Browse Articles

Alcohol and cancer risk: A review of messaging content on websites of medical, cancer, government, and nonprofit organizations and alcohol companies.

e22508 Background: Alcohol is a risk factor for multiple cancers, yet there is limited awareness of this risk relationship among the public in the United States (US) 1-5 . In December 2024 to January 2025, the US Surgeon General issued a health advisory 6 , emphasizing the link between alcohol and cancer, and the National Academies of Sciences, Engineering and Medicine (NASEM) 7 and the Interagency Coordinating Committee on the Prevention of Underage Drinking (ICCPUD) 8 both concluded that even low levels of alcohol use pose carcinogenic risks. Our study aimed to characterize the current alcohol and cancer risk-related messaging across websites of US health organizations in order to understand the extent to which public communication may be influencing the lack of awareness regarding alcohol and cancer risk. Methods: The websites of 157 medical, cancer, governmental, and non-profit organizations and alcohol-funded entities were selected. A standardized coding process identified the presence of alcohol and cancer risk-related messaging. Review of websites was conducted by two independent reviewers over the period between August to December 2025. Results: The analysis of 126 websites of National Medical societies revealed 72 of them had alcohol-related content. Among those 72 websites, 46 (64%) mentioned health risks associated with alcohol use, and 21 (29%) mentioned the cancer risk from alcohol. Notably, only six (13%) websites stated that there is no safe level of alcohol consumption. All governmental websites included statements about both health and cancer risks associated with alcohol, and that there is no safe level of alcohol consumption. Among eight cancer organization websites, six mentioned the cancer risk from alcohol and five stated that there is no safe level of alcohol consumption. Among seven alcohol-related nonprofit organizations, four mentioned the cancer risk from alcohol. Among eleven alcohol company websites, five mentioned the cancer risk associated with alcohol consumption. Conclusions: The lack of mention of the cancer risks associated with alcohol use may be contributing to the lack of knowledge about the cancer risks associated with alcohol use among the general public. These results suggest that there is an opportunity to improve the public’s knowledge regarding the harms of alcohol use, including its associated cancer risk.

Trends in U.S. pneumonia-related admissions among patients with multiple myeloma (2006–2022).

e19535 Background: Despite improvements in survival due to novel therapies, patients with multiple myeloma (MM) experience a disproportionate burden of infectious complications. Pneumonia (PNA) is a leading cause of hospitalization and death in this population, but contemporary data describing outcomes and costs of PNA-related admissions (ADMs) are limited. Methods: We conducted a serial cross-sectional study of PNA-related ADMs using the US Healthcare Cost and Utilization Project National Inpatient Sample (HCUP-NIS) from 2006 - 2022. ADMs among adults (≥ 18 years) with and without MM (PNA+MM and PNA-MM) were identified using ICD-9/-10 codes for PNA or MM in any position. National estimates of ADMs were obtained by applying discharge weights provided by HCUP-NIS. Parity between ICD-9 and -10 codes was approximated via general equivalency mapping. Similar cohorts were formed for patients with COVID-19 infection (COVID+MM and COVID-MM) for 2020 - 2022. Endpoints included number of ADMs, in-hospital mortality, length of stay (LOS), and inflation-adjusted median cost (2020 USD). Endpoints were stratified by age, sex, race, income quartile, payer, discharge disposition, and hospital characteristics. The Hospital Frailty Risk Score (HFRS) was used to explore frailty as a predictor of outcomes (2016+). ChatGPT 5.2 (OpenAI) was accessed from 11/2025 - 01/2026 to assist with R code used for cohort formation and data analysis. Study was reviewed and approved by HCUP (01/2025). Results: From 2006 - 2022, 284,450 weighted ADMs in the PNA+MM cohort were identified (mortality: 11.4%, median LOS: 6 days, median cost $48,119). Median cost in the PNA+MM cohort increased from $36,361 to $50,905 while in-hospital mortality decreased from 13.9% to 9.1% from 2006 - 2019. Inflation-adjusted total annual cost in the PNA+MM cohort increased from $855 million to $2.4 billion (181.5%) from 2006 - 2019. From 2006 - 2019, the proportion of PNA+MM ADMs relative to all PNA ADMs increased by 66.4% (r 2 = 0.95, p < 0.001), and the proportion of total costs attributable to PNA+MM ADMs relative to all PNA ADMs increased by 53.9% (r 2 = 0.67, p < 0.001). From 2020 - 2022, 22,935 weighted ADMs were identified in the COVID+MM cohort (mortality: 17.5%, median LOS: 6 days, median cost: $59,032). Compared with non-MM cohorts, the PNA+MM and COVID+MM cohorts were more likely to be older, male, and non-white (p < 0.001). The PNA+MM and COVID+MM cohorts were frailer and displayed higher mortality compared with non-MM cohorts (p < 0.001). Conclusions: PNA-related ADMs among patients with MM are increasingly costly, and there persist demographic and comorbidity-dependent disparities. Study limitations include mid-study change in ICD coding systems and variable clinician coding practices. Research around interventions aimed at reducing admission rates in this population (e.g. antimicrobial prophylaxis, intravenous immunoglobulin, vaccination) is warranted.

