Browse Articles

Landscape of therapeutically actionable HRR gene variants in Moroccan breast and ovarian cancer: A systematic review and meta-analysis.

e15075 Background: Hereditary breast and ovarian cancers (HBOC) in Morocco remain underexplored, despite a steady increase in access to germline genetic testing over the past decade. While BRCA1 and BRCA2 are well-established drivers of hereditary risk, the contribution of other homologous recombination repair (HRR) genes is increasingly recognized, particularly in the era of targeted therapies such as PARP inhibitors. A comprehensive synthesis of available data is therefore needed to clarify the prevalence, distribution, and clinical relevance of HRR gene variants in the Moroccan population. Methods: We performed a systematic review and meta-analysis in accordance with PRISMA guidelines. PubMed, Cochrane, and Scopus databases were searched for studies reporting germline HRR gene variants in Moroccan patients with breast and/or ovarian cancer. Eligible studies were assessed for methodological quality using the Newcastle–Ottawa Scale. Variant prevalence estimates were pooled across cohorts using meta-analytic models, with heterogeneity evaluated to ensure robustness of the findings. Results: Thirteen studies met the inclusion criteria, encompassing 762 patients screened for BRCA1, 641 for BRCA2, and 110 for additional HRR genes. The pooled prevalence of pathogenic or likely pathogenic variants was 9.3% (95% CI: 5.8–12.8%) for BRCA1 and 10.0% (95% CI: 8.3–11.6%) for BRCA2, yielding a combined BRCA1/2 prevalence of 14.7% (95% CI: 12.0–17.4%). Inclusion of the broader HRR gene spectrum increased the overall prevalence to 19.7%, highlighting a substantial proportion of patients harboring non-BRCA actionable variants. Recurrent alterations, including BRCA1:c.5309G > T, BRCA1:c.3279delC, and BRCA2:c.1310_1313delAAGA, were repeatedly observed across independent cohorts, supporting the hypothesis of founder or population-enriched mutations. Conclusions: Approximately one in five Moroccan patients fulfilling HBOC testing criteria carries a pathogenic HRR variant with potential diagnostic and therapeutic relevance. The identification of recurrent and population-enriched mutations provides a strong evidence base for optimizing genetic testing strategies in the Moroccan population. These findings support the design of population-adapted HRR gene panels, informed by local mutational spectra, with the aim of improving risk assessment, genetic counseling, and the implementation of precision oncology in resource-constrained settings.

Cadonilimab or ivonescimab plus axitinib in metastatic mucosal melanoma: Results from a phase Ib trial.

9516 Background: Mucosal melanoma (MM) is a rare and aggressive cancer with a poor prognosis. Unlike its more common counterpart, cutaneous melanoma, MM responds poorly to standard single-agent immunotherapy, such as anti-PD-1 monoclonal antibodies (mAbs), limiting treatment options. This clinical challenge underscores the need for more effective strategies, making combination therapies a clear future direction. The aim is to overcome resistance by targeting multiple pathways simultaneously. The main objective of this study is to explore and compare two distinct immunotherapy backbones—targeting the PD-1/CTLA-4 and VEGF pathways—for their potential in future rational combinations against MM. Methods: This is a phase Ib, dose de-escalation and cohort-expansion, open-label trial. The combination dose of axitinib will utilize a "3+3" dose de-escalation design. The starting dose is 5 mg. If dose-limiting toxicity (DLT) is observed at this dose level, the dose for subsequent cohorts will be reduced to 3 mg. Patients with metastatic melanoma receive 10 mg/kg cadonilimab Q3W (cohort 1) or 20 mg/kg ivonescimab Q3W (cohort 2) in combination with 5 mg axitinib BID. Primary endpoints are rate of dose-limiting toxicities (DLTs) and safety. Secondary endpoints are objective response rate (ORR) by investigator assessment, duration of response (DOR), progression-free survival (PFS), overall survival (OS) and overall safety. Results: As of December 25, 2025, a total of 28 patients have been enrolled in the study (14 cohort 1; 14 cohort 2). Median follow-up of 13 months. The distribution of primary tumor sites was as follows: oral cavity 14.3%, esophagus17.9%, rectum 32.1%, and cervix 35.7%. 21% had elevated LDH. BRAF status, known for 71% of patients, was positive in 25%. No DLT or treatment related death was observed，Axitinib was selected as 5 mg BID. 71.4% patients experienced treatment related AE (TRAE) and 17.9% patients experienced Grade 3-4 TRAEs. In cohort 1，the most common TRAEs include hypothyroidism, hypertension, diarrhea, anorexia and rash. In cohort 2, the most common TRAEs include proteinuria, myalgia, hyperthyroidism, hypertension, abnormal transaminase levels, anorexia and hyperlipidemia. No patient discontinued treatment due to TRAE. ORR, DOR, DCR, and median PFS for all cohorts are shown in Table. Conclusions: The combination therapy involving cadonilimab or ivonescimab plus axitinib demonstrates a promising safety and efficacy profile. These findings suggest that such dual-antibody-based regimens could provide synergistic anti-tumor activity with an improved therapeutic index, warranting further clinical investigation. Clinical trial information: NCT06424626 . Secondary Endpoint Cohort 1 （n=14） Cohort 2 （n=14） PR 5 5 SD 4 5 PD 5 4 ORR 35.7% 35.7% DCR 64.3% 71.4% DOR (mo) 6.7 6.4 mPFS (mo) 5.8 8.7 mOS (mo) NA NA

