Safety and Tolerability of Online Adaptive High-Field Magnetic Resonance-Guided Radiotherapy.

Importance: In 2018, the first online adaptive magnetic resonance (MR)-guided radiotherapy (MRgRT) system using a 1.5-T MR-equipped linear accelerator (1.5-T MR-Linac) was clinically introduced. This system enables online adaptive radiotherapy, in which the radiation plan is adapted to size and shape changes of targets at each treatment session based on daily MR-visualized anatomy. Objective: To evaluate safety, tolerability, and technical feasibility of treatment with a 1.5-T MR-Linac, specifically focusing on the subset of patients treated with an online adaptive strategy (ie, the adapt-to-shape [ATS] approach). Design, Setting, and Participants: This cohort study included adults with solid tumors treated with a 1.5-T MR-Linac enrolled in Multi Outcome Evaluation for Radiation Therapy Using the MR-Linac (MOMENTUM), a large prospective international study of MRgRT between February 2019 and October 2021. Included were adults with solid tumors treated with a 1.5-T MR-Linac. Data were collected in Canada, Denmark, The Netherlands, United Kingdom, and the US. Data were analyzed in August 2023. Exposure: All patients underwent MRgRT using a 1.5-T MR-Linac. Radiation prescriptions were consistent with institutional standards of care. Main Outcomes and Measures: Patterns of care, tolerability, and technical feasibility (ie, treatment completed as planned). Acute high-grade radiotherapy-related toxic effects (ie, grade 3 or higher toxic effects according to Common Terminology Criteria for Adverse Events version 5.0) occurring within the first 3 months after treatment delivery. Results: In total, 1793 treatment courses (1772 patients) were included (median patient age, 69 years [range, 22-91 years]; 1384 male [77.2%]). Among 41 different treatment sites, common sites were prostate (745 [41.6%]), metastatic lymph nodes (233 [13.0%]), and brain (189 [10.5%]). ATS was used in 1050 courses (58.6%). MRgRT was completed as planned in 1720 treatment courses (95.9%). Patient withdrawal caused 5 patients (0.3%) to discontinue treatment. The incidence of radiotherapy-related grade 3 toxic effects was 1.4% (95% CI, 0.9%-2.0%) in the entire cohort and 0.4% (95% CI, 0.1%-1.0%) in the subset of patients treated with ATS. There were no radiotherapy-related grade 4 or 5 toxic effects. Conclusions and Relevance: In this cohort study of patients treated on a 1.5-T MR-Linac, radiotherapy was safe and well tolerated. Online adaptation of the radiation plan at each treatment session to account for anatomic variations was associated with a low risk of acute grade 3 toxic effects.

Evaluation of Deep Learning Clinical Target Volumes Auto-Contouring for Magnetic Resonance Imaging-Guided Online Adaptive Treatment of Rectal Cancer.

Purpose: Segmentation of clinical target volumes (CTV) on medical images can be time-consuming and is prone to interobserver variation (IOV). This is a problem for online adaptive radiation therapy, where CTV segmentation must be performed every treatment fraction, leading to longer treatment times and logistic challenges. Deep learning (DL)-based auto-contouring has the potential to speed up CTV contouring, but its current clinical use is limited. One reason for this is that it can be time-consuming to verify the accuracy of CTV contours produced using auto-contouring, and there is a risk of bias being introduced. To be accepted by clinicians, auto-contouring must be trustworthy. Therefore, there is a need for a comprehensive commissioning framework when introducing DL-based auto-contouring in clinical practice. We present such a framework and apply it to an in-house developed DL model for auto-contouring of the CTV in rectal cancer patients treated with MRI-guided online adaptive radiation therapy. Methods and Materials: The framework for evaluating DL-based auto-contouring consisted of 3 steps: (1) Quantitative evaluation of the model's performance and comparison with IOV; (2) Expert observations and corrections; and (3) Evaluation of the impact on expected volumetric target coverage. These steps were performed on independent data sets. The framework was applied to an in-house trained nnU-Net model, using the data of 44 rectal cancer patients treated at our institution. Results: The framework established that the model's performance after expert corrections was comparable to IOV, and although the model introduced a bias, this had no relevant impact on clinical practice. Additionally, we found a substantial time gain without reducing quality as determined by volumetric target coverage. Conclusions: Our framework provides a comprehensive evaluation of the performance and clinical usability of target auto-contouring models. Based on the results, we conclude that the model is eligible for clinical use.

