Browse Articles
Discover research articles across all indexed journals
Clinical impact of succinate-dehydrogenase B protein expression in clear cell renal cell carcinoma.
e16511 Background: Succinate-deficiency renal tumors are considered a new entity since WHO 2016 classification. These tumors are mostly characterized by germline mutations in succinate-dehydrogenase gene and a good prognosis. However, previous works suggested that lack of expression of succinate dehydrogenase is related to a worse prognosis in clear cell renal cell carcinoma (ccRCC) histology. In this study, we characterized the protein expression of succinate dehydrogenase B (SDHB) and its relation to survival in a cohort of ccRCC patients. Methods: Paraffin samples from one hundred and nine ccRCC patients were analyzed by mass-spectrometry proteomics. Samples were divided into SDHB-low and high according to the median expression of SDHB protein. Kaplan-Meier and Cox regression were used for survival analyses. Differential biological processes and metabolic pathways according to protein expression in SDHB-low and high tumors were characterized using probabilistic graphical models and flux balance analysis. The obtained results were validated using two ccRCC cell lines, one SDHB-low and one SDHB-high, and one normal renal cell line. Cells were treated with a glycolytic inhibitor and viability studies were performed. Results: One hundred and nine ccRCC patients from La Paz University Hospital were included in this study, 40 (37%) female; 46 (42%) stage Ib, 6 (6%) stage II, 57 (52%) stage III; 9 (8%) grade 1, 39 (36%) grade 2, 17 (16%) grade 3, 7 (6%) grade 4, and 37 (34%) unknown. We found that low levels of SDHB protein were related to worse disease-free survival and overall survival. However, when this population is classified according to Keynote-564 criteria, not all low-SDHB tumors are classified as eligible for pembrolizumab despite their worse prognosis. Moreover, SDHB-low tumors had a higher expression of proteins related to glycolysis, indicative of an elevated Warburg effect, whereas SDHB-high tumors presented higher mitochondrial metabolism and adhesion. This elevated Warburg effect suggested the utility of glycolytic inhibitors for the treatment of SDHB-low tumors. We have tested this hypothesis in vitro. We treated three cell lines (a SDHB-low ccRCC cell line, a SDHB-high ccRCC cell line and a normal renal cell line) with a glycolytic inhibitor, which provoked a significant decrease in the SDHB-low ccRCC cell line viability when compared with the other two. Conclusions: Levels of SDHB protein seem to be a good prognostic biomarker in ccRCC. In addition, in those SDHB-low tumors that are not classified as eligible for pembrolizumab treatment an adjuvant treatment should be considered, with glycolytic inhibitors being an option.
A machine learning (ML)–based six-gene signature for risk stratification and therapeutic target identification in multiple myeloma.
7558 Background: Multiple myeloma(MM) accounts for 10-15% of hematologic malignancies with significant heterogeneity. Current risk stratification relies on clinical variables (ISS, sex, race, age, cytogenetics) but has limited prognostic accuracy (C-index ~0.63).TP53mt MM has worse outcomes, present in only 5% of newly diagnosed patients and increase to 25% at progression. Methods: We performed differential gene expression (DGE) and gene set enrichment analysis (GSEA) between TP53mt (36) and TP53wt (717) MM from (MMRF CoMMpass). Intersection of DGE and GSEA identified 46 genes, which along with clinical variables were modeled using Cox regression and 4 ML algorithms: LASSO, CoxBoost, XGBoost and Random Forest. Models were trained on 70% and tested on 30% of data, with external validation on GSE24080 (n=559).The final 6-gene signature was functionally evaluated in drug sensitivity data from 16 MM cell lines (DepMap). Performance was assessed using concordance index (C-index) and hazard ratios (HR). Results: CoxBoost model 6-gene signature (UBE2C, STMN1, TK1, DSCC1, PTTG1, CENPF) outperformed all other models. Combined with clinical variables, it achieved C-index 0.742 in discovery (95% CI: 0.708-0.776) versus 0.629 for clinical variables alone (+18.0% improvement, p<0.001).Minimal overfitting was observed (training: 0.748, test: 0.742, gap: 0.006). In multivariable Cox analysis, the CoxBoost risk score remained highly significant (HR 1.95, 95%CI:1.51-2.52,p<0.001) independent of ISS stage (HR 2.89, p<0.001) and age (HR 1.39 per decade, p<0.001).External validation on GSE24080 showed improved performance (C-index 0.683,+6.8% over clinical variables).The signature showed 50% gene overlap with LASSO (3/6 genes) and outperformed UAMS-70 (C-index 0.742 vs 0.631).Drug sensitivity analysis identified three compounds (MACIMORELIN, MW-150, BROFAROMINE) with selective activity against high-risk MM cells (p<0.001),validated across 16 cell lines. Conclusions: TP53 and its downstream pathways are central to MM progression. CoxBoost modeling produced a 6-gene signature that improved MM risk stratification over clinical variables, identifies therapeutic vulnerabilities, and is computationally efficient model suitable for clinical deployment, warranting prospective validation in MM treatment stratification. Performance of clinical-only and clinic-genomic machine learning models. Model Train C-Index Test C-Index Overfitting Improvement vs Clinical %Improvement Clinical 0.69 0.72 -0.03 0 0 Random Forest 0.57 0.59 -0.02 -0.13 -18.56 LASSO 0.70 0.68 0.02 -0.04 -5.89 XGBoost 0.81 0.65 0.16 -0.06 -8.95 CoxBoost 0.75 0.74 0.01 0.02 3.31
Accelerated failure time as an alternative to Cox proportional hazards for parathyroid carcinoma in the Surveillance, Epidemiology, and End Results database.