Development and implementation of a pilot prescreening tool for potential clinical trial participants for an immune checkpoint inhibitor trial.

e13655 Background: Clinical trial accrual can be challenging in a community-based cancer setting. During limited appointment time with patients, it is difficult for physicians to quickly access available clinical trials and determine eligibility. Additionally, time constraints also limit the ability for clinical research coordinators (CRCs) to conduct patient chart reviews via oncologists’ schedule to identify potential trial candidates. Under these circumstances, we developed a patient prescreening tool to help CRCs narrow down potential subjects in a community-based cancer institute. We implemented the pilot program to study SWOG S2013 (I-CHECKIT) at our sites. Methods: The prescreening tool was based on Epic Electronic Health Record (EHR) system data and Tableau dashboard. First, we worked with the research unit leader to transform eligibility criteria into extractable data elements. Second, we built Oracle Structured Query Language (SQL) queries to retrieve data from the Epic Clarity database. The query captured solid tumor patients who were scheduled for their first immune checkpoint inhibitor treatment, as well as any chemotherapy and upcoming appointments. Finally, we visualized the query result to Tableau dashboard. The dashboard was refreshed daily with an updated patient list and information. Only authorized study personnel had access to this tool. Results: I-CHECKIT cohort 1 was open to accrual at our sites on December 12, 2023. It took about two months to build SQL query and Tableau dashboard, and it was implemented on February 28, 2024. CRCs reported that the pilot tool saved a significant amount of time on prescreening patients, and eligible subjects were identified successfully. Conclusions: This solution is replicable for hospitals with Epic EHR system and Tableau platform. It helps reduce time and resources required to identify eligible patients for clinical trials. It expands the range of potential subjects beyond referrals from oncologists. We can implement this pilot tool with additional trials to further investigate its feasibility and usability.

Reducing manual chart review in oncology: Validation of an AI-based monitoring tool for lung cancer staging and biomarker tracking.

e13662 Background: Continuous monitoring of evolving staging and molecular data is essential for clinical decision support to capture shifts in an oncology patient's care journey. However, current systems struggle to synchronize fragmented electronic health records and extensive clinical documentation. This leaves clinicians with an incomplete picture, hindering their ability to follow complex and dynamic NCCN protocols without manual chart reviews. Methods: We implemented an AI-based continuous monitoring system (AI-CM) that uses embeddings from large language models, fine-tuned on clinical notes, to extract clinical events and flag patients whose care deviates from NCCN guidelines. Monitoring begins at the initial oncology ICD code, where the system tracks follow-ups that determine stage, histology, and site (diagnostic features). Subsequently, it continuously monitors documents for diagnosis-specific NCCN-recommended biomarker testing (ordered or performed). We validated AI-CM in a manually annotated cohort of 7,259 patients across eight institutions (2024–2025). The evaluation focused on biomarker testing for early-stage (n = 1,155) and advanced (n = 4,448) non-small cell lung cancer (eNSCLC, aNSCLC), specifically targeting EGFR, ALK, PD-L1, and NGS. Results: AI-CM identified eNSCLC and aNSCLC diagnostic features with precision of 0.77 and 0.85 and recall of 0.91 and 0.94, respectively, and biomarker testing with precision of 0.75 and recall of 0.81. Incorporating monitoring for diagnostic features reduced false positive diagnoses by 25% relative to relying solely on ICD codes. The median time to diagnosis was 41 days for eNSCLC and 16 days for aNSCLC, with a median time to test from diagnosis of 9 days. Within 6 weeks of the initial ICD code and diagnosis, AI-CM identified diagnosis and biomarker testing, on average, for 82% and 88% of patients, respectively (Table 1). This enables flagging for full review of a much smaller proportion of patients (18%) with a missed diagnosis, or who are non-adherent (12%) to NCCN guidelines for testing. By reviewing only the AI-recommended document to confirm diagnosis for the 89% of patients correctly identified as aNSCLC, and all documents for the remaining patients, the average number of documents requiring review within the first 6 weeks to determine these diagnoses was reduced from 5.5 to 1.3 per patient (a 76% reduction). Conclusions: AI-based continuous monitoring improves timely identification of patients deviating from guidelines and reduces manual review by prioritizing documents for human verification. Performances of AI-CM at 6 weeks from the previous clinical event. Target Patients Identified (%) Mean Documents to Review per Patient AI-CM Mean Documents to Review per Patient AI-CM Documents to Review Reduction (%) eNSCLC 76 7.3 2.4 67 aNSCLC 89 5.5 1.3 76 eNSCLC Testing 89 5.2 1.5 71 aNSCLC Testing 88 7.5 1.8 76

Neoadjuvant immunochemotherapy plus thymalfasin in locally advanced gastric cancer: A prospective clinical trial.

