Browse Articles

Changes in metastatic breast cancer (MBC) bone lesions and SREs/fractures via a single intratumoral injection for a first-of-its-kind drug: Preliminary phase 2a results.

589 Background: Lytic bone metastases develop in ~70% of breast cancer patients that lead to pain, functional decline, and skeletal-related events (SREs) that reduce quality of life (QoL) and survival. Prophylactic denosumab/bisphosphonate treatment offers limited protection, and current local palliative interventions, EBRT and surgery, don’t reverse lytic destruction. Zeta-BC-003 is a first-in-class intratumoral (IT) injectable biomaterial infused with N-allyl noroxymorphone, a small molecule drug that acts through the p21 pathway to drive bone healing and local tumor control. Reports on two Compassionate Use patients (7 lesions, 2-yr follow-up; Palma et al, Pain Manag 2023) showed no SREs/fractures or tumor activity coupled with neo-trabecular bone regeneration and therapeutic spread to other lesions within treated bone, which align with these Phase 2a results. Methods: ZGMBC (NCT05280067), an open-label Phase 2a study, enrolled women (N=10; 8/23-3/25) with MBC lytic lesions and a Spinal Instability Numeric Score (SINS) ≥3 - ≤9. A single intratumoral (IT) Zeta-BC-003 injection was administered via fluoroscopic guidance. CT & MRI were obtained at Days 0, 84, and 180. Primary endpoints: SREs/fractures; defect volume; pain via the Numeric Rating Scale (NRS); post-op pain control via the Morphine Equivalent Dose (MED); and AE/SAEs. Secondary endpoints: QoL (SF-12v2); SINS; and tumor response. Results: Subjects (mean age 52) had the following MBC subtypes: HR+ (N=6), HR+/HER2+ (N=2), HER2+/HR- (N=1), and TNBC (N=1). Breakthrough lesions were seen in 5 subjects despite bisphosphonate therapy and 1 subject died from pleural edema prior to study end. Zeta-BC-003 was injected into 10 subjects with 11 lesions with additional therapeutic effect seen in 4 adjacent untreated lesions. A CR was shown for all lesions and there were no SREs/fractures. The bone defect volume decreased 65.4% (±20.5%; p=0.0003) and 84.1% (±13.1%; p<0.0001) at Days 84 and 180. NRS pain scores decreased 4.2% (p<0.05) and MED decreased ≥33% in opioid-treated subjects. SINS scores improved 18.5% (p<0.05), indicating increased stability. Increased PCS (24%) and MCS (12%) scores showed increased QoL. There were no treatment emergent SAE/AEs. Conclusions: Zeta-BC-003 prevented SREs/fractures in all lesions, reduced lytic defect volume, improved spinal stability, decreased pain/opioid use, improved QoL, ceased tumor activity, and demonstrated therapeutic spread to untreated lesions in the same vertebral body. These findings contrast with historical SRE rates of 53% in MBC and align with the durable CR seen in our Compassionate Use patients. Zeta-BC-003 may represent a first-of-its-kind intratumoral drug that ceases lytic activity, stimulates neo-trabecular bone growth, and improves QoL by eliminating SREs/fractures while also increasing overall survival. Clinical trial information: NCT05280067 .

Early-onset pancreatic cancer in the United States (1991–2021): Temporal trends, geographic variation, and gender disparities.

e16444 Background: Early-onset pancreatic cancer (EOPC), defined as pancreatic cancer diagnosed before the age of 50, remains a growing public health concern. This study examines national trends in EOPC incidence and mortality, highlighting disparities by state and gender. Methods: Using the Global Burden of Disease (GBD) Study 2021 Results Tool, Age-standardized mortality rates (ASMR) and age-standardized incidence rates (ASIR) for patients aged 15–49 years were extracted and data were stratified by year, sex, and location. Mortality and incidence trends were analyzed using Joinpoint regression, and annual percentage change (APC) was calculated Results: The Joinpoint regression analysis of EOPC mortality in the U.S. (1991–2021) shows an ASR increase from 1991 to 2002 (APC: +2.13%), stabilization from 2002 to 2007 (APC: -0.13%), and a significant decline starting in 2007, with ASR decreasing at -2.85% per year until 2013. This downward trend continued from 2013 to 2017 (APC: -2.02%) and slowed from 2017 to 2021 (APC: -1.80%) The Joinpoint regression analysis of EOPC in the U.S. (1991–2021) shows an ASIR increase from 1991 to 2002 (APC: +2.57%), stabilization from 2002 to 2008 (APC: -0.23%), a sharp decline from 2008 to 2012 (APC: -3.11%), and a slower decline from 2012 to 2021 (APC: -0.93%). In 2021, the highest EOPC mortality rates were observed in Mississippi (1.73 per 100,000), West Virginia (1.65), and New Mexico (1.49), while the lowest were in Utah (0.70), Massachusetts (0.73), and Rhode Island (0.78). ASIR followed a similar trend, with Mississippi (2.22), West Virginia (2.16), and New Mexico (1.99) reporting the highest case rates, while Utah (1.01), Massachusetts (1.04), and California (1.07) had the lowest. APC analysis showed that EOPC mortality increased most in West Virginia (2.56), whereas the steepest declines were seen in the District of Columbia (-1.65). Similarly, incidence rates increased most in Mississippi (4.06), while Utah (0.02) exhibited the lowest growth. Sex-specific trends revealed a decrease in EOPC mortality for males by 8.12% (from 1.29 per 100,000 in 1991 to 1.18 in 2021), while female mortality increased by 3.71% (from 0.77 to 0.79). Incidence rates declined slightly in males by 1.96% (from 1.62 to 1.58), but increased by 11.50% in females (from 1.00 to 1.11). Conclusions: From 1991 to 2021, early-onset pancreatic cancer incidence and mortality in the United States increased initially, stabilized, and then declined beginning in the late 2000s. Despite overall improvement, substantial geographic and sex-based disparities persist. Mortality and incidence have declined in males but increased in females, highlighting the need for targeted prevention and early detection strategies in high-burden regions and at-risk populations.

