Objectives: Debate continues as to whether stereotactic radiosurgery (SRS) or fractionated stereotactic radiotherapy (FSRT) should be chosen for patients with non-functioning pituitary adenomas (NFPAs) touching or even compressing the optic chiasm. We describe our long-term follow-up results with SRS for such patients.

Methods: This institutional review board-approved, retrospective study used our prospectively accumulated database including 27 patients (14 females, 13 males, mean age; 61 [range; 19-85] years) who underwent gamma knife SRS between 1998 and 2006 for NFPAs touching/compressing the optic chiasm. Twenty-six patients had undergone surgical removal; once in 16, twice in eight and four times in two. The median tumor volume was 5.2 (range; 1.8-50.8) cc. To avoid excess irradiation to the optic apparatus (OA), the tumor was not totally covered with the selected peripheral doses. Instead, the lower part of the tumor was covered with a 50% or a 60% isodose gradient; i.e., 49-98% (mean; 84%, median; 88%) of the entire tumor received the selected doses. Median doses at the tumor periphery and the OA were 7.6 Gy and 11.0 Gy (ranges; 3.2-10.9 Gy and 8.7-12.9 Gy), respectively.   

Results: Seven patients (26%) were confirmed to be deceased due to unrelated diseases at a median post-SRS period of 150 (range: 15-174, IQR: 83-154) months. Follow-up MRI showed tumor growth in two patients (7.4%) at the 11th and 134th post-SRS month; the former underwent surgery and the other SRS. Excluding these two patients, the latest follow-up MRI examinations which were performed 13-216 (mean: 139, median: 156, IQR: 116-172) months after SRS showed no size changes in five (18.5%) and shrinkage in 22 patients (74.1%), i.e., the crude incidence of tumor growth control was 92.6% and cumulative incidences of tumor growth-free survival estimated with a competing risk analysis were 96.3%, 96.3% and 91.7% at the 60th, 120th and 180th post-SRS month. Neither SRS-induced optic neuropathy nor endocrinological impairment occurred.

Conclusion: In patients with NFPAs touching or even compressing the optic chiasm, SRS achieves good long-term results. The opinion that FSRT is the only appropriate treatment is thus unwarranted. 

Introduction:

Non-vestibular schwannomas (NVS) are rare intracranial tumors representing <10% of all cranial nerve schwannomas. For small to medium-size tumors, radiosurgical treatment in a single session is often preferred to surgical resection. The aim of the present study is to analyze the long-term tumor control and morbidity of a series of 30 NVS treated in our center and followed prospectively.

Materiel & Methods:

The data of a series of 30 patients were analyzed. Affected cranial nerves included cranial nerve III (n=1), V (n=8), VI (n=2), VII (n=8), IX-X-XI (n=9), and XII (n=2). Median age was 47. Two patients had neurofibromatosis. A functional deficit of the cranial nerve was present at the time of irradiation for 23 patients (77%). Seven patients (23%) had undergone previous microsurgical resection. All patients were treated by a single-session procedure with a Gamma Knife C or Perfexion. The median tumor volume and margin dose were 1.1cc (range, 9.9-0.02cc) and 12Gy (range, 10-15Gy).

Results:

The median follow-up was 6.2y (range, 2-11.5y). We found a significant (i.e., >10% of the initial volume) reduction of the tumor volume during follow-up for 17 patients (57%), and a stable volume for 13 patients (43%). No patient had MR-defined tumor growth during follow-up. We observed a transient worsening of the function of the cranial nerve in the first months after radiosurgery in 2 patients with a facial nerve schwannoma, which was treated successfully with corticoids. No patient developed permanent worsening of the affected cranial nerve, and no other morbidity occurred. Five patients had an improvement of their functional deficit: diplopia resolved for 2 patients with NVS of cranial nerve III and VI, and trigeminal neuralgia resolved for 3 patients with NVS of nerve V.

Conclusions:

Based on our experience of 30 patients successfully treated for a NVS by single-session radiosurgery, we concluded that this treatment is very efficient and can be safely delivered in a single session for patients with small to medium-size NVS.

Ependymomas are rare tumors of the central nervous system developing from ependymocytes. Surgery is the reference treatment, supplemented by adjuvant radiotherapy in case of incomplete excision. At the time of the recurrence, a second surgery is preferred, but not always feasible. In these cases, the therapeutic options are limited and radiosurgery seems a good therapeutic alternative.

We present a retrospective series of 21 patients treated by radiosurgery for 33 post-surgical recurrences of ependymomas at the University hospital of Lille, France, between 2003 and 2015. The diagnosis of ependymoma was carried out according to the WHO criteria and the indication of radiosurgical treatment validated by a multidisciplinary staff. The treatments were performed with a Leksell Gamma Knife model 4C. For each patient, the data collected were age, sex, WHO grade of the tumor, number of surgeries, location of recurrence, time between surgery and radiosurgery, prescribed dose and target volume . The overall survival after radiosurgery is 100% at 1 year, 91% after 3 years. Progression-free survival was 85% at 1 year, 80% at 2 years, 64% at 3 years and 51% at 5 years. The local control of lesions treated by radiosurgery is 100% at 1 year, 85% at 2 years, 76% at 5 years and 55% at 10 years. The average duration of follow-up was 4.2 years.Among the different variables analyzed, only the WHO grade of the tumor had a significant impact on local and regional control of the disease (p 0.03 and p <0.001).

In our series, radiosurgery is a treatment of choice of ependymoma recurrences with a good rate of local control. Patients with Grade 2 ependymomas have better local and regional control of the disease.

Objective: Endocrine deficits are the most frequent complication after pituitary adenoma radiosurgery (SRS).  The clinical and dosimetric factors associated with pituitary insufficiency remain unclear despite more than 30 years of clinical usage.

Methods: Retrospective review of 97 patients having single-fraction SRS from 2007 until 2014.  Eligible patients had no history of prior radiation, normal age and gender specific pituitary function before SRS, and at least 24 months of endocrine follow-up.  Forty patients (41%) had hormone secreting tumors; 57 patients had non-secreting tumors (59%).  The median prescription isodose volume was 2.8 cm³ (range, 0.5-30.5); the median tumor margin dose was 20 Gy (range, 12-27.5).  The median follow-up after SRS was 48 months (range, 24-107).

Results: Twenty-seven patients (28%) developed pituitary insufficiency at a median of 22 months (range, 4-69) after SRS.  The 2-year and 5-year rates of new endocrine deficits were 17% and 31%, respectively.  Multivariate analysis found men (risk ratio, 2.38; P=0.04), smaller gland volume (risk ratio, 0.99; P=0.02), and increasing mean gland dose (risk ratio, 1.31; P<0.0001) to be predictive of new endocrine deficits.  Further analysis using receiver operating curves showed the least to most sensitive dosimetric variable for predicting new endocrine deficits were gland volume (AUC 0.65, P=0.01), volume of gland receiving 12 Gy (Vgland12) (AUC 0.68, P=0.02), volume of gland receiving 14 Gy (Vgland14) (AUC 0.71, P=0.01), volume of gland receiving less than 14 Gy (Vgland<14) (AUC 0.83, P<0.0001), mean gland dose (AUC 0.83, P<0.0001), and volume of gland receiving less than 12 Gy (Vgland<12) (AUC 0.85, P<0.0001).  The incidence of new deficits based on mean gland dose were <10 Gy, 1/40 (3%); 10-15 Gy, 12/35 (34%); ≥15 Gy, 14/22 (64%).  The incidence of new deficits based on Vgland<12 were >300 mm³, 1/31 (3%); 100-300 mm³, 12/49 (25%); <100 mm³, 14/17 (82%).   

Conclusion: Endocrine deficits after pituitary adenoma SRS increase in a time and dose dependent manner.  Reducing the radiation exposure to the identifiable gland whenever feasible should lower the incidence of new hormonal deficits after pituitary adenoma SRS.  The primary weakness of this analysis is the high probability of inter-observer variability in pituitary gland definition.          

Objectives: Robotic technology, image guidance based on sophisticated tracking software and small-field dosimetry issues are incorporated into the CyberKnife® Robotic Radiosurgery System (CK M6). Upon commissioning and external beam auditing (by Equal ESTRO in our case), performance evaluation procedures should be carried out for each of the CK subsystems. Standardization of a quality assurance (QA) program of CyberKnife for suitable circumstances in Chile has not been established. In this research, we investigated the development of a QA program for CyberKnife and evaluated the feasibility of its application.

Methods: Considering all the subsystems involved in producing a CK M6 treatment, a list based on bibliographical recommendations was established and divided depending on the periodicity of QA tests to be carried out. All these developed Quality Check (QC) lists were later categorized into two groups: machine QA and patient-specific QA. An IBA (myQA®) comprehensive SQL quality control management database was used to schedule every single task and tabulate its expected results and tolerances including daily QA, monthly QA and yearly QA. Purpose and outcome are described and tracked for each test. In order to verify the validity of the established QA program, this QC list was applied strictly during the past year of operation. The acceptable tolerance was based on the careful comparison of values required by the CyberKnife manufacturer and QA results in different publications. The acquired measurement results were evaluated for the analysis of the current QA status and for the verification of the propriety for the developed QA program.

Acquired x-ray images were fed into 6D target locating software to calculate patient translations and rotations. A head-neck phantom, placed at different predefined positions on a treatment couch, was used to evaluate accuracy and precision of the target locating software for the fiducial, 6D skull, Xsight™ spine tracking methods. The dosimetric characteristics of the 6MV beam were also measured. Finally, beam delivery precision and total clinical accuracy were evaluated for the fiducial, 6D skull, Xsight™ spine tracking methods.

Results: The current QA status of the CyberKnife was evaluated from the accuracy of all measurements in relation with the application of the established QA program. Each measurement result was verified to have good agreement within the acceptable tolerance limit of the developed QA program. The Target locating software was found able to define the position of the imaging objects for translations and rotations respectively with an accuracy of 0.2mm and 0.2. The results revealed sub-millimeter beam-delivery precision and dose placement total accuracy for the fiducial, skull and XsightTM spine tracking methods.

Conclusion: Performance evaluation procedures were carried out for a CK M6. The system was controlled for mechanical accuracy of the manipulator, image quality, kV parameters of the TLS as well as for linac 6MV beam characteristics and beam output parameters. The results revealed sub-millimeter beam-delivery precision and dose placement total accuracy for the fiducial, skull and XsightTM spine tracking methods. It is considered that the developed QA program in this research could establish the standardization of QC methods for CyberKnife and confirms the accuracy and stability for image-guided stereotactic radiotherapy. 

Background: Large petroclival meningiomas are extremely challenging tumors because of their intimate location with critical neurovascular structures. Microsurgical resection is often associated with significant morbidity. For this, combined surgical/radiosurgical treatment is advocated. In this study, we evaluate the outcomes of petroclival meningiomas treated with hypofractionated CyberKnife SRS as a primary treatment modality.

Methods: A series of 32 patients with a petroclival and sphenopetroclival meningiomas with a volume >8cc were treated with upfront hypofractionated radiosurgery between 2007 and 2016 at the CyberKnife Center of the University of Messina, ITALY. Patients received a dose of 25-45 Gy in 5-15 fractions. Outcome measures included progression-free survival, neurological and quality of life outcome.

Results: At mean follow up of 56 months (range 6-114), tumor volumes increased in 12% of patients, remained stable in 44%, and decreased in 44%. Kaplan-Meier actuarial progression free survival rates at 3, 5, 8, and 10 years were 100, 89, 82, and 80% respectively. No complications were recorded except from medically controlled trigeminal neuralgia that occurred in 15.6% of patients. At last clinical follow-up, 90.6% of patients demonstrated no change or improvement in their neurological condition and quality of life.

Conclusions. In our series, we used upfront frameless stereotactic radiotherapy for the most challenging tumor of the skull base. Tumor control was achieved in 87.5% of patients without impairment of neurological status or quality of life. Even though further studies are necessary, hypofractionated stereotactic radiation therapy may change the paradigm of treatment of skull base meningioma.

The treatment of paraganglioma of the jugular foramen remains controversial due to its high morbidity, given the anatomical position, the presence of major vessels and cranial nerves that are often found within the lesion. 

Although benign, these tumors sometimes present rapid growth and the neurological symptoms that accompany are hearing loss, cerebral nerves palsies and tinnitus.

For many years surgery, with or without previous embolization has been the treatment of choice, but the inability to do a complete resection, and complications have made surgeons think of alternative treatments such as radiotherapy and most recently radiosurgery. One of the major uncertainties that have always accompanied radiosurgery as treatment for jugular glomus is its effectiveness and long-term results. We present a series of 29 patients treated with radiosurgery, over 50% of patients had undergone surgery, with a mean of 73 months (> 6 years), a symptomatic tumor control 96.5%, and morbidity post irradiation of 13% (Grade I RTOG). According to our data and that found in contemporary literature, radiosurgery has replaced surgery as the primary treatment of jugular foramen paragangliomas. 

Objective: The optimal doses of radiosurgery for skull base meningioma have been reported 13 - 15 Gy. We have been intended to perform low dose radiosurgery for treating close to critical structure of skull base. We evaluate long-term efficacy of low dose (≦12 Gy) gamma knife radiosurgery (GKS).

Methods: 295 patients treated with GKS (median treatment volume: 6.9 cm³, median prescribed marginal dose: 12 Gy, range 8 -12 Gy) were followed for a mean of 104 months (range 6 - 252 months). 22 patients with large volume tumors were treated by two-staged radiosurgery (volume fractions).

Results: Local tumor growth control was 85%, Actuarial rates of progression-free survival at 5, 15, and 15 years were 90%, 82%, and 74%, respectively. 6% were malignant transformation of tumors. Univariative analysis revealed only without preGKS surgery was positive factor and age, sex tumor volume, treatment dose, staged surgery were not significant for tumor control. The clinical outcome was improved in 15% and unchanged in 64% of patients. The radiation induced neuropathy occurred 7.8% (22 patients). The trigeminal neuralgia considering most devastating symptoms was improved or disappeared in 60% (6/10patients), but new trigeminal neuralgia was occurred in 3.7%(11 patients).

Conclusion: The low dose GKS (≦12 Gy) for skull base meningiomas can achieve long-term tumor growth control with low morbidity with a long-term follow-up period of > 8.5 years. But the malignant transformation was still difficult to control; this not related radiosurgery however this may be natural course. 

Rationale Hemangiopericytoma (HP) is well known as one of very rare intracranial tumors, and especially skull base one which is very close to the cavernous sinus should be too difficult to be cured by microsurgery alone because of much abundant feeding arteries. If there is residual tumor, it will have often opportunity to develop extracranial metastases. In addition, we should do stereotactic radiosurgery to prevent tumor regrowth. In our institute, we have tried to install and investigate pathological micro-anatomy in Gamma plan to grasp relationship between the tumor and surround vital structures, and hope to prevent tumor shrinkage without any neurological deficit.

Materials and Methods We already treated 11 HP cases with Gamma knife surgery (GKS) after surgical resection, and median prescribed dose was 12 (10-25)Gy. Among of them, 36.4%(4/11) could be controlled, but 54.5%(6/11) experienced extracranial metastases. The clinical results were failure, and then we have investigated micro-anatomy to detect tumor origin and its extension much more precisely to make complete dose planning. In detail, we irradiated higher dose (80% isodose area) to the tumor origin as a priority target, and also cover the tumor as much as possible to make tumor necrotizing due to feeding arterial occlusion pathologically.

Result We experienced another case whose tumor was located in the Meckel’s cave which was looked like trigeminal schwannoma, but the onset was very progressive abducens nerve palsy. At that moment, we investigated micro-anatomy in Gamma plan, but didn’t make clear diagnosis. So we decided to do surgical resection to define pathology with maximal volume reduction. We couldn’t remove sufficiently because of terrible bleeding from the tumor, and did GKS for residual with 24Gy as prescribed peripheral dose according to the micro-anatomical simulation. The patient experienced no deficit with tumor disappearance at 6 months later to GKS, and has to be observed much longer follow-up.

Conclusion We will progress the knowledge and clinical experiences of GKS based on pathological micro-anatomy to establish heavy particle microsurgery system to provide patients complete remission to prevent extracranial metastases.

Objective. To assess efficacy and safety of Leksell Gamma Knife Radiosurgery (LGKRS), in terms of tumor control (TC) and permanent symptomatic adverse radiation effects (AREs) (end points); to identify some prognosticators significantly influencing the end-points in patients affected with glomus jugular tumors (GJTs). 

Methods. Between April 1996 and December 2013, 51 consecutive patients with GJTs underwent LGKRS at our Department and were followed up for at least 3 years. Males/females ratio was 10/40; mean age was 56.6 years (20-82). In 21/50 (42%) patients, an endovascular embolization was performed and 19/50 (38%) cases underwent one or more surgical procedures before LGKRS. At the time of treatment, GJTs were classified according to the Glasscock–Jackson (G-J) grading as follows: Grade I 11 patients (23%), Grade II 23 (46%), Grade III 6 (11%), and Grade IV 10 (20%). Mean and range dose planning parameters were as follows: gross target volume (GTV: 0.22 mL, 0.06-0.70), prescription dose (PD: 16.8 Gy, 12.0-24.0), prescription isodose (PI: 49.70 %, 40-55), maximum dose (MD: 32.62 Gy, 24-48) and shot number (13.3, 2-33). On GK day, stereotactic localization was performed using volumetric and T1 fat saturated MRI sequences. Uni-multivariate logistic regression analysis was performed to evaluate which of the following independent variables – G-J grading, pre-LGKRS endovascular treatment, pre-LGKRS surgery, GTV and PD, adjusted for age and sex – could potentially influence the end points.