Clinical profile, treatment patterns, and survival outcomes of multiple myeloma patients in the Middle East: Real-world data from Bahrain.

e19576 Background: Real-world data from the Middle East in myeloma is limited, despite unique demographic factors that may affect outcomes. In this study we aimed to describe clinical profile, treatment patterns, and predictors of survival in myeloma patients treated in routine practice in Bahrain. Methods: We conducted a retrospective, cohort study of adult patients diagnosed with myeloma in the national oncology center in Bahrain from 2017 to 2023. Baseline clinical, and laboratory parameters, and treatment patterns (induction regimens, use of autologous stem cell transplantation (ASCT), and maintenance strategies), were recorded. Progression-free survival (PFS), overall survival (OS), and factors associated with survival were estimated. Results: One hundred and fifty-five patients were included in the study, the median age at presentation was 62 years (range 25–86 years). Most patients had advanced disease at presentation: 52% were ISS Stage III, and only 23% were Stage I. High-risk cytogenetics were identified in 15% of those tested, and circulating plasma cells were seen in 7% at diagnosis. The predominant immunoglobulin subtype was IgG (67%), followed by light-chain–only disease (19%), and IgA (13%). Kappa light chain restriction was more common (68%) than lambda (32%). Induction regimens demonstrated evolution over time in accordance with global paradigms- VCD, VTD and VRD regimens predominated pre-2020, and DVRD and DVD regimens predominated post-2020. Responses to first-line therapy were favorable: 76% complete response (CR), 12% very good partial response (VGPR), and 12% partial response (PR). 68% patients were ASCT-eligible, and 51% underwent ASCT. Depth of response improved further following transplantation: 90% CR, 8.5% PR, and 1.4% stable disease. Post-transplant maintenance therapies included IMiD (65%), proteasome inhibitor (15%), dual maintenance (11%), or single anti-CD38–based maintenance (9.1%). At a median follow-up of 3.4 years, the median PFS in R-ISS stages I, II and III patients were 12.8, 6, and 2.6 years respectively. On multi-variable analysis (MVA), baseline renal impairment had worse OS (HR 3.29, 95% CI, 1.16 – 9.34), while ASCT was strongly associated with improved OS (HR 0.11 95% CI, 0.03 – 0.44). Patients with circulating plasma cells at baseline, and IgA subtype showed trend toward inferior survival (p=0.05). R-ISS stage III was associated with inferior PFS (HR 6.27, 95% CI 1.50 – 26.28) on MVA. Conclusions: In this real-world cohort of Middle Eastern ethnicity, myeloma frequently presented at advanced stages, however modern treatment approaches, including ASCT, resulted in comparable survival outcomes to that described in literature. Our study provides region-specific evidence that complements clinical trial data, and supports the effectiveness of modern myeloma treatments in Middle Eastern populations.

Influence of genetic ancestry on UV mutational signatures linked to immunotherapy response in melanoma, beyond TMB.

3087 Background: Melanoma is a UV-driven malignancy with substantial biological heterogeneity, yet genomic profiling efforts have historically underrepresented diverse ancestral backgrounds, limiting insights into population-level differences in tumor biology and treatment response. Methods: In this retrospective study, a supervised machine learning framework was developed to infer genetic ancestry across 483,454 tumors profiled by tumor-only whole-exome sequencing (WES) at Caris Life Sciences (Phoenix, AZ), using 36,962 ancestry-informative variants derived from globally diverse Genome Aggregation Database (gnomAD) reference individuals (n=4,150). Survival outcomes inferred from claims data were evaluated using Cox proportional hazard ratios (HR) with 95% confidence intervals (CI) and log-rank p-values. Associations between features were evaluated using standard statistical tests and regression models. Results: We applied this framework to cutaneous melanoma (N=8,872). The UVB-related COSMIC mutational signatures SBS7a and SBS7b were markedly enriched in tumors from patients (pts) of European (EUR) ancestry (93.2% and 75.6%, respectively) but were uncommon in African ancestry tumors (6.9% and 5.2%). Within EUR ancestry pts, higher EUR ancestry fraction was associated with higher SBS7a and SBS7b prevalence (p=9.0x10 -4 and 1.5x10 -4 ), consistent with known differences in pigmentation biology and melanoma etiology. Among 3,587 melanoma pts treated with first-line immune checkpoint inhibitors (ICI), SBS7a positivity was strongly associated with improved overall survival (OS HR [95% CI] = 0.55 [0.46–0.66]) and time to next treatment (TTNT HR=0.58 [0.49–0.69]; all p<0.001), outperforming tumor mutational burden (TMB) as a prognostic biomarker. SBS7a+ remained prognostic in a multivariable Cox analysis accounting for TMB, age, sex, and the first two ancestry principal components. Furthermore, both SBS7a+ TMB-high and SBS7a+ TMB-low tumors demonstrated superior outcomes compared to SBS7a‒ TMB-low tumors (OS HR=0.49 [0.41–0.60] and 0.62 [0.51–0.74], respectively; all p<0.001). Similar patterns were observed in cutaneous squamous cell carcinoma. To assess specificity for UV-driven biology, we evaluated two non-sun-exposed malignancies: uveal melanoma and renal cell carcinoma (RCC). SBS7a was virtually absent in uveal melanoma and RCC (6.7% and 6.9%, respectively, among EUR ancestry pts) and did not stratify ICI outcomes in either cancer. These negative controls suggest that SBS7a does not arise without UV exposure and is not a general immunotherapy biomarker. Conclusions: These findings establish a scalable genetic ancestry framework, reveal ancestry-linked differences in melanoma mutagenesis, and demonstrate that SBS7a provides clinically meaningful, TMB-independent prognostic information specific to UV-driven tumors.

