Patients on chemotherapy often use multiple resources to obtain information on wide-ranging issues, including prognosis, treatments, diet, exercise, and safety precautions. Websites from national and international cancer organizations are a common source of information. In this commentary, we discuss the strength of evidence associated with some common recommendations and provide examples of the unintended consequences of some of this advice. We hope that this commentary will stimulate more careful consideration of the risks and benefits when drafting these patient-facing materials.

10022 Background: Countries of the South Asian Association for Regional Cooperation (SAARC) are home to approximately 600 million children aged 0–14 years, representing a quarter of the global child population. We report incidence and mortality for childhood cancers among South Asia's eight nations to inform priority setting for research, programming for health services, and policy for childhood cancers. Methods: Incidence and mortality of new childhood cancer diagnoses for all cancers combined and for selected leading diagnostic groups were retrieved from the GLOBOCAN 2022 database of Afghanistan, Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, and Sri Lanka. Age-standardized incidence rates (ASR) and mortality-to-incidence ratios (MIR) were calculated, together with statistical examination of cross-country variation in diagnostic structure and survival trends. Results: In SAARC nations, 37,716 new childhood cancer diagnoses and 17,700 deaths were estimated in 2022. India contributed 68.7% of subregional cases (25,939), and Pakistan contributed 20.8% (7,841). Age-standardized rates differed and were greatest in Sri Lanka (10.4/100,000) and Pakistan (10.1) and lowest in Bhutan (2.3) and Bangladesh (3.7). Boys represented 60% of cases overall, with significant cross-country variation in sex distribution (χ²=18.0, p=0.012). Leukemia was the most common cancer in all countries (35–50%), followed by central nervous system (CNS) tumors. MIR ranged from 0.30 (Sri Lanka) to 0.57 (Afghanistan), reflecting extreme survival differences, with CNS tumors demonstrating disproportionately high mortality relative to incidence. Conclusions: SAARC childhood cancer trends reflect both population size influences and health system capacity, with a preponderance of leukemia but alarming survival gaps, particularly for CNS cancers. These findings highlight actionable priorities, including strengthening population-based childhood cancer registries, centralizing care pathways for CNS tumors, expanding standardized leukemia treatment protocols with abandonment-prevention strategies, and improving equitable access to specialty care through subregional collaboration.

Abstract Profiling the tumor microenvironment (TME) is fundamental to understanding immune, stromal, and tumor interactions influencing cancer diagnosis, progression, and therapeutic response. Physical immunohistochemistry (IHC) remains essential but is limited by reagent dependency, labor-intensive workflows, and tissue exhaustion. This study aimed to develop and validate a virtual IHC platform to reproduce biomarker staining directly from hematoxylin and eosin (H&E) slides, enabling high-fidelity TME characterization while conserving tissue. Using the ViewsML deep learning platform, a virtual TME panel was trained to predict five key biomarkers: CD31 (endothelial/vasculature), CD45 (pan-leukocyte), CD68 (macrophage), smooth muscle actin/SMA (activated fibroblasts and pericytes), and cytokeratin AE1/3 (tumor epithelium). 316 digitized lung cancer biopsy slides (Aperio GT450) encompassing over 150 million annotated cells were used. Data were split into training (70%), validation (15%), and test (15%) sets. Neural networks were optimized for biomarker-specific predictions, and model performance was evaluated using ROC AUC metrics and blinded pathologist review. Virtual biomarkers showed strong concordance with physical IHC, achieving AUCs of CD31=0.91, CD45=0.90, CD68=0.93, SMA=0.91, and CK AE1/3=0.90. Pearson’s correlations were 0.73, 0.72, 0.76, 0.73, and 0.76 (p<0.0001). Virtual stains preserved spatial and morphological features, including CD31-positive vascular frameworks at tumor-stroma boundaries, CD45-positive immune infiltration, CD68-positive macrophage aggregates, SMA-positive stromal reaction patterns, and CK-positive tumor epithelium. Quantitative per-cell biomarker expression enabled automated precise cell fraction, cell ratio, spatial proximity, and immune and macrophage clustering analysis. These findings demonstrate that ViewsML’s virtual IHC technology can reproduce physical staining from standard H&E slides, offering a scalable solution for digital TME profiling and conserving tissue. This approach supports translational research, virtual biomarker screening in drug development, and clinical application by enabling immune versus stroma-dominant tumor stratification to guide immunotherapy. Future work will extend this platform to more tumor types and incorporate multiplex virtual staining for broader digital pathology integration. Citation Format: Kenneth To, Christopher Jackson, Louis Vaickus, Lawrence Schobs, Rohan Kamra Lyons, Rafay Azhar. From single H&E to virtual immunohistochemical biomarker staining in the lung tumor microenvironment [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 75.

