Objectives: This retrospective study aims to develop a multiomics approach that integrates radiomics, dosiomics, and delta features to predict treatment responses in brain metastasis (BM) patients undergoing PULSAR. Methods: A retrospective study encompassing 39 BM patients with 69 lesions treated with PULSAR was undertaken. Radiomics, dosiomics, and delta features were extracted from both pre-treatment and intra-treatment MRI scans alongside dose distributions. Six individual models, alongside an ensemble feature selection (EFS) model, were evaluated. The classification task focused on distinguishing between two lesion groups based on whether they exhibited a volume reduction of more than 20% at follow-up. Performance metrics, including sensitivity, specificity, accuracy, precision, F1 score, and the area under the receiver operating characteristic (ROC) curve (AUC), were assessed. Results: The EFS model integrated the features from pre-treatment radiomics, pre-treatment dosiomics, intra-treatment radiomics, and delta radiomics. It outperformed six individual models, achieving an AUC of 0.979, accuracy of 0.917, and F1 score of 0.821. Among the top nine features of the EFS model, six features came from post-wavelet transformation and three from original images. Conclusions: The study demonstrated the feasibility of employing a data-driven multiomics approach to predict treatment outcomes in BM patients receiving PULSAR treatment. Integrating multiomics with intra-treatment decision support in PULSAR shows promise for optimizing patient management and reducing the risks of under- or over-treatment.

Purpose of the Report: The PRIMARY score standardizes interpretation of PSMA PET for intraprostatic assessment of clinically significant prostate cancer (csPCa). Although early studies are promising, no meta-analysis has systematically evaluated its accuracy. We aimed to assess the diagnostic performance of PRIMARY and compare it with PI-RADS. Material and Methods: A systematic review and meta-analysis were performed according to PRISMA-DTA guidelines. Studies assessing the PRIMARY score on PSMA PET/CT or PET/MRI for csPCa detection were included. The primary analysis focused on biopsy-naïve patients; secondary analyses incorporated all studies irrespective of prior histology, evaluated thresholds (≥3 vs. ≥4), compared PRIMARY with PI-RADS, and examined combined performance. Pooled sensitivity and specificity were estimated with a bivariate random-effects model. Results: Thirteen studies (n=1974) were included. In biopsy-naïve cohorts (6 studies), PRIMARY ≥3 achieved pooled sensitivity/specificity of 90% (95% CI: 86%–93%) and 63% (95% CI: 57%–69%). In an expanded analysis of 9 studies regardless of prior biopsy, values were 87% (82%–90%) and 56% (46%–65%). Across 5 studies directly comparing thresholds, raising the cutoff from ≥3 to ≥4 reduced sensitivity from 89% to 81% but increased specificity from 53% to 72%. Three head-to-head studies showed comparable accuracy between PRIMARY and PI-RADS. Conclusions: PRIMARY demonstrates high sensitivity and moderate specificity for csPCa, comparable to PI-RADS. In biopsy-naïve men it shows robust performance, and combination with mpMRI further improves sensitivity. Threshold selection should reflect clinical intent, supporting PRIMARY as a tool for pre-biopsy risk stratification and integrated diagnostic workflows.

