Search Results (2,688)

Breast cancer cells exhibit a reversed pH gradient and metabolic plasticity that promote proliferation, invasion, and resistance to therapy. Natural products such as chlorogenic acid (CGA) and cinnamaldehyde (CA) have shown emerging anticancer potential. However, their effects on intracellular pH and metabolic transport systems remain undefined. Therefore, the aim of this study was to characterize these parameters in breast cancer and non-tumorigenic breast cells. This study evaluated the physiochemical properties of CGA and CA using LC–MS, under pH conditions (pH 1.2, 7.4, and 9.0) mimicking the gastrointestinal track (GIT). Additionally, LC–MS-based human liver microsome (HLM) assays with NADPH were used to evaluate susceptibility to CYP-mediated metabolism to evaluate first-pass metabolic stability. Intracellular uptake kinetics were quantified at multiple time points using LC–MS. Following CGA:CA treatment, intracellular pH (pH_i) was measured in cancerous MDA-MB-231 and non-tumorigenic MCF-10A breast cell lines using SNARF-1 targeted ratio-metric fluorescence approach. Expression of OATP1B1, GLUT1, and MCT1 were analyzed by Western and immunofluorescence respectively, to assess potential cellular uptake of CGA:CA through OATP1B1 and their effects on glucose uptake and lactate and proton transport. Physiochemical results demonstrated that the compounds ranged from fully stable (pH 1.2 and 7.4) to completely unstable (pH 9.0). HLM incubation indicated no CYP-mediated hepatic metabolism. Treatment results showed that there was rapid intracellular uptake of CGA and CA in cancer cells and that CGA:CA lowered pH_i in both MDA-MB-231 and MCF-7 cells, while pH_i remained mostly unchanged in MCF-10A cells. Protein analysis revealed that CGA:CA treatment downregulated GLUT1 and MCT1 expression in cancer cells, suggesting impaired glycolytic activity and lactate shuttling. OATP1B1 expression was significantly suppressed in cancer cells, suggesting feedback inhibition of the solute carrier protein. Collectively, these findings indicate that CGA and CA exhibit favorable biochemical stability and disrupt intracellular pH regulation and metabolic transporter expression in breast cancer cells. Importantly, normal cells are not significantly affected. Thus, CGA:CA demonstrates therapeutic potential for breast cancer through pH_i and metabolic modulation. Full article

Background: Physical exercise is a vital component of cancer rehabilitation, with demonstrated improvements in cancer health-related outcomes, relapse rates, mortality and overall survival, yet global participation remains low. In Singapore, uptake of community cancer rehabilitation is limited despite high prevalence of treatment-related impairments. Methods: A cross-sectional study was conducted among adults (≥21 years) enrolled in the Singapore Cancer Society Rehabilitation Center between December 2021 and March 2023. Clinical data was collected from medical records and assessments included the Distress Thermometer (DT), Brief Illness Perception Questionnaire (Brief IPQ), and modified Bandura’s Exercise Self-Efficacy (ESE) scale. Adequate exercise was defined as ≥150 min/week of moderate intensity aerobic exercise and ≥2 days/week of resistance training. Regression analyses were performed to identify factors associated with meeting recommended moderate-intensity aerobic exercise levels. Results: The median age of our study cohort was 60.0 (IQR: 51.0–66.3) years. Of 132 analyzed participants, only 29.5% met recommended aerobic exercise levels and 9.1% met resistance training level recommendations. The median duration of moderate intensity aerobic exercise per week was 85.0 (30.0–180.0) minutes. The three most common cancer diagnoses amongst the participants were breast (53.8%), gastrointestinal (11.4%) and gynecological (7.6%) cancers. Clinically significant distress (DT ≥ 5) was present in 40.9% of patients. Multivariate regression analyses revealed that having lower ESE was the only significant psychosocial factor associated with a lower likelihood of meeting recommended aerobic exercise activity levels (OR = 0.856; 95% CI = 0.787–0.931; p < 0.001). Conclusions: Our study found a low prevalence of meeting recommended physical activity levels amongst our cohort of Asian cancer survivors enrolled in a community rehabilitation program in Singapore. Reinforcing the need for future longitudinal studies involving Asian cancer survivors to explore additional factors associated with physical inactivity and whether behavioral interventions targeting ESE will improve physical activity participation. Full article

