Browse Articles

Clinician readiness and ethical thresholds for agentic AI autonomy in oncology care: A multi-specialty survey.

e13699 Background: Artificial intelligence (AI) systems capable of autonomous reasoning areemerging in oncology. This study assessed oncology provider attitudes, ethicalboundaries, and readiness for clinical adoption of varying levels of AI autonomy. Methods: A cross-sectional anonymous online survey was administered to cliniciansin 01/2026 at a comprehensive cancer center, across surgical (SO), medical (MO), andradiation oncology (RO). The survey assessed AI exposure, readiness for autonomousworkflows, impact on burnout, and ethical concerns. Respondents rated comfort acrossfive levels of increasing AI autonomy (L1: support; L2: task handler; L3: collaborator; L4:clinical guide; L5: fully autonomous) using 5-point Likert scales. Subgroup analyses byspecialty, role (Physician vs. APP), and experience were performed. Results: Seventy-nine clinicians responded (32 SO, 29 MO, 17 RO; 55% Attendings,44% APPs; 37% response rate). Personal daily/weekly AI use was 49%, while clinicaluse was 33%. Comfort was high for AI L1–L3 (60–68%) but declined sharply at L4(34%) and L5 (5%). SO clinicians were least comfortable with L4 (16% vs. 41-50% forRO/MO) and full autonomy L5 (0% vs 7%-12%).While 93% affirmed that human judgment is fundamentally necessary in oncology care,a “human touch threshold” emerged: SO felt the human component of oncology wouldbe adversely impacted by Level 4 (collaborative), while RO/MO by Level 5(autonomous).An “experience paradox” was observed: among physicians who predicted that > = 50% oftheir workflow could be automated by AI, late-career clinicians constituteddisproportionately (1.7x) large share of respondents. Optimistically, 46% of cliniciansexpected that AI would improve burnout risk within next 5 years.Compared with physicians, APPs reported greater readiness to adopt AI systems within5 years at L3-L4 than physicians (82% vs 69% at 5 years), but with lower comfortscores (41% vs 54%). Both groups demonstrated low comfort with fully autonomous AI(L5), with similar ethical boundaries on AI use. Fully autonomous Level 5 systems wereviewed as unlikely to be personally adopted within 5 years. Regarding accountability for adverse outcomes, 51% favored clinician responsibility, while 44% supported a shared-responsibility model.Primary drivers of AI adoption were felt to be hospitals/health systems (35%), followedby physicians (22%), and technology vendors (15%). Conclusions: Oncology clinicians support assistive and collaborative AI but draw ethical and professional boundary as AI approaches guided decision-making andstrongly reject fully autonomous clinical agents. Differences across specialties, roles, and experience levels indicate that AI implementation will need to be tailored and supported by clear governance structures.These findings establish a baseline for comparison with future changes in clinician perspectives as AI use evolves.

Impact of cooking-related household air pollution (HAP) on never-smoker NSCLC risk in women: Systematic review and meta-analysis.

e20690 Background: Among pollutants, exposure to cooking-related household air pollution (HAP) is a significant environmental health concern, particularly problematic for women who are non-smokers but still face a rising incidence of non-small cell lung cancer (NSCLC). However, globally, tobacco is the leading cause of cancers, but a substantial number of lung cancer cases have been reported among women, especially from low-income countries, who never smoked. Ultimately, our systematic review and meta-analysis focus on investigating the association between various cooking-related pollutants, such as cooking oil fumes and biomass fuel combustion, and risk for the development of NSCLC among these women. The pooled data will provide clarification of the impact of indoor smoking environmental pollutants on lung cancer etiology in never-smoking women. Methods: This systematic review and meta-analysis followed PRISMA guidelines and registered with PROSPERO. A total of 1901 articles were retrieved using databases including PubMed, Scopus, and EMBASE from inception to January 2026. Eligible studies evaluating the impact of exposure to cooking-related household air pollution (HAP) on the occurrence of non-small cell lung carcinoma among non-smoking adult women ≥18 years, compared with women unexposed to HAP were included. Data were analyzed using RevMan version 5.4 with a random-effects model to estimate pooled odd ratios (ORs) with 95% confidence intervals. A p-value < 0.05 was considered statistically significant. Results: Our systematic review and meta-analysis pooled results demonstrated a visible link between indoor pollution exposure and lung cancer. HAP exposure showed no significant association with lung cancer vs. healthy controls (OR: 0.66; 95% CI: 0.25-1.76, P = 0.41). Although limited by high heterogeneity (I2 = 99%), these results reflect variation in individuals across the globe. From the structural analysis of event sizes, we deduced that HAP exposure led to 1.51 times more chance of acquiring lung cancer (OR: 1.51; 95% CI: 1.25, 1.81; I2 = 20%; P < 0.0001). Our findings were verified from the reported HR: 2.04 (95% CI: 1.47, 2.82; I2 = 0%; P < 0.0001), indicating that exposed individuals had twice the chance of lung cancer being diagnosed. Conclusions: This systematic review and meta-analysis substantiates cooking-related household air pollution as an important etiologic factor in lung cancer among never-smoking women, implicating indoor combustion byproducts and supporting targeted public health interventions.

