Citations

3593 Background: CUB domain-containing protein 1 (CDCP1) is a transmembrane protein shown to be over-expressed in colorectal cancer (CRC), with an important role in tumor initiation, growth and metastasis. Through interactions with EGFR, CDCP1 promotes nuclear translocation of the key regulators of Wnt signaling, β-catenin and E-cadherin. High CDCP1 expression has been associated with poor survival outcomes in multiple solid tumors but data in CRC remains limited. We investigated associations between CDCP1 level and survival outcomes in patients with metastatic CRC (mCRC) in the phase III clinical trial CALGB/SWOG 80405 (Alliance). Methods: Data from 433 mCRC patients in the CALGB/SWOG 80405 (Alliance) trial were analyzed. Patients received bevacizumab (n = 226) or cetuximab (n = 207) plus chemotherapy as 1st-line treatment. Tumor RNA was extracted from FFPE samples and processed through the HiSeq 2500 (Illumina) platform. Overall survival (OS) and progression free survival (PFS) curves were evaluated in all treatment subgroups and stratified according to high or low CDCP1 expression. Likelihood ratio tests, hazard ratios (HRs), and 95% confidence intervals (CIs) were calculated using Cox proportional hazards multivariate models, adjusting for age, sex, ECOG performance status, tumor location, number of metastatic sites, KRAS status, consensus molecular subtypes, and treatment arm. Results: Overall, low CDCP1 expression was associated with better OS compared to high expression (HR = 1.38; 95% CI 1.07-1.77; 35.8 vs 25.0 months, p = 0.045). Subgroup analysis re-demonstrated this pattern in males (HR = 1.64; 95% CI 1.64-2.28; 35.9 vs 23.6 months, p = 0.01), patients with MSS tumors (HR = 1.56; 95% CI 1.17-2.08; 36.7 vs 25.1 months, p = 0.0094), and patients treated with cetuximab+FOLFOX (HR = 1.79; 95% CI 1.16-2.77; 40.8 vs 24.0 months, p = 0.03). The effect of CDCP1 expression on PFS was less pronounced. Relative to high expression, low expression was associated with better PFS in males (HR = 1.49; 95% CI 1.10-2.03; 13.1 vs 9.8 months, p = 0.027) and left-sided tumors (HR = 1.50; 95% CI 1.10-2.04; 14.4 vs 10.1 months, p = 0.034). Conclusions: Our findings represent real-world data of the predictive role of CDCP1 in patients with mCRC undergoing anti-EGFR or anti-VEGF treatment. Consistent with studies of other cancers, we show that low CDCP1 expression is predictive of better survival outcomes, with subgroup analysis revealing associations in males, MSS tumors, and left-sided tumors. Among treatment groups, low CDCP1 expression is predictive of better OS in cetuximab+FOLFOX-treated patients, a key finding given CDCP1’s close mechanistic ties to EGFR. Further studies should explore whether these associations are seen in other treatment modalities and whether CDCP1 blockade could represent a viable strategy in patients with high expression.