A prospective single-arm study of irinotecan liposome II combined with adebrelimab, gemcitabine, and nab-paclitaxel for conversion therapy in locally advanced pancreatic cancer.

TPS4265 Background: Locally advanced pancreatic cancer (LAPC) has an extremely poor prognosis, with only a minority of patients eligible for radical surgery at initial presentation. Conversion therapy is critical for improving resectability and survival. Currently, no standard conversion regimen exists for LAPC, and gemcitabine plus nab-paclitaxel (AG regimen) is widely used. Irinotecan liposome has favorable tumor targeting and manageable toxicity; the PAN-HEROIC-1 study showed that irinotecan liposome II combined with 5-fluorouracil/leucovorin (5-FU/LV) significantly prolonged overall survival in Chinese advanced pancreatic cancer patients. Adebelimab, a programmed death-ligand 1 (PD-L1) inhibitor, exerts synergistic antitumor effects with chemotherapy in solid tumors. Herein, we evaluate the efficacy and safety of Irinotecan Liposome II plus Adebelimab and AG regimen as conversion therapy for LAPC to improve surgical conversion rates and survival benefits. Methods: This prospective, single-arm, open-label study plans to enroll 45 patients with pathologically confirmed locally advanced pancreatic ductal adenocarcinoma. Key inclusion criteria: 18-75 years old; Eastern Cooperative Oncology Group (ECOG) PS 0-1; adequate major organ function; no central nervous system metastasis; effective contraception for reproductive-aged subjects (pregnancy excluded); informed consent. Exclusion criteria: non-ductal pancreatic cancer; active autoimmune diseases; severe infections; major surgery or investigational drugs within 4 weeks prior to enrollment; other investigator-judged ineligibility. Primary endpoint is surgical conversion rate; secondary endpoints include event-free survival, overall survival, R0/R1 resection rate, objective response rate, pathological complete response rate, and safety profile. Treatment Regimen: Irinotecan Liposome II (40 mg/m², IV infusion 90±30 min, Day 1, q2w) + Adebelimab (1200 mg, IV infusion 30-60 min, Day 1, q4w) + gemcitabine (1000 mg/m², IV infusion > 30 min, Days 1,8,15, q4w) + nab-paclitaxel (125 mg/m², IV infusion > 30 min, Days 1,8,15, q4w). Registration Number: ChiCTR 2500108269. Research Sponsor: Hengrui Pharma. Clinical trial information: ChiCTR2500108269.

Radiomics feature robustness of photon-counting CT and DECT in chest tumor phantom: A dual-center, multi-device study.

e20011 Background: Robustness of radiomics features across CT platforms is essential for clinical translation of quantitative imaging biomarkers. This study evaluated radiomics feature robustness derived from photon-counting detector CT (PCD-CT) and multiple dual-energy CT (DECT) systems using standardized chest tumor phantoms. Methods: Twelve chest tumor phantoms with different sizes and texture patterns were constructed using a CRIS chest phantom. All phantoms were scanned at two centers on four energy CT platforms: PCD-CT, dual-layer DECT, dual-source DECT, and rapid kV-switching DECT. Images were reconstructed as conventional images, virtual monoenergetic images at 40, 70, and 100 keV, and virtual non-contrast images. Tumor-specific regions of interest were delineated, and radiomics features were extracted after standardized preprocessing. Inter-scanner robustness was assessed using concordance correlation coefficient, coefficient of variation, and quartile coefficient of dispersion. Results: Radiomics feature robustness varied substantially by reconstruction type. In conventional images, the highest inter-scanner agreement was observed between dual-source and rapid kV-switching DECT, whereas most other device pairs showed limited robustness. Among texture phantoms, acceptable agreement was additionally observed between PCD-CT and dual-layer DECT. In virtual non-contrast images, robustness between PCD-CT and DECT platforms was consistently poor, while agreement among DECT systems remained acceptable. Low-energy virtual monoenergetic imaging at 40 keV significantly improved robustness between PCD-CT and DECT platforms, whereas robustness decreased at higher energy levels. Similar trends were observed across different radiation dose settings. Conclusions: Radiomics feature robustness across modern energy CT platforms is highly dependent on reconstruction strategy. Low-energy virtual monoenergetic imaging improves inter-scanner robustness, including for PCD-CT, whereas virtual non-contrast reconstruction compromises harmonization between PCD-CT and DECT systems. These findings support reconstruction selection and radiomics standardization in multicenter thoracic oncology studies.

