Browse Articles

Bridging science and sustainability: Applications of nanomaterials in wastewater remediation

Electrically Modulated Semi‐Convertible Hydrogel Microneedles for Programmable Insulin Delivery

ABSTRACT Insulin delivery systems that mimic pancreatic secretion offer promising improvements in type 1 diabetes management. However, current systems struggle with sustained and precise release. Inspired by the dynamic secretion mechanisms of the pancreas, a wearable electrically modulated semi‐convertible hydrogel microneedle system for insulin delivery has been developed for long‐term self‐regulation in the treatment of type 1 diabetes. This system combines a non‐covalent electro‐responsive silk fibroin network and a polyethylene glycol/chitosan matrix, integrated with flexible electrodes that exhibit a partial gel‐sol phase transition under 1.2 V electrical stimulation. In hyperglycemic conditions, the electric field induces silk fibroin phase transition, triggering insulin release through electrostatic interactions and polymer network expansion. Once blood glucose normalizes, insulin is released passively via diffusion when the electric field is turned off. In type 1 diabetic mice, a single microneedle patch provides segmented control, with a 10‐min stimulation effect lasting up to 8 h, extendable to 16 h with further stimulation. This system offers a versatile, sustained, and precise drug delivery strategy with significant potential for chronic disease management.

Superior Intermetallic Pt‐Co/C Catalysts With Optimized Triply Metal‐Loading, Size, and Ordering‐Degree for High‐Efficiency and Stable H ₂ –Air Fuel Cells

ABSTRACT Pt‐based intermetallic compounds (IMCs) are promising electrocatalysts for oxygen reduction reaction (ORR). This work presents small‐sized (2–4 nm) Pt‐Co IMCs with both high Pt loading (40–60 wt.%) and ordering degree (42%–91%) via an optimized L‐cysteine hydrochloride‐assisted impregnation. Besides heating rate and calcination temperature, metal loading is identified as a critical factor governing IMCs formation. Quantitative XRD analysis reveals that the IMCs fraction reaches 78.8% (29.4% L1 0 ‐PtCo‐IMC + 50.4% L1 2 ‐Pt 3 Co‐IMC) at 40% loading, peaks at 94.5% (68.4% L1 0 ‐PtCo‐IMC + 26.1% L1 2 ‐Pt 3 Co‐IMC) at 50% loading, and declines to 31.6% (solely L1 0 ‐PtCo‐IMC) at 60% loading. The 50% Pt‐Co‐IMCs/C with the highest L1 0 ‐PtCo content delivers the best ORR performance, achieving a mass activity of 0.80 A mg Pt −1 —1.7 times that of commercial PtCo‐TKK. In membrane electrode assembly tests under H 2 ‐air with a low Pt usage of 0.16 mg cm − 2 , a voltage of 0.603 V is achieved at 2.0 A cm − 2 , with only 11.4% voltage decay after 30 000 cycles. Density functional theory calculations corroborate the activity trend of L1 0 ‐PtCo >L1 2 ‐Pt 3 Co >disordered PtCo, thereby rationalizing the loading‐dependent ORR activity. The synergistic effect between optimized * OH desorption kinetics in the rate‐determining step and strong Pt‐Co d ‐ d orbital coupling jointly enhances the activity and stability.

Psychosocial interventions in the care of individuals with hereditary cancer risk: A scoping review.

e22660 Background: Individuals with hereditary cancer predisposition syndromes experience significant psychological distress throughout the genetic counseling pathway. However, the integration of structured psychosocial care into genetic services remains unclear. Methods: This review mapped psychosocial interventions integrated into hereditary cancer genetic counseling process using PRISMA-ScR, searching Embase, PubMed, Cochrane, Scopus, CINAHL and SciELO. Inclusion criteria encompassed studies reporting psychosocial interventions delivered as part of genetic counseling/testing process that target patients with possible or confirmed hereditary cancer predisposition syndromes. From 621 publications identified, 8 met inclusion criteria after full-text review. Interventions targeted distress, decision-making, family communication, and health behaviors. Most interventions were executed by physicians, with limited mental health specialist involvement. Results: Psychosocial interventions were associated with reductions in depression, anxiety, and distress, better understanding of genetic risk, improved genetic testing uptake and counseling acceptance, and healthier lifestyle behaviors, although follow-up rarely extended beyond 12 months and outcome measures varied. Reporting of intervention components, training requirements, and procedures was often incomplete, limiting reproducibility and implementation in routine practice. Conclusions: Findings highlight the need for clearly described, psychologist-led or co-led protocols, integrated outcome metrics and longer follow-up to support scalable, evidence-based psychosocial care in hereditary cancer genetic services. Study characteristics. Syndrome Intervention Duration Professionals Key outcomes 1 HBOC Tailored phone counseling (CBT) 12 months Not reported Higher Cancer Genetic Risk Assessment uptake 2 BRCA1/2 Psychosocial telephone counseling + workbooks 6–12 months Mental health counselors, oncologists Reduced depression and testing distress; lower clinically significant anxiety at 6 months 3 HBOC Web-based “Family Gene Toolkit” Cross-sectional Physicians High satisfaction; 78% would have preferred access at time of own testing 4 HBOC Spanish video + navigation 3 months Nurse + other staff 72% attended counseling; 70% completed testing 5 BRCA1/2 Telephone vs in-person counseling + mailed visual aids (CBT) 3 months Physicians Noninferior for knowledge, distress, satisfaction; reduced cost per patient 6 BRCA1/2 One-day psychoeducational retreat 6 months Physicians 61% increased screening or prophylactic surgery; 50% improved exercise or diet 7 HBOC Problem-solving training (CBT) 6–9 months Physicians Greater reduction in depressive symptoms than client-centered counseling 8 HBOC Group therapy 6 months Psychiatrist Reduced depression, anxiety, avoidance; improved grief processing

Leveraging spatial proteomics to identify epithelial cell subtypes driving poor prognosis in prostate cancer.