Quantitative MRI on MR-Linacs: Towards Biological Image-Guided Adaptive Radiotherapy.

Recognizing the potential of quantitative imaging biomarkers (QIBs) in radiotherapy, many studies have investigated the prognostic value of quantitative MRI (qMRI). With the introduction of MRI-guided radiotherapy systems, the practical challenges of repeated imaging have been substantially reduced. Since patients are treated inside an MRI scanner, acquisition of qMRI can be done during each fraction with limited or no prolongation of the fraction duration. In this review paper, we identify the steps that need been taken to move from MR as an imaging technique to a useful biomarker for MRI-guided radiotherapy (MRgRT).

A network score-based metric to optimize the quality assurance of automatic radiotherapy target segmentations.

Background and purpose: Existing methods for quality assurance of the radiotherapy auto-segmentations focus on the correlation between the average model entropy and the Dice Similarity Coefficient (DSC) only. We identified a metric directly derived from the output of the network and correlated it with clinically relevant metrics for contour accuracy. Materials and Methods: Magnetic Resonance Imaging auto-segmentations were available for the gross tumor volume for cervical cancer brachytherapy (106 segmentations) and for the clinical target volume for rectal cancer external-beam radiotherapy (77 segmentations). The nnU-Net's output before binarization was taken as a score map. We defined a metric as the mean of the voxels in the score map above a threshold (λ). Comparisons were made with the mean and standard deviation over the score map and with the mean over the entropy map. The DSC, the 95th Hausdorff distance, the mean surface distance (MSD) and the surface DSC were computed for segmentation quality. Correlations between the studied metrics and model quality were assessed with the Pearson correlation coefficient (r). The area under the curve (AUC) was determined for detecting segmentations that require reviewing. Results: For both tasks, our metric (λ = 0.30) correlated more strongly with the segmentation quality than the mean over the entropy map (for surface DSC, r > 0.65 vs. r < 0.60). The AUC was above 0.84 for detecting MSD values above 2 mm. Conclusions: Our metric correlated strongly with clinically relevant segmentation metrics and detected segmentations that required reviewing, indicating its potential for automatic quality assurance of radiotherapy target auto-segmentations.

Recommendations for improved reproducibility of ADC derivation on behalf of the Elekta MRI-linac consortium image analysis working group.

BACKGROUND AND PURPOSE: The apparent diffusion coefficient (ADC), a potential imaging biomarker for radiotherapy response, needs to be reproducible before translation into clinical use. The aim of this study was to evaluate the multi-centre delineation- and calculation-related ADC variation and give recommendations to minimize it. MATERIALS AND METHODS: Nine centres received identical diffusion-weighted and anatomical magnetic resonance images of different cancerous tumours (adrenal gland, pelvic oligo metastasis, pancreas, and prostate). All centres delineated the gross tumour volume (GTV), clinical target volume (CTV), and viable tumour volume (VTV), and calculated ADCs using both their local calculation methods and each of the following calculation conditions: b-values 0-500 vs. 150-500 s/mm2, region-of-interest (ROI)-based vs. voxel-based calculation, and mean vs. median. ADC variation was assessed using the mean coefficient of variation across delineations (CVD) and calculation methods (CVC). Absolute ADC differences between calculation conditions were evaluated using Friedman's test. Recommendations for ADC calculation were formulated based on observations and discussions within the Elekta MRI-linac consortium image analysis working group. RESULTS: The median (range) CVD and CVC were 0.06 (0.02-0.32) and 0.17 (0.08-0.26), respectively. The ADC estimates differed 18% between b-value sets and 4% between ROI/voxel-based calculation (p-values < 0.01). No significant difference was observed between mean and median (p = 0.64). Aligning calculation conditions between centres reduced CVC to 0.04 (0.01-0.16). CVD was comparable between ROI types. CONCLUSION: Overall, calculation methods had a larger impact on ADC reproducibility compared to delineation. Based on the results, significant sources of variation were identified, which should be considered when initiating new studies, in particular multi-centre investigations.

