MR-guided Adaptive Radiotherapy Research Articles

Analysis of online plan adaptation for 1.5T magnetic resonance-guided stereotactic body radiotherapy (MRgSBRT) of prostate cancer.

PurposeTo analyze and characterize the online plan adaptation of 1.5T magnetic resonance-guided stereotactic body radiotherapy (MRgSBRT) of prostate cancer (PC).MethodsPC patients (n = 107) who received adaptive 1.5 Tesla MRgSBRT were included. Online plan adaptation was implemented by either the adapt-to-position (ATP) or adapt-to-shape (ATS) methods. Patients were assigned to the ATS group if they underwent ≥ 1 ATS fraction (n = 51); the remainder were assigned to the ATP group (n = 56). The online plan adaptation records of 535 (107 × 5) fractions were retrospectively reviewed. Rationales for ATS decision-making were determined and analyzed using predefined criteria. Statistics of ATS fractions were summarized. Associations of patient characteristics and clinical factors with ATS utilization were investigated.ResultsThere were 87 (16.3%) ATS fractions and 448 ATP fractions (83.7%). The numbers of ATS adoptions in fractions 1–5 were 29 (29/107, 27.1%), 18 (16.8%), 15 (14.0%), 16 (15.0%), and 9 (8.4%), respectively, with significant differences in adoption frequency between fractions (p = 0.007). Other baseline patient characteristics and clinical factors were not significantly associated with ATS classification (all p > 0.05). Underlying criteria for the determination of ATS implementation comprised anatomical changes (77 fractions in 50 patients) and discrete multiple targets (15 fractions in 3 patients). No ATS utilization was determined using dosimetric or online quality assurance criteria.ConclusionsThis study contributes to facilitating the establishment of a standardized protocol for online MR-guided adaptive radiotherapy in PC.

A0962 - MR-guided adaptive versus CT-guided stereotactic radiotherapy for prostate cancer: Where is the best cost-benefit balance?

A0962 - MR-guided adaptive versus CT-guided stereotactic radiotherapy for prostate cancer: Where is the best cost-benefit balance?

Magnetic resonance linear accelerator technology and adaptive radiation therapy: An overview for clinicians.

Radiation therapy (RT) continues to play an important role in the treatment of cancer. Adaptive RT (ART) is a novel method through which RT treatments are evolving. With the ART approach, computed tomography or magnetic resonance (MR) images are obtained as part of the treatment delivery process. This enables the adaptation of the irradiated volume to account for changes in organ and/or tumor position, movement, size, or shape that may occur over the course of treatment. The advantages and challenges of ART maybe somewhat abstract to oncologists and clinicians outside of the specialty of radiation oncology. ART is positioned to affect many different types of cancer. There is a wide spectrum of hypothesized benefits, from small toxicity improvements to meaningful gains in overall survival. The use and application of this novel technology should be understood by the oncologic community at large, such that it can be appropriately contextualized within the landscape of cancer therapies. Likewise, the need to test these advances is pressing. MR-guided ART (MRgART) is an emerging, extended modality of ART that expands upon and further advances the capabilities of ART. MRgART presents unique opportunities to iteratively improve adaptive image guidance. However, although the MRgART adaptive process advances ART to previously unattained levels, it can be more expensive, time-consuming, and complex. In this review, the authors present an overview for clinicians describing the process of ART and specifically MRgART.

Master protocol trial design for technical feasibility of MR-guided radiotherapy

The master protocol trial design aims to increase efficiency in terms of trial infrastructure and protocol administration which may accelerate development of (technical) innovations in radiation oncology. A master protocol to study feasibility of techniques/software for MR-guided adaptive radiotherapy with the MR-Linac is described and discussed.

MR-Guided Adaptive Radiotherapy for Head and Neck Cancer: Prospective Evaluation of Migration and Anatomical Changes of the Major Salivary Glands.

