Browse Articles

Reirradiation with NBTXR3/SBRT in combination with nivolumab or pembrolizumab for the treatment of patients with recurrent or metastatic head & neck squamous cell carcinoma (HNSCC) in the phase I trial study 1100.

6034 Background: HNSCC reirradiation with stereotactic body radiation therapy (SBRT) has emerged as a potential option for the management of locoregionally recurrent (R) or metastatic (M) HNSCC, however treatment toxicity remains a limiting factor. NBTXR3 is a novel intratumoral radioenhancer composed of functionalized hafnium oxide nanoparticles which locally amplifies radiation therapy (RT). In pre-clinical models, NBTXR3/RT has demonstrated an ability to generate cell killing effects at low RT doses as well as trigger local and systemic immune responses. Here we report outcomes in pts with R or R+M HNSCC who were treated with reirradiation with NBTXR3/SBRT followed by immune checkpoint inhibitors (ICIs). Methods: A phase I dose escalation/expansion trial [NCT03589339] evaluating NBTXR3/SBRT followed by ICI (nivolumab or pembrolizumab) included a subgroup with R or R+M HNSCC that were either naïve or resistant to prior ICI. Pts received an intra-tumoral NBTXR3 injection at dose of 22% or 33% of gross tumor volume (GTV), SBRT (35 Gy in 5 fractions), and ICI. Primary objective was safety and establishing RP2D of NBTXR3/SBRT/anti-PD-1 combination. The expansion part tested the RP2D (33% of GTV). Secondary objectives include efficacy. Results: From June 2019 to February 2025, 30 pts were treated: 16 ICI naïve, 14 ICI resistant with median age of 67 years, 90% ECOG 0-1, and 55% HPV negative. 20 pts (66.7%) had R disease, 10 (33.3%) pts had R+M disease. All injected lesions were in a previously irradiated H&N field. Median time from end of prior radiotherapy to NBTXR3 injection was 21.3 [5-229] months. Median GTV was 16.1 [3-110] mL. 4 pts (13.3%) experienced G≥3 injection-related AEs, 8 (26.7%) G≥3 RT-related AEs, and 6 (20%) G≥3 NBTXR3-related AEs. Carotid injury was not observed. Most frequent injection or NBTXR3-related AEs were injection site pain (10%), oropharyngeal pain (6.7%), tumor pain (6.7%), dysphagia (6.7%), and soft tissue necrosis (6.7%). 27 pts (90%) were evaluable for efficacy. For all disease (i.e. injected and non-injected), the objective response rate (ORR) was 48.1% (13/27) and disease control rate (DCR) was 77.8% (21/27). In R pts, ORR was 64.7% (11/17) and DCR was 88.2% (15/17). In R+M pts, ORR was 20% (2/10) and DCR was 60% (6/10). Survival outcomes in reirradiation pts with R and R+M HNSCC will be presented. Conclusions: Reirradiation with NBTXR3 with SBRT and anti-PD1 was feasible in R or R+M HNSCC with an adverse effect profile expected for this clinical setting. Encouraging preliminary efficacy outcomes have been observed, thus warranting further investigation. Clinical trial information: NCT03589339 .

From trials to the clinic: Real-world experience with mirvetuximab soravtansine in platinum-resistant ovarian cancer and the "Lazarus effect.

e17579 Background: Platinum-resistant ovarian cancer (PROC) has limited treatment options, with response rates of 10–15% and median progression-free survival (PFS) of ~ 3.5 months. Mirvetuximab soravtansine (MIRVE), a folate receptor alpha (FRα-)directed antibody-drug conjugate, is the first targeted therapy approved for FRα-positive PROC. This single-center real-world study reports on its effectiveness and safety, including a unique case demonstrating a ‘Lazarus effect’ and stable brain metastases. Methods: Seventeen FRα+ PROC patients treated with MIRVE at the University Hospital of Florence (February-December 2025) were retrospectively analyzed. Data included prior therapies, FRα expression, PFS, response, and toxicity (CTCAE v5.0). PFS was defined as time from treatment initiation to progression or last follow-up. FOLR1 positivity was defined as ≥75% of viable tumor cells with moderate (2+) or strong (3+) membrane staining.Primary objectives were real-world effectiveness and safety; secondary objectives included factors associated with response and toxicity characterization. Results: The median patient age was 60.5 years (range 39–80). Patients were heavily pretreated: 35% had received more than two prior lines of therapy, 53% more than three, and 12% received MIRVE in the first-line platinum-resistant setting. FRα expression was histologically scored as 2+ in 3 patients(18%), and 3+ in 14 patients (82%). Median PFS was 3 months (range 1–10), with 8 progression events and 9 censored observations at the time of analysis. Grade ≥3 treatment-related adverse events occurred in 6% of patients (peripheral neuropathy). Grade 1–2 toxicities included ocular events (24%), neurotoxicity (35%), neutropenia (12%), and thrombocytopenia (6%). One patient harboring BRCA1 mutation and FRα 3+, previously treated with first-line carboplatin plus paclitaxel followed by olaparib maintenance and second-line carboplatin plus gemcitabine, achieved near-complete hepatic response, a partial response at other disease sites, and stable brain metastases. This exceptional clinical benefit was consistent with a marked “Lazarus effect” and was associated with a progression-free survival of 10 months. This patient also experienced the only grade 3 treatment-related toxicity (neuropathy), which led to treatment discontinuation. Conclusions: Mirvetuximab Soravtansine demonstrated a manageable safety and PFS outcomes consistent with those reported in clinical trials. High FRα expression (3+) correlated with improved clinical outcomes. Notably, the association of the most severe treatment-related toxicity with exceptional response suggests a potential association between higher-grade adverse events and therapeutic benefit. These real-world data complement data from pivotal trials and provide clinically relevant insights for patient management.

