Daily Adaptive Radiotherapy Research Articles

Deep learning-based dose prediction for magnetic resonance-guided prostate radiotherapy.

Daily adaptive radiotherapy, as performed with the Elekta Unity MR-Linac, requires choosing between different adaptation methods, namely ATP (Adapt to Position) and ATS (Adapt to Shape), where the latter requires daily re-contouring to obtain a dose plan tailored to the daily anatomy. These steps are inherently resource-intensive, and quickly predicting the dose distribution and the dosimetric evaluation criteria while the patient is on the table could facilitate a fast selection of adaptation method and decrease the treatment times. In this work, we aimed to develop a deep-learning-based dose-prediction pipeline for prostate MR-Linac treatments. Two hundred twelve MR-images, structure sets, and dose distributions from 35 prostate patients treated with 6.1Gy for 7 or 6 fractions at our MR-Linac were included, split into train/test partitions of 152/60 images, respectively. A deep-learning segmentation network was trained to segment the CTV (prostate), bladder, and rectum. A second network was trained to predict the dose distribution based on manually delineated structures. At inference, the predicted segmentations acted as input to the dose prediction network, and the predicted dose was compared to the true (optimized in the treatment planning system) dose distribution. Median DSC values from the segmentation network were 0.90/0.94/0.87 for CTV/bladder/rectum. Predicted segmentations as input to the dose prediction resulted in mean differences between predicted and true doses of 0.7%/0.7%/1.7% (relative to the prescription dose) for D98%/D95%/D2% for the CTV. For the bladder, the difference was 0.7%/0.3% for Dmean/D2% and for the rectum 0.1/0.2/0.2 pp (percentage points) for V33Gy/V38Gy/V41Gy. In comparison, true segmentations as input resulted in differences of 1.1%/0.9%/1.6% for CTV, 0.5%/0.4% for bladder, and 0.7/0.4/0.3 pp for the rectum. Only D2% for CTV and Dmean/D2% for bladder were found to be statistically significantly better when using true structures instead of predicted structures as input to the dose prediction. Small differences in the fulfillment of clinical dose-volume constraints are seen between utilizing deep-learning predicted structures as input to a dose prediction network and manual structures. Overall mean differences<2% indicate that the dose-prediction pipeline is useful as a decision support tool where differences are>2%.

Stereotactic Magnetic Resonance–Guided Daily Adaptive Radiation Therapy for Localized Prostate Cancer: Acute and Late Patient-Reported Toxicity Outcomes

ObjectiveTo report acute and late bowel, urinary and sexual dysfunction patient reported outcome measures (PROMS), amongst localised prostate cancer patients who underwent Stereotactic MRI-guided daily adaptive radiotherapy (SMART). MethodsAll patients who completed a baseline 12 item Patient-Reported Outcomes Version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE), before undergoing 36.25 Gy in 5 fractions by SMART, were subsequently followed up with the same graded questionnaire at set time points. Latest PSA levels were recorded. The percentage of patients who reported no change from their baseline AE or reported a new ≥ ‘frequent’ or’ severe’ AE grade during follow up was calculated. The maximum PRO-CTCAE grade for each item was recorded for each patient. The percentage of toxicity levels for each separate AE item at set time points were calculated. ResultsThe total number of patients was 69 with a median follow-up of 27 months. Median age of cohort 73 (range 54-85). The median pre-treatment PSA, T stage and Gleason score was 7.5 mmol/l (4.5-32 mmol/l), T2b (T2-T3b) and Gleason Score of 7 (3+4, range: 6-9) respectively. No patient had biochemical failure during follow up. Regarding bowel symptoms, >80% of men reported no change from baseline toxicity during follow up. New ≥ frequent diarrhoea was reported in 9% of patients. ‘Almost constant’ diarrhoea peaked at 1 month but was absent at >33 months. Regarding urinary symptoms, increased urinary urgency was the most common complaint (39%). 20% of men reported new ≥ frequent urinary urgency incidence peaking at 1 month but absent at >33 months. New “severe” sexual dysfunction was seen in 26% of patients and was persistent at >33 months. ConclusionOur study is one the largest patient reported outcomes study post prostate SMART. It shows acceptable levels of toxicity even up to 2 years post treatment.

Uncertainty estimation and evaluation of deformation image registration based convolutional neural networks

Objective. Fast and accurate deformable image registration (DIR), including DIR uncertainty estimation, is essential for safe and reliable clinical deployment. While recent deep learning models have shown promise in predicting DIR with its uncertainty, challenges persist in proper uncertainty evaluation and hyperparameter optimization for these methods. This work aims to develop and evaluate a model that can perform fast DIR and predict its uncertainty in seconds. Approach. This study introduces a novel probabilistic multi-resolution image registration model utilizing convolutional neural networks to estimate a multivariate normal distributed dense displacement field (DDF) in a multimodal image registration problem. To assess the quality of the DDF distribution predicted by the model, we propose a new metric based on the Kullback–Leibler divergence. The performance of our approach was evaluated against three other DIR algorithms (VoxelMorph, Monte Carlo dropout, and Monte Carlo B-spline) capable of predicting uncertainty. The evaluation of the models included not only the quality of the deformation but also the reliability of the estimated uncertainty. Our application investigated the registration of a treatment planning computed tomography (CT) to follow-up cone beam CT for daily adaptive radiotherapy. Main results. The hyperparameter tuning of the models showed a trade-off between the estimated uncertainty’s reliability and the deformation’s accuracy. In the optimal trade-off, our model excelled in contour propagation and uncertainty estimation (p <0.05) compared to existing uncertainty estimation models. We obtained an average dice similarity coefficient of 0.89 and a KL-divergence of 0.15. Significance. By addressing challenges in DIR uncertainty estimation and evaluation, our work showed that both the DIR and its uncertainty can be reliably predicted, paving the way for safe deployment in a clinical environment.

