MR-guided Radiotherapy Research Articles

Treatment of glioblastoma using MRIdian® A3i BrainTx™: Imaging and treatment workflow demonstration.

For patients with newly diagnosed glioblastoma, the current standard-of-care includes maximal safe resection, followed by concurrent chemoradiotherapy and adjuvant temozolomide, with tumor treating fields. Traditionally, diagnostic imaging is performed pre- and post-resection, without additional dedicated longitudinal imaging to evaluate tumor volumes or other treatment-related changes. However, the recent introduction of MR-guided radiotherapy using the ViewRay MRIdian A3i system includes a dedicated BrainTx package to facilitate the treatment of intracranial tumors and provides daily MR images. We present the first reported case of a glioblastoma imaged and treated using this workflow. In this case, a 67-year-old woman underwent adjuvant chemoradiotherapy after gross total resection of a left frontal glioblastoma. The radiotherapy treatment plan consisted of a traditional two-phase design (46 Gy followed by a sequential boost to a total dose of 60 Gy at 2 Gy/fraction). The treatment planning process, institutional workflow, treatment imaging, treatment timelines, and target volume changes visualized during treatment are presented. This case example using our institutional A3i system workflow successfully allows for imaging and treatment of primary brain tumors and has the potential for margin reduction, detection of early disease progression, or to detect the need for dose adaptation due to interfraction tumor volume changes.

Evaluation of Treatment Plan Quality between Magnetic Resonance-Guided Radiotherapy and Volumetric Modulated Arc Therapy for Prostate Cancer

Evaluation of Treatment Plan Quality between Magnetic Resonance-Guided Radiotherapy and Volumetric Modulated Arc Therapy for Prostate Cancer

Dosimetric Evaluation of Synthetic Computed Tomography Technique on Position Variation of Air Cavity in Magnetic Resonance-Guided Radiotherapy

Dosimetric Evaluation of Synthetic Computed Tomography Technique on Position Variation of Air Cavity in Magnetic Resonance-Guided Radiotherapy

Stereotactic MR-Guided Radiotherapy for Adrenal Gland Metastases: First Clinical Results.

Stereotactic MR-guided Radiotherapy (MRgRT) is an interesting treatment option for adrenal gland metastases (AGM). We reviewed data from 12 consecutive patients treated with MRgRT for an AGM in our center between 14 November 2019 and 17 August 2021. Endpoints were tolerance assessment, the impact of adaptive treatment on target volume coverage and organs at risk (OAR) sparing, local control (LC), and overall survival (OS). The majority of patients were oligometastatic (58.3%), with 6 right AGM, 5 left AGM and 1 left and right AGM. The prescribed dose was 35 to 50 Gy in 3 to 5 fractions. The median PTV V95% on the initial plan was 95.74%. The median V95% of the PTVoptimized (PTVopt) on the initial plan was 95.26%. Thirty-eight (69%) fractions were adapted. The PTV coverage was significantly improved for adapted plans compared to predicted plans (median PTV V95% increased from 89.85% to 91.17%, p = 0.0478). The plan adaptation also significantly reduced Dmax for the stomach and small intestine. The treatment was well tolerated with no grade > 2 toxicities. With a median follow-up of 15.5 months, the 1−year LC and OS rate were 100% and 91.7%. Six patients (50%) presented a metastatic progression, and one patient (8.3%) died of metastatic evolution during the follow-up. Adaptation of the treatment plan improved the overall dosimetric quality of MRI-guided radiotherapy. A longer follow-up is required to assess late toxicities and clinical results.

Quality assurance assessment of intra-acquisition diffusion-weighted and T2-weighted magnetic resonance imaging registration and contour propagation for head and neck cancer radiotherapy.

