Clinical and Translational Radiation Oncology

BACKGROUND AND PURPOSE: ROAM/EORTC-1308 is an international (Europe and Australasia) phase III randomised controlled trial (RCT) comparing radiotherapy (60Gy/30#) to observation after surgery for atypical meningioma. The treatment plans of all the patients randomised to radiotherapy were reviewed and approved prior to treatment delivery, either by the UK RTTQA group (European sites) or the Australian TROG (Australia and New Zealand). Pretrial credentialling included outlining and planning a benchmark case (DR). MATERIALS AND METHODS: Variations from the guidelines were noted in quality assurance (QA) reports both for DR and for the on-trial Individual Case Reviews (ICRs) and classed as major or minor. After recruitment had finished, a random 10% of the ICRs were independently reviewed for audit purposes. We used the QA reports to analyse the variations in the DR and in the ICRs RESULTS: 56 sites (20 UK, 25 EORTC, 11 TROG) undertook the DR of which 13 (23%) had major variations. During the trial 64 patients at 28 sites received radiotherapy, the median being 2 patients per site. Overall, 25 (39%) patient cases needed resubmitting: 22 (36%) sets of outlines (2 cases twice) and 12 (14%) treatment plans (1 plan twice). CONCLUSIONS: For complex radiotherapy of rare tumours, a DR is insufficient and prospective ICRs of all patients is required. The consistent high-quality radiotherapy in ROAM/EORTC1308 ensures the primary outcome (progression free survival) will be a robust assessment and any difference between treatment arms cannot be attributed to variation in radiotherapy treatment.

Background: Radiotherapy is a cornerstone of treatment for pediatric central nervous system (CNS) tumors, but dose selection remains largely uniform despite substantial interpatient variability in tumor radiosensitivity. This limitation is particularly consequential in children, in whom radiation-associated toxicity has lifelong impact. The genomic-adjusted radiation dose (GARD), which integrates tumor genomics with delivered radiation dose, quantifies the biological effect of radiotherapy and has been validated across multiple adult malignancies. Its relevance in pediatric CNS tumors remains unknown. Methods: We performed a retrospective cohort study using gene expression and clinical data from 246 pediatric patients with high-grade glioma, medulloblastoma, or ependymoma from the Childrens Brain Tumor Network. GARD was calculated using a sequencing-adapted radiosensitivity index integrated with radiation dose via the linear-quadratic model. Associations between GARD, physical radiation dose, and clinical outcomes (event-free survival and overall survival) were evaluated using Cox proportional hazards models stratified by tumor type and anatomic location. Patients who did not receive radiotherapy were analyzed as a negative control cohort (sham-GARD). Results: Among patients receiving radiotherapy, physical radiation dose was relatively uniform, yet GARD demonstrated substantial interpatient variability in predicted biological effect. Higher GARD was significantly associated with improved event-free survival (hazard ratio [HR] 0.90, 95% CI 0.83-0.97; p=0.004) and overall survival (HR 0.90, 0.83-0.99; p=0.018). By contrast, physical radiation dose was not associated with either endpoint. In patients who did not receive radiotherapy, sham-GARD was not associated with outcomes, supporting its role as a treatment-specific predictor rather than a general prognostic biomarker. Conclusions: In pediatric CNS tumors, the biological effect of radiotherapy as quantified by GARD is associated with clinical outcomes, whereas physical dose alone is not. These findings challenge the current paradigm of uniform radiotherapy dosing and support a genomically informed approach to dose individualization. Prospective evaluation of GARD-guided radiotherapy is warranted to optimize tumor control while minimizing long-term toxicity in children.