Neoadjuvant chemoradiotherapy with or without PD-1 blockade in pMMR/MSS low rectal cancer patients (CHOICE II): A multi-center, open-label, randomized controlled trial.

3640 Background: Neoadjuvant chemoradiotherapy (nCRT) combined with PD-1 blockade has shown great potential for improving the clinical complete response (cCR) rate in low rectal cancer patients in a previous exploratory clinical trial. This study aimed to further verify the efficacy and safety of nCRT with PD-1 blockade in patients with pMMR/MSS low rectal cancer. Methods: We conducted a prospective, multicenter, randomized, open-label trial (NCT05215379) with two parallel groups across 11 centers. Patients with confirmed cT 1-3a N 0-1 M 0 rectal adenocarcinoma of the pMMR/-MSS type, with an inferior margin of 5 cm from the anal verge, were randomly assigned to either the intervention group, receiving 50 Gy (2 Gy/d*25d) radiotherapy, 4 cycles of PD-1 blockade (sintilimab) and 2 cycles of CAPOX, or the control group, receiving 50 Gy (2 Gy/d*25d) radiotherapy and 2 cycles of CAPOX. After completion of neoadjuvant treatment, patients were reassessed to determine clinical efficacy. If patients achieved cCR, the Watch-and-Wait (W&W) approach was performed. If near-cCR (ncCR) was achieved, patients were re-evaluated by a multidisciplinary team to determine whether W&W, local excision, or total mesorectal excision (TME) should be performed. Patients assessed with an incomplete clinical response (iCR) received TME. The primary endpoint was the cCR rate, and the key secondary endpoints included the organ preservation (W&W and local excision) rate, disease free survival (DFS), and overall survival (OS). Results: From July 2022 to December 2024, a total of 201 patients with low rectal cancer were screened at 11 medical centers and 180 patients were finally randomly assigned. A total of 169 patients were included in the primary analysis (85 patients in the intervention group and 84 in control group). The cCR rates were 36.5% in the intervention group compared to 17.9% in the control group (p=0.007). There were 48 and 51 patients who underwent surgery in the intervention and control groups, respectively, and the pathological complete response (pCR) rates were 25.0% and 15.7% (p=0.249). The overall complete response (cCR+pCR) rates were 50.6% and 27.4% (p=0.002) in the intervention and control groups. The overall occurrence rates of adverse events (AEs) were 72.9% in the intervention group compared to 66.3% in the control group (P>0.05), and the occurrence of severe AEs (Grade III or IV) was comparable between the two groups (7.1% vs 3.8%, p=0.34). Due to the limited follow-up duration, the DFS and OS data are immature. Conclusion: In this trial, the results showed that nCRT combined with sintilimab significantly improved the cCR rate in patients with pMMR/MSS low rectal cancer with a tolerable safety profile. Clinical trial information: NCT05215379 .

Real-world usability of a clinical trial knowledge platform in a community cancer network.

e13712 Background: Clinical trial participation in the United States remains suboptimal, and limited point-of-care trial awareness contributes to missed opportunities for enrollment. Although public registries are comprehensive, they are not optimized for rapid use during clinical encounters. A clinical trial knowledge platform that consolidates protocol information, surfaces eligibility criteria, and generates concise AI-assisted summaries was previously evaluated at a comprehensive cancer center and demonstrated high feasibility and user engagement. We assessed the real-world usability of this platform in a large community cancer network. Methods: A mixed-methods usability evaluation was conducted, including 13 semi-structured interviews and 33 usability surveys. Thirteen participants completed a standardized 30-minute guided session using the platform, followed by a 30-minute interview, during which they performed trial-search, eligibility-review, and summary-review tasks reflecting typical point-of-care use. All participants completed a 10-item modified Mobile Health App Usability Questionnaire (mMAUQ; Likert scale 1–7; favorable ≥5). Quantitative results were summarized descriptively. Qualitative data were analyzed using grounded-theory and interpreted through the PACMAD usability model. Results: Usability scores were consistently high across core domains, including ease of navigation, learnability, interface clarity, and rapid access to relevant trial information (mean scores 6.5–6.7). Participants reported strong perceived clinical utility—supporting trial discussions, referral awareness, and willingness to recommend the tool—with workflow integration emerging as the primary area for further optimization (mean 5.9). Qualitative findings aligned with quantitative results: participants identified rapid access to eligibility criteria for prescreening during time-limited visits as the most clinically valuable feature and described the platform as intuitive, easy to recall, and cognitively light. Suggested refinements focused on workflow integration, including interoperability with clinical systems and potential ambient support during patient encounters. Conclusions: In a community oncology setting—where limited access to trial information can exacerbate disparities in enrollment—a clinical trial knowledge platform demonstrated excellent usability and strong perceived clinical utility. Findings suggest that practical integration of trial knowledge into time-limited clinical encounters—beyond trial matching alone—may be a critical determinant of whether digital clinical trial tools translate into provider behavior change and increased participation in research. Evaluation of the platform’s effect on trial accrual and equitable access to research opportunities across diverse care environments is underway.