Feasibility and preliminary efficacy of a peer community navigator (PCN)–led intervention for minoritized individuals considering early-phase clinical trials (EPCTs).

e23152 Background: EPCTs assess the safety and early clinical activity of new cancer therapies. Underrepresentation of members of racial and ethnic minority groups (REMG) in EPCTs limits accessibility to novel therapies and is influenced by limited knowledge about CTs, mistrust, and structural racism. We conducted a pilot study to assess the feasibility and preliminary efficacy of a peer community navigator (PCN)-led educational intervention for REMG patients with advanced solid tumors (AST) considering EPCT participation. Methods: Participants were English speaking and ≥18 with AST who self-identified as REMG and were referred to the Mount Sinai Early Phase Trials Unit (EPTU). The PCN identifies as Black and is a former EPCT participant and trained peer navigator. The intervention consisted of a virtual call with the PCN to provide an introduction to EPCTs and identify barriers to and concerns about participation before the initial EPTU visit followed by a structured exit interview. The primary endpoint of feasibility was assessed using Bowen’s Framework. Secondary endpoints included pre- and post-intervention Clinical Trials Knowledge and Beliefs Scale (CHEKS) and the Group-Based Medical Mistrust (GBMM) Scale. Wilcoxon signed-rank tests were used for comparison. Analyses were conducted using PRISM v8.0.2. Exit interviews were analyzed by 3 researchers who iteratively developed a codebook and independently coded transcripts using ATLAS.ti to identify emergent themes. Results: Among participants, 7 were male and 11 female; 9 identified as Black, 7 non-Black Hispanic, and 2 multiracial. Median age was 61. Median annual household income was $30K. Of 24 approached, 18 agreed to participate (75%), demonstrating high acceptability. PCN-EPTU communication occurred in all cases (100%), indicating strong integration. 16 of 18 participants completed the intervention (88.89%), exceeding the 80% remote feasibility benchmark. Median CHEKS scores increased from 105.0 (IQR 86.0-108.0) to 117.0 (IQR 111.0-121.0), median delta value 13.0, p < 0.01. No significant differences were observed in median GBMM Scale scores (25.0 [IQR 18.0–34.0] vs. 24.0 [IQR 18.0–25.0]). In exit interviews, all participants indicated they would work with a PCN again. Common themes included increased knowledge and comfort about EPCTs, rapport with the PCN, and increased likelihood to participate. Conclusions: This study supports the feasibility and acceptability of a PCN-led educational intervention to enhance REMG awareness of and interest in EPCTs. Significant increases in CHEKS scores indicated improved CT knowledge. Participant experiences were positive and indicated increased comfort with EPCTs and interest in enrolling. Next iterations will be multilingual and guided by focus group feedback from patients, caregivers, and clinicians.

Impact of COVID-19 pandemic on breast cancer incidence and immunohistochemical subtype distribution: A 10-year retrospective analysis from a regional Irish cancer center.

e12584 Background: The COVID-19 pandemic disrupted cancer diagnostic pathways and screening programmes. The long-term impact on breast cancer presentation patterns and immunohistochemical (IHC) subtype distribution remains incompletely characterised. This study evaluated temporal trends in breast cancer presentations at a regional Irish cancer centre. Methods: A retrospective longitudinal cohort study was conducted including approximately 80% of breast cancer cases from the South East region of Ireland treated at University Hospital Waterford. Patients diagnosed between January 2015–December 2019 (pre-pandemic) and January 2020–December 2024 (pandemic/post-pandemic) were analysed. Tumours were classified by IHC subtype (ER-positive, HER2-positive, triple-negative) and metastatic status. The primary endpoint was change in subtype distribution over time. Secondary endpoints included total case volume and incidence of de novo metastatic disease. Results: Breast cancer diagnoses declined during peak pandemic periods, followed by recovery with a 10% increase in post-pandemic case volume compared with pre-pandemic levels. Subtype analysis demonstrated a 4% increase in ER-positive disease, a 6% reduction in HER2-positive disease, and a 3% increase in triple-negative breast cancer (TNBC). TNBC showed a sustained upward trend following service recovery. The incidence of de novo metastatic disease doubled compared with pre-pandemic baseline. Conclusions: Post-pandemic recovery has been associated with increased breast cancer presentations and a shift towards more aggressive disease phenotypes, including rising TNBC and metastatic disease rates. These findings highlight the downstream impact of diagnostic delays and disrupted screening pathways and support the need for national-level analyses to guide future healthcare service planning.