Adequate image registration of anatomical and functional magnetic resonance imaging (MRI) scans is necessary for MR-guided head and neck cancer (HNC) adaptive radiotherapy planning. Despite the quantitative capabilities of diffusion-weighted imaging (DWI) MRI for treatment plan adaptation, geometric distortion remains a considerable limitation. Therefore, we systematically investigated various deformable image registration (DIR) methods to co-register DWI and T2-weighted (T2W) images. We compared three commercial (ADMIRE, Velocity, Raystation) and three open-source (Elastix with default settings [Elastix Default], Elastix with parameter set 23 [Elastix 23], Demons) post-acquisition DIR methods applied to T2W and DWI MRI images acquired during the same imaging session in twenty immobilized HNC patients. In addition, we used the non-registered images (None) as a control comparator. Ground-truth segmentations of radiotherapy structures (tumour and organs at risk) were generated by a physician expert on both image sequences. For each registration approach, structures were propagated from T2W to DWI images. These propagated structures were then compared with ground-truth DWI structures using the Dice similarity coefficient and mean surface distance. 19 left submandibular glands, 18 right submandibular glands, 20left parotid glands, 20 right parotid glands, 20 spinal cords, and 12 tumours were delineated. Most DIR methods took <30s to execute per case, with the exception of Elastix 23 which took ∼458s to execute per case. ADMIRE and Elastix 23 demonstrated improved performance over None for all metrics and structures (Bonferroni-corrected p<0.05), while the other methods did not. Moreover, ADMIRE and Elastix 23 significantly improved performance in individual and pooled analysis compared to all other methods. The ADMIRE DIR method offers improved geometric performance with reasonable execution time so should be favoured for registering T2W and DWI images acquired during the same scan session in HNC patients. These results are important to ensure the appropriate selection of registration strategies for MR-guided radiotherapy.

A pilot study of MRI radiomics for high-risk prostate cancer stratification in 1.5 T MR-guided radiotherapy.

To investigate the potential value of MRI radiomics obtained from a 1.5 T MRI-guided linear accelerator (MR-LINAC) for D'Amico high-risk prostate cancer (PC) classification in MR-guided radiotherapy (MRgRT). One hundred seventy-six consecutive PC patients underwent 1.5 T MRgRT treatment were retrospectively enrolled. Each patient received one or two pretreatment T2 -weighted MRI scans on a 1.5 T MR-LINAC. The endpoint was to differentiate high-risk from low/intermediate-risk PC based on D'Amico criteria using MRI-radiomics. Totally 1023 features were extracted from clinical target volume (CTV) and planning target volume (PTV). Intraclass correlation coefficient of scan-rescan repeatability, feature correlation, and recursive feature elimination were used for feature dimension reduction. Least absolute shrinkage and selection operator regression was employed for model construction. Receiver operating characteristic area under the curve (AUC) analysis was used for model performance assessment in both training and testing data. One hundred and eleven patients fulfilled all criteria were finally included: 76 for training and 35 for testing. The constructed MRI-radiomics models extracted from CTV and PTV achieved the AUC of 0.812 and 0.867 in the training data, without significant difference (P=0.083). The model performances remained in the testing. The sensitivity, specificity, and accuracy were 85.71%, 64.29%, and 77.14% for the PTV-based model; and 71.43%, 71.43%, and 71.43% for the CTV-based model. The corresponding AUCs were 0.718 and 0.750 (P=0.091) for CTV- and PTV-based models. MRI-radiomics obtained from a 1.5 T MR-LINAC showed promising results in D'Amico high-risk PC stratification, potentially helpful for the future PC MRgRT. Prospective studies with larger sample sizes and external validation are warranted for further verification.

Advancing the treatment of localized prostate cancer with MR-guided radiotherapy.

External beam radiotherapy (EBRT) is an important cornerstone in the treatment of localized prostate cancer. Current image-guided radiotherapy (IGRT) techniques allow for more accurate and precise delivery of radiation treatment by the use of imaging before each fraction. Magnetic resonance guided radiotherapy (MRgRT) is the next step in IGRT with hybrid systems combining linear accelerators with MRI-scanners. With MRgRT, it is possible to visualize pelvic anatomy in great detail and subsequently perform replanning of the radiation dose distribution before each radiotherapy fraction. This technique has the potential to increase the therapeutic window of EBRT, by improved normal tissue sparing due to margin reduction and more accurate target dose delivery. This is particularly promising for prostate cancer, with its biology lending itself to ultra-hypofractionation, reducing radiotherapy treatment to as little as five fractions. Also, recent studies have shown that focal dose escalation to the intraprostatic tumor to high ablative doses can substantially increase disease-free survival. In this article, we discuss these unique opportunities as well as the potential future benefits of MRgRT in prostate cancer treatment.

