Relating pre-treatment non-Gaussian intravoxel incoherent motion diffusion-weighted imaging to human papillomavirus status and response in oropharyngeal carcinoma.

Background and purpose: Diffusion-weighted imaging (DWI) is a promising technique for response assessment in head-and-neck cancer. Recently, we optimized Non-Gaussian Intravoxel Incoherent Motion Imaging (NG-IVIM), an extension of the conventional apparent diffusion coefficient (ADC) model, for the head and neck. In the current study, we describe the first application in a group of patients with human papillomavirus (HPV)-positive and HPV-negative oropharyngeal squamous cell carcinoma. The aim of this study was to relate ADC and NG-IVIM DWI parameters to HPV status and clinical treatment response. Materials and methods: Thirty-six patients (18 HPV-positive, 18 HPV-negative) were prospectively included. Presence of progressive disease was scored within one year. The mean pre-treatment ADC and NG-IVIM parameters in the gross tumor volume were compared between HPV-positive and HPV-negative patients. In HPV-negative patients, ADC and NG-IVIM parameters were compared between patients with and without progressive disease. Results: ADC, the NG-IVIM diffusion coefficient D, and perfusion fraction f were significantly higher, while pseudo-diffusion coefficient D* and kurtosis K were significantly lower in the HPV-negative compared to HPV-positive patients. In the HPV-negative group, a significantly lower D was found for patients with progressive disease compared to complete responders. No relation with ADC was observed. Conclusion: The results of our single-center study suggest that ADC is related to HPV status, but not an independent response predictor. The NG-IVIM parameter D, however, was independently associated to response in the HPV-negative group. Noteworthy in the opposite direction as previously thought based on ADC.

Clinical practicality and patient performance for surface-guided automated VMAT gating for DIBH breast cancer radiotherapy.

BACKGROUND AND PURPOSE: To evaluate the performance of automated surface-guided gating for left-sided breast cancer with DIBH and VMAT. MATERIALS AND METHODS: Patients treated in the first year after introduction of DIBH with VMAT were retrospectively considered for analysis. With automated surface-guided gating the beam automatically switches on/off, if the surface region of interest moved in/out the gating tolerance (±3 mm, ±3°). Patients were coached to hold their breath as long as comfortably possible. Depending on the patient's preference, patients received audio instructions during treatment delivery. Real-time positional variations of the breast/chest wall surface with respect to the reference surface were collected, for all three orthogonal directions. The durations and number of DIBHs needed to complete dose delivery, and DIBH position variations were determined. To evaluate an optimal gating window threshold, smaller tolerances of ±2.5 mm, ±2.0 mm, and ±1.5 mm were simulated. RESULTS: 525 fractions from 33 patients showed that median DIBH duration was 51 s (range: 30-121 s), and median 4 DIBHs per fraction were needed to complete VMAT dose delivery. Median intra-DIBH stability and intrafractional DIBH reproducibility approximated 1.0 mm in each direction. No large differences were found between patients who preferred to perform the DIBH procedure with (n = 21) and without audio-coaching (n = 12). Simulations demonstrated that gating window tolerances could be reduced from ±3.0 mm to ±2.0 mm, without affecting beam-on status. CONCLUSION: Independent of the use of audio-coaching, this study demonstrates that automated surface-guided gating with DIBH and VMAT proved highly efficient. Patients' DIBH performance far exceeded our expectations compared to earlier experiences and literature. Furthermore, gating window tolerances could be reduced.

A fast and robust constraint-based online re-optimization approach for automated online adaptive intensity modulated proton therapy in head and neck cancer.