4103 Background: Neoadjuvant immunochemotherapy has emerged as a promising strategy for locally advanced gastric and gastroesophageal junction (G/EGJ) adenocarcinoma, but a substantial proportion of patients derive limited benefit. Thymalfasin is an immunomodulatory peptide that may amplify antitumor immunity while attenuating toxicity. We conducted a phase II trial to evaluate anti-PD-1 plus SOX immunochemotherapy combined with thymalfasin as neoadjuvant treatment for G/EGJ adenocarcinoma. Methods: The prospective trial enrolled patients aged 18–75 with cStage III G/EGJ adenocarcinoma, ECOG 0–1, and adequate organ function. Treatment included three 21-day cycles of serplulimab (anti-PD-1) plus SOX (S-1 and oxaliplatin) and nine weeks of thymalfasin, followed by curative gastrectomy. The primary endpoint was pathological complete response (pCR). Secondary endpoints included major pathological response (MPR), safety, survival, and other efficacy measures. Peripheral immune remodeling was assessed by flow cytometry, and bulk RNA sequencing of peripheral blood mononuclear cells (PBMCs) interrogated thymalfasin-associated transcriptional programs. Results: Thirty patients were enrolled and all underwent curative-intent minimally invasive gastrectomy. pCR was achieved in 30.0% (9/30), and MPR in 56.7% (17/30). ypN0 status was observed in 63.3% (19/30), with N-stage downstaging in 80.0% (24/30). The objective response rate was 73.3% and the disease control rate was 100.0%. At a median follow-up of 14.0 months (range 10.0-17.2), only one retroperitoneal nodal relapse had occurred at 14.4 months after diagnosis; no deaths were documented. Any-grade adverse events (AEs) occurred in 93.3% of patients, grade ≥3 AEs in 26.7%, and immune-related AEs in 23.3%, with no grade ≥3 irAEs. Flow cytometry showed expansion of CD8⁺ T cells with increased CD69 expression and a concurrent reduction in HLA-DR-positive T cells, suggesting dynamic remodeling from broad systemic activation toward a more focused effector or memory response. RNA-seq revealed thymalfasin-associated upregulation of genes involved in antigen processing and presentation, type I interferon signaling, innate immune sensing, and immune metabolic reprogramming, and identified immune co-expression modules linked to treatment and response. Conclusions: Neoadjuvant serplulimab, SOX, and thymalfasin produced encouraging pathological response, substantial nodal clearance, and an acceptable safety profile in stage III G/EGJ adenocarcinoma. Peripheral immune and transcriptomic profiling are consistent with a hypothesis in which thymalfasin may help preserve and coordinate systemic antitumor immunity without excessive toxicity. These findings warrant confirmation in larger randomized trials. Clinical trial information: NCT06461910 .

The key role of an intramolecular non-classical hydrogen bond of vinylboron monomer for stereoselective polymerization

Incidence of cardiac lymphoma, mesothelioma, and sarcoma: A population-based study of primary malignant cardiac tumors.

e24000 Background: Primary malignant cardiac tumors (PMCTs) are considered extremely rare neoplasms, with limited population-level data on incidence patterns. This study aims to describe the epidemiology of PMCTs in the United States, focusing on three main histologic subtypes: sarcomas, lymphomas, and mesotheliomas. Methods: We extracted cases of pathologically confirmed PMCTs diagnosed between 2000 and 2022 from the Surveillance, Epidemiology, and End Results (SEER) 17 Registries database. Incidence rates (IRs) were age-adjusted to the 2000 US Standard Population. Long-term trends over the years were identified by Joinpoint regression software (version 5.4) through determination of joinpoints by permutation test (4499 permutations) at an alpha of p < 0.05. The annual percentage change (APC) and average annual percentage change (AAPC) were computed. Results: From 2000 to 2022, the overall incidence of PMCTs significantly increased with an Average Annual Percent Change (AAPC) of 4.9%. This trend was characterized by two distinct periods of rapid growth (2000-2004 and 2012-2015) separated by intervals of stability. When analysed by subtype, both lymphoma and mesothelioma demonstrated consistent and significant increases over the entire period, with AAPCs of 5.2% (p < 0.001) and 3.1% (p = 0.019), respectively. In contrast, the incidence of sarcoma remained stable, showing no statistically significant change throughout the 22-year study period (AAPC = 0.2%, p = 0.792). Conclusions: The rising incidence of primary malignant cardiac tumors is not uniform across histological subtypes. The overall increase is primarily driven by significant and steady growth in cardiac lymphoma and mesothelioma, while the incidence of sarcoma has remained stable. These divergent trends highlight the need for subtype-specific research to understand the distinct etiologies and risk factors contributing to the changing landscape of cardiac malignancies.

Compassionate release for incarcerated patients with cancer: A nationwide policy review.

