A Statistical Analysis of the Effect of Different Embolization Materials on Gamma Knife Arteriovenous Malformation Dose Distributions
Ozlem Dagli, Erkan Bostanci, O. Hakan Emmez, Fatih Ekinci, Emrah Celtikci
11 A Statistical Analysis of the Effect of Different Embolization Materials on Gamma Knife Arteriovenous Malformation Dose Distributions Ozlem DAGLI , Erkan BOSTANCI , O. Hakan EMMEZ , Fatih EKINCI and Emrah ÇELTİKÇİ [email protected], [email protected], [email protected], [email protected], [email protected] Gazi University, Faculty of Medicine, Department of Neurosurgery Gamma Knife Unit, Ankara, Turkey Ankara University, Computer Engineering Department, Golbasi, Ankara, Turkey Gazi University, Faculty of Medicine, Department of Neurosurgery, Ankara, Turkey Gazi University, Physics Department, Besevler, Ankara, Turkey, Gazi University, Faculty of Medicine, Department of Neurosurgery, Ankara, Turkey
Abstract
Gamma Knife Radiosurgery (GKRS) is a treatment choice in newly diagnosed AVMs and residual AVMs following endovascular interventions. Aim of this study is to determine if commercially available liquid embolic agents reduce the radiation dose to the target due to the attenuation of the Co beam. Doses accumulated by three different embolization materials, namely: Onyx, cyanoacrylate and polyvinyl alcohol (PVA) were analysed. A collimator helmet size of 8mm was employed in a Monte-Carlo simulation implemented in Geant4 simulation toolkit. Obtained dose accumulations were evaluated and results demonstrated statistically significant differences in the dosimetries calculated.
Keywords : Arteriovenous Malformation, Gamma Knife, Embolization, Radiosurgery, Statistical evaluation
1. Introduction
Brain disorders which are not accessible for conventional surgery can be treated using Leksell Gamma Knife ® which is a radio-surgical device. This device allows the treatment of different cases such as benign or malignant tumours, arteriovenous malformations and trigeminal neuralgia can be treated using a technique known as stereotactical radiosurgery. By using this widely adopted technique all around the world, high and sharp dose gradients can be administered in order to minimize the side-effect of the radiation on critical brain tissues in a single accurate session (10). Arteriovenous malformations (AVMs) are defects in the circulatory system, especially in the brain area, which are due to abnormal tangling of arteries and veins. AVMs have a risk of between 2% and 4% annually and are usually discovered in situations of intracerebral hemorrhage (1,7). The combined rate of morbidity and mortality are given as 50% (2). A number of treatment strategies can be followed in order to eliminate the risk by obliterating the AVM, namely microsurgery, radiosurgery and embolization (4,8,12,13,16). Gamma Knife radiosurgery (GKRS) has gained prominent role for a long time (since 1950s) in the treatment of cerebral AVMs. This technique has several advantages which can be specified in terms of obliteration rate, complication rate. Comparing with microsurgery, GKRS is also a minimally invasive operation type (6,9,14). Obliteration rates with the Gamma Knife are related to prescription dose; at a prescription dose of 25 Gy, AVM obliteration occurs about 60%–80% of the time. The obliteration rate decreases as the prescription dose to the nidus is decreased or the nidus volume is increased (9). In cases of large volume AVMs as shown in Figure 1, options of multiple or incremental treatments should be considered. Endovascular embolization followed by GKRS is one of these treatment methods. The purpose of this approach is to decrease blood flow rate and reduce the volume of the nidus to ≤10 cm and hence to improve the efficiency of the GKRS treatment. For nidus sizes less than 3 cm, embolization is not required for reducing the volume before radiosurgery. More successful results can be achieved for larger AVMs with correctly administered embolization. Conventionally, N butylcyanoacrylate (NBCA), a fast-polymerizingliquid adhesive, was often used as an embolization agent. Studies have shown that it is hard to work with this agent due to the unpredictability of the flow and the speed of polymerization (3,5,11,15,17). Recently, Onyx (Ethylene vinyl alcohol), which is a less adhesive and more slowly polymerizing agent, has gained prominence. This agent is an ethylene vinyl alcohol copolymer dissolved in dimethyl sulfoxide (18,19). (a) (b) (c) (d) Figure 1:
Female patient presented with seizure and she received pre-GKS embolization followed by GKS (Post-embolization angiography (anterior (a) and lateral (b)), angiography (anterior (c) and lateral (d)) with GKS This study investigated the dose perturbation that can occur due to use of different embolization materials, namely: onyx, cyanoacrylate and polyvinyl alcohol (PVA). Using the Geant4 simulation system, dose accumulations were computed for brain and the embolization materials. We compared the differences and aimed to decide which embolization materials show similar accumulation results to that of the human brain. The rest of the paper is structured as follows: Section 2 describes the methods used in the study, followed by Section 3 where the results are analyzed. The paper draws conclusions in Section 4 where the most suitable embolization material is identified.