Successful immune checkpoint inhibitor (ICI) therapy occurs in only a fraction of melanoma patients, and yet all patients are susceptible to potentially serious ICI-related side-effects. No current biomarkers robustly predict ICI treatment response in melanoma patients. In this study we sought to identify methylome and transcriptome markers which have the potential to predict immunotherapy response in melanoma patients ahead of treatment with anti-PD1 ICI monotherapy. Using Infinium MethylationEPIC microarrays, we analysed DNA methylation profiles of >850,000 CpG sites in pre-treatment melanoma tissues from patients administered anti-PD-1 monotherapy as first-line treatment. In addition, we analysed transcriptomes using RNA-seq. DNA methylation and gene expression data were then statistically compared to patient response to anti-PD1 therapy. We identified 2579 DNA hypomethylation and hypermethylation alterations correlating with melanoma response to anti-PD1 therapy. An integrative analysis of DNA methylomes and transcriptomes identified a subset of 35 loci, 13 of which were significantly differentially methylated in both initial discovery and external validation datasets. Functional enrichment analysis of hypomethylated sites (p-value <0.05) in non-responders was associated with "Formation of the cornified envelope", "Regulation of epithelial cell proliferation", and "Purine-containing compound metabolic process". We have identified novel integrated DNA methylation and gene expression markers, which correlate with anti-PD1 treatment response in melanoma patients. These findings suggest a relationship between tumour-associated genomic DNA methylation, gene expression patterns, and anti-PD1 ICI immunotherapy response in melanoma patients.

BACKGROUND:CheckMate 8KX was a phase I/II study investigating the pharmacokinetics and safety of subcutaneous (SC) nivolumab±recombinant human hyaluronidase PH20 (rHuPH20). METHODS:Patients with advanced solid tumors who were immune-checkpoint inhibitor-naïve were included. In parts A and B, patients received one single dose of nivolumab SC 720 mg or 960 mg±rHuPH20 2000 U/mL followed by nivolumab intravenous (IV) 480 mg every 4 weeks (Q4W). Patients receiving nivolumab IV could switch to nivolumab SC 1200 mg+rHuPH20 2000 U/mL Q4W in part C. Patients in part D received nivolumab SC 1200 mg+rHuPH20 Q4W. The primary objective was to describe the pharmacokinetics of nivolumab SC±rHuPH20. Secondary objectives included assessing safety and immunogenicity of nivolumab SC. Exploratory objectives included evaluating patient experience and preference and characterizing biomarker measures of immune function (pretreatment and on-treatment peripheral blood and tumor biopsies). RESULTS:A total of 103 patients with various solid tumor types were treated between December 4, 2018 and September 7, 2022. Mean duration of injection was <5 min for nivolumab SC±rHuPH20. Nivolumab exposures over the first dosing interval increased with increasing dose; geometric-mean time to maximum concentration across doses was 117-167 hours (5-7 days). Nivolumab SC+rHuPH20 was well tolerated; grade 3/4 treatment-related adverse events occurred in 0%-11.1% of patients; antidrug antibodies were reported in some patients but were not associated with altered safety; no neutralizing antibodies were detected. Overall, patients preferred SC delivery over IV or had no preference. Increased tumor and peripheral pharmacodynamic biomarkers were observed postnivolumab SC, consistent with historical data for nivolumab IV. CONCLUSION:The pharmacokinetics of nivolumab SC have been well characterized and enabled dose selection for further study. Nivolumab SC has an acceptable safety profile, allows for rapid administration, and is preferred by more patients than nivolumab IV. These results encourage large-scale investigation of nivolumab SC. TRIAL REGISTRATION NUMBER:NCT03656718.

