Journal of Neuro-Oncology最新文献

{"title":"Meningiomas classified as grade 2 due to brain invasion alone have better progression-free survival than other atypical meningiomas.","authors":"Salem M Tos, Sofia Kollia, Georgios Mantziaris, Georgios A Maragkos, Carson Brantley, Jared H Chung, Maria-Beatriz Lopes, Ashok R Asthagiri","doi":"10.1007/s11060-026-05783-1","DOIUrl":"10.1007/s11060-026-05783-1","url":null,"abstract":"<p><strong>Background: </strong>Brain invasion (BI) was incorporated as a major criterion for grade 2 meningiomas in the 2016 WHO guidelines. However, its prognostic value remains controversial due to variability in surgical sampling and histopathologic interpretation, as well as BI may be the consequence of mechanical forces related more with the size of the tumor than the biologic presence of an innately more aggressive biologic lesion. Despite advances in molecular characterization of meningiomas, the relevance of BI as an independent prognostic indicator remains unresolved. We sought to determine whether BI alone, taken out of context of other parameters, particularly tumor size, truly represents a more aggressive standalone feature, by comparing time-to-progression (TTP) in atypical meningiomas classified as such due to BI alone versus other atypical meningiomas.</p><p><strong>Methods: </strong>A single-institutional retrospective cohort study of patients with resected meningiomas from 2003 to 2023 was performed. Pathology reports were reviewed and tumors were re-graded using current WHO criteria. Only WHO Grade 2 patients were included in the study. Patients were then divided into two cohorts: The BIOB (Brain Invasion Otherwise Benign) group and \"other criteria\" (OC) group. Primary outcome was TTR and association of factors with TTR were assessed using Cox regression.</p><p><strong>Results: </strong>131 patients with a mean age of 59.2 ± 15.9 years were included. Median (IQR) preoperative tumor volume was 34.3 cm³ (16-57.9). Gross total resection was achieved in 69 (53%) tumors. 18 patients (14%) had adjuvant radiation. BI was noted in 78 (60%) patients, 45 (34%) of which were classified as BIOB. Mean radiographic follow-up was 74 ± 56 months. Progression was noted in 40 (31%) patients at a mean of 40 ± 39 months post-resection. Progression occurred in 17.8% of BIOB patients (8/45) and 37.2% of OC patients (32/86). Baseline demographic, radiologic, and treatment characteristics did not differ significantly between BIOB and OC groups (all P > 0.3). On univariate Cox regression, TTR was associated with the BIOB group (Hazard Ratio [HR] = 0.44, P = 0.036), GTR (HR = 0.29, P < 0.001), and residual volume (HR = 1.04, P = 0.002). On multivariable Cox regression, BIOB was associated with longer TTR (HR = 0.44, P = 0.041), when controlling for GTR and posterior fossa location.</p><p><strong>Conclusions: </strong>Atypical meningiomas secondary to brain invasion alone were associated with significantly longer TTR than tumors with other Grade 2 criteria. This was especially robust for larger BI-only tumors. Further research is needed to delineate if brain invasion should truly be a standalone criterion for Grade 2 classification.</p>","PeriodicalId":16425,"journal":{"name":"Journal of Neuro-Oncology","volume":"179 2","pages":""},"PeriodicalIF":3.4,"publicationDate":"2026-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13569586/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148945003","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Assessing the personal financial impact of vestibular schwannoma management.","authors":"Robert J Macielak, William Liu, Nicholas S Andresen, Yin Ren, Edward E Dodson, Daniel M Prevedello, Oliver F Adunka, Kyle C Wu, Desi P Schoo","doi":"10.1007/s11060-026-05797-9","DOIUrl":"10.1007/s11060-026-05797-9","url":null,"abstract":"<p><strong>Purpose: </strong>To investigate the personal financial impact of vestibular schwannoma (VS) management through a survey hosted by the Acoustic Neuroma Association (ANA).</p><p><strong>Methods: </strong>A cross-sectional survey was created to ask participants with a VS questions relating to their personal finances and employment with a specific emphasis on financial toxicity, with this toxicity being measured through the COmprehensive Score for financial Toxicity (COST) assessment and the Financial Distress Questionnaire (FDQ).