Health-related social needs screening in community oncology settings: Results from the WF-2303CD context assessment of NCORP clinics.

e23181 Background: Health-related social needs (HRSN; i.e., health-harming conditions like food insecurity and housing instability) are often unaddressed among individuals with cancer, contributing to worse health outcomes. Although increasingly recognized as key to quality care delivery, HRSN screening may be challenging in low-resource community oncology settings. In this study (WF-2303CD), we sought to understand HRSN screening implementation in a national sample of clinics in the NCI Community Oncology Research Program (NCORP). Methods: This mixed methods observational study was conducted with the Wake Forest NCORP Research Base. We trained site staff to assess current HRSN screening processes and contextual factors influencing implementation using Context-Driven Co-Design methods, including a key informant interview and 3 hours of clinic observation per clinic. The study team analyzed interview transcripts and observation notes using template analysis to inform clinic classification based on HRSN screening implementation quality and comprehensiveness. Results: We assessed HRSN screening processes and factors influencing implementation in 41outpatient oncology clinics across 18 NCORP Community Sites. Preliminary results suggest wide implementation variation in the 37 clinics that self-reported HRSN screening. Most clinics reported using the NCCN Distress Thermometer (n = 16) or Epic SDOH Wheel (n = 9), while others relied on verbal screening or non-validated homegrown tools (n = 12). Further evaluation revealed few clinics had high-quality, systematic screening processes and even fewer had robust processes for following up on reported needs or needs reassessment. Identified implementation challenges included reliance on staff assumptions or patients raising concerns unprompted, processes reaching only a subset of patients (e.g., patients receiving infusions), or only focused on specific HRSN domain(s). Contextual factors identified as influencing HRSN screening are highlighted in Table 1. Conclusions: This is the first national, in-depth study of HRSN screening in community oncology settings. Findings illuminate the current capacity of diverse clinics to identify patients with unmet HRSN, screening tools and approaches in use, implementation gaps, and potential levers of change. These insights will inform future efforts to enhance HRSN screening in oncology. Contextual factors influencing HRSN screening implementation. Facilitators Barriers Leadership investment Accreditation process Staff perceptions of importance Integration of social work/navigation with primary care team Trust (patient-staff, staff-staff) Staff turnover Staff time constraints Need for multiple workflows to reach all patients Lack of awareness of community resources Challenges accessing HRSN responses from other settings

Trends and disparities in appendiceal cancer mortality in the United States from 1999 to 2023.

e15678 Background: Appendiceal cancer (AC), once rare, is now recognized as a complex disease with rising mortality. At diagnosis, 74% of AC cases are already metastatic. No standardized chemotherapy exists for advanced cases, despite an alarming 232% increase in incidence in the United States, predominantly in individuals under 50 years. We analyzed U.S. mortality patterns by demographics and geography to identify high-risk groups. Methods: A retrospective analysis was performed using the CDC WONDER database. Deidentified death certificate data were extracted through ICD-10 code C18.1 to identify Appendiceal cancer–related deaths from 1999 to 2023. We analyzed mortality disparities by gender, race/ethnicity, states, and urban/rural classification. Age-adjusted mortality rates (AAMR) per 100,000 with 95% CI were reported. Trends and annual percent change (APC) were assessed using Joinpoint regression. Results: Between 1999 and 2023, 15,928 appendiceal cancer–related deaths were recorded in the United States. AAMRs increased significantly from 0.08 per 100,000 in 1999 to 0.24 in 2017 (APC: 5.61; 95% CI: 5.14–6.07), followed by stabilization through 2023 (APC: 1.03; 95% CI: −0.27–2.35). Females accounted for 8,741 deaths versus 7,187 among males and consistently had higher AAMRs. Female AAMRs rose from 0.10 to 0.25 (APC: 5.86; 95% CI: 5.23–6.51), while male AAMRs increased from 0.08 to 0.26 (APC: 6.55; 95% CI: 5.41–7.70), with both stabilizing after 2017. NH-White individuals had the highest AAMR (3.95), followed by NH-Black individuals (3.73), with Hispanics/Latinos lowest (1.43). NH-Whites showed the steepest rise (APC: 6.31; 95% CI: 5.62–7.00), while NH-Black individuals increased steadily (APC: 5.94; 95% CI: 4.36–7.53). State-level AAMRs ranged from 0.08 in Nevada to 0.30 in Delaware, with top-decile states having 3-fold higher rates than bottom-decile states. Rural areas had the highest AAMR in 2020 (0.27) and showed a marked post-2004 increase (APC: 7.72; 95% CI: 6.34–9.12), exceeding metropolitan trends. Conclusions: Appendiceal cancer deaths are increasing in the United States, highlighting the need for targeted efforts to address this trend and reduce the risk of a growing health burden.

Defining second-line sequencing in advanced renal cell carcinoma: A systematic review and meta-analysis.

