Search

Purpose: DNA-dependent protein kinase (DNA-PK) plays a key role in the repair of DNA double strand breaks via nonho- mologous end joining. Inhibition of DNA-PK can enhance the effect of DNA double strand break inducing anticancer thera- pies. Peposertib (formerly “M3814”) is an orally administered, potent, and selective small molecule DNA-PK inhibitor that has demonstrated radiosensitizing and antitumor activity in xenograft models and was well-tolerated in monotherapy. This phase 1 trial (National Clinical Trial 02516813) investigated the maximum tolerated dose, recommended phase 2 dose (RP2D), safety, and tolerability of peposertib in combination with palliative radiation therapy (RT) in patients with thoracic or head and neck tumors (arm A) and of peposertib in combination with cisplatin and curative-intent RT in patients with squamous cell carcinoma of the head and neck (arm B). Methods and Materials: Patients received peposertib once daily in ascending dose cohorts as a tablet or capsule in combina- tion with palliative RT (arm A) or in combination with intensity modulated curative-intent RT and cisplatin (arm B). Results: The most frequently observed treatment-emergent adverse events were radiation skin injury, fatigue, and nausea in arm A (n = 34) and stomatitis, nausea, radiation skin injury, and dysgeusia in arm B (n = 11). Based on evaluations of dose- limiting toxicities, tolerability, and pharmacokinetic data, RP2D for arm A was declared as 200 mg peposertib tablet once daily in combination with RT. In arm B (n = 11), 50 mg peposertib was declared tolerable in combination with curative-intent RT and cisplatin. However, enrollment was discontinued because of insufficient exposure at that dose, and the RP2D was not for- mally declared. Conclusions: Peposertib in combination with palliative RT was well-tolerated up to doses of 200 mg once daily as tablet with each RT fraction. When combined with RT and cisplatin, a tolerable peposertib dose yielded insuffic

Hypoxia is a common feature of solid tumours affecting their biology and response to therapy. One of the main transcription factors activated by hypoxia is hypoxia- inducible factor (HIF), which regulates the expression of genes involved in various aspects of tumourigenesis including proliferative capacity, angiogenesis, immune evasion, metabolic reprogramming, extracellular matrix (ECM) remodelling, and cell migration. This can negatively impact patient outcomes by inducing therapeutic resistance. The importance of hypoxia is clearly demonstrated by continued research into finding clinically relevant hypoxia biomarkers, and hypoxia-targeting therapies. One of the problems is the lack of clinically applicable methods of hypoxia detection, and lack of standardisation. Additionally, a lot of the methods of detecting hypoxia do not take into consideration the complexity of the hypoxic tumour microenvironment (TME). Therefore, this needs further elucidation as approximately 50% of solid tumours are hypoxic. The ECM is important component of the hypoxic TME, and is developed by both cancer associated fibroblasts (CAFs) and tumour cells. However, it is important to distinguish the different roles to develop both biomarkers and novel compounds. Fibronectin (FN), collagen (COL) and hyaluronic acid (HA) are important components of the ECM that create ECM fibres. These fibres are crosslinked by specific enzymes including lysyl oxidase (LOX) which regulates the stiffness of tumours and induces fibrosis. This is partially regulated by HIFs. The review highlights the importance of understanding the role of matrix stiffness in different solid tumours as current data shows contradictory results on the impact on therapeutic resistance. The review also indicates that further research is needed into identifying different CAF subtypes and their exact roles; with some showing pro-tumorigenic capacity and others having anti- tumorigenic roles. This has made it difficult to fully elucidate the

Personalized treatment strategies based on non‑invasive biomarkers have potential to improve patient management in patients with newly diagnosed glioblastoma (GBM). The residual tumour burden after surgery in GBM patients is a prognostic imaging biomarker. However, in clinical patient management, its assessment is a manual and time‑consuming process that is at risk of inter‑rater variability. Furthermore, the prediction of patient outcome prior to radiotherapy may identify patient subgroups that could benefit from escalated radiotherapy doses. Therefore, in this study, we investigate the capabilities of traditional radiomics and 3D convolutional neural networks for automatic detection of the residual tumour status and to prognosticate time‑to‑recurrence (TTR) and overall survival (OS) in GBM using postoperative [11C] methionine positron emission tomography (MET‑PET) and gadolinium‑enhanced T1‑w magnetic resonance imaging (MRI). On the independent test data, the 3D‑DenseNet model based on MET‑PET achieved the best performance for residual tumour detection, while the logistic regression model with conventional radiomics features performed best for T1c‑w MRI (AUC: MET‑PET 0.95, T1c‑w MRI 0.78). For the prognosis of TTR and OS, the 3D‑DenseNet model based on MET‑PET integrated with age and MGMT status achieved the best performance (Concordance‑Index: TTR 0.68, OS 0.65). In conclusion, we showed that both deep‑ learning and conventional radiomics have potential value for supporting image‑based assessment and prognosis in GBM. After prospective validation, these models may be considered for treatment personalization.

This repository contains the outputs and result data of our deep-learning-based experiments for the approximation of Monte-Carlo-simulated linear energy transfer distributions, which build the foundation for the corresponding article. The Pytorch checkpoint of our finally chosen SegResNet architecture trained on the UPTD dose distributions is located at dd_pbs/Dose-LETd/clip_let_below_0.04/segresnet/all_trainvalid_data/training/lightning_logs/version_6358843/checkpoints/last.ckpt. Moreover, we provide an exemplary data sample from a water phantom for trying our analysis pipeline.

Radiomics analyses provide a promising avenue for enabling personalized radiotherapy. Most frequently, prognostic radiomics models are based on features extracted from medical images that are acquired before treatment. Here, we investigate whether combining data from multiple timepoints during treatment and additionally from multiple imaging modalities can improve the predictive ability of radiomics models. We extracted radiomics features from computed tomography (CT) images acquired before treatment as well as two and three weeks after the start of radiochemotherapy for 55 patients with locally advanced head and neck squamous cell carcinoma (HNSCC). Additionally, we obtained features from FDG-PET images taken before treatment and three weeks after start of therapy. Cox proportional hazards models were then built based on features of the different image modalities, treatment timepoints and combinations thereof using two different feature selection methods in a five-fold cross-validation approach. Based on the cross-validation results, feature signatures were derived and their performance was independently validated. Discrimination regarding loco-regional control was assessed by the concordance index (C-index) and log-rank tests were performed to assess risk stratification. The best prognostic performance was obtained for timepoints during treatment for all modalities. Overall, CT was the best discriminating modality with an independent validation C-index of 0.78 for week two and week two and three combined. However, none of these models achieved a statistically significant patient stratification. Models based on FDG features from week three provided both, satisfactory discrimination (C-index=0.61 and 0.64) and a statistically significant stratification (p=0.044 and p<0.001) but produced highly imbalanced risk groups. After independent validation on larger data sets, the value of (multimodal) radiomics models combining several imaging timepoints should be prospective

Purpose: Adjuvant radio(chemo)therapy (RT) is part of the standard treatment of gliomas. Safety margins ensuring the coverage of microscopic tumour expansion of diffusely infiltrating gliomas and compensating for systematic positioning errors inevitably result in the normal-appearing (NA) brain tissue surrounding the tumour to be affected by radiation. The aim of the study was to investigate dose- and time-dependent diffusion alterations of NA white matter (WM) structures following RT using tract-based spatial statistics (TBSS). Methods: As part of a prospective, longitudinal study, magnetic resonance imaging (MRI) data of 24 grade II-IV glioma patients treated with photons, protons or mixed-modality therapy were acquired. MRIs before RT and 3-monthly during follow-up obtained up to three years after RT included diffusion tensor images (DTI) (TR/TE=6500/66ms, 2×2×2mm³, 32 directions, b=1000mm/s²). Fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD) were calculated from the DTI data. Corresponding radiation dose maps and clinical target volume (CTV) contours were aligned to MRI using ANTs. NA tissue was defined as brain tissue, excluding the CTV and areas of T2-hyperintensities. All FA images were nonlinearly registered to the “FMRI58B-FA atlas” from FSL with ANTs before applying parts of the TBSS algorithm to create a FA skeleton (Figure 1). The FA skeleton was combined with the “JHU-ICBM-labels-1mm atlas” to measure the diffusion in 19 WM structures. Relative signal changes of each WM structure were calculated as the difference between follow-up and the corresponding baseline signal and evaluated using a paired t-test. A multivariate linear mixed effects model was applied to determine diffusion changes as function of time after RT and mean dose delivered to the corresponding structure. Data from paired structures of the right and left hemispheres were combined for the analysis. Structures containing less than 50 vo