Towards Response ADAptive Radiotherapy for organ preservation for intermediate-risk rectal cancer (preRADAR): protocol of a phase I dose-escalation trial.

INTRODUCTION: Organ preservation is associated with superior functional outcome and quality of life (QoL) compared with total mesorectal excision (TME) for rectal cancer. Only 10% of patients are eligible for organ preservation following short-course radiotherapy (SCRT, 25 Gy in five fractions) and a prolonged interval (4-8 weeks) to response evaluation. The organ preservation rate could potentially be increased by dose-escalated radiotherapy. Online adaptive magnetic resonance-guided radiotherapy (MRgRT) is anticipated to reduce radiation-induced toxicity and enable radiotherapy dose escalation. This trial aims to establish the maximum tolerated dose (MTD) of dose-escalated SCRT using online adaptive MRgRT. METHODS AND ANALYSIS: The preRADAR is a multicentre phase I trial with a 6+3 dose-escalation design. Patients with intermediate-risk rectal cancer (cT3c-d(MRF-)N1M0 or cT1-3(MRF-)N1M0) interested in organ preservation are eligible. Patients are treated with a radiotherapy boost of 2×5 Gy (level 0), 3×5 Gy (level 1), 4×5 Gy (level 2) or 5×5 Gy (level 3) on the gross tumour volume in the week following standard SCRT using online adaptive MRgRT. The trial starts on dose level 1. The primary endpoint is the MTD based on the incidence of dose-limiting toxicity (DLT) per dose level. DLT is a composite of maximum one in nine severe radiation-induced toxicities and maximum one in three severe postoperative complications, in patients treated with TME or local excision within 26 weeks following start of treatment. Secondary endpoints include the organ preservation rate, non-DLT, oncological outcomes, patient-reported QoL and functional outcomes up to 2 years following start of treatment. Imaging and laboratory biomarkers are explored for early response prediction. ETHICS AND DISSEMINATION: The trial protocol has been approved by the Medical Ethics Committee of the University Medical Centre Utrecht. The primary and secondary trial results will be published in international peer-reviewed journals. TRIAL REGISTRATION NUMBER: WHO International Clinical Trials Registry (NL8997; https://trialsearch.who.int).

Radiogenomics Analysis Linking Multiparametric MRI and Transcriptomics in Prostate Cancer.

Prostate cancer (PCa) is a highly prevalent cancer type with a heterogeneous prognosis. An accurate assessment of tumor aggressiveness can pave the way for tailored treatment strategies, potentially leading to better outcomes. While tumor aggressiveness is typically assessed based on invasive methods (e.g., biopsy), radiogenomics, combining diagnostic imaging with genomic information can help uncover aggressive (imaging) phenotypes, which in turn can provide non-invasive advice on individualized treatment regimens. In this study, we carried out a parallel analysis on both imaging and transcriptomics data in order to identify features associated with clinically significant PCa (defined as an ISUP grade ≥ 3), subsequently evaluating the correlation between them. Textural imaging features were extracted from multi-parametric MRI sequences (T2W, DWI, and DCE) and combined with DCE-derived parametric pharmacokinetic maps obtained using magnetic resonance dispersion imaging (MRDI). A transcriptomic analysis was performed to derive functional features on transcription factors (TFs), and pathway activity from RNA sequencing data, here referred to as transcriptomic features. For both the imaging and transcriptomic features, different machine learning models were separately trained and optimized to classify tumors in either clinically insignificant or significant PCa. These models were validated in an independent cohort and model performance was used to isolate a subset of relevant imaging and transcriptomic features to be further investigated. A final set of 31 imaging features was correlated to 33 transcriptomic features obtained on the same tumors. Five significant correlations (p < 0.05) were found, of which, three had moderate strength (|r| ≥ 0.5). The strongest significant correlations were seen between a perfusion-based imaging feature-MRDI A median-and the activities of the TFs STAT6 (-0.64) and TFAP2A (-0.50). A higher-order T2W textural feature was also significantly correlated to the activity of the TF STAT6 (-0.58). STAT6 plays an important role in controlling cell proliferation and migration. Loss of the AP2alpha protein expression, quantified by TFAP2A, has been strongly associated with aggressiveness and progression in PCa. According to our findings, a combination of texture features extracted from T2W and DCE, as well as perfusion-based pharmacokinetic features, can be considered for the prediction of clinically significant PCa, with the pharmacokinetic MRDI A feature being the most correlated with the underlying transcriptomic information. These results highlight a link between quantitative imaging features and the underlying transcriptomic landscape of prostate tumors.