4539 Background: In the absence of head-to-head comparative trials, the optimal sequencing of second-line systemic therapy for advanced renal cell carcinoma (RCC) remains undefined. As immune checkpoint inhibitor (ICI) based regimens now represent the dominant first-line standard, second-line treatment decisions increasingly rely on indirect data. We conducted a systematic review and meta-analysis to assess the comparative clinical reliability, defined as consistency of benefit across trials, of commonly adopted second-line strategies in RCC. Methods: Randomized phase II–III and prospective trials evaluating second-line systemic therapies in advanced RCC were identified through a systematic literature search. 13 trials encompassing approximately 5,000 patients were included, predominantly post TKI and post immunotherapy. Treatments were grouped by mechanism of action, including tyrosine kinase inhibitor (TKI) based regimens, mTOR inhibitor based therapies, immune-based approaches, and targeted combinations. Hazard ratios (HRs) for overall survival (OS) and progression-free survival (PFS) were pooled using inverse-variance random-effects models, with heterogeneity assessed using the I² statistic. Results: TKI based strategies demonstrated the most consistent benefit across endpoints, with significant improvement in OS (pooled HR 0.77, 95% CI 0.67–0.87; p<0.0001) and favorable PFS (HR 0.56, 95% CI 0.50-0.63; p<0.0001). Targeted combination strategies achieved consistent PFS benefit (HR 0.66, 95% CI 0.53–0.81; p<0.0001), which did not translate into OS advantage. In contrast, mTOR inhibitor based regimens were associated with modest PFS improvement but failed to demonstrate OS benefit (HR 1.13, 95% CI 0.95–1.35), while immune based approaches showed heterogeneous outcomes. To contextualize these findings within contemporary clinical practice, analyses restricted to patients previously treated with first line ICI based regimens demonstrated a consistent treatment pattern, with TKI based therapies retaining a robust PFS benefit (HR 0.57, 95% CI 0.51–0.64), while targeted combinations maintained a significant PFS improvement without a corresponding OS signal (HR 0.66, 95% CI 0.51–0.85). Conclusions: In the absence of head to head sequencing trials, this systematic synthesis of prospective evidence supports TKI based therapies as the most reliable reference second line strategy in advanced RCC, showing consistent benefit across both OS and PFS, including the immunotherapy first setting. In contrast, mTOR inhibitor based approaches and targeted combinations were associated with greater heterogeneity and did not demonstrate a consistent survival benefit despite improvements in PFS. These findings provide clinically actionable guidance for second-line treatment selection after ICI failure and underscore the need for prospective sequencing focused trials.

Conversion therapy with PD-1/PD-L1 inhibitors plus gemcitabine-based chemotherapy for initially unresectable locally advanced gallbladder cancer.

e16272 Background: Locally advanced, initially unresectable gallbladder cancer (GBC) is associated with a poor prognosis. Although immunotherapy combined with chemotherapy is recommended by NCCN guidelines, its potential to enable conversion surgery remains underexplored. This study evaluated the efficacy and safety of an intensified regimen combining PD-1/PD-L1 inhibitors with gemcitabine-based chemotherapy. Methods: This single-center real-world retrospective study enrolled 47 patients with locally advanced, initially unresectable GBC treated between February 2021 and July 2025, with follow-up through December 31, 2025. The cohort comprised 33 females (70.2%) and 14 males (29.8%). Baseline unresectability was confirmed by a multidisciplinary team (MDT) for all patients. All patients received at least two cycles of immunochemotherapy consisting of PD-1 or PD-L1 inhibitor combined with gemcitabine-based regimens including GAP (n = 32), AG (n = 5), GC (n = 4), GS (n = 3), GEMOX (n = 2), and GCS (n = 1). The primary endpoint was the conversion to surgical resection rate. Secondary endpoints included objective response rate (ORR), R0 resection rate, tumor regression grade (TRG), treatment-related adverse events (TRAEs), and survival outcomes. Results: Patients received a median of 2 conversion therapy cycles (range, 2-10) and were followed up for a median of 15.8 months (range, 4-46 months). 40.4% of patients (19/47) achieved conversion to surgical resection, with an R0 resection rate of 94.7% (18/19). Immunochemotherapy resulted in a median progression-free survival (PFS) of 14.5 months (1-year rate 51.2%, 2-year rate 30.9%) and median overall survival (OS) of 17.3 months (1-year rate 68.6%, 2-year rate 39.4%). The objective response rate (ORR) was 46.8% (22/47), including 1 complete response (CR), 21 partial responses (PR), 15 stable disease (SD), and 10 progressive disease (PD). Surgical patients demonstrated a significant survival benefit compared with non-surgical patients, with superior median OS [not reached (NR) vs. 11.4 months, P < 0.001] and PFS (NR vs. 6.4 months, P < 0.001). In the surgical cohort, a major pathologic response (TRG 0-1) was observed in 15.7% (3/19) of surgical patients. Two patients experienced Clavien-Dindo grade ≥3 postoperative complications, with no postoperative deaths. Grade 3-4 TRAEs occurred in 48.9% (23/47) of all patients, with similar incidence between surgical and non-surgical cohorts. Conclusions: The intensified multi-agent immunochemotherapy regimen demonstrated encouraging survival outcomes and manageable toxicity as conversion therapy for initially unresectable locally advanced GBC, supporting its potential as a therapeutic strategy.

Organotropic and genomic characterization of pancreatic neuroendocrine carcinoma.

4172 Background: Pancreatic neuroendocrine carcinoma (PNEC) is a rare, aggressive malignancy. Neuroendocrine carcinomas (NEC) of other primary sites, such as small cell lung cancer (SCLC), have a propensity to metastasize to the brain. Incidence of brain metastases in PNEC is poorly characterized, and most staging recommendations for extrapulmonary NEC do not include routine brain imaging. We evaluated the incidence of brain metastases in PNEC, identified factors associated with brain metastases, and characterized the genomic profile of PNEC. Methods: We reviewed the genomic profile of patients with PNEC, pancreatic ductal adenocarcinoma (PDAC), SCLC, and pulmonary large cell neuroendocrine carcinoma (LCNEC), who underwent next generation sequencing (NGS) at Memorial Sloan Kettering Cancer Center (MSKCC). We queried the Surveillance, Epidemiology, and End Results (SEER) database to identify all cases of metastatic PDAC and small cell or large cell PNEC diagnosed 2010-2022. We performed multivariable logistic regression to identify demographic and clinical factors associated with brain metastases. We used chi-squared testing to compare metastatic site frequency between tumor types. Results: The SEER query identified 412 cases of metastatic PNEC and 60,668 cases of metastatic PDAC with documented site of metastasis at diagnosis. Compared to PDAC, PNEC had higher rates of metastasis to the brain (8% vs 0.8%, p<0.001), bone (16% vs 9%, p<0.001), and distant lymph nodes (15% vs 8%, p<0.001) but lower rates to the lung (19% vs 25%, p=0.03). The incidence of liver metastases was similar in both tumor types (88% vs 87%). Within the PNEC cohort, lung metastasis was positively associated with brain metastasis (odds ratio [OR] 5.4, 95% confidence interval [CI] 2.4-12.5, p<0.001), but liver metastasis was negatively associated with brain metastasis (OR 0.2, 95% CI 0.1-0.5, p<0.001). The presence of brain metastases did not impact overall survival. Twenty eight patients with PNEC underwent NGS at MSKCC, including 12 with mixed ductal-neuroendocrine carcinoma or mixed acinar-neuroendocrine carcinoma. TP53 alterations were common (>50%) in all tumor types. RB1 alterations were seen frequently in SCLC (78%), PNEC (54%), and LCNEC (24%), but not in PDAC (2%). KRAS alterations were highly prevalent in PDAC (84%), uncommon in SCLC (3%), and with intermediate frequency in PNEC (36%). Conclusions: We found a high incidence of brain metastasis in PNEC, especially in patients with lung metastases, supporting neuroimaging in many patients with this disease. The strong association between lung and brain metastases is also noted in breast cancer, possibly explained by sequential seeding or shared features driving organotropism to these organs. PNEC demonstrated genomic features of both PDAC and SCLC. More research is needed to determine how these genomic features can inform management practices in individual patients.