Einleitung In der Radioonkologie hat sich die Protonentherapie als Alternative zur Photonentherapie etabliert. Die relative biologische Wirksamkeit (RBW) von Protonen wird bislang klinisch als konstant angenommen, hängt jedoch u.a. von der physikalischen Dosis, dem Gewebetyp und dem linearen Energietransfer (LET) ab. Die Vernachlässigung dieser Variabilität kann zu einer relevanten Unterschätzung von therapiebedingten Nebenwirkungen führen. Die Berechnung des LET ist in einigen Bestrahlungsplanungssystemen möglich, steht jedoch in anderen Systemen sowie für retrospektive Daten nicht immer zur Verfügung. Deep-Learning-basierte Ansätze könnten in diesen Fällen helfen den LET, und somit die variable RBW, abzuschätzen. In dieser Arbeit wird das Potenzial von 3D Convolutional Neural Networks (CNN) zur Berechnung des dosisgemittelten LET (LETd) auf Basis der geplanten Dosisverteilung untersucht. Material & Methoden An einer Kohorte von 115 an der Universitäts Protonen Therapie Dresden behandelten Patienten mit primärem Hirntumor wurden drei Netzwerkarchitekturen (UNet, UNETR, SegResNet) mittels einer fünffachen Kreuzvalidierung trainiert und verglichen. Das leistungsfähigste Modell wurde an 28 Patienten aus dem Westdeutschen Protonentherapiezentrum Essen validiert. Dabei wurde die mittels Monte-Carlo (MC)-Simulation generierte LETd-Verteilung anhand der geplanten Dosisverteilung vorhergesagt. Die Genauigkeit des Resultats wurde für einzelne Regionen (u.a. Hirn, Hirnstamm, CTV) bewertet. Durch die RBE-gewichtete Dosis wurden therapiebedingte Nebenwirkungen anhand veröffentlichter Modelle abgeschätzt. Ergebnisse Das SegResNet zeigte die qualitativ besten Ergebnisse in der Kreuzvalidierung. In der externen Validierung dieses Modells ergaben sich für den mittleren LETd im Hirn, dem Hirnstamm und dem CTV jeweils eine Wurzel der mittleren quadratischen Abweichung (RMSE) von 0,22, 0,33 und 0,25 keV/µm. Der gepaarte Wilcoxon-Vorzeichen-Rang-Test wies für die drei Regionen keine s

Purpose: Adjuvant radio(chemo)therapy (RT) is part of the standard treatment of gliomas. Safety margins ensuring the coverage of microscopic tumour expansion of diffusely infiltrating gliomas and compensating for systematic positioning errors inevitably result in the normal-appearing (NA) brain tissue surrounding the tumour to be affected by radiation. The aim of the study was to investigate dose- and time-dependent diffusion alterations of NA white matter (WM) structures following RT using tract-based spatial statistics (TBSS). Methods: As part of a prospective, longitudinal study, magnetic resonance imaging (MRI) data of 24 grade II-IV glioma patients treated with photons, protons or mixed-modality therapy were acquired. MRIs before RT and 3-monthly during follow-up obtained up to three years after RT included diffusion tensor images (DTI) (TR/TE=6500/66ms, 2×2×2mm³, 32 directions, b=1000mm/s²). Fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD) were calculated from the DTI data. Corresponding radiation dose maps and clinical target volume (CTV) contours were aligned to MRI using ANTs. NA tissue was defined as brain tissue, excluding the CTV and areas of T2-hyperintensities. All FA images were nonlinearly registered to the “FMRI58B-FA atlas” from FSL with ANTs before applying parts of the TBSS algorithm to create a FA skeleton (Figure 1). The FA skeleton was combined with the “JHU-ICBM-labels-1mm atlas” to measure the diffusion in 19 WM structures. Relative signal changes of each WM structure were calculated as the difference between follow-up and the corresponding baseline signal and evaluated using a paired t-test. A multivariate linear mixed effects model was applied to determine diffusion changes as function of time after RT and mean dose delivered to the corresponding structure. Data from paired structures of the right and left hemispheres were combined for the analysis. Structures containing less than 50 vo

Aims: The number of Proton Therapy (PT) facilities is still limited worldwide, and the access to treatment could be characterized by patients’ logistic and economic challenges. Aim of the present survey is to assess the support provided to patients undergoing PT across Europe. Methods: Through a personnel contact, an online questionnaire (62 multiple-choice and open-ended questions) via Microsoft Forms was administered to 10 European PT centers. The questionnaire consisted of 62 questions divided into 6 sections: i) personal data; ii) general information on clinical activity; iii) fractionation, concurrent systemic treatments and technical aspects of PT facility; iv) indication to PT and reimbursement policies; v) economic and/ or logistic support to patients vi) participants agreement on statements related to the possible limitation of access to PT. A qualitative analysis was performed and reported. Results: From March to May 2022 all ten involved centers filled the survey. Nine centers treat from 100 to 500 patients per year. Paediatric patients accounted for 10–30%, 30–50% and 50–70% of the entire cohort for 7, 2 and 1 center, respectively. The most frequent tumours treated in adult population were brain tumours, sarcomas and head and neck carcinomas; in all centers, the mean duration of PT is longer than 3 weeks. In 80% of cases, the treatment reimbursement for PT is supplied by the respective country’s Health National System (HNS). HNS also provides economic support to patients in 70% of centers, while logistic and meal support is provided in 20% and 40% of centers, respectively. PT facilities offer economic and/or logistic support in 90% of the cases. Logistic support for parents of pediatric patients is provided by HNS only in one-third of centers. Overall, 70% of re- spondents agree that geographic challenges could limit a patient’s access to proton facilities and 60% believe that additional support should be given to patients referred for PT care.

LAPC is associated with a poor prognosis and requires a multimodal treatment approach. However, the role of radiation therapy in LAPC treatment remains controversial. This systematic review aimed to explore the role of proton and photon therapy, with varying radiation techniques and fractionation, in treatment outcomes and their respective toxicity profiles. Methods: Clinical studies published from 2012 to 2022 were systematically reviewed using PubMed, MEDLINE (via PubMed) and Cochrane databases. Different radiotherapy-related data were extracted and analyzed. Results: A total of 31 studies matched the inclusion criteria. Acute toxicity was less remarkable in stereotactic body radiotherapy (SBRT) compared to conventionally fractionated radiotherapy (CFRT), while in proton beam therapy (PBT) grade 3 or higher acute toxicity was observed more commonly with doses of 67.5 Gy (RBE) or higher. Late toxicity was not reported in most studies; therefore, comparison between groups was not possible. The range of median overall survival (OS) for the CFRT and SBRT groups was 9.3–22.9 months and 8.5–20 months, respectively. For the PBT group, the range of median OS was 18.4–22.3 months. Conclusion: CFRT and SBRT showed comparable survival outcomes with a more favorable acute toxicity profile for SBRT. PBT is a promising new treatment modality; however, additional clinical studies are needed to support its efficacy and safety.

