Stereotactic radiosurgery and radiotherapy for brainstem metastases: An international multicenter analysis.

Brainstem metastases (BSM) present a significant neuro-oncological challenge, resulting in profound neurological deficits and poor survival outcomes. Stereotactic radiosurgery (SRS) and fractionated stereotactic radiotherapy (FSRT) offer promising therapeutic avenues for BSM despite their precarious location. This international multicenter study investigates the efficacy and safety of SRS and FSRT in 136 patients with 144 BSM treated at nine institutions from 2005 to 2022. The median radiographic and clinical follow-up periods were 6.8 and 9.4 months, respectively. Predominantly, patients with BSM were managed with SRS (69.4%). The median prescription dose and isodose line for SRS were 18 Gy and 65%, respectively, while for FSRT, the median prescription dose was 21 Gy with a median isodose line of 70%. The 12-, 24-, and 36-month local control (LC) rates were 82.9%, 71.4%, and 61.2%, respectively. Corresponding overall survival rates at these time points were 61.1%, 34.7%, and 19.3%. In the multivariable Cox regression analysis for LC, only the minimum biologically effective dose was significantly associated with LC, favoring higher doses for improved control (in Gy, hazard ratio [HR]: 0.86, p < .01). Regarding overall survival, good performance status (Karnofsky performance status, ≥90%; HR: 0.43, p < .01) and prior whole brain radiotherapy (HR: 2.52, p < .01) emerged as associated factors. In 14 BSM (9.7%), treatment-related adverse events were noted, with a total of five (3.4%) radiation necrosis. SRS and FSRT for BSM exhibit efficacy and safety, making them suitable treatment options for affected patients.

Robotic stereotactic body radiotherapy for localized prostate cancer: final analysis of the German HYPOSTAT trial.

PURPOSE: We report results of the first German prospective multicenter single-arm phase II trial (ARO 2013-06; NCT02635256) of hypofractionated robotic stereotactic body radiotherapy (SBRT) for patients with localized prostate cancer (HYPOSTAT). METHODS: Patients eligible for the HYPOSTAT study had localized prostate cancer (cT1­3 cN0 cM0), Gleason score ≤ 7, prostate-specific antigen (PSA) ≤ 15 ng/ml, prostate volume ≤ 80 cm3, and an International Prostate Symptom Score (IPSS) ≤ 12. Initially, inclusion was limited to patients ≥ 75 years or patients 70-74 years with additional risk factors. The trial protocol was later amended to allow for enrolment of patients aged ≥ 60 years. The treatment consisted of 35 Gy delivered in 5 fractions to the prostate and for intermediate- or high-risk patients, also to the proximal seminal vesicles using the CyberKnife system (Accuray Inc., Sunnyvale, CA, USA). Primary endpoint was the rate of treatment-related gastrointestinal or genitourinary grade ≥ 2 toxicity based on the RTOG scale 12-15 months after treatment. Secondary endpoints were acute toxicity, late toxicity, urinary function, quality of life, and PSA response. RESULTS: From July 2016 through December 2018, 85 eligible patients were enrolled and received treatment, of whom 83 could be evaluated regarding the primary endpoint. Patients mostly had intermediate-risk disease with a median PSA value of 7.97 ng/ml and Gleason score of 7a and 7b in 43.5% and 25.9% of patients, respectively. At the final follow-up 12-15 months after treatment, no patient suffered from treatment-related gastrointestinal or genitourinary grade ≥ 2 toxicity. Acute toxicity was mostly mild, with three grade 3 events, and the cumulative rate of grade ≥ 2 genitourinary toxicity was 8.4% (95% CI 4.1-16.4%). There were no major changes in urinary function or quality of life. The median PSA value dropped to 1.18 ng/ml 12-15 months after treatment. There was one patient who developed distant metastases. CONCLUSION: Robotic SBRT with 35 Gy in 5 fractions was associated with a favorable short-term toxicity profile. Recruitment for the HYPOSTAT­2 trial (ARO-2018­4; NCT03795337), which further analyses the late toxicity of this regimen with a planned sample size of 500 patients, is ongoing.

Single-fraction prostate-specific membrane antigen positron emission tomography- and multiparametric magnetic resonance imaging-guided stereotactic body radiotherapy for prostate cancer local recurrences.