Results. Median survival was 53.9 months (37.7-183.50). At last clinical follow-up, an improved or stable neurological examination was observed on the whole series with no symptomatic AREs. TC was achieved in 96.0% of cases (48/51) with an actuarial progression-free survival rate of 97% at 10 years. No malignant transformation or radiation induced tumors were observed. Statistical analysis showed that G-J Grade II (p=0.043) was the only prognosticator positively affecting the probability of TC in such patients.  

Conclusion. Despite the particularly critical location of GJTs, long-term outcomes show that LGKRS reveals to be an extremely effective and safe primary or adjuvant treatment in such tumors, as well.

Objectives

WHO grade 2 meningiomas are aggressive tumors associated with a high recurrence rate leading to repeated surgical procedures, which can seriously worsen a patient’s neurological status. Although radiosurgery is an increasingly popular technique, its role in the management of grade 2 meningiomas has yet to be defined. In this study the authors aimed to evaluate radiosurgery in achieving control of proven tumor progression occurring after resection of grade 2 meningiomas.

Methods 

This retrospective study included consecutive patients, between 2000 and 2015, treated by gamma knife radiosurgery (GKRS) for radiologically proven progression of a previously surgically treated grade 2 meningioma.

Results

Thirty patients were eligible for analysis. There were 10 men and 20 women with a mean age of 59 years. The mean radiation dose was 15.2 Gy (range 12–21 Gy), and the mean target volume was 5.4 cm3 (range 0.194–14.2 cm3). Thirty-eight radiosurgical procedures were performed in the 30 patients. The mean progression-free survival after radiosurgery was 32.4 months among those with progression in a target irradiated volume and 26.4 months among those with progression in any intracranial meninges. With a mean follow-up of 56.4 months (range 12–108 months), the 12-, 24-, and 36-month actuarial local control rates for all patients were 75%, 52%, and 40%, respectively, and the regional control rates were 75%, 48%, and 33%. A single case of transient hemiparesis completely resolved without sequelae.

Conclusions 

Radiosurgery appears to be a safe and effective treatment for the local control of delayed progression after resection of a WHO grade 2 meningioma. Higher radiation doses similar to those applied for malignant tumors should be recommended when possible.

The ventro-intermediate thalamic nucleus (Vim) is a commonly used target in functional neurosurgery, such as Deep Brain Stimulation and Gamma Knife thalamotomy, for treatment of drug-resistant tremor.  Unfortunately, the routinely used clinical Magnetic Resonance Imaging (MRI) does not allow its direct visualization. Consequently, the targeting strategies are indirect, such as atlas-based registration and stereotactic coordinates. Recent findings have shown the potential of the susceptibility-weighted imaging (SWI) acquired at 7T for imaging the thalamic nuclei with Vim being one of them. The aim of this study is to compare the localization of the target points defined by the quadrilateral of Guiot, as used in daily clinical practice, with the visual Vim-area on SWI at ultra-high field.

Data was collected from five young healthy subjects (25±2 y.o., 3 males). At 3T (TimTrio SIEMENS Scanner) we acquired the standard imaging protocol for Gamma Knife thalamotomy (GKT) including T2-weighted (TR/TE=3200/402 ms, 0.5x0.5x1mm³)  and T2-weighted CISS (TR/TE=6.18/2.75 ms, 0.4375x0.4375x0.44 mm³) images, and at 7T (Siemens Medical Solutions) we acquired the SWI sequence (TR/TE=28/20 ms, 0.375x0.375x1mm³). Guiot targeting was performed bilaterally, six times for each subject using the 3T images. The left and the right Vim-area were manually delineated based upon the intensity variation observed from the SWI as well as the Schaltenbrand and Wahren stereotactic atlas. Both of these tasks were carried out in MITK 3M3 software.

The Vim was outlined for nine out of ten thalami, while the last one was difficult to discriminate due to a presence of a blood vessel. The volumes of the delimitated Vim-area are in the interval [76.3, 83.3] mm³ conforming its size expectations. The quadrilateral of Guiot showed to be highly reproducible with a maximum intra-subject variability of 1.1mm. Additionally, these points were always inside the manually delineated Vim and predominantly in the ventral part of the outlined volume showing a tendency of their localization.

This study reports for the first time, to the best of our knowledge, a validation of the clinical targeting against subject-related imaging reference. Moreover, we observed that the clinically used Guiot targeting points are confined in the ventral part of the visually distinguishable Vim as provided by SWI acquired at ultra-high field. Further studies with larger datasets, such as tremor patients and their electrophysiological confirmation, should validate these findings.

The difficulty to directly visualize the Ventro-intermediate thalamic nucleus (Vim) on the currently routine MRI leads to the application of indirect targeting methods in stereotactic neurosurgical procedures, such as Gamma Knife Surgery (GKS) and Deep Brain Stimulation (DBS), for a treatment of drug-resistant tremor.

There has been a growing effort for automated targeting in the image-processing community based on diffusion-MRI. This technique enables the depiction of the different structural-connectivity properties and therefore, the specific fiber orientation inside each thalamic nucleus.

Our group proposed an automated and robust method across healthy subjects and tremor patients for parceling the thalamus in seven main groups of nuclei while exploring the local diffusion information from the spherical harmonics representation of the orientation distribution functions (ODFs) in k-means clustering framework. One of the resulting parcels is the Ventral-Lateral-Ventral group (VLV) enclosing all motor-related nuclei including the Vim. We aim at further automatic subdivision of the VLV cluster. To this end, within the VLV, we first built a k-nn graph with edges corresponding to the respective ODFs distances and then, we performed a partition in 3 sub-clusters using the NCut algorithm.

The proposed subdivision was compared to the radiological response in the follow-up images of 17 patients treated for tremor with GKS unilaterally. The MRI protocol included pre-operative diffusion-weighted images with 64 (or 72) gradient direction, b=1000s/mm² and voxel-size: 2.2x2.2x2.2mm³, pre-operative and post-operative MPRAGEs with isotropic resolution of ~1mm³ - all of them acquired at 3T. In general, the sub-partition followed a spatial-distribution pattern and for 12 out of 17 cases we observed that one specific sub-cluster encloses entirely or the major part of the contrast enhancement corresponding to the GKS target appearing on the follow-up images.

The evaluation of the sub-partition outcome was further extended in one additional tremor patient treated with DBS, bilaterally. Similar pre-operative MRI data, as for the GKS-treated patients, was acquired. We observed that in each hemisphere both initial targeting point and the final electrode position are inside the anticipated sub-cluster. The final left and right position are 5 and 1 mm distant from the corresponding sub-clusters’ centroids, respectively.

This study shows that the diffusion-MRI-based sub-clustering of the VLV thalamic nuclei could potentially allow, in an automated manner, to narrow the area of Vim’s localization. Our preliminary results will be further investigated in a larger patients dataset, treated by either GKS or DBS, and on diffusion images with higher spatial resolution.

Objective: Gamma knife thalamotomy (GKT) is a well-established treatment for medically refractory tremor patients who are at risk for invasive procedures. The purpose of this study was to evaluate whether staged bilateral GKT provides benefit and acceptable risk to patients suffering from disabling bilateral tremor.

Methods: At a median interval of 22 months, 11, patients underwent staged bilateral GKT during a 17 year period (1999-2016). Eight patients had essential tremor (ET), two had Parkinson’s disease (PD) related tremor, and one had Multiple-sclerosis (MS) related tremor. For the 1^st GKT, a median maximum dose of 140 Gy was delivered to the posterior-inferior region of the nucleus ventralis intermedius (VIM) through a single isocenter with 4-mm collimators. Patients who benefitted from unilateral GKT were eligible for a contralateral GKT. For the 2^nd GKT, a median dose maximum of 130 Gy was delivered to the opposite VIM nucleus to a single 4-mm isocenter. We used the Fahn-Tolosa-Marin (FTM) clinical tremor rating scale to score tremor, drawing, writing and drinking before and after GKT. We used the Karnofsky performance scale (KPS) to grade quality of life and activities of daily living before and after the GKT.

Results: All patients had improvement in tremor after 1^st GKT. The median time to last follow-up after the 1^st GKT was 35 months (range, 11-70 months). No patients had tremor recurrence or diminished tremor relief. One patient experienced new temporary neurological deficit (contralateral lower extremity hemiparesis) from the 1^st GKT which improved on corticosteroids. The median time to last follow-up after the 2^nd GKT was 12 months (range, 2-70 months). Nine patients had improvement in at least one FTM score after the 2^nd GKT. Two patients had tremor arrest and complete resolution of function. No patient experienced tremor recurrence or diminished tremor relief after the 2^nd GKT. No patient experienced new neurological or radiological adverse effect from the 2^nd GKT. Statistically significant improvements were noted in the KPS following the 1^st and 2^nd GKT.

Conclusions: Staged bilateral GKT provides effective relief for medically refractory, disabling, bilateral tremor without increased risk of neurological complications. It is an appropriate strategy for carefully selected medically refractory bilateral tremor patients ineligible for deep brain stimulation.  

Objective: This study aims at reporting the correlation between the clinical results and the one-year postoperative MR neuro-imaging characteristics of the thalamic lesion after Gammaknife radiosurgery for  tremor.

Methods: Between April 2004 and March 2015, a Vim Gammaknife thalamotomy was performed in 319 patients for essential or Parkinsonian tremor in Marseille University hospital with a very stereotyped procedure. A neuro-imaging and clinical assessment was performed at one year FU for 253 patients. The volume of the lesion defined as the whole area of post-contrast enhancement was calculated for each patient in mm³, the pattern of lesion determined and the amount of edema evaluated according to a semi-quantitative scale. A comprehensive clinical evaluation by expert neurologists was performed at the same time.  Statistical analysis was performed using R software (Version 1.0.136/2016 RStudio, Inc)

Results: Imaging data were analyzable and reviewed for a total of 169 patients at one year follow-up. Among these patients, data from neurological clinical evaluation were obtained for 91 patients. The median percentage of tremor reduction was 70% (0-100%, SD:30%).  The median volume of the lesion at 12 months FU (+/- 3 months) was 91,45 mm³ (Mean = 104, Min:0, Max :1120, SD:284) . A correlation was established between the volume of the lesion and the percentage of tremor reduction (Pearson's coefficient of correlation r =+ 0,26 (p=0,0178). In patients regarded as clinical failure (< 45% of tremor improvement), the lesion volume was significantly smaller than in patients deemed responders (> 45% tremor reduction) ,p <0,0001).The amount of edema surrounding the lesion was found to be significantly related to the clinical improvement (p = 0.022). The “cocade” pattern enhancement type was strongly related to good outcome (p<0,001) and the absence of enhancement to the absence of improvement (p<0,00001 ,62% versus 0,07%).

Conclusions: These data confirm our previous results derived from 50 patients with blinded analysis of clinical outcome (Witjas and al. Neurology, 2015). Even though a significant correlation does exist between lesion volume, edema and clinical improvement, concordance is far from being very strong and linear between the imaging and clinical responses. These findings prompt to look for additional factors in order to better characterize the effects of Gammaknife that might also rest upon a delayed non-lesional neuromodulatory mechanism. These fascinating questions are of utmost importance and currently under investigation in our department.

Constantin Tuleasca^1,3,9, MD, Tatiana Witjas^4,5, MD, Elena Najdenovska², MSc, Antoine Verger⁶, MD, Nadine Girard⁷, MD-PHD, Jerome Champoudry⁸, MSc, Jean-Philippe Thiran^3,9, PHD, Meritxell Bach Cuadra², PHD, Marc Levivier^1,9, MD, PHD, Eric Guedj*⁶, MD, AND Jean Régis*⁸, MD

Objective: To assess for the first time structural brain changes, by voxel-based morphometry (VBM), before and after unilateral Gamma Knife thalamotomy (GKT) for drug-resistant tremor. To identify differences between clinical responders and non-responders to GKT.

Methods: Thirty-eight patients (mean age 71.8 years) with severe refractory right essential tremor (ET) were treated with unilateral left GKT. Targeting of ventro-intermediate nucleus (Vim) was performed with Leksell Gamma Knife using a single 4-mm collimator and 130 Gy.  Neurological, neuropsychological and neuroimaging (3 Tesla, including 3D T1 weighted) assessment had been done at baseline and 1 year after GKT. Clinical responders were considered those improved in tremor score (Fahn-Tolosa-Marin) with at least 45%.

Results: Thirty-one (81.6%) patients were responders (R) and 7 (18.4%) non-responders (NR). With regard to GM changes after GKT, independently of clinical answer, atrophy was present in extensive areas (right globus pallidus, left putamen, left thalamus, right anterior and medio-dorsal thalamus, cerebellar, right premotor and supplementary motor area, left and right visual association cortex, right ventral temporal, left parahippocampal and posterior cingulate gyrus). The interaction between R - NR with time showed brain plasticity in R remote areas, within left temporal pole (BA 38) and cluster including left occipital cortex (BA 19), visual areas V4 and V5, parahippocampal place area (p_unc<0.005, k>120).

Conclusions: Our results show brain plasticity after unilateral left GKT. Responders present changes in areas involved in motion, mainly locomotor monitoring towards the local and distant environment, suggesting the requirement to recruit in the targeting specific visuomotor networks.

Objective: To assess the feasibility and tolerance of bilateral Gamma Knife thalamotomy (GKT) in Essential Tremor (ET)

Background: Unilateral GKT is an established treatment for severe tremors. However, essential tremor is usually bilateral. The persistence of contralateral tremor may induce an impairment in activities of daily living (ADL). Bilateral procedures with thermocoagulation were contraindicated because of the risk of balance, cognitive or speech problems. As the lesion induced by radiosurgery within the VIM is progressive and limited, we proposed a study on bilateral GKT. Here are the preliminary results.

Methods: 15 patients (8 women) with severe ET who had benefit from a first GKT and who had a severe permanent contralateral tremor were included. Patients were included if there was no impairment in their balance or speech and if the neuropsychological assessment was stable. The 2^nd GKT was performed at least 18 months after the first GKT. Patients were assessed before and quarterly for at least 12 months after GKT2, with tremor rating scale, neuropsychological and gait/balance assessments and MRI. VIM lesioning was performed with Leksell Gamma unit with a single exposure through a 4mm collimator. Radiosurgical dose was 130Grays.

Results:  here are the preliminary results for 9 patients who completed the study at 1 year. Tremor score on the treated hand was improved by 57%. The improvement of ADL was 95%. Cognitive score and gait assessment were stable. No patient had hypophonia or dysarthria. Two patients were not significantly improved. One patient had a side effect related to GKT2. She developed hemiataxia and dysarthria induced by a hyperresponse pattern 11months after GKT.

Conclusions: These preliminary results on bilateral GKT for severe ET in a selected cohort of patients shows that the procedure is feasible without a major risk of cognitive or balance problems. However, a longer follow-up is needed

Objectives

Stereotactic radiosurgery (SRS) after surgical resection of brain metastases has been shown to be an effective alternative to whole brain radiation for local control.  In this study we assessed local failure patterns after post-op SRS to the surgical bed of completely resected brain metastases.

Methods and Materials

In a phase III randomized study (NCT00950001), 132 patients, who had no more than 3 brain metastasis with at least one completely resected were randomized to SRS or observation (OBS).  Complete resection was verified by postoperative magnetic resonance imaging (MRI) and SRS was delivered to the cavity by post-op day 30.  Using the Elekta Perfexion Gamma Knife unit, SRS was delivered with a 1 mm margin and the  prescription dose was target volume dependent which ranged from 12 to 16 Gy.  Tumor recurrence in the resection cavity or its immediate vicinity was identified on routine follow up MRI.  The first MRI that identified recurrence was imported into the Leksell GammaPlan software and registered with the cavity treatment planning MRI.  The recurrence was contoured and its position was evaluated relative to the prescription isodose line.  The location of recurrence was classified into three categories; 1) “in-field”, 2) “cross-field” and 3) “marginal” if the recurrence volume was within, crossed, or adjacent to the prescription isodose line, respectively.  The minimum dose delivered to the area of recurrence was quantified.

Results

12 of the 64 patients randomized to SRS had local recurrence after SRS.  Five cases failed in-field, 4 cases were cross-field and 3 cases were marginal.  The minimum dose received by the volume showing in-field failure during SRS delivery ranged from 12 to 20 Gy.  All marginal failures were notable for having dural involvement with the recurrence occurring adjacent to the prescription isodose line.

Conclusion

The factors contributed to failure after SRS are complicated.  In-field failures may be due to an inadequate prescribed dose or a biologic resistance to radiotherapy.  Marginal failures maybe related to inadequate margins.  A larger margin, particularly along the dura may help increase local control with SRS after surgery.  Cross-field failures are likely multifactorial due to factors allowing to both in-field and marginal failures.  Evaluation of the clinical factors and comparison to the failures within the OBS arm are ongoing.

Objectives: For patients with up to 10 brain metastases (BM), studies increasingly support excellent disease control and toxicity outcomes with stereotactic radiosurgery (SRS) in lieu of whole brain radiotherapy (WBRT) or as salvage therapy for recurrence after WBRT. Outcomes for patients with 10 or more metastases treated with SRS remain unclear. We report our institutional experience of treatment of patients with 10 or more brain metastases at first SRS treatment session.

Methods:  We identified 97 patients with 10 or more brain metastases (BM) receiving fixed-frame SRS (FFSRS) without WBRT for newly-diagnosed or recurrent BMs at our institution between September 1998 and December 2013. We reviewed treatment-related outcomes of overall survival (OS), treated lesion freedom from progression (FFP), freedom from new metastases (FFNM), and adverse radiation effect (ARE).