5126 Background: Epithelial cell features, such as morphology and cellular organization within the tumor architecture (e.g. Gleason Score), have long been known to be prognostic in prostate cancer. However, the specific epithelial cell subtypes driving clinical outcomes in patients remains unclear. Traditional tumor profiling techniques lack the resolution needed to interrogate complex single-cell biomarkers. Imaging Mass Cytometry (IMC) now enables high-dimensional identification and detailing of epithelial, stromal, and immune cell subtypes within patient tumors. Using a custom IMC panel, we profiled tumor biopsies from a large prospective biopsy cohort with long-term (median >12-year) clinical follow-up to identify potentially targetable epithelial cell populations associated with adverse outcomes in localized prostate cancer. Methods: Spatially resolved protein expression profiling was performed on primary prostate cancer tumor biopsies using a custom IMC assay enriched for targetable tumor markers. Cell segmentation and single-cell expression measurement were performed using the published ‘steinbock’ toolkit. Biochemical progression-free survival (bPFS) and cancer-specific survival (CSS) were pre-specified clinical endpoints. Univariate and multivariable survival analyses stratified by cell abundance tertiles were performed using a Cox proportional hazards model. The Mann-Whitney U test was used to assess for pairwise differences in cell abundance between patient groups. All significance testing was performed using a two-tailed significance level of 0.05. Results: Protein co-expression patterns in >3.4 million cells comprising 573 biopsy samples obtained from 385 patients were measured. Single-cell analysis revealed 15 cell clusters representing luminal prostate cancer cells (including PSMA-high, PSMA-intermediate, and PSMA-low), basal epithelial cells, and lymphocytes. Patients with tumors enriched for PSMA-high epithelial cells demonstrated impaired bPFS ( P <0.001) and CSS ( P =0.016). PSMA-high epithelial cells were enriched in high-grade (Gleason Score 8+) tumors (P<0.05). Multivariable analysis revealed PSMA-high epithelial cell enrichment to be prognostic of bPFS independently of Gleason Score and serum PSA at time of diagnosis ( P <0.05). Co-expression analysis demonstrated that these prognostic PSMA-high cells also expressed high levels of the cell-surface targets KLK2, B7-H3, and protein. Conclusions: We identified an epithelial single-cell biomarker associated with adverse clinical outcomes in localized prostate cancer. If validated through ongoing experiments in an independent cohort, our findings support new potential strategies for treatment intensification using targeted therapies in select high-risk patients.

Quantitative results from PSMAtrack: A prospective study evaluating changes in PSMA-PET during initial systemic therapy for metastatic hormone-sensitive prostate cancer (mHSPC).

5102 Background: PSA <0.2 ng/mL after 6 months (m) of therapy with androgen deprivation therapy (ADT) and an androgen receptor pathway inhibitor (ARPI), ± docetaxel is strongly prognostic in patients (pts) with mHSPC. This prospective, single-institution pilot study aimed to investigate PSMA-PET for assessing early dynamic changes in PSMA expression and tumor volume in response to treatment for mHSPC and to determine if changes in quantitative PSMA-PET parameters are associated with 6m PSA response. Methods: Key eligibility was mHSPC by conventional imaging with plan to start ADT and ARPI ± docetaxel. Primary or metastasis-directed radiotherapy was not permitted during the first 6m. F18-flotufolastat PSMA-PET was performed at baseline within 21 days of ADT start and after 6m of ADT+ARPI±docetaxel. PSMA-PET scans were qualitatively reviewed for PSMA-avid prostate cancer, (i.e., PSMA uptake > background blood pool in a pattern consistent with prostate cancer). Quantitative PSMA-PET parameters were collected including PSMA-avid total tumor volume (TTV, voxel SUV>3), TTV SUVmean and TTV SUVmax (aPromise, EXINI Diagnostics; LesionID MIM software, GE Healthcare). PSA values at baseline and after 6m of therapy were collected. Descriptive statistical analyses, Wilcoxon test, and Pearson correlations were used. Results: 20/21 enrolled pts (median age 70 years [range 60-95]) had evaluable baseline and 6m PSMA-PET scans. 12 (60%) received ADT+ARPI, 8 (40%) received ADT+ARPI+docetaxel, and median baseline PSA was 55ng/mL (range 2-3,696); 13 (65%) had high-volume mHSPC by CHAARTED criteria. At 6m, median PSA was 0.29 ng/mL (range <0.02-177), and 9 pts (45%) had a PSA ≤0.2 ng/mL. All 20 pts had residual PSMA-avid disease at 6m by qualitative review, with most (n=17, 85%) having residual disease outside the prostate; 2 pts (10%) had new lesions seen on 6m PSMA-PET. At 6m, median TTV and SUVmax, but not SUVmean, were significantly lower for those with PSA ≤0.2 than PSA >0.2 ng/mL (Table); baseline TTV was strongly correlated with 6m PSA (r=0.74, p<0.01). PSA and TTV were well correlated at baseline (r=0.68, p=0.001) and at 6m (r=0.81, p<0.01), as were changes in PSA and TTV from baseline to 6 m (r=0.67, p=0.001). Conclusions: Persistent PSMA-avid disease was seen in all pts with mHSPC after 6m of ADT+ARPI±docetaxel, regardless of PSA at 6m. Pts with 6m PSA >0.2ng/mL had significantly higher TTV and SUVmax at 6m PSMA-PET, and TTV best correlated with PSA levels at baseline and 6 m. Using 6m PSMA-PET could identify mHSPC pts with suboptimal initial response to standard therapy for consolidative therapeutic strategies. Clinical trial information: NCT06479187 . 6m PSA≤0.2 ng/mL, median (range) 6m PSA>0.2 ng/mL, median (range) P TTV (mL) 7.7 (0.4-79.9) 147 (5.9-3440) <0.01 SUVmax 8.7 (4.8-39.2) 47.4 (11.7-162) <0.01 SUVmean 5.3 (4.0-8.8) 6.4 (4.2-16.8) 0.14

Clinical validity of ultrasensitive single-digit parts per million ctDNA detection in non–small cell lung cancer.