e24040 Background: Cancer is the leading cause of death among incarcerated individuals in the United States, who experience higher cancer incidence and mortality than the general population. Compassionate release allows incarcerated individuals with severe medical illness, including terminal cancer, to be released from prison; however, it remains underutilized. Understanding policy-level barriers to compassionate release is critical to improving end-of-life care for incarcerated patients with cancer. Methods: We reviewed compassionate release policies in all 50 U.S. states and the District of Columbia as they pertain to incarcerated individuals with terminal cancer. Laws were evaluated based on application initiation, eligibility criteria, prognosis requirements, treatment considerations, release planning requirements, and the availability of appeals. Results: One state, Iowa, did not have a compassionate release law. Among the remaining 49 states and Washington, DC, 15 required compassionate release applications to be initiated by the department of corrections, while 32 permitted initiation by attorneys, family members, or incarcerated individuals; three states did not specify an initiating party. Prognostic eligibility varied widely: 12 states required a prognosis of less than six months, 13 required a prognosis less than one year, and seven states required a prognosis less than two years, while 18 states did not specify an explicit prognosis timeline. Limited functional status was a common eligibility criterion, and the ability to receive improved medical care outside the correctional system was considered in 20 states. Cancer and hospice care were infrequently referenced, with cancer explicitly mentioned in six states and hospice in 20. Eight states allowed appeals of denied applications; 13 did not allow appeals, 16 did not specify an appeals process, and 13 allowed reapplication if clinical circumstances changed. Conclusions: Compassionate release policies for incarcerated individuals with terminal cancer vary substantially across states and frequently include unclear or restrictive eligibility criteria. Future policy reform should focus on expanding who may initiate applications, clarifying and broadening eligibility criteria, and increasing awareness among clinicians and incarcerated individuals. Such efforts may facilitate timely, medically appropriate end-of-life care for incarcerated patients with cancer.

Role of oxidative stress and neutrophil reprogramming in mediating venous thrombosis in colorectal cancer.

e15733 Background: Colorectal cancer associated venous thrombosis (CRC-VT) is a major cause of mortality in CRC patients, yet the mechanisms linking CRC to thrombosis remain unclear. NETosis, a process in which neutrophils release prothrombotic extracellular traps (NETs), is elevated in various cancers. However, the mechanisms of cancer-associated NETosis and its contribution to CRC-VT remain unknown. We hypothesized that CRC associated reactive oxygen species (ROS) reprogram neutrophil toward a pro-NETotic phenotype promoting a prothrombotic state. Methods: C57BL/6J wild-type (WT) mice were subcutaneously injected with murine CRC cells (MC38). Neutrophil subpopulations were characterized by spectral flow cytometry. Spontaneous and agonist induced NETosis in bone marrow-derived neutrophils was quantified by immunostaining. Pharmacological and genetic approaches were used to assess contribution of ROS to NETosis. Venous thrombosis susceptibility was evaluated using inferior vena cava (IVC) stenosis model. DNase I treatment for NET degradation, Ly6G-mediated neutrophil depletion, and Nox2-KO mice were used to assess causality. Nox2-NADPH oxidase is the key superoxide generating enzyme in neutrophils and drives NETosis. Results: C57BL/6J mice bearing CRC-tumors exhibited elevated circulating neutrophils and H3Cit (a marker of NETosis). Spectral flow cytometry showed reduced maturation marker CD16 but elevated migration/activation markers (CD11b, CD62L), co-localized with H3Cit, indicating a CRC-driven-pro-NETotic reprogramming. Neutrophils displayed altered nuclear morphology with loss of lobulation, suggesting expansion of immature subsets. Bone marrow-derived neutrophils showed increased spontaneous and agonist-induced NETosis, suppressed by Nox2 inhibition or antioxidants. Finally, in IVC stenosis model, tumor-bearing mice developed larger thrombi, which were significantly reduced by DNase I treatment or Ly6G-mediated neutrophil depletion. Furthermore, tumor-bearing Nox2-KO mice showed reduced NET formation and venous thrombosis, suggesting that Nox2 mediated NETosis contributes to thrombosis in CRC model. Conclusions: ROS-driven neutrophil reprogramming links CRC to venous thrombosis. Our data may support ROS/NETs as potential targets to reduce CRC-VT burden.

Circadian immune coherence and readiness as time-aware early endpoints in a real-world NSCLC cohort treated with immune checkpoint inhibitors.

e23114 Background: Immune biomarkers guiding immune checkpoint inhibitor (ICI) therapy are commonly assessed as static baseline measurements, despite strong biological evidence that immune function is dynamically regulated over time. Circadian organization influences immune cell trafficking, checkpoint molecule expression, and T-cell effector function, potentially affecting ICI efficacy and toxicity. We investigated whether circadian immune coherence (CIC) and circadian immune readiness (CIR), derived from longitudinal routine hematology, could serve as time-aware early endpoints in real-world patients with non–small cell lung cancer (NSCLC) receiving ICIs. Methods: We performed a retrospective real-world analysis of 90 patients with advanced NSCLC treated with PD-1/PD-L1 inhibitors. Longitudinal complete blood count–derived immune indices were analyzed using AI-enhanced Fourier decomposition to extract oscillatory features, including phase stability across treatment cycles. CIC quantified inter-cycle phase coherence of immune rhythms, while CIR integrated coherence with oscillatory amplitude. Unsupervised clustering identified immune coherence phenotypes. Associations with objective response rate (ORR), progression-free survival (PFS), infusion timing, and immune-related adverse events (irAEs) were evaluated descriptively. Results: Detectable circadian immune oscillations were identified in 76% of patients. Two dominant phenotypes emerged: high circadian immune coherence (n=54) and low coherence (n=36). Patients with high CIC demonstrated longer median PFS (11.8 vs 6.4 months) and higher ORR (48% vs 22%) compared with low-coherence patients. Alignment of immunotherapy infusion with individual coherence peaks was associated with more favorable outcomes, whereas low-coherence patients showed limited sensitivity to timing effects. High CIC was also associated with a lower incidence of grade ≥2 irAEs (19% vs 33%). CIC and CIR provided clinically informative stratification beyond baseline inflammatory indices and static biomarkers. Conclusions: Circadian immune coherence and readiness capture biologically meaningful temporal organization of immune responses during ICI therapy. Early-emerging CIC/CIR phenotypes represent novel, time-aware biomarkers with potential utility for early dynamic risk stratification and optimization of real-world immunotherapy strategies in NSCLC.

