Browse Articles

Combination of tulmimetostat and PD-1 blockade for patients with advanced non–small cell lung cancer who progressed from first or second line of treatments.

TPS8664 Background: Immune checkpoint inhibitors (ICIs) have transformed advanced non–small cell lung cancer (NSCLC) treatment, yet many patients exhibit primary ICI resistance or progression after initial response. Epigenetic dysregulation, including aberrant EZH2-mediated trimethylation of histone H3 (H3K27me3), has been implicated in tumor immune evasion and reduced responsiveness to ICIs. Preclinical studies suggest that EZH2 inhibition may enhance antitumor immunity and potentiate PD-1 blockade. Tulmimetostat, a potent dual EZH2/EZH1 inhibitor, demonstrates favorable tolerability and antitumor activity across multiple solid tumors. We hypothesize that the addition of tulmimetostat to pembrolizumab may reverse the resistance to pembrolizumab in patients with ICI-refractory NSCLC. Methods: The primary objective of this open-label, single-arm, phase Ib/II study is to evaluate the safety, tolerability, and preliminary efficacy of combining tulmimetostat with pembrolizumab in Veterans with advanced NSCLC who have progressed after first- or second-line therapy, one of which must have included ICI. Key eligibility criteria include measurable disease per RECIST v1.1 and provision of adequate tumor tissue (archive or new). Tulmimetostat (200 mg or 300 mg) is administered orally once daily, beginning with a 7-day run-in period during cycle 1, followed by pembrolizumab 200 mg IV every 3 weeks. Treatment continues until confirmed disease progression, unacceptable toxicity, or withdrawal. Safety assessments include clinical exams, laboratory evaluations, ECGs, and CTCAE v5.0 grading of adverse events. Imaging is performed at baseline and every 9 weeks. Secondary endpoints include disease control rate, progression-free survival, and duration of response. Up to 18 patients will be enrolled in Phase Ib using a BOIN method to determine the dose-limiting toxicities (DLTs) and the recommended phase II dose (RP2D) of tulmimetostat. Phase II is a Simon optimal two-stage design to assess objective response rate (ORR) per RECIST v1.1. Subjects treated at the RP2D during phase Ib will be included in the efficacy analysis. In total, up to 66 patients will be enrolled across multiple VA sites. Optional on-treatment biopsies will be obtained at select timepoints. Biomarker analyses will include an assessment of H3K27me3 (Methyl Mark) and PD-L1 expression in tumor biopsy samples to explore epigenetic and immunologic correlates of treatment response. Correlative biomarker analyses, including H3K27me3 and PD-L1 expression in tumor biopsies, will offer insight into epigenetic and immunologic determinants of treatment response. Findings from this trial may support the development of a novel therapeutic approach for patients with NSCLC who have progressed after on standard immunotherapy or chemo-immunotherapy. Enrollment was initiated in January 2026. Clinical trial information: NCT05467748 .

Inpatient outcomes among early-onset breast cancer hospitalizations with metabolic dysfunction–associated liver disease: A national estimate.

e12729 Background: The incidence of early-onset breast cancer(EOBC) continues to rise globally. Metabolic dysfunction-associated liver disease (MASLD) is prevalent in younger populations, yet its impact on inpatient outcomes among patients hospitalized with EOBC remains unknown. Methods: Using the National Inpatient Sample (2017–2022), we identified EOBC hospitalizations among patients ≤50 years with ICD-10-CM codes for invasive or in situ breast cancer (C50*, D05*). MASLD was identified using validated NIS-based algorithms. Survey-weighted multivariable regression adjusted for demographics, insurance status, sepsis, Charlson comorbidity category, and hospital characteristics. The primary outcome was all-cause in-hospital mortality; secondary outcomes included mechanical ventilation, discharge disposition, breast surgery during admission, anemia, parenteral nutrition, length of stay (LOS), and hospital costs. Results: Among 190,655 estimated EOBC hospitalizations, 4,825 (2.53%) involved MASLD. Compared with EOBC hospitalizations without MASLD, MASLD was associated with higher odds of in-hospital mortality (adjusted odds ratio [aOR] 1.74, 95% CI 1.37–2.20), mechanical ventilation (aOR 1.41, 95% CI 1.04–1.91), discharge to a facility (aOR 1.48, 95% CI 1.17–1.87), anemia (aOR 1.20, 95% CI 1.04–1.38), and the composite of death or ventilation (aOR 1.64, 95% CI 1.32–2.04; all p≤0.03). MASLD was also associated with longer LOS (~17%) and higher total charges (~8%), and lower odds of undergoing breast surgery during the index admission (aOR 0.38, 95% CI 0.29–0.51). No association was observed with parenteral nutrition. Conclusions: MASLD identifies a high-risk inpatient phenotype among patients hospitalized with EOBC, characterized by significantly greater mortality, morbidity, and resource utilization independent of overall comorbidity burden. Recognition of MASLD during hospitalization may improve risk stratification and support early multidisciplinary management. In-Hospital outcomes of EOBC hospitalizations with vs without MASLD, NIS 2017–2022. Domain Outcome MASLD Effect Estimate 95% CI p-value Cohort size EOBC hospitalizations 190,655 total — — MASLD prevalence 2.53% — — Mortality / Severity In-hospital mortality aOR 1.74 1.37–2.20 <0.001 Mechanical ventilation aOR 1.41 1.04–1.91 0.027 Composite (death or ventilation) aOR 1.64 1.32–2.04 <0.001 Disposition Discharge to facility/rehab aOR 1.48 1.17–1.87 0.001 Procedures Breast surgery during admission aOR 0.38 0.29–0.51 <0.001 Parenteral nutrition aOR 1.08 0.87–1.35 0.46 Complications Anemia aOR 1.20 1.04–1.38 0.011 Acute kidney injury ↑ (unadjusted prevalence higher) — — Sepsis ↑ (adjusted in models) — — Venous thromboembolism ↑ (unadjusted prevalence higher) — — Resource utilization Length of stay +17% — <0.001 Total hospital charges +8% — 0.011

