Browse Articles
Discover research articles across all indexed journals
Implementing targeted thromboprophylaxis as a supportive care model: Mapping implementation preferences across the cancer care pathway.
e13552 Background: Risk-directed supportive care models aim to target interventions to patients most likely to benefit, yet implementation in routine oncology practice is often limited by workflow complexity, unclear role delineation, and uncertainty in patient selection. Cancer-associated thromboembolism (TE) exemplifies this challenge, with guideline-recommended thromboprophylaxis for high-risk ambulatory patients inconsistently delivered. The Australian phase III TARGET-TP trial (ACTRN12618000811202) demonstrated that biomarker-directed thromboprophylaxis reduced TE and mortality. This study mapped implementation preferences and identified barriers and enablers to inform scalable risk-directed care. Methods: This qualitative implementation study used semi-structured focus groups with patients/carers (n = 14) and healthcare professionals (HCPs; n = 20) across six groups. Process mapping elicited preferred and alternative approaches to delivering thromboprophylaxis across the care pathway. Implementation preferences were specified using the AACTT framework (Action, Actor, Context, Target, Time) and compared between consumers and HCPs. Deductive analysis applied AACTT and the Theoretical Domains Framework (TDF). Results: Six core implementation actions (ordering, risk assessment, documentation, prescribing, supply, and education) were discussed with 33 participants. HCPs favoured oncologist-led risk assessment embedded within tumour-specific electronic order sets with active prompts, supported by allied health stewardship to oversee appropriateness, documentation, and continuity. Junior doctors were positioned as implementers when plans were pre-documented, reducing cognitive and workflow burden. Patients and carers expressed heterogeneous preferences for consent and communication, ranging from passive integration of risk testing into routine blood collection with notification only if action was required, to preference for upfront discussion emphasising purpose and a clear management plan. TDF mapping showed HCP barriers clustered in early workflow stages, driven by Memory, Attention and Decision Processes, Environmental Context and Resources, and Social/Professional Role and Identity. Patient barriers were fewer and centred on education and consent, primarily within Knowledge, Beliefs about Consequences, and Emotion. Across groups, treatment modality (injectable vs oral) influenced acceptability. Conclusions: AACTT and TDF identified role-specific, workflow-sensitive strategies to support delivery of risk-directed supportive care in routine oncology practice. These findings highlight transferable implementation principles aligning digital systems, workforce roles, and patient preferences, with potential to reduce missed opportunities for evidence-based thromboprophylaxis.
Comparative analysis of depression, anxiety, suicidality, substance use, and palliative care use in pain-afflicted multiple myeloma patients across White, Black, and Hispanic populations.
e19501 Background: Bone pain affects 93% of multiple myeloma (MM) patients, significantly impairing quality of life and triggering psychological distress, potentially leading to depression, anxiety, suicidal ideation, and substance use disorders. We explored these outcomes across racial groups. Methods: White, Black, and Hispanic patients with MM and neoplastic pain were identified from the National Inpatient Sample (2016-2022). We compared demographics and evaluated odds of depression, anxiety, suicidality, substance use disorders, and palliative care utilization across racial groups using multivariable regression with White patients as reference. STATA 18.0 was used, accounting for NIS survey design. Results: We evaluated 49,400 MM patients with pain: 32,835 (66.47%) White, 11,585 (23.45%) Black, and 4,980 (10.08%) Hispanic. White patients were the oldest (mean 65.97 years) compared to Black (62.41 years) and Hispanic patients (62.00 years, p<0.001). White (58.25%) and Hispanic (55.08%) patients were predominantly male, while Black patients were predominantly female (46.8% male).Black patients had lower odds of depression (aOR 0.59, 95% CI 0.50–0.70, p<0.001) and anxiety (aOR 0.47, 95% CI 0.41–0.54, p<0.001) versus White patients. Hispanic patients showed similar patterns for depression (aOR 0.62, 95% CI 0.49–0.79, p<0.001) and anxiety (aOR 0.67, 95% CI 0.56–0.80, p<0.001). Suicidality did not differ significantly (Black: aOR 0.96, 95% CI 0.43–2.14, p=0.914; Hispanic: aOR 0.20, 95% CI 0.02–1.54, p=0.121). Black patients had higher odds of cocaine use (aOR 6.61, 95% CI 3.00–14.57, p<0.001) and cannabis use (aOR 2.12, 95% CI 1.41–3.18, p<0.001), while Hispanic patients showed no differences (cocaine: aOR 0.42, 95% CI 0.05–3.59, p=0.431; cannabis: aOR 0.59, 95% CI 0.27–1.27, p=0.178). Opioid use (Black: aOR 0.94, 95% CI 0.72–1.24, p=0.679; Hispanic: aOR 1.05, 95% CI 0.73–1.51, p=0.800) and alcohol abuse (Black: aOR 1.12, 95% CI 0.74–1.68, p=0.603; Hispanic: aOR 0.75, 95% CI 0.40–1.43, p=0.389) were comparable. Palliative care use (Black: aOR 1.08, 95% CI 0.96–1.22, p=0.201; Hispanic: aOR 1.12, 95% CI 0.92–1.35, p=0.255) and mortality (Black: aOR 0.93, 95% CI 0.73–1.19, p=0.569; Hispanic: aOR 1.02, 95% CI 0.74–1.40, p=0.923) were similar across groups. Conclusions: Our study found significant racial disparities in mental health and substance use patterns among MM bone pain patients. Despite these differences, palliative care utilization and in-hospital mortality were comparable among racial groups. These findings emphasize the importance of culturally sensitive screening for mental health disorders among MM patients with bone pain and call for future research to assess targeted interventions to enhance care for patients across all ethnic backgrounds.
Long-term follow-up of satricabtagene autoleucel (satri-cel) as sequential therapy after first-line treatment for advanced gastric cancer: A subgroup analysis.