Results: Among the 97 FFSRS-treated patients, the median age was 56 (range 17-84) and median KPS 80 (range 40-90). Primary histologies within the cohort were breast (38), lung (28), melanoma (22), and other (9). Median number of BM per patient was 12 (IQR 11-16). Median total treatment volume was 4.3 cc  (IQR 2.3-11.5 cc). Forty-three patients received FFSRS without WBRT as upfront BM radiotherapy and 54 as salvage therapy after prior WBRT, with median OS 7.8 and 8.8 mo, respectively. Sixty patients had available follow up imaging. FFP at 1 year with 95% confidence intervals for upfront vs. salvage FFSRS was 70% (49-84%) vs. 55% (30-74%) by patient and 94% (91-96%) vs. 86% (81-90%) by lesion. FFNM at one year for upfront vs. salvage FFSRS was only 9% (2-22%) vs. 14% (4-30%). At progression, 40 patients received additional FFSRS, 20 (21%) had WBRT, and one had partial brain radiotherapy. Symptomatic ARE was observed in 1% of 1018 treated lesions over the patients’ disease course. Seven patients experienced symptomatic ARE.

Conclusions: Our institutional experience demonstrates excellent local control following FFSRS for patients with 10 or more BM as upfront therapy or at recurrence following prior WBRT. Rates of ARE were modest, consistent with prior studies. Expectedly, these patients with large burden metastatic disease had a high rate of new metastases often requiring salvage SRS or WBRT.

Objectives:  To determine the implications of HER2/neu-amplification (HER2) status on local control and adverse radiation effect (ARE) after stereotactic radiosurgery for the treatment of brain metastases from breast cancer.

Methods:  We retrospectively reviewed all Gamma Knife radiosurgery cases from 1998-2013 for the treatment of brain metastases from breast cancer at our institution. Newly-diagnosed brain metastases treated with radiosurgery alone and new or recurrent brain metastases after prior radiotherapy were included, and resection cavities, lesions treated with prior radiosurgery, and lesions with no imaging follow-up were excluded from this analysis. Hormone receptor status was determined based on pathology reports from the primary breast tumor. Freedom from progression (FFP) and cumulative incidence of ARE for all treated lesions were determined using the Kaplan-Meier method with censoring at last imaging. Univariate analysis by lesion was performed using the log-rank test.

Results:  1314 newly-diagnosed or recurrent brain metastases in 204 breast cancer patients were included for analysis. The median imaging follow-up was 9.0 months. Actuarial 1- and 2-year FFP probabilities by lesion were 89% and 79% overall. HER2 positive status was associated with worse FFP; 1-year and 2-year FFP probabilities were 85% and 75% for HER2 positive tumors versus 95% and 86% for HER2 negative tumors (p<0.0001). Further stratification of HER2-positive lesions by ER status showed no significant difference in FFP or ARE, whereas among HER2 negative tumors, 1- and 2-year FFP probabilities were 97% and 91% for ER positive versus 92% and 80% for ER negative metastases. A dose-response relationship was seen in HER2-positive lesions (p<0.001). In HER2 positive lesions, the use of systemic therapy within 1 month of radiosurgery was associated with improved tumor control (p = 0.0007) and did not increase the risk of ARE. The cumulative incidence of ARE at 1 and 2 years was low at 2% and 6% for HER2 positive tumors versus 2% and 3% for HER2 negative tumors (p=0.11).

Conclusions:  Radiosurgery for HER2-positive brain metastases from breast cancer appears to be associated with poorer tumor control with low incidence of ARE.  Treatment intensification with dose escalated radiosurgery or concurrent treatment with systemic therapy should be considered.

Objectives:  To analyze radiographic and survival outcomes in a large cohort of patients with brain metastases who received hypofractionated stereotactic radiotherapy (HSR) with a focus on short-term and long-term tumor response, complications and mortality.  

Methods: Patients with brain metastases who underwent HSR between 2010 and 2016 were included in the study. Radiation treatments were performed with Gamma Knife 4C and Perfexion (Elekta AB, Stockholm, Sweden), Cyber Knife (Accuray, Sunnyvale, CA, USA) and linear accelerator TrueBeam STX (Varian Medical Systems, Palo Alto, CA). The indications for HSR were determined by the presence of large volume tumors or proximity to critical brain structures. Patients with multiple brain metastases were subjected to a combination of HSR and stereotactic radiosurgery (SRS). Radiation schemes were selected depending on the number of metastases, size, location, proximity to critical brain structures, histological type of primary cancer and patient general condition. Following treatment the patients underwent control MRI examination with standard protocols (2 mm T2 and 1 mm T1 with double contrast enhancement) at 8 weeks and then every 3 months. The median follow-up time after HSR was 9 months.

Results: The study revealed that the application of HSR for the treatment of large volume or critically located brain metastases provided a high level of local control (12-month local control rate was 86 %).  Shrinkage of tumor volume by more than 50% was observed in a vast majority of patients with radiosensitive tumor histology, which resulted in considerable improvement of the patients’ neurological condition. Complications in the form of radiation necrosis occurred in 20% of patients at a median of 7.6 months after HSR. History of previous brain irradiation increased the risk of radiation necrosis (HR=3.4, p<0.001). For the entire cohort 12-month and 24-month overall survival rates after HSR were 45% and 24% respectively. There was no statistically significant difference in the median survival of the patients receiving HSR alone and those receiving HSR plus SRS. Mortality within 2 months after HSR was 10% and was associated with neurological deterioration or systemic disease progression. The best survival results were obtained in patients belonging to the first RPA-class who achieved one-year survival in 73% of the cases.

Conclusion: HSR and its combination with SRS is an effective treatment strategy for patients with brain metastases having at least one large unresectable lesion or a lesion located in/near critical brain structures.

Background: The diagnosis specific graded prognostic assessment (ds-GPA) model is one of the most prevalent methods of prognostication for patients with cerebral metastases. The ds-GPA models for lung, RCC, and melanoma each contain number of metastases as one of the prognostic inputs. Recent literature has revealed that the impact of number of metastases has on patient outcomes is often overshadowed by the impact of cumulative intracranial volume (CITV) and other clinical variables. With this in mind, we wished to test if a simplification of the ds-GPA model for lung, renal cell carcinoma (RCC), and melanoma was non-inferior to the original model.

Objective: We wished to examine the hypothesis that binarizing the number of metastases to 1 and greater than 1 is non-inferior to the three bin model used for patients with lung, melanoma, and RCC cerebral metastases after accounting for the other components of their respective ds-GPA model.

Methods: Our patient cohort of interest consisted of 4348 patients (3745 lung, 321 RCC, 282 melanoma) from the United States, Japan, and Australia who underwent SRS for one or more brain metastases We used the statistical metrics net reclassification index (NRI) and integrated discrimination improvement (IDI) in order to examine the ds-GPA model for lung, RCC, and melanoma against a similar model with the only change being the number of metastases category binarized. Through use of these statistical measures we were able to compare the effect of replacing the 1, 2-3, and >3 metastases grouping with 1 and >1.

Results: We found that for lung, RCC, and melanoma no difference between the original ds-GPA and the simplified version was detectable even after controlling for the other elements of each pathology’s respective model (NRI and IDI CI’s capturing 0 with p >.05 for each tumor type). In fact, when subgroups of patients with similar tumor pathology were further stratified by institution/location, we found instances where the simplified model was superior to the original ds-GPA.  

Conclusion: We found that simplifying the number of metastasis portion of ds-GPA model for lung, RCC, and melanoma was non-inferior to their respective, original ds-GPA models.

Introduction

Gamma Knife Radiosurgery (GKR) of large numbers of micrometastases can be time consuming. Beam on time can often become the limiting factor in being able to treat every lesion in a particular patient. One technique to reduce beam on time is to use a larger collimator to a high isodose. This controversial technique decreases the dose gradient inside and outside the target. However, it is uncertain whether this makes a clinical difference, when target and treatment volumes are so small.

Objective

To investigate whether there is a difference in clinical efficacy between treating micro-metastases with GKR using the 4mm or 8mm collimator.

Methods

So far 33 patients undergoing GKR for brain metastases have consented to participate in this trial. Micro-metastases (defined here as under 0.14cc and maximum diameter under 6mm) were randomised to a single shot with either the 4mm or the 8mm collimator. Brainstem lesions and lesions within 11mm of one another were excluded. A marginal dose of 25Gy was prescribed for all lesions

Primary outcome was radiological local control at 12 months post GKR, or at the last imaging follow-up if death occurred during the first year post-GKR.

Results

72 eligible lesions were identified in 20/33 subjects and underwent randomisation. 38 lesions were randomly assigned to the 4 mm collimator and 34 to the 8 mm collimator.

A marginal dose of 25Gy was prescribed for all lesions. The mean Prescription Isodose was 73.4% (43-93) for the 4mm group and 95.7% (88-99) for the 8mm. 

The outcome has been reached by 66% of the micro-metastases and at least one follow up was available for 81% of the lesions. The mean follow-up time was 7.6 months (2.1-18.9). No lesion has shown definite enlargement (up to 30% TV increase) and no radiological evidence of oedema or radionecrosis has been detected. In the 4mm group, 40% of the lesions remained unchanged and 60% got reduced or disappeared. Similarly, in the 8mm group 45% were stable and 55% reduced/disappeared.

Seven patients developed new metastatic lesions at follow up and one of them underwent WBRT. 60% (12) patients have died, three of them from progressive intracranial metastatic disease, at a mean time of 6 months (0.3-11.3) post-GKR.

Conclusions

Initial results suggest that optimisation of treatment time for small lesions may be achieved by using the 8mm collimator without compromising clinical efficacy.  Further recruitment and follow up is needed.

Objective: We present long-term follow-up data after Gamma Knife radiosurgical (GKRS) treatment of acoustic neuroma.

Patient and Methods: Six-hundred and eighteen patients were radiosurgically treated for acoustic neuroma between 1992 and 2016 at the Department of Neurosurgery, Medical University Vienna. Patients with neurofibromatosis and patients who could not yet have a one year follow-up were excluded from the study. Thus, we present data of 557 patients with spontaneous acoustic neuroma and long-term follow-up data on 426 patients with a minimum follow-up of two years. Koos grades at time of diagnosis and at time of GKRS were evaluated. Patients were assessed according to the Gardner-Robertson hearing scale and House-Brackmann facial weakness scale prior to GKRS, and at times of follow-up. Data were evaluated retrospectively.

Results: 452 patients (81%) were treated radiosurgically alone and 105 patients (19%) were treated combined microsurgically-radiosurgically. Whereas the combined treatment was favored especially prior to 2002, the percentage of only radiosurgically treated cases has significantly increased since then. The overall complication rate after GKRS was low. The complication rate after GKRS further declined in the last decade. A significant enlargement of the ventricular system (hydrocephalus) after GKRS was only observed among patients harboring Koos grade III or IV tumors. One case of malignant transformation after GKRS was diagnosed (0.2%). Radiological outcome after GKRS revealed stable or decreased neuromas in the vast majority of cases including all Koos grades. The rate of non-functional hearing was already rather high prior to GKRS. At last follow-up, preservation of functional hearing was achieved in 52% of patients classified as Gardner-Robertson grade I or II prior GKRS.

Conclusion: GKRS is a safe and effective treatment in patients of all Koos grades. Advancements in the radiosurgical treatment especially over the last decade have led to a low complication rate and excellent outcome.

Objective Vestibular schwannoma (VS) is a benign tumor with associated morbidities and reduced quality of life. Except for mutations in NF2, the genetic landscape of VS remains to be elucidated. Little is known about the effect of Gamma Knife radiosurgery (GKRS) on the VS genome. The aim of this study was to characterize mutations occurring in this tumor to identify new genes and signaling pathways important for the development of VS. In addition, the authors sought to evaluate whether GKRS resulted in an increase in the number of mutations.

Methods Forty-six sporadic VSs, including 8 GKRS-treated tumors and corresponding blood samples, were subjected to whole-exome sequencing and called tumor-specific DNA variants. Pathway analysis was performed using the Ingenuity Pathway Analysis software. In addition, multiplex ligation-dependent probe amplification was performed to characterize copy number variations in the NF2 gene and microsatellite instability testing was done to investigate for DNA replication error.

Results With the exception of a single sample with an aggressive phenotype that harbored a large number of mutations, most samples showed a relatively low number of mutations. A median of 14 tumor-specific mutations in each sample were identified. The GKRS-treated tumors harbored no more mutations than the rest of the group. A clustering of mutations in the cancer-related axonal guidance pathway was identified (25 patients), as well as mutations in the CDC27 (5 patients) and USP8 (3 patients) genes. Thirty-five tumors harbored mutations in NF2 and 16 tumors had 2 mutational hits. The samples without detectable NF2 mutations harbored mutations in genes that could be linked to NF2 or to NF2-related functions. None of the tumors showed microsatellite instability.

Conclusions The genetic landscape of VS seems to be quite heterogeneous; however, most samples had mutations in NF2 or in genes that could be linked to NF2. The results of this study do not link GKRS to an increased number of mutations

Background: Modern clinical studies have converged upon “maximal safe resection” as the optimal surgical strategy for moderate-to-large vestibular schwannomas, however data on the long-term outcomes of residual tumor is lacking and the best treatment strategy for residual disease is an open question. The goal of this study was to review our rates of residual tumor growth, and the response to salvage versus upfront SRS.

Methods: This retrospective single-institution study included all sporadic vestibular schwannomas treated primarily with subtotal resection at our institution from 2002 – 2015. Patients with less than 1 year of follow-up imaging were excluded. Patients treated with upfront SRS after surgery were analyzed separately. The primary outcome was tumor stability or growth requiring salvage treatment, and response to salvage treatment (surgery or SRS).

Results: 295 patients underwent primary surgery for vestibular schwannoma at our institution between 2002 and 2015. A subtotal resection was performed in 140 of these cases. 49 cases were excluded due to <1 year follow-up imaging. 17 cases received upfront SRS after surgery (12 cases Gamma Knife, 12Gy; 5 cases CyberKnife, 25Gy in 5 fractions). There were 2 failures requiring salvage surgery 1.6 and 2.6 years after SRS (88% control rate at 1.8 median follow-up after SRS). Of the remaining 74 residual tumors managed with observation after STR, 57 (77%) remained stable at a median 4.1 years after surgery.  17 tumors (23%) progressed and required salvage treatment at a median of 2.8 years after STR. 11 were treated with Gamma Knife (12.5 Gy) at a median interval of 2.6 years after surgery, and all remained clinically stable over a median follow-up of 2.3 years after salvage SRS. One was treated with Cyber Knife (18Gy in 3 fractions) and went on to require a second surgery 3 years later. 4 residual tumors were treated with salvage surgery at a median interval of 5 years after STR, and one was lost to follow-up.  The control rate of salvage SRS was 92% at 2.3 years median follow-up.

Conclusion:  A majority (77%) of residual vestibular schwannomas will remain stable after a primary subtotal resection. Both upfront and salvage SRS had good control rates (~90%) in our series. Further study is needed to define the role of upfront versus salvage SRS after a primary STR.

Introduction
Prognostic factors of tumor control after Gamma Knife Radiosurgery (GKRS) for vestibular schwannoma (VS) are largely unknown. Recently, it has been reported that the growth rate of VS before treatment is indicative of the chance that radiosurgery achieves tumor control. Such findings may have important implications for treatment strategies and may lead to advise for either microsurgery or higher marginal doses for fast growing tumors. However, studies on this important aspect are limited and show conflicting results. Moreover, the available studies are hampered by methodological limitations such as limited patient numbers and follow-up and two-dimensional assessment of tumor size. The objective of this study is to identify a possible correlation between pre-treatment growth rate and tumor control after GKRS in a large database with sufficient follow-up and volumetric tumor assessments.

Methods
In the prospectively collected database of the Gamma Knife Center Tilburg, 445 patients with VS, treated between 2002 and 2014, that showed documented growth before treatment and who have had a minimum follow-up of 2 years after treatment, were identified. Tumor volumes before, at and after treatment were assessed. GKRS was performed in a uniform way, with a dose of 12-13Gy prescribed to the isodose line covering 90-99% of the target. Failures were defined as tumor progression on 2 consecutive MRI’s beyond 2 years after GKRS, or as judged by the radiosurgical team. Volume doubling times (VDT) before treatment were correlated with the observed tumor control rates and volumetric responses after treatment.

Results
Until now 266 of the 445 patients with documented pre-treatment tumor growth have been analyzed. 25 Patients were lost to follow-up. The median follow-up was 4 years. 25 Patients showed a radiological failure. The 5- and 10-year actuarial control rates were 91% and 78% respectively. VDT varied from 3 to 344 months, with a median of 16 months. Using the Mann-Whitney-U test, the VDT of tumors that showed tumor control is significantly higher than those that failed (p=0.01). After stratifying for VDT at the median, slow growing tumors showed a 5- and 10-year actuarial control rate of 97% and 89%, where the fast growing tumors had a 5- and 10-year control rate of 85% and 68% (p=0.009).

Conclusion
This study clearly shows that the pre-treatment growth rate correlates with the observed tumor control after GKRS. Fast growing tumors are less likely to show tumor control. This finding might justify alterations in the management of VS.

Background: Class I level evidence for treatment of Vestibular Schwannoma is generally lacking. One particular issue that has been put forward as an argument against radiosurgery for this tumor is that it is so quiescent that the effect of radiosurgery only reflects its natural course. In 2014, we introduced a randomized trial to measure the effect of radiosurgery against natural course.

Study design: The study will include 100 patients with newly diagnosed unilateral VS with diameter 5-20mm, age 19-69. Patients are randomized to  up-front Gamma Knife radiosurgery or observation by serial MRI. They are followed annually for 4 years. The study is blinded to observer (study radiologist and study physician). The primary endpoint is tumor growth measured as relative to baseline size and volume doubling time at four years. Secondary endpoints include Gardner Robertson hearing Class, need of additional treatment  and adverse effects. Patients fill out the PANQOL and EQ-50 forms at each control.

Results: The study has included 80 patients per February 2017 and has so far lost one patient who withdrew. One patient who was randomized to radiosurgery declined such treatment and is followed according to intention-to-treat principle.  The practical setup and the experiences learned from the study design will be presented.

