Citations

Lay Summary Glioblastoma is an aggressive brain cancer, and it can be difficult to tell whether changes seen on scans after radiation treatment represent tumor growth or treatment-related effects. The authors of this study wanted to evaluate whether a specialized imaging test called PSMA PET could help define tumor boundaries for radiation planning and distinguish tumor progression from treatment-related changes. To do this, they performed PSMA PET scans in 27 patients with high-grade gliomas, including patients undergoing radiation planning and patients being evaluated for possible tumor recurrence. They found that tumors showed higher PSMA PET signal than surrounding tissue, that PSMA PET sometimes identified tumor beyond what was visible on standard MRI, and that it correctly distinguished tumor progression from treatment-related changes in most patients.

Background Glioblastoma (GBM) remains the most aggressive primary adult brain cancer attributed to its immunosuppressive nature. Radiation therapy (RT) and concurrent temozolomide chemotherapy are the standard treatments for GBM. Emerging evidence indicates that metformin has potential as an anti-tumor agent that can reshape the immune landscape across various malignancies. We thus postulate that metformin may enhance the anti-tumor effect of RT by modulating the immunosuppressive milieu in GBM. Methods We first explore multiple in vitro conditions, with or without pre-induction, and mimicked a tumor microenvironment to demonstrate a highly context-dependent effect of metformin on the polarization of bone marrow-derived macrophages (BMDMs). We then investigated the antitumor activity and immune-modulatory effects of metformin in combination with RT in GBM-bearing mice. Results The in vitro experiment showed that metformin inhibited the immunosuppressive effects of IL-4/IL-13 but also the immunostimulatory effects of Lipopolysaccharide (LPS) on BMDMs, demonstrating bidirectional immunomodulatory properties that depend on the baseline inflammatory stimulus. In tumor cell co-culture environment, metformin exhibited context-dependent immunoregulatory effects, predominantly promoting M1-associated activation. Notably, concurrent metformin treatment counteracted M2 polarization typically induced by the tumor microenvironment. In a high m-MCSF inducted M2-dominated condition, metformin preferentially shifted the highly polarized M2 phenotype toward an M1-like state in a time and dose-dependent manner but with limited promotion of M1 subtype maturation. RNAseq results on BMDM treated with metformin showed increased homeostasis. Metformin showed dose-dependent immunomodulatory effects when combined with RT in vitro . In syngeneic GBM mouse models, concurrent metformin + RT significantly prolonged survival and reduced tumor burden by reprogramming tumor-associated macrophages (TAMs), elevating intratumoral CD8+ T-cell infiltration and the CD8+/Treg ratio, increasing circulating CD8+ T cells, reversing RT-induced expansion of monocytic myeloid-derived suppressor cells (mMDSCs), and expanding CD4+ and CD8+ effector memory populations in peripheral blood. Conclusion Metformin demonstrates inhibitory effects on M2 phenotype of macrophages. However, it is not a simple M2 inhibitor; instead, it functions primarily as a neutralizer of highly polarized macrophage states by promoting transcriptomic homeostasis. This context-dependent activity enables metformin to relieve the profoundly immunosuppressive TME of GBM to potentiate the anti-tumor effects of RT and by enhancing systemic immune memory.

Primary central nervous system choriocarcinoma is a non-germinomatous germ cell tumor with a highly malignant character, comprising up to 5% of all central nervous system germ cell tumors. Locations affected can include suprasellar or pineal locations most commonly, with others including the basal ganglia and lateral ventricles. Guidelines do not specify an optimal radiation regimen but recommend proceeding with maximally safe resection if feasible and then chemotherapy before radiotherapy. Radiation is often comprised of craniospinal irradiation with subsequent tumor bed boost. We present a case of an adult male with a primary central nervous system choriocarcinoma who underwent reversed sequence radiation therapy with photon tumor bed boost before proton craniospinal irradiation due to recurrent intracranial hemorrhages with long-term control of cancer and no subsequent hemorrhages. This case highlights the excellent outcome and feasibility of reversed sequence radiotherapy on choriocarcinoma in hemorrhage and tumor control.