Optimizing PRRT sequencing after progression on somatostatin analogue in neuroendocrine tumors: A real-world analysis.

e16330 Background: Optimal sequencing of peptide receptor radionuclide therapy (PRRT) following progression on somatostatin analogues (SSAs) in patients with neuroendocrine tumors (NETs) remains undetermined. Comparative data to guide timing of PRRT relative to other therapies are limited. Methods: We conducted a retrospective cohort study using the TriNetX Global Collaborative Network, including patients with metastatic G1-G2 well-differentiated NETs of gastroenteropancreatic origin who received SSAs as first-line therapy. Two cohorts were compared: (1) patients who received PRRT as second-line therapy without any prior systemic treatment (P2), and (2) patients who received other systemic therapies followed by PRRT (PX). Propensity score matching for baseline characteristics was performed to maintain parity. Primary outcome was overall survival (OS). Secondary analyses included treatment-related toxicities using diagnosis and laboratory-based adverse event signals within 12 months of index therapy. Results: A total of 4755 patients were included (P2: n = 3773; PX: n = 982). Approximately 80% of patients in both cohorts were Non-Hispanic White with Black patients representing around 12% in each group. Patients receiving PRRT after SSA failure without other intervening systemic treatments demonstrated improved survival outcomes compared with those receiving PRRT after other systemic therapies. The survival probability was 79.31% in P2 vs 70.48% in PX at the end of 5 years (p < 0.001), 65.75% in P2 vs 43.23% in PX at the end of 10 years (p < 0.001). Median OS favored second-line PRRT, with consistent benefit across sensitivity analyses. Toxicity analyses revealed that P2 showed lower rates of anemia, thrombocytopenia, neutropenia, and renal injury compared with PX, while rates of liver enzyme elevation were similar in both groups. Conclusions: In this large real-world analysis, use of PRRT earlier in the treatment sequence following SSA progression was associated with improved clinical outcomes and a favorable toxicity profile. These findings support consideration of PRRT as a preferred second-line option in appropriately selected patients and highlight the importance of treatment sequencing in optimizing long-term outcomes. In keeping with real-world EHR based database limitations, robust subgroup analyses could not be performed based on tumor site and grade owing to heterogeneous ICD coding practices. Further investigation incorporating more granular data is warranted. Adverse events in each cohort. Adverse Event PX P2 Hazard Ratio (95% CI) p value Neutropenia 6.2% 1.9% 3.26 (1.95–5.45) 0.54 Acute Kidney Injury 14.8% 8.3% 1.82 (1.39–2.39) 0.008 Elevated Liver Enzymes 2.0% 1.1% 1.81 (0.87–3.79) 0.40 Thrombocytopenia 18.9% 6.9% 2.90 (2.19–3.82) 0.66 Anemia 35.2% 24.3% 1.55 (1.32–1.83) 0.59

From chemotherapy to checkpoints: Survival trends in rectal adenocarcinoma.

e15594 Background: Colorectal cancer is the fourth most common malignancy worldwide and the second leading cause of cancer-related mortality in the United States. Adenocarcinoma accounts for approximately 90% of cases, with nearly one-third arising in the rectum. Management of rectal adenocarcinoma has evolved over the past two decades with advances in multimodality therapy and the introduction of immunotherapy. Pembrolizumab was approved in 2017 for select molecular subtypes of colorectal cancer. Our aim was to evaluate overall survival differences before (2005–2016) and after (2017–2022) 2017 and to assess associated sociodemographic and clinical factors. Methods: We conducted a retrospective cohort study using the SEER Plus database (December 2025 release) and identified patients diagnosed with rectal adenocarcinoma between 2005 and 2022 using ICD-O-3 code 8140/3 and site code C20.9. We stratified patients into pre-2017 and post-2017 cohorts, extracting 55,833 and 40,572 cases, respectively. We then used multivariable Cox proportional hazards regression models to assess associations between overall survival and age, sex, race, stage, median household income, and treatment modalities (surgery, chemotherapy, radiotherapy). Statistical analyses were performed using GraphPad Prism version 10.6.1, with p values < 0.05 considered statistically significant. Results: Each 1-year increase in age was associated with a 3.6% higher risk of death pre-2017 and a 3.4% higher risk post-2017 (p < 0.0001). Female sex was associated with an 11% lower risk of death pre-2017 and a 13% lower risk post-2017 compared to males (p < 0.0001). Distant stage was associated with a 3.5-fold higher risk of death in both eras (p < 0.0001). Higher income ( > 100K) was associated with a 14% lower risk pre-2017 and 16% lower risk post-2017 compared to income < 100K (p < 0.0001). Radiotherapy was associated with a 7% lower risk pre-2017 and 14% lower risk post-2017 versus no radiotherapy (p < 0.0001). Chemotherapy was associated with a 30% lower risk pre-2017 and 50% lower risk post-2017 versus no chemotherapy (p < 0.0001). Surgery was associated with a 60% lower risk pre-2017 and 74% lower risk post-2017 versus no surgery (p < 0.0001). Conclusions: Overall survival for rectal adenocarcinoma improved modestly in the post-2017 era, with greater gains across sex, income, and treatment subgroups, while age and advanced stage remained dominant predictors. These improvements likely reflect advances in systemic therapy, local control, screening, and multidisciplinary care. Although the post-2017 period coincides with the adoption of pembrolizumab, we did not assess immunotherapy exposure directly, precluding attribution of benefit specifically to checkpoint inhibition. Further studies incorporating molecular and treatment-specific data are needed to define the impact of immunotherapy and address socioeconomic disparities in rectal cancer outcomes.