Building a real-time cancer data mart (CDM) for an integrated health system, leveraging traditional natural language processing (NLP) methods, curated data sources, and large language models (LLMs).

e13654 Background: Timely cancer case ascertainment is critical for supporting clinical trials, research, and operational workflows. Despite being the gold standard, information on people diagnosed with cancer from accredited cancer registries is often delayed well over one-year post-diagnosis. Advances in LLMs offer promising solutions but require high computational and GPU power. A novel strategy combining rule-based NLP with curated data sources and selective LLM use may offer a practical alternative. We developed a real-time CDM and evaluated two approaches: an NLP-only method and a hybrid method for identifying incident cancer cases and classifying key oncology characteristics. Methods: We identified 230,500 patients with no cancer history who had pathology reports or cancer diagnoses between 07/01/2023 and 12/31/2023 in Kaiser Permanente Northern California (KPNC). All records were processed using eMaRC, a CDC rule-based NLP model widely used by cancer registries. For the hybrid method, pathology reports indicating malignancy or suspicious findings were further analyzed using MedGemma, a generative LLM trained on medical datasets and applied using structured prompts. Both models classified malignancy, primary site, histology, and behavior. When models disagreed on, MedGemma results were prioritized. We evaluated accuracy by comparing cancer cases against those recorded in KPNC Cancer Registry, which conforms to SEER/NAACCR standards. Sensitivity, specificity, PPV, and NPV were calculated for 1631 breast, 851 colorectal, 677 lung, 687 melanoma, and 1358 prostate cancers cases, which together represent 60% of annual cases recorded in the registry. Results: Both methods – eMaRC only and hybrid approach – demonstrated high specificity ( > 99%). Adding MedGemma to eMaRC resulted sensitivities of 97%–99% across all cancer types, an average 2% increase over eMaRC alone. Sensitivity improved the most for prostate cancer (95% to 99%), followed by lung cancer (94% to 97%). Among 230,500 patients, true case prevalence ranged from 0.3% to 0.7%. Using the hybrid approach, was 91% for breast and prostate, ranged from 84%-86% for melanoma and colorectal, and lowest for lung at 71%. In practice, cases classified as breast and prostate are generally reliable, while lung cancer classification needed further review due to a higher rate of false positives. Conclusions: A hybrid method combining rule-based NLP, curated data sources, and selective LLM use achieved overall high sensitivity and specificity. Although less detailed than the cancer registry and prone to misclassification of certain cancers, the CDM’s rapid and automated process is highly scalable and efficient, making it valuable for operational use and research requiring rapid case identifications.

Central nervous system metastases in gynecologic cancers: Real-world outcomes and prognostic factors.

e14012 Background: CNS metastases from gynaecologic cancer (GC) are uncommon and evidence guiding management is limited. We characterized a real-world cohort and evaluated outcomes and prognostic factors. Methods: We conducted a retrospective cohort study including patients with primary GC who developed radiologically confirmed parenchymal brain metastases and/or leptomeningeal disease between January 2013 and June 2025 at a comprehensive cancer center. Overall survival (OS) was assessed from primary diagnosis and from CNS involvement. Survival analyses used Kaplan–Meier methods and Cox regression. Prespecified multivariable models included intracranial disease burden (high: ≥3 lesions or leptomeningeal disease vs low: ≤2 lesions) and presence of extracranial metastatic disease at CNS diagnosis. Results: Among 13.505 GC patients, 33 (0.24%) developed CNS metastases. Median age at primary diagnosis was 59.2 years (34.5-76.4) and primary sites were endometrial (45.4%, 15/33), ovarian (21.2%, 7/33), cervical (18.2%, 6/33), and other/mixed (15.2%, 5/33). Median interval from diagnosis to CNS metastasis was 34.3 months (3.1–227.0). Most patients (93.9%, 31/33) were symptomatic and 90.9% (30/33) had extracranial metastases at CNS diagnosis. CNS involvement was parenchymal in 90.9% (30/33) and leptomeningeal in 9.1% (3/33). Among parenchymal cases, 56.7% (17/30) had multiple lesions and with ≥3 lesions 40.0% (12/30). Brain-directed management was supportive care (48.5%, 16/33), radiotherapy alone (24.2%, 8/33), multimodal therapy - surgery plus adjuvant radiotherapy/SRS (21.2%, 7/33), or surgery alone (6.1%, 2/33). Median OS from primary diagnosis was 40.9 months (95% CI: 16.4–65.45 months) and after CNS metastasis 2.17 months (6- and 12-month OS, 27.3% and 18.2%, respectively). Survival differed by strategy, with longer OS in patients receiving multimodal therapy compared with radiotherapy alone or supportive care (median 19.75 vs 2.96 vs 1.05 months). On multivariable analysis, higher CNS disease burden was associated with shorter OS (adjusted HR 2.72, 95% CI 1.24–5.97; P=0.012). Conclusions: CNS metastases from gynaecologic cancers were rare but associated with extremely poor prognosis, typically occurring in the setting of widespread extracranial disease. Intracranial disease burden and brain-directed treatment strategy were key prognostic factors. Selected patients with low CNS disease burden may achieve meaningful survival benefit from multimodal therapy.

Optec/Optal: A phase 2 study to evaluate outpatient (OP), step-up administration of teclistamab (Tec) or talquetamab (Tal) with prophylactic tocilizumab (prophyToci) in patients (pts) with relapsed/refractory multiple myeloma (RRMM).