BACKGROUND: Patients with locally advanced rectal cancer face substantial costs related to multimodal therapy. However, the extent of this burden is poorly characterized. OBJECTIVE: To characterize the prevalence and predictors of patient-reported financial toxicity after treatment for locally advanced rectal cancer. DESIGN: Multicenter retrospective cohort study with a cross-sectional survey administered to assess patients’ financial concerns as well as clinicodemographic data. SETTING: Three high-volume tertiary academic centers. PATIENTS: Patients who completed treatment for pathological stage II to III rectal adenocarcinoma between 2000 and 2019 and were alive. MAIN OUTCOME MEASURES: Patient-reported financial toxicity and risk factors associated with financial toxicity. RESULTS: A total of 523 patients were identified, 190 (36.3%) responded to the survey, and 2 were excluded for incomplete surveys. Of the responders, 163 (86.7%) were identified as having financial toxicity related to their cancer diagnosis, and most (96.8%) had insurance. Patients with financial toxicity were less likely to have fully paid medical leave (39.8% vs 75%; p = 0.03) compared to patients without financial toxicity. They also reported increased difficulty paying bills ( p < 0.001) and decreased monthly household income ( p < 0.001) due to their diagnosis. Furthermore, these patients reported increased difficulty paying for transportation to medical appointments ( p = 0.002) and were less likely to be completely satisfied with their insurance coverage ( p = 0.003). Feeling well informed regarding treatment-related costs was associated with decreased odds of financial toxicity (OR 0.14; 95% CI, 0.02–0.78; p = 0.02), whereas taking unpaid leave for treatment was associated with increased odds of financial toxicity (OR 4.71; 95% CI, 1.36–16.34; p = 0.01). LIMITATIONS: Cross-sectional design and moderate response rate may introduce selection and recall bias. CONCLUSIONS: Patients undergoing treatment for locally advanced rectal cancer, including those insured, are at risk for financial toxicity due to direct and indirect costs. Effective communication between these patients and physicians regarding the economic implications of treatment is essential to reducing financial toxicity. See Video Abstract . TOXICIDAD FINANCIERA TRAS EL TRATAMIENTO DEL CÁNCER RECTAL LOCALMENTE AVANZADO ANTECEDENTES: Los pacientes con cáncer de recto localmente avanzado enfrentan costos sustanciales relacionados con la terapia multimodal. Sin embargo, la magnitud de esta carga está poco caracterizada. OBJETIVO: Caracterizar la prevalencia y los predictores de la toxicidad financiera reportada por los pacientes después del tratamiento para el cáncer de recto localmente avanzado. DISEÑO: Estudio de cohorte retrospectivo multicéntrico con una encuesta transversal para evaluar las preocupaciones financieras de los pacientes, así como datos clínico-demográficos. ESCENARIO: Tres centros académicos terciarios de alto volumen. PACIENTES: Pacientes que completaron el tratamiento para el adenocarcinoma de recto en estadio patológico II-III entre 2000 y 2019 y estaban vivos. PRINCIPALES MEDIDAS DE RESULTADOS: Toxicidad financiera reportada por los pacientes y factores de riesgo asociados con la toxicidad financiera. RESULTADOS: Se identificó a un total de 523 pacientes, 190 (36,3%) respondieron a la encuesta y 2 fueron excluidos por encuestas incompletas. De los encuestados, 163 (86,7 %) presentaron toxicidad financiera relacionada con su diagnóstico de cáncer, y la mayoría (96,8 %) contaba con seguro médico. Los pacientes con toxicidad financiera