Enzyme-responsive macrocyclic metal complexes for biomedical imaging.

Metal chelator-based contrast agents are used as tumor navigators for cancer diagnosis. Although approved metal chelators show excellent contrast performance in magnetic resonance imaging (MRI), large doses are required for cancer diagnoses due to rapid clearance and nonspecific accumulation throughout the body, which can compromise safety. The present study describes an enzyme-responsive metal delivery system, in which enzyme overexpressed in the tumor microenvironment selectively activates the tumor uptake of gadolinium (Gd). Gd was loaded into enzyme-responsive macrocyclam (ErMC) modified with a PEGylated enzyme-cleavable peptide resulting in Gd@ErMC. The PEGylated shell layer protected Gd@ErMC from nonspecific binding in the blood, increasing the half-life of the contrast agent. Specific cleavage of the PEGylated shell layer by the enzyme selectively liberated Gd from Gd@ErMC at the tumor site. Evaluation of the in vivo distribution of Gd@ErMC in tumor-bearing mice by MRI and positron emission tomography (PET) showed that Gd@ErMC had an extended half-life and was highly specific. Histological and serological analysis of Gd@ErMC-treated mice showed that this agent was safe. This novel enzyme-responsive contrast agent delivery system shows promise as specific theranostic agent for MR-guided radiotherapy.

Stereotactic MR-Guided Adaptive Radiotherapy for Pancreatic Tumors: Updated Results of the Montpellier Prospective Registry Study.

Stereotactic MR-guided Adaptive RadioTherapy (SMART) is a novel process to treat pancreatic tumors. We present an update of the data from our prospective registry of SMART for pancreatic tumors. After the establishment of the SMART indication in a multidisciplinary board, we included all patients treated for pancreatic tumors. Primary endpoints were acute and late toxicities. Secondary endpoints were survival outcomes (local control, overall survival, distant metastasis free survival) and dosimetric advantages of adaptive process on targets volumes and OAR. We included seventy consecutive patients in our cohort between October 2019 and April 2022. The prescribed dose was 50 Gy in 5 consecutive fractions. No severe acute SMART related toxicity was noted. Acute and late Grade ≤ 2 gastro intestinal were low. Daily adaptation significantly improved PTV and GTV coverage as well as OAR sparing. With a median follow-up of 10.8 months since SMART completion, the median OS, 6-months OS, and 1-year OS were 20.9 months, 86.7% (95% CI: (75-93%), and 68.6% (95% CI: (53-80%), respectively, from SMART completion. Local control at 6 months, 1 year, and 2 years were, respectively, 96.8 % (95% CI: 88-99%), 86.5 (95% CI: 68-95%), and 80.7% (95% CI: 59-92%). There was no grade &gt; 2 late toxicities. Locally Advanced Pancreatic Cancers (LAPC) and Borderline Resectable Pancreatic Cancers (BRPC) patients (52 patients) had a median OS, 6-months OS, and 1-year OS from SMART completion of 15.2 months, 84.4% (95% CI: (70-92%)), and 60.5% (95% CI: (42-75%)), respectively. The median OS, 1-year OS, and 2-year OS from initiation of induction chemotherapy were 22.3 months, 91% (95% CI: (78-97%)), and 45.8% (95% CI: (27-63%)), respectively. Twenty patients underwent surgical resection (38.7 % of patients with initially LAPC) with negative margins (R0). To our knowledge, this is the largest series of SMART for pancreatic tumors. The treatment was well tolerated with only low-grade toxicities. Long-term OS and LC rates were achieved. SMART achieved high secondary resection rates in LAPC patients.