e18139 Background: Cox proportional hazards regression is a common model for multivariate survival analysis. This model relies on proportional hazards (PH), meaning hazard ratios for predictor variables must remain constant over time. Failure to report PH violation testing is prevalent in published studies; consequences of violation include decreased power and unreliable modeling. Large databases such as the Surveillance, Epidemiology, and End Results (SEER) cancer database allow clinicians to draw conclusions regarding treatment for rare cancers such as parathyroid carcinoma (PC), as there are no large-scale studies. Cox regression is commonly used to analyze this database. Multiple studies advocate for alternative models such as Accelerated Failure Time (AFT) when PH is violated; however, no studies have assessed PH violations or used the AFT model specifically for PC. The goal of this study is to assess PH violations in the SEER database for PC patients and demonstrate utility of AFT as an alternative to Cox regression. Methods: We conducted a systematic review of three databases to identify papers reporting multivariate survival analyses of PC patients in the SEER database. We extracted individual variables used for multivariate analysis in the included papers and used these variables to independently analyze patients in the SEER database diagnosed with PC from 2002 to 2022. PH violation was assessed via visual inspection of univariate Kaplan-Meier curves. Multivariate analysis was conducted using Cox regression and AFT models, and the results of these two models were compared. Results: Our systematic review returned 783 results, and 14 studies met criteria for inclusion. Most studies (92%, n=13) reported Cox regression results, but none of the 13 papers reported any testing for PH violation. We extracted 14 unique predictor variables from these papers. We then independently identified 666 PC patients in the SEER database; in this subset, a majority of the 14 variables violated PH in univariate analysis of overall (79%) and disease-specific (86%) survival. There were notable differences between the Cox regression and AFT models: simple (p = 0.046), complete (p = 0.02), and radical (p = 0.003) surgical intervention; negative lymph node metastases (p < 0.001); and above-average income (p = 0.04) were significant predictors of longer disease-specific survival in the AFT model but not in Cox regression. Other significant predictors of shorter disease-specific survival in the AFT model included Black race (p = 0.001), age > 72 years (p = 0.004), tumor size > 35mm (p < 0.001), regional disease (p = 0.047) and distant disease (p < 0.001). Conclusions: AFT is a more appropriate multivariate statistical model than Cox regression for PC patients in the SEER database. The AFT model identifies several clinically relevant survival predictors in this patient subset.
Evidence implicating human papillomavirus in bladder carcinogenesis: A case-control study.
4594 Background: The etiology of high-risk oncogenic human papillomavirus (HPV) subtypes with signature cytomorphological changes such as koilocytosis in bladder cancer (BCa) carcinogenesis still remains controversial and insufficiently investigated. We evaluated the diagnostic concordance rate between HPV-DNA test with urine HPV related cytomorphological findings in BCa urine samples. Methods: A retrospective case-control study was conducted. A total of 172 HPV-positive urine cytology specimens from patients with primary BCa treated at our hospital between January 2020 and December 2024 were included as the study group. During the same period, of 172 HPV - negative urine cytology specimens from primary BCa were selected as the control group. HPV DNA was detected by in situ hybridization was utilized. Immunohistochemistry (IHC) was performed to assess p16 INK4a protein expression as a surrogate marker of functional HPV activity. Cytomorphologic changes that had concurrent liquid-base cytology were identified using The Paris System for Reporting Urinary Cytology (TPS) criteria. Statistical analyses were performed using chi-square tests and fisher's exact tests. Results: Overexpression of p16 INK4a was observed in 82.2% (140/172) of the HPV-positive group and showed a high level of concordance with HPV infection status (Kappa=0.812, p<0.001). Among the HPV-positive group, high-risk HPV types accounted for 88.9% (152/172) with HPV-16 (114/152,75.6%) and HPV-18 (38/152,24.4%) being the predominant subtypes. A p16INK4a protein expression from HPV-negative group was detected which was significantly lower than in the positive group (13/172, 7.56%, p<0.001). HPV infection cytomorphology was identified, with 48.6% (83/172) koilocytes (52/172, 30.23%) either alone or in addition to basaloid clusters (16/172, 9.3%) and atypical squamous cells (ASCs) (15/172, 8.72%) identified using TPS from the HPV-positive group, and a significantly lower rate of 6.1% (10/172, p<0.001) was identified from the HPV-negative group. Conclusions: HPV plays an etiological role in carcinogenesis and contributes to a worse prognosis for patients with BCa development. The identification of cytomorphologic changes in urine cytology is a significant and underutilized piece of etiological evidence supporting a role for HPV in bladder carcinogenesis. It provides a reliable direct and visible link between the viral cytopathic effect and the urothelial cells. Integrating routine cytological assessment with modern molecular testing in urine samples could better define its role as an oncogenic driver in specific BCa variants. Recognizing this association has implications for risk stratification, screening in high-risk populations, and potential therapeutic avenues.
A multicenter, prospective real-world study of irinotecan liposome (II)–based combination regimens in patients with pancreatic ductal adenocarcinoma.
e16399 Background: Pancreatic ductal adenocarcinoma (PDAC) is a highly lethal malignancy with limited therapeutic options in advanced stages. Building upon the foundational evidence from the PAN-HEROIC-1 trial which demonstrated the clinical benefits of irinotecan liposome (II) and supported its approval for unresectable locally advanced or metastatic PDAC in China, this study aimed to evaluate the real-world safety and effectiveness of irinotecan liposome (II)-based combination regimens in patients with PDAC. Methods: This multicenter, prospective, real-world study included patients with pathologically or radiologically confirmed PDAC who were considered eligible by investigators for irinotecan liposome (II)-based combination regimens incorporating chemotherapy, chemoradiation, radiotherapy, immunotherapy, or targeted therapy. Patients were assigned to three cohorts according to treatment stage at enrollment: perioperative, first-line, and later-line treatment. Observational data were collected including adverse events, baseline characteristics, therapeutic regimens, and tumor response. The primary endpoint was treatment-emergent adverse events (TEAEs); secondary endpoints included overall survival (OS) and real-world progression-free survival (rwPFS), real-world time to progression (rwTTP), real-world objective response rate (rwORR), real-world disease control rate (rwDCR), real-world disease-free survival (rwDFS) and real-world duration of treatment (rwDOT). Results: As of August 30, 2025, a total of 236 patients were enrolled (perioperative, n = 23; first-line, n = 98; later-line, n = 115). The median age was 62.0 years (range, 33.0–83.0), with males and females accounting for 57.2% and 42.8%, respectively. Tumors were located in pancreatic head (25.4%), pancreatic body (11.4%), pancreatic tail (11.4%) and other pancreatic sites (51.7%). The most frequent metastasis site was the liver (39.0%). Baseline CA19-9 ≥37 U/mL was observed in 81.8% of patients. Any-grade TEAEs occurred in 36.9% of patients, with the most common being anemia (7.2%), nausea (6.8%), diarrhea (6.8%) and leukopenia (5.1%); grade ≥3 TEAEs occurred in 7.2% of patients. Median rwPFS was not reached in any cohort; the 6-month rwPFS rates were 83.3%, 66.4%, and 41.1% for the perioperative, first-line, and later-line cohorts, respectively. The median OS was not reached in any of the three cohorts. Conclusions: This real-world study revealed that the combination regimen based on irinotecan liposome (II) showed a manageable safety profile and encouraging effectiveness signals. Follow-up and patient enrollment are ongoing.
