Purpose/Objective

Brain metastases are common in patients with metastatic melanoma. Poor local control and cognitive consequences mean that the indications for Radiosurgery continue to increase. We analyze the effectiveness of treatment with radiosurgery. We try to evaluate the safety and initial response of brain metastases treated with immunotherapy or a combination of targeted therapy and radiosurgery with gamma knife.

Material/Methods

We present our experience with 25 treatment sessions in the Leksell gamma unit in 8 patients with metastatic melanoma, 7 of whom underwent treatment concurrently with immunotherapy or a combination of targeted therapies.

Results

The population was composed of 7 men and 1 woman. The average age of the patients was 64 years (age range, 42-75 years). The median time from diagnosis of primary melanoma to discovery of brain metastasis was 35 months (range, 1-132 months). At the time of diagnosis of the brain disease, 50% of the patients had neurological symptoms, only one patient debuted with seizures. 95% developed lymph node metastases. Eighty-six percent of the lesions were cortical, 13% were cerebellar, 1% were thalamic. 88% of the sessions (22 of 25 sessions) were treated in a single session. The mean treatment volume was 3 cc, with a mean prescription of 22 Gy up to the mean 60% isodose line. Median survival was 42 months from the time of diagnosis of primary melanoma and 7 months from Gamma knife radiosurgery. No complications occurred within 24 hours after the procedure; 3 of the patients presented with mild headache, nausea and transient vomiting. Our series has a short following. There was 1 death due to intracranial hemorrhage in the third month of treatment and in relation to disease progression. In the magnetic resonance images, 3 patients with at least 5 treated lesions progressed after the first 3-month control, the rest of them remain with stable disease. Tolerance to treatment was good.

Conclusion

Concurrent treatment with immunotherapy or a combination of targeted therapy and gamma knife radiosurgery does not seem to increase toxicity; in our series, those patients who have progressed the earliest are those with 5 or more lesions.

[Objectives] Leksell Gamma Knife Icon enables us to apply new methods of immobilization using mask fixation and the option of fractionated treatment. We compared the treatment results of brain metastases with those of the primary tumor.

[Methods] We retrospectively analyzed 1635 patients (a total of 2394 treatments) with brain metastases who underwent Gamma Knife Icon using mask fixation between September 25th, 2017 and December 31th, 2023 at Rakusai Shimizu Hospital. Patients with small, few, newly diagnosed, and non-eloquent area tumors were treated in a single session. If the tumor volume was larger than 5.0 ml, recurrence, or the location was in an eloquent area, we applied a fractionated schedule. For large tumors, we selected fractionated schedules as follows; 4.2-4.7Gy x 10Fr (5-20ml), 3.7-4.2Gy x 10Fr (20-30ml), 3.2-3.7Gy x 10Fr (30ml-). If the tumor number was large, we selected a multisession schedule.

[Results] The most common origin was lung (1103 patients, 1646 times), followed by breast (183, 287), gastro-intestinal (GI) tracts (176, 221), urogenital (104, 150), and others (69, 90). Tumor volume tended to be larger in GI tracts and urogenital. Single session tended to be more common in urogenital, and fractionated schedule tended to be more common in GI tract. The median survival time after icon therapy was 27.6 months for lung, 26.6 months for breast, 6.9 months for GI tract, and 14.4 months for urogenital. There were no differences in neurological death, local control, ADL maintenance, and serious complications among the groups.

[Conclusions] Although these results are limited to short periods, survival rates, local control rates and qualitative survival rated in patients unsuitable for stereotactic radiosurgery, such as those with large, recurrent, and eloquent site lesions, were within the acceptable ranges among various primary sites.

Title: Stereotactic Radiosurgery using the Mask System of Leksell Gamma Knife Icon for Patients with Over Ten Brain Metastases.

Authors: Yuta Oi, Gaku Fujiwara, Takuya Kawabe, Manabu Sato
Maizuru Medical Center, Rakusai Shimizu Hospital

Purpose: Following the JLGK0901 report, guidelines for multiple brain metastases have been revised, easing limitations on stereotactic radiotherapy based on tumor numbers. We present initial results of treatment utilizing the Gamma Knife Icon's mask system for over 10 metastatic brain tumors.

Methods: We retrospectively reviewed 280 patients (comprising 329 treatments) with brain metastases who underwent Gamma Knife Icon treatment using mask fixation between September 25th, 2017, and December 31st, 2023, at Rakusai Shimizu Hospital. The cohort included 146 males and 134 females, with a median age of 69 years (range: 20-93). The most prevalent primary tumor sites were lung (207 patients, 245 treatments), followed by breast (40, 47), gastro-intestinal (17, 19), and others (16, 18). The median number of tumors was 16 (range: 11-64), with a median maximum tumor volume of 1.3 mL (IQR: 0.3-5.0) and a median cumulative tumor volume of 3.4 mL (IQR: 1.1-10.4).

Patients with large, recurrent, or eloquent lesions received fractionated irradiation. Irradiation time with mask fixation was divided into multiple sessions, approximately 30 minutes each, based on patient comfort. The median treatment time was 31.1 minutes (IQR: 25.8-38.9).

Results: Median survival post-Icon therapy was 9.9 months, with only 3/8/11% experiencing neurological death at 6/12/24 months post-treatment. Local control failure rates were 7/13/23% at 6/12/24 months after treatment. New lesions emerged in 16/56/67/72/76/80% of patients at 3/6/9/12/15/18 months post-treatment, requiring early intervention as deemed necessary. Preservation of neurological function measured at 6/12/24 months post-treatment was 92/86/83%. Serious complications were observed in 0/1/1% at 6/12/24 months post-treatment.

Conclusions: While the medium- to long-term efficacy requires further follow-up, our findings suggest that highly accurate fractionated irradiation and multiple irradiation sessions are feasible using the Gamma Knife Icon, potentially leading to reduced treatment complications.

Purpose: We have developed a novel GK-specific knowledge-based planning (KBP) pipeline utilizing 3-dimensional dose prediction in conjunction with inverse optimization (IO) for the generation of deliverable treatment plans.

Methods: Data was obtained for 349 patients treated for either brain metastases or intracranial schwannomas at Sunnybrook Health Sciences Centre. The data from 322 patients was modified using a GK-specific data modification method, then used to train a neural network model for GK dose prediction. The trained model was then applied to predict dose predictions for 27 out-of-sample patients.

Subsequently, we developed a generalized IO model, based on an established inverse planning model¹, to learn objective function weights from dose predictions. This model was solved using the obtained dose predictions for the out-of-sample patients. The resulting weights were then used in the inverse planning model to generate deliverable treatment plans.

The quality of the resulting KBP plans was compared to manual clinical plans and plans resulting from a dose mimicking (DM) model using standard GK quality metrics and overall treatment time. Finally, we evaluated the overall average usage time of the pipeline and plan delivery characteristics to help determine its potential applicability in a clinical setting.

Results: Across all quality metrics, plans generated using the KBP pipeline performed at least as well as or better than the respective clinical plans. The average conformity and gradient of IO plans were 0.737 ± 0.158 and 3.356 ± 1.030, respectively, compared to 0.713 ± 0.124 and 3.452 ± 1.123 for the clinical plans. IO plans also outperformed DM plans for five of the six quality metrics. Additionally, plans generated using the IO pipeline had an average treatment time comparable to clinical plans.

The average time required to generate deliverable plan using the pipeline was 5 minutes 43 seconds and varied depending on target complexity. Compared to clinical plans, KBP plans utilized block sectors significantly more frequently and 4 mm collimators significantly less frequently. Additionally, KBP plans favor using multiple shots per isocenter in contrast to manual clinical plans, which are based on one shot per isocenter.

Conclusion: Plans resulting from an IO KBP pipeline consistently match or surpass the quality of manual plans. The results demonstrate the potential for the usage of KBP to generate GK treatment plans with minimal human intervention. 

Background:    The safety of a radiosurgery plan may be estimated based upon metrics of radiation dose (e.g. V12Gy) which are only available after a plan has been created and may be a function of the planning system, platform, or planner.   Using a standardized automated treatment planning system for single isocenter multiple target radiosurgery (HyperArc^TM) which produces consistent plan quality generally independent of planning expertise, we hypothesize that geometric metrics of brain metastases will predict V12 and inform the treating team to select radiosurgery prescription prior to planning.

Methods:        HyperArc^TM clinical plans utilizing a single isocenter for all targets were queried to investigate the utility of various pre-planning geometric metrics to predict V12 (prescription doses 20 Gy and 24 Gy) and V18 (9Gy x 3).   A total of 1717 clinical radiosurgery plans included 3399 targets. These plans were generated without an explicit target margin and planned for treatment delivery using an Edge^TM linear accelerator with a high-definition multi-leaf collimator (central resolution 2.5mm).   The target hotspot was not penalized in the optimizer.   The volume of the target was included in the V12 calculations.   V12 was calculated per lesion unless this isodose volume bridged between targets, in which the targets were excluded from the analysis.   Potential predictive geometric measures included target volume, equivalent sphere diameter, largest axial diameter, mesh surface area, and pseudo surface area.

Results:           All the pretreatment geometric metrics had some utility to predict V12.  The best fit to predict V12 (and V18 for 3 fractions) was for target volume.   Linear and power models were generated for various radiation dose schedules.  An example linear equation to predict V12Gy for a 20 Gy prescription is shown below:

V12Gy(cc) = 2.22*targetvolume + 1.84

This linear equation predicts V12Gy of 10cc occurs with a target volume of ~3.7 cc for a 20 Gy prescription for a HyperArc^TM plan treated with the HD-MLC.

Conclusions:   Target volume and other geometric predictors can be utilized to predict dosimetric measures of radiosurgery toxicity for automated single isocenter VMAT (HyperArc^TM) radiosurgery.   This knowledge prior to planning allows the treating team to run a single plan iteration with the optimal prescription.

Background:

Brain metastases can significantly increase patient morbidity. Stereotactic radiosurgery (SRS) is an effective technique for treatment and can improve quality of life. Earlier detection of brain metastases when compared to standard T1 magnetic resonance imaging (MRI) may lead to improved outcomes.

Purpose:

We created a machine learning (ML) model using longitudinal diffusion weighted images (DWI) and radiomics to improve detection of brain metastases in SRS patients.

Methods:

We analyzed 117 patients who had received multiple imaging sessions prior to SRS. Apparent diffusion coefficient (ADC) maps, contrast enhanced Gd-T1 MRI, and clinical computed tomography images from all time points were registered for each patient. Radiomic maps were calculated for every ADC map. Data features were extracted by generating spherical binary masks with a radius of 1 cm and sampling both healthy and metastatic regions within the brain. Difference features were added to the dataset by calculating the radiomic change between imaging sessions. Radiomic stability was used to select features by sampling healthy brain tissue with no known abnormalities. Features that were unstable (mean intraclass correlation coefficient < 0.75) were excluded from further analysis. Clinical features (age, gender, and primary cancer) were included for ML training. Final output labels were based on whether a metastasis was clinically confirmed within a sampled region during imaging.

The dataset was split 80/20 training-validation with stratification. XGBoost was used for training, with hyperparameters tuned using five-fold group cross validation, prioritizing macro-averaged recall. Standard classification metrics were calculated on the unseen validation dataset to assess model performance in detecting metastatic growth from patient ADC maps. A cerebellum-only ML model was created and tested, with cerebrum data to be calculated and trained in the future.

Results:

XGBoost was able to correctly identify metastatic tissue within the cerebellum prior to manifestation on Gd-T1 (balanced accuracy: 74.8±0.4%, recall: 84.7±0.4%). The area under the receiver operating characteristic was 86.0±0.4%. Local ADC intensity variance was the most important radiomic feature. The patient’s age and whether they received a melanoma diagnosis were the most important clinical features for the model to classify potential metastases.

Conclusions:

A DWI-based radiomics model was developed and trained using longitudinal SRS imaging data. Our cerebellum model results suggest that ML can be effective in detecting brain metastases. This can aid clinicians in deciding whether increased monitoring via imaging is recommended. Future work includes training the cerebrum model and prospective testing on patients with high metastatic incidence rates.

Purpose: Advances in systemic therapy for driver mutation-positive lung cancer brain metastases have prolonged prognosis. We investigated treatment strategies using Gamma Knife Icon.

Objective: We retrospectively analyzed 229 patients (a total of 388 treatments) with brain metastases who underwent Gamma Knife Icon using mask fixation between September 25th, 2017 and August 31th, 2023 at Rakusai Shimizu Hospital. Patients with small, few, newly diagnosed, and non-eloquent area tumors were treated in a single session. If the tumor volume was larger than 5.0 ml, recurrence, or the location was in an eloquent area, we applied a fractionated schedule. If the tumor number was large, we selected a multisession schedule. The most common driver mutation was EGFR mutation (193 cases), followed by ALK fusion (23), and others (KRAS, BRAF, ROS-1, RET, MET, ERBB2) (13). There were 87 males and 142 females with a mean age of 69 (34-89) years. Seventy patients had both primary tumor and brain metastasis, and 159 patients had brain metastasis after prior treatment of the primary tumor. The median number of metastases was 4 (IQR: 1-8), and the median maximum lesion volume was 0.6 (IQR: 0.2-2.7) mL.

