European Journal of Cancer

Distant melanoma metastasis at the time of diagnosis is uncommon, but has major implications for patient prognosis and treatment selection. However, few tools can reliably predict the risk of distant metastasis at initial presentation. Here, we developed and evaluated machine learning models to predict distant melanoma metastasis using routinely captured clinicopathologic and demographic variables across all histologic subtypes. Using the National Cancer Institute Surveillance, Epidemiology, and End Results (SEER) program from 2010-2022, we identified adults aged 20 to 90 years with melanoma as the first and only primary malignancy (n=51,285). Explainable Boosting Machine achieved a strong balance of discrimination and precision (AUROC = 0.947, AUPRC = 0.610, Precision = 0.793, Brier = 0.015). At 90% sensitivity, specificity was 0.843 with consistent performance across cross-validation folds. Clinicopathologic variables, including T stage, Breslow thickness, ulceration, and mitotic activity, contributed the largest share of predictive signal across descriptive, regression-based, and SHAP analyses, with smaller contributions from demographic factors. Decision curve analysis supported clinical utility, showing a net reduction of 88.3 per 100 patients and a standardized net benefit of 0.541. This model could be used to identify patients at sufficiently elevated risk to justify staging PET/CT despite otherwise localized clinical presentation. Cost-consequence analysis further showed that imaging true- and false-positive patients at 85% to 95% sensitivity threshold nearly doubled downstream imaging cost. We deployed the final model as an online calculator to support exploration of individualized risk estimates (https://melanoma-calculator.streamlit.app/).

Abstract: Background: More than half of metastatic melanomas arise from patients initially diagnosed with early-stage melanoma. Objective biomarkers are needed to better identify high-risk patients. Objective: To evaluate the prognostic value of multiple histopathological characteristics in predicting distant metastasis risk, in early-stage melanoma. Methods: Using data from discovery set (n=442) and a population-based validation cohort (n=306, sampled from 5,815 patients) of the Dutch Early-Stage Melanoma (D-ESMEL) study, we investigated 14 histopathological characteristics of melanoma and their tumor micro-environment (TME) in an unprecedented integration, by expert pathologist scoring and automated quantitative measurements derived from a validated automated segmentation. Results: Increased immune infiltrates (40% in cases vs. 50% in controls) were associated with lower risk of metastasis. Automated immune cell density was predictive in both the discovery set and the validation cohort, outperforming the manual pathological tumor infiltrating lymphocytes. The remaining histopathological features, including mitotic activity, did not retain independent value after controlling for current staging variables. Limitations: TME evaluation in standard Hematoxylin-Eosin slides. Conclusion: TME reaction is an important determinant of melanoma progression. The automated quantification of immune cell density appears to be a biomarker for distant metastasis risk. Further investigation into specific immune cell subtypes is required to facilitate clinical integration.

Background: Cases requiring 13 or more tissue sections in Mohs micrographic surgery (MMS) demand extended operative time, additional resources, and often specialised closure techniques. Pre-operative identification of such cases would improve surgical scheduling, resource allocation, and patient counselling. We aimed to develop and validate a machine learning prediction tool using pre-operative clinical features to identify cases likely to require13 sections. Objectives: To develop and validate machine learning models for predicting which Mohs procedures will require 13 sections, using pre-operative clinical features, and to identify key predictive factors. Methods: We analysed 408 consecutive Mohs procedures with 16 pre-operative clinical variables. Thirty machine learning algorithms were evaluated, including ensemble methods (Stacking, Voting), gradient boosting (XGBoost, LightGBM, CatBoost), neural networks (3-7 layers), support vector machines, and traditional classifiers. Model performance was assessed using 5-fold stratified cross-validation and independent test set evaluation. Feature importance was determined using SHAP (SHapley Additive exPlanations) analysis. Results: The stacking ensemble achieved the highest cross-validation AUC of 0.891 (95% CI: 0.849-0.934) and test AUC of 0.884. Tumour area (cm2), calculated using the ellipse formula to approximate clinical tumour morphology, emerged as the strongest predictor (SHAP importance: 0.141), followed by tumour size dimensions (0.086 and 0.068), aggressive histopathology (0.046), and recurrence status (0.035). Wide neural network architectures (5-layer) outperformed deeper configurations (7-layer). The model demonstrated 70.7% high-confidence predictions with uncertainty <15%. Conclusions: Machine learning models using pre-operative clinical features can accurately predict which Mohs procedures will require 13 or more sections. The stacking ensemble approach provides robust predictions suitable for clinical decision support. External validation in multi-centre cohorts with diverse patient populations and practice patterns is warranted to assess model generalisability.

