Adaptive fractionation at the MR-linac.

Objective. Fractionated radiotherapy typically delivers the same dose in each fraction. Adaptive fractionation (AF) is an approach to exploit inter-fraction motion by increasing the dose on days when the distance of tumor and dose-limiting organs at risk (OAR) is large and decreasing the dose on unfavorable days. We develop an AF algorithm and evaluate the concept for patients with abdominal tumors previously treated at the MR-linac in 5 fractions.Approach. Given daily adapted treatment plans, inter-fractional changes are quantified by sparing factorsδtdefined as the OAR-to-tumor dose ratio. The key problem of AF is to decide on the dose to deliver in fractiont, givenδtand the dose delivered in previous fractions, but not knowing futureδts. Optimal doses that maximize the expected biologically effective dose in the tumor (BED10) while staying below a maximum OAR BED3constraint are computed using dynamic programming, assuming a normal distribution overδwith mean and variance estimated from previously observed patient-specificδts. The algorithm is evaluated for 16 MR-linac patients in whom tumor dose was compromised due to proximity of bowel, stomach, or duodenum.Main Results. In 14 out of the 16 patients, AF increased the tumor BED10compared to the reference treatment that delivers the same OAR dose in each fraction. However, in 11 of these 14 patients, the increase in BED10was below 1 Gy. Two patients with large sparing factor variation had a benefit of more than 10 Gy BED10increase. For one patient, AF led to a 5 Gy BED10decrease due to an unfavorable order of sparing factors.Significance. On average, AF provided only a small increase in tumor BED. However, AF may yield substantial benefits for individual patients with large variations in the geometry.

PET/CT radiomics for prediction of hyperprogression in metastatic melanoma patients treated with immune checkpoint inhibitors.

Purpose: This study evaluated pretreatment 2[18F]fluoro-2-deoxy-D-glucose (FDG)-PET/CT-based radiomic signatures for prediction of hyperprogression in metastatic melanoma patients treated with immune checkpoint inhibition (ICI). Material and method: Fifty-six consecutive metastatic melanoma patients treated with ICI and available imaging were included in the study and 330 metastatic lesions were individually, fully segmented on pre-treatment CT and FDG-PET imaging. Lesion hyperprogression (HPL) was defined as lesion progression according to RECIST 1.1 and doubling of tumor growth rate. Patient hyperprogression (PD-HPD) was defined as progressive disease (PD) according to RECIST 1.1 and presence of at least one HPL. Patient survival was evaluated with Kaplan-Meier curves. Mortality risk of PD-HPD status was assessed by estimation of hazard ratio (HR). Furthermore, we assessed with Fisher test and Mann-Whitney U test if demographic or treatment parameters were different between PD-HPD and the remaining patients. Pre-treatment PET/CT-based radiomic signatures were used to build models predicting HPL at three months after start of treatment. The models were internally validated with nested cross-validation. The performance metric was the area under receiver operating characteristic curve (AUC). Results: PD-HPD patients constituted 57.1% of all PD patients. PD-HPD was negatively related to patient overall survival with HR=8.52 (95%CI 3.47-20.94). Sixty-nine lesions (20.9%) were identified as progressing at 3 months. Twenty-nine of these lesions were classified as hyperprogressive, thereby showing a HPL rate of 8.8%. CT-based, PET-based, and PET/CT-based models predicting HPL at three months after the start of treatment achieved testing AUC of 0.703 +/- 0.054, 0.516 +/- 0.061, and 0.704 +/- 0.070, respectively. The best performing models relied mostly on CT-based histogram features. Conclusions: FDG-PET/CT-based radiomic signatures yield potential for pretreatment prediction of lesion hyperprogression, which may contribute to reducing the risk of delayed treatment adaptation in metastatic melanoma patients treated with ICI.

Intrafractional stability of MR-guided online adaptive SBRT for prostate cancer.