GLP-1 receptor agonist therapy and risk of small intestinal neoplasms: Real-world evidence from a global network.

e16499 Background: As indications for GLP-1 RAs expand, their long-term oncologic safety remains a key concern. Supraphysiologic GLP-1 receptor stimulation has raised concerns for tumorigenesis in receptor-expressing tissues, including the gastrointestinal tract. While the safety of GLP-1 RAs in thyroid and pancreatic malignancies has been evaluated, their impact on small intestinal neoplasms remains poorly understood despite a rising incidence of these cancers. Emerging real-world data suggest potential anti-inflammatory and metabolic benefits of GLP-1 RAs with reduced colorectal cancer risk; however, it is unclear whether such effects extend to the small intestine, which harbors a high density of GLP-1- secreting L cells. We examined the association between GLP-1 RA therapy and the incidence of small intestinal malignancies compared with other standard therapies. Methods: We conducted a retrospective cohort study using the TriNetX Global Collaborative Network including adults (18–80 years) with DM or overweight/obesity. Patients prescribed a GLP-1 RA within 1 year of diagnosis were compared with those receiving other standard therapies, excluding individuals with pre-existing GI or hereditary cancer risk conditions. 1:1 PSM was performed for demographics, BMI, HbA1c and lifestyle factors. Primary and secondary small intestinal neoplasm outcomes were assessed at 3, 5 and 10 years with ARR and 95% CIs calculated. Results: At 3 years, GLP-1 RA use was associated with a lower absolute risk of malignant small intestinal neoplasms compared with non-GLP-1 therapy (1.4% vs 3.8%; ARR 2.4%, 95% CI 2.0–2.8; p< 0.0001). Similar risk reductions persisted at 5 years (ARR 2.6%, 95% CI 2.2%–3.0%) and 10 years (ARR 3.2%, 95% CI 2.7%–3.6%; both p< 0.0001). Lower absolute risks were also observed for benign neoplasms and carcinoid tumors across all time points. No consistent differences were observed for small intestinal lymphoma or GIST at earlier follow-up, though modest risk reductions for lymphoma emerged at 10 years. Conclusions: Despite theoretical concerns regarding incretin-based therapies and gastrointestinal tumorigenesis, our study found no evidence of increased risk for small intestinal malignancies among GLP-1 RA users. Instead, lower absolute risk was observed compared to controls concurring oncologic safety. Given limitations of retrospective analyses mechanistic and prospective studies are needed to further study the drivers of this observed inverse association. OUTCOMES at 10 years GLP1 Receptor agonist (%) No GLP1 therapy (%) ARR(95% CI) P VALUE Malignant neoplasms of the small intestine 0.021 0.053 3.2 (2.7 – 3.6) <0.0001 Benign neoplasms of the small intestine 0.051 0.0066 1.6 (1 - 2.1) <0.0001 Lymphoma of small intestine 0.014 0.018 0.4 (0.1 – 0.7) 0.0126 Carcinoid tumor of the small intestine 0.011 0.020 0.9 (0.6 – 1.2) <0.0001 GIST Small intestine 0.005 0.006 0.1 (0.1 – 0.3) 0.6920

Implementation of an expedited cancer workup clinic for uninsured patients at a metropolitan safety net hospital.

1573 Background: Timely evaluation of patients with suspected cancer is essential; yet delays in diagnosis and treatment remain common in safety-net health systems, particularly for uninsured or underinsured patients. Such delays are associated with advanced stages of disease at the time of diagnosis and increased mortality. While inpatient evaluation can facilitate multidisciplinary coordination, it is associated with higher costs. Outpatient delivery models designed to expedite cancer workups have been demonstrated to reduce time to diagnosis and number of hospitalizations. We describe outcomes from an Expedited Workup Clinic (EWC) in a large Los Angeles County safety-net hospital. Methods: The EWC was designed to coordinate outpatient imaging, biopsy, and oncology specialty referrals for vulnerable patients with suspected malignancy. All patients were uninsured and lacked a primary care provider. Referrals originated from the Emergency Department (ED), Urgent Care, or inpatient admissions. We retrospectively analyzed outcomes from 2022 to 2025 using real-world data extracted from the Electronic Health Record. Results: 255 patients were referred to the EWC encompassing 153 men (60%) and 102 women (40%). 155 patients (60%) were Hispanic/Latino, with other represented ethnicities including 28 (10.9%) Asian/Pacific Islander, 22 (8.6%) Black/African American, and 14 (5.5%) White/Caucasian. Spanish was the preferred language of 150 patients (58.8%). The majority of patients (87.9%) were referred from the ED with a median time to first EWC appointment of 11 days (IQR 7-14). From the first EWC appointment, the median time to imaging was 6 days (IQR 1-14), to biopsy was 21 days (IQR 12-54), and to initiation of oncologic treatment was 51 days (IQR 34-83). 92 patients (36.1%) ultimately had a negative malignancy workup, and 38 patients (14.9%) were lost to follow-up. Among patients with confirmed malignancy, the most commonly identified primary sites were colorectal (19.2%), lung (10.6%), hematologic (9.6%), pancreatobiliary (9.6%), prostate (6.7%), head and neck (6.7%), hepatic (6.7%), renal (4.8%), gynecological (3.8%), melanoma (3.8%) and breast (3.8%). Conclusions: The EWC demonstrates the feasibility of providing timely malignancy evaluation for high-risk, uninsured patients within a resource-limited setting. These findings support the use of an outpatient rapid evaluation clinic, while surfacing opportunities for improvement including reducing the number of patients lost to follow-up. Further evaluation of cost-effectiveness and access barriers may continue to strengthen the design of rapid workup clinics meant to serve our most vulnerable patients.