7510 Background: Tec, a B-cell maturation antigen x CD3 bispecific antibody (bsAb), and Tal, a G protein-coupled receptor class C group 5 member D x CD3 bsAb, are approved for the treatment of pts with triple-class exposed (TCE) RRMM. Cytokine release syndrome (CRS) is a common adverse event associated with T-cell therapies. CRS occurred in 72% and 73-79% of pts in the MajesTEC-1 (TEC-1) and MonumenTAL-1 (TAL-1) studies, respectively. One cohort of pts in TEC-1 received prophyToci, resulting in a lower rate of CRS (26%) without negatively impacting safety or efficacy. A cohort of pts in TAL-1 received prophyToci, with 18% experiencing CRS; safety and efficacy were comparable to the overall TAL-1 population. Methods: This nonrandomized, multicenter, prospective study evaluates prophyToci in RRMM pts treated with Tec (Arm A) or Tal (Arm B) using the approved step-up dose (SUD) schedule in the OP setting. ProphyToci 8 mg/kg was administered before SUD 1 of either Tec or Tal. Specific recommendations (home monitoring, IV immunoglobulin [IVIG] for IgG levels <400 mg/dL) and restrictions (caregiver requirements and travel distance) are also provided for pt safety. Primary endpoint is incidence of any grade CRS in cycles 1-2. Secondary endpoints include safety and efficacy measures. Safety parameters encompass incidence of Grade ≥2 CRS, any grade recurrent CRS, any grade infections, neurotoxicity (NT), and neutropenia throughout the study. Efficacy endpoints include overall response rate (ORR), best overall response (BOR) and progression-free survival (PFS). Results: As of 06 January 2026, 43 pts enrolled in Arm A, and 7, in Arm B. In Arm A, the CRS rate was 5%, all G1. In Arm B, G1 CRS occurred in 14% of pts. Adverse events G≥3 included neutropenia (G3, n=4; G4, n=6), anemia (G3, n=1), and febrile neutropenia (G3, n=1). Twenty pts (47%) in Arm A experienced infections, with 19% ≥G3. Among Arm A patients who experienced infections, 10 (50%) exhibited IgG levels <400 mg/dL, of whom 2 (20%) received IVIG. In Arm B, 4 pts (57%) experienced infections, with 29% (n=2) ≥G3. BOR in Arm A included 3 stringent complete responses (sCRs), 6 CRs, 11 very good partial responses (VGPR), 8 PRs, 1 stable disease, 5 progressive disease (PD), and 9 non-evaluable pts. Among the 34 evaluable pts in Arm A, the ORR was 82%. BOR in Arm B included 1 sCR and 6 non-evaluable pts. After a median follow-up of 11.8 months, 74.4% of pts (n=32) in Arm A did not experience progression. Conclusions: A single dose of prophyToci before SUD 1 of Tec or Tal reduced the incidence of CRS, with no impact on safety or efficacy, and supports administration of these bsAbs in the OP setting. The protocol was amended to add an arm evaluating the effect of prophylactic oral dexamethasone on CRS in pts treated with Tec (Arm C). Enrollment is ongoing (NCT05972135). Clinical trial information: NCT05972135 .

Large‐Scale Profiling of Kinase Degradation by Using Norbornene‐Based Hydrophobic Tag (HyT) Strategy

ABSTRACT Hydrophobic tags (HyTs) are promising bifunctional protein degraders that mimic misfolded proteins to trigger quality control‐mediated target degradation, offering key advantages over traditional heterobifunctional degraders such as proteolysis‐targeting chimeras (PROTACs). However, the scope and generality of this targeted protein degradation (TPD) strategy across the human kinome remain unexplored. In this study, we first addressed this gap by developing two general HyTs on the basis of a pan‐kinase scaffold capable of large‐scale proteome‐wide studies of kinase degradation by using quantitative chemoproteomics. We subsequently mapped the degradable kinome landscape by using a HyT‐based strategy, leading to the successful identification of 169 HyT‐degradable human kinases. Leveraging this comprehensive kinase degradome map, we next rationally designed norbornene‐based HyT degraders against ABL and AURKA, obtaining two optimized HyTs with potent degradation capabilities. We further elucidated the detailed mechanistic insight of these novel degraders. To overcome the inherent poor water solubility of HyTs, we next engineered HyT‐loaded, tumor microenvironment (TME)‐responsive self‐assembled nanoparticles (NPs), which showed improved tumor accumulation and therapeutic efficacy in vivo. With key advantages including rapid target identification and a modular NP‐based delivery platform, our work herein provides a comprehensive framework for future development of potential kinase therapeutics based on HyT degraders.

Structural integrity and optical properties of MgO nanoparticles synthesized via solution combustion

A global helium clock for groundwater residence time

Pharmacological targeting of the mycothiol cysteine ligase MshC in Mycobacterium abscessus

Defining prostaglandin-driven dysregulation in <i>PIK3CA</i> -mutant colorectal cancers: A real-world multi-omic analysis.