Accurate risk stratification of pigmented skin lesions is critical for early melanoma detection and for reducing unnecessary excisions. Artificial intelligence (AI) is increasingly applied to dermoscopic image analysis, but its diagnostic performance relative to standard dermoscopy in real-world clinical settings remains uncertain. To address this gap, we conducted a systematic review and meta-analysis of prospective clinical studies directly comparing AI alone, dermoscopy, and AI-assisted clinicians for malignancy risk assessment of pigmented skin lesions. We systematically searched PubMed, Embase, Web of Science, and Cochrane Library from inception to January 2026. Ten studies with 17 diagnostic arms (10 dermoscopy arms, 6 AI-alone arms, and 1 AI-assisted clinician arm) were included. Pooled sensitivity and specificity were 0.773 (95% CI, 0.648-0.863) and 0.793 (95% CI, 0.673-0.877) for dermoscopy, and 0.757 (95% CI, 0.428-0.928) and 0.859 (95% CI, 0.619-0.958) for standalone AI. Summary ROC curves showed overlapping performance, indicating that autonomous AI is broadly comparable to dermoscopy but does not demonstrate a consistent advantage. Heterogeneity in AI performance was driven almost entirely by threshold effects rather than by differences in inherent model capacity. AI-assisted clinicians showed promising results (sensitivity 1.000, specificity 0.837) in a single study, but more evidence is needed. Our findings suggest that, at present, AI should be viewed as a complementary decision-support tool rather than a replacement for dermoscopic evaluation. The study provides valuable evidence for clinicians, guideline developers, and researchers working on AI integration into melanoma diagnostic pathways.

Background The clinical significance of clonal hematopoiesis of indeterminate potential (CHIP) in melanoma remains incompletely defined, particularly with respect to CHIP genotype, clone size, and somatic mutations (e.g BRAF mutations). We integrated human cohort data and a syngeneic melanoma mouse model to evaluate whether CHIP is associated with melanoma risk, tumor growth, and differential clinical outcomes. Methods We analyzed CHIP prevalence and survival in a large treatment-unselected melanoma cohort (n=2,480), evaluated tumor growth in a syngeneic BRAF-mutant (BRAFmut) melanoma murine model of TET2-CHIP and DNMT3A-CHIP, and assessed survival outcomes in an immune checkpoint inhibitor (ICI)-treated advanced melanoma cohort (n=361). Associations with progression-free survival (PFS) and overall survival (OS) were evaluated using Kaplan-Meier analyses and multivariable Cox proportional hazards models. Results CHIP was enriched among patients with treatment-unselected melanoma compared with age/sex-matched healthy controls, and larger CHIP clone size showed an age-adjusted association with inferior OS. In a syngeneic BRAFmut melanoma murine model, TET2-CHIP, but not DNMT3A-CHIP, was associated with significantly increased primary melanoma tumor growth. Among patients with ICI-treated advanced melanoma, CHIP was associated with worse OS compared with patients without CHIP. TET2-CHIP had the strongest adverse association with survival, whereas DNMT3A-CHIP was not significantly associated with PFS or OS. Conclusions CHIP is enriched in melanoma and exploratory analyses demonstrate genotype-specific differences in melanoma tumor growth and clinical outcomes. These findings support further investigation of genotype-specific CHIP profiling as a potential biomarker for melanoma risk stratification and immunotherapy outcomes.

