Preoperative Partial Breast Irradiation in Patients with Low-Risk Breast Cancer: A Systematic Review of Literature.

BACKGROUND: Preoperative instead of standard postoperative partial breast irradiation (PBI) after breast-conserving surgery (BCS) has the advantage of reducing the irradiated breast volume, toxicity, and number of radiotherapy sessions and can allow tumor downstaging. In this review, we assessed tumor response and clinical outcomes after preoperative PBI. PATIENTS AND METHODS: We conducted a systematic review of studies on preoperative PBI in patients with low-risk breast cancer using the databases Ovid Medline, Embase.com, Web of Science (Core Collection), and Scopus (PROSPERO registration CRD42022301435). References of eligible manuscripts were checked for other relevant manuscripts. The primary outcome measure was pathologic complete response (pCR). RESULTS: A total of eight prospective and one retrospective cohort study were identified (n = 359). In up to 42% of the patients, pCR was obtained and this increased after a longer interval between radiotherapy and BCS (0.5-8 months). After a maximum median follow-up of 5.0 years, three studies on external beam radiotherapy reported low local recurrence rates (0-3%) and overall survival of 97-100%. Acute toxicity consisted mainly of grade 1 skin toxicity (0-34%) and seroma (0-31%). Late toxicity was predominantly fibrosis grade 1 (46-100%) and grade 2 (10-11%). Cosmetic outcome was good to excellent in 78-100% of the patients. CONCLUSIONS: Preoperative PBI showed a higher pCR rate after a longer interval between radiotherapy and BCS. Mild late toxicity and good oncological and cosmetic outcomes were reported. In the ongoing ABLATIVE-2 trial, BCS is performed at a longer interval of 12 months after preoperative PBI aiming to achieve a higher pCR rate.

Prediction of pathologic complete response after single-dose MR-guided partial breast irradiation in low-risk breast cancer patients: the ABLATIVE-2 trial-a study protocol.

BACKGROUND: Partial breast irradiation (PBI) is standard of care in low-risk breast cancer patients after breast-conserving surgery (BCS). Pre-operative PBI can result in tumor downstaging and more precise target definition possibly resulting in less treatment-related toxicity. This study aims to assess the pathologic complete response (pCR) rate one year after MR-guided single-dose pre-operative PBI in low-risk breast cancer patients. METHODS: The ABLATIVE-2 trial is a multicenter prospective single-arm trial using single-dose ablative PBI in low-risk breast cancer patients. Patients ≥ 50 years with non-lobular invasive breast cancer ≤ 2 cm, grade 1 or 2, estrogen receptor-positive, HER2-negative, and tumor-negative sentinel node procedure are eligible. A total of 100 patients will be enrolled. PBI treatment planning will be performed using a radiotherapy planning CT and -MRI in treatment position. The treatment delivery will take place on a conventional or MR-guided linear accelerator. The prescribed radiotherapy dose is a single dose of 20 Gy to the tumor, and 15 Gy to the 2 cm of breast tissue surrounding the tumor. Follow-up MRIs, scheduled at baseline, 2 weeks, 3, 6, 9, and 12 months after PBI, are combined with liquid biopsies to identify biomarkers for pCR prediction. BCS will be performed 12 months after radiotherapy or after 6 months, if MRI does not show a radiologic complete response. The primary endpoint is the pCR rate after PBI. Secondary endpoints are radiologic response, toxicity, quality of life, cosmetic outcome, patient distress, oncological outcomes, and the evaluation of biomarkers in liquid biopsies and tumor tissue. Patients will be followed up to 10 years after radiation therapy. DISCUSSION: This trial will investigate the pathological tumor response after pre-operative single-dose PBI after 12 months in patients with low-risk breast cancer. In comparison with previous trial outcomes, a longer interval between PBI and BCS of 12 months is expected to increase the pCR rate of 42% after 6-8 months. In addition, response monitoring using MRI and biomarkers will help to predict pCR. Accurate pCR prediction will allow omission of surgery in future patients. TRIAL REGISTRATION: The trial was registered prospectively on April 28th 2022 at clinicaltrials.gov (NCT05350722).