Potential of urine liquid biopsy in detecting of clinically relevant genomic alterations in advanced genitourinary cancers.

e17133 Background: Liquid biopsy has transformed molecular profiling through non-invasive detection of tumor-derived genomic alterations; however, plasma-based assays may fail due to low circulating tumor DNA (ctDNA) burden, clonal evolution, or treatment-related effects. Urine is an emerging complementary biofluid that may potentially capture renal, urothelial and prostatic tumor DNA fragments and may detect clinically relevant alterations when plasma ctDNA is below detection thresholds. This study aims at evaluating the potential of urine biospecimen to detect molecular alterations in genitourinary cancers, especially renal, urothelial and prostatic cancers. Methods: Molecular spectrum analysis was retrospectively performed for 23 patients in multi-specimen (Plasma and Urine) sequenced by OncoIndx Next generation sequencing panel. ctDNA detection was performed for concordance and discordance across cancer types and sampling timepoints with urine findings as the critical aspect. Results: Among 23 patients with bladder or prostate cancer undergoing multi-specimen genomic testing, ctDNA positivity was detected in 39.1% (9/23) using plasma alone and 17.4% (4/23) using urine alone. Combined plasma and urine analysis increased overall ctDNA detection to 56.5% (13/23), identifying an additional 17.4% of ctDNA-positive cases that would have been missed by plasma-only testing. This enhanced detection supports earlier molecular identification of relapse and more timely clinical intervention. Conclusions: Urine-based liquid biopsy contributes to additional ctDNA detection rate when combined with plasma-based ctDNA detection alone. These findings support urine as a non-invasive, repeatable complementary or first-line genomic testing strategy, particularly when plasma ctDNA is undetectable.

Factors associated with clinical trial failure and termination in non–small cell lung cancer immunotherapy: A ten-year review.

11116 Background: Immunotherapy (IO) has transformed NSCLC management, yet trial efficiency and strategy-level success patterns across settings remain incompletely described. We evaluated NSCLC IO trials for termination trends, outcome accessibility, shifts in strategies over time, and factors associated with endpoint success and failure to inform future trial designs. Methods: We identified interventional NSCLC IO trials from ClinicalTrials.gov from 2015–2024 with completed/terminated status. Trials were categorized by setting (metastatic 1L and beyond 1L; perioperative; other/mixed), intervention strategy (PD-(L)1 monotherapy; PD-(L)1+chemotherapy; PD-(L)1+CTLA-4; PD-(L)1+targeted/VEGF/ADC; RT/CRT+ICI; other IO), phase group, sponsor type (industry vs non-industry), and disease stage. Statistic analysis was done using chi-square/Fisher tests and multivariable logistic regression. Among completed trials, success/failure was defined as meeting/not meeting the primary endpoint when outcomes were assessable. Results: Among 198 NSCLC IO trials, 67 (33.8%) were terminated, 131 (66.2%) completed. Termination increased over time (24.0% 2015–2017, 48.1% 2018–2019, 57.1% 2020–2022; p = 0.0005) and persisted after accounting for setting, phase, sponsor type (2018-2019 vs 2015-2017: OR 2.87, p = 0.0037; 2020-2022 vs 2015-2017: OR 3.22, p = 0.0277). Metastatic disease trials shifted from late (56.0%, 2015–2017;42.9%, 2020–2022) to 1L (28.8% - 47.6%). Endpoint outcomes were assessable in 72/131 (55.0%) completed trials - more in metastatic 1L vs later-line (75.0% vs 42.3%, p = 0.0086; OR 0.30, p = 0.013) and in industry vs non-industry trials (71.9% vs 41.9%; p = 0.0012, OR 0.38, p = 0.030). Among these, 47 (65.3%) met primary endpoints. Success varied by strategy: PD-(L)1+chemo 73.1%, PD-(L)1 mono 60.9%, PD-(L)1+CTLA-4 50.0%, PD-(L)1+targeted 20.0%, RT+ICI 85.7%; PD-(L)1+chemo outperformed PD-(L)1+targeted (Fisher p = 0.0416). Strategy shifted over time (p = 0.034), with decreased PD-(L)1 mono and increased RT+ICI and PD-(L)1+targeted combinations in later years. Proportion of PD-(L)1 monotherapy was higher for perioperative trials (47.4%) vs metastatic 1L (16.7%) and later-line settings (26.7%). Cell-(n = 8) and vaccine/virus-based (n = 12) approaches were uncommon (term. 25% and 50%) with limited assessable outcomes. Conclusions: Treatment strategies shifted away from PD-(L)1 monotherapy toward combination approaches and, in metastatic disease, from late-line to 1st line settings. PD-(L)1+chemotherapy and RT+ICI showed higher success rates than PD-(L)1+targeted combinations, though some subgroups were small. Other IO agents are rarely used in NSCLC and have insufficient reporting. Incorporating predictive risk models into trial planning may reduce low-yield studies and accelerate identification of effective treatments.