Demographics and outcomes of individuals with young-onset small cell lung cancer.

8104 Background: A diagnosis of lung cancer before the age of 50 is commonly driven by oncogenic alterations and observed in patients without a history of tobacco use. Little is known about the characteristics and outcomes of individuals with young-onset small cell lung cancer (YO-SCLC). Methods: We retrospectively reviewed the charts of patients diagnosed with SCLC treated in the Indiana University (IU) Health System from 2018 to 2024. Patients were stratified by age <50 or >50 years at diagnosis of SCLC for statistical comparison of key variables and outcomes. Fisher’s exact and Wilcoxon rank sum tests were applied to categorical and continuous variables, respectively. Results: Among 416 individuals with SCLC, we identified 23 patients with YO-SCLC. The incidence of YO-SCLC was 5.5 cases per 100 cases of SCLC (95% CI, 3.7-8.2) with a median age of 47 (40-49). None of the YO-SCLC were detected by lung cancer screening (LCS), compared to 17% in the >50 cohort (Table 1). Patients with YO-SCLC are more likely to have active tobacco use but significantly less pack years, compared to >50. Majority of YO-SCLC (74%) were diagnosed with extensive-stage SCLC, analogous to >50. Extensive-stage YO-SCLC patients had similar platinum-sensitivity and PFS with platinum-based therapy with or without immunotherapy but trended to have improved median OS (24.7 vs 11.1 months, HR 1.67, 95% CI 0.88-3.15, P = 0.12). Limited-stage YO-SCLC patients were more likely to receive prophylactic cranial irradiation (PCI, 67% vs 25%, P = 0.04) but demonstrated no significant difference in PFS or OS with guideline-directed therapy, compared to >50. Conclusions: YO-SCLC represents a unique under-recognized population with trends toward improved OS, despite comparable platinum-sensitivity and PFS with chemoimmunotherapy for ES-SCLC. Such patients are currently excluded from LCS due to age. These results suggest eligibility criteria for LCS could be expanded to include individuals <50 with active tobacco use (>20 PY) and a family history of lung cancer to increase detection of YO-SCLC. The molecular underpinnings of YO-SCLC remain to be elucidated. Investigation into the genomic and neuroendocrine subtype composition of YO-SCLC is underway. Characteristic Age <50 (n = 23) Age >50 (n = 393) Overall (n = 416) P value Sex (female) 15 (65%) 231 (59%) 246 (59%) 0.7 Stage (ES-SCLC) 17 (74%) 257 (65%) 274 (66%) 0.5 BMI 33 (23, 38) 27 (23, 32) 27 (23, 32) 0.046 COPD 7 (30%) 207 (53%) 214 (51%) 0.052 Current tobacco use 20 (87%) 247 (63%) 267 (64%) 0.025 Pack years 30 (25, 50) 45 (30, 60) 45 (30, 60) 0.043 Eligible for lung cancer screening 0 (0%) 301 (77%) 301 (72%) <0.001 Detected by lung cancer screening 0 (0%) 68 (17%) 68 (16%) 0.021 +FH lung cancer 9 (39%) 103 (26%) 112 (27%) 0.14

Merkel cell carcinoma incidence, trends, and survival rates among adults from United States cancer statistics during 2000-2022.

e21560 Background: Merkel cell carcinoma (MCC) is a rare form of skin cancer that frequently metastasizes and is associated with low survival rates. Studies that describe the epidemiology of MCC are often limited by small sample size and short duration of follow-up. Methods: This is a retrospective, population-based epidemiologic study of all cases of MCC from the Surveillance, Epidemiology, and End Results (SEER) database between 2000 and 2022. Incidence and mortality rates were calculated using SEER*Stat software and SEER 17 (2000-2022) that covers approximately 26.5% of the U.S. population (based on 2020 census). Results: In total, 12,291 patients were identified between 2000-2022. Most cases are diagnosed at 2021 (716 cases) averaging 0.7 per 100 000 person-years. MCC incidence rates increased with increasing age among male and female with the majority of patients age 85+ (p < .05). There is a higher incidence in male (7,857 total cases) than female (4,434 total cases). Incidence is also higher among white patients (11,662 cases) (94.90%) than black patients (168 cases). Most of MCC are located in head and neck region with about (4319 cases). The annual percent change in incidence rates was highest among 65–69-year-olds and black people. Observed five years survival rate 48.70% (95% CI 47.5%-49.9%). Diagnosis between 2017-2022 was associated with improved survival. Conclusions: This study demonstrates that most patients were white, male, and age 85+.Head and neck region is the commonest site of tumour. Recent diagnosis was associated with improved survival.