Characterizing the genomic landscape of Langerhans cell histiocytosis using the AACR Project GENIE database.

6572 Background: Langerhans cell histiocytosis (LCH) is a rare neoplasm characterized by the abnormal clonal proliferation of bone marrow-derived Langerhans cells and occurs predominantly in pediatric patients. Clinical manifestations range from localized bone lesions to multisystem involvement. Given the high frequency of MAPK pathway mutations, particularly BRAF V600E, next-generation sequencing is recommended for LCH patients to identify actionable targets. This study utilizes the American Association for Cancer Research (AACR) Project Genomic Evidence Neoplasia Information Exchange (GENIE) database to address existing gaps in the genomic profiling of LCH, aiming to inform long-term risk stratification, uncover prognostic markers, and guide avenues for targeted therapy. Methods: AACR Project GENIE was accessed via cBioPortal (v18.0-public) on November 18, 2025 to identify all LCH patients. Gene mutations and demographic variables were tabulated, and statistical correlations were assessed using two-sided t-tests and non-parametric analyses with Benjamini–Hochberg false discovery rate correction. Entries with unknown values were excluded from analysis; discrepancies in percentage reflect unreported data. Results: This study identified 326 LCH samples from 290 patients, of whom 140 (48.3%) were female and 139 (47.9%) were male. By race, 140 (48.3%) identified as White, 9 (3.1%) as Black, and 9 (3.1%) as Asian. Of the samples, 152 (46.6%) originated from primary tumors and 20 (6.1%) were from metastatic tumors. The cohort included samples from 113 pediatric patients (34.7%) and 212 (65.0%) adults. Frequent mutations included BRAF (n=126; 38.7%), MAP2K1 (n=57; 17.5%), TET2 (n=17; 5.2%), and FAT1 (n=8; 2.5%). No significant difference was identified between the genomic profiles of males and females (p>0.05), although comparisons by race revealed several unique somatic point mutations in Asian and Black patients. Among Asian patients, mutations in MUTYH , PDGFRB , and TERT were significantly enriched, while BORCS8-MEF2B , BRCA1 , and DDR2 mutations were enriched in Black patients. Within this cohort, pediatric patients were more likely to present with PTEN mutations, while adults ≥34 years old were more likely to have VTI1A or ATM mutations. BRAF mutations were mutually exclusive with MAP2K1 mutations (n=124/124; p<0.001), while TET2 mutations frequently co-occurred with ASXL1 mutations (n=2/18, p<0.05). There were no significant differences in mutations between primary and metastatic samples (p>0.05). Conclusions: Our findings support the classification of BRAF and MAP2K1 mutations as key driver mutations and identify several additional mutations that are significantly enriched across demographic groups. Further research into these genomic associations is crucial to address existing knowledge gaps and translate into novel treatment modalities for patients with LCH.

A context-constrained oncology agent for automated, guideline-concordant treatment decision support in precision oncology.

1620 Background: Precision oncology requires biologically plausible, clinically executable decisions aligned with regulatory approvals. General-purpose Large Language Models (LLMs) often produce recommendations rejected by molecular tumor boards due to semantic drift, inappropriate agent extrapolation, and ignored clinical constraints, raising significant safety concerns for point-of-care adoption. Methods: GenOnco is a constrained oncology agent integrating NCCN, ASCO, and ESMO clinical practice guidelines; FDA/EMA regulatory approvals; NEJM seminal articles and high-impact publications; curated clinicogenomic resources (e.g., Molecular Oncology Almanac); comprehensive clinical trials knowledge (ClinicalTrials.gov, landmark phase III results); and practice-changing trial evidence. The system utilizes a dual-layer architecture: a reasoning LLM paired with a deterministic, non-LLM clinical decision layer that enforces line-of-therapy requirements and actionability filters independent of generative output. We evaluated GenOnco on 102 multi-institutional oncologist-derived real-world solid tumor queries, independently scored by two blinded oncologists (κ = 0.87). Performance was benchmarked against published datasets: MEREDITH (200 cases) and Google DeepMind AMIE (100 cases). Primary endpoints were guideline concordance and clinical executability, analyzed via McNemar's test (N = 162). Results: GenOnco achieved 95.1% accuracy (95% CI: 91.2%–98.4%) on synthetic benchmarks and 93.1% (95% CI: 89.1%–96.5%) on real-world queries, eliminating 100% of expert-rejected outputs from comparators. GenOnco significantly outperformed GPT-4o (48.1%), Claude 3.7 (49.4%), Gemini 2.5 (50%), ASCO Guidelines Assistant (67.3%), OpenEvidence, MEREDITH, and Google DeepMind AMIE (all P < .001; N = 162), with gains of 15-47 percentage points. Conclusions: Constrained reasoning prioritizing executable pathways outperforms unconstrained LLMs for precision oncology support. Limitations include solid tumor focus and query diversity; diverse population validation planned. GenOnco demonstrates potential for point-of-care integration.