Background The current management of lung cancer patients has reached a high level of complexity. Indeed, besides the traditional clinical variables (e.g., age, sex, TNM stage), new omics data have recently been introduced in clinical practice, thereby making more complex the decision-making process. With the advent of Artificial intelligence (AI) techniques, various omics datasets may be used to create more accurate predictive models paving the way for a better care in lung cancer patients. Methods The LANTERN study is a multi-center observational clinical trial involving a multidisciplinary consortium of five institutions from different European countries. The aim of this trial is to develop accurate several predictive models for lung cancer patients, through the creation of Digital Human Avatars (DHA), defined as digital representations of patients using various omics-based variables and integrating well-established clinical factors with genomic data, quantitative imaging data etc. A total of 600 lung cancer patients will be prospectively enrolled by the recruiting centers and multi-omics data will be collected. Data will then be modelled and parameterized in an experimental context of cutting-edge big data analysis. All data variables will be recorded according to a shared common ontology based on variable-specific domains in order to enhance their direct actionability. An exploratory analysis will then initiate the biomarker identification process. The second phase of the project will focus on creating multiple multivariate models trained though advanced machine learning (ML) and AI techniques for the specific areas of interest. Finally, the developed models will be validated in order to test their robustness, transferability and generalizability, leading to the development of the DHA. All the potential clinical and scientific stakeholders will be involved in the DHA development process. The main goals aim of LANTERN project are: i) To develop predictive models

The EU Horizon 2020 Framework-funded Standardized Treatment and Outcome Platform for Stereotactic Therapy Of Re-entrant tachycardia by a Multidisciplinary (STOPSTORM) consortium has been established as a large research network for investigating STereotactic Arrhythmia Radioablation (STAR) for ventricular tachycardia (VT). The aim is to provide a pooled treatment database to evaluate patterns of practice and outcomes of STAR and finally to harmonize STAR within Europe. The consortium comprises 31 clinical and research institutions. The project is divided into nine work packages (WPs): (i) observational cohort; (ii) standardization and harmonization of target delineation; (iii) harmonized prospective cohort; (iv) quality assurance (QA); (v) analysis and evaluation; (vi, ix) ethics and regulations; and (vii, viii) project coordination and dissemination. To provide a review of current clinical STAR practice in Europe, a comprehensive questionnaire was performed at project start. The STOPSTORM Institutions' experience in VT catheter ablation (83% ≥ 20 ann.) and stereotactic body radiotherapy (59% > 200 ann.) was adequate, and 84 STAR treatments were performed until project launch, while 8/22 centres already recruited VT patients in national clinical trials. The majority currently base their target definition on mapping during VT (96%) and/or pace mapping (75%), reduced voltage areas (63%), or late ventricular potentials (75%) during sinus rhythm. The majority currently apply a single-fraction dose of 25 Gy while planning techniques and dose prescription methods vary greatly. The current clinical STAR practice in the STOPSTORM consortium highlights potential areas of optimization and harmonization for substrate mapping, target delineation, motion management, dosimetry, and QA, which will be addressed in the various WPs. © The Author(s) 2023. Published by Oxford University Press on behalf of the European Society of Cardiology.

Background and purpose: There is no consensus about an ideal robust optimization (RO) strategy for proton therapy of targets with large intra-fractional motion. We investigated the plan robustness of different RO strategies regarding setup/range errors, interplay effects and interfractional anatomical changes. Materials and methods: For eight non-small cell lung cancer patients with primary and/or nodal clinical target volume (CTVp/CTVn) with motion >5mm, different RO approaches were investigated: 3DRO considering the average CT (AvgCT) with a target density override, 4DRO considering three/all 4DCT phases, and 4DRO considering the AvgCT and three/all 4DCT phases. Realistic interplay scenarios were reconstructed based on patient breathing and machine logfile data for deliveries with/without layered rescanning. Robustness against setup/range errors, interplay effects and interfractional anatomical changes were analyzed for target coverage and OAR sparing. Results: All nominal plans fulfilled the clinical requirements, while 4DRO without AvgCT generated the most conformal dose distributions. Robustness against setup/range errors was best for 4DRO with AvgCT. No RO strategy was sufficient in countervailing fraction-wise dose distortions caused by interplay effects. Irrespective of rescanning, target coverage was restored in all cases when accumulating four interplay scenarios. 4DRO with AvgCT showed higher CTVp robustness against interfractional changes, but plan adaptations are necessary for all RO strategies in case of relevant anatomical changes. Conclusion: All RO strategies are clinically acceptable but exhibit equally low robustness against interplay effects. To ensure fraction-wise target coverage, additional motion mitigation is required for CTVs with large motion amplitudes and interfractional changes need to be monitored.

Radiomics analyses provide a promising avenue for enabling personalized radiotherapy. Most frequently, prognostic radiomics models are based on features extracted from medical images that are acquired before treatment. Here, we investigate whether combining data from multiple timepoints during treatment and additionally from multiple imaging modalities can improve the predictive ability of radiomics models. We extracted radiomics features from computed tomography (CT) images acquired before treatment as well as two and three weeks after the start of radiochemotherapy for 55 patients with locally advanced head and neck squamous cell carcinoma (HNSCC). Additionally, we obtained features from FDG-PET images taken before treatment and three weeks after start of therapy. Cox proportional hazards models were then built based on features of the different image modalities, treatment timepoints and combinations thereof using two different feature selection methods in a five-fold cross-validation approach. Based on the cross-validation results, feature signatures were derived and their performance was independently validated. Discrimination regarding loco-regional control was assessed by the concordance index (C-index) and log-rank tests were performed to assess risk stratification. The best prognostic performance was obtained for timepoints during treatment for all modalities. Overall, CT was the best discriminating modality with an independent validation C-index of 0.78 for week two and week two and three combined. However, none of these models achieved a statistically significant patient stratification. Models based on FDG features from week three provided both, satisfactory discrimination (C-index=0.61 and 0.64) and a statistically significant stratification (p=0.044 and p<0.001) but produced highly imbalanced risk groups. After independent validation on larger data sets, the value of (multimodal) radiomics models combining several imaging timepoints should be prospective

Neural-network-based outcome predictions may enable further treatment personalization of patients with head and neck cancer. The development of neural networks can prove challenging when a limited number of cases is available. Therefore, we investigated whether multitask learning strategies, implemented through the simultaneous optimization of two distinct outcome objectives (multi-outcome) and combined with a tumor segmentation task, can lead to improved performance of convolutional neural networks (CNN) and vision transformers (ViT). Model training was conducted on two distinct multicenter datasets for the endpoints loco-regional control (LRC) and progression-free survival (PFS), respectively. The first dataset consisted of pre-treatment computed tomography (CT) imaging for 290 patients and the second dataset contained combined positron emission tomography (PET)/CT for 224 patients. Discriminative performance was assessed by the concordance index (C-index). Risk stratification was evaluated using log-rank tests. Across both datasets, CNN and ViT model ensembles achieved similar results. Multitask approaches showed favorable performance in most investigations. Multi-outcome CNN models trained with segmentation loss were identified as the optimal strategy across cohorts. On the PET/CT dataset, an ensemble of multi-outcome CNNs trained with segmentation loss achieved the best discrimination (C-index: 0.29, 95% confidence interval (CI): 0.22-0.36) and successfully stratified patients into groups with low and high risk of disease progression (p=0.003). On the CT dataset, ensembles of multi-outcome CNNs and of single-outcome ViTs trained with segmentation loss performed best (C-index: 0.26 and 0.26, CI: 0.18-0.34 and 0.18-0.35, respectively), both with significant risk stratification for LRC in independent validation (p=0.002 and p=0.011). Further validation of the developed multitask-learning models is planned based on a prospective validation study, which is currently

Purpose: Radio(chemo)therapy is used as standard treatment for glioma patients. The surrounding normal tissue is inevitably affected by the irradiation. The aim of this longitudinal study was to investigate perfusion alterations in the normal-appearing tissue after proton irradiation and assess the dose sensitivity of the normal tissue perfusion. Methods: In 14 glioma patients, a sub-cohort of a prospective clinical trial (NCT02824731), perfusion changes in normal-appearing white matter (WM), grey matter (GM) and subcortical GM structures, i.e. caudate nucleus, hippocampus, amygdala, putamen, pallidum and thalamus, were evaluated before treatment and at three-monthly intervals after proton beam irradiation. The relative cerebral blood volume (rCBV) was assessed with dynamic susceptibility contrast MRI and analysed as the percentage ratio between follow-up and baseline image (∆rCBV). Radiation-induced alterations were evaluated using Wilcoxon signed rank test. Dose and time correlations were investigated with univariate and multivariate linear regression models. Results: No significant ∆rCBV changes were found in any normal-appearing WM and GM region after proton beam irradiation. A positive correlation with radiation dose was observed in the multivariate regression model applied to the combined ∆rCBV values of low (1-20Gy), intermediate (21-40Gy) and high (41-60Gy) dose regions of GM (p<0.001), while no time dependency was detected in any normal-appearing area. Conclusion: The perfusion in normal-appearing brain tissue remained unaltered after proton beam therapy. In further studies, a direct comparison with changes after photon therapy is recommended to confirm the different effect of proton therapy on the normal-appearing tissue.

