Certain rare cancers such as ovarian or pancreatic cancer would benefit if detected early at a stage when they are resectable. Unfortunately, approved biomarkers for these cancers are not adequate for screening the general population, and it is unlikely that a single marker will meet the performance criteria for screening. Determining a combination of biomarkers for early detection of rare cancers is a challenge. Often model selection suffers from overfitting in the discovery phase, which leads to poor performance upon validation. Since ovarian cancer has a poor prognosis, we aim to identify biomarkers that perform robustly in early cancer detection discovery and validation phases. Stability selection methods have been used to prevent overfitting and to reliably select truly expressed biomarkers. Ensemble learning methods provide robust prediction results in the face of model misspecification. We present a novel framework with a biomarker selection stage with stability selection and prediction stage using an ensemble of machine learning (ML) methods, namely the stability selection ensemble learning (STABEL). The ensemble consists of random forest (RF), logistic regression (LR), linear discriminant analysis (LDA), and support vector machine (SVM). Simulated data, along with ovarian cancer data from Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trials are used to test the methodology. Our results demonstrate the power of integrating ensemble learning with stability selection to select cancer biomarkers and achieve strong predictive performance. The STABEL technique outperforms or performs similarly to the existing methods in terms of sensitivity, specificity, prediction accuracy, and area under the curve.

BACKGROUND:Pancreatic ductal adenocarcinoma (PDAC) surveillance is recommended for high-risk individuals (HRI) carrying pathogenic variants (PV) in PDAC susceptibility genes or meeting criteria for familial pancreatic cancer (FPC). Cigarette smoking drives earlier PDAC onset in general populations, but this effect is not well-characterized among HRIs. We hypothesized that smoking accelerates the age at PDAC diagnosis for HRIs. METHODS:We identified PDAC patients enrolled in the Pancreatic Adenocarcinoma Gene Environment Risk (PAGER) study at the University of Pittsburgh who underwent multi-gene panel testing. Chi-square tests, t-tests, and analysis of variance were used to assess relationships between smoking history and PDAC diagnosis age. RESULTS:Of 3,602 patients with PDAC, 1,266 met eligibility criteria. 148 carried PVs in PDAC susceptibility genes, 154 had FPC, and 964 had sporadic PDAC. Among PV carriers, current smokers were diagnosed 4.1 years earlier than never smokers (P = 0.16). Compared to sporadic patients, PV carriers were diagnosed 1.6 years earlier (P = 0.55) among current smokers, 3.4 years earlier (P = 0.07) among past smokers, and 3.8 years earlier (P = 0.01) among never smokers. PV carriers who had quit smoking more than 10 years prior were diagnosed 9.0 years later than current smokers (P = 0.01) while those who had quit smoking within 10 years had no significant difference in diagnosis age. CONCLUSIONS:Cigarette smoking is associated with earlier development of PDAC among PV carriers, but smoking cessation of 10 years or more mitigates this effect. If confirmed in larger studies, these findings suggest smoking history should factor into the age at which PDAC surveillance begins.

Background: Although physical activity (PA) offers substantial physical and psychosocial benefits, engagement remains suboptimal among cancer survivors. A theory-informed understanding of survivors' perceived barriers, facilitators, and recommendations is needed to inform patient-centered PA about survivorship interventions. Objective: This study aimed to explore perceived barriers, facilitators, and recommendations for PA engagement among adult cancer survivors using the Theoretical Domains Framework (TDF). Methods: A phenomenological qualitative design was used. Eighteen cancer survivors from Nebraska participated in semi-structured interviews via Zoom or telephone. Semi-structured interviews (guided by open-ended questions with flexibility for probing) were transcribed verbatim, imported into MAXQDA 2024, and analyzed using TDF to identify themes and subthemes. Results: Three overarching themes emerged: barriers, facilitators, and recommendations related to PA engagement. Barriers included individual factors (low motivation and self-efficacy, limited awareness of PA guidelines, time constraints, and physical limitations due to treatment and comorbidities), social factors (limited support from family, friends), clinical factors (limited PA guidance from healthcare providers), and environmental factors (restricted access to resources and unfavorable weather). Facilitators included individual factors (PA knowledge, motivation, goals, and health benefits), social factors (support from family, friends), and clinical factors (encouragement from healthcare providers), and environmental factors (favorable weather and available community PA resources). Recommendations emphasized the need for tailored education, supportive counseling, and structured PA programs within survivorship care. Conclusions: Cancer survivors described multilevel determinants of PA engagement across individual, social, and environmental contexts. Findings highlight the importance of theory-informed, patient-centered strategies that enhance PA guideline awareness, strengthen social and clinical support, and improve access to community resources to promote sustained PA during cancer survivorship.