Results: Single session was performed 118 times, fractionation 156 times, and multisession 114 times. The median survival after Icon therapy was 51.8 (95%CI:47.6-NA) months, including 53.3 months for EGFR-positive patients and 51.8 months for ALK-positive. Neurological death was only 2/4/4/7% at 12/24/36/48 months after treatment. Local control failure was 18/31/38% at 12/24/36 months after treatment. New lesions appeared in 49/65/71% at 12/24/36 months after treatment. Preservation of neurological function was 93/88/86% at 12/24/36 months post-treatment. Serious complications were only 1/1/1% at 12/24/36 months after treatment.

Conclusions: Although these results are limited to short periods, survival rates, local control rates and qualitative survival rated in patients unsuitable for stereotactic radiosurgery, such as those with large, recurrent, and eloquent site lesions, were within the acceptable ranges.

Objective:

We compare the clinical outcome of patients with <10 radiosurgically treated brain metastases (BM) to patients with ≥10 BM.

Methods:

A retrospective analysis of all patients with an age >18 years, at least one Gamma Knife radiosurgical treatment (GKRS) for at least one BM between 2012 and 2022 and at least one follow-up was performed. Based on the number of BM on the planning MRI, the patients were divided into two groups: 1) <10 BM and 2) ≥10 BM.

Results:

In our study population, 1253 patients with radiosurgically treated BM from different primary tumors (lung cancer = 795/1253, 63%; melanoma = 261/1253, 16% and breast cancer = 197/1253, 21%) could be identified. 

At the time of first GKRS treatment (GKRS1), 115/1253 (9%) patients had more than 10 BM. The estimated median survival after GKRS1 did not show any significant differences between patients with <10 BM and ≥10 BM, even after analyzing for each primary tumor. Furthermore, even in patients with worse clinical condition, defined as a Karnofsky Performance Status Scale of <80%, the estimated median survival after GKRS1 did not differ between patients with <10 and ≥10 BM.

Conclusion: 

GKRS represents an effective treatment option for patients with multiple BM, even with more than 10 BM. 

Objectives: The management of large posterior fossa metastases presents a unique challenge in neuro-oncology, demanding an approach that balances efficacy, safety, and preservation of neurological function. Traditionally, the treatment paradigm for large brain metastases, particularly those in the posterior fossa, has heavily relied on surgical resection. In recent years, hypofractionated Gamma Knife radiosurgery (hf-GKRS) has emerged as a promising modality, offering a targeted, minimally invasive approach with a favorable side-effect profile. This retrospective, single-center study evaluated patient outcomes of upfront frameless hf-GKRS for large posterior fossa metastases. 

Methods: Thirty-one patients with 37 large (>4 cm³) posterior fossa metastases were included for analysis. There were 20 male patients, and the median age of the patients was 64 years (range, 26-83 years). The most common primary diagnosis was non-small cell lung cancer (n=12). The median target volume was 8.1 cm³ (range, 4.10 cm³-34.80 cm³). hf-GKRS was administered in 3 daily fractions for 11 lesions (median volume=6.7 cm³) and 5 daily fractions for 26 lesions (median volume=8.25 cm³). The median total dose to the margin was 30 Gy (range, 24-30 Gy), with a dose per fraction of 6 Gy (range, 5-9 Gy). Key outcomes assessed included local control, distant progression-free survival, overall survival, and associated toxicities.

Results: The mean follow-up was 12.6 months (range, 2-44 months). LC was achieved in 89.2% of metastases. LC estimates at 6, 12, and 24 months were 100%, 92.9%, and 69.6%, respectively. Distant progression-free survival rates were 73.3% at six months, decreasing to 55.9% at one year. At the end of the follow-up, 83.9% of patients were alive. Radiation necrosis occurred in 2 patients (8.1%), while no cases of leptomeningeal disease were observed.

Conclusions: A high tumor control rate was achieved over sufficient follow-up, which demonstrates the efficacy and safety of upfront hypofractionation in unresected, large posterior fossa metastases in selected patients.

Prupose:

SRS offers a non-invasive alternative to surgery. Linac-based SRS typically uses DCA and VMAT. Currently no unified guidelines exist for selecting the best technique based on lesion geometry, crucial for high-conformity single lesion treatments. This study provides radiation therapists with a tool to evaluate DCA and VMAT for various tumor sizes and shapes, thereby optimizing technique selection to improve outcomes.

Material and method:

75 brain lesions from 19 patients were analyzed. For patients with multiple lesion, separate plans were created. The objective was to assess the percentage differences in the Conformity Index (DeltaCI) and the Spillage Index (DeltaDSI) as benchmarks for technical selection in the treatment of brain lesions, while also examining the influence of lesion size and asymmetry. This method enabled a detailed analysis of the percentage differences in CI and DSI for individual lesions and the identification of discrepancies between both techniques (Tab.1). Effects such as increased CI with smaller volumes or enhanced spillage with larger volumes were thus minimized. The lesions were categorized based on their volumes into 4 categories, category 1 up to 1 ml, category 2 1-2 ml, category 3 3-4 ml, and category 4 >4 ml, therefore identical field geometry was used for both VMAT and DCA-plans to ensure a fair comparison. Geometric parameters such as sphericity and an innovative asymmetry index Qasym (relates the maximum diameter of the lesion to the effective diameter derived from the lesion volume) were utilized for an in-depth examination, to understand and potentially improve how lesion size and shape influence technical selection.

Results:

Scatter plot (Fig. 1) illustrates the relationships between DeltaCI, DeltaDS90%, DeltaDS50%, and DeltaDS25% relative to Qasym across four defined volume categories. These visuals emphasize the correlation between Qasym and dosimetric differences. Our findings indicate that for smaller lesions (categories 1 and 2) with Qasym values ≥ 1.2, VMAT shows better dose conformity and less spillage than DCA. In larger lesions (>2 ml, categories 3 and 4), VMAT consistently outperforms DCA, highlighting its benefits for treating both larger lesions and smaller lesions with higher Qasym values.

Conclusion:

Our study provides guidance for choosing between DCA and VMAT in treating intracranial lesions. For lesions >2 ml or with a Qasym over 1.2, VMAT is preferred due to DCA's limitations. For smaller lesions up to a Qasym of 1.2, DCA is recommended. This approach helps assess the most suitable technique based on lesion size and asymmetry.

Objective: Multi-fraction, mask-based stereotactic radiosurgery (SRS) expands the patient cohort amenable to Gamma Knife (GK) SRS, specifically large lesions and lesions in close proximity to eloquent areas of the brain.  However, mask-based GK Icon (GKI) treatment could also potentially allow more patient inter- and intra-fraction motion which could result in compromising the original treatment objectives, including loss of target coverage, excess organ at risk doses, etc.  In this study, a patient-based motion model using pre-treatment cone beam CT (CBCT) was created.  The dosimetric effects of this uncertainty model are presented and any correlation to lesion characteristics investigated.

Methods: A retrospective cohort study of 227 lesions in 100 patients receiving five fraction GKI SRS between April 2018 to May 2022 was conducted.  Each of the five daily set up CBCT taken prior to treatment delivery was retrospectively defined as the stereotactic reference coordinate system and the treatment dose distribution was calculated on this shifted reference system.  These five shifted three-dimensional doses were subtracted from the original dose distribution and summed to create a mean and standard dose uncertainty for each treatment plan.  Dose volume histograms were extracted to determine the effect of the dosimetric uncertainty on target coverage and dose falloff.

Results: The percent dosimetric uncertainty for the dose covering 99% and 1% of the target was 0.04±0.83% and -0.54±0.69% (average and 95% confidence interval), respectively.  All dosimetric uncertainties were less than 0.75% for both single and multiple lesions treated in a session.  The percent dosimetric uncertainty was showed no dependence on target size, obliquity, volume or position.  Linear fits of the uncertainty versus these variables all yielded fitting parameters consistent with zero.  In this dosimetric uncertainty model, the dose to 95% of a one-millimeter shell inside of target was found to be 104% the target prescription dose and average dose to a one millimeter shell outside the target was 101% the prescribed dose. 

Conclusions: Within the patient-specific dosimetric uncertainty model, five fraction GKI SRS treatments caused minimal deviations from the intended dose distribution when both single and multiple targets are treated.  The dosimetric uncertainty was independent of any lesion characteristic, and also demonstrated that target coverage and dose falloff were not compromised.  This model approximates the maximum inter- and intra-fraction motion during mask-based treatment, and that five fraction can be delivered safely.  However, any uncertainty model needs to include patient outcomes to determine true clinical significance.

Purpose: The irradiated metastatic brain tumor (BM) has been investigated with cell-type based approach. Single-cell level discoveries revealed the cell-states have a critical role in tumor biology and, its cell-state-specific response was not reported in the irradiated BM. Preoperative stereotactic radiosurgery (SRS) is a new paradigm, and it enabled post-irradiation radiation biology study. Here, we aimed to find biological responses and cell-state specific dynamics after preoperative SRS: especially focusing on preoperative gamma knife surgery (or neoadjuvant gamma knife surgery, neoGKS).

Methods and Materials: From 2008 to 2022, a retrospective analysis was done on a total of 120 patietns treated at a single institution. Among them, we examined the transcriptomic data with deconvolution analysis. Irradiated neoGKS samples were validated with immunohistochemistry, western blot, RNA-sequencing (neoGKS n = 9 vs control n = 10).

Results: neoGKS group showed apoptosis and DNA damage responses. Transcriptomic analyses confirmed the post-irradiation change with the overexpression of CDKN1A, MDM2, and B2M in the neoGKS group than the control group (P < 0.01). Deconvolution revealed that neoGKS reduced the tumor-cell-state score (P < 0.01) and elevated immune-cell-state score after neoGKS (P = 0.012).

Conclusions: Transcriptome revealed neoGKS irradiation-associated gene expression and cellular-state-wise dynamics. The significantly reduced tumor-cell-state score after neoGKS may support post-

resection survival benefit after neoGKS. Transcriptome-based cell-state-specific changes would be applied to compare different preoperative SRS modalities and make optimized irradiation plans.

Purpose: Single isocenter dynamic conformal arc stereotactic radiosurgery using Elements software and Exactrac system (BrainLab^TM) allows to treat multiple brain metastases. Factors that may influence the effectiveness of this method should be evaluated. Our study aimed was to assess the effectiveness of a linac-based single-isocenter SRS (SI-MM-SRS) for multiple brain metastases in relation to various clinical factors

Methods: The analysis included a group of 123 patients with MBM lesions (median 4, range 2-12) treated at the Department of Neurooncology and Radiosurgery at Franciszek Lukaszczyk Oncology Center in Bydgoszcz between 02.08.2018 r. and 15.09.2020 r. A total of 560 brain metastases were treated. The minimum follow-up was 12 months and the median follow-up was 23 months. All patients were treated with the BrainLab Elements MultiMets software using single isocenter Dynamic Conformal Arcs. The ExacTrac system was used to monitor the position during SRS. Thirty-six patients with 195 metastatic lesions had follow-up MRI 6 months after treatment, 36% received immunotherapy within 4 months of SRS. Local control was analyzed with RANO criteria.

Results: Sixteen percent of patients was still alive in time of analysis (>3 years of fu). The 6- and 12-months rate was 60% and 33%. In the multivariate
analysis the sum of PTV volumes (p=0.0007) but not a number of lesions was related to survival. Patients whose sum of PTV volume was less than
10 cm3 had longer survival than patients with a volume above 10 cm3 (p=0,007, Fig.1). Surprisingly, patients treated with 5 or more metastases lived
statistically longer than patients with 4 or below brain metastases (p=0,041, Fig.2). Local control was achieved in 93% of the lesions. Lesions with
a margin of at least 0.5 mm had better local control 6 months after treatment (p=0,049; Fig 3.). A better response was also associated with a conformality
index (CI) below 1.42 (p=0,0006; Fig.4) and with the use of immunotherapy within 4 months of SRS (p=0,026). No correlation was found between
DTI (p=0,419) and GI (p=0,599) parameters and local control of metastatic lesions.

Conclusions: SI-MM-SRS is a highly effective method of treating multiple brain metastases. The survival of patients depends on the sum of the volume of metastatic lesions, not their number. Parameters such as margin and CI seem to influence the effectiveness of treatment.

Objectives:

The goal of this study is to evaluate an automatic lesion detection algorithm available as a beta test version software at our institution.

Methods:

The automatic lesion detection algorithm is integrated into BrainLAB Element treatment planning software packages. It is based on neural network and powered by GPU. It will automatically delineates multiple cranial tumors based on contrast-enhanced T1-weighted MRI scans. The results comprise the contrast-enhancing tumor core (active tumor with cystic or necrotic portions). The auto-detection process is automatically started once a new MRI is loaded using the “SmartBrush” Element.

In this study, we used 30 patients’ MRI containing a total of 190 clinical contoured brain metastases as the test sample (range 2 ~ 23). The number of automatic detected targets, matching targets, missing targets were recorded and reported. The number of new found targets is also reported and then further categorized into two groups: sizable target (>=0.05 cc), and tiny target (<0.05 cc). This is because for the tiny targets, clinical decision may have varied depending on specific patient situations. Therefore, there is lack of basis for the comparison.

Results:

A total of 204 objects were contoured by the auto-detection. 173 objects are matching with the clinical contoured targets, and 31 targets were new found (10 sizable targets and 21 tiny targets). The overall successful detection rate is 91%. 17 (9%) clinical contoured targets were not identified by the auto-detection. They are mostly concentrated on two specific MRI images where the image contrast is visibly lower than normal. Visible examples of such images and missing targets will be presented.