8017 Background: Circulating tumor DNA (ctDNA) is a promising biomarker for detecting molecular residual disease (MRD) and predicting recurrence after curative treatment in non-small cell lung cancer (NSCLC). Ultrasensitive personalized assays have demonstrated clinical validity at detection thresholds of ~80-100 parts per million (PPM). As this assay can detect ctDNA at concentrations an order of magnitude lower (below 10 PPM), we extend these analyses to explore the clinical outcomes of patients with single-digit PPM ctDNA. In the future, this may guide adjuvant treatment decision-making. Methods: We analyzed longitudinal plasma samples from 431 patients with stage IA–IIIB NSCLC enrolled in the TRACERx study and treated with upfront surgery. We achieved ultrasensitive MRD testing with ctDNA limits of detection as low as 1-3 PPM across 2994 plasma samples using NeXT Personal, a personalized liquid biopsy assay which tracks up to ~1,800 patient-specific somatic variants. Plasma samples were classified as preoperative, postoperative landmark (10–120 days), or longitudinal follow-up. Associations with recurrence-free survival (RFS) and overall survival (OS) were assessed. Results: Preoperative ctDNA detection below 10 PPM was associated with inferior recurrence-free survival (RFS) and overall survival (OS) compared to undetectable ctDNA in individuals with lung adenocarcinoma (<10 PPM vs. undetected: HR: 3.25, 95% CI 1.52-6.94, p = 0.0024 and HR: 3.49, 95% CI 1.24-9.81, p = 0.018 for RFS and OS). Postoperative ctDNA was detected in 21% of patients. Detection below 10 PPM occurred in 16.7%, 15.5%, and 11.5% of stage I, II, and III patients, respectively. ctDNA detection preceded imaging-confirmed recurrence in 120 individuals. In postoperative landmark plasma samples, 18% of ctDNA detections occurred below 10 PPM and were associated with increased risk for recurrence and death versus undetectable ctDNA (RFS: HR 3.49, 95% CI 1.70–7.17, p=0.00067; OS: HR 2.58, 95% CI 1.09–6.14, p=0.032). Within ctDNA-positive patients, <10 PPM was associated with improved outcomes compared with ≥10 PPM (HR: 0.38, 95% CI 0.19-0.78, p = 0.0086, HR: 0.37, 95% CI 0.16-0.86, p = 0.021). Conclusions: Sub-10 PPM ctDNA detection identifies a clinically high-risk subgroup of early-stage NSCLC patients. Notably, this included stage I patients who would not typically be classified as high-risk or considered for adjuvant therapy; ctDNA positivity in this group therefore reveals high-risk disease that would otherwise go unrecognized. Reliable detection of ctDNA at sub-10 PPM levels may refine post-surgical risk stratification beyond conventional staging and imaging, enabling earlier identification of molecular relapse and potentially guiding treatment escalation in future practice.

Progress report of a phase 2 study of BPM31510 (a lipid nano dispersion of oxidized CoQ10) with vitamin K in combination with standard of care (SOC) RT and TMZ in glioblastoma multiforme (GBM) patients without prior therapy.

TPS2101 Background: BPM31510 is an investigational, drug-lipid conjugate nanodispersion. Preclinical studies demonstrate that intravenous infusion can achieve plasma concentrations of oxidized CoQ10 approximately 10,000 times higher than baseline physiological levels with its accumulation in organs, including brain. Preclinical studies show that supraphysiological concentrations of CoQ10 induce metabolic shifts in a variety of cancer cells and generate Reactive Oxygen Species (ROS) leading to apoptosis. In Phase 1, the main adverse events were related to mild hepatotoxicity and coagulopathy. All coagulation related issues were alleviated by coadministration of Vitamin K. The Phase 2 trial in newly diagnosed patients with glioblastoma tests whether treatment with BPM31510 results in improved efficacy of standard of care temozolomide (TMZ) and radiotherapy (RT). Methods: BPM31510IV-11; NCT04752813 is a single-arm, non-randomized, open-label, phase 2 study of intravenous (IV) BPM31510 + prophylactic subcutaneous (SC) Vitamin K1 with standard chemoradiation and temozolomide (TMZ) in newly diagnosed GBM patients. The primary endpoint is Progression Free Survival at 6 months (PFS6) and 12 months (PFS12) while the secondary endpoints are Overall S (OS) survival and safety. Patients with recent hemorrhage, coagulopathy, or requiring anticoagulants are excluded. SC Vit K and IV BPM31510 are initiated 2-4 weeks after biopsy or tumor resection. After two weeks, standard of care chemoradiation and TMZ are started in combination with weekly IV BPM31510 for 8 weeks, after which IV BPM31510 treatment and standard RT are discontinued with subjects continuing to receive maintenance TMZ treatment for up to 12 cycles as per standard of care. Of the approximately 50 patients enrolled and treated upon the recommendation of an independent safety monitoring committee, 25 will be available for PFS6 efficacy endpoint analysis and these data will be presented. Clinical trial information: NCT04752813 .

PRISM: A transparent, self-correcting LLM system for reproducible therapy recommendations in neuroendocrine tumors.

e13657 Background: The clinical application of Large Language Models (LLMs) in oncology is currently limited by opaque reasoning and the potential for "hallucinations," which pose safety risks for Multidisciplinary Tumor Boards (MTBs). Neuroendocrine tumors (NETs) specifically require precise interpretation of complex guidelines and recent primary evidence from clinical trials that are not already implemented in the guidelines. To address this need, we present PRISM (Personalized Recommendations via Integrated Synthesis & Modeling), a multi-agent LLM architecture designed to improve the quality, traceability, and reproducibility of therapy recommendations in NET care through a transparent, self-correcting workflow. Methods: PRISM employs a structured seven-stage fully automated, LLM-based workflow decomposing clinical reasoning into explicit steps, including guideline selection (e.g. ENETS & ESMO), initial patient case analysis, and search for matching trial evidence. A validation agent autonomously verifies matched evidence against existence (e.g., trial numbers), triggering correction loops to correct hallucinations or citation errors. In a feasibility study of 15 complex NET cases, we compared PRISM against a standard single-pass LLM (baseline). Performance was assessed by an expert using a structured framework evaluating clinical quality and safety given the patient’s organ functions, together with an overall quality rating (1–10) and implementation willingness (Yes/Maybe/No). Results: PRISM autonomously identified an average of 12 clinical trials per patient with potentially relevant primary evidence for treatment options, from which 66.7% were used in the treatment recommendations. The validation agent rejected 93% (14/15) of initial drafts due to hallucinations or citation errors. Autonomous re-processing (mean 2.6 iterations) led to full pass of 73% (11/15) of cases. In a blinded head-to-head comparison, the baseline model produced unsafe recommendations in 13% of cases, leading to hard implementation rejections (“No”) and expert-identified hallucinations. In contrast, PRISM reduced unsafe outputs to 6.7% and eliminated all hard rejections, achieving 100% implementation potential (60% “Yes”, 40% “Maybe”). PRISM achieved higher consistency in output quality (range: 5–9 vs. 1–10). Conclusions: PRISM addresses key limitations of standard LLMs by replacing opaque generation with a transparent, self-correcting validation loop. It systematically identifies relevant guidelines and primary evidence from clinical trials. In this feasibility study, the system ensured trustworthiness by enforcing strict evidence traceability. These findings support the prospective evaluation of multi-agent LLM systems as decision-support tools for treatment decisions in oncology, especially to bridge the expertise gap in non-specialized settings.

