Citations

PURPOSE Cancer recurrence is a critical outcome for patients and physicians. Retrospective cancer recurrence data can evaluate recurrence-directed treatment and generate novel interventions targeting recurrent cancers. However, large cancer databases do not provide recurrence-related information, stymying study at scale, and consequently require significant manual record review. Automated evaluation of records may allow for the rapid generation of easily analyzed data sets, accelerating the evaluation of recurrence altogether. METHODS Patients treated with radiation therapy at one tertiary referral center from 2010 to 2018 with a verified cancer status (cancer recurrence v no cancer recurrence) were identified. Patients with recurrent disease were initially identified through manual record review, and the associated pathology report was collected. Google Automated Machine Learning with Natural Language Processing (AutoNLP) and Google Gemini 1.5 Pro were used to generate a model for binary classification, with comparison to the gold-standard manually developed data set. RESULTS A total of 7,054 patients were identified. 3,431 (48.6%) were female, with a median age of 64 years. Head and neck (1,482, 21%), breast (1,480, 21%), upper GI (1,307, 18.5%), and lung/thorax (1,107, 15.7%) were the most common disease sites. Recurrence was verified for 1,546 patients (21.9%) using pathology reports, of which 1,249 positive cases were paired with 651 negative pathology reports for model development. Google Gemini 1.5 Pro consistently outperformed AutoNLP across all measurements of accuracy, generating a greater absolute difference in precision, recall, negative predictive value, and specificity, and a higher likelihood of correct classification at the individual level, rendering Gemini superior in recurrence status extraction. CONCLUSION AutoNLP and Google Gemini 1.5 Pro are promising tools for identifying recurrence from pathology reports, with the latter demonstrating superior overall performance, making it particularly suitable for clinical translation.

IMPORTANCE Neoadjuvant chemoimmunotherapy has high rates of pathologic response and improved prognosis among solid tumors. Its role in head and neck squamous cell carcinoma (HNSCC) remains understudied. OBJECTIVE To summarize the efficacy associated with neoadjuvant chemoimmunotherapy regimens in HNSCC and to compare the efficacy associated with neoadjuvant chemoimmunotherapy vs immunotherapy alone in locoregionally advanced resectable HNSCC prior to definitive surgery. DATA SOURCES A librarian-led systematic review was conducted using multiple electronic databases, including Medline, EMBASE, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and Scopus, following Preferred Reporting Items for Systematic Reviews and Meta- Analyses (PRISMA) guidelines. Databases were queried from inception to October 2024. STUDY SELECTION Inclusion criteria included prospective interventional trials that have completed accrual, with results published in English. Only trials of resectable, treatment-naive HNSCC were included. Exclusion criteria included lack of reporting on pathologic response or Response Evaluation Criteria In Solid Tumors (RECIST) version 1.1 response. DATA EXTRACTION AND SYNTHESIS Two investigators independently conducted title, abstract, and full-text reviews per PRISMA guidelines. Primary outcomes extracted were major pathologic response (MPR) and complete pathologic response (CPR) at surgery and complete response (CR) on imaging by RECIST 1.1. Secondary outcomes included 1-year overall survival (OS) and toxic effects. MAIN OUTCOMES AND MEASURES A meta-analysis was performed using a binary random-effects model to determine the pooled proportion of primary outcomes in the immunotherapy and chemoimmunotherapy groups. Heterogeneity was reported as I 2 . RESULTS A total of 751 patients (502 [77%] male; age range, 27-87 years) across 23 studies were included in the meta-analysis, including 357 patients (47%) who underwent chemoimmunotherapy, 102 patients (14%) who underwent dual-agent immunotherapy, and 292 patients (39%) who underwent single-agent immunotherapy. Pooled rates of MPR plus CPR were 66% (95% CI, 58%-73%; I 2 = 0.59) for chemoimmunotherapy, 18% (95% CI, 6%-29%; I 2 = 0.65) for dual-agent immunotherapy, and 6% (95% CI, 3%-9%; I 2 < 0.01) for single-agent immunotherapy. Across all groups, 1-year OS rates ranged from 88% to 96% for single-agent immunotherapy, 88% to 96% for dual-agent immunotherapy, and 88% to 100% for chemoimmunotherapy. Grade 3 to 5 adverse events were reported in 61 of 210 patients (29%), 2 of 67 patients (3.0%), and 36 of 210 patients (17%) among the reporting studies in the single-agent immunotherapy, dual-agent immunotherapy, and chemoimmunotherapy groups, respectively. CONCLUSIONS AND RELEVANCE These findings suggest that neoadjuvant chemoimmunotherapy regimens were associated with higher pathologic and radiographic response rates compared with immunotherapy alone in locoregionally advanced HNSCC. These results call for phase 3 trials comparing these neoadjuvant regimens head to head.