Conclusions:

An automatic lesion detection algorithm is evaluated with 30 brain metastases patients’ MRI scans. It apparently showed the detection accuracy is highly depending on the MRI quality and contrast enhancing. 2/3 of the new found targets are in the range of <0.05cc in volume. We will look further into those targets on patients’ follow-up MRIs to determine if it was not clinical identified due to their tiny size. It would be helpful if a confidence level measure for each auto-detected targets being available to help with the clinical teams' decision making when reviewing the auto detection results.

Objective:

Linac-based stereotactic radiosurgery (SRS) and HyperArc technique has gained increasing popularity for managing brain metastasis due to the greater accessibility, standardized process and high delivery efficiency from single-isocenter multiple-targets (SIMT) treatments. However, due to the small field condition and increasing modulation complexity, the plan dosimetry highly relies on the modeling of multi-leaf collimator (MLC). In  Eclipse v18, an enhanced leaf model (ELM) is introduced by constructing the actual rounded leaf end design and attenuation, to replace the current dosimetric leaf gap (DLG). We intended to investigate its impact to SIMT HyperArc plans and compare it against the conventional DLG-based V16 model.

Methods:

22 multi-mets patients received Linac-based SRS treatments were retrospectively selected. The number of metastases ranges from 2 to 19 (average 6, median 6), volumes of PTVs ranges from 0.03 to 17.12cc (average 1cc, median 0.33cc), and distance of PTV centers to isocenter ranges from 1.2 to 9.5cm (average 5cm, median 5cm). The original clinical plans were created in Eclipse v16 using HyperArc technique for an Edge Linac with HD-MLC, using 10FFF energy, Analytical Anisotropic Algorithm(AAA), and DLG optimized for SRS treatment (AAA-SRS-16). We recalculated clinical plans using the original AAA-16 models, whose DLG was for conventional treatments (AAA-16). The recalculation with ELM model was performed in a test Eclipse v18 environment. The original beam data was from Eclipse v16 and the ELM was configured from ion chamber and solid water measurement (AAA-18). We also repeat the comparison with 6FFF plans made retrospectively.

Results:

For 10FFF, AAA-18 SIMT plans are similar to AAA-SRS-16 plans, with an average 1%/1mm gamma rate of 100%.  The average dose difference between them is -3.9%, and between AAA-18 and AAA-16 is -7.1%. For 6FFF, we observed an average dose difference of 7.4% between AAA-18 and AAA-16. The main difference occurs at peak dose and valley dose region, while the dose falloff region is similar. The configuration of DLG in AAA-SRS-16 model involved several rounds of adjustment from AAA-16, recalculation and revalidation, while the ELM(AAA-18) configuration was one-time effort, which greatly improves the efficiency of commissioning and reduces the uncertainties and user variability.

Conclusion:

The new enhanced leaf model introduced in Eclipse v18 substantially improves the efficiency of dose algorithm modeling. It showed similar dosimetry compared with finely tuned DLG for SRS in the current MLC model for HyperArc SIMT plans, while changes the doses about 7% from the DLGs tuned for conventional treatments.

Objectives:

To ascertain the degree of variability in stereotactic radiosurgery (SRS) dosing for breast and renal cancer metastases in the central nervous system (CNS), in centers across Latin America and Spain, through a survey conducted among radio-oncologists and neurosurgeons

Materials and Methods:

Responses from a survey conducted through Google Drive among 106 SRS specialists in Latin America and Spain were reviewed. Specific questions about SRS dosing for CNS metastases of breast and renal cancer were selected, focusing on single-fraction SRS practice. Descriptive statistics were calculated to demonstrate variability among respondents in the reported doses for the same clinical case with identical histology, and the Student's t-test for independent samples was used to detect statistically significant differences between the prescriptions for both histology.

Results:

Among the surveyed participants, 93.4% were from LATAM and 6.6% from Spain, with an 85% participation from LATAM countries. Respondents included 87% radiation oncologists and 13% neurosurgeons. The technologies used were: LINAC (70%), Gamma Knife (15%), CyberKnife (7%), Halcyon (5%), and ZAP (3%).

The average dose for breast metastases was 20.48 Gy [15-25 Gy], with a standard deviation of 1.04 Gy. Most prescriptions were concentrated in the 20-21 Gy range, accounting for 54.43% of the total.

The average dose for renal metastases was 21.91 Gy [15-25 Gy], with a slightly higher variability (standard deviation of 1.06 Gy). A more balanced distribution was observed in the higher ranges, with 31.11% of doses in the 24-25 Gy range and 21.11% in 22-23 Gy.

The difference in average dose between the histological types was statistically significant (t = -4.35, p = 0.000024), indicating a trend to prescribe higher doses for renal cancer.

Conclusions:

The variability found in dose prescription for the same histological type suggests the need for a consensus in SRS practice for CNS metastases, at least in LATAM. The results highlight the importance of establishing international guidelines for standardization in SRS dosing for single-fraction CNS metastases. Uniformity in prescription would allow more homogeneous comparisons between studies and technologies, yielding more robust results.

Background and Objectives

Triple negative breast cancer (TNBC) remains an aggressive disease with a poor prognosis. Although ongoing studies are assessing the efficacy of new systemic therapies for patients with TNBC, the overwhelming majority have excluded patients with brain metastases (BM). Therefore, we aim to characterize systemic therapies and outcomes in a cohort of patients with TNBC and BM managed with stereotactic radiosurgery (SRS) and delineate predictors of increased survival.

Methods

We used our prospective patient registry to evaluate data from 2012-2023. We included patients who received SRS for BM. A competing risk analysis with the Fine and Gray method was conducted to assess local and distant control where death was the competing risk.

Result

Forty-three patients with 262 tumors were included. The median OS was 16 months (95% CI 13-19 months). Predictors of increased OS after initial SRS include modified Breast GPA score >1 (HR= 0.183, 95% CI 0.088-0.496, p<0.001) and use of immunotherapy such as pembrolizumab (HR= 0.360, 95% CI 0.175-0.830, p=0.011). The median time on immunotherapy was 8 months (IQR 4.4, 11.2). The cumulative rate for development of new CNS metastases after initial SRS at 6 months, 1 year, and 2 years was 23%, 40%, and 70%, respectively. The quotient of total tumor volume to the sum of tumors at initial SRS (adjusted tumor burden) was developed to predict new CNS metastasis. An adjusted tumor burden of ≥3 was a significant and reliable negative predictor of development of new CNS metastasis (SHR 0.813, 95% CI 0.696- 0.949,  p=0.009).

Conclusions

TNBC patients with BM can achieve longer survival than might have been previously anticipated with median survival now surpassing one year. The use of immunotherapy is associated with increased median overall survival of 23 months and the adjusted tumor burden may be considered as a useful predictive tool for determining distant CNS tumor progression.

Background/Objectives:

The diagnosis of brain metastases (BM) places a major burden on both patients and providers. Stereotactic radiosurgery (SRS) is a primary or adjuvant option for this disease, but underlying socioeconomic factors may delay access to SRS and subsequent follow-up care. Patients with higher household income, higher educational attainment, and who are enrolled in an insurance plan are more likely to receive SRS. This study explores the impact of social determinants of health on time to SRS delivery after diagnosis of brain metastases.  

Methods:

This is a retrospective study of patients with brain metastases who underwent Gamma Knife® (GK) SRS as the primary modality at a single institution from 2008-2023. Patients with prior surgical resection or whole brain radiotherapy (WBRT) were excluded. Time from first BM diagnosis to initial SRS was analyzed across patient demographics, median household income based on U.S. Census Bureau 5-year estimates, and insurance carrier (private, Medicare, Medicaid). 

Results:

1216 patients with 4576 brain metastases were included in the analysis. The median time from diagnosis of BM to SRS was 15 days (IQR:20). White patients had a significantly lower time to SRS (13 days, IQR:17) compared to Black (21 days, IQR:24, p < 0.001) and Asian patients (20 patients, IQR:23, p < 0.001). Medicare patients had a significantly lower time to SRS (12 days, IQR:15) compared to private insurance (15.5 days, IQR:21, p = 0.006) and Medicaid (20 days, IQR:24, p < 0.001). Patients with median household income > $75,000 had a shorter time to SRS (14 days, IQR:18) compared to those below this income bracket (17.5 days, IQR:22, p = 0.004).

Conclusion:

There are differences in time from diagnosis of brain metastases to first-line SRS across patients of different ethnicities and socioeconomic strata. Efforts to reduce healthcare disparities are critical in ensuring timely SRS delivery to brain metastasis patients.

Introduction/Objectives

In the era of targeted therapies, antibody drug conjugates (ADCs) are being used more frequently in patients with breast cancer brain metastases (BCBM) treated with SRS. A recent report raised the possibility of increased risk of symptomatic necrosis when ADCs are used concurrent with SRS. Therefore, we investigated if similar risk is observed in our institutional experience.

Methods

We queried our prospective patient registry from 2012-2023 to identify BCBM patients with a minimum of three-months of follow-up who received at least one dose of trastuzumab-emtansine, trastuzumab-deruxtecan, ladiratuzumab-vedotin, or sacituzumab-govitecan and underwent concurrent SRS. Adverse radiation effects (AREs) were determined via radiographic follow-up with peritumoral patchy enhancement with a mismatch on the long relaxation time images was coded as an inflammatory change consistent with ARE. Concurrent use of ADC was noted if SRS was done 7 days before or 30 days after ADC delivery. A competing risk analysis with the Fine and Gray method was conducted.

Results

In total, 46 BCBM patients that received ADC with 290 tumors were included. The median age was 56.5 (IQR, 48-63) and the median follow-up time was 23 months (IQR, 15-42). At the time of analysis, 19 patients (41%) were alive while 22 patients (48%) were deceased due to non-neurologic causes and 5 patients (11%) were deceased due to neurologic causes. Sixteen patients (35%) received whole-brain radiotherapy (WBRT) prior to SRS and ADC treatments. Twenty-seven patients (59%) received ADC concurrently with SRS while 19 patients (41%) received ADC sequentially. The median marginal dose was 16 Gy (IQR, 15-18) and the median total tumor volume was 1.8 cm³ (IQR, 0.48-6). Amongst the entire cohort, the median number of SRS treatments was 2 (IQR, 2-3). Overall, 6 tumors (2%) exhibited ARE and the 12 and 24-month cumulative incidence of ARE for the entire cohort were 1% and 2%, respectively. Five of the tumors were symptomatic requiring a short course of corticosteroids and no further sequalae. The cumulative Concurrent ADC was not associated with increased risk of ARE (SHR, 0.024 [95% CI, 0-248]; P=0.428). Local tumor control was 96% throughout the follow-up period. 

Conclusions

Analysis of our institutional experience did not identify an increased risk of symptomatic ARE with concurrent SRS and ADC. Notably, our median marginal dose was lower than previous reports. A larger multi-institutional study may shed additional light on the incidence of AREs with the use of concurrent ADCs.

As the only neuroradioloigst at the Gamma Knife Center of the Pacific since 1998 in Honolulu, Hawaii, and with a 'captured' Island population, I have the resposibility of reviewing all of the pre and post GK MRI scans and have discovered an interesting post treatment phenomenon of complete enhancment resolution of certain metastsis, consistent with successful GK tumor obliteration, but then with delayed recurrent enhancment at the treatement site between 6 and 35 months later, mimicking recurrent tumor but actually being benign post GK dealyed blood brain barrier breakdown and not active tumor. This pattern needs to be understood by all treating GK physiicians and interpreting radiologists to avoid potentially dangerous retreatment or changing ongoing and otherwise successful chemo or immunotherapy.  This is now more important due to the success of GK and the mobility of patients, often seen at a other medical facilities distant to the original treatment location.

Objectibe: 

Pilocytic astrocytomas (PA) are the most common gliomas (WHO I) in children. in current protocols, The irradiation usuallyt is not used due to the potential risks of long-term complications and  postponed until recurrence. Burdenko Neurosurgical Institute has the greatest experience in modern stereotactic radiotherapy for this pathology.

Materials and  Methods:

152 patients with Optic Pathway PA were irradiated at Burdenko Neurosurgical Institute between April 2005 and January 2018. The study group consisted of 38 adults and 114 children.  The median age was 13,7 years. 80 (52,6%) patients had a prior histological tumor verification (tumor resection or biopsy). In 72 (47,4%) patients the diagnosis was based on clinical evidence and radiological data. Neurofibromatosis type 1 was detected in 28 patients (18.4%). In 111 (73%) patients, radiotherapy was the primary treatment (58 patients) or was performed immediately after non-radical surgery (53 patients) . In 41 (27%) patients, treatment was due to continued tumour growth after non-radical surgery (20 patients) or after polychemotherapy (vincristine+carboplatin) (21 patients). Endocrine disorders were detected in 23 (55%) patients (out of 42) examined. Most of the patients (129 pts – 84,8%) underwent SRT in standard fractionation (1,8 Gy/fr, mean dose 54 Gy), 23 patients (15,2%) underwent hypofractionated SRT (5-6 Gy per fraction, mean dose 25-30 Gy).

Results:

149 patients (98%) were available for the follow-up. The median follow-up period was 75 months (11-197) after the patients were diagnosed PA. At the end of the follow-up (9.2023) 144  patients (97,5%) were alive. The median follow-up period after irradiation was 57 months (range, 6-196 months). Fifteen (10.4%) patients developed pseudoprogression followed by spontaneous regression or partial removal/emptying of cysts. Recurrences (3 local and 3 distant) occurred in 6 (4.2%) patients. Finally tumor control or regression was achieved in 138 patients (95.8%). Five-year recurrence-free survival was 97.7%. Endocrine function decline occurred in 33% of the examined patients. No malignant transformation, radiation necrosis, secondary tumors, hearing impairment or moya-moya disease were observed in patients. 