Choosing Wisely Pakistan: A call for value-based oncology practice in resource-limited settings.
e13532 Background: The Choosing Wisely initiative promotes evidence-based oncology care by reducing low-value interventions. While several low- and middle-income countries (LMICs) have successfully launched national Choosing Wisely oncology campaigns, Pakistan currently lacks a structured framework. With a rising cancer burden, limited diagnostic and treatment resources, financial constraints, and delays in radiotherapy access, there is a critical need to prioritize high-value cancer care. This study describes the development of a multidisciplinary roadmap for initiating Choosing Wisely Pakistan. Methods: This initiative is informed by real-world clinical practice in tertiary cancer care and review of existing Choosing Wisely oncology recommendations from high-income countries and LMICs. A multidisciplinary working group comprising radiation oncologists, medical oncologists, surgical oncologists, radiologists, and nuclear medicine physicians was formed. Common low-value practices in Pakistan were identified across diagnosis, treatment planning, imaging, and radiotherapy access. A consensus-driven process was followed including shortlisting practices based on frequency, cost–benefit ratio, clinical impact, and feasibility, followed by expert voting and structured multidisciplinary discussions. Engagement with relevant national professional societies is underway to seek formal support and facilitate implementation. Results: Several potentially low-value practices were identified, including inappropriate use of FNAC in breast cancer, initiation of surgery without receptor status or multidisciplinary input, overuse of PET-CT where it does not alter management, delays in lung cancer diagnosis due to empirical treatment, underutilization of cervical cancer screening, biopsy avoidance in head and neck cancers, and lack of prioritization for patients awaiting radical radiotherapy. These practices represent opportunities for improved and value-based cancer care. Specialty specific recommendations have been made and will be shared on various platforms in national and local languages after endorsement from official societies. Conclusions: Choosing Wisely Pakistan represents a feasible, multidisciplinary, and nationally scalable approach to optimizing oncology care in a resource-limited setting. Formal endorsement by professional societies and integration into clinical practice could reduce unnecessary interventions and improve equitable access to cancer care in Pakistan.
Association of circulating tumor cells with PD-L1 expression and clusters in confirmative tumor thrombus in selective solid cancers.
e16191 Background: Tumor thrombus (TT) is often an incidental, leading direct extension of tumor cells into a blood vein. It is commonly observed in renal cell carcinoma (RCC), hepatocellular carcinoma (HCC), and Wilms tumor. TT significantly worsens the prognosis and alters staging. Further the evidence of TT is often observed at locations like renal vein, inferior vena cava, and portal vein. Thus it requires multidisciplinary evaluation due to its aggressive nature and potential for obstruction or embolization. To differentiate tumor thrombus, diagnostic approaches revolve aroundimaging, such as CT or MRI, from a "bland" thrombus (a blood clot).We evaluated if there is any association and role of circulating tumor cells (CTCs) expressing immune-relevant markers e.g. PD-L in TT. Thus presence of CTCs originating from TT margins may refine risk stratification and therapeutic decision-making. Methods: In an observational study, total of 12 patients with HCC (n = 4), pancreatic (n = 3), liver (n = 3), RCC (n = 1), gallbladder (GB) (n = 1), etc. with confirmative with TT (age 51-80) underwent blood analysis for presence of CTCs with PD-L1 expression at baseline (B) and 3 patients with a follow up (FU). Samples were processed using OncoDiscover CTC enrichments CDSCO approved technology on an automated Zeiss microscope by confirmative EpCAM⁺, CK18⁺, DAPI⁺, CD45⁻ and PD-L1⁺. Results: Total of 16 CTCs were detected in 10 patient’s (83.33%) 1.5 ml of blood samples ranging 1-6 CTCs. At baseline, HCC, pancreatic, RCC, GB, and other cancers showed presence of CTCs with PD-L1 expression. In FU samples revealed the persistent CTC positivity, although PD-L1 positive CTCs were reduced. The mean CTC distribution was observed to be 1.33 with CK 18 expression, while expression of PD-L1 + ve CTCs were observed in a substantial subset, with mean distribution 0.67 (9 CTCs /12 patients). These cells are indicative with immune-evasive potential. CTC clusters were rare and detected in only 1 HCC patient, but persisted even at FU. Both male and female patients demonstrated comparable CTC positivity. Conclusions: The presence of CTCs in peripheral blood highlights active dissemination in TT margins. Although CTC clusters were infrequent, their occurrence may indicate heightened metastatic risk. For the first time we showed presence of CTCs and shedding from thrombus margins into blood circulation. More studies in this direction are suggested. Clinical trial information: IESC/FP /24/2023 (Acronym: USGCTC).
Characteristics of patients managed with active surveillance versus treatment for prostate cancer within a large nationwide health system.