BACKGROUND: MR-guided online adaptive stereotactic body radiation therapy (SBRT) for prostate cancer aims to reduce toxicity by full compensation of interfractional uncertainties. However, the process of online adaptation currently takes approximately 45 min during which intrafractional movements remain unaccounted for. This study aims to analyze the dosimetric benefit of online adaptation and to evaluate its robustness over the duration of one treatment fraction. METHODS: Baseline MR-scans at a MR-linear accelerator were acquired for ten healthy male volunteers for generation of mock-prostate SBRT plans with a dose prescription of 5 × 7.25 Gy. On a separate day, online MR-guided adaptation (ViewRay® MRIdian) was performed, and thereafter MR images were acquired every 15 min for 1 h to assess the stability of the adapted plan. RESULTS: A dosimetric benefit of online MR-guided adaptive re-planning was observed in 90% of volunteers. The median D95CTV- and D95PTV-coverage was improved from 34.8 to 35.5 Gy and from 30.7 to 34.6 Gy, respectively. Improved target coverage was not associated with higher dose to the organs at risk, most importantly the rectum (median D1ccrectum baseline plan vs. adapted plan 33.3 Gy vs. 32.3 Gy). The benefit of online adaptation remained stable over 45 min for all volunteers. However, at 60 min, CTV-coverage was below a threshold of 32.5 Gy in 30% of volunteers (30.6 Gy, 32.0 Gy, 32.3 Gy). CONCLUSION: The dosimetric benefit of MR-guided online adaptation for prostate SBRT was robust over 45 min in all volunteers. However, intrafractional uncertainties became dosimetrically relevant at 60 min and we therefore recommend verification imaging before delivery of MR-guided online adapted SBRT.

Improving interinstitutional and intertechnology consistency of pulmonary SBRT by dose prescription to the mean internal target volume dose.

PURPOSE: Dose, fractionation, normalization and the dose profile inside the target volume vary substantially in pulmonary stereotactic body radiotherapy (SBRT) between different institutions and SBRT technologies. Published planning studies have shown large variations of the mean dose in planning target volume (PTV) and gross tumor volume (GTV) or internal target volume (ITV) when dose prescription is performed to the PTV covering isodose. This planning study investigated whether dose prescription to the mean dose of the ITV improves consistency in pulmonary SBRT dose distributions. MATERIALS AND METHODS: This was a multi-institutional planning study by the German Society of Radiation Oncology (DEGRO) working group Radiosurgery and Stereotactic Radiotherapy. CT images and structures of ITV, PTV and all relevant organs at risk (OAR) for two patients with early stage non-small cell lung cancer (NSCLC) were distributed to all participating institutions. Each institute created a treatment plan with the technique commonly used in the institute for lung SBRT. The specified dose fractionation was 3â€¯× 21.5 Gy normalized to the mean ITV dose. Additional dose objectives for target volumes and OAR were provided. RESULTS: In all, 52 plans from 25 institutions were included in this analysis: 8 robotic radiosurgery (RRS), 34 intensity-modulated (MOD), and 10 3D-conformal (3D) radiation therapy plans. The distribution of the mean dose in the PTV did not differ significantly between the two patients (median 56.9 Gy vs 56.6 Gy). There was only a small difference between the techniques, with RRS having the lowest mean PTV dose with a median of 55.9 Gy followed by MOD plans with 56.7 Gy and 3D plans with 57.4 Gy having the highest. For the different organs at risk no significant difference between the techniques could be found. CONCLUSIONS: This planning study pointed out that multiparameter dose prescription including normalization on the mean ITV dose in combination with detailed objectives for the PTV and ITV achieve consistent dose distributions for peripheral lung tumors in combination with an ITV concept between different delivery techniques and across institutions.

High-dose re-irradiation of intracranial lesions - Efficacy and safety including dosimetric analysis based on accumulated EQD2Gy dose EQD calculation.

INTRODUCTION: The use of cranial re-irradiation is growing with improving overall survival and the advent of high-precision radiotherapy techniques. Still the value of re-irradiation needs careful evaluation regarding safety and efficacy. We analyzed dosimetric and clinical data of patients receiving cranial re-irradiation using EQD2 sum plans. METHODS AND MATERIAL: We retrospectively analyzed the data of 76 patients who received repeated cranial radiotherapy from 02/2013 to 09/2016. 34 patients suffered from recurrent primary brain tumors, 42 from brain metastases. Dosimetric analysis was performed accumulating EQD2 dose distributions based on rigid image registration. Clinical and radiological data was collected at follow-ups including toxicity, local control and overall survival. RESULTS: In total 76 patients had at least 2 courses of intracranial radiotherapy. The median accumulated prescription EQD2 dose was 96.5 Gy2 for all radiation courses combined. The median D(0.1 cc) of the brain for patients receiving more than 100 Gy2 was 114 Gy2 with a highest dose of 161.5 Gy2. 74% of patients suffered from low grade (G1-G2) acute toxicity, only two high grade (>G3) toxicities were recorded.Median overall survival from the time of first re-irradiation was 57 weeks (range 4-186 weeks). The median time to local failure for patients with a primary brain tumor was not reached and 24 weeks (range 1-77 weeks) for patients with brain metastases. CONCLUSION: Repeated radiotherapy appears both safe and efficient in patients with recurrent primary or secondary brain tumors with doses to the brain up to 120 Gy2 EQD2, doses below 100 Gy2 for brainstem and doses below 75 Gy2 EQD2 to chiasm and optic nerves.