Purpose Currently, there is an intense debate on variations in intra-cerebral radiosensitivity and relative biological effectiveness (RBE) in proton therapy of primary brain tumours. Here, both effects were retrospectively investigated using late radiation-induced brain injuries (RIBI) observed in follow-up after proton therapy of patients with diagnosed glioma. Methods In total, 42 WHO grade 2–3 glioma patients out of a consecutive patient cohort having received (adjuvant) proton radio(chemo)therapy between 2014 and 2017 were eligible for analysis. RIBI lesions (symptomatic or clinically asymptomatic) were diagnosed and delineated on contrast-enhanced T1-weighted magnetic resonance imaging scans obtained in the first two years of follow-up. Correlation of RIBI location and occurrence with dose (D), proton dose-averaged linear energy transfer (LET) and variable RBE dose parameters were tested in voxel- and in patient-wise logistic regression analyses. Additionally, anatomical and clinical parameters were considered. Model performance was estimated through cross-validated area-under-the-curve (AUC) values. Results In total, 64 RIBI lesions were diagnosed in 21 patients. The median time between start of proton radio(chemo)therapy and RIBI appearance was 10.2 months. Median distances of the RIBI volume centres to the cerebral ventricles and to the clinical target volume border were 2.1 mm and 1.3 mm, respectively. In voxel-wise regression, the multivariable model with D, D × LET and periventricular region (PVR) revealed the highest AUC of 0.90 (95 % confidence interval: 0.89–0.91) while the corresponding model without D × LET revealed a value of 0.84 (0.83–0.86). In patient-level analysis, the equivalent uniform dose (EUD11, a = 11) in the PVR using a variable RBE was the most prominent predictor for RIBI with an AUC of 0.63 (0.32–0.90). Conclusions In this glioma cohort, an increased radiosensitivity within the PVR was observed as well as a spatial correlation of RIBI

Background and Purpose: Motion mitigation is of crucial importance in particle therapy (PT) of patients with abdominal tumors to ensure high-precision irradiation. Magnetic resonance imaging (MRI) is an excellent modality for target volume delineation and motion estimation of mobile soft-tissue tumors. Thus, the aims of this study were to develop an MRI- and PT-compatible abdominal compression device, to investigate its effect on pancreas motion reduction, and to evaluate patient tolerability and acceptance. Materials and Methods: In a prospective clinical study, 16 patients with abdominal tumors received an individualized polyethylene-based abdominal corset. Pancreas motion was analyzed using time- and phase resolved MRI scans (orthogonal 2D-cine and 4D MRI) with and without compression by the corset. The pancreas was manually segmented in each MRI data set and the population-averaged center-of-mass motion in inferior-superior (IS), anterior-posterior (AP) and left-right (LR) directions was determined. A questionnaire was developed to investigate the level of patient acceptance of the corset, which the patients completed after acquisition of the planning computed tomography (CT) and MRI scans. Results: The corset was found to reduce pancreas motion predominantly in IS direction by on average 47 % - 51 % as found in the 2D-cine and 4D MRI data, respectively, while motion in the AP and LR direction was not significantly reduced. Most patients reported no discomfort when wearing the corset. Conclusion: An MRI- and PT-compatible individualized abdominal corset was presented, which substantially reduced breathing-induced pancreas motion and can be safely applied with no additional discomfort for the patients. The corset has been successfully integrated into our in-house clinical workflow for PT of tumors of the upper abdomen.

Radiomics analyses provide a promising avenue for enabling personalized radiotherapy. Most frequently, prognostic radiomics models are based on features extracted from medical images that are acquired before treatment. Here, we investigate whether combining data from multiple timepoints during treatment and additionally from multiple imaging modalities can improve the predictive ability of radiomics models. We extracted radiomics features from computed tomography (CT) images acquired before treatment as well as two and three weeks after the start of radiochemotherapy for 55 patients with locally advanced head and neck squamous cell carcinoma (HNSCC). Additionally, we obtained features from FDG-PET images taken before treatment and three weeks after start of therapy. Cox proportional hazards models were then built based on features of the different image modalities, treatment timepoints and combinations thereof using two different feature selection methods in a five-fold cross-validation approach. Based on the cross-validation results, feature signatures were derived and their performance was independently validated. Discrimination regarding loco-regional control was assessed by the concordance index (C-index) and log-rank tests were performed to assess risk stratification. The best prognostic performance was obtained for timepoints during treatment for all modalities. Overall, CT was the best discriminating modality with an independent validation C-index of 0.78 for week two and week two and three combined. However, none of these models achieved a statistically significant patient stratification. Models based on FDG features from week three provided both, satisfactory discrimination (C-index=0.61 and 0.64) and a statistically significant stratification (p=0.044 and p<0.001) but produced highly imbalanced risk groups. After independent validation on larger data sets, the value of (multimodal) radiomics models combining several imaging timepoints should be prospective

This dataset contains the model checkpoints, predictions and performance metrics for the multitask neural networks presented in the corresponding manuscript.

Proton therapy is a promising option for cancer treatment, even though its radiobiological properties are not yet fully considered in clinical practice. In this context, the relative biological effectiveness (RBE) of protons is the most important quantity, which is strongly related to their linear energy transfer (LET). LET distributions can be provided by commercial treatment-planning systems based on Monte Carlo simulations. However, such systems require a considerable amount of computational resources, are not yet available in every proton-therapy centre and may not be applicable to assess retrospective patient data. Here, we provide proof-of-concept for inferring LET distributions using convolutional neural networks (CNN) based on proton therapy radiation dose distributions and treatment-planning computed tomography (CT). We further evaluate established models for estimating treatment-related side effects after proton therapy of brain tumours and observe good agreement between CNN and MC based outputs.

Background: Radiotherapy in patients with primary brain tumors may affect hippocampal structure and cause dyscognitive side-effects. Patients and methods: Using structural MRI and comprehensive neurocognitive evaluation, we investi- gated associations between hippocampal structure and memory deficits in 15 patients with WHO grade 3 and grade 4 gliomas receiving standard radio(chemo)therapy. Results: We did not find changes in hippocampal thickness or cognitive abilities three months after com- pleting radiotherapy. However, subjective memory impairment was associated with symptoms of depression, but not with objective memory performance, cortical thickness of the hippocampus or radi- ation dose. Conclusions: Irrespective of whether there is a bidirectional relationship between affective changes and subjective cognitive dysfunction in these patients, depressive symptoms remain a target for intervention to improve their quality of life. The results of our pilot study highlight that future assessment of side effects of radiotherapy concerning memory should include assessments of depressive symptoms.

Despite technological advances, normal tissue sparing in photon beam irradiation is still challenging. Since in esophageal cancer this may inflict damage on the lungs, heart and bone marrow, possibly impacting on outcome, the aim of this study was to investigate the association of normal tissue dose and blood parameters on the survival of patients having undergone neoadjuvant radiochemotherapy (RCTx) followed by surgery. This retrospective study included 125 patients irradiated to 40–41.4 Gy with photons or protons combined with concurrent chemotherapy. On initial and restaging 18F-FDG-PET/CT, the lungs and heart were contoured as organs at risk for which standardized uptake values (SUV) were evaluated. The mean radiation dose (Dmean) to the lungs and heart, the volume of the lungs receiving at least 20 Gy (V20Gy_lung) and various pre- and per-treatment blood parameters were included in the Cox regression analyses. Results: The median follow-up time was 19.8 months and median overall survival 37 months (95% confidence interval: 16–58.9 months). In multivariate analysis, higher radiation doses to the lungs and heart were statistically significantly associated with decreased overall survival (Dmean_lung: p < 0.001; V20Gy_lung: p < 0.002; Dmean_heart: p = 0.005). Neither the 18F-FDG-PET nor blood parameters were predictive for overall survival. In patients with locally advanced esophageal cancer treated with RCTx, the radiation dose to the heart and lungs was significantly associated with overall survival.