Background. Treatment of locally advanced rectal cancer (LARC) increasingly varies in radiotherapy use, sequence, and intensity. Pretreatment biomarkers for the mismatch-repair-proficient majority remain limited: the biopsy-adapted Immunoscore predicts neoadjuvant response and recurrence risk, but no available biomarker estimates intrinsic tumor radiosensitivity or guides radiotherapy dose, use, or sequence. Genomic Adjusted Radiation Dose (GARD) combines the biopsy-derived Radiosensitivity Index (RSI) with the prescribed dose-fractionation schedule through the linear-quadratic model to estimate tumor-specific modeled radiation effect. We sought to evaluate whether pretreatment GARD is prognostic for outcomes in radiotherapy-treated LARC. Patients and methods. We performed a retrospective pooled analysis of 497 patients with LARC drawn from four prospective clinical trial and institutional cohorts; 335 patients (67%) were prospectively enrolled in clinical trials. The cohort spanned induction chemotherapy followed by chemoradiotherapy (CRT) (n=142), CRT followed by consolidation chemotherapy (n=120), and CRT without sequential chemotherapy (n=235). Pretreatment gene expression was measured by microarray or RNA sequencing and harmonized across platforms before GARD calculation. The primary endpoint was disease-free survival (DFS). GARD was evaluated continuously using cohort-stratified Cox regression and dichotomized at the outcome-blind pooled-cohort median of 19.3. Multivariable models adjusted for age, sex, and clinical stage. Results. Median follow-up was 5.3 years. Among 456 patients evaluable for DFS, 99 experienced an event. Higher GARD was associated with longer DFS as a continuous variable (hazard ratio [HR] per 1-unit increase, 0.92; 95% CI, 0.86-0.99; p=0.027) and at the median threshold (GARD >19.3 versus <19.3: HR, 0.62; 95% CI, 0.41-0.92; p=0.021). Five-year DFS was 81% versus 73%, and 10-year DFS was 80% versus 66%, respectively. GARD remained independently associated with DFS after adjustment for age, sex, and clinical stage (HR, 0.92; p=0.023). Overall survival (OS) was directionally consistent but not statistically significant (HR per 1-unit increase, 0.94; p=0.18). Among 445 patients with evaluable Neoadjuvant Rectal (NAR) scores, higher-GARD patients had lower median NAR scores (8.4 versus 15.0; p=0.004), were more frequently classified as low risk (34% versus 21%), and were less frequently classified as high risk (23% versus 31%). Among 460 patients evaluable for pathologic complete response (pCR), the pCR rate was numerically higher with higher GARD (22% versus 15%; odds ratio per 1-unit increase, 1.07; p=0.09). Conclusions. Pretreatment GARD, a biology-based model of tumor-specific radiation effect, stratified DFS independently of clinical stage and was associated with NAR-defined pathologic response across contemporary treatment sequences. These findings provide multicohort evidence of prognostic validity but do not establish prediction of radiotherapy benefit. Prospective GARD-stratified trials should test whether incorporating tumor radiosensitivity into decisions about radiotherapy use, dose, and sequence improves tumor control and organ preservation while reducing treatment-related morbidity.

Background: Brachytherapy delivers localized radiation from sources placed within or adjacent to tissue at risk. In brain tumor surgery, intracavitary implantation can begin at resection and concentrate dose along the cavity wall, where many recurrences arise. Current preoperative GammaTile tools estimate tile requirements but not patient-specific placement, resulting dose, or adjacent anatomy. We developed an integrated software framework for patient-specific virtual GammaTile planning. Methods: The framework incorporates AI-assisted tumor and cavity segmentation with user approval, converts accepted masks into patient-derived surfaces, supports virtual tile placement, calculates lifetime Cs-131 dose using TG-43, maps dose distributions to HCP-MMP cortical parcels and normative HCP-1065 white-matter bundles, and supports AI-assisted synthesis of the resulting complex anatomical and connectomic output for clinician review. We retrospectively applied it to three patients with glioblastoma. Tile requirements were compared with the GammaTile Cavity Surface Area Calculator. For Patient 1, calculated 60- and 80-Gy volumes were compared with the clinical plan. Results: Preoperative layouts required 6.5, 7, and 4 tile equivalents for Patients 1-3; calculator estimates were higher by 1.5, 4, and 1 tiles. Postoperative differences narrowed to 1, 1, and 0 tiles. The plans generated patient-specific lifetime dose distributions at 60, 80, and [&ge;]120 Gy. In Patient 1, calculated and clinical 60- and 80-Gy volumes were similar despite incomplete spatial overlap. Atlas mapping revealed distinct cortical and white-matter exposure patterns across cases beyond lobar description. Conclusions: Patient-specific virtual GammaTile planning can connect anticipated tile placement with calculated dose and adjacent neuroanatomy. These cases demonstrate feasibility and establish a framework for prospective validation and future comparison of candidate implant arrangements.