Organ-specific endocrine toxicities and serious outcomes associated with immune checkpoint inhibitors: A contemporary pharmacovigilance analysis (2020–2024).

e24197 Background: Endocrine immune-related adverse events (irAEs) are recognized complications of immune checkpoint inhibitors (ICIs), yet contemporary real-world data on organ-specific severity and serious outcomes are limited. We conducted a pharmacovigilance study to characterize organ-specific endocrine toxicity signals and associated serious outcomes in the current ICI era. Methods: We performed a retrospective analysis of adverse events reported to the FDA Adverse Event Reporting System (FAERS) from January 2020 through December 2024. ICI exposure was identified using drug role codes with comprehensive capture of all current PD-1, PD-L1, and CTLA-4 inhibitors, administered as monotherapy or in combination. Primary suspected (PS) ICIs were used for primary analyses, with PS plus secondary suspected (PS+SS) analyses assessing robustness and combination regimens.Endocrine adverse events were categorized into thyroid, pituitary, adrenal, and diabetes-related groups using MedDRA preferred terms. Disproportionality was assessed using reporting odds ratios (RORs) with 95% confidence intervals (CIs) relative to all other drugs. Serious outcomes were defined using FAERS outcome codes (death, hospitalization, life-threatening, disability, or other medically serious condition). Results: Among 7.47 million deduplicated FAERS cases, 104,128 PS ICI-exposed cases were identified, including 7,844 endocrine irAEs. Compared with non-ICI therapies, ICIs showed significantly increased reporting of endocrine toxicities, strongest for pituitary (ROR 41.5, 95% CI 39.0–44.2) and adrenal events (ROR 24.5, 95% CI 23.3–25.7), followed by thyroid (ROR 6.2, 95% CI 6.0–6.5) and diabetes-related toxicities (ROR 3.4, 95% CI 3.3–3.6). Findings were consistent in PS+SS analyses.Despite lower frequency, pituitary and adrenal irAEs carried the highest severity burden, with hospitalization rates of 58.1% and 58.7% versus 37.5% for thyroid events, and life-threatening outcomes in 11.8% and 10.7% versus 6.1%, respectively. Overall mortality among endocrine irAEs was 8.8%. Combination ICI therapy (21,369 PS+SS cases) showed higher endocrine toxicity burden than monotherapy (12.3% vs 6.3%) and higher hospitalization (58.9% vs 51.8%) and life-threatening event rates (12.1% vs 9.6%), with similar mortality. Conclusions: ICI-associated endocrine toxicities exhibit distinct organ-specific risk–severity profiles. Pituitary and adrenal irAEs, though less frequent than thyroid dysfunction, show disproportionately high reporting odds and serious-outcome burden, consistent with a “rare but severe” phenotype. Combination ICI therapy further increases endocrine toxicity risk and severity, underscoring the need for heightened vigilance and proactive monitoring, especially for pituitary and adrenal complications.

Expression of Concern: RNA-Seq analysis revealed genes associated with drought stress response in kabuli chickpea (Cicer arietinum L.)

Asymmetric Vacancies Efficiently Couple O ₂ Activation With Reactive Oxygen Species Evolution for Enhanced Photocatalytic Methane Conversion

ABSTRACT Photocatalytic oxidation mediated by reactive oxygen species (ROS) provides an effective platform for a wide range of important chemical transformations. However, conventional oxygen‐vacancy engineering strategy, while enhancing O 2 activation, often hampers subsequent ROS evolution due to overly strong adsorption, thereby limiting oxidation kinetics. Here, we demonstrate that asymmetric vacancies in ZnGa 2 O 4 , characterized by a Zn Td −O v −Ga Oh configuration, can overcome this intrinsic limitation by synergistically coupling O 2 activation with efficient ROS evolution. Specifically, the dynamic Ga Oh site preferentially promotes O 2 adsorption and activation, whereas the Zn Td site interacts weakly with oxygen‐derived species, facilitating ROS release and vacancy replenishment, thereby achieving an optimal balance between these critical steps. Consequently, Ag/ZnGa 2 O 4 delivers the highest turnover number (TON) reported to date among Ag‐based catalysts for the photocatalytic oxidative coupling of methane via a ROS‐mediated pathway. The general effectiveness of this asymmetric‐vacancy strategy is further validated in other representative photocatalytic reactions, including hydrogen peroxide production and the oxidative coupling of benzyl alcohol.

Exploring multi-component nanofiber architectures of ZnO, curcumin, and andrographis paniculata for advanced wound care applications

Intelligent Acousto‐Electrical Metamaterials (IAM) for Sound Source Detection

ABSTRACT Acoustic transducers are essential for object localization and environmental sensing. Conventional transducers rely on piezoelectric crystals, whose acoustic‐electric response is fixed by the crystal lattice's inherent asymmetry and orientation. This results in static coupling behavior, necessitating bulky arrays of rigid elements with complex wiring and high computational demands for directional sensing. Here, we report a fundamentally new class of acoustic‐electric coupling that emerges from topology‐governed charge transport in 3D micro‐architected piezoelectric metamaterials. Unlike single crystals, these architected materials exhibit dynamic, geometry‐driven electromechanical responses. Acoustic waves excite multiple coupled vibration modes, enabling selective amplification, suppression, or reversal of charge flow based on the incident wave's frequency, direction, and the material's topology. This tunable, symmetry‐breaking response is encoded not in the chemistry but in the architecture—representing a shift from crystal‐defined to structure‐programmed piezoelectricity. We further demonstrate that a single metamaterial transducer can perform frequency‐dependent beam shaping without changing aperture size or requiring mechanical adjustment. Combined with machine learning and 3D printing, these intelligent acousto‐electrical metamaterials (IAM) enable real‐time localization of multiple moving sound sources. This approach lays the foundation for compact, adaptive, and intelligent acoustic sensing systems across a range of applications—from autonomous vehicles to medical imaging and underwater robotics.

