ABSTRACT

Introduction:  Survival prognostication is an important aspect of personalizing oncologic care for patients with melanoma brain metastasis (BM).  We previous demonstrated the utility of a cumulative intracranial tumor volume modified diagnosis-specific graded prognostic assessment scale (CITV-dsGPA) for SRS-treated melanoma BM patients. Pertinent prognostic variables in this model included age, Karnofsky performance status (KPS), and CITV.  Here we determined whether the incorporation of BRAF mutation status into this CITV-modified scale further enhanced its prognostic accuracy.   

Methods:  We collated the survival pattern of 331 melanoma BM patients with known BRAF mutation status treated with stereotactic radiosurgery (SRS) and validated our findings in an independent cohort of 174 patients. All patients with BRAF mutation were treated with BRAF inhibitors. The prognostic utility of the model with and without BRAF mutation information was compared using the net reclassification index (NRI > 0) and integrated discrimination improvement (IDI) metric.

Results:   BRAF mutation status is an important determinant of clinical survival in both univariate analysis (Hazard Ratio for death for BRAF mutated melanomas (HR) = 0.74, p<0.001 as well as a multi-variate Cox proportional hazard model that included age, KPS, and CITV (HR for BRAF mutated melanoma = 0.72, p < 0.001).  Addition of BRAF mutation status to the CITV-ds-GPA model for melanoma significantly improved its prognostic value, with NRI > 0 of 0.294 (p=0.01) and IDI of 0.017 (p=0.02). We validated these the prognostic utility of this model in an independent cohort of 174 melanoma patients.  

Conclusions:  Optimal survival prognostication for SRS-treated patients with melanoma BM requires an integrated assessment of age, KPS, CITV, and BRAF mutation status.

Objective:

Trigeminal Schwannomas are second most common intracranial Schwannomas. They have been traditionally treated by microsurgery which is associated with significant morbidity, and complete excision is challenging. Gamma knife radiosurgery (GKRS) is a minimally invasive alternative. This study evaluates the radiological and clinical outcome in a series of Trigeminal schwannomas’ patients treated with Gamma Knife radiosurgery.

Material and Methods:

Thirty patients were treated with Leksell Gamma Knife between May 2008 till  Dec.2018. Mean age at treatment was 43.4 yrs (Range 21-65Yrs).GKRS was used as initial treatment in 24 patients (80%) after initial subtotal resection in 4 patients(20%).The tumor volume ranged from 0.5 cc to19.3 cc (Men, 3.8cc).Mean prescription  dose was 13.2 Gy at 50 % isodose line (range. 12.5Gy to 14 Gy).

Results:

Average follow up was 48.4months ( range 6-124 months).Tumor size remained static in 16 (53.3%) and showed radiological evidence of shrinkage in 12 (40%).Tumor progression occurred in 2 (6%) patients. No patient had worsening of pre-existing neurologic symptoms or development of new cranial nerves deficits at the last follow up.

Conclusion:

GKRS is a safe and effective treatment alternative for patients with Trigeminal nerve schwannoma. There is not only long term tumor control but also functional preservation.

Cranial nerve schwannomas are radiosensitive tumors that are commonly managed by stereotactic radiosurgery (SRS). There is a large body of literature supporting the use of SRS for vestibular and trigeminal schwannomas. Schwannomas affecting the oculomotor nerves (cranial nerves III, IV and VI) are rare tumors. They are skull base tumors in close proximity the brainstem and often involving the cavernous sinus, for which resection can cause significant morbidity. As for other schwannomas, SRS can be used to manage these tumors, but only a handful of cases have been published so far, often among other uncommon schwannoma location reports. 

The goal of this study was to collect retrospective multicenter data on tumor control, clinical evolution and morbidity after SRS. This study was performed through the International Radiosurgery Research Foundation (IRRF). To be included, patients had to be treated with single fraction SRS for an oculomotor nerve schwannoma. The diagnosis was based on either diplopia or ptosis as the main presenting symptom as well as anatomic location on the trajectory of the presumed cranial nerve of origin, or prior surgical resection confirming diagnosis.

7 institutions submitted data for a total of 25 patients. There were 11 CN III schwannomas, 11 CN IV schwannomas and 3 CN VI schwannomas. Data analysis is ongoing, and further results will be available at the meeting.

Objective  This study was performed to analyze the long-term outcome of Gamma Knife radiosurgery (GKRS) in a series of 28 patients with cavernous hemangioma of the orbital apex. Methods  Twenty-eight patients with cavernous hemangioma of the orbital apex were treated with GKRS between March 2005 and June 2014. The series included 11 male and 17 female patients with an average age of 40.5 years (range 22–65 years). The diagnoses were confirmed by histology in 1 cases and presumed in accordance with clinical and radiological findings in 27 cases. The mean volume of the lesion at GKRS was 1.9±1.1cm³ (range 0.2~8.9cm³). The prescription peripheral dose ranged from 10.0 to 14.0 Gy. All patients had no history of radiation therapy. Results The median duration of follow-up was 52.5 months (range 24–120 months).Periodically scheduled MRI/CT and clinical follow-up showed evidence of tumor shrinkage in 26 patients (92.9%). Visual acuity (VA) was preserved in all cases. Thirteen patients (46.4%) experienced vision improvement of varying degrees, and VA was stable in 12 cases (42.9%) . Deterioration in VA was observed in only 3 cases (10.7%), including 2 patients had transient visual impairment within two weeks after GKRS. Exophthalmos disappeared on clinical ophthalmic examination in 14 cases. Only 4 cases (14.3%) had a transient chemosis. No recurrence was found during the follow-up examinations.  Conclusions This retrospective investigation indicates that GKRS provides an long-term effective management strategy in patients with cavernous hemangioma of the orbital apex, with a high rate of visual function preservation.