357 Background: Prostate cancer (PCa) is the most incident cancer in adult males in the United States, and the second highest in the world. CTCAE is the standardized method to grade the severity of treatment-related adverse events (AEs), but are tedious to collect and subject to inter-observer variability. This study aimed to evaluate the performance of LLMs in automatically extracting and grading CTCAE toxicities from clinical notes and patient-reported outcomes of PCa patients from a clinical trial. Methods: A total of 55 patients from NCT02874014 trial, undergoing proton therapy (PT) for localized PCa, were included. The ground truth of CTCAE graded toxicities was obtained from the trial’s results. Clinical notes within 3 days, and the most recent patient questionnaires within 90 days of toxicity assessment were retrieved from the electronic health record. A comprehensive prompt was developed to extract and CTCAE grade toxicities from the retrieved records. LLMs used for this analysis were OpenAI GPT-4o and Gemini 2.0 Flash. Phase 1 evaluated both LLMs' performance against the ground truth. To ensure the metrics reflected the models’ performance based only on the information available in the text provided, cases where both LLMs contradicted the ground truth were manually reviewed. Phase 2 recalculated LLMs' performance metrics using this adapted ground truth. A conservative stepwise ensemble model was developed to leverage the complementary strengths of both LLMs. Results: In the conservative stepwise ensemble model, Gemini 2.0 Flash served as the initial screening platform to identify AEs in step 1. All cases with identified toxicity events were then reviewed and confirmed using OpenAI GPT-4o in step 2. This effectively reduced false positive rates while preserving high sensitivity. The ensemble model achieved strong performance metrics: binary accuracy (0.954), sensitivity (0.961), specificity (0.953), F1 score (0.836), and grade accuracy (0.932). Conclusions: This study demonstrates the feasibility of the use of LLMs for extraction and CTCAE toxicity grading in PCa patients treated with RT. The ensemble model achieved strong performance metrics. Further validation is needed for other cancer diagnosis and treatment modalities. Phase Model Binary Accuracy Precision Sensitivity Specificity F1 Score Grade Accuracy 1 Gemini 0.946 0.672 1.000 0.939 0.804 0.919 1 OpenAI 0.940 0.785 0.637 0.978 0.703 0.895 2 Gemini 0.956 0.735 1.000 0.950 0.848 0.927 2 OpenAI 0.954 0.917 0.679 0.991 0.780 0.903

OBJECTIVE Brain metastases significantly impact neurocognitive function and overall survival. Stereotactic radiosurgery (SRS) is a cornerstone of treatment for patients with limited metastases and expected survival beyond 3 months. Despite current guidelines, up to 20% of patients with brain metastases undergoing SRS have been reported to die within 90 days. This study retrospectively evaluated prognostic factors associated with 90-day survival after SRS, aiming to improve patient selection. METHODS The authors retrospectively analyzed a cohort of 1546 patients who underwent Gamma Knife SRS for brain metastases at their institution between 2015 and 2023. One hundred seventy patients who survived less than 90 days after SRS were identified and case matched to 170 patients who survived over 90 days. Measured variables included patient demographic characteristics, tumor characteristics, treatment history, functional status, and control of the primary cancer. The authors modeled post-SRS 90-day survival using binomial and multivariate logistic regression. RESULTS Multivariate analysis highlighted Karnofsky Performance Score (KPS) < 70 (OR 17.4, p < 0.001), prior whole-brain radiation (OR 6, p = 0.004), and focal neurological deficits (OR 3.02, p = 0.003) as significant predictors of poor survival, whereas CNS progression before SRS (OR 0.22, p < 0.001) and control of systemic cancer (OR 0.556, p = 0.002) were associated with survival < 90 days. The predictive model demonstrated acceptable performance with an area under the curve (AUC) of 0.85, accuracy of 80%, sensitivity of 87%, and specificity of 71%. CONCLUSIONS Key predictors of 90-day survival after SRS for brain metastases include functional status (KPS), control of systemic cancer, CNS progression status, and focal neurological deficits. These findings are complementary factors that can assist in making decisions and SRS patient selection.

