1. Introduction
Breast cancer is one of the most common cancers affecting women. Approximately 1 in 8 American women will develop breast cancer over the course of her lifetime [1] . Breast cancer mortality is second only to lung cancer. In order to minimize the risk of breast cancer recurrence, breast cancer patients are often treated using adjuvant radiation therapy after surgery [2] . However, the radiation therapy can cause adverse skin reactions, such as painful acute desquamation and chronic fibrosis [3] . Skin reaction tends to be most severe for post- mastectomy chest wall cases treated using bolus. Therefore, the assessment of skin dose is important to evaluate the risk of side effects from radiation treatment. According to the International Commission on Radiological Protection (ICRP)’s Publication No.60 [4] , the skin dose should be assessed for the dermis and epidermis layers of the skin. Although we are able to estimate dose to a patient using the radiation treatment planning system (TPS) [5] [6] [7] , a number of studies have demonstrated that surface and near-surface doses estimated by TPS can be inaccurate [8] [9] [10] . One reason is the steep dose gradient in this region. The ultimate check of the actual dose delivered to a patient in radiotherapy can be achieved by using in-vivo dosimetry [9] . The in-vivo dosimetry is applied to assess the actual dose delivered to accessible critical organs such as rectum, vagina and bladder; or in difficult geometries organs where the dose is hard to predict from the treatment plan such as head-and-neck and breast cancers [11] .
Many techniques have been used for in-vivo dosimetry, such as semiconductor diodes, thermoluminescence dosimeter, and metal oxide semiconductor field effect transistor (MOSFET) [12] [13] [14] [15] . In this study, the dosimetry estimate using nanoDots (Landauer Inc., Glenwood, Ill.) and Gafchromic EBT2 film (Ashland Inc., Wayne, NJ) was carried out to measure chest wall dose distribution for parallel opposed beams using treatment planning.
Cardiac toxicity is a well-established late effect of radiation therapy to the chest wall, such as coronary artery disease, which may occur years after radiation treatment. Lung toxicity is not as commonly reported, but irradiation of larger volumes can lead to radiation pneumonitis. Dose in heart and left lung should be closely monitored in post-mastectomy radiation therapy [16] [17] [18] [19] [20] . Certain organs are inaccessible for in vivo dosimetry, such as the heart and lung. Therefore, a phantom was used to measure delivered dose to these organs.
2. Methods and Materials
2.1. NanoDot Calibration
Compared with other types of optically stimulated luminescence dosimeters (OSLDs), the nanoDotis the most feasible dosimeter for single point measurements of skin dose assessment in radiotherapy application due to its small size. The detector material is aluminum oxide doped with carbon (Al2O3:C) with 1.2 mm thickness and 5 mm diameter encased in 10 mm × 10 mm × 1.8 mm light tight plastic holder, as shown in Figure 1. The plastic holder has a density of 1.03 g/cm3, and the leaf thickness covering the front and back of the nanoDots is 0.36 mm [21] .
Before using nanoDots for dose measurements, they were calibrated under full buildup conditions at 100 cm SAD for 6 MV photons. The calibration was carried out using solid water slabs (Gammex Inc., Middleton, WI) and 0.5 cm bolus. The slabs consisted of various thickness slabs from 0.2 to 5 cm with dimensions of 30 cm × 30 cm and having a density of 1.045 g/cm3. The nanoDots were placed on 10 cm thick slabs, positioned on the central axis at depth of maximum dose, 1.5 cm for 6 MV photons beam (Truebeam, Varian Inc., Palo Alto, CA). A half centimeter bolus and 1.0 cm solid water phantom were placed above the nanoDot. The linear accelerator was calibrated for a SAD of 100 cm and field size of 10 cm × 10 cm for an output of 1 cGy/MU. The nanoDots were irradiated for absorbed dose of 2 cGy, 5 cGy, 8 cGy, 50 cGy, 100 cGy, 150 cGy, 200 cGy and 300 cGy. One dosimeter was kept as a control for background measurement. The nanoDots were read using the microStarii reader (Landauer Inc., Glenwood, Ill.). Results were presented as a calibration curve of nanoDots reading value versus absorbed dose to water. A linear relationship of dose/reading for nanoDots was determined.