Background: Immune checkpoint inhibitors have transformed cancer treatment, yet a large number of patients fail to respond. Identifying molecular characteristics that predict response before treatment initiation remains an unmet need. Towards that end, this study presents a large-scale integrative analysis of existing single-cell and bulk tissue datasets, aimed at identifying predictive features while providing insights into their cellular origin and potential function within the tumor microenvironment. Methods: A stepwise analysis was performed using single-cell RNA-sequencing data from 60 melanoma patients at baseline, separated into discovery (n=41) and validation (n=19) sets. An integrated bulk transcriptomics dataset (n=128) from melanoma patients and a bladder cancer dataset (n=298) were used for further validation. Results: Integrative analysis of melanoma single-cell datasets revealed that responders exhibit distinct molecular profiles across multiple cell types compared to non-responders. Notably, these included downregulation of the TNFR superfamily and other immunosuppressive genes (TNFRSF18, TNFRSF9, TNFRSF4, LGALS1, BATF, IL12RB2, LINGO1, DUSP4, SDC4, VCAM1) in T-cells. By investigating the findings from the immune cell populations in the bulk tumor context, 13 transcripts were found to be consistently associated with response across all cohorts. These were differentially expressed in T-cells (SELL, EPB41, CD96, UHFR2, LINGO1, LGALS1), B-cells (ALDH5A1), NK cells (PLEC, PDGFRB) and Monocytes (TLR10, ST6GAL1, IKZF1, MPRIP). A predictive model based on these features effectively discriminated responders from non-responders in melanoma (AUC=0.73). The model maintained significant predictive power in an independent bladder cancer dataset (IMvigor210; AUC=0.64). Of high clinical relevance, it demonstrated enhanced performance in identifying responders among patients with low tumor mutational burden (AUC=0.75). Conclusion: Our study reveals pre-treatment molecular features related to immune-cancer crosstalk that are associated with response to immunotherapy. A 13-gene model demonstrates potential added clinical value in stratifying responders, particularly in patients with low tumor mutational burden, meriting further validation.

Background: The All of Us Research Program represents a rich resource for cancer epidemiology research, with over 400,000 participants with whole genome sequences linked to electronic health records (EHR). Large cancer datasets often focus exclusively on cases without controls and neglect pre-diagnosis healthcare occurrences. Here, we perform a phenome-wide association study (PheWAS) of EHR data at least 1 year pre-diagnosis between cancer cases and matched controls, revealing co-occurring and mutually exclusive phenotypes. Methods: We identified 55,000+ cancer cases across 21 cancer types in All of Us version 8. To eliminate age-related confounding, we implemented a two-stage matching and censoring strategy: loose matching on demographics to establish index dates and cohort comparability, followed by right-censoring of EHR data (excluding 1 year pre-diagnosis/index), then 1:2 matching to address residual demographic imbalance. We tested associations between 23,193 cancer cases, 46,386 matched controls and approximately 1,600 clinical phenotypes using logistic regression adjusted for sex at birth, self-reported race, age at diagnosis/index date, and two censored EHR metrics: observation window and unique condition count, with Bonferroni correction for multiple testing. Results: Our analysis identified 232 significantly associated phenotypes, confirming established cancer risk factors including elevated prostate specific antigen (OR = 2.92, 95% CI: 2.65-3.23; p-value=1.8x10-101) and multinodular goiter (OR = 1.73, 95% CI: 1.56-1.91; p-value=6.7x10-27). Further investigation into the relationship between several phenotypes with seeming inverse effects is warranted. Conclusions: This PheWAS of EHR data at least 1 year pre-diagnosis leveraged the diversity of All of Us to examine how clinical phenotypes prior to cancer diagnosis vary across cancer types and racial groups. Our findings validate All of Us as a robust platform for cancer epidemiology research, confirming established risk factors at scale across diverse populations. This work provides methodological insights for EHR-based susceptibility analyses and demonstrates the value of agnostic phenome-wide approaches for generating hypotheses in precision medicine.