Empirical analysis of a plan-of-the-day strategy to approximate daily online reoptimization for prostate CBCT-guided adaptive radiotherapy.

Adaptive radiotherapy (ART) can improve the dose delivered to the patient in the presence of anatomic variations. However, the required time, effort, and clinical resources are intensive. This work analyzed a plan-of-the-day (POD) approach on clinical patients treated with online ART to explore implementations that balance dosimetric benefit and clinical resource cost. Eight patients treated to the prostate and proximal seminal vesicles with 26 fractions of CBCT-guided, daily online ART were retrospectively analyzed. With a plan library composed of daily adaptive plans from the initial week of treatment and the original plan, the effect of a POD approach starting the following week was investigated by simulating use of these previously generated plans under 3- and 6-degree-of-freedom patient alignment. The plan selected for each treatment was that from the library that maximized the Dice similarity coefficient of the clinical target volume with that of the current treatment fraction. The resulting distribution of several target coverage and organ-at-risk dose metrics are described relative to those achieved with the daily online reoptimized adaptive technique. The values of target coverage and organ-at-risk dose metrics varied across patients and metrics. The POD schemas closely approximated the reference values from a fully reoptimized adaptive plan yet required less than 20% of the reoptimization effort. The POD schemas also had a much greater effect on target coverage metrics than 6-degree-of-freedom registration did. Organ-at-risk dose metrics also varied considerably across patients but did not exhibit a consistent dependence on the particular schema. POD schemas were able to achieve the vast majority of the dosimetric benefit of daily online ART with a small fraction of the online reoptimization effort. Strategies like this might allow for more practical and strategic implementation of ART so as to benefit a greater number of patients.

Acute Toxicity and Efficiency Outcomes in the DARTBOARD Randomized Trial of Daily Adaptive Radiotherapy for Head and Neck Squamous Cell Carcinoma

Acute Toxicity and Efficiency Outcomes in the DARTBOARD Randomized Trial of Daily Adaptive Radiotherapy for Head and Neck Squamous Cell Carcinoma

Daily Adaptive vs. Non-Adaptive External Beam Radiation Therapy with Concurrent Chemotherapy for Locally Advanced Non-Small Cell Lung Cancer (NSCLC): A Prospective Randomized Trial of an Individualized Approach for Toxicity Reduction (ARTIA-Lung)

Radiotherapy (RT) with concurrent chemotherapy is a standard treatment for Stage III NSCLC. However, radiation pneumonitis incidence increases rapidly with volume of lung irradiated, and esophagitis increases with dose to the esophagus. Both conditions can have adverse impacts on patients' quality of life. Daily online adaptive RT (ART) may allow for reduced normal tissue doses due to smaller margins around the target, as well as target size reduction with tumor response. This prospective, randomized controlled trial tests the hypothesis that the proportion of study participants who experience score ≥3 PRO-CTCAE cough, dyspnea, or dysphagia will be at least 20% lower with daily ART than with non-adapted RT. Enrollment goals are 244 subjects at up to 10 cancer centers worldwide. Eligibility criteria include stage IIIA-IIIC (AJCC v8) NSCLC; baseline grade 0-2 dyspnea, cough, and dysphagia; and no contralateral hilar or supraclavicular/cervical lymph node involvement. Subjects will be randomized (1:1) to CBCT-based daily ART or non-adapted RT using IMRT or VMAT delivering 60-66 Gy in 30-33 fractions with concurrent platinum doublet chemotherapy. Adjuvant immunotherapy with durvalumab is permitted. Follow-up for study participants will be for 1-year post-completion of chemoRT. Study endpoints include: frequency of PRO-CTCAE score ≥3 cough, dyspnea, or dysphagia from randomization to 30 days post-chemoRT; patient-reported quality of life (FACT-L and EQ-5D-5L questionnaires); percentage of lung receiving ≥20 Gy; mean doses to lung, heart, and esophagus; primary tumor response on CT or PET-CT (RECIST v1.1); local disease progression; and incidence of grade ≥2 pneumonitis within 1 year. Stratification factors are the treating institution and the presence of contralateral mediastinal lymph node metastases (associated with increased volume of irradiated lung). Interim analyses for futility and superiority will be performed when the primary endpoint data have been collected for 50% of evaluable participants. This study opened to enrollment on 20 October 2022 and is expected to be completed in approximately 3 years. This prospective, randomized clinical trial rigorously evaluates the impacts of daily online ART on radiation pneumonitis, esophagitis, and quality of life in patients with advanced NSCLC. It will collect standard tumor response and disease progression metrics to assure that reduced margins do not have an adverse impact on outcomes. Online ART is emerging as an innovative approach enabling increased sparing of normal tissues. The results of this clinical study will support evidence-based clinical decisions around ART technologies.