Factors affecting outcomes in early-onset vs average-onset head and neck cancers: A multicenter retrospective database study.

e18108 Background: The incidence of early-onset (EO) head and neck squamous cell carcinoma (HNSCC) is increasing and appears to be associated with improved survival compared with average-onset (AO) disease, suggesting potential biological differences. However, data are limited. We compared demographics, primary tumor sites, metastatic patterns, and outcomes between EO and AO HNSCC. Methods: We performed a multicenter retrospective study using the TriNetX federated, de-identified electronic medical record network. Patients with HNSCC were identified with diagnosis as the index event. EO was defined as diagnosis before age 50 and AO as age ≥51. Primary and metastatic sites were identified using ICD-O and ICD-10 codes. Propensity score matching (1:1 greedy nearest-neighbor, caliper 0.1) was performed for sex, race, primary site, HPV status, and comorbidities. Kaplan–Meier survival analysis and comparative statistics were performed within TriNetX. Results: Among 468,496 patients, 41,596 (8.9%) had EO HNSCC. EO patients were more often female (43.1% vs 29.6%), Asian (12.2% vs 7.2%), and African American (8.2% vs 7.1%), and less often White (40.4% vs 61.7%) (all p<0.0001). Primary tumors were more frequently nasopharyngeal (15.0% vs 4.8%) and parotid (7.7% vs 5.3%), and less frequently tonsillar (5.3% vs 7.6%), oropharyngeal (4.6% vs 6.2%), and base of tongue (2.0% vs 5.7%) (all p<0.0001). After matching, each cohort included 39,535 patients. Median overall survival was not reached in the EO cohort (3,450 deaths, 8.7%) and was 6,011 days in the AO cohort (7,823 deaths, 19.8%) (HR 0.49, 95% CI 0.47–0.51; p<0.0001). Stage IV disease developed less frequently in EO patients (12.6% vs 16.5%, p<0.0001). EO patients had fewer lung (3.1% vs 4.0%), lymph node (11.4% vs 13.5%), liver (1.6% vs 2.1%) all p<0.0001, and bone metastases (3.5% vs 3.7%, p=0.03), with no difference in brain metastases. There was no difference in receiving radiation treatment and those that received surgery within 6 months of diagnosis. Conclusions: EO HNSCC differs significantly from AO disease in demographics, tumor site distribution, metastatic patterns, and survival. EO disease is associated with fewer metastases and superior overall survival, suggesting distinct tumor biology and supporting the need for further mechanistic studies.

Prognostic potential of PROSTest for predicting outcome in ADT-treated Black South African prostate cancer patients.

5104 Background: Prostate cancer (PCa) is the second most common cancer in men with disproportionately aggressive disease observed in black African and Caribbean men. Androgen deprivation therapy (ADT) is commonly used as treatment but there is no personalized strategy to guide patient selection. We investigated the prognostic value of PROSTest before ADT initiation and after the first ADT cycle. Methods: We performed a descriptive study of PROSTest with cross-sectional and longitudinal arms in black PCa patients receiving or intended to receive ADT. An ADT-naïve PCa cohort ( n =50) intended to start ADT was enrolled as well as an ADT-exposed cohort ( n =80) 6-12, 12-24 or >24 months after starting ADT. Goserelin acetate alone or in combination with cyproterone acetate was administered. The cohorts were followed up 4-monthly for 36 months with evaluation of survival and treatment response (responding, in remission, recurrence, progression, castration resistance). PSA levels were measured using the Architect Total PSA chemiluminescent immunoassay. Whole blood samples were collected into deproteinizing RNA-buffered tubes for determination of the PROSTest activity score using a multiplexed PCR assay. Statistical analysis used non-parametric statistics and Fisher’s exact test in GraphPad Prism v8. Results: Untreated metastatic PCa patients had the highest baseline PROSTest scores (71.8-85.3). Untreated patients with organ-confined disease were subdivided into group A (PROSTest scores ≤32) and group B (PROSTest scores ≥70). ADT-treated patients enrolled after starting ADT were subdivided into group C (PROSTest scores 9.8-58.3) and group D (PROSTest scores 70.0 – 90.2). Groups A and C showed improved survival and lower rates of recurrence, progression and castration resistance at 36 months compared to groups B and D (p<0.05). Patients that required treatment re-initiation or escalation due to rising PSA all had high enrolment PROSTest scores ≥70. The groups did not differ significantly in terms of age, CAPRA score, Gleason grade, PSA levels, nadir PSA, time to nadir or SUVMAX ( p >0.05). Conclusions: The PROSTest activity score determined pre- or post-ADT (first cycle) identifies PCa patients at risk for poor treatment outcomes.

Validation of SNHG11 as a prognostic and predictive biomarker for anti-EGFR benefit in colorectal cancer using real-world data.