Efficacy outcomes by patterns of progression in patients with advanced renal cell carcinoma from the phase 3 CLEAR trial.

4527 Background: In CLEAR, lenvatinib + pembrolizumab (L+P) significantly improved efficacy vs sunitinib (S) in treatment-naïve patients with advanced renal cell carcinoma (Motzer 2021, Motzer 2024). We summarize efficacy by patterns of progressive disease (PD) for the L+P arm and report survival per a modified PD classification system (based on patients receiving immunotherapy; data cutoff 31 July 2022). Methods: Treatment‐naïve patients (n=1069) who had clear‐cell advanced renal cell carcinoma were randomized (1:1:1) to receive: L 20 mg PO daily + P 200 mg IV Q3W; or L + everolimus; or S 50 mg PO daily (4 wks on/2 wks off). Methods for post-progression survival/OS analyses were previously reported (Grünwald 2025). A modified system of classifying PD (Saal 2025) is used in new analyses of the L+P arm reported here: low risk (progression of existing lesions); intermediate risk (new lesions without progression of existing lesions); high risk (progression of existing lesions + new lesions). The number of patients with PD was determined by independent imaging review at the data cutoff date. Results: At the time of PD, the median tumor burden of target lesions was lower with L+P (−48.1%) vs S (−17.4%). Patients in the L+P arm with larger % decreases in sums of target lesion diameters at PD had longer median post-progression survival (≤-61% [n=59], 35.6 mos [95% CI 28.4−39.2]; >-61%-≤-34% [n=58], 24.4 mos [95% CI 15.5−34.5]; >-34% [n=59], 20.2 mos [95% CI 16.4−26.9]) and OS (Grünwald 2024, 2025). In the L+P arm, most patients had low/intermediate risk PD (Table). Patients with low (HR 0.39; 95% CI 0.22-0.70) or intermediate risk (HR 0.47; 95% CI 0.27-0.81) PD had improved OS vs patients with high risk PD. Patients with low/intermediate risk PD stayed on 1 st subsequent anticancer medication longer than patients with high risk PD (Table). Conclusions: In the L+P arm, PD was mainly characterized by progression of existing lesions or formation of new lesions without progression of existing lesions (low/intermediate risk PD); these patients had improved OS (vs high risk PD) suggesting prognostic/predictive value of patterns of PD. Patients with low/intermediate risk PD stayed on 1 st subsequent anticancer medication longer than patients with high risk PD, implying that 2L therapy can be used effectively upon progression with L+P. These results, together with lower tumor burden seen at PD in the L+P arm (Grünwald 2024), indicate that L+P has robust tumor control, supporting its use as a standard 1L therapy in advanced renal cell carcinoma. Clinical trial information: NCT02811861 . Low risk PD(n=69) Intermediate risk PD(n=91) High risk PD(n=30) Patients with any subsequent systemic anticancer medication during survival follow-up, n (%) 49 (71.0) 58 (63.7) 15 (50.0) Time to discontinuation of 1 st anticancer medication during survival follow-up, median (Q1, Q3) (months) 14.4 (5.6, not estimable) 7.3 (3.1, 20.3) 4.3 (1.5, 13.7) Included data is from the L+P arm.

Real-world chemotherapy decision pathways and survival in early HR+/HER2- breast cancer: Oncotype DX recurrence score–guided versus clinician-guided treatment in NCDB (2010–2022).