Reducing CTV to PTV Margins with Daily Adaptive Radiotherapy in the Postoperative Treatment of Endometrial and Cervical Cancer

In the postoperative treatment of endometrial and cervical cancer, a PTV expansion of 1.5cm or greater would be generally recommended to ensure adequate coverage of CTV with no or minimal image guidance. Online adaptive radiotherapy (oART) has demonstrated to be feasible to reduce inter-fractional radiotherapy errors as it re-optimizes treatment plan every fraction. In this study, we hypothesized using daily cone-beam computed tomography (CBCT)-guided oART would reduce CTV to PTV margins in the postoperative treatment of endometrial and cervical cancer. Seven patients from a single-center with postoperative endometrial and cervical cancer treated with daily CBCT-guided oART were retrospectively reviewed. A total dose of 45Gy is prescribed in CTV-vaginal cuff (CTV-V) and CTV- regional lymph nodes (CTV-N) with daily fractions of 1.8Gy for five fractions per week. A total of 175 radiotherapy fractions of daily pre-treatment CBCTs and post-treatment CBCTs scans were uploaded to oART emulator for CTV-V and CTV-N contouring by a single observer. CTVpre to PTVpre margins of 5, 7, 10, 12, 15mm was used in all directions. Each post-treatment CBCT was rigidly registered to the pre-treatment CBCT with respect to bony anatomy. The CTVpost was projected onto the PTVpre and assess required planning margins to encompass the CTVpost. The average total treatment time (post-treatment CBCTs - pre-treatment CBCTs) was 23 min 14s (range: 18min 53s ∼ 28min 53s). For all fractions, the adapted plan was selected, and the volume of the PTV-V and PTV-N receiving 100% of the prescribed dose or more (V100%) was more than required 95%. A uniform three-dimensional CTV-Npre to PTV-Npre margin of 5mm could encompass CTV-Npost in all fractions (100%; 175/175f). A uniform three-dimensional CTV-Vpre to PTV-Vpre margin of 10mm could encompass CTV-Vpost in all fractions (100%; 175/175f) and 5mm could encompass 98% fractions (172/175f). In the fractions of 5mm margins that failed to encompass CTV-Vpost, the volume of CTV missed three times in anterior-posterior directions and once in lateral direction. Compared with scheduled plan, adapted plan with 5mm margins significantly reduced the mean dose to bladder (Dmean: 103.97 cGy vs 107.20 cGy, P<0.05) and rectum (Dmean: 117.82 cGy vs 123.67 cGy, P<0.05), and achieved better PTV-N (V100%: 97.8% vs 90.0%, P<0.05) and PTV-V (V100%:96.8% vs 85.4%, P<0.05) dose coverage in 175 fractions. In this study, 5mm CTV to PTV margins was adequate to encompass 100% fractions of CTV-N and 98% fractions of the CTV-V. Adapted plan with 5mm margins significantly reduced the dose to critical organ at risks while improving target coverage.

1.5T MR-Guided Daily Adaptive Radiotherapy: Preliminary Clinical Report of the First 5000 Fractions

1.5T MR-Linac improves target volume and adjacent OARs visualization, ensuring high precision in radiation treatment delivery. Daily MR-imaging allows on-table adapted planning and real-time intra-fraction imaging without additional exposure to radiation. Our department implemented the first high field MRgRT system in Italy and herein we present the preliminary report of the first 5000 fractions delivered. We aim to describe the clinical implementation, feasibility, toxicity and patient tolerability of daily adapted RT. Since October 2019, 1.5T MR-Linac has been available in our department. A prospective observational study for the clinical use of 1.5T MR-Linac is currently ongoing in our department: patients affected by prostate adenocarcinoma, pancreatic cancer, oligometastases and patients requiring retreatments were included. Two different workflows were used depending on the OARs daily anatomy: Adapt To Position (ATP) where the reference plan position is adjusted rigidly to match the position of the targets and OARs, and Adapt To Shape (ATS) where a new plan is created to better match the anatomy of the day. Toxicity and quality of life were assessed at baseline and after treatment using the CTCAE v5.0. Patient-reported outcomes of prostate cancer patients were investigated by means IPSS, ICIQ-SF, IIEF-5, EPIC-26, EORTC-QLQ-C30 and PR-25 questionnaires. Between October 2019 and June 2022, 590 patients with 675 target sites were treated with MR-guided radiation therapy in 5000 total fractions. Median patient age was 70 years (39-86). Among 675 tumor sites, the most frequently treated region was pelvis (n = 549, 81%). The most common diagnosis was prostate cancer (n = 326). On-table adaptive radiation therapy was used at every treatment session: ATP workflow in 63 fractions (1%) and ATS workflow in 4937 fractions (99%), respectively. Median prescribed dose was 35 Gy (20-67.5 Gy) in median 5 fractions (5-30). Mean total treatment time was 43 minutes (20-56). Treatments were well-tolerated and no acute G>3 toxicities were reported. Concerning the PROMS, all questionnaires showed no relevant deterioration between the pre-, post-RT and follow-up evaluations. MR-guided radiation treatment using 1.5T MR-Linac has been successfully implemented into clinical routine at our department. The data reported support an optimal profile of tolerability of daily on-table adaptive radiation therapy in acceptable time slots. These results are confirmed by PROMs.

