Variations in radiosensitivity of breast cancer and normal breast cell lines using a 200MeV clinical proton beam

Abstract

Background: Breast cancer is one of the most commonly diagnosed among woman in South Africa, and a more resilient effort should be focused on treatment improvements. Worldwide, proton therapy is increasingly used as a radiation treatment alternative to photon therapy for breast cancer, mostly to decrease the risk for radiation-induced cardiovascular toxicity. This in vitro study aims to determine a better understanding of the radiosensitivity of both tumour and normal breast cell lines to clinical proton irradiation. In addition, we propose to investigate whether the increase in linear energy transfer (LET) towards the distal part of the proton beam results in an increase in relative biological effectiveness (RBE) for both cell lines. Methods: Malignant (MCF-7) and non-malignant (MCF-10A) breast cells were irradiated at different water equivalent depths in a 200 MeV proton beam at NRF iThemba LABS using a custom-made Perspex phantom: the entrance plateau, 3 points on the Bragg peak, the D80% and the D40%. A cytokinesis-block Micronucleus (CBMN) assay was performed and Micronuclei (MNi) were manually counted in binucleated cells (BNCs) using fluorescent microscopy. Reference dosimetry was carried out with a Markus chamber and irradiations were performed with a clinical proton beam generated at NRF iThemba LABS that was degraded to a R50 (half-value depths) range of 120 mm, with a field size of 10 cm x 10 cm and a 50 mm SOBP. The phantom could be adjusted to accommodate different perspex plates depending on the depth required within the proton beam. Cells were then exposed to 0.5, 1.0, 2.0, 3.0 and 4.0 Gy doses for each cell line independently and for each dose point. Results and Discussion: For the CBMN results, a program was developed on Matlab platform to calculate the 95% confidence ellipse on the co-variance parameters α and β. These values were determined by fitting the linear quadratic dose response curve to the average number of radiation induced MNi per 1000 BN cells. The ellipse region around a coordinate (the average MN frequency) for both MCF-7 and MCF-10A cells at the plateau region was defined by the mean estimate of the α-value and the β-value that were plotted on the X-axis and Y-axis respectively. The ratio of the two parameters, α/β, is a measure of the impact of fractionation to determine the biological effective dose. In fractionated proton therapy, the MCF10A cells will repair less between two fractions compared to the MCF7 cells. This is not an indication of therapeutic gain from a fractioned treatment protocol. For this reason, the hypofractionated stereotactic treatment protocols that can be applied with protons could be to the befit of the breast cancer patient. The above argument is based only on the radiosensitivity of the two cell lines exposed in the plateau region. Further analysis of the 95% confidence ellipse of both cell lines also showed a clear increase of the alpha value toward the distal portion of the beam and indicates an increase in energy transfer in this region. The gradual increase in α and β parameters with depth for protons for both cells is of clinical importance, since it implicates a non-homogeneous dose within the targeted area and an unwanted high dose behind the targeted area. Distal energy modulation could be investigated especially with larger breast tumours. RBE was calculated as the ratio of the dose at the different positions to the dose at the entrance plateau position (reference) to obtain an equal level of biological effect. A statistically significant difference in radiosensitivity could be observed between malignant and non-malignant cells at all positions (p<0.05). The variation in RBE was between 0.99 to 1.99 and 0.92 to 1.6 for the MCF-7 and MCF10A cell respectively. Conclusions: There is a variation in RBE along the depth-dose profile of a clinical proton beam. In addition, there is difference in radiosensitivity between the cancerous cells and the normal breast cells. While this study highlights a variation in sensitivity between cells it could be used by the modelling community to further develop biologically motivated treatment planning for proton therapy.