Correlation and regression investigation
We performed correlation analysis in each treatment modality group comparing the ISVs with the biodosimetric values measured at different time points. After Benjamini–Hochberg correction of multiple testing we found that 9 months after LDR therapy the chromosome aberration values (dicentrics + rings, chromatid deletions, total aberration and aberrant cell values) correlated positively and significantly with the largest ISVs, namely V1%, V1Gy. The correlations were weak, between 0.37 and 0.45 (Table 2). In HDR therapy, 3 months after radiotherapy dicentric plus rings correlated positively and weakly with most ISVs (Table 3). Aberrant cell number correlated with V150% at this time point as well (Spearman correlation coefficient 0.41). We found no correlations in any other time points of the follow-up. In the EBRT treatment group, the ISVs showed significant positive correlations with various chromosome aberrations directly after the radiotherapy, at the 3rd, 6th and at the 12th month. The aberrations which correlated with the most ISVs are shown in Table 4. Furthermore, the Spearman correlation coefficient for V10Gy are 0.52 and 0.51 for dicentrics + rings and aberrant cells at 3 months, respectively. The frequency of chromatid breaks and aberrant cells correlated with V10Gy (Spearman correlation coefficient 0.50 and 0.56) and aberrant cell number with V10% (0.49) at 6 months. Total aberrations (0.48) and aberrant cell number (0.43) showed correlation with V10Gy at 12 months.
Univariate regression analysis was performed to explain the variance of chromosome aberrations by variables of ISVs. Multiple regression cannot be used, because of the high collinearity of the variables. We sumini–Hochberg correction in Tables 4 and 5. In case of LDR therapy, we found, that none of the ISVs were significant predictors of chromosome aberrations (p > 0.2 R 2 < 6.4% at the ninth month after therapy). However, at the third month after HDR the predictive value (R 2 ) varied between 26.0 and 31.7, the best predictor being V100% (p = 0.004, R 2 = 31.7%) (Table 5). We also found significant regression models 12 months after the HDR irradiation, where we received no significant correlations. The R 2 ranged between 15.5% and 20.5% and the concerned aberrations were chromatid breaks, total aberrations and aberrant cell values (Table 5). After EBRT, the data of the regression models of time-points directly after radiotherapy and 9 months later are shown in Table 6. Furthermore, we also found significant regression models at 6 months after the therapy (V10% and chromatid deletions: R 2 = 19.5) and 12 months after the therapy (dicentrics and rings with V1% and V1Gy, R 2 = 20.1 and 18.7, respectively). On the other hand, less regression relation was found than with rank correlation. R 2 was between 17.3 and 49.4%, the highest value was calculated for the relationship of V10% and the total aberrations 9 months after the therapy.
Brachytherapy, especially LDR treatment therapy is very hard to design from inside the cellphone countries while the amount beginning out-of LDR cures persists as much as one year. Additionally, different ranges on numerous sources and the effect of the latest circulation are hard to replicate when you look at the inside vitro requirements. While the base level knowledge of the latest guidelines from radiobiology is also recognized in this scenario, brand new summarised impact of your own several different factors-times, serving rates, cures time, irradiated volume-to your chromosome aberrations are unfamiliar. Within analysis, we compared new physical aftereffect of various radiation therapy strategies off prostate malignant tumors therapy into the mobile top, having fun with lymphocytes result from the newest circulation. Using this type of model, hence, we are able to prohibit minder the latest bias effect of the many irradiated volumes.