Comparing adaptation strategies in MRI-guided online adaptive radiotherapy for prostate cancer: Implications for treatment margins.

PURPOSE: To quantify the difference in accuracy of adapt-to-position (ATP), adapt-to-rotation (ATR) and adapt-to-shape (ATS) workflows used in MRI-guided online adaptive radiotherapy for prostate carcinoma (PCa) by evaluating the margins required to accommodate intra-fraction motion of the clinical target volumes for prostate (CTVpros), prostate including seminal vesicles (CTVpros + sv) and gross tumor volume (GTV). MATERIALS AND METHODS: Clinical delineations of the CTVpros, CTVpros + sv and GTV of 24 patients with intermediate- and high-risk PCa, treated using ATS on a 1.5 T MR-Linac, were used for analysis. Delineations were available pre- and during beam-on. To simulate ATP and ATR workflows, we automatically generated the structures associated with these workflows using rigid transformations from the planning-MRI to the daily online MRIs. Clinical GTVs were analyzed as ATR GTVs and only ATP GTVs were simulated. Planning target volumes (PTVs) were generated with isotropic margins ranging 0.0-5.0 mm. The volumetric overlap was calculated between these PTVs and their corresponding clinical delineation on the MRI acquired during beam-on and averaged over all treatment fractions. RESULTS: The PTV margin required to cover > 95% of the CTVpros was equal (2.5 mm) for all workflows. For the CTVpros + sv, this margin increased to 5.0, 4.0 and 3.5 mm in the ATP, ATR and ATS workflow, respectively. GTV coverage improved from ATP to ATR for margins up to 4.0 mm. CONCLUSION: ATP, ATR and ATS workflows ensure equal coverage of the CTVpros for the current clinical margins. For the CTVpros + sv, ATS showed optimal performance. GTV coverage improves by additional adaptations to prostate rotations.

Deep learning for segmentation of the cervical cancer gross tumor volume on magnetic resonance imaging for brachytherapy.

BACKGROUND: Segmentation of the Gross Tumor Volume (GTV) is a crucial step in the brachytherapy (BT) treatment planning workflow. Currently, radiation oncologists segment the GTV manually, which is time-consuming. The time pressure is particularly critical for BT because during the segmentation process the patient waits immobilized in bed with the applicator in place. Automatic segmentation algorithms can potentially reduce both the clinical workload and the patient burden. Although deep learning based automatic segmentation algorithms have been extensively developed for organs at risk, automatic segmentation of the targets is less common. The aim of this study was to automatically segment the cervical cancer GTV on BT MRI images using a state-of-the-art automatic segmentation framework and assess its performance. METHODS: A cohort of 195 cervical cancer patients treated between August 2012 and December 2021 was retrospectively collected. A total of 524 separate BT fractions were included and the axial T2-weighted (T2w) MRI sequence was used for this project. The 3D nnU-Net was used as the automatic segmentation framework. The automatic segmentations were compared with the manual segmentations used for clinical practice with Sørensen-Dice coefficient (Dice), 95th Hausdorff distance (95th HD) and mean surface distance (MSD). The dosimetric impact was defined as the difference in D98 (ΔD98) and D90 (ΔD90) between the manual segmentations and the automatic segmentations, evaluated using the clinical dose distribution. The performance of the network was also compared separately depending on FIGO stage and on GTV volume. RESULTS: The network achieved a median Dice of 0.73 (interquartile range (IQR) = 0.50-0.80), median 95th HD of 6.8 mm (IQR = 4.2-12.5 mm) and median MSD of 1.4 mm (IQR = 0.90-2.8 mm). The median ΔD90 and ΔD98 were 0.18 Gy (IQR = -1.38-1.19 Gy) and 0.20 Gy (IQR =-1.10-0.95 Gy) respectively. No significant differences in geometric or dosimetric performance were observed between tumors with different FIGO stages, however significantly improved Dice and dosimetric performance was found for larger tumors. CONCLUSIONS: The nnU-Net framework achieved state-of-the-art performance in the segmentation of the cervical cancer GTV on BT MRI images. Reasonable median performance was achieved geometrically and dosimetrically but with high variability among patients.