Simple SummaryDuring radiotherapy of head and neck cancer patients, the radiation dose for the major salivary glands (parotids and submandibular glands) should be kept as low as possible to reduce toxicity risks. Nevertheless, volume changes and positional shifts of salivary glands during treatment may result in an increased dose, so adapting the treatment plan is recommended. While previous studies primarily used one or two CTs or CBCTs for imaging during treatment, frequent MRI allows for improved soft tissue contrast and more exact contour adaptation without an additional imaging dose. In this study, twelve patients were treated with MRI-guided radiotherapy with weekly offline adaptations. Significant parotid- and submandibular gland shrinkage, as well as a medial shift of the parotids, were observed. These results support the rationale of MR-guided radiotherapy for head and neck cancer, emphasizing the relevance of future clinical evaluation of toxicity to optimize the benefits of MRI-guided radiotherapy.The aim of this study was to quantify anatomical changes of parotids and submandibular glands and evaluate potential dosimetric advantages during weekly adaptive MR-guided radiotherapy (MRgRT) for the definitive treatment of head and neck cancer (HNC). The data and plans of 12 patients treated with bilateral intensity-modulated radiotherapy for HNC using MR-linac, with weekly offline adaptations, were prospectively evaluated. The positional and volumetric changes of the salivary glands were analyzed by manual segmentation in weekly MRI images and the dosimetric impact of these anatomical changes on the adapted treatment plans was assessed. The mean volume change in parotid and submandibular gland volume was −31.9% (p < 0.0001) and −29.7% (p < 0.0001) after five weeks, respectively. The volume change was significantly correlated with the cumulative dose for the respective gland at the time of volume measurement. Inter-parotid distance changed by −5.4% (6.5 mm) on average after five weeks (p = 0.0005). The distance became significantly smaller only in the left-right direction. The inter-submandibular gland distance changed by 0.7 mm (p = 0.38). This study demonstrated significant changes in salivary gland volumes and position following daily MR guidance and weekly plan adaptation. Ongoing clinical trials will provide data on the clinical impact of these changes and novel MR-based adaptation strategies.

A Deep Learning-Based Automatic Contour Quality Assurance Pipeline for Complex Anatomy on MRI

Despite the recent significant improvement on deep learning (DL)-based auto-segmentation, review and editing on auto-segmented contours are often needed to avoid errors. Such a manual time-consuming process is not desirable, particularly for recently introduced MR-guided adaptive radiation therapy (MRgART). To eliminate this manual process, we have previously proposed automatic contour quality assurance (ACQA) methods to automatically verify contour accuracy and have demonstrated that the ACQA methods worked well for most of organs, except for complex anatomy (e.g., stomach, bowels). This study aims to develop a DL-based ACQA pipeline specifically for complex organs that have inherent anatomy variations and intensity heterogeneity on MRIs.The proposed pipeline includes the following major parts: 1) image pre-processing consisting of bias correction, normalization, intensity clipping, gas/high intensity contents filling and image cropping, 2)generation of three-channel RGB images by combining 30% core, inner shell and 4mm outer shell, 3) data augmentation through randomized 3D and 2D rotation, scaling and translation, 4) abnormal detection using statistical geometrical features extracted from ground truth contours (geo-check), and 5) training of a convolutional neural network (CNN) image classification model using samples passing geo-check. The DL auto-segmented contours of stomach on T2-weighted abdominal MRIs from 54 patients were utilized (40 for training and five-fold cross validation and 14 for independent test). A slice was labeled as inaccurate if the contour on the slice has mean distance to agreement (MDA) ≥2 mm or 95% Hausdorff distance (HD) ≥10 mm as compared to the ground truth. The labels were then manually checked to avoid mislabeling. The model performance was evaluated using area under curve (AUC) from receiver operating characteristic curve (ROC), predicting accuracy and F1 score.A total of 139,230 sample images were generated for model training and 4826 for testing. The percentage of accurate slices identified were 64% and 72%, respectively, for the training and testing datasets. The average AUC/accuracy/F1 were 0.94/0.88/0.94 for cross validation, and 0.9/0.85/0.9 and 0.92/0.88/0.92 for the DL without or with geo-check over all testing cases. For each testing patient, the accuracy/F1 score were in the range of 0.78-0.98/ 0.71-0.99 with a median value of 0.89/0.91. The execution time for a testing contour was less than 1 second on a I7 CPU and a GTX 1060 GPU.The proposed DL-ACQA method can quickly and automatically identify accurate and inaccurate contour slices after auto-segmentation of complex anatomy, e.g., stomach. With further development, this method may be integrated into current workflow, facilitating efficient and accurate segmentation in routine treatment planning or MRgART.

Deep Learning Auto-Segmentation on Multi-Sequence MRI for MR-Guided Adaptive Radiation Therapy