Investigating the potential of deep learning for patient-specific quality assurance of salivary gland contours using EORTC-1219-DAHANCA-29 clinical trial data.

INTRODUCTION: Manual quality assurance (QA) of radiotherapy contours for clinical trials is time and labor intensive and subject to inter-observer variability. Therefore, we investigated whether deep-learning (DL) can provide an automated solution to salivary gland contour QA. MATERIAL AND METHODS: DL-models were trained to generate contours for parotid (PG) and submandibular glands (SMG). Sørensen-Dice coefficient (SDC) and Hausdorff distance (HD) were used to assess agreement between DL and clinical contours and thresholds were defined to highlight cases as potentially sub-optimal. 3 types of deliberate errors (expansion, contraction and displacement) were gradually applied to a test set, to confirm that SDC and HD were suitable QA metrics. DL-based QA was performed on 62 patients from the EORTC-1219-DAHANCA-29 trial. All highlighted contours were visually inspected. RESULTS: Increasing the magnitude of all 3 types of errors resulted in progressively severe deterioration/increase in average SDC/HD. 19/124 clinical PG contours were highlighted as potentially sub-optimal, of which 5 (26%) were actually deemed clinically sub-optimal. 2/19 non-highlighted contours were false negatives (11%). 15/69 clinical SMG contours were highlighted, with 7 (47%) deemed clinically sub-optimal and 2/15 non-highlighted contours were false negatives (13%). For most incorrectly highlighted contours causes for low agreement could be identified. CONCLUSION: Automated DL-based contour QA is feasible but some visual inspection remains essential. The substantial number of false positives were caused by sub-optimal performance of the DL-model. Improvements to the model will increase the extent of automation and reliability, facilitating the adoption of DL-based contour QA in clinical trials and routine practice.

Accuracy of deformable image registration-based intra-fraction motion management in Magnetic Resonance-guided radiotherapy.

Background and Purpose: Intra-fraction motion management is key in Stereotactic Ablative Radiotherapy (SABR) gated delivery. This study assessed the accuracy of automatic tumor segmentation in the delivery of MR-guided radiotherapy (MRgRT) by comparing it to manual delineations performed by experienced observers. Materials and Methods: Twenty patients previously treated with MR-guided SABR for thoracic and abdominal tumors were included. Five observers with at least two years of experience in MRgRT manually delineated the gross tumor volume (GTV) for 20 patients on 240 frames of a cine MRI on a sagittal plane. Deformable Image Registration (DIR) based GTV contours were propagated using four different algorithms from a reference frame to subsequent frames.Geometrical analysis based on the Dice Similarity Coefficient (DSC), centroid distance and Hausdorff Distance (HDD) were performed to assess the inter-observer variability and the accuracy of automatic segmentation. A Confidence Value (CV) metric for the reliability of the tumor auto-contouring was also calculated. Results: Inter-observer delineation variability resulted in mean DSC of 0.89, HDD of 5.8 mm and centroid distance of 1.7 mm. Tumor auto-contouring by the four DIR algorithms resulted in an excellent agreement with the manual delineations by the experienced observers. Mean DSC for each algorithm across all patients was greater than 0.90, whereas the HDD and centroid distances were below 4.0 mm and 1.5 mm, respectively. The CV showed a strong correlation with the DSC. Conclusions: DIR-based auto-contouring in MRgRT exhibited a high level of agreement with the manual contouring performed by experts, allowing accurate gated delivery.

Single Ultra-High Dose Rate Proton Transmission Beam for Whole Breast FLASH-Irradiation: Quantification of FLASH-Dose and Relation with Beam Parameters.

