Evaluation by Monte Carlo Simulation of Doses Distribution in Tumors with Hypoxia

1. Introduction

The modality of treatment with ionization radiation, specifically, for cancer can be considered routinely in several hospital centers, which must have human and technological resources capable of conforming the radiation dose into the target volumes and trying to avoid high toxicity in the adjuvant tissues. Although all these resources exist, the treatment planning is based on prior knowledge of the structures of the patient that can be obtained through any kind of medical images. Once the tumor has been diagnosed, appropriate treatment is indicated without, often, considering some factors that may influence treatment success, due the limitation of the medical images. Factors known as repair of sublethal DNA damage, cell repopulation, redistribution of cells, and reoxygenation are not considered [1].

Experimental data showed that oxygen is the most component that modified the radiation sensitivity and hypoxic cells that can be 2–3 times more resistant to ionizing radiation, which would imply administering doses higher than doses to achieve the same effect in oxygenated cells under normal conditions [2, 3, 4].

An important concept in clinical radiobiology is that the tumor may have subpopulations in hypoxic areas, thus leading to success of radiotherapy. Still, there are concepts related to acute or chronic hypoxic cells that may also alter treatment outcomes [5]. The low concentration of oxygen or hypoxia in the tumor tissues is a radiobiological phenomenon that has been observed since the beginning of the twentieth century [6].

The dependence of oxygen in tissue is related with the generation of free radicals, which came from interacting between the radiation energy and tissues. The quantity that measures the likelihood of this interacting is the cross section, which decreases when the beam energy increases [7]. Thus, few radicals will interact with the DNA of the tumor cell and consequently decreasing the chain reaction. This dependence is known as the oxygen effect that is ignored in the accuracy of ionization radiation treatments [8, 9], which is why hypoxia characterizes a tissue as radioresistant. Therefore, hypoxia and radiosensitivity are related to lower oxygen concentration rate; there will be higher survival cell rate, in postirradiation. On the other hand, there is an optimal oxygenation value wherewith the radiosensitivity increases [10]. Concentrations of oxygen in some tumors can be found in different regions in the same tumor complex. In this case, the tumor hypoxia may occur in a chronic or acute form. A form of oxygen concentration is due to the accelerated growth of the tumor in the most central parts, which are usually originated by the lack of adequate blood supply. In this way, in an axial section of one tumor, have concentric circles of regions of different oxygenation, whose central areas are necrotic [4]. One way to express the decreased radiosensibilty of cell, due to hypoxia, is through the parameter oxygen enhancement ratio (OER), which is defined as the amount of dose reduction required for cell of a given oxygenation level compared to cell with no oxygen to obtain the same effect [11].

One way to study the oxygen effect in tumors is through computational modeling. Thus, different modeling of oxygen effect has been proposed such as a voxel-based multiscale tumor response model [12]. The model used in this work was written in C++ simulating a virtual tumor with considering biological parameters as vascular fraction that is related to the oxygenation of the tumor.

Laura Antonovic and co-authors [13] used a treatment planning system TRiP [14] to simulated spherical tumors in silico based on a biological model of oxygen diffusion [15]. The beam used was carbon ion.

Another publication showed the use of a computational model for trans-vascular oxygen transport and blood vessel networks in tumors [16, 17].

The paper title Dose prescription and optimisation based on tumor hypoxia [18] proposed a method to prescribe dose distributions in radioresistant tumors.

The Monte Carlo HYP-RT model was used to simulate tumors considering the repopulation and reoxygenation for hypoxic head and neck tumors [19].

A 4D cellular model was applied to simulate head and neck cancer with oxygenation varying with vascularity and blood oxygenation [20].

An algorithm implemented on Geant4-DNA (codes based on Monte Carlo) was developed to show the effect of oxygen on DNA [21].

The Monte Carlo simulation, specifically the codes based on this method, can also be an effective dosimetric tool for the study of dose deposited. The dose–response from the codes shows an advantage of providing detailed studies in different conditions that involve procedures which are lengthy, complex, and expensive. The most commonly used Monte Carlo simulation codes in radiotherapy simulations are EGS, MCNP, and PENELOPE [22, 23, 24, 25, 26, 27, 28]. The quality of the results provided by the different codes is directly related to the accuracy of the implemented transport model and its data libraries associated with the cross section of the transported particles. Thus, the mixed charged particle transport algorithm, implemented by the penetration and energy loss of positron and electrons (PENELOPE) code [29], led to its intense use in radiotherapy [30, 31, 32, 33, 34, 35].

This chapter presents a study of dose distribution in simulated cranial tumor with different concentrations of oxygen [18, 36, 37, 38, 39] through the PENELOPE simulation code, which is based on the Monte Carlo method [40, 41]. The tumor with different concentrations of oxygen will be compared with one under normal oxygen conditions.