Compliance modulation of a soft robotic atrioventricular model of heart failure with preserved ejection fraction

Mortality trends of the leading cancers in Brazil: Real-world evidence from a national data platform.

e23421 Background: Cancer is a leading cause of mortality worldwide, with substantial variation by tumor type and region. Understanding national mortality patterns is essential to inform public health planning and cancer control strategies. We evaluated mortality trends of the leading cancers in Brazil using nationwide real-world data. Methods: This retrospective, population-based study analyzed deaths due to malignant neoplasms (ICD-10 C00–C97) recorded in the Brazilian Mortality Information System (SIM/DATASUS) between 2008 and 2024, using the Nodian platform, which aggregates, standardizes, and structures nationwide public health databases. Age-standardized mortality rates (ASRs) were estimated using the direct method, standardized to the Brazilian 2010 population. Analyses were stratified by sex, age group, and geographic macroregion. Temporal trends were assessed using average annual percent change (AAPC) estimated by linear regression and classified as increasing, decreasing, or stable (p < 0.05). Results: A total of 3,560,646 cancer-related deaths were identified nationwide. Lung cancer was the leading cause of cancer mortality (448,270 deaths; 12.6%), followed by breast (275,147; 7.7%), prostate (251,365; 7.1%), stomach (239,445; 6.7%), and colorectal cancer (192,023; 5.4%). Together, these five cancer types accounted for approximately 39% of all cancer deaths. Breast cancer represented the highest proportion of cancer deaths among women, while prostate cancer predominated among men. Most deaths occurred among individuals aged ≥60 years. Substantial regional heterogeneity was observed in the distribution of leading cancer types. Trend analyses showed declining mortality for stomach cancer (AAPC −1.6%, p < 0.001), increasing mortality for pancreatic cancer (AAPC +1.0%, p = 0.04), and relatively stable trends for breast and prostate cancer. Conclusions: Nationwide real-world data reveal marked heterogeneity in cancer mortality patterns in Brazil by sex, age, region, and time. Integrated population-level data platforms enable scalable and timely assessments to support cancer surveillance and evidence-based public health decision-making.

Trends in breast cancer stage at diagnosis over two decades in Côte d’Ivoire.

e22510 Background: In Côte d’Ivoire, late-stage breast cancer diagnosis remains a leading cause of poor outcomes. National cancer control plans have been implemented to promote early detection. This study aimed to evaluate the evolution of breast cancer stage at diagnosis over the last two decades. Methods: We conducted a retrospective study of patients diagnosed with histologically confirmed breast cancer at Treichville University Hospital between 2005 and 2024. Sociodemographic, clinical, and tumour stage (I-IV) data were collected. Trends in stage distribution across four time periods (2005-2009, 2010-2014, 2015-2019, 2020-2024) were analysed using chi-square tests (significance set at p<0.05). Results: A total of 699 patients were included (99.1% female; mean age 49.4±11.8 years). Stage III disease predominated across all periods (64.9% in 2005-2009 vs 51.5% in 2020-2024). Early-stage diagnoses (I-II) increased from 4.3% to 10.4% during the same interval. Metastatic cases fluctuated (19.1%, 11.1%, and 18.6% for 2010-2014, 2015-2019, and 2020-2024, respectively; p=0.037). The mean delay from symptom onset to diagnosis was 11.7±4.7 months, with a recent increase to 13.3±10 months (2020-2024), possibly due to COVID19-related healthcare disruptions. Conclusions: Our findings indicate gradual progress toward earlier breast cancer diagnosis in Côte d’Ivoire over two decades, reflecting the initial impact of national cancer control strategies. Nevertheless, advanced-stage presentations remain dominant. Strengthening organised screening programs, public education, and access to oncology services remains essential to improving survival outcomes.

Efficacy and safety of CAR T-cell therapy in hematologic malignancies: A systematic review and meta-analysis.

e18575 Background: CAR T-cell therapy has shown promise in treating relapsed/refractory hematologic malignancies. To comprehensively evaluate its benefits and risks, we performed a systematic review and meta-analysis of published studies. Methods: Literature search identified 47 studies (31 trials, 10 observational, 6 case series; n per study 12–450; median follow-up 18 months). Pooled overall and complete response rates (ORR and CR), along with 12-month progression-free and overall survival (PFS and OS), were estimated. Rates of cytokine release syndrome (CRS), neurotoxicity, cytopenias, infections, hypogammaglobulinemia, and treatment-related mortality were pooled. Between-study heterogeneity (I²) and publication bias were assessed. Results: Pooled ORR was 75% (95% CI 70–80%) and CR 45% (95% CI 40–50%). B-cell malignancies had higher response (ORR 82%, CR 53%) than T-cell (ORR 63%, CR 34%). At 12 months, PFS was 60% (95% CI 55–65%) and OS 72% (95% CI 67–77%). CRS occurred in 85% (grade ≥3: 20%) and neurotoxicity in 40% (grade ≥3: 15%). Other toxicities included cytopenias (70%), infections (35%), hypogammaglobulinemia (25%), and treatment-related mortality (5%). Heterogeneity was high (I² ORR 75%, CR 68%); sensitivity analyses reduced it (65%, 60%). No significant publication bias was detected. Longer follow-up (>24 months) was associated with higher relapse and lower PFS/OS; patients with prior stem-cell transplant had slightly lower ORR and CR. Patient-reported quality of life improved post-therapy despite acute toxicities. Conclusions: This meta-analysis confirms that CAR T-cell therapy achieves high response rates (ORR 75%, CR 45%) and favorable 12-month survival (PFS 60%, OS 72%) in hematologic malignancies. Toxicities were common (CRS 85%; grade ≥3: 20%; neurotoxicity 40%, grade ≥3: 15%), as were cytopenias (70%) and infections (35%), but these were generally manageable. Treatment-related mortality was 5%. Quality of life improved overall despite acute toxicity. Heterogeneity was substantial, but sensitivity analyses and absence of publication bias support the robustness of these findings. Longer follow-up revealed higher relapse rates and reduced survival, and prior stem-cell transplant was linked to marginally lower responses. These results underscore that CAR T-cell therapy offers significant benefits along with considerable toxicity, and highlight the need to optimize patient selection and management.