Supplementary methods, tables and figures

This study examined national trends in adherence to aerobic physical activity (PA) guidelines and identified demographic, socioeconomic, behavioral, and health-related factors associated with aerobic PA adherence among adult cancer survivors in the United States. A repeated cross-sectional analysis was conducted using data from the National Health Interview Survey (NHIS) 2010, 2015, and 2020 cycles. The analytic sample included 9,752 adult cancer survivors. Adherence to aerobic PA guidelines was defined as engaging in ≥ 150 min/week of moderate-intensity activity, ≥ 75 min/week of vigorous-intensity activity, or an equivalent combination. Survey-weighted descriptive analyses and multivariable logistic regression models were conducted to estimate prevalence and adjusted odds ratios (aORs) with 95

Ca2+-dependent repair of plasma membrane breaches is essential for animal cell viability. An initial passive influx of extracellular Ca2+ triggers the formation of a protein plug that rapidly seals breaches. However, the mechanism of extracellular Ca2+ requirement for subsequent repair remains undefined. EHD2 protein stabilizes the plasma membrane caveolae, which sustain membrane repair, and maintains high surface levels of the caveolae-resident Ca2+ channel Orai1. We establish the requirement of both Orai1 and EHD2 for repair of plasma membrane lesions induced by mechanical injury or by a model bacterial pore-forming toxin. We demonstrate rapid EHD2 recruitment and Orai1-mediated Ca2+ entry at plasma membrane sites of localized mechanical stimulus, the latter requiring EHD2 and CAV1. EHD2 and Orai1 are necessary for mechanosensitive YAP/TAZ-TEAD activation and positive feedback for CAV1 expression that promotes membrane repair. Our studies establish EHD2 and Orai1 as novel components of mammalian plasma membrane repair and mechanoadaptation.

PURPOSE:The Pancreatic Cancer Detection Consortium (PCDC) performed a blinded Early Detection Research Network-defined phase II biomarker bakeoff study of blood-based biomarker panels. The aims were to evaluate panel performance, to compare the panels' performance with that of cancer antigen 19-9 (CA19-9) alone, and to evaluate the performance of new combinations of the individual biomarkers. EXPERIMENTAL DESIGN:Ten biomarkers representing eight biomarker panels and CA19-9 were evaluated using plasma, serum, and germline DNA from 140 stage I to IV pancreatic ductal adenocarcinoma (PDAC) cases and 140 controls from three tertiary care institutions, with controls frequency matched to cases on age and sex. LASSO regression was employed to explore new biomarker combinations. The primary metric was area under the receiver operating characteristic curve (AUC). RESULTS:The study population was 51% female, with median age 67.3 (minimum: 45, maximum: 90) years. Biomarker panel AUCs ranged from 89.9 to 96.3; the AUC for serum CA19-9 alone was 91.7 [95% confidence interval (CI), 87.8-95.6]. Two panels had significantly higher AUCs than serum CA19-9 alone, the CA19-9/FUT2/3 panel (AUC = 96.3, P = 0.002), and the tissue factor pathway inhibitor/tenascin C (TFPI/TNC-FNIII-C) panel (AUC = 95, P = 0.01). Exploratory models to recombine biomarkers retained all but two biomarkers [optimism-corrected AUC = 96.4 (94-98.9)]. CONCLUSIONS:The CA19-9/FUT2/3 panel was the best performing panel in this biomarker bakeoff. Its evaluation in larger studies is warranted. Biomarker bakeoffs are an effective strategy for comparing the performance of promising biomarkers for pancreatic cancer early detection, and the PCDC is well poised to conduct such studies. Recommendations for performing such studies are provided.

