Validation of a methylation-based signature for subventricular zone involvement in glioblastoma.

PURPOSE: Glioblastomas (GBM) with subventricular zone (SVZ) contact have previously been associated with a specific epigenetic fingerprint. We aim to validate a reported bulk methylation signature to determine SVZ contact. METHODS: Methylation array analysis was performed on IDHwt GBM patients treated at our institution. The v11b4 classifier was used to ensure the inclusion of only receptor tyrosine kinase (RTK) I, II, and mesenchymal (MES) subtypes. Methylation-based assignment (SVZM ±) was performed using hierarchical cluster analysis. Magnetic resonance imaging (MRI) (T1ce) was independently reviewed for SVZ contact by three experienced readers. RESULTS: Sixty-five of 70 samples were classified as RTK I, II, and MES. Full T1ce MRI-based rater consensus was observed in 54 cases, which were retained for further analysis. Epigenetic SVZM classification and SVZ were strongly associated (OR: 15.0, p = 0.003). Thirteen of fourteen differential CpGs were located in the previously described differentially methylated LRBA/MAB21L2 locus. SVZ + tumors were linked to shorter OS (hazard ratio (HR): 3.80, p = 0.02) than SVZM + at earlier time points (time-dependency of SVZM, p < 0.05). Considering the SVZ consensus as the ground truth, SVZM classification yields a sensitivity of 96.6%, specificity of 36.0%, positive predictive value (PPV) of 63.6%, and negative predictive value (NPV) of 90.0%. CONCLUSION: Herein, we validated the specific epigenetic signature in GBM in the vicinity of the SVZ and highlighted the importance of methylation of a part of the LRBA/MAB21L2 gene locus. Whether SVZM can replace MRI-based SVZ assignment as a prognostic and diagnostic tool will require prospective studies of large, homogeneous cohorts.

DNA-methylome-assisted classification of patients with poor prognostic subventricular zone associated IDH-wildtype glioblastoma.

Glioblastoma (GBM) derived from the "stem cell" rich subventricular zone (SVZ) may constitute a therapy-refractory subgroup of tumors associated with poor prognosis. Risk stratification for these cases is necessary but is curtailed by error prone imaging-based evaluation. Therefore, we aimed to establish a robust DNA methylome-based classification of SVZ GBM and subsequently decipher underlying molecular characteristics. MRI assessment of SVZ association was performed in a retrospective training set of IDH-wildtype GBM patients (n = 54) uniformly treated with postoperative chemoradiotherapy. DNA isolated from FFPE samples was subject to methylome and copy number variation (CNV) analysis using Illumina Platform and cnAnalysis450k package. Deep next-generation sequencing (NGS) of a panel of 130 GBM-related genes was conducted (Agilent SureSelect/Illumina). Methylome, transcriptome, CNV, MRI, and mutational profiles of SVZ GBM were further evaluated in a confirmatory cohort of 132 patients (TCGA/TCIA). A 15 CpG SVZ methylation signature (SVZM) was discovered based on clustering and random forest analysis. One third of CpG in the SVZM were associated with MAB21L2/LRBA. There was a 14.8% (n = 8) discordance between SVZM vs. MRI classification. Re-analysis of these patients favored SVZM classification with a hazard ratio (HR) for OS of 2.48 [95% CI 1.35-4.58], p = 0.004 vs. 1.83 [1.0-3.35], p = 0.049 for MRI classification. In the validation cohort, consensus MRI based assignment was achieved in 62% of patients with an intraclass correlation (ICC) of 0.51 and non-significant HR for OS (2.03 [0.81-5.09], p = 0.133). In contrast, SVZM identified two prognostically distinct subgroups (HR 3.08 [1.24-7.66], p = 0.016). CNV alterations revealed loss of chromosome 10 in SVZM- and gains on chromosome 19 in SVZM- tumors. SVZM- tumors were also enriched for differentially mutated genes (p < 0.001). In summary, SVZM classification provides a novel means for stratifying GBM patients with poor prognosis and deciphering molecular mechanisms governing aggressive tumor phenotypes.

Prognostic Value of microRNA-221/2 and 17-92 Families in Primary Glioblastoma Patients Treated with Postoperative Radiotherapy.

MicroRNAs (miRs) are non-coding master regulators of transcriptome that could act as tumor suppressors (TSs) or oncogenes (oncomiRs). We aimed to systematically investigate the relevance of miRs as prognostic biomarkers in primary glioblastoma multiforme (GBM) treated with postoperative radio(chemo)therapy (PORT). For hypothesis generation, tumor miR expression by Agilent 8x15K human microRNA microarrays and survival data from 482 GBM patients of The Cancer Genome Atlas (TCGA cohort) were analyzed using Cox-PH models. Expression of candidate miRs with prognostic relevance (miR-221/222; miR-17-5p, miR-18a, miR-19b) was validated by qRT-PCR using Taqman technology on an independent validation cohort of GBM patients (n = 109) treated at Heidelberg University Hospital (HD cohort). In TCGA, 50 miRs showed significant association with survival. Among the top ranked prognostic miRs were members of the two miR families miR-221/222 and miR-17-92. Loss of miR-221/222 was correlated with improved prognosis in both cohorts (TCGA, HD) and was an independent prognostic marker in a multivariate analysis considering demographic characteristics (age, sex, Karnofsky performance index (KPI)), molecular markers (O-6-methylguanine-DNA methyltransferase (MGMT) methylation, IDH mutation status) and PORT as co-variables. The prognostic value of miR-17-92 family members was ambiguous and in part contradictory by direct comparison of the two cohorts, thus warranting further validation in larger prospective trials.