Background/Objectives: Colonoscopy is the standard examination for many symptomatic patients referred for lower-gastrointestinal investigation, but it is resource-intensive and may be a burdensome experience. Colon capsule endoscopy (CCE) offers a minimally invasive, sedation-free first-line examination, although clinically important findings, inadequate cleansing or incomplete transit can generate downstream colonoscopy. DanCap aims to compare the costs and clinical consequences of a CCE-first pathway with routine conventional colonoscopy (CC). Methods: DanCap is a single-centre, cluster-allocated crossover trial at Odense University Hospital, Denmark. General practice clinics follow CCE or CC according to the parity of their pre-existing provider number, with pathways crossing after 200 consecutive CCE participants; no trial-generated random allocation sequence is used. Eight hundred symptomatic adults aged >18 years referred for expedited lower-gastrointestinal investigation are planned. CCE participants with suspected cancer, any polyp ≥6 mm, inadequate cleansing or an incomplete examination are referred for colonoscopy. The primary outcome is pathway cost. Registered secondary outcomes and prespecified process measures include polyp and colorectalcancer detection, examination quality, reinvestigation and patient-reported consequences. FIT and microbiome are registered secondary outcomes; participation in the substudy is optional and the analyses are exploratory within the CCEpathway. Primary analyses will follow the assigned pathway and account for GP clinic clustering, period and allocation sequence. Expected Results: The trial will quantify the resource use and clinical consequences of implementing CCE in routine symptomatic practice. Conclusions: DanCap is intended to inform decisions about CCE pathway implementation rather than to establish unbiased whole-cohort test sensitivity or specificity. Trial registration: ClinicalTrials.gov NCT06475560; first submitted 20 June 2024 and first posted 26 June 2024. Recruitment began on 27 November 2024; the registry listed the study as recruiting when last updated on 19 March 2026. Full article