Triggering daily online adaptive radiotherapy in the pelvis: Dosimetric effects and procedural implications of trigger parameter-value selection.

Online adaptive radiotherapy (ART) can address dosimetric consequences of variations in anatomy by creating a new plan during treatment. However, ART is time- and labor-intensive and should be implemented in a resource-conscious way. Adaptive triggers composed of parameter-value pairs may direct the judicious use of online ART. This work analyzed our clinical experience using CBCT-based daily online ART to demonstrate how a conceptual framework based on adaptive triggers affects the dosimetric and procedural impact of ART. Sixteen patients across several pelvic sites were treated with CBCT-based daily online ART. Differences in standardized dose metrics were compared between the original plan, the original plan recalculated on the daily anatomy, and an adaptive plan. For each metric, trigger values were analyzed in terms of the proportion of treatments adapted and the distribution of metric values. Target coverage metrics were compromised due to anatomic variation with the average change per treatment ranging from -0.90 to -0.05 Gy, -0.47 to -0.02 Gy, -0.31 to -0.01 Gy, and -12.45% to -2.65% for PTV D99%, PTV D95%, CTV D99%, and CTV V100%, respectively. These were improved using the adaptive plan (-0.03 to 0.01 Gy, -0.02 to 0.00 Gy, -0.03 to 0.00 Gy, and -4.70% to 0.00%, respectively). Increasingly strict triggers resulted in a non-linear increase in the proportion of treatments adapted and improved the distribution of metric values with diminishing returns. Some organ-at-risk (OAR) metrics were compromised by anatomic variation and improved using the adaptive plan, but changes in most OAR metrics were randomly distributed. Daily online ART improved target coverage across multiple pelvic treatment sites and techniques. These effects were larger than those for OAR metrics, suggesting that maintaining target coverage was our primary benefit of CBCT-based daily online ART. Analyses like these can determine online ART triggers from a cost-benefit perspective.

PO-1485 1.5T MR-guided daily adaptive radiotherapy: preliminary clinical report of the first 5000 fractions

PO-1485 1.5T MR-guided daily adaptive radiotherapy: preliminary clinical report of the first 5000 fractions

Do we need patient-specific QA for adaptively generated plans? Retrospective evaluation of delivered online adaptive treatment plans on Varian Ethos.

The clinical introduction of dedicated treatment units for online adaptive radiation therapy (OART) has led to widespread adoption of daily adaptive radiotherapy. OART allows for rapid generation of treatment plans using daily patient anatomy, potentially leading to reduction of treatment margins and increased normal tissue sparing. However, the OART workflow does not allow for measurement of patient-specific quality assurance (PSQA) during treatment delivery sessions and instead relies on secondary dose calculations for verification of adapted plans. It remains unknown if independent dose verification is a sufficient surrogate for PSQA measurements. To evaluate the plan quality of previously treated adaptive plans through multiple standard PSQA measurements. This IRB-approved retrospective study included sixteen patients previously treated with OART at our institution. PSQA measurements were performed for each patient's scheduled and adaptive plans: five adaptive plans were randomly selected to perform ion chamber measurements and two adaptive plans were randomly selected for ArcCHECK measurements. The same ArcCHECK 3D dose distribution was also sent to Mobius3D to evaluate the second-check dosimetry system. All (n=96) ion chamber measurements agreed with the planned dose within 3% with a mean of 1.4% (±0.7%). All (n=48) plans passed ArcCHECK measurements using a 95% gamma passing threshold and 3%/2mm criteria with a mean of 99.1% (±0.7%). All (n=48) plans passed Mobius3D second-check performed with 95% gamma passing threshold and 5%/3mm criteria with a mean of 99.0% (±0.2%). Plan measurement for PSQA may not be necessary for every online-adaptive treatment verification. We recommend the establishment of a periodic PSQA check to better understand trends in passing rates for delivered adaptive treatments.

