICE group received a collaborative hybrid telecare model involving a lean team (dietitian and social worker), comprising an initial in person education followed by app-based monitoring and adaptive telephone counseling. Primary Endpoint: The incidence of cancer cachexia (defined as involuntary weight loss >5%, or >2% if BMI <20 kg/m²) over 6 months. Secondary Endpoints: Longitudinal changes in nutritional status (PG-SGA, NRS 2002) and body weight. Exploratory end points included the QOL (using the EORTC QLQ-C30), psychology characteristics (using DT, PHQ-9 and HADS) and caregiver burden (using ZBI). Assessments were conducted at monthly intervals through the six months endpoint. Results: A total of 160 patients were enrolled (80 to T-NICE, 80 to SNC). Cachexia incidence was significantly lower in T-NICE vs. SNC (22.4% VS. 60.8%; adjusted odds ratio [aOR], 0.18; 95%CI, 0.09-0.37; P < .001). After six months, weight stabilized in T-NICE group compared to loss in SNC (mean change, +0.86kg [95% CI, -0.20-1.93] VS −3.61 kg [95% CI, -4.70 to -2.52]), resulting in a mean difference of -4.56 kg (95% CI, -5.73 to -3.39; P < .001). The T-NICE group also had lower PG-SGA scores (5.10 [95%CI, 4.14-6.06] VS 6.50 [95%CI, 5.54-7.46]; P = .015). T-NICE participants had higher EORTC QLQ-C30 Global Health scores (74.42 vs. 66.67; P = .004) and improved various functional status. Psychological benefits and reduced caregiver burden were also observed. Conclusions: The T-NICE trial demonstrated that an integrated tele-nutrition model significantly reduced cachexia incidence, improved QOL, nutritional status, psychology status and alleviated caregiver burden. Key strengths included the model’s scalability and low staffing requirements, maximizing resource efficiency while delivering substantial clinical benefits. Clinical trial information: NCT06332664 .

RP2 oncolytic immunotherapy alone and in combination with nivolumab (nivo) in patients with advanced solid tumors: Final safety, efficacy, and biomarker results from the phase 1 first-in-human (FIH) study.

2504 Background: RP2 is an HSV-1-based oncolytic immunotherapy expressing a fusogenic glycoprotein (GALV-GP-R – ), human GM-CSF, and a human anti–CTLA-4 antibody. We present final safety, efficacy, and biomarker data from the FIH study of RP2 in patients (pts) with advanced solid tumors (NCT04336241). Methods: Enrolled pts had advanced/metastatic solid tumors with ≥1 measurable, injectable lesion (≥1 cm) and had progressed following, or could not tolerate, available standard therapy. Pts received RP2 monotherapy via intratumoral injection in dose escalation (Part 1; 5 doses Q2W). Pts in Parts 2 (RP2 + nivo) and 3 (RP2 monotherapy) received up to 8 doses of RP2 (Q2W × 8 or Q2W × 4 followed by Q4W × 4 doses) at the RP2D. In Part 2, nivo was administered starting at Cycle 2 or 4 for ~20 to 22 months. Pts could receive an additional course of RP2 per protocol. Results: As of 01DEC2025, 85 pts (median age 59 y) were enrolled and treated in RP2 (n = 25) or RP2 + nivo (n = 60) cohorts. Tumor types included uveal or cutaneous melanoma (UM, n = 17; CM, n = 11); colorectal (n = 14), head and neck (n = 13), and pancreatic (n = 12) cancers; and sarcoma (n = 7); 42% of pts received prior immune checkpoint inhibitor treatment. All pts have completed treatment. Treatment-related adverse events (TRAEs) in > 10% of pts included pyrexia, chills, fatigue, influenza-like illness, hypotension, pruritus, and nausea. Grade ≥3 TRAEs occurred in 20% of pts (no single event occurred in > 2 pts). A majority of pts (58/85) received injections into deep/visceral lesions, mainly in liver and lung, which were well tolerated. Among pts with ≥1 post-baseline scan (75/85), the ORR was 17.3% (13/75; 1 unconfirmed) with a median DOR of 22.1 months and DCR (CR+PR+SD) of 44.0% (33/75). Responses occurred in 5/15 (33.3%) pts with UM. RP2 monotherapy resulted in confirmed responses in 4/21 (19%) pts (1 esophagogastric adenocarcinoma, 1 UM, 1 chordoma, 1 mucoepidermoid), with best overall response of CR and median DOR not reached. Tumor regression occurred in both injected and non-injected lesions, including all 3 pts with monotherapy responses with non-injected lesions. A robust intratumoral immune response with increased CD8+ T-cell infiltrates, PD-L1 upregulation, and inflammatory immune pathway activation were observed, as well as epitope spreading with novel virus- and cancer-associated T-cell clones. Conclusions: RP2 ± nivo treatment was well tolerated in pts with both superficial and deep visceral metastases. Durable responses were observed with RP2 monotherapy and RP2 + nivo in heavily pretreated pts with diverse tumors with evidence of systemic anti-tumor immune activity, including tumor reduction in distant non-injected lesions. This study supports the ongoing evaluation of RP2 in pts with UM (NCT06581406) and the planned evaluation in other solid tumors. Clinical trial information: NCT04336241 .