1673 Background: Survivorship care plans (SCPs) summarize cancer treatments and provide evidence-based recommendations for surveillance, screening, and management of treatment-related complications. Despite the availability of guidelines and templates, SCPs remain underutilized in routine practice due to the time-intensive and error-prone nature of manual creation. Empowering cancer survivors to use publicly available large language models (LLM)–based chatbots, such as ChatGPT, may offer a feasible solution to this problem. Methods: Standardized patient profiles were developed for six common cancers: breast, colorectal, non–small cell lung, small cell lung, prostate cancer, and diffuse large B-cell lymphoma. Profiles were reviewed by a primary care physician and an oncologist to ensure clinical accuracy and representativeness of common survivorship scenarios. Using lay language to simulate real-world patient input, ChatGPT (version 5.2, publicly available at the time of the study) was prompted with each profile using the question, “What follow-up care do I need?” Common survivorship symptoms for each cancer were also tested. Responses were evaluated across five domains: medical accuracy (concordance with current ASCO or NCCN guidelines), thoroughness, patient safety and risk framing, clarity and patient comprehension, and actionability. Five-point Likert scales were used (1 = not at all; 5 = very much). Scores were averaged across two independent physician raters. Readability was assessed using the Flesch Reading Ease Score. Results: ChatGPT-generated SCPs demonstrated high medical accuracy (mean score 4.25/5), with no factual errors or hallucinations identified on manual review. Patient safety and risk framing scored 3.75/5; symptom red flags and care escalation guidance were appropriately stated in most scenarios. Thoroughness was moderate (3.50/5), with key survivorship elements, such as smoking cessation counseling or genetic risk considerations; sometimes omitted unless explicitly prompted. Readability was limited, with a mean Flesch Reading Ease Score of 40.5, corresponding to a college reading level and exceeding recommended readability standards for patient education materials. Clarity (3.38/5) and actionability (3.38/5) were similarly constrained due to dense medical language. Conclusions: ChatGPT demonstrated high medical accuracy in generating survivorship care plans across six common cancers, supporting the feasibility of using publicly available LLMs to guide survivorship care. However, improvements in comprehensiveness, readability, and patient-centered actionability are necessary before such tools can be safely integrated into clinical survivorship workflows.

1110 Background: CDK4/6 inhibitors with endocrine therapy are widely used for HR+/HER2– metastatic breast cancer (BC). While studies have shown similar progression-free outcomes across agents, overall survival (OS) data are limited. Since OS is a definitive endpoint and BMI may influence outcomes, this study examines BMI-dependent OS differences among CDK4/6 inhibitors. Methods: A retrospective cohort study was conducted using the Mayo Clinic Platform full AMC1 dataset. Postmenopausal women with BC receiving palbociclib/ribociclib/abemaciclib plus aromatase inhibitor (AI) were included. Patients with prior use of fulvestrant, anthracyclines, radiation therapy, GnRH agonists, or history of MACE, heart failure, myocardial infarction, or stroke before BC diagnosis were excluded. Patients were stratified by BMI and CDK4/6 inhibitor. The primary outcome was 5-year OS. Results: Among 1,989 patients (palbociclib: 868; ribociclib: 465; abemaciclib: 656), palbociclib was associated with significantly worse OS compared to ribociclib in BMI <25 (HR 2.91; 95% CI: 1.76–4.79; P < 0.01) and BMI ≥30 (HR 1.91; 95% CI: 1.21–3.01; P < 0.01). Compared to abemaciclib, palbociclib showed worse OS: BMI <25 (HR 1.73; P < 0.01), BMI 25–30 (HR 1.56; P < 0.01), and BMI ≥30 (HR 1.82; P < 0.01). No significant OS differences were observed between ribociclib and abemaciclib. Palbociclib patients were older, particularly in BMI <25 (Cohen's D = 0.39 vs. ribociclib; 0.27 vs. abemaciclib) and BMI ≥30 (0.28 vs. ribociclib; 0.23 vs. abemaciclib). In the