Healthy tissue-sparing effects of FLASH (≥40 Gy/s, ≥4-8 Gy/fraction) radiotherapy (RT) make it potentially useful for whole breast irradiation (WBI), since there is often a lot of normal tissue within the planning target volume (PTV). We investigated WBI plan quality and determined FLASH-dose for various machine settings using ultra-high dose rate (UHDR) proton transmission beams (TBs). While five-fraction WBI is commonplace, a potential FLASH-effect might facilitate shorter treatments, so hypothetical 2- and 1-fraction schedules were also analyzed. Using one tangential 250 MeV TB delivering 5 × 5.7 Gy, 2 × 9.74 Gy or 1 × 14.32 Gy, we evaluated: (1) spots with equal monitor units (MUs) in a uniform square grid with variable spacing; (2) spot MUs optimized with a minimum MU-threshold; and (3) splitting the optimized TB into two sub-beams: one delivering spots above an MU-threshold, i.e., at UHDRs; the other delivering the remaining spots necessary to improve plan quality. Scenarios 1-3 were planned for a test case, and scenario 3 was also planned for three other patients. Dose rates were calculated using the pencil beam scanning dose rate and the sliding-window dose rate. Various machine parameters were considered: minimum spot irradiation time (minST): 2 ms/1 ms/0.5 ms; maximum nozzle current (maxN): 200 nA/400 nA/800 nA; two gantry-current (GC) techniques: energy-layer and spot-based. For the test case (PTV = 819 cc) we found: (1) a 7 mm grid achieved the best balance between plan quality and FLASH-dose for equal-MU spots; (2) near the target boundary, lower-MU spots are necessary for homogeneity but decrease FLASH-dose; (3) the non-split beam achieved >95% FLASH for favorable (not clinically available) machine parameters (SB GC, low minST, high maxN), but <5% for clinically available settings (EB GC, minST = 2 ms, maxN = 200 nA); and (4) splitting gave better plan quality and higher FLASH-dose (~50%) for available settings. The clinical cases achieved ~50% (PTV = 1047 cc) or >95% (PTV = 477/677 cc) FLASH after splitting. A single UHDR-TB for WBI can achieve acceptable plan quality. Current machine parameters limit FLASH-dose, which can be partially overcome using beam-splitting. WBI FLASH-RT is technically feasible.

Magnetic resonance imaging-guided radiotherapy for intermediate- and high-risk prostate cancer: Trade-off between planning target volume margin and online plan adaption.

Magnetic resonance-guided radiotherapy with daily plan adaptation for intermediate- and high-risk prostate cancer is time and labor intensive. Fifty adapted plans with 3 mm planning target volume (PTV)-margin were compared with non-adapted plans using 3 or 5 mm margins. Adequate (V95% ≥ 95%) prostate coverage was achieved in 49 fractions with 5 mm PTV without plan adaptation, however, coverage of the seminal vesicles (SV) was insufficient in 15 of 50 fractions. There was no insufficient coverage for prostate and SV using plan adaptation with 3 mm. Hence, daily adaptation is recommended to obtain adequate SV-coverage when using 3 mm PTV.

Same-day consultation, simulation and lung Stereotactic Ablative Radiotherapy delivery on a Magnetic Resonance-linac.

Background and Purpose: Magnetic resonance-guided radiotherapy (MRgRT) with real-time intra-fraction tumor motion monitoring allows for high precision Stereotactic Ablative Radiotherapy (SABR). This study aimed to investigate the clinical feasibility, patient satisfaction and delivery accuracy of single-fraction MR-guided SABR in a single day (one-stop-shop, OSS). Methods and Materials: Ten patients with small lung tumors eligible for single fraction treatments were included. The OSS procedure consisted of consultation, treatment simulation, treatment planning and delivery. Following SABR delivery, patients completed a reported experience measure (PREM) questionnaire. Prescribed doses ranged 28-34 Gy. Median GTV was 2.2 cm3 (range 1.3-22.9 cm3). A gating boundary of 3 mm, and PTV margin of 5 mm around the GTV, were used with auto-beam delivery control. Accuracy of SABR delivery was studied by analyzing delivered MR-cines reconstructed from machine log files. Results: All 10 patients completed the OSS procedure in a single day, and all reported satisfaction with the process. Median time for the treatment planning step and the whole procedure were 2.8 h and 6.6 h, respectively. With optimization of the procedure, treatment could be completed in half a day. During beam-on, the 3 mm tracking boundary encompassed between 78.0 and 100 % of the GTV across all patients, with corresponding PTV values being 94.4-100 % (5th-95th percentiles). On average, system-latency for triggering a beam-off event comprised 5.3 % of the delivery time. Latency reduced GTV coverage by an average of -0.3 %. Duty-cycles during treatment delivery ranged from 26.1 to 64.7 %. Conclusions: An OSS procedure with MR-guided SABR for lung cancer led to good patient satisfaction. Gated treatment delivery was highly accurate with little impact of system-latency.