2557 Background: Claudin18.2 (CLDN18.2) has emerged as a new target for the treatment of gastric cancer in the first-line (1L) setting with the approval of zolbetuximab. This long-term analysis reports the extended efficacy and safety of satri-cel (autologous CLDN18.2-specific CAR T cells) as sequential therapy after 1L treatment in patients with advanced gastric/gastroesophageal junction (G/GEJ) cancer, after the results of all cohorts published in 2024 (Qi C, et al. Nat Med. 2024;30(8):2224-2234. NCT03874897). Methods: This trial is an open-label, multi-cohort, phase 1 trial, which evaluated the safety and efficacy of satri-cel in patients with CLDN18.2-positive advanced gastrointestinal cancers. Cohort 3 in dose-expansion stage enrolled patients with advance G/GEJ cancer and were given satri-cel as sequential treatment after 1L therapy. The primary endpoint was safety; secondary endpoints included efficacy, pharmacokinetics and immunogenicity. Results: As of October 18, 2025, 5 patients with CLDN18.2-positive G/GEJ cancer received satri-cel infusion(s) of 250×106 cells, sequentially after 1L therapy. Each patient received a total of one (n=1), two (n=1), and three doses (n=3) of satri-cel. Patients had received a median of 5 cycles (range, 4-11) of 1L chemotherapy before satri-cel infusion, with 1 (20%) treated with PD-1 inhibitor. Notably, only 1 patient achieved PR after first-line therapy. Three patients (60%) were Lauren diffuse type and 1 (20%) with mixed type, 4 (80%) had signet ring cell carcinoma, and 4 (80%) had peritoneal metastases. Median follow-up from initial 1L therapy was 54.6 months (reverse KM, 95% CI: 51.1, NE). Among 4 patients with target lesions, confirmatory objective response rate was 100%, and median duration of response was not reached. One has maintained SD for 20.9 months and 2 received surgical resection after satri-cel therapy. Median progression-free survival and median overall survival since 1L therapy was 20.9 months (95% CI: 10.8, NE) and 22.1 months (95% CI: 10.8, NE), respectively. Two were still alive as of cutoff date, with a follow-up of 58.1 months and 51.1 months. Safety was manageable. No grade 3 or higher cytokine release syndrome (grade 1: n=1, 20%; grade 2: n=4, 80%), any grade immune effector cell-associated neurotoxicity syndrome, or treatment-related deaths occurred. Despite common hematologic toxicities, no severe infections (grade ≥3) or febrile neutropenia were reported. Conclusions: With an extended follow-up exceeding 4.5 years, satri-cel as first-line sequential treatment continues to demonstrate durable survival benefit with a manageable safety profile in patients with advanced G/GEJ cancer, supporting its highly promising potential in earlier lines of therapy. Clinical trial information: NCT03874897 .
Descriptive comparison of primary debulking surgery (PDS-ACT) versus neoadjuvant chemotherapy with interval debulking (NACT-IDS) in stage III–IV epithelial ovarian cancer: Single-center experience.
e17598 Background: Optimal cytoreduction remains a cornerstone of management for advanced epithelial ovarian cancer (EOC). We compared tumor burden, operative complexity, perioperative morbidity, and cytoreductive outcomes between primary debulking surgery followed by adjuvant chemotherapy (PDS-ACT) and neoadjuvant chemotherapy followed by interval debulking surgery (NACT-IDS) in a retrospective single-center cohort. Methods: This is a study conducted at Al-Bairouni University Hospital—Syria’s national cancer center serving 65–70% of the country’s cancer patients. Institutional ethical approval was obtained to perform the study. We conducted a retrospective cohort chart review of patients with stage III–IV EOC treated with PDS-ACT or NACT-IDS. Data were abstracted from Al-Bairouni medical records and operative documentation and harmonized for analysis. Results: Baseline tumor spread markers were more frequent in the PDS group (large ascites >500 mL 24.4% vs 6.9%; multiple mesenteric nodules 61.0% vs 22.8%; multisectional intestinal infiltration ≥3 segments 65.9% vs 24.1%; multiple liver metastases 31.7% vs 15.2%; fused enlarged lymph nodes 39.0% vs 8.3% for PDS vs IDS). Operative PCI >10 occurred in 61.0% of PDS vs 37.9% of IDS. IDS involved more multivisceral procedures (rectal resection 24.1% vs 4.9%; any peritonectomy 31.0% vs 12.2%). Open-and-close occurred in 9.8% (PDS) vs 5.5% (IDS). Early postoperative morbidity was higher after PDS (long hospitalization 24.4% vs 10.3%; acute renal failure 12.2% vs 1.4%). Completeness of cytoreduction favored IDS: complete/optimal cytoreduction 75.9% (110/145) vs 39.0% (16/41) in PDS. Conclusions: In this retrospective single-center cohort, patients undergoing PDS had higher baseline tumor burden and higher operative PCI, whereas NACT-IDS achieved higher complete/optimal cytoreduction despite more frequent multivisceral procedures. Prospective studies including survival outcomes are needed to contextualize these findings. Key outcomes in PDS-ACT vs NACT-IDS (stage III–IV EOC). Variable PDS-ACT (n=41) NACT-IDS (n=145) Large ascites >500 mL 10/41 (24.4%) 10/145 (6.9%) Multiple mesenteric nodules 25/41 (61.0%) 33/145 (22.8%) Intestinal infiltration ≥3 segments 27/41 (65.9%) 35/145 (24.1%) Operative PCI >10 25/41 (61.0%) 55/145 (37.9%) Rectal resection 2/41 (4.9%) 35/145 (24.1%) Any peritonectomy 5/41 (12.2%) 45/145 (31.0%) Long hospitalization 10/41 (24.4%) 15/145 (10.3%) Acute renal failure 5/41 (12.2%) 2/145 (1.4%) Complete/optimal cytoreduction 16/41 (39.0%) 110/145 (75.9%) Values are n/N (%). PDS-ACT: primary debulking surgery + adjuvant chemotherapy; NACT-IDS: neoadjuvant chemotherapy + interval debulking surgery.
PROPEL: Capturing the lived experience of lung cancer for patients and caregivers.
e23255 Background: A limited understanding of the lived experience of a lung cancer diagnosis hinders the development of a resource framework that effectively supports people with the diagnosis (patients) and their support systems (caregivers). Addressing this knowledge gap is essential to identifying ways to enhance quality of life in both within the healthcare setting and in everyday life. Methods: PROPEL (Patient Reported Outcomes of People Experiencing Lung Cancer) is a real-world, non-interventional study designed examining the lived experience of patients and caregivers. Participants self-enrolled and completed tailored series of validated patient-reported outcome measures (PROMs) capturing the multidimensional impacts of a lung cancer diagnosis on health, wellbeing, and functioning. To assess health-related quality of life, patients completed the EORTC QLQ-C30 and selected PROMIS scales, while caregivers completed the Caregiver Roles and Responsibilities (CRRS) questionnaire. Results: 57 patients have enrolled; most were female (63%), aged 55-75 years (61%), reported Stage IV disease (70%), and were undergoing treatment (84%). 21% identified as a visible minority. The EORTC QLQ-C30 Global Health Scale (scale range 0-100) indicated moderate quality of life ( = 62.7; σ: 20.1), low symptom burden ( < 40), and moderate-high range functioning subscales. Role Functioning (daily activities) was highest ( = 81.5; σ: 23.7) and Emotional Functioning lowest ( = 67.0; σ: 24.4). PROMIS Psychological Impact of Illness Scales reinforced this finding, where compared to before their diagnosis, participants reported increased feelings of disconnection (30% vs 2%) and worry about the future (43% vs 12%), and decreased belief that their life had meaning (68% vs 82%). 52 caregivers have enrolled; most were female (86%) and provided care to a spouse/partner (52%). Most cared for someone with Stage IV disease (90%), with limited ability for self-care (49% ECOG ≥ 2), in active treatment (76%), and provided a median 20 hours/week of direct care. 58% of caregivers with employment were on leave. CCRS (scale range 0-152) indicated moderate levels of functioning among caregivers ( = 89.1; σ: 22.3), with largest deficits in functioning seen in Self-Care (scale range 0-24; = 12.9; σ: 5.2) and Emotional (scale range: 0-36; = 18.4; σ: 7.6) subscales. Common themes included taking on additional responsibilities (39%), not prioritizing self-care (78%), feeling a lack of recognition from cancer care teams (78%), that the future is on hold (44%), and overwhelmed (39%). All PROMs demonstrated strong internal consistency (Cronbach’s α > 0.75). Conclusions: PROPEL is the first Canadian study to collect comprehensive PROMs data on lung cancer, capturing the experiences of patients and caregivers and highlighting significant burdens. This builds a knowledge base to guide the development of support services for comprehensive wrap-around care.