Reducing PTV Margins with Daily Adaptive Radiotherapy to the Prostatic Fossa

<h3>Purpose/Objective(s)</h3> In modern trials, traditional planning target volume (PTV) margins for postoperative prostate radiotherapy have been large (7-10 mm) to account both for daily changes in patient positioning / motion and target deformation (from changes in adjacent bladder fill and/or rectal distension). With daily adaptive radiotherapy, these inter-fractional changes could be minimized, potentially reducing the margins required for treatment and improving adjacent normal tissue dosimetry. <h3>Materials/Methods</h3> A single-center retrospective study was conducted from 03/2021 to 11/2021. Patients receiving conventionally fractionated radiotherapy to their prostatic fossa +/- elective nodes with pre- and post-treatment CBCT imaging (CBCTpre and CBCTpost, respectively) were included. Clinical target volumes (CTV) were contoured by a single observer on all CBCTs (only the prostatic fossa CTV was analyzed in this study). CBCTpre and CBCTpost images were registered using the same shifts used on the day of treatment. Motion was calculated from the center of mass of CBCTpre to that of CBCTpost, which was used to determine predicted margins with Van Herk's formula. Adequate coverage with the proposed margins was verified by performing a symmetric expansion on all CBCTpre scans (PTVpre) and evaluating the overlap with CBCTpost. <h3>Results</h3> Twenty patients were included with 263 image sets available and with 248 image sets included after exclusion. The average time between scans was 10.7 ± 3.0 minutes. The estimated margins predicted to achieve ≥95% coverage for 90% of the population were 1.6 mm, 2.0 mm, and 2.2 mm (x, y, and z axes, respectively). After symmetrically expanding all CBCTpre's CTVs by 1mm, 2mm, 3mm, and 4mm, the percent of image sets achieving ≥95% coverage were 73.6%, 92.2%, 97.1%, and 98.8%. <h3>Conclusion</h3> Using daily adaptive radiotherapy could significantly decrease the required PTV margin expansions for postoperative prostate radiotherapy, as a symmetric expansion of 3 mm achieved excellent coverage in this cohort which compared favorably to historical data. Given these results were obtained from CBCTpre and CBCTpost scans in the non-adaptive setting, these results will need to be validated in a prospective cohort.

Dosimetric Impact of Simulated Daily Adaptive Radiotherapy with Significantly Reduced Setup Margins in the Definitive Treatment of Head and Neck Cancer

<h3>Purpose/Objective(s)</h3> Even when treating head and neck cancer (HNC) with daily image-guided radiation therapy, setup margins of 3-5 mm are required to ensure adequate coverage of disease due to setup error. These margins lead to additional normal tissue dose that may compromise acute and long-term quality-of-life. The use of daily adaptive radiotherapy (ART) using an online adaptive system significantly reduces the need for a PTV margin, since the targets and organs-at-risk (OAR) are recontoured each fraction. In this study, we hypothesized using daily ART (simulated) with 1 mm margins based on our involved nodal radiotherapy (INRT) paradigm would improve coverage and reduce OAR dose. <h3>Materials/Methods</h3> Ten patients with HNC treated on our institutional phase II INRT trial were selected for this study. Cone-beam CT scans were uploaded to the ETHOS emulator for daily adaptive replanning which was simulated using fractions 5, 10, 15, 20, 25, 30, and 35. Two plans were generated with a 1 mm (reduced) and a 5 mm (conventional) PTV expansion. The dose grid from the delivered, conventional margin plan was transposed onto the CBCT and adapted contours. Separately, a reduced target margin plan optimized on the online adaptive system was created. PTV coverage (70 Gy-gross/primary disease, 66.5 Gy-suspicious nodes, 63 Gy-extended volume surrounding the primary disease) and D1 were calculated for target structures along with the dose to OARs. Comparisons performed via paired two tailed t-test. <h3>Results</h3> Six patients with oropharynx cancer and four patients with larynx cancer were included. Median V100% coverage for 70 Gy (96% vs 95.6%), 66.5 Gy (98.5% vs 76.5%), and 63 Gy (98.9% vs 74.9%) were improved with the adaptive planning (all p-values<0.03). Plan homogeneity was improved in the ART plan (e.g., PTV70 median D1% 72.8 Gy vs 74.8 Gy, all p-values<0.02). The total median dose reduction was 10.8 Gy (interquartile range, IQR 8.0-12.6 Gy) for the ipsilateral submandibular gland (SMG), 12.2 Gy (IQR 7.3-23.5 Gy) for the contralateral SMG, 8.3 Gy (IQR 3.8-12.6 Gy) for the ipsilateral parotid, 8.1 Gy (IQR 4.4-11.0 Gy) for the contralateral parotid, 7.0 Gy (IQR 4.5-9.1 Gy) for larynx, 6.6 Gy (IQR 2.7-8.4 Gy) for oral cavity, and 8.6 Gy (IQR 5.9-12.1 Gy) for constrictors. <h3>Conclusion</h3> Simulated daily ART using a 1 mm PTV margin provided significant improvement over conventional PTV margins in coverage and critical OAR dosimetry. This small proof-of-concept study suggests this approach may lead to improved quality-of-life while preserving or improving tumor coverage. Further evaluation is needed to assess the resources needed to implement daily adaptations. A phase II randomized trial of daily ART is underway to determine if dosimetric improvements lead to similar gains in patient-reported outcomes.

Time Dependence of Coverage of the Prostatic Fossa: Implications for Daily Adaptive Radiotherapy