Objective. In head-and-neck cancer intensity modulated proton therapy, adaptive radiotherapy is currently restricted to offline re-planning, mitigating the effect of slow changes in patient anatomies. Daily online adaptations can potentially improve dosimetry. Here, a new, fully automated online re-optimization strategy is presented. In a retrospective study, this online re-optimization approach was compared to our trigger-based offline re-planning (offlineTBre-planning) schedule, including extensive robustness analyses.Approach. The online re-optimization method employs automated multi-criterial re-optimization, using robust optimization with 1 mm setup-robustness settings (in contrast to 3 mm for offlineTBre-planning). Hard planning constraints and spot addition are used to enforce adequate target coverage, avoid prohibitively large maximum doses and minimize organ-at-risk doses. For 67 repeat-CTs from 15 patients, fraction doses of the two strategies were compared for the CTVs and organs-at-risk. Per repeat-CT, 10.000 fractions with different setup and range robustness settings were simulated using polynomial chaos expansion for fast and accurate dose calculations.Main results. For 14/67 repeat-CTs, offlineTBre-planning resulted in <50% probability ofD98%≥ 95% of the prescribed dose (Dpres) in one or both CTVs, which never happened with online re-optimization. With offlineTBre-planning, eight repeat-CTs had zero probability of obtainingD98%≥ 95%Dpresfor CTV7000, while the minimum probability with online re-optimization was 81%. Risks of xerostomia and dysphagia grade ≥ II were reduced by 3.5 ± 1.7 and 3.9 ± 2.8 percentage point [mean ± SD] (p< 10-5for both). In online re-optimization, adjustment of spot configuration followed by spot-intensity re-optimization took 3.4 min on average.Significance. The fast online re-optimization strategy always prevented substantial losses of target coverage caused by day-to-day anatomical variations, as opposed to the clinical trigger-based offline re-planning schedule. On top of this, online re-optimization could be performed with smaller setup robustness settings, contributing to improved organs-at-risk sparing.

A review of the clinical introduction of 4D particle therapy research concepts.

Background and purpose: Many 4D particle therapy research concepts have been recently translated into clinics, however, remaining substantial differences depend on the indication and institute-related aspects. This work aims to summarise current state-of-the-art 4D particle therapy technology and outline a roadmap for future research and developments. Material and methods: This review focused on the clinical implementation of 4D approaches for imaging, treatment planning, delivery and evaluation based on the 2021 and 2022 4D Treatment Workshops for Particle Therapy as well as a review of the most recent surveys, guidelines and scientific papers dedicated to this topic. Results: Available technological capabilities for motion surveillance and compensation determined the course of each 4D particle treatment. 4D motion management, delivery techniques and strategies including imaging were diverse and depended on many factors. These included aspects of motion amplitude, tumour location, as well as accelerator technology driving the necessity of centre-specific dosimetric validation. Novel methodologies for X-ray based image processing and MRI for real-time tumour tracking and motion management were shown to have a large potential for online and offline adaptation schemes compensating for potential anatomical changes over the treatment course. The latest research developments were dominated by particle imaging, artificial intelligence methods and FLASH adding another level of complexity but also opportunities in the context of 4D treatments. Conclusion: This review showed that the rapid technological advances in radiation oncology together with the available intrafractional motion management and adaptive strategies paved the way towards clinical implementation.

Evaluating AI-generated CBCT-based synthetic CT images for target delineation in palliative treatments of pelvic bone metastasis at conventional C-arm linacs.

PURPOSE: One-table treatments with treatment imaging, preparation and delivery occurring at one treatment couch, could increase patients' comfort and throughput for palliative treatments. On regular C-arm linacs, however, cone-beam CT (CBCT) imaging quality is currently insufficient. Therefore, our goal was to assess the suitability of AI-generated CBCT based synthetic CT (sCT) images for target delineation and treatment planning for palliative radiotherapy. MATERIALS AND METHODS: CBCTs and planning CT-scans of 22 female patients with pelvic bone metastasis were included. For each CBCT, a corresponding sCT image was generated by a deep learning model in ADMIRE 3.38.0. Radiation oncologists delineated 23 target volumes (TV) on the sCTs (TVsCT) and scored their delineation confidence. The delineations were transferred to planning CTs and manually adjusted if needed to yield gold standard target volumes (TVclin). TVsCT were geometrically compared to TVclin using Dice coefficient (DC) and Hausdorff Distance (HD). The dosimetric impact of TVsCT inaccuracies was evaluated for VMAT plans with different PTV margins. RESULTS: Radiation oncologists scored the sCT quality as sufficient for 13/23 TVsCT (median: DC = 0.9, HD = 11 mm) and insufficient for 10/23 TVsCT (median: DC = 0.7, HD = 34 mm). For the sufficient category, remaining inaccuracies could be compensated by +1 to +4 mm additional margin to achieve coverage of V95% > 95% and V95% > 98%, respectively in 12/13 TVsCT. CONCLUSION: The evaluated sCT quality allowed for accurate delineation for most targets. sCTs with insufficient quality could be identified accurately upfront. A moderate PTV margin expansion could address remaining delineation inaccuracies. Therefore, these findings support further exploration of CBCT based one-table treatments on C-arm linacs.

