Department of Radiation Oncology

Background

Objective

Methods

Results

Conclusions

PURPOSE: The MIRAGE (Magnetic resonance imaging-guided versus computed tomography-guided stereotactic body radiotherapy for prostate cancer) trial (NCT04384770) showed that aggressive planning target volume (PTV) margin reduction, achieved by leveraging the advanced image-guidance of magnetic resonance imaging guided radiation therapy, led to lower toxicity after stereotactic body radiation therapy (SBRT) to the prostate compared with computed tomography guided SBRT (CTgSBRT) with a standard PTV margin. It is unknown whether the dosimetric benefit of reducing PTV margins alone-independent of the other components of advanced image guided radiation therapy that allowed margin reduction-would be predicted to replicate this decrease in toxicity. METHODS AND MATERIALS: We retrospectively replanned 16 CTgSBRT patients with 3- and 2-mm PTV margins from the 4 mm used in MIRAGE to compare with magnetic resonance imaging guided SBRT (MRgSBRT) patients' radiation therapy plans (n = 79), which used a 2-mm PTV margin. We compared target and organs-at-risk dosimetry and performed multivariable linear regression to control for potentially related covariates. We estimated normal tissue complication probability to assess the potential impact of PTV margin reduction on the risks of rectal and bladder toxicity. RESULTS: Compared with MRgSBRT patients, replanned CTgSBRT patients had similar (3 mm) or statistically significantly improved (2 mm) rectal dosimetry and similar bladder dosimetry (2 mm). These trends persisted on multivariable linear regression. Normal tissue complication probability modeling demonstrated a statistically significant improvement in rectal toxicity whereby margin reduction from 4→3 mm and 4→2 mm resulted in relative reductions in toxicity probabilities of 56% and 78%, respectively. For bladder toxicity, reducing margins from 4 to 3 or 2 mm reduced the probability of bladder toxicity by 96%, although these relative differences were not statistically significantly. CONCLUSIONS: Reducing PTV margins for CTgSBRT plans produced similar to improved organs-at-risk dosimetry compared with the clinical MRgSBRT plans, which in turn would be predicted to reduce toxicity when compared to larger margin CTgSBRT plans. This improvement in dosimetry could explain the reduction in modeled rectal toxicity, but the impact on bladder toxicity is unclear, with large, but nonsignificant relative differences underscoring our incomplete understanding of the drivers of post-SBRT urinary toxicity.

PurposeThe prognosis for pediatric high-grade gliomas associated with mutations in the H3-3A gene is very poor. To investigate whether tumor lysate-pulsed dendritic cells (DC) together with checkpoint blockade might be a potential treatment modality for diffuse hemispheric glioma H3 G34-mutant (DHG), we have developed a novel syngeneic mouse model.MethodsWe used the RCAS/tv-A system to target the expression of H3G34R and PDGFβ and knock out p53 in neural progenitors in C57BL/6 neonatal mice. Three independent cell lines were obtained that expressed transcripts associated with oligodendrocyte and interneuron lineages. Lethal tumor developed following intracranial injection.ResultsTwo cycles of DC vaccination with PD-1 blockade decreased tumor burden and increased survival. In treatment resistant tumors we found higher expression of several genes involved in remodeling the extracellular matrix compared with tumors from untreated animals, suggesting a causal link to resistance to immunotherapy in this tumor model.ConclusionImmunotherapy involving autologous dendritic cells pulsed with tumor lysate and combined with anti-PD-1 antibody might be an effective treatment for DHG. Treatment failure in our tumor model is associated with increased expression of genes implicated in remodeling extracellular matrix in the tumor microenvironment.

Understanding disorders of consciousness (DOC) remains one of the most challenging problems in neuroscience, hindered by the lack of experimental models for probing mechanisms or testing interventions. Here, to address this, we introduce a generative adversarial artificial intelligence (AI) framework that pits deep neural networks—trained to detect consciousness across more than 680,000 ten-second neuroelectrophysiology samples and validated on 565 patients, healthy volunteers and animals—against interpretable, machine learning-driven neural field models. This adversarial architecture produces biologically realistic simulations of both conscious and comatose brains that recapitulate empirical neurophysiological features across humans, monkeys, rats and bats. Without explicit programming, the AI model retrodicts known DOC responses to brain stimulation and generates testable predictions about the mechanisms of unconsciousness. Two such predictions are validated here: selective disruption of the basal ganglia indirect pathway, supported by diffusion magnetic resonance imaging in 51 patients with DOC, and increased cortical inhibitory-to-inhibitory synaptic coupling, supported by RNA sequencing of resected brain tissue from 6 human patients with coma and a rat stroke model. The model also identifies high-frequency stimulation of the subthalamic nucleus as a promising intervention for DOC, supported by electrophysiological data from human patients. This work introduces an AI framework for causal inference and therapeutic discovery in consciousness research, as well as in complex systems more broadly.