Representing people's experience of cancer and its treatment (ResPECT) pilot study: A global collaborative study to improve cancer care and outcomes as part of the Lancet Commission on Cancer and Health Systems.
1592 Background: Measurement of cancer burden increasingly informs survivorship research, health system design, and policy. Existing approaches rely on structured patient-reported outcome measures, which are costly to administer, culturally constrained, and difficult to scale globally. Advances in artificial intelligence (AI), particularly large language models (LLMs), offer the opportunity to systematically analyze unstructured patient narratives at scale. The ResPECT pilot study evaluates whether an LLM-driven framework can assess multidimensional cancer burden from patient-reported free text. Methods: Adult cancer patients and survivors provided structured burden ratings (Likert scale, 1–10) and open-text narratives describing physical, psychological, social, financial, and other impacts of cancer and its treatment. Unstructured text was analyzed using a retrieval augmented generation (RAG) architecture combining dense semantic embeddings (Nomic Embed), an open-weight LLM (Mistral 7B), and an ensemble retriever integrating vector similarity search and BM25 retrieval. Performance was evaluated across three predefined domains: (1) analytical concordance with quantitative analyses across 23 hypotheses as well as (2) contextual correctness and (3) hallucination rate over 50 questions, assessed by two independent reviewers. Results: A total of 1,642 participants from 23 countries were analyzed; 66.9% identified as female and 22.0% identified themselves to be part of an underrepresented group. Overall, impact on psychological well-being was ranked significantly higher compared to physical, social, or financial well-being (all p≤0.001). Younger participants and individuals identifying with underrepresented groups reported significantly higher overall and domain-specific burden, while U.S. participants reported greater financial burden than UK participants (p≤0.001). The LLM-based analyses reproduced quantitative findings in 21 of 23 hypotheses (91.3% concordance). Contextual correctness was 90% (90 of 100), and hallucinations occurred in 3% (3 of 100) of generated responses, predominantly involving minor paraphrasing of patient quotations. The LLM consistently identified key themes, particularly related to psychological well-being, including emotional stress, depression, relationship strain, isolation, guilt, and hope. Conclusions: The ResPECT initiative demonstrates that an LLM-driven framework can reliably provide and effectively scale qualitative insights into cancer burden from a patient perspective. While we acknowledge that the current pilot is not representative of cancer patients worldwide, the ResPECT approach may be scalable to derive comparable and actionable insights to inform investment in cancer services around the world.
Optimizing biliary drainage in malignant biliary obstruction: A systematic review and meta-analysis of transpapillary versus suprapapillary stent positioning.
4155 Background: Malignant biliary obstruction causes significant morbidity and poor survival with 5-year survival rates below 10%. Effective biliary drainage is essential to relieve cholestasis, prevent infection, and improve quality of life. Endoscopic stenting is a standard intervention, but the optimal stent position remains unclear. Placement of a stent across the duodenal papilla can lead to duodeno-biliary reflux, subsequent stent dysfunction, while suprapapillary placement may reduce reflux and improve stent patency. We performed a systematic review and meta-analysis to compare the clinical outcomes of trans-papillary versus suprapapillary biliary stent placement in patients with malignant biliary obstruction. Methods: We conducted a comprehensive search of major electronic databases to identify studies comparing suprapapillary and transpapillary biliary stent placement. All relevant studies published up to November 2025 were considered. The clinical outcomes assessed included stent occlusion, stent patency, post-ERCP pancreatitis, cholangitis, and overall adverse events. Meta-analysis was performed, and the choice of a fixed-effects or random-effects model was determined based on the presence of heterogeneity across studies. Results: A total of 14 studies, including randomized controlled trials and retrospective cohort studies, encompassing 1,169 patients, compared suprapapillary with transpapillary biliary stent placement. Suprapapillary stenting was associated with a significantly lower risk of post-ERCP pancreatitis (OR 0.23, 95% CI 0.11–0.49; I² = 0%). Stent occlusion demonstrated a non-significant trend favoring suprapapillary placement (OR 0.68, 95% CI 0.46–1.01; I² = 46%). Stent patency duration was longer with transpapillary stents (SMD 0.66, 95% CI 0.27–1.05). Rates of cholangitis were similar between groups (OR 0.86, 95% CI 0.54–1.38). Overall, adverse events were significantly lower with suprapapillary stenting (OR 0.67, 95% CI 0.47–0.94; I² = 25%). Conclusions: Suprapapillary biliary stent placement appears to be safer than transpapillary stenting, with lower rates of post-ERCP pancreatitis and overall adverse events while still providing effective biliary drainage. However, its superiority over transpapillary placement in terms of stent occlusion and stent patency remains uncertain. These findings highlight the potential advantages of suprapapillary stenting but underscore the need for further large-scale randomized controlled trials before it can be established as the standard of care.
Automated abstraction of colorectal cancer registry data using AI: Accuracy and implementation insights.