Objective: Cavernous sinus hemangiomas (CSHs) are rare vascular tumors. This study aimed to clarify the 10 year  outcomes of multi-fraction stereotactic radiosurgery  for the treatment of large CSHs（10 cm3 <tumor volume</=40 cm3）.  Methods: Between January 2008 and January 2018,  Eighty-eight patients with large CSHs （10 cm3 < tumor volume</=40cm3） underwent multi-fraction Cyberknife radiosurgery . Eighty-five(96%) patients underwent multi-fraction stereotactic radiosurgery as the primary management for their CSHs based on clinical and imaging criteria, and the other three patients had previous operation before multi-fraction stereotactic radiosurgery. The median volume of the CSHs was 23.9  cm3 (range, 10.3-40.0 cm3). Multi-fraction stereotactic radiosurgery was delivered in 3 fractions. The median marginal dose was 20.3 Gy (range, 19.5–21 Gy) prescribed to a median 64% isodose line. Results:  The median follow-up period was 56 months (range, 12–123 months). Tumor control was achieved in all patients (100%) during the follow-up period. At 12 months after cyberknife radiosurgery, MRI revealed a mean of 80% tumor volume reduction (range, 60%-99%). The last MRI showed a mean of 90% tumor volume reduction. Sixty-six( 75%)patients who had cranial neuropathies before radiosurgery demonstrated improvements in their neurological deficits （improvement of vision, facial numbness）, 16(18%) patients initially asymptomatic kept the same clinical status, 4 patients developed mild facial numbness.  One patient reported a stroke 3 years post radiosurgery because of hypertension. One elder patient felt reduction of memory post radiosurgery. No patient had visual function deterioration, and other adverse radiation effects during the follow-up period. Conclusion: Our experience confirms that multi-fraction stereotactic radiosurgery is a safe and an effective management strategy for large CSHs. Considering the risk involved in microsurgery, multi-fraction SRS may serve as the primary treatment option for patients with large CSHs.

OBJECTIVE  Jugular foramen schwannomas (JFSs) are rare lesions and controversy regarding their management still exists. Complete resection is possible but may be associated with significant morbidity. Stereotactic radiosurgery (SRS) is a minimally invasive alternative or adjunct to microsurgery. The authors reviewed clinical and imaging outcomes of Cyberknife SRS for patients with these tumors.

METHODS   Fifty-nine patients with JFSs underwent hypofractionated Cyberknife radiosurgery between January 2008 and January 2015. Thirteen patients had previous microsurgical resection, one patient had recurrent tumor post Gamma Knife radiosurgery, the rest 45 patients underwent Cyberknife radiosurgery based on their neuroimaging and clinical manifestations. Fifty-four patients had preexisting cranial nerve (CN) symptoms and signs. The median tumor volume was 15.1 cm3 (range 2.6-36.0 cm3), and 39 of them was larger than 10cm3 in volume. The radiation dose prescribed to the tumor margin and the number of fractions depend on the tumor volume. Twelve patients with large tumors were treated in 4 fractions, 31 patients were treated in three fractions and 16 patients in two fractions. The median margin dose was 19.2 Gy/2 Fx ( fractions),  21.1Gy/3Fx  and 24.5Gy/4Fx. Patients with neurofibromatosis were excluded from this study.

RESULTS  The median follow-up was 58 months (range 24-105 months). Tumors regressed in 41(69%) patients, remained stable in 14 and progressed in 4. The progression-free survival (PFS) was 93% at 5 years, Preexisting cranial neuropathies improved in 32 patients, remained stable in 10 patients, and worsened in 17 patients.  Four patients underwent resection at a median of 14 months after Cyberknife SRS (range 8-30 months).

CONCLUSIONS  Cyberknife radiosurgery proved to be a safe and effective primary or adjuvant management approach for JFSs. Long-term tumor control rates and stability or improvement in CN function were confirmed.

Stereotactic radiosurgery (SRS) is a viable treatment modality for spine metastases. SRS is increasingly being used in the multimodal management of these patients. Results of post-operative stereotactic radiosurgery following separation surgery has been reported but clinical outcomes and local control for a more heterogenous surgical sample (i.e., anterior approaches, anterior column reconstruction, revision surgery after previous SRS) is lacking in the literature. We present data on clinical outcomes and local tumor control at a major cancer center following contemporary surgical approaches for spine metastases.

After IRB approval, retrospective review of patients between 2012 and 2017. Demographic information, tumor histology, survival rates, recurrence rates, clinical outcomes and complications were recorded and analyzed

The mean age of patients in our cohort was 64 years (range 44 to 85 years), with 32% female.The radiation dose was 18 Gy in 1 fraction using 6 MV photons with a 24Gy boost to the gross tumor volume.The follow up period range was 3-84 months, with average time between surgery (11.1% anterior approach, 88.9% posterior approach) and SRS 3 weeks. Radiographic evaluation following SRS was every 3 months after treatment with CT or MRI. The 1- and 2-year survival rates were 57% and 38% respectively. The overall rate of local recurrence was 12.7% within the follow up period. Multivariate analysis revealed tumor location (thoracic) and histology (lung carcinoma, colon adenocarcinoma, or melanoma) as significant prognostic factors for local control and overall survival. The overall surgical and medical complication rates were 14.3% and 19.0% respectively.  The most common complication after SRS was an acute pain flare. The rate of hardware failure was 6.3%, with 3 patients developing procedure-related neurological deficits, but there were no cases of radiation myelopathy. Eighteen patients required additional surgery for metastatic disease at adjacent or distant spinal levels. There were no differences in local control if a patient had anterior column reconstruction or not.

SRS is an effective treatment modality following all types of surgery for metastases, not just separation surgery. SRS should be considered in the post-operative management for spinal metastases given the low complications, and local control roughly 87% irrespective of histology.