e12754 Background: Despite widespread adoption of the 21-gene Recurrence Score (Oncotype DX RS) to guide adjuvant chemotherapy in early-stage HR+/HER2- breast cancer, real-world practice shows substantial chemotherapy use without genomic testing, relying on clinician judgment and clinicopathologic features. Prior studies show RS identifies low-risk patients who can safely omit chemotherapy, but comparative survival among chemotherapy-treated patients by RS-guided vs clinician-guided decisions remains underexplored. Methods: Using NCDB (2010–2022), we identified adult women with resected T1b–T2 N0 HR+/HER2- breast cancer who received adjuvant chemotherapy and endocrine therapy (excluding neoadjuvant therapy or missing key data). RS-guided chemotherapy (RS-CT) was defined as chemotherapy with RS > 26, and clinician-guided chemotherapy (Clin-CT) as chemotherapy based on clinician judgment without a documented RS record. Baseline differences were assessed by standardized mean differences (SMD). Overall survival (OS) was evaluated with Kaplan–Meier and stabilized inverse probability of treatment weighting (IPTW) with weighted Cox models. Absolute differences were quantified using IPCW-restricted mean survival time (RMST). Results: Among 59,247 patients (median follow-up 75.0 months), 28,349 (47.9%) were RS-guided and 30,898 (52.1%) were clinician-guided. IPTW achieved balance (post-weighting SMD < 0.10). RS-CT was associated with improved OS (adjusted HR: 0.927, 95% CI 0.878–0.979; p = 0.006). Benefit was observed in patients > 55 years among RS-CT group (HR: 0.870, 95% CI 0.816–0.928, p < 0.001; RMST +2.640 months), whereas the association was attenuated in those < 55 years (HR: 1.133, 95% CI 1.023–1.254, p = 0.016). By race, a significant benefit was seen only in Black patients (HR: 0.821, 95% CI 0.706–0.956; p = 0.011; RMST +3.554 months). Similar patterns of better survival in RS-CT group were observed across subgroups defined by T stage (lower T stage) and histologic grade (higher grade). Conclusions: In chemotherapy-treated early HR+/HER2- breast cancer, RS-guided selection was associated with superior survival vs clinician-guided decisions, with greater benefit in patients > 55 years and Black patients. These findings suggest that reliance solely on clinicopathologic features for chemotherapy decisions may not optimally identify biologic risk in the treated population, potentially exposing patients to chemotherapy toxicity without commensurate benefit, whereas the Oncotye (RS) based on genomic profiling may identify chemotherapy-eligible patients whose underlying biologic risk is not fully captured by clinicopathologic features and clinician judgment in routine practice.

Oncology social work outreach as a predictor of oncology appointment completion among cancer patients with transportation needs.

e13525 Background: Cancer patients typically have multiple treatment and evaluation appointments throughout their cancer therapy. Missing just one appointment could negatively impact or delay a patient’s treatment trajectory, and the challenges and impacts compound the more frequently a patient is scheduled for treatment 1 . Access to reliable transportation is imperative to support frequent oncology appointments and subsequently treatment adherence, yet it is the most prevalent barrier cancer patients face 2 . At our NCI-Designated Comprehensive Cancer Center, patients across seven treatment locations are assessed for transportation needs and other access barriers via a Health-related Social Needs screen (HRSNS) shortly after beginning cancer therapy. Oncology social workers (OSW) then outreach patients with identified needs. Transportation interventions by OSW may include transportation subsidies, travel vouchers, and rideshare programs. Methods: We examined the cancer patients with a transportation need identified on their HRSNS (n = 298) in 2025 and determined if an OSW outreached the patient to assist with transport needs. We then compared the average completion rate of scheduled oncology appointments (office visits, infusion visits, telehealth, procedures) among patients with a documented OSW outreach and those without. Completion rate is the percentage of completed appointments among all oncology appointments scheduled within the calendar year. Results: This subset of patients had between one and 149 scheduled oncology appointments in 2025, with an average of 20 scheduled appointments per patient. The results of the paired t-test analysis showed a significant difference between the mean values of appointment completion rate among patients with a social work outreach (n = 68, x̄=0.74) and those without (n = 230, x̄ =0.68), with a corresponding p-value of 0.046. Conclusions: Cancer transportation needs interfere with a cancer patient’s ability to keep oncology appointments, but interventions by OSW may increase a patient’s ability to keep scheduled oncology appointments.

Upfront ultrafast multigene panel testing for NCCN guideline–recommended biomarker detection in advanced NSCLC.

e15094 Background: NCCN Guidelines for advanced non–small cell lung cancer (NSCLC) recommend comprehensive biomarker assessment using multigene panel testing (MGPT) prior to initiating systemic therapy. Although rapid testing with subsequent MGPT is acknowledged in the guidelines, the optimal testing strategy remains undefined. Sequential single-gene testing (SGT) is associated with prolonged turnaround times (TAT), increased tissue consumption, and a higher risk of incomplete biomarker evaluation. Ultrafast next-generation sequencing (NGS)–based MGPT assays have emerged as potential upfront alternatives; however, the Centers for Medicare & Medicaid Services National Coverage Determination restricting reimbursement to a single NGS assay emphasizes the importance of initial test selection. Methods: Concordance between an ultrafast targeted MGPT (uMGPT; 50 genes) and a large MGPT (lMGPT; 523 genes) was evaluated for NCCN-recommended biomarkers with OncoKB Level 1 evidence. Archival clinical specimens from patients with advanced NSCLC (n = 104), previously analyzed by lMGPT, were retested using uMGPT. Concordance was assessed using positive and negative percent agreement. Laboratory TAT was defined as the time from specimen receipt to NGS result reporting following pathology review and nucleic acid extraction. Results: 37 NCCN-recommended biomarkers were identified across 104 samples. The uMGPT demonstrated complete concordance with lMGPT for the biomarkers tested, including 100% positive percent agreement for EGFR mutations (n = 14), KRAS G12C (n = 24), ALK fusions (n = 2), and MET exon 14 skipping alterations (n = 2), with 100% negative percent agreement. Median laboratory TAT for uMGPT was 1 day. Conclusions: uMGPT enabled rapid and comprehensive detection of the NCCN-recommended biomarkers evaluated in this study for advanced NSCLC while conserving limited tissue. Compared with sequential SGT, uMGPT enables simultaneous identification of multiple actionable alterations, reducing the risk of incomplete biomarker assessment. Although uMGPT does not encompass the full range of biomarkers assessed by lMGPT, it may function as a complementary rapid, upfront approach alongside lMGPT to identify therapeutically relevant alterations in NSCLC. Consideration of assay performance, operational feasibility, and clinical timelines may help guide optimal rapid biomarker testing strategies and support the integration of uMGPT into routine clinical workflows.