<h3>Purpose/Objective(s)</h3> Online daily adaptive radiotherapy (ART) is an attractive modality for reducing PTV margins and improving the therapeutic ratio through elimination of inter-fractional uncertainty. However, daily adaptation is more time consuming than traditional image-guided radiation, and consequently, this increased time from initial CBCT acquisition until treatment delivery may introduce additional setup errors. For example, when delivering postoperative radiotherapy to the prostatic fossa, the adjacent bladder and rectum can change shape in the time required to adapt patients, resulting in deformation of the target and potentially a geographic miss. Therefore, we hypothesized that decreased time between scans would be associated with better coverage. <h3>Materials/Methods</h3> A single-center retrospective study was conducted from 03/2021 to 11/2021. Patients receiving conventionally fractionated radiotherapy to their prostatic fossa +/- elective nodes with pre- and post-treatment CBCT imaging before and after the same fraction (CBCTpre and CBCTpost, respectively) were included. Clinical target volumes (CTVpre and CTVpost) were contoured by a single observer on all CBCTs (only the prostatic fossa CTV was analyzed in this study). CBCTpre and CBCTpost images were registered using the same shifts from the day of treatment. Overlap of CTVpre and CTVpost was calculated and correlated with time with Pearson's correlation coefficient. The time threshold at which the maximum difference in overlap of CTVpre and CTVpost was identified iteratively. Subgroups above and below this threshold were compared, evaluating the percent overlap between CTVpre with 1, 2, and 3 mm expansions with CTVpost (here, overlap serves as a surrogate of dose coverage). <h3>Results</h3> When evaluating overlap between CTVpre and CTVpost, there was a trend towards more overlap with shorter intervals between scans (r=-0.091). A threshold of 13.5 minutes was identified with scans <13.5 minutes apart (n=210) having an average overlap of 91.4% vs ≥13.5 minutes (n=32) being 89.2% (p = 0.026). However, this difference lost significance when using 3 mm PTV margins (p=0.134), which was previously shown to be the optimal margin for this cohort. Still, there was a suggestion of better coverage for intra-scan times <13.5 minutes for all margins evaluated: 1mm (75.7% vs 59.4% achieved 95% coverage), 2mm (93.3% vs 84.4%), and 3mm (98.1% vs 90.6%). <h3>Conclusion</h3> Shorter times utilized for daily adaptation of the prostatic fossa may be associated with better coverage, likely secondary to less displacement of the target from patient movement and deformation from changes in adjacent normal tissue. Given image sets with intra-scan times >13.5 minutes were limited in this study, these findings will need to be validated in larger prospective cohorts.

Adaptive on Demand: Leveraging Cone Beam Computed Tomography-Based Adaptive Online Radiotherapy System as Simulation-Free Replan Platform

<h3>Purpose/Objective(s)</h3> Artificial intelligence powered cone beam computed tomography (CBCT) based online adaptive radiotherapy (ART) system offers a streamlined and efficient process for daily ART as the default. However, patients with gradual anatomy change from tumor response or weight loss may not significantly benefit from the daily ART. Instead, it will be more practical to adapt when changes are observed. In this work, we present a workflow and our early experience to utilize CBCT based ART system as a quick replan platform and the adapted plan is used for image guided radiotherapy (IGRT) until next adaptation. <h3>Materials/Methods</h3> Adaptive on demand is designed for patients treated with conventional fractionated dose and planned with conventional margins. The preplan is designed with 12-18 static IMRT fields with consideration of potential geometry/density change in beam arrangement and optimization to ensure the robustness of ART plan quality. Robustness evaluation is performed as needed. A plan revision process is required in the ART treatment planning system (TPS) to switch between adaptive and non-adaptive fractions in a treatment course. The plan is converted to the adaptive treatment upon physician's request. After each ART, the adapted plan is imported to TPS and physicist checks the quality of the synthetic CT (SCT), which is a deformed planning CT to CBCT geometry, and determines SCT/CBCT to be used as reference for the upcoming IGRT fractions. <h3>Results</h3> A total of 23 HN patients and 13 lung patients were treated with ART on demand from 09/2021-01/2022. 110 on-demand ART fractions analyzed. On average, each HN/lung patient received 1.9 (range 1-6)/5.2 (range 4-6) ART and 90.6%/85.3% frequency that the adapted plan was selected for treatment. Overall, the adapted plans show statistically significant improvements in tumor coverage and dose conformity comparing to original plan and meet the OAR objectives. Detailed dosimetric benefits are presented in other abstracts. Given the SCT is generated via deformable registration, the distortion of the SCT was observed in some cases. While dosimetric impact are negligible for the majority of cases, the geometric distortions of bony anatomy can lead to setup difficulty for the IGRT fractions and two patients were rescanned due to this reason. Cutting out the mask is recommended when it becomes loose to avoid SCT picking up mask as body and results in unsatisfactory planning target volume cropping from skin. The conversion to adaptive fraction requires 15 minutes while setting up adapted plan for IGRT fraction takes 25 minutes. Physician's time at console time is 18/14 minutes for HN and lung. <h3>Conclusion</h3> Leveraging CBCT based ART system as replan platform is feasible and satisfactory in plan quality. Physicist check on the SCT quality is crucial to prevent potential target generation uncertainties and setup error. This process significantly improved the turnaround time for replan and saves a trip for rescan.