The purpose of this study was to use highly complete data from publicly and privately funded radiation therapy centers to describe national-level receipt of radiation therapy for those diagnosed with lung cancer in Aotearoa New Zealand. We linked national health datasets to New Zealand Cancer Registry lung cancer registrations from the period 2012–2019 (n = 18,081) to describe the radiation therapy delivery overall and in terms of treatment intent patterns, for all lung cancer registrations and by tumor type. We use marginal standardization and regression modeling to describe the extent to which these treatment factors vary by patient characteristic and population sub-group, independent of other factors. Around 40

Sample sizes of Sites 1-4 across ten experiments

Artificial intelligence applications in biomedicine face major challenges from data privacy requirements. To address this issue for clinically annotated tissue proteomic data, we developed a federated deep learning approach (ProCanFDL), training local models on simulated sites containing data from a pan-cancer cohort (n = 1,260) and 29 cohorts held behind private firewalls (n = 6,265), representing 19,930 replicate data-independent acquisition mass spectrometry runs. Local parameter updates were aggregated to build the global model, achieving a 43% performance gain on the hold-out test set (n = 625) in 14 cancer subtyping tasks compared with local models and matching centralized model performance. The approach's generalizability was demonstrated by retraining the global model with data from two external, data-independent acquisition mass spectrometry cohorts (n = 55) and eight acquired by tandem mass tag proteomics (n = 832). ProCanFDL presents a solution for internationally collaborative machine learning initiatives using proteomic data, for example, for discovering predictive biomarkers or treatment targets while maintaining data privacy. SIGNIFICANCE:A federated deep learning approach applied to human proteomic data, acquired using two distinct proteomic technologies from 40 tumor cohorts across eight countries, enabled accurate cancer histopathologic subtyping while preserving data privacy. This approach will enable the privacy-compliant development of large-scale proteomic artificial intelligence models, including foundation models, across institutions globally.

Distribution of cancer subtypes across Sites 2-4 and Experiments 1-10

Inherited dihydropyrimidine dehydrogenase (DPD) deficiency is a risk factor for severe 5-fluorouracil toxicity. We report a phenotyping approach (thymine challenge test) to prospectively determine DPD activity and the association with severe adverse events. The primary aim of this prospective study was to determine whether a thymine challenge test could prospectively identify patients at risk of severe toxicity from treatment with 5-fluorouracil/capecitabine in combination chemotherapy schedules or monotherapy. The focus was prediction of those at risk of ≥ grade 3 gastrointestinal toxicity. DPD activity was determined from the thymine/dihydrothymine (THY/DHT) ratio measured in a urine sample after a thymine test dose (250 mg, oral). Of the 166 patients, 11.7

Cell type enrichment analysis of top SHAP-ranked proteins across cancer subtypes

Per class AUROC for models using ProCan Compendium

Formalin-fixed paraffin-embedded (FFPE) tissues present an important resource for cancer proteomics. They are more readily available than fresh frozen (FF) tissues and can be stored at ambient temperature for decades. FFPE blocks are largely stable for long-term preservation of tumour histology, but the antigenicity of some proteins in FFPE sections degrades over time resulting in deteriorating performance of immunohistochemistry (IHC). It is not known whether FFPE sections that have previously been cut from blocks and used for liquid chromatography-mass spectrometry (LC–MS) analysis at a later time are affected by storage time or temperature. We determined the stability of FFPE sections stored at room temperature (RT) versus − 80 °C over 48 weeks. The stored sections were processed at different timepoints (n = 11) and compared to sections that were freshly cut from FFPE blocks at each timepoint (controls). A total of 297 sections (rat brain, kidney and liver stored at RT, − 80 °C or freshly cut) were tryptically digested and analysed on TripleTOF 6600 mass spectrometers in data-dependent acquisition (DDA) mode. Kidney and liver digests were also analysed in data-independent acquisition (DIA) mode. The number of proteins and peptides identified by DDA with ProteinPilot and some common post-translational modifications (PTMs) were unaffected by the storage time or temperature. Nine of the most common FFPE-associated modifications were quantified using DIA data and all were unaffected by storage time or temperature. Therefore, FFPE tissue sections are suitable for proteomic studies for at least 48 weeks from the time of sectioning.