Brain radiotherapy and risk of wound healing following Ommaya reservoir placement in leptomeningeal disease (LMD).

e13579 Background: Leptomeningeal disease (LMD) is a late-stage sequelae of solid and hematologic disease progression. Development of LMD generally portends a poor prognosis. Aggressive treatment with cranial spinal radiotherapy and intrathecal therapy (IT) shows promise in extending survival; whereas delays in treatment could negatively impact survival. Ommaya reservoirs are the most common implant for IT therapy. The timing of brain radiation is often delayed following Ommaya placement due to concerns of wound infection or delayed wound healing. We seek to determine whether whole brain radiation (WBRT) negatively impacts wound healing following Ommaya reservoir placement. Methods: The authors performed a retrospective chart review on all patients who underwent Ommaya reservoir placement for LMD at an NCI-accredited comprehensive cancer center between 2007 and 2025. Patient demographics, clinical characteristics, tumor characteristics, IT regimen, radiotherapy regimen, infection status, and mortality data were collected. Results: A total of 198 patients were identified who underwent Ommaya reservoir placement and perioperative irradiation. Of those, 165 (83.3%) patients underwent WBRT, with the remainder undergoing targeted radiotherapy. 68% of the patients were female. Breast cancer was the most common histology (46%), followed by melanoma (15%) and lung cancer (14%). 43% of patients were active smokers and 13% had diabetes mellitus. 28% of patients received WBRT within 7 days after Ommaya reservoir placement, 26% received WBRT within 7 days prior to surgery, 25% received WBRT 15-30 days after surgery, and 12% received WBRT 15-30 days prior to surgery. The rate of infection among these patients was 2.4%. 50% of infections occurred 30-90 days after surgery, while the other half occurred more than 90 days after surgery. There was no correlation between infections and timing of radiation. The overall cohort 30- and 90-day mortality rates were 10.6% and 12.1%, respectively. Among patients who were infected, there were no 30-day mortalities, and there was one 90-day mortality. Conclusions: WBRT does not appear to increase the risk of wound infection before or after Ommaya reservoir placement. These findings offer opportunities for future studies examining the timing of WBRT and IT directed therapy in LMD.

Efficacy and safety of the pertuzumab biosimilar BCD-178 versus the originator pertuzumab in patients with HER2-positive locally advanced breast cancer: Results from the international, double-blind, randomized, phase III PREFER trial.

613 Background: Pertuzumab, a HER2-targeting monoclonal antibody, significantly improves outcomes for patients with HER2-positive breast cancer (HER2+ BC) in different settings in combination with another HER2-agents with/without chemotherapy. BCD-178 is a biosimilar of the originator pertuzumab (OP). Preclinical and phase I clinical studies have demonstrated its equivalence to OP. The phase III PREFER trial compared the efficacy and safety of BCD-178 and OP as neoadjuvant and adjuvant therapy in HER2+ BC. Methods: The study hypothesized that neoadjuvant therapy containing BCD-178 would yield a pathological complete response (pCR) rate equivalent to that of therapy containing OP in women aged 18–75 years with stage II–III HER2+, hormone receptor-negative invasive BC and a primary tumor > 2 cm. Patients were randomized 1:1 to receive neoadjuvant therapy with either BCD-178 or OP, combined with trastuzumab and chemotherapy (TCHP regimen). The primary endpoint was the pCR rate (ypT0/is, ypN0) assessed by central independent review. Safety was monitored throughout, with adverse events (AEs) graded by CTCAE v5.0. Immunogenicity and pharmacokinetics were also evaluated (NCT05802225). Results: Of 398 randomized patients, 201 received BCD-178 and 197 received OP. The median age was 53 years; over one-third of patients in each group had stage IIIB/IIIC disease. pCR rates were 75.0% with BCD-178 and 72.2% with OP (risk ratio [RR] 2.7, 95% CI: -6.0 to 11.6). The 95% CI for the between-group difference in pCR rate fell within the predefined equivalence margins, confirming equivalent efficacy. The safety profiles were comparable. The most common AEs (≥10%) were diarrhea and anemia. Immunogenicity was low and similar between groups: binding antibodies were detected in 1.6% (BCD-178) vs. 1.1% (OP) of patients, and neutralizing antibodies in 1.1% in both groups. The concentration-time profiles and pharmacokinetic profiles (Cmin, Ceoi) were also comparable. Conclusions: In patients with HER2+, hormone receptor-negative locally advanced breast cancer, the biosimilar BCD-178 demonstrated equivalent efficacy, safety, immunogenicity, and pharmacokinetics to the originator pertuzumab within a neoadjuvant TCHP regimen. Clinical trial information: NCT05802225 .

Prospective, randomized, placebo-controlled phase 2b trial of SurVaxM plus adjuvant temozolomide in newly diagnosed glioblastoma (SURVIVE).