Background: Circulating tumor DNA (ctDNA) is predictive of recurrence in resected stage III melanoma, yet its role in the neoadjuvant setting for clinical stage III (cSIII) is unclear. Objective: Assess the association between ctDNA and outcomes following neoadjuvant immunotherapy (IO) + targeted therapy (TT) in cSIII melanoma. Design: Patients in the NeoACTIVATE study were treated with neoadjuvant IO + TT, underwent lymphadenectomy, and received adjuvant immunotherapy. Methods: Patients with ctDNA testing performed at baseline, pre- and post-operation were analyzed. Baseline positron emission tomography–computed tomography volumetrics and surgical, major pathological responses (MPR) were assessed. Results: Thirteen patients had serial ctDNA, 10 (77%) were detectable at baseline, and 9/10 (90%) had ctDNA clearance. Seven (54%) achieved MPR, all with undetectable preoperative ctDNA, yet 3/7 (43%) had disease recurrence. Conclusion: ctDNA and MPR are poor predictors of recurrence following neoadjuvant IO + TT for cSIII melanoma patients. Further studies are warranted to define the role of ctDNA in this setting.

Abstract Brain metastases significantly impact neurocognitive function and overall survival. Stereotactic radiosurgery (SRS) is a cornerstone of treatment for patients with limited metastases and expected survival beyond three months. Despite current guidelines, up to 20% of patients with brain metastases undergoing SRS have been reported to die within 90 days. This study retrospectively evaluates prognostic factors associated with 90-day survival after SRS, aiming to improve patient selection. We retrospectively analyzed a cohort of 1,546 patients who underwent Gamma Knife SRS for brain metastases at our institution between 2015 and 2023. One hundred seventy patients who survived fewer than 90 days post-SRS were identified and case-matched to 170 patients who survived over 90 days. Measured variables included patient demographics, tumor characteristics, treatment history, functional status, and control of the primary cancer. We modeled post-SRS 90-day survival using binomial and multivariate logistic regression. Multivariate analysis highlighted Karnofsky Performance Status (KPS) &lt; 70 (OR: 17.4, p &lt; 0.001), prior whole brain radiation (OR: 6, p = 0.004), and focal neurological deficits (OR: 3.02, p = 0.003) as significant predictors of early deaths, while CNS progression before SRS (OR: 0.22, p &lt; 0.001) and control of systemic cancer (OR: 0.556, p = 0.002) were associated with a less than 90-day survival. The predictive model demonstrated acceptable performance with an area under the curve (AUC) of 0.85, accuracy of 80%, sensitivity of 87%, and specificity of 71%. Key predictors of 90-day survival following SRS for brain metastases include functional status (KPS), control of systemic cancer, CNS progression status, and focal neurologic deficits. These findings are complementary factors that can assist in making decisions and SRS patient selection.

Abstract Background Deep learning auto segmentation (DLAS) tools are widely used for radiation therapy planning, yet limited information exists on how patient positioning impacts their performance, particularly in head and neck (HN) computed tomography (CT) imaging. Purpose This study investigates the impact of abnormal neck tilt on the accuracy of DLAS‐generated contours in HN CT scans. We hypothesize that abnormal positioning degrades auto segmentation performance, particularly for organs at risk (OARs) influenced by cervical spine orientation. Methods A total of 35 HN CT scans were retrospectively analyzed. Neck tilt was quantified using principal component analysis of the spinal cord contours from C1 to C4, with patients stratified into normal and abnormal tilt groups based on percentile thresholds. Seven FDA‐cleared DLAS tools were evaluated for OAR segmentation accuracy across brainstem, parotid glands, submandibular glands, brachial plexus, and optic nerves. Gold‐standard contours were curated from manually created clinical contours and compared to DLAS output using Sorenson's volumetric dice similarity coefficient (DSC), surface Dice similarity coefficient with a 2 mm threshold (sDSC), mean distance to agreement (MDA) and difference in mean dose. Statistical differences were assessed using Wilcoxon rank‐sum tests ( p < 0.05). Results Among the evaluated OARs, the parotid glands showed consistent and statistically significant degradation in contouring accuracy for patients with abnormal neck tilt across all metrics (DSC, sDSC, MDA) and across all seven DLAS tools. Changes in contouring accuracy resulted in significant differences in both absolute and signed difference in mean dose. No consistent differences were observed for other structures across multiple DLAS tools. Conclusions Abnormal neck tilt is associated with reduced DLAS accuracy and greater dose variability for parotid gland segmentation in HN radiotherapy planning. These findings underscore the need for enhanced patient quality assurance strategies for patients with abnormal positioning when using clinical DLAS workflows. Future work should explore automated detection of anatomical outliers and site‐specific model retraining to ensure accurate contouring delineation for patients with atypical positioning or anatomy.