Cohort details of ProCan Compendium

728 Background: Gemcitabine (GEM) plus nab-paclitaxel (NAB-PAC) is a standard first-line chemotherapy regimen for advanced/metastatic pancreatic ductal adenocarcinoma (PDAC), with median Progression Free Survival (mPFS) and Overall Survival (mOS) of 5.5 & 8.7 in the MPACT trial and 5.6 & 9.2 months (mo) in the NAPOLI 3 trial, respectively. Certepetide (formerly LSTA1 or CEND-1) is a novel cyclic peptide that improves targeted penetration of co-administered drugs into tumor and stroma, leading to potentially increased anti-neoplastic activity. ASCEND is a randomized phase II trial designed to investigate the impact of adding certepetide to GEM/NAB-PAC. (NCT05042128). Methods: Eligible participants (pts) with histologically confirmed advanced PDAC and ECOG 0-1 were randomized 2:1 to receive GEM/NAB-PAC plus certepetide (3.2 mg/kg) or placebo (PLA) on days 1, 8, & 15 of a 28-day cycle. Stratification was by age (<65/≥65 years), ECOG (0/1), presence of liver metastasis (Y/N) and trial site. The primary objective was to determine the effect of adding certepetide to GEM/NAB-PAC on PFS. Objective tumor response rate (OTRR), safety and OS were secondary objectives. This non-comparative phase II design targeted a PFS increase of 17% at 6 mo from 47% to 63%. A sample size of 65 patients in the certepetide arm was expected to have 80% power, with 95% confidence to exclude an uninteresting 6-mo PFS rate of 47%. Results: 95 pts (66 certepetide, 29 PLA) were enrolled between May 2022 to December 2023. 6-mo PFS in the certepetide and PLA groups was 49.0% (95% CI 36.4%, 60.5%) and 40.8 (95% CI 22.6%, 58.3%) respectively, with mPFS of 5.5 mo in both groups. mOS was 12.42 mo for the certepetide and 9.72 mo for the PLA. An OTRR of 38.3% (certepetide) and 26.9% (PLA) was observed. Of note, 4 complete responses were observed in the certepetide group vs. 0 in PLA group. In subjects with ECOG 0, 6 mo PFS was 68.1% (95% CI 48.7%, 81.4%) in the certepetide compared to 36.4% (95% CI 11.2%, 62.7%) in the PLA. Grade ≥3 toxicities were similar in both groups at 16% (certepetide) and 15% (PLA). Conclusions: The addition of certepetide is safe and despite showing no improvement in 6-mo PFS, a possible signal of benefit in OS and OTRR, including the 4 complete responses was observed, warranting further investigation. A further cohort of the ASCEND study evaluating the addition of a second dose of certepetide is ongoing. Clinical trial information: NCT05042128 .

background: The prevalence of diabetes has been increasing in Aotearoa New Zealand by approximately 7% per year, and is three times higher among Māori and Pacific peoples than in Europeans. The depth of the diabetes epidemic, and the expansive breadth of services required for its management, elevate the need for high-quality evidence on the projected future burden of this complex disease. methods: In this manuscript we have projected the prevalence of diabetes (type 1 and type 2 combined) out to 2040–2044 using age-period-cohort modelling. National-level data from central government on diabetes prevalence (Virtual Diabetes Register) were used to describe recent diabetes prevalence trends (2006–2019) by age group, calendar period and birth cohort, with these trends used to project diabetes prevalence out from 2020 to 2044. results: Aotearoa New Zealand will experience a significant increase in the absolute volume of prevalent diabetes, rising by nearly 90% to more than 500,000 by 2044. The age-standardised prevalence of diabetes will increase from around 3.9% of the population (268,248) to 5.0% overall (502,358). The prevalence and volume of diabetes diagnoses will increase most drastically for Pacific peoples, most notably Pacific females for whom diabetes prevalence is projected to increase to 17% of the population by 2044. conclusions: The increases in the future burden of diabetes mellitus projected here will heighten pressure on health services. Immediate action is required to reduce new cases of diabetes and other obesity-related illnesses. Fiscal policies to prevent these diseases, coupled with population-level interventions to more effectively manage and control diabetes, are effective tools for reducing disease burden.