BACKGROUND: In 2021, the United States Preventive Services Task Force (USPSTF) lowered the recommended starting age for colorectal cancer (CRC) screening from 50 to 45 for average-risk individuals. However, screening uptake among younger adults has been slow, and little is known about how social factors influence screening behaviors in this early-midlife cohort. This study examined associations between unmet social needs and CRC testing and modality among adults aged 45-49 following the updated USPSTF recommendation. METHODS: This cross-sectional analysis of 2022 Behavioral Risk Factor Surveillance System data included adults aged 45-49 years. Outcomes included any CRC testing since the USPSTF recommendation update and initial modality (stool-based test vs colonoscopy) among tested individuals. Primary predictors included past-year housing, transportation, or food insecurity and total number of unmet social needs. Weighted binary logistic regression models were adjusted for sociodemographic and clinical confounders. RESULTS: Among 13 251 respondents aged 45-49, 22.5% had initiated CRC testing since the USPSTF recommendation update. Food insecurity and increased number of unmet social needs were associated with lower CRC testing uptake in unadjusted models. However, there were no significant differences in CRC testing by social needs status in adjusted models. Among those tested, transportation insecurity was associated with lower use of colonoscopy in adjusted models. CONCLUSIONS: Among younger, newly screening-eligible adults, there were no significant differences in CRC testing uptake after controlling for sociodemographic and clinical covariates. Further investigation is needed to understand barriers and facilitators to CRC screening uptake in this population and guide interventions to reduce early-onset CRC morbidity and mortality.

BackgroundPediatric high-grade gliomas (pHGG) are among the most aggressive childhood brain tumors, with limited treatment options and poor prognosis. Vaccine-based immunotherapy offers a promising strategy by leveraging tumor-specific or associated antigens to stimulate durable anti-tumor immune responses with minimal toxicity.DiscussionThis review outlines the scientific rationale for vaccine therapies in pHGG, detailing key targets such as glioma-associated antigens (EphA2, IL-13Rα2, survivin), driver mutation–derived neoantigens (H3.3K27M, TP53, IDH1), and viral antigens (CMV pp65). We evaluate current vaccine platforms, including peptide vaccines, dendritic cell vaccines, mRNA-based vaccines, and neoantigen-personalized approaches, highlighting early-phase clinical trial results that demonstrate safety and immunogenicity. Despite encouraging preliminary data, several challenges hinder clinical translation, including the distinct immune environment in the central nervous system, intratumoral heterogeneity, low mutational burden, immunosuppressive microenvironments, steroid use, and logistical hurdles in vaccine production and trial design. Future research must address these barriers through optimized antigen selection, combinatorial therapies, novel delivery systems, and pediatric-specific immune profiling.ConclusionWith continued multidisciplinary collaboration, vaccine therapies may emerge as a meaningful addition to the therapeutic arsenal for children with pHGG.

PurposeStereotactic radiosurgery (SRS) offers less neurotoxicity and comparable survival to whole-brain radiation therapy (WBRT) for brain metastases (BM). Current SRS practice patterns are understudied. We examined national trends in SRS and WBRT utilization.MethodsWe queried the National Cancer Database for patients with BM from twelve cancers (≥ 18 years; diagnosed 2004–2020) treated with radiotherapy. Patients were grouped by first-course radiotherapy modality (SRS:1–5 fractions; WBRT:5–15 fractions). Multivariable logistic regression assessed SRS predictors, adjusting for sociodemographic and clinical variables. A race*diagnosis year interaction evaluated temporal trends. Difference-in-differences analysis assessed Medicaid expansion impact.ResultsOf 89,984 patients, 24,174 (27%) received SRS. SRS utilization rose from 8 to 54% between 2004 and 2020 (P < 0.001). SRS was more likely in patients diagnosed in recent years (aOR = 3.85 [95% CI = 3.70–4.01]), who received prior chemotherapy (aOR = 1.17 [1.13–1.21]) or surgery (aOR = 2.25 [2.11–2.40]), and those with colorectal (aOR = 1.93 [1.64–2.26]), lung (aOR = 1.37 [1.24–1.50]), melanoma (aOR = 2.76 [2.46–3.10]), thyroid (aOR = 2.17 [1.36–3.46]), or kidney/bladder cancer (aOR = 2.76 [2.44–3.12]) versus breast cancer. SRS was less likely in patients with lower income (aOR = 0.88 [0.85–0.92]) or educational attainment (aOR = 0.88 [0.85–0.92]), Medicaid/Medicare (aOR = 0.86 [0.83–0.90]), no insurance (aOR = 0.49 [0.44–0.53]), or treatment at community (aOR = 0.31 [0.29–0.34]), comprehensive community (aOR = 0.56 [0.54–0.58]), or integrated facilities (aOR = 0.77 [0.73–0.80]). Race and ethnicity were not overall associated with SRS use. Medicaid expansion had no impact (aOR = 0.99 [0.88–1.10]).ConclusionsSRS utilization increased between 2004 and 2020. However, disparities persist for patients with lower socioeconomic status, uninsured/public insurance, or non-academic center treatment, potentially reflecting access disparities or differences in disease burden. Targeted efforts are needed to ensure equitable access to advanced cancer therapies, particularly in the context of potential additive effects of disparities in disease burden.