Applicability of radiomics in interstitial lung disease associated with systemic sclerosis: proof of concept.

OBJECTIVE: To retrospectively evaluate if texture-based radiomics features are able to detect interstitial lung disease (ILD) and to distinguish between the different disease stages in patients with systemic sclerosis (SSc) in comparison with mere visual analysis of high-resolution computed tomography (HRCT). METHODS: Sixty patients (46 females, median age 56 years) with SSc who underwent HRCT of the thorax were retrospectively analyzed. Visual analysis was performed by two radiologists for the presence of ILD features. Gender, age, and pulmonary function (GAP) stage was calculated from clinical data (gender, age, pulmonary function test). Data augmentation was performed and the balanced dataset was split into a training (70%) and a testing dataset (30%). For selecting variables that allow classification of the GAP stage, single and multiple logistic regression models were fitted and compared by using the Akaike information criterion (AIC). Diagnostic accuracy was evaluated from the area under the curve (AUC) from receiver operating characteristic (ROC) analyses, and diagnostic sensitivity and specificity were calculated. RESULTS: Values for some radiomics features were significantly lower (p < 0.05) and those of other radiomics features were significantly higher (p = 0.001) in patients with GAP2 compared with those in patients with GAP1. The combination of two specific radiomics features in a multivariable model resulted in the lowest AIC of 10.73 with an AUC of 0.96, 84% sensitivity, and 99% specificity. Visual assessment of fibrosis was inferior in predicting individual GAP stages (AUC 0.86; 83% sensitivity; 74% specificity). CONCLUSION: The correlation of radiomics with GAP stage, but not with the visually defined features of ILD-HRCT, implies that radiomics might capture features indicating severity of SSc-ILD on HRCT, which are not recognized by visual analysis. KEY POINTS: • Radiomics features can predict GAP stage with a sensitivity of 84% and a specificity of almost 100%. • Extent of fibrosis on HRCT and a combined model of different visual HRCT-ILD features perform worse in predicting GAP stage. • The correlation of radiomics with GAP stage, but not with the visually defined features of ILD-HRCT, implies that radiomics might capture features on HRCT, which are not recognized by visual analysis.

FDG PET versus CT radiomics to predict outcome in malignant pleural mesothelioma patients.

BACKGROUND: Careful selection of malignant pleural mesothelioma (MPM) patients for curative treatment is of highest importance, as the multimodal treatment regimen is challenging for patients and harbors a high risk of substantial toxicity. Radiomics-a quantitative method for image analysis-has shown its prognostic ability in different tumor entities and could therefore play an important role in optimizing patient selection for radical cancer treatment. So far, radiomics as a prognostic tool in MPM was not investigated. MATERIALS AND METHODS: This study is based on 72 MPM patients treated with surgery in a curative intent at our institution between 2009 and 2017. Pre-treatment Fluorine-18 fluorodeoxyglucose (FDG) PET and CT scans were used for radiomics outcome modeling. After extraction of 1404 CT and 1410 FDG PET features from each image, a preselection by principal component analysis was performed to include only robust, non-redundant features for the cox regression to predict the progression-free survival (PFS) and the overall survival (OS). Results were validated on a separate cohort. Additionally, SUVmax and SUVmean, and volume were tested for their prognostic ability for PFS and OS. RESULTS: For the PFS a concordance index (c-index) of 0.67 (95% CI 0.52-0.82) and 0.66 (95% CI 0.57-0.78) for the training cohort (n = 36) and internal validation cohort (n = 36), respectively, were obtained for the PET radiomics model. The PFS advantage of the low-risk group translated also into an OS advantage. On CT images, no radiomics model could be trained. SUV max and SUV mean were also not prognostic in terms of PFS and OS. CONCLUSION: We were able to build a successful FDG PET radiomics model for the prediction of PFS in MPM. Radiomics could serve as a tool to aid clinical decision support systems for treatment of MPM in future.

Re-irradiation in the thorax - An analysis of efficacy and safety based on accumulated EQD2 doses.