Purpose: To quantify the dosimetric effect of residual intrafraction motion during gated MR-guided prostate stereotactic body radiotherapy with simultaneous integrated boost to the dominant intraprostatic lesion, and to derive and evaluate target-specific, population-based margins designed to mitigate motion-induced coverage loss. Materials and Methods: Thirty patients were treated on a 1.5-T MR-Linac with 36.25, 40, and 45 Gy prescribed to the planning target volume, clinical target volume (CTV), and gross tumor volume (GTV), in 5 fractions. Across 150 fractions, segment-level dose matrices synchronized with time-resolved target displacements were combined to reconstruct per-fraction and cumulative dose. Directional prescription-isodose deviations were used to derive internal target volume margins, which were evaluated through offline replanning, gating-envelope analysis, and dose reconstruction. Results: Median 95th-percentile motion ranges were 0.5, 1.4, and 2.9 mm in the left-right, anterior-posterior, and superior-inferior directions; these exceeded 3 mm in 0.7%, 13.2%, and 47.7% of fractions, respectively. The median treatment time was 13.2 min (IQR, 12.0-15.1 min) with a median gating duty cycle of 93.2% (80.6%-97.8%). Per-fraction median GTV deviations were -1.2% (-2.4% to -0.1%) for D95 and -5.6% (-15.3% to -0.2%) for V45Gy; corresponding CTV deviations were -0.5% (-1.1% to -0.1%) and -2.1% (-4.1% to -0.4%). Cohort-derived margins of 2/2/1/2/1/2 mm for the GTV and 1/1/1/2/1/2 mm for the CTV in the left/right/anterior/posterior/superior/inferior directions achieved sufficient coverage in approximately 94% of fractions in each direction. Gated delivery using the proposed margins improved coverage but may reduce duty cycle by an average of 6.9 +/- 4.7 %. Conclusion: Residual intrafraction motion produced directionally asymmetric coverage loss, with the largest effect on the boosted GTV. Motion-inclusive dose reconstruction enabled derivation of practical asymmetric margins that improved robustness while identifying the tradeoff between target coverage and delivery efficiency.

Background: Prostate-specific membrane antigen (PSMA) positron emission tomography (PET) is increasingly used to characterize newly diagnosed prostate cancer on the assumption that uptake highlights aggressive disease. However, maximum standardized uptake value (SUVmax) varies widely across localized tumors. Recent genomic assessments have suggested SUVmax reflects tumor biology even when accounting for other tumor indices. Here, we examined highly annotated pathologic, imaging and clinical correlates of intraprostatic SUVmax at the level of individual tumor foci in a prospective multimodal cohort. Methods: Exploratory lesion-level analysis of the radical prostatectomy arm of a single-center prospective diagnostic-accuracy trial (NCT04461509). Patients underwent simultaneous 18F-DCFPyL PSMA PET, multiparametric magnetic resonance imaging (MRI) and high-resolution MRI before prostatectomy, with central blinded pathology and imaging review. Grade Group (GG) [&ge;]2 foci measuring [&ge;]10 mm were eligible. The primary outcome was focus-level SUVmax. Correlates spanning histology, tumor microenvironment, MRI and clinical indices were examined by Spearman correlation and rank-based group comparison, followed by separate linear regressions of log(1+SUVmax) each adjusted for GG and tumor size. A digital-pathology sensitivity analysis quantified immune, stromal and tumor pixel-area fractions on hematoxylin and eosin (H&E) sections from a subset of foci using a trained QuPath classifier. Results: Twenty-seven patients contributed 34 tumor foci, of which 21 (61.8%) were GG2. Median SUVmax was 5.7 (interquartile range [IQR] 3.9 - 12.1; range 0.0 - 57.9). SUVmax rose across the cribriform/intraductal carcinoma (Crib/IDC) gradient (median 3.9, 4.8 and 10.2 for absent, focal and extensive; p = 0.010), correlated with % Gleason pattern 4 ({rho} = +0.47, p = 0.005) and correlated inversely with stromal content ({rho} = - 0.40, p = 0.020). Grade Group (p = 0.115), Prostate Imaging Reporting and Data System (PI-RADS) 4 versus 5 (p = 0.705), standard apparent diffusion coefficient (ADC) ({rho} = - 0.27, p = 0.130), baseline prostate-specific antigen (PSA) ({rho} = +0.04, p = 0.816) and PSA density ({rho} = - 0.02, p = 0.918) were not associated with SUVmax. In separate models of log(1+SUVmax) adjusted for Grade Group and tumor size, % Gleason pattern 4 remained associated with uptake (standardized {beta} = +0.76, 95% CI +0.11 to +1.41), Crib/IDC ({beta} = +0.36, - 0.03 to +0.74), and stromal content ({beta} = - 0.30, - 0.62 to +0.02). On 18 slides re-quantified by pixel classifier, no pixel-based compartment fraction correlated with SUVmax (immune {rho} = - 0.06, stroma {rho} = +0.02, tumor {rho} = - 0.04; all p > 0.5). Conclusions: In this prospective lesion-level cohort, intraprostatic PSMA SUVmax was associated with increased Gleason pattern 4 and cribriform/intraductal burden but not with other common risk indices. SUVmax therefore appears to carry information additive to conventional risk factors. Further investigations examining the tumor biology that reflects SUVmax are warranted.