OBJECTIVE: To analyse the efficacy and safety of focal prostate-specific membrane antigen positron emission tomography (PSMA-PET)- and multiparametric magnetic resonance imaging (mpMRI)-guided single-fraction stereotactic body radiotherapy (SBRT) for the treatment of prostate cancer (PCa) local recurrences. PATIENTS AND METHODS: Patients with PSMA-PET-positive PCa local recurrences treated with single-fraction SBRT between 2016 and 2020 were included. Identification for subsequent recurrences or metastatic spread based on increasing prostate-specific antigen (PSA) levels were evaluated using PSMA-PET imaging. RESULTS: A total of 64 patients were identified. Patients received various treatments before SBRT (31 patients with radical prostatectomy [RP], 18 external beam radiotherapy [EBRT] with RP, five EBRT, and the remaining 10 other combinations). The median follow-up was 21.6 months. The median PSA level before SBRT was 1.47 ng/mL. All patients received a single-fraction treatment with a median prescription dose and isodose line of 21 Gy and 65%, respectively. At the time of SBRT, six patients (9%) received an androgen deprivation therapy (ADT). PSA levels decreased after SBRT (P = 0.03) and three local recurrences were detected during the follow-up. The progression-free survival after 1-, 2-, and 3-years was 85.3%, 65.9%, and 51.2%, respectively. Six patients (9%) started ADT after SBRT due to disease progression. The rates of newly started ADT after 1-, 2-, and 3-years were 1.8%, 7.3%, and 22.7%, respectively. Grade 1 or 2 toxicities occurred in six patients (9%); no high-grade toxicity was observed. CONCLUSION: While the available data for SBRT in the PCa local recurrence setting describe outcomes for fractionated irradiations, the findings of this first analysis of single-fraction, PSMA-PET- and mpMRI-guided focal SBRT are encouraging. Such treatment appears to be a safe, efficient, and time-saving therapy even in intensively pretreated patients. Recurrence-directed treatments can delay the use of ADT and could avoid prostate bed irradiation in selected patients.

Proton beam radiation therapy for vestibular schwannomas-tumor control and hearing preservation rates: a systematic review and meta-analysis.

OBJECTIVE: Proton beam therapy is considered, by some authors, as having the advantage of delivering dose distributions more conformal to target compared with stereotactic radiosurgery (SRS). Here, we performed a systematic review and meta-analysis of proton beam for VSs, evaluating tumor control and cranial nerve preservation rates, particularly with regard to facial and hearing preservation. METHODS: We reviewed, using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) articles published between 1968 and September 30, 2022. We retained 8 studies reporting 587 patients. RESULTS: Overall rate of tumor control (both stability and decrease in volume) was 95.4% (range 93.5-97.2%, p heterogeneity= 0.77, p<0.001). Overall rate of tumor progression was 4.6% (range 2.8-6.5%, p heterogeneity < 0.77, p<0.001). Overall rate of trigeminal nerve preservation (absence of numbness) was 95.6% (range 93.5-97.7%, I2 = 11.44%, p heterogeneity= 0.34, p<0.001). Overall rate of facial nerve preservation was 93.7% (range 89.6-97.7%, I2 = 76.27%, p heterogeneity<0.001, p<0.001). Overall rate of hearing preservation was 40.6% (range 29.4-51.8%, I2 = 43.36%, p heterogeneity= 0.1, p<0.001). CONCLUSION: Proton beam therapy for VSs achieves high tumor control rates, as high as 95.4%. Facial rate preservation overall rates are 93%, which is lower compared to the most SRS series. Compared with most currently reported SRS techniques, proton beam radiation therapy for VSs does not offer an advantage for facial and hearing preservation compared to most of the currently reported SRS series.

Implementation of model explainability for a basic brain tumor detection using convolutional neural networks on MRI slices.

PURPOSE: While neural networks gain popularity in medical research, attempts to make the decisions of a model explainable are often only made towards the end of the development process once a high predictive accuracy has been achieved. METHODS: In order to assess the advantages of implementing features to increase explainability early in the development process, we trained a neural network to differentiate between MRI slices containing either a vestibular schwannoma, a glioblastoma, or no tumor. RESULTS: Making the decisions of a network more explainable helped to identify potential bias and choose appropriate training data. CONCLUSION: Model explainability should be considered in early stages of training a neural network for medical purposes as it may save time in the long run and will ultimately help physicians integrate the network's predictions into a clinical decision.