INTRODUCTION: Thoracic re-irradiation remains a challenge regarding the balance of local efficacy and acceptable toxicities. In this retrospective analysis we analyzed dosimetrical and clinical data of patients treated with thoracic re-irradiation based on accumulated EQD2Gy doses. METHODS AND MATERIAL: We retrospectively analyzed the data of 42 consecutive single-institutional patients treated with repeated courses of thoracic radiotherapy from 12/2011 to 01/2017. Accumulated EQD2 dose distributions were calculated and dose parameters for organs at risk and target volumes were analysed. RESULTS: The median prescription dose was 42.2 Gy (10-70.6 Gy) for all RT courses. The median Dmean of both lungs was 10.1 Gy3 (range: 1.9 Gy3-17.9 Gy3) with a maximum D0.1 cc of 253.86 Gy3. The median D0.1 cc of the esophagus was 62.2 Gy3 with a maximum of 103.78 Gy3. The maximum D0.1 cc for the bronchial tree was 187.33 Gy3 (median 74.35 Gy3) and for the Aorta 216.1 Gy3 (median 70.9 Gy3). Median OS after first re-irradiation was 19 months (range 1-45 months). 12-month local control after a course of re-irradiation was 52.6%. 80% of patients suffered from a G1-G2 toxicity, most frequently coughing. One patient suffered from a G5 complication probably unrelated to re-irradiation. CONCLUSION: Even though several organs at risk received maximum accumulated doses of >100 Gy3, thoracic reirradiation resulted in an acceptable toxicity profile. Local tumor control and overall survival remained encouraging even after multiple courses of thoracic radiotherapy.

Repeated Courses of Radiosurgery for New Brain Metastases to Defer Whole Brain Radiotherapy: Feasibility and Outcome With Validation of the New Prognostic Metric Brain Metastasis Velocity.

Purpose: Stereotactic radiosurgery (SRS) is the preferred primary treatment option for patients with a limited number of asymptomatic brain metastases. In case of relapse after initial SRS the optimal salvage treatment is not well defined. Within this retrospective analysis, we investigated the feasibility of repeated courses of SRS to defer Whole-Brain Radiation Therapy (WBRT) and aimed to derive prognostic factors for patient selection. Materials and Methods: From 2014 until 2017, 42 patients with 197 brain metastases have been treated with multiple courses of SRS at our institution. Treatment was delivered as single fraction (18 or 20 Gy) or hypo-fractionated (6 fractions with 5 Gy) radiosurgery. Regular follow-up included clinical examination and contrast-enhanced cMRI at 3-4 months' intervals. Besides clinical and treatment related factors, brain metastasis velocity (BMV) as a newly described clinical prognostic metric was included and calculated between first and second treatment. Results: A median number of 1 lesion (range: 1-13) per course and a median of 2 courses (range: 2-6) per patient were administered resulting in a median of 4 (range: 2-14) metastases treated over time per patient. The median interval between SRS courses was 5.8 months (range: 0.9-35 months). With a median follow-up of 17.4 months (range: 4.6-45.5 months) after the first course of treatment, a local control rate of 84% was observed after 1 year and 67% after 2 years. Median time to out-of-field-brain-failure (OOFBF) was 7 months (95%CI 4-8 months). WBRT as a salvage treatment was eventually required in 7 patients (16.6%). Median overall survival (OS) has not been reached. Grouped by ds-GPA (≤ 2 vs. >2) the survival curves showed a significant split (p = 0.039). OS differed also significantly between BMV-risk groups when grouped into low vs. intermediate/high risk groups (p = 0.025). No grade 4 or 5 acute or late toxicity was observed. Conclusion: In selected patients with relapse after SRS for brain metastases, repeat courses of SRS were safe and minimized the need for rescue WBRT. The innovative, yet easy to calculate metric BMV may facilitate treatment decisions as a prognostic factor for OS.

Correction to: Planning comparison of five automated treatment planning solutions for locally advanced head and neck cancer.

Following publication of the original article [1], the authors reported that one of the authors' names was processed incorrectly.

Planning comparison of five automated treatment planning solutions for locally advanced head and neck cancer.