<h3>Purpose/Objective(s)</h3> Robust auto-segmentation based on daily MRI is highly desirable for MR-guided adaptive radiation therapy (MRgART). However, it is generally challenging particularly if the daily MRI is acquired with multiple sequences or if it is for imaging complex anatomy, e.g., abdomen. To overcome these issues, we developed a deep learning algorithm to segment abdominal organs quickly and automatically on multi-sequence MRI. <h3>Materials/Methods</h3> The deep learning auto-segmentation model was developed using a modified U-net, ResUnet3D, with end-to-end mapping, and was trained with 85 daily MRI sets acquired during MRgART on a 1.5T MR-Linac for 37 patients with abdominal tumors. The daily MRI sets were motion averaged images derived from 4D MRI acquired with multiple sequences, including Fast Field Echo (FFE), Balanced Turbo Field Echo (BTFE) and TFE. Each daily MRI was paired with a CT. All MRIs were pre-processed using bias field correction and intensity standardization. The ground truth contours were created manually based on existing guidelines. For the training datasets, field-of-view extension via online random 3D elastic transform/deformation was utilized to augment and improve the quality of datasets. During the model training, a patient-based z-score normalization method was employed to standardize the multi-contrast images. Two models were trained, one for 11 organs: aorta, duodenum, kidneys, liver, pancreas, spinal cord, spleen, stomach, large bowel, and small bowel, and another for 10 organs with combined large and small bowel. Performance of the obtained models was evaluated using 11 test datasets in terms of dice similarity coefficient (DSC), mean distance to agreement (MDA) and time efficiency using an Intel Xeon CPU (2.4 GHz, 64GB RAM) hardware. <h3>Results</h3> For the challenging bowels, the auto-segmentation accuracy for the model trained with the combined bowel was better, with MDA/DSC of 2.1/0.86 versus 6.8/0.76 (large bowel) and 6.1/0.68 (small bowel) for the separate-bowel model. Average DSC for the obtained contours of aorta, left kidney, right kidney, liver, spinal cord, spleen, and stomach was 0.93, 0.93, 0.94, 0.94, 0.83, 0.92 and 0.88, respectively from both models. While no significant difference in model performance was observed with BTFE versus TFE testing datasets, a slightly improved segmentation was seen on the BTFE images for pancreas and duodenum, DSC 0.71 and 0.72. This improvement may be attributed to the enhanced contrast seen in BTFE images when compared to the TFE and FFE images. The time required to generate all organs on an MRI was 20-24 seconds. <h3>Conclusion</h3> We have developed deep learning auto-segmentation models based on multi-sequence MRI sets for abdominal organs. The models performed reasonably well. With further development using large datasets, the models may be integrated in MRgART for fast, fully automated, and accurate contouring of abdominal organs on daily MRI.

Guided Deep Learning Auto-Correction of Suboptimal Auto-Segmentation of Complex Structures for MRI-Guided Adaptive Radiotherapy

Fast and accurate auto-segmentation is desirable for MR-guided adaptive radiotherapy (MRgART), however is challenging particularly for complex structures, e.g., bowels. Anticipating that auto-segmentation, particularly deep learning (DL)-based, will be substantially improved in the near future, the auto-segmentation for these complex structures would become acceptable or suboptimal, implying a need for correcting such suboptimal contours. This study aims to develop a guided DL method to automatically and quickly correct for the suboptimal auto-segmented bowel contours for MRgART of abdominal tumors.Auto-segmentation contours of both small and large bowels created by standard deep convolutional neural network models on 46 T2-MRI sets were used as initial contours to demonstrate the feasibility of the proposed method. As preprocessing, MRI images were standardized (bias corrected, contrast enhanced and normalized) and cropped into multiple 2D subregions using the initial auto-segmented bowel contours, with each including a bowel loop. A total of 1320 suboptimal initial subregion contours with Dice similarity coefficients (DSC) in the range of 0.8∼0.9, as compared to the ground truth (manually delineated contours), were selected to train and test a U-net model for contour auto-correction. The model was trained using 75% of the dataset with two input channels: 1) the pre-processed MRI; 2) initial suboptimal contours used as a guidance as these contours were close to the ground truth. Data augmentation (rotation, flipping and scaling) was performed during the training, with the training objective of minimizing the binary cross entropy loss between the U-net output and ground truth. The obtained U-net auto-correction model was tested using the rest 25% of the dataset (325 subregions from 11 scans) and its performance was measured using DSC and 95% Hausdorff distance (95% HD).The U-net model performance is summarized in the table below. The model effectively corrected for the suboptimal bowel contours with more than 1/3 contours reaching DCS > 0.9 after correction. The execution time to correct for a testing contour was less than 1 second on a GTX 1060 GPU.A U-net model was developed to automatically and quickly correct for suboptimal auto-segmented bowel contours. With further development using large datasets, the U-net model could be integrated in conjunction with an auto-segmentation pipeline to improve the accuracy of suboptimal contours for complex anatomy, minimizing the labor-intensive and time-consuming manual editing and accelerating the process of MRgART.

On the Development of a Robust Deep Learning Synthetic CT Generation Method for MR-Guided Adaptive Radiation Therapy of Abdominal Tumors