Melanoma remains one of the most treatment-refractory malignancies due to immune evasion, high mutational burden, and profound tumor heterogeneity. Although immune checkpoint inhibitors have transformed frontline management, a substantial proportion of patients develop resistance or experience relapse, underscoring the need for alternative and complementary immunotherapeutic strategies. Tumor-infiltrating lymphocyte (TIL) therapy and engineered viral vector-based immunotherapies represent mechanistically distinct yet clinically promising approaches for advanced melanoma. This systematic review and Bayesian meta-analysis evaluated the comparative efficacy of TIL therapy and engineered viral vector immunotherapies in advanced melanoma. A structured search of PubMed, Embase, Scopus, and Web of Science (2015-2025) identified 13 eligible studies, including four randomized controlled trials and nine prospective single-arm studies, reporting objective response rate (ORR), progression-free survival (PFS), overall survival (OS), and treatment-related adverse events. Eight studies met criteria for inclusion in the Bayesian quantitative synthesis of ORR outcomes. Risk of bias and certainty of evidence were assessed using Cochrane and GRADE frameworks. TIL therapy demonstrated substantial standalone efficacy, particularly in PD-1-refractory populations, with reported ORRs reaching 49%, median PFS of 7.2 months, and OS extending to 25.8 months. Viral vector-based therapies, including talimogene laherparepvec (T-VEC) and RP1, showed more modest monotherapy activity but demonstrated improved responses when combined with immune checkpoint inhibitors. Among the studies included in the Bayesian quantitative synthesis, the pooled ORR estimate was 37.8% (95% highest density interval [HDI]: 30.6%-45.3%). Sensitivity analysis excluding the small-sample Cui et al. (2022) study yielded a similar pooled estimate of 38.3% (95% HDI: 30.4%-46.2%). Exploratory meta-regression supported the overall robustness of the findings. Certainty of evidence for ORR was moderate, whereas survival and safety outcomes were downgraded due to heterogeneity, sparse reporting, and inconsistent endpoint definitions. Collectively, these findings support complementary rather than competing roles for TIL and engineered viral vector immunotherapies within evolving melanoma treatment paradigms. The results further highlight the potential importance of biomarker-guided sequencing strategies, including viral immune priming followed by adoptive cellular therapy, as a framework for optimizing personalized immunotherapy in both refractory and earlier-line melanoma settings.

Population-based mammography screening improves breast cancer outcomes, but its impact on real-world treatment pathways and quality indicators (QIs) remains incompletely described. We conducted a retrospective nationwide cohort study using linked data from the Czech National Cancer Registry and the National Registry of Reimbursed Health Services. Women aged [&ge;]18 years with a first breast cancer diagnosis between 2017 and 2024 were classified as screen-detected (SCR) or diagnostically-detected (DIG) according to the imaging modality preceding histological verification. Outcomes included stage distribution, untreated cases, first-line treatment, main treatment modality, time to treatment, multidisciplinary team discussion (MDT), centralization to Comprehensive Cancer Centres (COCs), and survival patterns. The verified cohort included 47,648 women: 26,817 SCR cases (56.3 %) and 20,831 DIG cases (43.7 %). In this nationwide analysis, SCR breast cancer was associated with earlier stage at diagnosis and better survival patterns, but also with longer time to treatment and longer time to MDT discussion than DIG-detected disease. Although treatment rates were high and centralization improved over time, substantial regional variation persisted in care pathways, MDT use, and access to COCs. These findings support continued strengthening of screening participation, monitoring of care intervals, and quality assurance of MDT reporting and regional oncology care delivery.

Purpose: There are no clinically validated biomarkers to assess recurrence risk and guide treatment de-escalation in Basal-like and HER2-enriched breast cancer. Taxane-based chemotherapy remains a cornerstone of treatment despite significant toxicity. We evaluated the prognostic and predictive utility of the MHCII Immune Activation Score (IA Score) in these subtypes. Experimental Design: We retrospectively analyzed Basal-like and HER2-enriched breast cancers from the NCIC CTG MA.21 trial, which randomized patients with node-positive or high-risk node-negative disease to adjuvant chemotherapy with or without taxanes. MA.21 predated immune checkpoint inhibitors and routine HER2-targeted therapy. Subtype was previously assigned by PAM50. The 36-gene MHCII-IA assay used RNA from formalin-fixed, paraffin-embedded tissue. Multivariable Cox and Kaplan-Meier analyses evaluated associations between IA Score, clinicopathologic variables, tumor-infiltrating lymphocytes (TILs), relapse-free survival (RFS), and taxane benefit. Results: Among Basal-like (N=317) and HER2-enriched (N=155) tumors, higher IA Score was associated with improved RFS independent of lymph node status and provided stronger prognostic discrimination than TILs. Node-negative patients with high IA Score had excellent outcomes (8-year RFS >90%) versus those with low IA Score (8-year RFS <76%). In node-positive disease, high IA Score increased 8-year RFS by >10% relative to low IA Score. IA Score stratified taxane benefit: node-positive IA-low patients benefited, whereas IA-high tumors had favorable outcomes regardless of regimen. Conclusions: MHCII Immune Activation Score is a prognostic and predictive biomarker in Basal-like and HER2-enriched breast cancer. High IA Score identified patients with excellent outcomes before pembrolizumab, trastuzumab, and taxane-based treatment escalation, providing a rationale for prospective risk-adapted de-escalation strategies.