Background and purpose: In times of high-precision radiotherapy, the accurate and precise definition of the primary tumour localisation and its microscopic spread is of enormous importance. In glioblastoma, the microscopic tumour extension is uncertain and therefore population-based margins for clinical target volume (CTV) definition are clinically used, which could either be too small leading to increased risk of loco-regional recurrences or too large thus enhancing the probability of normal tissue toxicity. Therefore, the aim of this project is to investigate an individualized definition of the CTV in preclinical glioblastoma models, based on specific biological tumour characteristics. Material and methods: The microscopic tumour extensions of two different orthotopic brain tumour models (U87MG_mCherry; G7_mCherry) were evaluated before and during fractionated radiotherapy and correlated with corresponding histological data. Representative tumour slices were analysed using Matrix-assisted Laser Desorption/Ionization (MALDI) and stained for putative cancer stem cell markers as well as invasion markers (Nestin, MMP14, Musashi 1, CD44). Results: The edges of the tumour are clearly shown by the MALDI segmentation via unsupervised clustering of mass spectra and are consistent with the histologically defined border in H&E staining in both models. MALDI component analysis supposed specific peaks as potential markers for normal brain tissue (e.g. 1339 m/z), whereas other peaks demarcated the tumours very well (e.g. 1562 m/z for U87MG_mCherry) irrespective of treatment. MMP14 staining revealed only a few positive cells mainly in the tumour border, which could reflect the invasive front in both models. Conclusion: The results of this study indicate that a step towards an individualized CTV definition based on biological tumour characteristics, especially using MALDI information, in glioblastoma models seems possible. Visualization of tumour volume and protein heterogeneity

Patients with locally advanced head and neck squamous cell carcinoma (HNSCC) may benefit from personalised treatment, requiring biomarkers that characterize the tumour and predict treatment response. We integrate pre-treatment CT radiomics and whole-transcriptome data from a multicentre retrospective cohort of 206 patients with locally advanced HNSCC treated with primary radiochemotherapy to classify tumour molecular subtypes based on radiomics, develop surrogate radiomics signatures for gene-based signatures related to different biological tumour characteristics and evaluate the potential of combining radiomics features with full-transcriptome data for the prediction of loco-regional control (LRC). Using end-to-end machine-learning, we developed and validated a model to classify tumours of the atypical subtype (AUC [95% confidence interval]: 0.69 [0.53-0.83]) based on CT imaging, observed that CT-based radiomics models have limited value as surrogates for six selected gene signatures (AUC<0.60), and showed that combining a radiomics signature with a transcriptomics signature consisting of two metagenes representing the hedgehog pathway and E2F transcriptional targets improves the prognostic value for LRC compared to both individual sources (validation C-index [95% confidence interval], combined: 0.63 [0.55-0.73] vs radiomics: 0.60 [0.50-0.71] and transcriptomics: 0.59 [0.49-0.69]). These results underline the potential of multi-omics analyses to generate reliable biomarkers for future application in personalized oncology.

Correction to: Strahlenther Onkol 2021 https://doi.org/10.1007/s00066-021-01812-2

Radiomics analyses commonly apply imaging features of different complexity for the prediction of the endpoint of interest. However, the prognostic value of each feature class is generally unclear. Furthermore, many radiomics models lack independent external validation that is decisive for their clinical application. Therefore, in this manuscript we present two complementary studies. In our modelling study, we developed and validated different radiomics signatures for outcome prediction after neoadjuvant chemoradiotherapy (nCRT) in patients with locally advanced rectal cancer (LARC) based on computed tomography (CT) and T2-weighted (T2w) magnetic resonance (MR) imaging datasets of 4 independent institutions (training: 122, validation 68 patients). We compared different feature classes extracted from the gross tumour volume for the prognosis of tumour response and freedom from distant metastases (FFDM): morphological and first order (MFO) features, second order texture (SOT) features, and Laplacian of Gaussian (LoG) transformed intensity features. Analyses were performed for CT and MRI separately and combined. Model performance was assessed by the area under the curve (AUC) and the concordance index (CI) for tumour response and FFDM, respectively. Overall, intensity features of LoG transformed CT and MR imaging combined with clinical T stage (cT) showed the best performance for tumour response prediction, while SOT features showed good performance for FFDM in independent validation (AUC = 0.70, CI = 0.69). In our external validation study, we aimed to validate previously published radiomics signatures on our multicentre cohort. We identified relevant publications on comparable patient datasets through a literature search and applied the reported radiomics models to our dataset. Only one of the identified studies could be validated, indicating an overall lack of reproducibility and the need of further standardization of radiomics before clinical application.

Ten new organs at risk (OARs) were recently introduced in the updated European Particle Therapy Network neurological contouring atlas. Despite the use of the illustrated atlas and descriptive text, interindividual contouring variations may persist. To further facilitate the contouring of these OARs, educational films were developed and published on www.cancerdata.org.

Background and purpose: Temporary alopecia is a common side-effect in brain tumour patients receiving cranial radiotherapy with a significant psychological burden for the affected patient. The purpose of this study was to generate a method in our treatment planning system (TPS) to visualize the expected radiation-induced alopecia 4 weeks after treatment, in order to inform the patients thereupon before the start of radiotherapy. Material and methods: A pilot study was conducted in ten patients receiving hypo- (HF) or conventionally fractionated (CF) photon beam Volumetric Modulated Arc Therapy (VMAT) for an intracranial lesion. Dose calculations were correlated to visible alopecia four weeks after the end of treatment to create a structure predictive of alopecia in our TPS. These alopecia structures for both fractionation schedules were validated in two cohorts of 69 HF and 78 CF patients undergoing radiotherapy between 2016 and 2019. Results: In the pilot cohort, a total physical dose of 4 Gy for HF and 12.6 Gy for CF radiotherapy were found to be predictive of alopecia 4 weeks after treatment. Applying these doses to our validation cohort, we found an accurate prediction of alopecia in 59/69 (86%) HF and 73/78 (96%) CF patients. For the total patient group of 147 patients, the predicted amount of alopecia was accurate in 90% of the cases. All inaccurate predictions overestimated the expected extent of alopecia. Conclusion: The presented straightforward method to visualize predicted alopecia 4 weeks after treatment has proven to predict the extent alopecia highly accurate in the vast majority of patients. Sharing these results with the patients pre-treatment may result in stress reduction before cranial irradiation.

Purpose: Treatment-related toxicity after irradiation of brain tumours has been underreported in the lit- erature. Furthermore, there is considerable heterogeneity on how and when toxicity is evaluated. The aim of this European Particle Network (EPTN) collaborative project is to develop recommendations for uni- form follow-up and toxicity scoring of adult brain tumour patients treated with radiotherapy. Methods: A Delphi method-based consensus was reached among 24 international radiation-oncology experts in the field of neuro-oncology concerning the toxicity endpoints, evaluation methods and time points. Results: In this paper, we present a basic framework for consistent toxicity scoring and follow-up, using multiple levels of recommendation. Level I includes all recommendations that are considered minimum of care, whereas level II and III are optional evaluations in the advanced clinical or research setting, respectively. Per outcome domain, the clinical endpoints and evaluation methods per level are listed. Where relevant, the organ at risk threshold doses for recommended referral to specific organ specialists are defined.

In ‘‘Joint EANM/SNMMI/ESTRO Practice Recommendations for the Use of 2-[18F]FDG-PET/CT External Beam Radiation Treatment Planning in Lung Cancer V1.0” clinical indications for PET-CT in (non-)small cell lung cancer are highlighted and selective nodal irradiation is discussed. Additionally, concepts about target definition, target delineation and treatment evaluation are reviewed.