A pilot study of same-day MRI-only simulation and treatment with MR-guided adaptive palliative radiotherapy (MAP-RT)

We conducted a prospective pilot study evaluating the feasibility of same day MRI-only simulation and treatment with MRI-guided adaptive palliative radiotherapy (MAP-RT) for urgent palliative indications (NCT#03824366). All (16/16) patients were able to complete 99% of their first on-table attempted fractions, and no grades 3–5 toxicities occurred.

A multi-institutional comparison of acceptance testing data for a 0.35 T MRI scanner

Objective. To present and quantify the variability in the acceptance testing data for the imaging component of the 0.35 T magnetic resonance-linear accelerator (MR-linac). Approach. The current acceptance testing protocol by the MR-linac vendor was described along with the equipment and scanner parameters utilized throughout the process. The Bo field homogeneity, SNR/uniformity of the combined and individual receiver coils, American College of Radiology (ACR) image quality testing, and spatial integrity of the imaging data were collected from twelve different institutions. The variability in the results was accentuated and the ramifications of the results were discussed in the context of MR-guided radiation therapy. Main Results. The Bo field homogeneity was found to have a large gantry dependence with the median values being <4 ppm for all gantry angles. The SNR and uniformity were found to be well above the vendor-specified thresholds with a relatively small institutional-dependence. All institutions passed the ACR image uniformity tests. The largest institutional variability was noted to be for the slice positional accuracy test. The spatial fidelity was calculated to be <1.0 and <2.1 mm within a 100 and a 175 mm radius from the isocenter. Significance. The results from this study can be used to set the tolerances and formal guidelines for MR-linacs imaging quality assurance. Additionally, the multi-institutional data reported in this work will aid in future MR-linac acceptance and commissioning.

Magnetic Resonance-Guided Adaptive Radiation Therapy for Prostate Cancer: The First Results from the MOMENTUM study—An International Registry for the Evidence-Based Introduction of Magnetic Resonance-Guided Adaptive Radiation Therapy

Magnetic resonance (MR)-guided radiation therapy (MRgRT) is a new technique for treatment of localized prostate cancer (PCa). We report the 12-month outcomes for the first PCa patients treated within an international consortium (the MOMENTUM study) on a 1.5T MR-Linac system with ultrahypofractionated radiation therapy. Patients treated with 5×7.25 Gy were identified. Prostate specific antigen-level, physician-reported toxicity (Common Terminology Criteria for Adverse Events [CTCAE]), and patient-reported outcomes (Quality of Life Questionnaire PR25 and Quality of Life Questionnaire C30 questionnaires) were recorded at baseline and at 3, 6, and 12 months of follow-up (FU). Pairwise comparative statistics were conducted to compare outcomes between baseline and FU. The study included 425 patients with localized PCa (11.4% low, 82.0% intermediate, and 6.6% high-risk), and 365, 313, and 186 patients reached 3-, 6-, and 12-months FU, respectively. Median prostate specific antigen level declined significantly to 1.2 ng/mL and 0.1 ng/mL at 12 months FU for the nonandrogen deprivation therapy (ADT) and ADT group, respectively. The peak of genitourinary and gastrointestinal CTCAE toxicity was reported at 3 months FU, with 18.7% and 1.7% grade ≥2, respectively. The QLQ-PR25 questionnaire outcomes showed significant deterioration in urinary domain score at all FU moments, from 8.3 (interquartile range [IQR], 4.1-16.6) at baseline to 12.4 (IQR, 8.3-24.8; P=.005) at 3 months, 12.4 (IQR, 8.3-20.8; P=.018;) at 6 months, and 12.4 (IQR, 8.3-20.8; P=.001) at 12 months. For the non-ADT group, physician- and patient-reported erectile function worsened significantly between baseline and 12 months FU. Ultrahypofractionated MR-guided radiation therapy for localized PCa using a 1.5T MR-Linac is effective and safe. The peak of CTCAE genitourinary and gastrointestinal toxicity was reported at 3 months FU. Furthermore, for patients without ADT, a significant increase in CTCAE erectile dysfunction was reported at 12 months FU. These data are useful for educating patients on expected outcomes and informing study design of future comparative-effectiveness studies.