We developed a new interactive inverse planning approach based on a fully convex framework. The system has been first implemented to be used with Gamma Knife (GK) radiosurgery.

The convex framework is based on the precomputation of a dictionary composed of the individual dose distributions of all possible shots, considering all their possible locations, sizes, and shapes inside the target volume. The convex problem is solved to determine the plan, i.e., which shots and with which weights, that will actually be used, considering a sparsity constraint on the shots to fulfill the constraints while minimizing the beam-on time. The system is called IntuitivePlan and allows data from the generated dose plans to be transferred into the GK treatment planning software for further dosimetry evaluation and treatment.

The system has been very efficiently implemented and an optimal plan is usually obtained in less than 1 minute, depending on the complexity of the problem, on a desktop computer or in only a few minutes on a high-end laptop. Dosimetry data from clinical cases were generated with IntuitivePlan. The dosimetry characteristics are very satisfactory and adequate in terms of conformity, selectivity, gradient, protection of organs at risk, and yield to much shorter beam-on (treatment) time. Moreover the system allows to interact and move intuitively, in a very user-friendly manner, the isodose lines and modify them directly, to relax or increase constraints, and automaically generate a new optimal plan.

The possibility of using optimal interactive real-time inverse planning in conjunction with GK opens new perspectives in radiosurgery, especially considering the potential use of the full capabilities of its latest generations. This approach gives new users the possibility of using the system for easier and quicker access to good-quality plans with a shorter technical training period and opens avenues for new planning strategies for expert users. The use of a convex optimization approach allows an optimal plan to be provided in a very short processing time. The versatility of the system will allow to develop similar approaches for other radiosurgery/radiotherapy plateforms. Independent quantitative prospective evaluation comparing inverse planned and expert planned cases are underway to validate this novel and promising treatment planning approach. 

Purpose: Historically, Leksell Gamma Knife® (LGK) procedures have been regarded as single fraction exposures with no account taken as to the potential influence of the repair of sub-lethal radiation damage during treatment. However, given the large variations in treatment time and dose-rate during treatment delivery, this basic assumption has to be re-evaluated. This study aims at quantifying two possible sources of variations: unscheduled gaps during treatment and the order in which the individual shots are delivered.

Methods: An established biologically effective dose (BED) model, which includes bi-exponential repair parameters, has been used to analyse LGK treatments in a cohort of 15 patients with vestibular schwannomas. These patients were treated using a prescription dose of either 12 or 13 Gy with the LGK model Perfexion® (PFX). For comparisons, the BED values were evaluated for the physical prescription dose iso-surface.

A single gap of 15 minutes was introduced at different times throughout the course of treatment to assess the impact on BED. After evaluating all possible combinations of shot sequences for a subset of the cohort, a heuristic approach was taken to estimate the achievable range of BED values when determining the order of delivery of the individual iso-centres. This is achieved by either grouping or distributing shots according to their average dose-rate on the prescription dose iso-surface.

Results: In regard to the temporal position of the gap in the treatment delivery, mean BED values were shown to decrease between 0.1% and 9.9%, depending on the individual average dose-rate profiles. Depending on the number of iso-centres used for the treatment, the application of a heuristic approach to sorting the shots results in BED variations of up to 14.2%, relative to the mean BED for the original sequence.

Conclusions: The influence of treatment variables, like dose-rate profiles and shot sequence, on the BED should be considered during treatment planning for LGK radiosurgery. Further studies on the predictive value of BED in treatment outcome should be carried out.

Purpose: Historically, Leksell Gamma Knife® (LGK) procedures were regarded as single fraction exposures with no account taken as to the potential influence of the repair of sub-lethal radiation damage during treatment. However, given the large variations in treatment time and dose-rate distributions, this basic assumption needs to be re-evaluated. This study aims at demonstrating the feasibility of efficiently calculating biologically effective dose (BED) distributions to aid with the treatment planning process.

Methods: An established biologically effective dose (BED) model [1], which includes bi-exponential repair parameters, has been implemented in MATLAB [2]. A custom version of the treatment planning system (TPS) Leksell GammaPlan® 10.1 (LGP) was used to export the physical dose distribution matrices and the corresponding beam-on times for every iso-centre. Using the parameters for the BED model (alpha/beta ratio, repair rates, partition coefficient) allows for the calculation of 3D BED distribution matrices from the 4D dose distribution matrices (combination of one 3D matrix per iso-centre).

Results: The processing time in MATLAB for a single 31×31×31 BED matrix calculation, as it is used in LGP for the physical dose, does not exceed 0.05 s on a regular laptop (Intel® Core™ i7-7700HQ, 16 GB RAM). This was measured using the MATLAB built-in function timeit() for a 20-iso-centre treatment plan. The resulting BED distributions can be visualised side-by-side with the corresponding physical dose distribution.

Conclusions: Calculating the BED distribution for a given physical dose treatment plan could be implemented into the TPS to better estimate the effectiveness of a treatment and its dependence on parameters like the treatment time, the number, location and shape of iso-centres and their order of delivery. Since the additional computational cost is limited, the BED could be updated and visualised in real time, similar to the physical dose visualisation that is used in LGP now.

References:

[1] W. T. Millar et al., ‘The role of the concept of biologically effective dose (BED) in treatment planning in radiosurgery.’, Phys. Medica PM Int. J. Devoted Appl. Phys. Med. Biol. Off. J. Ital. Assoc. Biomed. Phys. AIFB, vol. 31, no. 6, pp. 627–33, Sep. 2015, http://dx.doi.org/10.1016/j.ejmp.2015.04.008

[2] BED model on GitHub: https://github.com/klinge-th/modelBED

Purpose:  The Leksell Gamma Knife^® radiosurgery platform has excellent dose characteristics that admits the delivery of high quality plans. To utilize its full potential, a new inverse planning approach has been developed that both resolves the shortcomings of earlier approaches and unlocks new capabilities.