3667 Background: PIK3CA mutations occur in a meaningful subset of colorectal cancers (CRC) and have been associated with improved clinical outcomes in patients receiving aspirin and celecoxib. PIK3CA activation has been linked to upregulation of PTGS2 and downstream prostaglandin E2 signaling, promoting tumor-associated inflammation, angiogenesis, and immune suppression. Aspirin-mediated inhibition of prostaglandin synthesis may therefore counteract this biology in PIK3CA -mutant CRC, although the underlying molecular mechanisms remain incompletely characterized. Methods: A total of 823 CRC tumor samples underwent DNA-based next-generation sequencing, with matched whole-transcriptome sequencing (20,802 genes) available for 263 tumors. PIK3CA mutation prevalence, hotspot distribution, and co-mutation patterns were evaluated. A predefined prostaglandin–inflammation gene signature ( PTGS2, PTGES, PLA2G4A, IL6, STAT3, PTGER4, VEGFA, IDO1 ) was assessed (log₂ fold change &gt;2). Coordinated signature activation was quantified. Results: PIK3CA mutations were detected in 12.8% (105/823) of CRCs and were more frequent in right than left-sided tumors (15.2% vs 11.1%). Among PIK3CA -mutant tumors, 25.7% were KRAS/NRAS/BRAF wild type (WT), while KRAS co-mutation predominated (65.7%); NRAS and BRAF co-mutations were rare (1.9% and 5.7%), as summarized in the Table. ERBB2 alterations were rare in PIK3CA -mutant tumors (4.8%) and in KRAS/NRAS/BRAF -mutant tumors (3.2%,), while they were enriched in the KRAS/NRAS/BRAF WT subgroup (8.9%). Among PIK3CA -mutant tumors, key prostaglandin biosynthesis genes PTGS2, PTGES , and PLA2G4A were overexpressed in 8.8% (3/34), while downstream inflammatory and immune-modulatory genes including IL6, STAT3, IDO1 , and CD274 were overexpressed in 11.8% (4/34). Prostaglandin–inflammation signature activation was more frequent in KRAS WT than KRAS -co-mutant PIK3CA tumors (57.1% [4/7] vs 51.9% [14/27]). Among microsatellite-stable (MSS) tumors, 47.6% (10/21) showed signature activation. Tumors with combined PIK3CA mutation and prostaglandin pathway activation showed enrichment of inflammatory, angiogenic, and epithelial–mesenchymal transition–related programs. Conclusions: PIK3CA -mutant CRC represents a biologically distinct subset characterized by coordinated activation of prostaglandin-driven inflammatory and immune-modulatory pathways. These features provide a mechanistic rationale for the reported aspirin benefit in PIK3CA -mutant CRC and support integrated genomic–transcriptomic profiling to guide biomarker-driven therapeutic strategies. Prevalence of PIK3CA alterations and co-mutations in CRC. Biomarkers Prevalence (%; n=105) PIK3CA mutation (overall) 12.8 (105/823) E545K 21.9 E542K 19 H1047R 16.2 H1047L 6.7 Q546K 5.7 PIK3CA + KRAS 65.7 PIK3CA + NRAS 1.9 PIK3CA + BRAF 5.7 PIK3CA + KRAS/NRAS/BRAF WT 25.7

Mapping the ovarian cancer diagnostic journey in Israel: Results from a qualitative patient and clinician study.

e17570 Background: Ovarian cancer is the most lethal gynecologic malignancy worldwide, with survival declining sharply between early and advanced stages, making timely diagnosis critical. Globally, the median interval from symptom onset to diagnosis remains prolonged (approximately 31 weeks), with substantial variation across countries. Patient journey mapping has improved diagnostic timelines in other disease areas and studies by the World Ovarian Cancer Advocacy Coalition emphasizes both the scarcity of research on ovarian cancer mapping and the variation in patient experiences and outcomes between countries. However, no systematic assessment of the ovarian cancer diagnostic journey has been conducted in Israel. This two-part study aims to identify barriers to timely diagnosis in Israel, with Part 1 qualitative interviews informing the development of a national patient journey questionnaire in Part 2. This abstract reports Part 1 findings. Objective: The overall objective of this study is to characterize the ovarian cancer diagnostic journey in Israel. Part 1, detailed here, uses patient and clinician interviews to develop a diagnostic framework to inform the development of a national patient journey questionnaire used in Part 2. Methods: In-depth semi-structured interviews were conducted with 11 women diagnosed with ovarian cancer (10 without known BRCA mutation at symptom onset; 1 with known BRCA mutation), and a multidisciplinary group of clinicians, including gynecologic oncologists, gynecologists, family physicians, a urogynecologist, and a genetic counselor. Interviews were analyzed using thematic analysis to construct diagnostic trajectories and identify potential deviations. Results: Interviews revealed a typical diagnostic pathway beginning with symptom recognition, followed by evaluation in community-based primary or gynecologic care, preliminary diagnostic testing, hospital referral, and eventual gynecologic oncology consultation. Both patients and clinicians described frequent deviations from this pathway (“diagnostic sidetracks”). Common sidetracks included normalization or minimization of symptoms, misdiagnosis, unclear referral pathways, limited access to diagnostic testing, and delays in appointment availability. Facilitators of faster diagnosis included prior awareness of BRCA mutation status and strong patient persistence in healthcare relationships, suggesting a more direct diagnostic route for women aware of their genetic risk. Conclusions: These Part 1 findings establish a diagnostic journey framework to guide the development of a nationwide ovarian cancer patient journey questionnaire. By identifying common diagnostic deviations and facilitators, this work supports future system-level interventions aimed at reducing diagnostic delays and enabling more efficient detection of ovarian cancer.

A multicenter single-arm study of induction EC/EP plus camrelizumab followed by maintenance camrelizumab plus apatinib in treatment-naive advanced or metastatic extrapulmonary neuroendocrine carcinoma.