Background: In recent decades, the incidence of colorectal cancer (CRC) has been rising worldwide. CRC ranks second in cancer-related mortality. The identification of reliable biomarkers for early diagnosis and prognosis prediction, along with a deeper understanding of the underlying molecular events, holds substantial promise for improving patient outcomes. The tissue inhibitor of the metalloproteinase 1 (TIMP1) gene is overexpressed in various gastrointestinal malignancies and contributes to tumor progression. However, its role in regulating the CRC tumor immune microenvironment (TIME) and its potential as a clinically actionable prognostic biomarker remain unclear. Methods: To probe how TIMP1 acts as a prognosis-related candidate biomarker in colorectal carcinoma, TCGA-derived datasets were adopted to conduct Kaplan–Meier survival assessment. We also investigated the connection between the expression abundance of TIMP1 and the infiltration of immune populations and intratumoral lymphocytes; furthermore, immune checkpoint-related genes were systematically assessed across multiple tumor types via the TISIDB and TIMER2.0 platforms, with particular emphasis on CRC. We adopted the ESTIMATE scoring system to figure out how TIMP1 gene expression correlates with the phenotypic properties of the colorectal-cancer TIME. We relied on the limma toolkit for the screening of differential transcripts from high-TIMP1 and low-TIMP1 cohorts. Enrichment assessments covering Gene Ontology terms and Kyoto Encyclopedia of Genes and Genomes entries were then carried out to predict the potential biological pathways associated with TIMP1. We constructed the protein–protein interaction map for TIMP1-interacting partners via the STRING repository. To further explore TIMP1-correlated genes, we performed Venn diagram intersection analysis combined with Spearman’s correlation test. Finally, quantitative reverse-transcription PCR was then implemented to detect TIMP1 messenger-RNA abundance inside the RKO colorectal carcinoma cell line as well as normal colonic epithelial CCD-18Co cells, which offered in vitro experimental verification for our bioinformatic outcomes. Results: According to outcome data, TIMP1 transcripts were markedly up-regulated in CRC specimens and cell lines relative to normal samples. Elevated TIMP1 expression served as a poor-prognosis indicator for overall survival (hazard ratio [HR] = 0.43, 95% confidence interval [CI] = 0.29–0.64, p < 0.001) and disease-specific survival (HR = 0.39, 95% CI = 0.22–0.68, p = 0.001) among colorectal-carcinoma patients. TIMP1-high and TIMP1-low groups exhibited notable differences in immune cell infiltration (CD8⁺ T, macrophage, mast, neutrophil, B, monocyte, dendritic, and CD4⁺ T cells). TIMP1 expression was also significantly correlated with tumor-infiltrating lymphocytes, key immune checkpoint genes (e.g., CD274 [PD-L1] and CTLA4), and immunomodulatory chemokines (e.g., CCL3 and CCL5). Twelve TIMP1-interacting DEGs were selected: COL5A1, FN1, PRG4, and a cluster of nine MMPs (MMP1/2/3/7/8/9/11/13/14), all of which showed significant positive correlations with TIMP1 (r = 0.31–0.63, all p < 0.001). Conclusions: TIMP1 expression correlates with features of the tumor immune microenvironment and extracellular matrix remodeling in CRC, suggesting that TIMP1 shows potential as a candidate biomarker. However, its potential as a therapeutic target warrants further experimental investigation. Full article

Gastrointestinal (GI) cancers represent a diverse group of malignancies that remain a major cause of cancer-related morbidity and mortality worldwide. Their management is increasingly complex, reflecting differences in tumor biology, anatomical location, stage, and molecular profile. In recent years, advances in molecular diagnostics, immunotherapy, and targeted treatment have moved clinical decision-making beyond a purely organ- and stage-based approach toward more individualized, biomarker-guided care. This narrative review summarizes established and emerging biological and targeted therapies used in esophageal, gastric and gastroesophageal junction, colorectal, pancreatic, hepatocellular, and biliary tract cancers. It focuses on immune checkpoint inhibitors targeting PD-1, PD-L1, and CTLA-4; HER2-directed monoclonal antibodies and antibody–drug conjugates; antiangiogenic and anti-EGFR therapies; and newer strategies involving CLDN18.2, FGFR2b, and tumor-agnostic alterations such as NTRK fusions. The review also considers the predictive biomarkers used to guide treatment selection and the growing integration of systemic therapy with surgery in neoadjuvant, perioperative, adjuvant, and conversion settings. However, clinical efficacy alone does not determine whether new treatments become part of routine practice. Regulatory approval, reimbursement, access to molecular testing, and the availability of specialized multidisciplinary care are equally important. The rapidly evolving treatment landscape for GI cancers therefore requires clinical decisions that account for tumor biology, anatomical resectability, molecular eligibility, expected benefit, treatment-related toxicity, and local access to therapy. Expanding access to comprehensive biomarker testing and effective molecularly guided treatments will be essential to translate progress in precision oncology into more personalized and equitable care for patients with GI cancers. Full article