High-dose chemotherapy (HDCT) and peripheral-blood stem-cell transplant (PBSCT) for relapsed seminoma: A single-center cohort.

e17013 Background: The optimal salvage therapy for patients with relapsed seminoma remains unsettled. We evaluated clinical characteristics, outcomes, and toxicity in patients with relapsed seminoma treated with HDCT and PBSCT as second-line or subsequent therapy. Methods: We performed a retrospective analysis of consecutive patients with relapsed seminoma who received salvage HDCT and PBSCT at Indiana University between 2001-2026. HDCT consisted of two planned tandem cycles of high dose carboplatin and etoposide. Patient, disease, and treatment characteristics were abstracted from institutional records. Primary endpoints included progression-free survival (PFS) and overall survival (OS) after HDCT estimated with the Kaplan-Meier method. Secondary endpoints included the number of cycles of HDCT received, dose modifications, and grade ≥3 toxicities. Results: A total of 147 patients were included with a median age of 38 years (range 20-70) and median follow-up of 3.5 years from date of start of HDCT (range 0.0-19.1). Primary tumor sites were testicular (91.8%), retroperitoneal (6.8%), and mediastinal (1.4%). Metastatic sites included lungs in 13.6%, non-pulmonary visceral metastases (NPVM) in 21.1% (bone 12.9%, liver 6.1%, brain 4.1%), and lymph nodes in all 147 patients. HDCT was administered as second-line therapy in 81.0% and as ≥third-line therapy in 19.0%. 9.5% of patients had platinum refractory disease defined as progression within 4 weeks of first-line cisplatin-based chemotherapy. Both planned tandem HDCT cycles were completed in 91.2% of patients and dose reductions occurred in 10.9%. Grade ≥3 toxicity occurred in 57.8%, most commonly gastrointestinal (42.7%) and infectious (27.4%). The estimated 2-year PFS was 89.4% (95% CI 82.7%-93.6%) and 5-year PFS was 86.3% (77.9%-91.7%). The estimated 2-year OS was 93.1% (87.1%-96.4%) and 5-year OS was 87.6% (78.9%-92.3%). Five-year PFS for HDCT given as 2nd line compared to ≥3rd line was 90.2% (82.3%-94.6%) vs 70.8% (43.3%-86.8%). Five-year OS for HDCT given as 2nd line compared to ≥3rd line was 93.0% (85.7%-97.4%) vs 67.3% (39.1%-94.6%). Death occurred in 11.6%, with a median time from HDCT to death of 13 months (range 0-191). Treatment related death occurred in 2 patients (1.4%). At last follow-up, 85.7% of patients had no evidence of disease. Conclusions: HDCT + PBSCT is an effective salvage therapy for patients with relapsed seminoma. HDCT as initial salvage chemotherapy in seminoma provided unprecedented PFS and OS and remained effective as ≥3rd line therapy. Given the very high cure rates observed with HDCT alone, our experience suggests that routine use of standard-dose salvage chemotherapy prior to HDCT may not be necessary in appropriately selected patients.