Methods: We present an inverse planning approach, where the isocenter positions are determined and fixed prior to the sector-duration optimization step, which is done using linear programming. In sector duration optimization irradiation times for each sector and collimator at each isocenter are the optimization variables. In the framework of  linear programming, we describe two methods for problem size reduction: dualization and representative subsampling. Furthermore, we propose an efficient penalization of beam-on-time promoting the total irradiation time of each sector to be equally long, which is advantageous since they can irradiate simultaneously.  We can constrain dose to organs at risk and we study the effect of beam-on time penalization on the trade-off between plan quality and beam-on time.

Results:  Compared to a naïve beam-on time penalization, we find that our efficient beam-on time penalization reduces the beam-on time by a factor 2-3. Both explicit dualization and representative subsampling lead to optimization time-savings by a factor 5-20. Overall, in a comparison on 75 clinical plans we find that it is always possible to find plans with similar coverage and better selectivity and beam-on time. In addition, in 44 of these cases, the plans have an improved gradient index. On a standard Leksell GammaPlan^® workstation, the optimization times for typical cases are less than a minute. 

Conclusion: We present a combination of techniques that enables sector-duration optimization, which renders clinically acceptable plans, in a clinically feasible time frame. 

Purpose: The major importance of the effects, related to the repair of sublethal radiation damage - as treatment duration varies, are a current controversy in radiosurgery. Cell survival studies have been performed to verify the importance of this effect in relation to established models. 

Methods: Mammalian V79-4 cells were irradiated in vitro with γ-rays, either as an acute exposure, where the effects of sublethal irradiation damage repair can be ignored, or as protracted exposures over 15 – 120 min.  Protraction was achieved either by introducing a variable time gap between two doses of 7 Gy, or as a continuous exposure at lower dose rates so that a range of doses were delivered in fixed times of 30, 60 or 120 min. 

Results: For all doses there was a progressive reduction in efficacy with increasing overall treatment time. This was illustrated by the progressive increase in clonogenic cell survival with the progressive increase in exposure time, with a resulting shift of the cell survival curves.  Cell survival curves for irradiations given either as an acute exposure (6.1 Gy/min), over fixed times (30, 60 and 120 min) were well fitted by the LQ model, giving an α/β ratio of  4.0 Gy and because of the limited data set a single repair half-time of 31.5 min.

Conclusions: The present results are consistent with published data with respect to the response of solid tumors and normal tissues, whose response to both continuous and fractionated irradiation is also well described by the LQ model. This suggests the need for dose compensation in radio-surgical treatments, where dose is delivered over a similar range of protracted overall treatment times, perhaps as a prerequisite to full Biological Effective Dose treatment planning.

Objectives: Cyberknife treatments with Small and Irregular MLC-formed fields are widely used in SRS and SBRT. The current technique is to perform film or ion chamber measurements to confirm the dose accuracy, which is time-consuming and prone to errors. The purpose of this study is to apply state-of-the-art machine learning techniques for accurate dose verification of MLC-based SRS and SBRT treatments.

Methods: The commissioning and clinical measurement data were collected from our institution and the datasets were randomly split into training and testing data. The measured dose distributions were treated as outcomes of a deep neuro network (DNN) based estimator with inputs of different fields with detailed MLC positions. Dose results for different MLC fields were predicted using models trained with regularization added to the cost functions. The predicted dose distributions for small and irregular fields were evaluated using percentage relative error regarding measured data at the depth of 1.5cm and 5cm.

Results: The packages we used were Tensorflow and scikit-learn in python. With augmentation techniques, datasets of field sizes ranging from 7mm x 7mm up to 115mm x 100mm were tested for model training and dose output prediction. The dose of small and irregular SRS treatment field was accurately predicted with the proposed machine learning methods.  The mean relative error between the predicted and the measured dose is 0.11% with a maximum error of 0.4%.

Conclusions: The proposed method could potentially be used for dose reconstruction with recorded MLC and robot positions during the delivery which will shorten the patient QA time and enhance treatment efficacy. 

Target delineation is an important step in radiosurgery (RS)  treatment planning. Routinely the targets are delineated through slice-by-slice manual segmentation on MR images. This process is time-consuming, operator - dependent and could lead to treatment delays. The aim of this study was to investigate the speed up of the tumor delineation within the radiosurgery treatment planning using contours generated by a deep convolutional neural network (CNN).

The MR images of ten patients treated with Gamma Knife RS  were selected from routine clinical practice. The dataset consisted of four cases of meningioma, two cases of vestibular schwannoma and four cases of multiple brain metastases. We compared the times needed for two contouring techniques: manual delineation of the tumors and a user adjustment of the CNN generated contours of the tumors. The time spent on each task was recorded. The tasks were performed in Leksell Gamma Plan (version 11.1, Elekta AB) and iPlan (version 4.5, BrainLab) by four experts. The 3D - Unet architecture with residual connections, trained with custom loss function and sampling procedure [Krivov et al, 2018], optimized for metastases segmentation was used for automatic brain tumor segmentation. The automatic contours were generated within five seconds.  The time required to import these contours to the treatment planning systems was less than one minute.

The generated contours were acceptable with no or minor corrections.The total median time needed to delineate a tumor manually was 9.15 min. (ranged from 3.15 min. to 29.18 min). The median times saved were 6.54 min. (range 40 sec. - 17.06 min.), 2.16 min. (range 48 sec.- 8.20 min.), 9 min. (range 1 min. -  26 min.), 5.27 min. (range 3 min - 17.35 min) for User 1, User 2, User 3, User 4 respectively. The Wilcoxon signed-rank test was used to compare results (p < 0.05, r > 0.6). On average, the automatic algorithm speeds up the process of the delineation in 2.30 times.