Object: Although several small individual series on repeat Gamma Knife radiosurgery (GKS) for recurrent vestibular schwannoma (VS) following prior GKS have been published, we aim to systematically aggregate data from the literature as well as from our own institutions to better understand the safety and efficacy of repeat GKS for VS.

Methods: All patients that underwent repeat GKS of sporadic VS at two tertiary academic referral centers between 2006 and 2016 were eligible for study. An aggregated dataset of previously published cases plus our own data were analyzed. A cohort of patients treated with salvage microsurgery (MS) following failed GKS were used as comparison. 

Results: Sixteen patients from our own institutions were included for analysis. Five patients were previously treated with MS. Mean age at first treatment was 60.1 years, mean time between first and second treatment was 50 months (4.2 years). Mean tumor volume at the first treatment was 0.539 cm³, at the second treatment 1.407 cm³. Mean follow-up after the second GKS treatment was 34 months. Five patients had serviceable hearing (AAO-HNS=A/B) before the initial treatment, none of these had preserved their serviceable hearing after the second treatment. Thirteen patients had good facial nerve function as evaluated by the House-Brackmann grade (HB≥2) before the initial treatment. None of the 13 patients with HB≥2 had worsened facial nerve function after the first GKS treatment. Of 10 patients with a complete dataset, one had worsened facial nerve function after the second GKS treatment. There were no new reported cases of trigeminal neuralgia. Overall, in 12 patients with complete follow-up: 5 tumors grew despite the second GKS treatment, 7 patients have tumor control so far. The combined dataset with patients previously reported in the literature included 88 patients: 86.4% accomplished tumor control after a mean follow-up of 43.2 months, 18.5% maintained functional hearing, 97.4% maintained good facial nerve function, 4.4% developed trigeminal neuralgia. All patients treated with salvage microsurgery accomplished tumor control 26 months after salvage, 73% had good facial nerve outcome. 

Conclusion: Salvage treatment of vestibular schwannoma is challenging. Repeat GKS treatment provide poorer tumor control than what can be expected from initial GKS treatment and poorer tumor control than salvage microsurgery, few patients maintain good hearing. The risk of facial and trigeminal nerve dysfunction however, is low - and lower than what can be expected from salvage microsurgery.

STUDY'S OBJECTIVE :

The use of upfront stereotactic radiosurgery (SRS) as the sole treatment for patients with more than five brain metastasis (BM) is still a matter of debate. However, this approach seems to gain momentum. In this study we report our results using SRS alone in patients with ≥ 5 BM as the initial treatment.  

METHODS :

103 patients underwent SRS between 2005 - 2016 for the treatment of ≥ 5 BM at our institution. 30% were male, 70% were female and the median age at SRS was 58. Primary histology was NSCLC in 57% of patients, breast cancer in 28%, melanoma in 12% and colorectal in 3%. All patients were divided by their KPS and by the RPA classification;12% were class 1, 82% class 2 and 6%class 3. 55 (53%) patients had previous irradiation for 1-4 BM. The mean number of treated BM was 7 (5 – 19) and the mean cumulative BM volume treated was 3.6 cc (0.06 – 27.7). We performed subgroups analyses based on these characteristics. Extracranial disease status prior to SRS was classified as stable in 28% versus progressive in 72%.

RESULTS :

Median follow-up after SRS was months 5.1 months (0- 57). Local control, based on RANO criteria, was achieved in 75% of patients. Median overall survival (OS) was 6 months. 72% developed new distant metastases. Multivariate analyses revealed that cumulative volume of treated BM (p=0.0128), stable extracranial disease status (p=0.00195) and RPA (p=0.0221) were independent prognostic factors for OS. Specifically, patients with a cumulative volume of treated BM £ 6 cc (OR: 2.54, p=0.006, IC95: 1.3 – 4.99) had better prognosis. The total number of BM had no impact on survival (p= 0.206). No factor was found to be predictor for local recurrence. RPA was also significant (p=0.0265) in terms of distant recurrence in multivariate analyses.

CONCLUSION :

This study suggests that SRS is a reasonable option for the management of patients with ≥5 brain metastases, especially when the cumulative treatment volume is £ 6 cc, as it was associated with favorable OS and local control.

Introduction: Linac-based Stereotactic Radiosurgery (SRS) for brain metastases may be influenced by the time interval between treatment preparation and delivery, related to risk of anatomical changes. We studied tumor position shifts and its relations to peritumoral volume edema changes over time, as seen on MRI.

Methods: Twenty-six patients who underwent SRS for brain metastases in our institution were included. We evaluated the occurrence of a tumor shift between the diagnostic MRI and radiotherapy planning MRI. For 42 brain metastases the tumor and peritumoral edema were delineated on the contrast enhanced T1weighted and FLAIR images of both the diagnostic MRI and planning MRI examinations. Center of Mass (CoM) shifts and tumor border were evaluated. We evaluated the influence of steroids on peritumoral edema and tumor volume and the correlation with CoM and tumor border changes.

Results: The median values of the CoM shifts and of the maximum distances between the tumor borders obtained from the diagnostic MRI and radiotherapy planning MRI were 1.3 mm (maximum shift of 5.0 mm) and 1.9 mm (maximum distance of 7.4 mm), respectively. We found significant correlations between the absolute change in edema volume and the tumor shift of the CoM (p<0.001) and tumor border  (p=0.040). Patients who received steroids did not only had a decrease in peritumoral edema, but also had a median decrease in tumor volume of 0.02cc while patients who did not receive steroids had a median increase of 0.06cc in tumor volume (p=0.035).

Conclusion: Our results show that large tumor shifts of brain metastases can occur over time. Because shifts may have a significant impact on the local dose coverage, we recommend minimizing the time between treatment preparation and delivery for Linac based SRS.

Objectives:  We seek to establish the feasibility of using SSRS allowing for spinal cord dose constraint relaxation as the primary management of MESCC in inoperable patients monitoring for radiation related toxicity and radiographic local control (LC).

Methods: Patients with MESCC in the thoracic spine deemed inoperable with no prior history of radiation at the site of interest were enrolled on this prospective Phase 1 single institution protocol.  Single fraction SSRS was delivered to a histology dependent prescription dose of 18 or 24 Gy.  Spinal cord constraint relaxation was performed from an initial allowable Dmax cohort of 10 Gy only if tumor progression occurred.  If the risk of radiation induced spinal cord myelopathy (RM) remained lower than the risk of tumor progression, then the cord Dmax was elevated in 2 Gy increments to a maximum of 16 Gy in the final cohort.  Patients were monitored every 3 months with follow-up visits, MRI scans and validated patient reported outcome surveys.

Results: Thirty-two patients enrolled on the trial of which 4, 12, 8 and 8 were in the 10 Gy, 12 Gy, 14 Gy and 16 Gy cord Dmax cohorts, respectively.  The most common histology was renal cell carcinoma (n=12).  The most common GTV prescription dose was 18 Gy (n=17) followed by 24 Gy (n=15).  The median age was 62.7 yrs (range 35-81 yrs).  At baseline, there were 10 sites with MESCC Grade 1B, 10 sites with Grade 1C, 9 sites with Grade 2, 2 sites with Grade 1A, and 1 site with Grade 3 epidural extension of disease.

The median overall survival of the cohort was 28.6 mos (95% CI 19.6, NR).  Of the 32 patients treated with SSRS, 4 were lost to follow-up without post-SSRS evaluation.  Of the remaining 28 patients, the 1-year LC was 80.5% and median LC was not met.  With a median clinical follow-up of 17.2 months (range 3-35.7 mo), there were no cases of RM.  In the cohort receiving a cord Dmax of 16 Gy, there were no cases of RM with a median follow-up of 15.4 mo (range 6.4-21.0 mo).

Conclusions: SSRS is a safe and effective tool in patients with MESCC.  Cord constraint relaxation may be considered in inoperable patients with MESCC.

Objectives:  Our institution utilizes conservative stereotactic radiosurgery (SRS) dosing and reduces dose by ~1-2 Gy for motor region metastases because of toxicity concerns, but this may compromise local control.  Here we evaluate freedom from progression (FFP) and risk of adverse radiation effect (ARE) after Gamma Knife SRS for primary motor cortex or frontoparietal brain metastases to consider dosing recommendations.

Methods:  Within a cohort of brain metastases with follow-up imaging treated at our institution 1998-2013 with SRS alone upfront or SRS for recurrence after prior radiotherapy, we compared actuarial FFP and risk of ARE for motor cortex or frontoparietal (“motor”) vs. non-motor frontal or parietal (“non-motor”) lesions.  FFP and ARE were measured from the date of SRS with censoring at last follow-up imaging.

Results:  Among 3920 brain metastases treated with SRS alone upfront or SRS for recurrence, 143 were coded as motor cortex, 75 frontoparietal, and 1727 non-motor frontal or parietal.  The median dose was 18 Gy for 218 “motor” metastases vs. 19 Gy for 1727 “non-motor” metastases (Wilcoxon rank-sum p < 0.0001).  Lesion quadratic mean diameter (QMD) was highly significantly associated with both FFP and ARE.  For newly-diagnosed motor vs. non-motor metastases with QMD <0.75 cm, 0.75-2.0 cm, or >2.0 cm, 1-year FFP probabilities with 95% confidence intervals were 95% (85-98%) vs. 94% (91-96%), 88% (73-95%) vs. 88% (83-91%), and 71% (39-89%) vs. 62% (48-72%) with corresponding 1-year ARE probabilities of 2% (0-11%) vs. 1% (0-3%), 0% vs. 5% (3-9%), and 8% (1-43%) vs. 0%.  For recurrent motor vs. non-motor metastases with QMD <0.75 cm, 0.75-2.0 cm, or >2.0 cm, 1-year FFP probabilities were 79% (60-89%) vs. 94% (89-97%), 63% (24-86%) vs. 80% (70-86%), and 50% (1-91%) vs. 70% (32-89%) with 1-year ARE probabilities of 0% vs. 1% (0-6%), 32% (12-69%) vs. 5% (2-11%), and 0% vs. 23% (8-52%).  For recurrent metastases, motor location was associated with significantly worse FFP (Cox proportional hazards p = 0.029 and hazard ratio = 2.07, stratified by QMD category), with a trend toward higher risk of ARE (p = 0.09; hazard ratio = 2.76).  At least 18.5 Gy was needed to give equivalent FFP for recurrent motor-region metastases.

Conclusions:  Local control was worse for recurrent motor-region vs. non-motor metastases, and disappointing for lesions > 2 cm.  Given the consequences of uncontrolled tumor and generally low risk of symptomatic ARE, we suggest radiosurgical management of motor-region metastases with at least 18.5 Gy, or hypofractionation.

Purpose: To evaluate the efficacy and safety of adrenal oligometastases treated with hypofractionated stereotactic radiosurgery in a single institution.

Methods and Materials: Between August 2013 and September 2016, we treated 34 adrenal metastases from 29 patients. All lesions except one was treated with linear accelerators, the residual 1 lesion was treated with robotic radiosurgery. Motion management strategy for linear accelerator based treatments was the internal tumor volume formation with 4 dimensional computed tomography and fiducial placement and tracking with the robotic radiosurgery. The prescription dose was 50 Gy in 5 fractions for all patients. The response evaluations were performed with PET-CT after 3 months and upper abdominal CT or MRI evert 3 months afterwards. Local control was defined as metabolic response of the first PET-CT imaging and no progression in size afterwards.

Results: The local control and survival analyses were performed on the patients who have at least 3 months of follow up. The majority of the patients had lung cancer primary. The median age of the patients were 58 (43-84), all but one patients were male. The median follow up was 24 months (2-40 months). In 1 patient local progression was detected after 11 montjs, the rest are still under local control. Six, 12 and 24 months local control rates were 100%, 96% and 96% respectively. Twelve patients were alive during the analysis. Overall survival rates of 6, 12 and 24 months were 84%, 57% and 10% respectively from the time of the first diagnosis and 93%, 83% and 69% from the diagnosis of adrenal metastases. None of the patients experienced any acute of late toxicity.

Conclusion: Stereotactic radiosurgery for adrenal oligometastases is a non-invasive, safe and effective local treatment strategy without any compromise in systemic therapy. It is being used as a standard approach in our hospital.  

Purpose: the aim of this study is to evaluate local control, overall survival and pattern of toxicity for oligometastatic patients with isolated lymphnode recurrence treated with Stereotactic Body Radiation Therapy (SBRT) and Volumetric Modulated Arcs (VMAT).

Materials and methods:  Seventy-one patients were treated for a total of 79 lesions from 2009 to 2015. Dose prescription was 45Gy in 6 daily fractions for all patients. Delivery was performed with VMAT and flattening filter free beams (FFFs). Dosimetric analysis was carried on the treatment plans while clinical outcome was assessed by means of actuarial analysis. Treatment response was assessed by means of the RECIST criteria. Toxicity was recorded according to the common toxicity criteria version 4.0.

Results:  with a  median follow of 1.5 years (range: 0.2-6.2), 45 patients (63.3%) had solitary metastasis while 26 (36.6%) had multiple lesions.  Primary tumour was located in the gastrointestinal tract in 59.1% of patients. Local control was achieved in 97.5% of the lesions with an actuarial rate at 1 year of 83%.  Progression free survival at 1 year was 86% while overall survival was 93%.  Only 2 patients (3%) developed grade 2 gastro-intestinal toxicity, no grade 3 cases were observed.

Conclusion: SBRT with VMAT technique and FFFs can be considered a well tolerated treatment with adequate clinical response in this group of patients confirming its appropriateness in the oligometastatic setting.

Objectives: Geometric accuracy of stereotactic MR imaging is one of major requirements for any successful intracranial stereotactic procedure. In this study we want to evaluate image distortion for three different Siemens MR scanners and also test and compare three different phantoms and methodology of measurement.

Methods and materials: Following three different phantoms are used for MR image distortion measurement in this study: 1) in-house made cylindrical Perspex phantom with 59 and 63 glass rods for axial and coronal inserts, respectively, 2) commercial PTGR phantom consisting 21 three-dimensional cross-hairs filled with contrast medium and 3) CIRS 3D Anthropomorphic Skull Phantom filled with matrix of 3 mm diameter rods spaced 1.5. cm apart. For all three phantoms rods or points with rigid geometrical positions are well imaged by different scanning MR protocols. All phantoms can be also fixed in the Leksell stereotactic frame and thus stereotactic imaging procedures can be reproduced following exactly the same steps as for a real patient, including also the stereotactic image definition in the Leksell GammaPlan. Three different Siemens MR scanners were measured in this study: 1.5 T Avanto, 1.5 T Symphony and 3T Skyra.  

Results: The measured distortions proved satisfactory accuracy precision for stereotactic localization for all three scanners. The mean radial (total in all three X, Y, Z coordinates) distortion for these MR scanners for a major imaging protocol (T1 weighted 3D imaging) measured by PTGR phantom were 0.8 mm, 1.1 mm and 1.1 mm, respectively. Similar results were obtained also by other two phantoms. Detailed comparison of all three phantoms and different methodology of measurement is work in progress.

Conclusion: Generally in all experiments with various phantoms there was detected dependence of the MR image distortion on the type of the MR scanner, slice orientation and imaging protocol. Image distortions are also property of each particular scanner, the worst distortion was observed for 3T Skyra. In the case of a clinical case with obvious or suspected MR image distortion we suggest always CT stereotactic imaging as an independent method.

This study was supported by Ministry of Health, Czech Republic - conceptual development of research organization (Nemocnice Na Homolce - NNH, 00023884)

Purpose: The accuracies of co-registration between the stereotactic images and the cone beam CT (CBCT) images of Gamma Knife Icon were assessed and parameters to enhance the accuracy were studied.

Materials and Methods: The accuracy of the co-registration procedure was evaluated by co-registering CBCT images taken at various situations to a reference CBCT. Then, stereotactic CT images of an anthropomorphic head phantom were obtained and co-registered to the CBCT images taken at arbitrarily moved positions. The coordinates of fifteen landmarks inside the phantom were measured. The co-registration accuracy between CBCT and stereotactic MR images were studied using the images of 41 patients who had CBCT images and stereotactic MR images together. Positions of anterior commissure (AC) and posterior commissure (PC) were measured both in the fiducial marker based system and the CBCT based system. In order to assess the effect of image distortion in the MR images, co-registration was performed with four different ranges in the patient head and their accuracies were compared.

Results: Co-registration between CBCT images showed deviations of 0.2 +/- 0.1 mm. After co-registration of stereotactic CT images to the CBCT images, the mean and standard deviation of the coordinate values in the left-right (x-axis), anteroposterior (y-axis), and craniocaudal (z-axis) direction was 0.0 +/- 0.3 mm, -0.3 +/- 0.2 mm, and 0.0 +/- 0.2 mm, respectively. The overall mean three dimensional deviation was 0.4 +/- 0.1 mm and it was not related with the distance from the center (p = 0.685). Co-registration of MR images to CBCT images were related with larger errors. The three dimensional difference of AC coordinates was 1.1 +/- 0.3 mm and that of PC was 0.9 +/- 0.3 mm. These differences were statistically correlated with the movement of coordinate systems calculated by the co-registration procedure. The target coverage ratios in the CBCT based system was lower than those in the fiducial marker based system, 92.2 +/- 7.1 % versus 97.9 +/- 1.7 % (p = 0.000). A region of co-registration (ROC) covering skull base area produced smaller co-registration error than other regions (p = 0.000).

Conclusion: Image co-registration error of GK Icon CBCT was similar with the registration error of the fiducial markers and the resolution of the images. In order to reduce the co-registration error, a portion of the MR images including skull base area is recommended for co-registration.