Sentinel lymph node biopsy (SLNB) is increasingly omitted in early-stage breast cancer, often prompting whole-breast irradiation (WBI). We evaluated partial-breast irradiation (PBI) without axillary surgery among 78 clinically node-negative patients (median age 75) treated from 2014 to 2022. After 53-month median follow-up, no ipsilateral, regional, or distant recurrences occurred. These results demonstrate excellent outcomes and suggest PBI is a feasible, safe alternative to WBI when SLNB is omitted.

Importance: Enfortumab vedotin (EV) is an antibody-drug conjugate approved for the treatment of locally advanced or metastatic urothelial cancer (la/mUC). Cutaneous adverse events (cAEs) are common during EV therapy, with prior studies suggesting an association between EV-related cAEs and improved survival; however, there is insufficient data to delineate the survival benefit of EV-induced cAEs from those associated with concurrent immune checkpoint inhibitors (ICIs). Objective: This study aims to evaluate the association of EV-induced cAEs and survival, and to characterize the timing and morphology of EV-induced cAEs. Design: We conducted a multi-institutional retrospective study of patients with la/mUC treated with EV between 2020 and 2025. Setting: Multicenter academic referral center. Participants: A total of 449 EV-treated patients were included. Patient characteristics were extracted manually, and likelihood scoring was used to attribute cAEs to either EV or other etiologies. Exposure: EV treatment. Main Outcomes and Measures: We estimated progression-free (PFS) and overall (OS) survival using Kaplan-Meier method. Multivariable time-varying and landmark Cox regression models were used to evaluate associations between EV-induced cAE and survival. Sensitivity analyses were performed at landmarks from 15 to 105 days. Results: Of 449 patients, 206 (45.9%) developed a cAE; 39 (18.9%) were high-grade and 127 (61.7%) were attributed to EV. The most common cAEs were pruritus (41.3%), unspecified and desquamating dermatitis (37.3%), and morbilliform dermatitis (27.7%). Across all treatment groups, survival was longer in patients with EV-induced cAEs. Developing an EV-induced cAE was protective across all examined landmark times, with hazard ratio (HR) 0.60 (95% CI: 0.43-0.82, p<0.001) for PFS and HR 0.46 (95% CI: 0.31-0.67, p<0.001) for OS at primary landmark time of 30 days. Early-onset EV-induced cAEs were protective at all landmark times and high-grade EV-induced cAEs were not associated with worse survival. Conclusions and Relevance: EV-induced cAEs were independently associated with improved PFS and OS in patients with la/mUC, even after accounting for immortal time bias and ICI exposure. Distinguishing EV-induced cAEs from other etiologies in timeline and morphology may help guide oncology and dermatology management.

Background. The European Society for Medical Oncology (ESMO) Scale for Clinical Actionability of molecular Targets (ESCAT) ranks genomic alterations by the evidence supporting the predictive value of the molecular target for response to targeted therapies. No openly available, systematically curated set of standard care biomarkers mapped to the ESCAT framework exists to support clinical decision-making or harmonize biomarker interpretation. Methods. We mapped all OncoKBTM Level 1 biomarkers to ESCAT tiers using evidence cited by OncoKBTM, excluding abstract-only data. Eight board-certified oncologists and hematologists independently assigned ESCAT tiers, with discrepancies resolved through structured consensus meetings. Recurring evidence scenarios that did not correspond to any existing ESCAT tier informed a set of a priori defined modifications, which were subsequently applied to biomarkers that could not be classified using native ESCAT criteria. Results. Of 188 OncoKBTM Level 1 biomarkers, 16 were excluded due to abstract-only evidence. Using native ESCAT criteria, 51% of the remaining biomarkers were classified as Tier 1, 3% Tier 2, 18% Tier 3, 6% Tier X and 22% could not be assigned to any tier. Applying the modified ESCAT criteria resolved all previously unclassifiable biomarkers and increased Tier 1 assignments to 73%. Inter-rater reliability (Krippendorffs alpha) was moderate (0.586) and 62% of classifications required consensus discussions. Comparison with ESCAT tiers reported in ESMO Clinical Practice Guidelines showed improved concordance when using the modified criteria. Conclusions. The native ESCAT criteria are highly stringent, resulting in many FDA-recognized, clinically validated biomarkers that are currently assigned level 1 by OncoKBTM not mapping to any existing tier. Our predefined modifications improved alignment with OncoKBTM Level 1 designations and with published ESMO clinical practice guidelines. The mapped set of standard care biomarkers are provided on the OncoKBTM website, offering a practical resource that harmonizes ESCAT tiers of evidence with a widely adopted levels of evidence schema.