CyberKnife robotic spinal radiosurgery in prone position: dosimetric advantage due to posterior radiation access?

CyberKnife spinal radiosurgery suffers from a lack of posterior beams due to workspace limitations. This is remedied by a newly available tracking modality for fiducial-free, respiration-compensated spine tracking in prone patient position. We analyzed the potential dosimetric benefit in a planning study. Fourteen exemplary cases were compared in three scenarios: supine (PTV=CTV), prone (PTV=CTV), and prone position with an additional margin (PTV=CTV+2 mm), to incorporate reduced accuracy of respiration-compensated tracking. Target and spinal cord constraints were chosen according to RTOG 0631 protocol for spinal metastases. Plan quality was scored based on four predefined parameters: dose to cord (D0.1cc and D1cc), high dose (V10Gy), and low dose (V4Gy) volume of healthy tissue. Prescription dose was 16 Gy to the highest isodose line encompassing 90% of the target. Results were related to target size and position. All plans fulfilled RTOG 0631 constraints for coverage and dose to cord. When no additional margin was applied, a majority of eight cases benefitted from prone position, mainly due to a reduction of V4Gy by 23% ± 26%. In the 2 mm prone scenario, the benefit was nullified by an average increase of V10Gy by 43% ± 24%, and an increase of D1cc to cord (four cases). Spinal cord D0.1cc was unchanged (< ± 1 Gy) in all but two cases for both prone scenarios. Conformity (nCI) and number of beams were equivalent in all scenarios, but supine plans used a significantly higher number of monitor units (+16%) than prone. Posterior beam access can reduce dose to healthy tissue in CyberKnife spinal radiosurgery when no additional margin is applied. When a target margin of 2 mm is added, this potential gain is lost. Relative anterior-posterior position and size of the target are selection criteria for prone treatment.

Self-Shielding Gyroscopic Radiosurgery for Uveal Melanoma: A First Case Report.

Stereotactic radiosurgery (SRS) is a well-established treatment modality for the management of uveal melanoma, achieving high tumor control and eye retention rates. There are several SRS treatment platforms available, including the recently developed self-shielding gyroscopic radiosurgery (GRS) system. We report the first use of GRS in the treatment of uveal melanoma. We report the treatment of a 63-year-old female patient with a left-sided uveal melanoma. Akinesia of the ocular globe in the orbit was achieved by retrobulbar anesthesia. The treatment plan used six isocenters (three with the 10 mm and three with the 7.5 mm apertures) and 140 beams to cover 99.2% of the planning target volume (PTV) with 21 Gy at the 54% isodose line. Treatment was delivered in a single session with the GRS device. The total workflow time from retrobulbar anesthesia to completion of treatment was 122 minutes. The procedure was flawless, clinically well tolerated by the patient, and reliably performed in an outpatient setting, thus comparable to our published experience with robotic SRS. The evaluation of new radiosurgery treatment platforms is critical to maintaining quality standards and refining future treatments.

Tumor Control Probability and Time-Dose Response Modeling for Stereotactic Radiosurgery of Uveal Melanoma.

INTRODUCTION: Uveal melanoma (UM), while a rare malignancy, stands as the most prevalent intraocular malignancy in adults. Controversies persist regarding the dose dependency of local control (LC) through radiotherapy. This study seeks to elucidate the significance of the prescription dose by employing time-dose response models for UM patients receiving photon-based stereotactic radiosurgery (SRS). MATERIALS AND METHODS: Inclusion criteria comprised UM patients treated between 2005 and 2019. All patients underwent single-fraction SRS. Datapoints were separated into three dose groups, with Kaplan-Meier analysis performed on each group, from which time-dose response models for LC were created at 2, 4, and 7 years using maximum-likelihood fitted logistic models. RESULTS: Outcomes from 594 patients with 594 UM were used to create time-dose response models. The prescribed doses and the number of patients were as follows: 17-19 Gy (24 patients), 20 Gy (122 patients), 21 Gy (442 patients), and 22 Gy (6 patients). Averaged over all patients and doses, LC rates at 2, 4, and 7 years were 94.4%, 88.2%, and 69.0%, respectively. Time-dose response models for LC demonstrated a dose-dependent effect, showing 2-year LC rates of more than 90% with 20 Gy and 95% with 22 Gy. For four years and a LC of 90%, a dose of approximately 21 Gy was required. After seven years, the 21 Gy prescription dose is predicted to maintain a LC above 70%, sharply declining to less than 60% LC with 19 Gy and less than 40% with 18 Gy. CONCLUSION: In contrast to prior findings, the time-dose response models for UM undergoing photon-based SRS emphasize the critical role of the prescription dose in achieving lasting LC. The dose selection must be carefully balanced against toxicity risks, considering tumor geometry and individual patient characteristics to tailor treatments accordingly.

