Ionizing radiation has a crucial role in the treatment of the cancer diseases. It is used both as a local field and a systemic modality. Conventionally external beams of radiation are used to explicitly expose large solid tumors recognized in the patient's body. However, it is very hard to precisely target small and disseminated tumors with external beams of radiation. When dealing with the small tumors such as breast and prostate micrometastases, systemic treatment is the only option available [1].

Systemic radiation therapy at present, confine to radionuclide therapy in which particle emitting radionuclides (mainly β-emitters) are attached to tumor-seeking agents in order to target the cancer cells [2]. Monoclonal antibodies are practically the most specific targeting agent available and are used in radioimmunotherapy, an advanced form of radionuclide therapy [3]. Majority of available monoclonal antibodies, target the antigens at the cell-surface and normally do not pass through intact cellular or subcellular membranes in living cells [4]. While the most susceptible part of the cells to radiation damage i.e. DNA, is located in cell nuclei. During the last two decades, there has been significant progress in production of new monoclonal antibodies and recently cell-penetrating monoclonal antibodies are under development [5]. It suggests, a novel therapeutic strategy in radionuclide therapy using cell-penetrating monoclonal antibodies to carry Auger-emitting radionuclides into the cells. Cell-penetrating pharmaceutical agents and very short range particles (i.e. Auger electrons) exclusively expose the nucleus of the target cells. The dense shower of short-range Auger electrons and highly localized energy deposited around the decay site can be very toxic to targeted cells with minor cross radiation to surrounding normal tissues [6]. The ratio of self-dose to cross radiation and therefore biological response however, will depend upon the intercellular localization of Auger-emitting radionuclides (cytoplasm, nucleus). There are several Auger emitting radionuclides e.g. 111In (t1/2, 2.1 d), 123I (t1/2, 13.3 h), and 125I (t1/2, 60.5 d) with potential clinical applications, however, based on their half-life 111In is the most suitable for Auger therapeutic purposes and was considered in the present investigation.

Experimental dosimetry at this level is almost impossible and the only options available are analytical and Monte Carlo calculations. One of the first attempt in analytical calculation of dosimetric characteristics of auger electron performed by Medical Internal Radiation Dose (MIRD) committee of the American society of nuclear medicine [7]. However, it is very hard to consider all the transport characteristics of charged particles such as energy-loss straggling and secondary electron production in analytical calculations [8-9].

Monte Carlo is a stochastic method for solving complex, mathematical, statistical, and physical problems including the transport of particles in nonhomogeneous materials. Knowledge of the stochastic interactions of particles with matter is essential for estimation of the particle energy loss and absorbed energy to the materials along the particle's track [10]. Therefore, by evident limitation of analytical methods, Monte Carlo is the most suitable methods for estimation of absorbed dose at microscopic level. Variety of Monte Carlo codes with various degrees of sophistication in tracking the particle transport are available and have been used for cellular dosimetry [11]. An exact dose estimation requires considering the detailed spectra of radiation and exact relative abundance of the radiations emitting from the radionuclides. It is therefore useful to carry out Monte Carlo simulation of Auger emitting radionuclides in order to assess the sensitivity of the results with respect to transport approximations generally used in Monte Carlo codes.

In this study, we used Evaluated Nuclear Data File (ENDF/B-IV) cross-sections to establish S values of 111In to the nucleus for a single cell and report in 36 situation cell model that contain two nested spheres in 5-12 and 2-11 micrometers as cell and nucleus radius respectively. Geant4 Monte Carlo simulation was performed and S-values for 111In were calculated by using different physics model (Standard, Livermore, Penelope and Geant4-DNA) and compared with MIRD S-values.

We used Geant4 (Geant4.11.0.patch02) as the Monte Carlo simulator [12]. It includes several C++ class libraries that provide functions for all types of electromagnetic processes. It provides 4 physics models to choose based on the energy of particles including: standard, low-energy Penelope, low-energy Livermore and very low-energy Geant4-DNA. They can cover electron interaction down to 1 keV, 250 eV, 250 eV and 10 eV for standard, Penelope, Livermore and DNA models.

