This study aimed to construct and validate a machine learning triage model based on combined routine complete blood count (CBC) and cell population data (CPD) parameters, for stratifying non-Hodgkin lymphoma (NHL) risk among adult hematology inpatients at first admission.

A total of 339 NHL inpatients and 417 hematology inpatients with other hematological disorders (negative controls, NC) were retrospectively enrolled, split into training cohort (n=510) and temporal internal validation cohort (n=246). All laboratory data were routine blood test results obtained at the time of first hospital admission. Univariate analysis was performed to screen differential CBC and CPD parameters, followed by LASSO regression feature selection conducted exclusively on the training cohort to avoid feature selection leakage. Finally, 8 indicators were retained for modeling: Age, HGB, MONO, NEUT%, BASO%, NE-SSC, LY-WX, and MO-WZ. Five machine learning algorithms (logistic regression [LR], naive Bayes [NB], support vector machine [SVM], neural network [NNET], k-nearest neighbor [KNN]) were established. Model performance was evaluated via ROC curves, calibration curves, and decision curve analysis (DCA). SHapley Additive exPlanations (SHAP) analysis was applied to interpret feature importance of the optimal SVM model.

Sixteen differential parameters between NHL and NC groups were identified, and 8 independent predictive variables were retained after LASSO regularization. The SVM model constructed with the 8 combined CBC and CPD indicators demonstrated the optimal predictive efficacy and robustness across both cohorts, with AUC=0.894 (training, 95%CI: 0.858-0.923) and AUC=0.803 (validation, 95%CI: 0.748-0.850). SHAP analysis verified that three core CPD parameters (MO-WZ, NE-SSC, LY-WX) all ranked among the top six most influential predictors, providing independent predictive contribution beyond conventional CBC indicators. DCA demonstrated theoretical net clinical benefit for risk threshold probabilities between 0.1-0.8.

The SVM triage model integrating routine CBC and CPD parameters from first-admission blood tests achieves reliable risk stratification for NHL among hematology inpatients. The combined application of conventional blood count indicators and cell population morphological parameters can provide effective reference for inpatient pre-biopsy triage. Further multi-center prospective external validation is required to improve model generalizability.

Total mesorectal excision is currently the standard surgical treatment for rectal cancer; however, the optimal extent of proximal lymph node dissection remains a subject of ongoing debate. The No.253 lymph node is the third station node along the inferior mesenteric artery system. Whether dissection of this region is necessary for all cases of rectal cancer remains inconclusive. Our previous retrospective study demonstrated that the incidence of No.253 lymph node metastasis is extremely low in rectal patients without clinical high-risk factors. Based on these findings, we have designed a randomised controlled trial to evaluate the oncologic safety of omitting No.253 lymph node dissection in this specific patient population.

This is a multicentre, randomised controlled clinical trial aimed at evaluating the safety of resection for rectal cancer without No.253 lymph node dissection. A total of 1384 patients with clinically confirmed low-risk rectal cancer will be enrolled. Participants will be randomly assigned to either the No. 253 lymph node dissection group or the non-dissection group. The primary endpoint is 3-year disease-free survival rate, with a pre-specified non-inferiority margin of 7%. Secondary endpoints include the total number of lymph nodes dissected, incidence of intraoperative and 30-day postoperative complications, intraoperative blood loss, operative time, postoperative hospital stay and 3-year overall survival (OS) rate.

This study protocol has been approved by the Ethics Committee of the National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College (reference no. 23/467-4210), and has been registered at ClinicalTrials.gov (NCT06146946). The findings of this trial will be disseminated through publication in peer-reviewed international medical journals. De-identified individual participant data will not be shared.

NCT06146946.

Immune checkpoint inhibitors (ICIs) are increasingly used before colorectal cancer surgery, particularly for mismatch repair-deficient or microsatellite instability-high tumours. Their surgery-specific safety, especially the risk of anastomotic leakage (AL), remains uncertain. We aim to synthesise AL and other postoperative outcomes after preoperative or perioperative ICI therapy.