An R package for an integrated evaluation of statistical approaches to cancer incidence projection.

BACKGROUND: Projection of future cancer incidence is an important task in cancer epidemiology. The results are of interest also for biomedical research and public health policy. Age-Period-Cohort (APC) models, usually based on long-term cancer registry data (> 20 yrs), are established for such projections. In many countries (including Germany), however, nationwide long-term data are not yet available. General guidance on statistical approaches for projections using rather short-term data is challenging and software to enable researchers to easily compare approaches is lacking. METHODS: To enable a comparative analysis of the performance of statistical approaches to cancer incidence projection, we developed an R package (incAnalysis), supporting in particular Bayesian models fitted by Integrated Nested Laplace Approximations (INLA). Its use is demonstrated by an extensive empirical evaluation of operating characteristics (bias, coverage and precision) of potentially applicable models differing by complexity. Observed long-term data from three cancer registries (SEER-9, NORDCAN, Saarland) was used for benchmarking. RESULTS: Overall, coverage was high (mostly > 90%) for Bayesian APC models (BAPC), whereas less complex models showed differences in coverage dependent on projection-period. Intercept-only models yielded values below 20% for coverage. Bias increased and precision decreased for longer projection periods (> 15 years) for all except intercept-only models. Precision was lowest for complex models such as BAPC models, generalized additive models with multivariate smoothers and generalized linear models with age x period interaction effects. CONCLUSION: The incAnalysis R package allows a straightforward comparison of cancer incidence rate projection approaches. Further detailed and targeted investigations into model performance in addition to the presented empirical results are recommended to derive guidance on appropriate statistical projection methods in a given setting.

cnAnalysis450k: an R package for comparative analysis of 450k/EPIC Illumina methylation array derived copy number data.

MOTIVATION: Detailed copy number (CN) variation data can be obtained from 450k or EPIC Illumina methylation assays. However, the effects of different preprocessing strategies (normalization, transformation and selection of gain/loss cutoff values) on variant calling have not been evaluated systematically. RESULTS: We provide an R package which allows to directly compare any preprocessed CN data. It provides its own CN alteration detection methodology: segments are identified through detection of changes in variance of CN data and are subsequently filtered for significance. Meaningful cutoffs for gain/loss definition can be identified automatically through analysis of the resulting ΔCN distributions of all analyzed samples. Three exemplary datasets (2x450k, 1xEPIC) were selected for comparative analyses of Raw, Illumina, SWAN, Quantile, Noob, Funnorm and Dasen normalizations. Importantly, all CN data distributions were skewed (-0.66 to -1.2) therefore requiring different gain/loss cutoffs. Depending on the normalization method, prominent baseline differences between samples could be observed. We present a workflow, which alleviates both issues: Z-transformation removes baseline differences between samples, and automatic cutoff selection circumvents the problems accompanying the skewed distributions. Additional filtering of candidates by significance yields comparable results for most enumerated normalization methods except for SWAN. In contrast, manual cutoff determination results in highly variable numbers of variant calls, highly dependent on the selected normalization method. Taken together, we present a workflow which allows to robustly identify copy number alterations in methylation array data fairly independent of the applied normalization. AVAILABILITY AND IMPLEMENTATION: The cnAnalysis450k package is available on github ( https://github.com/mknoll/cnAnalysis450k ). CONTACT: m.knoll@dkfz.de. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Sow serenity: automatic long-term measurement of lying behavior in crates and free-farrowing pens using 3D accelerometers.