Supplementary Figure 5. MUC5AC silencing inhibits cMET and CD44v6 expression and Astrocytes induce MUC5AC expression in breast cancer cells. A-B. Western blots showing knockdown of MUC5AC in MDA-231BR HER2 (A) and MUC5AC knockout in HCC1954BR (B) reduces expression of cMET and CD44v6. C. IHC staining for cMET and CD44v6 on brain tissue sections from mice injected with shSCR and shMUC5AC MDA-231BR HER2 cells. D. SKBR3 cells were treated with human astrocyte conditioned media for 24h and cells were lysed and subjected for MUC5AC expression. HGF induced signaling in MDA-231BR HER2 (E) and HCC1954BR (F); cells were treated with HGF (100ng/mL) for indicated times. After treatment cells were lysed and subjected for cMET and p-cMET expression using Western blot.

BACKGROUND AND OBJECTIVE:Anastomotic leak (AL) is a serious complication following esophagectomy and is often linked to poor perfusion of the gastric conduit (GC) and esophageal stump (EC). The aim of this study is to compare the efficacy of intraoperative Indocyanine green fluorescence angiography (ICG-FA) versus visual assessment VA) to assess perfusion status and its impact on the rate of AL. METHODS:Fifty-eight esophageal or gastroesophageal junction carcinoma patients were randomized to ICG-FA (28) and VA (30) groups. Perfusion status was assessed with VA alone in the VA group and with VA followed by ICG-FA in the ICG-FA group. RESULTS:The ICG-FA group had a lower leak rate of 4% when compared to 27% in the VA group (p = 0.03). ICG-FA identified nine cases where VA misjudged the GC tip vascularity, thereby avoiding unnecessary resections. ICG-FA necessitated revision of the GC tip in one case missed by VA and also identified poor perfusion of ES tip in three cases mandating revision which were deemed well-perfused by VA. CONCLUSION:ICG-FA demonstrated superiority over VA in assessing perfusion adequacy of the GC and ES, which resulted in a statistically significant decrease in the rate of anastomotic leaks.

Mentors: Elizabeth Bradford Bell, Tony Richa Program: Otolaryngology Type: Original Research Background: Barriers to healthcare have been associated with poor outcomes in patients with cancer. This study includes the largest sample size in a United States study aimed at better understanding if the rurality of patients’ residences is associated with the stage of oral cavity or oropharyngeal squamous cell carcinoma (SCC) at initial presentation. Methods: This 2024 hospital-based multicenter retrospective observational review utilized the National Cancer Database (NCDB). The rurality of each patient’s residence was binned into Metro, Urban, and Rural. The main outcome of this study was the AJCC 8th Edition clinical stage assigned to the patient at initial presentation. All models were adjusted for the year of initial presentation. Results: A total sample of 42,258 patients was evaluated, including 43.2% with oral cavity and 56.8% with oropharyngeal SCC diagnoses. Patients with oropharyngeal SCC from urban areas had higher odds of having a Stage 3/4 cancer (AOR = 1.16; 95% CI: 1.05, 1.28) relative to metro patients. Participants from urban areas had an increased risk of death for both the oral cavity SCC group (AHR = 1.11; 95% CI: 1.01, 1.21) and oropharyngeal SCC group (AHR = 1.23; 95% CI: 1.10, 1.37) compared to patients residing in metro areas. Conclusion: Urban patients were more likely to have a higher stage of oropharyngeal SCC at initial presentation, along with worse survival for either type of SCC, compared to the metro groups. Rural outcomes did not statistically differ from metropolitan outcomes. Future local or regional database analyses would better characterize the more nuanced patterns present in a particular region.