A generalized model for monitor units determination in ocular proton therapy using machine learning: A proof-of-concept study.

Objective.Determining and verifying the number of monitor units is crucial to achieving the desired dose distribution in radiotherapy and maintaining treatment efficacy. However, current commercial treatment planning system(s) dedicated to ocular passive eyelines in proton therapy do not provide the number of monitor units for patient-specific plan delivery. Performing specific pre-treatment field measurements, which is time and resource consuming, is usually gold-standard practice. This proof-of-concept study reports on the development of a multi-institutional-based generalized model for monitor units determination in proton therapy for eye melanoma treatments.Approach.To cope with the small number of patients being treated in proton centers, three European institutes participated in this study. Measurements data were collected to address output factor differences across the institutes, especially as function of field size, spread-out Bragg peak modulation width, residual range, and air gap. A generic model for monitor units prediction using a large number of 3748 patients and broad diversity in tumor patterns, was evaluated using six popular machine learning algorithms: (i) decision tree; (ii) random forest, (iii) extra trees, (iv) K-nearest neighbors, (v) gradient boosting, and (vi) the support vector regression. Features used as inputs into each machine learning pipeline were: Spread-out Bragg peak width, range, air gap, fraction and calibration doses. Performance measure was scored using the mean absolute error, which was the difference between predicted and real monitor units, as collected from institutional gold-standard methods.Main results.Predictions across algorithms were accurate within 3% uncertainty for up to 85.2% of the plans and within 10% uncertainty for up to 98.6% of the plans with the extra trees algorithm.Significance.A proof-of-concept of using machine learning-based generic monitor units determination in ocular proton therapy has been demonstrated. This could trigger the development of an independent monitor units calculation tool for clinical use.

Nine years of plan of the day for cervical cancer: Plan library remains effective compared to fully online-adaptive techniques.

BACKGROUND AND PURPOSE: Since 2011, our center has been using a library-based Plan-of-the-Day (PotD) strategy for external beam radiotherapy of cervical cancer patients to reduce normal tissue dose while maintaining adequate target coverage. With the advent of fully online-adaptive techniques such as daily online-adaptive replanning, further dose reduction may be possible. However, it is unknown how this reduction relates to plan library approaches, and how the most recent PotD strategies relate to no adaptation. In this study we compare the performance of our current PotD strategy with non-adaptive and fully online-adaptive techniques in terms of target volume size and normal tissue sparing. MATERIALS AND METHODS: Treatment data of 376 patients treated with the PotD protocol between June 2011 and April 2020 were included. The size of the Planning Target Volumes (PTVs) was reconstructed for different strategies: full online adaptation, no adaptation, and the latest clinical version of the PotD protocol. Normal tissue sparing was estimated by the difference in margin volume to construct the PTV and the volume overlap of the PTV with bladder and rectum. RESULTS: The current version of our PotD approach reduced the PTV margin volume by a median of 250 cm3 compared to no adaptation. Bladder-PTV overlap decreased from a median of 142 to 71 cm3, and from 39 to 16 cm3 for rectum-PTV. Fully online-adaptive approaches could further decrease the PTV volume by 144 cm3 using a 5 mm margin for residual errors. In this scenario, bladder-PTV overlap was reduced to 35 cm3 and rectum-PTV overlap to 11 cm3. CONCLUSION: The current version of the PotD protocol is an effective technique to improve normal tissue sparing compared to no adaptation. Further sparing can be achieved using fully online-adaptive techniques, but at the cost of a more complex workflow and with a potentially limited impact. PotD-type protocols can therefore be considered as a suitable alternative to fully online-adaptive approaches.

Robust optimization of a radiotherapy pretreatment preparation workflow.