BMI 25–30 group, palbociclib group was also older than ribociclib group (Cohen's D = 0.26). Conclusions: In postmenopausal BC patients on AIs and CDK4/6 inhibitors, palbociclib was associated with worse OS in BMI < 25 and BMI ≥ 30 groups, while ribociclib and abemaciclib showed similar outcomes across BMI groups. Palbociclib patients were older than those on other drugs, which may partially confound results; however, in the BMI 25–30 group, despite similar age differences, no survival difference was observed between palbociclib and ribociclib, suggesting age alone may have limited impact. These findings indicate a potential interaction between BMI and drug choice, and highlight the need for prospective studies to confirm and adjust for these variables. Limitations include the observational design and inability to perform multivariable regression analyses to adjust for confounders. Overall survival comparisons by CDK4/6 inhibitor and BMI subgroup. Comparison BMI Group Hazard Ratio (95% CI) P-value Palbociclib vs Ribociclib <25 2.91 (1.76 – 4.79) < 0.01 25 – 30 1.34 (0.87 – 2.071) 0.17 ≥30 1.91 (1.21 – 3.01) < 0.01 Palbociclib vs Abemaciclib <25 1.73 (1.29 – 2.32) < 0.01 25 – 30 1.56 (1.09 – 2.23) < 0.01 ≥30 1.82 (1.15 – 2.88) < 0.01 Ribociclib vs Abemaciclib <25 1.27 (0.81 – 1.99) 0.29 25 – 30 0.87 (0.54 – 1.43) 0.59 ≥30 1.05 (0.63 – 1.75) 0.85

10585 Background: Prior studies suggest that thyroid cancer (TC) survivors may have an elevated risk of secondary malignancies. However, whether this risk differs by TC histology, specifically with breast cancer (BC), remains unclear. We aimed to investigate the association between TC subtype—papillary thyroid carcinoma (PTC) versus medullary thyroid carcinoma (MTC)—and subsequent BC risk, adjusting for demographic factors. Methods: We conducted a retrospective cohort study using the SEER (2000–2022), including adults with a primary, microscopically confirmed PTC or MTC. Subsequent BC occurrence was defined as a binary outcome (yes/no). Multivariable logistic regression was used to estimate adjusted odds ratios (aORs) for MTC versus PTC, controlling for age (<50 vs ≥50 years), sex, and race. Stratified analyses by age and race were also performed. Stata 15 was used for statistical analyses. Results: 14,840 patients (14,563 PTC, 277 MTC) were included. MTC patients were older (mean 59.08±12.66 vs 56.16±12.98 years, p<0.001) and had a higher proportion of males (44.4% vs 29.4%, p<0.001). The overall incidence of secondary BC was lower in MTC patients (7.6% vs 22.5%, p<0.001). In unadjusted logistic regression, MTC was associated with significantly lower BC risk (OR=0.28, 95% CI: 0.18–0.44, p<0.001). After adjustment, MTC remained significantly protective (aOR=0.34, 95% CI: 0.22–0.55, p<0.001). Stratified analyses demonstrated consistent findings across age groups (<50: aOR=0.10, 95% CI: 0.03–0.42; ≥50: aOR=0.34, 95% CI: 0.21–0.54), , as well as white and Hispanic patients. Conclusions: Survivors of MTC have a substantially lower risk of developing secondary BC compared to PTC survivors. This association persists after adjustment for age, sex, and race, and is consistent across age and most racial subgroups. However, different BC types were not assessed due to limited number of BC patients among MTC survivors, and the small MTC sample size and potential survival differences between groups may affect the results. These findings highlight a possible effect of TC subtype on BC risk, suggesting the potential influence of underlying biological or genetic mechanisms. Baseline demographic characteristics of patients with papillary and medullary thyroid carcinoma. Characteristic Papillary Thyroid Carcinoma (PTC) (N=14,563) Medullary Thyroid Carcinoma (MTC) (N=277) P-value Age at Diagnosis 0.0002* Mean ± SD, years 56.16 ± 12.98 59.08 ± 12.66 Median, years 57 61 Sex, n (%) <0.001† Female 10,288 (70.6) 154 (55.6) Male 4,275 (29.4) 123 (44.4) Race, n (%) 0.015† White 10,263 (70.5) 197 (71.1) Black 921 (6.3) 15 (5.4) American Indian 77 (0.5) 1 (0.4) Asian 1,379 (9.5) 13 (4.7) Hispanic 1,903 (13.1) 50 (18.1) SD: Standard Deviation. *P-value from two-sample t-test (equal variances assumed). †P-value from chi-square test for the overall distribution across categories.