Rapid and accurate generation of synthetic CT (sCT) from daily MRI, required in MR-guided adaptive radiotherapy (MRgART), is challenging in abdomen due to the air volumes that can change quickly and randomly (thus, no paired CT available) and hard to automatically segment on daily MRI. This work aims to develop a novel structure-preservation deep learning method to quickly create sCT from a special MRI sequence.The sCT model was based on the generative adversarial networks (GANs) technology with innovations including extra deformable layers in sub-networks and mutual information loss terms, which were added to effectively guide the network to preserve true structures from MRI for those highly deformed organs and air pockets in abdomen. The model was developed to use air scan, a specially designed MRI sequence to image air in abdomen, which a 3D FLASH Cartesian sequence that had optimized RF pulses to achieve a minimum echo time of 1 msec with heavy acceleration to keep scan times under 9 seconds to avoid motion artifacts. Daily air scans acquired along with daily plan MRI on a 1.5T MR-Linac during MRgART for 21 patients with abdominal tumors were used to create the sCT model (sCT-DL). The sCT-DL was compared with the results from: (1) sCT-DIR, generated via deformable image registration (DIR) of reference CT and daily plan MRI, and (i) sCT-Gdiff, a previously reported method to automatically segment air volumes on daily MR by a threshold in a union of (i) deformed air-containing organs (e.g., bowels) and (ii) a expansion to account for DIR inaccuracy (calculated by taking the root mean square error between primary and deformed secondary images, divided by the gradient of primary). In addition, the air volumes on sCT-DL obtained with a threshold of HU < -300 and on sCT-Gdiff were compared to those manually delineated on the air scans (Air-manual) based on Dice similarity coefficient (DSC). Dose calculated for a MR-Linac plan on sCT-DL was compared to those calculated on sCT-DIR and sCT-Gdiff. Dosimetric accuracy was measured using the fractional volume with dose disagreement < 3% (FV3). The bone volumes from sCT-DL were compared to those on sCT-DIR (ground truth) based on DSC and FV3.The sCT-DL creation was very fast (< 1.0 sec with a hardware of 28 CPUs and P2000 GPU). The air volume DSC was 0.49 ± 0.1 for sCT-DL and 0.89 ± 0.06 for sCT-Gdiff, as compared to the Air-manual volumes. For dose accuracy, the FV3 was 0.86 ± 0.03 for sCT-DL versus 0.90 ± 0.01 for sCT-Gdiff. For the bone volumes on sCT-DL, DSC was 0.54 ± 0.04, and FV3 was 0.87 ± 0.01 as compared to the ground truth of sCT-DIR.It is promising to use the proposed novel structure-preservation deep learning method to automatically generated sCT in abdomen based on this proof-of-principle study. The sCT can be generated within 10 sec including the special MRI acquisition. With further development using large datasets, the novel sCT method may be implemented for MRgART of abdominal tumors.

Establishing the Aortic Maximum Tolerated Dose (MTD) for Ablative 5-Fraction Stereotactic MR-Guided Adaptive Radiation Therapy (SMART)

Stereotactic MR-guided adaptive radiotherapy (SMART) can facilitate the safe delivery of ablative dose (BED > 100 Gy10) for inoperable pancreas cancer while sparing gastrointestinal (GI) OARs. Initially restricted to small volume pancreatic targets, SMART has now been expanded to provide ablative elective nodal irradiation (ENI). With this evolution, the proximity of the aorta to the ENI region becomes a concern. Given that the aortic maximum tolerated dose (MTD) for an ultra-hypofractioned course is poorly understood, our working hypothesis is that the MTD for aortic toxicities within the first year after a 5 fraction SMART course is > 52.5 Gy (Dmax). We tested this on our in-house clinical datasets.Twenty-nine inoperable pancreatic cancer patients were treated with 50 Gy in 5 fractions daily (QD) SMART course from April 2019 to March 2020. All patients underwent adaptive radiotherapy in which the dose was minimized to GI OARs, but no intentional aortic sparing was undertaken. For this study, the cumulative dose from all 5 SMART fractions was summed onto a TRUFI MR fractional scan (i.e., fraction 1). To this end, a rigid transformation was performed from each fractional scan (fraction 2-5) to fraction 1 frame of reference (FOR) based on alignment of aorta adjacent to the PTV. Fractional dose was transformed to fraction 1 FOR followed by summation. A contouring ring (2 cm in SI, 3 cm in AP/LR from PTV) was delineated. The aorta was segmented within the contouring ring. The aortic wall (AW) was delineated based on reported variable thickness of 1-2.5 mm in the literature on TRUFI MR sequence, and denoted as AW#mm.The median follow-up of patients was 17.4 months from SMART. 90% of patients (27 of 29) received ENI, typically including a 5-10 mm margin around celiac axis and superior mesenteric artery. Using the following maximum aortic dose parameter (Dmax/D0.03cc/D0.05cc/D1cc), the respective mean ± SD values in Gy were 56.7 ± 6.1/55.1 ± 6.6/49.3 ± 7.5/44.2 ± 8.3 (AW1mm), 56.7 ± 6.1/55.2 ± 6.6/51.1 ± 7.6/48.0 ± 8.0 (AW1.63mm), 56.9 ± 6.0/55.5 ± 6.4/51.8 ± 7.3/49.2 ± 7.6 (AW2mm), 56.8 ± 6.2/55.4 ± 6.6/51.8 ± 7.5/49.6 ± 7.8 (AW2.5mm). For D5cc/D10cc the respective mean ± SD dose in Gy was 13.5 ± 8.1/3.1 ± NA (AW1mm), 28.0 ± 8.0/8.1 ± 7.0 (AW1.63mm), 33.0 ± 8.6/14.4 ± 10.2 (AW2mm), 35.9 ± 9.3/21.3 ± 9.6 (AW2.5mm). No grade 3 or higher toxicities were observed during this follow-up period. 83% of patients (24 of 29) had Dmax > 52.5 Gy; 48% of patients (14 of 29) had D1cc > 47.5 Gy; 48% of patients (14 of 29) had D0.5cc > 50.4 Gy for all AW thicknesses. 72% (21 of 29), 59% (17 of 30) and 28% (8 of 29) of patients had aorta Dmax greater than 53.5 Gy, 55 Gy, and 60 Gy, respectively.An aorta Dmax of 53.5 Gy in 5 fractions was found to be safe for pancreatic cancer patients receiving daily SMART (n = 21 patients)-an increase from the RTOG 0813 MTD limit (i.e., 52.5 Gy). For ablative stereotactic pancreatic patients with inclusion of ENI, aorta dose should be tracked for further understanding of the aortic MTD.K.E. Mittauer: Research Grant; ViewRay Inc. Honoraria; ViewRay Inc. Travel Expenses; ViewRay Inc. Partnership; MR Guidance LLC.R. Herera: None. T. Romaguera: None. D. Alvarez: None. M.P. Mehta: None. A. Gutierrez: None. M.D. Chuong: Research Grant; ViewRay, AstraZeneca. Honoraria; ViewRay, Sirtex, Accuray. Speaker's Bureau; ViewRay, Sirtex, Accuray. Advisory Board; ViewRay, Advanced Accelerator Applications. Travel Expenses; ViewRay, AstraZeneca, Accuray; Particle Therapy Cooperative Group. Manages the operations and activities within the proton center; Miami Cancer.