Treatment patterns for adrenal metastases using surgery and SABR during a 10-year period.

We studied treatment patterns for adrenal metastases using surgery or SABR at a single institution during a 10-year period. The number of patients undergoing SABR doubled since 2016, without a change in numbers undergoing surgery. Both treatments resulted in low rates of acute toxicity and similar survivals.

Deep Learning for Automated Elective Lymph Node Level Segmentation for Head and Neck Cancer Radiotherapy.

Depending on the clinical situation, different combinations of lymph node (LN) levels define the elective LN target volume in head-and-neck cancer (HNC) radiotherapy. The accurate auto-contouring of individual LN levels could reduce the burden and variability of manual segmentation and be used regardless of the primary tumor location. We evaluated three deep learning approaches for the segmenting individual LN levels I−V, which were manually contoured on CT scans from 70 HNC patients. The networks were trained and evaluated using five-fold cross-validation and ensemble learning for 60 patients with (1) 3D patch-based UNets, (2) multi-view (MV) voxel classification networks and (3) sequential UNet+MV. The performances were evaluated using Dice similarity coefficients (DSC) for automated and manual segmentations for individual levels, and the planning target volumes were extrapolated from the combined levels I−V and II−IV, both for the cross-validation and for an independent test set of 10 patients. The median DSC were 0.80, 0.66 and 0.82 for UNet, MV and UNet+MV, respectively. Overall, UNet+MV significantly (p < 0.0001) outperformed other arrangements and yielded DSC = 0.87, 0.85, 0.86, 0.82, 0.77, 0.77 for the combined and individual level I−V structures, respectively. Both PTVs were also significantly (p < 0.0001) more accurate with UNet+MV, with DSC = 0.91 and 0.90, respectively. The accurate segmentation of individual LN levels I−V can be achieved using an ensemble of UNets. UNet+MV can further refine this result.

Clinical adoption patterns of 0.35 Tesla MR-guided radiation therapy in Europe and Asia.

BACKGROUND: Magnetic resonance-guided radiotherapy (MRgRT) utilization is rapidly expanding, driven by advanced capabilities including better soft tissue imaging, continuous intrafraction target visualization, automatic triggered beam delivery, and the availability of on-table adaptive replanning. Our objective was to describe patterns of 0.35 Tesla (T)-MRgRT utilization in Europe and Asia among early adopters of this novel technology. METHODS: Anonymized administrative data from all 0.35T-MRgRT treatment systems in Europe and Asia were extracted for patients who completed treatment from 2015 to 2020. Detailed treatment information was analyzed for all MR-linear accelerators (linac) and -cobalt systems. RESULTS: From 2015 through the end of 2020, there were 5796 completed treatment courses delivered in 46,389 individual fractions. 23.5% of fractions were adapted. Ultra-hypofractionated (UHfx) dose schedules (1-5 fractions) were delivered for 63.5% of courses, with 57.8% of UHfx fractions adapted on-table. The most commonly treated tumor types were prostate (23.5%), liver (14.5%), lung (12.3%), pancreas (11.2%), and breast (8.0%), with increasing compound annual growth rates (CAGRs) in numbers of courses from 2015 through 2020 (pancreas: 157.1%; prostate: 120.9%; lung: 136.0%; liver: 134.2%). CONCLUSIONS: This is the first comprehensive study reporting patterns of utilization among early adopters of a 0.35T-MRgRT system in Europe and Asia. Intrafraction MR image-guidance, advanced motion management, and increasing adoption of on-table adaptive RT have accelerated a transition to UHfx regimens. MRgRT has been predominantly used to treat tumors in the upper abdomen, pelvis and lungs, and increasingly with adaptive replanning, which is a radical departure from legacy radiotherapy practices.