Objective. Increasing cancer incidence, staff shortage and high burnout rate among radiation oncologists, medical physicists and radiation technicians are putting many departments under strain. Operations research (OR) tools could optimize radiotherapy processes, however, clinical implementation of OR-tools in radiotherapy is scarce since most investigated optimization methods lack robustness against patient-to-patient variation in duration of tasks. By combining OR-tools, a method was developed that optimized deployment of radiotherapy resources by generating robust pretreatment preparation schedules that balance the expected average patient preparation time (Fmean) with the risk of working overtime (RoO). The method was evaluated for various settings of an one-stop shop (OSS) outpatient clinic for palliative radiotherapy.Approach. The OSS at our institute sees, scans and treats 3-5 patients within one day. The OSS pretreatment preparation workflow consists of a fixed sequence of tasks, which was manually optimized for radiation oncologist and CT availability. To find more optimal sequences, with shorterFmeanand lowerRoO, a genetic algorithm was developed which regards these sequences as DNA-strands. The genetic algorithm applied natural selection principles to produce new sequences. A decoder translated sequences to schedules to find the conflicting fitness parametersFmeanandRoO. For every generation, fitness of sequences was determined by the distance to the estimated Pareto front ofFmeanandRoO. Experiments were run in various OSS-settings.Main results. According to our approach, the expectedFmeanof the current clinical schedule could be reduced with 37%, without increasingRoO. Additional experiments provided insights in trade-offs betweenFmean,RoO, working shift length, number of patients treated on a single day and staff composition.Significance. Our approach demonstrated that OR-tools could optimize radiotherapy resources by robust pretreatment workflow scheduling. The results strongly support further exploration of scheduling optimization for treatment preparation also outside a one-stop shop or radiotherapy setting.

Multi-institutional consensus on machine QA for isochronous cyclotron-based systems delivering ultra-high dose rate (FLASH) pencil beam scanning proton therapy in transmission mode.

BACKGROUND: The first clinical trials to assess the feasibility of FLASH radiotherapy in humans have started (FAST-01, FAST-02) and more trials are foreseen. To increase comparability between trials it is important to assure treatment quality and therefore establish a standard for machine quality assurance (QA). Currently, the AAPM TG-224 report is considered as the standard on machine QA for proton therapy, however, it was not intended to be used for ultra-high dose rate (UHDR) proton beams, which have gained interest due to the observation of the FLASH effect. PURPOSE: The aim of this study is to find consensus on practical guidelines on machine QA for UHDR proton beams in transmission mode in terms of which QA is required, how they should be done, which detectors are suitable for UHDR machine QA, and what tolerance limits should be applied. METHODS: A risk assessment to determine the gaps in the current standard for machine QA was performed by an international group of medical physicists. Based on that, practical guidelines on how to perform machine QA for UHDR proton beams were proposed. RESULTS: The risk assessment clearly identified the need for additional guidance on temporal dosimetry, addressing dose rate (constancy), dose spillage, and scanning speed. In addition, several minor changes from AAPM TG-224 were identified; define required dose rate levels, the use of clinically relevant dose levels, and the use of adapted beam settings to minimize activation of detector and phantom materials or to avoid saturation effects of specific detectors. The final report was created based on discussions and consensus. CONCLUSIONS: Consensus was reached on what QA is required for UHDR scanning proton beams in transmission mode for isochronous cyclotron-based systems and how they should be performed. However, the group discussions also showed that there is a lack of high temporal resolution detectors and sufficient QA data to set appropriate limits for some of the proposed QA procedures.

Cochlear-optimized treatment planning in photon and proton radiosurgery for vestibular schwannoma patients.

Objective: To investigate the potential to reduce the cochlear dose with robotic photon radiosurgery or intensity-modulated proton therapy planning for vestibular schwannomas. Materials and Methods: Clinically delivered photon radiosurgery treatment plans were compared to five cochlear-optimized plans: one photon and four proton plans (total of 120). A 1x12 Gy dose was prescribed. Photon plans were generated with Precision (Cyberknife, Accuray) with no PTV margin for set-up errors. Proton plans were generated using an in-house automated multi-criterial planning system with three or nine-beam arrangements, and applying 0 or 3 mm robustness for set-up errors during plan optimization and evaluation (and 3 % range robustness). The sample size was calculated based on a reduction of cochlear Dmean > 1.5 Gy(RBE) from the clinical plans, and resulted in 24 patients. Results: Compared to the clinical photon plans, a reduction of cochlear Dmean > 1.5 Gy(RBE) could be achieved in 11/24 cochlear-optimized photon plans, 4/24 and 6/24 cochlear-optimized proton plans without set-up robustness for three and nine-beam arrangement, respectively, and in 0/24 proton plans with set-up robustness. The cochlea could best be spared in cases with a distance between tumor and cochlea. Using nine proton beams resulted in a reduced dose to most organs at risk. Conclusion: Cochlear dose reduction is possible in vestibular schwannoma radiosurgery while maintaining tumor coverage, especially when the tumor is not adjacent to the cochlea. With current set-up robustness, proton therapy is capable of providing lower dose to organs at risk located distant to the tumor, but not for organs adjacent to it. Consequently, photon plans provided better cochlear sparing than proton plans.