1.5T MR-Guided Daily-Adaptive SBRT for Prostate Cancer: Preliminary Report of Toxicity and Quality of Life of the First 100 Patients.

Purpose: The present study reports the preliminary outcomes in terms of adverse events and quality of life in the first 100 patients treated with 1.5T MR-guided daily-adaptive stereotactic body radiotherapy for prostate cancer. Methods: From October 2019 to December 2020, 100 patients, enrolled in a prospective study, received MR-guided SBRT for prostate cancer. Rectal spacer insertion was optional and administered in 37 patients. In total, 32 patients received androgen deprivation therapy in accordance with international guidelines. A prospective collection of data regarding toxicity and quality of life was performed. Results: The median age was 71 years (range, 52-84). The median total dose delivered was 35 Gy (35-36.25 Gy) in five sessions, either on alternate days (n = 25) or consecutive days (n = 75). For acute toxicity, we recorded: seven cases of acute G2 urinary pain and four cases of G2 gastrointestinal events. The median follow-up was 12 months (3-20), recording three late G2 urinary events and one G3 case, consisting of a patient who required a TURP 8 months after the treatment. For gastrointestinal toxicity, we observed 3 G ≥ 2 GI events, including one patient who received argon laser therapy for radiation-induced proctitis. Up to the last follow-up, all patients are alive and with no evidence of biochemical relapse, except for an M1 low-volume patient in distant progression two months after radiotherapy. QoL evaluation reported a substantial resolution of any discomfort within the second follow-up after radiotherapy, with the only exception being sexual items. Notably, after one year, global health items were improved compared to the baseline assessment. Conclusions: This study reports very promising outcomes in terms of adverse events and QoL, supporting the role of 1.5T MR-guided SBRT for prostate cancer. To date, this series is one of the first and largest available in the literature. Long-term results are warranted.

High grade glioma radiation therapy on a high field 1.5 Tesla MR-Linac - workflow and initial experience with daily adapt-to-position (ATP) MR guidance: Afirst report.

This study reports the workflow and initial clinical experience of high grade glioma (HGG) radiotherapy on the 1.5 T MR-Linac (MRL), with a focus on the temporal variations of the tumor and feasibility of multi-parametric image (mpMRI) acquisition during routine treatment workflow. Ten HGG patients treated with radiation within the first year of the MRL's clinical operation, between October 2019 and August 2020, were identified from a prospective database. Workflow timings were recorded and online adaptive plans were generated using the Adapt-To-Position (ATP) workflow. Temporal variation within the FLAIR hyperintense region (FHR) was assessed by the relative FHR volumes (n = 281 contours) and migration distances (maximum linear displacement of the volume). Research mpMRIs were acquired on the MRL during radiation and changes in selected functional parameters were investigated within the FHR. All patients completed radiotherapy to a median dose of 60 Gy (range, 54-60 Gy) in 30 fractions (range, 30-33), receiving a total of 287 fractions on the MRL. The mean in-room time per fraction with or without post-beam research imaging was 42.9 minutes (range, 25.0-69.0 minutes) and 37.3 minutes (range, 24.0-51.0 minutes), respectively. Three patients (30%) required re-planning between fractions 9 to 12 due to progression of tumor and/or edema identified on daily MRL imaging. At the 10, 20, and 30-day post-first fraction time points 3, 3, and 4 patients, respectively, had a FHR volume that changed by at least 20% relative to the first fraction. Research mpMRIs were successfully acquired on the MRL. The median apparent diffusion coefficient (ADC) within the FHR and the volumes of FLAIR were significantly correlated when data from all patients and time points were pooled (R=0.68, p<.001). We report the first clinical series of HGG patients treated with radiotherapy on the MRL. The ATP workflow and treatment times were clinically acceptable, and daily online MRL imaging triggered adaptive re-planning for selected patients. Acquisition of mpMRIs was feasible on the MRL during routine treatment workflow. Prospective clinical outcomes data is anticipated from the ongoing UNITED phase 2 trial to further refine the role of MR-guided adaptive radiotherapy.