Background and purpose: Radiotherapy (RT) is an adjuvant treatment option for glioma patients. Side effects include tissue atrophy, which might be a contributing factor to neurocognitive decline after treatment. The goal of this study was to determine potential atrophy of the hippocampus, amygdala, thalamus, putamen, pallidum and caudate nucleus in glioma patients having undergone magnetic resonance imaging (MRI) before and after RT. Materials and methods: Subcortical volumes were measured using T1-weighted MRI from patients before RT (N = 91) and from longitudinal follow-ups acquired in three-monthly intervals (N = 349). The volumes were normalized to the baseline values, while excluding structures touching the clinical target volume (CTV) or abnormal tissue seen on FLAIR imaging. A multivariate linear effects model was used to determine if time after RT and mean RT dose delivered to the corresponding structures were significant predictors of tissue atrophy. Results: The hippocampus, amygdala, thalamus, putamen, and pallidum showed significant atrophy after RT as function of both time after RT and mean RT dose delivered to the corresponding structure. Only the caudate showed no dose or time dependant atrophy. Conversely, the hippocampus was the structure with the highest atrophy rate of 5.2 % after one year and assuming a mean dose of 30 Gy. Conclusion: The hippocampus showed the highest atrophy rates followed by the thalamus and the amygdala. The subcortical structures here found to decrease in volume indicative of radiosensitivity should be the focus of future studies investigating the relationship between neurocognitive decline and RT. © 2022 The Authors

Recently, a number of clinical studies have explored links between possible Relative Biological Effectiveness (RBE) elevations and patient toxicities and/or image changes following proton therapy. Our objective was to perform a systematic review of such studies. We applied a "Problem [RBE], Intervention [Protons], Population [Patients], Outcome [Side effect]” search strategy to the PubMed database. From our search, we retrieved studies which: (a) performed novel voxel-wise analyses of patient effects versus dose and LET (n= 13), and (b) compared image changes between proton and photon cohorts with regard to proton RBE (n=9). For each retrieved study, we extracted data regarding: primary tumour type; size of patient cohort; type of image change studied; image-registration method (deformable or rigid); LET calculation method, and statistical methodology. We compared and contrasted their methods in order to discuss the weight of clinical evidence for variable proton RBE. We concluded that clinical evidence for variable proton RBE remains statistically weak at present. Our principal recommendation is that proton centres and clinical trial teams collaborate to standardize follow-up protocols and statistical analysis methods, so that larger patient cohorts can ultimately be considered for RBE analyses.

Progression of prostate-specific antigen (PSA) values after curative treatment of prostate cancer patients is common. Prostate-specific membrane antigen (PSMA-) PET imaging can identify patients with metachronous oligometastatic disease even at low PSA levels. Metastases-directed local ablative radiotherapy (aRT) has been shown to be a safe treatment option. In this prospective clinical trial, we evaluated local control and the pattern of tumor progression. Between 2014 and 2018, 63 patients received aRT of 89 metastases (MET) (68 lymph node (LN-)MET and 21 bony (OSS-)MET) with one of two radiation treatment schedules: 50 Gy in 2 Gy fractions in 34 MET or 30 Gy in 10 Gy fractions in 55 MET. The mean gross tumor volume and planning target volume were 2.2 and 14.9 mL, respectively. The median follow-up time was 40.7 months. Local progression occurred in seven MET, resulting in a local control rate of 93.5% after three years. Neither treatment schedule, target volume, nor type of lesion was associated with local progression. Regional progression in the proximity to the LN-MET was observed in 19 of 47 patients with at least one LN-MET (actuarial 59.3% free of regional progression after 3 years). In 33 patients (52%), a distant progression was reported. The median time to first tumor-related clinical event was 16.6 months, and 22.2% of patients had no tumor-related clinical event after three years. A total of 14 patients (22%) had another aRT. In conclusion, local ablative radiotherapy in patients with PSMA-PET staged oligometastatic prostate cancer may achieve local control, but regional or distant progression is common. Further studies are warranted, e.g., to define the optimal target volume coverage in LN-MET and OSS-MET.

Purpose 2-[18F]FDGPET/CT is of utmost importance for radiation treatment (RT) planning and response monitoring in lung cancer patients, in both non-small and small cell lung cancer (NSCLC and SCLC). This topic has been addressed in guidelines composed by experts within the feld of radiation oncology. However, up to present, there is no procedural guideline on this subject, with involvement of the nuclear medicine societies. Methods A literature review was performed, followed by a discussion between a multidisciplinary team of experts in the different fields involved in the RT planning of lung cancer, in order to guide clinical management. The project was led by experts of the two nuclear medicine societies (EANM and SNMMI) and radiation oncology (ESTRO). Results and conclusion This guideline results from a joint and dynamic collaboration between the relevant disciplines for this topic. It provides a worldwide, state of the art, and multidisciplinary guide to 2-[18F]FDG PET/CT RT planning in NSCLC and SCLC. These practical recommendations describe applicable updates for existing clinical practices, highlight potential faws, and provide solutions to overcome these as well. Finally, the recent developments considered for future application are also reviewed.

Purpose 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) is utilized for staging and treatment planning of head and neck squamous cell carcinomas (HNSCC). Some older publications on the prognostic relevance showed inconclusive results, most probably due to small study sizes. This study evaluates the prognostic and potentially predictive value of FDG-PET in a large multi-center analysis. Methods Original analysis of individual FDG-PET and patient data from 16 international centers (8 institutional datasets, 8 public repositories) with 1104 patients. All patients received curative intent radiotherapy/chemoradiation (CRT) and pre-treatment FDG-PET imaging. Primary tumors were semi-automatically delineated for calculation of SUVmax, SUVmean, metabolic tumor volume (MTV) and total lesion glycolysis (TLG). Cox regression analyses were performed for event-free survival (EFS), overall survival (OS), loco-regional control (LRC) and freedom from distant metastases (FFDM). Results FDG-PET parameters were associated with patient outcome in the whole cohort regarding clinical endpoints (EFS, OS, LRC, FFDM), in uni- and multivariate Cox regression analyses. Several previously published cut-off values were successfully validated. Subgroup analyses identified tumor- and human papillomavirus (HPV) specific parameters. In HPV positive oropharynx cancer (OPC) SUVmax was well suited to identify patients with excellent LRC for organ preservation. Patients with SUVmax of 14 or less were unlikely to develop loco-regional recurrence after definitive CRT. In contrast FDG PET parameters deliver only limited prognostic information in laryngeal cancer. Conclusion FDG-PET parameters bear considerable prognostic value in HNSCC and potential predictive value in subgroups of patients, especially regarding treatment de-intensification and organ-preservation. The potential predictive value needs further validation in appropriate control groups. Further research on advanced imaging appro

Purpose: To compare dose distributions and robustness in treatment plans from eight European centres in preparation for the European randomized phase-III PROTECT-trial investigating the effect of proton therapy (PT) versus photon therapy (XT) for oesophageal cancer. Materials and methods: All centres optimized one PT and one XT nominal plan using delineated 4DCT scans for four patients receiving 50.4 Gy (RBE) in 28 fractions. Target volume receiving 95% of prescribed dose (V95%iCTVtotal) should be >99%. Robustness towards setup, range, and respiration was evaluated. The plans were recalculated on a surveillance 4DCT (sCT) acquired at fraction ten and robustness evaluation was performed to evaluate the effect of respiration and inter-fractional anatomical changes. Results: All PT and XT plans complied with V95%iCTVtotal >99% for the nominal plan and V95%iCTVtotal >97% for all respiratory and robustness scenarios. Lung and heart dose varied considerably between centres for both modalities. The difference in mean lung dose and mean heart dose between each pair of XT and PT plans was in median [range] 4.8 Gy [1.1;7.6] and 8.4 Gy [1.9;24.5], respectively. Patients B and C showed large inter-fractional anatomical changes on sCT. For patient B, the minimum V95%iCTVtotal in the worst- case robustness scenario was 45% and 94% for XT and PT, respectively. For patient C, the minimum V95%iCTVtotal was 57% and 72% for XT and PT, respectively. Patient A and D showed minor inter- fractional changes and the minimum V95%iCTVtotal was >85%. Conclusion: Large variability in dose to the lungs and heart was observed for both modalities. Inter- fractional anatomical changes led to larger target dose deterioration for XT than PT plans.

letter to editor

Background and purpose Major differences exist among proton therapy (PT) centres regarding PT delivery in adult cancer patient. To obtain insight into current practice in Europe, we performed a survey among European PT centres. Materials and methods We designed electronic questionnaires for eight tumour sites, focusing on four main topics: 1) indications and patient selection methods; 2) reimbursement; 3) on-going or planned studies, 4) annual number of patients treated with PT. Results Of 22 centres, 19 (86%) responded. In total, 4,233 adult patients are currently treated across Europe annually, of which 46% consists of patients with central nervous system tumours (CNS), 15% head and neck cancer (HNC), 15% prostate, 9% breast, 5% lung, 5% gastrointestinal, 4% lymphoma, 0.3% gynaecological cancers. CNS are treated in all participating centres (n=19) using PT, HNC in 16 centres, lymphoma in 10 centres, gastrointestinal in 10 centres, breast in 7 centres, prostate in 6 centres, lung in 6 centres, and gynaecological cancers in 3 centres. Reimbursement is provided by national health care systems for the majority of commonly treated tumour sites. Approximately 74% of centres enrol patients for prospective data registration programs. Phase II-III trials are less frequent, due to reimbursement and funding problems. Reasons for not treating certain tumour types with PT are lack of evidence (30%), reimbursement issues (29%) and/or technical limitations (20%). Conclusion Across European PT centres, CNS tumours and HNC are the most frequently treated tumour types. Most centres use indication protocols. Lack of evidence for PT and reimbursement issues are the most reported reasons for not treating specific tumour types with PT.