The usage of deep learning generated contours accelerates delineation more than twofold. Though the automatically generated contours were almost identical to the manual ones,  further investigation is needed to quantify these differences and compare it with inter-rater reliability.

1.   Krivov E. et al. Tumor Delineation For Brain RS by a ConvNet and Non-Uniform Patch Generation // 3rd Int. Workshop Patch - MI . 2018. 8p.

Objective: A major challenge in the follow-up of patients managed with stereotactic radiosurgery (SRS) for brain metastases (BM) is to differentiate pseudoprogression (PP) from tumor recurrence (TR). A clinical score based on tumor and treatment related factors would be valuable when selecting appropriate treatment.

Material and methods: Follow-up images of 97 consecutive patients treated with SRS for 406 BM were analyzed. Of these 100 (24.6 %) BM in 48 (49.5 %) patients responded either with TR (delayed growth; 53 (13.1 %) BM) or PP (temporary volume increase; 47 (11.5 %) BM). Differences between the 2 groups were analyzed and used to develop a PP risk assessment score (PP-RAS).

Results: Significant factors associated with a higher incidence of PP versus TR were: prior radiation SRS or WBRT (p = 0.001, π = 13.7), target cover ratio > 98 % (p = 0.031, π = 4.7), BM volume ≤ 2cc (or BM ≤ 1.5 cm in diameter, p = 0.039, π = 4.3), and primary lung cancer vs. other primaries (p = 0.084, π = 3.0). Based on the presence (0) or not (1) of these 5 parameters, a risk assessment score for PP versus TR was established. A PP-RAS score of 0 corresponds with high likelihood of PP vs. TR, whereas a score of 5 corresponds with a high risk of TR.

A score of ≤ 1 point was associated with 100 % PP, 2 points with 57 % PP  and 43 % TR, 3 points with 57 % TR and 43 % PP, whereas ≥ 4 points were associated with 84 % TR and 16 % PP , π=24.6, df =4, p < 0.001).

Conclusion: Based on 5 readily available parameters at the time of SRS our risk assessment score could robustly differentiate between PP versus growth following SRS. The score is user-friendly, intuitive and cost-free. It may be a useful tool to guide the decision making whether to retreat or observe at appropriate follow-up intervals.

The aim of this study was to investigate the clinical relevance of hypofractionated stereotactic radiosurgery (SRS) specifically in treating medium-sized brain metastases (BMs) of 2.5 to 3 cm compared with single-fraction SRS.

Between 2011 and 2015, a total of 100 patients with newly diagnosed BMs (n=105) of 2.5 to 3 cm had been treated with either single-fraction SRS using the Gamma Knife (GK; n=67; median 58 years) or hypofractionated SRS using the CyberKnife (CK; n=38; median 64 years) at our institution. Primary cancers originated from the lung (n=56, 53.3%), the breast (n=22, 21.0%), the gastrointestinal tract (n=14, 13.3%), and others (n=13, 12.4%). A median marginal dose 21 Gy (range, 18-23 Gy) was delivered for single-fraction GK and a median cumulative dose 35 Gy (range, 27-41 Gy) was delivered in median 5 daily fractions (range, 3-5 fractions) for hypofractionated CK. None of the patients received any prior or upfront whole brain radiotherapy. In each patient, treatment outcome was measured by local tumor control (LTC), overall and progression-free survival (OS and PFS), and the occurrence of radionecrosis (RN).

With a median follow-up of 14 months (3-59 months), significant differences were observed in the incidence of RN (29.9% versus 5.3%, P=0.004) and LTC (LTC rates at 1 year 66.6% versus 92.4%, P=0.035) between the single-fraction versus hypofractionated SRS groups. There were no differences in PFS (median 6 months versus 6 months, P=0.368) and OS (median 13 months versus 18 months, P=0.234) between the groups. Treatment related adverse events (≥ grade 2 toxicity by CTCAE ver 4.0) were more frequently occurred in single-fraction group than hypofractionated group, but the difference was not statistically significant (56.3% versus 36.1%, P=0.084).

These findings suggest a better safety and efficacy profile of hypofractionated SRS compared with single-fraction SRS for the treatment of medium-sized BMs. Further prospective studies are needed to address definitive conclusions.

Purpose: To evaluate the efficacy and safety of multi-fraction stereotactic radiosurgery(MF-SRS) in patients with giant oligometastatic brain metastases.

Methods: All patients with giant (defined as ≥20 cm³) brain oligometastases who had been referred for MF-SRS at the Huashan Hospital between July 2009 until December 2016 were identified. The radiosurgical dose, isodose curve and fraction selection were based on various factors including tumor volume, tumor characteristic, location and original pathology. The BED of MF-SRS was higher than 40 Gy, corresponding to a single dose of about 16 Gy. All patients had routine clinical and radiologic follow-up at 30 days next to the last fraction and then 3-month intervals. Overall survival was evaluated using the Kaplan-Meier method. Multivariate analyses were performed with the multivariate Cox proportional hazard model.

Results: Between July 2009 and December 2016, 27 patients with 29 intracranial giant metastases were included in this study.The radiosurgical margin dose (median 30 Gy, range 26.1–36 Gy) was prescribed at an isodose curve of 64%–75% (median 68%) with multiple fractions (median 3 fractions, range 3-5), targeting a median tumor volume of 35.51 cm3 (range, 20.7-98.1 cm3). The overall median survival time (MST) was 20 months. The 1-, 2-, and 5-year survival rates were 66.7%, 43.7%, and 23.6%, respectively.Local tumor control rates were 85.7%, 66.9% and 50.2% at 1-, 2-, and 3-year, respectively.On multivariate analysis extracranial metastasis (p = 0.021, HR: 3.76, 95% CI:1.22–11.56), was confirmed as associated with worsened overall survival.The median KPS improved significantly from 40 to 70 (p<0.05, paired t test).No patient obtained a worsened KPS.