Evaluation of intravenous IPI201 in combination with anti–PD-1 therapy in a colorectal cancer mouse model.

2625 Background: Immune checkpoint inhibitors targeting PD-1 improve outcomes in colorectal cancer (CRC) but are limited by intrinsic and acquired resistance. Resorcinyl isoprenyl benzene derivatives have previously shown anti-tumor activity in multiple models, including remodeling of the tumor microenvironment (TME). IPI201 is a novel, synthetic resorcinyl isoprenyl benzene derivative administered intravenously. Here, we evaluated whether IPI201 combined with anti-PD-1 therapy alters tumor and immune outcomes in a murine model of CRC. Methods: MC38 colon tumors were implanted subcutaneously in mice and treated with vehicle, IPI201 (i.v.), anti-PD-1 (BE0156; i.p.), 5-fluorouracil (5-FU) (i.p.), or IPI201 plus anti-PD-1 at varying dose combinations once the implanted tumor reached a size of 75 mm 3 . Change in tumor volume and subject survival were assessed over 30 days. Tumor necrosis was evaluated by blinded histopathology. Tumor gene expression was analyzed using NanoString nCounter immune-focused panels with pathway and network analyses comparing combination therapy to monotherapies and vehicle controls. Results: While IPI201 or anti-PD-1 monotherapy demonstrated limited activity, combination treatment produced enhanced anti-tumor effects. The optimal dose combination resulted in a 66% reduction in tumor growth compared with anti-PD-1 alone, with a significant treatment interaction ( P =0.03). Combination therapy significantly improved survival relative to monotherapies and 5-FU (50% vs 0%, adjusted P <0.02), yielding the highest survival among treatment groups. Tumor necrosis scores were increased with combination therapy, with necrosis twofold higher than anti-PD-1 alone ( P <0.01). Transcriptomic profiling revealed differential regulation of 771 genes across 51 biological pathways, including VEGF, EGFR, MAPK/ERK, and PI3K/Akt/mTOR signaling, as well as immune-associated pathways related to antigen presentation, cytotoxic lymphocyte activation, and myeloid cell recruitment. This transcriptional signature is consistent with a pro-immune, anti-tumor TME. Conclusions: IPI201 and PD-1 blockade combination improves survival, promotes tumor necrosis, and remodels the tumor microenvironment in a preclinical CRC model. These findings support further development of IPI201 as a combination immunotherapy strategy and provide a mechanistic rationale for evaluating resorcinyl isoprenyl benzene-based enhancers of checkpoint inhibition in CRC.

Utilisation of the Henry score in malignant bowel obstruction and characterisation of management approaches and outcomes: Experience from a New Zealand regional centre.

12049 Background: Malignant Bowel Obstruction (MBO) is a common, life-threatening complication of advanced malignancy associated with a high symptom burden and poor prognosis. Henry et al developed the Henry Score to predict 30-day mortality in MBO patients. It has not been routinely used to guide clinical decision-making regarding the management of MBO or validated in populations outside the United States of America. This study was designed to assess the utility of the Henry Score in predicting 30-day mortality in palliative patients presenting to hospital with malignant bowel obstruction in a New Zealand regional centre and to characterise the management approaches and outcomes of surgical versus non-surgical interventions. Methods: Retrospective cohort study of 235 patients admitted with MBO to Palmerston North Hospital (2010-2024). The Henry Score (0-5) incorporating five clinical and laboratory factors was calculated for each patient. Management approaches and survival outcomes were analysed. Results: Complete Henry Score data was available for 82.1% of patients. Sixty-four underwent surgical management, 171 non-surgical interventions. Median overall survival was 82.5 days (surgical: 130.5 days; non-surgical: 52 days; p=0.005). The 30-day mortality rate was 33.2%. The Henry Score demonstrated prognostic discrimination, with 30-day mortality of 28.0% for low scores (0-2) vs 43.0% for high scores (3-5) (p=0.034), and corresponding median survivals of 83 vs 39 days. Cox analysis identified hypoalbuminemia as the only factor significantly associated with mortality risk (HR 1.67, p=0.001). Māori patients had higher Henry Scores (68.2% vs 41.9% with scores ≥3, p=0.025) and shorter median survival (37 vs 97.5 days) than New Zealand European patients. Conclusions: The Henry Score effectively predicts 30-day mortality in MBO patients in New Zealand. Higher scores were associated with increased 30-day mortality, supporting the potential use of this measure for prognostication and advanced care planning. Hypoalbuminemia was the strongest predictor of mortality. Future prospective studies are needed to determine if clinical implementation of the Henry Score improves patient outcomes.