PURPOSE:Breast cancer brain metastasis remains a significant clinical problem. Mucins have been implicated in metastasis; however, whether they are also involved in breast cancer brain metastasis remains unknown. We queried databases of patients with brain metastasis and found mucin 5AC (MUC5AC) to be upregulated and therefore sought to define the role of MUC5AC in breast cancer brain metastasis. EXPERIMENTAL DESIGN:In silico dataset analysis, RNA-sequence profiling of patient samples and cell lines, analysis of patient serum samples, and in vitro/in vivo knockdown experiments were performed to determine the function of MUC5AC in breast cancer brain metastasis. Coimmunoprecipitation was used to unravel the interactions that can be therapeutically targeted. RESULTS:Global in silico transcriptomic analysis showed that MUC5AC is significantly higher in patients with breast cancer brain metastasis. Analysis of archived breast cancer brain metastasis tissue further revealed significantly higher expression of MUC5AC in all breast cancer subtypes, and high MUC5AC expression predicted poor survival in HER2+ breast cancer brain metastasis. We validated these observations in breast cancer brain metastatic cell lines and tissue samples. Interestingly, elevated levels of MUC5AC were detected in the sera of patients with breast cancer brain metastasis. MUC5AC silencing in breast cancer brain metastatic cells reduced their migration and adhesion in vitro and in brain metastasis in the intracardiac injection mouse model. We found high expression of cMET and CD44v6 in breast cancer brain metastasis, which increased MUC5AC expression via hepatocyte growth factor signaling. In addition, MUC5AC interacts with cMET and CD44v6, suggesting that MUC5AC promotes breast cancer brain metastasis via the cMET/CD44v6 axis. Inhibition of the MUC5AC/cMET/CD44v6 axis with the blood-brain barrier-permeable cMET inhibitor bozitinib (PLB1001) effectively inhibits breast cancer brain metastasis. CONCLUSIONS:Our study establishes that the MUC5AC/cMET/CD44v6 axis is critical for breast cancer brain metastasis, and blocking this axis will be a novel therapeutic approach for breast cancer brain metastasis.

The Pancreatic Cancer Detection Consortium (PCDC) performed a blinded EDRN defined phase 2 Biomarker Bakeoff study of eight promising biomarker panels using pancreatic cancer ductal adenocarcinoma (PDAC) case and general population healthy control samples from three institutions. The scientific questions were to determine whether panels could discriminate PDAC cases from controls with Receiver Operating Characteristic Area Under the Curve (AUC) > 0.8, compare panel performance head-to-head via 95% confidence intervals (CI), and determine whether the panel improved upon serum CA19-9 alone. A secondary goal was to perform modeling to discover a new panel consisting of individual biomarkers. Eligible biomarkers were required to have had a prior EDRN defined phase 2 validation study published and be ready for validation in a sample set from multiple institutions. Three institutions submitted a total of 140 cases and 140 controls frequency matched to cases on age and sex. Four consortium member institutions submitted for validation eight panels consisting of ten plasma, serum and DNA genotyping biomarkers in addition to CA19-9. Locked panel formulas were submitted prior to specimen distribution. The primary metric was AUC with bootstrapped 95% CIs. LASSO penalized regression was used for exploratory panel building. Nested cross validation and internal-external validation were performed, and optimism-corrected performance estimates were computed. The overall sample (N=280) was 51% female, median age of 67 years, and 94% were of non-Hispanic white race and ethnicity. Controls had median BMI 27.3, 56% never smokers, and 12% type II diabetes mellitus (DM2); cases (staged I to IV) had median BMI 26.6, 42% never smokers, and 26% DM2. Overall panel AUC estimates ranged from 89.9 to 96.3 and all were significantly larger than the null hypothesis value of 0.8 (p<0.05). The apparent serum CA19-9 AUC was 91.7 (87.8, 95.6). Two panels were significantly different from serum CA19-9, a panel with FUT2/3 genotype with serum CA19-9 (p=0.002) and a panel containing tissue factor pathway inhibitor (TFPI), tenascin C (TNC), and plasma CA19-9 (p=0.01). In subset analyses by PDAC stage, no panel improved upon the null hypothesis value of 0.8 in stage I (p>0.05), all panels improved upon the null hypothesis value of 0.8 in stage II and III (p<0.05), and only serum CA19-9 did not improve upon the null hypothesis value of 0.8 in stage IV. All biomarkers but two microRNAs were retained in a LASSO model that had optimism corrected AUC=96.4 (95% CI: 94.0, 98.9). All submitted panels performed well overall in this blinded study, with the panel containing FUT2/3 genotype performing the best. Panel discovery improved upon serum CA19-9, though optimism corrected AUC 95% CIs overlap with those of apparent CA19-9 performance. Further work is needed to prioritize promising biomarkers and to improve detection of early stage PDAC. Ann L. Oberg, William R. Bamlet, Grant Izmirlian, Seetharaman Balasenthil, Surinder K. Batra, Masataka Hayashi, Daniel Herman, Michael A. Hollingsworth, Maneesh Jain, Ann M. Killary, Brianna M. Krusen, Suyu Liu, Gopalakrishnan Natarajan, Subrata Sen, Lynette M. Smith, Sudhir Srivastava, Brian M. Wolpin, Kenneth S. Zaret, Michael G. Goggins. Pancreatic Cancer Detection Consortium biomarker bake-off: A consortium Phase 2 blinded biomarker validation and panel discovery study [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Pancreatic Cancer Research—Emerging Science Driving Transformative Solutions; Boston, MA; 2025 Sep 28-Oct 1; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2025;85(18_Suppl_3):Abstract nr B070.