We will search PubMed/MEDLINE, Embase, Web of Science, Scopus, CENTRAL, ClinicalTrials.gov, CNKI and the WHO International Clinical Trials Registry Platform from inception to July 2026 without language restrictions. We will include randomised and non-randomised comparative studies, cohort and case-control studies and single-arm studies reporting postoperative outcomes after preoperative or perioperative ICI therapy; case reports and case series with fewer than five surgically treated patients will be excluded. The primary outcome is AL among patients who underwent bowel resection with a constructed intestinal anastomosis. Two reviewers will independently screen records, extract data and assess risk of bias using design-appropriate validated tools. Comparative and non-comparative studies will be analysed separately: risk ratios from comparative studies will be pooled in random-effects models, and single-arm AL rates will be pooled as proportions using a binomial-logit generalised linear mixed model. Heterogeneity, prespecified subgroup analyses, sensitivity analyses and the certainty of the evidence (Grading of Recommendations Assessment, Development and Evaluation) will be reported.

Ethics approval is not required because only published aggregate data will be used. Results will be submitted to a peer-reviewed journal and presented at scientific meetings. Extraction materials, analytical code and the derived dataset will be shared in an open repository with the completed review.

CRD420261426134.

Chronic inflammation links cancer aetiologically to other diseases; therefore many cancer survivors live with long-term effects and comorbidities. While it becomes common place to outline long-term care in survivorship plans, liver health is often overlooked, with potential underdiagnosis of metabolic-dysfunction-associated steatotic liver disease (MASLD). Detecting early signs of MASLD is crucial, with new therapeutics able to prevent and even reverse liver damage. There is a need to understand its prevalence in cancer survivors, which is now possible with non-invasive tools such as vibration-controlled transient elastography (VCTE), which makes screening viable. Therefore, our study aims to address whether cancer survivors constitute a high-risk population for MASLD.

The EVALUATE Study is an observational, single-centre, cross-sectional study co-designed with patient and public involvement. It aims to assess the prevalence of fibrotic MASLD (fMASLD) and precursor conditions (non-fibrotic MASLD and steatotic liver disease) among 200 participants who were diagnosed with breast cancer or colorectal cancer 5-10 years ago, without recurrence. Participants' risk of fMASLD will be determined using VCTE and blood-based biomarkers. A subset with indeterminate risk will also undergo MRI of the liver. Data will be analysed using logistic regression (univariable and multivariable) to examine patient-related, tumour-related and treatment-related risk factors for fMASLD.

This study involves human participants and was approved by the Cantonal Ethics Committee of Bern (2025-00890). Participants gave informed consent to participate in the study before taking part. The results will be communicated through peer-reviewed publications and academic conferences.

NCT07280143/DRKS00038398.

Cancer is a leading cause of morbidity and mortality worldwide. Previous research has observed disparities in cancer diagnosis, care and outcomes among people who have been marginalised, including incarcerated populations. Evidence suggests an excess burden of cancer among people who experience incarceration. Currently, no reviews have synthesised quantitative evidence on all types of cancer prevalence, incidence, mortality and survival, stratifying by both cancer type and sex and gender, among people who experience incarceration, compared with those who do not, globally. The objective of this review will be to describe cancer prevalence, incidence, mortality and survival among people who experience incarceration and assess the epidemiological association between incarceration and these outcomes.

Peer-reviewed studies that will be included must report on individuals of at least one sex or gender who are currently or formerly incarcerated in correctional facilities under any legal status and report on cancer prevalence, incidence, mortality, stage at diagnosis and/or survival with or without a comparator group. Studies to be included must employ original quantitative or mixed-methods designs in which quantitative data are reported separately. Searches will be conducted in January 2027 in MEDLINE, EMBASE, Cochrane Library, CINAHL, Web of Science and Sociological Abstracts, and we will review reference lists. Two independent reviewers will conduct title, abstract and full-text screening and, for included studies, complete data extraction and assess methodological quality using the Newcastle-Ottawa Scale. We will calculate pooled prevalence rates, incidence rates and effect estimates using a random-effects model. Heterogeneity and publication bias will be assessed by I² and visual inspection of the funnel plot, respectively.