Introduction Trials investigating hypofractionated radiotherapy in carcinoma cervix have demonstrated an increased risk of acute gastrointestinal toxicity. Patients with node negative disease have a very low risk of pelvic nodal relapse, a dose painting approach integrating differential doses to the primary and elective nodal clinical target volumes may allow safe hypofractionated radiotherapy by reducing the acute gastro-intestinal toxicity. Methods and Analysis Patients with Human Papillomavirus (HPV) associated squamous cell carcinoma who have no pelvic nodal involvement as documented on PET CT and MRI will be eligible. Patients will be randomly allocated to receive standard whole pelvic radiotherapy (CRT) covering the primary clinical target volume (CTVp) and the nodal clinical target volume (CTVn) to a dose of 45 Gy (25 fractions, 1.8 Gy per fraction) with concurrent weekly cisplatin at a dose of 40 mg/m2 weekly for five cycles. In the experimental arm (HDRT), a dose of 42.5 Gy (2.125 Gy per fraction) and 36 Gy (1.8 Gy per fraction) will be delivered in 20 fractions to the CTVp and CTVn respectively. Concurrent cisplatin will be delivered at a dose of 50 mg/m2 weekly for four cycles. The primary objective will be to compare the proportion of patients with grade 2 or higher acute gastrointestinal toxicity (nausea, vomiting and diarrhea). Secondary outcomes will include disease free survival, quality of life and late toxicity. The primary statistical analysis will be reported on the intention-to-treat population. The primary outcome will be analysed using a covariate adjusted log binomial model to compare the relative risk of grade 2 or higher acute gastrointestinal toxicity. We assume that HDRT will reduce the toxicity rate from 35% to 21% (14% absolute risk reduction). Using an asymmetric two-sided group sequential design with three planned interim analyses, the target sample size is 274 patients for 80% power with a 5% (1-sided) type I error. Ethics and dissemination The trial will be initiated at the participating centers after institutional review board approval at the respective centers. All patients will provide written informed consent. The results will be presented in peer reviewed literature as well as academic conferences. Trial Registration The trial has been prospectively registered in the clinical trial registry of India.

Background: Biochemical recurrence (BCR) occurs in up to 40% of men following radical prostatectomy (RP). Current risk models rely primarily on clinicopathologic variables and may not fully capture the biological heterogeneity associated with recurrence. The Decipher Genomic Classifier (DGC), prostate-specific membrane antigen positron emission tomography (PSMA-PET), and multiparametric magnetic resonance imaging (mpMRI) provide complementary prognostic information that may improve prediction. Objective: To develop and evaluate machine learning (ML) models integrating DGC, PSMA-PET, and mpMRI for preoperative prediction of BCR following RP. Methods: This retrospective study included patients with available preoperative DGC, PSMA-PET, mpMRI, and clinicopathologic data. Logistic regression (LR), random forest (RF), and XGBoost models were developed using single- and multimodality feature combinations. Early- and intermediate-fusion strategies were evaluated. Performance was assessed using an area under the receiver operating characteristic curve (AUC) and accuracy. Clinical utility was evaluated using decision curve analysis. Results: XGBoost consistently outperformed LR and RF. DGC achieved the highest single-modality performance (AUC 0.94, accuracy 86.7%). Among multimodal models, DGC combined with PSMA-PET using intermediate fusion achieved the best overall performance (AUC 0.93, accuracy 87.0%). Addition of mpMRI reduced performance (AUC 0.85, accuracy 83.0%). Decision curve analysis demonstrated positive net benefit across clinically relevant thresholds. Conclusion: XGBoost-based multimodal fusion improved preoperative BCR prediction following RP. DGC was the strongest individual predictor, while integration with PSMA-PET provided the best overall performance, supporting the potential of radiogenomic ML models for personalized risk stratification.

Abstract Purpose: To compare scalar circulating blood dose summaries, assess whether associations with grade 3 or higher lymphopenia and overall survival persisted after planning target volume (PTV) adjustment, and distinguish scalar from exploratory dynamic blood dose analyses. Methods and Materials: The assembled dataset included 133 patients from 2 retrospective thoracic radiation cohorts; 121 entered the dosimetric main analysis and 116 passed a post hoc dosimetric gate. Six scalar exposures were compared: an ICE3 (Immune Circulation radiation Exposure Estimator Engine) 10 compartment mean-dose metric, corrected effective dose to immune cells (EDIC), body remainder dose, a hematological dose (HEDOS) derived organ-mean approximation, mean lung dose, and mean heart dose. Logistic base models included baseline absolute lymphocyte count, concurrent chemotherapy, and cohort; PTV was then added. Overall survival used cohort-stratified Cox models on an endpoint-specific common set. Benjamini Hochberg correction was applied within prespecified 6exposure families and, separately, across 3 selected post hoc bootstrap contrasts. Results: The lymphopenia analysis included 94 patients and 69 events. After PTV adjustment, the HEDOS derived approximation remained nominally associated (odds ratio, 2.54; 95% confidence interval, 1.18-5.45; P=.017; q=.102), but no exposure survived false-discovery-rate control. The survival set contained 92 patients and 42 deaths; no PTV adjusted scalar exposure was associated with survival (all q>=.531). In post hoc analyses, the standardized association of mean heart dose with overall survival was more positive than that of ICE3 (difference in log hazard ratios, 0.44; 95% CI, 0.13 0.94; multiplicity-adjusted q=.024). The corresponding contrast with the HEDOS derived approximation did not meet the adjusted significance threshold (q=.053). These coefficient contrasts do not establish superior predictive performance or causality. Exported ICE3 kinetic summaries had no false-discovery-rate-significant residual associations. A 13 case HEDOS bDVH audit showed little change under one continuous versus 10 second gap perturbation. Conclusions: PTV adjustment attenuated scalar blood-dose associations with severe lymphopenia, and no scalar exposure retained a multiplicity-robust survival association. The selected mean heart dose coefficient contrast is hypothesis generating and does not establish superior prediction. The primary cohort comparison evaluated a HEDOS derived organ-mean approximation rather than the full dynamic HEDOS framework; therefore, these findings should not be interpreted as evidence against the potential value of particle level blood dose distributions or time-dependent blood-flow modeling.