Indications for Plan Adaptation in Patients with Adrenal Metastases Treated With Stereotactic MR-Guided Adaptive Radiation Therapy (SMART)

Stereotactic body radiation therapy (SBRT) is increasingly utilized in patients with oligometastatic and oligoprogressive disease; however, concerns about interfractional changes in gastrointestinal (GI) anatomy limit broad adoption and dose-escalation in the context of adrenal metastases. The objective of this study is to evaluate the need and indications for daily adaptive replanning in this patient population.Consecutive patients with adrenal metastases treated using respiratory-gated stereotactic magnetic resonance (MR)-guided adaptive radiation therapy (SMART) at a single institution were evaluated. Breath-hold 3D-MRI was acquired prior to each fraction to visualize the anatomy-of-the-day. To quantify the indication for adaptation, dose prediction (i.e., calculation of initial plan on the anatomy of the day) was performed. If adaptation was indicated (i.e., critical organ at risk [OAR] dose constraints were exceeded and/or desired target coverage was not achieved), the plan was reoptimized using the original beam parameters with adjustment of individual beam fluence. Comparisons between the predictive and adaptive dose were performed to evaluate adaptive replanning decision-making based on pre-defined institutional criteria. Dose differences between the predictive and adaptive doses were compared by paired sample t test.16 patients with adrenal metastases underwent a total of 80 SMART fractions to a median dose of 50 Gy in 5 fractions. The median PTV was 58.9 cc (Range: 16.1 - 377.4 cc). Plan adaptation was performed in 15/16 patients at least once during the treatment course due to predefined GI OAR dose constraints being exceeded, after accounting for the anatomy of the day. The indications for adaptation were to improve target coverage in 19/80 fractions (24%), or reduce dose to at least one critical GI OAR in 33 fractions (41%), and for both conditions in 9 fractions (11%). The most commonly affected GI OAR for plan adaptation was the stomach for left-sided metastases (6/7 patients, 86 %) and duodenum for right-sided metastases (3/9 patients, 33 %). Plan adaptation allowed for dose-escalation to the GTV with statistically significant increase in the minimum dose (42 Gy vs. 34 Gy, P < 0.001). No patients experienced acute or late grade ≥3 CTCAE v4.0 treatment-related toxicities.SMART enabled dose-escalation with simultaneous OAR sparing in response to changes in the anatomy-of-the-day for patients with adrenal metastases. SMART using an ablative prescription dose should be explored in future prospective studies for targets in proximity to historically dose-limiting GI OARs.

How to Treat Double Synchronous Abdominal Metastases With Stereotactic MR-Guided Adaptive Radiation Therapy (SMART)?