BACKGROUND: Automated treatment planning and/or optimization systems (ATPS) are in the process of broad clinical implementation aiming at reducing inter-planner variability, reducing the planning time allocated for the optimization process and improving plan quality. Five different ATPS used clinically were evaluated for advanced head and neck cancer (HNC). METHODS: Three radiation oncology departments compared 5 different ATPS: 1) Automatic Interactive Optimizer (AIO) in combination with RapidArc (in-house developed and Varian Medical Systems); 2) Auto-Planning (AP) (Philips Radiation Oncology Systems); 3) RapidPlan version 13.6 (RP1) with HNC model from University Hospital A (Varian Medical Systems, Palo Alto, USA); 4) RapidPlan version 13.7 (RP2) combined with scripting for automated setup of fields with HNC model from University Hospital B; 5) Raystation multicriteria optimization algorithm version 5 (RS) (Laboratories AB, Stockholm, Sweden). Eight randomly selected HNC cases from institution A and 8 from institution B were used. PTV coverage, mean and maximum dose to the organs at risk and effective planning time were compared. Ranking was done based on 3 Gy increments for the parallel organs. RESULTS: All planning systems achieved the hard dose constraints for the PTVs and serial organs for all patients. Overall, AP achieved the best ranking for the parallel organs followed by RS, AIO, RP2 and RP1. The oral cavity mean dose was the lowest for RS (31.3 ± 17.6 Gy), followed by AP (33.8 ± 17.8 Gy), RP1 (34.1 ± 16.7 Gy), AIO (36.1 ± 16.8 Gy) and RP2 (36.3 ± 16.2 Gy). The submandibular glands mean dose was 33.6 ± 10.8 Gy (AP), 35.2 ± 8.4 Gy (AIO), 35.5 ± 9.3 Gy (RP2), 36.9 ± 7.6 Gy (RS) and 38.2 ± 7.0 Gy (RP1). The average effective planning working time was substantially different between the five ATPS (in minutes): < 2 ± 1 for AIO and RP2, 5 ± 1 for AP, 15 ± 2 for RP1 and 340 ± 48 for RS, respectively. CONCLUSIONS: All ATPS were able to achieve all planning DVH constraints and the effective working time was kept bellow 20 min for each ATPS except for RS. For the parallel organs, AP performed the best, although the differences were small.

Modelling the immunosuppressive effect of liver SBRT by simulating the dose to circulating lymphocytes: an in-silico planning study.

BACKGROUND: Tumor immune-evasion and associated failure of immunotherapy can potentially be overcome by radiotherapy, which however also has detrimental effects on tumor-infiltrating and circulating lymphocytes (CL). We therefore established a model to simulate the radiation-dose delivered to CL. METHODS: A MATLAB-model was established to quantify the CL-dose during SBRT of liver metastases by considering the factors: hepatic blood-flow, -velocity and transition-time of individual hepatic segments, as well as probability-based recirculation. The effects of intra-hepatic tumor-location and size, fractionation and treatment planning parameters (VMAT, 3DCRT, photon-energy, dose-rate and beam-on-time) were analyzed. A threshold dose ≥0.5Gy was considered inactivating CL and CL0.5 (%) is the proportion of inactivated CL. RESULTS: Mean liver dose was mostly influenced by treatment-modality, whereas CL0.5 was mostly influenced by beam-on-time. 3DCRT and VMAT (10MV-FFF) resulted in lowest CL0.5 values of 16 and 19%. Metastasis location influenced CL0.5, with a mean of 19% for both apical and basal and 31% for the central location. PTV-volume significantly increased CL0.5 from 27 to 67% (10MV-FFF) and from 31 to 98% (6MV-FFF) for PTV-volumes ranging from 14cm3 to 268cm3. CONCLUSION: A simulation-model was established, quantifying the strong effects of treatment-technique, tumor-location and tumor-volume on dose to CL with potential implications for immune-optimized treatment-planning in the future.

Interobserver variability in target volume delineation of hepatocellular carcinoma : An analysis of the working group "Stereotactic Radiotherapy" of the German Society for Radiation Oncology (DEGRO).

BACKGROUND: Definition of gross tumor volume (GTV) in hepatocellular carcinoma (HCC) requires dedicated imaging in multiple contrast medium phases. The aim of this study was to evaluate the interobserver agreement (IOA) in gross tumor delineation of HCC in a multicenter panel. METHODS: The analysis was performed within the "Stereotactic Radiotherapy" working group of the German Society for Radiation Oncology (DEGRO). The GTVs of three anonymized HCC cases were delineated by 16 physicians from nine centers using multiphasic CT scans. In the first case the tumor was well defined. The second patient had multifocal HCC (one conglomerate and one peripheral tumor) and was previously treated with transarterial chemoembolization (TACE). The peripheral lesion was adjacent to the previous TACE site. The last patient had an extensive HCC with a portal vein thrombosis (PVT) and an inhomogeneous liver parenchyma due to cirrhosis. The IOA was evaluated according to Landis and Koch. RESULTS: The IOA for the first case was excellent (kappa: 0.85); for the second case moderate (kappa: 0.48) for the peripheral tumor and substantial (kappa: 0.73) for the conglomerate. In the case of the peripheral tumor the inconsistency is most likely explained by the necrotic tumor cavity after TACE caudal to the viable tumor. In the last case the IOA was fair, with a kappa of 0.34, with significant heterogeneity concerning the borders of the tumor and the PVT. CONCLUSION: The IOA was very good among the cases were the tumor was well defined. In complex cases, where the tumor did not show the typical characteristics, or in cases with Lipiodol (Guerbet, Paris, France) deposits, IOA agreement was compromised.