1056 Background: Resistance to HER2-directed therapies remains a major challenge in HER2-positiveMBC. In preclinical models, HER2+ tumors overexpress poly (ADP-ribose) polymerase (PARP), and PARP inhibition induces apoptosis through NF-κB suppression even in the absence of pathogenic mutations in BRCA 1/2 . We evaluated the safety, efficacy, and translational correlates of niraparib combined with trastuzumab in patients (pts) with refractory HER2+ MBC. Methods: This open-label Phase 1b/2 trial (NCT03368729) enrolled pts with HER2+ MBC (IHC 3+ and/or FISH amplified), measurable disease, LVEF ≥50%, and progression on ≥1 prior HER2-targeted therapy; stable CNS disease was allowed. Pts with germline BRCA1/2 mutations or prior PARP inhibitor exposure were excluded. Niraparib 300 mg orally was administered on Days 1–21 of a 21-day cycle with trastuzumab 6 mg/kg IV every 3 weeks (loading dose 8 mg/kg on cycle 1). The Phase 2 portion used a Gehan/Simon two-stage design with early stopping for futility (P0=24%); ≥1 response among the first 20 patients triggered expansion to 40 patients. ORR was assessed by RECIST v1.1, meeting the primary endpoint and allowing response rate estimation with <7% standard error and a two-sided 95% exact CI width of 0.28. Gene expression profiling of paired tumor tissue and peripheral blood mononuclear cells (PBMCs) was performed using the NanoString nCounter Tumor Signaling 360 panel at pretreatment compared to cycle 2. Results: Forty-six pts were enrolled (9/2019–2/2025); 41 were evaluable for response. Pts were heavily pretreated (median 3 prior regimens), including trastuzumab deruxtecan in 22 pts; 33% had brain metastases. Treatment-related adverse events were predominantly grade 1–2 and manageable. Clinical benefit rate (CBR) was 44% (18/41) including 1 complete response, 2 partial responses, and 15 stable disease. Eleven patients achieved disease control ≥ 6 months; 1 patient remained on treatment at data cutoff. Tumor profiling revealed enrichment of a B-cell–driven tertiary lymphoid structure signature with reduced hypoxia signaling in responders. Matched PBMC analyses demonstrated coordinated downregulation of interferon signaling, DNA damage response, metabolic activation, angiogenesis, and resistance pathways in clinical responders, whereas non-responders retained signatures of chronic stress and immune exhaustion. Conclusions: Niraparib plus trastuzumab demonstrated a favorable safety profile and durable clinical benefit in heavily pretreated HER2+ MBC. Integrated tumor and blood-based analyses identify immune-organized tumor microenvironments and systemic immune normalization as key correlates of prolonged response, supporting further investigation of PARP inhibition strategies. Clinical trial information: NCT03368729 .

BACKGROUND: Atherothrombosis, which underlies most acute coronary syndromes and is driven by intraplaque thrombosis, preferentially occurs in regions of disturbed blood flow (d-flow). Although LATS1/2 (large tumor suppressor kinases 1 and 2) are known regulators of endothelial mechanotransduction, the mechanisms by which d-flow connects endothelial senescence, proliferation, and intraplaque thrombosis remain poorly understood. METHODS: We investigated endothelial cell (EC)-specific roles of Lats1/2 using inducible EC-specific knockout mice in a partial carotid ligation model. Spatial multiomics of human and mouse plaques was performed using imaging mass cytometry, COMET sequential immunofluorescence, and spatial metabolomics. RESULTS: Tamoxifen-induced deletion of both Lats1 and Lats2 (homozygous) in ECs caused fatal edema and increased vascular permeability. In contrast, Lats1 het ( ± ) / Lats2 homo ( −/−) -EC-specific knockout mice survived and developed spontaneous atherothrombotic plaques with neovascularization. Spatial proteomics revealed that LATS1/2 loss induced a senescence-associated stemness phenotype driven by CD38 upregulation. Spatial metabolomics showed sulfite and taurine accumulation, indicating SUOX (sulfite oxidase) deficiency. CD38 suppressed SUOX, demonstrated a switch into the reverse mode of mitochondrial complex V, increased succinate dehydrogenase activity, and promoted ATP consumption. Despite ATP depletion, glutamate metabolism and the citric acid cycle flux increased, sustaining EC proliferation under energetic stress. This senescence-associated stemness state promoted both proliferation and senescence, leading to fragile, leaky neovessels and intraplaque thrombotic lesions. Pharmacological CD38 inhibition attenuated these phenotypes. Similar EC states were observed in human plaques. CONCLUSIONS: Loss of Lats1/2 in ECs induces a CD38-associated senescence-associated stemness-like phenotype that promotes intraplaque thrombosis through mitochondrial metabolic reprogramming, including changes consistent with reverse-mode operation of mitochondrial complex V. These findings define a mechanistic link between disturbed flow, endothelial metabolic reprogramming, and intraplaque thrombosis and hemorrhage.