e15516 Background: Cancer registries support surveillance, quality measurement, and research, yet abstraction of high-value clinical variables remains heavily manual. Human registrars must synthesize information across structured fields and unstructured documentation, including pathology, operative reports, imaging, and oncology notes, resulting in time-intensive workflows that are difficult to standardize and scale. Automated chart abstraction of registry variables directly from the electronic medical record (EMR), particularly from unstructured text, could improve efficiency, timeliness, and consistency of registry data capture and reporting. Methods: We evaluated Synapsis AI, an AI-based abstraction system developed by Dyania Health, Inc. to automate extraction of colorectal cancer (CRC) registry variables from the EMR. Adult patients (18–90 years) treated at Cleveland Clinic with the ICD-10 codes C18-C20 first recorded during Q1 2024 were included, capped at 200 cases. The system was instructed to abstract 33 registry fields corresponding to 60 questions per patient, including mainly unstructured data, such as staging, lymph node metrics, biomarkers, treatments, and surgical details, as defined in STORE and SSDI guidelines. AI outputs were compared with registrar-abstracted tumor registry data. Results: Overall accuracy across all fields was 94.78%. Question-level accuracy reached 100% for lymph node yield and positivity, pathologic and post-therapy TNM stage, mismatch repair and microsatellite instability status, immunotherapy regimens, and surgical procedures. The lowest accuracy (28%) occurred for residual colorectal tumor presence. This was primarily due to ambiguity in field definition rather than model error: when no relevant documentation was identified, the system returned “NA,” reflecting a weak negative inference, whereas registry standards required a “No” response. Additional contributors to inaccuracies included incomplete access to imaging reports and insufficient upfront alignment on field definitions and response conventions. Accuracy exceeded 72% across all remaining fields. Conclusions: These findings demonstrate that AI can accurately abstract CRC registry data at scale. Importantly, performance limitations were largely attributable to scoping of fields and access to relevant data, rather than model capability. Proper upfront scoping to align definitions, explicit communication of registrar-developed abstraction conventions, and health-system defined prioritization of sources of truth are critical to achieving registrar-aligned automation. When these elements are addressed, AI systems can function as reliable, scalable extensions of cancer registry workflows.
Adjuvant and neoadjuvant evaluation of a structural variant-based MRD assay in early breast cancer: A real-world cohort.
e12607 Background: Despite curative intent surgery and standard adjuvant therapy, patients (pts) with ER+, node+ stage II–III breast cancer (BC) remain at risk for systemic recurrence. Conventional surveillance uses imaging and clinical assessment, which may detect relapse only after overt disease. Tumor-informed circulating tumor DNA (ctDNA) assays enable detection of molecular residual disease (MRD), providing a molecular surveillance approach that may identify residual disease not captured by standard imaging. Pathlight is an ultrasensitive structural variant (SV)-based, tumor-informed ctDNA assay designed to detect low-level MRD from plasma. This study describes the real-world performance, feasibility, and clinical context of Pathlight MRD testing in routine BC care. Methods: An observational real-world cohort study of up to 100 pts from a single institution between May 1 and Dec 31 2025, through retrospective data analysis (no clinical interventions were made). Adult pts (18+) with histologically confirmed ER+ BC, stage II or III with node+ disease, treated with curative intent surgery with/without adjuvant therapy, who underwent Pathlight MRD testing through routine care were included. Pts with metastatic disease or HER2+ at time of testing were excluded. Personalized ctDNA assays were generated using up to 16 SVs from tumor tissue. Descriptive analysis evaluated MRD detection rates, assay feasibility, and clinicopathologic characteristics. Results: Pathlight MRD testing was performed in 69 unique pts, and 150 total monitoring tests. Detectable MRD was observed in 19% (13/69) pts. Median age was 55 yrs (range 33-80), with 61% (42/69) and 28% (19/69) at stages II and III respectively. 62/69 tests were initiated for adjuvant surveillance monitoring. Median turnaround time for initial SV validation was 20 days (range 6-41) and 3 days for ctDNA monitoring tests (range 1-7). Pts had median 13 SVs tracked (range 6-16). Detectable MRD was observed across all stages; I (8%, 1/13), II (54%, 7/13) and III (38%, 5/13). Tests were initiated for neoadjuvant (54%, 7/13) and adjuvant (46%, 6/13) settings. 92% (12/13) were ER+. All pts with proliferation data available were > 20% Ki-67 (11/11). Tumor fingerprints were generated from both biopsy specimens (61%, 8/13) and surgical specimens (38%, 5/13). Ultrasensitive detection was required in most cases: 77% (10/13) detected tumor fractions < 0.01% (range 0.000013%–0.025%). Conclusions: MRD detection using Pathlight was feasible and operationally efficient. MRD positivity occurred at ultra-low ctDNA levels (most below 0.01% VAF), underscoring the analytical sensitivity required for monitoring. MRD+ cases were enriched for higher-risk clinicopathologic features (stage II–III disease, elevated Ki-67). These real-world findings support further prospective evaluation of MRD surveillance in early BC.
A phase 1, first-in-human, open-label, dose-escalation and -expansion study of gamitrinib, a first-in-class, mitochondria-directed inhibitor of the molecular chaperone heat shock protein-90 (Hsp90) in patients with advanced cancer.
TPS3180 Background: Mitochondria are subcellular organelles whose signaling and bioenergetics functions are exploited in cancer. This requires tight control of intra-organelle protein folding, which is made possible by the accumulation of molecular chaperones of the Heat Shock Protein-90 (Hsp90) family in mitochondria of tumors, but not normal tissues. This provides a unique therapeutic opportunity to disable multiple tumor pathways at once, but previously developed Hsp90 inhibitors do not accumulate in mitochondria, and systemic inhibition of chaperone activity is toxic in humans. Gamitrinib is a first-in-class anticancer agent specifically designed to disrupt protein folding in mitochondria, combining the Hsp90 inhibitory backbone of 17-allylamino-geldanamycin (17-AAG) to the mitochondria-targeting moiety, triphenylphosphonium. Gamitrinib selectively accumulates in mitochondria, does not affect the function of cytosolic Hsp90, and irreversibly disrupts organelle protein folding with activation of multiple cell death pathways. In preclinical studies, Gamitrinib shows potent, cytotoxic activity against multiple cancer cell types (IC 50 0.16–29 µM), inhibits primary and metastatic tumor growth in xenograft and genetic mouse models, has slower clearance (85.6 ± 5.8 mL/min/kg) and longer half-life (12.2 ± 1.55 h) than unconjugated 17-AAG, and is well-tolerated with no alterations in clinical-chemistry parameters, organ function, or tissue histology at dose levels of up to 5 (rats)- and 12 (dogs)-fold higher than therapeutically effective doses in mice (Cancer Biol Ther 23:117). Methods: Based on preclinical activity, favorable safety, and unique mechanism of anticancer activity, a first-in-human, open-label, dose-escalation phase 1 trial of Gamitrinib in sequential cohorts of adult patients with advanced solid tumors or lymphoma, for whom no standard therapy is available, was initiated (NCT04827810). Per study protocol, Gamitrinib is administered as a 1-h weekly intravenous (IV) injection on a 28 day-cycle with a 3+3 dose-escalation design and dosing continuing until disease progression, unacceptable toxicity, or patient discontinuation for any other reason. Trial objectives include determination of MTD and/or RP2D, overall safety, PK profile, correlative studies of target engagement in paired pre- and post-treatment biopsies in a 6-patient expansion cohort at MTD and documentation of anti-tumor activity by RECIST 1.1 and RECIL 2017 criteria. As of 2026, 18 evaluable patients have been accrued over 5 escalating Gamitrinib dose levels (10-85 mg) with no DLTs as assessed by NCI CTCAE v5.0 grading. 336 plasma samples have been collected for PK analysis. The study is currently open and enrollment is ongoing. Clinical trial information: NCT04827810 .