3138 Background: Long non-coding RNAs (lncRNAs) regulate colorectal cancer (CRC) initiation, progression, and therapeutic response through interactions with oncogenic and microenvironmental pathways. SNHG11 has been implicated in CRC through epigenetic and c-Myc–driven transcriptional programs. In CALGB/SWOG 80405, higher tumor SNHG11 expression correlated with improved outcomes, particularly anti-EGFR therapy (Bartolini 2025, ASCO 43; 16_suppl). We comprehensively evaluated molecular correlates with SNHG11 and validated the prognostic and predictive value in an independent real-world CRC cohort. Methods: In total, 15,456 CRC underwent DNA (592-gene or whole exome) and RNA (whole transcriptome) sequencing at Caris Life Sciences. Tumors were stratified by SNHG11 expression quartiles, comparing low (Q1) vs high (Q4). CMS classification was assessed using RNAseq. Associations were assessed using X 2 /Fisher’s exact with p-values adjusted for multiple comparisons (q<0.05). Clinical outcome was obtained from insurance claims. Overall survival (OS) was calculated from tissue collection to last contact and time-on-treatment (ToT) from treatment start to end. Cox proportional hazards model generated hazard ratio (HR) and log-rank p-values. Results: SNHG11 Q1 tumors were associated with higher right-sided prevalence (28% vs 18%, p<0.001), lower left-sided prevalence (50% vs 62%, p<0.001), and older age than Q4 (64 vs 62, p<0.001). Consistent with the findings in CALGB/SWOG 80405, Q4 were associated with CMS2 (53% vs 19%, p<0.001) while Q1 with CMS4 (49% vs 21%, p<0.001). Q1 tumors showed increased MSI-H/dMMR, TMB-H and PD-L1 expression (10% vs 3%, 15% vs 5% and 5% vs 2%, respectively, all p<0.001). TP53 and APC mutations were enriched in Q4 vs Q1 (84% vs 69% and 85% vs 72%, p<0.001) whereas KRAS (49% vs 41%), BRAF (13% vs 7%), ARID1A (12% vs 8%), RNF43 (6% vs 2%) and CTNNB1 (3% vs 2%) mutations were higher in Q1 vs Q4 (p<0.001). Q4 was associated with longer OS vs Q1 (mOS 30.8 vs 27.8 months; HR 0.93; 95% CI 0.89–0.97, p<0.001). After adjusting for MSI status, tumor sidedness, age, race, and anti-EGFR therapy, Q4 remained a significant independent predictor of improved OS compared to Q1 (adjusted HR=0.93; 95% CI: 0.89-0.96; p<0.001). Q4 was significantly associated with longer ToT in patients receiving anti-EGFR therapy vs Q1 (7.4 vs 5.8 months, HR 0.86; 95% CI 0.79–0.93, p<0.001) with no association observed with bevacizumab ToT. Conclusions: In this large independent real-world cohort, we provided additional molecular insight into a key lncRNA SNHG11 in CRC. Consistent with previous findings in a large randomized controlled trial, we confirm that elevated SNHG11 expression was associated with CMS2, left-sided tumors, longer OS and longer TOT of anti-EGFR therapy, supporting SNHG11 as a promising prognostic biomarker and a candidate predictive biomarker for anti-EGFR benefit in CRC.

Predicting neoadjuvant chemoradiotherapy response and postoperative recurrence in locally advanced rectal cancer using a preoperative large language model.

3614 Background: Neoadjuvant chemoradiotherapy (nCRT) is standard for locally advanced rectal cancer (LARC), yet responses are heterogenous and early identification of responders remains challenging. We developed a Transformer-based large language model that analyzes fragment-level patterns in circulating cell-free DNA (cfDNA) to predict pathological complete response (pCR) and postoperative recurrence risk independently of somatic mutation calling. Methods: A total of 510 plasma samples were collected from 102 LARC patients at five perioperative time points. The model was trained on post-nCRT cfDNA data from a training cohort (N = 62) to predict pCR, evaluated in an internal validation cohort (N = 40) and an external validation cohort (N = 96). Associations between model prediction scores, nCRT response, and recurrence-free survival (RFS) were assessed across time points. Fragment-level features contributing to model predictions were analyzed to provide biological interpretability. Results: The model demonstrated robust predictive performance, with AUCs of 0.909 (95% confidence interval [CI]: 0.812-1.000) in the training cohort, 0.903 (95% CI: 0.785-1.000) in the validation cohort, and 0.835 (95% CI: 0.754-0.916) in the external cohort. Patients achieving pCR had significantly higher prediction scores than non-pCR patients (Wilcoxon p < 0.001), and scores inversely correlated with tumor regression grade (Jonckheere-Terpstra test p < 0.001). At the predefined cutoff corresponding to 95% specificity in the training cohort, sensitivities were 0.73, 0.64, and 0.44 across datasets. High post-nCRT and post-surgical prediction scores were associated with improved RFS (log-rank p = 0.012 and p = 0.024, respectively), and all relapsed patients had consistently low post-nCRT scores. The top 1% of cfDNA fragments, ranked by model importance, were predominantly 50-100 bp with ~ 40% GC content, and the proportion of subnucleosomal particles was significantly higher in pCR versus non-pCR samples, providing biological insights into features driving the model. Conclusions: We developed a robust, mutation-independent framework for predicting nCRT response and postoperative recurrence risk in LARC, supporting individualized treatment decisions and post-surgical surveillance strategies.