Accounting for fractionation and heterogeneous dose distributions in the modelling of osteoradionecrosis in oropharyngeal carcinoma treatment.

BACKGROUND AND PURPOSE: Osteoradionecrosis (ORN) of the mandible is a severe complication following radiotherapy (RT). With a renewed interest in hypofractionation for head and neck radiotherapy, more information concerning ORN development after high fraction doses is important. The aim of this explorative study was to develop a model for ORN risk prediction applicable across different fractionation schemes using Equivalent Uniform Doses (EUD). MATERIAL AND METHODS: We performed a retrospective cohort study in 334 oropharyngeal squamous cell carcinoma (OPSCC) patients treated with either a hypofractionated Stereotactic Body Radiation Therapy (HF-SBRT) boost or conventional Intensity Modulated Radiation Therapy (IMRT). ORN was scored with the CTCAE v5.0. HF-SBRT and IMRT dose distributions were converted into equivalent dose in 2 Gy fractions (α/ß = 0.85 Gy) and analyzed using EUD. The parameter a that led to an EUD that best discriminated patients with and without grade ≥ 2 ORN was selected. Patient and treatment-related risk factors of ORN were analyzed with uni- and multivariable regression analysis. RESULTS: A total of 32 patients (9.6%) developed ORN grade ≥ 2. An EUD(a = 8) best discriminated between ORN and non-ORN (AUC = 0.71). In multivariable regression, pre-RT extractions (SHR = 2.34; p = 0.012), mandibular volume (SHR = 1.04; p = 0.003), and the EUD(a = 8) (SHR = 1.14; p < 0.001) were significantly associated with ORN. CONCLUSION: Risk models for ORN based on conventional DVH parameters cannot be directly applied to HF-SBRT fractionation schemes and dose distributions. However, after correcting for fractionation and non-uniform dose distributions using EUD, a single model can distinguish between ORN and non-ORN after conventionally fractionated radiotherapy and hypofractionated boost treatments.

The impact of bone marrow sparing on organs at risk dose for cervical cancer: a Pareto front analysis.

Background and purpose: To quantify the increase in bladder and rectum dose of a bone marrow sparing (BMS) VMAT strategy for primary treatment of locally advanced cervical cancer (LACC). Materials and methods: Twenty patients with stage IB-IVA cervical cancer were selected for this study. The whole Pelvic Bones (PB) was taken as substitute for bone marrow. For every patient, Pareto-optimal plans were generated to explore the trade-off between rectum, bladder, and PB mean dose. The PB mean dose was decreased in steps of 1 Gy. For each step, the increase in rectum and bladder mean dose was quantified. The increase in mean dose of other OAR compared to no BMS was constrained to 1 Gy. Results: In total, 931 plans of 19 evaluable patients were analyzed. The average [range] mean dose of PB without BMS was 22.8 [20.7-26.2] Gy. When maximum BMS was applied, the average reduction in mean PB dose was 5.4 [3.0-6.8] Gy resulting in an average mean PB dose of 17.5 [15.8-19.8] Gy. For <1 Gy increase in both the bladder and the rectum mean dose, the PB mean dose could be decreased by >2 Gy, >3 Gy, >4 Gy, and >5 Gy for 19/19, 13/19, 5/19, and 1/19 patients, respectively. Conclusion: Based on the comprehensive three-dimensional Pareto front analysis, we conclude that 2-5 Gy BMS can be implemented without a clinically relevant increase in mean dose to other OAR. If BMS is too dominant, it results in a large increase in mean dose to other OAR. Therefore, we recommend implementing moderate BMS for the treatment of LACC patients with VMAT.

