Pain due to bone metastasis is one of the prevalent complications of cancer [
In the treatment of metastatic spinal bone tumor, combination therapy consisting of vertebroplasty and radiotherapy has been performed [
Bone cement used in percutaneous vertebroplasty contains about 30% barium, a radio-opaque agent [
Four water-equivalent phantoms (Toughwater Phantom, 1-cm thickness, 1.018 g/cm3, Kyoto Kagaku CO., Kyoto, Japan) were prepared by tracing the shape of the anthropomorphic phantom (RANDO phantom, Kyoto Kagaku CO., Kyoto, Japan) at the second lumbar vertebra (between number 23 and 24) according to ICRU Report number 44.
With two water-equivalent phantoms, the area corresponding to the second lumbar vertebral body was scooped out as shown in Figure
The middle phantom is a number 23 RANDO phantom. The upper and lower phantoms are water-equivalent phantoms modeled after the anthropomorphic phantom, and a test hole corresponding to the second lumbar vertebral body has been hollowed out. Testing sample such as PMMA and Toughwater Phantom is set in the hole.
The water-equivalent phantom modeled after the second lumbar vertebra was subjected to CT (Asteion 4; Toshiba Medical Systems, Tokyo, Japan) (imaging condition: 120 kV; 250 mA; 5 mm slice thickness). CT image data was transferred to a radiation treatment planning system (Eclipse Treatment Planning System, version 8.1; Varian Medical Systems, CA, USA) to plan radiotherapy. An isocenter for radiotherapy was set inside the spinal cord on a target slice, and a 10 cm × 10 cm irradiation field was established around the isocenter for posterior single field irradiation. The reference point for radiotherapy was set at the isocenter. Using 4-MV X-ray, 1 Gy of radiation was irradiated. Using the treatment planning system with clinical setting, a dose distribution map was prepared for the water-equivalent phantoms with and without bone cement.
Additionally, we imported the bulk density (1.303 g/cm3) of the bone cement into the treatment planning system to reduce error from high CT values, and a dose distribution map was again prepared for the water-equivalent phantoms with and without bone cement.
A film (EDR2; Eastman Kodak Co., Rochester, NY, USA) was placed between two water-equivalent phantoms (Figure
A film (EDR2) is located between two water-equivalent phantoms. They were firmly set in the anthropomorphic phantoms, and radiation was irradiated.
Effects of bone cement on irradiation as assessed by the film dosimetry were investigated by comparing dose distribution between water-equivalent phantoms with and without bone cement. Dose distribution as calculated by RTPS was compared to the dose distribution measured by the film dosimetry.
For the water-equivalent phantom without bone cement, dose distribution calculated using RTPS both with clinical setting and with importing the bulk density of the bone cement was even and undistorted (Figure
The calculated dose distribution maps using RTPS and the dose distribution maps measured by film dosimetry. (a) with the phantom without bone cement, the calculated dose distribution was undistorted. (b) with the phantom with bone cement, using RTPS with clinical setting, a depression in the isodose curve was seen for areas corresponding to inside the cement and the ventral side of the cement. (c) with the phantom with bone cement, using RTPS with importing the bulk density of the bone cement, dose inside the cement was the same as in the surrounding area. A depression in the isodose curve was seen in the ventral side of the cement however, the distortion with importing the bulk density of the bone cement (c) was less than with clinical setting (b). (d) with the phantom without bone cement, no distortion and no difference in the isodose curve were seen between calculated dose distribution (a) and measured dose distribution (d). (e) with the phantom with bone cement, a dose distribution map drawn by the film-based dose distribution analysis system. Dose increases were seen within and around bone cement. No significant dose decrease was seen behind bone cement.