Conclusion: 

Stereotactic irradiation (SRT, SRT hypo) is an effective method of treatment for optic pathway PA in patients with residual tumors and patients with progressive disease. The method provides the highest rates of tumor growth control compared to other treatment methods. With the available follow-up period, we did not identify any complications that may justify postponing the radiation treatment. Further study of the results of stereotactic irradiation and revision of the indications for this treatment are required.

Purpose: Radiation necrosis (RN) is a major late toxicity after radiation therapy for brain metastases, with oral corticosteroids being the primary but suboptimal long-term management due to side effects and drug interactions. Boswellia serrata (BS), known for its anti-inflammatory properties, has shown promise in reducing cerebral edema post-brain radiation therapy. This study evaluates the effectiveness of BS in patients with brain metastases treated with stereotactic radiosurgery (SRS) who subsequently developed RN.

Methods: We analyzed patients who developed RN post-SRS for brain metastases between 2020-2022 at our institution and were treated with BS (4.2-4.5g daily). Follow-up MRI was conducted every 2-3 months, with responses assessed using Response Assessment in Neuro-Oncology (RANO) criteria. The primary endpoint was a ≥25% decrease in edema volume on T2-FLAIR MRI from baseline. Patients were censored for tumor progression, repeat RT, or death.

Results: Among 50 patients treated with BS for Grade 1-3 RN, median age was 62.8 years, and median RT dose was 24 Gy in 3 fractions. Median time to RN onset post-SRS was 10 months, with a follow-up period of 6 months. Out of 40 patients with follow-up MRIs, 15% achieved complete response (CR), 40% partial response (PR), 35% had stable disease, and 10% progressive disease. Median time to response was 9 months for CR and 6 months for PR. Symptomatic improvement was seen in 35.7% of patients using BS alone, while 64% required steroids. Salvage treatments included steroids, surgery, Bevacizumab, or hyperbaric oxygen therapy. Side effects were minimal, with 6% experiencing mild gastrointestinal issues. Two patients discontinued BS due to enrollment in an immunotherapy trial.

Conclusion: BS demonstrated over 50% response rates in treating Grade 1-3 RN post-SRS, with a significant portion avoiding long-term steroid use. BS emerges as a safe, accessible, and promising alternative for RN management, warranting further prospective studies.

Gamma knife radiosurgery （GKRS） is recommended as the first-line treatment for brain metastases of lung adenocarcinoma （LUAD） in many guidelines, but its specific mechanism is unclear. We aimed to study the changes in the proteome of brain metastases of LUAD in response to the hyperacute phase of GKRS and further explore the mechanism of differentially expressed proteins (DEPs). Cancer tissues were collected from a clinical trial for neoadjuvant stereotactic radiosurgery before surgical resection of large brain metastases (ChiCTR2000038995). Five brain metastasis tissues of LUAD were collected within 24 hours after GKRS. Five brain metastasis tissues without radiotherapy were collected as control samples. Proteomics analysis showed that 163 proteins were upregulated and 25 proteins were downregulated. GO and KEGG enrichment analyses showed that the DEPs were closely related to ribosomes. Fifty-three of 70 ribosomal proteins were significantly overexpressed, while none of them were underexpressed. The risk score constructed from 7 upregulated ribosomal proteins (RPL4, RPS19, RPS16, RPLP0, RPS2, RPS26 and RPS25) was an independent risk factor for the survival time of LUAD patients. Overexpression of ribosomal proteins may represent a desperate response to lethal radiotherapy. We propose that targeted inhibition of these ribosomal proteins may enhance the efficacy of GKRS.

Aim

Recent advanced technologies allowed different treatment modalities and different dose calculation alghorithms with different output dose characteristics. The aim of this study was to compare and to evaluate dosimetric aspects of stereotactic radiotherapy through the use of two techniques: LINAC-based versus helical tomotherapy (HT).

Material and methods

Eight patients with solitary brain metastasis received stereotactic radiotherapy and were included in the analysis. Patients were subjected to 1-mm slice thickness computed tomography simulation with Gross Tumor Volume (GTV) defined by contouring the visible lesions on MRI images and Planning Treatment Volume (PTV) obtained by 2 mm isotropic extension of the GTV. Each contouring dataset of the patients was planned with both tomotherapy (Raystation v11B TPS) and LINAC-based treatment planning system (BrainLab Elements v3.0 TPS). The LINAC-based modality was realized with 6 MV FFF beams and no-coplanar arcs. The delivered dose was 27 Gy in three fractions for each treatment plan. The Paddick Conformity Index (PCI), the inverse Paddick Conformity Index (iPCI), the Gradient Index (GI), the PTV-coverage, the beam-on time and the volume receiving 18Gy (V18) were calculated and compared for both treatment modalities. Results were analyzed with Wilcoxon signed-rank test.

Results

The median volume of lesions was 3 cc. PTV coverage, PCI and iPCI were similar for both treatment modalities: mean values were respectively 95.7%, 0.84, 1.2 for LINAC-based and 95.7%, 0.81, 1.2 for tomotherapy.

GI and beam-on time were statistically significantly lower with LINAC, with a mean value of GI of 4.0 versus 6.6 for HT and with a beam-on time of 199 seconds for LINAC versus 517 seconds for tomotherapy.

Also V18 improved with LINAC, with a median value of 6.7 cc compared to 8 cc with HT.

Conclusion

In our analysis, the LINAC-based system offered the best dose gradient with similar values of PTV coverage, PCI and iPCI compared to tomotherapy. Also the beam-on time obtained with LINAC system was lower respect to the tomotherapy one. The LINAC-based approach also provided significantly better V18 values compared with HT improving toxicity profile with the same efficacy. The outcome of our preliminary analysis has encouraged us to preferably treat patients with LINAC-based modality in order to obtain better dose distribution improving toxicity profile and shortening treatment time.

Objective: To describe the experience and results from coordinated and closely scheduled Gamma Knife radiosurgery (GKRS) and cochlear implantation (CI) in a vestibular schwannoma (VS) cohort.  Unfortunately, studies of this patient population have demonstrated the negative influence of non-functional hearing as well as minimal hearing changes on quality-of-life measures. Further, subjective testing shows that VS patients with hearing loss experience notable functional deficits in comparison to binaural hearing controls.  Thus, current hearing function and potential for rehabilitation are critical considerations when discussing tumor management via GKRS given the detrimental effects of non-functional hearing.

 Methods:  Data were retrospectively collected from patients undergoing cochlear implantation immediately (within 24 hours) after GKRS from December 2003 to August 2022 at a single, large tertiary center. 

Main Outcome Measures: Tumor control defined by tumor growth on post-treatment surveillance and audiometric outcomes including Consonant-Nucleus-Consonant (CNC) words and AzBio sentences in quiet.

Results: In total, 6 patients were identified that met inclusion criteria, with an age range of 38- to 69-years-old and tumor sizes ranging from 2.0 to 16.3 mm.  Margin dose was 13 Gy and maximum dose was 26 Gy.   One patient was local and the remaining five lived 105 to 1447 miles from our center.  Four patients had NF2-associated schwannomatosis.  All patients successfully underwent GKRS and CI on the same or next day. Postoperatively, all patients obtained open-set speech recognition. CNC word scores ranged from 40 to 88% correct, and AzBio scores ranged from 44 to 94% correct.  During post-treatment MRI surveillance, which ranged from 12 to 68 months, all tumors were noted to be adequately visualized, and no tumor progression was noted.  This mirrors our previous experience with CI following GKRS in 17 patients (18 ears implanted).

Conclusions: Coordinated GKRS and CI can be safely performed in patients with VS on the same day or immediately subsequent day , serving to decrease burden on patients and increase access to this vital rehabilitative strategy.  

In severe and refractory cases of Obsessive-Compulsive Disorder (OCD), neurosurgical procedures may be proposed as a therapeutic option. Ventral Anterior Capsulotomy using Gamma Rays (GVC) is one of the psychosurgery options for OCD. The aim of this study was to evaluate tractographic differences between refractory OCD patients who underwent GVC and healthy controls. This involved a non-probabilistic convenience sample of refractory OCD patients who underwent GVC, as well as randomly selected healthy controls matched for gender, age, and imaging apparatus/protocol. Pre-processing steps were conducted using MRtrix3 software, and tractography was processed using constrained spherical deconvolution (CSD) to enable segmentation of the Anterior Limb of the Internal Capsule (ALIC) based on connectivity with the frontal cortex. After pre-processing, tractographies were automatically segmented using a convolutional neural network called Tractseg into 72 fiber bundles. Tractometric profiles for the control and OCD groups were constructed for each bundle, considering three metrics (fractional anisotropy, mean diffusivity, and peak length). In this study, 27 participants were included in a 1:2 ratio, comprising 9 OCD patients (5 with 3T MRI) and 18 controls (10 with 3T MRI). The mean age of the OCD patients was 36.9 years (95% CI 32.4 – 46.5), with 7 (77.8%) being male. Brodmann areas 11 and 47 have predominantly ventral distribution, whereas BA06 and BA08 are located dorsally in the ALIC. BA09, BA10, and BA46 have an intermediate craniocaudal distribution, with BA46 predominantly lateral. There were no differences between the OCD and control groups regarding the topographical distribution of fibers in the ALIC. All participants had their diffusion images processed using the CSD algorithm, with subsequent automated segmentation of the 72 fiber bundles as predicted by Tractseg. Differences between the OCD and control groups were identified in at least one segment of the tractometric profile for all considered metrics, in tracts diffusely distributed in both hemispheres and not restricted to the cortico-striato-thalamo-cortical pathway. Despite the apparent topographical distribution similarity, refractory OCD patients who underwent GVC, when compared to the control group, exhibited discrepancies in their tractometric profiles.

Background: In single-session stereotactic radiosurgery (SRS) for treating cerebral metastases, the tumor size primarily constrains the dosage, and guidelines recommend a fractionated treatment approach for metastases that exceed 3 cm in diameter or have a volume surpassing 10 cm³. Standard fractionated regimens include 27 Gy delivered in three fractions over three days or 30 Gy administered in five fractions from Monday to Friday within the same week. However, these daily regimens may not always align with the scheduling constraints of the radiosurgical unit, other planned treatments, or reimbursement policies. Our objective was to investigate the outcomes of hypo-fractionated SRS with a two to three-week interval between fractions instead of the daily regimens.

Methods: We analyzed patients with at least 1-year follow-up who received either double- or triple-session SRS for the same cerebral metastasis within four weeks. In addition to background data, we reviewed the time between fractions, tumor volumes, 1-year tumor control rate, latest tumor control rate, perilesional edema, radionecrosis, and any relevant clinical worsening.

Results: We are currently collecting data. The typical fractionated treatment regimen was 12-14 Gy x 2 administered with a three-week split between the fractions. The goal of the first fraction was to decrease the tumor size, allowing for a second fraction to achieve improved long-term tumor control.

Conclusion: We have good experience with a two- or three-staged SRS over several weeks instead of the recommended daily treatment regimen. Two- or three-staged SRS is an effective treatment technique for large brain metastasis that significantly reduces tumor volume at the later SRS stages, and the long-term tumor control is likely comparable to high-dose, single-session SRS. The final results from our analyses will be presented. 

Background: Brain metastases has been the ideal target for stereotactic radiosurgery (SRS) for several decades, whereas new techniques have been implemented such as “hyper arc” (HA) by using the EDGE VARIAN SRS system. With this presentation we are reporting our experience with the first implementation of HA technique in Greece, in Radiation Oncology center of Mediterraneo General Hospital.

Patients and Methods: We retrospectively analyzed 42 patients with metastatic brain lesions. The SRS with HA was delivered via the EDGE VARIAN system and the HA technique. The primary was as following: breast (25 patients), lung (14 patients) and rectum (3 patients).

Results: The range of lesions was between one and ten The delivery of treatment was realized in single fraction for lesions with diameter less than 3cm. Three to five fractions were used with multiple lesions, whereas the criterion of v12 for normal brain tissue was not met. There were no acute or late toxicities from the skin or cognitive affairs related to the CNS.   Local control (LC) was achieved in 100% of patient at the time of the first follow-up and the projected 6-month local progression-free survival (LPFS) was 95%.

Conclusion: High LC and LPFS can be achieved with SRS for brain metastatic lesions with HA technique. The study continues to recruit patients to obtain mature results in the following years.

Background: Symptomatic patients' outcomes following motor cortex brain metastases (BMs) treated with Stereotactic Radiosurgery (SRS) are not well-described in the literature. Most cohorts include mixed cohorts of symptomatic as well as asymptomatic patients, various sized lesions, treated with either SRS or surgery. The available evidence is, therefore, inconclusive.  

Methods: Here, we studied the data of 70 patients, treated with SRS, combined either with or without surgery in Sheba Medical Center between the years 2010 to 2022. Patients were diagnosed with BMs located within the motor cortex or adjacent to it and presented accordingly with hemiparesis or hemiplegia. Patients' demographics, and clinical and oncological outcomes, were retrieved using a novel institutional AI algorithm software. SRS and surgical treatment paradigms as well as their associated outcomes were collected. BMs were, in turn, classified according to their location with respect to the motor cortex, and their volumetric data was measured and documented.