Bronchialkarzinome sind in Deutschland die dritthäufigste Tumorerkrankung. Trotz erheblicher therapeutischer Fortschritte ist die Prognose der Lungentumoren nach wie vor ungünstig, die relative 5-Jahres- Überlebensrate nach Diagnose eines Bron- chialkarzinoms beträgt lediglich 16 % [3]. Hinsichtlich der Tumorart werden nicht- kleinzellige (NSCLC) und kleinzellige Bron- chialkarzinome unterschieden. Bei fortgeschrittenen NSCLC besteht die Standardtherapie aus einer gleichzei- tig durchgeführten Radiochemotherapie, gefolgt von einer Immuntherapie. In Metaanalysen konnte die Überlegenheit der simultanen Radiochemotherapie ge- genüber einem sequenziellen Vorgehen gezeigt werden.

Purpose: We tested the hypothesis that gene expressions from biopsies of locally advanced head and neck squamous cell carcinoma (HNSCC) patients can supplement dose-volume parameters to predict dysphagia and xerostomia following primary radiochemotherapy (RCTx). Material and methods: A panel of 178 genes previously related to radiochemosensitivity of HNSCC was considered for nanoString analysis based on tumour biopsies of 90 patients with locally advanced HNSCC treated by primary RCTx. Dose-volume parameters were extracted from the parotid, subman- dibular glands, oral cavity, larynx, buccal mucosa, and lips. Normal tissue complication probability (NTCP) models were developed for acute, late, and for the improvement of xerostomia grade ≥2 and dysphagia grade ≥3 using a cross-validation-based least absolute shrinkage and selection operator (LASSO) approach combined with stepwise logistic regression for feature selection. The final signatures were included in a logistic regression model with optimism correction. Performance was assessed by the area under the receiver operating characteristic curve (AUC). Results: NTCP models for acute and late xerostomia and the improvement of dysphagia resulted in optimism-corrected AUC values of 0.84, 0.76, and 0.70, respectively. The minimum dose to the contra-lateral parotid was selected for both acute and late xerostomia and the minimum dose to the larynx was selected for dysphagia improvement. For the xerostomia endpoints, the following gene expressions were selected: RPA2 (cellular response to DNA damage), TCF3 (salivary gland cells development), GBE1 (glycogen storage and regulation), and MAPK3 (regulation of cellular processes). No gene expression features were selected for the prediction of dysphagia. Conclusion: This hypothesis-generating study showed the potential of improving NTCP models using gene expression data for HNSCC patients. The presented models require independent validation before potential application in clini

Proton therapy is a promising option for cancer treatment, even though its radiobiological properties are not yet fully considered in clinical practice. In this context, the relative biological effectiveness (RBE) of protons is the most important quantity, which is strongly related to their linear energy transfer (LET). LET distributions can be provided by commercial treatment-planning systems based on Monte Carlo simulations. However, such systems require a considerable amount of computational resources, are not yet available in every proton-therapy centre and may not be applicable to assess retrospective patient data. Here, we provide proof-of-concept for inferring LET distributions using convolutional neural networks (CNN) based on proton therapy radiation dose distributions and treatment-planning computed tomography (CT). We further evaluate established models for estimating treatment-related side effects after proton therapy of brain tumours and observe good agreement between CNN and MC based outputs.

We include the input data, analysis scripts, analysis results and scripts to create the visualizations and plots used in the manuscript and supplement to our article "Longitudinal and multimodal radiomics models for head-and-neck cancer outcome prediction".

RDS data of models developed and reported for the article entitled "Integrated radiogenomics analyses allow for subtype classification and improved outcome prognosis of patients with locally advanced HNSCC". Software package version is also available in repository

Introduction: Adjuvant radio(chemo)therapy (RT) is part of the treatment of primary brain tumor patients. In order to capture microscopic tumor extension and to compensate for random and systematic positioning uncertainties, it is inevitable that tumor-surrounding normal brain tissue is irradiated. The aim of this study was to determine relative cerebral blood volume (rCBV) changes in glioma patients before and after RT in normal appearing white matter (WM) and grey matter (GM). Methods: As part of an ongoing study, anatomical and functional MRI data of glioma patients undergoing gross tumor resection followed by RT is being collected. The analysis of a subset of this cohort, 17 glioma patients (3 grade II, 11 grade III, 3 grade IV, average age 46.9y ± 13.2y) is presented here. Two patients were treated with photon therapy, 14 patients with proton therapy and one patient received treatment modalities. MRI scans acquired prior to RT and at least one follow-up MRI obtained 3, 6 and 9 months after RT was evaluated. All MRI data were collected on a 3T Philips Ingenuity PET/MR scanner (Philips, Eindhoven, The Netherlands) using an 8 channel head coil and included anatomical T1w-images (3D-GRE, TR/TE=10/3.7ms, FA=20°, voxel size 1×1×1mm3) and dynamic susceptibility contrast (DSC) imaging using a PRESTO sequence (TR/TE=15/21ms, 60 dynamics, dynamic scantime=1.7s, voxel size 3.6×3.6×3.5mm3) with intravenous gadolinium contrast agent (0.1mol/kg, 4ml/s, 7s delay) followed by a saline flush (20ml, 4ml/s). The CBV-map (fig. 1D) was calculated as the area under curve (AUC) of the voxelwise time course of the 4D PRESTO image. Computed tomographies (CTs) used for planning, radiation dose (fig.1C) and clinical target volume (CTV) contours were registered to the T1w images using ANTs1 . T1w-images were segmented into GM and WM using SPM122 and the corresponding 95% tissue probability maps were rigidly registered with ANTs1 to the CBV-map. B1 inhomogeneities as well as voxels with st

Introduction Currently, there is an intense debate on the need to consider a variable clinical relative biological effectiveness (RBE) in proton therapy. Here, a variable clinically derived RBE-model was applied in-silico to predict the risk for late radiation-induced brain injuries (RIBI) in benign brain tumor patients having undergone proton therapy. Materials & Methods In total, 23 patients with benign brain tumors of WHO grade I-II, who received (adjuvant) proton radio(chemo)therapy between 2015 and 2017, were analyzed. Dose and linear energy transfer distributions were retrospectively simulated and used to calculate variable RBE-weighted dose in brain tissue. The variable RBE-model was previously derived for RIBI observed after proton therapy in grade II-IV gliomas. RBE-weighted dose, dose-volume parameters and normal tissue complication probabilities (NTCP) were calculated in brain tissue within and outside the clinical target volume (CTV) using the variable RBE-model and the clinically applied RBE of 1.1, excluding brainstem, gross tumor volume and surgical cavity from analysis. Results The average difference in maximum RBE-weighted dose between the variable and constant RBE-model was 12 Gy(RBE) [range: 7.9-15 Gy(RBE)] and 14.8 Gy(RBE) [8.9-19.2 Gy(RBE)] within and outside the CTV. In the same regions, values of 9.2 Gy(RBE) [5.8-12.6 Gy(RBE)] and 1.0 Gy(RBE) [0.2-3.2 Gy(RBE)] were obtained for the difference of the mean dose. Using the variable RBE-model the cohort average D4ml (minimum dose to the hottest 4 ml) and NTCP increased by 10.8 Gy(RBE) [7.4-16.0 Gy(RBE)] and 25.4% [0.6-53.4%], respectively (figure 1). Summary A substantial increase in high dose and predicted RIBI risk was found in normal and normal-appearing brain tissue using the assumption of a variable RBE-model instead of a generic RBE of 1.1. After correlation of predicted with occurring RIBI on follow-up MRI scans, our results may help to verify and extend clinical RBE-models established for