</p><p><strong>Results: </strong>At the time of study completion, 41 respondents had fully completed the provided survey, with the majority being > 50-years-old at the time of intervention (24 participants, 59%) and female (33 participants, 80%). Twenty-nine participants (71%) reported undergoing microsurgical resection of their VS, 9 (22%) documented undergoing radiation therapy, and 3 (7%) described undergoing a combination of both modalities. After initiation of management, 11 respondents (27%) experienced a change in employment and 17 (41%) experienced a change in their duties at work, including change in hours, reduced activity, or new unemployment. The median score on the COST questionnaire was 32 (IQR 15-39). On the FDQ, only 4 respondents (10%) were noted to have moderate financial toxicity and 2 (5%) reported severe financial toxicity.</p><p><strong>Conclusions: </strong>Among surveyed VS patients after treatment, change in employment as well as job duties can be observed; however, the ultimate financial toxicity of these changes appears limited, with few patients experiencing substantial financial hardship. Further study of this point through prospective study could better identify risk factors associated with personal financial difficulty and improve generalizability.</p>","PeriodicalId":16425,"journal":{"name":"Journal of Neuro-Oncology","volume":"179 2","pages":""},"PeriodicalIF":3.4,"publicationDate":"2026-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148929751","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Contrast enhancement: the missing imaging variable in IDH-wildtype gliomas with historical WHO grade 2 and 3 histology.","authors":"Alexandre Roux, Marc Zanello, Johan Pallud","doi":"10.1007/s11060-026-05791-1","DOIUrl":"10.1007/s11060-026-05791-1","url":null,"abstract":"","PeriodicalId":16425,"journal":{"name":"Journal of Neuro-Oncology","volume":"179 2","pages":""},"PeriodicalIF":3.4,"publicationDate":"2026-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148920037","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Stereotactic radiosurgery for brain metastases from male breast cancer: a multicenter retrospective study.","authors":"Ali H Duzkalir, Mehmet O Askeroglu, Dogu C Yildirim, Kristen Marciniuk, Kenneth Bernstein, Salem M Tos, Jheremy S Reyes, Roman Liscak, Gabriela Simonova, Takuma Sumi, Hideyuki Kano, Turker Kilic, Deniz Kilic, Greg N Bowden, Shachar Z Shemesh, Tehila Kaisman-Elbaz, Cheng-Chia Lee, Huai-Che Yang, Tugce Kutuk, Joshua D Palmer, David Mathieu, Constantinos G Hadjipanayis, Ajay Niranjan, Jason P Sheehan, Douglas Kondziolka, L Dade Lunsford, Selcuk Peker","doi":"10.1007/s11060-026-05795-x","DOIUrl":"10.1007/s11060-026-05795-x","url":null,"abstract":"<p><strong>Purpose: </strong>Brain metastases (BMs) from male breast cancer (MBC) are exceptionally rare, and the efficacy and safety of stereotactic radiosurgery (SRS) in this population remain poorly characterized.</p><p><strong>Methods: </strong>We conducted a multicenter retrospective cohort study across 12 institutions. Eligible patients had histologically confirmed MBC with histopathologically or radiologically documented BMs treated with SRS. Lesion-level outcomes were analyzed using generalized estimating equations (GEE) with robust variance estimation. Overall survival (OS) was estimated by the Kaplan-Meier method. Neurological symptom change was assessed with McNemar's exact test, and discriminative ability of dosimetric variables for local failure was evaluated using receiver operating characteristic (ROC) analysis.</p><p><strong>Results: </strong>Nineteen patients harboring 185 SRS-treated lesions were included. Median age at SRS was 59.0 years. The overall crude local control rate was 97.3%. Median OS was 28.0 months. Patients with controlled systemic disease at SRS demonstrated significantly longer median OS than those without (40.0 vs. 22.0 months; p = 0.032). Target volume was the only significant discriminator of local failure on ROC analysis (p < 0.001; optimal threshold 0.98 cc) and independently predicted adverse radiation effects on GEE analysis (p = 0.013). Neurological symptom burden decreased significantly after SRS (63.2% to 26.3%; p = 0.016), and median Karnofsky Performance Status remained stable.</p><p><strong>Conclusion: </strong>SRS achieves high local control with a favorable safety and neurological profile in MBC BMs. Systemic disease control at the time of SRS is the primary determinant of survival, while target volume is the principal dosimetric predictor of both local failure and radiation toxicity. These findings support the application of established SRS practice parameters in this rare population.