Early detection of suspicious moles remains the most effective means of reducing mortality from skin cancer, yet systematic screening is constrained by the time and expertise required for manual mole assessment. This paper presents an end-to-end computational pipeline that utilizes wide-angle skin photographs (including consumer-grade smartphone images) and produces quantitative ABCD (Asymmetry, Border irregularity, Color variegation, Diameter) feature scores for every detected mole. The pipeline operates in four stages: mole detection via adaptive thresholding and blob analysis, super-resolution enhancement using EDSR, false-positive filtering using a brightness-based statistical criterion, and lesion segmentation using the Boundary Attention Mapper (BAM). BAM generates high-resolution segmentation masks by fusing early-layer activations with GradCAM heatmaps from a trained EfficientNet-B7 classifier, achieving 90.45% accuracy on the ISIC2017 dataset, outperforming both conventional GradCAM (87.78%) and dedicated segmentation architectures, including DeepLabv3 and SAM v2 by more than 5 percentage points in Dice score. The EfficientNet-B7 backbone achieves a micro-average AUC of 0.97 across eight lesion classes, with a melanoma AUC of 0.99. Color quantification uses K-means clustering with a threshold calibrated on the PH2 dataset (MSE = 1.425). Applied to 87 wide-angle images, the mole detection module achieved an F1 score of 86%. The system outputs a structured CSV of per-lesion ABCD scores suitable for clinical triage and longitudinal tracking. A clinical validation study with dermatologists and surgeons is underway to assess concordance between automated and expert assessments.

Background The NEO-EXCEL trial hypothesised that aromatase inhibitor (AI)-activity as neoadjuvant endocrine therapy for early-stage breast cancer in postmenopausal women may be enhanced in combination with cyclooxygenase-2 (COX-2) inhibition. Methods NEO-EXCEL was a phase III, placebo-controlled, randomised trial in postmenopausal women with oestrogen receptor (ER)-positive resectable breast cancer with tumours [&ge;]2cm. Women were randomised (1:1:1:1): exemestane (25mg od) plus celecoxib (400mg bid), exemestane (25mg od) plus placebo (bid), letrozole (2.5mg od) plus celecoxib (400mg bid), or letrozole (2.5mg od) plus placebo (bid). Primary endpoint was clinical response (complete/partial) measured by callipers at 16 weeks; a standard assessment method at the time of trial inception. Sixteen-week ultrasound-determined response was the main secondary outcome to verify the calliper-based primary. Analysis was intention-to-treat. Results Due to slow accrual the trial design was redesigned from a definitive 2x2, 1000 patient trial to one randomising 269 patients between 20-Nov-2007 and 29-Apr-2014; 34.9% were human epithelial growth factor receptor 2-positive. AI+celecoxib produced a significantly greater objective clinical response than AI+placebo (72.9% vs 55.6%, P=0.003), which remained after adjustment for AI type and stratification factors (odds ratio = 2.3; 95% CI 1.3-3.8, P=0.003). Ultrasound-determined response was however not significantly enhanced (48.7% [AI+celecoxib] vs 41.2% [AI+placebo], P=0.34). Progression free survival and overall survival remained similar (median follow-up = 5.1 years [range 0.1-7.1]). Conclusions NEO-EXCEL is the first completed, phase III double-blind, placebo-controlled trial testing the addition of celecoxib to AI as neoadjuvant endocrine therapy in early breast cancer. Clinical response showed significant improvement but there was no significant ultrasound-determined response improvement nor any surgical or long-term outcome evidence of AI+COX-2 inhibition improving treatment outcomes for ER+ early resectable postmenopausal breast cancers. Use of short-term celecoxib at 400mg bd for 16 weeks was safe with no excess cardiotoxicity observed.