Accuracy of robotic radiosurgery in renal cell carcinoma.

PURPOSE: Although emerging clinical evidence supports robotic radiosurgery as a highly effective treatment option for renal cell carcinoma (RCC) less than 4 cm in diameter, delivery uncertainties and associated target volume margins have not been studied in detail. We assess intrafraction tumor motion patterns and accuracy of robotic radiosurgery in renal tumors with real-time respiratory tracking to optimize treatment margins. METHODS: Delivery log files from 165 consecutive treatments of RCC were retrospectively analyzed. Five components were considered for planning target volume (PTV) margin estimation: (a) The model error from the correlation model between patient breath and tumor motion, (b) the prediction error from an algorithm predicting the patient breathing pattern, (c) the targeting error from the treatment robot, (d) the inherent total accuracy of the system for respiratory motion tracking, and (e) the margin required to cover potential target rotation, simulated with PTV rotations up to 10°. RESULTS: The median tumor motion was 10.5 mm, 2.4 mm and 4.4 mm in the superior-inferior, left-right, and anterior-posterior directions, respectively. The root of the sum of squares of all contributions to the system's inaccuracy results in a minimum PTV margin of 4.3 mm, 2.6 mm and 3.0 mm in the superior-inferior, left-right and anterior-posterior directions, respectively, assuming optimal fiducial position and neglecting target deformation. CONCLUSIONS: We have assessed kidney motion and derived PTV margins for the treatment of RCC with robotic radiosurgery, which helps to deliver renal treatments in a more consistent manner and potentially further improve outcomes.

Self-Shielding Gyroscopic Radiosurgery: A Prospective Experience and Analysis of the First 100 Patients.

Background Stereotactic radiosurgery is a well-established treatment option for the management of various benign and malignant brain tumors. It can be delivered with several treatment platforms, usually requiring shielded radiation vaults to meet regulatory safety requirements. Recent technical advances have led to the first self-shielding platform enabling the delivery of gyroscopic radiosurgery (GRS). Given the limited number of GRS treatment platforms, the novelty of its characteristics, and the lack of available data, we report our prospective experience with the first 100 patients treated with GRS. Materials and methods Patients undergoing GRS for the treatment of intracranial tumors were enrolled in this prospective study. Patient and treatment characteristics, including patient satisfaction, were collected and analyzed. Results A total of 100 patients with 155 tumors were treated. The most commonly treated tumors comprised brain metastases (BM) (49%), vestibular schwannomas (31%), and meningiomas (14%). The median prescription dose for malignant and benign tumors was 20 and 13 Gy, respectively. The median prescription isodose line was 56%. Gross tumor volumes were small, with a median of 0.37 cc for BM and 0.92 cc for the other entities. The median total treatment time was 40 minutes. Dosimetric performance indices showed median values of 1.20 (conformity index), 1.24 (new conformity index), 1.74 (homogeneity index), and 3.13 (gradient index). Volumetric assessment of the treated tumors showed an overall decrease in size at the first available follow-up. Most patients were satisfied with the treatment experience. Conclusion Our first prospective experience of the use of GRS is favorable. Analyses of the dosimetric performance, treatment times, volumetric assessment, and patient satisfaction demonstrate its suitability for stereotactic treatments of intracranial tumors. Further prospective clinical and dosimetric analyses for GRS are pending.

Towards fast adaptive replanning by constrained reoptimization for intra-fractional non-periodic motion during robotic SBRT.