</p>","PeriodicalId":16425,"journal":{"name":"Journal of Neuro-Oncology","volume":"179 2","pages":""},"PeriodicalIF":3.4,"publicationDate":"2026-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148919969","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Diffusion histogram analysis predicts progression-free survival in contrast enhancing recurrent IDH mutant gliomas treated with bevacizumab.","authors":"Elizabeth Loxterkamp, Audrey Luo, Francesco Sanvito, Collin T Le, Catalina Raymond, Chuyin Yang, Terry J Prins, Addison Fisher, Noriko Salamon, Linda M Liau, Robert A Chong, Phioanh L Nghiemphu, David A Nathanson, Timothy F Cloughesy, Albert Lai, Benjamin M Ellingson","doi":"10.1007/s11060-026-05782-2","DOIUrl":"10.1007/s11060-026-05782-2","url":null,"abstract":"<p><strong>Background: </strong>Pretreatment apparent diffusion coefficient (ADC) histogram analysis-specifically the lower Gaussian peak mean (ADC-L)-is a predictive biomarker of progression-free survival (PFS) and overall survival (OS) in recurrent IDH-wildtype glioblastoma receiving anti-VEGF treatment including bevacizumab. IDH-mutant gliomas differ biologically from glioblastoma, as they exhibit lower tumor cellularity, longer survival, and distinct sensitivity to a variety of therapies. Whether ADC-L retains predictive utility in IDH-mutant recurrent glioma has not been established.</p><p><strong>Methods: </strong>In this retrospective, single-center study, sixty patients with IDH-mutant recurrent glioma (41 astrocytoma, 19 oligodendroglioma) with measurable contrast enhancement and available pretreatment diffusion-weighted MRI who received bevacizumab were analyzed. Univariate and multivariate Cox regression and Kaplan-Meier analyses were performed for PFS and OS.</p><p><strong>Results: </strong>The optimal ADC-L threshold was 1.19 μm²/ms. High ADC-L (≥1.19 μm²/ms) was associated with significantly longer PFS (median 5.46 vs. 2.76 months; p = 0.006) but not OS. In multivariate analysis, low ADC-L (HR = 2.054; p = 0.0228) and astrocytoma histology (HR = 2.295; p = 0.0158) were independent predictors of shorter PFS when accounting for number of prior recurrences. ADC-L was also significant predictor of PFS in both astrocytoma and oligodendroglioma patients, independently, after accounting for number of recurrences.</p><p><strong>Conclusions: </strong>Pretreatment ADC-L predicts PFS in IDH-mutant recurrent glioma receiving bevacizumab. Supporting prospective evaluation of ADC-L as a biomarker for bevacizumab patient selection in this population.</p>","PeriodicalId":16425,"journal":{"name":"Journal of Neuro-Oncology","volume":"179 2","pages":""},"PeriodicalIF":3.4,"publicationDate":"2026-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13558324/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148920047","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Treatment selection or treatment effect? Interpreting functional outcomes in motor cortex metastases.","authors":"Volkan Demircan","doi":"10.1007/s11060-026-05787-x","DOIUrl":"10.1007/s11060-026-05787-x","url":null,"abstract":"<p><p>The optimal management of brain metastases involving the primary motor cortex requires balancing local tumor control with preservation of neurological function. Kraus et al. provide valuable evidence supporting microsurgical resection (MSR) in appropriately selected patients; however, comparisons between MSR and stereotactic radiotherapy (SRT) warrant careful interpretation. Treatment allocation was influenced by baseline performance status, motor function, disease burden, and lesion characteristics. Patients undergoing surgery had greater baseline motor impairment and therefore greater opportunity for measurable improvement, whereas preservation of intact or near-intact function after SRT may be categorized as stability despite being clinically meaningful. Conversely, 17 of the 41 patients initially allocated to SRT underwent salvage surgery for progression or a new neurological deficit and were excluded from the comparative SRT cohort. This 41% crossover both selects the analysed SRT cohort for patients in whom radiotherapy was successful and identifies a clinically important subgroup with poor outcomes after initial SRT. Furthermore, postoperative radiotherapy was administered to 95% of surgically treated patients with available data, indicating that the study compares a multimodality strategy of MSR followed by postoperative irradiation with definitive SRT rather than isolated surgery with radiotherapy. The unreported extent of resection further complicates interpretation of local control and the contribution of postoperative treatment. Collectively, the findings support surgery for appropriately selected patients, particularly those with large symptomatic lesions requiring rapid decompression, but do not establish better functional outcomes with surgery than with SRT after accounting for treatment selection. Future studies should report outcomes from initial treatment allocation through salvage therapy and evaluate individualized strategies incorporating lesion volume, baseline neurological status, corticospinal tract involvement, extent of resection, systemic disease burden, and the need for immediate decompression.