This review will use aggregate data from published studies; therefore, ethical approval is not required. Findings will be shared through peer-reviewed publications and conference presentations. The protocol has been registered with the International Prospective Register of Systematic Reviews (PROSPERO; CRD420261408922).

Early diagnosis of childhood cancer remains challenging because initial symptoms are common, nonspecific and have low predictive value in primary healthcare (PHC). We assessed whether socioeconomic differences shape pre-diagnostic PHC pathways before childhood cancer diagnosis.

We conducted a register-based matched cohort study using PHC data from three Swedish regions between 2001 and 2022, covering 54% of the general paediatric population in Sweden. Children aged 0-19 years with a first cancer diagnosis were matched 1:25 to cancer-free population references by age, sex and region. In-person PHC physician contacts in the year before the index date were analysed. Logistic and negative binomial regression models were used to assess socioeconomic differences in PHC utilisation and PHC contact frequency, respectively stratified by diagnostic group.

The study included 3711 childhood cancer patients and 93 181 references. Children with cancer had more PHC contacts than population references in the year before diagnosis (mean 2.28 vs 1.27, mean difference 1.01, 95% CI 0.94 to 1.08; p<0.001), although 22%-49% had no PHC contact 3 months before diagnosis depending on diagnostic group. Maternal educational gradients in PHC use were present in the general paediatric population but largely absent among children with cancer. An exception was central nervous system (CNS) tumours, for which children of mothers with low educational attainment had 28% more prediagnostic PHC contacts than those with highly educated mothers.

Socioeconomic differences in PHC utilisation by maternal educational level largely diminish as children approach a cancer diagnosis, except for CNS tumours, where increased contacts among children of mothers with low educational attainment may indicate a prolonged or delayed diagnosis.

Among the 18 million cancer survivors in the United States, many report significant unmet physical, psychosocial, and practical needs that existing healthcare systems struggle to address at scale. This conceptual perspective examines how artificial intelligence (AI)-powered platforms can serve as adjuncts to care by addressing unmet needs, supporting symptom management, and delivering psychoeducation and psychosocial support. Drawing on evidence from oncology AI and cancer survivorship, we consider how AI may improve the quality, personalization, and accessibility of care, distinguishing applications supported by prospective evidence from those that remain in early stages. Preliminary evidence suggests AI has the potential to deliver personalized psychoeducation, symptom monitoring, and psychosocial support as adjuncts to clinical care. To guide responsible development and implementation of these tools, we propose the PRECISION-AI Adjunctive Supportive Care Development Framework, a systematic model for developing, validating, and implementing AI in supportive oncology care. As a primary case example, we describe My Wellness Support, an agentic AI-enabled, risk-stratified supportive care platform designed to augment standard survivorship care, including its conceptual framework and development, ethical considerations, data protections, transparency, and bias mitigation. A proof-of-concept study in breast cancer patients is currently underway. Outcome data are not yet available, and no effectiveness or safety claims are made for the platform. AI in cancer survivorship has potential to expand access to survivorship support and improve health-related quality of life, but realizing this potential requires rigorous prospective validation, equitable implementation across diverse populations, and sustained collaboration among healthcare providers, developers, survivors, ethicists, and policymakers.

Designing phase III oncology trials to demonstrate overall survival (OS) superiority may be challenging due to sample size, follow-up time, resources, and costs as compared to trials evaluating intermediate endpoints, which may or may not be reliable surrogates of OS. In this study, we investigate at scale the relative gains in OS achieved in modern phase III oncology trials using OS primary endpoints in OPTIMA, a prospectively registered, pre-specified meta-analysis leveraging a large assembly of reconstructed individual patient-level data (PROSPERO: CRD420251162778). After screening 1184 trials and excluding 937 without an OS primary endpoint, 159 trials enrolling 104 740 patients met the inclusion criteria. In the two-stage Cox model, randomization to the experimental arm was associated with improvement in OS (random effects HR, 0.87; 95% CI, 0.85-0.89; P < .0001). In the one-stage Cox model stratified by trial identity, OS was also longer with the experimental arm (HR, 0.88; 95% CI, 0.85-0.90; P < .0001). Taken together, the data from OPTIMA should encourage renewed enthusiasm among trialists, funders, regulatory agencies, and policymakers for prioritizing phase III oncology trials evaluating OS primary endpoints.