BACKGROUND: Periodic and slow target motion is tracked by synchronous motion of the treatment beams in robotic stereotactic body radiation therapy (SBRT). However, spontaneous, non-periodic displacement or drift of the target may completely change the treatment geometry. Simple motion compensation is not sufficient to guarantee the best possible treatment, since relative motion between the target and organs at risk (OARs) can cause substantial deviations of dose in the OARs. This is especially evident when considering the temporally heterogeneous dose delivery by many focused beams which is typical for robotic SBRT. Instead, a reoptimization of the remaining treatment plan after a large target motion during the treatment could potentially reduce the actually delivered dose to OARs and improve target coverage. This reoptimization task, however, is challenging due to time constraints and limited human supervision. PURPOSE: To study the detrimental effect of spontaneous target motion relative to surrounding OARs on the delivered dose distribution and to analyze how intra-fractional constrained replanning could improve motion compensated robotic SBRT of the prostate. METHODS: We solve the inverse planning problem by optimizing a linear program. When considering intra-fractional target motion resulting in a change of geometry, we adapt the linear program to account for the changed dose coefficients and delivered dose. We reduce the problem size by only reweighting beams from the reference treatment plan without motion. For evaluation we simulate target motion and compare our approach for intra-fractional replanning to the conventional compensation by synchronous beam motion. Results are generated retrospectively on data of 50 patients. RESULTS: Our results show that reoptimization can on average retain or improve coverage in case of target motion compared to the reference plan without motion. Compared to the conventional compensation, coverage is improved from 87.83 % to 94.81 % for large target motion. Our approach for reoptimization ensures fixed upper constraints on the dose even after motion, enabling safer intra-fraction adaption, compared to conventional motion compensation where overdosage in OARs can lead to 21.79 % higher maximum dose than planned. With an average reoptimization time of 6 s for 200 reoptimized beams our approach shows promising performance for intra-fractional application. CONCLUSIONS: We show that intra-fractional constrained reoptimization for adaption to target motion can improve coverage compared to the conventional approach of beam translation while ensuring that upper dose constraints on VOIs are not violated.

Meningioma involving the superior sagittal sinus: long-term outcome after robotic radiosurgery in primary and recurrent situation.

Objective: Treatment for meningiomas involving the superior sagittal sinus (SSS) is challenging and proved to be associated with higher risks compared to other brain locations. Therapeutical strategies may be either microsurgical (sub-)total resection or adjuvant radiation, or a combination of both. Thrombosis or SSS occlusion following resection or radiosurgery needs to be further elucidated to assess whether single or combined treatment is superior. We here present tumor control and side effect data of robotic radiosurgery (RRS) in combination with or without microsurgery. Methods: From our prospective database, we identified 137 patients with WHO grade I meningioma involving the SSS consecutively treated between 2005 and 2020. Treatment decisions were interdisciplinary. Patients underwent RRS as initial/solitary treatment (group 1), as adjuvant treatment after subtotal resection (group 2), or due to recurrent tumor growth after preceding microsurgery (group 3). Positive tumor response was assessed by MRI and defined as reduction of more than 50% of volume. Study endpoints were time to recurrence (TTR), time to RRS, risk factors for decreased survival, and side effects. Overall and specific recurrence rates for treatment groups were analyzed. Side effect data included therapy-related morbidity during follow-up (FU). Results: A total of 137 patients (median age, 58.3 years) with SSS meningiomas WHO grade I were analyzed: 51 patients (37.2%) in group 1, 15 patients (11.0%) in group 2, and 71 patients (51.8%) in group 3. Positive MR (morphological response) to therapy was achieved in 50 patients (36.4%), no response was observed in 25 patients (18.2%), and radiological tumor progression was detected in 8 patients (5.8%). Overall 5-year probability of tumor recurrence was 15.8% (median TTR, 41.6 months). Five-year probabilities of recurrence were 0%, 8.3.%, and 21.5% for groups 1-3 (p = 0.06). In multivariate analysis, tumor volume was significantly associated with extent of SSS occlusion (p = 0.026) and sex (p = 0.011). Tumor volume significantly correlated with TTR (p = 0.0046). Acute sinus venous thrombosis or venous congestion-associated bleedings did not occur in any of the groups. Conclusion: RRS for grade I meningiomas with SSS involvement represents a good option as first-line treatment, occasionally also in recurrent and adjuvant scenarios as part of a multimodal treatment strategy.

Automated isocenter optimization approach for treatment planning for gyroscopic radiosurgery.