Importance Adjuvant radiotherapy can improve locoregional control and survival in patients with olfactory neuroblastoma (ONB), particularly with advanced-stage and histologic-grade disease. Standard radiotherapy treatment is with intensity-modulated radiotherapy (IMRT). Proton beam radiotherapy (PBRT) provides theoretical advantages in greater sparing of dose to uninvolved organs at risk. Objective To investigate if there are differences in the effectiveness and radiation treatment–related adverse events (RTAEs) between adjuvant IMRT and PBRT for patients with ONB. Design, Setting, and Participants This propensity score–matched cohort study included patients with ONB treated between February 2005 and April 2021 with either IMRT or PBRT at 9 academic tertiary care centers in North America. Patients were matched 1:2 based on age, modified Kadish stage, and Hyams grade. Data were analyzed from July 2024 to January 2025. Exposure Adjuvant IMRT or adjuvant PBRT. Main Outcomes and Measures Local recurrence-free survival (RFS), any RFS, and overall survival (OS). RTAEs, ie, grade 2 events or higher based on Common Terminology Criteria for Adverse Events, were recorded for both modalities. Results Of 54 included patients, 27 (50%) were female, and the mean (SD) age was 46.2 (15.4) years. A total of 18 were treated with PBRT and 36 were treated with IMRT. Most patients had modified Kadish stage C disease (33 of 54 [61%]), and 24 patients (44%) had Hyams grade III or IV disease. The RTAE rate was 20% (8 of 40); IMRT had a rate of 21% (6 of 29), and PBRT had a rate of 18% (2 of 11). The difference in the point estimates for 10-year RFS showed a potential clinical benefit favoring IMRT, although the wide confidence interval indicates uncertainty (10-year RFS: IMRT, 63.3%; 95% CI, 44.6-89.8; PBRT, 37.8%; 95% CI, 14.2-100; difference, 25.5 percentage points; 95% CI, −17.6 to 68.6). There were no clinically meaningful differences in 10-year local RFS (IMRT, 75.6%; 95% CI, 59.8-95.4; PBRT, 72.7%; 95% CI, 45.2-100; difference, 2.9 percentage points; 95% CI, −35.9 to 41.7) or 10-year OS (IMRT, 61.8%; 95% CI, 42.8-89.1; PBRT, 57.1%; 95% CI, 24.3-100; difference, 4.7 percentage points; 95% CI, −49.2 to 58.6), although wide confidence intervals indicate considerable uncertainty. Conclusions and Relevance Due to the imprecision of estimates, no definitive conclusions can be made regarding the comparative effectiveness of IMRT vs PBRT for patients with ONB. These preliminary data may inform the design of appropriately powered prospective studies evaluating the efficacy of PBRT vs IMRT in this population.

ABSTRACT Background As head and neck cancer survival rates improve, there is growing demand for quality survivorship care and tools to predict patient adherence; this study aims to evaluate an advanced practice provider ( APP )‐led survivorship clinic and assess these key metrics. Methods We performed a retrospective review of 213 head and neck cancer patients treated with curative radiation therapy between 2016 and 2023, to assess follow‐up adherence in an advanced practice provider–led survivorship clinic. Results With a median follow‐up of 2.4 years, follow‐up care metrics identified high completion percentages (> 90%) for lung cancer screening, smoking cessation, thyroid screening, dental evaluation, xerostomia, dysphagia, and dysgeusia. The HOUsing ‐based SocioEconomic Status ( HOUSES ) index at diagnosis showed a significant adherence difference (median percentile, 78% vs. 55%, adherent vs. nonadherent; p = 0.03). Conclusions High adherence rates with guideline‐defined care metrics are achievable in an advanced practice provider‐led head and neck cancer survivorship clinic. The HOUSES index may assist in predicting nonadherence; however, further studies are needed.