Introduction: Adjuvant radio(chemo)therapy (RT) is part of the treatment of primary brain tumor patients. In order to capture microscopic tumor extension and to compensate for random and systematic positioning uncertainties, it is inevitable that tumor-surrounding normal brain tissue is irradiated. The aim of this study was to determine relative cerebral blood volume (rCBV) changes in glioma patients before and after RT in normal appearing white matter (WM) and grey matter (GM). Methods: As part of an ongoing study, anatomical and functional MRI data of glioma patients undergoing gross tumor resection followed by RT is being collected. The analysis of a subset of this cohort, 17 glioma patients (3 grade II, 11 grade III, 3 grade IV, average age 46.9y ± 13.2y) is presented here. Two patients were treated with photon therapy, 14 patients with proton therapy and one patient received treatment modalities. MRI scans acquired prior to RT and at least one follow-up MRI obtained 3, 6 and 9 months after RT was evaluated. All MRI data were collected on a 3T Philips Ingenuity PET/MR scanner (Philips, Eindhoven, The Netherlands) using an 8 channel head coil and included anatomical T1w-images (3D-GRE, TR/TE=10/3.7ms, FA=20°, voxel size 1×1×1mm3) and dynamic susceptibility contrast (DSC) imaging using a PRESTO sequence (TR/TE=15/21ms, 60 dynamics, dynamic scantime=1.7s, voxel size 3.6×3.6×3.5mm3) with intravenous gadolinium contrast agent (0.1mol/kg, 4ml/s, 7s delay) followed by a saline flush (20ml, 4ml/s). The CBV-map (fig. 1D) was calculated as the area under curve (AUC) of the voxelwise time course of the 4D PRESTO image. Computed tomographies (CTs) used for planning, radiation dose (fig.1C) and clinical target volume (CTV) contours were registered to the T1w images using ANTs1 . T1w-images were segmented into GM and WM using SPM122 and the corresponding 95% tissue probability maps were rigidly registered with ANTs1 to the CBV-map. B1 inhomogeneities as well as voxels with st

Purpose/Objective The inclusion of breathing-induced variation using 4D robust plan optimization (RO) may yield more robust proton plans for non-small cell lung cancer (NSCLC) patients. Here, we analysed the benefit of 4DRO in terms of robustness against different motion effects. Material/Methods Five NSCLC patients with relevant intrafractional motion in the primary (CTVp; on average 3.2-11.1mm) and nodal clinical target volumes (CTVn; 0.9-7.4mm) were included. CTVs and organs at risk (OARs) were contoured on the planning (pCT) and up to two control 4DCTs (cCT). In RayStation 7.99 (RaySearch, Sweden), we optimized three robust normo-fractionated plans [dose: 66Gy(RBE)]with our clinical criteria of 5mm setup and 3.5%+2mm range uncertainty: RO on the average CT with density override of the primary integral gross tumour volume (3DRO); RO on the average, minimum, maximum and mid inspiration CT image (4DRO3); and RO on the average CT and all eight 4DCT phases (4DRO8). On each of the average, minimum and maximum inspiration pCT, 16 setup and range error scenarios were analysed. To assess the influence of intrafractional changes, a 4D dose was calculated for the pCT and compared to those for the cCTs assuming equal weights of all breathing phases. Interplay effects were simulated by 4D dynamic dose (4DDD) scenarios on the pCT using a logfile-based dose reconstruction with machine logfiles from mock treatments with and without 5 layered rescans and the breathing signals from 4DCT acquisition. To account for a possible fractionation effect within the first fractions, we accumulated 4DDD scenarios with 4 different starting times. Results All nominal plans fulfilled target coverage (D98%>95%) and OAR sparing; 3DRO achieved lower mean lung dose [up to 0.3 Gy(RBE)] in 4 patients and lower V5Gy of contralateral lung (up to 4pp). CTVp/CTVn coverage failed setup and range robustness on average in 7%/17% (3DRO), 9%/10% (4DRO3) and 9%/12% (4DRO8) of the scenarios, respectively. 4D d

Introduction: Adjuvant radio(chemo)therapy (RT) is part of the treatment of primary brain tumor patients. In order to capture microscopic tumor extension and to compensate for random and systematic positioning uncertainties, it is inevitable that tumor-surrounding normal brain tissue is irradiated. The aim of this study was to determine relative cerebral blood volume (rCBV) changes in glioma patients before and after RT in normal appearing white matter (WM) and grey matter (GM). Methods: As part of an ongoing study, anatomical and functional MRI data of glioma patients undergoing gross tumor resection followed by RT is being collected. The analysis of a subset of this cohort, 17 glioma patients (3 grade II, 11 grade III, 3 grade IV, average age 46.9y ± 13.2y) is presented here. Two patients were treated with photon therapy, 14 patients with proton therapy and one patient received treatment modalities. MRI scans acquired prior to RT and at least one follow-up MRI obtained 3, 6 and 9 months after RT was evaluated. All MRI data were collected on a 3T Philips Ingenuity PET/MR scanner (Philips, Eindhoven, The Netherlands) using an 8 channel head coil and included anatomical T1w-images (3D-GRE, TR/TE=10/3.7ms, FA=20°, voxel size 1×1×1mm3) and dynamic susceptibility contrast (DSC) imaging using a PRESTO sequence (TR/TE=15/21ms, 60 dynamics, dynamic scantime=1.7s, voxel size 3.6×3.6×3.5mm3) with intravenous gadolinium contrast agent (0.1mol/kg, 4ml/s, 7s delay) followed by a saline flush (20ml, 4ml/s). The CBV-map (fig. 1D) was calculated as the area under curve (AUC) of the voxelwise time course of the 4D PRESTO image. Computed tomographies (CTs) used for planning, radiation dose (fig.1C) and clinical target volume (CTV) contours were registered to the T1w images using ANTs1 . T1w-images were segmented into GM and WM using SPM122 and the corresponding 95% tissue probability maps were rigidly registered with ANTs1 to the CBV-map. B1 inhomogeneities as well as voxels with st

Background: We analyse the benefit of 4D robust plan optimization (RO) for proton therapy treatments of NSCLC patients with pronounced breathing-induced anatomical variations. Methods: Eight NSCLC patients with relevant maximal intrafractional motion of the primary (CTVp; 5.6-24.5mm) and/or nodal CTV (CTVn; 5.8-17.3mm) on the planning 4DCT (pCT) were included. We optimized three robust normo-fractionated plans with a criterion of 5mm setup and 3.5%+2mm range uncertainty: RO on the AverageCT with iGTVp density override (3DRO); RO on the AverageCT and three 4DCT phases (4DRO3); and RO on the AverageCT and all eight 4DCT phases (4DRO8). Setup and range error scenarios were analysed on the pCT. To assess robustness against intra- and inter-fractional changes, 4D doses were calculated on the pCT and up to two control 4DCTs (cCTs) assuming equal weights between breathing phases. Interplay effects were simulated on the pCT from patient breathing signals and machine logfiles of plan deliveries with and without 5 layered rescans. Fractionation effects were emulated by accumulating four interplay scenarios with different starting times. Results: All nominal plans fulfilled target coverage (D98%>95%) and OAR sparing; CTVp/CTVn coverage failed setup and range robustness in 12%/35% (3DRO), 14%/18% (4DRO3) and 13%/20% (4DRO8) of the scenarios (Fig1a), respectively. 4D dose target coverage on the pCT was >94% for all plans; interfractional changes in the cCTs reduced the CTVp coverage by about 2pp (Fig1b). Interplay analyses (Fig1c) revealed a mean CTVp/CTVn coverage loss by 3.4pp/2.4pp, 3.0pp/2.3pp, and 2.9pp/3.2pp in single scenarios of 3DRO, 4DRO3 and 4DRO8 plans, respectively. D98% values were improved on average by 1.0pp with rescanning, but were even worsened in some scenarios. Irrespective of rescanning, the simulated fractionation fullfilled the target coverage in all cases. Conclusion: 3DRO and 4DRO showed similar robustness against different motion effects. 4DRO provides