Purpose: Quality assurance results of Gamma Knife Icon^TM (GKI) for one year were analyzed to assess the accuracy of frameless Gamma Knife surgery and stability of the system.

Methods: Routine QA works on the radiological part such as absolute dose rate measurement, verification of couch movement, absolute comparison between calculated and measured dose distribution, end-to-end test of positional accuracy were performed. The image quality of the cone-beam CT (CBCT) of GKI was measured biweekly with Catphan^® 503 phantom. The positional accuracy of CBCT was assessed daily using four ball bearings on a manufacture provided phantom. The accuracy of the high definition motion management (HDMM) system was also assessed with a home-made device using a micrometer.

Results: Measured dose rates coincided with calculated values with mean error of 0.68 +/- 0.08% and measured half-life of cobalt 60 was 5.301 +/- 0.040 years. Gamma index pass rates with 3.0%/1.0mm criterion were greater than 99.1% for all single shots. A virtual target plan showed 98.6 +/- 1.6% gamma index pass rates at various locations in an anthropomorphic phantom. Positional accuracy of the radiation focus at the center and at an extreme position were 0.08 +/- 0.06mm and 0.06 +/- 0.05mm, respectively. For one year, the mean offset of the focus was 0.1 +/- 0.0 mm. The mechanical accuracy of the CBCT coordinate system was stable for one year (0.06 +/- 0.02mm deviation). For 2.5 mGy CTDI images, contrast to noise ratio was 1.13 +/- 0.3, and uniformity was 14.6 +/- 0.7%. For 6.3 mGy CTDI images, they were 1.78 +/- 0.08 and 14.5 +/- 0.7%, respectively. The slope of the HDMM values with respect to the real movement was 1.03 +/- 0.01 along the x-axis and 1.00 +/- 0.03 along the z-axis. The end-to-end test on the positional accuracy of the mask based irradiation was 0.9 +/- 0.3mm in an anthropomorphic phantom. Image co-registration showed 0.4 +/-0.1 mm deviation between CT and CBCT images and 1.0 +/- 0.3mm deviation between MR and CBCT images.

Conclusions: Overall accuracy of a millimeter order was verified for a frameless gamma knife surgery by various quality assurance works.

Objectives: Main purpose of this study was to perform an assessment of extracranial patient doses received during treatment on Leksell Gamma Knife (LGK) Perfexion. Results were compared with published data from previous gamma knife systems (model B and C).

Methods and materials: Extracranial doses were measured for 80 patients treated on the LGK Perfexion. Thermoluminescent dosimeters (TLDs) were positioned on patients at seven different following locations: eyes, thyroid, chest, abdomen, pelvis, knee and ankle. Measured data were evaluated and analyzed in terms of parameters that may affect extracranial doses. Following parameters were considered for analyses: prescribed dose, total irradiation time, distance between isocentre and position of TLDs, volume of prescribed isodose, total integral dose in target volume and total integral dose in brain.

Results: Mean extracranial doses delivered to patients in this study were: eyes (151.2 mGy), thyroid (10.1 mGy), chest (4.1 mGy), abdomen (1.2 mGy), pelvis (0.73 mGy), knee (0.30 mGy) and ankle (0.11 mGy). Significant dependence of extracranial doses was observed on total irradiation time, distance between isocentre and position of TLDs, volume of prescribed isodose, total integral dose in target volume and total integral dose in brain. In comparison with previous LGK models (B and C), there was observed a significant decrease of the extracranial doses in LGK Perfexion in the range of two to twenty times lower (depending on measured anatomical location).

Conclusion: Measured extracranial doses are generally very low and thus safe for treated patients. Observed measured doses are far below dose limits for deterministic effects. In comparison with previous LGK systems, Perfexion appears to be much safer with significantly lower extracranial doses.

Purpose: The Leksell Gamma Knife ICON introduces a mask fixation capability for patient setup and an optical tracking system for patient position monitoring. The purpose of this study is to evaluate the treatment target positioning accuracy at different stages of the mask-based Gamma Knife radiosurgery (GKRS) procedures.

Methods: CBCT imaging was applied to 11 patients who underwent multi-session mask-based GKRS and 7 patients with single session frame-based GKRS. A reference CBCT image set was obtained for each patient before the first session using the 6.3mGy dose setting. Setup CBCT images were acquired for each mask patient before each fraction using the 2.5mGy dose setting and repeated whenever a treatment pause was triggered by the motion surveillance system that tracks a fiducial marker attached to the patient nose. The treatment target positioning accuracy within the workflow of a mask-based GKRS is analyzed in terms of: 1) the registration change between the planning MR images and the reference CBCT images as obtained from the frame based GKRS with CBCT imaging;  2) the inter-fraction patient positioning accuracy determined from the registration changes for different fractions of a mask-based GKRS; 3) the intra-fraction patient positioning accuracy calculated as the difference between the registration changes for the initial setup scan and the intra-fraction CBCT scan following a treatment pause.

Results: The averaged values of the absolute translational changes in the X,Y,Z directions and the rotational changes along the X,Y,Z axes from the MR/CBCT image registrations for the 7 frame-based cases are 0.16mm,0.1mm,0.28mm and 0.41°,0.19°,0.12°, respectively. The corresponding numbers for the 41 inter-fraction registrations are 0.82mm,0.72mm,1.48mm,0.74°,0.78°, and 1.37°. The averaged values of the absolute differences in the translational and rotational changes between the 17 intra-fraction scans and the corresponding initial setup scans are 0.38mm,0.44mm,0.70mm and 1.12°,0.35°,0.86°, respectively. The smallest set of differences observed for the 17 treatment pauses are 0.01mm,0.03mm,0.03mm, and -0.16°,0.1°, -0.09°. This indicates minimum head movement despite some movement of the patient nose.

Conclusions: The registration changes between the planning MR images and the reference CBCT images are in general much smaller than those between the reference and the setup CBCT scans. The largest discrepancies in the CBCT registrations are usually seen in the Z direction for the translation changes or along the Z axis for the rotational changes. The optical tracking system may report a false alarm in some cases, owing to the inaccurate correlation between the nose reflector and the patient’s skull.

“A monolithic silicon detector array for small field dosimetry in Stereotactic Radiotherapy: DUO”

Introduction:  Stereotactic radiosurgery (SRS) commissioning and quality assurance (QA) are challenging as the technique uses extremely small, highly collimated photon beams, which require high geometric precision and dosimetric accuracy¹. Silicon diode arrays have a number of advantages including: real time operation (compared to film) and high spatial resolution and small size (compared to ionizing chambers) ², which make their implementation advantageous for SRS QA. This work aims to characterize the monolithic silicon diode array named “DUO” designed for stereotactic QA.

Methods: DUO is a silicon monolithic detector manufactured on a p-type substrate, designed by CMRP at UOW as shown in figure 1. The pixels are arranged in two cross linear arrays with 256 individually readout pixels for each arm. The pixel pitch is 0.2 mm and the overall detector area is 52 × 52 mm². DUO is placed on a 0.5 mm thick tissue equivalent printed circuit board. Characterization of DUO was performed, and used to measure 6MV beam profiles, percent depth dose and output factor for Elekta SRS cone collimators from 5 to 50 mm diameter on an Axesse ELEKTA Linear accelerator with Agility head. The DUO measurements were compared with results obtained with EBT3 films and Stereotactic field diode (SFD).

Results: The output factor agrees within 1% when compared with EBT3, and 2% with SFD for all cone sizes. The profiles of SRS cones show agreement in the FWHM and (20-80) % penumbra with EBT3 within 1% and 0.6 mm, respectively. The measured depth dose response agreed to within 1.5%, compared to EBT3 for depths beyond the build-up region.

Conclusions: DUO is a suitable detector for stereotactic dosimetry as it has excellent resolution 0.2 mm in a direction of steepest dose gradient, on time data analysis and provides both in-plan and cross-plan measurements. The good agreement with EBT3 films measurements confirms its accurate and precise data.

References

1. Alfonso, R., Andreo, P., Capote, R., Huq, M. S., Kilby, W., Kjäll, P. & Ullrich, W. (2008). A new formalism for reference dosimetry of small and nonstandard fields. Medical physics, 35(11), 5179-5186.
2. Wong, J. H. D., Knittel, T., Downes, S., Carolan, M., Lerch, M. L. F., Petasecca, M., & Rosenfeld, A. B. (2011). The use of a silicon strip detector dose magnifying glass in stereotactic radiotherapy QA and dosimetry. Medical physics, 38(3), 1226-1238.

OBJECTIVE To determine pain control and side effects after gamma knife radiosurgery (GKRS) for classical idiopathic trigeminal neuralgia (TN) with or without neurovascular compression (NVC).

METHODS This study included 47 patients with type 1 idiopathic TN and Barrow Neurological Institute (BNI) pain class IV or V that were treated with GKRS as a first-treatment modality between January 2005 and March 2015. A retrospective analysis was made of NVC status, pain control, side effects, recurrence, and cross-sectional area.

RESULTS During the median 21.5 months follow-up (range, 3–119 months), 36 (76.6%) patients showed good outcomes (improved to below BNI class IIIa). Twenty two patients did not show NVC (group A) and 25 had NVC present (group B). Good outcomes were not different in two groups (group A: 19/22; group B: 17/25)(p = 0.138). The numbers of cases in BNI class I or II, and recurrences were not significantly different between the two groups (p = 0.532, 0.786, respectively). The mean areas, measured at the target coordinate, were 8.64 ± 2.59 mm³ (range, 2.81–12.74 mm³) in non-deviated cases (n = 27) and 2.59 ± 1.68 mm³ (range, 0.80–5.29 mm³) in deviated (n = 10). Side effects were significantly more frequent in deviated cases (8/10; 80%) than in non-deviated (7/27; 25.9%) (p = 0.003).

CONCLUSION NVC is not a predictive factor for pain control after GKRS for the treatment of idiopathic TN. Side effects may occur more frequently in patients with NVC at the target coordinate when a dorsal root entry zone is used, but the subjective symptoms are not always bothersome.

Objective: CyberKnife stereotactic radiosurgery (SRS) is routinely used for treatment of trigeminal (TG) neuralgia by non-isometric, conformational high-dose administration to the corresponding nerve root-core. Given its proximity to the brain stem, nerve-core treatment optimization has required careful selection of dose and distance to achieve efficacy without compromising safety. We consider how the cisternal nerve length contributes to divergent outcomes in two treatment plans. 

Methods: We conducted a retrospective, single-institution review of 91 patients treated with CyberKnife for TG neuralgia in during 1/2005 to 5/2007 (Plan A, N = 47) and 1/2009 to 8/2013 (Plan B, N=44). Plan A (vs B) conservatively targeted the nerve core at 2 mm (vs 2.5 mm) and set the maximum brainstem dose to 10 Gy. The primary outcome was pain control at follow-up. Select secondary outcomes included changes in Barrow Neurologic Institute (BNI) pain and numbness scores. Length of the cisternal TG nerve, max brainstem dose, prior treatments, SRS treatment planning, and demographics were recorded. Univariate, multivariate and receiver operating characteristic curve analyses were performed.

Results: Patients were followed for a mean 23 months, and 62% were naïve to prior treatment. CK was administered in one fraction (92%) to a median and max dose of 60 and 75 Gy, respectively. Plans A and B exhibited treatment failure, durable improvement, and pain recurrence in 17.6%, 29.4%, and 52.9% vs 8.5%, 70.2%, 21.3%, respectively. BNI pain scores improved for 55.8 vs 78.7%, respectively while BNI numbness scores increased for 65.1% vs 38.3%, respectively. Nerve length, volume, and max brainstem dose were 7.48 vs 10.3 mm, 0.034 vs 0.036 cc, and were 33.7 and 48.8 Gy, respectively. Length significantly predicted any pain improvement under Plan A, but not Plan B, which was better predicted by treatment history (AUC = 0.82). Under Plan A, shorter nerves demonstrated a corresponding decrease in treated volume and under Plan B, longer nerves were less likely to have bothersome post-SRS numbness.

Conclusion: Our analysis demonstrates the ongoing challenge with predicting CyberKnife SRS treatment outcomes for TG neuralgia, and highlights how individualized consideration of TG nerve anatomy should assume an increased role in future patient selection. While patients who present with longer TG nerves may already be optimal candidates, those with shorter lengths and additional favorable history, will possibly benefit from a more aggressive dosing protocol.

Objective: Glossopharyngeal neuralgia (GPN) is a very rare pathology (0.7-0.8/100.000). Patients usually describe short episodes of paroxysmal pain, beginning at the base of the tongue and pharynx and irradiating towards the neck and the internal ear. We aim at reviewing our bicentric experience (Marseille and Lausanne University Hospital) in patients treated with Gamma Knife surgery (GKS) for idiopathic GPN.

Methods: Between 2003 and 2015, 21 patients were treated with 25 procedures. Eleven were women and 10 were men. All cases fulfilled the pharmaco-resistance criteria. Were analyzed patients with at least 6 months follow-up. GKS using a Gamma Knife (model B or C or Perfexion) was performed, based on both MRI and computer tomography targeting. A single 4-mm isocenter was positioned in the cisternal portion of the glossopharyngeal nerve at a mean distance of 14.6 +/- 3mm (range 9.3-23.5) anteriorly to the emergence of the nerve. The target was placed in the cisternal part for 2 and close to the glossopharyngeal meatus in 23 procedures. The mean maximal dose was 81.4+/-6.7 Gy (range 60-90). Three cases have had previous microvascular decompression (MVD), which was effective for 2, 8 and 13 years, respectively.

Results: The mean follow-up period was 5.2 +/- 3 years (range 0.9-12.1). At 3 months follow-up, 91.6% of the cases were pain free (BNI classes I to IIIA). At one year, 81.8% were still pain free (BNI classes I to IIIA), with 60% of them being BNI class IA. Recurrence appeared in 59.1%, in a mean time of 13.6 +/-10.4 months (range 3.1-36.6). Of them, 35% were controlled with medication and 25% (3 cases, 4 procedures) underwent a new radiosurgical procedure after 7, 17, 19 and 30 months, respectivelly. From these cases, two needed another open surgical procedure, with one undergoing a termocoagulation and another a neurotomy. At last follow-up, 16 cases (80%) were still pain free (BNI I-IIIA, 60% BNI IA). No complication was reported.

Conclusion: As in all cranial neuralgias, the reference technique remains MVD, as it addresses the cause (e.g. the neurovascular conflict). Radiosurgery is a valuable alternative, less invasive, with a very high rate of efficacy, in the absence of complications. The most important aspect is that the fifth nerve is easily identifiable, while the ninth nerve remains more challenging, so as its targeting. A multidisciplinary approach including a neurologist and neuroradiologist might be necessary, both for diagnosis and imaging purposes.

Introduction : A cohort of 207 patients affected by Trigeminal Neuralgia (TN) was treated by Cyberknife radiosurgery and regularly followed for at least 36 months .

Methods: Patients with typical TN and severe drug-resistant pain (grade IV-V on the Barrow Neurological Institute [BNI] scale and numerical rating scale [NSR] scores >5) were selected and treated with image-guided robotic radiosurgery (Cyberknife, Accuray Inc., Sunnyvale, Ca). The treatment was delivered in single session using a non-isocentric technique, a 5-mm collimator and the trigeminal node set.  A 5-6 mm segment of the nerve was contoured as treatment target. A 58/60 Gy dose was prescribed to the 80% isodose. Clinical re-evaluation was performed at 3, 6, 12, 18, 24 and 36 months. BNI and NRS scales have been used to assess the pain level before the treatment and during the follow-up. The BNI facial numbness scale was used to assess the development of sensory disturbances following the treatment.  

Results: 6 months after the procedure 191 out of 207 patients (92%) were pain-free (BNI I-IIIa). Mild hypoesthesia (grade II BNI) was reported by 14 (6.7%) and severe bothersome hypoesthesia (BNI grade IV) by 1 (0.5%). 11 patients reporting no improvement after treatment (5.3%) underwent a second procedure. Twenty-five out of 207 (12%) pain-free patients who experienced recurrent pain within 3 years from the treatment underwent retreatment with restoration of analgesia. Peak of recurrent pain was found 12 months after the first procedure (13 patients), while other 7 recurred after 18 months and 5 after 24 months. Overall, 36 patients underwent second treatment (17%). All of them developed stable pain relief while the rate of moderate and somewhat bothering (grade III BNI) versus to severe and very bothersome hypoesthesia (grade IV BNI) was, respectively, 13 and 6. Follow-up MR imaging showed focal contrast enhancing restricted over the target region in 16 patients without significant association with pain control or hypoesthesia. Brainstem edema was found in one case (the only patient developing BNI grade IV hypoesthesia). Actuarial pain control rate  after 6, 12, 24 ,36 months was, respectively, 92%, 90%, 77% and 71%. Rate of moderate and severe sensory disturbance was 9.2%.

Conclusions: Cyberknife radiosurgery targeting a 5-6 mm segment of the TN with a prescribed dose of 58-60 Gy is a safe and effective treatment for TN providing high pain control rates with an acceptable risk of sensory complications, which are typically found after re-irradiation.

Objective:

To analyze the long-term outcomes for medically refractory trigeminal neuralgia (TN) treated by Gamma Knife Radiosurgery (GKRS) in our institution.

Methods:

We included 309 patients treated consecutively by GKRS for medically refractory TN, between 2011 and 2014, in Lille University Hospital. Clinical baseline, treatment parameters and functional outcomes (using Barrow Neurological Institute (BNI) Pain Intensity Score) were reported.

Results:

The mean age was 62 y/o (range, 23 to 86). 13% of patients harbored a pre-GKRS hypoesthesia. A vascular conflict was reported in 230 (82%) patients. The mean dose was 90 Gy at the 100% isodose line, delivered at a mean distance to the root-entry zone of 9.6 mm. The mean dose to the first 10 mm³of the brainstem was 9.7 Gy. The mean delay before clinical improvement was 4.8 weeks. Patients with BNI

Conclusion 

GKRS is effective to treat medically refractory TN, with half of patient free of medication at last follow-up, with a very low morbidity.