Toxicity and efficacy characterization of doxorubicin and trabectedin association versus doxorubicin alone in patients over 70 years with leiomyosarcoma: Analysis from LMS02 and LMS04 studies.

11515 Background: LMS02 (phase II) and LMS04 (randomized phase III) trials demonstrated the superiority of a combination regimen with Doxorubicin (Doxo, 60 mg/m²) and Trabectedin (Trab, 1.1 mg/m²) D1-D21 (Doxo+Trab) versus Doxorubicin (Doxo, 75 mg/m²) alone in the first line setting for metastatic or unresectable leiomyosarcomas with median Overall Survival (mOS) of 33 months (95% CI 26-48 months) versus 24 months (95% CI 19-31 months) (HR for death 0.65; 95% CI 0.44-0.95), respectively. The incidence of adverse events (AE), notably Grade 3-4 AE, was higher with the combination regimen compared to Doxo alone (97% versus 56%, p<0.001, respectively), although these were manageable. As more than one-third of soft-tissue sarcomas occur in patients (pts) over 65-year-old (yo), a better understanding of the risks and benefits of Doxo+Trab association in elderly patients is needed. Methods: Data from case report forms of pts aged ≥70 years enrolled in LMS02 and LMS04 were retrospectively analyzed to assess efficacy and toxicity of Doxo+Trab versus Doxo monotherapy. Results: 44 pts of 70 yo and older were included: 13 pts in LMS02 (Doxo+Trab 6 cycles, without maintenance), 15 pts in LMS04 arm B (Doxo+Trab : 6 cycles, followed by Trab maintenance for up to 17 cycles), and 16 pts in LMS04 arm A (Doxo monotherapy). The mean age was 74 yo (range 70-86), all pts had ECOG PS 0-1, and lung was the most frequent metastatic site. Median OS was 24 months with Doxo, 39 months with Doxo + Trab, and 48.9 months with Doxo + Trab-Trab (LMS04 cohort). Median PFS was 6.7 months with Doxo, 13.2 months in the Doxo + Trab arm, and 20.5 months with Doxo + Trab - Trab Toxicity analysis included patients from LMS02 and LMS04 as follows: 28 pts treated with Doxo+Trab (from both LMS02 and LMS04) and 16 pts treated with Doxo monotherapy in LMS04. The median number of received cycles was 6 in both groups. Twenty-two pts (80%) treated with the combination experienced Grade ≥3 AE, including 13 pts with Grade 4 events. These toxicities were predominantly hematologic; with 12% febrile neutropenia and 9% asthenia. In the monotherapy arm, ten pts (63%) had Grade ≥3 AE, including 5 pts with Grade 4 events, with asthenia (17%) and febrile neutropenia (17%) being the most frequent. The lack of statistical significance (p=0.2) should be interpreted cautiously due to limited sample size. Conclusions: For pts over 70 yo, the efficacy of the combination Doxo+Trab followed by trabectedin maintenance in advanced LMS in the first line setting, as measured by PFS and OS, did not differ from that in younger patients. Toxicity analysis suggests that age alone is not a limiting factor for this combination. Therefore, age-related frailty should not prevent patients from receiving a potentially life-prolonging treatment. A global geriatric evaluation and active supportive care may make treatment safer.

GEC-DTSP: A GNN–RL-based Edge–Cloud Digital Twin framework for real-time traffic forecasting and adaptive signal control

Urban traffic networks exhibit highly dynamic and nonlinear spatiotemporal interactions that require predictive modeling and adaptive control mechanisms capable of operating under low-latency constraints. This study proposes a GNN–RL-based Edge–Cloud Digital Twin framework for real-time traffic forecasting and adaptive signal control. At the network edge, multi-source traffic data collected from roadside sensors are processed on distributed edge devices to perform multi-step prediction of traffic flow, vehicle density, and congestion states. The forecasting module integrates Graph Convolutional Networks (GCNs) to capture spatial dependencies across the road topology with Long Short-Term Memory (LSTM) units and a Transformer-based predictor to model short- and long-range temporal dynamics. These predicted traffic states are transmitted to a cloud-level Digital Twin Engine, which performs data fusion, state estimation, calibration, and scenario-based simulation to maintain a continuously updated virtual representation of the physical traffic network. Using the forecasted states as inputs, a deep reinforcement learning optimization module performs adaptive signal phase control to minimize average vehicle delay and maximize intersection throughput. The overall framework operates as a closed feedback loop integrating edge-level spatiotemporal forecasting, cloud-level synchronization and simulation, and reinforcement learning–based control policy optimization. Experimental evaluation demonstrates a 17% reduction in average vehicle waiting time and significant improvements in forecasting performance measured using MAE and RMSE, with strong robustness to missing and noisy data conditions. The proposed architecture provides a scalable and low-latency solution for data-driven traffic prediction and signal control within an edge–cloud digital twin environment.