Orthoptera as emerging nutritional resources: comparative analysis of protein and mineral composition

Identification of novel ubiquitin receptors on the 26S proteasome by photo-crosslinking mass spectrometry

Real-world outcomes of 7+3 versus azacitidine/venetoclax in CML transformed to AML.

6579 Background: Although tyrosine kinase inhibitors (TKIs) have markedly improved outcomes in chronic myeloid leukemia (CML), transformation to acute myeloid leukemia (AML) remains a devastating event with poor prognosis. Intensive induction chemotherapy with cytarabine and anthracycline (7+3) is standard, but many patients are older or frail and unable to tolerate intensive therapy. Azacitidine plus venetoclax (aza-ven) has emerged as a low-intensity alternative; however, real-world comparative outcome data in CML-transformed AML remain limited. We conducted a large real-world analysis to compare survival and early clinical outcomes with 7+3 versus aza-ven. Methods: Adult patients with CML transformed to AML who received induction therapy between January 1, 2010, and June 1, 2025, were identified using the TriNetX Research Network. Patients were classified by receipt of 7+3 or azacitidine plus venetoclax within one year before or after allogeneic hematopoietic stem cell transplantation. Overall survival (OS) was evaluated using Kaplan–Meier analysis with hazard ratios (HRs) and log-rank testing. Early clinical outcomes were compared using z-tests with risk differences (RD), risk ratios (RR), and 95% confidence intervals (CI). Results: A total of 280 patients were included, of whom 222 received 7+3 and 58 received azacitidine plus venetoclax (aza-ven). Patients treated with aza-ven had a higher comorbidity burden and lower baseline albumin, consistent with a frailer population. Early infectious complications within 30 days occurred in 53.2% of patients receiving 7+3 and 58.6% receiving aza-ven (risk difference [RD] −5.47%, 95% CI −19.74% to 8.81%; risk ratio [RR] 0.91, 95% CI 0.71–1.16; p=0.46). One-year overall survival was similar between groups (66.2% vs 68.1%; hazard ratio [HR] 1.03, 95% CI 0.62–1.72; log-rank p=0.91). Hospitalization within 30 days occurred less frequently with 7+3 than with aza-ven (71.6% vs 84.5%; RD −12.86%, 95% CI −23.91% to −1.82%; RR 0.85, 95% CI 0.74–0.97; p=0.046), while hospitalization between days 31–180 was identical in both groups (50.0% vs 50.0%; p=1.00). Infection between days 14–30 occurred in 34.7% versus 43.1%, respectively (RD −8.42%, 95% CI −22.62% to 5.78%; p=0.24). Conclusions: In this real-world cohort of patients with CML transformed to AML, azacitidine plus venetoclax achieved comparable 1-year survival and early infectious outcomes to intensive 7+3 induction despite being used in a frailer population. Larger multicenter studies are needed to confirm comparative effectiveness and refine patient selection.

Induction endocrine therapy in older and very old patients with breast cancer.

e12656 Background: Management of breast cancer (BC) in older adults is complicated by a high burden of comorbidity and substantial competing mortality, increasing the risk of both overtreatment and undertreatment. In routine practice, preoperative endocrine therapy with aromatase inhibitors (AIs) is often used as a “bridging” strategy during diagnostic workup and surgical planning. Assessment of changes in proliferative activity (Ki-67) may inform subsequent treatment individualization. Methods: We conducted a retrospective single-center analysis of 152 women aged ≥75 years who underwent surgical treatment at the Loginov Moscow Clinical Scientific Center between 2015 and 2020. All patients received preoperative AI therapy during the workup period; treatment duration before surgery ranged from 2 weeks to 1 year. We evaluated clinicopathologic characteristics, type of surgery, Ki-67 based on postoperative immunohistochemistry, early oncologic outcomes (locoregional recurrence, 1-year OS and DFS), and causes of death. Categorical variables are reported as n (%) with exact 95% confidence intervals (Clopper–Pearson). For zero-eventoutcomes, the upper bound of the 95% CI was calculated. Reported 1-year OS/DFS rates were supplemented with exact 95% CIs for proportions. Results: Early-stage disease (stage I–IIa) was observed in 137/152 (90.1%; 95% CI 84.2–94.4), while locally advanced disease (stage IIb–III) accounted for 15/152 (9.9%). Tumor size ≤3.0 cm was present in 133/152 (87.5%; 95% CI 81.2–92.3). Invasive carcinoma of no special type was diagnosed in 107/152 (70.4%), and invasive lobular carcinoma in 20/152 (13.2%); all tumors were hormone receptor–positive subtypes. With respect to surgical approach, mastectomy was performed in 90/152 (59.2%; 95% CI 51.0–67.1) and breast-conserving surgery in 62/152 (40.8%; 95% CI 32.9–49.0). On postoperative immunohistochemistry, a > 62% reduction in Ki-67 was documented in 82% of cases; mean Ki-67 decreased from 30% to 17% after induction endocrine therapy. No locoregional recurrences were observed in the early follow-up period (0/152; upper 95% CI 2.4%). One-year overall survival (OS) was 140/152 (92.1%; 95% CI 86.6–95.9) and one-year disease-free survival (DFS) was 147/152 (96.7%; 95% CI 92.5–98.9). Death due to BC progression occurred in 5/152 (3.3%; 95% CI 1.1–7.5), while deaths from other causes occurred in 7/152 (4.6%; 95% CI 1.9–9.3). Conclusions: In women aged ≥75 years with hormone receptor–positive BC, preoperative induction endocrine therapy with AIs administered during diagnostic workup was associated with a marked reduction in proliferative activity (Ki-67) and favorable early clinical outcomes, with no documented locoregional recurrences. The observed mortality pattern underscores the clinical importance of competing causes of death in this age group and should be considered when determining the intensity of systemic therapy.