4176 Background: Extrapulmonary neuroendocrine carcinoma (EP-NEC) is a poorly differentiated and highly aggressive malignancy with poor prognosis. Platinum (cisplatin or carboplatin) plus etoposide (EC/EP) remains the standard first-line regimen. However, clinical outcomes are suboptimal, highlighting the need for more effective strategies. Camrelizumab is a programmed cell death-1 (PD-1) inhibitor, and apatinib is a vascular endothelial growth factor receptor-2 (VEGFR2) inhibitor. Both have demonstrated antitumor activity in multiple solid tumors. Therefore, we investigated a sequential regimen consisting of induction EC/EP plus camrelizumab followed by maintenance camrelizumab plus apatinib in treatment-naive patients with advanced or metastatic EP-NEC. Methods: This multicenter, single-arm trial enrolled patients with previously untreated advanced or metastatic EP-NEC. Patients received 4–6 cycles of induction therapy with cisplatin (25 mg/m² iv, d1-3, q3w) or carboplatin (AUC = 5 iv, d1, q3w) plus etoposide (100 mg/m² iv, d1-3, q3w) in combination with camrelizumab (200 mg iv, d1, q3w). Patients without disease progression received maintenance camrelizumab (200 mg iv, q3w) plus apatinib (250 mg, qd) until progression or unacceptable toxicity. The primary endpoint was objective response rate (ORR). Secondary endpoints included disease control rate (DCR), progression-free survival (PFS), overall survival (OS), and safety. Tumor response was assessed every 6 weeks per RECIST v1.1. The trial is registered on ClinicalTrials.gov, NCT05142865. Results: As of January 2026, 30 patients were enrolled (median age of 59.5 years, range 30–73), and all 30 patients were evaluable for efficacy. The ORR was 66.7% and the DCR was 83.3%. Median PFS was 9.87 months (95% CI, 6.23–NA). Median OS was not reached, and the 1-year OS rate was 74.6%. All patients experienced adverse events (AEs), with grade ≥3 AEs observed in 43.3% (13/30). Grade ≥3 AEs were mainly hematologic toxicities, including anemia (13.3%), neutropenia (13.3%) and thrombocytopenia (13.3%). Common non-hematologic AEs included elevated transaminases (43.3%), mainly grade 1–2, with grade ≥3 events in 10.0%. Conclusions: Induction EC/EP plus camrelizumab followed by maintenance camrelizumab plus apatinib showed encouraging activity and manageable safety in treatment-naive advanced or metastatic EP-NEC. Further studies are warranted to confirm these findings. Clinical trial information: NCT05142865 .

Clinicopathologic and molecular characterization of <i>ARID1A</i> -mutant solid tumors using paired tissue and liquid biopsy sequencing.

e15054 Background: ARID1A is a commonly altered chromatin-remodeling gene across solid tumors, with roles in DNA damage repair, tumor immunogenicity and therapeutic response. The clinical significance of ARID1A alterations detected in liquid biopsy, particularly at low variant allele frequency (VAF), remains uncertain. We performed a large clinicopathologic and molecular analysis to define tissue–plasma concordance and clinically relevant VAF thresholds. Methods: Solid tumor cases with tumor tissue and/or liquid biopsy next-generation sequencing (NGS) performed on the MDA MAPP platform at MD Anderson Cancer Center (2019–2025) were analyzed. Tissue–plasma concordance was evaluated in paired samples. Concordance, associations, and overall survival were assessed using standard statistical methods. Results: Among 17,092 patients with NGS data, 1,033 (6.0%) harbored ARID1A mutations in liquid biopsies across multiple tumor types, including colorectal (n = 218, 21.1%), lung carcinomas (n = 228, 22.1%), breast invasive ductal (n = 107, 10.4%), and cholangiocarcinoma (n = 88, 8.1%). Across ARID1A variants, 48.9% were pathogenic/likely pathogenic and 51.1% were variants of unknown significance/benign. ARID1A -mutant tumors showed high genomic complexity, with co-alterations in TP53 (81.4%), APC (41.3%), PIK3CA (33.5%), and KRAS (27.6%). Across 1,303 plasma-detected ARID1A variants, circulating tumor DNA (ctDNA) VAF was distributed as follows: 40.1% &lt; 1%, 30.1% 1–5%, 19.9% 5–20%, and 10.0% &gt; 20%. A total of 303 patients had paired tissue/plasma sequencing. At the patient level, tissue-plasma concordance was complete in 66.7%, partial in 22.4%, and absent in 10.9% of patients. At the variant level, tissue–plasma concordance increased stepwise with higher VAF: 55.7% at &lt; 1% (44/79), 84.7% at 1–5% (61/72), 89.7% at 5–20% (52/58), and 86% at &gt; 20% (26/30). Using a plasma VAF cutoff of 1%, variants with VAF ≥1% had significantly higher tissue–plasma concordance than those with VAF &lt; 1% (85.6% vs 55.7%; OR 4.74, 95% CI 2.60–8.65; p = 8.96×10⁻⁷ ). Concordance by tumor type was compared between variants with VAF &gt; 1% and &lt; 1%, with highest concordance observed in colorectal adenocarcinoma (85.0% vs 59.3%; p = 0.01 ), lung adenocarcinoma (90.6% vs 61.1%; p = 0.02 ), and cholangiocarcinoma (95.0% vs 66.7%; p = 0.15 ). In exploratory analyses, tumors with ARID1A mutations had inferior overall survival compared with ARID1A –wild-type tumors (log-rank p = 0.0002 ). Conclusions: ARID1A alterations detected in liquid biopsy are common but variably concordant with tumor sequencing. Plasma ctDNA VAF is a determinant of reliability, with a ≥1% threshold identifying ARID1A variants with high tissue concordance, particularly in colorectal cancer and lung adenocarcinoma. Integrating VAF, pathogenicity, and tumor context is essential for clinical interpretation of ARID1A liquid biopsy results.

Survival outcomes of cabozantinib treatment with and without immune checkpoint inhibition in patients with heavily pretreated advanced sarcoma.