TPS2106 Background: Survivin is widely expressed in glioblastoma (GBM) and is implicated in tumor cell survival and treatment resistance, supporting survivin-targeted immunotherapy in newly diagnosed disease. The survivin long-peptide conjugate vaccine SVN53-67/M57-KLH (SurVaxM) has shown feasibility, immunogenicity, an acceptable safety profile and promising efficacy, in phase 1 and 2 trials in newly diagnosed GBM. We hypothesize that SurVaxM combined with standard adjuvant temozolomide (TMZ) post chemo radiation will improve outcomes in newly diagnosed GBM and that correlative biomarkers will define patient selection. Methods: SURVIVE is a prospective, randomized, double-blind, placebo-controlled, multicenter phase 2b trial evaluating SVN53-67/M57-KLH plus standard adjuvant TMZ versus placebo plus standard adjuvant TMZ in adults with newly diagnosed GBM. Patients are randomized 3:2 to investigational vaccine (Arm A) or placebo (Arm B), with stratification by Karnofsky performance status, MGMT promoter methylation status, and IDH status. Study intervention consists of four priming SurVaxM injections administered every 2 weeks followed by maintenance dosing every 8 weeks, administered alongside standard adjuvant TMZ (5 days every 28-day cycle; typically, 6-12 cycles per standard practice). The vaccine regimen is delivered as a survivin peptide conjugate with immunologic adjuvants; blinding is maintained via matched placebo injections. MRI surveillance is performed per protocol to support response assessment and the planned imaging correlatives. Key eligibility criteria include age ≥18 years; Karnofsky performance status ≥70; pathologically confirmed cerebral GBM meeting 2021 WHO criteria with available MGMT and IDH status; gross-total or near-total resection with minimal residual enhancement on immediate postoperative MRI (≤72 hours); completion of chemoradiation with ≥75% of planned concurrent temozolomide; no evidence of progression on post-chemoradiation MRI; randomization within approximately 16–18 weeks of resection; and dexamethasone ≤4 mg/day at enrollment. The planned evaluable sample size is 228 patients (n=137 Arm A; n=91 Arm B). The primary analysis will be conducted in the intention-to-treat population using a two-sided alpha of 0.05, with stratified time-to-event comparisons to account for key baseline factors. The alternate hypothesis assumes a 1-year survival rate of 75% in the SurVaxM group (Arm A; comparable to a median overall survival (mOS) of 29 months assuming an exponential distribution) and 60% in the control group (Arm B; comparable to a mOS of 16 months assuming an exponential distribution). Clinical trial information: NCT05163080 .

Machine learning–based prognostic models for mortality prediction after unplanned hospitalization in cancer patients.

11122 Background: Unplanned hospitalizations in oncology patients often indicate acute clinical deterioration and are associated with high mortality. Prognostic assessment at admission remains a challenge, particularly in resource-limited settings. Machine learning (ML) methods can integrate clinical and laboratory features to generate individualized mortality predictions, although the data is scarce. We aimed to develop and validate ML-based models to predict in-hospital, 30-day, and 90-day all-cause mortality in adult cancer patients with unplanned hospitalizations. Methods: We retrospectively analyzed patients with cancer who were admitted for unplanned hospitalizations at Hacettepe University Oncology Hospital between January 2018 and August 2024. Class imbalance was addressed using SMOTE, and ANOVA-based feature selection retained the top 20 predictors. Nine ML algorithms were tested, including SVM, kNN, decision trees, neural networks, and ensemble models. Model performance was assessed via 5-fold cross-validation using accuracy, precision, recall, F1-score, and AUC. Results: A total of 1646 cancer patients with unplanned hospitalizations were included. The most common cancers were gastrointestinal malignancies (32.8%), lung cancer (17.4%), and breast cancer (9.8%). For in-hospital mortality prediction, Random Forest achieved the best performance on the original dataset (AUC-ROC: 0.89; accuracy: 84.6%). After SMOTE, SVM yielded the superior results (AUC-ROC: 0.98; accuracy: 93.6%). For 30-day mortality prediction, Random Forest was optimal pre-SMOTE (AUC-ROC:0.85; accuracy:83.6%), whereas LightGBM reached the highest post-SMOTE performance (AUC-ROC:0.97; accuracy:91.8). For 90-day mortality prediction, XGBoost demonstrated the best performance on the original dataset (AUC-ROC: 0.84; accuracy: 80.4%). Following the application of SMOTE, LightGBM outperformed the other models (AUC-ROC:0.90; accuracy:82.3%). Conclusions: This first study focusing on predicting in-hospital mortality in cancer patients with unplanned admissions shows that ML models based on routine admission data can reliably predict early- and long-term mortality. Incorporating such models into workflows may support early risk stratification, shared decision-making, and individualized care planning.

Trends and disparities in lung cancer and atrial fibrillation/flutter–related mortality in the United States, 2000–2023.

e23367 Background: Lung cancer remains a leading cause of mortality among U.S. adults, frequently complicated by cardiovascular comorbidities that exacerbate disease severity and limit treatment options. This study aims to analyze and interpret annual mortality trends and disparities among adults in the United States from 2000 to 2023, for various demographic and geographic factors. Methods: The mortality data from the CDC WONDER multiple cause of death files for adults aged ≥25 years were used to analyze age-adjusted and crude mortality rates (AAMRs and CMRs) per 1,000,000 through ICD 10 code: C34 (lung cancer) and ICD 10 code: I48 (atrial fibrillation/flutter), stratified by year, gender, race/ethnicity, place of death and geography. Joinpoint regression was used to estimate average annual percent change (AAPC) and annual percent change (APC) with 95% confidence intervals (CIs). Statistical significance was defined as p < 0.05. Results: From 2000 to 2023, a total of 99,451 deaths were attributed to lung cancer with coexisting atrial fibrillation/flutter, most occurring among medical facility inpatient settings. The overall AAMR increased from 11.87 in 2000 to 25.83 in 2023 (AAPC: 3.38; 95% CI: 2.80 to 3.96), with the most significant increase observed between 2018 and 2021 (APC: 6.90; 95% CI: 2.85 to 11.11). Men had a higher AAMR than women (25.07 vs 12.71). However, the rise in mortality was more pronounced among women than among men (AAPC: 4.15 vs 2.71). Adults aged 65 and above experienced the highest CMR (80.50), though the increase was more rapid among adults aged 45–64 years (AAPC: 5.00 vs 3.14). The highest AAMR was observed among non-Hispanic (NH) Whites (20.32), while the lowest was noted among Hispanics or Latinos (5.74). Geographic disparities were evident, with the Midwest having the highest AAMR (19.34) and the West having the lowest AAMR (17.03). Non-metropolitan areas showed a sharper rise in mortality (AAPC: 4.43 vs 3.07) and a higher AAMR than metropolitan areas (19.47 vs 16.44). At the state level, Vermont and Nebraska ranked highest, falling in the top 90th percentile during 2000–2020 and 2021–2023, respectively. Conclusions: Mortality trends related to lung cancer and atrial fibrillation/flutter have increased over the past two decades, placing a disproportionate burden among men, NH White individuals, and those residing in the Midwest region and non-metropolitan areas. This underscores the need for targeted interventions and delivery of equitable healthcare for the high-risk population. Average annual percent change (AAPC) of age-adjusted mortality rates for Lung Cancer and Atrial Fibrillation/Flutter in the United States, 2000 to 2023. Variable Deaths AAPC (95%CI) Overall 99,451 3.38 (2.80 to 3.96) Male 58,818 2.71 (1.82 to 3.60) Female 40,633 4.15 (3.73 to 4.56) Non-metropolitan areas 16,108 4.43 (4.10 to 4.76) Metropolitan areas 62,053 3.07 (2.92 to 3.22)