tuvieron menos probabilidades de tener licencia médica con goce de sueldo completo (39,8 % frente a 75 %; p = 0,03) en comparación con los pacientes sin toxicidad financiera. También informaron mayor dificultad para pagar facturas ( p < 0,001) y una disminución de los ingresos familiares mensuales ( p < 0,001) debido a su diagnóstico. Además, estos pacientes informaron mayor dificultad para pagar el transporte a las citas médicas ( p = 0,002) y menor probabilidad de estar completamente satisfechos con su cobertura médica ( p = 0,003). Sentirse bien informado sobre los costos relacionados con el tratamiento se asoció con una menor probabilidad de toxicidad financiera (OR: 0,14; IC del 95 %: 0,02-0,78; p = 0,02), mientras que tomar una licencia sin sueldo para el tratamiento se asoció con una mayor probabilidad de toxicidad financiera (OR: 4,71; IC del 95 %: 1,36-16,34; p = 0,01). LIMITACIONES: El diseño transversal y la tasa de respuesta moderada pueden introducir sesgos de selección y de recuerdo. CONCLUSIONES: Los pacientes en tratamiento para el cáncer de recto localmente avanzado, incluidos los asegurados, corren el riesgo de sufrir toxicidad financiera debido a los costos directos e indirectos. Una comunicación eficaz entre estos pacientes y los médicos sobre las implicaciones económicas del tratamiento es esencial para reducir la toxicidad financiera. ( AI-generated translation )

Abstract Glioblastoma (GBM) remains the most aggressive primary adult brain cancer with limited therapeutic options. Our prior research demonstrated that androgen receptor antagonists (ARAs) enhance survival in GBM mouse models by preferentially suppressing glioma stem cells. This study investigates the potential of ARAs as radiosensitizers in combination with radiotherapy (RT). Combined effects of ARAs and RT were evaluated using an orthotopic GBM mouse model. RNA-Seq and TCGA analyses delineated AR-associated regulatory networks, while integrated cellular and molecular approaches utilizing human and mouse GBM cell lines, as well as in patient-derived primary high-grade glioma cultures, interrogated signaling and immune paradigms. ARA treatment induced G2/M cell cycle arrest, apoptosis, and downregulated DNA repair genes, with AR expression correlating with the DNA repair and TGF-β pathway. In both immortalized GBM cell lines and primary high-grade glioma cultured cells, ARAs in combination with RT showed only a modest enhancement of radiosensitivity. However, in an orthotopic GBM mouse model, ARA + RT demonstrated a strong synergy, achieving 100% long-term survival, compared to less than 50% with ARA alone and 0% with RT alone. Mechanistically, ARA modulated the TGF-β pathway, durably switching from Smad3 linker to c-terminal phosphorylation (pSmad3C) and inhibiting LIF/STAT3 axis. Distinct TGF-βs ligand expression patterns were observed in ARA-treated GBM cells. A protein–protein interaction noted between AR and Smad3, which was disrupted following ARA treatment, leading to elevated protein levels and nuclear localization of pSmad3C (S423/425), indicating normalization/activation of the TGF-β/pSmad3C-dependent anti-tumorigenic cascade. Comprehensive analyses in GBM cell lines and mouse model tissues demonstrated pathway reprogramming characterized by elevated TGF-β2 and increased pSmad3C. Tissue analysis revealed immune activation in the tumor microenvironment, while peripheral blood and spleen showed systemic immune responses following ARA + RT. Our study provides novel insights into how ARAs enhance RT efficacy through immunomodulation involving TGF-β/pSmad3C cascade, offering therapeutic implications in GBM.