For the water-equivalent phantom with bone cement, dose distribution calculated with clinical setting was distorted for the areas corresponding to inside the cement and on the ventral side of the cement. In other words, dose inside the cement was lower than in the surrounding area. Dose for the ventral side (the side through which radiation had passed through the cement) was lower than that for the dorsal side (the side through which radiation had yet to pass through the cement). A dose distribution map showed that dose was lower for the area after the cement (Figure
For the water-equivalent phantom with bone cement, dose distribution calculated with importing the bulk density of the bone cement was not distorted for the areas corresponding to inside the cement but distorted on the ventral side of the cement. Dose inside the cement was the same as in the surrounding area. Dose for the ventral side was lower than that for the dorsal side; however, the distortion with importing the bulk density of the bone cement was less than with clinical setting (Figure
For the equivalent phantom without bone cement, dose distribution was undistorted (Figure
In the
Dose distribution was analyzed in the
With the equivalent phantom without bone cement, no difference and no distortion were seen between dose distribution calculated by RTPS and that measured using the film-based dose distribution analysis system.
With the equivalent phantom with bone cement, differences were seen between dose distribution calculated by RTPS and that measured by the film-based dose distribution analysis system. Dose at the bone cement as calculated by RTPS with clinical setting was lower than that in the surrounding area, and with importing the bulk density of the bone cement was similar to that in the surrounding area, but dose at the bone cement as calculated by the film dosimetry was higher than that in the surrounding area. Dose for the area after bone cement was lower than that for adjacent area according to RTPS with clinical setting, but the dose distribution calculated by RTPS with importing the bulk density of the bone cement and the dose distribution measured using the dose distribution analysis system did not confirm any dose reduction after bone cement.
Three-dimensional radiotherapy planning is based on CT, but when performing radiotherapy following percutaneous vertebroplasty, bone cement containing barium in vertebral bodies can affect the dose distribution of radiotherapy. The present basic study using water-equivalent phantoms was performed to clarify the effects of bone cement containing barium in vertebral bodies by percutaneous vertebroplasty on dose distribution during radiotherapy and to ascertain differences between dose distribution calculated by RTPS and actual dose distribution.
The RTPS cannot handle the conversion from these high CT values to electron densities, and the barium could have generated CT artifacts making the contouring of the area with bone cement less accurate. Therefore, the authors imported the bulk density of the bone cement into the RTPS. By assuming that the bulk density of the cement is also the electron density relative to water, probably a much more realistic dose distribution will be obtained. This is in fact the procedure recommended for other situations where high-
In our study, dose increases were seen within and before and to the lateral sides of bone cement. Dose increase could have been caused by scattering of radiation by the cement. However, no significant dose increase was confirmed after cement. In particular, dose increase before cement could have been caused by backscattering of irradiation by the cement. The distance from vertebral bodies to spinal cord is about 2 mm. In the present study, the area with dose increase outside bone cement caused by back scattering of irradiation was 2.33 mm, and the degree of dose increase outside bone cement was 1.35 cGy. Therefore, spinal cord dose can appear to exceed planned dose.
In treatment planning using RTPS, the dose decrease was calculated after passing through bone cement. However, no significant dose decrease was seen by the film dosimetry.
In radiotherapy for metastatic bone tumors in the thoracic and lumbar vertebrae, posterior single-field irradiation is generally performed. When performing posterior single field irradiation for thoracolumbar vertebrae, caution must be exercised on spinal cord dose. As reported by Emami et al. [
In recent years, new radiotherapeutic techniques such as intensity-modulated radiation therapy (IMRT) and body stereotactic radiotherapy have been performed [
The difference between the dose distribution calculated using RTPS and the dose distribution measured by the film dosimetry could be explained as follows. First, bone cement has as high as 30% barium concentration. The RTPS did not have data for the effects of an artificial material (bone cement containing barium) on dose distribution. Secondly, bone cement containing barium could have generated CT artifacts. Hence, if bone cement exists within an irradiation field during treatment planning, expected dose distributions may differ from actual dose distribution. Importing the bulk density of the bone cement into RTPS could decrease the difference between the dose distribution calculated using RTPS and the dose distribution measured by the film dosimetry.
In conclusion, the dose distribution of an area containing bone cement calculated using RTPS differs from actual dose distribution.
The author’s deepest appreciation goes to Dr. Motoo Nomura who provided carefully considered feedback and valuable comments. This paper was supported by the research grant from Kansai Medical University. This paper was supported by KAKENHI (21791222).