Background: MRI is widely used in radiotherapy planning due to its high soft-tissue contrast, but geometric distortions can compromise target localization accuracy. Objective: This study aimed to develop an accessible method for assessing geometric distortion in MRI using two phantoms - a commercial anthropomorphic phantom and a custom-made phantom fabricated from ABS plastic. Approach: CT imaging was used as the reference standard. Distortion was assessed through linear measurements of periodic structures in a DICOM viewer, followed by statistical analysis. Significance: The study evaluates the clinical impact of distortion on radiotherapy planning and proposes a cost-effective solution for routine quality assurance in resource-limited settings.

Background Optic pathway hypothalamic glioma (OPHG) can cause irreversible visual impairment. Radiotherapy and radiosurgery are used for progressive, recurrent, or treatment-refractory disease, but their effects on visual outcomes remain incompletely defined. Objective To synthesize visual outcomes following radiotherapy or radiosurgery for OPHG, with visual preservation as the primary outcome. Methods This prospectively registered systematic review and meta-analysis (PROSPERO CRD420261440136) followed the PRISMA 2020 statement. PubMed, Scopus, Web of Science, Embase, the Cochrane Library, Google Scholar, and ClinicalTrials.gov were searched from inception to 1 June 2026. Eligible studies included patients with OPHG or related optic pathway or hypothalamic gliomas treated with radiotherapy or radiosurgery and reporting extractable visual outcomes. Visual preservation was defined as stable or improved vision. Random-effects meta-analyses of logit-transformed proportions were performed. Results Forty-nine studies were included in the systematic review, of which 19 contributed 494 visual outcome observations to the meta-analysis. The pooled visual preservation rate was 75.6% (95% CI 65.1% to 83.7%; I2 = 71.6%). Pooled rates of visual improvement, stability, and worsening were 24.7% (95% CI 19.5% to 30.8%), 46.7% (95% CI 37.0% to 56.6%), and 20.6% (95% CI 12.7% to 31.6%), respectively. Preservation was higher in pure radiotherapy or radiosurgery cohorts than in mixed-treatment cohorts (85% vs 66%; subgroup p = 0.0104), although differences in cohort composition and treatment attribution may have influenced this finding. Conclusion Radiotherapy and radiosurgery were associated with preservation of vision in approximately three-quarters of evaluable outcomes, predominantly through visual stability rather than recovery. Standardized visual outcome definitions are needed to improve future comparisons and treatment evaluation.