Safety and pharmacokinetics (PK) of lurbinectedin (lurbi) in pediatric patients (pts) with relapsed/refractory (R/R) solid tumors and preliminary antitumor activity in pediatric and young adult pts with R/R Ewing sarcoma (EwS): Results from a phase 1 study.

11518 Background: About 30% of pediatric cancers are solid tumors. EwS is the second most common bone tumor among children and young adults; 5-year survival rates for R/R EwS are <15%. Lurbi is a trabectedin analogue that alkylates DNA and blocks transcription of EWS-FLI1, a primary driver of EwS. In a phase 2 basket trial, lurbi demonstrated an objective response rate (ORR; complete response [CR] or partial response [PR]) of 14% and disease control rate (DCR; CR, PR, or stable disease) of 57% in 28 adults (median age, 33 years [y]) with R/R EwS. Here, we report phase 1 results from a multicenter, open-label study (NCT05734066) of lurbi in pediatric and young adult pts with R/R solid tumors including EwS. Methods: Part 1 evaluated the safety, tolerability, PK, and recommended phase 2 dose (RP2D) of lurbi monotherapy in pediatric pts (aged 2–18 y) with R/R solid tumors. RP2D selection used a Bayesian optimal interval design starting with the approved adult dosage, 3.2 mg/m 2 lurbi IV over 1 hour every 3 weeks. Two additional cohorts were treated at the RP2D: a safety cohort for pts aged <18 y and a histology-specific safety and efficacy cohort for pts aged 2–30 y with R/R EwS. Tumor responses were assessed every 6 weeks per RECIST v1.1. Results: As of Oct 22, 2025, 21 pts received ≥1 dose of lurbi (3.2 mg/m 2 , n = 17; 4.0 mg/m 2 , n = 4). Median (range) age was 16 (10–25) y; 13 (62%) pts had received ≥3 prior lines of treatment (Tx). Median (range) lurbi cycles administered was 3 (1–18); 3 (14%) pts were still receiving lurbi; 18 (86%) discontinued Tx. At 4 mg/m 2 , dose-limiting pulmonary edema and rhabdomyolysis/acute kidney injury were observed. Among pts receiving lurbi 3.2 mg/m², 8 (47%) had grade ≥3 Tx-related adverse events (TRAEs); anemia (5 [29%]) and platelet count decrease (5 [29%]) were most common (Table). For pts aged ≥6 y, the RP2D was 3.2 mg/m 2 ; accrual of pts aged <6 y is ongoing to determine RP2D. Although geometric mean (Geo CV%) C max (163 µg/L [105%]) and AUC inf (1893 µg∙h/L [126%]) were higher in children/young adults vs adults in the phase 2 basket trial (increased 53% and 244%, respectively), there was substantial overlap in drug exposures. Among 11 evaluable pts with R/R EwS receiving lurbi 3.2 mg/m², the confirmed ORR was 18%, and the DCR was 55%; 2 responders were progression free at their last assessments (5.7 and 6.9 months). Conclusions: Based on these data, lurbi had a manageable safety profile, predictable PK properties, and encouraging preliminary clinical activity in pts with R/R EwS. Phase 2 expansion in pts ≤30 y with R/R EwS is ongoing. Clinical trial information: NCT05734066 . TRAEs in all pts and by dose. n (%) All N = 21 3.2 mg/m 2 n = 17 4.0 mg/m 2 n = 4 Any TRAE 21 (100) 17 (100) 4 (100) Grade ≥3 12 (57) 8 (47) 4 (100) Serious 4 (19) 2 (12) 2 (50) Led to Tx discontinuation 1 (5) 0 1 (25) Led to death 1 (5) 0 1 (25) a a Due to acute kidney injury.

Prognostic value of early NLR normalization in MSI-H mCRC: A real-world analysis.

e15625 Background: Immune Checkpoint Inhibitors (ICIs) are the standard of care first line for Microsatellite Instability-High (MSI-H) metastatic colorectal cancer (mCRC). Baseline Neutrophil Lymphocyte Ratio (NLR) and Absolute Lymphocyte Count (ALC) have been proposed as prognostic markers for ICI response across solid tumors. However, it remains unclear if early kinetic/dynamic changes in the NLR affect the long term survival outcomes. We evaluated whether early NLR kinetics predict Overall Survival (OS) independent of baseline nutritional status. Methods: We used the TriNetX global network database to identify patients with metastatic colorectal cancer treated with first line ICI therapy. To strictly isolate the MSI-High population, we excluded any patient who received cytotoxic chemotherapy within one year prior to or one month after the index date. We stratified patients into two cohorts based on inflammatory kinetics during the first 3–10 weeks of treatment: Responders (High Baseline NLR normalized to < 5) and Non-Responders (High Baseline NLR remained > 5). We used 1:1 Propensity Score Matching to balance key characteristics, including age, sex, steroid use, and baseline nutritional status (albumin). Results: A total of 768 propensity matched patients were analyzed (384 per cohort). Baseline characteristics were well balanced, including serum albumin (P = 0.07), which suggests comparable nutritional status. Early NLR normalization was associated with a significantly superior Median Overall Survival of 43 months compared to 21 months for Non-Responders. This survival benefit persisted over a decade, with 5-year and 10-year survival rates of 46.7% and 34.8% for Responders, compared to 30.1% and 18.4% for Non-Responders (HR 0.59; P < 0.001). Time to Next Treatment did not differ significantly between groups. Conclusions: Our study provides compelling real world evidence that early improvement in systemic inflammation as reflected by NLR normalization, is a critical determinant of long-term survival in patients with MSI-H mCRC receiving first-line ICI. Importantly, this survival benefit is durable and independent of baseline nutritional status. These findings validate NLR kinetics as a robust, cost effective biomarker for early evidence of therapeutic efficacy. Overall survival and hazard ratio (responders vs non-responders). Outcome Measure Responders (NLR > 5 → ≤ 5) Non-Responders (NLR > 5 → > 5) HR (95% CI) P-Value N (Propensity Matched) 384 384 Median Overall Survival 43 Months 21 Months 0.59 (0.47–0.74) <0.001 5-Year Survival Rate 46.70% 30.10% 0.584 (0.467–0.731) <0.001 10-Year Survival Rate 34.80% 18.40% 0.592 (0.474–0.738) <0.001 Abbreviations: HR, Hazard Ratio; NLR, Neutrophil-to-Lymphocyte Ratio; CI, Confidence Interval.

