Browse Articles

Evaluation of the impact of GLP-1 receptor agonists (RAs) on stage of PDAC diagnosis: A retrospective matched cohort study.

e16462 Background: Pancreatic ductal adenocarcinoma (PDAC) is an aggressive malignancy that is most frequently diagnosed at either the locally advanced or metastatic setting. Most patients will not have the opportunity for curative-intent surgical resection. The early symptoms of pancreatic cancer, including weight loss, early satiety, and back and/or abdominal pain, provide clinical clues that can trigger a medical evaluation and diagnosis of PDAC. GLP-1 (glucagon-like peptide-1) receptor agonists (RAs) are widely used for type 2 diabetes and obesity management. However, the side effects of GLP-1 RAs include early satiety, diarrhea, and weight loss. This potential overlap in drug effects and cancer symptoms raises the question of delayed PDAC diagnosis in patients on GLP-1 RAs. We conducted a retrospective matched cohort study of patients diagnosed with PDAC to evaluate whether GLP1 RAs were associated with a later stage of diagnosis. Methods: Eighty-eight patients with PDAC on a GLP-1 RA at the time of diagnosis were matched 2:1 by age, ECOG status, and diagnosis date to 176 patients with PDAC who were not on a GLP-1 RA at the time of diagnosis. Extensive clinical information was gathered, including stage at diagnosis [resectable, borderline resectable (BRPC), locally advanced (LAPC), or metastatic (mPDAC)], diagnosis of diabetes, treatments, and outcomes. Risks of diagnosis at each stage were estimated using multinomial logistic regression. Results: GLP-1 RA users had 21.7 percentage point higher risk of diagnosis with BRPC than resectable PDAC, as compared to non-GLP-1 RA users (p = .035), and were 12.0 and 12.5 percentage points more likely to be diagnosed as LAPC or mPDAC than resectable disease, respectively. After weighting adjustment for diabetes mellitus diagnosis, GLP-1 RA users were 18.6 percentage points more likely to be diagnosed with BRPC vs. resectable disease (p = 0.09). 39% of non-GLP-1 RA users underwent curative-intent surgical resection compared to 27% of users (p = 0.065). Conclusions: In this small retrospective matched cohort study, we observed that patients who were taking GLP-1 RA agonists at the time of PDAC diagnosis were more likely to be diagnosed with non-resectable disease compared to a matched group of patients who were not taking GLP-1 RAs. This observation may be due to an overlap in symptoms between GLP-1 RA and PDAC, leading clinicians to attribute satiety and weight loss to the agent. Clinicians should be attentive to the symptoms of patients on GLP-1 RAs, particularly if symptoms are out of proportion to what is expected with the use of this drug class. Stage Difference in risk differences (GLP-1 minus non-GLP-1) Std. Err. p-value BRPC vs. resectable 0.217 0.080 0.035 LAPC vs. resectable 0.120 0.087 0.217 mPDAC vs. resectable 0.125 0.101 0.263

PRaG 5.0: Personalized thymosin alpha-1 and radioimmunotherapy for advanced refractory solid tumors—Data from an expanded cohort.

2516 Background: Treatment-related lymphopenia and inadequate anti-tumor immune activation remains major barriers to improving outcomes of combination immunoradiotherapy, especially for advanced refractory solid tumors. PRaG 5.0 innovatively integrates personalized thymosin alpha-1 (Tα-1) into the PRaG regimen (PD-1 inhibitor + radiotherapy + GM-CSF) to preserve lymphocytes and enhance anti-tumor immunity. We report updated efficacy and safety data of PRaG 5.0 in advanced refractory solid tumors. Methods: This is a prospective, single-armed, phase II study enrolled 43 heavily pretreated patients. Eligible patients were stratified by baseline T-lymphocyte counts to receive personalized Tα-1 dosing: 7-day loading dose for low baseline T-lymphocyte counts, and maintenance Ta-1 (1.6 mg, thrice weekly) for those with normal counts. Tα-1 was combined with the PRaG regimen for at least 2 cycles. After PRaG cycles, patients continued PD-1 inhibitor plus Tα-1 until disease progression or intolerable toxicity. The primary endpoint was objective response rate (ORR) per RECIST v1.1. Secondary endpoints included median progression-free survival (mPFS), disease control rate (DCR), safety, and immune cell dynamics (assessed by flow cytometry, single-cell sequencing, and TCR sequencing). Results: Among the 43 screened subjects, 39 received treatment, of whom 30 required loading dose thymosin α1 (Tα1) therapy, 36 were treated with PRaG regimen and 17 entered the maintenance phase with PD-1 inhibitor combined with Tα1. The objective response rate (ORR) of the 39 patients was 30.8% (95%CI 17.0%, 47.6%), including 4 cases of complete response (CR) and 8 cases of partial response (PR); the median progression-free survival (PFS) was 4.2 months (95%CI 2.3, 6.9). No Grade 3 or above TRAE was reported. Immune analysis revealed significant increases in CD8+ T cells, NK cells, and CD4+TEM cells post-Tα-1 loading, with a notable decrease in Tregs (p = 0.001). Single-cell sequencing demonstrated higher clonality in GZMK-CD8+T in responders, while TRDV2-CD8+T cells expanded in non-responders. Conclusions: Integrating of Tα-1 into the PRaG regimen is a promising therapeutic strategy, offering enhanced efficacy and a favorable safety profile. Detailed immune profiling identified distinct T-cell dynamics associated with treatment response, highlighting potential biomarkers for clinical application. Further validation in larger cohorts is warranted. Clinical trial information: NCT05790447 .