Evaluation of the safety, efficacy and dosimetry of TRC003 in metastatic castration-resistant prostate cancer: A prospective, open-label, single-arm study.

5037 Background: TRC003 (also named as 225 Ac-PSMA-313) is a novel molecule of targeted alpha therapy (TAT) that administers alpha-particle radiation to cancer cells expressing Prostate Specific Membrane Antigen (PSMA). In this study, we plan to investigate preliminary safety, efficacy and dosimetry of TRC003 in patients with metastatic castration-resistant prostate cancer (mCRPC). Methods: This is a prospective, open-label, single-arm, single center study in accordance with ICH GCP guidelines. The patients with progressive PSMA-positive mCRPC who have been treated with androgen receptor pathway inhibitors (ARPIs) and/or taxane-based chemotherapy were recruited and received TRC003 treatment intravenously at 200 µCi (7.4 MBq) doses administered every 8 weeks per cycle for 4 cycles followed by expanded 4 cycle treatment if the participants continued to be benefit from the treatment. Adverse events, ORR, PSA response, bPFS, rPFS and OS were evaluated. In addition, the SPECT images were acquired with noncircular orbit, 120 project views over 360° at 4, 24, 48 and 96 hours after administration for dosimetry analysis. Results: From Feb 2024 to Jan 2026, 15 patients were included with an average age of 67.1 yr and median baseline PSA of 98.6 ng/mL. 11 (73.3%) participants received ARPI and taxane-based chemotherapies while 4 (26.7%) participants received only ARPI previously. Median baseline SUVmax and SUVmean on PSMA-PET were 32.7 (14.7-60.26) and 18.2 (9.51-34.73), respectively. A total of 63 cycles of TRC003 were administered, with a median of 4 cycles per patient. The most common AE was anemia (80%) and dry mouth (73.3%). Grade 3 treatment-related adverse events occurred as anemia (20%), lymphocytopenia (13.3%), weight loss (13.3%), thrombocytopenia (6.7%), and vomiting (6.7%). No treatment-related Grade ≥4 AE were observed. No kidney related toxicities were observed. The ORR of all patients was 45.5%, whereas the ORR reached 57.1% in the subset of patients who had received prior treatment of ARPIs and taxanes. The median tumor absorbed dose was 2.40±3.33 Gy (soft lesions: 5.03±5.51 vs bone: 1.65±2.19 Gy). The absorbed doses of kidneys, liver and spleen were 1.56±0.38, 1.07±0.19 and 0.54±0.29 Gy, respectively. The absorbed dose coefficients of tumors, kidney and liver were 0.36±0.50 (soft lesions 0.75±0.82 vs bone 0.25±0.33), 0.23±0.07, 0.16±0.03 Gy/MBq. Conclusions: TRC003 was safe and well-tolerated, and exhibited efficacy for mCRPC patients who were resistant to ARPI or chemotherapy. Overall TRC003 is a promising radiopharmaceutical agent for the treatment of mCRPC. Clinical trial information: ChiCTR2400083275.

Applying learning science principles (LSP) in a workshop (WS) for hematology and oncology (HemOnc) trainees (HOT) about how to use artificial intelligence (AI) in medical education (MedEd) and clinical practice (CP).