Patterns of utilization and clinical adoption of 0.35 Tesla MR-guided radiation therapy in the United States – Understanding the transition to adaptive, ultra-hypofractionated treatments

Purpose/ObjectiveMagnetic resonance-guided radiation therapy (MRgRT) utilization is rapidly expanding worldwide, driven by advanced capabilities including continuous intrafraction visualization, automatic triggered beam delivery, and on-table adaptive replanning (oART). Our objective was to describe patterns of 0.35Tesla(T)-MRgRT (MRIdian) utilization in the United States (US) among early adopters of this novel technology. Materials/MethodsAnonymized administrative data from all US MRIdian treatment systems were extracted for patients completing treatment from 2014 to 2020. Detailed treatment information was available for all MRIdian linear accelerator (linac) systems and some cobalt systems. ResultsSeventeen systems at 16 centers delivered 5736 courses and 36,389 fractions (fraction details unavailable for 1223 cobalt courses), of which 21.1% were adapted. Ultra-hypofractionation (UHfx) (1–5 fractions) was used in 70.3% of all courses. At least one adaptive fraction was used for 38.5% of courses (average 1.7 adapted fractions/course), with higher oART use in UHfx dose schedules (47.7% of courses, average 1.9 adapted fractions per course). The most commonly treated organ sites were pancreas (20.7%), liver (16.5%), prostate (12.5%), breast (11.5%), and lung (9.4%). Temporal trends show a compounded annual growth rate (CAGR) of 59.6% in treatment courses delivered, with a dramatic increase in use of UHfx to 84.9% of courses in 2020 and similar increase in use of oART to 51.0% of courses. ConclusionsThis is the first comprehensive study reporting patterns of utilization among early adopters of MRIdian in the US. Intrafraction MR image-guidance, advanced motion management, and increasing adoption of adaptive radiation therapy has led to a substantial transition to ultra-hypofractionated regimens. 0.35 T-MRgRT has been predominantly used to treat abdominal and pelvic tumors with increasing use of on-table adaptive replanning, which represents a paradigm shift in radiation therapy.

Patient-specific daily updated deep learning auto-segmentation for MRI-guided adaptive radiotherapy

Background and purposeDeep Learning (DL) technique has shown great potential but still has limited success in online contouring for MR-guided adaptive radiotherapy (MRgART). This study proposed a patient-specific DL auto-segmentation (DLAS) strategy using the patient’s previous images and contours to update the model and improve segmentation accuracy and efficiency for MRgART. Methods and materialsA prototype model was trained for each patient using the first set of MRI and corresponding contours as inputs. The patient-specific model was updated after each fraction with all the available fractional MRIs/contours, and then used to predict the segmentation for the next fraction. During model training, a variant was fitted under consistency constraints, limiting the differences in the volume, length and centroid between the predictions for the latest MRI within a reasonable range. The model performance was evaluated for both organ-at-risks and tumors auto-segmentation for a total of 6 abdominal/pelvic cases (each with at least 8 sets of MRIs/contours) underwent MRgART through Dice Similarity Coefficient (DSC) and 95% Hausdorff Distance (HD95), and was compared with deformable image registration (DIR) and frozen DL model (no updating after pre-training). The contouring time was also recorded and analyzed. ResultsThe proposed model achieved superior performance with higher mean DSC (0.90, 95 % CI: 0.88–0.95), as compared to DIR (0.63, 95 %CI: 0.59–0.68) and frozen DL models (0.74, 95 % CI: 0.71–0.79). As for tumors, the proposed method yielded a median DSC of 0.95, 95 % CI: 0.94–0.97, and a median HD95 of 1.63 mm, 95 % CI: 1.22 mm-2.06 mm. The contouring time was reduced significantly (p < 0.05) using the proposed method (73.4 ± 6.5 secs) compared to the manual process (12 ∼ 22 mins). The online ART time was reduced to 1650 ± 274 seconds with the proposed method, as compared to 3251.8 ± 447 seconds using the original workflow. ConclusionThe proposed patient-specific DLAS method can significantly improve the segmentation accuracy and efficiency for longitudinal MRIs, thereby facilitating the routine practice of MRgART.