Phase 1 study of venetoclax combined with a 10-day regimen of decitabine in subjects with high-risk acute myeloid leukemia.

6547 Background: We and others have reported the antileukemic activity of a 10-day regimen of the hypomethylating agent (HMA), decitabine (DEC), in high-risk acute myeloid leukemia (AML). Venetoclax (VEN), at a standard (std) dose of 400 mg, combined with HMA, at a std schedule, has improved AML outcomes. We aimed to investigate the safety of adding VEN to the 10-day regimen of DEC and explore the effects of BCL2 inhibition on the immune repertoire. Methods: This was a multicenter, investigator-initiated, dose-escalation study in adults with high-risk AML defined as, disease with a TP53 mutation, adverse-risk cytogenetics or relapsed/refractory (R/R). Patients (pts) were naïve to HMA and VEN. The primary endpoint was to establish the maximum tolerated dose (MTD) of VEN when combined with a 10-day DEC schedule. Dose-limiting toxicities (DLT) were evaluated during cycle 1. In the absence of a DLT, 3-6 evaluable pts were enrolled per dose level (DL) or until the 400mg DL of VEN was reached (Table 1). Peripheral blood mononuclear cells (PBMCs) were isolated before and after treatment. Shifts in the immune landscape and BCL2 family expression were profiled with spectral flow cytometry. Results: 14 pts were enrolled, with a median age of 66.5 years (range 19-77); including 9 pts (64%) with R/R AML, 8 (57%) with adverse-risk cytogenetics and 6 (43%) with a TP53 mutation. One pt had a DLT at DL1, due to grade (G) 4 neutropenia and persistent marrow aplasia in the absence of leukemia beyond day 42. One patient with R/R AML and pre-existing elevated alanine aminotransferase (ALT) secondary to known graft-versus-host disease developed G2 ALT elevation. 3 pts were then enrolled at DL-1, before reescalation back up to DL1. 2 pts at DL1 were deemed not eligible for DLT evaluation, requiring an additional 2 pts enrolled at DL1, without DLT, to confirm safety. The MTD was determined to be DL2 (400 mg). 5 pts (36%) had a complete remission (CR) or CR with incomplete count recovery (CRi). Non-hematologic or infectious G1-2 toxicities included fatigue (43%), diarrhea (43%) and decreased appetite (36%). 1 pt died of suspected sepsis outside of the DLT evaluation period. 7 pts had serial PBMCs available for multiparameter flow cytometric immune and BCL2 family profiling. Of these, 3 responders had strong induction of BCL2 in CD4+/CD8+ lymphocytes and B cells after 1 cycle and were enriched for a population of “senescent” CD8+ T cells that had relatively low expression of coinhibitory markers (PD-1, LAG3, TIM3). Conclusions: A 10-day DEC+VEN regimen is feasible in high-risk AML, with myelosuppression being the major DLT. The recommended phase 2 dose is 400mg VEN (DL2), but responses were also seen at 200mg. The immune profile of responders have demonstrated intriguing patterns that merit investigation in a larger HMA+VEN cohort. Clinical trial information: NCT03844815 . DL DEC dose(D1-10)mg/m2 VEN dose (D1-21)mg # of pts DLT CR CRi No Response -1 15 100 3 0 0 0 3 1 20 200 8 1 1 3 4 2 20 400 3 0 0 1 2