Decision-making and regret after radical cystectomy in patients with bladder cancer: An international cross-sectional survey.

e16605 Background: Radical cystectomy (RC) is the recommended treatment for Bacillus Calmette-Guérin-unresponsive high-risk non-muscle invasive bladder cancer (HR-NMIBC) and muscle invasive bladder cancer (MIBC). This survey explored the patient RC decision-making process, level of regret, and current impact of RC on their life. Methods: Adults with HR-NMIBC/MIBC who underwent RC (≥6 months ago) from France, Germany, Italy, Japan, the United States, Brazil, and Mexico participated in an online survey. The survey included the Decision Regret Scale (DRS) and questions on patient decision-making before RC, regret after RC, and the current impact of RC on their daily life. Quantitative data were analyzed descriptively; qualitative data were analyzed thematically. Results: Patients (n = 222) were a mean age of 50 years, 70% male, a mean of 2.3 years after RC, and 23% had metastatic disease at time of data collection. Patients reported collaborative and supportive relationships with healthcare providers (HCPs) and 84% felt “very” or “somewhat” satisfied with information provided for decision-making. Before RC, 93% of patients reported being informed of alternative treatments, though 35% and 55% reported not receiving information on systemic chemotherapy or radiation, respectively. Following RC, 90% of patients experienced some level of regret; with 36% reporting moderate/severe regret. The most common reasons for regret were impact on physical/mental health, quality of life, and issues with urinary diversion. Most patients (93%) reported bothersome current symptoms or functional impacts; > 50% patients reported impact on their sex drive and pleasure, energy level, body image, sleep quality, mental health, ability to travel and do day-to-day activities, and their financial situation. Patients with moderate/severe regret reported significantly more impact to their daily life than patients with mild regret. After a mean of 2.3 years post RC, 41% of patients have not returned to their usual activities and 48% have not returned to their usual health before diagnosis. Most patients (95%) relied on caregivers for support. Conclusions: Patients undergoing RC felt well-informed and made shared decisions with their HCP. However, most (90%) patients experienced some level of decision regret, mainly due to the physical or psychological impacts of RC. Patients still experience substantial impact to their daily life and rely on caregivers for support, at a mean of 2.3 years after RC. These results highlight the need to better understand the patient experience of RC beyond typical assessment methods and to balance decision-making with newer bladder-preserving treatment options.

<i>UGT1A1</i> gene variants associated with irinotecan toxicity in patients from northeastern Mexico.

e15142 Background: Genetic variants in the UGT1A1 gene affect Irinotecan metabolism and are associated with an increased risk of treatment-related toxicity. Data on the prevalence of clinically relevant UGT1A1 variants and their association with Irinotecan toxicity in Mexican populations are limited. This study aimed to characterize the pharmacogenetic profile of UGT1A1 in Northeastern Mexico. Methods: We conducted an observational case–control study with ongoing enrollment. Cases were patients with gastrointestinal malignancies treated with Irinotecan-based chemotherapy, while controls were cancer-free individuals from the general population with no prior exposure to Irinotecan. UGT1A1 genotyping for alleles *28, *36, and *37 was performed by Sanger sequencing, and allele *6 was analyzed using a qPCR-based assay. Toxicities were graded according to CTCAE criteria. This analysis represents an interim evaluation of the study cohort. Results: At the time of this interim analysis, 211 individuals were included (11 cases and 200 controls). UGT1A1*28 was the most frequent variant (allele frequency: 0.30). The rare alleles UGT1A1*36 (frequency: 0.0075) and UGT1A1*37 (frequency: 0.0025) were identified, representing the first report of these variants in the Mexican mestizo population. No UGT1A1*6 alleles were detected. Genotype-predicted phenotypes suggested a high prevalence of reduced Irinotecan metabolism, with 8.5% poor metabolizers, 43% intermediate metabolizers, and 1.5% ultrarapid metabolizers. Among Irinotecan-treated patients, 27.3% developed grade ≥3 treatment-related toxicity. The most common adverse events were gastrointestinal and hematologic, predominantly grade 1–2. Grade ≥3 toxicities included nausea, diarrhea, vomiting, and fatigue (each 9.1%). Carriers of UGT1A1 toxicity-associated variants showed a higher incidence of severe toxicity compared with non-carriers (37.5% vs 0%), although this association remains exploratory due to the limited sample size. Conclusions: The mestizo population of northeastern Mexico presents a distinctive UGT1A1 allele frequency profile characterized by a high prevalence of variants associated with Irinotecan toxicity, differing from those reported in other Latin American populations. These findings support the relevance of population-specific pre-treatment UGT1A1 genotyping and highlight the need for regional dosing guidelines to reduce severe toxicity and optimize Irinotecan therapy in underrepresented populations.