Using Spatial Probability Maps to Highlight Potential Inaccuracies in Deep Learning-Based Contours: Facilitating Online Adaptive Radiation Therapy.

PURPOSE: Contouring organs at risk remains a largely manual task, which is time consuming and prone to variation. Deep learning-based delineation (DLD) shows promise both in terms of quality and speed, but it does not yet perform perfectly. Because of that, manual checking of DLD is still recommended. There are currently no commercial tools to focus attention on the areas of greatest uncertainty within a DLD contour. Therefore, we explore the use of spatial probability maps (SPMs) to help efficiency and reproducibility of DLD checking and correction, using the salivary glands as the paradigm. METHODS AND MATERIALS: A 3-dimensional fully convolutional network was trained with 315/264 parotid/submandibular glands. Subsequently, SPMs were created using Monte Carlo dropout (MCD). The method was boosted by placing a Gaussian distribution (GD) over the model's parameters during sampling (MCD + GD). MCD and MCD + GD were quantitatively compared and the SPMs were visually inspected. RESULTS: The addition of the GD appears to increase the method's ability to detect uncertainty. In general, this technique demonstrated uncertainty in areas that (1) have lower contrast, (2) are less consistently contoured by clinicians, and (3) deviate from the anatomic norm. CONCLUSIONS: We believe the integration of uncertainty information into contours made using DLD is an important step in highlighting where a contour may be less reliable. We have shown how SPMs are one way to achieve this and how they may be integrated into the online adaptive radiation therapy workflow.

Markerless Real-Time 3-Dimensional kV Tracking of Lung Tumors During Free Breathing Stereotactic Radiation Therapy.

PURPOSE: Accurate verification of tumor position during irradiation could reduce the probability of target miss. We investigated whether a commercial gantry-mounted 2-dimensional (2D) kilo-voltage (kV) imaging system could be used for real-time 3D tumor tracking during volumetric modulated arc therapy (VMAT) lung stereotactic body radiation therapy (SBRT). Markerless tumor tracking on kV fluoroscopic images was validated using a life-like moving thorax phantom and subsequently performed on kV images continuously acquired before and during free-breathing VMAT lung SBRT. METHODS AND MATERIALS: The 3D-printed/molded phantom containing 3 lung tumors was moved in 3D in TrueBeam developer mode, using simulated regular/irregular breathing patterns. Planar kV images were acquired at 7 frames/s during 11 Gy/fraction 10 MV flattening filter free VMAT. 2D reference templates were created for each gantry angle using the planning 4D computed tomography inspiration phase. kV images and templates were matched using normalized cross correlation to determine 2D tumor position, and triangulation of 2D matched projections determined the third dimension. 3D target tracking performed on cone beam computed tomography projection data from 18 patients (20 tumors) and real-time online tracking data from 2 of the 18 patients who underwent free-breathing VMAT lung SBRT are presented. RESULTS: For target 1 and 2 of the phantom (upper lung and middle/medial lung, mean density -130 Hounsfield units), 3D results within 2 mm of the known position were present in 92% and 96% of the kV projections, respectively. For target 3 (inferior lung, mean density -478 Hounsfield units) this dropped to 80%. Benchmarking against the respiratory signal, 13/20 (65%) tumors (10.5 ± 11.1 cm3) were considered successfully tracked on the cone beam computed tomography data. Tracking was less successful (≤50% of the time) in 7/20 (1.2 ± 1.5 cm3). Successful online tracking during lung SBRT was demonstrated. CONCLUSIONS: 3D markerless tumor tracking on a standard linear accelerator using template matching and triangulation of free-breathing kV fluoroscopic images was possible in 65% of small lung tumors. The smallest tumors were most challenging.

Magnetic Resonance-guided Stereotactic Radiotherapy for Localized Prostate Cancer: Final Results on Patient-reported Outcomes of a Prospective Phase 2 Study.