e17129 Background: Active surveillance (AS) is broadly endorsed as preferred management for low-grade prostate cancer and has been increasingly adopted within the VA Health Care System (VAHCS). This study analyzes the characteristics of VAHCS patients managed with AS rather than immediate treatment, as well as those that receive confirmatory biopsy (CBx) within 24 months of diagnostic biopsy (DBx), in order to evaluate the factors that direct physicians to utilize AS for low-grade prostate cancer. Methods: A retrospective cohort was identified using patients listed in the VA Informatics and Computing Infrastructure (VINCI) Prostate Data Core. Using natural language processing (NLP), the cohort was limited to patients with a DBx between 2005 and 2024 indicating Gleason Grade Group (GG) 1 prostate cancer or GG 2 prostate cancer with < 50% positive cores at DBx. Patients were stratified by whether they received treatment (radiation, prostatectomy, and/or orchiectomy/ADT) during the first 15 months or remained on AS or watchful waiting (WW). Patients with CBx prior to treatment or at least one PSA > 1 ng/mL were classified as managed with AS/WW. Patients were censored due to death, progression to metastatic status, no recorded PSA labs within 15 months, or a reported PSA value < 1 ng/mL. Hierarchical mixed-effects logistic models were used to analyze AS usage and CBx administration within 24 months of DBx with VA facility as a random effect. Results: There were 173,725 patients with a positive biopsy for prostate cancer between 2005 and 2024. Of those diagnosed with GG 1 prostate cancer or GG 2 with < 50% positive cores, 40,341 patients (49.7%) were managed with AS/WW while 40,872 patients (50.3%) received treatment during the first 15 months after diagnosis. Since 2015, the proportion managed with AS was 64%. Black or African American patients were less likely to be managed by AS (OR: 0.84; 95% CI: 0.80-0.88; p < 0.001) or receive CBx within 24 months of DBx (OR: 0.84; 95% CI: 0.79-0.88; p < 0.001). AS management was also less likely among Hispanic or Latino patients (OR: 0.83, 95% CI: 0.74-0.93; p < 0.001), while older patients were more likely to be managed by AS (OR: 1.26 per decade, 95% CI: 1.23-1.30; p < 0.001). Patients in zip codes with a higher area deprivation index (ADI) also had lower AS utilization (OR: 0.96 per quartile; 95% CI: 0.94-0.98; p < 0.001) and CBx receipt (OR: 0.96; 95% CI: 0.93-0.98; p < 0.001). Results are robust for age, race, and ethnicity when limiting the cohort to diagnoses between 2015 – 2024 and stratification by GG. Conclusions: Prostate cancer care in the VAHCS reflects the best uptake of AS/WW for favorable risk disease in the country. By analyzing the characteristics of patients managed by AS and receiving quality AS through CBx receipt, further efforts can be made to adapt these management practices for low-grade prostate cancer.
Potential alternative dosing strategies for FDA-approved oncology biologics based on reanalysis of exposure-response data.
e15135 Background: Many modern oncology biologics were developed using historical maximum tolerated dose paradigms that are poorly aligned with targeted/immunotherapy PK/PD features including long half-life, early efficacy plateaus, and target saturation. Therefore, labeled regimens may exceed the exposure needed for maximal benefit and could be candidates for lower doses and/or longer dosing intervals, with potential to reduce avoidable toxicity and financial burden. Regulatory initiatives (e.g., Project Optimus) emphasize selecting optimal rather than maximal doses through prospective robust dose exploration and PK/exposure-response (E-R) integration. Methods: FDA-approved large-molecule oncology drugs were identified using Drugs@FDA reports. Labeling, FDA multidisciplinary and clinical pharmacology reviews, and published literature were systematically reviewed. Pharmacokinetic and population PK parameters, covariates, and E–R evidence were extracted. Alternative dosing regimens (dose reduction and/or interval extension) were proposed when dosing appeared potentially non-optimal, defined as: (1) no credible E–R relationship or likely false-positive E–R, due to confounding, with limited lower-dose exploration; (2) dosing exceeding that required for maximal response or target saturation; or (3) dosing frequency excessive relative to drug PK. Drug cost impact was estimated using average wholesale price (AWP; Micromedex Red Book), assuming no vial sharing or storage. Results: Twenty-nine FDA-approved oncology mAbs/immunotherapies were screened; 21/29 (72%) had candidate alternative dosing with potential maximum dose savings of 25–67%. 19/21 (90%) had estimable annual per-patient impact, with maximum annual per-patient savings ranging from $39,645–$240,583. For drugs with available 2024 annualized global sales, the theoretical upper-bound global annual savings (sales × maximum dose-saving %) totaled ~$31.1B/year, led by pembrolizumab (~$14.7B), trastuzumab including biosimilars (~$3.7B), bevacizumab including biosimilars: (~$3.63B), atezolizumab (~$3.0B), and nivolumab (~$2.7B), followed by durvalumab (~$1.57B) and ipilimumab (~$1.31B). The highest maximum dose-saving percentages were observed for atezolizumab (67%) and toripalimab (66.6%), with several additional agents showing ~50% potential dose savings, including pembrolizumab, nivolumab, trastuzumab, bevacizumab, dostarlimab, penpulimab, and ipilimumab. Conclusions: A substantial proportion of marketed oncology biologics may be administered at doses and/or intervals exceeding those needed to maintain pharmacologically active exposure. PK/E–R–informed dose optimization aligns with Project Optimus principles and may reduce drug use and financial toxicity without compromising efficacy.
Clinical characteristics, treatment patterns, and outcomes of common solid tumors in Arab American patients.
e23407 Background: Arab Americans are frequently misclassified as White in U.S. cancer registries, limiting the study of cancer presentation, treatment patterns, and outcomes in this understudied population. Michigan hosts the largest Arab American community in the U.S., providing an opportunity to characterize cancer epidemiology in this group. We aimed to describe clinicopathologic features, treatment patterns, and outcomes across solid tumors in an institutional Arab American cohort to support future disparities-focused investigations. Methods: We created a database by identifying patients with Arabic or Arab-appearing surnames and confirming Arab ethnicity through chart review based on self-reported language and/or country of origin. Patients with any solid tumor at any stage were included; those with unconfirmed or non-Arab origin were excluded. Demographics, tumor characteristics, treatments, and outcomes were collected. Treatment patterns were analyzed by cancer type and stage. Survival was assessed using Kaplan–Meier methods, with multivariable Cox regression performed to evaluate prognostic factors for overall survival. Results: A total of 96 patients were included (2004-2024): breast (n = 40), colorectal (n = 20), prostate (n = 20), and lung (n = 16) cancers. The median age was 56 years and 64.6% presented with early-stage disease (AJCC 8 th Edition, I–II). Surgery was performed in 68.8%, chemotherapy in 40.6%, radiation in 44.8%, hormone therapy in 40.6%, immunotherapy in 3.1%, and targeted therapy in 4.2%. There was significant overlap reflecting multimodality therapy in stage I–III disease, whereas stage IV patients were predominantly treated with systemic therapy alone (73.3%). At a median follow-up of 56 months, median OS was not reached (95% CI: 129 months to NR), with 5-year OS of 80.2% (95% CI: 71.8%–89.6%). Cancer-specific 5-year OS varied: breast 91.9%, prostate 93.3%, lung 59.6%, and colorectal 58.2%. Median PFS for all stages was 80 months (95% CI: 50-NR), with stage III–IV patients having median PFS of 13 months (95% CI: 9–45). In multivariable analysis, age < 65 years was associated with improved survival (HR 0.163, 95% CI: 0.038–0.705, p = 0.015), while advanced stage (III–IV) showed a trend toward worse outcomes (HR 3.374, 95% CI: 0.910–12.517, p = 0.069). Indirect comparison to SEER-published 5-year relative survival estimates revealed lower 5-year OS for colorectal cancer (58.2% vs ~65%) in our cohort, while breast, prostate, and lung cancer survival were similar or higher than population benchmarks. Conclusions: This is the first descriptive study to characterize treatment patterns and oncologic outcomes among Arab American patients. This preliminary analysis highlights the need for expanded comparative and molecular studies of site-specific cancers to better understand racial disparities in outcomes.