Quizartinib in combination with decitabine and venetoclax for newly diagnosed and relapsed/refractory <i>FLT3</i> -mutated AML.
6504 Background: Hypomethylating agents (HMAs; azacitidine/decitabine) plus venetoclax (VEN) are standard for newly diagnosed acute myeloid leukemia (AML) unfit for intensive chemotherapy; however FMS-like tyrosine kinase-3 internal tandem duplication ( FLT3-ITD ) confers resistance and poor overall survival (OS). Addition of quizartinib (QUIZ), a potent FLT3 inhibitor, may improve outcomes. Methods: Phase I/II study of QUIZ+decitabine+VEN in two cohorts: newly diagnosed FLT3-ITD AML pts unfit for intensive chemotherapy and relapsed/refractory (R/R) FLT3-ITD AML pts. Primary endpoints were maximum tolerated dose (MTD) and overall response rate (ORR). Results: Eighty-eight patients (pts) were enrolled (frontline N=42; R/R N=46). QUIZ 26.5mg/day was selected as the recommended phase II dose (RP2D). In the frontline cohort, median age was 70y (62–85); 15 pts (36%) were ≥75y; 25 (60%), 12 (N=28%), and 5 (12%) had de novo, secondary, and therapy-related AML, respectively. DNMT3A (43%), NPM1 (33%), and RUNX1 (31%) were the most common co-mutations. Complete remission (CR), CR with incomplete count recovery (CRi), morphologic leukemia-free state (MLFS), end-of-cycle-1 (EOC1) measurable residual disease (MRD) negativity by multicolor flow cytometry (MFC; 0.01%), and next-generation sequencing (NGS) FLT3-ITD negativity (5×10⁻⁵) were 72% (N=29), 17% (N=7), 2% (N=1), 58% (19/32), and 27% (5/18), respectively; 2 pts in cycle 1 were not evaluable. Best MRD negativity by MFC and FLT3-NGS were 68% (23/34) and 83% (15/18). Median time to absolute neutrophil count (ANC) >500, ANC >1000, and platelets >50K were 40 (19–72), 41 (14–72), and 35 (16–71) days, respectively. The median relapse-free survival (RFS) and OS were 24.7 and 36.5mo. Fifteen pts (36%) proceeded to allogeneic stem cell transplant (ASCT) in CR1, with OS not reached vs 36.6mo without ASCT (p=0.65, landmark analysis). Median 3 cycles (1–39) were delivered; 11 pts remain on study. Thirty-one pts discontinued protocol due to ASCT (N=15), relapse (N=9), physician/pt choice (N=3), induction death (N=1), death in CR (N=1), death with disease (N=1), or hospice (N=1). Grade ≥3 non-hematologic adverse events (>5%) included febrile neutropenia (37%), pneumonia (33%), infections (17%), pain (10%), ALT increase (10%), fracture (10%), sepsis (9%), hypertension (7%), hypotension (7%), hyponatremia (7%), gait disturbance (7%), oral mucositis (7%), and pleural effusion (7%). In the R/R cohort, CR/CRi was 28% (N=13) with 33% (N=15) MLFS; 37% (N=17) proceeded to ASCT; median OS was 6.3mo (further data will be provided at the time of presentation). Conclusions: QUIZ+decitabine+VEN combination resulted in high CR/CRi, deep MRD responses, and encouraging survival in older pts with newly diagnosed FLT3-ITD AML. Clinical trial information: NCT03661307 .
Spatial transcriptomic profiling to identify stroma-based prognostic groups in oncogene-addicted lung cancer.
8611 Background: Driver genomic aberrations determine prognosis and therapy in non-small cell lung cancer (NSCLC). In this context, some other genomic features such as TP53 mutations confer poor outcomes. However, other mechanisms leading to worse prognosis in patients (pts) sharing the same genomic profile remain incompletely characterized. Methods: NSCLC pts from three cohorts based on their molecular profile ( EGFR -mutated [EGFRm], KRAS -mutated [KRASm] and ALK/ROS1/RET fusions) were included. Baseline formalin-fixed paraffin-embedded (FFPE) tumor samples were analyzed using NanoString GeoMx Digital Spatial Profiling. Regions of interest (ROIs) were selected based on histopathological features and fluorescently labeled antibodies for tumor (Tm) and stromal (St) areas. RNA expression of >1,800 genes from selected ROIs was assessed using GeoMx Cancer Transcriptome Atlas. Leiden algorithm was used for clustering, and differential gene expression and other statistical analyses were performed using R software. Results: A total of 189 pts (100 EGFRm, 56 KRASm, 33 fusions [17 ALK , 7 ROS1 and 9 RET rearrangements]) were identified. Separate analysis of Tm and St compartments revealed an association between genotype and transcriptome in the first, whereas the St transcriptome failed to correlate with the molecular profile. Unsupervised clustering of the Tm compartment identified four subgroups with distinct gene expression. Tm Cluster 4 (30 pts) was mainly composed of KRASm pts (86.7%) and presented the poorest prognosis, with a median overall survival (mOS) of 7.5 months (mo) ( p <0.0001). Within St compartment, four prognostic clusters were also identified. St Cluster 1 (76 pts: 43.4% EGFRm, 40.8% KRASm, 15.8% fusions) and St Cluster 4 (23 pts: 21.7% EGFRm, 52.2% KRASm, 26.1% fusions) included pts from all three cohorts and showed significantly decreased OS (17.4 mo and 15.6 mo, respectively; p <0.0001). Genes involved in matrix remodeling and epithelial-mesenchymal transition signaling were overexpressed in St Cluster 1, while St Cluster 4 exhibited a highly immunogenic profile; no correlation with TP53 mutation status was observed. Conclusions: Spatially resolved transcriptomic profiling suggests that stromal composition could identify oncogene-addicted NSCLC pts with poor prognosis regardless of molecular subtype, potentially guiding the development of novel therapeutic strategies. Further validation studies are warranted.