Online Adaptive Radiotherapy and Opportunity Cost

<h3>Purpose/Objective(s)</h3> Changes in patient anatomy and tumor geometry pose a challenge to ensuring consistent target coverage and organs-at-risk (OAR) sparing; online adaptive radiotherapy (ART) has been used to account for these inter-fractional changes by facilitating replanning prior to each treatment. This project explores the opportunity cost of CT-based online ART by evaluating the time and human resource requirements for adaptive versus conventional image-guided radiation therapy (IGRT). This study seeks to quantify time requirements for ART and, using time driven activity-based costing (TDABC), determine the cost of this time to determine if the dosimetric benefit is worthwhile. <h3>Materials/Methods</h3> Online ART for pelvic disease sites was evaluated using Ethos based planning technology. Time, which at our facility is best represented by the duration between two cone beam CT (CBCT) scans taken for each adaptive fraction, was used as a broadly applicable and transferable metric, representing the additional time required for adaptive treatment on top of standard IGRT. Dosimetric impact was also considered. All parties (dosimetrist, therapists, physicist, physician) need to be present for the duration of treatment planning given the rapid response time required. Using recently validated TDABC at this facility, the per fraction cost of ART was determined, reflecting the added cost of ART on top of IGRT. <h3>Results</h3> Preliminary data suggests a time difference of approximately 7 versus 24 minutes for daily IGRT and ART, respectively. Each adaptive fraction requires more staffing for the duration of this time delaying completion of other tasks and functionally increasing resource utilization. <h3>Conclusion</h3> While online ART decreases the uncertainty of anatomical shifts, the added human resources required per fraction complicate the workflow of daily treatment and increase expense. This has an even greater impact on the implementation of ART in the context of the alternative payment model. Moreover, online ART removes physicians from clinical time. This opportunity cost is likely to be similar across different disease sites, unlike toxicity benefits which may vary more depending on anatomic location and require further longitudinal studies. Regardless, the implementation of progressively complex radiotherapy technologies, like ART, requires consideration of the time and human resource requirements and subsequent opportunity cost.

CT-Guided Adaptive Radiotherapy (CT-ART) Improves Target Coverage in Hypofractionated Radiation Therapy for Bladder Cancer

<h3>Purpose/Objective(s)</h3> We assess the feasibility and dosimetric characteristics of CT-guided daily adaptive radiotherapy (CT-ART) for patients with muscle-invasive bladder cancer (MIBC). Compared to the scheduled treatment, we hypothesized that CT-ART would allow us to optimize target coverage and dose to organs at risk (OARs) by accounting for inter-fraction motion of the bladder and OARs. <h3>Materials/Methods</h3> Four patients with cT2a-T4aN0M0 MIBC treated with hypofractionated RT or chemoRT at 2.75 Gy/Fx (55Gy/20) with empty bladder using CT-ART were retrospectively reviewed. There were 42 total fractions delivered with CT-ART. For each Fx, cone beam CT (CBCT) scans were acquired with artificial-intelligent segmented target and OARs. Contours were manually edited and a patient-specific internal target volume (ITV) was applied for bladder filling during the adaptive process. An adaptive plan (P_A) was then created and optimized based on these volumes. Adaptive plan metrics for target/OARs were then compared to the initial plan (P_I) for each Fx. The PTV_5500 V₁₀₀ (%), V₉₅ (%), D_97% (Gy), and D_95% (Gy); rectum D_47Gy (cc); and non-sigmoid bowel OARs were compared between the P_I and P_A plans using paired samples t-test (Table). <h3>Results</h3> CT-ART improved PTV coverage, with mean per Fx improvement of 12.4% and 28.3% for V₉₅ and V₁₀₀ and 0.4 Gy/Fx for both D _97% and D_95% (<i>p</i><0.01 for all). The P_I plan did not meet PTV V₁₀₀ ≥95% in 39 of 42 Fxs (92.9%) vs. 4 of 42 Fx (9.5%) with the P_A plans. P_A plans did not result in increased dose to OARs (Table). Median CT-ART treatment duration improved over time [median 33 min (range 30-47) for the last patient treated]. <h3>Conclusion</h3> CT-based online adaptive radiation therapy was feasible and significantly improved PTV coverage without a corresponding increase in dose to OARs. This study demonstrates the potential value of CT-ART for MIBC. This approach will be tested in an upcoming phase II trial.

Patient-Specific Auto-Segmentation of Target and OARs via Deep Learning on Daily Fan-Beam CT for Adaptive Prostate Radiotherapy

<h3>Purpose/Objective(s)</h3> Contouring the target and organs at risk (OARs) of the treated patient in each fraction is a key step of daily adaptive radiotherapy. We designed a patient-specific deep learning study to improve the auto-segmentation performance for adaptive prostate radiotherapy, by leveraging the daily image-guidance fan-beam CT images of the specific patient available in the current IGRT workflow. <h3>Materials/Methods</h3> Our institution has set up a newly developed biology-guided radiotherapy (BgRT) system enabling high-quality fan-beam CT imaging for daily treatments. The data of the first prostate patient treated with this system including 26 daily fan-beam CT scans were utilized in this study. A population network was first trained for the auto-segmentation of the prostate and seven pelvic OARs, based on a population dataset containing 56 different patient cases treated on the regular Linac. The pre-trained population network was intentionally fine-tuned to adapt to this specific patient with a transfer learning method. The selected OARs of this patient were contoured by two radiation oncologists on the initial treatment planning and all sets of daily fan-beam CT images as the ground truth for the patient-specific learning. Three different fine-tuning methods, i.e., fine-tuning the last convolution layer, the last convolution block and the whole network were analyzed. A longitudinal study was conducted by exploring the relationship between the auto-segmentation performance and the number of the sequential prior data used for learning. The performance of the patient-specific network was compared with the population network as well as the clinical rigid registration method to highlight its efficiency in radiotherapy practice. <h3>Results</h3> The Dice similarity coefficient (DSC) results of the proposed patient-specific network were 0.882, 0.934, 0.864, 0.827, 0.952, 0.954, 0.921 and 0.904 for the prostate, bladder, rectum, seminal vesicle, left and right femoral head, bowel and lymph nodes, respectively, outperforming the population network (mean DSC of 0.627) and the clinical registration method (mean DSC of 0.716). The whole network fine-tuning method outperformed the other two fine-tuning methods, showing it is more effective in this study. The contouring accuracy of the patient-specific network gradually increased with the increment of longitudinal training cases and approached saturation with more than four training cases. <h3>Conclusion</h3> This study proposes an accurate auto-segmentation method based on the patient-specific deep learning for adaptive prostate radiotherapy. By utilizing the image and contour similarity between the treatment fractions, the patient-specific auto-segmentation could outperform the common population network and the clinical registration method by a large margin, and thus is promising to facilitate more precise radiotherapy based on the new BgRT system.