Results: Patients' demographics showed that their median age-at-diagnosis was 65 years (range, 38-89), male-to-female ratio was 2.3:1, and median follow-up duration was 7 months (range, 0-154). BMs' originated as follows: lung, n=33, melanoma, n=17, breast n=5, gastrointestinal, n=10, others, n=5, and the median duration of time from primary cancer diagnosis to BMs diagnosis was 11 months (range, 0-199). Patients were treated with SRS alone (n=37), or SRS combined with tumor resection or Ommaya reservoir insertion (n=33). The mean radiation dosage was 19 Gy (range, 12-32), delivered in 1-5 fractions according to the acceptable treatment protocols. The entire cohort's median overall survival (OS) was 9.8 months (95% CI 7.3-13.5). The median OS of patients treated with SRS and surgery was 14.9 months (n=32, 95% CI 9.9-23.3) while patients treated with SRS alone demonstrated OS of 6.8 months (n=38, 95% CI 3.4-9.1), p=0.0012. 41% of the patients demonstrated motor deficit improvement, demonstrated to be related to favorable OS (p=0.05). Due to low numbers, preliminary analysis showed that it was not possible to perform further analysis regarding clinical improvement and specific treatment types. The average tumor volume treated was 5.7 cc (range, 0.5-42), and was not correlated with patient's outcomes. 

Conclusion: This study aims to comprehensively explore the clinical and oncological outcomes of a homogenous cohort of symptomatic motor cortex BMs patients treated with SRS alone or SRS combined with surgery. Favorable OS was demonstrated in patients treated with SRS and surgery and in patients who exhibited post-treatment clinical improvement.  

Background and Objectives: Involvement of the cranial nerves is a rare feature of cancer and is a marker of poor survival. Cranial nerve metastases (CNM) can be primary via leptomeningeal metastasis or secondary by spread due to head and neck cancers or due to distant tumor metastasis to the skull base. In addition, cranial nerve metastases may cause cranial nerve-related symptoms that can impact patient quality of life.

Methods: We performed a single-center retrospective cohort study of all patients with CNM treated with SRS at our institution between April 2003 and February 2021. Demographic and clinical information were retrieved from the electronic medical record. Median follow-up was 12.9 months.

Results: Our study cohort consisted of 9 patients with primary CNM and 8 with secondary CNM – for a total of 17 patients, with a total of 33 lesions (23 primary, 10 secondary). Eleven patients (64.7%) had symptoms caused by cranial nerve metastases. Symptoms were resolved in 5 of 11 patients (45.5%) after SRS. Patients with secondary CNM were more likely to have cranial nerve symptoms and more likely to have resolution of symptoms following SRS. The median time between SRS and symptom improvement was 3 months. Local tumor control was achieved in 30 of 33 lesions (90.9%). Local tumor control at 6 months and at 1 year were 100%. Overall survival at 6 months was 76.5% for the entire cohort, with 55.6% and 100% for the primary and secondary subgroups, respectively. Only one patient had an adverse event (5.9%).

Conclusion: Our study suggests that SRS may be a safe and effective treatment for cranial nerve metastasis providing 90.9% local tumor control at final follow-up, and symptomatic stability or improvement in 90.9% of symptomatic cases. Patients with secondary CNM may stand to benefit more from a symptom management standpoint.

Background: Pineocytomas are a rare type of tumor that arise from the parenchyma of the pineal gland. Gross total resection can potentially cure these benign lesions but is associated with significant risks of morbidity. Stereotactic radiosurgery (SRS) is thought to provide adequate tumor control, but the current literature is mostly limited to small single institution case series. This study was designed to provide multi-institutional data to strengthen the evidence related to the use of SRS for pineocytomas.

Methods: Centers participating in the International Radiosurgery Research Foundation were asked to review their database and provide data for patients who had SRS for a histology confirmed grade 1 pineocytoma, for whom clinical and imaging follow-up of at least 6 months was available. 

Results: We identified 38 patients (23 male and 15 female) who underwent SRS as part of the management of a pineocytoma. Median age at SRS was 39 years (range 8-76). SRS was performed as primary treatment in 68%, adjuvant after partial resection 19%, and at recurrence in13% of patients. The median margin dose used was 15 Gy (range 11-25 Gy). The median treatment volume was 3.35 cc (range 0.1-17.9 cc).Local tumor control was achieved in 92% of pineocytomas after SRS treatment, with mean actuarial progression-free survival of 21.6 years. At last follow-up, 82% were still controlled, 8% had local recurrence and 10% had cerebrospinal fluid dissemination. The only significant factor associated with tumor control was the indication for treatment. Mean actuarial local control was 26.1 years for primary treatments, 4.8 years for residual tumors after partial resection, and 4.6 years for recurrent tumors (p=0.016). Five patients (13%) died during follow-up, all due to tumor progression. The actuarial mean survival duration was 24.3 years, with a 5-year survival rate of 91%, and an estimated rate of 76% at 29 years. Transient symptomatic adverse radiation effects (ARE) were observed in 4 patients (11%). No parameter was identified as a risk factor for death or ARE.

Conclusion: Stereotactic radiosurgery is a safe and effective treatment for pineocytomas. It can be offered to patients as a primary management option after histological confirmation of the diagnosis as an alternative to surgical resection.

Introduction. Oligodendrogliomas are primary brain tumors classified as IDH-mutant and 1p19q co-deleted in the 2021 WHO Classification. Surgery, fractionated radiotherapy and chemotherapy are well established treatments for these tumors, but there are few studies evaluating the efficacy of stereotactic radiosurgery (SRS). As these tumors are less infiltrative than astrocytomas and typically recur locally, they could be appropriate for local therapy such as SRS.

Methods. This was a retrospective multicenter study performed through the International Radiosurgery Research Foundation (IRRF). Adult patients were included if they underwent single-fraction SRS for a grade 2 or 3 histologically confirmed oligodendroglioma. Mixed tumors (formerly oligoastrocytomas) were excluded. The primary endpoints were progression-free survival (PFS) and overall survival (OS). Secondary endpoints included clinical evolution and occurrence of adverse radiation events or other complications. 

Results. Eight institutions submitted data for a total of 55 patients with a median clinical follow-up of 24 months. The median age at treatment was 46 years (range, 18-75) and median pre-treatment KPS was 90% (range, 60-100%). Prior surgical management included gross-total resection in 54.5%, partial resection in 25.5% and biopsy in 20%. Prior radiotherapy had been used in 58% of patients and chemotherapy in 71%. The median treatment volume was 4 cc (range, 0.1-27) and median marginal dose delivered was 15 Gy (range, 9-24). After SRS, the median PFS was 17 months, with actuarial rates of 60.1% at 1 year, 31% at 2 years and 24.4% at 5 years after SRS. The median OS post-SRS was 58 months, with actuarial rates of 91.5% at 1 year, 83.4% at 2 years and 49.3% at 5 years. The KPS remained stable post-SRS in 51% and worsened in 46.7% of patients, most often due to tumor progression (73.1%). Adverse radiation-induced imaging changes occurred in 29.6% of patients but were symptomatic in only 7.5%. Factors significantly associated with worse PFS were WHO grade 3, prior radiotherapy and chemotherapy and higher treatment marginal dose. Factors significantly associated with worse survival were WHO grade 3 and prior radiotherapy and chemotherapy. 

Conclusion. SRS appears to be a valuable management option for oligodendrogliomas.

Introduction: Radiomics promises to revolutionize clinical decision making in the management of melanoma brain metastases (MBM) treated with stereotactic radiosurgery (SRS). Radiomic features extracted from baseline (day-of-treatment) magnetic resonance imaging (MRI) can be integrated into clinicoradiological parameters to predict long term outcomes in order to tailor multimodal treatment strategies, permit individualized surveillance imaging frequency and facilitate early changes to therapy following SRS. We analyze the predictive accuracy of a baseline-only MRI radiomics machine-learning model in MBM patients to predict local failure following SRS.

Methods

Patients receiving single-fraction Cobalt-60 based SRS for MBM at a single Australian institution were analyzed. Progression of disease (PD) outcomes were defined either histologically or according to RANO-BM criteria. 3214 radiomic features were extracted from the pre-SRS day-of-treatment T1-weighted contrast-enhanced MPRAGE MRI sequences using in-house software developed in MATLAB. High and low pass wavelet filtering was applied and highly dependent radiomic features were selected using lease-absolute-shrinkage-and-selection-operator (LASSO) regression. Binary classifiers were trained and validated using a leave-one-out-cross-validation (LOOCV) technique to generate predictive models. Synthetic minority oversampling (SMOTE) was used to counter the effects of class imbalance. A multivariate model was additionally developed integrating radiomic features with baseline lesion volume, immunotherapy use and SRS dose. The final model was applied to a de-novo dataset to assess predictive accuracy.

Results: 101 MBM patients were treated with SRS. The median duration of follow-up was 29.2 months (IQR 19.7-39.8). Median dosage was 20Gy (IQR 18-20) in a single fraction. The median volume and diameter of the lesion at baseline were 0.24cc (IQR 0.06-1.02) and 7.7mm (IQR 4.8-12.2), respectively. 77.0% of patients received immunotherapy concurrently (4 weeks pre to 4 weeks post-SRS). Overall local control in the cohort was 87.1%. Adjusting for concurrent immunotherapy status, dose and lesion volume, radiomics analysis demonstrated that utilizing baseline imaging alone, long term PD following SRS was accurately predicted with an 88.9% accuracy in the training dataset. Radiological texture heterogeneity radiomics markers were most strongly associated with local failure. When applied to a de-novo (untrained) dataset in the clinical setting, the model demonstrated a 73% predictive accuracy (95% CI 60.6-85.6%).

Conclusion:

A pre-treatment baseline-MRI radiomics model has a high degree of accuracy in predicting long-term local failure in melanoma brain metastases treated with SRS. Additional integration of radiomics models utilizing multiparametric imaging combined with patient and treatment characteristics will optimize the use of radiomic tools into the clinic.

Introduction:

SRS for the treatment of limited-brain-metastases (BM) is widely accepted, but there are still limitations in the management of numerous-BM. Frameless-single-isocenter-multitarget SRS is a novel technique that allows a rapid treatment delivery to multiple-BM. We report preliminary clinical and dosimetric outcomes of our experience with this technique.

Methods and Materials

We have reviewed clinical and dosimetric-outcomes of patients with intact BM treated with SRS using single-iso-single-target (if 1met) and single-iso-multi-target-technique(if≥2mets). Immobilization was based on an SRS-stereotactic-mask. Brainlab® SRS Elements software was used for registration, image fusion, target contouring and treatment planning. Exactrac System and a 6degree of freedom couch were used for monitoring, correcting position and assessing and applying residual-errors also when couch rotations.  Patient positioning was monitored in real-time using surface-tracking.

Results

From 19/05/2022 to 11/12/2023, we treated 60 patients with a total of 255 BM. Patients and treatment characteristics are described in Figure-1. The 67% of patients had at least 2BM treated and the average of treated-BM per-patient per-course was 3.6 (range1-13). The average total treated BM per-patient (sum of all courses) was 4.4. Lung cancer was the most frequent (63%) primary tumor.The 77% of cases were patients with a brain relapse and the remaining 23% had BM at diagnosis.

The 92.5% of BM were treated with single fraction. The most used fractionations were 20 (27.8%) and 21Gy (43.5%) respectively and the median PTV target volume (if single fraction) was 0,2cc(range 0,016-4,32cc).The median Cumulative Target Volume per isocenter and the sum of all SRS courses were 1.37 and 1.46cc respectively. The 100% of patients completed the SRS-treatment with no incidences.

With an average follow-up of 5.3 months (0.1-19months), we have not identified any local-relapse although 27% developed an intracranial-relapse that was treated again with SRS in 44% of cases. We didn´t find any relation between overall-survival and the presence of any driver-mutation (p=0.97), BM at diagnosis vs. recurrences (p=0.113), number of SRS courses (p=0.688), number of isocenters (p=0.679) or number of treated-BM (1 vs. 2-3 vs. ≥4; p=0.7). Healthy-normal tissue constraints were adequately accomplished with a median V12 (if single-dose) and V20 (if 5-fractions) of 0.2 and 5cc respectively. No acute-toxicity >Grade2 was reported.

Conclusion:

Based on our preliminary experience and limited by the short follow-up, we find single isocenter and single and multi-target-SRS technique is feasible, well tolerated and allows excellent local control. Overall survival didn´t show differences regarding the number of treated BM.

BACKGROUND: The uncertainty surrounding the selection of an appropriate treatment for patients with a limited number (<=3) of large volume brain metastases (LBM) persists. Recent researches have indicated that staged or fractionated stereotactic radiosurgery yields a notable response rate and tolerable toxicity levels in such patients. This study aimed to assess the effectiveness and safety of hyperfractionated CyberKnife radiosurgery as a novel treatment approach for limited number of large volume brain metastases patients.

METHODS: Patients with LBM treated with hyperfractionated CyberKnife radiosurgery were included in this study. Hyperfractionated stereotactic radiosurgery (FSRS) dose was 21-34 Gy (3-5 fractions) with 64%-70% isodose line by CyberKnife according to the brain tumor volume, site, and previous dose. The primary objective was to identify the overall survival after salvage treatment. Secondary objectives included progression-free survival (PFS), clinical response (Karnofsky performance scale), imaging response (Magnetic Resonance Imaging, MRI) and treatment-related adverse events.