2. Material and method 2.1 Material
In the study, a Gamma Knife device, brain phantom was simulated using Geant4. The Gamma Knife Device (GK) used was the 4C model with 201 sources. Phantoms are materials that are equivalent to human tissue and which are used to examine dose distributions in tissue. A large percentage of the human body is made up of water, and the atomic number Z of muscles and soft tissues are almost equivalent to water. Because of this, the basic dose distributions are usually made in solid water phantom, as they are reproducible, are very close to the radiation absorption-scattering properties of muscles and soft tissues and can be measured in three dimensions. Since using liquid water as phantom is not always practical, more useful solid water phantoms have been developed. Ideally, mass density, number of electrons per mass and effective atomic number must be equal to water, in order to ensure that the material to be used is tissue or water equivalent. In this study, a 160 mm diameter brain phantom was used. The embolization agents are selected as Onyx, Cyanoacrylate and Polyvinyl alcohol. The material compositions of these agents obtained from ICRP (20) are shown in Table 1.
Table 1
Embolization materials used in the analysis
Component Name Chemical Formula Density Onyx ( C H O-C H ) x Cyanoacrylate C H NO Polyvinyl alcohol ( C H O) x Geant4 toolkit allows the simulation of particles passing through the matter (21). While developing the system, first, the structural element need to be defined. These elements include the materials, volumes and their locations in the simulation environment. Next stage is the definition of the relevant physics elements such as which particles and physical processes will be used. The mechanisms defining the formation of physical particles also need to be specified. Then, the geometrical structure is displayed along with the particle traces after the simulation is run and the simulation outputs are collected.
The aim here was to identify the gamma beam profile and to reveal the most convenient and useful structure for obtaining the beam profile using the existing beam-collimator materials. The geometric structure of the simulation environment is as follows: the patient is defined as a target with 90x90x90 mm detector plate placed in a room of size 4 m x 4 m with indoor air. The detector plate consists of voxels (small cubes) and is used to store the dose accumulations.
The modeling processes and physical calculations can be described as follows: The Gamma knife device consists of a semi-spherical iron-plated unit with 201 Co-60 sources. In the simulation, a single source was used from 201 different directions and the doses are collected on the target. Gamma rays came from different directions and focused on the target as shown in Figure 2. Calculations were made for 8mm collimator helmet size.
Figure 2.
Focusing the beams from a single source from 201 directions. The patient source distance was 401mm, and two gamma rays of 1.17 MeV and 1.33 MeV were released from Co-60 decay. When modeling, brain phantom of 160 mm in diameter, was used for dosimetric measurements. The material of the phantom was then changed with the embolization agents and comparisons were made. The simulations were performed 4 times with 10 histories.
3. Results
This study investigated the dosimetric effects of three embolization agents used in arteriovenous malformations. Figures 3-5 show individual comparisons of the embolization agents with the brain tissue. The relative dose averages were normalized in range 0-1.
Figure 3.
A comparison of radial doses in onyx embolization material and brain from the 8 mm collimator helmet of the Leksell Gamma Knife
Figure 4.
A comparison of radial doses in Polivinil alcohol embolization material and brain from the 8 mm collimator helmet of the Leksell Gamma Knife R e l a t i v e d o s e ( a . u ) Voxel x dimention (mm)
BRAINONYX R e l a t i v e d o s e ( a . u ) Voxel x dimention (mm)
BRAINPVA
Figure 5.
A comparison of radial doses in cyanoacrylate embolization material and brain from the 8 mm collimator helmet of the Leksell Gamma Knife A statistical evaluation was performed using t-test on the simulation results. As shown in Table 2, Onyx and Cyanoacrylate have resulted in statistically significant differences (2.9433>1.7531, p=0.0050 and 3.8673>1.7531, p=0.0008 for Onyx and Cyanoacrylate, respectively) in the maximum dose regions when compared to the brain tissue. On the other hand, results for the PVA was not found to be statistically significant (0.4905<1.7531, p=0.3154).
Table 2
Statistical evaluation of dose differences of embolization materials with brain with t-test
Onyx Cyanaocrylate PVA t Stat
4. Conclusion
There are studies claiming that the embolization materials result in the scattering of rays (Shielding effect), though it is shown that the effect is minimal. Literature also shows a dose decrease of 10-15% when Cyanoacrylate is used for X-rays and 0.01%-0.2% when Onyx is used with gamma rays (23). Our results are compatible with previous findings in that the difference between onyx and the brain tissue is lower than that of cyanoacrylate and brain. It is also interesting to see that dose accumulations for PVA was not found to be statistically significant with the dose accumulations for brain.
References
1. Barrow DL, Reisner A: Natural history of intracranial aneurysms and vascular malformations. Clin Neurosurg 40: 3–39, 1993. 2. van Beijnum J, Lovelock CE, Cordonnier C, Rothwell PM, Klijn CJM, Al-Shahi Salman R, SIVMS Steering Committee and the Oxford Vascular Study: Outcome after spontaneous and arteriovenous malformation-related intracerebral haemorrhage: population-based studies. Brain 132: 537–543, 2009. 3. Chun DH, Kim MS, Kim ST, Paeng SH, Jeong HW, Lee WH: Embolization with gamma knife radiosurgery of giant intracranial arteriovenous malformations. Turkish -0.200.20.40.60.811.2 0 20 40 60 80 100 R e l a t i v e d o s e ( a . u ) Voxel x dimention (mm)