2. Material and methods

The Monte Carlo code, PENELOPE^®, version 2008 was used to achieve the main objective to simulate tumors with hypoxia. Since the code allows the “construction” of materials by the compound’s chemical composition, the soft tissue material from the code, number 262, was modified by adding different concentrations tumors with hypoxia, from knowledge that chemical compounds of the normal tissue are approximately equal to the tumor. The geometry of simulation used was a parallelepiped of 8 × 15 × 21 cm³ containing six identical spherical tumors of 1.2 cm radius, as shown in Figure 1 [47]. In that work 2 × 10⁹ primary particles and 0.1mm² pixel size and photon spectra at 6 MV [48], which was applied in the input file of the code, were used. The tumors were located before the buildup region for the 6 MV beam at 10 cm from the top of the phantom. The simulated material of the parallelepiped was a soft tissue, available in the code, while the tumors had different pressures of oxygen, from 5 to 70 mmHg. Simulation responses were obtained through the values average in the whole target or spherical tumors.

This geometry used to simulate the radiation conditions was 22 × 22 cm² radiation field, 100 cm source-tumor distance, with a 6 MV energy beam. The obtained results from the simulation were analyzed in terms of deposited energy and values of OER relative to the pressure O₂ in terms of mmHg, following the OER equation proposed for [55].

The second part of that work published [47] analyzed the oxygen effect in the dose distributions in simulated cranial tumors: one with different oxygen concentrations and the other with normal oxygenation [18, 36, 37, 38, 39]. A cylinder with 18 cm in diameter and 20 cm in height was simulated to represent the head of an adult, containing concentric spheres of radii of 0.5–1.2 cm that can represent the dimensions of a glioblastoma tumor. Due the characteristics of localization of this kind of tumor, the simulation tumors were centralized at 10 cm height, beyond the equilibrium range to charged particle for the energy used in this study. The bean incident was simulated in the direction to the phantom, parallel to the Z axis. In Figure 2 the geometry for radiation for a simulated tumor with different concentrations of oxygen from minor (radius sphere 0.5 cm) to greater is shown (radius sphere 1.2 cm).

The simulated material of the tumors and brain structures was soft tissue due to similar chemical compounds present. This material is available in the simulation code. In the spheres (tumors) these materials were modified with different concentrations of oxygen. The radiation conditions used were 1.2 × 1.2 cm radiation field, source-tumor distance of 100 cm, and photon spectra at 6 MV.

3. Results

From the results obtained to the six spheres inserted in the parallelepiped (Figure 1), the deposited energy was plotted for each spherical tumor containing different pressures of oxygen as shown in Figure 3. Values of the OER were obtained through Eq. (1) and plotted for each pressure of oxygen, as shown in Figure 4.

The penmain file of the code generates an output file with generic information, such as number of simulated primary showers, secondary-particle generation probabilities, average deposited energies, and statistical error, etc. Each simulation was written in a separate file. The program computes and delivers the statistical uncertainties (3σ) of all evaluated quantities and distributions. Thus, for all obtained results, an error at least 3.58% was reported by the code used.

Figure 5a shows the dose distributions for a simulated tumor of 1.2 cm radius with different pressures of O₂, and Figure 5b shows the same tumor of Figure 5a with normal oxygenation.

These distributions were compared through a dose profile, as shown in Figure 6, along the center of the dose maps.

4. Discussions and conclusions

Statistical uncertainties at least 3.58% was reported by the code for all simulations. From the results shown in Figure 3, we can observe that the deposited energy have an approximately linear behavior with the increases in pressure until 50 mmHg of O₂; from this pressure the deposited energy shows a constant trend for higher pressures than 60 mmHg of O₂. This behavior was also equivalent to those presented in the literature [36, 49].

The obtained values of the OER relative to pressure of O₂ shown in Figure 4 have a behavior similar to an increasing logarithm function as found in literature [50, 51, 52, 53, 54]. After the 60 mmHg of pressure of O₂, the OER values have a trend to value constant. Figures 3 and 4 confirm the increasing of the deposited energy for tumors with a high concentration of oxygen, because of the oxygen effect that is disregarded in accuracy of the ionization radiation treatment [53, 55].

The dose distributions shown in Figure 5 visually show almost the same distribution, but in the field profile of both distributions, shown in Figure 6, a difference of 7.29% was found at a radial distance of 0.6 mm of the tumor. The code allowed evaluated the influences of effect of oxygen for tumors with hypoxia that is related with the outcome of treatment with radiotherapy. However, the hypoxic tumor cells are resistant to radiation [8]. From the comparisons of the dose distribution in the central plane of the phantom, it was observed that there are regions where the concentration of oxygen was lower (that of sphere of less radius), thus, the energy deposited was lower, unlike the spheres with higher oxygen concentrations.

Despite the fact that the code cannot simulate the physiological factor, which it can modulate a variety of normal developmental and metabolic processes that cause injury to the tumor cell, the present study of tumors with hypoxia plays an important role in dose distribution that can compromise the treatment outcomes and individual prognosis.

5. Considerations

There are a number of simulation and numerical methods and codes that demonstrate the importance of the low concentration oxygen effect in response to ionizing radiation treatments. This chapter shows the work Study of the distribution of doses in tumors with hypoxia through the PENELOPE code which showed physical parameters with results similar to the literature through the PENELOPE-Monte Carlo code.

Acknowledgments

The authors thank professor Ph.D. Patricia Nicolucci for her orientation and suggestions for all the works, which are part of our doctorate work.