In all simulations, photoelectric effect, Compton interaction and Rayleigh scattering were consi-dered for photon transport. For electron transport, bremsstrahlung interaction, atomic ionization and atomic scattering were considered. Auger electron and x-ray production were also activated in all simulations. Simulations were performed on a PC (Intel® Core™ i7 Processors) operating on Linux fedora 19. We performed 6 simulations simultaneously on different cores. No variance- reduction method was used in the simulations.

Radiation emitting from 111In were set exactly based on MIRD: Radionuclide Data and Decay Schemes [13]. Radionuclide 111In decays by electron capture (100%) and each nuclear transition results in 13 Auger electrons with mean energy Em=0.926 keV and the total yield of Yr=7.431. The decay scheme also includes 12 conversion electrons (Em=176.100 keV, Yr=0.158), 42 low energy x-ray photons (Em=2.105 keV, Yr=9.498) and 2 gamma photons (Em=209.000, keV, Yr=1.847). As per GEANT4 procedure; each type of radiation was defined using a discrete histogram (his point spectrum) at the energy resolution of 1 eV. The relative yields reported in MIRD was assumed to be significant up to 3 decimal number. (Figures 1)

The cell model used for the present calculations consisted of two nested spheres as cells and its nucleus. The radius of the cell and nucleus ranged from 5 to 12 mm and 2 to 11 mm, respectively. This cell sizes were selected in order to compare the results with published data [14]. As in other papers published we assumed the cells are composed of unit density water (G4_WATER) [15]. Radionuclide (111In) was assumed to be uniformly distributed in one of the following regions and simulation was performed independently; inside the cytoplasm (Cy), over the cell surface (CS), and inside the cell nucleus (N). In each simulation 104 decays was considered and the absorbed energy in target region (rk) from the radiation in source organs (rh) was determined. Based on the MIRD schema the corresponding s-values were calculated.

They only considered conversion electrons (145–245 keV) and Auger electrons (8.5 eV- 25.5 keV) released during 111In decay for dose calculation. They considered 104 electrons history to achieve standard deviations (or uncertainty) smaller than 1%. The energy cut-off was 1 keV that is the electrons energy lower than 1 keV was assumed locally absorbed in the spot. Relative differences (RD) in percent were calculated accordiong to the following formula:

Where SD and SH represent the S-values calculated using Monte Carlo and MIRD respectively.

Figures 2-5 includes the scatter plot of the relative difference between four series of data. This figure demonstrates the Bland-Altman plot to reveal the RD versus reference values (MIRD S-values). Relative differences are up to 11.46% for self-absorption but higher (36.34%) for cross-absorption. The average value of relative differences named bias demonstrates a systematic difference in Monte Carlo simulation. As the plot shows although 85% of the data points are inside the limits of agreement (average of RD ± 1.96 × standard deviation of RD) and the statistical difference is acceptable (≈5%) but there is a high bias (5.27%) between data.

There are currently five techniques to decontaminate equipment with oil and gas NORM, but the most widely used method worldwide is ultra-high-pressure water jetting abrasive decontamination. Despite being used to minimize the dispersion of particles due to forced precipitation with the use of water, the dispersion of radioactive particles in the air is still observed (Figure 8), and it will require an individual internal monitoring for an effective evaluation of the radiological risk of these workers.

Although internal monitoring is very important, the lack of published scientific articles involving NORM oil and gas decontamination makes very difficult the discussion about this area. Despite having few articles, two were found that provide information on the decontamination of NORM oil and gas equipment and showed the real need for internal monitoring of workers.

Since radiation exposure and activity concentrations of oil and gas NORM are very different from other NORM industries (e.g. coal, mineral sands, fertilizers, construction, etc.), internal individual monitoring will become necessary to assess potential doses to organs and tissues, especially during equipment decontamination work.

With the implementation of a fully radiological risk assessment for equipment decontamination work, the companies would perform a correct radiological monitoring of their employees. This will be one of an important action to ensure the health and safety conditions to workers involved.

Furthermore, countries, with their regulatory authorities, would develop specific regulations to ensure full radiological protection for workers exposed to oil and gas NORM during equipment decontamination.