System-dependent image distortion related to gantry positions of a 0.35 T MRgRT: Characterization and the corresponding correction.

MR-guided radiotherapy with high accuracy treatment planning requires addressing MR imaging artifacts that originate from system imperfections. This work presents the characterization and corresponding correction of gantry-related imaging distortions including geometric distortion and isocenter shift in a 0.35 T magnetic resonance imaging (MRI)-guided radiotherapy (MRgRT) system using distortion vector fields (DVFs). Two phantoms, the magnetic resonance imaging distortion in 3D (MRID3D ) phantom and the Fluke phantom, along with a human volunteer were imaged at different gantry angles on a 0.35 T MR-Linac. The geometric distortion and isocenter shift were characterized for both phantom images. DVFs with a field of view extended beyond the physical boundary of the MRID3D phantom were extracted from images taken at 30° gantry angle increments, with vendor-provided distortion correction turned on and off (DstOff). These extended DVFs were then applied to the relevant phantom images to correct their geometric distortions and isocenter shift at the respective gantry angles. The extended DVFs produced from the MRID3D phantom were also applied to Fluke phantom and human MR images at their respective gantry angles. The resampled images were evaluated using structural similarity index measure (SSIM) comparison with the vendor corrected images from the MRgRT system. Geometric distortion with "mean (±SD) distortion" of 3.2±0.02, 2.9±0.02, and 1.8±0.01mm and isocenter shift (±SD) of 0.49±0.3, 0.05±0.2, and 0.01±0.03mm were present in the DstOff MRID3D phantom images in right-left (RL), anterior-posterior (AP), and superior-inferior (SI) directions, respectively. After resampling the originally acquired images by applying extended DVFs, the distortion was corrected to 0.18±0.02, 0.09±0.01, 0.15±0.01mm, and isocenter shift was corrected to 0.14±0.05, -0.02±0.04, and -0.07±0.05mm in RL, AP, and SI directions, respectively. The Fluke phantom average geometric distortion with "mean (±SD) distortion" of 2.7±0.1mm was corrected to 0.2±0. 1mm and the average isocenter shift (±SD) of 0.51±0.2mm, and 0.05±0.03 was corrected to -0.08±0.03mm, and -0.05±0.01 in RL and AP directions, respectively. SSIM (mean±SD) of the original images to resampled images was increased from 0.49±0.02 to 0.78±0.01, 0.45±0.02 to 0.75±0.01, and 0.86±0.25 to 0.98±0.08 for MRID3D phantom, Fluke phantom, and human MR images, respectively, for all the gantry angles compared to the vendor corrected images. The gantry-related MR imaging distortion including geometric distortion and isocenter shift was characterized and a corresponding correction was demonstrated using extended DVFs on 0.35 T MRgRT system. The characterized gantry-related isocenter shift can be combined with geometric distortion correction to provide a technique for the correction of the full system-dependent distortion in an MRgRT system.

Initial clinical applications treating pediatric and adolescent patients using MR-guided radiotherapy.

To demonstrate the clinical applications and feasibility of online adaptive magnetic resonance image guided radiotherapy (MRgRT) in the pediatric, adolescent and young adult (AYA) population. This is a retrospective case series of patients enrolled onto a prospective study. All pediatric (age < 18) and AYA patients (age< 30), treated on the Elekta Unity MR linear accelerator (MRL) from 2019 to 2021 were enrolled onto a prospective registry. Rationale for MRgRT included improved visualization of and alignment to the primary tumor, re-irradiation in a critical area, ability to use smaller margins, and need for daily adaptive replanning to minimize dose to adjacent critical structures. Step-and-shoot intensity-modulated radiation treatment (IMRT) plans were generated for all Unity patients with a dose grid of 3 mm and a statistical uncertainty of < 1% per plan. A total of 15 pediatric and AYA patients have been treated with median age of 13 years (range: 6 mos - 27 yrs). Seven patients were <10 yo. The clinical applications of MRgRT included Wilms tumor with unresectable IVC thrombus (n=1), Ewing sarcoma (primary and metastatic, n=3), recurrent diffuse intrinsic pontine glioma (DIPG, n=2), nasopharyngeal carcinoma (n=1), clival chordoma (n=1), primitive neuroectodermal tumor of the pancreas (n=1), recurrent gluteo-sacral germ cell tumor (n=1), C-spine ependymoma (n=1), and posterior fossa ependymoma (n=1). Two children required general anesthesia. One AYA patient could not complete the MRgRT course due to tumor-related pain exacerbated by longer treatment times. Two AYA patients experienced anxiety related to treatment on the MRL, one of which required daily Ativan. No patient experienced treatment interruptions or unexpected toxicity. MRgRT was well-tolerated by pediatric and AYA patients. There was no increased use of anesthesia outside of our usual practice. Dosimetric advantages were seen for patients with tumors in critical locations such as adjacent to or involving optic structures, stomach, kidney, bowel, and heart.