Pharmacokinetics and pharmacodynamics of the GCN2 activator HC-7366 in a phase 1a study in patients with advanced solid tumors.

e15103 Background: HC-7366 is a novel, selective and potent activator of general control nonderepressible 2 (GCN2), a core kinase in the integrated stress response (ISR). GCN2-driven ISR hyperactivation broadly affects metabolic pathways important for tumor progression. HC-7366 has shown preclinical anti-tumor activity alone and in combination with varied standard of care therapies across diverse cancer types. HC-7366 was evaluated in a Ph1a study to determine its maximum tolerated dose and safety in patients with advanced solid tumors. Herein, we describe the pharmacokinetic (PK) and pharmacodynamic (PD) effects of HC-7366 in this study. Methods: The Ph1a first-in-human, open-label study (NCT05121948) enrolled 34 US patients across five QD HC-7366 dose cohorts: 10 mg (3), 20 mg (3), 40 mg (12), 75 mg (10) and 125 mg (6). Most patients had advanced colorectal cancer, with other tumor types also included. Blood samples for PK/PD analyses were collected at baseline and throughout treatment, and paired tumor biopsies were collected at screening and after treatment. Results: In PK analyses, HC-7366 exposures were dose-proportional across 10–75 mg cohorts after single and multiple doses. Steady state was reached within 7–14 days, with a median T MAX of 2 hours, a geometric mean T 1/2 of ~13.5 hours, and an accumulation ratio at steady state of 1.68. HC-7366 showed no substantial time dependence, with observed linearity ratios near 1 across cohorts. Free-drug exposures at 40 and 75 mg were within effective ranges from preclinical models and were well below severely toxic levels in GLP toxicology studies. Biopsy pairs were collected from patients in 10 (n = 2), 20 (n = 1), 40 (n = 1), 75/40 (1 cycle/dose, n = 1), and 75 mg (n = 1) cohorts. Consistent with preclinical effects, IHC showed pathway engagement through asparagine synthetase induction in 6/6 biopsy pairs (avg positive: pre = 3.6%, post = 26.4%) and reduced nuclear HIF1α in both tumor cells (5/6, 85-99% reduction) and in myeloid cells (6/6, 53-99% reduction). Serum PD markers showed a transient reduction of EPO by day 8 in the 75 mg (21%) and 125 mg (53%) cohorts, which may be due to inhibition of HIF2α at higher doses as observed preclinically. HC-7366 also induced changes in metabolic proteins relevant to anti-tumor effects, including increased adiponectin after 3 weeks (40 mg: 25%, 75 mg: 43%, 125 mg: 70%), decreased chemerin after 1 week (40 mg: 28%, 75 mg: 26%, 125 mg: 46%), and decreased IGF-1 after 3 weeks (40 mg: 28%, 75 mg: 55%, 125 mg: 72%). Conclusions: PK/PD evaluation of HC-7366 showed a favorable PK profile, with dose-proportional exposure and a half-life suitable for QD dosing. Drug exposure, pathway engagement, and mechanistic biomarkers suggested 40-75 mg as the optimal dose range. Importantly, the metabolic effects suggest suitability for development in metabolically rewired cancers, including renal cell carcinoma (NCT06234605). Clinical trial information: NCT05121948 .

COPERNICUS, a pragmatic phase 2b study of first-line (1L) subcutaneous (SC) amivantamab (ami) + lazertinib (laz) with supportive care in <i>EGFR</i> -mutated advanced NSCLC: Early safety results.