Introduction : A cohort of 207 patients affected by Trigeminal Neuralgia (TN) was treated by Cyberknife radiosurgery and regularly followed for at least 36 months .

Methods: Patients with typical TN and severe drug-resistant pain (grade IV-V on the Barrow Neurological Institute [BNI] scale and numerical rating scale [NSR] scores >5) were selected and treated with image-guided robotic radiosurgery (Cyberknife, Accuray Inc., Sunnyvale, Ca). The treatment was delivered in single session using a non-isocentric technique, a 5-mm collimator and the trigeminal node set.  A 5-6 mm segment of the nerve was contoured as treatment target. A 58/60 Gy dose was prescribed to the 80% isodose. Clinical re-evaluation was performed at 3, 6, 12, 18, 24 and 36 months. BNI and NRS scales have been used to assess the pain level before the treatment and during the follow-up. The BNI facial numbness scale was used to assess the development of sensory disturbances following the treatment.  

Results: 6 months after the procedure 191 out of 207 patients (92%) were pain-free (BNI I-IIIa). Mild hypoesthesia (grade II BNI) was reported by 14 (6.7%) and severe bothersome hypoesthesia (BNI grade IV) by 1 (0.5%). 11 patients reporting no improvement after treatment (5.3%) underwent a second procedure. Twenty-five out of 207 (12%) pain-free patients who experienced recurrent pain within 3 years from the treatment underwent retreatment with restoration of analgesia. Peak of recurrent pain was found 12 months after the first procedure (13 patients), while other 7 recurred after 18 months and 5 after 24 months. Overall, 36 patients underwent second treatment (17%). All of them developed stable pain relief while the rate of moderate and somewhat bothering (grade III BNI) versus to severe and very bothersome hypoesthesia (grade IV BNI) was, respectively, 13 and 6. Follow-up MR imaging showed focal contrast enhancing restricted over the target region in 16 patients without significant association with pain control or hypoesthesia. Brainstem edema was found in one case (the only patient developing BNI grade IV hypoesthesia). Actuarial pain control rate  after 6, 12, 24 ,36 months was, respectively, 92%, 90%, 77% and 71%. Rate of moderate and severe sensory disturbance was 9.2%.

Conclusions: Cyberknife radiosurgery targeting a 5-6 mm segment of the TN with a prescribed dose of 58-60 Gy is a safe and effective treatment for TN providing high pain control rates with an acceptable risk of sensory complications, which are typically found after re-irradiation.

Introduction

During the 1990’s other companies (OUR) in Asia begun developing stereotactic gamma ray machines that rotated, needing a significant less amount of cobalt 60 sources (25 or 30) with the same dosimetry characteristics than the Gamma Knife units of that time. Recently the first fully automated rotating gamma ray unit called the Infiniâ by Masep (Schenzen, China) has been installed in El Salvador Central America and 204 patients have been treated so far. The purpose of the current communication is to evaluate the technical singularities of this system and to briefly review basic dosimetry and patient treatment process in order to better understand this relatively new intracranial stereotactic radiosurgery machine.

Technical characteristics.

There are six rows containing 5 cobalt sources each, that are focused at the isocenter rotating at one cycle per minute. The mechanical precision of the machine was 0.4 mm on average of all collimators tested (0.11 for the 4mm, 0.16 for the 8mm, 0.17 for the 14mm and 0.20 for the 18mm collimator) the offset registered at 15Gy on all axis was 0.0 as measured along the different profiles. Initial dose rate was 3.89 Gy per minute, dose at the center of the target at the anthropomorphic phantom was 0.97 (criteria 0.95-1.05) with a treated volume of 0.95 (criteria 0.75-1.25) with a minimum dose to target of 1.05 (criteria >0.90). The treatment bed is fully automatized as well as collimator change on all 4 diameters, it has a treatment range of 180,180,230mm in the X,Y,Z respectively and with dose sculpturing capabilities do to independent beam switch to off position every 5⁰.

Pathology and treatment characteristics.

Thus far 204 patients have been treated: 85 (42%) benign tumors, 44 (22%) with malignant tumors, 38 (19%) vascular lesions of diverse types and 34 (17%) with “functional” indications such as trigeminal neuralgia, tremor and pain.

Conclusions.

Infini® is a reliable machine with mechanical characteristics at least comparable to its better known predecessor, the Gammaknife®. By obtaining the same results with substantially less amount of cobalt sources makes the machine more efficient.  

Purpose: Synchrony Tracking System (STS) is used to predict and create online correlation model for respiratory induced internal target motion and external markers on the chest wall during treatment. In this study we aimed to evaluate the limits and uncertainties of correlation model system with respect to dosimetric deviations for several plans of Cyberknife radiation therapy system (CKS) by using end to end (E2E) test.

Methods: Isocentric plan with fiducial and E2E centroid error results in mm were assumed as reference. Next, isocentric and non-isocentric synchrony plans with phase difference (PD) and with no PD were achieved. The plans with no PD were only irradiated in synchrony phantom. On the other hand, plans with PD were irradiated in synchrony phantom and external markers were placed on the second respiratory phantom in order to create random phase difference during irradiations. Phase shift was obtained by changing the velocity of inhale and exhale periods of second respiratory phantom after linear or nonlinear correlation model generated by the system. E2E centroid error results were analyzed for all plans in order to find out the dosimetric deviations with respect to each other.

Results: The total targeting error (TTE) of isocentric fiducial plan from E2E test was 0.12 mm as reference result for no motion induced irradiation. TTE of isocentric and non-isocentric synchrony plans with no PD were 0.35 mm and 0.99 mm respectively for linear correlation model (LCM). TTE of isocentric and non-isocentric synchrony plans with PD were 3.95 mm, 4.17 mm respectively for LCM.

Conclusion: Incoherent internal target translation and orientation movement or global body inconsistency cause variations in dosimetric parameters such as source skin distance and depth. This can affect wider and inaccurate irradiation volume during non-isocentric synchrony irradiation even with no PD. Also suboptimal correlation models and predictions between internal target and external markers during PD induced respiratory motion cause much more inaccurate and dosimetric deviated irradiations. These results offer to create accurate, optimal and correct linear or nonlinear correlation models between internal target and external markers during clinical treatments with synchrony tracking system.

Objective: To report quality assurance (QA) procedures and results for the newly released InCise2™ Multileaf Collimator (MLC) installed on our CyberKnife M6™ System.

 Methods: Accuray recently released its second version of Multileaf Collimator (MLC), InCise™2 MLC, for clinical use on CyberKnife M6™ System. As one of the test sites, we not only did a thorough evaluation of InCise™2 MLC during commissioning, but also generated a system of QA procedures to ensure the MLC performance and short term and long term stability. Our MLC QA program includes daily, monthly and annual tests. A patient-specific QA for each case treated with MLC is also performed with a pinpoint chamber and film measurements. A Picket Fence test in a standard perch position is performed daily for a quick check of MLC positioning accuracy. For monthly QA, we performed a MLC Garden Fence test for leaf / bank positioning in standard (A/P) and clinically relevant non-standard positions. Total system and delivery accuracy with MLC is also assessed with End-to-End tests monthly. In annual QA, besides the tests in monthly QA, we also check the consistency of dosimetric parameters, including MLC leaf transmission, MLC beam profiles, output factors and tissue-phantom ratios, etc. Data for more than one and half years was analyzed to assess the MLC short term and long term stability.

 Results: No significant MLC positioning errors (>0.5mm) were observed with visual inspection from daily Picket Fence tests. Based on monthly Garden Fence tests, mean leaf position offsets were -0.05±0.13mm for X1 leaf bank and -0.08 ± 0.12mm for X2 leaf bank. No significant difference on mean leaf positioning offsets was observed between different leaf orientations. The change of mean leaf position offsets with time was less than 0.2mm, indicating a stable MLC positioning accuracy. Total system accuracy with MLC was 0.43±0.21mm as shown in the monthly End-to-End tests. All measurements for the dosimetric parameters were stable and well within the manufacturer specifications. Point dose measurements for more than 30 patients agreed with calculation within 3%, and all film measurements passed 3%/2mm Gamma evaluation for more than 95% of the measurement points.

 Conclusion: QA procedures were setup for the Incise™2 MLC installed on CyberKnife M6™. Our QA results indicate that the Incise™2 MLC has a good short term and long term stability.

Objective:

SBRT and SRS plans require highly conformal dose to the planning target volume (PTV) to spare adjacent normal tissues. Conformal dynamic arc technique can achieve very sharp dose falloff and efficeint delivery. It is routinely used in our institution for SBRT and SRS treatments. However, the conformality is usually not as good as inverse planned volumetric modulated arc therapy (VMAT) technique. Therefore, a simple revised PTV method is presented in this study to improve the conformality of dynamic arc plans.

Materials and Methods:

Twelve SBRT (target volume range 8 ~ 108cc) and thirteen brain SRS patients (target volume range 0.7 ~ 47cc) were selected in this study. SBRT plans were created using coplanar arcs and SRS plans were created using three to six non-coplanar arcs using Varian Eclipse treatment planning system (version 13.5). The linear accelerator (LINAC) was a Varian Truebeam equipped with 5mm multi-leaf collimator (MLC). For each case, an original plan was generated by fitting MLC to the PTV and normalized to “100% prescription dose covers 95% of PTV”. Then the 100% isodose volume was converted into a structure (100%IDV). At any radial angle ϴ from contour geometry center in each axial slice, the radius of PTV is defined as R1(ϴ) and radius of 100%IDV is defined as R2(ϴ). The distance between them is calculated as D(ϴ) = R2(ϴ)-R1(ϴ). Then a new PTV radius is calculated as R’(ϴ) = R1(ϴ) - D(ϴ). A revised PTV contour was then generated, and a new plan was developed by fitting the MLC to the revised PTV. Paddick conformity index (PCI) and gradient index(GI) were compared for each case between the original and revised plans. GI is defined as the ratio of 50%IDV/100%IDV.

Results:

For the SBRT plans, the revised plan improved PCI by 10% on average (PCI increased from 0.79 ± 0.05 to 0.87 ± 0.04), and no change on GI (4.11 ± 0.25 vs 4.10 ± 0.33 ). For SRS plans, both original and revised plan were renormalized to 99% PTV coverage. PCI improved by 11% on average (PCI increased from 0.73 ± 0.06 to 0.81 ± 0.07), and no change on GI (2.94 ± 0.51 vs. 2.92 ± 0.55 ).

Conclusion:

Revised PTV method is simple and effective to improve conformality of conformal dynamic arc plans for most SBRT and SRS patients, except for some very irregular concave shape target.

Objectives

Today’s procedures to perform patient-specific quality assurance (QA) for multi-leaf collimator (MLC) based CyberKnife stereotactic radiotherapy are generally based on dose measurements for every treatment plan. This work describes an alternative approach, which makes use of Monte Carlo (MC) techniques to independently calculate dose distributions.

Methods

A vendor-independent MC based dose calculation framework using the EGSnrc MC simulation code system was developed and validated for CyberKnife MLC treatments. Each beam of the treatment plan with its corresponding MLC field shape compiled to an EGS++ geometry is simulated using EGSnrc with a phase space based beam model, creating a pre-patient phase space file for each beam. The framework then samples particles from all beams (weighted according to the treatment plan) and performs energy deposition scoring in the patient CT using DOSXYZnrc. MC calculations were validated against measured depth dose curves (DD) and dose profiles in water in units of cGy per Monitor Unit for several MLC fields. The framework was then validated by comparing MC calculated dose to film measurement for a clinical prostate treatment plan applied to a solid water phantom. Further, for another clinical prostate case, MC dose calculations were compared to TPS dose calculation using a finite size pencil beam algorithm. Both film and TPS dose were compared to MC calculations by gamma analysis with a 10% (global) low‑dose threshold using 2% (global) dose difference / 1 mm distance to agreement criteria.

Results

Measured and MC calculated dose profiles and DD in water agreed within 3% and 1 mm. MC calculations showed good agreement to the film measurement with 93.4% of voxels passing gamma evaluation. Comparing TPS dose to MC calculated dose showed a gamma pass rate of 93.1% despite dose differences of up to ±10% (global) near bones and metal fiducials.

Conclusion

An alternative to measurements for patient-specific QA was successfully developed and validated against measurements.

Objective:

To define the skull volume for Gamma Knife radiosurgery (GKRS) dosimetry, there are two possibilities in Icon^TM: manual skull scaling measurements and image-based skull definition. We evaluate if dose calculation significantly differs depending on these 2 techniques.

Methods

We included 48 GKRS treatments performed from July 2016 to January 2017 in Lausanne University Hospital, distributed among four groups: convexity lesions (18), parenchymal deep-seated lesions (13), vestibular schwannomas (11) and trigeminal neuralgias (6). For each treatment, we recorded the beam-on time (min), target volume coverage (%), prescription isodose volume (cm³) and maximal dose (Gy) to the nearest organ at risk if relevant (e.g. cochlea) according to each of the 2 skull definition techniques. The image-based contours were performed using CT-scan, which provides more reproducible segmentation than MRI. During this period, the mean dose rate was 3.72 Gy/min.

Results

Between the manual measures and image-based contours, the beam on time varied of +1.27% (corresponding to 27 sec) (p <0.001 Wilcoxon signed rank test), the target volume coverage varied of -0.04% (0.6 mm³) (p=0.09), the prescription isodose volume varied of -0.07% (12 mm³) (p=0.22) and the maximal dose to organ at risk varied of -0.37% (0.016 Gy) (p=0.6). Using image-based contours, the mean increase of dose delivery was theoritically of +1.674 Gy per treatment (if no blocked sectors).

Conclusion

The beam on time is significantly increased using image-based contours in comparison with manual skull measurements, resulting in an increase of the total dose delivery per treatment. The other dosimetric parameters did not differ significantly.

Introduction

The concept of "de novo" arteriovenous malformations (AVM) remains a poorly understood condition. It has been documented, observing follow-up radiological studies, that AVMs are dynamic lesions. However, this concept is exceptional in the literature after radiosurgery treatment.

Material and methods

108 pediatric patients have been treated for their AVM with Gamma Knife stereotactic radiosurgery. Seven have developed "de novo" AVM on the periphery of the treatment performed in a retrospective study.

Results and conclusions

The mean follow-up period after the radiosurgical treatment of the 7 AVMs with "de novo" AVM was 14.5 years, diagnosing these entities at 5 years of Gamma Knife in 71% of the patients. The mean age at treatment was 7.45 years (3.4-13.6), 5 males and 2 females. About the location of AVMs, one was profound and the rest superficial, with eloquence in 5 of them. The Spetzler-Martin grades were II, III and IV in 2, 3 and 2 patients respectively. In the first treatment the mean volumen of the nidus is 2.6 cc and the mean radiosurgical dose is 20 Gy. Only one patient had hemorrhage at 7 years of radiosurgery. All but one patient has been treated subsequently. The mean volumen of "de novo" AVMs is 8.2 cc. This study corroborates the possibility of developing "de novo" AVM predominantly in childhood, after achieving obliteration of the volumen treated with radiosurgery.

Objective:

Evaluate the obliteration rate of arteriovenous malformations (AVM) and clinical results after radiosurgery in pediatric patients with and without previous embolization.

Method:

Of the hundred patients undergoing radiosurgery with Gamma Knife, thirty patients had been embolized prior to radiosurgery. All patients have a minimim follow-up of three years. The rate of obliteration, hemorrhage and clinical outcome after radiosurgery were analyzed between the two groups under study.

Results and conclusion:

Nidus obliteration was achieved in 63% of patients in the non-embolized AVM group (Group A) and in 59% of the previously embolized AVM group (Group B). During the first three years after radiosurgery, three patients in group A suffered hemorrhage, without sequelae; and one patient in group B, with clinical worsening. In group A, three more patients presented hemorrhage at 7,8 (death) and 10 years of treatment, and in group B one patient at 10 years, passing away. The ratio of the obliteration rate to the AVM volume has been in group A: < 3 cc: 70%, 3-10 cc: 63% and > 10 cc: 20%; and in group B: 73%,60% and 0% for the same volumes, respectively. The clinical situation regarding treatment has worsened in five patients in group A and in two patients in group B, remaining stable or improving in the rest. The relation of appearance of "de novo" AVM between group A and B has been 5 to 1, producing delayed hemorrhage in two of these patients in group A.

Partially embolized AVM in pediatric patients are susceptible to successful treatment with Gamma Knife, without significant differences in the obliteration rate between the two groups.

Object.

We retrospectively analyzed our experience with time-staged Gamma Knife radiosurgery (GKRS) for large arteriovenous malformation (AVM)s.

Methods.

Between 1998 and 2016, 835 patients were treated with GKRS for cerebral AVMs. Among the 835 patients, 113 patients had large AVMs with volumes larger than 14 cm³.

After exclusion of patients who followed up less than 3 years, a total of 89 patients were enrolled in this study. All patients were treated with a planned time-staged GKRS.

The median age was 32 years (range, 4-60). The most common presentation was seizure (n=22). Fifty-four patients underwent a second GKRS and 11 patients underwent a third GKRS.

The median volume was 22 cm³ (range, 14-59) at first GKRS, 11 cm³ (range, 0.4-33.8) at second GKRS and 2.5 cm³ (range, 0.3-17.4) at third GKRS.

The median marginal dose was 13 Gy at first GKRS, 12Gy at second GKRS and 16 Gy at third GKRS. Nidus obliteration of AVMs was confirmed using transfemoral cerebral angiography (TFCA).

The median clinical follow-up after first GKRS was 76 months.

Results.