Glioblastoma multiforme (GBM) is the most common and aggressive primary brain tumor in adults, with poor survival despite advancements in treatment. Adaptive stereotactic radiation therapy (RT) using a magnetic resonance imaging linear accelerator is an emerging approach for patients with newly diagnosed GBM eligible for conventional fractionation. We hypothesize that adaptive stereotactic RT can provide comparable outcomes with conventional fractionation while reducing treatment burden. We retrospectively reviewed 96 adults with newly diagnosed GBM treated at our institution between 2018 and 2024. Inclusion criteria included the age of 18 years, confirmed GBM diagnosis, and completed treatment. Patients with prior brain irradiation or incomplete treatment were excluded. Propensity score matching was performed to balance demographics, tumor characteristics, and treatment protocols across 5-fraction, 15-fraction, and 30-fraction groups. Statistical analyses included the Fisher exact test, Mann-Whitney U test, Cox proportional hazards models, and Kaplan-Meier survival curves. After propensity score matching, 17 pairs were matched for 5 versus 30 fractions and 14 pairs for 5 versus 15 fractions. Median overall survival was 21.1 versus 18.2 months (5 vs 15 fractions, P = .77) and 11.7 versus 14.6 months (5 vs 30 fractions, P = .5). Median progression-free survival was 9.0 versus 7.9 months (5 vs 15 fractions, P = .89) and 8.9 versus 9.7 months (5 vs 30 fractions, P = .97). Local failure and grade 3 toxicity rates were similar across groups. O6-methylguanine-DNA-methyltransferase unmethylated status, higher Eastern Cooperative Oncology Group scores, and age 60 years were associated with worse progression-free survival and overall survival. Median travel distances were lower in the 5-fraction group, with a median of 220 miles compared with 877.5 (15 fractions) and 1638 miles (30 fractions). Adaptive RT allowed for real-time tumor monitoring but volumetric changes did not correlate with clinical outcomes. Adaptive 5-fraction RT demonstrates comparable survival outcomes with conventional fractionation while reducing treatment-related travel burden. Further prospective studies are needed to validate its role in GBM management.

Purpose Identifying potential participants for clinical trials using traditional manual screening methods is time-consuming and expensive. Structured data in electronic health records (EHR) are often insufficient to capture trial inclusion and exclusion criteria adequately. Large language models (LLMs) offer the potential for improved participant screening by searching text notes in the EHR, but optimal deployment strategies remain unclear. Methods We evaluated the performance of GPT-3.5 and GPT-4 in screening a cohort of 74 patients (35 eligible, 39 ineligible) using EHR data, including progress notes, pathology reports, and imaging reports, for a phase 2 clinical trial in patients with head and neck cancer. Fourteen trial criteria were evaluated, including stage, histology, prior treatments, underlying conditions, functional status, etc. Manually annotated data served as the ground truth. We tested three prompting approaches (Structured Output (SO), Chain of Thought (CoT), and Self-Discover (SD)). SO and CoT were further tested using expert and LLM guidance (EG and LLM-G, respectively). Prompts were developed and refined using 10 patients from each cohort and then assessed on the remaining 54 patients. Each approach was assessed for accuracy, sensitivity, specificity, and micro F1 score. We explored two eligibility predictions: strict eligibility required meeting all criteria, while proportional eligibility used the proportion of criteria met. Screening time and cost were measured, and a failure analysis identified common misclassification issues. Results Fifty-four patients were evaluated (25 enrolled, 29 not enrolled). At the criterion level, GPT-3.5 showed a median accuracy of 0.761 (range: 0.554–0.910), with the Structured Out-put + EG approach performing best. GPT-4 demonstrated a median accuracy of 0.838 (range: 0.758–0.886), with the Self-Discover approach achieving the highest Youden Index of 0.729. For strict patient-level eligibility, GPT-3.5’s Structured Output + EG approach reached an accuracy of 0.611, while GPT-4’s CoT + EG achieved 0.65. Proportional eligibility performed better over-all, with GPT-4’s CoT + LLM-G approach having the highest AUC (0.82) and Youden Index (0.60). Screening times ranged from 1.4 to 3 minutes per patient for GPT-3.5 and 7.9 to 12.4 minutes for GPT-4, with costs of $0.02–$0.03 for GPT-3.5 and $0.15–$0.27 for GPT-4. Conclusion LLMs can be used to identify specific clinical trial criteria but had difficulties identifying patients who met all criteria. Instead, using the proportion of criteria met to flag candidates for manual review maybe a more practical approach. LLM performance varies by prompt, with GPT-4 generally outperforming GPT-3.5, but at higher costs and longer processing times. LLMs should complement, not replace, manual chart reviews for matching patients to clinical trials.