8613 Background: In MARIPOSA, intravenous (IV) ami + laz significantly prolonged overall survival vs osimertinib (HR, 0.75; P =0.005) in 1L EGFR -mutated (exon 19 deletion [Ex19del]/L858R) advanced NSCLC. However, extended infusion times, infusion-related reactions (63%), venous thromboembolism (VTE; 36%), and dermatologic adverse events (AEs; paronychia [68%], rash [62%]) were observed, potentially leading to discontinuations of ami due to AEs (34%). Several studies have since identified ways to optimize ami + laz administration. In PALOMA-3/-2, SC ami coformulated with hyaluronidase (rHuPH20) enhanced patient (pt) experience by reducing administration-related reactions (ARRs) and time, as well as VTE with prophylactic anticoagulation, leading to FDA/EMA approval. In COCOON, an enhanced dermatologic regimen reduced grade ≥2 dermatologic AEs vs standard of care. Methods: COPERNICUS (NCT06667076) is the first study to combine SC ami, optimized supportive care and a pragmatic design to broaden the pt population and better resemble real-world usage. This is an early report from Cohort 1 on pt demographics and safety of SC ami every 4 weeks (Q4W) + laz daily in pts with 1L EGFR Ex19del/L858R advanced NSCLC receiving VTE/dermatologic AE prophylaxis. Pragmatic design included partnering with academic/community sites to enhance pt diversity, allowing 1 cycle of 1L chemotherapy while awaiting biomarker results and using SC ami Q4W to reduce visit frequency. Pts received prophylactic anticoagulation for the first 4 months of treatment. Dermatologic prophylaxis aligned with the regimen described in COCOON. Here we report safety (key secondary endpoint), including incidence/severity of VTE, ARRs and dermatologic AEs. All comparisons to MARIPOSA are descriptive. Results: As of data cutoff (02 Jan 2026), Cohort 1 had enrolled 190 pts in the US (target enrollment, 300; median [range] follow-up, 3.9 [0.1–11.7] mo); 92% were still ongoing in the study. Median age was 66 y, with 55% of pts ≥65 y and 21% ≥75 y; 28% were Asian and 9% African American, reflecting broad enrollment. 6 pts had received 1 chemotherapy cycle. AEs were mostly grade 1–2, with no new safety signals; 5% discontinued ami due to AEs. With dermatologic prophylaxis, paronychia and rash were 26% and 22%, respectively, showing numerical reductions vs MARIPOSA. ARRs and VTE (both grouped terms) were also numerically lower at 9% and 7%, respectively. Conclusions: Compared with MARIPOSA, SC ami and dermatologic/VTE prophylaxis in COPERNICUS substantially reduced ARRs, dermatologic AEs, VTE, and ami discontinuations, highlighting the impact of early supportive care interventions. Using a pragmatic design, these early safety data support wide use of SC ami Q4W + laz in a diverse population. Given limited follow up, pts will continue to be evaluated for safety and efficacy. Clinical trial information: NCT06667076 .

Evaluation of genomic DNA stability and white blood cell integrity in TFLB and Streck blood collection tubes for oncology diagnostics application.

e22563 Background: Reliable stabilization of genomic DNA (gDNA) and cellular integrity is critical for oncology clinical trials and molecular diagnostics, specifically for germline mutations. The stability of gDNA is required to extend transport times. TFLB is a single-tube solution designed to stabilize intact cells, nucleic acids, and proteins for five days or more. This study aimed to characterize the long-term stability and extractability of gDNA and the white blood cell (WBC) integrity characterization via flow cytometry from buffy coat samples stored in TFLB compared to Streck Nucleic Acid (NA) and Streck cfDNA tubes at ambient temperature. Methods: Whole blood was collected into TFLB, Streck Nucleic Acid (NA), and Streck cfDNA tubes and stored at 22°C. gDNA was extracted from buffy coat fractions at Days 1, 3, 5, and 7. In a parallel evaluation, WBC stability is being assessed via flow cytometry for 24-hour intervals up to 96 hours (T0–T4). Flow parameters include Forward/Side Scatter for cell morphology and CD45+ gating to monitor apoptosis and membrane integrity. Results: TFLB demonstrated high and consistent gDNA recovery through the first 72 hours (~60 ng/µl), followed by a significant increase in extractable concentration to ~110 ng/µl by Day 7. In contrast, both Streck NA and Streck cfDNA tubes exhibited a catastrophic decline in extractable gDNA after Day 3. Yields for Streck NA dropped from ~54 ng/µl at Day 3 to ~4 ng/µl at Day 5, while Streck cfDNA yields fell from ~36 ng/µl to ~4 ng/µl in the same period. Streck BCT variants maintained near-zero extractable gDNA levels through Day 7. Flow cytometry analysis determined that the gDNA correlates with the stabilization of WBCs in TFLB for up to 96 hours. Streck’s loss of gDNA concentration correlates with aggressive cell fixation or early-stage apoptosis in Streck tubes. Conclusions: TFLB effectively maintains gDNA extractability from buffy coat for up to 7 days at ambient temperature, whereas Streck tubes showed a loss of extractable DNA after 72 hours. The sharp decline in Streck tubes likely reflects aggressive cell fixation or DNA cross-linking rather than DNA degradation, making them unsuitable for delayed gDNA extraction from cellular fractions using standard protocols. The integration of flow cytometry further defines how these tubes preserve cellular surface markers and viability. These findings suggest that TFLB provides a more robust platform for clinical trials requiring integrated cellular and molecular analysis from a single blood draw.

Structural mimicry of the TOP1 cleavage complex by intratumoural bacterial proteins as a possible mechanism for geographical variation in TOP1 inhibitor efficacy in non-small cell lung cancer.

e15039 Background: We previously demonstrated that microbiome-derived proteins with sequence homology to human topoisomerase I (TOP1) are differentially enriched between US and Japanese populations and associated with discordant clinical trial outcomes (Lam et al., 2024). While these findings suggested a possible role for bacterial modulation of TOP1 inhibitor (TOP1i) efficacy, the underlying mechanism remained unclear. Here, we present the results of a deep structural analysis to determine whether TOP1-like bacterial proteins mimic the drug-interacting region of the TOP1-DNA cleavage complex, potentially contributing to differences in the efficacy of TOPi-based interventions, such as antibody-drug conjugates (ADCs). Methods: We extended our previous analysis by integrating intratumoural microbiome datasets from geographically diverse adult populations. Candidate bacterial proteins were prioritised based on sample prevalence and sequence identity with human TOP1. Crystallographic TOP1-DNA-drug ternary complexes of TOP1i analogues, including camptothecin and topotecan, were used to define the binding region at the protein-DNA interface surrounding the cleavage site. Structural alignment and protein folding predictions to evaluate geometric compatibility at the TOP1-DNA interface were performed using BioCorteX’s proprietary engines version: 20260114_172801. Results: Following sequence-based prioritisation and structural filtering, 13 bacterial proteins predicted to preserve key features of the human TOP1 catalytic core and drug-proximal protein-DNA interface were identified. Across &gt; 2,000 tumour whole-genome sequenced samples, non-small cell lung cancer (NSCLC) demonstrated the highest prevalence of these proteins (24%), significantly exceeding that observed in other epithelial solid tumours, which generally showed low-to-intermediate prevalence (median: 8.3%), and in non-epithelial malignancies (median: 3.4%), where detection was infrequent. This NSCLC-enrichment was not proportional to overall microbial burden. Within NSCLC, substantial geographic heterogeneity was observed across international cohorts, including samples from the United States, United Kingdom, and China, indicating population-level variation in exposure to these structurally relevant bacterial proteins. Conclusions: Building on our prior population-level analyses, these findings support a lung-predominant, tumour microbiome-linked mechanism by which intratumoural bacterial proteins structurally homologous to TOP1 may influence local interactions with TOP1 inhibitors at the tumour site. This framework may contribute to previously observed geographical differences in effective exposure to TOP1-targeted therapies, including TOP1 inhibitor-based chemotherapy and ADCs.