Genetic and transcriptional alterations in cancer cells shape their interactions with immune and stromal compartments, thereby influencing tumor progression, immune evasion, and response to immune checkpoint inhibitors (ICIs). Yet, how these interactions are spatially organized within melanoma and how they relate to clinical outcome remains incompletely understood.

We used spatial multi-omic profiling on melanoma tissues from patients, compared those with loss-of-function mutations in the Neurofibromin 1 (NF1) tumor suppressor gene to those with intact NF1, and integrated these results with functional studies and subsequent gene expression changes observed in patient-derived melanoma models and an immunotherapy-resistant syngeneic mouse model.

Spatial multi-omic analysis of melanoma tissues from patients identified 12 meta-niches composed of distinct cell populations with unique molecular features. Although all tumors contained these meta-niches, those enriched in immunosuppressive cancer-associated fibroblasts (CAFs) and macrophages were significantly more common in tumors with NF1 loss. Conversely, niches enriched in cytotoxic CD8⁺ T cells were significantly reduced. In both human and mouse melanoma, NF1 loss was associated with impaired antigen-presentation programs and reduced CD8⁺ T cell infiltration. Mechanistically, EGFR signaling emerged as a prominent feature of NF1-deficient melanoma ecosystems and correlated with reduced expression of antigen-presentation genes and T cell exclusion. EGFR inhibition restored the expression of antigen-presentation programs and enhanced antitumor immune responses in a syngeneic Nf1 knockdown model resistant to ICIs.

These findings define spatially organized, functionally distinct meta-niches enriched in NF1-mutant melanoma that may contribute to their aggressive disease features and poor ICI response. Our study links an understudied melanoma driver to specific immune-evasion mechanisms and identifies EGFR inhibition as a candidate strategy to improve outcomes in patients with NF1-mutant melanoma resistant to immunotherapy.

Overall survival (OS) is considered the reference-standard endpoint for assessing treatment efficacy in oncology but requires prolonged follow-up evaluation. This study assessed progression-free survival (PFS) as a surrogate endpoint for OS in patients with localized pancreatic ductal adenocarcinoma (PDAC).

This retrospective international multicenter study included patients with a diagnosis of localized PDAC who were treated with at least one cycle of (m)FOLFIRINOX (2012-2022). Surrogacy between PFS and OS was assessed at the individual level using the Kendall rank correlation coefficient (Kendall's τ), with a τ of 0.8 considered the threshold for good surrogacy.

This study analyzed 1783 patients with a median age of 64.0 years, and 43.4% underwent surgical resection. The median OS was 22.2 months, and the median PFS was 12.8 months. Kendall's τ was 0.647 (95% confidence interval [CI], 0.626-0.668) overall and was higher for patients with lower baseline tumor burdens (primarily resectable [0.730], borderline resectable [0.640], locally advanced [0.601]), and more favorable pathologic responses (from 0.822 for ypT0 to 0.535 for ypT3-4, and from 0.755 for ypN0 to 0.541 for ypN2). The median post-progression survival was 6.7 months (interquartile range, 3.3-13.4 months).

Individual-level surrogacy between PFS and OS did not meet the prespecified threshold for good surrogacy due to considerable interpatient variability in post-progression survival, supporting the continued use of OS as the gold-standard efficacy endpoint in clinical trials. However, surrogacy was stronger among patients with lower pretreatment tumor burdens and more favorable pathologic responses, potentially facilitating perioperative treatment development through shorter follow-up periods in pancreatic cancer trials.