Background The role of dose escalation of radiotherapy (RT) in unresectable stage III NSCLC followed by immunotherapy is unclear. The objective of the current study is to investigate if intermediate dose escalation (IDE) is beneficial in stage III NSCLC patients who receive definitive concurrent chemoradiation (dcCRT) followed by immunotherapy. Materials and methods The study used data from the National Cancer database. Multivariable Cox regression analysis was used to assess the all-cause mortality of patients who received standard RT dose (SD) (60 Gy ± 10%) and IDE (64–74 Gy). Results 47,315 patients were diagnosed in the era before immunotherapy and received dcCRT only, while 4,947 patients were treated with dcCRT and immunotherapy. In the cohort with dcCRT only, patients who received SD had statistically significant worse mortality but clinically minimal difference compared to patients with IDE (HR: 1.09, 95% CI: 1.07-1.12; p&lt;0.0001). In the era of immunotherapy, SD was still associated with worse mortality compared to IDE (HR: 1.17, 95% CI: 1.03-1.33; p=0.02). However, the survival benefit associated with IDE was only restricted to patients who started immunotherapy within six weeks after RT completion (HR: 1.26, 95% CI: 1.05-1.6; p=0.01). There was no difference in mortality between SD and IDE among patients who started immunotherapy between 7–10 weeks (HR: 1.13, 95% CI: 0.88-1.45; p=0.35) or &gt;10 weeks after RT completion (HR: 0.74, 95% CI: 0.51-1.07; p=0.11). Conclusion IDE of RT is not needed for patients diagnosed with stage III unresectable NSCLC who receive immunotherapy &gt; six weeks after dcCRT.

Abstract Background: Recent advances in computer vision have given rise to image foundation models (FMs), large-scale neural networks pre-trained on millions of unlabeled histopathology patches, which can learn generalizable representations of tissue morphology. In many cancers, these FMs have predicted genetic mutations and prognoses directly from histology, suggesting that morphology encodes “hidden” molecular information. In lymphoma, this field is nascent but growing.General-purpose histopathology FMs have limited representation of lymphoid morphology. We developed and evaluated LymphoVision, a lymphoma-specialized image foundation model pretrained with self-supervised learning, to extract clinically meaningful features from routine hematoxylin and eosin (H&E) histology. Methods: LymphoVision was pre-trained using knowledge distillation and consistency (DINOv2) on over 37 million multi-resolution patches (5x, 10x, 20x, 40x) sampled using a density-aware strategy from 31,211 H&E whole slide images (WSIs) from 9,155 archival lymphoma and reactive lymphoproliferative cases. To ensure diverse, representative samples, our patch sampling method uses an efficient graph-based clustering algorithm and prioritizes morphologically informative regions by clustering patch embeddings and performing density sampling within each cluster. The pretrained encoder was evaluated using a clustering-constrained attention multiple instance learning (MIL) framework on three diagnostic tasks: (1) cell-of-origin (COO) classification in diffuse large B-cell lymphoma (DLBCL), distinguishing germinal center B-cell (GCB) from non-GCB subtypes based on the Hans immunohistochemistry algorithm; (2) follicular lymphoma (FL) grading, differentiating low-grade (grades 1–2) from high-grade (grade 3); and (3) multiclass lymphoma subtype classification across six categories: DLBCL, FL, marginal zone lymphoma (MZL), mantle cell lymphoma (MCL), classical Hodgkin lymphoma (CHL), and benign/reactive lymphoid conditions. WSI-level labels were derived from pathology reports, and none of the evaluation cases were included in pretraining. Models used a ViT-Giant LymphoVision backbone to generate 20x patch embeddings, followed by MIL using 50% training data and 2-fold cross-validation (25% validation, 25% test). Performance was assessed using area under the receiver operating characteristic curve (AUC), accuracy, recall (sensitivity) and F1 score. Results: For DLBCL COO classification (total n=358 cases), LymphoVision achieved a mean test AUC of 0.93 (range 0.92–0.94), accuracy of 84.5% and F1 of 0.85 across two independent test folds (n=91 per fold, 50% GCB). These results outperformed general-purpose FMs on the same task and cases, including Virchow-v2 (AUC 0.85) and UNI-v2 (AUC 0.88), despite those models being pretrained on 10–100 times more images. For FL grading (total n=660 cases), the test set in each fold included 166 cases (18% grade 3), yielding a mean test AUC of 0.88 (range 0.87–0.90), accuracy of 89.2% and F1 of 0.80. In multiclass lymphoma subtype classification (total n=1,176 cases), the test set included 296 cases per fold. LymphoVision achieved a mean test AUC of 0.98 (range 0.97-0.99), mean test accuracy of 87.1% and mean F1 of 0.84. Class-wise recall was 92.0% for DLBCL (n=112), 88.2% for FL (n=72), 77.2% for MZL (n=29), 70.8% for MCL (n=35), 87.5% for CHL (n=24), and 95.8% for benign/reactive lymphoid conditions (n=24). Conclusion: LymphoVision is a lymphoma-specialized FM trained using self-supervision that achieves high diagnostic accuracy across multiple clinically meaningful tasks using weak supervision and standard H&E images, suggesting the encoder has learned fine-grained lymphoma morphology. Based on historical benchmarks from the 1997 International Lymphoma Study Group classification project, expert hematopathologists achieved diagnostic agreement rates of 55% to 84% using H&E morphology alone for the non-Hodgkin lymphoma (NHL) entities included in our multiclass task. LymphoVision approached or exceeded these benchmarks. These findings underscore the value of our approach using disease-specialized pretraining and intelligent patch sampling for data-efficient histopathology learning. LymphoVision has the potential to augment diagnostics, enable scalable AI-powered classification in real-world workflows, and serve as a foundation for prognostic and theragnostic applications. External validation and multimodal integration are ongoing.