Robustness analysis of CTV and OAR dose in clinical PBS-PT of neuro-oncological tumors: prescription-dose calibration and inter-patient variation with the Dutch proton robustness evaluation protocol.

Objective. The Dutch proton robustness evaluation protocol prescribes the dose of the clinical target volume (CTV) to the voxel-wise minimum (VWmin) dose of 28 scenarios. This results in a consistent but conservative near-minimum CTV dose (D98%,CTV). In this study, we analyzed (i) the correlation between VWmin/voxel-wise maximum (VWmax) metrics and actually delivered dose to the CTV and organs at risk (OARs) under the impact of treatment errors, and (ii) the performance of the protocol before and after its calibration with adequate prescription-dose levels.Approach. Twenty-one neuro-oncological patients were included. Polynomial chaos expansion was applied to perform a probabilistic robustness evaluation using 100,000 complete fractionated treatments per patient. Patient-specific scenario distributions of clinically relevant dosimetric parameters for the CTV and OARs were determined and compared to clinical VWmin and VWmax dose metrics for different scenario subsets used in the robustness evaluation protocol.Main results. The inclusion of more geometrical scenarios leads to a significant increase of the conservativism of the protocol in terms of clinical VWmin and VWmax values for the CTV and OARs. The protocol could be calibrated using VWmin dose evaluation levels of 93.0%-92.3%, depending on the scenario subset selected. Despite this calibration of the protocol, robustness recipes for proton therapy showed remaining differences and an increased sensitivity to geometrical random errors compared to photon-based margin recipes.Significance. The Dutch proton robustness evaluation protocol, combined with the photon-based margin recipe, could be calibrated with a VWmin evaluation dose level of 92.5%. However, it shows limitations in predicting robustness in dose, especially for the near-maximum dose metrics to OARs. Consistent robustness recipes could improve proton treatment planning to calibrate residual differences from photon-based assumptions.

HYpofractionated, dose-redistributed RAdiotherapy with protons and photons to combat radiation-induced immunosuppression in head and neck squamous cell carcinoma: study protocol of the phase I HYDRA trial.

BACKGROUND: Radiotherapy (RT) is the standard of care for most advanced head and neck squamous cell carcinoma (HNSCC) and results in an unfavorable 5-year overall survival of 40%. Despite strong biological rationale, combining RT with immune checkpoint inhibitors does not result in a survival benefit. Our hypothesis is that the combination of these individually effective treatments fails because of radiation-induced immunosuppression and lymphodepletion. By integrating modern radiobiology and innovative radiotherapy concepts, the patient's immune system could be maximally retained by (1) increasing the dose per fraction so that the total dose and number of fractions can be reduced (HYpofractionation), (2) redistributing the radiation dose towards a higher peak dose within the tumor center and a lowered elective lymphatic field dose (Dose-redistribution), and (3) using RAdiotherapy with protons instead of photons (HYDRA). METHODS: The primary aim of this multicenter study is to determine the safety of HYDRA proton- and photon radiotherapy by conducting two parallel phase I trials. Both HYDRA arms are randomized with the standard of care for longitudinal immune profiling. There will be a specific focus on actionable immune targets and their temporal patterns that can be tested in future hypofractionated immunoradiotherapy trials. The HYDRA dose prescriptions (in 20 fractions) are 40 Gy elective dose and 55 Gy simultaneous integrated boost on the clinical target volume with a 59 Gy focal boost on the tumor center. A total of 100 patients (25 per treatment group) will be recruited, and the final analysis will be performed one year after the last patient has been included. DISCUSSION: In the context of HNSCC, hypofractionation has historically only been reserved for small tumors out of fear for late normal tissue toxicity. To date, hypofractionated radiotherapy may also be safe for larger tumors, as both the radiation dose and volume can be reduced by the combination of advanced imaging for better target definition, novel accelerated repopulation models and high-precision radiation treatment planning and dose delivery. HYDRA's expected immune-sparing effect may lead to improved outcomes by allowing for future effective combination treatment with immunotherapy. TRIAL REGISTRATION: The trial is registered at ClinicalTrials.gov; NCT05364411 (registered on May 6th, 2022).

Development of a local dose-response relationship for osteoradionecrosis within the mandible.