Assessing treatment patterns and outcomes among patients (pts) with metastatic renal cell carcinoma (mRCC): A British Columbia (BC) population-based analysis.
e16544 Background: Current first-line treatments for mRCC include immune check point inhibitors (IO) and tyrosine kinase inhibitors (TKIs). There is limited data on attrition rate between lines of therapy and clinical outcomes in relation to the institutional volume and/or hospital setting. Objective: To analyze treatment patterns, attrition rate and clinical outcomes in mRCC pts following first-line therapy in high vs. lower volume or community setting within BC. BC Cancer currently operates six regional cancer centres providing care for pts across BC. Methods: BC Cancer’s centralized pharmacy data were screened to identify 547 mRCC pts treated with first-line IO combinations or TKIs from May 2019 to December 2024. Patient charts were reviewed for baseline characteristics, relevant treatment data, reasons for treatment choices, and outcomes. Treatments received in each line of therapy and attrition rate were tabulated using frequencies and percentages. Results: In the first 327 pts, median age was 65, with 79.2% male. 86 percent of pts received first-line IO combinations (193 IO-IO, 103 IO+TKI). 60 pts have not yet progressed on first line therapy. 76% of pts treated at the highest volume centre received second-line therapy, compared with 55% of pts treated at other regional centres and 48% of pts treated in the community. After first line IO-IO, 28% of pts treated in the highest volume centre had third-line therapy versus 25% of pts treated at other regional centres and only 6% in the community. Following first-line IO-TKI, 23% of pts from the highest volume centre had third-line therapy versus only 14% at the other regional centres and in the community. Due to the lack of funding for third-line therapy after IO/TKI start in BC, assessment of attrition rate to third-line therapy has to be interpreted with caution. The most common therapy in subsequent lines included TKI monotherapy. Conclusions: In mRCC pts treated with first-line IO combinations and TKIs, the most prevalent subsequent regimens are TKI-based therapies. The attrition rate with each additional line of therapy remains substantial. Lower attrition rates were observed among pts treated in high-volume specialized settings. These findings reflect the increasing complexity of RCC management and emphasizes the need to manage these pts in or in cooperation with specialized, high-volume centers within a multidisciplinary setting.
Gaps in access to chimeric antigen receptor T-cell (CAR-T) therapy post leukapheresis: Waiting time and post-leukapheresis treatment patterns in relapsed/refractory multiple myeloma (RRMM)—Real-world evidence from U.S. claims.
7530 Background: B-cell maturation antigen (BCMA)-directed CAR-T therapy is a highly effective treatment with potential for long-term remission of RRMM with improved outcomes vs daratumumab-based regimens. CAR-T suitability may be limited by delays/manufacturing failure and changes in pt condition, financial situation, or access to specialized treatment centers. We characterized pts with RRMM who underwent leukapheresis to prepare for CAR-T and pt treatment waiting time, to understand the delays incurred prior to CAR-T infusion. Methods: A real-world study in the US using the PharMetrics Plus claims database (July 2020–September 2024) was performed. Eligible pts were aged ≥18 years with RRMM, had their first leukapheresis claim (index date) during the index period (March 2021–August 2024), were continuously enrolled in medical/pharmacy benefits for ≥180 days prior to and ≥30 days after index date, and were not clinical trial participants. Pt characteristics, rate of CAR-T receipt, and time to CAR-T receipt were assessed during the variable follow-up. Results: Overall, 337 pts who received leukapheresis were identified. Median (interquartile range) age was 62 (55–67) years, and 59.1% of pts were male. Mean National Cancer Institute (NCI) comorbidity index was 0.52. Most pts received care in the Midwest (31.8%) or Northeast (24.6%) regions. Most pts (95.9%) received leukapheresis in an outpatient setting. Median available follow-up time was 255 days (mean 337 days). During follow-up, 183 pts (54.3%) did not receive CAR-T after leukapheresis. Among pts who had ≥12 months of follow-up (n=143), the proportion of pts who did not receive CAR-T was similar at 53.1% (n=76). Among pts who received CAR-T during the follow-up (n=154; 45.7%), the median waiting time was 54.0 days (mean 56.7 days). The cohort of pts who did not receive CAR-T had a higher proportion of females (48.6%/31.8%), were more likely to have Medicaid (29.5%/7.8%), had higher NCI comorbidity index (scores 0.55/0.47), and most commonly received care in the Northeast (35.5%/11.7%) while pts who received CAR-T most commonly received care in the Midwest (23.0%/42.2%). The most common first classes of multiple myeloma medications after leukapheresis but prior to CAR-T infusion in pts who did/did not receive CAR-T were an alkylating agent (65.6%/36.1%), proteasome inhibitor (40.3%/30.6%), steroid (38.3%/39.3%), immunomodulatory drug (22.7%/27.3%), anti-CD38 (20.8%/20.8%), and stem cell transplant (0%/17.5%). Conclusions: Over half of pts receiving leukapheresis did not receive CAR-T infusion, highlighting a persistent gap between eligibility and treatment delivery. These data emphasize an urgent need for broadly accessible BCMA-targeted therapies that provide deep and durable responses.
Identification of nutritional risk in chimeric antigen receptor T-cell (CAR-T) therapy: Protein-energy malnutrition and its impact on patient outcomes.
12158 Background: Protein-energy malnutrition (PEM), an imbalance between protein and energy intake and the body’s requirements, is highly prevalent in cancer, affecting 20–70% of patients, with an overall prevalence of approximately 41%. PEM and cachexia impair therapeutic efficacy, increase treatment-related toxicities, and worsen survival outcomes. Although CAR T-cell therapy has revolutionized oncology, its associated complications—such as cytokine release syndrome (CRS) and immune effector cell–associated neurotoxicity syndrome (ICANS) further affects nutritional status, creating a complex bidirectional relationship between PEM and CAR-T complications. This study utilizes the National Inpatient Sample (NIS) to assess nutritional status in CAR T-cell therapy recipients and potential impact on outcomes and treatment-related complications. Methods: A retrospective cohort study was conducted using the 2021–2023 National Inpatient Sample (NIS) database. Adult patients (aged ≥18 years) who were hospitalized and received CAR-T therapy were identified using ICD-10 codes and stratified according to the presence of PEM. The primary outcome was the inpatient mortality rate, while secondary outcomes included length of hospital stay, total hospital charges, incidence of CRS, ICANS, and other medical complications. Results: Among 10,690 hospitalized adult CAR-T recipients, 2,190 (20%) had PEM. In-hospital mortality in this group was 190 (9%). The PEM cohort was predominantly male (79%) and White (74%), with a mean age of 61.5 years. Compared to patients without PEM, those with PEM experienced higher in-hospital mortality (8% vs 1.2%; OR 7.2; P < 0.001), longer hospital stays (LHS) (24 vs 15 days; Coefficient 9.2; P < 0.001), and higher costs ($1,428,895 vs $1,106,642; Coefficient 322,253; P < 0.001) and more CAR-T-related complications, including CRS (44% vs 39%; OR 1.3; P < 0.05), ICANS (12% vs 7%; OR 1.7; P < 0.05), sepsis (OR 3.13; P < 0.001), stroke (OR 4; P < 0.001), and cardiovascular events (OR 1.6; P < 0.001). Conclusions: This analysis identifies PEM as a significant predictor of mortality, prolonged hospitalization, higher costs and increased complications among CAR-T recipients. This highlights baseline nutritional status as a critical outcome determinant. This emphasizes the need for pre-treatment nutritional screening and early, targeted interventions to break the vicious cycle of malnutrition, improve treatment tolerance, reduce adverse events, and optimize outcomes. Further studies should evaluate tailored nutritional strategies and supportive care to enhance CAR-T safety and efficacy.
Pan-cancer computational analysis of adaptive immune and vascular changes in primary versus metastatic tumors.