Accelerometers are useful in analyzing lying behavior in farm animals. The effect of the farrowing system on sow lying behavior has been studied around parturition, but not long-term. In a natural environment, sows increase activity 14 d post parturition, which we expected to be also evident in housed sows when they can move freely. The objective of this study was (1) to validate the methodology to automatically measure sow lying bouts and duration with accelerometers and (2) to apply it to crated and free-farrowing sows 24-h pre-parturition until weaning. We used videos with manual behavior coding as the gold standard for validation and calculated the agreement with an intraclass correlation coefficient (ICC), which was 0.30 (95% CI: -0.10 to 0.64) for the number of lying bouts. When transitional sitting bouts were excluded from the video dataset, the ICC for lying bouts increased to 0.86 (95% CI: 0.40 to 0.95). For lying duration, the ICC was 0.93 (95% CI: 0.26 to 0.98). We evaluated the effects of housing, day relative to parturition, and time of day on lying using the accelerometer data and linear mixed models. In crated sows, the number of lying bouts increased toward parturition, peaking at about five bouts per 6 h, and decreased to almost zero bouts after parturition. Then, it increased again (P = 0.001). In free-farrowing sows, the number of lying bouts gradually decreased from a high level towards parturition and was lowest after parturition. It remained constant, as in the crated sows, until day 15, when the number of bouts increased to eight bouts on day 20 (P = 0.001). Sows in both systems were lying almost all of the time between 18:00 and 00:00 hours and on all days (P = 0.001). The crated sows showed a very similar pattern in the other three-quarters of the day with a reduced lying time before parturition, a peak after parturition, reduced lying time from days 5 to 20, and an increase again towards weaning (P = 0.001). Free-farrowing sows had a similar pattern to the crated sows from 00:00 to 06:00 hours, but without the reduction in lying time from days 5 to 20. They showed an increase in lying time toward parturition, which remained constant with a final decrease toward weaning, especially during the day (P = 0.001). This study proves the accuracy of accelerometer-based sow lying behavior classification and shows that free-farrowing systems benefit lactating sows around parturition but also towards weaning in the nest-leaving phase by facilitating activity.

We analyzed lying behavior of sows using sensors, focusing on crated versus free-farrowing sows from pre-parturition to weaning. Lying behavior varies in this time following the needs of the sow and her litter. In a natural environment, sows increase activity 14 d post parturition, which we expected to be also evident in housed sows when they are allowed to move freely. Validation with video data showed excellent agreement for duration and frequency of lying. In crated sows, the number of lying bouts peaked around parturition, decreased after parturition, and then gradually increased. In free-farrowing sows, lying down occurred less often before parturition, but increased by day 20 compared to crated sows. Both housing systems showed prolonged lying periods from 18:00 to 00:00 hours. Crated sows had reduced lying times before parturition and lied longest post-parturition, which decreased until day 5 and then increased toward weaning. Free-farrowing sows had similar nocturnal patterns but persistent lying times that increased prior to parturition and decreased prior to weaning. Overall, the study highlighted the accuracy of accelerometer-based lying behavior classification and showed that free-farrowing systems benefit lactating sows not only around parturition but also toward weaning, facilitating activity during the nest-leaving phase.

Clinical Workflow of Cone Beam Computer Tomography-Based Daily Online Adaptive Radiotherapy with Offline Magnetic Resonance Guidance: The Modular Adaptive Radiotherapy System (MARS).

PURPOSE: The Ethos (Varian Medical Systems) radiotherapy device combines semi-automated anatomy detection and plan generation for cone beam computer tomography (CBCT)-based daily online adaptive radiotherapy (oART). However, CBCT offers less soft tissue contrast than magnetic resonance imaging (MRI). This work aims to present the clinical workflow of CBCT-based oART with shuttle-based offline MR guidance. METHODS: From February to November 2023, 31 patients underwent radiotherapy on the Ethos (Varian, Palo Alto, CA, USA) system with machine learning (ML)-supported daily oART. Moreover, patients received weekly MRI in treatment position, which was utilized for daily plan adaptation, via a shuttle-based system. Initial and adapted treatment plans were generated using the Ethos treatment planning system. Patient clinical data, fractional session times (MRI + shuttle transport + positioning, adaptation, QA, RT delivery) and plan selection were assessed for all fractions in all patients. RESULTS: In total, 737 oART fractions were applied and 118 MRIs for offline MR guidance were acquired. Primary sites of tumors were prostate (n = 16), lung (n = 7), cervix (n = 5), bladder (n = 1) and endometrium (n = 2). The treatment was completed in all patients. The median MRI acquisition time including shuttle transport and positioning to initiation of the Ethos adaptive session was 53.6 min (IQR 46.5-63.4). The median total treatment time without MRI was 30.7 min (IQR 24.7-39.2). Separately, median adaptation, plan QA and RT times were 24.3 min (IQR 18.6-32.2), 0.4 min (IQR 0.3-1,0) and 5.3 min (IQR 4.5-6.7), respectively. The adapted plan was chosen over the scheduled plan in 97.7% of cases. CONCLUSION: This study describes the first workflow to date of a CBCT-based oART combined with a shuttle-based offline approach for MR guidance. The oART duration times reported resemble the range shown by previous publications for first clinical experiences with the Ethos system.

Daily AI-Based Treatment Adaptation under Weekly Offline MR Guidance in Chemoradiotherapy for Cervical Cancer 1: The AIM-C1 Trial.