Advances in healthcare have led to substantial growth in the global population of cancer survivors, yet young adult cancer survivors (YACS) aged 18–39 years remain an underrepresented population in cancer survivorship research. This review examines the health challenges and psychosocial factors that shape nutritional status in YACS, with the aim of informing age-appropriate nutritional care. YACS face significant financial and psychosocial demands unique to this life stage, balancing education, early careers and family responsibilities. They report greater unmet nutritional and health needs than other age cohorts, in addition to concerns regarding long-term side effects, financial strain, fear of cancer recurrence and increased responsibility for managing their own care. Cancer treatment can cause nutrition impact symptoms (e.g., nausea, taste changes, fatigue, dysphagia and gastrointestinal (GI) disturbances) that may persist for years, and YACS are vulnerable to late effects such as endocrine dysfunction, cardiometabolic disease, chronic pain and osteoporosis, all of which can be influenced by diet. Poor dietary patterns have been observed in YACS, indicating low intakes of fruits, vegetables, fibre and dairy, and higher consumption of saturated fat, sodium and processed foods. This narrative review evaluated fourteen nutritional intervention studies targeting YACS aged 18–39. Most existing interventions demonstrate minimal recruitment and retention rates, small sample sizes, and a reliance on self-report methods rather than objective nutritional measures. These limitations highlight the need for a deeper understanding of YACS’ specific nutritional needs and more effective strategies to engage this cohort. Future research should prioritise larger, more representative samples, incorporate objective nutritional and clinical measures, and explicitly address psychosocial and age-related barriers to healthy eating in young adulthood. Full article

Background: Greater familiarity between surgeons and anaesthetists is associated with better teamwork and patient safety. We investigated whether the volume of major UGI cancer resections the same surgeon and anaesthetist perform together influences post-operative outcomes at our tertiary specialist cancer centre. Methods: A 10-year retrospective cohort study was conducted via propensity score matching analysis (PSM). Consecutive adults undergoing elective surgical resection for primary UGI carcinoma and colorectal liver metastases from January 2014 to December 2024 were included. Cases were performed by surgeon–anaesthetist dyads composed of five Consultant Surgeons and 34 Consultant Anaesthetists. The primary Consultant Surgeon and Consultant Anaesthetist, the Charlson co-morbidity (CCM) score, and Clavien–Dindo (CD) complications were recorded. Cases were matched for age and CCM score. Results: A total of n = 792 cases were included, which became n = 546 after PSM. The median number of cases performed per dyad was 23, and the median CCM score = 5. Dyads were stratified into high (≥23 cases/dyad) and low volume (<23 cases/dyad), and cases into high (CCM ≥ 5) and low risk (CCM < 5). High-volume dyads had a lower incidence of post-operative complications when adjusted for PSM (35.2%) compared to low-volume dyads (42.8%) (p = 0.829). Probability of CD grade III complications among low-volume dyads increased with a rise in CCM (p = 0.224), but not in high-volume dyads. Conclusions: Greater familiarity between surgeons and anaesthetists may develop by operating on more cases together. This study has highlighted a trend suggesting that high-volume dyads could safely perform UGI resections on higher-risk cases without significant increases in post-operative complications. Dyad volume should be considered in operative scheduling for UGI cancer surgery. Full article

Background: Modern radiotherapy techniques have improved the safety of prostate cancer treatment, but acute gastrointestinal and genitourinary toxicity remains a common clinical problem. Early identification of patients who are likely to develop treatment-related adverse effects may help optimize supportive care. This study evaluated the relationships between clinical characteristics, radiotherapy planning parameters, systemic inflammatory biomarkers, and acute pelvic toxicity during hypofractionated radiotherapy. Materials and Methods: We retrospectively analyzed 142 patients with histologically confirmed prostate cancer who underwent external beam radiotherapy at the Provincial Integrated Hospital in Płock, Poland, between October 2024 and March 2026. Clinical, laboratory, and radiotherapy planning data were obtained from institutional electronic records. Laboratory analyses included CRP, creatinine, total leukocyte, neutrophil, lymphocyte, erythrocyte and platelet counts, hemoglobin, and hematocrit. Rectal and bladder toxicity was prospectively graded according to the RTOG/EORTC criteria at 2 and 4 weeks after treatment initiation. Correlation and multivariable regression analyses were performed to identify factors associated with acute treatment-related toxicity. Results: Radiotherapy significantly reduced leukocyte and lymphocyte counts (both p < 0.001) and produced smaller reductions in erythrocyte count, hemoglobin, hematocrit, and absolute neutrophil count (all p < 0.05). Patients who reported rectal or urinary symptoms early during treatment generally continued to experience toxicity later in the treatment course. Rectal and bladder toxicities were positively correlated, indicating that patients with toxicity affecting one pelvic organ frequently developed symptoms in the other. Routine inflammatory biomarkers and most dosimetric parameters showed limited independent value for predicting clinically relevant acute toxicity. Conclusions: Early gastrointestinal and genitourinary toxicity appears to be a useful clinical indicator of persistent treatment-related symptoms during hypofractionated prostate radiotherapy. Early rectal and bladder toxicity was associated with toxicity later during treatment, whereas routine laboratory markers and most dosimetric parameters showed limited independent associations with acute pelvic toxicity. Full article