9030 Background: AI use is ubiquitous among HOT and healthcare workers, but with variable base knowledge and comfort level. Its indiscriminate use may be problematic. To address this educational gap, we developed a WS for HOT to assess their use of AI and educate them on how to employ it in MedEd and patient-centered CP based on LSP. Methods: A faculty member developed a WS on how to use AI in MedEd and CP based on LSP of analogy, contrasting cases, elaboration, generation, and question driven learning. The WS was for HOT, but faculty could join. It had 3 sections: introduction to basic concepts of AI; principles of the Health Insurance Portability and Accountability Act and protected health information; AI in MedEd; and AI in CP. Sections had real use case examples of AI, for participants to interact, ask questions, and share their personal use cases. HOT were invited to compare and generate examples during the WS, based on their educational and clinical experience and needs. Participants’ AI use and self-perceived and objective knowledge about it were assessed in a pre activity survey, a post WS survey, and a post 8 week survey to evaluate long term learning. Results: A 2 hour WS with 9 HOT and 2 faculty was completed successfully. 10 participants completed the pre WS survey, and 11 completed the post WS and the post 8 week surveys. 100% of them used generative AI tools before the WS, while 10% had prior training in AI. 90% used AI in personal life and MedEd, 70% used it in CP, and 30% in academic research. Table 1 shows participants’ self-assessed knowledge of AI, comfort level in using AI for MedEd and CP, self-assessed preparedness to implement AI in MedEd and CP, and objective knowledge. 8 weeks post WS, 91% of participants reported increased AI use in CP, leading anywhere from slight to significant decreases in time spent on documentation tasks for 91% of them. In MedEd, 73% reported increased AI use, and 82% described slight to significant improvement in the quality of time dedicated to it. 82% described the WS as extremely relevant to their daily responsibilities, and 91% reported confidence navigating ethical and privacy risks of AI in HemOnc. All participants were interested in having a similar WS in the future. Conclusions: An LSP based WS about AI increased the self-perceived and objective knowledge of HOT about using this tool in MedEd and CP. It also increased HOT comfort level, preparedness, and efficiency using AI. With the ubiquitous use of AI by HOT, ongoing education about it focused on supporting high-quality patient care should be a part of fellowship curriculum. Pre WS % Post WS % Post 8 week % Moderately/Very knowledgeable about AI 50 91 100 Comfortable using AI in MedEd 60 82 82 Prepared to implement AI in MedEd 30 91 73 Comfortable using AI in CP 40 73 82 Prepared to implement AI in CP 0 91 73 Correct answers on objective assessment 53 68 71

Safety and efficacy of GT719, an off-the-shelf CAR-NKT therapy, in adults with relapsed or refractory CD19-positive B-cell malignancies.

7078 Background: GT719 is an off-the-shelf CAR-NKT therapy derived from umbilical cord blood hematopoietic stem cells and genetically engineered to co-express an anti-CD19 CAR, an invariant natural killer T (NKT)-specific TCR, and secreted interleukin-15 (IL-15). This design enables CD19-targeted tumor killing, avoids risk of graft-versus-host disease (GVHD), and enhanced cellular expansion, persistence, and immune function through IL-15 signaling. We report clinical data from 7 adults with relapsed or refractory (R/R) CD19-positive B-cell hematologic malignancies enrolled in two open-label, single-arm studies (NCT06948981, NCT07131254) evaluating the safety and preliminary efficacy. Methods: The primary endpoint was safety, including treatment-emergent adverse events (TEAEs), graded per CTCAE v5.0. Secondary endpoints included 3-month overall response rate (ORR), best overall response (BOR), duration of response (DOR), progression-free survival (PFS), and overall survival (OS). Efficacy was assessed using Lugano 2014 criteria and the 2024 Chinese guidelines for adult acute lymphoblastic leukemia. Results: As of December 31, 2025, 7 patients were treated (median age 51 years; median two prior therapies), including 3 with B-ALL and 4 with B-NHL. Following FC lymphodepletion, all patients received a single GT719 infusion (5×10⁷ to 1×10⁹ viable cells). Most adverse events (AEs) were Grade 1-2. Grade ≥3 AEs were limited to lymphodepletion-related cytopenia and resolved or improved to Grade ≤2 within 14 days, excluding disease-related cytopenia. One patient experienced Grade 1 cytokine release syndrome that resolved within 1 day. No ICANS, neurotoxicity, or GVHD, severe infections, or serious TEAEs were observed. Efficacy was evaluable in all patients across B-ALL, follicular lymphoma (FL, grade 3a), and diffuse large B-cell lymphoma (DLBCL). The 3-month ORR was 50% (3/6), with all responders achieving CR; one patient was excluded due to insufficient follow-up. The disease control rate (DCR) was 71.4% (5/7). Two of 3 FL patients achieved CR, while one had a PR and remained under follow-up. One of three B-ALL patients achieved stringent complete remission, maintained through 24 weeks of relapse-free follow-up. CAR transgene analysis demonstrated robust in vivo expansion, with peak levels occurring between Days 7-10 in B-ALL and within two months in FL. GT719 persistence was observed for up to six months post-infusion. Conclusions: GT719 demonstrated a favorable safety profile and encouraging antitumor activity in adults with R/R CD19-positive B-cell malignancies. The absence of severe GT719-related toxicities and evidence of durable cellular persistence support further clinical investigation of this off-the-shelf CAR-NKT therapy. Clinical trial information: NCT06948981 , NCT07131254 . Research Sponsor: Grit Biotechnology.