(1) Background: External beam radiotherapy (EBRT) and concurrent chemotherapy, followed by brachytherapy (BT), offer a standard of care for patients with locally advanced cervical carcinoma. Conventionally, large safety margins are required to compensate for organ movement, potentially increasing toxicity. Lately, daily high-quality cone beam CT (CBCT)-guided adaptive radiotherapy, aided by artificial intelligence (AI), became clinically available. Thus, online treatment plans can be adapted to the current position of the tumor and the adjacent organs at risk (OAR), while the patient is lying on the treatment couch. We sought to evaluate the potential of this new technology, including a weekly shuttle-based 3T-MRI scan in various treatment positions for tumor evaluation and for decreasing treatment-related side effects. (2) Methods: This is a prospective one-armed phase-II trial consisting of 40 patients with cervical carcinoma (FIGO IB-IIIC1) with an age ≥ 18 years and a Karnofsky performance score ≥ 70%. EBRT (45-50.4 Gy in 25-28 fractions with 55.0-58.8 Gy simultaneous integrated boosts to lymph node metastases) will be accompanied by weekly shuttle-based MRIs. Concurrent platinum-based chemotherapy will be given, followed by 28 Gy of BT (four fractions). The primary endpoint will be the occurrence of overall early bowel and bladder toxicity CTCAE grade 2 or higher (CTCAE v5.0). Secondary outcomes include clinical feasibility, quality of life, and imaging-based response assessment.

MR Intensity Normalization Methods Impact Sequence Specific Radiomics Prognostic Model Performance in Primary and Recurrent High-Grade Glioma.

PURPOSE: This study investigates the impact of different intensity normalization (IN) methods on the overall survival (OS) radiomics models' performance of MR sequences in primary (pHGG) and recurrent high-grade glioma (rHGG). METHODS: MR scans acquired before radiotherapy were retrieved from two independent cohorts (rHGG C1: 197, pHGG C2: 141) from multiple scanners (15, 14). The sequences are T1 weighted (w), contrast-enhanced T1w (T1wce), T2w, and T2w-FLAIR. Sequence-specific significant features (SF) associated with OS, extracted from the tumour volume, were derived after applying 15 different IN methods. Survival analyses were conducted using Cox proportional hazard (CPH) and Poisson regression (POI) models. A ranking score was assigned based on the 10-fold cross-validated (CV) concordance index (C-I), mean square error (MSE), and the Akaike information criterion (AICs), to evaluate the methods' performance. RESULTS: Scatter plots of the 10-CV C-I and MSE against the AIC showed an impact on the survival predictions between the IN methods and MR sequences (C1/C2 C-I range: 0.62-0.71/0.61-0.72, MSE range: 0.20-0.42/0.13-0.22). White stripe showed stable results for T1wce (C1/C2 C-I: 0.71/0.65, MSE: 0.21/0.14). Combat (0.68/0.62, 0.22/0.15) and histogram matching (HM, 0.67/0.64, 0.22/0.15) showed consistent prediction results for T2w models. They were also the top-performing methods for T1w in C2 (Combat: 0.67, 0.13; HM: 0.67, 0.13); however, only HM achieved high predictions in C1 (0.66, 0.22). After eliminating IN impacted SF using Spearman's rank-order correlation coefficient, a mean decrease in the C-I and MSE of 0.05 and 0.03 was observed in all four sequences. CONCLUSION: The IN method impacted the predictive power of survival models; thus, performance is sequence-dependent.

MR-Class: A Python Tool for Brain MR Image Classification Utilizing One-vs-All DCNNs to Deal with the Open-Set Recognition Problem.

BACKGROUND: MR image classification in datasets collected from multiple sources is complicated by inconsistent and missing DICOM metadata. Therefore, we aimed to establish a method for the efficient automatic classification of MR brain sequences. METHODS: Deep convolutional neural networks (DCNN) were trained as one-vs-all classifiers to differentiate between six classes: T1 weighted (w), contrast-enhanced T1w, T2w, T2w-FLAIR, ADC, and SWI. Each classifier yields a probability, allowing threshold-based and relative probability assignment while excluding images with low probability (label: unknown, open-set recognition problem). Data from three high-grade glioma (HGG) cohorts was assessed; C1 (320 patients, 20,101 MRI images) was used for training, while C2 (197, 11,333) and C3 (256, 3522) were for testing. Two raters manually checked images through an interactive labeling tool. Finally, MR-Class' added value was evaluated via radiomics model performance for progression-free survival (PFS) prediction in C2, utilizing the concordance index (C-I). RESULTS: Approximately 10% of annotation errors were observed in each cohort between the DICOM series descriptions and the derived labels. MR-Class accuracy was 96.7% [95% Cl: 95.8, 97.3] for C2 and 94.4% [93.6, 96.1] for C3. A total of 620 images were misclassified; manual assessment of those frequently showed motion artifacts or alterations of anatomy by large tumors. Implementation of MR-Class increased the PFS model C-I by 14.6% on average, compared to a model trained without MR-Class. CONCLUSIONS: We provide a DCNN-based method for the sequence classification of brain MR images and demonstrate its usability in two independent HGG datasets.