Phase II study of sintilimab combined with nab-paclitaxel in the treatment of platinum-resistant recurrent ovarian cancer.

5560 Background: Platinum-resistant recurrent ovarian cancer (PROC) remains a major therapeutic challenge with limited effective treatment options. This single-arm phase II trial aimed to evaluate the efficacy, safety, and tolerability of sintilimab combined with nab-paclitaxel in patients with platinum-resistant recurrent epithelial ovarian, fallopian tube, or primary peritoneal cancer. Methods: A total of 28 patients were enrolled and received intravenous sintilimab (200 mg, day 1) plus nab-paclitaxel (260 mg/m², day 1) every 3 weeks. The primary endpoint was objective response rate (ORR); secondary endpoints included progression-free survival (PFS), disease control rate (DCR), overall survival (OS), and safety. Results: The median age was 57 years (range: 37–74 years), with 96.4% having high-grade serous histology, 67.9% being BRCA wild-type, 57.1% with prior bevacizumab exposure, and 50% with a PD-L1 combined positive score (CPS) ≥1. After a median follow-up of 11.8 months, the ORR was 35.7% (10 partial responses), DCR was 64.3%, median PFS was 4.28 months, and median OS was 15.83 months. Patients with CPS ≥1 had a higher ORR (42.9% vs. 28.6%) and DCR (71.4% vs. 57.1%) than those with CPS <1, but no significant difference in PFS (3.83 vs. 4.13 months, p = 0.18). Similar trends were observed for CPS ≥5 and CPS ≥10, with ORRs of 44.4% and 50.0%, respectively, but no significant PFS differences (all p > 0.05).Notably, patients with tumor-infiltrating lymphocytes (TILs) ≥10% exhibited a significantly higher ORR (72.7% vs. 12.5%, p = 0.005) and longer median PFS (15.57 months vs. 3.4 months, p = 0.005). Treatment-related adverse events (TRAEs) were mainly leukopenia (28.6%), neutropenia (25.0%), anemia (46.4%), and peripheral neuritis (28.6%), with rare grade 3–4 toxicities. Conclusions: Sintilimab combined with nab-paclitaxel shows promising efficacy and acceptable safety in PROC, with TILs ≥10% and a high CD8⁺/CD4⁺ ratio emerge as potential predictive biomarkers, while PD-L1 CPS shows limited utility. This regimen represents a potential new therapeutic option for this difficult-to-treat population and warrants further investigation in randomized controlled trials. Clinical trial information: 2000031890.

Endocrine therapy recommendation after breast-conserving surgery for ER+ ductal carcinoma in-situ and impact on clinical outcomes.

e12619 Background: NCCN guidelines recommend consideration of 5 years of adjuvant endocrine therapy (ET) after breast conserving surgery (BCS) for ER+ DCIS to lower risk of recurrence. While rates of patient acceptance and adherence to ET are known to be low, rates of provider recommendation for ET are less understood. We aim to identify patient and tumor factors associated with provider recommendation for ET and variables associated with recurrence in patients with ER+ DCIS. Methods: Patients with ER+ DCIS treated with BCS at a single institution from 2017-2019 were included. Data was collected on clinicopathologic variables, local therapy, and recommendation and adherence to adjuvant therapy. Recommendation for ET was defined as referral to medical oncology and documented recommendation for ET. Univariate and multivariate logistic regression analyses were used to identify factors associated with recurrence. Results: Of 242 patients with DCIS,124 met inclusion criteria. Median follow up was 6.5 yrs. Most patients were non-Hispanic (79.8%), White (58.1%), and postmenopausal (70.2%) with median age of 64. Median DCIS span was 16 mm, 96.8% were unifocal, and 53.2% were intermediate grade. Most patients (70.2%) completed radiation therapy. ET was recommended to 108/124 (87.1%) patients: 46/108 (42.6%) declined, 37/108 (34.3%) took ET for 5 yrs, and 25/108 (23.1%) took ET for less than 5 yrs. Recommendation for ET was not associated with patient or tumor factors, however there was a trend toward omission of ET for current users of HRT (p = 0.09), and a trend toward ET recommendation for larger size of DCIS (p = 0.06). There were 11 (8.9%) recurrences at median 3.8 yrs, including 6 (54.5%) ipsilateral DCIS, 4 (36.4%) ipsilateral invasive carcinomas, and 1 (9.1%) contralateral invasive carcinoma. No patient experienced distant recurrence or death from breast cancer. Receipt of RT was associated with lower recurrence than no RT (3.4% vs 21.6%, p = 0.003). Use of ET for any duration was also associated with lower recurrence (0% vs 17.7%, p < 0.001). Among those who did not take ET, after adjusting for DCIS span and RT, those who declined ET were more likely to recur than those not offered ET (OR 7.94, 95% CI 1.08-176.0, p = 0.0409). Conclusions: In this series of patients with ER+ DCIS treated with BCS, providers recommended ET to a majority of patients, but uptake and adherence to ET was low with about one-third of patients completing 5 years of therapy. Any duration of ET appeared protective, and patients who declined ET were more likely to experience local recurrence than those who were not prescribed ET. While this suggests that providers seem to be safely omitting ET for some patients, future studies are needed to develop precise markers of recurrence risk that could be useful in identifying subsets of patients with ER+ DCIS who could benefit from shorter durations of ET and those who can safely omit ET altogether.