3124 Background: Metastasis drives cancer mortality, yet tumor microenvironment (TME) evolution during systemic spread remains poorly understood. Current paradigms often assume uniform immunosuppression at metastatic sites or seek a universal pro-metastatic signature. We investigated whether a conserved immune signature exists across metastatic sites or if remodeling follows tissue-specific trajectories that could inform precision therapeutic strategies. Methods: We performed computational meta-analysis from TCGA using xCell enrichment scores across >20,000 transcriptomic samples from primary tumors and distant metastases in five solid tumor types: Lung (n=1,622), Breast (n=1,543), Kidney (n=1,541), Brain (n=1,243), and Thyroid (n=1,016). Mann-Whitney U testing with Bonferroni correction identified cell populations with differential enrichment (padj<0.05). Cohorts were balanced for age, gender, and race (all p>0.05). Results: We identified two convergent metastatic trajectories with opposite directional changes in a four-cell plasticity module: B-cells, class-switched memory B-cells, CD4+ effector memory T-cells, and platelets. The Immune-Vascular Acquisition Trajectory (Breast, Brain, Thyroid) showed metastatic enrichment of all four cell types (Mean Difference [Primary-Metastatic]: -0.62; padj<0.001). The Immune-Vascular Depletion Trajectory (Lung, Kidney) showed significant depletion of all four populations at metastatic sites (Mean Difference [Primary-Metastatic]: +0.74; padj<0.001). Platelet dynamics consistently mirrored adaptive immune trajectories across all cancer types. Conclusions: These findings challenge two prevailing assumptions: uniform immunosuppression at metastatic sites and a single universal pro-metastatic immune signature. Instead, metastatic immune remodeling follows organ-specific convergent trajectories driven by adaptive and vascular plasticity. The identification of shared trajectories among biologically distinct cancers suggests that metastatic site microenvironment dictates TME evolution. The de novo acquisition of immune signatures in brain metastases challenges the "immunologically cold metastasis" paradigm, while depletion in naturally immunogenic tumors (Lung, Kidney) indicates TME "cooling" for immune evasion. The invariant coupling of platelet and adaptive immune dynamics highlights vascular remodeling as critical to metastatic niche establishment. These distinct trajectories may explain differential responses to immunotherapy and anti-angiogenic agents across metastatic sites, supporting the need for organ-specific therapeutic strategies. Future studies are needed to investigate whether primary tumors harbor early signatures of this module predictive of metastatic trajectory, enabling stratified surveillance and intervention.
Value of risk-directed thromboprophylaxis in ambulatory lung and gastrointestinal cancer care: Economic evaluation of the TARGET-TP randomized trial.
12137 Background: Thromboembolism (TE) is a major and preventable hematological complication in patients with cancer, contributing to substantial adverse health impacts for patients and the health system. Targeted Thromboprophylaxis in Ambulatory Patients Receiving Anticancer Therapies (TARGET-TP; ACTRN12618000811202) was an Australian phase III randomized trial of biomarker risk-directed enoxaparin thromboprophylaxis. The TARGET-TP intervention reduced TE (HR 0.31, p = 0.005) and six-month mortality (HR 0.48, p = 0.03) versus usual care (no prophylaxis). This study evaluated the value for money of the TARGET-TP intervention versus usual care. Methods: A within-trial and modelled economic evaluation was conducted from the Australian health system perspective. Resource use collected within the trial reflected risk stratification, thromboprophylaxis, side-effect management, and treatment of TE events. Health benefits were quantified as quality-adjusted life years (QALYs), utilities derived from SF-12v2 data. Bootstrapping was used to evaluate sampling uncertainty. A Markov model estimated long-term costs and outcomes over a 5-year horizon. Scenarios assessed the use of oral anticoagulants and stewardship, while probabilistic sensitivity analysis (PSA) tested parameter uncertainty. Results: Within-trial analysis showed the TARGET-TP intervention was associated with lower costs, while being more effective (i.e., dominant), reducing costs by $1,095 per patient and generating an additional 0.0097 QALYs, driven mainly by fewer hospitalizations. The TARGET-TP intervention was dominant in > 83% of bootstrap iterations. Modelled analyses supported these findings, with cost savings of $3,551 per patient and a gain of 0.4 QALYs. PSA demonstrated dominance in > 95% of iterations (Figure 1). The oral anticoagulant and stewardship scenarios remained dominant for rivaroxaban, while apixaban was highly cost-effective, driven by bleeding events and drug costs. Conclusions: Biomarker-driven primary thromboprophylaxis in patients with lung and gastrointestinal cancer was less costly and more effective from the healthcare system perspective, supporting its implementation in routine care. Clinical trial information: ACTRN12618000811202.
SPARK-Lung: A study of patterns and outcomes in young-onset lung cancer.
8070 Background: Young onset lung cancer patients diagnosed before age 50 appear to have distinct features compared to their older counterparts, including a higher proportion of never-smokers and increased frequency of targetable driver mutations. Yet, the molecular characteristics and survival outcomes of young-onset non-small cell lung cancer (NSCLC) remain poorly defined in U.S. populations. Improved understanding of the demographic and molecular profile of this distinct disease is needed to guide diagnostic testing strategies and personalize treatment. Methods: We conducted a retrospective cohort study using the US-based, electronic health record-derived deidentified Flatiron Health Research Database, which included adults diagnosed with NSCLC between 2013 and 2025 who underwent next-generation sequencing. Young-onset NSCLC was defined as age < 50 years and older-onset as ≥50 years. Demographics, stage at diagnosis, and prevalence of driver mutations and fusions were compared between age groups. Differences are reported using standardized mean differences (SMD) with SMD ≥0.2 indicating significant difference. Treatment patterns and overall survival (OS) were evaluated in metastatic patients adjusting for sex, histology, smoking status, performance status, insurance, and TP53 status. Results: Among 3792 young-onset and 120,161 older-onset patients with NSCLC across all stages, young-onset patients were more often female (54.2% vs 51.2%), Asian (5.3% vs 2.4%), Black (12.4% vs 8.1%), or Hispanic/Latino (8.5% vs 3.6%) (SMD 0.38). They were more often treated in academic settings (24.8% vs 16.0%; SMD 0.26) and more likely to have no tobacco history (36.6% vs 13.8%; SMD 0.65). Non-squamous histology predominated in younger patients (85.1% vs 73.1%; SMD 0.29) and actionable driver alterations were more common (47.8% vs 40.5%), particularly ALK (9.3% vs 1.2%; SMD 0.66) and EGFR (17.6% vs 11.0%; SMD 0.21). Young-onset patients were more likely to present with metastatic disease (69.8% vs 49.2%; SMD 0.43). After adjustment, treated young-onset patients received more lines of therapy than older patients (median no. 2 vs 1; incident rate ratio 1.16; 95% CI 1.12–1.20; p < 0.0001) and had longer median treatment duration (23.7 vs 14.3 mos, p < 0.0001). Among patients with metastatic disease at diagnosis, unadjusted median OS was longer in young-onset versus older-onset patients (2.0 vs 1.1 years), with higher 1-year (72.6% vs 57.2%) and 2-year (56.7% vs 41.8%) OS rates. Conclusions: Young-onset NSCLC represents a clinically distinct population in the United States, characterized by higher rates of advanced stage at diagnosis, a higher percentage of actionable molecular alterations and improved survival in the metastatic setting. Recognition of these differences may inform screening, testing, treatment, and survivorship strategies for younger adults with NSCLC.