Purpose. In radiotherapy, dose distribution conformity and compactness are critical to patient outcomes. Advanced techniques like4πradiotherapy leverage non-coplanar beams for superior dosimetry by exploring additional degrees of freedom. However,4πplanning is computationally intensive due to large dose-loading matrices for candidate beams. This work presents an ultra-high performance parallel (UHPP) framework to accelerate high-dimensional treatment planning.Methods. For dose calculation, we developed: (1) a two-step total energy released per unit mass (TERMA) computation module calculating the TERMA array once per beam, enabling reuse across convolution directions; (2) a synchronized dose calculation module based on collapsed-cone convolution superposition (CCCS), arranging rays in dedicated sequences to preserve thread efficiency and minimize memory access; (3) a scattering-based coordinate transformation mapping dose from beamlet to patient Cartesian coordinates, eliminating aliasing without atomic operations. The framework includes CCCS exponential kernel calculation for varying LINAC spectra. For beam orientation optimization, we employed fast iterative shrinkage-thresholding algorithm with group sparsity regularization, accelerated using cuSPARSE library on GPUs. We benchmarked against Monte Carlo (MC) simulations for dose accuracy and compared computational performance to state-of-the-art (SOTA) methods. Plan quality was evaluated across four approaches: UHPP, SOTA, clinical VMAT plans, and MC calculations based on UHPP plans.Results. Compared to MC simulations, UHPP achieved minimum 98% gamma passing rates under 1.5%/1.5 mm criterion for water and slab phantoms, and average 97.35% and 92.18% under 3%/3 mm criterion for pancreas and head-and-neck patients, respectively. UHPP delivered 8.86× and 6.99× speedups in dose calculation and plan optimization while maintaining comparable or superior plan quality. Both UHPP and SOTA consistently produced4πplans outperforming clinical VMAT plans in organ-at-risk sparing and target coverage.Conclusion. The UHPP framework delivers high dose accuracy and substantial computational speedup without sacrificing4πplanning's dosimetric advantages, supporting practical adoption of advanced4πradiotherapy in clinical workflows.

Tomography recovers internal volume from projection measurements. Formulated as inverse problems, classic computed tomography generally reconstructs attenuation property in a preset cartesian grid coordinate. While this is intuitive and convenient for digital display, such discretization leads to forward-backward projection inconsistency, and discrepancy between digital and effective resolution. We take a different perspective by considering the image volume as continuous and modelling forward projection as a hybrid continuous-to-discrete mapping from volume to detector elements, which we call "ray bundles". The ray bundle can be regarded as an unconventional heterogenous coordinate. Projections are modeled as line integrations along ray bundles in the continuous volume space and approximated by numerical integration using customized sample points. This modeling approach is conveniently supported with an implicit neural representation approach. By representing the volume as a function mapping spatial coordinates to attenuation properties and leveraging ray bundle projection, this approach reflects transmission physics and eliminates the need for explicit interpolation, intersection calculations, or matrix inversions. A novel sampling strategy is further developed to adaptively distribute points along the ray bundles, emphasizing high gradient regions to allocate computational resources to heterogenous structures and details. We call this system T-ReX to indicate Transmission Ray bundles for X-ray geometry. We validate T-ReX through comprehensive experiments across three scenarios: simulated full-fan projections with primary signal only, half-fan setups with simulated scatter and noise, and an in-house dataset with realistic acquisition conditions. These results highlight the effectiveness of T-ReX in sparse view X-ray tomography.

Chordomas remain one of the most difficult-to-treat of skull base tumors. The best chance of survival and cure to date is with en bloc radical surgical resection, followed by adjuvant radiotherapy. Oftentimes, tumor infiltration into functionally critical, or difficult-to-access areas, precludes optimal resection. The median overall survival for these tumors is 116 months. Cold atmospheric plasma (CAP) is generated by applying a high voltage electric field to helium or argon feed gases, resulting in reactive atmospheric species. Over the past decade, CAP has been applied experimentally in a number of oncologic conditions and has demonstrated anti-tumor effects both in vitro and in vivo. Additionally, CAP has been shown to increase cancer cells’ sensitivity to radiation and could therefore be a useful tool in improving chordoma recurrence rates as an intraoperative adjuvant therapy to the current standard of care. To date, there are no studies in the literature examining the efficacy of CAP in inducing cytotoxicity in chordoma. We treated CH2, CH7 and UM-Chor1 chordoma cells with CAP, measuring resulting cell viability and intracellular ROS accumulation. Here, we show a dose-dependent increase in intracellular ROS and cell death with direct CAP exposure in vitro, finding an exquisite sensitivity of chordoma cells to CAP-mediated cytotoxicity.