Prevalence, genomic landscape, and outcomes of HRD and HRR alterations in solid tumors: Real-world data from western India.

e15207 Background: HRD and HRR alterations are key predictive and prognostic biomarkers in solid tumors, guiding targeted therapy selection. Limited Indian real-world data exist on their prevalence, genomic profile, and outcomes across cancer types. Methods: The retrospective data of cancer patients who have undertaken the aforementioned test from Jan 2018 to July 2025 were evaluated. Treatment patterns, PARP inhibitor use, and survival outcomes were analyzed; mOS was estimated using Kaplan–Meier and compared by log-rank test. Results: A total of 566 patients underwent HRD and HRR testing with a median age of 61 years (IQR 52–70) and the most common cancer type tested were gynecological (29.3%,n = 166), breast (23.5%,n = 133). Next Generation Sequencing-based genomic profiling revealed HRD positivity in 26.7% (n = 151), with the highest prevalence in gynecological cancers (39.1%), followed by breast (15.9%), genitourinary and hepatobiliary (8.6% each). Among positive cases, 67 were HRR-positive and 84 were HRD-positive, with comprehensive LOH, TAI, and LST scoring available for 81 of the HRD-positive patients and TAI was the predominant determinant of HRD positivity. While the most common HRR genes alteration detected were BRCA1 (17.3%), TP53 (15.8%), and ATM (11.4%). Among HRD-positive cases, 64.9% (98/151) had positive mutation, most commonly involving BRCA in 29.6%, TP53 in 27.6% and 5.1% exhibiting concurrent BRCA and TP53 mutations however 35.1% had HRD positivity without any mutation based on genomic scar score. PARP inhibitors were used in 30.5% of patients, most frequently olaparib (82.6%), with an overall response rate 69.8%. Among HRD-positive patients, mOS was not reached for BRCA-mutated cases compared to 36.8 months for non-BRCA mutated cases (p = 0.012). Conclusions: This study reinforces the pivotal role of universal HRD testing in informing biomarker-driven therapeutics, maximizing survival gains, and narrowing gaps in implementing precision cancer care. HRD positivity with BRCA positivity have better response rate as compared to HRD with non BRCA mutated patients.

Feasibility of constructing an external control arm for early efficacy assessment in phase 1a: Methodological framework from the INVOKE trial.

11521 Background: Phase 1a dose-escalation trials are primarily designed for safety but often lack sufficient visibility into efficacy signals due to small sample sizes and the absence of a comparator arm. This uncertainty can delay decision-making and increase late-stage failure rates. We hypothesized that a synthetic External Control Arm (ECA) aggregating real-world data (RWD) and historical clinical trials could be established to contextualize efficacy readouts from the ongoing OKN4395 Phase 1a study (INVOKE; NCT06789172; Trial in Progress abstract submitted to this meeting). Methods: We curated a retrospective dataset of 298 solid tumor patients derived from two complementary sources, RWD and recent clinical trials (inclusion date ≥ 2016). To ensure comparability with the INVOKE population, patient selection adhered strictly to the study’s inclusion and exclusion criteria, with only individuals who had exhausted all standard-of-care (SoC) options being retained. We further refined the cohort to ensure exact overlap, retaining only external patients who shared an identical baseline covariate profile with at least one patient in the INVOKE cohort. Finally, we applied statistical adjustment methods, including Inverse Probability of Treatment Weighting (IPTW) and G-Computation, to align the remaining cohort on major prognostic variables (e.g. ECOG status, prior lines of therapy, inter alia). Results: A total of 13 INVOKE patients with available endpoint data were eligible for ECA matching. The selection process yielded a final robust ECA cohort of 40 patients. In this refined population, we achieved robust covariate alignment between the ECA and the INVOKE cohort, with absolute Standardized Mean Difference (SMD) < 0.13 for all major prognostic variables. Despite the methodological differences between IPTW and G-computation, both statistical approaches yielded highly consistent estimates. Data from INVOKE, a study in early stages, is yet to publish data (TIP presented at this meeting), comparative efficacy data will be disclosed at a later date. Conclusions: We demonstrated that constructing a synthetic ECA generating outcomes directly comparable to those of the population of an active Phase 1a dose-escalation study is methodologically feasible. The establishment of this matched external cohort and first application to the INVOKE Phase 1a study provides a proof of concept framework for early efficacy assessment. This "de-risking" tool is now positioned to provide rapid, concurrent contextualization of emerging data from the INVOKE trial upon data maturity.

T-cell stem-like memory cells as a novel biomarker to identify patients with sustained disease control during treatment with PD-(L)1 immunotherapy.

2578 Background: Reliable, clinically easily testable biomarkers to monitor the benefit of treatment with immune checkpoint inhibitors (ICI) remain limited. We performed comprehensive analysis of markers defining differentiation and functional status of different populations of peripheral blood T-cell in order to identify biomarkers associates to treatment response in treated patients. Methods: Peripheral blood was analyzed from metastatic patients receiving PD-(L)1–based therapy: anti–PD-(L)1 monotherapy (n=102), combination therapies (n=35), or chemo-immunotherapy (n=14). Sampling time on therapy was heterogeneous, recorded, and included in sensitivity analyses. All patients provided informed consent under IRB approval. A variety of CD4 andCD8 cell populations (naive, stem-like memory [SCM], central memory [CM], effector memory [EM], EMRA/TEMRA). were analyzed using -standardized multiparameter flow cytometry. Pre-specified stem-like–to–effector balance ratios were computed within CD4 and CD8 (SCM/CM and SCM/EM; SCM/EMRA exploratory). Markers analysis included CD45RA, CCR7, CD95, TCF1/TCF7, TOX, TIGIT, and Ki-67). Clinical benefit was assessed by disease control rate (DCR=CR+PR+SD) versus progressive disease ( PD). Biomarkers were analyzed using Mann–Whitney U tests and clustered using unsupervised 1D k-means. Time-to-event analyses used univariable Cox models for time to disease control loss (DCL), defined from ICI initiation to first PD (or discontinuation for progression). Results: Among 151 patients, DCR n=89 and PD n=62, disease control was characterized by a stem-like/early memory blood T-cell profile. DCR showed higher CD4 and CD8 naive and SCM frequencies (p<0.05), whereas PD showed enrichment in differentiated T cell populations including CD4 CM/EM and CD8 EM (p≤0.01 to p<0.0001). Stem-like–to–effector cell ratios were higher in DCR for both CD4 and CD8 (SCM/CM p<0.001; SCM/EM p<0.01) and for CD4 SCM/EMRA (p<0.05). In total CD3 T cells, DCR had higher CD45RA (p<0.001), CCR7 (p<0.05), and TCF1/TCF7 (p<0.05), with lower CD95 (p<0.05) and reduced TIGIT (p<0.01), TOX (p<0.05), and Ki-67 (p<0.05). In Cox models for DCL (inverse-coded; HR>1 favorable), significant associations were observed across subsets, ratio, and marker features (HR range 1.7–4.8; all p<0.05), with the strongest effects for stem-like–to–effector cell ratios and differentiated subset metrics. Associations were observed across PD-(L)1–based regimens. Conclusions: Stem-like memory CD4 and CD8 T cells represent a new marker associated with efficacious response to ICI and DCR while a dominant effector/terminal differentiation profile to DCL and PD. These findings underscore the importance of immune monitoring to identify patients at risk of early loss of treatment benefit and disease progression and to guide treatment modifications.