BACKGROUND: The recent introduction of magnetic resonance-guided radiation therapy (MRgRT) has allowed improved treatment planning and delivery of stereotactic body radiotherapy (SBRT) in prostate cancer (PC). The health-related quality of life (HRQoL) outcomes using this novel approach are important in shared decision making for patients. OBJECTIVE: To report HRQoL using both patient- and clinician-reported outcomes at 1 yr following stereotactic MRgRT for patients with localized PC. DESIGN, SETTING, AND PARTICIPANTS: A prospective phase 2 trial included 101 patients with localized PC. INTERVENTION: All patients received 36.25Gy in five fractions of MRgRT delivered within 2 wk. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: HRQoL was prospectively assessed at baseline, at the last fraction, at 6 wk, and at 3, 6, 9, and 12 mo after treatment, by patient-reported outcome measures using European Organization for Research and Treatment of Cancer QLQ-C30 and QLQ-PR25 questionnaires, and International Prostate Symptom Score. At the same time points, clinicians reported on symptomatic adverse events (AEs). Effect sizes for changes in HRQoL were calculated with repeated measures analysis of variance. RESULTS AND LIMITATIONS: Availability of HRQoL data exceeded 95% at all study time points. From both questionnaires and the recorded AEs, the largest treatment effects on urinary and bowel symptoms were recorded in the first 6 wk of follow-up. Thereafter, all symptoms decreased and returned to baseline values at 12 mo. No grade ≥3 toxicity was reported. No patient reported any relevant limitation due to urinary symptoms, and only 2.2% of patients reported a relevant impact on daily activities due to bowel problems at 1 yr. The majority of patients had intermediate- or high-risk PC for which androgen deprivation therapy (83.2%) was prescribed, thereby precluding study of MRgRT on sexual function. Longer follow-up is awaited in order to evaluate the oncological outcome. CONCLUSIONS: Delivery of MRgRT for SBRT resulted in low toxicity at 1 yr. PATIENT SUMMARY: All patients completed magnetic resonance-guided radiation therapy, which was well tolerated with only transient early urinary and bowel symptoms, which resolved 1 yr after treatment, as confirmed by patient-reported outcome measures.

Impact of daily plan adaptation on organ-at-risk normal tissue complication probability for adrenal lesions undergoing stereotactic ablative radiation therapy.

INTRODUCTION: Stereotactic ablative radiotherapy (SABR) can achieve good local control for metastatic adrenal lesions. Magnetic resonance (MR)-guidance with daily on-table plan adaptation can augment the delivery of SABR with greater dose certainty. The goal of this study was to quantify the potential clinical benefit MR-guided daily-adaptive adrenal SABR using the normal tissue complication probability (NTCP) framework. METHODS: Patients treated with adrenal MR-guided SABR at a single institution were retrospectively reviewed. Lyman-Kutcher-Burman NTCP models were used to calculate the NTCP of upper abdominal organs-at-risk (OARs) at simulation and both before and after daily on-table plan adaptation. Differences in OAR NTCPs were assessed using signed-rank tests. Potential predictors of the benefits of adaptation were assessed by linear regression. RESULTS: Fifty-two adrenal MR-guided SABR courses were analyzed. The baseline simulation plan underestimated the absolute stomach NTCP by 10.0% on average (95% confidence interval: 4.7-15.2%, p < 0.001). Daily on-table adaptation lowered absolute NTCP by 8.7% (4.2-13.2%, p < 0.001). The most significant predictor of the benefits of adaptation was lesion laterality (p = 0.018), with left-sided lesions benefitting more (13.3% [6.3-20.4%], p < 0.001) than right-sided lesions (2.1% [-1.6-5.7%], p = 0.25). Sensitivity analyses did not change the statistical significance of the findings. CONCLUSION: NTCP analysis revealed that patients with left adrenal tumors were more likely to benefit from MR-guided daily on-table adaptive SABR using current dose/fractionation regimens due to reductions in predicted gastric toxicity. Right-sided adrenal lesions may be considered for dose escalation due to low predicted NTCP.

Renal atrophy following gated delivery of stereotactic ablative radiotherapy to adrenal metastases.

