Investigating the biochemical response of proton minibeam radiation therapy by means of synchrotron-based infrared microspectroscopy

Abstract

The biology underlying proton minibeam radiation therapy (pMBRT) is not fully understood. Here we aim to elucidate the biological effects of pMBRT using Fourier Transform Infrared Microspectroscopy (FTIRM). In vitro (CTX-TNA2 astrocytes and F98 glioma rat cell lines) and in vivo (healthy and F98-bearing Fischer rats) irradiations were conducted, with conventional proton radiotherapy and pMBRT. FTIRM measurements were performed at ALBA Synchrotron, and multivariate data analysis methods were employed to assess spectral differences between irradiation configurations and doses. For astrocytes, the spectral regions related to proteins and nucleic acids were highly affected by conventional irradiations and the high-dose regions of pMBRT, suggesting important modifications on these biomolecules. For glioma, pMBRT had a great effect on the nucleic acids and carbohydrates. In animals, conventional radiotherapy had a remarkable impact on the proteins and nucleic acids of healthy rats; analysis of tumour regions in glioma-bearing rats suggested major nucleic acid modifications due to pMBRT.

Figure 1

In vitro study. PCA in the A+FP regions of CTX-TNA2 (astrocytes, top) and F98 (glioma, bottom) cells irradiated with the indicated doses; for each cell line, the PCA scores (upper row) and loadings (lower row) are included. Explained variances by the PCs are about 45% and 55% for the healthy and tumour rat cell lines, respectively. Each point of the PCA scores represents a cell spectrum, and colours correspond to the irradiation configurations: blue for control (non-irradiated), red for BB, green for MB_peak and orange for MB_valley. Vertical lines in the loadings indicate the most relevant peaks for each PC: dashed and pink for PC-1, and dot-dashed and purple for PC-2. The mean doses for BB and pMBRT irradiations were 2, 5 and 10 Gy for astrocytes, and 5, 10 and 20 Gy for glioma cells. For pMBRT, the specific peak and valley doses were 6.5 ± 0.3 Gy and 0.70 ± 0.05 Gy ($\overline{D}$ = 2.1 ± 0.1 Gy), 15.2 ± 0.8 Gy and 1.6 ± 0.1 Gy ($\overline{D}$ = 5.2 ± 0.3 Gy), 32 ± 1 Gy and 3.0 ± 0.2 Gy ($\overline{D}$ = 10.0 ± 0.5 Gy), and 64 ± 3 Gy and 5.5 ± 0.3 Gy ($\overline{D}$ = 19.8 ± 0.9 Gy). MBs were generated by means of a divergent collimator of 15 slits with a width of 400 $\mathrm{\mu }$m, separated a c-t-c distance of 4 mm.

Figure 2

In vitro study. Intensity distributions of the (from left to right) AI/AII, PhI/AII, PhII/AII, _as(CH₂/CH₃) and C=O/_asCH₃ spectral ratios of CTX-TNA2 (astrocytes, top) and F98 (glioma, bottom) cells. Each row corresponds to one dose. Irradiation modalities are coloured in blue for control, red for BB, green for MB_peak and orange for MB_valley. The mean doses for BB and pMBRT irradiations were 2, 5 and 10 Gy for astrocytes, and 5, 10 and 20 Gy for glioma cells. For pMBRT, the specific peak and valley doses were 6.5 ± 0.3 Gy and 0.70 ± 0.05 Gy ($\overline{D}$ = 2.1 ± 0.1 Gy), 15.2 ± 0.8 Gy and 1.6 ± 0.1 Gy ($\overline{D}$ = 5.2 ± 0.3 Gy), 32 ± 1 Gy and 3.0 ± 0.2 Gy ($\overline{D}$ = 10.0 ± 0.5 Gy), and 64 ± 3 Gy and 5.5 ± 0.3 Gy ($\overline{D}$ = 19.8 ± 0.9 Gy). MBs were generated by means of a divergent collimator of 15 slits with a width of 400 $\mathrm{\mu }$m, separated a c-t-c distance of 4 mm.

Figure 3

In vitro study. PCA in the HW region of CTX-TNA2 (astrocytes, top) and F98 (glioma, bottom) irradiated with the indicated doses; for each cell line, the PCA scores (upper row) and loadings (lower row) are included. Explained variances by the PCs are about 85% for both cell lines. Each point of the PCA scores represents a cell spectrum, and colours correspond to the irradiation configuration: blue for control (non-irradiated), red for BB, green for MB_peak and orange for MB_valley. Vertical lines in the loadings indicate the most relevant peaks for each PC: dashed and pink for PC-1, and dot-dashed and purple for PC-2. The mean doses for BB and pMBRT irradiations were 2, 5 and 10 Gy for astrocytes, and 5, 10 and 20 Gy for glioma cells. For pMBRT, the specific peak and valley doses were 6.5 ± 0.3 Gy and 0.70 ± 0.05 Gy ($\overline{D}$ = 2.1 ± 0.1 Gy), 15.2 ± 0.8 Gy and 1.6 ± 0.1 Gy ($\overline{D}$ = 5.2 ± 0.3 Gy), 32 ± 1 Gy and 3.0 ± 0.2 Gy ($\overline{D}$ = 10.0 ± 0.5 Gy), and 64 ± 3 Gy and 5.5 ± 0.3 Gy ($\overline{D}$ = 19.8 ± 0.9 Gy). MBs were generated by means of a divergent collimator of 15 slits with a width of 400 $\mathrm{\mu }$m, separated a c-t-c distance of 4 mm.

Figure 4

In vivo study. Violin plots assessing the probability density of the (from left to right) AI/AII, PhI/AII, PhII/AII, _as(CH₂/CH₃) and C=O/_asCH₃ spectral ratios of healthy rat brain sections 24 h (top) and 2 h (middle) post-RT in the cortex region, and of F98-bearing rat brain section 24 h post-RT in the tumour region (bottom). The colours indicate the irradiation configuration: control in blue, BB in red and MB in green. The mean dose for BB and pMBRT irradiations was 30 Gy. For pMBRT, the specific peak and valley doses were 59 ± 2 Gy and 14.5 ± 1.0 Gy. MBs were generated by means of a divergent collimator of 5 slits with a width of 400 $\mathrm{\mu }$m, separated a c-t-c distance of 2.8 mm.