Purpose/Objective There is paucity of data reporting outcomes of breast cancers with initial internal mammary nodal involvement and no visceral metastases, treated with curative hypofractionated radiotherapy . We report the outcomes from a tertiary centre alongside spatial patterns of recurrences in the above group Material/Methods For this retrospective cross-sectional study, consecutive patients contoured as per the ESTRO 2013 guidelines, treated between 2016-2022 were eligible if their diagnostic imaging demonstrated involvement of the internal mammary nodes. Radiotherapy (40 Gy/15#/3 weeks) was delivered to the residual breast / thoracic wall, SCF region corresponding to the ESTRO lymph node level 4 and internal mammary chain nodes. Residual IMN/ level 4 nodes received a boost of 10Gy/5#. Spatial mapping of sites of recurrence at the local site and three nodal sites (axilla, SCF and IMN) was performed using deformable image registration. Sites of recurrence at the local site and three nodal levels were contoured separately. Volumetric intersection of the recurrent gross tumour volume (GTV_recurrence) with treated clinical target volume (CTV) was calculated. Actuarial overall (OS), disease free survival (DFS) & cumulative incidence of local (LR), regional (RR) and loco-regional recurrence(LRR) were calculated using Kaplan Meier method. Univariate comparison of outcomes with or without residual disease was performed using the log rank test. Results The median age of the 61 eligible women was 49 years. 77% received neoadjuvant chemotherapy and the rest adjuvant chemotherapy. 82% patients had a mastectomy. Axillary lymph node dissection was done in 96.7%. Boosts to residual IMN and SCF nodes were delivered to 21(34.4%) and 2 (3.3%) respectively. Median follow up was 3.6 years. Out of the 61 patients, 42 patients were disease free with an estimated 3 year disease free survival of 75% (95% CI 64, 88%). Spatial mapping of locoregional recurrence was possible in all but 1 patient with local (only) recurrence who was lost to follow-up after mammogram only. Among the patients with loco regional recurrence 1 had recurrence in local site + SCF +axilla, 3 had recurrence in the SCF+axilla, 2 in the SCF+IMN and 1 in the axilla+SCF+IMN. Only one patient had isolated axillary recurrence or isolated SCF recurrence. There were no IMN only recurrences. Among the 8 patients with nodal recurrence, a total of 27 individual GTV_recurrence were identified in the axilla(n=11), SCF(n=11) and IMN (n=5). IMN recurrences showed complete or partial overlap with CTV. SCF recurrences were a mix with predominantly in-field recurrences while axillary recurrences occurred outside the treated volume.Four (6.6%) patients had Grade 2 lymphoedema as documented late side effect. Conclusion Aggressive treatment of IMN disease with adjuvant radiation is effective with good locoregional control. Systemic recurrences are common and may benefit from intensification strategies.

Background. Accurate delineation of the gross tumour volume (GTV) - primary tumour (GTVp) and nodal disease (GTVn) - on FDG-PET/CT is a critical step of head and neck radiotherapy planning. Comparisons between lightweight custom networks and the auto-configured nnU-Net v2 are usually reported as end-to-end pipelines, conflating the contribution of the network with that of the inference-time post-processing applied on top of it. We separated the two. Methods. MiniUNet3D (custom 3D U-Net, 18.3 M parameters) and nnU-Net v2 (3d_fullres, 88.2 M parameters) were trained on the same 578 FDG-PET/CT cases (85/15 author-defined split of the HECKTOR 2025 Task 1 set, 8 centres) and evaluated on the same internal cohort. Three arms were compared pairwise: MiniUNet3D raw output at a fixed 0.5 threshold, MiniUNet3D with a locked adaptive post-processing pipeline, and nnU-Net v2. Comparisons used paired Wilcoxon tests with bootstrap confidence intervals, Bonferroni and Benjamini-Hochberg correction, and Cohen's d; catastrophic failure (Dice < 0.01) was compared with an exact McNemar test. Cases with an empty reference for a given target were excluded from that target's analysis (n = 98 GTVp, n = 93 GTVn). Results. With post-processing matched off, nnU-Net v2 was superior: median GTVp Dice 0.799 versus 0.592 (mean difference -0.244, 95 % CI -0.300 to -0.191; d = -0.88) and GTVn 0.774 versus 0.598 (d = -0.82). Post-processing raised MiniUNet3D to 0.800 (GTVp) and 0.738 (GTVn), recovering 79 % of that difference. Post-processed, MiniUNet3D matched nnU-Net v2 on GTVp Dice (p = 0.113) but remained inferior on nodal disease after Bonferroni correction (Dice p = 0.041; surface Dice p = 0.049). Catastrophic GTVp failures were 25/98 raw, 8/98 post-processed and 1/98 for nnU-Net v2 (McNemar p = 0.016). Inference took 34 s versus 78 s per case on the same GPU. Conclusions. Post-processing recovered most, but not all, of the difference between the two models, and it did not confer robustness: an eight-fold higher rate of empty contours on small primaries persisted, which is the more consequential difference for planning safety. Pipeline comparisons reported without a post-processing ablation risk attributing to a network what post-processing supplied.

Head and neck radiation oncology, with its complex locality and epidemiology, requires various imaging techniques to achieve sufficient visualization. Through collaboration with medical physicists, radiation therapists, radiation oncologists, and neuroradiologists, our group clinically implemented a standard set of magnetic resonance imaging (MRI) sequences on an MR-simulation scanner for three anatomical regions: skull base, oral cavity, and larynx/pharynx. A total of 41 patients with head and neck cancer and 3 volunteers were scanned on either a 1.5T or 3T MR-simulation scanner. Sequences described include Dixon-based approaches, SPAIR-based approaches, short T2-based approaches, and quantitative imaging methods. We describe and outline the steps to optimize these notable MRI sequences to aid in successful implementation in a radiation oncology clinic.