Abstract Background Accurate and efficient dose calculation is essential for online adaptive planning in proton therapy. Deep learning (DL) has shown promising dose prediction results for pencil beam scanning proton therapy (PBSPT) in recent years, but existing DL‐based dose prediction methods still suffer from limited generalizability and an inability to effectively handle outlier clinical cases. This may lead to inaccurate dose delivery to targets or excessive irradiation to organs at risk (OARs), thereby compromising the safety and efficacy of online adaptive proton therapy. Purpose To design a physics‐aware and generalizable AI‐based PBSPT dose prediction method that incorporates underlying physics to enhance generalizability, particularly in handling outlier clinical cases. Methods This study analyzed PBSPT plans of 103 prostate (93 for training and 10 for testing) and 78 lung cancer patients (68 for training and 10 for testing) from our institution, with each case comprising CT images and structure sets. Using the doses generated by our Monte Carlo‐based dose engine as the reference standard, we compared three methods: the region of interest (ROI)‐based method, the beam mask and sliding window method, and the proposed noisy probing dose method, which rapidly generates a low‐statistics dose via uniformly weighted spots on an expanded spot‐placement target volume without optimization. To evaluate the generalizability of these methods to rare treatment planning scenarios, 12 cases with uncommon beam angles or prescription doses were used to assess their performance, which was evaluated using dose‐volume histogram (DVH) indices, 3D Gamma passing rates (3%/2 mm/10%), and Dice coefficients for dose agreement, while prediction times were measured to gauge model efficiency. Results The proposed noisy probing dose method consistently outperformed the ROI‐based and beam mask baselines across all evaluated metrics, with more accurate dose agreement and superior generalizability. For DVH indices, the noisy probing dose method achieved the smallest deviation in clinical target volume (CTV) dose coverage: in prostate cancer, CTV D98 deviation was reduced by 45% (from 0.53 ± 0.22 Gy [RBE] for ROI‐based) and 29% (from 0.41 ± 0.28 Gy [RBE] for beam mask) to 0.29 ± 0.06 Gy [RBE]; in lung cancer, similar improvements were observed, with CTV D98 deviations reduced to 0.34 ± 0.12 Gy [RBE]. The 3D Gamma passing rates improved to 99.65% ± 1.15% for prostate targets and 97.04% ± 1.17% for lung targets. The dice coefficients of the 90% iso‐dose lines were also the highest with the noisy probing dose method (prostate: 0.983 ± 0.005; lung: 0.967 ± 0.01). For the 12 outlier cases, the noisy probing dose method maintained superior generalizability, yielding higher 3D Gamma passing rates (prostate targets: 96.79% ± 0.83%, OARs: 94.29% ± 1.01%; lung targets: 93.38% ± 1.34%, OARs: 93.95% ± 1.32%), demonstrating robust generalizability to rare clinical scenarios. The dose predictions for all testing cases were completed within 0.3 seconds. Conclusions A novel noisy probing dose method was proposed for PBSPT dose prediction in prostate and lung cancer patients. By embedding more proton‐specific physics, this method demonstrated an improvement in the generalizability of dose prediction.