Cadonilimab, a bispecific anti–PD-1/CTLA-4 antibody, for patients with dMMR/MSI-H metastatic colorectal cancer after progression on anti–PD-(L)1 therapy: A multicenter, single-arm, phase 2 trial (CSWOG-C03).

3577 Background: Immune checkpoint blockade targeting PD-1 is the standard of care for patients with mismatch repair–deficient or microsatellite instability–high (dMMR/MSI-H) metastatic colorectal cancer (mCRC). However, approximately 37%–45% of patients experience disease progression within 1 year when treated with anti-PD-1 monotherapy, and effective subsequent treatment options remain undefined. The CSWOG-C03 study evaluated the efficacy and safety of cadonilimab, a bispecific anti-PD-1/CTLA-4 antibody, in patients with dMMR/MSI-H mCRC after progression on prior anti-PD-(L)1 therapy. Methods: This multicenter, single-arm, phase 2 study enrolled patients with centrally confirmed dMMR/MSI-H mCRC and iRECIST-defined disease progression after prior anti-PD-(L)1 monotherapy or anti-PD-(L)1-based combination therapy. Patients received intravenous cadonilimab at 6 mg/kg every 2 weeks until disease progression or unacceptable toxicity. The primary endpoint was the 12-month progression-free survival (PFS) rate, assessed by RECIST v1.1. Results: A total of 24 patients were enrolled and received study treatment. The median age was 54 years; 29% were female, 17% had Lynch syndrome, 92% had peritoneal metastases, and 33% had liver metastases. Overall, 63% of patients had received ≥2 prior lines of systemic therapy; all had prior exposure to anti-PD-(L)1 therapy, 71% had received fluoropyrimidines and oxaliplatin, and 29% had received irinotecan. The best response to prior anti-PD-(L)1 therapy was partial response in 2 patients, stable disease in 17, and progressive disease in 5. With a median follow-up of 14.8 months at data cutoff (January 22, 2026), the 12-month PFS rate was 44.6% (95% CI, 23.4–63.9), meeting the prespecified primary endpoint of 40%. Median PFS was 6.1 months (95% CI, 2.2–10.1). Median overall survival (OS) was not reached and the 12-month OS rate was 68.6% (95% CI, 42.5–84.7). The confirmed objective response rate (ORR) was 20.8% (95% CI, 7.1–42.2), and the disease control rate (DCR) was 75.0% (95% CI, 53.3–90.2). One patient completed the protocol-specified 2-year treatment without progression, and 7 remained on study treatment. Treatment-related adverse events of any grade occurred in 15 patients (63%), with grade 3 events observed in 3 patients (13%). Conclusions: Cadonilimab provided meaningful disease control with modest antitumor activity and a manageable safety profile in patients with dMMR/MSI-H mCRC after progression on prior anti-PD-(L)1 therapy. Clinical trial information: NCT05426005 . N=24 12-month PFS rate 44.6% (95% CI, 23.4–63.9) Best overall response PR 5 (21) SD 13 (54) PD 5 (21) NE 1 (4) Objective response rate 20.8% (95% CI, 7.1–42.2) Disease control rate 75.0% (95% CI, 53.3–90.2)

Small cell prostate cancer: Risks for metastatic disease.

e17036 Background: Small cell carcinoma of the prostate is a rare but highly aggressive subtype of prostate cancer, frequently presenting with distant metastases and poor survival outcomes. Despite its severity, population-level data examining predictors of metastatic disease at presentation remain limited. Understanding these factors may improve early detection and risk stratification. Methods: A retrospective, cross-sectional analysis was conducted using the SEER 17 registries (2000–2022). Patients with microscopically confirmed SCCP were identified using ICD-O-3 histology code 8041/3 and primary site code C61.9. Demographic, clinical, and metastatic variables were extracted. Univariate and multivariable logistic regression models were used to evaluate predictors of overall and site-specific metastases. Variables meeting a univariate threshold of p < 0.25 were included in multivariable models. Model assumptions were assessed using VIF, Tolerance, and Box–Tidwell tests. Results: A total of 541 patients were identified, of whom 71.7% presented with metastatic disease. The most common metastatic sites were bone (35.9%), liver (22.6%), lung (14.0%), and brain (3.9%). On multivariate analysis, a higher percentage of positive biopsy cores was independently associated with increased odds of metastatic disease. Younger age was associated with higher odds of brain metastasis. Lower household income and residence in metropolitan counties were associated with increased likelihood of metastatic disease, particularly bone metastases. Brain metastasis was strongly associated with concurrent liver and lung metastases, suggesting a pattern of widespread systemic involvement. Conclusions: Patients with SCCP frequently present with metastatic disease, and several clinical and socioeconomic factors influence metastatic risk. Higher tumor burden, lower income, metropolitan residence, and younger age were associated with a greater likelihood of metastasis at diagnosis. These findings highlight the importance of early diagnostic evaluation in high-risk groups and provide a foundation for improved prognostication and targeted care strategies in this rare and aggressive malignancy.

Time-toxicity of first-line systemic therapy for hepatocellular carcinoma in clinical practice.