In addition, a study on the angular dependence of nanoDots was carried out. NanoDots were irradiated with 200 MU/min at gantry angle of 0˚, 30˚, 60˚, 90˚, 270˚, 300˚ and 330˚, the setup is shown in Figure 2. Results are for relative dose versus gantry angle.
2.2. Film Calibration
The Gafchromic EBT2 film (Ashland Inc., Wayne, NJ) was calibrated under the same conditions as the nanoDot. The calibration consisted of 12 dose points spanning 0 to 400 cGy. Singular films for each dose level were positioned at the isocenter plane. Guides were marked on the slab to aid film placement. For improved accuracy of the EBT2 film calibration, each exposure was repeated twice. The final calibration was derived from the average response of the films.
After film calibration, the percentage depth dose (PDD) values were measured using films loaded on the slab stack at the following depths: 0 (surface), 1.5, 3, 7,
Figure 1. AnanoDot, material used is Aluminum oxide doped with Carbon (Al2O3:C) with 1.2 mm thickness and 5 mm diameter encased in 10 mm × 10 mm × 1.8 mm light tight plastic holder.
Figure 2. Schematic diagram of angular dependence on nanoDots. The gantry angles were at 30˚, 60˚, 90˚, 330˚, 300˚ and 370˚.
and 10 cm. Films were orientated perpendicular to the radiation beam. Note that in reality these depths increase slightly when the finite thickness of the film (approximately 0.3 mm) is accounted for. For comparison, PDD curves were also measured from films aligned vertically along the central beam axis, aligned using side lasers. Following the proper procedure suggested by the manufacture, each film was scanned with 300 dpi using a film scanner (Epson expression 10000 XL, Long Beach, CA). The optical density was computed by subtracting the pre-scan from the post-scan [22] [23] . To determine the dose at the certain depth the polynomial fitting equation obtained from film calibration was employed to convert optical density to dose. The ImageJ (NIH, Bethesda, Maryland) software package was used to analyze the PDD.
2.3. Treatment Planning Using an ART Phantom
A female ART phantom was used in this study, as shown in Figure 3(a). The portion utilized in this work was the left chest wall. The phantom was constructed with a natural human skeleton cast inside soft tissue simulating material. Three tissue-simulating materials are the soft tissue material with a density of 0.997 g/cm3, designed to have the same absorption as human tissue. The skeleton’s bones possess a density of 1.610 g/cm3, and lungs with a density of 0.330 g/cm3. Film was placed between transverse slabs of ART phantom to measure dose to left lung and heart. Two layers of 0.5 cm thick bolus were used to cover the left chest. Four nanoDots were placed above the bolus, between the layers, and below the bolus, as shown in Figure 3(b) and Figure 3(c). Three control nanoDots were placed on chest wall to quantify imaging dose [24] . The phantom was scanned using CT simulator (Somatom, Siemens Medical Systems Inc., Malvern, PA) with both nanoDots and film in place. Three 3D conformal radiotherapy plans with 200 cGydose per fraction using 6 MV photon medial and lateral tangential fields were performed using planning system (Pinnacle3 9.10, Philips, Andover, MA). All dosimeters were contoured on CT and dose was extracted. The ART phantom was positioned on the treatment couch at the same set-up
(a) (b) (c)
Figure 3. (a) Femal ART phantom; (b) Axial view of the scanned ART phantom; (c) Sagittal view. Positions of 12 nanoDots are labeled.
during the CT simulation, and irradiated with tangential fields (Truebeam, 6 MV, 200 cGy). The irradiation was performed three times using three sets of dosimeters for each plan. Irradiated nanoDots were read using microStarii reader. Film was scanned with a resolution of 300 dpi using Epson expression film scanner and analysed using DoseLab (Mobius, Houston, TA). The red channel data was used during film analysis, color correction was disabled as recommended by manufacture. All films were scanned approximately 24 hours after their irradiation.
3. Results
3.1. Nanodot Calibration
The nanoDot calibration curve for 6 MV photons for the range of 0 cGy to 300 cGy was established as shown in Figure 4. The graph shows that nanoDots signal is linearly proportional to absorbed dose to water. The linear equation obtained from the graph will be used in extrapolating the surface dose. Figure 5 shows the variation of surface dose for different angle of incident beam. The measured dose was normalized to the surface dose of perpendicular beam incidence. The curve is plotted with the least order polynomial. It shows that the incidence of beam entry angle on the phantom surface has an effect on surface dose; the more oblique the beam angle the higher the surface dose. This is possible due to the relative increase of thickness of nanoDot cover at oblique angles. The maximum angular response is given by 90˚ gantry rotation. The application of angular correction factor is recommended for nanoDot when involving oblique radiation beams.