Dosimetric and feasibility evaluation of a CBCT-based daily adaptive radiotherapy protocol for locally advanced cervical cancer.

Evaluate a cone-beam computed tomography (CBCT)-based daily adaptive platform in cervical cancer for multiple endpoints: (1) physics contouring accuracy of daily CTVs, (2) CTV coverage with adapted plans and reduced PTV margins versus non-adapted plans with standard-of-care (SOC) margins, (3) dosimetric improvements to CTV and organs-at-risk (OARs), and (4) on-couch time. Using a Varian Ethos™ emulator and KV-CBCT scans, we simulated the doses 15 retrospective cervical cancer patients would have received with/without online adaptation for five fractions. We compared contours and doses from SOC plans (5-15mm CTV-to-PTV margins) to adapted plans (3mm margins). Auto-segmented CTVs and OARs were reviewed and edited by trained physicists. Physics-edited targets were evaluated by an oncologist. Time spent reviewing and editing auto-segmented structures was recorded. Metrics from the CTV (D99%), bowel (V45Gy, V40Gy), bladder (D50%), and rectum (D50%) were compared. The physician approved the physics-edited CTVs for 55/75 fractions; 16/75 required reductions, and 4/75 required CTV expansions. CTVs were encapsulated by unadapted, SOC PTVs for 56/75 (72%) fractions-representative of current clinical practice. CTVs were completely covered by adapted 3mm PTVs for 71/75 (94.6%) fractions. CTV D99% values for adapted plans were comparable to non-adapted SOC plans (average difference of -0.9%), while all OAR metrics improved with adaptation. Specifically, bowel V45Gy and V40Gy decreased on average by 87.6 and 109.4 cc, while bladder and rectum D50% decreased by 37.7% and 35.8%, respectively. The time required for contouring and calculating an adaptive plan for 65/75 fractions was less than 20 min (range: 1-29 min). Improved dose metrics with daily adaption could translate to reduced toxicity while maintaining tumor control. Training physicists to perform contouring edits could minimize the time physicians are required at adaptive sessions improving clinical efficiency. All emulated adaptive sessions were completed within 30 min however extra time will be required for patient setup, image acquisition, and treatment delivery.

Dosimetric comparison of automatically propagated prostate contours with manually drawn contours in MRI-guided radiotherapy: A step towards a contouring free workflow?

BackgroundThe prostate demonstrates inter- and intra- fractional changes and thus adaptive radiotherapy would be required to ensure optimal coverage. Daily adaptive radiotherapy for MRI-guided radiotherapy can be both time and resource intensive when structure delineation is completed manually. Contours can be auto-generated on the MR-Linac via a deformable image registration (DIR) based mapping process from the reference image. This study evaluates the performance of automatically generated target structure contours against manually delineated contours by radiation oncologists for prostate radiotherapy on the Elekta Unity MR-Linac. MethodsPlans were generated from prostate contours propagated by DIR and rigid image registration (RIR) for forty fractions from ten patients. A two-dose level SIB (simultaneous integrated boost) IMRT plan is used to treat localised prostate cancer; 6000 cGy to the prostate and 4860 cGy to the seminal vesicles. The dose coverage of the PTV 6000 and PTV 4860 created from the manually drawn target structures was evaluated with each plan. If the dose objectives were met, the plan was considered successful in covering the gold standard (clinician-delineated) volume. ResultsThe mandatory PTV 6000 dose objective (D98% > 5580 cGy) was met in 81 % of DIR plans and 45 % of RIR plans. The SV were mapped by DIR only and for all the plans, the PTV 4860 dose objective met the optimal target (D98% > 4617 cGy). The plans created by RIR led to under-coverage of the clinician-delineated prostate, predominantly at the apex or the bladder-prostate interface. ConclusionPlans created from DIR propagation of prostate contours outperform those created from RIR propagation. In approximately 1 in 5 DIR plans, dosimetric coverage of the gold standard PTV was not clinically acceptable. Thus, at our institution, we use a combination of DIR propagation of contours alongside manual editing of contours where deemed necessary for online treatments.