Stereotactic ablative radiotherapy (SABR) planning for adrenal metastases aims to minimize doses to the adjacent kidney. Renal dose constraints for SABR delivery are not well defined. In 20 patients who underwent MR-guided breath-hold SABR in five daily fractions of 8-10 Gy, ipsilateral renal volumes receiving ≥20 Gy best correlated with loss of renal volumes, with median renal volume reduction being 6% (range: 3%-11%, 10th-90th percentiles). Organ function did not deteriorate in 18 patients, who had post treatment renal function tests available. This suggests that the ipsilateral renal volume receiving 20 Gy can be used as partial organ dose constraint for SABR to targets in the upper abdomen.

Strategies to improve deep learning-based salivary gland segmentation.

BACKGROUND: Deep learning-based delineation of organs-at-risk for radiotherapy purposes has been investigated to reduce the time-intensiveness and inter-/intra-observer variability associated with manual delineation. We systematically evaluated ways to improve the performance and reliability of deep learning for organ-at-risk segmentation, with the salivary glands as the paradigm. Improving deep learning performance is clinically relevant with applications ranging from the initial contouring process, to on-line adaptive radiotherapy. METHODS: Various experiments were designed: increasing the amount of training data (1) with original images, (2) with traditional data augmentation and (3) with domain-specific data augmentation; (4) the influence of data quality was tested by comparing training/testing on clinical versus curated contours, (5) the effect of using several custom cost functions was explored, and (6) patient-specific Hounsfield unit windowing was applied during inference; lastly, (7) the effect of model ensembles was analyzed. Model performance was measured with geometric parameters and model reliability with those parameters' variance. RESULTS: A positive effect was observed from increasing the (1) training set size, (2/3) data augmentation, (6) patient-specific Hounsfield unit windowing and (7) model ensembles. The effects of the strategies on performance diminished when the base model performance was already 'high'. The effect of combining all beneficial strategies was an increase in average Sørensen-Dice coefficient of about 4% and 3% and a decrease in standard deviation of about 1% and 1% for the submandibular and parotid gland, respectively. CONCLUSIONS: A subset of the strategies that were investigated provided a positive effect on model performance and reliability. The clinical impact of such strategies would be an expected reduction in post-segmentation editing, which facilitates the adoption of deep learning for autonomous automated salivary gland segmentation.

Bringing FLASH to the Clinic: Treatment Planning Considerations for Ultrahigh Dose-Rate Proton Beams.

PURPOSE: Preclinical research into ultrahigh dose rate (eg, ≥40 Gy/s) "FLASH"-radiation therapy suggests a decrease in side effects compared with conventional irradiation while maintaining tumor control. When FLASH is delivered using a scanning proton beam, tissue becomes subject to a spatially dependent range of dose rates. This study systematically investigates dose rate distributions and delivery times for proton FLASH plans using stereotactic lung irradiation as the paradigm. METHODS AND MATERIALS: Stereotactic lung radiation therapy FLASH-plans, using 244 MeV scanning proton transmission beams, with the Bragg peak behind the body, were made for 7 patients. Evaluated parameters were dose rate distribution within a beam, overall irradiation time, number of times tissue is irradiated, and quality of the FLASH-plans compared with the clinical volumetric-modulated arc therapy (VMAT) plans. RESULTS: Sparing of lungs, thoracic wall, and heart in the FLASH-plans was equal to or better than that in the VMAT-plans. For a spot peak dose rate (SPDR, the dose rate in the middle of the spot) of 100 Gy/s, ∼40% of dose is delivered at FLASH dose rates, and for SPDR = 360 Gy/s this increased to ∼75%. One-hundred percent FLASH dose rate cannot be achieved owing to small contributions from distant spots with lower dose rates. The total irradiation time varied between 300 to 730 ms, and around 85% of the dose-receiving body volume was irradiated by either 1 or 2 beams. CONCLUSIONS: Clinical implementation of FLASH using scanning proton beams requires multiple treatment planning considerations: dosimetric, temporal, and spatial parameters all seem important. The FLASH efficiency of a scanning proton beam increases with SPDR. The methodology proposed in this proof-of-principle study provides a framework for evaluating the FLASH characteristics of scanning proton beam plans and can be adapted as FLASH parameters are better defined. It currently seems logical to optimize plans for the shortest delivery time, maximum amount of high dose rate coverage, and maximum amount of single beam and continuous irradiation.

Effect of COVID-19 pandemic on practice in European radiation oncology centers.