4145 Background: Combination immunotherapy has become a standard first-line systemic treatment for hepatocellular carcinoma (HCC). While efficacy outcomes have been well described, the treatment-related time burden experienced by patients, including frequent hospital visits and admissions, has not been sufficiently evaluated in real-world clinical practice. Aims: To evaluate time toxicity as a patient-centered measure of treatment burden and to identify baseline factors associated with high time toxicity in patients receiving first-line combination immunotherapy for HCC. Methods: This retrospective observational study included patients with HCC who received first-line combination immunotherapy (atezolizumab + bevacizumab or durvalumab + tremelimumab) at our institution by April 2025. The observational period was defined as the treatment duration plus 28 days. Time toxicity (TT) was calculated as the proportion of days requiring hospital contacts, including scheduled visits, management of adverse events, and hospital admissions. High time toxicity (hTT) was defined as TT ≥20%, representing a clinically meaningful high treatment burden. Factors associated with hTT were examined using multivariate logistic regression analysis. Results: A total of 126 patients were included, with a median age of 74 years (IQR 67–81). The median observational period was 178 days (IQR 75-287), the median total hospital contacts was 20 days (IQR 13-30) and the median inpatient days was 7 days (IQR 4-14). Moreover, the median TT was 11.8% (IQR 9.1–17.9) and 23.0% of all patients experienced hTT. In multivariate analysis, impaired liver function (mALBI grade 2b–3; OR 7.17, 95% CI 2.60–19.80) and AFP ≥100 ng/mL (OR 3.03, 95% CI 1.20–7.67) were independently associated with hTT. Conclusions: Patients with impaired liver function and elevated AFP levels experienced a significantly higher treatment-related time burden during first-line combination immunotherapy for HCC. Assessment of time toxicity may provide clinically meaningful information to support shared decision-making, particularly in elderly or vulnerable patient populations.

Estimating individual benefit from adjuvant chemotherapy in hormone receptor–positive breast cancer using causal multimodal artificial intelligence.

551 Background: Adjuvant chemotherapy decisions in HR+/HER2– early breast cancer are often guided by prognostic gene expression assays. However, assessment of recurrence risk may not yield a predictive biomarker that accurately estimates treatment benefit. Here, we combine advances in deep learning and causality research to address this problem using routine histopathology and clinical data. We develop and validate a causal multimodal artificial intelligence (AI) model that predicts the patient-specific benefit of adding chemotherapy to endocrine therapy in HR+/HER2– early breast cancer. Methods: De-identified whole-slide histopathology images (WSIs) and clinical covariates from 9,269 patients across 12 observational cohorts were used to develop the Ataraxis Breast model (ATX). The locked model was then externally validated in a cohort from the University of Chicago (n = 435, stage I-III HR+/HER2– breast cancer, median age = 56, median follow-up = 7.2 years), grouped by adjuvant therapy received (endocrine therapy [ET] n = 322, chemoendocrine therapy [CET] n = 113). The primary endpoint was 5-year recurrence-free interval (RFI). WSIs were encoded via a pathology foundation model and integrated with clinical variables (including T/N stage, age, ductal vs lobular histology). ATX predicted counterfactual RFIs, assuming CET or ET, with the difference taken as the predicted treatment-benefit score. Discrimination was evaluated using Harrell’s C-index. Multivariate Cox proportional hazards models were fitted, adjusting for clinicopathological factors (age and T/N stage), to estimate hazard ratios (HR). Subgroup differences were assessed using inverse-propensity-weighted Kaplan-Meier estimates and two-sided log-rank tests. Results: In the external validation cohort, ATX was significantly associated with the primary endpoint (adjusted HR = 1.65, 95% CI 1.05-2.6; p = 0.029), and demonstrated good discrimination at the patient-level (C-index = 0.704, 95% CI 0.588-0.821). Notably, ATX stratified patients into subgroups with differential chemotherapy effects: patients in the top tertile of predicted treatment benefit experienced improved RFI with CET (ET RFI = 0.744, CET RFI = 0.980; p < 0.001), with no difference observed in patients with low predicted treatment benefit (ET RFI = 0.975, CET RFI = 0.970; p = 0.85). The interaction between ATX and the magnitude of chemotherapy benefit, when adjusted for clinicopathological factors, was significant (p-interaction = 0.005). Conclusions: A causal multimodal AI model accurately estimates chemotherapy benefit for patients with HR+/HER2– early breast cancer, meeting the criteria for a predictive biomarker. The causal AI methodology presented here may provide a generalizable framework to optimize scalable, predictive biomarkers for other therapies and cancer types.

Exploratory associations between incretin-based therapies and plasma cell disorders in patients with type 2 diabetes: A real-world study.