ImportanceIdentification of factors associated with variation in outpatient surgery may further quality improvement efforts to safely reduce postoperative hospital length of stay nationally.ObjectivesTo explore variation in the use of outpatient surgery, incorporating patient, geographic, and hospital factors.Design, Setting, and ParticipantsThis retrospective cross-sectional study used deidentified administrative claims data from OptumLabs Data Warehouse. Participants included adults who underwent 1 of 10 general, urological, or gynecological operations between January 1, 2015, and June 30, 2021, in the US. Patients who underwent combined procedures or reoperations or had at least 15 Elixhauser comorbidities were excluded. Data were analyzed from July 26 to December 16, 2023.ExposureInpatient or outpatient surgical procedures.Main Outcomes and MeasuresMultilevel logistic regression assessed variation in the use of outpatient surgery rates by hospital characteristics (bed size, presence of trainees, and rural referral center status) and hospital census division, adjusting for patient factors (age, sex, number of Elixhauser comorbidities, year, and rural-urban commuting area). This multilevel model allowed for the sources of variability to be quantitatively attributed to patient characteristics, geography, and hospital characteristics.ResultsA total of 330 424 (72.3%) of 456 954 included patients underwent outpatient surgery. The median age was 54 (IQR, 41-67) years, and of those with data available, most patients were female (268 692 of 414 193 [64.9%]). The likelihood of outpatient surgery varied significantly by hospital census division for all 10 operations (eg, MIS salpingo-oophorectomy range, 29.6%-58.8%; P &amp;lt; .001). Variation in hospital census division contributed most to outpatient surgery for 8 of 10 operations compared with other patient and hospital characteristics. Hospital census division contributed the greatest degree to the variation in outpatient simple mastectomy (20.6%) and the least to outpatient open ventral hernia repair (0.7%). Multivariable analysis showed that the odds of outpatient surgery for patients from metropolitan areas were higher for minimally invasive salpingo-oophorectomy (odds ratio [OR], 1.62; 95% CI, 1.34-1.95) and open ventral hernia repair (OR, 1.16; 95% CI, 1.09-1.24). Hospitals with 400 or more beds were independently associated with decreased odds of outpatient surgery compared with hospitals with 50 to 199 beds for 4 of 7 operations (MIS paraesophageal hernia repair [OR, 0.58; 95% CI, 0.47-0.71; P &amp;lt; .001]; MIS cholecystectomy [OR, 0.73; 95% CI, 0.68-0.78; P &amp;lt; .001]; open ventral hernia [OR, 0.51; 95% CI, 0.46-0.57; P &amp;lt; .001]; MIS ventral hernia repair [OR, 0.66; 95% CI, 0.56-0.77; P &amp;lt; .001]). The presence of a residency training program was independently associated with increased odds of outpatient surgery for simple mastectomy (OR, 1.35; 95% CI, 1.16-1.58; P &amp;lt; .001) and mastectomy with reconstruction (OR, 1.50; 95% CI, 1.27-1.77; P &amp;lt; .001) and decreased odds of outpatient surgery for minimally invasive cholecystectomy (OR, 0.96; 95% CI, 0.92-1.00; P = .04), open ventral hernia repair (OR, 0.93; 95% CI, 0.86-1.00; P = .04), and total thyroidectomy (OR, 0.84; 95% CI, 0.71-1.00; P = .04).Conclusions and RelevanceIn this cross-sectional study, significant variation existed in the use of outpatient surgery in the US and appeared to be driven primarily by hospital geography. Addressing these variations may improve the use of resources.