Purpose: Clinical assessment of vertebral lesion quality (osteolytic, osteoblastic, mixed) remains subjective, with limited interobserver reliability. This study evaluated a novel application of 3D convolutional neural networks (3D-CNNs) for classifying lesion quality from CT volumes in metastatic cancer patients. Materials and Methods: This retrospective study used CT data from 151 cancer patients planned for radiotherapy for metastatic spine disease (September 2020-July 2024). Leveraging vertebra-level expert annotations, we introduced an unconventional U-Net-based strategy converting coarse voxel-wise predictions into vertebra-level lesion classifications. The final dataset comprised 2,125 vertebrae across four classes (no lesion, osteolytic, osteoblastic, mixed), split into a 3-fold cross-validation set and an independent holdout test set. Model performance was benchmarked against a DenseNet121 baseline and a musculoskeletal radiologist, with Cohen's kappa assessing inter-rater agreement. Results: The 3D model achieved an ensemble accuracy of 84.7%, outperforming DenseNet121 (72.1%), with substantial gains in F1 score, precision, and balanced accuracy. It showed high concordance with the radiologist (Cohen's kappa = 0.76) and comparable sensitivity and specificity across all lesion subtypes. We found both models and the radiologist to struggle with osteolytic lesions, reflecting the difficulty of distinguishing this class from age-related changes in vertebral bone density and architecture caused by benign bone lesions, age-related systemic skeletal disorders and cancer treatments. Conclusions: 3D-CNNs trained with vertebra-level labels can accurately and reliably classify vertebral metastatic lesion quality from CT scans, offering a scalable path toward automated characterization of metastatic spine disease to support clinical decision-making and large-scale radiomics research.

Purpose: Digital-twin frameworks can potentially offer a pathway toward individualized treatment but remain largely untested in radiation oncology. We present a causal digital-twin model for head-and-neck (H&N) cancer that estimates patient-specific survival benefit from chemoradiotherapy (CRT) versus radiotherapy (RT) alone, stratified by HPV status. Methods: Using the public RADCURE cohort (n = 3,346), we developed a causal digital-twin framework in which patients were divided into HPV-positive and HPV-negative groups and assigned to training (RT start year [&le;] 2007) and test sets (RT start year > 2007). Bayesian-network structure learning identified each cohort's Markov Blanket around the 2-year survival endpoint (SVy2), yielding the minimal causally relevant feature set. A causal-forest model (double machine learning, DML framework) estimated the individualized treatment effect {tau}=P(SVy_2|CRT,X)-P(SVy_2|RT,X), with {tau} denoting the model-estimated effect. Overlap restrictions were applied to mitigate extrapolation. Predictive discrimination (AUC, C-index), average treatment effects (ATE), individualized effects (ITE), and decision-curve analyses were computed. Results: For HPV-negative disease, internal evaluation demonstrated strong discriminative performance (test AUC = 0.72; C-index = 0.81, p < 0.01). The estimated test-set ATE favored CRT with a modest but positive improvement in SVy2 (mean {tau} = 0.13). Patterns in {tau} were clinically coherent, including increased benefit in node-positive and good-performance patients and reduced benefit in high-burden or poor-performance presentations. Decision-curve analysis showed that a {tau} -guided selective-CRT strategy provided higher net benefit than uniform treatment policies. In HPV-positive disease, treatment-effect heterogeneity was attenuated, consistent with its more uniformly favorable prognosis. Conclusion: This study demonstrates a proof-of-concept causal digital-twin framework for interpretable, patient-specific survival simulation in H&N cancer. The model captures clinically plausible treatment-effect heterogeneity and establishes a foundation for multi-cohort validation and future integration into individualized treatment decision support.

Introduction: Resection margins in oral squamous cell carcinoma (OSCC) are frequently inadequate. This study investigated the accuracy and diagnostic performance of ex-vivo 7T MRI against histology. Methods: Fifty OSCC specimens underwent ex-vivo 7T MRI. Tumours were segmented on corresponding histology and twice on MRI: blinded to histology (radiologist segmentation; clinical practice) and using histology (histology-MR segmentation; MR-based ground truth). Resection margins were measured in five regions and compared with histology. Results: Anterior and posterior margins were generally smaller on MRI than on histology (maximum mean difference -2.5 mm), whereas cranial, caudal and deep margins were generally larger on MRI (mean differences <1 mm). For close margins (<5 mm), sensitivity was 95% for histology-MR segmentations and 91% for radiologist segmentations, with 75% specificity for both. Conclusion: Ex-vivo 7T MRI accurately measures resection margins and detects close margins with high sensitivity. A future clinical trial will determine whether intraoperative implementation reduces inadequate margins and adjuvant treatment.