The 2026 primary care oncology toolkit: An evolving shared-care framework for the implementation of prospective survivorship evidence.

11076 Background: As cancer survivorship transitions into a chronic disease model, a significant "implementation gap" exists between landmark 2025 clinical data and frontline primary care. Survivors face a 2.39-fold higher risk of cardiovascular mortality, yet standardized protocols for PCP-led surveillance remain in development. This review identifies an evolving "Shared-Care Framework" designed to integrate prospective 2025 evidence into the internal medicine setting, anticipating a shift toward decentralized, longitudinal survivorship care. Methods: A narrative synthesis was performed to develop a pilot-ready framework for primary care. We analyzed practice-changing evidence published between January 2025 and January 2026, including the Barac Risk Model, the Bhalraam et al. SGLT2i meta-analysis , and emerging cervical screening strategies. Findings were synthesized into thematic surveillance blocks to identify opportunities for prospective guideline-concordant implementation. Results: Synthesis of recent landmark data identified several opportunities for prospective adoption in the internal medicine setting. The framework evaluates evolving cardio-metabolic surveillance strategies, including the integration of the Barac Model (2025) for 10-year heart failure (HF) risk prediction and the proactive use of SGLT2 inhibitors, which have demonstrated a 51% reduction in HF hospitalizations among survivors. Furthermore, the review identifies prospective clinical blocks for PCP-led surveillance of stable survivors, emphasizing proactive bone health monitoring for patients on aromatase inhibitors and metabolic screening for those on androgen deprivation therapy (ADT). This includes the potential adoption of low-dose oxybutynin as a new standard for managing ADT-induced hot flashes. Finally, the framework addresses health equity paradigms by evaluating the transition to in-office HPV self-sampling and automated EMR alerts for age 45 colorectal screening to mitigate the rising incidence of early-onset disease. Conclusions: This evolving shared-care framework offers a scalable model for the future of survivorship. By aligning primary care protocols with prospective 2025 landmark data, clinicians can optimize long-term coordination and address health disparities ahead of formal guideline updates.

Expression of Concern: Deciphering Transcriptional Programming during Pod and Seed Development Using RNA-Seq in Pigeonpea (Cajanus cajan)

Plant-mediated synthesis of Ag/AgO nanocomposites: Physicochemical characterization and photocatalytic remediation of petroleum effluents

A Low‐Voltage Multi‐Band Tunable Smart Window for Self‐Adaptive Thermoregulation

ABSTRACT As a crucial smart window technology, polymer‐dispersed liquid crystal (PDLC) faces two major performance limitations in applications: high driving voltage and limited spectral regulation. Herein, we develop a PDLC‐based smart window that operates at low voltage and enables multi‐band modulation. By incorporating p–n heterojunction nanoparticles, the electro‐optical performance of the PDLC film is significantly enhanced. Specifically, a 35.1% reduction in saturation voltage (to 16.3 V) and an increased contrast ratio of 130 were achieved at a film thickness of 20 µm. Subsequent integration of W‐VO 2 /PMMA resulted in a V‐PDLC smart window. The resulting device demonstrates switchable long‐wave infrared emissivity (ε LWIR = 0.75 at high temperature; ε LWIR = 0.41 at low temperature), enabling passive radiative cooling while maintaining high visible light transmittance and efficient near‐infrared modulation. Field test and energy simulations confirmed that the system not only provides effective temperature regulation (approximately 8°C cooling during the day and 2°C insulation at night) but also demonstrates significant energy‐saving potential across various climatic zones. This research offers valuable insights for developing smart windows with adaptive and on‐demand adjustment capabilities.

A deep learning approach for keratoconus detection using spatio-temporal features from corneal imaging

Abstract Keratoconus is a progressive corneal disease that requires early and accurate detection to prevent severe visual impairment. This study presents a deep learning-based classification model for distinguishing between healthy and keratoconic eyes using dynamic corneal imaging data from the CORVIS system. A hybrid CNN-RNN architecture was developed, combining a fine-tuned InceptionV3 network for spatial feature extraction with a recurrent LSTM module to capture temporal patterns across image sequences. To ensure robust evaluation, a 10-fold stratified cross-validation strategy was employed, with data splits performed at the patient level to avoid data leakage. The model achieved an average accuracy, precision, recall, and F1-score of approximately 0.90 across folds, demonstrating strong generalization performance. Boxplot visualizations of metric distributions further confirmed model stability and revealed minimal performance variance. Class-wise analysis showed high effectiveness in detecting both healthy and keratoconic cases, although slightly greater variability was observed in the classification of healthy eyes. These results indicate that the proposed method is a promising tool for keratoconus screening and may complement existing diagnostic workflows. Further validation on external datasets is recommended prior to clinical deployment.

Zinc-induced exposure of an LxL motif drives clathrin-mediated endocytosis of the zinc transceptor ZIP4