ESTRO surveyed European radiation oncology department heads to evaluate the impact of COVID-19. Telemedicine was used in 78% of the departments, and 60% reported a decline in patient volume. Use of protective measures was implemented on a large scale, but shortages of personal protective equipment were present in more than half of the departments.

Clinical Outcomes of Stereotactic MR-Guided Adaptive Radiation Therapy for High-Risk Lung Tumors.

PURPOSE: Magnetic resonance (MR)-guided SABR was performed for patients with lung tumors in whom treatment delivery was challenging owing to tumor location, motion, or pulmonary comorbidity. Because stereotactic MR-guided adaptive radiation therapy (SMART) is a novel approach, we studied clinical outcomes in these high-risk lung tumors. METHODS AND MATERIALS: Fifty consecutive patients (54 lung tumors) underwent SMART between 2016 and 2018 for either a primary lung cancer (29 patients) or for lung metastases (21 patients). Eligible patients had risk factors that could predispose them to toxicity, including a central tumor location (n = 30), previous thoracic radiation therapy (n = 17), and interstitial lung disease (n = 7). A daily 17-second breath-hold MR scan was acquired in treatment position, and on-table plan adaptation was performed using the anatomy of the day. Gated SABR was delivered during repeated breath-holds under continuous MR guidance. RESULTS: All but 1 patient completed the planned SMART schedule. With daily plan adaptation, a biologically effective dose ≥100 Gy to 95% of the planning target volume was delivered in 50 tumors (93%). Median follow-up was 21.7 months (95% confidence interval, 19.9-28.1). Local control and overall and disease-free survival rates at 12 months were 95.6%, 88.0%, and 63.6%, respectively. Local failures developed in 4 patients: in 2 after reirradiation for a recurrent lung cancer and in 2 patients with a colorectal metastasis. Overall rates of any grade ≥2 and ≥3 toxicity were 30% and 8%, respectively. Commonest toxicities were grade ≥2 radiation pneumonitis (12%) and chest wall pain (8%). No grade 4 or 5 toxicities were observed. CONCLUSIONS: Use of MR-guided SABR resulted in low rates of high-grade toxicity and encouraging early local control in a cohort of high-risk lung tumors. Additional studies are needed to identify patients who are most likely to benefit from the SMART approach.

Stereotactic MR-guided adaptive radiation therapy for peripheral lung tumors.

BACKGROUND AND PURPOSE: We studied the benefits of using stereotactic MR-guided adaptive radiation therapy (SMART) for delivery of SABR in peripherally located lung tumors. METHODS AND MATERIALS: Twenty-three patients (25 peripheral lung tumors) underwent SMART in 3-8 fractions on an MR Linac or Cobalt-60 system. Before each fraction, a breath-hold MR scan was acquired, followed by on-table plan adaptation based on the anatomy-of-the-day. Breath-hold gated delivery was performed under continuous MR-guidance using an in-room monitor. Benefits of on-table adaptation were studied by comparing 112 «predicted¼ plans, which are the baseline plans recalculated on the anatomy-of-the-day, with the on-table reoptimized plans. RESULTS: The full SMART procedure took a median of 48 and 62 minutes on the MR Linac and Cobalt-60 system, respectively. Median SMART-PTVs were 9.5 cm3 (range, 3.1-55.6). In 14 patients who had undergone a free-breathing 4DCT, SMART-PTVs measured 53.7% (range, 31.9-75.0) of PTVs that would have been generated using a motion-encompassing internal target volume approach. On-table adaptation improved prescription dose coverage of the PTV from a median of 92.1% in predicted plans, to 95.0% in reoptimized ones, thereby increasing the proportion of fractions delivering ≥100 Gy (BED10Gy) to 95% of PTV, from 90.2% to 100.0%. CONCLUSION: Delivery of gated breath-hold SABR using MR-guidance resulted in significantly smaller target volumes than would have been the case with an ITV-based approach. Although on-table adaptation ensured delivery of ablative doses in all fractions, the dosimetric benefits were modest, suggesting that daily online plan adaptation may not benefit most patients with peripheral lung tumors.