3.2. Film Calibration
A film calibration curve was obtained from films placed at isocenter plane and irradiated with 10 × 10 cm2 field to doses within range of 0 - 400 cGy, as shown
Figure 4. NanoDot calibration curve for 6 MV photon beam.
Figure 5. Response of nanoDot with the variation of beam direction.
in Figure 6(a). Figure 6(b) shows measured PDD for 10 × 10 cm2 fields. The solid curve is PDD measured using EBT2 films vertically aligned with beam central axis. The markers indicate the PDD obtained from axial films at multi-depths.
3.3. Dose Distribution to the Chest Wall
The comparison of surface dose for parallel opposed treatment plans using nanoDots were made. In order to assess the reproducibility of the measurements, the surface dose for each plan was measured three times. In total, 36 nanoDots for each plan were evaluated in this study. The average dose for surface nanoDots at positions #9 - 12 for plan 1 was: 197.2 cGy, 194.6 cGy, 207.6 cGy and 202.2 cGy, and the dose deviation compared with prescribed dose was: 3.8%, 2.7%, 9.8% and 5.0%, as shown in Figure 7. NanoDots at lateral positions, with greater beam obliquity, had larger variance than those at medial positions. A similar trend was observed for other nanoDots as shown in the Figure 7; #5 - 8 were between the bolus layers, #1 - 4 were below the bolus layers. We also noted
(a) (b)
Figure 6. (a) Film calibration curve; (b) PDD comparison with film is vertical and axial to photon beam.
Figure 7. Comparison of measured and prescribed dose using nanoDots at different locations at chest wall, as shown in Figure 3.
that the surface dose at lateral tangential angle is higher than medial tangential angle after applying angle correction. This might be due to effect of exit dose, which may contribute to the surface dose, particularly for small separation body sections such as the chest wall. Thus, the surface dose that was measured was actually the combination of entrance and exit dose.
The measured dose using film for heart and the left lung was extracted along the red line drawn in Figure 8(c); a 6.2% and 7.5% maximum difference from prescribed dose was shown in Figure 8(d).
4. Discussion
In this study, we conducted measurements to describe the relationship between the prescribed dose and the measured dose to chest wall, left lung and heart. The dose distribution study using nanoDots and film was investigated using an ART phantom. Three parallel-opposed 3D conformal radiotherapy treatments were planned and delivered onto the ART phantom. The surface dose for the lateral
tangential field resulted in a higher measured dose than medial tangential field due to the effect of the radiation beam angle incidence and opposed beam exit dose.
Based on the results of this study, if nanoDots are used to assess for multi-field treatment, an angular dependence correction factor must be in place.
There are certain limitations in this preliminary study. First, we did not perform dose comparison for IMRT plans [25] [26] [27] . Second, beyond the advantageous characteristics of nanoDots as mentioned above, there are a number of other characteristics that must be accounted for when measuring, such as temperature, energy dependence, and linear energy transfer dependence [28] [29] [30] [31] . Last, plans using electron beams, which are commonly used to treat the chest wall, were not performed. It should be noted that for CT X-ray energies, the response of nanoDot is increased significantly [24] [30] . This is due to the relatively high Z value of Al2O3, prompting more photoelectric photon interactions. Three control nanoDots were used when ART phantom was scanned using a CT simulator.
5. Conclusion
The dosimetry provided by an advanced treatment planning system was verified using nanoDots and film. Both nanoDots and film provided reasonable estimation of dose distribution in post-mastectomy radiation therapy. Although surface dose measured using nanoDots does not give an exact estimate of skin dose due to the finite size of the detector, it is a useful method for in-vivo estimation of skin dose. There is no established method for accurately quantifying treatment-associated coronary artery and lung diseases, which may occur years after radiation treatment. Dose distribution estimate in heart and lung using film and ART phantom is a useful tool in the evaluation of breast cancer patients after treatment.