circRNA signatures associated with BRCA mutation status in ovarian cancer and a RASSF8-derived candidate biomarker.
e17548 Background: Clinically actionable variants (CAVs) in BRCA1 and BRCA2 determine eligibility for targeted therapy in high grade serous ovarian cancer (HGSOC). However, RNA based biomarkers indicating BRCA status remain limited. Circular RNAs (circRNAs) are non coding RNAs with regulatory roles in gene expression. Their stability and detectability in human biofluids make circRNAs attractive liquid biopsy candidates. This study examined whether circRNA expression varies by BRCA background and assessed translational potential. Methods: This was a translational biomarker study examining circRNA expression across distinct BRCA backgrounds in high grade serous ovarian cancer. Three HGSOC cell lines representing BRCA1 mutant, BRCA2 mutant, and BRCA wildtype backgrounds were analysed. The primary endpoint was identification of circRNAs differentially expressed according to BRCA status. circRNA profiling was performed using the Arraystar circRNA microarray platform. Differential expression was assessed using normalised signal intensity, a fold change threshold ≥2, and associated statistical outputs, with adjustment for multiple testing. Candidate circRNAs were prioritised based on fold change, reproducibility, and annotation quality. Selected circRNAs were validated using divergent primers and quantitative PCR across the HGSOC panel, a chemotherapy resistant model, and clinical tumour samples stratified by BRCA status. Results: Heatmaps and volcano plots showed circRNA profiles were driven mainly by cell line specific patterns. However, at the individual transcript level, distinct circRNAs demonstrated BRCA associated differences. Five circRNAs were prioritised for validation. Three circRNAs (hsa_circ_400223, hsa_circ_008026, and hsa_circ_003300) were upregulated in both BRCA1 and BRCA2 mutant models compared with wildtype. hsa_circ_100059 was predominantly expressed in the BRCA1 mutant cell line, while hsa_circ_025636 was highly expressed in the BRCA2 mutant cell line. In clinical samples, hsa_circ_025636, a RASSF8 derived circRNA, showed higher expression in a subset of BRCA mutant cases relative to wildtype (p<0.05). Conclusions: This study provides evidence that circRNAs may serve as biomarkers of BRCA status in HGSOC. While global circRNA profiles were not defined by BRCA genotype, specific circRNAs demonstrated BRCA associated expression. The RASSF8 derived circRNA hsa_circ_025636 is a promising candidate and warrants evaluation in larger cohorts alongside functional studies.
Improving clinical trial enrollment rates among patients with operable invasive breast cancers: An ASCO Quality Training Program initiative.
e23305 Background: Clinical trial participation is essential for advancing cancer care and recommended in National Comprehensive Cancer Network guidelines. At Kaiser Permanente San Francisco (KPSF), about 5% of patients with early-stage breast cancer enrolled in a clinical trial in 2024, similar to community standards. Through ASCO’s 2025 Quality Training Program, we identified potential barriers to trial enrollment and tested interventions to reach a goal enrollment rate of 20%, more typical of National Cancer Institute-designated cancer centers, in 2025. Methods: After process mapping and surveying key stakeholders including administrators, oncologists, nurses, and patients, we identified a lack of patient and provider awareness of available trials as the biggest perceived barrier to clinical trial enrollment and, using PDSA methodology, tested multiple changes to target this. We modified printed care plans to integrate clinical trial participation and developed a tracking process for a single trial in January 2025, created a standardized electronic note template in March for providers to document clinical trial discussions, and began tracking all early-stage breast cancer trials in June, including patient eligibility and reasons for non-enrollment. Our outcome measure was the percentage of all patients with newly diagnosed operable invasive breast cancer first seen in our oncology clinic from 1/1/2025 to 8/31/2025 who enrolled in a clinical trial at KPSF. Results: In total, 184 new patients were seen in 2024 and 97 in 2025 (through August). Key patient characteristics, including age, race/ethnicity, BMI, clinical stage, and comorbidity burden, were similar between years. Clinical trial enrollment increased from 4.9% in 2024 to 33.0% in 2025, with 6 consecutive months (March through August) exceeding the mean 2024 rate (Table 1). Of 70 non-enrolled patients in 2025, 41.4% were due to lack of trial awareness or tracking; this decreased from 72.7% in January to 0% in August. The trials discussion note template was used by breast oncologists in 49.4% of new encounters since its introduction. Conclusions: A formal quality improvement process can help identify and reduce local barriers to early-stage breast cancer clinical trial enrollment by improving trial awareness and tracking. Future efforts should assess long-term sustainability of our process and generalizability to other cancer types. Clinical trial enrollment rates over time in patients with newly diagnosed operable invasive breast cancers. Year Month (PDSA Cycle) Enrollment Rate 2024 Jan—Dec 4.9% (9/184) 2025 Jan (modified care plans, single trial tracking) 16.7% (2/12) Feb 0% (0/8) Mar (trials discussion note template) 41.7% (5/12) Apr 16.7% (1/6) May 55.6% (5/9) Jun (expanded tracking) 50.0% (9/18) Jul 28.6% (6/21) Aug 27.3% (3/11)
Efficacy and safety of SPH4336 plus endocrine therapy in HR-positive/HER2-negative metastatic breast cancer with brain metastases: A multicenter, single-arm, phase II study.