ABSTRACT Leptomeningeal disease (LMD) is a rapidly fatal complication of systemic cancer for which sensitive diagnostic tools and informative biomarkers remain limited. Here, we introduce CSF-Seq, a method for whole-transcriptome sequencing of cell-free RNA (cfRNA) from human cerebrospinal fluid (CSF), designed to enable molecular profiling of LMD and other central nervous system (CNS) conditions. Using a prospectively collected CSF biobank, we analyzed 125 samples spanning multiple pathologies, including breast and lung LMD, glioblastoma, traumatic brain injury, and non-cancer neurological controls. Through optimized RNA extraction, library preparation, and deep sequencing, CSF-Seq generated robust and reproducible transcriptome-wide profiles despite the low abundance and fragmentation of cfRNA in CSF. CSF transcriptomes exhibited disease-specific expression, separating LMD from non-cancer controls and from non-LMD cancers, independent of CSF collection modality. Tumor-associated epithelial transcripts, including CEACAM6 and MUC1, were consistently enriched in LMD samples, whereas immune and CNS-associated transcripts were broadly detected across disease states, consistent with contributions from both tumor and non-tumor sources. Cross-site processing of matched samples demonstrated high concordance, indicating preservation of sample-specific transcriptional signatures across independent workflows. Importantly, we identified a collection method– independent LMD gene expression signature that was significantly associated with overall survival, supporting its potential prognostic relevance. Together, these findings establish CSF-Seq as a technically robust and clinically informative platform for transcriptomic biomarker discovery in CNS metastatic disease, offering a minimally invasive approach for disease characterization, risk stratification, and longitudinal monitoring in patients with LMD.

ABSTRACT Richter transformation of Chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL) into classic Hodgkin lymphoma (CHL-RT) is rare and remains incompletely understood. Two histologic subtypes are recognized: type 1 (CLL/SLL with scattered Hodgkin/Reed-Sternberg (HRS) cells) and type 2 (HRS cells within a polymorphous inflammatory background). In this multi-institutional study of 77 patients with CHL-RT (27 type 1 and 50 type 2), we characterized immune evasion markers, PD-L1 / PD-L2 copy number alterations, tumor microenvironment, and performed targeted next-generation sequencing on 37 CLL/SLL samples. HRS cells in CHL-RT displayed immune-evasion phenotypes similar to de novo CHL, though PD-L1 expression was lower in type 1 cases. PD-L1/PD-L2 gain/polysomy were frequent (83.3%). CLL/SLL with CHL-RT harbored increased mutations in XPO1, FBXW7, BIRC3, TRAF3 , and HLA-A versus reference CLL/SLL. Similar mutational profiles, demographics, and survival outcomes support a biological continuum between type 1 and type 2 CHL-RT, with distinct genetic features in CLL/SLL predisposing to CHL transformation. STATEMENT OF SIGNIFICANCE Classic Hodgkin lymphoma-type Richter transformation arises from convergent Hodgkin-like immune evasion on a genetically primed CLL/SLL background. Through integrated genomic and tumor microenvironment analyses in a large multicenter cohort, we establish this as a biologically coherent entity and provide rationale for PD-L1/PD-L2-directed therapeutic strategies.

Abstract Heterogeneous treatment effects (HTE) are key to precision medicine, but most real-world studies lack the scale and diversity needed to detect them. While multi-site analyses offer a potential solution, data-sharing constraints often prevent access to patient-level information across institutions. We introduce POLARIS, a federated framework for time-to-event target trial emulation. POLARIS converts each site’s weighted Cox risk function into a compact tensor shared once with the coordinating center, enabling lossless reproduction of pooled estimates without sharing patient-level data. We applied POLARIS across five U.S. health systems to study risk of gastrointestinal outcomes after GLP-1 receptor agonist (GLP-1RAs) initiation versus sodium-glucose cotransporter 2 inhibitors (SGLT2is) and dipeptidyl peptidase 4 inhibitors (DPP4is). Results showed that GLP-1RAs were consistently associated with higher risks of nausea and vomiting, particularly among men, individuals with higher baseline HbA1c (≥ 8.5%), and lipid therapy. POLARIS provides a scalable solution for distributed target trial emulation and fine-grained assessment of HTE across diverse health systems.