Carbon-ion radiotherapy (CIRT) as treatment of pancreatic cancer at HIT: initial radiation plan analysis of the prospective phase II PACK-study.

PURPOSE: To analyze the dose objectives and constraints applied at the prospective phase II PACK-study at Heidelberg ion therapy center (HIT) for different radiobiological models. METHODS: Treatment plans of 14 patients from the PACK-study were analyzed and recomputed in terms of physical, biological dose and dose-averaged linear energy transfer (LETd). Both LEM-I (local effect model 1) and the adapted NIRS-MKM (microdosimetric kinetic model), were used for relative biological effectiveness (RBE)-weighted dose calculations (DBio|HIT and DBio|NIRS). A new constraint to the gastrointestinal (GI) tract was derived from the National Institute of Radiological Science (NIRS) clinical experience and considered for plan reoptimization (DBio|NIRS-const_48Gy and DBio|NIRS-const_50.4Gy). The Lyman-Kutcher-Burman (LKB) model of Normal Tissue Complication Probability (NTCP) for GI toxicity endpoints was computed. Furthermore, the computed LETd distribution was evaluated and correlated with Local Control (LC). RESULTS: Only two patients showed a LETd98% in the GTV greater than 44 keV/µm. A HIT-dose constraint to the GI of [Formula: see text] was derived from the NIRS experience, in alternative to the standard at HIT Dmax = 45.6 GyRBEHIT. In comparison with the original DBio|HIT,DBio|NIRS-const_48GyandDBio|NIRS-const_50.4Gy resulted in an increase in the ITV's D98% of 8.7% and 11.3%. The NTCP calculation resulted in a probability for gastrointestinal bleeding of 4.5%, 12.3% and 13.0%, for DBio|NIRS, DBio|NIRS-const_48Gy and DBio|NIRS-const_50.4Gy, respectively. CONCLUSION: The results indicate that the current standards applied at HIT for CIRT closely align with the Japanese experience. However, to enhance tumor coverage, a more relaxed constraint on the GI tract may be considered. As the PACK-trial progresses, further analyses of various clinical endpoints are anticipated.

Effective Reprogramming of Patient-Derived M2-Polarized Glioblastoma-Associated Microglia/Macrophages by Treatment with GW2580.

PURPOSE: Targeting immunosuppressive and pro-tumorigenic glioblastoma (GBM)-associated macrophages and microglial cells (GAM) has great potential to improve patient outcomes. Colony-stimulating factor-1 receptor (CSF1R) has emerged as a promising target for reprograming anti-inflammatory M2-like GAMs. However, treatment data on patient-derived, tumor-educated GAMs and their influence on the adaptive immunity are lacking. EXPERIMENTAL DESIGN: CD11b+-GAMs freshly isolated from patient tumors were treated with CSF1R-targeting drugs PLX3397, BLZ945, and GW2580. Phenotypical changes upon treatment were assessed using RNA sequencing, flow cytometry, and cytokine quantification. Functional analyses included inducible nitric oxide synthase activity, phagocytosis, transmigration, and autologous tumor cell killing assays. Antitumor effects and changes in GAM activation were confirmed in a complex patient-derived 3D tumor organoid model serving as a tumor avatar. RESULTS: The most effective reprogramming of GAMs was observed upon GW2580 treatment, which led to the downregulation of M2-related markers, IL6, IL10, ERK1/2, and MAPK signaling pathways, while M1-like markers, gene set enrichment indicating activated MHC-II presentation, phagocytosis, and T-cell killing were substantially increased. Moreover, treatment of patient-derived GBM organoids with GW2580 confirmed successful reprogramming, resulting in impaired tumor cell proliferation. In line with its failure in clinical trials, PLX3397 was ineffective in our analysis. CONCLUSIONS: This comparative analysis of CSF1R-targeting drugs on patient-derived GAMs and human GBM avatars identified GW2580 as the most powerful inhibitor with the ability to polarize immunosuppressive GAMs to a proinflammatory phenotype, supporting antitumor T-cell responses while also exerting a direct antitumor effect. These data indicate that GW2580 could be an important pillar in future therapies for GBM.

DNA-Methylome-Based Tumor Hypoxia Classifier Identifies HPV-Negative Head and Neck Cancer Patients at Risk for Locoregional Recurrence after Primary Radiochemotherapy.