Peripheral whole-blood gene expression profiling offers a minimally invasive route to lung cancer detection, but high-dimensional transcriptomic data are prone to optimistic bias when preprocessing and model selection are not properly separated from performance evaluation. We applied L1-penalised (LASSO) logistic regression to 303 peripheral whole-blood microarray profiles (123 lung cancer cases and 180 healthy controls; Gene Expression Omnibus accession GSE252168; Illumina HumanHT-12 v4) within a leakage-free nested cross-validation framework (5 outer and 3 inner folds), in which all data-dependent steps (imputation, univariate feature screening by ANOVA F-test with k = 500, and standardisation) were confined strictly to training partitions. Statistical significance was assessed by permutation testing (B = 100), and feature selection stability was quantified across outer folds. LASSO was compared with ridge logistic regression, linear support vector machines, and random forest under the same framework. The LASSO model identified a sparse 29-probe signature with a pooled out-of-fold area under the ROC curve (AUC) of 0.990 (nested estimate 0.989 +/- 0.015), accuracy 97.4%, sensitivity 94.3%, and specificity 99.4% at a 0.50 threshold; permutation testing confirmed significance (p = 0.0099). Six probes, including CDC42, U2AF1, and RPS15A, were selected in all five outer folds, forming a stable core, and all classifiers exceeded AUC 0.987, indicating a strong, algorithm-independent signal. A leakage-free nested cross-validation framework enables unbiased performance estimation and reproducible feature selection in blood-based lung cancer classification. The 29-probe panel is an internally validated candidate requiring prospective, multicentre external validation before clinical use.

Objective Accurate prognostication of recurrence risk in HR+/HER2- early breast cancer is central for therapeutic decision-making, including identifying patients who may safely avoid adjuvant systemic therapy. However, the performance of existing prognostic tools remains insufficient for effective clinical stratification, motivating the development of artificial intelligence (AI)-based methods to improve risk stratification. Methods Ataraxis Breast CTX (ATX) is a multi-modal AI test that integrates H&E-stained whole-slide images with clinicopathologic features to predict risk of recurrence for individual patients. This study aims to validate ATX in an external dataset enriched for clinically low-risk patients from Dordrecht, the Netherlands. ATX scores were generated for 892 women diagnosed with early HR+/HER2- breast cancer. Of the 892 patients, 299 did not receive adjuvant systemic therapy. The discriminative performance of ATX was assessed using C-index and its stratification ability was evaluated by log-rank tests comparing Kaplan-Meier survival curves across risk groups. Results ATX achieved a C-index of 0.71 and a 5-year time-dependent AUC of 0.71, demonstrating strong discrimination in predicting recurrence-free survival (RFS). Among 299 patients who received no adjuvant therapy, ATX achieved a C-index and time-dependent AUC of 0.78 and 0.81 respectively, suggesting ATX retains prognostic information in the absence of systemic therapy. ATX scores were used to stratify patients into risk groups using a pre-specified threshold, where 656 (74%) were classified as ATX low-risk and 236 (26%) were classified as high-risk. Notably, untreated and treated ATX low-risk patients had comparable 5-year RFS (untreated: 5-year RFS = 96%, 95% CI = 92-97%; treated: 5-year RFS = 96%, 95% CI = 93-97%) with near identical 10-year RFS (86%, 95% CI = 83-92% for both), suggesting ATX low-risk status may identify a subgroup with favorable prognosis independent of treatment exposure. Conclusion ATX provides robust prognostic stratification in an external cohort of clinically low-risk HR+/HER2- early breast cancer and identifies a subgroup of patients who did not receive systemic therapy with favorable observed outcomes. These results support prospective validation of ATX as a decision-support tool for adjuvant therapy de-escalation in HR+/HER2- early breast cancer.