Esophageal cancer is a common malignancy of the upper gastrointestinal tract that is associated with high incidence and mortality rates. Radiotherapy constitutes a cornerstone therapeutic modality for esophageal cancer. In radiotherapy, ionizing radiation is used to eliminate tumor cells through direct DNA damage and indirect reactive oxygen species (ROS)-mediated effects. However, clinical outcomes are frequently limited by interpatient heterogeneity and intrinsic tumor radioresistance. This review systematically describes the determinants of radiosensitivity in esophageal cancer within the established radiobiological framework of the “6Rs”: DNA damage repair (Repair), which is mediated by γ-H2AX phosphorylation, PARP family enzymes, and nonhomologous end joining (NHEJ) and homologous recombination (HR) pathways; cell cycle redistribution (Redistribution), which is regulated by G1/S and G2/M checkpoint dynamics; tumor repopulation (Repopulation), which is driven by cancer stem cell activity during fractionated treatment; reoxygenation (Reoxygenation), which is modulated through HIF-1α signaling and ROS homeostasis; intrinsic radiosensitivity (Radiosensitivity), which reflects interindividual and histopathological variability; and reactivation of antitumor immune responses (Reactivation), which enhances efficacy by remodeling the tumor immune microenvironment. Furthermore, regulated cell death mechanisms, including ferroptosis, autophagy, and apoptosis, significantly modulate radiotherapeutic responses. Elucidating these interconnected mechanisms provides a robust theoretical foundation for developing targeted interventions, identifying predictive biomarkers, and advancing precision radiotherapy strategies to optimize clinical outcomes for patients with esophageal cancer. Full article

Immune checkpoint inhibitors (ICIs) have significantly changed the treatment landscape of several gastrointestinal (GI) malignancies, particularly microsatellite instability-high (MSI-H) and mismatch repair-deficient (dMMR) tumors. However, most GI cancers remain resistant to current PD-1/PD-L1-based immunotherapy because of complex and highly suppressive tumor microenvironments characterized by immune stromal exclusion, myeloid-driven immune suppression, defective antigen presentation, and adaptive immune resistance mechanisms. Emerging evidence suggests that alternative inhibitory pathways, including lymphocyte activation gene-3 (LAG-3), T-cell immunoreceptor with immunoglobulin and ITIM domain (TIGIT), T-cell immunoglobulin and mucin-domain containing-3 (TIM-3), and V-domain Ig suppressor of T-cell activation (VISTA), contribute substantially to persistent T-cell dysfunction and resistance to checkpoint blockade. This review summarizes the immune landscape of GI malignancies and discusses the biological mechanisms underlying resistance to current ICIs. We highlight the evolving role of next-generation immune checkpoints, ongoing clinical development of novel inhibitors, and emerging combination strategies involving chemotherapy, anti-angiogenic therapy, radiation, bispecific antibodies, and tumor microenvironment modulation. In addition, we discuss current limitations in biomarker development and the growing role of circulating tumor DNA, spatial immune profiling, and multi-omics approaches in patient selection. Finally, we explore future directions aimed at improving precision immunotherapy and expanding durable responses across GI cancers. Full article