Among the 82 patients who underwent a TFCA following first GKRS, complete nidus obliterations were obtained in 12 patients.

Fifty-four patients underwent a second GKRS with a post-GKRS median interval of 39 months. Thirty-three of 54 patients had a 3-year follow-up TFCA.

Complete nidus obliteration was confirmed in 18 of 33 patients. Eleven of 33 patients underwent a third procedure with a median interval of 39 months.

Eight of 11 patients underwent a TFCA at 3 years after GKRS. Five of 8 patients had a complete nidus obliteration.

Therefore, the overall nidus obliteration rate in this study was 60% (35 of 58 eligible patients).

During follow-up period, a hemorrhage developed in 15 patients (17%) including 5 cases of major bleeding and 10 cases of minor bleeding.

Only one patient died of intracerebral hemorrhage after GKRS. Symptomatic adverse radiation effects were detected in 12 (13%) of 89 patients.

However, permanent morbidity rate was 1% at the last follow-up. No patients in this study developed a delayed cyst formation following GKRS.

Conclusions.

The management of large AVMs is still challenging. In this study, a time-staged GKRS for large AVMs shows a relative high obliteration rate and a low complication rate.

Although long-term follow-up and repeat GKRS are needed to achieve complete obliteration, a time-staged GKRS might be an effective and safe treatment option in the management of large AVMs. 

Objectives

The management of non-hemorrhagic arteriovenous malformations (AVMs) remains a greatly debated topic, even more so since the ARUBA trial. We report on the long-term outcomes and obliteration rates after Gamma Knife radiosurgery (GKRS) treatment for such AVMs.

Methods

We retrospectively analyze data from a series of 101 patients harboring unruptured AVMs treated by GKRS as first-line treatment in our University Hospital between April 2004 and September 2011. Inclusion criteria were: age > 18 years old, no clinical history suggestive of acute hemorrhage, no bleeding stigma on the pre-treatment MRI and/or CT scan, minimal follow-up > 3 years. Exclusion criteria were: pediatric population, volume-staged GKRS, prior embolisation or surgical management.

Results

Mean age at presentation was 38.9 years (range 19-64). The main initial symptoms were: epilepsy in 50% of patients and headache in 27% of patients. Mean follow-up was 9.9 years (range 3-13 years). Median target volume was 1.9 cm³ (IQR, 0.8-3.3 cm³), median Spetzler-Martin grade: 2 (IQR, 1 to 2), median Pollock-Flickinger score: 1.067 (IQR, 0.8-1.3), median Virginia score: 1 (IQR, 1 to 2). Median treatment dose was 24 Gy at 50%. 17 patients benefited from a second GKRS after 3 years follow-up without obliteration. Hemorrhage during the post-treatment follow-up was reported in 11 patients (annual risk of 1.1%). Transient post-GKRS morbidity was reported in 4.9% with persistent neurological deficit in 2.9% of patients. The obliteration rate was 72%, based on cerebral angiography and/or MRI. At last follow-up 95% of patients had a mRS ≤1 and 88% of patients were free of symptoms. Concerning epilepsy, 84% of patients were seizure-free at last follow-up.

Conclusions

GKRS as first-line treatment for unruptured cerebral AVMs achieves high obliteration rates while maintaining patient autonomy and even improving their clinical symptoms (e.g. epilepsy).

Introduction: It has been shown that higher margin dose given to the AVM nidus correlates with higher AVM obliteration rates, with the maximum obliteration rate observed at 25 gy. However, it is not always possible to prescribe such higher margin doses as these can lead to higher rates of adverse radiation effects (AREs) especially in large volume AVMs.  Theoretical dose modeling indicated that it is possible to expand specific isodose volumes (e.g. 70% of the maximum dose) closer to prescription isodose without a substantial increase in 12 gy volume. The purpose of this study was evaluate if a higher dose covering a higher volume of AVM nidus correlates with improved nidus obliteration.

Methods: For this retrospective, single-institution analysis, the authors reviewed their experience in 43 patients who had Gamma Knife surgery between 2007 and 2013. Patients with multiple AVMs, prior Gamma Knife treatment, planed stage treated AVMs, AVMs larger than 10 cm³, or lacking at least three years of follow-up data were excluded. Nidus volume, margin doses, 12 Gy volume, and absolute doses covering various percentages of nidus volume were determined using Leksell GammaPlan^® software. The average marginal dose was 20 gy (range: 16-23 gy), and the average AVM nidus volume was 3.44 cm³ (range: 0.2579-9.13 cm³).  AVM obliteration was confirmed by MRI and/or Angiography.  Comparisons between groups were performed using Mann-Whitney U Test and Pearson’s c² test of independence.

Results: Of the 43 patients, a total of 37 (86%) patients had complete obliteration of the AVM over an average follow-up time of 36 months (range: 11-79 months). Our analysis indicated that higher obliteration rates were achieved in patients who received greater than 23 gy to more than 75% of nidus volume (p = 0.036). Patents who received greater than 23 gy to 75% of nidus volume did not have increased risk of ARE (p=0.82). Similarly obliteration rates were significantly better (p=0.025) for patients with expanded 70% isodose volume (more than 43% nidus covered with 70% isodose line). Expanded 70% isodose volume was not associated with increased risk of AREs (p=0.56).

Conclusion: Higher dosage (23 gy or higher) delivered to at least 75% AVM nidus volume or expanded 70% isodose volume are associated with higher obliteration rates following AVM radiosurgery without increasing the likelihood of AREs. 

Background

Cavernous malformations (CMs) natural history has remained unclear several years. This lack of knowledge has made treatment decisions difficult. Indeed, the use of stereotactic radiosurgery (SRS) is nowadays controversial. The purpose of this paper is to analyze factors implicated in bleeding  and adverse radiation effects in patients treated with Gamma Knife Radiosurgery (GKRS).

Methods

The authors reviewed ninety-five cavernous malformations prospective database, 57 women and 38 men, underwent GKRS for high-surgical-risk CMs. The median malformation volume was 1570mm3. The median tumor margin dose was 11,87 Gy and the mean tumor maximum dose was 19,56 Gy. Statistical analysis was performed with IBM SPSS software version 20.0 and R Core Team software version 2013.

Results

Ninety-five cavernous were situated in: brainstem (64), thalamus/basal ganglia (12) and hemispheric eloquent areas (19). All patients had experienced at least one symptomatic bleeding before treatment. Imaging follow-up after SRS revealed lesion volume regression in 39 CMs. The pretreatment annual hemorrhage rate was 3,06% compared with 1,4% during first 3 years latency interval, and 0,16% thereafter (p-value = 0.004). Four patients developed new location-dependent neurological deficits and three patients had edema-related headache after radiosurgery. All of them presented full recovery.

In spite of univariate analysis didn't findstatistically significant association between marginal dose and post-treatment bleeding, Multiple Regression Model, with Akaike`s Information Criterion including sex, showed finally statistically significant association between lower marginal dose and bleeding (p-value=0,03) and no association between target volume and bleeding (p-value=0,13).

Volume and coverture dose show a weak negative correlation with adverse radiation effects (AREs) (ρ = -0.260, p-value = 0.011). Using Multiple Regression analysis, AREs couldn´t be related to prescribed radiation dose, brainstem location nor multiple pre-treatment hemorrhages.

In Multiple Regression Model, size reduction seemed to relate with patients age (p-value: 0,042) and maximum dose levels  (p-value: 0,028).

Conclusions

Marginal dose was the only variable that showed statistically high significant influence on bleeding rate in multiple regression analysis (p-value = 0.030). A lower marginal dose seems to be related with post-treatment rebleeding.

Highly conformal GKRS with lower margin dose average (margin dose average: 11.3 Gy) could be related to safety of this treatment in recent series. In spite of descriptive analysis showed a trend towards relating adverse radiations effects with higher margin dose, we didn't find statistically significant association in our study. This result could be due to the scarce number of AREs and post-treatment bleeding events in our series.

Objective: For some stereotactic radiosurgery (SRS) treatment sites, delivering a homogeneous dose to the target volume has been associated with reduced adverse effects and toxicities. Previous work has indicated that lowering the beam energy from the standard megavoltage range to the orthovoltage range results in the improvement of various plan quality metrics in SRS. Modulation of beam energy, even across a small field, may further serve to create homogeneous, conformal dose distributions in orthovoltage SRS delivery. The objective of this work was to build and characterize an orthovoltage energy-modulated SRS system in order to achieve dose distributions approaching rectangular functions.

Methods: Previous work described the design of cone-based, filtered orthovoltage energy-modulated SRS system using mathematical optimization techniques and Monte Carlo simulations. This system uses a NIST-traceable 250 kVp irradiator with an auxiliary cone/filter system that was constructed using both in-house machining techniques as well as the commissioning of an outside prototyping firm capable of binderjetting, an additive manufacturing technique, using tungsten. Four nondivergent cone collimators were constructed from free machining brass to be used optionally along with four epoxy-infiltrated bonded tungsten filters of variable thicknesses. Radiochromic EBT3 film measurements of the system were performed in a custom, thin-window water phantom to compare dose distributions derived from open cones to those from filtered cones. Films were scanned with a prototype laser densitometry system.

Results: Measured beam profiles showed that the modulated beams could more closely approach rectangular function dose profiles compared to the open cones, based on quantification of profile flatness and penumbra. This result remained consistent for all four cones and for the three different depths tested. This result confirms previous computational work indicating the benefit of orthovoltage energy fluence modulation.

Conclusion: Both computational and measurement results showed that filtered orthovoltage SRS prototypes are able to achieve dose distributions approaching rectangular function distributions at depth, therefore establishing the feasibility and efficacy of a full future treatment platform that relies on fluence modulation in the orthovoltage energy range to manipulate resulting dose distributions.

Metal artifact reduction with Dual Energy CT for Gamma Knife Radiosurgery in pacemaker patients

Purpose: Pacemaker patients with brain metastases who are to undergo Gamma Knife (GK) radiosurgery are typically simulated with CT as they are not usually MRI compatible. Since the patients are framed and fixated with either titanium (Ti) or aluminum (Al) pins to the skull, metal induced artifacts by the Ti or Al pins are unavoidable. This can result in obscuration and suboptimal evaluation of brain lesions. Conventional single-energy CT (SECT) with metal artifact reduction (MAR) and newer dual energy CT (DECT) with high keV mono-energy imaging are promising techniques for reduction of metal artifacts. The purpose of this study is to assess the effectiveness of these two methods in metal artifact reduction to enable optimal treatment planning in pacemaker patients undergoing GK radiosurgery.

Methods and Materials: An anthropomorphic head phantom (Radiology Support Devices) fixated with one pair of Ti and one pair of Al pins was scanned with SECT and DECT on GE HD 750 64-slice CT, and SECT on Philips large bore 16-slice CT. Both metal artifact reduction techniques of MAR and DECT on GE, and OMAR (orthopedic MAR) on Philips were compared. The scan and reconstruction parameters are as follows (in the order of [kVp/mA/Collimation (mm)/Pitch/CTDIVol (mGy)/Thickness (mm)/Recon]): Philips LB16 [120/482/16x0.75/0.438/65.4/1/UC & UC+OMAR]; GE HD750 [120/220/32x0.625/0.531/65.2/1.25/Standard & Standard+MAR]; and GE HD750 [(80,140]/375/32x0.625/0.531/67.0/1.25/140keV DECT].

Results: Metal artifact reduction when using Ti fixation was best achieved by DECT followed by MAR, then SECT and lastly OMAR (DECT> MAR > SECT > OMAR). In case of Al fixation DECT was again the best, SECT was as good as MAR, and OMAR was the worst (DECT > SECT = MAR, > OMAR). More metal artifacts were introduced by Ti than Al pins. For Al pins, while DECT could reduce artifacts, OMAR introduced more artifacts than SECT, and MAR was not effective for artifact reduction.

Conclusions: There were fewer artifacts introduced by Al than Ti. DECT was more effective than MAR in reducing metal artifacts for both Ti and Al. OMAR should not be used in CT acquired for simulation in GK pacemaker patients. Pacemaker patients undergoing GK radiosurgery may be best served by frame fixation using Al pins in combination with DECT because of the least amount of metal artifact generated, and therefore likely better visualization of metastases in the underlying brain parenchyma.

Objectives: Purpose of this study was to make a dosimetry audit after Leksell Gamma Knife (LGK) Co-60 sources reload. Comparison of TMR10 and Convolution calculation algorithms in the Leksell GammaPlan (LGP) was also made by measurement in heterogeneous anthropomorphic phantom.   

Methods and materials: Dosimetry audits were performed by two institutions: 1) National Radiation Protection Institute, Prague, Czech Republic (NRPI) (on-site audit) and 2) The MD Anderson Dosimetry Laboratory (MDADL), Houston, USA (postal audit). Measurements were made in three different phantoms: 1) ABS Elekta plastic spherical phantom, 2) adapted anthropomorphic Alderson Rando phantom and 3) Stereotactic Radiosurgery Head phantom from MDADL. Calibration of the LGK unit was verified in the Elekta phantom by two independent PTW 31010 ion chambers. Altogether six measurements in two different orientations were made. Then comparison between planned and delivered dose in anthropomorphic Alderson Rando phantom was done for a test treatment plan calculated by both TMR10 and Convolution algorithms. Mean dose in two PTW 31010 ion chambers positioned close to heterogeneous area in the phantom was measured. All these measurements were performed on-site by NRPI medical physicists. Additionally, irradiation of MDADL head phantom was made. The head phantom consisted of imaging insert with nylon ball target to obtain imaging for treatment planning and then the insert was exchanged to a dosimetry insert with TLDs and Gafchromic films for dosimetry measurements. After on-site irradiation, the phantom was sent back to MDADL for an evaluation.      

Results: Deviation between measured and reported calibration dose rate in the ABS plastic phantom was 0.7 %. Deviation in mean dose measured by ion chamber positioned within target volume in heterogeneous anthropomorphic head phantom was -1.1 % and 2.5 % for TMR10 and Convolution algorithms, respectively. Results from MDADL are not yet ready at the time of writing this abstract. Convolution algorithm generally calculated always longer irradiation times by 2-3 % on average compared to TMR10. This fact was also supported by measurement results. Based on results from this study the statement that Convolution algorithm provides more accurate calculation is not supported.      

Conclusion: To perform dosimetry independent audit after a new LGK installation or after Co-60 source reload belongs to a good medical physics practice. Both on-site and postal audits were used in this study. Very reasonable agreement was observed for reported calibration dose rate. Also measurements for target volume mean dose in anthropomorphic heterogeneous phantom for both algorithms showed reasonable results. 

The purpose of this work was to assess the dosimetric accuracy of SRS in the UK for linac-based (LB), Tomotherapy (TT), Cyberknife (CK) and Gamma Knife (GK) radiosurgery. 

The methodology developed for this assessment employed an anthropomorphic phantom with realistic tissue densities. The simulated scenario featured an irregular 8 cm³ lesion located anterior to the brainstem. The case was presented to 26 UK centres who developed 28 treatment plans: 16 LB, 7 GK, 4 CK and 1 TT. An end-to-end test was conducted for each plan, incorporating immobilisation, scanning, planning and treatment delivery following the local protocol. Previously characterised dosimeters (EBT-XD film and alanine pellets) were placed inside the phantom to measure absolute dose inside the target and brainstem, for comparison with Treatment Planning System (TPS) predictions. Film measurements were compared to TPS dose planes using gamma-analysis. 

Alanine measurements showed that LB (including TT) had the largest range in percentage difference to the TPS of 5.2% (-1.3% to +3.9%) with a mean of +0.5%. CK had a range of 2.6% (+1.4% to +4%), with the highest mean difference in comparison to the other platforms (+2.5%). GK showed the smallest range at 2.4% (-0.8% to +1.5%) being comparable to that of CK, with the smallest mean percentage difference (+0.4%) comparable to that of LB. Similar trends were observed in the brainstem with alanine measurements showing a range from -1% to +3.6% (mean= +1.3%), 0% to +1.9% (mean= +0.9%) and -1.1% to +0.9% (mean= +0.1%), for LB, CK and GK respectively. Film measurements showed comparable results between centres, regardless of the platform used. For 3%-2 mm Local-gamma, all except two films showed passing rates above 75%.  For 5%-1 mm Global-gamma, all except 2 films showed passing rates above 90%. Large variations were observed in prescription practices, delivery techniques and plan quality.

This audit enabled a comparison of all UK centres in terms of the dosimetric accuracy achieved during treatment delivery. The LB group showed the largest variations in agreement to the TPS, related to more heterogeneous practices within the group, compared to smaller variations seen in CK, and more consistent practices seen in GK. Good overall agreement with the TPS was observed with 2 centres falling above 3.6% (2sd). The results suggest that good agreement with predicted dose distributions is achievable by all modalities. The variations in prescription practices, techniques and plan quality highlight the need for standardisation in SRS practice.

Purpose

Shielding considerations for both, primary and secondary radiation must be revised when switching from a conventional linear accelerator to a Cyberknife (CK). In this context two important parameters to be investigated are the direction distribution of the primary radiation and the modulation factor (MF) of treatment plans, which is linked to secondary radiation. This work assesses the impact of a novel multi leaf collimator (MLC) on the required radiation shielding of the CK analyzing the clinically applied treatment beams.

Methods

For 364 patients (163 fixed cones, 180 iris collimator, 21 MLC) the delivered beams were projected onto the boundaries of a 9.5x5.9x3.9 m³ vault in order to obtain a monitor unit (MU)-weighted spatial distribution of the treatment beams. This was accomplished by a previously developed framework, which extracts the beam directions, the corresponding number of MUs and the employed collimator of all treatment beams delivered by the CK. In addition, the total delivered MUs and the prescribed dose were stored in the database for each treatment beam. Using this information, the MF, defined as the ratio of delivered MUs divided by the prescribed dose, was assessed for each treatment plan and used collimator type.