GLP-1 therapy and hormone receptor–positive breast cancer risk and survival: A real-world analysis.
10548 Background: Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) are increasingly used for weight management and cardiometabolic risk (CMR) reduction, yet their association with breast cancer (BC) incidence and survival in non-diabetic women with elevated BMI remains unclear. We evaluated the relationship between GLP-1 RA exposure and incident BC, as well as overall survival (OS), in a large real-world (RW) cohort of women without laboratory-defined diabetes. Methods: We conducted a retrospective cohort study using the TriNetX US Collaborative Network including female patients (pts) with BMI 25–35 kg/m². Age <50 vs ≥50 years served as proxies for pre- and postmenopausal status, with a ≥50 sensitivity analysis to assess age/menopause confounding. Pts with HbA1c ≥6.5% or prior BC were excluded. Cohorts were propensity score matched (PSM) on age at index, BMI, metformin and statin exposure, BRCA mutation status, hormonal contraceptive use, parity, breastfeeding history, and smoking status. The primary endpoint was incident HR+/HER2− BC, summarized as incidence proportion and risk reduction estimates; secondary analyses evaluated HR+/HER2+, HR−/HER2+, and triple-negative breast cancer (TNBC). OS was evaluated using Kaplan–Meier and Cox regression. Results: After PSM, 148,709 pts were included (mean age 45.8 ± 9.54 years in both cohorts). Over a median 36-month follow-up, incident HR+/HER2− BC was lower with GLP-1 RA vs controls (0.29% vs 0.33%), with an absolute difference −0.04% (95% CI −0.081 to −0.001%; P=0.046) and RR 0.88 (95% CI 0.77–0.99). OS was improved with GLP-1 RA vs controls (HR 0.75, 95% CI 0.65–0.86; P<0.0001). No difference in BC incidence was found for HR+/HER2+ (RR 0.98, 95% CI 0.86–1.12) or HR− subtypes, including HR−/HER2+ BC (RR 0.83, 95% CI 0.55–1.23) and TNBC (RR 0.68, 95% CI 0.41–1.12). Findings for HR+/HER2− were consistent across menopausal strata (Table). In the postmenopausal cohort (n=91,577; mean age 63.3 ± 6.88 vs 62.4 ± 6.52 years), GLP-1 RA exposure was also associated with lower incidence of HR+/HER2− BC (RR 0.93, 95% CI 0.86–0.99) and improved OS (HR 0.87, 95% CI 0.81–0.94; P=0.0003). Conclusions: In a large non-diabetic cohort of women with BMI 25–35 kg/m² and no prior BC, GLP-1 RA exposure was associated with a modest reduction in incident HR+/HER2− BC and improved OS. Despite limited follow-up and age-based menopausal proxies, these findings provide RW evidence supporting prospective validation with standardized exposure definitions and longer follow-up. Future studies should clarify whether GLP-1 RAs confer direct antitumor effects or indirect benefits mediated through weight loss and CMR reduction and identify pts most likely to derive BC risk and survival benefit. GLP-1 RA and HR+/HER2− BC outcomes. RR (Breast Cancer Incidence) HR (Overall Survival) Age <50 0.88 (0.77–0.99) 0.75 (0.65–0.86) Age ≥50 0.93 (0.86–0.99) 0.87 (0.81–0.94)
Real-world risk of interstitial lung disease and pneumonitis with trastuzumab deruxtecan vs trastuzumab in HER2-positive metastatic breast cancer: A propensity score–matched analysis.
e13030 Background: HER2-targeted antibody–drug conjugates (ADCs), such as trastuzumab deruxtecan (T-DXd), have substantially improved clinical outcomes in patients with HER2-positive metastatic breast cancer. However, these clinical benefits must be balanced with notable safety concerns, especially pulmonary toxicities such as interstitial lung disease (ILD) and pneumonitis. This study aimed to evaluate and quantify the risk of ILD and pneumonitis associated with T-DXd in comparison to trastuzumab using real-world data. Methods: The data used in this retrospective observational study was collected on 22 January 2026 from the TriNetX Dataworks – USA network, which provided access to electronic medical records from more than 120 million patients across 69 healthcare organizations. Adult patients with malignant breast neoplasms treated with T-DXd or trastuzumab were included. Propensity score matching was applied to balance baseline characteristics. Patients with prior ILD or pneumonitis were excluded. Outcomes assessed within one year of treatment initiation included newly diagnosed ILD and pneumonitis. Risk estimates were calculated as risk differences, risk ratios (RR), and odds ratios (OR) with 95% confidence intervals (CI). Results: In a real-world cohort of HER2-positive metastatic breast cancer patients from the TriNetX Dataworks–USA network, 1,908 patients treated with trastuzumab deruxtecan (T-DXd) and 1,986 with trastuzumab were analyzed after propensity score matching. Within one year of treatment initiation, ILD incidence was significantly higher in the T-DXd group (5.7%) compared to trastuzumab (2.3%) [RR: 2.47, 95% CI: 1.76–3.46; OR: 2.56, 95% CI: 1.80–3.63; p < 0.001]. Pneumonitis incidence was also elevated with T-DXd (11.1% vs 6.8%) [RR: 1.63, 95% CI: 1.31–2.01; OR: 1.70, 95% CI: 1.35–2.15; p < 0.001]. Conclusions: In this large, propensity-matched real-world study, T-DXd is associated with a significantly increased real-world risk of ILD and pneumonitis compared to trastuzumab. Our findings underscore the need for heightened clinical vigilance when using T-DXd, including early recognition and management of respiratory symptoms. Further studies – ideally prospective cohorts or registry-based analyses – are recommended to refine risk stratification for T-DXd–associated pulmonary toxicity. Comparison of ILD & pneumonitis risk between T-DXd and trastuzumab treatment groups. Outcome T-DXd (n=1,908) Trastuzumab (n=1,986) Risk Difference (95% CI) RR (95% CI) OR (95% CI) p-value ILD 5.7% 2.3% 3.4% (2.2–4.6) 2.47 (1.76–3.46) 2.56 (1.80–3.63) <0.001 Pneumonitis 11.1% 6.8% 4.3% (2.4–6.1) 1.63 (1.31–2.01) 1.70 (1.35–2.15) <0.001