Dimercaprol (BAL): Insights into conformational stability, fragmentation pathways via tandem LR-ESI, HR-EI mass spectrometry, and gas-phase thermochemical properties from quantum chemical calculations

Dimercaprol (British antilewisite, BAL) is a long-established chelating agent used in the treatment of heavy metal poisoning; however, its physicochemical and thermochemical properties have not yet been fully characterized. In this study, we combined gas-phase quantum chemical calculations with high-resolution mass spectrometry to investigate the conformational stability, fragmentation pathways, and thermochemical parameters of BAL. Fragmentation behavior was examined by gas chromatography/mass spectrometry-QTOF under electronic ionization conditions, and the resulting spectra were interpreted through proposed dissociation pathways involving water, hydrogen sulfide, and thiyl radical losses, supported by reaction enthalpies calculated at the theoretical level M06-2X/6–311++G( 3df,3pd ). Conformational analysis identified five low-energy structures (BAL-1 to BAL-5), where intramolecular hydrogen bonds and gauche/anti interactions play a key role in stability; BAL-3 was consistently predicted as the lowest-energy conformer. Vibrational frequencies calculated with the B3LYP, M06-2X, and MN15 functionals showed good agreement with experimental FTIR and Raman data. The thermochemical properties were further evaluated using G n composite methods (G3MP2B3, G3B3, G4MP2 and G4) which yielded an average standard enthalpy of formation at ΔH° ( f,298K ) of –45.6 ± 1.1 kcal/mol. This work provides a detailed experimental and theoretical characterization of dimercaprol, providing information on its possible fragmentation mechanism and conformational landscape, and offering a thermochemical framework that could support future pharmacological, toxicological and environmental applications.

Lattice orientation-dependent circular photogalvanic effect in hydrogenated TiO2

TiO2 is a widely utilized semiconductor material. This study demonstrates that hydrogenation can induce spin–orbit coupling (SOC) in TiO2, thereby generating a circular photogalvanic effect (CPGE). A significant CPGE is observed in hydrogenated TiO2 (H:TiO2) under illumination at wavelengths of 450, 532, and 780 nm. The crystal orientation is found to significantly influence the CPGE in H:TiO2. Regardless of hydrogenation duration or illumination wavelength, the H:TiO2 (001) exhibits the strongest CPGE. This effect is attributed to the larger lattice interstices along the [001] direction, which facilitate faster hydrogen diffusion. For H:TiO2 subjected to short-term hydrogenation, the role of hydrogen predominates. However, for H:TiO2 subjected to prolonged hydrogenation, oxygen vacancies are inevitably introduced, which alters the relative strength order of the CPGE among the H:TiO2(100), (110), and (111). X-ray photoelectron spectroscopy measurements reveal that a larger shift in the valence band maximum corresponds to a stronger CPGE. This study demonstrates that easily fabricated H:TiO2 exhibits pronounced SOC and can serve as a promising platform for both fundamental research and practical applications.

Deadly cures: Unlocking anticancer potential of reptile, amphibian, and arthropod venoms through molecular innovation and nanotechnology

Synergistic Modulation of Crystallization Kinetics and Interface Energy‐Level Alignment for Efficient Perovskite/Silicon Tandem Solar Cells

ABSTRACT Increasing bromine content is a common strategy for widening the bandgap of perovskite materials. However, this often accelerates crystallization, thereby degrading film quality and energy‐level alignment. To address this, we introduce the multifunctional molecule 2‐amino‐4‐cyanobenzoic acid (2A4CBA) into a 1.68 eV perovskite precursor to simultaneously modulate crystallization and energy‐level. The 2A4CBA interacts with perovskite precursors via coordination and hydrogen bonding through its amino, cyano, and carboxyl groups. These interactions significantly delay crystallization, broaden the processing window, and promote complete reaction between lead halide and salt, thereby reduced residual unreacted lead iodide and ultimately yielded high‐quality perovskite films with larger grains and lower defect state density. Moreover, 2A4CBA differentially modulates the work functions of top surface and buried interface, establishing a graded energy‐level alignment that facilitates separate extraction of electrons and holes. As a result, the single‐junction inverted wide‐bandgap (WBG) perovskite solar cell (PSC) achieved a power conversion efficiency (PCE) of 23.44% along with markedly improved storage and operational stability. Furthermore, a 1 cm 2 perovskite/silicon tandem solar cell (TSC) integrating this optimized WBG perovskite reached a PCE of 33.20% (certified 32.88%) and exhibited excellent operational stability.