7571 Background: Incretin-based therapies, including dipeptidyl peptidase-4 inhibitors (DPP-4i) and glucagon-like peptide-1 receptor agonists (GLP-1RAs), are widely used in the management of type 2 diabetes mellitus (T2DM). Experimental and clinical data suggest that metabolic and immune pathways may influence plasma cell dyscrasia development, yet the relationship between incretin-based therapies and monoclonal gammopathy of undetermined significance (MGUS) or multiple myeloma (MM) remains unclear. In this study, we aimed to evaluate the association between incretin-based therapy exposure and the risk of MGUS and MM. Methods: A multicenter, retrospective cohort study was conducted using the TriNetX Global Collaborative Network, a federated electronic health record database, between January 1, 2015, and December 31, 2025. Adult patients with T2DM were categorized into six exposure cohorts. Individuals treated with GLP-1RAs with a minimum of one year of therapy were compared with patients initiating sodium–glucose cotransporter-2 inhibitors (SGLT2i), sulfonylureas (SU), and thiazolidinediones (TZDs). The same cohort design was applied to DPP-4i users. A 1:1 propensity score matching (PSM) model was applied to minimize confounding by balancing baseline demographics, clinical comorbidities, laboratory parameters, and concurrent medication use. Individuals with a prior cancer diagnosis were excluded. The primary endpoint was progression to MM, and the secondary endpoint was the development of MGUS. Statistical analyses were performed within the TriNetX platform. Results: After 1:1 PSM, GLP-1RA therapy was associated with a significantly lower risk of both MGUS and MM compared with SU (n = 141,225 per group; MGUS RR 0.59, 95% CI 0.50–0.69; MM RR 0.84, 95% CI 0.77–0.92) and SGLT2i (n = 57,919 per group; MGUS RR 0.75, 95% CI 0.65–0.86; MM RR 0.62, 95% CI 0.49–0.79). No significant differences in MGUS or MM risk were observed when GLP-1RAs were compared with TZDs. In contrast, DPP-4i use was associated with aincreased risk of MM compared with TZDs (n = 29,462 per group; RR 1.93, 95% CI 1.40–2.66) and SGLT2i (n = 68,481 per group; RR 1.44, 95% CI 1.19–1.74).No significant associations were observed for MGUS across comparators, and no significant differences in MGUS or MM risk were seen between DPP-4i and SU (Table 1). Conclusions: In patients with T2DM, GLP-1RA exposure were associated with differential rates of MGUS and MM diagnosis across antihyperglycemic drug classes. GLP-1RA was associated with a lower rates of MGUS and MM compared with SU and SGLT2i, whereas DPP-4i use was associated with higher MM rates when compared with SGLT2i and TZDs. While these findings are clinically promising, prospective studies are needed to confirm causality and clarify underlying mechanisms.

CARBON-SUSTAIN study: Visit-sparing, low-emission hypofractionated radiation therapy in a traffic-dense Indian megacity.

e23117 Background: In megacities, serial radiation therapy (RT) visits markedly augment healthcare-related carbon emissions due to traffic congestion, yet this burden remains poorly quantified. Bengaluru city illustrates this challenge, where each additional visit amplifies emissions, travel time, and patient burden. CARBON SUSTAIN STUDY aimed to quantify climate impact of RT delivery, assess if evidence-based hypo-fractionation(HF) can enable guideline concordant, low emission, carbon efficient, climate sensitive care model without compromising treatment precision. Methods: Electricity related emissions were estimated from measured per-fraction energy consumption (kWh/fraction) using direct clamp-meter measurements at uninterruptible power supply output (415-V, three phase; power factor (0.9) and Bengaluru Electricity Supply Company grid emission factor (0.71 kg CO₂/kWh). Measured energy use per fraction was 13.14 kWh (Ethos), 20.15 kWh (TomoTherapy), and 12.30 kWh (CyberKnife). Travel-related emissions were modeled using home to hospital distance bands (≤5 km, 6–10 km, 11–20 km, >20 km) with congestion-adjusted emission factors for weekday peak traffic. Primary endpoint was total carbon dioxide equivalent (CO₂e) emissions from electricity and travel. Secondary endpoints included Care Access Days Saved (CADS), defined as fraction-visits avoided, and regimen-level carbon efficiency. Statistical analysis included descriptive statistics, CADS with 95% confidence intervals, Kruskal–Wallis testing for CO₂e comparisons, and multivariable linear regression with robust standard errors. Results: In a total of 786 patients, delivered fractions were 12,185 compared with 20,859 expected under conventional schedule, resulting in 8,674 fraction-visits avoided (CADS) and 41.6% reduction in hospital visits. Electricity-related CO₂e avoided was 37.1 metric tons. Travel-related CO₂e avoided was at 95.4 metric tons, yielding total reduction of 132.5 metric tons of CO₂e. On multivariable analysis, number of fractions was strongest independent predictor of total CO₂e (p < 0.001),while travel-distance band contributed additional variance (p < 0.001). Conclusions: In traffic-dense megacities, meaningful decarbonisation of RT is driven primarily by reducing visit frequency. Visit-sparing HFRT substantially lowers carbon emissions maintaining treatment precision, offering scalable, low-emission, climate-sensitive framework. These reductions were attained using guideline-concordant HF without altering oncologic outcomes. Fraction reduced and CADS by disease site. Cancer site Patient (n) Delivered fractions Conventional fractions (CADS) %reduction Breast ca 542 8,976 16,260 7,284 44.8 Prostate ca 53 961 2,014 1,053 52.3 Glottic ca 27 868 945 77 8.1 Bone metastasis 164 1,380 1,640 260 15.9 Total 786 12,185 20,859 8,674 41.6

Comparative evaluation of targeted RNA sequencing and RNA-exome sequencing for fusion detection in soft tissue sarcomas.