BACKGROUND: Radiosurgery is a well-established treatment for various intracranial tumors. In contrast to other established radiosurgery platforms, the new ZAP-X® allows for self-shielding gyroscopic radiosurgery. Here, treatment beams with variable beam-on times are targeted towards a small number of isocenters. The existing planning framework relies on a heuristic based on random selection or manual selection of isocenters, which often leads to a higher plan quality in clinical practice. PURPOSE: The purpose of this work is to study an improved approach for radiosurgery treatment planning, which automatically selects the isocenter locations for the treatment of brain tumors and diseases in the head and neck area using the new system ZAP-X® . METHODS: We propose a new method to automatically obtain the locations of the isocenters, which are essential in gyroscopic radiosurgery treatment planning. First, an optimal treatment plan is created based on a randomly selected nonisocentric candidate beam set. The intersections of the resulting subset of weighted beams are then clustered to find isocenters. This approach is compared to sphere-packing, random selection, and selection by an expert planner for generating isocenters. We retrospectively evaluate plan quality on 10 acoustic neuroma cases. RESULTS: Isocenters acquired by the method of clustering result in clinically viable plans for all 10 test cases. When using the same number of isocenters, the clustering approach improves coverage on average by 31 percentage points compared to random selection, 15 percentage points compared to sphere packing and 2 percentage points compared to the coverage achieved with the expert selected isocenters. The automatic determination of location and number of isocenters leads, on average, to a coverage of 97 ± 3% with a conformity index of 1.22 ± 0.22, while using 2.46 ± 3.60 fewer isocenters than manually selected. In terms of algorithm performance, all plans were calculated in less than 2 min with an average runtime of 75 ± 25 s. CONCLUSIONS: This study demonstrates the feasibility of an automatic isocenter selection by clustering in the treatment planning process with the ZAP-X® system. Even in complex cases where the existing approaches fail to produce feasible plans, the clustering method generates plans that are comparable to those produced by expert selected isocenters. Therefore, our approach can help reduce the effort and time required for treatment planning in gyroscopic radiosurgery.

Robotic stereotactic body radiotherapy for the management of adrenal gland metastases: a bi-institutional analysis.

PURPOSE: Adrenal gland metastases (AGMs) are a common manifestation of metastatic tumor spread, especially in non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC). In patients with a limited systemic tumor burden, effective treatments for AGMs are needed. Due to varying fractionation schemes and limited reports, short-course treatment results for stereotactic body radiotherapy (SBRT) for AGMs are lacking. This work analyzes the outcomes of short-course SBRT for AGMs. METHODS: Patients who underwent robotic SBRT for AGMs with one to five fractions were eligible for analysis. RESULTS: In total, data from 55 patients with 72 AGMs from two institutions were analyzed. Most AGMs originated from renal cell carcinoma (38%) and NSCLC (35%). The median follow-up was 16.4 months. The median prescription dose and isodose line were 24 Gy and 70%, respectively. Most patients (85%) received SBRT with just one fraction. The median biologically effective dose assuming an α/ß ratio of 10 (BED10) was 80.4 Gy. The local control and progression-free survival after 1 and 2 years were 92.9%, 67.8%, and 46.2%, as well as 24.3%, respectively. Thirteen patients (24%) suffered from grade 1 or 2 toxicities. The BED10 showed a significant impact on LC (p < 0.01). Treatments with a BED10 equal to or above the median were associated with a better LC (p < 0.01). CONCLUSION: Robotic SBRT is an efficient and safe treatment modality for AGM. Treatment-associated side effects are sporadic and manageable. Results suggest short-course SBRT to be a preferable and time-saving treatment option for the management of AGMs if an adequate BED10 can be safely applied.

AI-based optimization for US-guided radiation therapy of the prostate.

OBJECTIVES: Fast volumetric ultrasound presents an interesting modality for continuous and real-time intra-fractional target tracking in radiation therapy of lesions in the abdomen. However, the placement of the ultrasound probe close to the target structures leads to blocking some beam directions. METHODS: To handle the combinatorial complexity of searching for the ultrasound-robot pose and the subset of optimal treatment beams, we combine CNN-based candidate beam selection with simulated annealing for setup optimization of the ultrasound robot, and linear optimization for treatment plan optimization into an AI-based approach. For 50 prostate cases previously treated with the CyberKnife, we study setup and treatment plan optimization when including robotic ultrasound guidance. RESULTS: The CNN-based search substantially outperforms previous randomized heuristics, increasing coverage from 93.66 to 97.20% on average. Moreover, in some cases the total MU was also reduced, particularly for smaller target volumes. Results after AI-based optimization are similar for treatment plans with and without beam blocking due to ultrasound guidance. CONCLUSIONS: AI-based optimization allows for fast and effective search for configurations for robotic ultrasound-guided radiation therapy. The negative impact of the ultrasound robot on the plan quality can successfully be mitigated resulting only in minor differences.