PurposeTo develop and validate a multimodal recurrence-risk model integrating histology, genomic testing, and clinical variables. MethodsWe developed AI-Path, a whole-slide image biomarker for recurrence prediction trained in CALGB 9344, and validated it in three independent cohorts: TAILORx, a multi-site Chicago cohort, and the MDX-BRCA cohort. We then integrated AI-Path with Oncotype DX Recurrence Score (RS), tumor size, and nodal status into a Cox model, PathClinRS, fit using 60% of cases from TAILORx, with the remaining 40% held out for validation. The primary end point was distant recurrence-free interval. Performance was assessed using Harrells concordance index (C-index) and Kaplan-Meier analyses. ResultsA total of 12,418 patients were included. In TAILORx, AI-Path outperformed RS for distant recurrence (C-index, 0.682 vs 0.647; P = .038), driven by superior prediction of late recurrence (0.656 vs 0.567; P < .001). In node-negative disease, PathClinRS outperformed RSClin in the TAILORx fitting (0.72 vs 0.70; P = .016) and validation sets (0.74 vs 0.70; P = .004). In node-positive disease, PathClinRS outperformed RSClinN+ in Chicago (0.94 vs 0.74; P < .001) and MDX-BRCA (0.71 vs 0.66; P = .004) cohorts. Compared with NATALEE eligibility, PathClinRS identified nearly twice as many high-risk node-negative patients while maintaining a comparable 10-year distant recurrence risk (16.7% vs 16.6% per NATALEE eligibility in TAILORx fitting; 21.0% vs 19.4% in TAILORx validation). PathClinRS identified 68% of intermediate risk premenopausal patients as low-risk with no evidence of chemotherapy benefit, compared to only 36% identified as low risk by standard clinicopathologic criteria. ConclusionDigital histopathology provides prognostic information complementary to genomic assays and has the potential to personalize therapy beyond existing clinicogenomic tools.

Purpose: Chemotherapy-induced peripheral neuropathy (CIPN) affects many people receiving taxane treatment for breast cancer. Symptom trajectories vary, with some recovering, and others experiencing persistent, or delayed worsening (coasting) symptoms. The prevalence and predictors of these trajectories remain unclear. This study identified the prevalence and biopsychosocial predictors of CIPN persistence, improvement, and coasting within three months post-treatment. Methods: This secondary analysis included participants treated with taxanes for stage I-III breast cancer who completed the Functional Assessment of Cancer Therapy/Gynecologic Oncology Group-Neurotoxicity-4 (FACT/GOG-NTX-4) at baseline, post-chemotherapy, and three months later. A minimally important difference (MID) from baseline on the FACT/GOG-NTX-4 defined persistence, improvement, coasting, and no MID-CIPN (below the MID threshold at each assessment) trajectories. Baseline assessments included self-reported pain/well-being, sensory, balance, and lower limb physical functioning measures, and sociodemographic and treatment data were collected. Results: Among 102 participants (51.57{+/-}11.24 years), persistence occurred in 34.3%, improvement in 25.5%, coasting in 6.9%, and no MID-CIPN in 33.3%. Compared to no MID-CIPN, older age (OR=1.120; 95%CI: 1.026-1.222), higher expected pain (OR=1.630; 95%CI: 1.082-2.456), and cold hyperalgesia at the foot (OR=1.130; 95%CI: 1.018-1.254) predicted persistence. Lower fatigue predicted improvement (OR=0.904; 95%CI: 0.845-0.968). No predictors were identified for coasting. Conclusion: CIPN trajectories are heterogeneous. Age and pre-treatment pain expectations, cold hyperalgesia, and fatigue differentiate patients with persistent CIPN and those likely to improve from those with no CIPN. Implications for Cancer Survivors: Early identification of individuals at risk for persistent neurotoxicity may support risk stratification and guide targeted supportive care strategies.

Our study evaluated whether a deep learning auto segmentation model combined with machine learning triage can streamline radiotherapy clinical trial quality assurance (QA). We analyzed 107 stereotactic ablative radiotherapy (SABR) cases from a multi-institutional phase II clinical trial of neurovascular sparing prostate SABR, focusing on physician contours of the internal pudendal artery (IPA) as a novel organ-at-risk with substantial interobserver variability. Contours were scored by the trial principal investigator as Per-Protocol or Minor Deviation/Unacceptable. We applied a deep learning model for IPA auto-segmentation. Agreement between human and AI contours was then quantified using 14 overlap, distance, and surface metrics, and a supervised classifier was trained on these metrics to flag clinical trial protocol deviations. While AI segmentation achieved only modest geometric accuracy with mean Dice similarity coefficient of 0.446 and 95th percentile Hausdorff distance of 14.23, when incorporating all 14 metrics, a machine learning classifier yielded AUROC of 0.836, flagging all Minor Deviation/Unacceptable cases with 100% sensitivity on the 27 case hold-out set with 6 false positives and no false negatives. AI segmentation combined with metrics-based machine learning can triage protocol deviations within a multi-institution radiotherapy clinical trial, supporting prospective evaluation of AI-assisted trial QA.