1076 Background: Cyclin-dependent kinase 4/6 (CDK4/6) inhibitors plus endocrine therapy are standard first-line treatment for HR-positive/HER2-negative (HR + /HER2 - ) metastatic breast cancer (MBC). However, patients with brain metastases have poor prognosis and limited treatment options due to inadequate blood-brain barrier penetration of most agents and limited evidence for intracranial efficacy. SPH4336, a novel oral CDK4/6 inhibitor, exhibits potent CDK4/6 inhibition in preclinical models with favorable blood-brain barrier penetration. It has shown acceptable safety profiles and preliminary antitumor activities in advanced solid tumors, including HR + /HER2 - MBC. This study assessed the efficacy and safety of SPH4336 plus endocrine therapy in HR + /HER2 - MBC with brain metastases. Methods: This open-label, single-arm, phase II study enrolled patients aged 18-75 years with histologically confirmed HR + /HER2 - and radiologically confirmed brain metastases at 25 centers in China. Prior CDK4/6 inhibitor exposure before brain metastases was limited to <2 agents. Patients received SPH4336 (400 mg orally once daily) plus physician's choice of endocrine therapy (letrozole, fulvestrant, or exemestane) in 28-day cycles until disease progression or unacceptable toxicity. Primary endpoint was intracranial objective response rate (ORR) per RANO-BM criteria. Results: Between September 19, 2023, and August 22, 2025, 30 patients were enrolled. Two patients (6.7%) had a history of palliative radiotherapy for brain metastases. As of December 10, 2025, the median follow-up was 11.7 months (range: 1.6-26.7). Per RANO-BM criteria in 29 evaluable subjects, the intracranial ORR, disease control rate (DCR), and clinical benefit rate were 37.9% (95% CI: 20.7-57.7), 72.4% (95% CI: 52.8-87.3), and 44.8% (95% CI: 26.4-64.3), respectively. Based on RECIST 1.1, the extracranial ORR and DCR were 33.3% (95% CI:17.3-52.8) and 66.7% (95% CI: 47.2-82.7), respectively. The overall ORR was 33.3% (95% CI: 17.3-52.8), and the overall DCR was 60.0% (95% CI: 40.6-77.3). The median intracranial, extracranial, and overall progression-free survival were 8.4 months (95% CI: 1.9-not reached [NR]), 11.0 months (95% CI: 1.9-NR), and 8.1 months (95% CI: 1.9-11.0), respectively. Grade ≥3 treatment-related adverse events occurred in 70.0% of patients, most commonly γ-glutamyltransferase elevation (43.3%), aspartate aminotransferase elevation (20.0%), and neutrophil count decrease (20.0%). No treatment-related death occurred. Conclusions: SPH4336 plus endocrine therapy demonstrated clinically meaningful intracranial antitumor activity with a manageable safety profile in HR + /HER2 - MBC with brain metastases. These findings warrant further investigation in randomized studies for this patient population with high unmet medical need. Clinical trial information: NCT05872347 .
Knowledge-distilled EfficientNets for peripheral blood smear classification.
e14083 Background: Leukemias represent the most common malignancies in children and remain a major source of cancer-related morbidity across age groups. Microscopic examination of peripheral blood smears is central to diagnosis and classification but is labor intensive, operator dependent, and challenged by morphologic overlap between leukemic subtypes and reactive hematologic conditions. Delays or misclassification can directly affect treatment initiation and risk stratification. Although deep learning has demonstrated high accuracy in hematologic image analysis, many models are computationally intensive and difficult to deploy outside specialized centers. Scalable AI systems that preserve diagnostic fidelity while enabling broad clinical adoption are needed. Methods: We conducted a retrospective, multi-institutional analysis of digitized peripheral blood smear images curated from publicly available leukemia repositories, including ALL-IDB–derived datasets, encompassing acute lymphoblastic leukemia, other leukemic subtypes, and normal hematologic samples. Ground-truth labels were established through expert hematopathologist annotation with diagnostic confirmation. A high-capacity EfficientNetB7 reference model was trained to capture hierarchical cytomorphologic features across heterogeneous smear preparations. Structured knowledge distillation was then applied to train a lightweight EfficientNetB0 model, transferring discriminative capability while substantially reducing computational requirements. Extensive computational evaluation was accompanied with external validation across sites. Performance metrics included accuracy, sensitivity, specificity, F1 score, and AUROC. Results: The distilled EfficientNetB0 model achieved high diagnostic accuracy across leukemia and non-leukemia classes, with overall accuracy exceeding 97% and AUROC greater than 0.97 on external validation. Sensitivity for leukemic blast detection remained high across morphologically challenging cases, including low-blast-count samples. Performance was consistent across staining variations and imaging conditions. Mean inference time was under 0.1 seconds per image on standard hardware, supporting real-time clinical use. Expert reviewers reported the system to be valuable for diagnostic triage and workload prioritization. Conclusions: Knowledge-distilled EfficientNet modeling enables accurate, rapid, and computationally efficient leukemia classification from peripheral blood smears while preserving clinically meaningful performance. By addressing key barriers to deployment, this approach supports scalable integration of AI-assisted hematologic diagnostics across resource-diverse settings. Prospective studies are planned to evaluate impact on diagnostic turnaround time, interobserver variability, and treatment initiation.
Use of knowledge-distilled EfficientNet to enable rapid, scalable glioma diagnosis.