Results

Compared to all considered treatments, the MLC beams delivered slightly more MUs to the wall to the left and right of the patient (13.0% vs of 12.0%) and the floor (73.5% vs 71.0%). None of the analyzed MLC beams hit the wall at the patient’s feet in contrast to 5.3% of MUs for all treatments. However, comparing the MLC beams only with all extra-crainal beams, the differences for the wall at the patient’s feet vanish.

The mean MF for the treatments without MLC is 8.5 MU/cGy, while the mean MF for the treatment using the MLC is 6.3 MU/cGy. Taking only the extra-cranial treatments into account, the mean MF for the treatment with and without using the MLC is 7.1 MU/cGy and 10.4 MU/cGy, respectively.

Conclusion

The MLC, which was mainly used for extra-cranial treatments, has a minor influence on the direction distribution of the treatment beams compared with the differences arising from treating cranial or extra-cranial entities. If the MLC is employed, the larger field sizes available together with the segmented delivery lead to a reduction of the MF for the considered cases.

Purpose: The virtual cone is a standardized MLC control point sequence designed to mimic the dose distribution of a 4 mm cone. It was developed to facilitate rapid treatment planning and quality assurance for treatment of small lesions using an MLC equipped linear accelerator. We report on the first patient treatment using the virtual cone.

Methods: The virtual cone is comprised of an MLC control point sequence having multiple non-coplanar arcs and a dose rate that varies as a function of gantry angle. The patient was immobilized using an open-face thermoplastic mask and received a treatment planning CT scan having 1 mm slice spacing. The target volume, a solitary metastasis defined using MR images, was 0.021 cm³ (diameter 3.4 mm). The virtual cone sequence was imported and dose was calculated on a 1 mm grid using Eclipse AAA version 13.6 (Varian Medical Systems, Palo Alto, CA). The plan was normalized such that 99% of the target volume received at least 20 Gy. Measurement using radiochromic film in a phantom was done prior to treatment to confirm dosimetric and geometric accuracy.  The patient was localized prior to treatment using orthogonal kV images followed by cone-beam CT and was monitored during radiation delivery using optical surface monitoring.

Results: The ratio of the measured dose to the calculated dose was 1.02. The 2-dimensional magnitude of the difference between the measured and calculated dose distributions was 0.13 mm. After correction for the systematic offset, the maximum distance between the measured and calculated prescription isodose contour was 0.26 mm. These results were consistent with preclinical end-to-end studies using an anthropomorphic skull phantom.  The maximum dose was 143.9%, corresponding to a prescription isodose line of 69%. Dose calculation dominated the treatment planning time, requiring 22 min on a cluster of calculation servers, whereas importing the plan and placing isocenter was less than 5 min. Delivery of the plan for the QA measurement required 12 min. 

Conclusion:  The virtual cone is an efficient technique for treatment of small spherical dose target volumes. Because the control point sequence is standardized, patient specific QA measurements will not be necessary in routine clinical use.  Integration of the virtual cone directly into the treatment planning system along with improved dose calculation efficiency will make planning using this technique extremely efficient, potentially allowing for same day treatment. Planned clinical applications of the virtual cone include trigeminal neuralgia and treatment of essential tremor.

Background: Large acoustic schwannomas (VS) are tumors with a diameter of >3 cm or a volume >8 cm³. They are usually considered not amenable to conventional, single-fraction stereotactic radiosurgery (SRS) because a large section of the brainstem may be exposed to harmful doses of radiation. The problem can be avoided by using a hypofractionated irradiation scheme.

Methods: Twenty-five patients with VS of >8 cm³ (range 8-24 cm³, median 9.5 cm³) were treated from August 2007 to January 2016 at the CyberKnife center at the University of Messina, Italy. All patients underwent 3-fraction radiosurgery with a total dose ranging 18-19.5 Gy.

Results: Follow-up period ranged from 12 to 106 months (median 48 months). Radiological growth control was achieved in 88% of cases: 11 tumors (44%) displayed no relevant size variation; 11 (44%) showed a >50% volume shrinkage. Three patients (12%) needed salvage tumor resection. No patient presented worsening of trigeminal sensory disturbances or facial nerve dysfunction. No patient had serviceable hearing before treatment. Five patients (20%) developed hydrocephalus after treatment or showed deterioration of the preoperative ventricular enlargement with new neurological symptoms. All 5 patients were treated with ventriculo-peritoneal shunts with full recovery. Actuarial progression-free survival rates at 1 year and 5 years were 97% and 83%, respectively.

Conclusions: The current and other published results suggest that hypofractionation may extend the indication of SRS to VS larger than 8 cm³. The tumor control rate is not significantly different from smaller tumors. Hydrocephalus is the only complication recorded in our series. This complication is related to preoperative ventricular size and can be easily and effectively treated with minor surgery. Even though the limited experience and short follow-up currently available in the literature do not provide sufficient support for widespread application of hypofractionated SRS in younger patients with large VS, further studies on the issue are warranted.

Objective:

Acoustic neuromas (AN) have a close proximity to radiosensitive critical structures (i.e. inner ear, cranial nerves). Single fraction radiosurgical (SRS) treatment requires precise definition of the target volume as well as the surrounding critical structures (CS). Therefore, various imaging modalities are available with cranial computed tomography (CT) and different magnetic resonance imaging (MRI) sequences. The aim of this study was to evaluate to what extent this affects the definition of gross tumor volume (GTV) and identification of CS. There are no clear guidelines for the imaging modality to be used for delineation of GTV and CS in SRS, so far.

Methods:

The GTV, anterior-posterior and transverse diameter of the internal acoustic canal (IAC) were conducted with a variety of different image modalities (plain CT window widths, T1 TFE 3D, T2 TSE 2D, T1 FFE 3D, T2 DRIVE 3D) in 73 patients and compared with each other. Furthermore, the identification rate of CS (trigeminal nerve, labyrinthine artery) was evaluated. The obtained GTVs were compared to the respective T1 TFE 3D volume for every individual. Significance in volume changes were verified using Wilcoxon signed-rank test.

Results:

The average deviation from the GTV obtained in T1 TFE 3D imaging was 43.4 ± 23.1% (for CT brain window), 18.7% ± 21.3 (T2 TSE 2D), 27.7% ± 17.9 (T2 DRIVE 3D) and 15.5 ± 10.4 (T1 FFE 3D). All deviations were significant (p <0.0001). The anterior-posterior and transverse diameter of IAC showed significant (p<0.0061) differences between T1 TFE 3D imaging and CT brain and bone window and T2 weighted MRI. The rate of inner ear identification was 99% in T2-weighted sequences and 100% in CT bone window. The identification rate of the trigeminal nerve was 97% in T2-weighted MRI compared to 34% in CT brain window.

Conclusion:

Various imaging modalities are available for the definition of the GTV and CS. However, differences in the predefined GTV (up to 43% in CT and 18% in T2-MRI respectively) significantly depend on the image modalities in use. How far these differences affect dosimetry remains unclear and should be part of further investigations.

Objective. The treatment strategy for the patients affected by sporadic vestibular schwannomas is recently changing and the number of patients which undergo radiosurgery as a primary treatment modality for such lesions is continuously increasing.

While the question about the best treatment is waiting for more definitive results, the attention is actually focusing on the hearing function sparing.

The aim of the present study is to investigate about the potential advantages of multisession radiosurgery(mRS) compared to single session radiosurgery(sRS) in terms of hearing preservation.

Patients and methods. The present is an “ad interim” analysis of a prospective randomized clinical trial.

The primary end-point of the study is the difference in term of hearing preservation between patients treated with mRS and sRS because of a sporadic acoustic neuroma.

The conditions for patient eligibility are:

- sporadic acoustic neuroma diagnosis.

- Age≥18 years old

- KPS≥70

- Serviceable hearing(class A/B from the AAOHNS classification)

- Written consent

All the enrolled patients are clinically, radiologically and audiometrically evaluated.

The volumetric analysis of the tumor is always performed.

Results. At the time of the present analysis 52 patients have undergone a radiosurgical treatment. The mean follow-up period is 27 months. Twenty-nine patients had a sRS and 25 had a three fraction mRS.

In term of hearing preservation, no differences were observed between the two groups. A significant difference was observed between the patients that were class A at the treatment time compared to class B. Indeed, only 10% of the class A patients compared to 53% of the class B patients lost the serviceable auditory function during the follow-up period.

The volumetric analysis showed that most part of the tumors(86%) had a shrinkage or a stabilization. The 14% of the patients experienced a tumor enlargement, at least in the first two follow-up MRIs. No significant differences were observed between the patients treated with mRS and sRS.

Conclusions. At our knowledge, the present clinical trial is the first one comparing two different radiosurgical regimens in terms of hearing sparing.

While we are waiting for the definitive results of the present study, the preliminary ones suggest that mRS has no advantages compared to sRS in terms of hearing preservation.

Similarly to surgical studies, the data suggest that the better is the auditory function at the moment of the treatment, the more probable is the hearing preservation.

The volumetric analysis confirms the good tumor control rate.

Introduction:

Anatomical parameters of the petrous bone and tumor could perhaps predict clinical findings in patients with vestibular schwannoma. This information might be important to determine the optimal parameters of the dosimetric planning and to estimate hearing outcome.

Materiel & Methods:

We have retrospectively analyzed anatomical and clinical parameters on a series of 656 patients treated radiosurgically for a vestibular schwannoma. Hearing status was evaluated with the Gardner-Robertson classification. The high-resolution CT at bone windows and high-resolution MRI-T2 and MRI-T1gadolinium have been used to analyze linear and volumetric measurements of the IAC and the part of the tumor located into the IAC. We estimate the erosion of IAC bone by comparison of the ipsi- and controlateral IAC volume. We correlate the IAC volume and linear measurements with the % of IAC occupied by the tumor. We compare the bony erosion of the IAC and % of intrameatal tumor with hearing level before Gamma Knife irradiation and hearing status at last follow-up.  

Results:

The pre-treatment anatomical and audiological data of all patients were analyzed in LGP 10.0, and the patients were followed prospectively for tumor control and hearing outcome. The volume of IAC at the side of the schwannoma was increased in comparison with the controlateral side in 87% of cases, and was increased by more than 120% of the controlateral IAC volume in 66% of cases. We found an extremely significant association (p<0.0001) between the ratio Vol IACipsi/Vol IACcontra and the volume of tumor located into the IAC, as well as the % of tumor volume located into the IAC. The pre-radiosurgical ipsilateral hearing level (GR grade and useful/not useful hearing status) was significantly associated (p=0.034 and p=0.0032, respectively) with the volume and the % of tumor volume located into the IAC. Therefore, the ratio Vol IACipsi/Vol IACcontra and the % of tumor volume located into the IAC are parameters related to hearing loss and could be used in the decision process for treatment or wait-and-scan.

Conclusions:

The IAC can be eroded by the intracanalicular part of the schwannoma. Volumetric parameters of the IAC and intrameatal volume of the tumor are significantly related to patients hearing status and some cut-off of these parameters can be used to decide when the wait-and-scan attitude must be stopped in favor of radiosurgical treatment.

OBJECTIVE Vestibular schwannomas (VSs) represent a common indication of Gamma Knife surgery (GKS). While most studies focus on long-term morbidity and adverse radiation effects (AREs), none describe the acute clinical AREs that might appear on a short-term basis. These types of events are investigated, and their incidence, type, and outcomes are reported in the present paper. METHODS The included patients were treated between July 2010 and March 2016, underwent at least 6 months of follow-up, and presented with disabling symptom during the first 6 months after GKS that affected their quality of life. The timing of appearance, as well as type of main symptom and outcome, were noted. The prescribed dose was 12 Gy at the margin. RESULTS Thirty-five (22%) of 159 patients who fulfilled inclusion criteria had acute clinical AREs. The mean followup period was 30 months (range 6-49.2 months). The mean time of appearance was 37.9 days (median 31 days; range 3-110 days). In patients with de novo symptoms, more frequent were vertigo (n = 4; 11.4%) and gait disturbance (n = 3; 8.6%). The exacerbation of a preexisting symptom was more frequently related to hearing loss (n = 10; 28.6%), followed by gait disturbance (n = 7; 20%) and vertigo (n = 3, 8.6%). In the univariate logistic regression analysis, the following factors were statistically significant: age (p = 0.002; odds ratio [OR] 0.96), hearing at baseline by Gardner-Robertson (GR) class (p = 0.006; OR 0.21), pure tone average at baseline (p = 0.006; OR 0.97), and Koos at baseline (Koos Grade I used as reference) (for Koos Grade II, OR 0.17 and p = 0.002; for Koos Grade III, OR 0.42 and p = 0.05). Fractional polynomial regression analysis showed a nonlinear relationship between the outcome and the radiation dose rate (minimum reached at a cutoff of 2.5 Gy/minute) and the maximal vestibular dose (maximum reached at a cutoff of 8 Gy). The clinical acute AREs disappeared in 32 (91.4%) patients during the first 6 months after appearance. Permanent and somewhat disabling morbidity was found in 3 (1.9% from the whole series): 1 each with complete hearing loss (GR Class I before and V after), hemifacial spasm (persistent but alleviated), and dysgeusia. CONCLUSIONS Acute effects after radiosurgery for VS are not rare.  In most cases, none of these effects are permanent, and they will ultimately improve or disappear with steroid therapy. Permanent AREs remain very rare.

Introduction

Previous publications suggest that vestibular schwannomas (VS) which grow rapidly prior to stereotactic radiosurgery (SRS) are more likely to continue growing after treatment. However this is based either on potentially inaccurate tumour length or small patient numbers.  In University Hospitals Bristol (UHBristol) the majority of patients with VS have treatment after documented growth.  The aim of this study is to accurately describe the volumetric changes pre- and post-SRS of growing VS and investigate if the tumour’s growth kinetics are predictive of these changes.

Method

Patient cohort consisted of patients with VS treated with SRS who had a MRI scan within 24 months prior to SRS demonstrating growth and minimum 2 years MRI follow-up.  To ensure volumetric accuracy, MRI scans >1.5mm slice interval were excluded.  Scan closest to 1 year pre-treatment, treatment day and all post treatment scans were imported into Oncentra Planning System (Elekta, Stockholm).  VS was contoured to establish tumour volume at all timepoints (pre, day 0, and 1/2/3 year post SRS). The rate of volume change of the VS was calculated per month and statistical analysis utilised Pearson Correlation.

Results

60 consecutive patients with VS were treated at UHBristol with SRS 10/2013-12/2014.  3 were excluded due to lack of growth on scans 4-8 months prior to treatment and 14 due to inadequate pre-treatment scans. 43 patients had 173 scans contoured. Tumours were treated on Perfexion Gamma Knife (Elekta, Stockholm) with mean 12.3Gy to 50% isodose with 99% coverage, conformity index 0.83 and gradient index 2.85.

At 4-22 months pre-treatment the mean VS volume was 1.36cc; at SRS 1.94cc; at 1 year 1.19cc; and 2 year 1.08cc.  Pre-treatment tumours grew mean +8.2%/month [+0.29-+36.8%/month].  38 patients had adequate scans at 1 year and 43 at 2 year post-SRS.  At 1 year post treatment overall rate of growth was -2.9%/month [-6.7-+3.3%/month]: 86.8% smaller; 5.3% stable; 7.9% grew but rate of growth reduced from +9.3%/month to +2.2%/month.  At 2 years compared to treatment day mean growth was -1.9%/month [-3.4-+2.3%/month]: 88.4% shrank; 4.7% grew then stabilised; 4.7% initially shrank then grew; 2.3% continued to grow at a slower rate.

There was no correlation between rate of growth prior to after SRS, nor between rate of shrinkage at 1 year to 2 year.

Conclusion

Despite growth pre-treatment, most VS shrink in the first 2 years post SRS.  The rate of pre-treatment growth does not predict post treatment continued growth or degree of shrinkage.

Background: Facial nerve schwannomas are rare tumors and account for less than 2% of intracranial neurinomas, despite being the most common tumors of the facial nerve. The optimal management is currently under debate and includes observation, microsurgical resection, radiosurgery (RS) and fractionated radiotherapy. RS might be a valuable alternative, as a minimally invasive technique, in symptomatic patients and/or presenting with tumor growth.

Methods: We review our series of 4 consecutive cases, treated with Gamma Knife surgery (GKS), during the period July 2010 and January 2017. Clinical and dosimetric parameters were assessed. GKS was performed in all cases using the Leksell Gamma Knife Perfexion.

Results: The mean age at the time of GKS was 44 years (range 34-56). The mean follow-up period was 42 months (range 12-60). The first symptom was facial palsy in 2 (50%) cases and hemifacial spasm in 2 (50%). All had a facial palsy at baseline, one with House-Brackmann (HB) II, 2 with HB III and one with HB VI. The mean target volume at the time of GKS was 0.360 cc (range 0.030-0.638) and the mean prescription isodose volume was 0.462 cc (range 0.052-0.805). The mean maximal dose delivered was 12 Gy at the 50% isodose line. The mean dose received by the cochlea was 3.8 Gy (range 0.1-10). The mean number of isocenters was 6.5 (range 1-10). One patient benefited from a staged-volume GKS. At last follow-up, facial palsy remained stable in 2 cases (one HB II and one HB III), and improved in 2 (one from HB III to HB II; one from HB VI to HB II). Regarding hemifacial spasm, both patients presenting one at baseline had a decrease in its frequency and intensity.

Conclusion: In our experience, RS and particularly GKS appear to be an appropriate therapeutic option in the management of these tumors. However, it should be also accompanied by a rehabilitation program, in collaboration with specialized colleagues, so as to give the patients the best chances for recovery. Radiosurgery remains a minimally invasive technique and with a small risk of functional decline, which should be putted in balance with the patient’s baseline clinical status and tumor characteristics.