Self-Shielding Gyroscopic Radiosurgery-A First Clinical Experience, Case Series, and Dosimetric Comparison.

BACKGROUND: Self-shielding gyroscopic radiosurgery (GRS) represents a technical innovation in the field of stereotactic radiosurgery. GRS does not require a radiation vault and is optimized for radiosurgical treatments. Reports on its usage are limited. We describe the first clinical experience of GRS at our institution to assess the application of GRS in the treatment of cranial tumors. Moreover, we perform a dosimetric comparison to robotic radiosurgery (RRS) with vestibular schwannoma (VS) GRS patients. METHODS: Patients who were treated with GRS between July and November 2021 were included. Patient, tumor, and dosimetric characteristics were retrospectively summarized and analyzed. RESULTS: Forty-one patients with 48 intracranial tumors were included. Tumor entities mostly comprised VS, brain metastases, and meningiomas. The median prescription dose and isodose line were 13.5 Gy and 50.0% for benign neoplasia versus 20 Gy and 60.0% for malignant tumors, respectively. The mean planning target volume was 1.5 cubic centimeters. All patients received a single-fraction treatment without encountering any technical setup difficulties. Treatment plan comparisons with RRS revealed comparable plan characteristics, dose gradients, and organs at risk doses. Significant differences were detected concerning the new conformity index and number of monitor units per treatment (both P < 0.01). CONCLUSIONS: This case series provides more evidence on the usage of self-shielding GRS in the management of cranial tumors. Dosimetric comparisons for VS cases revealed mostly equivalent dosimetric characteristics to RRS. Further clinical and physical analyses for GRS are underway.

Propensity Score-Matched Analysis of Single Fraction Robotic Radiosurgery Versus Open Partial Nephrectomy in Renal Cell Carcinoma: Oncological Outcomes.

Introduction High-dose local stereotactic robotic radiosurgery (RRS) is a non-invasive alternative to surgery in renal masses and selected patients. We have, so far, limited its use to the elderly and patients at high risk from surgery. In this study, we matched patients with renal tumors who were treated with single fraction RRS to patients who underwent open partial nephrectomy (OPN). Methods Between January 2009 and December 2017, we included 571 consecutive patients undergoing OPN and 99 patients who underwent RRS in this retrospective analysis. Patients had to have a follow-up of at least six months and we were able to match 35 with a propensity score. Matching criteria were Eastern Cooperative Oncology Group (ECOG) status, age, clinical tumor, nodes, and metastases (TNM), and tumor diameter. Tumor response, renal function, survival, and adverse events were evaluated every three months until progression or death. Results Median age was 65 years for RRS (range 58-75) and 71 (range 56-76) for OPN (p=0.131). Median diameter of renal tumors was 2.8 cm (range 2.4-3.9) for RRS and 3.5 cm (2.8-4.5) for OPN, p=0.104. Median follow-up was 28.1 months (range 6.0-78.3 months). Local tumor control nine months after RRS and OPN was 98% (95% CI: 89-99%). Renal function remained stable with a median creatinine clearance (Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI)) at baseline of 76.8mlmin/1.73m2 (range 25.3-126.3) and 70.3ml/min/1.73m2 (range 18.6-127.3) at follow-up (p=0.89). Median overall survival was not reached. No difference in overall survival (OS) was seen in RRS compared to OPN (p=0.459). Conclusions Single fraction RRS is an alternative to OPN in patients unfit for surgery. Oncological and functional results are comparable to those of OPN. Further studies are needed to determine long-term results and limits of RRS in this setting and in younger patients.

Image-Guided Robotic Radiosurgery for the Treatment of Lung Metastases of Renal Cell Carcinoma-A Retrospective, Single Center Analysis.