RESULTS: Between January 2020 and December 2022, a total of 40 patients were included in the study. The one-year overall survival rate following FSRS was 75%. Positive imaging responses were observed in 36 patients, accounting for 90% of the cohort, with a T1 weighted contrast MRI volume range from 10.4 to 47.2 cm3. The study also demonstrated a significant clinical improvement, as evidenced by the best Karnofsky performance scale score (P < 0.05, paired t-test). Among the participants, 12 patients (30%) experienced Grade 1 or 2 fatigue, while 4 patients reported Grade 3 headache. Additionally, the median CNS PFS of patients with LBM from non-small-cell lung cancer (NSCLC) was significantly longer compared to other cancer types (24.5 months vs. 12.5 months, P = 0.03).

CONCLUSIONS: FSRS showed favorable clinical and radiologic control as a new treatment regimen for limited number large volume brain metastases. NSCLC patients appear to benefit more from the treatment. To further evaluate this conclusion, an ongoing multicenter prospective observational study is being conducted to assess the efficacy of FSRS for LBM from NSCLC.

Background and purpose: Glioblastomas all eventually relapse after initial treatment, and an option to treat these recurrences is fractionated stereotactic body reirradiation (fSRT). The location of recurrences after reirradiation have been studied, but not precisely after fSRT delivered by a dedicated stereotactic device. We aimed to analyze the patterns of these recurrences after fSRT, as there is limited data to sharpen the choice of safety margins and dose and fractionation regimen.

 Materials and Methods: We retrospectively analyzed the data of patients with glioblastoma recurrence reirradiated by fSRT between October 2010 and December 2020, in 25 Gy in 5 fractions delivered by a CyberKnife® at Institut de Cancérologie de Lorraine. We matched the images of the relapse post-fSRT with the stereotactic radiation treatment planning scan to determine the relapse location.

 Results: Among 62 patients, we found that the localization of recurrences after fSRT was “out-field” in 54.8%, “marginal” in 40.3% and “in-field” in 4.8%. The median PFS from fSRT was 3.4 months (95% CI 2.9 – 4.8 months). KPS score ≥ 70% at recurrence (HR = 0.27 [95% CI 0.08 – 0.93], p = 0.038), PTV volume ≥ 35cc (HR = 3.61 [95% CI 1.23 – 10.6], p = 0.02) and existence of one or more previous recurrences (HR = 2.32 [95% CI 1.07 – 5.05], p = 0.033) were significantly associated with PFS. The median OS from diagnosis was 25.7 months (95% CI 22.2 – 32 months), and from fSRT was 10.8 months (95% CI 8.97 – 14.8 months).

Conclusion: Reirradiation of glioblastoma by fSRT with 25 Gy in 5 fractions provides good local control, with recurrences occurring mostly outside of the reirradiated area.

Objectives

To report clinical outcomes and identify predictive factors associated with improved treatment results in Linac-based Stereotactic Radiosurgery (SRS) and fractionated Stereotactic Radiosurgery (fSRS) for single and multiple brain metastases (BM).

Methods

Between March 2020 and June 2022, 70 patients for a total of 129 BM with at least one-month follow-up were retrospectively included. Patients received either 15-21 Gy in a single fraction (n=59) or 27 Gy in three fractions (n=11) using single-isocenter coplanar FFF-VMAT technique. Post-treatment MRI scans were used to assess local control (LC) according to the RECIST (Response Evaluation Criteria in Solid Tumors) scale. Kaplan-Meier analysis was performed to evaluate in-field progression-free survival (ifPFS), brain progression-free survival (bPFS), and overall survival (OS) rates. Log-rank test and logistic regression analyses were carried out to identify predictive factors associated with better outcomes.

Results

The population consisted of 33 females and 37 males, with a median age of 66 years [30-85]. Lung (44%) and visceral (47%) were the most frequent tumor histology and extracranial metastases site, respectively. The median follow-up period was 9 months [1-41]. The 1-year and 2-year LC rates for all lesions were 94% and 90%, respectively, with 13 (19%) patients experiencing local recurrence in at least one treated BM. The median ifPFS was 7.8 months, while the corresponding 1-year and 2-year rates were 80% and 72%, respectively. The median bPFS was 3.9 months, with 1-year and 2-year bFPS rates of 40% and 20%, respectively. The same features for OS were 13 months, 52%, and 29%, respectively. Lung primary tumor histology and non-visceral extracranial metastases were significantly associated with increased OS and bPFS. No statistically significant differences in clinical outcomes (P>0.05) were found for number of treated lesions, total target volume, BM minimum dose and systemic therapy. Patient age and gender showed borderline significant correlations with bPFS (P=0.055) and ifPFS (P=0.060), respectively. Extended bPFS was observed in younger than 66 years patients (mean, 9.9 vs 6.4 months), while female patients had superior ifPFS (mean, 11.6 vs 9.1 months). At multivariate analysis, lung primary tumor histology was independently related to brain progression (OR, 0.35; 95% CI, 0.12–0.98; P=0.043). 

Conclusions

Linac-based SRS/fSRS treatments with single-isocenter coplanar FFF-VMAT technique were feasible and resulted in encouraging LC outcomes. Patient prognosis remains unfavorable, mostly dependent on histology and extracranial disease status, rather than on the radiation treatment. Further analyses on a larger patient population are currently underway to confirm these findings.

Background: Lung cancer (LC) is an important health problem for its incidence and mortality, with 9 of each 10 death related with metastatic dissemination, and 20-50% of NSCLC present brain metastases during the follow up, with < 12 months survival).

Recent advances in LC treatment have incorporated immunotherapy and molecular therapies, and radical local treatment for oligometastases. Radiosurgery (SRS) represents a non-invasive treatment suitable for patients with intracranial relapse.

Nevertheless, not all patients get a good response to SRS, and the identification of the probability of success has been recognized as necessary, allowing a stratification useful to avoid an eventual overtreatment and the over costs associated.

Mean objective: To determinate the clinical utility of the circulating tumoral cell (CTC) and ctDNA to predict the efficacy of local ablative treatment of brain oligometastases in clinical outcomes, and stratify the recurrence risk of failure. 1) To associate CTC presence and phenotypic characteristics with progression free survival; 2) To associate genetics perfils and ctDNA with progression free survival, and 3) To determinate the role of radiosurgery treatment over fragments of ctDNA detected.

Material and Methods.

In this exploratory research, a successive group of 30 brain mets from NSCLC patients, treated with SRS (18-22 Gy Gamma Knife single session), will be collected for CTC and ctDNA determination. A 30ml of peripheral blood will be processed, 10ml will be allocated for CTCs analysis by semi-automatic technologies based on Isoflux isolation and CTC characterization through the Ammnis platform.

The remaining 20ml will be dedicated to molecular analyses (epigenomic analysis by the TruSight™ Oncology 500 ctDNA kit by Illumina).  The sequencing will be executed on a NovaSeq 6000 (Illumina) at the our hospital.

For fragmentomic analyses, we will adopt a whole-genome sequencing (WGS) approach based on LIFE-CAN, applying it to both NSCLC and healthy donors.

The Bioinformatic unit of Fundación Progreso y Salud will responsible of the bioinformatic analyses

Results. Patients will be subjected to a usual follow-up protocol including clinical assessment and brain MRI evaluation at 6 week and 3 months after SRS. Relationship among CTC amounts and genetic signature, will be related with response, disease free survival, place of relapse and overall survival.

For this preliminary study, a basic statistic will be carried out, and a significance difference would allow us for a posterior protocol, including other brain metastases from other primary location like melanoma, breast and colon cancer.  

Background

For stereotactic radiosurgery (SRS) planning, linac-based dynamic conformal arc (DCA) or volumetric modulated arc therapy (VMAT) plan quality is highly dependent on the planner’s experience. Due to its superior normal tissue sparing, Gamma Knife is the gold standard for patients with 1-4 brain metastases, but treatment time could become prohibitively long for patients with many targets.

Aim

To develop a machine learning-based plan quality evaluation tool for linac-based multiple brain metastases SRS plans using the idealized Gamma Knife plans as the benchmark to assist treatment planning.

Methods

41 patients with multiple brain metastases (range: 5-22) treated with SRS were included in this retrospective study. Idealized Gamma Knife plans, without limiting treatment delivery time, were created for all patients to be used as the benchmark for model development. The dataset was split into 25/7/9 for training/validation/testing. A 3D U-Net was used to predict the benchmark dose around each target. The input is the target contour mask in a 6.4 cm wide region-of-interest (ROI) centered at a target, and the output is the 3D dose distribution in the ROI. To account for dose falloff outside the target and adjacent target outside the ROI, a 1-cm wide exponential falloff was added for each target. To focus on the dose falloff region, a modified mean absolute error was used as the loss function, which added a discount factor of 0.2 for voxels with benchmark dose below 2 Gy or above 16 Gy. The ROI dose for each target was predicted by the network and filled back in the patient volume to obtain a partially filled 3D dose distribution for evaluation.

Results

To avoid overfitting, the final model was trained for 21 epochs when minimal validation loss was reached. It was tested on 9 patients with 66 targets in total. The average PTV volume is 0.98±1.99 cc (range: [0.02,12.88]). The mean brain V12Gy, V8Gy, V4Gy errors were -1.07±0.65 cc, -0.51±0.52 cc, 2.53±8.22 cc for all patients and -0.22±0.28 cc, -0.19±0.39 cc, 0.77±1.36 cc for all targets. The dice coefficients for the benchmark and predicted 12 Gy, 8 Gy, 4 Gy isodose lines for all targets were 0.86±0.08, 0.86±0.11, and 0.81±0.10.

Conclusion

A deep learning model was trained to predict the dose distribution of an idealized Gamma Knife plan for multiple brain metastases SRS patients, which can be used as benchmark to guide treatment planning using other delivery techniques such as linac and CyberKnife.

Background and Purpose

Patients with EGFR-mutated NSCLC represent a unique subset of lung cancer patients with distinct clinical and molecular characteristics. Previous studies have shown a higher incidence of brain metastases (BM) in this subgroup of patients, and neurologic death has been reported to be as high as 40% and correlates with leptomeningeal disease (LMD).

Methods

Between 2012 and 2021, a retrospective review of our prospective registry identified 606 patients with BM from NSCLC, with 170 patients having an EGFR mutation. Demographic, clinical, radiographic, and treatment characteristics were correlated to the incidence of LMD and survival.

Results

LMD was identified in 22.3% of patients (n = 38) at a median follow-up of 19 (2–98) months from initial SRS. Multivariate regression analysis showed targeted therapy and a cumulative number of metastases as significant predictors of LMD (p = 0.034, HR = 0.44), (p = .04, HR = 1.02).

The median survival time after SRS of the 170 patients was 24 months (CI 95% 19.1–28.1). In a multivariate Cox regression analysis, RPA, exon 19 deletion, and osimertinib treatment were significant predictors of overall survival. The cumulative incidence of neurological death at 2 and 4 years post initial stereotactic radiosurgery (SRS) was 8% and 11%, respectively, and correlated with LMD.

Conclusion

The study shows that current-generation targeted therapy for EGFR-mutated NSCLC patients may prevent the development and progression of LMD, leading to improved survival outcomes. Nevertheless, LMD is associated with poor outcomes and neurologic death, making innovative strategies to treat LMD essential.

Purpose

Over the last decade Stereotactic Radiosurgery (SRS) and Fractionated Stereotactic Radiotherapy (FSRT) became standard of care for limited brain metastases.  We now evaluated our cohort with robotic SRS/FSRT of the past 10 years.

Material and Methods

323 patients (157 male, 166 female, age 27-86) of different histologies (lung cancer 132, melanoma 81, breast cancer 56, other 54) with a total of 1164 brain metastases (BM) were treated in 500 series. Simultaneous systemic targeted therapies and/or immunotherapy (TT) were given in 81 (25.1%) cases. Number of BM was 1, 2-10 and >10 in 110, 195 and 18 cases, respectively and 81 patients had received Whole-Brain-Radiotherapy (WBRT) before SRS/FSRT. The median PTV was 0.45ccm (0.01-78.8ccm) with a GTV-PTV-margin of 0-1mm. Median D_98%, D_50% and D_2% of all PTV calculated as biological effective dose with an alpha/beta-value of 10 Gy (BED₁₀) averaged 51.2Gy₁₀ (20.1-63.5Gy₁₀), 75.2 Gy₁₀ (24.8-120.7Gy₁₀) and 106.4 Gy₁₀ (27.8-143.6Gy₁₀), respectively.

Results

Mean follow-up period was 14.8 (0-109) months and the median overall survival (OS) was 8.7 months with 12- and 24-months OS of 45% and 21.6%, respectively. Significant differences in the 12-months OS were seen for melanoma patients with 12-month OS of 50% vs. 45% (whole cohort, p=0.05). Prior WBRT was associated with a non-significant reduction of 12-months OS after SRS/FSRT (10.0 vs. 12.8 months with and without WBRT, p=0.054). The most significant prognostic factor for longer OS was Karnofsky Performance Status (KPS) of ≥90% (p=0.001). Overall PTV (OPTV) <2.6ccm was also associated with a longer OS of 15 months vs. 10 months with OPTV ≥2.6ccm (p<0.01). Simultaneous TT-application led to prolonged OS of 14.5 vs. 10.6 months (w and w/o TT, p=0.227). Local control (LC) after 12 months was 92.4%. A higher PTV D_98% lead to better LC (96.0% vs. 81.5% for BED ≥51.2Gy₁₀, p=0.024), whereas a higher PTV D_2% had no significant effect (92.5% vs. 92.1% for BED ≥106.4 Gy₁₀, p=0.701). Localization of relapses was at the edge of the PTV in 16 and inside the PTV in 29 cases. Bigger metastases (GTV ≥0.45ccm, corresponding diameter of 0.95cm) recurred more often than smaller ones (p=0.047). Rate of side-effects was low (grade ≥3 2%). In 1 case repeated SRS with simultaneous BRAF inhibition led to an fatal intracerebral bleeding (grade 5).