Conlusion: MF-SRS is a safe and effective option for patients with giant brain oligometastases and poor performance status.  Prospective studies are required to confirm the findings in this study.

Purpose/Objectives: Brain metastases from germ cell tumors (GCTs) are rare and mainly occur in young men whose functional status is otherwise unimpaired. Standard of care in GCT patients with brain metastases remains WBRT, but radiation-induced neurocognitive deficits have been reported in patients with germ cell metastases to the brain treated with whole brain radiation (WBRT). To the authors’ knowledge, there are no published reports evaluating the efficacy of stereotactic radiosurgery (SRS) as the primary treatment of GCT brain metastases. We hypothesize that SRS alone is a feasible alternative for the treatment of limited GCT brain metastases.

Materials/Methods: The records of 14 male patients with various GCT histologies treated between 2012 and 2017 at a single institution were retrospectively reviewed. All of the patients received gammaknife radiosurgery without WBRT. Two of the patients underwent SRS to a post-surgical resection cavity.

Results: The median follow-up of the cohort was 24 months (range 1-50 months) with a median age of 29 years (range, 17-56) and a median of 1.5 (range, 1-7) lesions treated. The 1-year OS was 71% with a median survival of 23 months. There were no local recurrences among the SRS-treated lesions resulting in a 1-year LC of 100%. Only one neurologically related death occurred, leading to a neuro-specific mortality rate of 7%. A total of 35% of patients experienced distant brain recurrence at a median time of 2.7 months, which resulted in a 1-year distant brain control of 71%. Of the 5 patients with distant failure, 3 were successfully salvaged with repeat SRS treatment without additional relapse on follow-up imaging. Only one patient received salvage WBRT 19 months post initial SRS, leading to a 2-year WBRT free survival of 93%.

Conclusions: Given that GCT brain metastases are often limited in number, and mainly affect young men with good functional status, avoidance of WBRT may provide control of intracranial disease with the goal of preserving neurocognitive function in these young patients. Our preliminary results suggest that SRS may safely replace WBRT as an initial treatment of choice patients with GCT brain metastases.

Objectives: The aim of this study is to evaluate the outcome after 2 fraction dose stage gammaknife radiosurgery (GKS) for large brain metastases.

Methods: A total of 53 large brain metastases in 42 patients were treated by 2 fraction dose stage GKS. The inclusion criterion was “large” brainstem located lesions or large lesions with a 12 Gy isodose volume of normal brain parenchyma exceeding 10 cm3. The mean tumor volume for the first ones was 4.4 cm3, for the latter ones was 12.7 cm3. For both fractions, the prescription margin dose was 12Gy on the 50% isodose line, with 2 weeks between them. Local control failure was define as an increase of more than 20% between initial and last tumor volume. The primary cancer was melanoma (8), pulmonary (19), breast (8), kidney (2), other (5). The mean age was 62 years old (31-88). The median Karnofsky score was 90. Nine patients had GPA (0-1), 20 patients had GPA (1.5-2), 7 patients had GPA (2.5-3), 6 patients had GPA (3.5-4).

Results: At the second fraction, mean tumor volume was 8.2 cm3. The mean percentage of volume variation for decreasing lesions was 29%. Only 3 lesions increased their volume. At last follow-up, mean tumor volume was 6.0 cm3; 46 lesions decreased volume with a mean percentage of volume variation of 68%; only 7 lesions increased volume. There was a significant correlation (p=0.05) regarding volume variation between 1rst and 2^nd GKS and between 1rst GKS and last follow-up. Rate local control at 6 months and 1 year was 87.3% and 68.3% respectively. The rate of adverse radiation effect (ARE) was 14.3%. No predictive factor of local control or ARE was found in an univariate analysis.

Conclusion: The new 2-fraction-dose-staged GKS concept seems to be a well-tolerated and effective treatment option for large BMs.

Introduction: Brain metastases (BMs) from solid tumors represent a topic of increasing interest for the higher incidence in the last years. Stereotactic radiosurgery (SRS) is the main effective local therapeutic approach used. We draw a phase III trial comparing Gammaknife (GK) and Linac based (Edge) SRS.

Materials and Methods: Patients with a maximum of 4 BMs up to ≤30 mm were included. Randomization was stratified according to age, presence of extracranial metastases, and number of BMs. For Arm A (GK) single dose of 20-24 Gy at 50% isodose was prescribed. For Arm B single dose of 24 Gy was prescribed to PTV. Radionecrosis was assessed mismatching T1/T2 MRI images, perfusion MRI, and in doubt cases Methionine-CT/PET.

Results: From October 2014 to September 2018, 202 patients for 354 BMs treated were included, 96 in Arm A and 106 in Arm B, for 182 and 172 metastases, respectively.  RN occurred in 28 (7.8%) cases, 12 in Arm A and 16 in Arm B; grade II in 9 cases of GK arm at a median time of 7 months and in 15 cases of Edge arm at a median time of 9 months; grade III RN was recorded in 4 cases, in 3 of GK arm, at a median time of 3 months, and in 1 of Edge arm at 37.5 months. The 12, and 18  months local control (LC) rates were 98.8% and 90.9% for arm  and 96.2% and 96.2% for Arm B (p=0.96).  The median, 12, and 18 months OS rates were 17.8  months, 74.1%, and 48.9%. The volume of BMs was impacting on radionecrosis occurrence (p value=0.005; p value=0.03).