Pulmonary metastases are the most frequent site of metastases in renal cell carcinoma (RCC). Metastases directed treatment remains an important treatment option despite advances in systemic therapies. However, the safety and efficacy of robotic radiosurgery (RRS) for the treatment of lung metastases of RCC remains unclear. Patients with metastatic RCC and lung metastases treated by RRS were retrospectively analyzed for overall survival (OS), progression-free survival (PFS), local recurrence free survival (LRFS) and adverse events. The Kaplan-Meier method was used for survival analysis and the common terminology criteria for adverse events (CTCAE; Version 5.0) classification for assessment of adverse events. A total of 50 patients were included in this study. Median age was 64 (range 45-92) years at the time of RRS. Prior to RRS, 20 patients (40.0%) had received either tyrosine kinase inhibitors or immunotherapy and 27 patients (54.0%) were treatment naïve. In our patient cohort, the median PFS was 13 months (range: 2-93). LRFS was 96.7% after two years with only one patient revealing progressive disease of the treated metastases 13 months after RRS. Median OS was 35 months (range 2-94). Adverse events were documented in six patients (12%) and were limited to grade 2. Fatigue (n = 4) and pneumonitis (n = 2) were observed within 3 months after RRS. In conclusion, RRS is safe and effective for patients with metastatic RCC and pulmonary metastases. Radiation induced pneumonitis is specific in the treatment of pulmonary lesions, but not clinically relevant and survival rates seem favorable in this highly selected patient cohort. Future directions are the implementation of RRS in multimodal treatment approaches for oligometastatic or oligoprogressive disease.

Intracranial Hemorrhage in Patients with Anticoagulant Therapy Undergoing Stereotactic Radiosurgery for Brain Metastases: A Bi-Institutional Analysis.

BACKGROUND: Stereotactic radiosurgery (SRS) is a well-established treatment modality for brain metastases (BM). Given the manifold implications of metastatic cancer on the body, affected patients have an increased risk of comorbidities, such as atrial fibrillation (AF) and venous thromboembolism (VTE), which includes pulmonary embolism (PE) and deep-vein thrombosis (DVT). These may require therapeutic anticoagulant therapy (ACT). Limited data are available on the risk of intracranial hemorrhage (ICH) after SRS for patients with BM who are receiving ACT. This bi-institutional analysis aimed to describe the bleeding risk for this patient subgroup. METHODS: Patients with ACT at the time of single-fraction SRS for BM from two institutions were eligible for analysis. The cumulative incidence of ICH with death as a competing event was assessed during follow-up with magnetic resonance imaging or computed tomography. RESULTS: Forty-one patients with 97 BM were included in the analyses. The median follow-up was 8.2 months (range: 1.7-77.5 months). The median and mean BM volumes were 0.47 and 1.19 cubic centimeters, respectively. The most common reasons for ACT were PE (41%), AF (34%), and DVT (7%). The ACT was mostly performed utilizing phenprocoumon (37%), novel oral anticoagulants (32%), or low-molecular-weight heparin (20%). Nine BM from a group of five patients with ICH after SRS were identified: none of them caused neurological or any other deficits. The 6-, 12-, and 18-month cumulative bleeding incidences per metastasis were 2.1%, 12.4%, and 12.4%, respectively. The metastases with previous bleeding events and those originating from malignant melanomas were found to more frequently demonstrate ICH after SRS (p = 0.02, p = 0.01). No surgical or medical intervention was necessary for ICH management, and no observed death was associated with an ICH. CONCLUSION: Patients receiving an ACT and single-fraction SRS for small- to medium-sized BM did not seem to have a clinically relevant risk of ICH. Previous bleeding and metastases originating from a malignant melanoma may favor bleeding events after SRS. Further studies are needed to validate our reported findings.

Convolutional Neural Networks to Detect Vestibular Schwannomas on Single MRI Slices: A Feasibility Study.

In this study. we aimed to detect vestibular schwannomas (VSs) in individual magnetic resonance imaging (MRI) slices by using a 2D-CNN. A pretrained CNN (ResNet-34) was retrained and internally validated using contrast-enhanced T1-weighted (T1c) MRI slices from one institution. In a second step, the model was externally validated using T1c- and T1-weighted (T1) slices from a different institution. As a substitute, bisected slices were used with and without tumors originating from whole transversal slices that contained part of the unilateral VS. The model predictions were assessed based on the categorical accuracy and confusion matrices. A total of 539, 94, and 74 patients were included for training, internal validation, and external T1c validation, respectively. This resulted in an accuracy of 0.949 (95% CI 0.935-0.963) for the internal validation and 0.912 (95% CI 0.866-0.958) for the external T1c validation. We suggest that 2D-CNNs might be a promising alternative to 2.5-/3D-CNNs for certain tasks thanks to the decreased demand for computational power and the fact that there is no need for segmentations. However, further research is needed on the difference between 2D-CNNs and more complex architectures.