e23524 Background: Soft tissue sarcomas are heterogeneous mesenchymal malignancies with overlapping histology, making diagnosis challenging. According to World Health Organization (WHO) classification of STS and bone sarcomas, >150 recurrent fusions have emerged as central diagnostic hallmarks for many sarcoma entities. RNA-based next-generation sequencing, particularly RNA-exome sequencing, enables sensitive detection of known and novel fusion transcripts with precise breakpoint characterization, and comprehensive molecular profiling even from limited tissue samples. The aim of this study was to assess the diagnostic utility of exome-capture RNA sequencing compared to targeted RNA next-generation sequencing for the detection of sarcoma-associated gene fusions, particularly in cases with challenging histology or negative results on targeted fusion panels. Methods: Total RNA was extracted from formalin-fixed, paraffin-embedded (FFPE) tissue using the MagMAX FFPE DNA/RNA Ultra Kit (Thermo Fisher Scientific). RNA-exome libraries were prepared with the NadPrep Total RNA to DNA – EZ DNA Library Preparation Kit (Nanodigmbio) and sequenced on the DNBSEQ-T7 platform (MGI Tech). Fusion transcript detection was performed using SeqPilot, complemented by bioinformatic tools STAR-Fusion and FusionInspector. In parallel, targeted RNA NGS libraries were generated using a custom sarcoma fusion panel, sequenced on the Ion GeneStudio S5 Prime System (Thermo Fisher Scientific) and analysed via the Torrent Suite software. Results: A total of 143 samples from patients with various histological subtypes of sarcoma were analyzed using a custom sarcoma panel (Ion AmpliSeq). Sarcoma-associated rearrangements were identified in 34% of cases (49/143). The most frequently detected alteration was the EWSR1::FLI1 fusion, observed in 7.7% of samples, a well-established molecular hallmark of Ewing sarcoma. Subsequently, 24 samples that were negative by the custom sarcoma panel were further analyzed using the RNA exome panel. In addition, in 17% (4/24) of cases, detection of TFE3::ASPSCR1 , HEY1::NCOA2 , FGFR1::WHSC1L1 , and EWSR1-NR4A3 fusions aided histological classification and confirmed the initial sarcoma diagnosis. Conclusions: RNA-exome sequencing provided additional diagnostic yield in panel-negative cases, enabling the detection of clinically informative fusions that supported or refined histological classification. These findings highlight the significant role of RNA-exome sequencing in the molecular diagnosis of STSs, particularly in diagnostically challenging cases, and support its integration into advanced diagnostic workflows. Nevertheless, prior review by sarcoma-expert pathologists remains essential to determine the appropriate indication for NGS testing.

Explainable AI to enable precision risk stratification for advanced melanoma in underrepresented Middle Eastern cohorts.

e21526 Background: Early identification of advanced-stage melanoma using baseline clinical and histopathological data could enable risk-stratified surveillance and earlier intervention, potentially improving outcomes. However, most predictive models are developed from Western populations, limiting their applicability to underrepresented Middle Eastern cohorts. We developed explainable machine learning (ML) models to identify actionable predictors of advanced-stage disease and brain metastasis using pre-treatment data from a Lebanese tertiary center cohort. Methods: We retrospectively analyzed 327 melanoma patients diagnosed at the American University of Beirut Medical Center. After excluding 132 patients with incomplete staging data, 195 patients were included for predictive modeling. The primary outcome was advanced-stage disease at diagnosis (stage III or IV; n = 72), with early-stage disease (stage 0–II; n = 123) as the comparator. The secondary outcome was brain metastasis during follow-up (n = 143; 9 events). We selected 20 clinically relevant baseline variables spanning demographics, medical history, and tumor characteristics, ensuring no data leakage by excluding post-diagnosis variables. We evaluated five machine learning algorithms using 5-fold stratified cross-validation with 1000 bootstraps. Feature importance was assessed using SHapley Additive exPlanations (SHAP), and robust predictors were identified by convergent evidence from SHAP rankings and bootstrap-validated univariable logistic regression. Results: The Random Forest model achieved the highest discriminative performance for advanced-stage prediction (AU-ROC 0.814, 90% CI 0.743–0.846; AU-PRC 0.781, 90% CI 0.636–0.827; F1-score 0.621). Gradient Boosting followed closely (AU-ROC 0.793). Univariable analysis identified five robust predictors of advanced stage, including Clark level V (OR 7.67, 90% CI 4.91–12.43, p = 0.018), high mitotic count (≥7 mitoses; OR 6.12, p = 0.033), lymphovascular invasion (OR 4.27, p = 0.017), nodular histology (OR 3.07, p = 0.017), and Charlson Comorbidity Index (CCI) (OR 1.41 per unit increase, p < 0.001). Conversely, radial growth phase was protective (OR 0.25, p = 0.016). SHAP explainable ML analysis ranked CCI as the most influential predictor (mean |SHAP| = 1.06). Notably, CCI was also significantly associated with brain metastasis risk (OR 1.35, 90% CI 1.10–1.65, p = 0.014). Conclusions: Our explainable ML framework enables immediate risk stratification at diagnosis using five convergent predictors: comorbidity burden (CCI), Clark level, mitotic count, lymphovascular invasion, and histology subtype (nodular), with radial growth phase showing a protective effect. These findings support targeted surveillance and management for high-risk patients, advancing equitable AI-driven precision oncology for underrepresented populations.

Zopapogene imadenovec-drba, a novel non-replicating adenoviral vector-based immunotherapy: Effects on complete and durable responses in recurrent respiratory papillomatosis pivotal trial.