11551 Background: Cabozantinib is an oral tyrosine kinase inhibitor (TKI), with key targets including MET, VEGFR2, AXL, RET, and others. It is commonly used to treat specific carcinomas, but its efficacy in treating sarcomas is still emerging, especially in combination with immune checkpoint inhibition. The benefit of cabozantinib with or without immune checkpoint inhibition (ICI) in previously heavily treated patients with advanced sarcoma is unclear, underscoring the need for further investigation. Methods: We report results from a retrospective study of heavily pretreated advanced sarcoma patients who received cabozantinib with or without ICI. Eligible patients included those who had locally advanced or metastatic sarcoma and were exposed to cabozantinib for a minimum of one day. Patients who participated in the clinical trial “Randomized phase II trial of cabozantinib combined with PD-1 and CTLA-4 inhibition versus cabozantinib in metastatic soft tissue sarcoma” were excluded from this study. Results: There were 105 eligible patients with thirty-one different histology types, a median age of 55 years old, of which 55% were female patients. The median lines of prior treatment were four. Approximately 40.9% of patients were treated with cabozantinib alone and 59.1% were treated with cabozantinib plus either nivolumab, or nivolumab plus ipilimumab, or pembrolizumab. The progression-free survival (PFS) of the full cohort was 3.3 months. Eight patients (7.6%) obtained a partial response (PR) and forty-two patients (40.0%) had stable disease (SD) longer than 3 months, with a clinical benefit rate (CBR) of 47.6%. Partial responses occurred across a variety of histologic subtypes, including small round cell tumor, UPS, osteosarcoma, leiomyosarcoma, GIST, liposarcoma, and angiosarcoma. Stable disease was observed in a variety as well, including but not limited to LMS, synovial sarcoma, ASPS, chondrosarcoma, and osteosarcoma. The median PFS and OS for the 47.6% of patients who obtained a clinical benefit (PR + SD) were 6.9 and 22.9 months, compared to the median PFS of 2.1 months (p &lt; 0.001) and median OS of 7.4 months (p = 0.001) for the patients with progressive disease. Several patients with aggressive histology including osteosarcoma, desmoplastic small round cell tumor, synovial sarcoma and others had PFS more than 12 months. Combination therapy with cabozantinib and an ICI accounted for the majority of PR’s observed, including the longest ongoing PR (&gt;15.3 months) in a patient with desmoplastic small round cell tumor. Conclusions: In a cohort of heavily treated patients with advanced soft tissue and bone sarcoma, cabozantinib alone or with ICI showed durable CBR in nearly 50% of patients, suggesting cabozantinib alone or with ICI could be of substantial benefits to certain sarcoma patients who have very limited treatment options remaining.

Real-world survival outcomes of surgical patients with squamous cell carcinoma of unknown primary of the head and neck: A National Cancer Database analysis.

e18142 Background: Squamous cell carcinoma of unknown primary (SC-CUP) of the head and neck is an infrequent disease, presenting with metastatic cancer in the cervical lymph nodes without an identifiable primary tumor site despite extensive work-up. Given the rarity of the disease, it is still unclear if conventional adverse prognostic factors in patients with head and neck squamous cell carcinoma with known primary, such as higher pathologic N stage, extranodal extension (ENE), lymphovascular invasion (LVI), perineural invasion (PNI), or positive resection margin, would similarly translate to those with SC-CUP. Therefore, we aim to explore this using an established real-world clinical registry from the National Cancer Database (NCDB). Methods: We included patients with SC-CUP of the head and neck with ICD-O-3 codes C00.0-C14.8 or C32.0-C32.9 in the 2022 NCDB registry who were diagnosed from 2004-2021. They must have a confirmed diagnosis of squamous cell carcinoma with pathologic T0, N1-3, and M0 stage with no prior cancer diagnosis, and have undergone wide, total, or radical excisions as part of the treatment. Patients with nasal cavity, middle ear, paranasal sinuses, or nasopharyngeal cancers, as well as those with clinical M1 disease, were excluded. Survival analysis was evaluated using univariate Cox regression analysis. Results: A total of 161 patients were included in the analysis. Most patients were male (85.1%) and White (91.3%). 54 patients received concurrent chemoradiation (CCRT), 48 received radiation therapy (RT) alone, and 40 received no chemotherapy or RT. For pathologic nodal staging, 115, 42, and 4 patients had N1, N2, and N3 disease, respectively. Patients with positive ENE or LVI had numerically poorer overall survival (OS) (hazard ratio (HR) for positive ENE 1.61, 95% confidence interval (CI) 0.69-3.75, p=0.27 and HR for positive LVI 1.67, 95%CI 0.18-15.25, p=0.65), along with those with pN3 disease (HR 1.70, 95% CI 0.22-12.91, p=0.61) and HPV-negativity (HR 4.96, 95% CI 0.45-55.04, p=0.19). However, patients with pN2 disease had outcomes similar to those with pN1 disease (HR 1.04, 95% CI 0.46-2.36, p=0.92). There was a trend toward improved OS in those who received radiation therapy alone (HR 0.43, 95% CI 0.13-1.38, p=0.16), whereas CCRT did not improve survival (HR 0.80, 95% CI 0.30-2.10, p=0.65). Of note, there is insufficient survival data on patients with positive resection margins or PNI for the analysis. Conclusions: This is the largest study to date that evaluated patients with surgically resected SC-CUP of the head and neck. Positive ENE, LVI, pN3, and HPV-negative disease are potential adverse risk factors. Adjuvant radiation therapy could potentially improve OS in this population, but concurrent chemotherapy with radiation was not beneficial. Further studies that evaluate the role of radiation therapy should be pursued.

HRS-8080, an oral selective estrogen receptor degrader (SERD), in ER+/HER2− advanced breast cancer (aBC) with <i>ESR1</i> mutations ( <i>ESR1</i> m): Results from a first-in-human phase 1 trial.