Conclusion

Robotic SRS/FSRT is safe and effective. In the context of central dose-optimization higher PTV D_98% improved LC. Caution is advised for simultaneous re-treatment with BRAF inhibition.

OBJECTIVE

Treatment with immune checkpoint inhibitors (ICIs) has demonstrated clinical benefit for a wide range of cancer types. The neutrophil-to-lymphocyte ratio (NLR) reportedly correlates with survival time or progression-free survival in patients treated with ICIs. However, NLR has not yet been assessed in patients with brain metastases (BMs) in the setting of stereotactic radiosurgery (SRS) combined with concurrent ICIs. The present study sought to investigate the predictive impact of NLR on the survival data of patients with BMs who received SRS with concurrent ICIs.

METHODS

The clinical records of patients who received SRS with concurrent ICIs for BMs between January 2015 and August 2023 were retrospectively analyzed. Neutrophil-to-lymphocyte ratio (NLR) was calculated by using the data obtained from the latest examination prior to SRS. The optimal NLR cutoff value was identified by receiver operating characteristic (ROC) curve analysis for time-to -event data (overall survival (OS) ≤ 18 months). OS and intracranial disease progression-free survival (IC-PFS) rates were compared between two NLR groups.

RESULTS

Of the 185 eligible patients included, 132 patients were male. The median age of the patients was 69 years (IQR 61–75 years). The primary cancers were lung, genitourinary, skin, breast, gastrointestinal, and other cancers in 132, 23, 22, 2, 2 and 4 patients, respectively. The post-SRS median OS and IC-PFS time for the entire cohort was 18.9 months (IQR 14.0–23.1 months) and  9.8 months (IQR 7.5–11.6 months), respectively. ROC curve analysis identified NLR cutoff value as 5.0 (area under the curve: 0.63, Youden index: 0.30). Kaplan-Meier analysis revealed that patients with high NLR (> 5) had a significantly shorter OS (median survival time 10.1 months for 48 patients vs. 22.2 months for 137 counterparts, HR 1.9, 95% CI 1.3–2.9, p = 0.002). Similarly, a significant difference in the median IC-PFS was found: 5.6 months with NLR > 5 vs. 11.3 months with NLR ≤ 5 (HR 1.7, 95% CI 1.2–2.6, p = 0.009).

CONCLUSIONS

The present study found that an elevated pre-SRS NLR (> 5) was associated with shorter survival and worse intracranial disease control after SRS with concurrent ICIs for BMs. NLR is a simple, cost-effective and widely accessible biomarker, and can be used in SRS treatment for patients with BMs being treated with concurrent ICIs. Further investigation in other large datasets is however required to validate these findings.

Our institution upgraded the Gamma Knife radiosurgery (GKRS) system from Model 4C to ICON in 2019. Because about 30% of the patients were treated using the mask immobilization technique and single or multiple fraction scheme with this new unit, we wanted to confirm that the treatment outcome was as good as that obtained by the 4C unit with the Leksell G-frame and single fraction. This paper presents the outcome analysis of 71 patients with GKRS on the ICON unit from 2019 to 2022.

The patient population consisted of 30 (42%) males and 41 (58%) females, with a mean age of 60.8 years old (range: 16 - 92). 23 patients were treated for benign tumors, including 12 meningiomas and 4 vestibular schwannomas. 45 patients with malignant tumors were treated for their metastatic lesions, with a mean number of lesions of 1.85 per person. The most common primary histology of metastatic cancer patients was lung cancer (19), melanoma (8), and breast cancer (5). Ten patients had previous radiotherapy to the brain. All patients were treated with a head immobilization mask, daily cone-beam CT, and real-time motion management. After radiation oncologists and neurosurgeons drew target contours on the MRI, a 1-mm margin was uniformly added to generate the treatment volume. The prescription dose was 24 Gy on average (6 – 30). The mean tumor volume was 4.61+/- 6.53 ml (0.0040-31.4). Our institutional policy is to use fractioned GKRS with a maximum dimension of tumor size greater than 2 cm (or 4 ml). There were 58 patients treated with 3 or 5 fractions. The patients had follow-up MRI scans every three months. The mean follow-up length was 412 days (28-1176). The Kaplan-Meier (KM) analysis was done using R.

The survival rates of patients were 87%, 73%, and 66% for 6 months, 1 year, and 2 years respectively. There was no statistically significant difference between males and females (p=0.83). The patients with two or fewer malignant tumors lived longer than those with more than two tumors (p = 0.0001), but the total tumor volume did not affect survival. It is notable that only two lesions out of 131 locally failed.

The preliminary data of mask-based GKRS by a single institution showed promising outcome results. In the future, the brain toxicity of the new GKRS protocol needs to be analyzed to further confirm its clinical efficacy compared with the single fraction GKRS with the G-frame.

Objectives

The aim of this study is to present the results of GKRS for patients with brain metastases who survived more than 5 years. 

Methods

52 patients with brain metastases, who underwent radiosurgery with Leksell Gamma Knife 4C or Perfexion (Elekta AB, Sweden) and survived at least 5 years after the treatment, were included in the study. There were 21 men and 31 women. The most common primary tumors were breast cancer, lung cancer, melanoma and renal cell carcinoma for 14, 13, 9 and 7 patients, respectively. 15 patients were diagnosed with primary cancer simultaneously with the diagnosis of brain metastases. For 32 patients the brain was the only site of tumor spread. GKRS was performed for 1 to 30 brain metastases. The prescribed radiation dose varied from 16 to 24 Gy at 40-85% isodose. After treatment, the patients underwent regular follow-up examinations (MRI and/ or PET with amino acids). Differential diagnosis of tumor recurrence and radiation necrosis was performed with the help of amino acid PET. Overall survival, local control and radiation necrosis were evaluated with the help of the Kaplan-Meier function. Significance between groups was calculated with the Log-Rank test. 

Results

The median follow-up time after GKRS was 85 months (mean – 93, range 60 – 160). 32 patients were alive at the time of analysis (December 2023) and 11 patients survived more than 10 years after GKRS. The median OS was 121 months. The actuarial survival rates at 6 years were 78.5 %, 8 years – 62.3% and 10 years – 55.4%. Statistically significant factors for OS were the patient’s gender and primary tumor type (p=0.024 and p=0.002). Other factors (age, KPS, presence of extracranial metastases, number of brain metastases) were not significant. Tumor recurrence was detected in 15 metastases (14 patients) within a median of 35 months after GKRS (mean 39, range 10 – 81). Primary tumor type, brain metastasis volume and radiation dose were significant factors affecting local control (p < 0.01). Signs of radiation necrosis were observed on MRI in 27 metastases (20 patients) within a median time of 13 months after GKRS (mean 18, range 4 – 60). Tumor volume and radiation dose were significant factors associated with radiation necrosis (p < 0.01).

Conclusions

Gamma Knife radiosurgery is a reliable treatment for patients with brain metastases, with a high level of efficacy and safety estimated over a long-term observation period. 

Introduction Brain metastases during pregnancy poses complex conundrum in management. Gamma Knife (GK) stereotactic radiosurgery (SRS) offers a valuable option to clinicians in this scenario. We describe the safety and effectiveness of GK SRS in treating a solitary cerebellar metastasis in a woman with recurrent breast cancer in the third trimester of pregnancy. Dosimetry readings during a trial run and actual treatment were recorded and follow-up MRI was performed after one month. A Literature review on similar cases were carried out.

Methods A 42-year-old woman presented with dizziness and unsteady gait during her third pregnancy at 28 weeks of gestation. She was a known case of triple negative breast carcinoma with local recurrence in 2021 and had completed second line chemotherapy 10-months prior to referral. Upon presentation, she was fully conscious with neurological examination showing right cerebellar signs. MRI brain showed solitary right cerebellar enhancing mass, 2x2.7x2.1cm with perilesional edema and hemosiderin rim likely represent hemorrhagic metastasis. Chest radiograph depicted multiple cannon ball lesions. Obstetrical assessment revealed singleton fetus with gestation appropriate growth parameters and an estimated fetal weight of 1kg. Following multidisciplinary discussion, she agreed for urgent single session SRS to the brain metastasis with 2 cycles of 3-weekly paclitaxel chemotherapy. During frame-based GK SRS, a trial run with dosimeters placed on a phantom showed radiation exposure way below the 100mSv dose limit. Actual treatment was performed with 16Gy at 50% isodose in 24 shots over 39.7 minutes beam on time. The treatment plan showed 98% coverage, 89% selectivity and gradient index 2.98.

Results Dosimeters placed near uterine fundus and suprapubic region (consistent with concomitant ultrasound localization of the fetal head) recorded 2.83mSv and 0.27mSv respectively. The patient successfully completed SRS treatment without complications. She safely delivered a healthy baby boy at 36 weeks of pregnancy. Follow-up MRI at three months interval showed total resolution of the lesion. Our literature review revealed one other similar case report which was a patient with melanoma brain metastasis in the second trimester of pregnancy who successfully completed the planned GK treatment.

Conclusions GK SRS is known for the lowest extracranial dose of all SRS modalities. It is safe and effective in treating pregnant patients with brain metastases. It allows concurrent chemotherapy, eliminates anesthetic risk while giving time to achieve adequate gestational age and fetal weight before birth. It improves quality of life and fetal outcome with lower perinatal risk and maternal morbidity.

Objectives

To describe the state-of-the-art of Stereotactic Radiosurgery (SRS) for brain metastases in centers across Latin America and Spain, through a survey conducted among radio-oncologists and neurosurgeons. This study details technological platforms, SRS protocols, and examines regional variabilities.

Materials and Methods

We conducted a specific SRS survey (26 questions) via Google Drive, targeting professionals from Latin America and Spain, utilizing the database of the Ibero-Latin American Radiosurgery Society.

Responses from 106 specialists were analyzed. The survey was designed to provide a comprehensive overview of SRS practice in the region.

Results

1. Participation: 93.4% of respondents were from LATAM, and 6.6% from outside LATAM. A participation of 85% of LATAM countries was obtained. 87% of respondents were Radio-Oncologists, and 13% were Neurosurgeons.

2. Experience and Certification: 62% of specialists had more than10 years of experience in SRS practice, and 80% worked in centers with some SRS certification.

3. Technologies Used for SRS: LINAC 70%, Gamma Knife 15%, CyberKnife 7%, Halcyon 5%, and ZAP 3%.

4. Prescription Dose and Adjustment by Histology: The average dose for an example case of breast histology was 22 Gy [18-26 Gy], while for renal histology it was 24 Gy [20-28 Gy]. 45% of participants adjusted the prescription dose according to histology.

5. Dose Adjustments & OARs: 60% of respondents adhere to RTOG 90-05 guidelines (dose prescripcion related to tumor size) to adjust the dose in situations of prior radiation or voluminous metastases.

6. Fractionated SRS: Used in 50% of cases, with an average dose of 24 Gy [18-30 Gy], especially in large metastases or those close to critical organs.

7. Use of V12 as a predictor of Radionecrosis: Used by 91.5% of respondents, but only 56.6% use it regardless of the total number of lesions.

8. Re-SRS for Recurrences: 41.51% prescribe the same dose initially used. 38.68% choose a lower dose, and 3.77% a higher dose.

9. Post-SRS Follow-up: Conducted every 3-6 months with MRI, being the common practice in 85% of cases.

Conclusions

SRS practice in Latin America shows variability that can be compared to that evidenced internationally in the literature.

The findings underscore the importance of generating internationally accepted protocols and regional consensus, to standardize SRS practice and ensure optimal outcomes for these patients.

Purpose) To examine the effectiveness of stereotactic radiation for intracranial solitary fibrous tumors, taking into account differences in modality. Background) Among solitary fibrous tumors, those that originate within the intracranial space often follow an aggressive course, and radiation treatment after surgery or for recurrent lesions is thought to play an important role. At our hospital, we use stereotactic radiosurgery (SRS) and hypo-fractionated stereotactic radiation therapy (SRT) using a Gamma Knife, or hyper-fractionated SRT using a Liniac, depending on the case. Methods) From April 2004 to September 2023, stereotactic radiation treatment was performed on 24 cases of intracranial solitary fibrous tumors. The average age of the patients at the time of first radiation treatment was 47.3 years. SRS or hypo-fractionated SRT using Gamma Knife was performed for small lesions, and hyper-fractionated SRT using Liniac was performed for large lesions and/or those were close to risk organs (16 cases using only Gamma Knife, only Linac hyper-fractionated SRT in 6 cases, Gamma knife and Linac hyper-fractionated SRT in 2 cases). The average lesion volume at the time of initial radiation treatment was 5.4 ml with Gamma Knife and 43.0 ml with Linac hyper-fractionated SRT. The irradiation doses were: Gamma Knife SRS with an average marginal dose of 16.4 Gy, Gamma Knife SRT with an average marginal dose of 31.2 Gy, and Linac hyper-fractionated SRT with an average marginal dose of 47.5 Gy in 15 -20 fractions. Results) An average follow-up period of 77.1 months was obtained. Progression-free survival rates were 94.7%, 60.9%, and 42.6%, respectively, at 1, 3, and 5 years after initial radiation treatment with Gamma Knife, and 100%, 83.3%, and 62.5%, respectively, with Linac hyper-fractionated SRT. At the time of final observation, 20 out of 24 patients were alive. Conclusion) Good control of the intracranial solitary fibrous tumor was obtained by performing stereotactic radiation taking into consideration the modality characteristics.