Background. Ex vivo machine perfusion (MP) has transformed organ preservation, offering significant benefits in liver transplantation (LT), particularly with high-risk donor grafts. However, adoption in the United States has been limited. We aimed to examine early adoption trends, surgeon perceptions, and barriers to implementing MP in the United States after Food and Drug Administration approval of MP platforms. Methods. A 23-question electronic survey was distributed to members of the American Society of Transplant Surgeons between October and November 2022, capturing attitudes and practices related to MP adoption. Responses from 96 surgeons representing 77 LT centers across 11 Organ Procurement and Transplantation Network regions were analyzed. Results. Forty-four respondents (48%) reported having an MP program at their institution. Adoption of MP was significantly more common in high-volume centers and those performing ≥20 donation after circulatory death (DCD) transplants annually (P < 0.001). MP utilization received strong support, with 88% endorsing its use for DCD liver allografts and 82% for donation after brain death allografts. Respondents cited MP’s ability to reduce ischemic cholangiopathy, enable graft repair, and facilitate viability assessment as key benefits. Normothermic MP was preferred for high-risk donor profiles, including DCD grafts, older donors, and steatotic livers, and was associated with an increased willingness to accept medically complex grafts compared with static cold storage. Barriers to MP utilization included program costs, personnel demands, and logistical complexities. Centers with higher proportions of privately insured patients were more likely to adopt MP. Despite these challenges, 84% of respondents expressed interest in future MP adoption. Conclusions. MP enhances graft utilization and outcomes, particularly for complex and high-risk donor livers, but widespread US adoption requires addressing financial and logistical barriers. Future efforts should focus on refining cost-effectiveness analyses, collaboration with organ procurement organizations and device companies, and developing standardized training to optimize MP integration and maximize its clinical impact on LT.

ABSTRACTIntroductionVon Willebrand factor (VWF) multimer analysis is essential for diagnosing and classifying von Willebrand disease (VWD) but requires expert interpretation and is subject to inter‐rater variability. We developed an automated image analysis pipeline using deep learning to improve the reproducibility and efficiency of VWF multimer pattern classification.MethodsWe trained a YOLOv8 deep learning model on 514 gel images (6168 labeled instances) to classify VWF multimer patterns into 12 classes. The model was validated on 192 images (2304 instances) and tested on an independent set of 94 images (1128 instances). Images underwent preprocessing, including histogram equalization, contrast enhancement, and gamma correction. Two expert raters provided ground truth classifications.ResultsThe model achieved 91% accuracy compared to Expert 1 (macro‐averaged precision = 0.851, recall = 0.757, F1‐score = 0.786) and 87% accuracy compared to Expert 2 (macro‐averaged precision = 0.653, recall = 0.653, F1‐score = 0.641). Inter‐rater agreement was very high between experts (κ = 0.883), with strong agreement between the model and Expert 1 (κ = 0.845) and good agreement with Expert 2 (κ = 0.773). The model performed exceptionally well on common patterns (F1 > 0.93) but showed lower performance on rare subtypes.ConclusionAutomated VWF multimer analysis using deep learning demonstrates high accuracy in pattern classification and could standardize the interpretation of VWF multimer patterns. While not replacing expert analysis, this approach could improve the efficiency of expert human review, potentially streamlining laboratory workflow and expanding access to VWF multimer testing.