With the progress of systemic therapies, the opportunity to treat synchronous multiple abdominal lesions of unresectable oligometastatic patients is increasing. With a limited number of fractions and high dose per fraction, stereotactic radiotherapy (SBRT) is a treatment of choice in many indications. Magnetic resonance image-guided radiation therapy (MRgRT) offers a satisfying image quality and contrast especially for abdominal diseases. Indeed, this technique makes possible a daily adaptative radiotherapy (ART) in order to manage inter-fraction motion and live tumor visualization and gating to take into account intra-fraction motion. This work presents our experience in the management of SMART for double synchronous abdominal lesions treatment.Three patients have received a SMART for double synchronous abdominal lesions on our MR-Linac. Patient 1: a 43-years-old woman with two liver metastases from breast cancer. Patient 2: a 44-years-old man with two pancreatic metastases from leiomyosarcoma. Patient 3: a 45-years-old man with one liver metastasis and the pancreatic head primitive cancer. The proximity of the lesions was from 3.5 to 6.5 cm. The prescribed dose was 30 to 35 Gy in 5 fractions for pancreatic lesions and 40 to 60 Gy in 5 or 6 fractions for liver lesions. Patients have been selected according to their ability to hold and reproduce apneas. Each lesion has been treated in breath hold position on alternate days. A step-and-shoot intensity modulated radiation therapy (IMRT) technique has been initially planned for each lesion including an optimization volume permitting to reduce as low as possible the contribution to the second lesion. Each lesion has been gated on a unique sagittal slice with limited margin in the direction of the second lesion.For each patient, the treatment has been carried out properly in terms of comfort, respect of the limit of dose contribution to the second lesion and volume detection for the gating of the tumor. For patient 1, 2 and 3, no acute toxicity neither in-field progression has been reported, respectively 8, 7 and 6 months after the end of the treatment.Thanks to SMART, we have been able to define a feasible and robust methodology permitting double synchronous abdominal metastases treatment. The three patients have successfully followed up the treatment without identified comfort or technical issues. After minimum 6 months, no toxicity nor progression have been reported.M. Rouffiac: None. C. Chevalier: None. D. Thibouw: None. M. Quivrin: None. K. Peignaux-Casasnovas: None. G. Truc: None. L. Aubignac: None. J. Boudet: None. A. Petitfils: None. I. Bessieres: None.

Cumulative Target Volume Dose and Locoregional Failure in Pancreatic Cancer Patients With Treated With Ablative Stereotactic MR-Guided Adaptive Radiation Therapy (SMART)

<h3>Purpose/Objective(s)</h3> Emerging data suggest that ablative stereotactic magnetic resonance-guided adaptive radiation therapy (SMART) for inoperable pancreas cancer (PCa) achieves favorable local control (LC) and overall survival (OS) compared to non-ablative dose. In the current SMART workflow, each adaptive fraction is reoptimized without accounting for previously delivered dose. Cumulative target volume (TV) dose from all SMART fractions has not previously been evaluated with respect to clinical outcomes. <h3>Materials/Methods</h3> We evaluated 44 inoperable PCa patients who were prescribed 50 Gy in 5 daily fractions on a 0.35T-MR LINAC. 33 patients (75%) were treated with elective nodal irradiation, typically including the celiac axis and superior mesenteric artery. PTV margin was 3 mm. On-table adaptive replanning was performed to primarily satisfy gastrointestinal (GI) organ at risk (OAR) constraints, and secondarily maximize TV coverage. Cumulative dose for all fractions was summed onto a TRUFI MR fractional scan (i.e., fraction 5). A rigid transformation was performed for fractions 1 to 4 scans onto the fraction 5 frame of reference (FOR) based on alignment of gross disease. Each fractional dose was transformed onto fraction 5 FOR and summed cumulatively for all 5 fractions. Locoregional failure (LRF) was delineated on follow-up imaging by a diagnostic radiologist with PCa expertise. Cumulative TV dose metrics (D80, D90, mean, max) were evaluated with respect to LRF. Diagnostic scans were rigidly registered to the fraction 5 FOR based on recurrence region to investigate recurrence characteristics and target dose metrics. A Spearman correlation was used to investigate LRF and (1) ENI vs. no ENI, (2) TV coverage metrics, (3) planning TV (PTV) volume. <h3>Results</h3> The median PTV volume for all patients, with ENI, and without ENI was 125.9 cc (range: 20.6-356.8), 137.3 cc (range: 72.8-356.8, and 55.4 (range: 20.6-280.15), respectively. LRF occurred in 12 patients (27.3%); the median LRF volume was 10.9 cc (range: 1.1-145.5cc). The median Dice similarity coefficient taken to characterize LRF overlap with PTV was 0.05 (range: 0-0.45). Moderate negative correlation coefficients were noted between recurrence to ENI [-0.59 (<i>P</i> < 0.001)], recurrence to PTV volume [-0.41 (<i>P</i> < 0.001)] and recurrence to max PTV dose [–0.47 (<i>P</i> < 0.001)]. No correlation was observed for cumulative D80, D90, mean dose to recurrence. <h3>Conclusion</h3> To our knowledge, this is the first analysis of cumulative TV dose from 5-fraction SMART for PCa. While use of ENI is controversial, moderate and statistically significant correlations were noted between LRF and ENI, in addition to PTV volume and maximum dose. Additional investigation is warranted to better understand whether specific dose thresholds exist for gross disease and electively treated volumes with respect to long-term LC and OS.