There has been a significant increase in the clinical utilization of stereotactic radiosurgery (SRS) for brain metastases (BM). Reducing the dose to the hippocampi in SRS holds promising implications for improved neurocognitive outcomes. HyperArc^® high-definition radiotherapy (HA)  is a single isocenter end-to-end solution for treating BM and other intracranial targets. This study aimed to evaluate dosimetric outcomes of patients receiving HA regarding factors that may impact symptomatic neurocognitive outcomes.

This is a retrospective study of patients who received HA treatment at selected Icon Cancer Centres in Australia between September 2018 and March 2022. Data included for analysis were previously uploaded to the HyperArc Registry (https://clinicaltrials.gov/study/NCT05270707), including demographics, target and organs at risk (OAR) dosimetry, primary tumour characteristics and patient outcomes (neurological symptoms and overall survival). Hippocampi were retrospectively contoured if not present in the original plan. Brain dose was defined as brain minus the total planning target volume (PTV). 

A total of 110 patients receiving 139 courses of radiation therapy were included for analysis. The median age at treatment was 67 years and tumor histology was predominantly non-small cell lung cancer, breast cancer or melanoma. The median number of treated metastases was four (interquartile range 3-6). Treatments were typically 24Gy/3# (n=83/139) and 30Gy/5# (n=35/139). Plans with a total PTV  >10 cm³ had significantly greater mean brain volume doses than PTV <10 cm³ (5.14 Gy vs. 2.14 Gy, p-value <0.001). Furthermore, the mean bilateral hippocampus dose exhibited a positive correlation with the total PTV. Plans with at least one target located within 2 cm of the hippocampi had markedly greater mean hippocampi doses compared to plans where all targets were greater than 2 cm from the hippocampi (4.83 Gy vs. 1.74 Gy, p-value <0.001). On multivariate analysis, the volume of the closest treated brain metastasis to the hippocampi was not predictive of hippocampal dose. Patients presenting with symptoms were more likely to have a greater disease volume (total PTV 23.86 cm³ vs. 17.15 cm³, p-value=0.001); however, the number of brain metastases was not predictive of symptoms. Patients with neurological symptoms at baseline were significantly more likely to experience neurological symptoms during follow-up (OR = 5.6, 95% CI 2.23-14.1).

Mean brain doses are correlated with total PTV. When the total PTV is greater than 10 cm³, the mean brain dose tends to be greater than 5 Gy. Hippocampi should be considered as organs at risk (OAR) and optimized in plans with targets

[Objective] We introduced the Leksell Gamma Knife Icon (ICON) in November 2016, and started fractionated irradiation for large metastatic brain tumors. In the present study, we investigated the efficacy of ICON in the treatment of large metastatic brain tumors. [Methods and Subjects] We included 178 patients who received ICON fractionated radiosurgery between December 1, 2016, and December 31, 2021, and who could be followed up for more than 1 year. [Results] The Gamma Knife fractionated irradiations were 30 Gy / 3 fx, 35 Gy / 5 fx, and 40-42 Gy / 8-10 fx. The number of patients divided into three groups by number of fractionations was 26, 94, and 58, respectively, and the mean volume of the irradiated object was 7.4 cm3, 11.8 cm3, and 25.2 m3, respectively. Median survival was 10.6 months overall and 16.6, 8.3, and 12.2 months for each fraction, respectively, with no significant differences. Kaplan-Meier analysis showed significantly longer MST in women, KPS ≥ 80, and primary breast and lung cancer, and these factors were significant in multivariate analysis. The overall cumulative recurrence rate was 6.9% at 1 year and 9.7% at 2 years. Competing risk analysis of the associated factors showed an increased recurrence rate for lesions larger than 14 cm3 in the five-fractionation group. The incidence of delayed radiation injury was 7.3% at 1 year and 9.7% at 2 years, with a trend toward higher incidence in the 3-fraction and 14 cm3 or greater groups, but the only significant factor was female gender. [Conclusion] With the introduction of ICON, effective and safe treatment of even large tumors is now possible. The number of cases that can be treated with gamma knife therapy is increasing, even in cases where craniotomy was previously considered an indication, and future treatment strategies for metastatic brain tumors should also be considered.

Surgical resection is used to treat brain metastases  may be associated with the risk of developing leptomeningeal disease (LMD). Stereotactic radiation therapy (SRT) is an effective strategy for the treatment of large brain metastases and may be an alternative to surgery or an adjuvant component.

Objective: To compare different techniques for the treatment of large brain metastases   examined rates and predictors of leptomeningeal disease.

Methods: 369 patients with large (≥2 cm in diameter) BMs  were underwent surgical treatment (S) alone (72 patients), or adjuvant SRT in hypofractionation (F) mode (126 patients), or neoadjuvant SRS (65 patients) or FSRT alone (106 patients) between 2011 and 2022. Among them were patients with non-small cell lung cancer (86), breast cancer (123), melanoma (59), kidney cancer (35), gastrointestinal cancer, (44)  and gynecologic cancer  (22) . Categorical baseline characteristics were compared using the χ2 test. LMD scores were assessed by the Kaplan-Meier (KM) method, and the log-rank test was used to compare subgroups.

Results: LMD was detected in 81 (21,9%) of 369 cases including 27.7%, 32.5%, 12.31% and 11.32% in S, S+FSRT, SRS+S and FSRT subgroups, respectively. The KM estimates of 12-month and 24-month LMD-free survival in the S, S+FSRT, SRS+S, and FSRT groups were 75.5% and 62%, 70.6% and 61.8%, 84% and 81%, 88.9% and 84.7%, respectively (P = 0.0031). The hazard ratio for developing LMD comparing with patients who received FSRT alone were 2.7(CI 1.42 to 5.23), 2.7 (CI 1.6 to 4.7), and 1.26 (CI 0.64 to 2.5) in the S, S+FSRT, and SRS+S groups. The 12-month LMD-free survival rates of large BMs in the non-small cell lung cancer, breast cancer, melanoma, kidney cancer, gastrointestinal cancer, and gynecologic cancer subgroups were 81.7%, 74%, 80.6%, 85.6%, 75.8%, and 63.8% respectively (P = 0.0388).

Conclusions: The risk of developing LMD depends on the primary focus, lower in the non-small cell lung cancer and kidney cancer subgroups. Surgery increase the risk of developing LMD compared to FSRT alone. SRS+S and FSRT have similar  low risk of developing LMD, and  may be the method of choice for patients with large BMs. 

Objective.

The impact of the stereotactic radiosurgery (SRS) prescription dose (PD) on local progression for small (≤ 1 cm) brain metastases was evaluated.

Methods.

An retrospective review was performed on 247 patients with brain metastases ≤ 1 cm (2070 tumors) who received  SRS with Gamma Knife  Icon between 2015 and 2022. Local progression were the primary end points. Each tumor was followed from the date of radiosurgery until one of the end points was reached or the last MRI follow-up. The median radiographic follow-up per lesion was 6,7 months.

Results. 

The median patient age was 57 years, and 58% of the patients were female. The most common primary pathology was breast cancer (36,8%) followed by non–small cell lung cancer (28,3%), melanoma (19%), renal cell carcinoma (12,5% %) and colorectal cancer (3,2%). The median tumor volume was 0,056 (95% CI 0,051–0,064) cm3. The PD for 1530 tumors (73,9%) was 24 Gy, for 233 tumors (11,2%) it was 22 Gy, and for 307 tumors (14,8%) it was 20 Gy.  

In total, 14 patients (5,6%) had local progression of 79 tumors (3,8%). The local progression for PD 24 Gy was in 61 tumors (3,9%), for PD 22 Gy it was in 6 tumors (2,6%) and for PD 20 Gy it was in 12 tumors (3,9%). In univariate analysis there was no statistical difference (P = 0,3532) in local progression for tumors with a dose of 24, 22 and 20 Gy. 

Conclusions. 

PD (within 20-24 Gy) is not an independent prognostic factor for local control of tumors smaller than 1 cm. Probably some pathologies and locations may also contribute to an increased risk of local progression. Further research is needed.

Objective.

Stereotactic radiosurgery (SRS) is an established primary treatment for newly diagnosed brain metastases with high local control rates. However, data about local re-irradiation in case of local failure after SRS are rare. We studied the effectiveness of treating local relapses with a repeated course of radiosurgery (re-SRS).

Methods.

We retrospectively evaluated patients with brain metastases treated with re-SRS for local tumor progression between 2015 and 2022. Patient and treatment characteristics as well as rates of tumor control and toxicity were analyzed.

Results

Overall, 110 locally recurrent brain metastases in 59 patients were irradiated with re-SRS. Median age at re-SRS was 53 years. The most common primary pathology was breast cancer (49,1%) followed by melanoma (22%), non–small cell lung cancer (16,9%), renal cell carcinoma (8,5% %) and colorectal cancer (3,4%). In the first SRS and in the re-SRS were treated with Gamma Knife. The median tumor volume for the first SRS and in the re-SRS was 0,82 and 1,43 cm3 respectively. Median prescription dose for the first SRS and in the re-SRS was 22 and 20 Gy respectively.

In total, 14 patients (23,7%) had local progression of 18 tumors (16,4%). The 1-year overall survival rate was 85,8% and the 1-year local control rate was 83,9%. The overall rate of radiological radio-necrosis was 28,2%.

Conclusions

A second course of SRS for locally recurrent brain metastases after prior local SRS appears to be feasible with acceptable toxicity and can be considered as treatment option for selected patients. Furthermore, further research is required to establish optimal fractionation regimens for repeat SRS in locally recurrent lesions.

INTRODUCTION:  Glioblastoma (GBM) are often relapse after preliminary removal with subsequent conventional radiation therapy, while often the optimal tactics for treating relapses has not been precisely determined. One of the option for the relapses of GBM relapses is  stereotactic radiotherapy with hypofractionation mode  (HFRT) .

OBJECT: To evaluate the role HFRT with GammaKnife (GK) in patients with recurrent of GBM after resection and fractionated radiation therapy (RT).

METHODS: From July 2018 till December 2023 at “Moscow GammaKnife Center”, which affiliated with Burdenko Neurosurgical Institute (National scientific research Center of neurosurgery named after N.N. Burdenko) 19 patients (8 males and 11 female) with recurrent of GBM was treated by HFRT with LGK Icon. Most patients were older than 55 years – 13 vs 6 pts. Median age at first GK procedure was 60 years (from 21 to 71). 18 patients underwent tumor repeat resection, chemoradiotherapy and adjuvant chemotherapy. One patient was treated without biopsy, after PT-CT with methionine for verification. The median time from initial surgery to GKRS was 17 months. 

RESULTS: The median target volume was 8.7 (from 2.2 to 72.2 cc) and the median dose to the tumor margin was 35 Gy (range 24-35 Gy) for 3 or 5 fractions. Total number of irradiated targets is 43. Average 19 patients followed at least 1 year (max 5 years). Progression-free survival after the initial GKRS was 68.4%, at 1 year. The distant tumor relapse rate despite RT and GKRS was 10.5% at 12 months respectively. Overall survival (OS) after HFRT was 65.5% at 1 year, and 2-year OS reaches 21.0%, respectively. Adverse radiation effects developed in 1 patient (5.2%).

CONCLUSIONS: HFRT by GK in different modes is the treatment of choice, along with reoperation, in patients with recurrent glioblastoma after initial combine treatment.

Abstract Background. One of the rare (1.5-2.5%) brain tumors is gliomas of the brain stem (BSG) in adults. At the time of detection of BSG, patients were about 30 years old, and the usually degree of malignancy of the disease was low (WHO grade I–II). The average life expectancy in adults varies from 30 to 40 months. Surgical treatment is not used in most cases, therefore radiation therapy is the main method of treatment

Objective. To evaluate the results of radiation therapy in adult patients with brain stem tumors and identify predictors of treatment effectiveness.

Material and methods. Radiation therapy was performed in 115 patients with brain stem tumors between 2005 and 2021. Patients under the age of 40 years (n=80), from 40 to 60 years (n=30) and older than 60 years (n=5). The average age was 34.45± 12,873. There were 67 men (58%) and 48 women (42%). Surgical intervention was performed in 44 (38.2%) people, while 71 (61.7%) did not. The functional state was assessed according to the Karnovsky index (IK). in IK 70% - 53 (46%), IK 80% - 45 (39.1%), IK 90% - 12 (10.4%) and IK 60% - 5 (4.5%). The average follow-up period after radiosurgery was 119.8 months. All patients received radiation therapy in the conventional mode, a single dose of 2 Gy, a total mean dose of 54 Gy.

Results. Radiation therapy for patients with brain stem tumors improved progression-free survival and overall survival. The overall cumulative survival rate at 12, 24, 36 and 60 months was 96.5%; 92.7%; 83.5% and 68.7%. The median disease-free survival was 43.47 months (95% CI from 30.5 to 56.4). The rate of disease-free survival for the research cohort of patients (total sample n=105) in the range of 12, 24.36 and 60 months was 79.5%; 72.4%; 58.9% and 38.4%, respectively. Significant factors influencing the outcome of treatment are: age, functional state and histological form of the tumor.

Conclusion.  Radiation therapy for adult patients with brain stem gliomas  in the standard fractionation mode at a total dose of 54 Gy is effective treatment method and is the method of choice for these patients.