PURPOSE: Tumor hypoxia is a paradigmatic negative prognosticator of treatment resistance in head and neck squamous cell carcinoma (HNSCC). The lack of robust and reliable hypoxia classifiers limits the adaptation of stratified therapies. We hypothesized that the tumor DNA methylation landscape might indicate epigenetic reprogramming induced by chronic intratumoral hypoxia. EXPERIMENTAL DESIGN: A DNA-methylome-based tumor hypoxia classifier (Hypoxia-M) was trained in the TCGA (The Cancer Genome Atlas)-HNSCC cohort based on matched assignments using gene expression-based signatures of hypoxia (Hypoxia-GES). Hypoxia-M was validated in a multicenter DKTK-ROG trial consisting of human papillomavirus (HPV)-negative patients with HNSCC treated with primary radiochemotherapy (RCHT). RESULTS: Although hypoxia-GES failed to stratify patients in the DKTK-ROG, Hypoxia-M was independently prognostic for local recurrence (HR, 4.3; P = 0.001) and overall survival (HR, 2.34; P = 0.03) but not distant metastasis after RCHT in both cohorts. Hypoxia-M status was inversely associated with CD8 T-cell infiltration in both cohorts. Hypoxia-M was further prognostic in the TCGA-PanCancer cohort (HR, 1.83; P = 0.04), underscoring the breadth of this classifier for predicting tumor hypoxia status. CONCLUSIONS: Our findings highlight an unexplored avenue for DNA methylation-based classifiers as biomarkers of tumoral hypoxia for identifying high-risk features in patients with HNSCC tumors. See related commentary by Heft Neal and Brenner, p. 2954.

Neuroprotective Effects of Ultra-High Dose Rate FLASH Bragg Peak Proton Irradiation.

PURPOSE: To investigate brain tissue response to ultra-high dose rate (uHDR, FLASH) and standard dose rate (SDR) proton irradiations in the Bragg peak region. METHODS AND MATERIALS: Active scanning uHDR delivery was established for proton beams for investigation of dose rate effects between clinical SDR and uHDR at ∼10 Gy in the Bragg peak region (dose-averaged linear energy transfer [LETD] ranging from 4.5 to 10.2 keV µm-1 ). Radiation- induced injury of neuronal tissue was assessed by studying the DNA double strand break repair kinetics surrogated by nuclear γH2AX staining (radiation induced foci [RIF]), microvascular density and structural integrity (MVD, CD31+ endothelium), and inflammatory microenvironmental response (CD68+ microglia/macrophages and high mobility group box protein 1[HMGB]) in healthy C57BL/6 mouse brains. RESULTS: Averaged dose rates achieved were 0.17 Gy/s (SDR) and 120 Gy/s (uHDR). The fraction of RIF-positive cells increased after SDR ∼10-fold, whereas a significantly lower fraction of RIF-positive cells was found after uHDR versus SDR (∼2 fold, P < .0001). Moreover, uHDR substantially preserved the microvascular architecture and reduced microglia/macrophage regulated associated inflammation as compared with SDR. CONCLUSIONS: The feasibility of uHDR raster scanning proton irradiation is demonstrated to elicit FLASH sparing neuroprotective effects compared to SDR in a preclinical in vivo model.

Whole Blood Transcriptional Fingerprints of High-Grade Glioma and Longitudinal Tumor Evolution under Carbon Ion Radiotherapy.

PURPOSE: To assess the value of whole blood transcriptome data from liquid biopsy (lbx) in recurrent high-grade glioma (rHGG) patients for longitudinal molecular monitoring of tumor evolution under carbon ion irradiation (CIR). METHODS: Whole blood transcriptome (WBT) analysis (Illumina HumanHT-12 Expression BeadChips) was performed in 14 patients with rHGG pre re-irradiation (reRT) with CIR and 3, 6 and 9 weeks post-CIR (reRT grade III:5, 36%, IV:9, 64%). Patients were irradiated with 30, 33, 36 GyRBE (n = 5, 6, 3) in 3GyRBE per fraction. RESULTS: WTB analysis showed stable correlation with treatment characteristics and patients tumor grade, indicating a preserved tumor origin specific as well as dynamic transcriptional fingerprints of peripheral blood cells. Initial histopathologic tumor grade was indirectly associated with TMEM173 (STING), DNA-repair (ATM, POLD4) and hypoxia related genes. DNA-repair, chromatin remodeling (LIG1, SMARCD1) and immune response (FLT3LG) pathways were affected post-CIR. Longitudinal WTB fingerprints identified two distinct trajectories of rHGG evolution, characterized by differential and prognostic CRISPLD2 expression pre-CIR. CONCLUSIONS: Lbx based WTB analysis holds the potential for molecular stratification of rHGG patients and therapy monitoring. We demonstrate the feasibility of the peripheral blood transcriptome as a sentinel organ for identification of patient, tumor characteristics and CIR specific fingerprints in rHGG.

Deep Learning-based Automatic Lung Segmentation on Multiresolution CT Scans from Healthy and Fibrotic Lungs in Mice.