Green tea extract/catechin supplementation for secondary prevention of metachronous colorectal adenomas after polypectomy.

e15670 Background: Green tea extracts (GTE) have proposed protective effects in colorectal neoplasia, but randomized evidence remains inconsistent. We performed a meta-analysis of randomized controlled trials (RCTs) evaluating oral GTE supplementation following colorectal adenoma resection. Methods: We systematically reviewed RCTs enrolling adults after complete adenoma removal and a surveillance colonoscopy confirming no residual adenomas. Interventions were oral GTE initiated post-polypectomy versus standard surveillance. The primary endpoint was colorectal adenomas (≥1) at surveillance colonoscopy; secondary endpoint was advanced adenomas. Risk ratios (RR) were pooled using random-effects models; heterogeneity was assessed with I². Risk of bias was evaluated using Cochrane RoB2. Absolute risk reduction (ARR) and number needed to treat (NNT) were calculated at specified surveillance intervals. Results: Three RCTs were included (n = 900; 441 GTE vs 459 control); 900 randomized participants completed surveillance colonoscopy and contributed to primary analysis. Overall, GTE was not associated with statistically significant reduction in adenomas (pooled RR 0.69, 95% CI 0.44–1.07; p = 0.10), with substantial heterogeneity (I² = 69%). Differences in follow-up duration and exposure likely contributed, including short-term supplementation (0.9–1.5g/day) versus prolonged standardized decaffeinated extract (300mg day) with endpoint colonoscopy at 26–44 months. In sensitivity analysis restricted to the two 12-month trials, GTE significantly reduced adenomas (RR 0.53, 95% CI 0.36–0.80; p = 0.002; I² = 0%), corresponding to recurrence rates of 19.7% (26/132) vs 36.8% (50/136) (ARR 17.1%; NNT≈6). In contrast, the 3-year MIRACLE trial showed attenuated benefit (51.1% vs 55.7%; RR 0.92; ARR 4.6%; NNT≈22), with no reduction in advanced adenomas (RR 0.97) [Table 1]. Conclusions: GTE supplementation did not significantly reduce colorectal adenomas overall, with substantial heterogeneity. Apparent benefits in short-term trials were not sustained with longer follow-up, and advanced adenoma rates were unchanged. Future trials with harmonized dosing, formulation, and surveillance intervals are needed to clarify clinical relevance. Study Experimental Events/Total Control Events/Total Weight (%) Risk Ratio 95% CI Seufferlein 2022 158/309 180/323 47.2 0.92 0.79–1.06 Shimizu 2008 9/60 20/65 22.1 0.49 0.24–0.99 Shin 2018 17/72 30/71 30.7 0.56 0.34–0.92 Total (Random Effects) 184/441 230/459 100.0 0.69 0.44–1.07 Heterogeneity — — — Tau² = 0.10; χ² = 6.39 (df = 2) P = 0.04; I² = 69% Test for overall effect — — — Z = 1.67 P = 0.10 Total (Random Effects) (excluding Seufferlein 2022) 26/132 50/136 100.0 0.53 0.36-0.80 Heterogeneity (excluding Seufferlein 2022) — — — Tau² = 0.00; χ² = 0.10 (df = 1) P = 0.76; I² = 0% Test for overall effect (excluding Seufferlein 2022) — — — Z = 3.03 P = 0.002

Phenotypic heterogeneity in primary tumors and brain metastases of small cell lung cancer: Insights into tumor microenvironment and metastatic behavior.

8114 Background: Small cell lung cancer (SCLC) is an aggressive malignancy with frequent brain metastases and pronounced phenotypic plasticity. How the tumor microenvironment and phenotypic heterogeneity relate to brain metastases remains poorly understood. Methods: We conducted a retrospective study of 124 patients with pathologically confirmed SCLC treated at a tertiary cancer center. In primary tumors, pathological features—including tumor-infiltrating lymphocytes (TILs), morphological patterns (necrosis, hyalinization, desmoplasia), and immunohistochemical characterization of lineage-associated SCLC markers—were assessed on H&E-stained sections. Immune-desert tumors were defined as having no detectable lymphocytic infiltration. Associations between these pathological features and baseline brain metastases were evaluated using appropriate statistical tests for categorical and continuous variables, with available brain metastasis samples also analyzed histopathologically and immunohistochemically to assess spatial heterogeneity and microenvironmental plasticity. Results: Among 124 patients, TIL levels were lower in those with baseline brain metastases than without (1.46% ± 4.54 vs 3.04% ± 5.29; p = 0.049), and the immune-desert phenotype was more frequent (88% vs 65%; p = 0.041), indicating a cold tumor microenvironment in patients with brain metastases. Immunohistochemical and morphological patterns were similar regardless of baseline brain metastases. Histopathological evaluation of the two available brain metastases showed marked intralesional heterogeneity. Regions had inflammatory infiltrates, coagulative necrosis, or stromal hyalinization. Tumor cells showed a spiculated growth pattern along the GFAP-positive astrocytic network, suggesting interaction with astrocytes. Synaptophysin expression was higher adjacent to lymphocytes, NEUROD1 predominant near necrotic zones, and POU2F3 higher in hyalinized regions. These findings indicate regional activation of different SCLC transcriptional programs, reflecting microenvironmental plasticity. Conclusions: An immune-desert phenotype in primary SCLC tumors is associated with baseline brain metastases, which exhibit pronounced spatial heterogeneity and region-specific tumor programs. These findings suggest that the immune context of the primary tumor may influence microenvironmental plasticity and contribute to brain metastasis development, warranting validation in larger cohorts.