2040 Background: Accurate preoperative classification of gliomas on MRI according to WHO 2021 integrated diagnosis is essential for surgical planning and treatment selection, yet conventional interpretation remains vulnerable to inter-observer variability. Deep learning shows promise for brain tumor classification but computational demands restrict deployment in resource-limited settings. Knowledge distillation, transferring discriminative knowledge from computationally intensive "mentor" networks to efficient "mentee" models, enables scalable AI systems maintaining diagnostic accuracy while enabling global translation. Methods: We analyzed 2,840 brain MRI studies sourced from six continents comprising gliomas (n=1,400; 700 WHO grade 2–3, 700 grade 4), meningiomas (n=900), and pituitary adenomas (n=540). Ground truth followed WHO 2021 integrated diagnosis with expert neuroradiologist consensus. Data were partitioned at the patient level with site-held-out external testing. A high-capacity EfficientNetB7 mentor model was developed using multi-sequence MRI inputs (T1, T2, FLAIR, post-contrast T1) and distilled into lightweight EfficientNetB0. Knowledge distillation preserved discriminative performance while enabling CPU deployment without GPU acceleration. Training: input 224×224 pixels, Adam optimizer (lr=0.001), Kullback-Leibler divergence loss (temperature T=4), twelve epochs. Performance metrics included accuracy, sensitivity, specificity, F1-score, and AUROC. Independent clinical feasibility involved 52 board-certified neuroradiologists across 11 geopolitical regions. Results: For WHO grade 2–3 versus grade 4 stratification, distilled EfficientNetB0 achieved 97.2% accuracy, 98.1% sensitivity, 96.3% specificity, and AUROC 0.993. Differential diagnosis accuracy: 95.3% (glioma vs. meningioma), 96.1% (glioma vs. pituitary). Site-held-out external validation showed consistent performance (93–97% grading, 94–96% differential). Mean inference time: 0.19 seconds on CPU. Distilled model: 3.9 million parameters (97% reduction), 2.1 billion FLOPs (94% reduction). Neuroradiologist assessment: 94.7% rated system valuable for diagnostic confidence and surgical triage. Inter-rater agreement: 96.8% (κ=0.951, 95% CI 0.941–0.961). Conclusions: Knowledge-distilled EfficientNetB0 achieves 97.2% accuracy for glioma grading and 95–96% for differential diagnosis while reducing parameters by 97% and enabling CPU deployment. Global validation across six continents and clinical assessment by 52 neuroradiologists establish real-world feasibility. The architecture supports translation as a decision-support tool for preoperative glioma stratification, particularly in resource-constrained settings. Prospective clinical trials evaluating impact on surgical decision-making are warranted.
Understanding the role of large language models in cancer mortality prediction: A real-world study.
e23177 Background: Although mortality risk prediction is essential in oncology, risk determinants vary widely across cancer types and patient characteristics, rendering population-level prediction challenging. While traditional statistical and machine learning (ML) approaches have shown encouraging results, the effectiveness of ML and large language models (LLMs) for near and long-term cancer mortality prediction remains unknown. We sought to assess the mortality prediction of these different models for multiple timepoints in a heterogeneous cancer population. Methods: We conducted a retrospective cohort study using electronic health record (EHR) data from Yale New Haven Health, with follow-up through 6/30/25. Patients with any first cancer diagnosis on or after 1/2019 were included. One post-diagnosis encounter was randomly selected as the index visit, excluding visits within 7 days of death. The final cohort was randomly divided into training, validation, and test sets. Structed demographics, laboratory, and diagnostic comorbidity codes were included in training models and derived from the 180 days preceding the index visit. Outcomes included mortality at 30 and 180 days after index visit. We assessed logistic regression (LR), ML model (XGBoost), and LLM (GPT-4o) for prediction of mortality. LR and XGBoost were trained on the full training set, whereas GPT-4o was not trained on patient data and evaluated using zero-shot inference and limited in-context learning (ICL) with five example patients, within a secure environment. Model performance was compared using area under the receiver operating characteristic curve (AUROC), sensitivity, and specificity with higher values indicating better performance. Results: Our cohort included 82,662 individuals with a median follow-up of 23.4 months after their index date. 15,152 (18.33%) of individuals died during our study period. The most common diagnoses were prostate (14.5%), breast (12.4%), and lung (6.8%). Fully trained ML models achieved the highest AUROC, while untrained LLMs (0-shot) showed meaningful performance and improved with minimal ICL, approaching traditional and ML methods (Table). Conclusions: Our findings highlight the importance of assessing both short- and longer-term mortality, as model performance differed across timepoints. LLMs captured clinically relevant risk signals from structured EHR data even without task-specific training, with further improvement using minimal ICL. While ML achieved the best performance after supervised training, LLMs show promise for mortality prediction and integration into EHR-based clinical decision support. Performance of models for mortality prediction. Model Time (days) AUROC Sensitivity Specificity LR 30 0.90 0.81 0.84 180 0.87 0.74 0.82 XGBoost 30 0.92 0.38 0.98 180 0.89 0.59 0.93 GPT-4o (0-shot) 30 0.87 0.90 0.70 180 0.83 0.81 0.73 GPT-4o (ICL) 30 0.88 0.85 0.80 180 0.84 0.75 0.80
Trends and disparities in prostate cancer mortality among men with hypertension in the United States: A CDC WONDER analysis (1999–2020).
e17064 Background: Prostate cancer is a major cause of cancer-related mortality among older men in the United States. Hypertension is highly prevalent in this population and has been associated with adverse oncologic outcomes. However, national trends and disparities in mortality among patients with prostate cancer and hypertension remain poorly characterized. Methods: Mortality data from the CDC Wide-ranging Online Data for Epidemiologic Research (CDC WONDER) database were analyzed for adults aged ≥55 years from 1999–2020. Deaths listing prostate cancer (ICD-10: C61) and hypertensive diseases (ICD-10: I10–I15) as underlying or contributing causes were included. Crude and age-adjusted mortality rates (AAMRs) per 100,000 persons were standardized to the 2000 U.S. population, and temporal trends were assessed using Joinpoint regression to estimate annual (APC) and average annual percent changes (AAPC), stratified by age, race/ethnicity and states. Results: From 1999–2020, 124,668 deaths occurred among individuals with prostate carcinoma and hypertensive diseases, with overall age-adjusted mortality rates (AAMRs) increasing from 4.0 to 11.0 per 100,000 (AAPC 3.77%, 95% CI 2.08–5.48; p = 0.000009). Age-stratified analysis demonstrated the highest mortality among adults aged ≥85 years, with crude mortality rates rising from 18.1 to 58.7 (AAPC 4.95%, 95% CI 3.93–5.97; p < 0.001), while the lowest mortality was observed in the 55–64-year age group (from 0.7 to 1.2; AAPC 3.72%, 95% CI 1.29–6.21; p = 0.0025). Sex-wise, mortality occurred exclusively among males, with AAMRs increasing from 11.0 to 26.8; males experienced a sharp increase from 1999–2001 (APC 32.47%, 95% CI 14.10–53.80; p = 0.001), followed by a non-significant decline from 2001–2018 (APC −0.39%, 95% CI −0.80–0.03; p = 0.069), and a renewed rise from 2018–2020 (APC 15.08%, 95% CI 3.20–28.32; p = 0.015). Race-stratified analysis showed the highest AAMRs among Non-Hispanic (NH) Black or African American individuals (overall AAMR 18.6), followed by NH White (7.1), Hispanic or Latino (6.4), NH American Indian or Alaska Native (5.9), and NH Asian or Pacific Islander populations (4.3). States with age-adjusted mortality rates above the 90th percentile included District of Columbia (17.1), Mississippi (16.4), North Dakota (12.0), Oklahoma (12.0), and Minnesota (11.3). Conclusions: Overall, mortality involving prostate cancer with concurrent hypertension increased significantly from 1999 to 2020 in the United States. Mortality burden was disproportionately higher among NH Black or African American individuals, older age groups and varied substantially across states, indicating persistent demographic and geographic heterogeneity over the study period. These patterns highlight population groups and regions that may benefit from targeted public health planning and resource allocation.