Purpose: To develop a model to accurately segment mouse lungs with varying levels of fibrosis and investigate its applicability to mouse images with different resolutions. Materials and Methods: In this experimental retrospective study, a U-Net was trained to automatically segment lungs on mouse CT images. The model was trained (n = 1200), validated (n = 300), and tested (n = 154) on longitudinally acquired and semiautomatically segmented CT images, which included both healthy and irradiated mice (group A). A second independent group of 237 mice (group B) was used for external testing. The Dice score coefficient (DSC) and Hausdorff distance (HD) were used as metrics to quantify segmentation accuracy. Transfer learning was applied to adapt the model to high-spatial-resolution mouse micro-CT segmentation (n = 20; group C [n = 16 for training and n = 4 for testing]). Results: The trained model yielded a high median DSC in both test datasets: 0.984 (interquartile range [IQR], 0.977-0.988) in group A and 0.966 (IQR, 0.955-0.972) in group B. The median HD in both test datasets was 0.47 mm (IQR, 0-0.51 mm [group A]) and 0.31 mm (IQR, 0.30-0.32 mm [group B]). Spatially resolved quantification of differences toward reference masks revealed two hot spots close to the air-tissue interfaces, which are particularly prone to deviation. Finally, for the higher-resolution mouse CT images, the median DSC was 0.905 (IQR, 0.902-0.929) and the median 95th percentile of the HD was 0.33 mm (IQR, 2.61-2.78 mm). Conclusion: The developed deep learning-based method for mouse lung segmentation performed well independently of disease state (healthy, fibrotic, emphysematous lungs) and CT resolution.Keywords: Deep Learning, Lung Fibrosis, Radiation Therapy, Segmentation, Animal Studies, CT, Thorax, Lung Supplemental material is available for this article. Published under a CC BY 4.0 license.

Radioresistance and Transcriptional Reprograming of Invasive Glioblastoma Cells.

PURPOSE: Infiltrative growth pattern is a hallmark of glioblastoma (GBM). Radiation therapy aims to eradicate microscopic residual GBM cells after surgical removal of the visible tumor bulk. However, in-field recurrences remain the major pattern of therapy failure. We hypothesized that the radiosensitivity of peripheral invasive tumor cells (peri) may differ from the predominantly investigated tumor bulk. METHODS AND MATERIALS: Invasive GBM populations were generated via debulking of the visible tumor core and serial orthotopic transplantation of peri cells, and sustained proinvasive phenotype of peri cells was confirmed in vitro by scratch assay and time lapse imaging. In parallel, invasive GBM cells were selected by transwell assay and from peri cells of patient-derived 3-dimensional spheroid cultures. Transcriptome analysis deciphered a GBM invasion-associated gene signature, and functional involvement of key pathways was validated by pharmacologic inhibition. RESULTS: Compared with the bulk cells, invasive GBM populations acquired a radioresistant phenotype characterized by increased cell survival, reduced cell apoptosis, and enhanced DNA double-strand break repair proficiency. Transcriptome analysis revealed a reprograming of invasive cells toward augmented activation of epidermal growth factor receptor- and nuclear factor-κB-related pathways, whereas metabolic processes were downregulated. An invasive GBM score derived from this transcriptional fingerprint correlated well with patient outcome. Inhibition of epidermal growth factor receptor and nuclear factor-κB signaling resensitized invasive cells to irradiation. Invasive cells were eradicated with similar efficacy by particle therapy with carbon ions. CONCLUSIONS: Our data indicate that invasive tumor cells constitute a phenotypically distinct and highly radioresistant GBM subpopulation with prognostic impact that may be vulnerable to targeted therapy and carbon ions.

The impact of tumor metabolic activity assessed by 18F-FET amino acid PET imaging in particle radiotherapy of high-grade glioma patients.

Background: Selective uptake of (18)F-fluoro-ethyl-tyrosine (18F-FET) is used in high-grade glioma (HGG) to assess tumor metabolic activity via positron emission tomography (PET). We aim to investigate its value for target volume definition, as a prognosticator, and associations with whole-blood transcriptome liquid biopsy (WBT lbx) for which we recently reported feasibility to mirror tumor characteristics and response to particle irradiation in recurrent HGG (rHGG). Methods: 18F-FET-PET data from n = 43 patients with primary glioblastoma (pGBM) and n = 33 patients with rHGG were assessed. pGBM patients were irradiated with photons and sequential proton/carbon boost, and rHGG patients were treated with carbon re-irradiation (CIR). WBT (Illumina HumanHT-12 Expression BeadChips) lbx was available for n = 9 patients from the rHGG cohort. PET isocontours (40%-70% SUVmax, 10% steps) and MRI-based treatment volumes (MRIvol) were compared using the conformity index (CI) (pGBM, n = 16; rHGG, n = 27). Associations with WBT lbx data were tested on gene expression level and inferred pathways activity scores (PROGENy) and from transcriptome estimated cell fractions (CIBERSORT, xCell). Results: In pGBM, median SUVmax was higher in PET acquired pre-radiotherapy (4.1, range (R) 1.5-7.8; n = 20) vs. during radiotherapy (3.3, R 1.5-5.7, n = 23; p = 0.03) and in non-resected (4.7, R 2.9-7.9; n = 11) vs. resected tumors (3.3, R 1.5-7.8, n = 32; p = 0.01). In rHGG, a trend toward higher SUVmax values in grade IV tumors was observed (p = 0.13). Median MRIvol was 32.34 (R 8.75-108.77) cm3 in pGBM (n = 16) and 20.77 (R 0.63-128.44) cm3 in rHGG patients (n = 27). The highest median CI was observed for 40% (pGBM, 0.31) and 50% (rHGG, 0.43, all tumors) isodose, with 70% (40%) isodose in grade III (IV) rHGG tumors (median CI, 0.38 and 0.49). High SUVmax was linked to shorter survival in pGBM (>3.3, p = 0.001, OR 6.0 [2.1-17.4]) and rHGG (>2.8, p = 0.02, OR 4.1 [1.2-13.9]). SUVmax showed associations with inferred monocyte fractions, hypoxia, and TGFbeta pathway activity and links to immune checkpoint gene expression from WBT lbx. Conclusion: The benefits of 18F-FET-PET imaging on gross tumor volume (GTV) definition for particle radiotherapy warrant further evaluation. SUVmax might assist in prognostic stratification of HGG patients for particle radiotherapy, highlights heterogeneity in rHGG, and is positively associated with unfavorable signatures in peripheral whole-blood transcriptomes.