From once-weekly to semi-weekly whole prostate gland stereotactic radiotherapy with focal boosting: Primary endpoint analysis of the multicenter phase II hypo-FLAME 2.0 trial.

BACKGROUND AND PURPOSE: The hypo-FLAME trial showed that once-weekly (QW) focal boosted prostate stereotactic body radiotherapy (SBRT) is associated with acceptable acute genitourinary (GU) and gastrointestinal (GI) toxicity. Currently, we investigated the safety of reducing the overall treatment time (OTT) of focal boosted prostate SBRT from 29 to 15 days. MATERIAL AND METHODS: Patients with intermediate- and high-risk prostate cancer were treated with SBRT delivering 35 Gy in 5 fractions to the whole prostate gland with an iso-toxic boost up to 50 Gy to the intraprostatic lesion(s) in a semi-weekly (BIW) schedule. The primary endpoint was radiation-induced acute toxicity (CTCAE v5.0). Changes in quality of life (QoL) were examined in terms of proportions achieving a minimal clinically important change (MCIC). Finally, acute toxicity and QoL scores of the BIW schedule were compared with the results of the prior QW hypo-FLAME schedule (n = 100). RESULTS: Between August 2020 and February 2022, 124 patients were enrolled and treated BIW. No grade ≥3 GU or GI toxicity was observed. The 90-days cumulative incidence of grade 2 GU and GI toxicity rates were 47.5% and 7.4%, respectively. Patients treated QW scored significant less grade 2 GU toxicity (34.0%, p = 0.01). No significant differences in acute GI toxicity were observed. Furthermore, patients treated QW had a superior acute bowel and urinary QoL. CONCLUSION: Semi-weekly prostate SBRT with iso-toxic focal boosting is associated with acceptable acute GU and GI toxicity. Based on the comparison between the QW and BIW schedule, patients should be counselled regarding the short-term advantages of a more protracted schedule. Registration number ClinicalTrials.gov: NCT04045717.

Online Adaptive MRI-Guided Radiotherapy for Primary Tumor and Lymph Node Boosting in Rectal Cancer.

The purpose of this study was to characterize the motion and define the required treatment margins of the pathological mesorectal lymph nodes (GTVln) for two online adaptive MRI-guided strategies for sequential boosting. Secondly, we determine the margins required for the primary gross tumor volume (GTVprim). Twenty-eight patients treated on a 1.5T MR-Linac were included in the study. On T2-weighted images for adaptation (MRIadapt) before and verification after irradiation (MRIpost) of five treatment fractions per patient, the GTVln and GTVprim were delineated. With online adaptive MRI-guided radiotherapy, daily plan adaptation can be performed through the use of two different strategies. In an adapt-to-shape (ATS) workflow the interfraction motion is effectively corrected by redelineation and the only relevant motion is intrafraction motion, while in an adapt-to-position (ATP) workflow the margin (for GTVln) is dominated by interfraction motion. The margin required for GTVprim will be identical to the ATS workflow, assuming each fraction would be perfectly matched on GTVprim. The intrafraction motion was calculated between MRIadapt and MRIpost for the GTVln and GTVprim separately. The interfraction motion of the GTVln was calculated with respect to the position of GTVprim, assuming each fraction would be perfectly matched on GTVprim. PTV margins were calculated for each strategy using the Van Herk recipe. For GTVln we randomly sampled the original dataset 20 times, with each subset containing a single randomly selected lymph node for each patient. The resulting margins for ATS ranged between 3 and 4 mm (LR), 3 and 5 mm (CC) and 5 and 6 mm (AP) based on the 20 randomly sampled datasets for GTVln. For ATP, the margins for GTVln were 10-12 mm in LR and AP and 16-19 mm in CC. The margins for ATS for GTVprim were 1.7 mm (LR), 4.7 mm (CC) and 3.2 mm anterior and 5.6 mm posterior. Daily delineation using ATS of both target volumes results in the smallest margins and is therefore recommended for safe dose escalation to the primary tumor and lymph nodes.