Conclusions: Gamma-knife and LINAC based SRS for BMs were comparable in terms of LC. The occurrence of GIII radionecrosis was greater and earlier in the GK arm respect to Edge arm.

ABSTRACT

Introduction

Gamma Knife Radiosurgery (GKRS) as monotherapy is often avoided for the treatment of large (≥ 2cm) posterior fossa brain metastasis (LPFM) based on a theoretical risk of increased peritumoral edema (PTE) and associated compression of the 4^th ventricle (4V) related to the treatment.

Methods

A single center, IRB approved, retrospective review of LPFM treated with GKRS from 2009-2017 was performed. Brainlab® iPlan software was used to evaluate the tumor, 4V and PTE volumes at initial treatment and all subsequent follow-ups. Statistical analysis was performed using Wilcoxon Signed Rank test and McNemar test.

Results

We identified 49 lesions in 47 consecutive patients; 55% males, median age 61.5 years, median KPS was 90 at the time of GKRS. Median number of LPFM and overall brain metastases was 1 and 2.5 respectively. The median overall tumor, PTE, and 4V volumes at diagnosis were 5.37 cm³, 17.11 cm³, 1.23 cm³ respectively with a median study follow-up of 7.28 months (range 1.87-56.23). At first follow-up, 2 months post treatment, median tumor volume decreased by 54.04 % [range -96.95, 48.69] (p <0.001), median PTE decreased by 55.66 % [range -99.92, 143.36] (p <0.001) and 4V increased (23.89 [range -50.16, 545.76] p 0.31). No patient required surgical intervention, external ventricular drainage or shunting between treatment and first follow-up. Post treatment, 65.95 % receive our routine steroid taper, 4.25% received no steroids and 29.78% required prolonged steroid treatment.  

Conclusions

Patients with LPFM treated with GKRS had in a significant reduction in tumor size and PTE (p <0.001) and marked opening of the 4V post treatment. This study suggests that GKRS was well tolerated and can be considered in the management of LPFM especially in patients for whom surgery in contraindicated.  

Purpose: To compare the incidence of pseudoprogression (PP), local tumor control (LC) and overall survival (OS) in patients treated with stereotactic radiotherapy (SRT) for a solitary brain metastasis using a 2mm Planning Target Volume (PTV) margin versus a 0mm margin. The development of pseudoprogression and the volume of the brain receiving 12Gy (V12) was assessed for single fraction treatment while V18 was assessed for treatment with 3 fractions.

Patients and methods: Patients were treated on the Novalis LINAC and had a minimal follow-up of 24 months. Dose was prescribed according to the PTV-volume: 1x21Gy, 1x18Gy or 3x8.5Gy. A 2mm CTV-PTV margin was used in 93 patients while a 0mm margin was used from 2015 onwards in 37 patients. Follow-up included a 3-monthly MRI-scan including a perfusion MRI-scan when PP was suspected. The V10-V18Gy of the brain was calculated.

Results: The mean CTV-volume was similar (p=0.3) while the mean PTV-volume was significantly larger in the 2mm group (15.2cc versus 7.6cc P<0.01). There was no significant difference in the incidence of PP between the 2mm and 0mm group at 1 year (31% versus 33%) and at 2 years (43% and 33% P=0.6). Symptomatic PP was significantly more frequent in the 0mm arm (4.1% vs 1.6%, P=0.02). The 2-year LC rate was similar in the 2mm and the 0mm group (78% and 82%, P=0.4) while the 2-year OS was significant better in the 0mm group (29% versus 13%, P=0.01). The V10-V18 of the brain was not predictive for PP.

Conclusion: PTV margin reduction from 2mm to 0mm did not reduce the incidence of PP in linac-SRT for single brain metastases. LC and OS rates were similar, indicating margin reduction is safe. V10-V18Gy were not associated for the incidence of pseudoprogression.

Objective: The number of brain metastases (BMs) plays an important role in the decision making between stereotactic radiosurgery (SRS) and whole-brain radiation therapy
Methods: We analyzed the survival of 457 SRS-treated patients with BM as a function of BM number. 113 (24.7%) patients were treated with repeat SRS for local recurrence and distant brain relapse. Survival analyses were performed with Kaplan-Meier analysis as well as univariate and multivariate Cox proportional hazards models.
Results: The median survival for all patients was 9.8 months (95% CI 8.5-11.3). Patients with BMs were categorized as those with 1, 2-4, 5-9 and ≥10 BMs. Median overall survival for patients with 1 BM was superior to those with 2-4, 5-10 and >10 BMs (14.3 months vs. 10.5 months vs. 6.6 months vs. 7.2 months, р= 0,0001). Survival of patients with 5-9 BMs did not differ from those with >10 BMs (6.6 months vs. 7.2 months, p=0.8288).
Аccording to multivariate analysis, age younger than 50 years (p = 0.005, HR 0.59, 95% CI 0.41 - 0.86), number of lesions ≥5 (p = 0.02, HR 1,44, 95% CI 1,05 - 1.95), Karnofsky performance status ≥80 (p < 0.0001, HR 0.52, 95% CI 0.38 - 0.69) аnd absence of extracranial metastases (p < 0.01, HR 0.56, 95% CI 0.35 - 0.90) had significant impacts on overall survival.
Conclusions: The contribution of BM number to overall survival is modest and should be considered as one of the many variables considered in the decision between SRS and whole-brain radiation therapy. Our finding confirm correlation between the survival outcome and the number of lesions treated SRS. According to our data, the median survival of patients with 2-4 metastases in the brain is significantly higher compared with a group of patients who have 5-10 or more lesions.