AAMP is a binding partner of costimulatory human B7-H3.

Background: Targeted immunotherapies are of growing interest in the treatment of various cancers. B7 homolog 3 protein (B7-H3), a member of the co-stimulatory/-inhibitory B7-family, exerts immunosuppressive and pro-tumorigenic functions in various cancer types and is under evaluation in ongoing clinical trials. Unfortunately, interaction partner(s) remain unknown which restricts the druggability. Methods: Aiming to identify potential binding partner(s) of B7-H3, a yeast two-hybrid and a mass spectrometry screen were performed. Potential candidates were evaluated by bimolecular fluorescence complementation (BiFC) assay, co-immunoprecipitation (co-IP), and functionally in a 3H-thymidine proliferation assay of Jurkat cells, a T-cell lineage cell line. Prognostic value of angio-associated migratory cell protein (AAMP) and B7-H3 expression was evaluated in isocitrate dehydrogenase 1 wildtype (IDH1wt) glioblastoma (GBM) patients from The Cancer Genome Atlas (TCGA)-GBM cohort. Results: Of the screening candidates, CD164, AAMP, PTPRA, and SLAMF7 could be substantiated via BiFC. AAMP binding could be further confirmed via co-IP and on a functional level. AAMP was ubiquitously expressed in glioma cells, immune cells, and glioma tissue, but did not correlate with glioma grade. Finally, an interaction between AAMP and B7-H3 could be observed on expression level, hinting toward a combined synergistic effect. Conclusions: AAMP was identified as a novel interaction partner of B7-H3, opening new possibilities to create a targeted therapy against the pro-tumorigenic costimulatory protein B7-H3.

modelBuildR: an R package for model building and feature selection with erroneous classifications.

BACKGROUND: Model building is a crucial part of omics based biomedical research to transfer classifications and obtain insights into underlying mechanisms. Feature selection is often based on minimizing error between model predictions and given classification (maximizing accuracy). Human ratings/classifications, however, might be error prone, with discordance rates between experts of 5-15%. We therefore evaluate if a feature pre-filtering step might improve identification of features associated with true underlying groups. METHODS: Data was simulated for up to 100 samples and up to 10,000 features, 10% of which were associated with the ground truth comprising 2-10 normally distributed populations. Binary and semi-quantitative ratings with varying error probabilities were used as classification. For feature preselection standard cross-validation (V2) was compared to a novel heuristic (V1) applying univariate testing, multiplicity adjustment and cross-validation on switched dependent (classification) and independent (features) variables. Preselected features were used to train logistic regression/linear models (backward selection, AIC). Predictions were compared against the ground truth (ROC, multiclass-ROC). As use case, multiple feature selection/classification methods were benchmarked against the novel heuristic to identify prognostically different G-CIMP negative glioblastoma tumors from the TCGA-GBM 450 k methylation array data cohort, starting from a fuzzy umap based rough and erroneous separation. RESULTS: V1 yielded higher median AUC ranks for two true groups (ground truth), with smaller differences for true graduated differences (3-10 groups). Lower fractions of models were successfully fit with V1. Median AUCs for binary classification and two true groups were 0.91 (range: 0.54-1.00) for V1 (Benjamini-Hochberg) and 0.70 (0.28-1.00) for V2, 13% (n = 616) of V2 models showed AUCs < = 50% for 25 samples and 100 features. For larger numbers of features and samples, median AUCs were 0.75 (range 0.59-1.00) for V1 and 0.54 (range 0.32-0.75) for V2. In the TCGA-GBM data, modelBuildR allowed best prognostic separation of patients with highest median overall survival difference (7.51 months) followed a difference of 6.04 months for a random forest based method. CONCLUSIONS: The proposed heuristic is beneficial for the retrieval of features associated with two true groups classified with errors. We provide the R package modelBuildR to simplify (comparative) evaluation/application of the proposed heuristic (http://github.com/mknoll/modelBuildR).