Bioinformatics models in drug delivery: Predicting biomaterial-biological interactions for targeted therapies

Heterogeneity, dynamics and organelle interactions of lipid droplets

Galvanic Replacement Synthesis Enabled by Gallium‐Based Liquid Metal: A Powerful Route for Material Design and Versatile Applications (Adv. Mater. 35/2026)

Wafer‐Scale 2D High‐Entropy Transition Metal Dichalcogenide Thin‐Film Catalysts for Efficient and Durable Photoelectrochemical Hydrogen Production

ABSTRACT Photoelectrochemical (PEC) performance of conventional 2D transition metal dichalcogenides (TMDs) in hydrogen evolution reaction (HER) is constrained by the limited selection of metal cations, predominantly MoS 2 , whose inert basal planes and unstable 1T phases hinder PEC efficiency. High‐entropy TMDs, in which local lattice distortion and charge redistribution occur within a van der Waals layered structure, are expected to overcome these intrinsic limitations by improving catalytic activity, photocarrier dynamics, and phase stability. Here, we demonstrate a wafer‐scale 2D high‐entropy (MoWTaNbRu)S 2 thin‐film catalyst with distorted 1T phase on p ‐Si photocathode for PEC‐HER. The high‐entropy effect induces substantial electronic redistribution, enhancing the contribution of d ‐orbitals near the Fermi level and optimizing hydrogen adsorption energetics. PEC kinetic analyses, including intensity‐modulated photocurrent spectroscopy, demonstrate that (MoWTaNbRu)S 2 markedly suppresses the recombination of photogenerated charge carriers, enabling more efficient charge extraction and accelerated interfacial reaction kinetics. Furthermore, the high‐entropy‐driven stabilization of the metastable 1T phase ensures excellent durability of the photocathode. As a result, the (MoWTaNbRu)S 2 /TiO 2 / p ‐Si photocathode shows a remarkable photocurrent density and stability for over 100 h, outperforming single‐metal TMDs. This study demonstrates how configurational entropy enhances catalytic activity, photocarrier transport, and phase stability of TMDs, establishing a general design principle for next‐generation PEC catalysts.

Cold Atmospheric Plasma‐Activated Decellularized Extracellular Matrix Gel as a Tumor‐Infiltrating Immunoactivation Platform for Post‐Surgical Cancer Immunotherapy

ABSTRACT Surgical resection is a frontline treatment for many solid tumors; however, residual tumor cells in the surgical cavity often lead to recurrence and metastasis. Adjuvant therapies are used to mitigate this risk but are frequently limited by systemic toxicity and variable efficacy. Here, we present an injectable, cold atmospheric plasma (CAP)‐loaded decellularized tumor extracellular matrix (DECM) hydrogel (denoted as CAP‐DECM gel) as a novel in situ tumor‐infiltrating immunoactivation platform for post‐surgical cancer immunotherapy. These gels combine two critical functionalities: the attraction of residual tumor cells by DECM‐derived chemokines and cytokines, and the local induction of immunogenic cell death (ICD) through CAP‐derived reactive species. Studies demonstrate that CAP‐DECM gels effectively recruit tumor cells, promote ICD hallmarks, activate various immune cells, including dendritic cells and macrophages, and elicit robust T‐cell responses. In murine post‐resection melanoma and breast cancer models, CAP‐DECM gels significantly suppressed tumor recurrence, reprogrammed the tumor microenvironment toward an immune‐supportive phenotype, and triggered systemic anti‐tumor immunity. Furthermore, combining CAP‐DECM gels with anti‐PD‐L1 checkpoint blockade therapy enhanced long‐term survival and conferred resistance to tumor rechallenge. Our results suggest that this tumor‐infiltrating immunoactivation platform transforms the surgical cavity into a self‐contained immune activation depot and offers a promising, personalized strategy for preventing tumor relapse.

Beyond Earth: Resilience of Quasi‐2D Perovskite Solar Cells in Space

ABSTRACT Perovskite solar cells (PSCs) offer unique advantages for space‐based energy harvesting, combining cost‐effective manufacturing with flexible, high power‐to‐weight devices that can reduce payload mass in deployable structures. Despite this promise, few reports have demonstrated the viability of this technology in realistic, space‐based scenarios, where they are subjected to large temperature variations and hard radiation. Here, we present a comprehensive analysis of PSC performance in low Earth orbit (LEO). The champion rigid cell exhibited relatively stable in‐orbit performance at ∼80% of initial efficiency over a 44‐day measurement interval that concluded nearly 100 days after launch, corresponding to ∼1600 orbital eclipse cycles and temperature ranges from −25 to 35°C. Mission data was systematically compared with laboratory measurements of rigid and ultrathin flexible PSCs across temperatures from −80 to +80°C and upon exposure to high‐energy proton radiation. Flexible devices retained over 92% efficiency after a proton dose equivalent to 50 years in orbit. Despite this radiation tolerance, mitigating pre‐flight environmental degradation remains a challenge for ultrathin substrates. Combined, this study bridges the gap between short suborbital demonstrations and long‐term orbital performance, highlighting the potential of PSCs as a low‐cost, resilient alternative for light harvesting, even in harsh space environments.