Introduction: Local control of brain metastases reduces risk of death from intracranial progression and improves survival, which can be extended by 6 months with whole brain radiation therapy (WBRT), but this treatment has a greater risk of a significant neurological decline compared with patients who receive more focal treatments, like stereotactic radiosurgery(SRS). The widespread of the volumetric modulated arc technique therapy (VMAT) which can deliver a fraction of WBRT and a simultaneous integrated boost (WBIB) to multiple brain metastases in a few minutes suspected possible advantages of local and regional control for this treatment and the studies are controversial about the real impact in neurocognition

Methods: Patients with confirmed one to six brain metastases submitted to either WBIB or SRS alone using VMAT were selected. WBIB group patients received WBRT in daily 4 Gy to a total of 20 Gy over one week and the boost reached the total dose of 40Gy with daily 8 Gy in each evident metastasis. For the SRS alone group, doses varied according to the metastasis sizes, based on the RTOG. We retrospectively accessed patient clinical and radiological data until death registration.

Results: 31 patients were selected, 20 in SRS group and 11 in the WBIB group, mean age of 63 years old. Majority of patients have secondary lung central nervous system metastasis. The mean overall survival after the radiotherapy treatment was 11,4 months in the SRS group and 8,2 months in the WBIB group (p =0,001). There was no difference in local control after 3 and 6 months between the groups, but the was a significantly greater incidence of new leasons in the group that received only SRS compared with WBIB( 38,1 versus 27,3 %  p<0,001). Subgroup analysis of the SRS group showed that 47,6% needed to receive WB treatment after a mean time of 6,8 months. Grade I and II toxicity incidence was greater in the WBIB and there was also a greater decreasement in the KPS status of theses patients after the 6 months reassessment

Conclusions: WBIB seems to achieve a similar local control to focal SNS leasons and better regional control to subclinical leasons compared with patients who receive only SRS, despite a greater toxicity. Afterwards, WB would be needed in almost 50% of the patients who received SRS and therefore the WBIB  treatment could propitiate a best cost effectiveness choice 

Background

Post-operative stereotactic radiosurgery (SRS) is the standard of care for resected brain metastases, but SRS techniques are not standardized. Although expert consensus guidelines recommend that the surgical corridor leading to resection cavity be included in SRS plan, this statement is not evidence-based. We analyzed failures and toxicity with post-resection SRS, with the hypothesis that the corridor needs not be targeted with SRS. 

Methods

In this IRB-approved retrospective review, from 428 lesions treated from 2005-2018 with post-resection SRS, 58 evaluable lesions had a ‘deep’ tumor with a surgical corridor, defined as ≥1.0cm from surface pre-operatively. SRS targeted the surgical corridor, defined as the surgical tract uninvolved by tumor on pre-operative imaging, in 33 (57%). Failure was defined as local (LF) if within the surgical cavity involved with tumor pre-resection, corridor (CF) if within the surgical tract leading to the cavity, distant (DF) if a new parenchymal tumor, or leptomeningeal (LMD) if new nodular/classical leptomeningeal enhancement. The cumulative incidences of failure and adverse radiation effect (ARE) were analyzed with death and whole brain radiotherapy as competing risks, with 95% confidence intervals.

Results

The median follow-up was 14 months. Not targeting the surgical corridor was associated with prior SRS or resection for other brain metastases (23% vs. 0%, p=0.01), deeper tumors (median 2.1 cm vs. 1.4 cm, p<0.01), and systemic treatment within 3 months (p =0.01), but not other factors (p>0.10). The 12-month failure rates, if the surgical corridor was not treated vs. was treated, respectively, were: CF 8% (1-24%) vs. 0% (p=0.12), LF 4% (0-17%) vs. 13% (4-27%) (p=0.32), LMD 40% (19-61%) vs. 10% (2-23%) (p=0.03), DF 65% (43-81%) vs. 35% (19-52%) (p=0.02), and ARE 8% (1-22%) vs. 13% (4-28%) (p=0.35). After adjusting for use of systemic therapy, all differences were not statistically significant (p>0.05).

 Conclusion

Omitting the surgical corridor in post-operative SRS for resected brain metastases was not independently associated with statistically significant differences in recurrences or adverse radiation effect.

Stereotactic radiosurgery (SRS) is a cornerstone treatment for brain metastases (BM). Despite its notably efficacy in tumor control, adverse radiation effect (ARE) remains a dreaded complication. Here, we perform a quantitative analysis to examine the relative contribution of various clinical and dosimetric factors to ARE.

We identified 214 patients with 1,106 BM who were treated with SRS at University of California San Diego (2007-17) and 148 patients with 1,760 BM who were SRS treated at Karolinska Institutet who had  >3 months of MRI follow-up. ARE was defined by post-SRS FLAIR involving >25% of the cross-sectional area of the centrum semiovale, third ventricle, temporal horns, or the fourth ventricle. Standard statistical measures were used to identify risk factors for ARE.

In the UCSD cohort, 62 patients (29%) suffered post-SRS ARE. In univariate models, the risk of ARE increased with 1) the number of treated tumors (OR 1.061 per additional tumor, p<0.001), 2) cumulative intracranial tumor volume (CITV) (OR 1.016 per cm³, p= 0.03), 3) the number of SRS sessions (OR 1.425 per additional session, p<0.001), and 4) previous WBRT (OR 4.601, p<0.001). In a multivariate model, ARE risk was associated with the number of SRS sessions (p=0.002) and prior WBRT (p=0.02). These findings were recapitulated in the cohort of 148 Karolinska patients. In a combined multivariable model accounting for the length of follow-up, the associations with number of SRS and WBRT remained robust: number of SRS (p=0.002), history of WBRT (p< 0.001). Variance analysis indicated that a history of WBRT contributes more to the risk of ARE than the number of SRS sessions.

In this analysis of our decade-long experience, we demonstrate that the number of SRS sessions and prior WBRT independently contribute to the odds of developing ARE, and that prior WBRT is the biggest contributor to ARE.