Impact of systemic delays in targeted therapy (TT) access for advanced NSCLC (aNCSLC) with actionable genomic alterations (AGA+) in Alberta, Canada.
e23028 Background: Approval and funding timelines for precision oncology drugs in Canada lag those of other OECD nations. Following proof of efficacy (POE) (clinical trial readout or publication), multiple steps including regulatory review, health technology assessment, price negotiation, and sequential jurisdictional (provincial) funding agreements must be completed before a novel TT is available and funded through the Canadian system. We assessed survival probabilities and potential years of life lost (PYLL) relative to this process among patients who received first-line (1L) palliative-intent TT approved and funded in the last 10 years (afatinib, alectinib, brigatinib, crizotinib, entrectinib, larotrectinib, lorlatinib, osimertinib, and selpercatinib) for an AGA+ aNSCLC diagnosis. Methods: The Alberta Glans-Look Lung Cancer Research (GLR) dataset was used to determine the outcomes (PFS and OS) of two population- level cohorts: 1) AGA+ receiving an approved and funded 1L TT, and 2) AGA- receiving a non-TT (chemotherapy and/or immune checkpoint inhibitor). Outcomes of AGA- patients were considered a proxy for outcomes of AGA+ patients for whom a TT is not available. Average time from POE to public funding in Alberta was determined using clinical trial publications and documentation from Health Canada, Canada’s Drug Agency, pan-Canadian Pharmaceutical Alliance, and the Alberta provincial drug formulary. Probability of survival from POE to funding was estimated using Kaplan-Meier methods. PYLL for both PFS and OS was calculated as the difference in observed time-to-event between the AGA+ and AGA- cohorts multiplied by the median POE to funding interval for 1L-TT. For PFS, this value was further multiplied by the estimated annual number of Alberta AGA+ aNSCLC diagnoses (from Statistics Canada), and for OS by the estimated annual deaths in Alberta from AGA+ aNSCLC (from Brenner et al. 2023; PMID 37999145). Results: 485 AGA- and 529 AGA+ (143 ALK, 337 EGFR, 7 RET, 42 ROS1) patients with aNSCLC receiving a 1L palliative intent systemic therapy were identified from 2015-2025. Average time from POE to public funding in Alberta was 42.9 mo [range: 18 - 71.8 mo]; 31.6, 28.2, 47, and 53 mo for ALK, EGFR, NTRK, RET, and ROS1 1L-TT respectively. Probability of survival to 42.9 months was 41% for AGA+ on 1L-TT and 0% for AGA- receiving non-TT. PYLL during this interval was 993.7 years for PFS and 914.22 years for OS. Conclusions: The time required for novel TT to traverse from POE to public funding in Alberta results in substantial lost years of life. Given these conditions, people living with aNSCLC today are unlikely to survive long enough to access a novel emerging TT in either the 1 st or later treatment lines. Incentivizing regulatory submission directly following POE and streamlining the approval and funding process is a current unmet need to improve equitable TT access and outcomes.
Tracing the journey: Patient engagement in long-term antiretroviral therapy care in Malawi
Background Consistent engagement in antiretroviral therapy (ART) care is crucial to improve health outcomes and reduce HIV transmission. This study examined ART engagement patterns among ART clients during their first two years at two public ART clinics in Lilongwe, Malawi. Retention support for new initiates is provided by ART “Buddies” or through a two-way texting (2wT) system for those with phones and interest. Methods ART engagement patterns were assessed over six-month intervals (>0–6, > 6–12, > 12–18, > 18–24) among clients stratified by retention support group: 1) Buddy without phone access; 2) Buddy with phone access; and 3) 2wT. Outcomes were based on ART program status at the end of each interval. Clients were classified as retained on ART or not retained (lost to follow-up (LTFU), transferred out, stopped, or died). Among those retained, engagement was further categorised as continuously engaged (attended all appointments within 13 days), cyclically engaged (returned late (14–59 days) at least once, or re-engaged (missed an appointment by ≥60 days but returned to care). Clients were not censored within interval and could re-enter care in subsequent intervals. Results Among 6,303 clients, 5,880(93%) received ART Buddy support and 423(7%) received 2wT. Of those in the Buddy support group, 1,030 (18%) had phone access. 2wT clients showed the highest continuous engagement up to 18 months, compared to Buddy clients with and without phone access (70% vs 57% vs 30% at 0–6 months, 72% vs 58% vs 33% at >6–12 months, and 81% vs 76% vs 62% at >12–18 months). At 24 months, 3,363 (53%) were retained on ART: 2,790 (58%) of Buddy with phone; 277(27%) of Buddy without phone, and 296 (70%) of 2wT. Of those retained at 24 months, 1,834 (55%) were continuously engaged, 1,029 (31%) had cyclically engaged, and 500 (15%) re-engaged after LTFU across intervals. Conclusion ART engagement was dynamic and heterogeneous over time. Tailored retention support based on engagement patterns and time on ART could improve long-term retention in ART care.
Experimental determination of nanoscale thermal transport properties in heat-assisted magnetic recording media using step-by-step time-domain thermoreflectance method
Heat-assisted magnetic recording (HAMR) enables ultra-high areal density by locally heating magnetic recording media with high magnetic anisotropy, but the writing performance critically depends on the behavior of heat flow in its complex multilayer structure. In this work, we establish an experiment-informed approach that combines time-domain thermoreflectance (TDTR) with a step-by-step fitting strategy to extract the thermal conductivities and interfacial thermal resistances in HAMR media. These experimentally determined properties are directly implemented in finite-element method simulations, which show good agreement with reported spin-stand measurements. The TDTR-based experiment-informed modeling established in this study reliably describes a device-level behavior of HAMR and provides quantitative guidance for recording media design.