(RW) post-progression outcomes following first-line (1L) ribociclib (RIB) + aromatase inhibitor (AI) versus AI alone in African American and low socio-economic status (SES) patients with hormone receptor-positive/human epidermal growth factor receptor 2-negative (HR+/HER2–) metastatic breast cancer (MBC) in US.

1073 Background: RIB + endocrine therapy (ET) demonstrated significant overall survival over ET alone in all three phase III MONALEESA trials for HR+/HER2− MBC. RW studies can provide insight into pt subgroups who may be under-represented in clinical trials. This is particularly important for African American pts and those with low SES, groups who often face disparities in access and outcomes. This study assessed RW post-progression outcomes among these under-represented populations in the US, to guide treatment (tx) for those affected by social determinants of health. Methods: This retrospective study used deidentified data from the Flatiron Health Research Database, including 5,649 US women with HR+/HER2− MBC who began 1L RIB + AI or AI monotherapy between Jan 1, 2020 and Jan 31, 2025, with follow-up through Jul 31, 2025. Analyses of RW progression-free survival 2 (rwPFS2; time from start of 1L to first disease progression during 2L, or death during 1L or 2L, whichever occurred first) and RW chemotherapy-free survival (rwCFS) used unadjusted and adjusted time-to-event methods (Kaplan-Meier and Cox regression), with baseline differences addressed using inverse probability tx weighting with stabilized weights (sIPTW). Key pt demographics and characteristics were used for adjustment; details to be provided in the full presentation. Results: A total of 520 African American pts were included in the analysis; 243 received RIB + AI and 277 received AI monotherapy. After sIPTW, median rwPFS2 was 19.4 months longer for the RIB + AI group vs AI monotherapy (median 36.8 vs 17.4 months; HR, 0.44; 95% CI, 0.29–0.67). Median rwCFS was also extended by 14.3 months with RIB + AI (median 39.5 vs 25.2 months; HR, 0.49; 95% CI, 0.33–0.72). Among 1805 pts with low SES index (pts area-level SES categorized in the lowest two quintiles to represent the lowest 40% of block groups in terms of SES), 810 received RIB + AI and 995 received AI monotherapy. The median rwPFS2 was 14.2 months longer for those receiving RIB + AI vs AI monotherapy (median 37.2 vs 23.0 months; HR, 0.54; 95% CI, 0.44–0.67) after sIPTW adjustment. Similarly, median rwCFS was prolonged by 22.0 months for the RIB + AI group vs AI monotherapy (median 51.4 vs 29.4 months; HR, 0.58; 95% CI, 0.47–0.71). Conclusions: This study demonstrates that treating African American pts and individuals with low SES using 1L RIB + AI, compared with AI alone, leads to longer rwPFS2 and rwCFS. These results support 1L RIB + AI as an effective tx choice for these pt groups and underscores the importance of ensuring access to optimal tx for better outcomes.

Uncertainty-aware multimodal imaging for lung metastasis risk stratification in extremity soft-tissue sarcoma.

e23543 Background: In the field of soft-tissue sarcoma (STS), imaging-based risk prediction has been based on Radiomics and Deep Learning Models that have been optimized on measures of discrimination, including accuracy or area under the curve. Although feasibility has been established, most of the studies are done on small cohorts with no assessment of predictive uncertainty, implicitly assuming identical reliability for all patients. This is a major gap in rare cancers like STS where the fundamental limitation to model’s confidence is limited by sample size and biological heterogeneity. In this study, uncertainty estimation has been incorporated in multimodal MRI and FDG-PET-based imaging for lung metastasis risk stratification, making the model’s output more useful for clinical decision support. Methods: Data for Pre-treatment MRI and FDG-PET imaging for extremity soft-tissue sarcoma patients was obtained from The Cancer Imaging Archive. Patients were included only if both modalities and confirmed lung metastasis outcomes were available, resulting in a cohort of 51 patients (19 metastatic, 32 non-metastatic). The preprocessed MRI and PET images were individually encoded with independent convolutional neural networks with latent features being combined at the representation stage and the binary cross-entropy loss being used to optimize the model. Monte Carlo dropout was used in estimating predictive uncertainty during inference. Strict leave-one-patient-out cross-validation was used for model training and evaluation to guarantee a complete patient-level separation. Performance measurement involved accuracy, area under the receiver operating characteristic curve, uncertainty error correlation and uncertainty-based deferral. Another secondary 10-patient exploratory analysis examined feasibility in case of extreme data scarcity. Results: The model achieved a leave-one-out accuracy of 62.8% with an AUC of 0.28, reflecting inherent heterogeneity of the lung metastasis prediction method in a rare and heterogeneous population of sarcomas. False Forecasting was linked with greater uncertainty and pulling out the cases with the high level of uncertainty resulted in improved accuracy among the remaining patients. Similar uncertainty-performance patterns were observed in the 10-patient exploratory analysis, where the accuracy reached 80% among low uncertainty cases. Conclusions: Although the previous studies on sarcoma imaging focus on a point-estimate, this study shows that under the worst conditions of severe data constraints, uncertainty-aware multimodal models prove to be safer in assisting clinical decision-making. The suggested approach focuses on defining situations when predictions cannot be trusted, which is why our approach fills one of the existing gaps in the literature of sarcoma AI and corresponds to the safety-critical oncology processes.