</p>","PeriodicalId":16425,"journal":{"name":"Journal of Neuro-Oncology","volume":"179 2","pages":""},"PeriodicalIF":3.4,"publicationDate":"2026-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148911552","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Prognostic value of the Evans index in neoplastic meningitis: a real-world cohort study.","authors":"Alipio González-Vázquez, Nydia A Lorenzana-Mendoza, Juan Armando Reyes Pérez, Bernardo Cacho-Díaz","doi":"10.1007/s11060-026-05780-4","DOIUrl":"10.1007/s11060-026-05780-4","url":null,"abstract":"<p><strong>Purpose: </strong>Neoplastic meningitis (NM) from solid tumors is associated with a grim outcome. The Evans Index (EI) has recently been proposed as a prognostic imaging biomarker in these patients, but external validation is lacking. We aimed to independently validate its prognostic performance in one of the largest real-world cohorts reported to date.</p><p><strong>Methods: </strong>Information from patients with NM seen from January 2010 to December 2024 at a single center with an imaging study available for review was retrospectively analyzed. Multivariable adjustments were made to measure the association between EI and overall survival (OS).</p><p><strong>Results: </strong>A total of 345 patients were included; the most prevalent primary tumors were breast (66%), lung (12%), genitourinary (6%), and head and neck (5%). The median OS (MOS) of the whole cohort was 6.9 months (95% CI, 5.5 to 8.3); the MOS for those with EI ≤ 0.3 was 7.36 months, compared with 2.99 months for those with EI > 0.3 (P = .020). After adjusting for age, sex, primary tumor, Karnofsky Performance Score (KPS), symptoms, extracranial disease activity, CSF protein, and CSF lactic dehydrogenase levels, the EI was significantly associated with mortality (EI > 0.3: HR 1.95 [95% CI, 1.09-3.50], P = .024).</p><p><strong>Conclusion: </strong>In this validation study, the baseline EI was associated with prognosis even after adjustment for confounding factors. The EI may have clinical utility in assessing outcomes in patients with NM.</p>","PeriodicalId":16425,"journal":{"name":"Journal of Neuro-Oncology","volume":"179 2","pages":""},"PeriodicalIF":3.4,"publicationDate":"2026-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148906039","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Integrating ex vivo platforms with AI to guide glioblastoma treatment.","authors":"Mainak Bardhan, Samanta Paul, Sirjana Pun, Manish Raj Aryal, Pier Paolo Claudio, Jason Binder, Riccardo Barrile, Daniel A Pomeranz Krummel","doi":"10.1007/s11060-026-05708-y","DOIUrl":"10.1007/s11060-026-05708-y","url":null,"abstract":"<p><strong>Purpose: </strong>Ex vivo platforms can rapidly and cost-effectively screen patient-derived tumor cells or tissue. Artificial intelligence (AI) algorithms can search and identify patterns in large datasets and provide predictions. This review focuses on integrating microphysiological platforms with AI to inform physician and patient decision-making.</p><p><strong>Methods and results: </strong>Combining efficacy, safety, and pharmacology results from drug screens with the output of extensive AI searches can yield insights to guide physician and patient decision-making and potentially improve a patient's prognosis. We detail ex vivo platforms at different stages of development that represent the diversity of approaches: a microphysiological system and a high-throughput screen that assesses drug cytotoxicity in both bulk and drug-tolerant tumor cells. We review AI approaches that can enhance the utility of microphysiological platforms.</p><p><strong>Conclusion: </strong>Integrating emerging microphysiological platforms with AI is expected to significantly impact physician and patient choice of treatment.</p>","PeriodicalId":16425,"journal":{"name":"Journal of Neuro-Oncology","volume":"179 2","pages":""},"PeriodicalIF":3.4,"publicationDate":"2026-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148887840","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Annexin A2 associated with bevacizumab resistance and infiltrative escape in glioblastoma.","authors":"Taketo Ezaki, Ryota Tamura, Yohei Yamamoto, Jun Takei, Akihiko Teshigawara, Kyoichi Tomoto, Yasuharu Akasaki, Yuichi Murayama, Masahiro Toda, Hikaru Sasaki, Keisuke Miyake, Toshihide Tanaka","doi":"10.1007/s11060-026-05772-4","DOIUrl":"10.1007/s11060-026-05772-4","url":null,"abstract":"<p><strong>Background: </strong>The clinical benefit of bevacizumab (Bev) in glioblastoma (GBM) is typically transient, as tumors develop adaptive resistance characterized by an infiltrative shift from angiogenesis dependent growth. This study investigated Annexin A2 (ANXA2), a regulator of both angiogenesis and invasion, as a potential biomarker of Bev resistance.