PURPOSE: Osteoradionecrosis (ORN) of the mandible is a severe complication following radiotherapy of the head and neck, but not all regions of the mandible may be equally at risk. Therefore our goal was to explore a local dose response relationship for subregions of the mandible. MATERIALS AND METHODS: All oropharyngeal cancer patients treated at our hospital between 2009 and 2016 were reviewed. Follow-up was cut-off at 3 years. For patients that developed ORN, the ORN volume was delineated on the planning CT. Each mandible was divided into 16 volumes of interest (VOIs) based on the location of the dental elements and the presence of ORN in each was scored. Generalized estimating equations were used to build a model for the probability of developing ORN in an element VOI. RESULTS: Of the 219 included patients, 22 developed ORN in 89 element VOIs. Mean dose to the element VOI (odds ratio (OR) = 1.05 per Gy, 95% confidence interval (CI): (1.04,1.07)), pre-radiotherapy extractions of an element ipsilateral to element of interest (OR = 2.81, 95% CI: (1.12,7.05)), and smoking at start of radiotherapy (OR = 3.37, 95% CI: (1.29,8.78)) were significantly associated with an increased probability of ORN in the VOI. CONCLUSION: The developed dose-response model indicates that the probability of ORN varies within the mandible and strongly depends on the local dose, the location of extractions, and smoking.

PTV-based VMAT vs. robust IMPT for head-and-neck cancer: A probabilistic uncertainty analysis of clinical plan evaluation with the Dutch model-based selection.

BACKGROUND AND PURPOSE: In the Netherlands, head-and-neck cancer (HNC) patients are referred for proton therapy (PT) through model-based selection (MBS). However, treatment errors may compromise adequate CTV dose. Our aims are: (i) to derive probabilistic plan evaluation metrics on the CTV consistent with clinical metrics; (ii) to evaluate plan consistency between photon (VMAT) and proton (IMPT) planning in terms of CTV dose iso-effectiveness and (iii) to assess the robustness of the OAR doses and of the risk toxicities involved in the MBS. MATERIALS AND METHODS: Sixty HNC plans (30 IMPT/30 VMAT) were included. A robustness evaluation with 100,000 treatment scenarios per plan was performed using Polynomial Chaos Expansion (PCE). PCE was applied to determine scenario distributions of clinically relevant dosimetric parameters, which were compared between the 2 modalities. Finally, PCE-based probabilistic dose parameters were derived and compared to clinical PTV-based photon and voxel-wise proton evaluation metrics. RESULTS: Probabilistic dose to near-minimum volume v = 99.8% for the CTV correlated best with clinical PTV-D98% and VWmin-D98%,CTV doses for VMAT and IMPT respectively. IMPT showed slightly higher nominal CTV doses, with an average increase of 0.8 GyRBE in the median of the D99.8%,CTV distribution. Most patients qualified for IMPT through the dysphagia grade II model, for which an average NTCP gain of 10.5 percentages points (%-point) was found. For all complications, uncertainties resulted in moderate NTCP spreads lower than 3 p.p. on average for both modalities. CONCLUSION: Despite the differences between photon and proton planning, the comparison between PTV-based VMAT and robust IMPT is consistent. Treatment errors had a moderate impact on NTCPs, showing that the nominal plans are a good estimator to qualify patients for PT.

Advances in radiotherapy and its impact on second primary cancer risk: A multi-center cohort study in prostate cancer patients.

BACKGROUND: Modelling studies suggest that advanced intensity-modulated radiotherapy may increase second primary cancer (SPC) risks, due to increased radiation exposure of tissues located outside the treatment fields. In the current study we investigated the association between SPC risks and characteristics of applied external beam radiotherapy (EBRT) protocols for localized prostate cancer (PCa). METHODS: We collected EBRT protocol characteristics (2000-2016) from five Dutch RT institutes for the 3D-CRT and advanced EBRT era (N = 7908). From the Netherlands Cancer Registry we obtained patient/tumour characteristics, SPC data, and survival information. Standardized incidence ratios (SIR) were calculated for pelvis and non-pelvis SPC. Nationwide SIRs were calculated as a reference, using calendar period as a proxy to label 3D-CRT/advanced EBRT. RESULTS: From 2000-2006, 3D-CRT with 68-78 Gy in 2 Gy fractions, delivered with 10-23 MV and weekly portal imaging was the most dominant protocol. By the year 2010 all institutes routinely used advanced EBRT (IMRT, VMAT, tomotherapy), mainly delivering 78 Gy in 2 Gy fractions, using various kV/MV imaging protocols. Sixteen percent (N = 1268) developed ≥ 1 SPC. SIRs for pelvis and non-pelvis SPC (all institutes, advanced EBRT vs 3D-CRT) were 1.17 (1.00-1.36) vs 1.39 (1.21-1.59), and 1.01 (0.89-1.07) vs 1.03 (0.94-1.13), respectively. Nationwide non-pelvis SIR was 1.07 (1.01-1.13) vs 1.02 (0.98-1.07). Other RT protocol characteristics did not correlate with SPC endpoints. CONCLUSION: None of the studied RT characteristics of advanced EBRT was associated with increased out-of-field SPC risks. With constantly evolving EBRT protocols, evaluation of associated SPC risks remains important.