Nomogram based overall survival prediction in stereotactic body radiotherapy for oligo-metastatic lung disease.

BACKGROUND: Radical local treatment of pulmonary metastases is practiced with increasing frequency due to acknowledgment and better understanding of oligo-metastatic disease. This study aimed to develop a nomogram predicting overall survival (OS) after stereotactic body radiotherapy (SBRT) for pulmonary metastases. PATIENTS AND METHODS: A multi-institutional database of 670 patients treated with SBRT for pulmonary metastases was used as training cohort. Cox regression analysis with bidirectional variable elimination was performed to identify factors to be included into the nomogram model to predict 2-year OS. The calibration rate of the nomogram was assessed by plotting the actual Kaplan-Meier 2-year OS against the nomogram predicted survival. The nomogram was externally validated using two separate monocentric databases of 145 and 92 patients treated with SBRT for pulmonary metastases. RESULTS: The median follow up of the trainings cohort was 14.3months, the 2-year and 5-year OS was 52.6% and 23.7%, respectively. Karnofsky performance index, type of the primary tumor, control of the primary tumor, maximum diameter of the largest treated metastasis and number of metastases (1 versus >1) were significant prognostic factors in the Cox model (all p<0.05). The calculated concordance-index for the nomogram was 0.73 (concordance indexes of all prognostic factors between 0.54 and 0.6). Based on the nomogram the training cohort was divided into 4 groups and 2-year OS ranged between 24.2% and 76.1% (predicted OS between 30.2% and 78.4%). The nomogram discriminated between risk groups in the two validation cohorts (concordance index 0.68 and 0.67). CONCLUSIONS: A nomogram for prediction of OS after SBRT for pulmonary metastases was generated and externally validated. This tool might be helpful for interdisciplinary discussion and evaluation of local and systemic treatment options in the oligo-metastatic setting. KEY MESSAGE: A nomogram for prediction of overall survival after stereotactic body radiotherapy (SBRT) for pulmonary metastases was developed and externally validated. This tool might be helpful for interdisciplinary discussion and evaluation of local and systemic treatment options in the oligo-metastatic setting.

Stability of radiomic features in CT perfusion maps.

This study aimed to identify a set of stable radiomic parameters in CT perfusion (CTP) maps with respect to CTP calculation factors and image discretization, as an input for future prognostic models for local tumor response to chemo-radiotherapy. Pre-treatment CTP images of eleven patients with oropharyngeal carcinoma and eleven patients with non-small cell lung cancer (NSCLC) were analyzed. 315 radiomic parameters were studied per perfusion map (blood volume, blood flow and mean transit time). Radiomics robustness was investigated regarding the potentially standardizable (image discretization method, Hounsfield unit (HU) threshold, voxel size and temporal resolution) and non-standardizable (artery contouring and noise threshold) perfusion calculation factors using the intraclass correlation (ICC). To gain added value for our model radiomic parameters correlated with tumor volume, a well-known predictive factor for local tumor response to chemo-radiotherapy, were excluded from the analysis. The remaining stable radiomic parameters were grouped according to inter-parameter Spearman correlations and for each group the parameter with the highest ICC was included in the final set. The acceptance level was 0.9 and 0.7 for the ICC and correlation, respectively. The image discretization method using fixed number of bins or fixed intervals gave a similar number of stable radiomic parameters (around 40%). The potentially standardizable factors introduced more variability into radiomic parameters than the non-standardizable ones with 56-98% and 43-58% instability rates, respectively. The highest variability was observed for voxel size (instability rate >97% for both patient cohorts). Without standardization of CTP calculation factors none of the studied radiomic parameters were stable. After standardization with respect to non-standardizable factors ten radiomic parameters were stable for both patient cohorts after correction for inter-parameter correlations. Voxel size, image discretization, HU threshold and temporal resolution have to be standardized to build a reliable predictive model based on CTP radiomics analysis.