Introduction Real-world implementation of the KEYNOTE-522 regimen presents substantial challenges, including high treatment-modification burden and adherence barriers, particularly in community and non-trial settings. Methods We conducted a retrospective cohort analysis of stage II-III TNBC patients treated at the University of Nebraska Medical Center from 2018–2024. Patients receiving neoadjuvant pembrolizumab with chemotherapy (KEYNOTE-522 regimen) (N = 50) were compared with a historical control group receiving dose-dense anthracycline-cyclophosphamide-taxane (ddACT) chemotherapy alone (N = 28). The primary focus was evaluating treatment-related toxicities, adherence challenges, pathologic complete response (pCR), and implications for surgical decision-making and clinical practice. Results Among 78 patients, 50 received the KEYNOTE-522 regimen and 28 received ddACT. High treatment-modification burden was more frequent in the pembrolizumab group, as reflected by treatment modifications or discontinuations (50.0% vs. 17.9%, p = 0.0051). pCR rates were higher with pembrolizumab (42.0% vs. 28.6%, p = 0.2391), and all pCRs corresponded to RCB 0. Recurrence occurred exclusively in non-pCR patients. At 12 months, recurrence and mortality were lower in the pembrolizumab group (16.0% vs. 44.4%, p = 0.0367; 8.0% vs. 29.6%, p = 0.0777), despite shorter median follow-up (17.7 vs. 40.4 months). EFS did not differ significantly by treatment (p = 0.1240), but was strongly associated with RCB class (p = 0.0118) and prognostic stage (p < 0.0001). In ddACT patients, treatment modification predicted worse EFS (p = 0.0215). No demographic or clinical variables were independently associated with pCR. Conclusions In this real-world cohort, the KEYNOTE-522 regimen was associated with high treatment-modification burden and adherence challenges, particularly among medically and socially complex patients. These findings suggest that implementing pembrolizumab outside clinical trials may require biomarker-guided patient selection and equity-focused implementation strategies before routine adoption in under-resourced settings.