CT radiomics models are unable to predict new liver metastasis after successful thermal ablation of colorectal liver metastases.

BACKGROUND: Patients with colorectal liver metastases (CRLM) who undergo thermal ablation are at risk of developing new CRLM after ablation. Identification of these patients might enable individualized treatment. PURPOSE: To investigate whether an existing machine-learning model with radiomics features based on pre-ablation computed tomography (CT) images of patients with colorectal cancer can predict development of new CRLM. MATERIAL AND METHODS: In total, 94 patients with CRLM who were treated with thermal ablation were analyzed. Radiomics features were extracted from the healthy liver parenchyma of CT images in the portal venous phase, before thermal ablation. First, a previously developed radiomics model (Original model) was applied to the entire cohort to predict new CRLM after 6 and 24 months of follow-up. Next, new machine-learning models were developed (Radiomics, Clinical, and Combined), based on radiomics features, clinical features, or a combination of both. RESULTS: The external validation of the Original model reached an area under the curve (AUC) of 0.57 (95% confidence interval [CI]=0.56-0.58) and 0.52 (95% CI=0.51-0.53) for 6 and 24 months of follow-up. The new predictive radiomics models yielded a higher performance at 6 months compared to 24 months. For the prediction of CRLM at 6 months, the Combined model had slightly better performance (AUC=0.60; 95% CI=0.59-0.61) compared to the Radiomics and Clinical models (AUC=0.55-0.57), while all three models had a low performance for the prediction at 24 months (AUC=0.52-0.53). CONCLUSION: Both the Original and newly developed radiomics models were unable to predict new CLRM based on healthy liver parenchyma in patients who will undergo ablation for CRLM.

Visualising spatial heterogeneity in glioblastoma using imaging habitats.

Glioblastoma is a high-grade aggressive neoplasm characterised by significant intra-tumoral spatial heterogeneity. Personalising therapy for this tumour requires non-invasive tools to visualise its heterogeneity to monitor treatment response on a regional level. To date, efforts to characterise glioblastoma's imaging features and heterogeneity have focussed on individual imaging biomarkers, or high-throughput radiomic approaches that consider a vast number of imaging variables across the tumour as a whole. Habitat imaging is a novel approach to cancer imaging that identifies tumour regions or 'habitats' based on shared imaging characteristics, usually defined using multiple imaging biomarkers. Habitat imaging reflects the evolution of imaging biomarkers and offers spatially preserved assessment of tumour physiological processes such perfusion and cellularity. This allows for regional assessment of treatment response to facilitate personalised therapy. In this review, we explore different methodologies to derive imaging habitats in glioblastoma, strategies to overcome its technical challenges, contrast experiences to other cancers, and describe potential clinical applications.

A margin recipe for the management of intra-fraction target motion in radiotherapy.

Background and purpose: Strategies to limit the impact of intra-fraction motion during treatment are common in radiotherapy. Margin recipes, however, are not designed to incorporate these strategies. This work aimed to provide a framework to determine how motion management strategies influence treatment margins. Materials and methods: Two models of intra-fraction motion were considered. In model 1 motion was instantaneous, before treatment starts and in model 2 motion was a continuous drift during treatment. Motion management strategies were modelled by truncating the underlying error distribution at cσ, with σ the standard deviation of the distribution and c a free parameter. Using Monte Carlo simulations, we determined how motion management changed the required margin. The analysis was performed for different number of treatment fractions and different standard deviations of the underlying random and systematic errors. Results: The required margin for a continuous drift was found to be well approximated by an average position of the target at ¾ of the drift. Introducing a truncation at cσ, the relative change in the margin was equal to 0.3c. This result held for both models, was independent of σ or the number of fractions and naturally generalizes to the situation with a residual (systematic) error. Conclusion: Treatment margins can be determined when motion management strategies are applied. Moreover, our analysis can be used to study the potential benefit of different motion management strategies. This allows to discuss and determine the most appropriate strategy for margin reduction.