1069 Background: HRS-8080 is a novel, potent oral SERD that binds ER with high affinity. We conducted a multi-part, first-in-human phase 1 trial of HRS-8080 in ER+/HER2- aBC, and here report findings in the ESR1 m subset from the dose-escalation/expansion parts for HRS-8080 monotherapy. Methods: This multicenter, open-label, single-arm phase 1 study enrolled patients (pts) with locally advanced/metastatic ER+/HER2- BC previously treated with ≥1 line of endocrine therapy (ET) and ≤2 lines of chemotherapy for aBC; ESR1 m status was centrally assessed by ctDNA. In Part A (i3+3 design), pts received HRS-8080 QD (50, 150, 300, 600, or 900 mg) or BID (300 mg); in Part B, 300 mg BID/600 mg QD were selected for expansion. Safety was a primary objective. Results: 40 pts with ESR1 m were treated (≥2 prior lines of therapy, 70.0%; ≥2 prior lines of ET for aBC, 42.5%; chemotherapy for aBC, 57.5%), including 32 at expansion doses. Overall, grade ≥3 treatment-related adverse events (TRAEs) occurred in 8 (20.0%) of 40 pts; the most common were decreased lymphocyte count and decreased neutrophil count (2 [5.0%] pts each). TRAEs led to dose reduction in 1 (2.5%) pt. There were no treatment-related deaths. Efficacy in the ESR1 m subset is shown in Table 1. At expansion doses of 300 mg BID/600 mg QD, the confirmed objective response rate was 34.4% (11/32; 95% CI 18.6–53.2) and the disease control rate was 75.0% (24/32; 95% CI 56.6–88.5). Median duration of response was 11.1 mo (95% CI 6.5–not reached), with over half of responses remained ongoing. As of data cutoff, 21 (65.6%) pts had disease progression or died; median progression-free survival was 7.2 mo (95% CI 3.6–11.1). Conclusions: HRS-8080 demonstrated a favorable safety and tolerability profile and encouraging antitumor activity in ESR1m ER+/HER2− aBC, supporting ongoing clinical development both as monotherapy and in combination regimens. Clinical trial information: NCT05189717 . Key baseline characteristics and efficacy in ESR1 m subset. * 50mg QD~300mg QD (n=6) 300mg BID+600mg QD (n=32) 900mg QD(n=2) All patients(n=40) ≥2 prior lines of therapy, n (%) 5 (83.3) 21 (65.6) 2 (100.0) 28 (70.0) ≥2 prior lines of ET for aBC, n (%) 3 (50.0) 13 (40.6) 1 (50.0) 17 (42.5) Chemotherapy for aBC, n (%) 5 (83.3) 16 (50.0) 2 (100.0) 23 (57.5) Confirmed ORR, %(n/N; 95% CI) 0 34.4(11/32; 18.6–53.2) 0 27.5(11/40; 14.6–43.9) Median DoR, mo (95% CI) – 11.1 (6.5–NR) – 11.1 (6.5–NR) Median PFS, mo (95% CI) 2.2 (1.7–5.5) 7.2 (3.6–11.1) NR (9.2–NR) 7.2 (3.6–10.9) Median follow-up, mo (range) 3.0 (0.3–24.6) 6.7 (1.5–34.3) 21.3 (10.4–32.2) 6.6 (0.3–34.3) Data cutoff, Dec. 18, 2025. * Patients relapsed after 24 months if receiving adjuvant ET, and had disease progression ≥6 months of the first-line ET in the advanced setting. aBC, advanced breast cancer; DoR, duration of response; ET, endocrine therapy; NR, not reached; ORR, objective response rate; PFS, progression-free survival.

Correlation of University of Texas Cancer Cachexia Staging (UTCCS) with clinical variables of systemic inflammation, nutritional status, and physical function in patients with cancer cachexia.

12046 Background: Cancer cachexia is a heterogenous syndrome characterized by ongoing loss of skeletal mass with or without adipose tissue, that is not entirely reversible. University of Texas Cancer Cachexia Staging (UTCCS) is a recently described novel model for staging cachexia using simple clinical parameters such as body mass index (BMI), serum albumin (alb), neutrophil to lymphocyte ratio (NLR) and resting heart rate (RHR). The validity of this model to correlate with functional status, nutrition and overall survival in patients with advanced or metastatic cancer has been assessed in current study. Methods: In a prospective study of newly diagnosed cancer patients with stage III/IV cancers, &gt;5% body weight loss and serum albumin &lt; 3.5g/dL biomarkers of cachexia were identified. UTCCS was developed using 4 clinical variables namely (BMI), (Alb), (NLR) and (RHR) using quartile as cut off. Based on these variable patients were categorized into stage I, stage II or stage III cancer cachexia. Blood biomarkers, dietary assessment and functional status were estimated. Kaplan-Meier survival curves were generated for patients in each stage. Results: Cancer distribution amongst 110 patients included lung cancer (n=50), GI cancers (n=42) and other cancers (n= 18). Cachexia stage distribution in patients was assessed (stage I= 45%, stage II =36% and stage III=18%). Patients with stage III cachexia had statistically significantly lower ECOG performance status (p=0.006), higher inflammatory state as measured by advanced lung cancer inflammation index (p&lt;0.001), lower simplified nutritional appetite questionnaire (SNAQ score) (p=0.01), lower skeletal muscle index (p=0.03) and 30 seconds chair rise time (p&lt;0.001). Median overall survival (OS) for stage I was 12.7 months, stage II 7.1 months and stage III 3.5 months (p=0.047). Conclusions: UTCCS system is a novel way of estimating the severity of cancer cachexia and correlates with inflammatory, dietary and functional parameters expected in patients with increasing severity of cancer cachexia.

Surgical management and survival in pelvic osteosarcoma.