Objectives: To estimate the absolute benefit of adding and prolonging androgen deprivation therapy (ADT) to radiation therapy (RT) for men with localized prostate cancer by integrating cancer-specific risk, other-cause mortality (OCM) risk, and relative treatment efficacy. Subjects and Methods: Individualized risks were estimated by integrating cause-specific hazard estimates from a validated staging system (STAR-CAP) with those from a validated non-cancer mortality model (OCCAM). ADT treatment efficacy was then incorporated using published hazard ratio estimates from the MARCAP meta-analysis. Model utility was illustrated using the Prostate, Lung, Colon, and Ovarian (PLCO) cancer screening trial cohort (n=5468 prostate cancer patients with complete covariates). The primary outcome was the estimated 10-year absolute risk reduction (ARR) in distant metastasis (DM) from adding short-term ADT (STADT) to RT or extending STADT to long term ADT (LTADT). Results: Within PLCO, model estimated risk of DM at 10 years ranged from <1% to 56% under guideline concordant care and OCM risk ranged from 3% to 80% independent of treatment. For patients with NCCN unfavorable intermediate-risk (n=1814), adding STADT to RT reduced estimated 10 year DM risk by a median of 4%, but the ARR estimates for individuals ranged from <1% to almost 15%. Similar variation was seen within NCCN high-risk patients (n=1289) with a median ARR of 9% [Range: 0.4%-17.2%] when prolonging ADT treatment. Moreover, 25% of NCCN unfavorable intermediate-risk patients have less than a 2.4% ARR while 25% of NCCN high-risk patients experience less than a 5% ARR. Conclusion: An integrated model accounting for both prostate cancer aggressiveness and comorbidities demonstrates substantial heterogeneity in the estimated absolute benefit of ADT within conventional risk groups. This approach provides individualized estimates to support treatment discussions; external validation is needed before clinical implementation.

BackgroundRadiotherapy can provide durable local control for optic pathway-hypothalamic glioma (OPHG), but its use is limited by concern regarding delayed vascular, endocrine, visual, oncological, and neurological toxicities. ObjectiveTo systematically characterize and quantify the safety of radiotherapy and radiosurgery for OPHG and explore clinically relevant modifiers of treatment-related toxicity. MethodsPubMed, Scopus, Web of Science, Embase, Cochrane, Google Scholar, and ClinicalTrials.gov were searched from inception through 1 June 2026. Eligible non-randomized studies reporting safety outcomes after radiotherapy or radiosurgery were included. Random-effects binomial-normal generalized linear mixed-effects models were used to pool proportions, with exact conditional models for sparse comparative analyses. ResultsThirty-five studies were included, of which 31 contributed event-level data to at least one quantitative safety outcome. The pooled incidence of any treatment-related toxicity was 8.46% (95% CI, 1.37-38.01%). Vasculopathy occurred in 9.44% (95% CI, 5.22-16.49%). Secondary neoplasms occurred in 5.41% (95% CI, 2.23-12.53%), decreasing to 2.83% under a strict malignant-event definition. Incident endocrinopathy had the highest pooled estimate at 21.19% (95% CI, 4.72-59.31%) and increased with longer follow-up. Treatment-related visual toxicity was 2.26%, whereas radiation-related mortality was 0.59%. Radiation necrosis, severe toxicity, and treatment-attributed neurocognitive toxicity were sparsely reported. ConclusionLate toxicity following radiotherapy for OPHG is heterogeneous, with endocrinopathy, vasculopathy, and secondary neoplasms representing the principal quantifiable safety concerns. Treatment decisions should therefore be individualized, with prolonged vascular, endocrine, visual, and oncological surveillance and further prospective evaluation of contemporary radiation techniques.

Patients with history of neck radiation for Human Papilloma Virus (HPV)-associated head and neck cancer experience rapid progression of carotid artery stenosis. The present study sought to determine whether screening and treating asymptomatic carotid artery stenosis in patients with a history of neck radiation is cost-effective. This study is a cost-utility analysis using a probabilistic Markov model over a thirty-year time horizon assessing carotid screening and treatment to avoid neurologic consequences of neck radiation for HPV-associated head and neck cancer. A strategy of no carotid surveillance was associated with a 14.8% cumulative risk of stroke and a strategy of ultrasound surveillance and treatment with TCAR was associated with a 3.0% cumulative risk of stroke. The latter had a median incremental cost of $1.04 million USD and provided a median 39.1 additional QALYs, resulting in a median incremental cost-effectiveness ratio of $26,556 per QALY. In conclusion, this study suggests that ultrasound surveillance and treatment with TCAR for asymptomatic carotid artery stenosis is likely to be cost-effective for patients who have been successfully treated with radiation therapy for HPV-associated head and neck cancer.