6104 Background: Recurrent respiratory papillomatosis (RRP) is a rare, neoplastic disorder caused by chronic human papillomavirus (HPV) type 6 or 11 infection. Significant morbidity can occur due to airway obstruction and transformation into malignant cancer. Repeat surgical debulking has historically been the most common treatment for RRP symptom management. Zopapogene imadenovec-drba (zopa), a novel adenoviral vector-based immunotherapy, is the first and only FDA-approved treatment for adults with RRP. Zopa is now recommended as the first-line treatment for adults with RRP in an RRP Foundation position statement authored by 16 key opinion leaders (Best et al. Laryngoscope 2026). Methods: The pivotal trial (NCT04724980) evaluated zopa in patients with RRP requiring ≥3 clinically indicated interventions 12 months (m) prior to treatment. 12m follow-up data was reported and demonstrated that 4 subcutaneous injections of zopa (5x10 11 particle units per injection; n=35) were well-tolerated, with no serious adverse events, no grade >2 treatment-related adverse events, and no early treatment discontinuations. The most common adverse events were injection-site reaction, fatigue, chills, fever, and myalgia. Robust efficacy was observed following zopa treatment with 51% (34 to 69; 95% CI) of patients achieving a complete response (CR), defined as no requirement for interventions in the 12m following treatment, and 86% (30/35) of patients experiencing a decrease in interventions in the year following treatment as compared to the year prior to treatment. Here we present data up to 51m of follow-up. Results: As of December 15, 2025, 83% (15/18) of patients who achieved a CR at 12m remain in CR with no recurrence of papilloma requiring surgical or medical intervention. The median duration of follow-up for patients in CR was 36m (range: 30-51m), with 3 patients having a response lasting more than 4 years. No new safety events were observed during long-term follow-up. Conclusions: Zopa treatment demonstrated significant clinical benefit with the vast majority of CR patients experiencing ongoing durable complete responses for up to 4 years with excellent long-term safety. Updated follow-up results for all patients achieving CR will be available at the time of presentation. Clinical trial information: NCT04724980 .

Customized NIR‐II Dual‐Phase Carbon Dots: Elucidating the Emission Mechanism for Smart Photonic Applications

ABSTRACT Carbon dots (CDs) exhibiting dual‐phase photoluminescence (PL) in the second near‐infrared window (NIR‐II, 900–1700 nm) region have shown considerable potential for applications in information encryption, lighting, and bioimaging. However, significant challenges remain owing to a lack of reliable design strategies. In this study, high‐brightness dual‐phase NIR‐II PL CDs (NIR‐CDs) were successfully prepared for the first time employing innovative strategies. NIR‐CDs possess a distinctive core–shell structure, and the donor–acceptor interactions between the carbon core and strong electron‐donating groups on the shell effectively induce a charge‐transfer state. The NIR‐CDs exhibited a fluorescence quantum yield (QY) of up to 5.91% at 920 nm in conjunction with excellent stability. Notably, the rich functional groups on the shell of NIR‐CDs provided hydrogen bonding sites, enabling a 960 nm fluorescence emission and high absolute QY of 3.82% in the solid state. The dual‐phase PL of NIR‐CDs enables their use in NIR‐II multilevel information encryption and fingerprint authentication. The NIR‐II flexible film based on NIR‐CDs exhibits excellent fluorescence stability under various external stimuli. Furthermore, the quick response code based on NIR‐CDs remains clearly identifiable beneath 3 mm porcine tissue, which highlights their potential for anti‐counterfeiting, encryption, and bioimaging sensing.

Synthesis of 2,3-Naphthalocyanine/ nickel hydroxide (2,3-Nc/Ni(OH)2) nanocomposite for uric acid electrochemical sensing

Novel opioid provides cleaner pain relief

Reliability of quantitative and qualitative chorioretinal uveitis lesion analysis on blue, green, and near-infrared fundus autofluorescence and color fundus photography

Abstract As there is a lack of reliable structural outcome parameters for posterior uveitis trials, we compared inter-rater reliability (IRR) of chorioretinal lesion area measurement (quantitative analyses) and characterization (qualitative analyses) on color fundus photography (CFP) and multimodal fundus autofluorescence (FAF). In this prospective cohort study, posterior uveitis eyes were imaged with CFP (Eidon, iCare, Vantaa), short- (swBAF, 450 nm; Eidon) and long-wavelength blue-light-autofluorescence (lwBAF, 488 nm), green-light-autofluorescence (GAF, 518 nm), and infrared-autofluorescence (IRAF, 787 nm) (all Spectralis, Heidelberg Engineering, Heidelberg). Lesion area, measured with ImageJ (National Institutes of Health), and image characteristics of lesions were graded on all modalities by two masked raters. A total of 318 lesions from 27 eyes (17 patients) were included. Absolute inter-rater differences in area measurement were 0.57, 0.78, 0.30, 0.36, and 0.55 in standardized 10 3 pixels for CFP, swBAF, lwBAF, GAF, and IRAF, respectively. IRR was high for all modalities for quantitative (intraclass correlation, 95% confidence interval (CI) [0.997–0.998]) and at least substantial for qualitative measures (unweighted Cohen’s kappa 0.84, 0.91, 0.89, 0.91, and 0.89 for CFP, swBAF, lwBAF, GAF, and IRAF, respectively, all p  < 0.0001, CIs [0.79–0.95]). Hence, FAF could be a reliable complementary imaging modality for posterior uveitis clinical trials, especially for lesion quantification.