e23502 Background: Osteosarcoma is the most common malignant bone tumor in adolescents, young adults and older adults. Current treatment management involves chemotherapy and surgical resection. However, pelvic osteosarcoma poorly responds to neoadjuvant chemotherapy and is difficult to obtain surgical margins due to pelvic anatomy, necessitating a hemipelvectomy. Studies addressing pelvic osteosarcoma surgical treatment modalities remain limited and necessary to characterize survival and clinical outcomes. Methods: This retrospective cohort study utilized the National Cancer Database (NCDB) to identify patients with osteosarcoma diagnosed between 2004-2021. Osteosarcoma of the pelvis was identified using International Classification of Diseases for Oncology (ICD-O-3), histology code 9180–9195 and primary site code C41.4. Only patients with osteosarcoma as their sole primary malignancy were included. Those with missing survival data, negative follow-up time or non-surgical management were excluded. Chi-square testing, Kaplan–Meier analysis, and Cox regression were performed on this sample, and statistical significance was defined as p &lt; 0.05. Results: A total of 377 individuals were included. Patients who underwent pelvic amputation had worse 5-year survival than radical excision (mean survival 92.3 vs 124.5 months, p&lt;0.001). Patients with positive surgical margins (HR 8.415, 95% CI 2.238–31.649, p=0.002), NCDB Stage IV (HR 10.550, CI 1.498–74.289, p=0.018), and increasing age (HR 1.097, 95% CI 1.035-1.163, p=0.002) were independently associated with worse survival. Patients who received radiation therapy (HR 0.140, 95% CI 0.027–0.724, p=0.019) and later year of diagnosis (HR 0.878, 95% CI 0.783–0.986, p=0.028) were associated with improved survival. Surgical procedure type was not an independent predictor of survival (HR: 0.478, 95% CI: 0.190–1.212, p=0.118). Patients who underwent pelvic amputation frequently had poorly differentiated tumors, NCDB stage II-IV, tumor size of 100 to &lt;150 mm, received chemotherapy and on medicare and medicaid. Patients who underwent radical excision tended to have well or moderately differentiated tumors, NCDB stage I, tumor size of 50 to &lt;100 mm, and have private insurance. Conclusions: Patients who underwent radical excision experienced better survival compared with pelvic amputation. However, when adjusting for clinical and tumor-related factors, surgical procedure type was not an independent predictor of survival, suggesting that outcomes are largely driven by underlying tumor biology and disease burden. Positive surgical margins, advanced NCDB stage, and increasing age were independently associated with worse survival, whereas patients who received radiation therapy and more recent year of diagnosis were associated with improved outcomes. Further studies should focus on anatomic extent, margin optimization and functional outcomes to guide surgical decision-making.

Progression-free survival as a surrogate for overall survival in ovarian cancer: Real-world evidence-based assessment and trial-level correlation.

e17587 Background: Progression-free survival (PFS) is widely used as a surrogate endpoint for overall survival (OS) in ovarian cancer, but its validity varies across clinical contexts. To evaluate PFS–OS surrogacy using individual-level agreement in a large real-world cohort and trial-level correlation across randomized studies. Methods: An institutional cohort of 4,269 patients was analyzed to compare factor-specific hazard ratios (HRs) for PFS and OS and to quantify individual-level surrogacy using censored Kendall’s τ. Trial-level HRs were extracted from phase II/III ovarian cancer trials. Surrogacy was assessed using Spearman correlation, Bland–Altman agreement, and weighted meta-regression, with stratified analyses by disease setting, treatment modality, publication era, maintenance therapy, and HR_PFS thresholds. Results: In the institutional cohort, HR_PFS and HR_OS showed near-perfect concordance (R = 0.98) with minimal bias. Censored Kendall’s τ indicated strong individual-level correlation (τ = 0.78; p &lt; 0.001), with 89% concordant event ordering. Across trials, PFS and OS demonstrated moderate correlation (ρ ≈ 0.62) with small mean bias but wide limits of agreement. Surrogacy was strongest in platinum-resistant disease and first-line cytotoxic chemotherapy, and weaker in platinum-sensitive and targeted therapy settings. Maintenance-therapy trials showed more stable agreement than non-maintenance trials. Temporal analyses revealed the most robust surrogacy in the 2010s, with attenuation in the 2020s as treatment heterogeneity increased. Conclusions: PFS is a practical but context-dependent surrogate for OS in ovarian cancer. Strong individual-level and HR-level concordance supports its use, although trial-level surrogacy varies by disease setting, treatment modality, and therapeutic era. Continued validation is essential as modern treatments reshape post-progression outcomes.

Lost in translation? Evaluating associations between English proficiency and next generation sequencing completion, and comparing understandability across different translation modalities among Spanish speakers with metastatic breast cancer.

1524 Background: Disparities in next-generation sequencing (NGS) utilization persist among patients with metastatic breast cancer (MBC), disproportionately affecting Hispanic/Latino populations. We hypothesized that the challenges in explaining NGS, particularly when mediated through translation, could be contributing to this observed discrepancy in NGS testing rates within the Hispanic/Latino populations. We evaluated the associations between limited English proficiency and NGS testing by comparing the understandability between English, certified Spanish translations, and artificial intelligence (AI)-generated Spanish translations of provider explanations of NGS testing. Methods: We analyzed 191 patients with recurrent MBC from the Dallas Metastatic Breast Cancer Study. Ethnicity was categorized as Hispanic vs non-Hispanic. NGS completion between 2014–2022 was assessed. Three versions of NGS explanations (original English, certified Spanish translation, and AI-generated Spanish translation) were evaluated using the Flesch-Kincaid Readability Score (English) and the Fernández-Huerta Index (Spanish) using 10 provider-patient interactions. Scores range from 0–100, with higher scores indicating greater understandability; CDC/NIH-recommended patient materials target scores of 60–70. Readability discordance &lt; 10 points was considered acceptable. Results: 81.7% (156) of patients were proficient in English and were 3.2 times more likely to complete NGS testing than Spanish speakers (95% CI 1.35-7.73, p &lt; 0.01). Mean readability score between the English, certified Spanish, and AI Spanish were 47.9, 51.4, and 54.1, respectively. The mean discordance was 6.1 points between English and certified Spanish, 5.9 points between certified Spanish and AI Spanish, and 8.6 points between English and AI Spanish. All ten original English explanations fell below CDC/NIH readability targets, and only 4 of 10 certified Spanish translations met the recommended standards. When refining the prompts to maximize patient understanding, AI-generated text generated perfect understandability scores. Conclusions: Although translation fidelity was preserved, most provider-generated explanations failed to meet recommended readability standards, suggesting that communication quality, not translation alone, may limit patient understanding. AI-generated explanations may serve as a tool to mitigate language-related barriers when certified interpreters or optimized explanations are unavailable and enhance informed decision-making across varied clinical discussions, not limited to NGS testing.