Patient-specific transfer learning for auto-segmentation in adaptive 0.35 T MRgRT of prostate cancer: a bi-centric evaluation.

Online adaptive radiation therapy (RT) using hybrid magnetic resonance linear accelerators (MR-Linacs) can administer a tailored radiation dose at each treatment fraction. Daily MR imaging followed by organ and target segmentation adjustments allow to capture anatomical changes, improve target volume coverage, and reduce the risk of side effects. The introduction of automatic segmentation techniques could help to further improve the online adaptive workflow by shortening the re-contouring time and reducing intra- and inter-observer variability. In fractionated RT, prior knowledge, such as planning images and manual expert contours, is usually available before irradiation, but not used by current artificial intelligence-based autocontouringapproaches. The goal of this study was to train convolutional neural networks (CNNs) for automatic segmentation of bladder, rectum (organs at risk, OARs), and clinical target volume (CTV) for prostate cancer patients treated at 0.35 T MR-Linacs. Furthermore, we tested the CNNs generalization on data from independent facilities and compared them with the MR-Linac treatment planning system (TPS) propagated structures currently used in clinics. Finally, expert planning delineations were utilized for patient- (PS) and facility-specific (FS) transfer learning to improve auto-segmentation of CTV and OARs on fractionimages. In this study, data from fractionated treatments at 0.35T MR-Linacs were leveraged to develop a 3D U-Net-based automatic segmentation. Cohort C1 had 73 planning images and cohort C2 had 19 planning and 240 fraction images. The baseline models (BMs) were trained solely on C1 planning data using 53 MRIs for training and 10 for validation. To assess their accuracy, the models were tested on three data subsets: (i) 10 C1 planning images not used for training, (ii) 19 C2 planning, and (iii) 240 C2 fraction images. BMs also served as a starting point for FS and PS transfer learning, where the planning images from C2 were used for network parameter fine tuning. The segmentation output of the different trained models was compared against expert ground truth by means of geometric metrics. Moreover, a trained physician graded the network segmentations as well as the segmentations propagated by the clinicalTPS. The BMs showed dice similarity coefficients (DSC) of 0.88(4) and 0.93(3) for the rectum and the bladder, respectively, independent of the facility. CTV segmentation with the BM was the best for intermediate- and high-risk cancer patients from C1 with DSC=0.84(5) and worst for C2 with DSC=0.74(7). The PS transfer learning brought a significant improvement in the CTV segmentation, yielding DSC=0.72(4) for post-prostatectomy and low-risk patients and DSC=0.88(5) for intermediate- and high-risk patients. The FS training did not improve the segmentation accuracy considerably. The physician's assessment of the TPS-propagated versus network-generated structures showed a clear advantage of thelatter. The obtained results showed that the presented segmentation technique has potential to improve automatic segmentation for MR-guidedRT.

P038 - Toxicity patterns and preliminary clinical outcomes of 1.5T MR-guided stereotactic body radiotherapy for prostate cancer

P038 - Toxicity patterns and preliminary clinical outcomes of 1.5T MR-guided stereotactic body radiotherapy for prostate cancer