Long-Term Multi-Institutional Outcomes of 5-Fraction Ablative Stereotactic MR-Guided Adaptive Radiation Therapy (SMART) for Inoperable Pancreas Cancer With Median Prescribed Biologically Effective Dose of 100 Gy10

Randomized trials have shown improved local control (LC) but no overall survival (OS) benefit with the addition of non-ablative radiation therapy (RT) dose compared to chemotherapy (CT) alone for pancreas cancer (PCa). Emerging data suggest that dose-escalated RT may improve LC and OS. A few studies suggest that stereotactic magnetic resonance-guided adaptive RT (SMART) can facilitate the safe delivery of ablative dose for inoperable PCa, although long-term outcomes are not well understood.Inoperable PCa patients who received SMART were identified from the RSSearch Registry. Patients with < 3 months (mo.) follow-up after SMART were excluded. LC, progression free survival (PFS), and OS were estimated using the Kaplan-Meier method. LC was evaluated according to RECIST 1.1 criteria. Acute toxicity was considered within 90 days of SMART and evaluated by CTCAE v4 criteria.A total of 148 PCa patients were treated on a 0.35T MR LINAC across 3 institutions between 2018-2020. Median age was 68 years (range 47-91), and 73.6% had ECOG 0-1 performance status. Patients had locally advanced (57.4%), borderline resectable (29.1%), or medically inoperable (13.5%) disease. Median CA19-9 at diagnosis was 202.1 U/mL (range 0.9-21,281). Induction CT was delivered to 89.2% for a median 3.9 mo (range 0.2-11.3); FOLFIRINOX (52.7%) or gemcitabine/nab-paclitaxel (23.4%) were common. Median prescribed RT dose was 50 Gy (range 40-50) in 5 fractions, mostly in consecutive days (96.6%) and in breath hold (95.3%). Median biologically effective dose (BED10) was 100 Gy10. All patients were treated with real-time tissue tracking and automated beam gating without fiducial markers. An elective target volume was rarely used (25%). Pancreaticoduodenectomy was performed in 23% at a median 46 days (range 34-304) after SMART. Median follow-up was 16 mo. from diagnosis for all patients (range, 4-39). Median, 1-year, and 2-year LC was not reached (NR), 94.6%, and 83%, respectively. Median, 1-year, and 2-year PFS was 18 mo., 72%, and 35.9%, respectively. Median, 1-year, and 2-year OS was 26 mo., 82%, and 52.7%, respectively. Acute and late grade 3 toxicity possibly related to SMART occurred in 4.1% and 12.8%, respectively. There was no reported grade 4+ toxicity.To our knowledge, this is the largest reported analysis of 5-fraction SMART for inoperable PCa. These data add to the evidence that ablative radiation dose may improve long-term outcomes including OS. Prospective evaluation of this novel approach is warranted with attention directed at optimizing patient selection, understanding the clinical significance of cumulative dose delivered across all adapted fractions, and assessing treatment response after the delivery of ablative dose.

Impact of Stereotactic MR-Guided Adaptive Radiation Therapy on Early Clinical and Dosimetric Outcomes in Patients With Pancreatic Cancer

Dose escalated stereotactic body radiotherapy (SBRT) may improve both local control and overall survival for patients with pancreatic cancer. However, escalated dose cannot be routinely offered because of the risk of injury to adjacent normal organs. Stereotactic magnetic resonance (MR)-guided adaptive radiation therapy (SMART) represents an innovation that may achieve safer delivery of ablative doses and improve outcomes. The goal of the present study is to identify specific dosimetric gains that can be achieved with MR-guided adaptive pancreatic SBRT and describe acute clinical outcomes.We performed a single institution IRB approved retrospective analysis of 37 consecutive pancreatic cancer patients treated with 40 Gy in 5 fractions with SMART on a MR-linear accelerator using a light inspiration breath hold. The adaptive workflow starts with an "original" plan which is then adapted daily based on anatomical variation for each fraction. "Predicted" plan dosimetry is generated for each fraction by re-calculating the original plan on the re-contoured pre-treatment anatomy, estimating what would have been delivered without adaptive capabilities. A "re-optimized" plan is developed based on the anatomy of the day to improve target coverage and meet critical organ-at-risk (OAR) metrics, including stomach, duodenum, large bowel and small bowel V33 < 1cc. Acute GI toxicity rates were assessed according to CTCAE v5.The median age of our cohort was 71 (range 35-87) with 23/37 (62%) patients having locally advanced disease and 14/37 (38%) having borderline resectable disease. Adaptive planning was used in all cases with a prescription goal of 40 Gy in 5 fractions prescribed to 90% of the target volume. The median CTV was 80.5 cc (range, 10-180cc) and the PTV was a 3mm expansion. Adaptive planning was required in 156/185 fractions to meet either the stomach (63.8%), duodenum (33.5%), large bowel (12.4%) or small bowel (11.4%) V33 < 1cc metric. Of the plans adapted for OAR sparing, 90 plans required adaptation for one gastrointestinal organ, 50 plans for two, 10 plans for three and 1 plan for four. For predicted plans exceeding the V33 < 1cc metric, the median V33 for duodenum was 2.475cc (range: 1.01-10.68cc), stomach was 2.67cc (range: 1.01-8.21cc), small bowel was 2.71cc (range: 1.14-9.54cc) and large bowel was 1.93cc (range: 1.05-8.36cc). All re-optimized plans met the V33 < 1cc metric. No local failure or acute grade 3 toxicities were observed with a median follow-up of 6 months (range 3-15).Online adaptive re-optimization was necessary for the majority of fractions and improved plan dosimetry was achieved specifically for duodenum, stomach, large and small bowel, achieving low rates of acute toxicity. The clinical benefits of MR-guided adaptive pancreatic SBRT for dose escalation warrant further investigation.