Role of mitochondrial translation in modulating inflammatory disease outcome: Current knowledge and future perspectives

Facilitators and barriers to Medicaid services for childhood cancer survivors: Perspectives from oncology providers and staff.

e22009 Background: Stable insurance coverage is critical for childhood cancer survivors who require lifelong medical care to manage long-term health risks. Healthcare provider perspectives on the care experiences of Medicaid-insured childhood cancer survivors, an economically disadvantaged survivor population, remain understudied. Methods: We conducted semi-structured interviews with 22 healthcare providers or staff caring for adult survivors of childhood cancer, including oncologists, advanced practice providers or nurses, social workers, and financial counselors from Medicaid expansion and non-expansion states. Most participants were female (91%), aged 40-54 years (50%; range: 32–57; median: 51), and non-Hispanic White (73%), with 41% operating within both pediatric and adult care settings. Transcripts were analyzed using deductive and inductive thematic approaches. Results: From the provider or staff perspective, six themes regarding barriers to survivor care emerged: 1) many survivors forgo primary care, relying instead on survivorship clinics due to greater trust in oncology teams; 2) some survivors avoid primary care due to desire to move on from medical visits; 3) primary care providers often lack expertise or confidence in managing complex survivorship needs; 4) specialty care access is limited, as many specialists do not accept Medicaid, largely due to low reimbursement, expired contracts, and/or falling outside of managed care networks; 5) survivors face long wait times and appointment scheduling difficulties, especially in adult care settings where providers are less likely to accept new Medicaid patients; and 6) young adult survivors often struggle to navigate the adult healthcare system due to limited knowledge and support. Four themes were identified as survivor care facilitators: 1) dedicated support staff - e.g., social workers, navigators, transition coordinators – who connect survivors to the providers they need; 2) specialized transition clinics that bridge pediatric and adult care; 3) lists of Medicaid-accepting providers, which can improve referral and care continuity; and 4) integrated health systems that facilitate patient access when specialty and primary care are housed within cancer centers or affiliated networks. When asked for recommendations, providers and staff emphasized the need for policies that recognize the unique long-term care needs of childhood cancer survivors, streamline or expand Medicaid eligibility for this population, and advance toward universal coverage models. Conclusions: Oncology providers and staff described multifactorial, system-level barriers and facilitators shaping care for Medicaid-insured childhood cancer survivors. Findings underscore the need for policy reforms and enhanced care navigation/transition infrastructure to improve equitable access to primary and specialty services.

Neoadjuvant hepatic arterial infusion chemotherapy (HAIC) plus tislelizumab (Tisle) combined with adjuvant tisle as a perioperative therapy for hepatocellular carcinoma (HCC) at high-risk of recurrence: A single-arm, phase II study.

e16284 Background: Adjuvant therapy alone after curative resection has failed to reduce recurrence or improve survival in HCC patients (pts) at high risk of relapse. A perioperative strategy integrating neoadjuvant and adjuvant treatment may improve outcomes. HAIC combined with PD-1 inhibition has demonstrated synergistic antitumor activity with favorable safety in HCC. This study evaluates a perioperative “sandwich strategy” consisting of neoadjuvant HAIC plus Tisle, followed by surgery and adjuvant Tisle, in high-risk resectable HCC. Methods: This ongoing prospective single-arm phase II trial (NCT06467799) plans to enroll 39 pts, with resectable CNLC Ib/IIa HCC beyond Milan criteria (single tumor > 5 cm or 2–3 tumors with largest > 3 cm), without portal vein thrombus, extrahepatic metastasis, or prior treatment. Eligible pts receive 2 cycles of neoadjuvant FOLFOX-HAIC plus Tisle, followed by surgery upon confirmed resectability, and 4 cycles of adjuvant Tisle. A Bayesian optimal phase II design was used, with 1-year recurrence-free survival (RFS) as the primary endpoint. The historical 1-year RFS rate for surgery alone was 60%, and 80% for the experimental group. With a one-sided type I error of 0.05, power of 0.9, and 10% dropout, the total sample size was 39. Interim analyse was performed at 10, 20, 30, 39 pts, maintaining an alpha of 0.05 and power of 0.86. Secondary endpoints include objective response rate (ORR, mRECIST), pathologic complete response(pCR), overall survival (OS) and safety. Results: As of January 1, 2026, 32 pts were enrolled; 26 completed neoadjuvant therapy and six remained on neoadjuvant treatment. The ORR to neoadjuvant therapy was 53.8%. Of the 26 patients completing neoadjuvant therapy, 25 underwent planned surgical resection. The pCR rate was 24.0%, and microvascular invasion was observed in only one patient. Sixteen pts had completed adjuvant therapy. With a median follow-up of 6.3 months, no disease recurrence or death was observed. The most common treatment-related adverse events were grade 1–2 included decreased appetite and vomiting, with no grade ≥3 adverse events reported. Conclusions: Perioperative HAIC plus Tisle demonstrated favorable safety and promising antitumor activity in pts with high-risk HCC. Comprehensive long-term outcoms, safety, pathological, and clinical response data will be provided. Clinical trial information: NCT06467799 . Baseline tumor characteristics and efficacy of neoadjuvant therapy. parameter All （n=32） Tumor size (mean, cm) 7.5 Tumor N.12-3 68.8 %31.2% mRECIST evaluable for neoadjuvant treatment (n=26) CRPRSDPD 5/269/2612/260 ORR(%) 53.8% DCR(%) 100.0% pCR(%) 24.0%(6/25) MVI(%) 4.0%%(1/25)