Circulating tumor DNA (ctDNA) is a promising non-invasive biomarker for detecting minimal residual disease (MRD) and predicting recurrence after curative treatment, yet evidence in esophageal squamous cell carcinoma (ESCC), esophageal adenocarcinoma (EAC), and gastric cancer has largely been examined within individual tumor types. In this scoping review, we mapped this evidence across all three malignancies and clarified key methodological and clinical considerations. Following the PRISMA-ScR guidelines, we searched PubMed, Scopus, and the Cochrane Library (3 April 2026) for studies linking ctDNA to disease-free, recurrence-free, or overall survival after curative-intent treatment. Twenty-seven studies (1746 patients; 10 ESCC, 4 EAC, 8 gastric, and 5 mixed) were included. Across every tumor type, postoperative ctDNA MRD was the most informative timepoint, with independent multivariable hazard ratios for disease-free, recurrence-free, or event-free survival of 2.8 to 21.8, whereas preoperative ctDNA was seldom prognostic. Serial monitoring further improved performance and flagged recurrence 78 to 278 days before imaging. Tumor-informed assays showed higher sensitivity than tumor-agnostic ones (80% vs. 35%), though direct comparisons were limited; correction for clonal hematopoiesis was essential for tumor-agnostic assays, and blood-based assays performed poorly in diffuse-type and peritoneal disease. Postoperative ctDNA MRD is a consistent, independent prognostic biomarker across upper gastrointestinal cancers that adds prognostic information beyond conventional staging and pathological response, supporting prospective interventional trials of ctDNA-guided management. Full article

Active molecules derived from traditional Chinese medicine (TCM) represent a valuable source of lead compounds for modern drug discovery, demonstrating considerable potential in the prevention and treatment of chronic diseases, cancer therapy, and immune modulation. Nevertheless, their clinical translation is often impeded by intrinsic limitations such as poor aqueous solubility, low bioavailability, inadequate chemical stability, and significant gastrointestinal irritation. Cyclodextrins (CDs) and their derivatives, characterized by a hydrophobic internal cavity, are capable of forming inclusion complexes with TCM and their derived active constituents, thereby improving their physicochemical and pharmacokinetic profiles. Over the past five years, substantial progress has been made in this domain. CD-based inclusion strategies have been shown to markedly enhance the solubility and dissolution rate of poorly water-soluble TCM compounds, including flavonoids, alkaloids, and terpenoids. Moreover, this approach contributes to improved drug stability, effective taste masking, and the achievement of modified release profiles, such as sustained or targeted delivery, ultimately leading to enhanced therapeutic efficacy and reduced adverse effects. The development of novel CD derivatives and smart delivery systems has further broadened their application potential. This review summarizes recent advances in the CD-based encapsulation of TCM active molecules, highlighting key formulation strategies that address persistent challenges in the modernization of TCM. It also provides a foundation for future research and development in natural product-based therapeutics. Full article

Background/Objectives: Treatment with adjuvant abemaciclib plus endocrine therapy (ET) has been demonstrated to improve outcomes in patients with HR+/HER2- early breast cancer at high risk of recurrence. However, real-world data regarding abemaciclib tolerability and treatment maintenance remain limited. In this study we describe the occurrence of adverse events (AEs) and their management through dose adjustments—dose reductions and temporary treatment interruptions—reflecting strategies that maximize compliance. Methods: We conducted a retrospective cohort study at the Oncology Department of ULS São João, in Porto, Portugal. Inclusion criteria comprised patients with HR+/HER2- early breast cancer treated, for at least three months, with abemaciclib, between January 2022 and January 2026. Clinical data were collected from electronic medical records and were analyzed using descriptive statistics, with continuous variables reported as median (IQR) and categorical variables as frequencies and percentages. Results: Fifty-six patients were included in this cohort. With the median follow-up of 14.5 months, 98.2% of patients experienced at least one AE. Most common AEs were gastrointestinal and hematological and occurred within the first 3 months of treatment. Dose reductions occurred in 46.4% of patients and temporary treatment interruptions in 25% of patients. However, no permanent treatment discontinuations were reported. Conclusions: In this cohort, AEs associated with adjuvant abemaciclib were common but generally low-grade and manageable. Early recognition of AEs allows appropriate dose adjustments and supportive measures, which support treatment maintenance. These findings highlight the importance of close clinical monitoring and patient counseling to enhance treatment compliance. Full article