</p><p><strong>Methods: </strong>We analyzed 66 tissue-specimens from 33 GBM patients using quantitative real-time PCR and immunohistochemistry/immunofluorescence. The cohort included 15 Bev-naïve cases and 18 neoadjuvant Bev (neoBev) cases. 33 refractory specimens-including unique paired samples from the same patients at different treatment phases (initial resection vs. refractory stage) were evaluated to correlate ANXA2 expression levels with progression-free survival (PFS), overall survival (OS), and MRI recurrence patterns.</p><p><strong>Results: </strong>Multivariate analysis identified high ANXA2 expression as a significant independent poor prognostic factor for both PFS and OS. In the neoBev group, patients with low ANXA2 mRNA expression levels demonstrated significantly superior survival outcome compared to those with high expression. While ANXA2 mRNA levels tended to increase at the refractory stage, a significant negative correlation was observed between the number of Bev cycles and ANXA2 expression levels. Histopathological analyses revealed intense ANXA2 expression in the tumor vasculature and stroma at the GBM leading edge, co-localization with the hypoxia marker hypoxia-inducible factor-1α.</p><p><strong>Conclusions: </strong>ANXA2 may serve as a predictive biomarker of the adaptive transition from an angiogenic to infiltrative phenotype with Bev therapy. These findings suggest that ANXA2 serves as a predictive biomarker for Bev clinical benefit and may represent a candidate therapeutic target requiring further validation in GBM.</p>","PeriodicalId":16425,"journal":{"name":"Journal of Neuro-Oncology","volume":"179 2","pages":""},"PeriodicalIF":3.4,"publicationDate":"2026-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13541986/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148887792","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Therapeutic hybrid intelligence with neural and knowledge-based expert reasoning for SRS (THINKERS): a mixture-of-experts framework for gastrointestinal brain metastases.","authors":"Jheremy S Reyes, Constantinos G Hadjipanayis, L Dade Lunsford, Ajay Niranjan","doi":"10.1007/s11060-026-05781-3","DOIUrl":"10.1007/s11060-026-05781-3","url":null,"abstract":"<p><strong>Background: </strong>Brain metastases from gastrointestinal (GI) malignancies are heterogeneous, while stereotactic radiosurgery dose selection remains guided by lesion size, location, prior treatment, and physician judgment rather than individualized estimates of local failure and toxicity. We developed THINKERS-GI, a three-expert mixture-of-experts framework for outcome prediction and dose-policy evaluation after Gamma Knife radiosurgery (GKRS).</p><p><strong>Methods: </strong>We performed a retrospective single-center study of 300 patients with 1,233 GI brain metastases. Pretreatment variables were used to train a three-expert mixture-of-experts neural network with discrete-time competing-risk survival modeling. Margin dose was incorporated as a queryable input for candidate-dose evaluation. Internal validation used grouped 5-fold cross-validation and a closed grouped holdout set. Performance was assessed using time-dependent AUC, C-index, Brier score, calibration, and MAE.</p><p><strong>Results: </strong>In grouped cross-validation, time-dependent AUC ranged from 0.808 to 0.894, with a C-index of 0.889 ± 0.034. The calibrated 12-month Brier score was 0.019 ± 0.013, and MAE for time to local failure was 1.8 ± 0.7 months. In the grouped holdout set, 12-month AUC was 0.806 (95% CI, 0.758-0.851), C-index was 0.781 (95% CI, 0.742-0.818), calibrated Brier score was 0.101, and MAE was 1.27 months (95% CI, 1.2-1.9). Dose sweeping selected a dose within ±2 Gy of the delivered dose in 91.8% of lesions and increased predicted 12-month local control from 82.3% to 84.4%, while reducing predicted adverse radiation effect risk from 7.9% to 7.6%.</p><p><strong>Conclusions: </strong>THINKERS-GI provides an internally validated framework for individualized local failure prediction, toxicity estimation, and dose-policy evaluation after GKRS for GI brain metastases.</p><p><strong>Clinical trial number: </strong>Not applicable.</p>","PeriodicalId":16425,"journal":{"name":"Journal of Neuro-Oncology","volume":"179 2","pages":""},"PeriodicalIF":3.4,"publicationDate":"2026-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148880802","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}