CIBERSORT analysis for HOXA10 and its driven signature

Advanced prostate cancer (PCa) remains a significant clinical challenge, and docetaxel plays a significant role in disease management. Despite the efficacy of docetaxel as a first-line chemotherapy, resistance often develops. We developed three clinically relevant in vitro PCa cell models and transcriptomic analysis identified that the Paf1/RNA polymerase II complex component (PAF1)-associated pluripotent-transcription factor (TF), SOX2, plays a crucial role in docetaxel resistance. The cancer stem cell (CSC) transcriptional master regulator PAF1 is significantly higher in PCa cell lines, tumor tissues, and docetaxel resistant (DR) PCa cells than in age-matched control cells. To determine the molecular underlying and functional characteristics of PAF1 in resistance mechanisms, we performed coimmunoprecipitation, embryonic stem cell network proteins, in vitro tumor-initiating ability, and 3D multicellular organoid growth using PAF1 knockdown cells. Tet-inducible PAF1 depletion reduced the drug-efflux phenotype, tumor-initiating frequencies, and three-dimensional organoid growth of the docetaxel-resistant PCa cell lines. Functional studies also showed restoration of docetaxel sensitivity in a 3D tumorsphere model upon PAF1 depletion. PAF1 depletion was also associated with decreased pluripotent TFs and other CSC markers. This study provides a novel regulatory mechanism of docetaxel resistance in PCa through PAF1.

Supplementary Figure 1. Expression of MUC5AC is high in breast cancer brain metastasis. A. Heatmap showing the expression of upregulated genes in BC BrM cells as compared to primary BC cells (MDA-231BR vs MDA-231P). B. Comparison of MUC5AC expression (z-score) between BC and brain metastasis samples, using online available Gene Expression Omnibus (GEO) public datasets, probe (214385_s_at). C. Heatmap showing the expression of goblet cells marker genes in BC and breast cancer brain metastasis cell lines (MDA-231BR vs MDA-231P). Data represents upregulation of marker proteins. D. Pie chart representing H-score for MUC5AC expression in BrM tissues originating from BC (Breast cancer), LC (Lung cancer) and other cancers.

Background The preoperative classification of pancreatic cysts and detection of advanced neoplasia (high-grade dysplasia/pancreatic ductal adenocarcinoma [PDAC]) represents a significant diagnostic challenge. A prospective, multi-institutional study found that DNA-based testing (PancreaSeq) of pancreatic cyst fluid (PCF) improved the assessment of pancreatic cysts. Notable imitations to PancreaSeq necessitated the development of a DNA/RNA-based panel called PancreaSeq Genomic Classifier (GC). We validated PancreaSeq GC on a prospective patient cohort. Patients and Methods PancreaSeq GC was blindly tested on a prospective cohort of 241 patients with diagnostic follow-up. The performance of PancreaSeq GC in this cohort, which included 186 mucinous cysts (97 with advanced neoplasia), was benchmarked against traditional diagnostics and its DNA-only predecessor, PancreaSeq. Results PancreaSeq GC achieved 94.6% sensitivity and 96.4% specificity (area under the curve [AUC] of 0.955) for mucinous cysts. In comparison, increased fluid viscosity, elevated carcinoembryonic antigen (CEA), and PancreaSeq testing had lower sensitivities (71.4-88.2%) and lower AUC (0.824-0.941); however, PancreaSeq specificity was 100%. McNemar's test demonstrated higher sensitivity of PancreaSeq GC versus PancreaSeq for mucinous cysts (p < 0.001). For advanced neoplasia, PancreaSeq GC had 86.6% sensitivity and 97.9% specificity (AUC of 0.923), with McNemar's test confirming improved sensitivity over PancreaSeq (p = 0.031). Worrisome features, malignant cytopathology, and high-risk stigmata had lower sensitivities (44.3-84.5%) and lower AUC (0.604-0.892), while PancreaSeq had identical specificity to PancreaSeq GC. PancreaSeq GC had high accuracy in classifying intraductal oncocytic papillary neoplasms (IOPNs), intraductal tubulopapillary neoplasms (ITPNs), and cystic pancreatic neuroendocrine tumors (cPanNETs), with 97.1-100% sensitivity and 100% specificity. Conclusions This validation study demonstrates statistically significant improvements in PancreaSeq GC for mucinous cysts and advanced neoplasia, establishing it as a clinically valuable tool for the preoperative evaluation of pancreatic cysts.