Discovery of gene set–based biomarkers and therapeutic targets for papillary thyroid cancer based on circulating DNA and tissue RNA.

6074 Background: Mutations in pathways known to promote oncogenesis in Papillary Thyroid Carcinoma (PTC) have been demonstrated in cell-free DNA (cfDNA) samples from patients with PTC, and at higher levels in patients with advanced and de-differentiated disease such as anaplastic thyroid cancer (ATC). In our study, the most frequent cfDNA mutations were associated with activation of BRAF -independent pathways and epigenetic regulation, suggesting a prominent role as drivers of disease in some patients. Also observed were mutations more frequently associated with dedifferentiated thyroid cancer, such as TP53 at 10% (5/49) in cfDNA vs 2% (1/47) in tissue samples. We hypothesized that nucleosome pattern analysis on cfDNA samples from a BRAF -independent subgroup would reveal evidence of heightened epigenetic signaling. Methods: We classified PTC into two subtypes based on cfDNA mutation profiles: BRAF -independent pathway (BI, 35% of patients with ≥1 mutation in KMT2A , ATM , or TP53 ) and BRAF /alternate-mediated pathway (B/a, 65% with ARID1A , BRAF , or others). These two gene groups were mutually exclusive by co-mutation and protein interaction networks. Chromatin structure was inferred from cfDNA fragment patterns using LIQUORICE. Gene signatures and pathway scores were then analyzed via GSEA and GSVA on the tissue RNA-seq data. Results: Significantly higher H3K4me1 signal (p = 0.047) was detected in cfDNA from the BI subgroup, indicating active epigenetic enhancer states. Additional gene expression analysis on tumor tissue revealed a unique six-gene signature (FDR = 0.089) enriched in the maternal-to-zygotic transition (MZT) pathway, with GSVA scores differing markedly between BI and B/a groups (p = 0.004). Cross-referencing these patients with the PTC cohort data from The Cancer Genome Atlas (TCGA) showed worse disease-free survival in the BI group (p = 0.008) but no difference in overall survival (p = 0.580) between the two groups, possibly due to early-stage diagnosis. Conclusions: Mutations in TP53 / ATM / KMT2A , associated with BRAF -independent pathways, correlate with active epigenetic states in PTC. This supports the hypothesis that epigenetic mechanisms are important drivers of progression in PTC. This study also supports a therapeutic strategy which targets these pathways, such as HDAC inhibitors previously validated in ATC cell lines and PDX models. Finally, six MZT-related genes may serve as biomarkers and regulators of epigenetic and transcriptional reprogramming important to the evolution of cancer in PTC patients.

Anti–PD-(L)1 plus chemotherapy in advanced PD-L1–negative nonsquamous non–small cell lung cancer: A systematic review and meta-analysis.

e20580 Background: Nonsquamous non-small cell lung cancer (nsqNSCLC) accounts for approximately 70–80% of all non-small cell lung cancer cases. Although immune checkpoint inhibitors improve outcomes in advanced nsqNSCLC, the true benefit of adding anti–PD-(L)1 to chemotherapy in patients with advanced PD-L1–negative nsqNSCLC remains unclear. Methods: We conducted a systematic review and meta-analysis of phase 3 randomized clinical trials comparing chemotherapy plus anti–PD-(L)1 versus chemotherapy with or without placebo in untreated, locally advanced (ineligible for concurrent chemoradiation or surgery) or metastatic PD-L1–negative nsqNSCLC. Overall response rate (ORR), progression-free survival (PFS), and overall survival (OS) were used as endpoints. Heterogeneity was assessed using the Cochran Q test (significant vs. nonsignificant) and the I² statistic (low, moderate, substantial, or considerable). Risk of bias was evaluated using RoB 2 and categorized as low, moderate, or high. Results: A total of 2,084 patients with advanced PD-L1–negative nsqNSCLC from 12 phase 3 randomized clinical trials were included: CameL, CheckMate 227 part 1, CheckMate 227 part 2, DOMAJOR, EMPOWER-Lung 3, GEMSTONE-302, IMpower130, IMpower132, KEYNOTE-189, ORIENT-11, POSEIDON, and RATIONALE-304. Of these, 1,196 (57%) received chemotherapy plus anti–PD-(L)1 (intervention arm), while 888 (43%) received chemotherapy with or without placebo (control arm). All included studies used platinum-doublet chemotherapy regimens. Chemotherapy plus anti–PD-(L)1 was associated with a higher probability of ORR (risk ratio, 1.61; 95% CI 1.26–2.05; P = 0.0001), improved PFS (HR 0.68, 95% CI 0.60–0.77; P &lt; 0.00001), and improved OS (HR 0.81, 95% CI 0.70–0.94; P = 0.006) compared with chemotherapy with or without placebo. Moderate inter-study heterogeneity was observed in the OS meta-analysis, with a significant Cochran Q test. A post hoc sensitivity analysis of OS excluding RATIONALE-304, which included a disproportionately high number of never-smokers, demonstrated low and nonsignificant heterogeneity. Another post hoc sensitivity analysis excluding the two studies with a high overall risk of bias (IMpower130 and IMpower132) showed a statistically significant improvement in OS with chemotherapy plus anti–PD-(L)1 (HR 0.81, 95% CI 0.72–0.92; P = 0.001), further supporting the robustness of the survival benefit compared with chemotherapy with or without placebo. Conclusions: These findings support the use of anti–PD-(L)1 plus chemotherapy as first-line therapy for patients with advanced PD-L1–negative nsqNSCLC. Prospective investigation is warranted to determine whether dual checkpoint inhibition, with or without chemotherapy, is superior to anti–PD-(L)1 plus chemotherapy in this patient population.