Early and late contrast enhancing lesions after photon radiotherapy for IDH mutated grade 2 diffuse glioma.

OBJECTIVE: The interpretation of new enhancing lesions after radiotherapy for diffuse glioma remains a clinical challenge. We sought to characterize and classify new contrast enhancing lesions in a historical multicenter cohort of patients with IDH mutated grade 2 diffuse glioma treated with photon therapy. METHODS: We reviewed all follow-up MRI's of all patients treated with radiotherapy for histologically confirmed, IDH mutated diffuse grade 2 glioma between 1-1-2007 and 31-12-2018 in two tertiary referral centers. Disease progression (PD) was defined in accordance with the RANO criteria for progressive disease in low grade glioma. Pseudoprogression (psPD) was defined as any transient contrast-enhancing lesion between the end of radiotherapy and PD, or any new contrast-enhancing lesion that remained stable over a period of 12 months in patients who did not exhibit PD. RESULTS: A total of 860 MRI's of 106 patients were reviewed. psPD was identified in 24 patients (23%) on 76 MRI's. The cumulative incidence of psPD was 13% at 1 year, 22% at 5 years, and 28% at 10 years. The mean of the observed maximal volume of psPD was 2.4 cc. The median Dmin in psPD lesions was 50.1 Gy. The presence of an 1p/19q codeletion was associated with an increased risk of psPD (subhazard ratio 2.34, p = 0.048). psPD was asymptomatic in 83% of patients. CONCLUSION: The cumulative incidence of psPD in grade 2 diffuse glioma increases over time. Consensus regarding event definition and statistical analysis is needed for comparisons between series investigating psPD.

A probabilistic deep learning model of inter-fraction anatomical variations in radiotherapy.

Objective. In radiotherapy, the internal movement of organs between treatment sessions causes errors in the final radiation dose delivery. To assess the need for adaptation, motion models can be used to simulate dominant motion patterns and assess anatomical robustness before delivery. Traditionally, such models are based on principal component analysis (PCA) and are either patient-specific (requiring several scans per patient) or population-based, applying the same set of deformations to all patients. We present a hybrid approach which, based on population data, allows to predict patient-specific inter-fraction variations for an individual patient.Approach. We propose a deep learning probabilistic framework that generates deformation vector fields warping a patient's planning computed tomography (CT) into possible patient-specific anatomies. This daily anatomy model (DAM) uses few random variables capturing groups of correlated movements. Given a new planning CT, DAM estimates the joint distribution over the variables, with each sample from the distribution corresponding to a different deformation. We train our model using dataset of 312 CT pairs with prostate, bladder, and rectum delineations from 38 prostate cancer patients. For 2 additional patients (22 CTs), we compute the contour overlap between real and generated images, and compare the sampled and 'ground truth' distributions of volume and center of mass changes.Results. With a DICE score of 0.86 ± 0.05 and a distance between prostate contours of 1.09 ± 0.93 mm, DAM matches and improves upon previously published PCA-based models, using as few as 8 latent variables. The overlap between distributions further indicates that DAM's sampled movements match the range and frequency of clinically observed daily changes on repeat CTs.Significance. Conditioned only on planning CT values and organ contours of a new patient without any pre-processing, DAM can accurately deformations seen during following treatment sessions, enabling anatomically robust treatment planning and robustness evaluation against inter-fraction anatomical changes.