Background/Objectives: Emergency presentation in colorectal cancer (CRC) is commonly regarded as a uniformly high-risk condition, although clinically distinct emergency phenotypes may carry substantially different prognostic implications. We aimed to characterise emergency diagnostic phenotypes, bowel obstruction, perforation, severe stenosis, and lower gastrointestinal bleeding, and to examine their associations with treatment allocation, perioperative mortality, and overall survival after adjustment for recorded covariates. Methods: We performed a retrospective, real-world cohort study of 1051 consecutive patients with CRC diagnosed between 2018 and 2021 at a tertiary referral centre. Patients were assigned to four mutually exclusive groups using a classification defined before outcome analysis (obstructive–perforative, stenotic, haemorrhagic, elective). Overall survival was assessed using Kaplan–Meier analysis and multivariable Cox regression, with formal testing of the proportional hazards assumption. Results: Emergency presentation occurred in 43.7% of patients but was markedly heterogeneous. Metastatic burden did not differ across phenotypes (p = 0.122). The obstructive–perforative phenotype showed the least favourable outcomes: perioperative mortality was 11.5%, and it remained associated with mortality after adjustment for recorded covariates (HR 1.58, 95% CI 1.25–2.00; p < 0.001), with hazard ratios ranging from 1.33 to 1.69 across six model specifications. Haemorrhagic presentation was rectal-predominant and behaved favourably, with survival comparable to elective diagnosis. Among 30-day survivors, the excess hazard was attenuated but persisted (HR 1.39, 95% CI 1.04–1.86). Conclusions: Emergency presentation in CRC is not a monolithic risk category. Recognising the presenting phenotype may offer a pragmatic framework for provisional risk characterisation at diagnosis, but external validation and adjustment for patient-level comorbidity and performance status are required before routine clinical use. Full article

Background: Machine learning (ML) is increasingly applied in medicine, underscoring the need for transparent and clinically relevant models. In gastrointestinal oncology, most ML studies rely on raw imaging data, which limits clinical adoption due to poor interpretability and the difficulty of collecting high-quality, large-scale video and image datasets in routine practice. Endoscopic ultrasound (EUS) plays a central role in the evaluation of pancreatic cancer; however, structured EUS features remain underused in predictive modeling. Objective: To assess the performance and interpretability of ML models for diagnosing pancreatic ductal adenocarcinoma (PDAC) using routinely collected EUS variables. Methods: We conducted a retrospective multicenter study using data from two Italian hospitals (n = 641) for model training and internal validation and from a third hospital (n = 120) for external validation, collected from 2015 to 2023. Decision trees, random forests, naïve Bayes and other classifiers were developed and evaluated. Model performance was assessed in terms of discriminative ability, calibration, and selective prediction. Results: All models demonstrated high discriminative performance (AUC ≥ 0.90). Decision trees provided the most favorable balance between interpretability and accuracy (balanced accuracy = 0.87; sensitivity = 0.89). Calibration and selective prediction analyses confirmed the robustness of the models. Conclusions: These findings demonstrate the feasibility of implementing interpretable yet high-performing ML models for PDAC diagnosis in real-life endoscopic settings. Full article