Dosimetric and Clinical Evaluation of Stereotactic MR-Guided Online Adaptive Radiation Therapy (SMART) for Treatment of Adrenal Metastases

Stereotactic MR-guided Adaptive Radiotherapy (SMART) is an appealing treatment option for irradiation of intra-abdominal tumors. Daily adaptation with SMART has the advantage of dose escalation while protecting organs at risk (OARs). We aimed to document dosimetric parameters during SMART and analyze local control for adrenal metastases.Fifteen patients treated between September 2018 and October 2020 for adrenal metastases were included in the study. The institutional constraints were used to evaluate the doses to OARs and target volumes. For each fraction, an adaptive plan was generated following registration and delineation of the daily disposition. A comparison was done between predicted plans (original plan on the daily anatomy) and adapted plans (new plan after re-optimization accounting the new contours) in terms of target volumes coverage and OARs doses. Patients were treated every other day.The median age was 63 years (range 36-74 years) and 73.3% of patients were male. The median delivered total dose was 50 Gy (range 35-50 Gy); with a median fraction number of 5 (range 3-5 fractions) and the median fraction dose was 10 Gy (range 7-15 Gy), corresponding to a median BED10 of 100 Gy (range 60-113 Gy). Nine patients presented with a left-sided lesion, five with right-sided and one patient with both left and right-sided lesions. All of the initial plans met target and OARs constraints. Among 70 fractions 56 fractions (80%) were re-optimized. The dosimetric findings for PTV and OARs on predicted and re-optimized fractions were compared. PTV V100 (90.9% vs 94.3%) and PTV D98 (40.8 Gy vs 44.5 Gy) were better with re-optimization and the differences were statistically significant (P < 0.05). The OARs doses for 0.5 cc, 1 cc, 5 cc and 10 cc were compared. The right and left sided lesions were analyzed separately. For the left sided metastases there was OARs dose violation for stomach in one patient and for bowel in one patient. For the right sided metastases there was OARs dose violation for duodenum in two patients and for bowel in two patients. After re-optimization all dosimetric parameters met the institutional constraints. Local control rate was 93.8% with a median follow up time of 11.7 months (range 4.5-26.4 months). One patient with progressive disease underwent adrenalectomy. All treatments were well tolerated. The most common radiation induced acute toxicities were grade 1-2 fatigue and nausea. There were no grade 3 or higher acute or late toxicities experienced during the treatment and follow-up course.The use of re-optimization with SMART improved the PTV coverage and OARs doses for treatment of adrenal metastases. SMART has some advantages when compared to other radiotherapy techniques due to its better visualization of soft tissue, real-time tumor tracking without fiducial markers and potentially reduced toxicity to OARs. These results need to be confirmed with more patients and longer follow-up.

PO-1208 Stereotactic MR-guided online adaptive radiation therapy for management of pancreatic malignancies

PO-1208 Stereotactic MR-guided online adaptive radiation therapy for management of pancreatic malignancies

PO-1499 1.5T MR-guided daily adaptive radiotherapy: preliminary clinical report of the first 2000 fractions.

PO-1499 1.5T MR-guided daily adaptive radiotherapy: preliminary clinical report of the first 2000 fractions.

PO-1375 Feasibility of neurovascular sparing MR-guided adaptive radiotherapy for prostate cancer

PO-1375 Feasibility of neurovascular sparing MR-guided adaptive radiotherapy for prostate cancer

PH-0501 Plan Adaptation in Patients with Abdominopelvic Tumors Treated with Stereotactic MR-Guided Adaptive Radiotherapy

PH-0501 Plan Adaptation in Patients with Abdominopelvic Tumors Treated with Stereotactic MR-Guided Adaptive Radiotherapy