Systematic modeling of tumor-microenvironment–mediated NK cell dysfunction in pancreatic cancer to identify KLRB1 as a target for cell engineering.

e14541 Background: Adoptively transferred natural killer (NK) cells fail in pancreatic ductal adenocarcinoma (PDAC) due to rapid functional suppression within the immunosuppressive tumor microenvironment (TME). A precise understanding of the dominant inhibitory mechanisms, especially those exerted by cancer-associated fibroblasts (CAFs) and soluble tumor-derived factors, is lacking. We hypothesized that a systematic, multi-model dissection of PDAC-mediated suppression could reveal conserved, targetable pathways for engineering resistant NK cells. Methods: To model the PDAC TME, we used two complementary in vitro suppression systems—co-culture of expanded human peripheral blood NK cells with patient-derived CAFs (cell-contact inhibition) and culture in conditioned medium from patient-derived PDAC organoids (soluble-factor inhibition)—and performed scRNA-seq to identify the most consistently upregulated inhibitory receptor, knocked it out in primary human NK cells using CRISPR–Cas9, and compared functional recovery of WT versus KO NK cells in both in vitro models and in an NSG mouse co-engrafted with human PDAC cells and CAFs. Results: scRNA-seq of NK cells exposed to either CAF co-culture or PDAC organoid CM revealed a shared and dominant upregulation of the inhibitory receptor KLRB1 (encoding CD161). CRISPR-mediated KLRB1-KO did not impair NK cell development or basal cytotoxicity. Critically, under both suppression models, KLRB1-KO NK cells maintained significantly higher levels of degranulation (CD107a+) and IFN-γ production compared to WT NK cells. Transcriptomically, KO cells preserved effector gene programs and exhibited attenuated induction of exhaustion signatures. In the in vivo co-engraftment model, adoptively transferred KLRB1-KO NK cells mediated superior tumor control. Single-cell analysis of NK cells recovered from tumors confirmed that KO cells maintained a higher frequency of a transcriptionally active effector cluster, whereas WT NK cells were predominantly driven into dysfunctional states. Conclusions: By systematically modeling both contact-dependent and soluble-mediated immunosuppression, we identify KLRB1 as a central, conserved "molecular brake" induced in NK cells by the PDAC TME. Its deletion generates "TME-resilient" NK cells with preserved effector function in situ. This work validates a discovery pipeline using complementary TME models and establishes KLRB1-KO as a promising strategy to engineer next-generation allogeneic NK cell products for overcoming microenvironmental suppression in solid tumors.

Clinical experience–based assessment of informed consent forms written by generative artificial intelligence.

e13717 Background: The OpenEvidence (OE) generative artificial intelligence (AI) platform has emerged as a purpose-built tool for clinical use, with recent expansion into patient-based communication capabilities. The implementation of AI in generating informed consent documents has not yet been described for chemotherapy or immunotherapy. We sought to evaluate generalized systemic cancer treatment consent forms generated by OE and GPT 5.1 (OpenAI). Methods: Forms were generated using the same prompt for both OE and GPT 5.1. We developed an 8-item assessment instrument to evaluate forms based on description, accuracy, clarity, detail, and length. The assessment compared AI generated forms to a standard baseline consent form. Responses were collected on a 5-point scale ranging from -2 (worse) to +2 (better) with 0 representing equivalence. The instrument was administered through the Qualtrics platform, with raters randomized to evaluate one of the forms in a blinded format. Additionally, raters were blind to the AI generated nature of the forms. Raters were pharmacists, nurses, and physicians in the hematology/oncology department at our institution. Anonymized information regarding raters’ clinical practice background was collected. Composite instrument scores were calculated by summing item-level responses for each rater. Overall composite scores between OE and GPT 5.1 were compared using the Wilcoxon rank-sum test. Differences in composite scores across years-in-practice categories were evaluated separately for each model using the Kruskal–Wallis rank-sum test. Results: A total of 29 complete form assessments were collected: 16 for GPT 5.1 and 13 for OE. 79% (n = 23) of raters had an outpatient specialization. 59% (n = 17) of raters were nurses, 34% (n = 10) were pharmacists, and 7% (n = 2) were physicians. 34% (n = 10) of raters had < 10 years in practice and 32% (n = 9) raters had > 20 years in practice. Median (IQR) composite instrument scores for GPT 5.1 and OE were 6 (2, 14) and 8 (5, 10) respectively, p = 0.63. Scoring distribution did not differ significantly by rater training background (GPT 5.1: p = 0.79, OE: p = 0.97). Composite scores varied across years-in-practice groups for both models. For GPT 5.1, median (IQR) composite scores were 2 (−3, 5) for < 10 years, 13 (9, 16) for 10–20 years, and 11 (4, 15) for > 20 years (p = 0.07). For OE, median (IQR) composite scores were 9.0 (8.0, 11.0), 8.0 (5.8, 14.0), and 3.0 (−1.0, 4.0), respectively (p = 0.05). Conclusions: Despite the medically oriented focus of OE, informed consent documentation did not differ significantly between OE and the publicly available GPT 5.1. We observed variation in perceptions of these forms based on clinical experience, which may inform implementation of this technology.