A phase 2 clinical trial in progress of (Z)-endoxifen plus goserelin as neoadjuvant therapy in premenopausal women with ER+/HER2– breast cancer (EVANGELINE).

TPS655 Background: Early (14-28 day) on-treatment suppression of tumor proliferation during neoadjuvant endocrine therapy (NET) is strongly associated with favorable long-term outcomes in estrogen receptor–positive (ER+) HER2– breast cancer. In premenopausal women, tamoxifen with or without ovarian function suppression (OFS) results in suboptimal Ki-67 suppression compared with aromatase inhibitor + OFS. (Z)-endoxifen (ENDX), the active metabolite of tamoxifen, dually inhibits ER signaling and PKCβ1-mediated AKT activation, supporting its evaluation as an alternative NET strategy in this population. Methods: EVANGELINE (NCT05607004) is an ongoing, multicenter, open-label Phase 2 study evaluating daily 40 mg ENDX plus goserelin administered every 28 days as neoadjuvant therapy in premenopausal women with ER+/HER2–, cT2–3, cN0–1 breast cancer. The primary objective is to determine the proportion of patients with baseline Ki-67 >10% who achieve Ki-67 ≤10% after 4 weeks of therapy. A Simon two-stage design is used to test whether the Week 4 Ki-67 ≤10% rate is at least 65%, with 20 patients enrolled into the first stage and, if promising, another 25 patients enrolled into the second stage (cohort A). A parallel cohort of 20 patients with baseline Ki-67 ≤10% (cohort B) is enrolled to assess objective response rate (ORR) at 24 weeks per RECIST v1.1. Secondary objectives include examining safety and tolerability, residual cancer burden, and PEPI score. Correlative analyses include examining effect of treatment on select tumor and plasma biomarkers. Clinical trial information: NCT05607004 .

Optimal subsequent-line strategies for <i>EGFR</i> -mutant NSCLC after TKI progression: A systematic review and network meta-analysis.

e20506 Background: Acquired resistance to epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) is inevitable in most patients with EGFR- mutant advanced non-small-cell lung cancer (NSCLC). In cases of systematic progression with unknown resistance mechanisms, post-TKI options vary, but the lack of direct comparisons hinders decision-making. We evaluated the efficacy and safety of subsequent-line treatments to identify optimal strategies. Methods: A systematic review and network meta-analysis (NMA) were conducted by searching PubMed, Embase, and the Cochrane Library up to October 10, 2025, plus conference abstracts from ASCO, ESMO, and WCLC (2015–2025). Randomized controlled trials of post-TKI treatments in EGFR -mutant NSCLC were included. Primary outcomes were progression-free survival (PFS) and overall survival (OS); secondary outcomes included objective response rate (ORR) and grade ≥3 adverse events (≥3 AEs). A Bayesian framework was employed to estimate hazard ratios (HRs) and odds ratios (ORs) with 95% confidence intervals (CIs), and the surface under the cumulative ranking curve (SUCRA) was used to rank treatments. Subgroup analyses were conducted by EGFR mutation subtypes and clinicopathological features. The protocol is registered with PROSPERO (CRD420251232072). Results: Ten RCTs (3,820 patients, 13 regimens) were analyzed. Compared with chemotherapy alone, amivantamab plus lazertinib and chemotherapy (AMI+Lazer+CT) showed the best PFS (HR 0.44, 95% CI 0.35–0.56). For OS, the TROP2 antibody-drug conjugate (ADC) sacituzumab tirumotecan (SACTMT) ranked first (HR 0.60, 95% CI 0.44–0.82). Atezolizumab plus bevacizumab and chemotherapy (ATE+BEV+CT) yielded the highest ORR (OR 3.26, 95% CI 1.79–5.94). Amivantamab-based regimens were associated with elevated toxicity; AMI+Lazer+CT had the highest risk of ≥ 3 AEs (OR 14.59, 95% CI 8.99–24.94). AMI+CT was optimal for L858R mutation (HR 0.30, 95% CI 0.17–0.54), whereas SACTMT was superior for 19DEL (HR 0.42, 95% CI 0.31–0.57). ATE+BEV+CT provided the best PFS benefit (HR 0.44) for T790M-negative patients, while ivonescimab plus chemotherapy (IVO+CT) ranked first for T790M-positive (HR 0.21). AMI+Lazer+CT ranked first for PFS in patients aged ≥ 65 years (HR 0.41), smokers (HR 0.45), and those without brain metastases (HR 0.42). Conversely, ATE+BEV+CT significantly improved PFS in patients with brain metastases (HR 0.32, 95% CI 0.19–0.53). Conclusions: AMI+Lazer+CT and SACTMT are the optimal regimens for PFS and OS for EGFR -mutant NSCLC after TKIs progression, respectively, though the toxicity of amivantamab-based regimens requires management. SACTMT and AMI+CT are preferred for 19DEL and L858R, while ATE+BEV+CT and IVO+CT are the respective treatments for patients with and without brain metastases, highlighting the need for personalized treatment.