Brachytherapy, especially LDR therapy is very hard to model in cell cultures because the dose delivery of LDR therapy lasts approximately one year. Furthermore, the different distances from the multiple sources and the effect of the blood circulation are hard to reproduce in in vitro conditions. Although the basic knowledge of the laws of radiobiology is also known in this scenario, the summarised impact of the multiple different factors - energy, dose rate, treatment time, irradiated volume – on chromosome aberrations is unknown. In our study, we compared the biological effect of the different radiotherapy modalities of prostate cancer treatment on cellular level, using lymphocytes originate from the blood circulation. With this model, therefore, we could exclude the bias effect of the different irradiated volumes.
Previously, most studies of the research field investigated the effect of organ doses on late radiation toxicities. However, the lymphocytes do not stay in one organ, therefore in biodosimetric investigations such irradiated volumes should be used that are independent of the organs. In five studies [19–23] total reference air kerma (TRAK) was compared with ISVs. The studies followed ICRU 38 recommendation [24] to use total reference air kerma for the dose and volume description in cervix brachytherapy. On the other hand, these studies applied much higher doses than what we used and they studied these factors in an intracavitary scenario. Furthermore, Barillot et al. showed, the reference volume enclosed by the 60 Gy isodose surface for cervix BT or with combined EBRT to be an independent predictor of the late complications [25]. They demonstrated the relationship between the volumes and rectal complication and soft tissue sequela in univariate and rectal toxicities. As such, the predictive value of ISVs was independent on the mean organ dose [25].
The dose on the lymphocytes can also be important, because radiation induces immunological changes. Relevant studies gained importance in the light of the increasingly employed immunotherapies in the patients previously treated with radiotherapy.
Jin et al. developed a model for estimating the “dose on the immune system” in order to investigate its relation with treatment outcome. They proposed that there were rapidly circulating immune cells in the heart, lung and blood vessels and the radiation dose is uniformly delivered to them. On the other hand, there are slowly circulating immune cells in the lymphatic systems and blood reservoirs and they are irradiated only if they are in the irradiated volume at the time of the dose delivery. On 464 non-small cell lung cancer patients they found that higher immune system dose was associated with poorer local progression free survival and overall survival [13]. Ladbury et al. also found the similarly modelled immune dose to be correlated with survival in an independent cohort [15].
To better estimate the “dose on blood”, a set of groups developed computational models [14–17]. These models were able to provide dose distribution information as well. As chromosome aberration technique provides information about the quantity of cells with certain aberration numbers, it reflects dose distribution. Therefore, our data can help to validate similar models in prostate. Considerations of model modification for brachytherapy can also be tested in our database.
For three treatment modalities of prostate cancer, we calculated volumes irradiated by extremely low doses (minimum was 0.7 Gy) and compared them with chromosome aberrations, which to our knowledge hasn’t been done before. Our work is also a special one with the analysis of LDR and HDR BT as monotherapy by biological dosimetry methods.
The limitation of our study, however, is that various treatment planning systems and multiple techniques were used in dose delivery in EBRT (RapidArc, IMRT, traditional and simultaneous integrated boost), but the small sample size did not allow us to make a subgroup analysis. On the other hand, there are less difference between the isodose volumes in the used different techniques in our cohort, than between the different therapeutic modalities. (The average V1Gy of 3D conformal therapy is 8968 ± 521 cm3, it is 7601 ± 511 cm3 for IMRT, the t-test is non-significant.) For the same reason, we could not stratify the HDR patients according to the prescribed dose of 19 and 21 Gy. Also, TPS-s may not calculate the low dose ISVs accurately [26]. Different treatment planning systems would calculate different volumes even on the same image sets, and this may cause higher effect on the larger ISVs. Furthermore, V1% was not obtainable in case of HDR therapy, which we had to omit in the analysis.
The evaluation of volumes in three different kinds of radiotherapy treatments resulted in substantial differences between BT and EBRT regarding both volumes and biological doses (Table 4 and Fig. 1). Collection of outcome and toxicity data is still in progress for further analysing clinical differences between the therapies.
We revealed positive correlations between the volumetric features of LDR BT and chromosomal aberrations in blood lymphocytes of patients only at the ninth month after the implantation, but not at other time points (Table 5). However, in the univariate regression analysis, there were no significant predictors of the examined biodosimetric values among the ISVs of different doses. It is important, that although the 98% of the total dose is delivered in LDR BT within one year, the dose rate continuously decreases. Furthermore, the description of the radiobiological effect is complicated by the tumour shrinkage and tumour repopulation which can cause complex models [27, 28]. Thus, it is not trivial to find linear or quadratic relations. We hypothesize that after nine months, the irradiation with the decreasing dose rate cannot counteract the death of the damaged lymphocytes, which decreases the aberration frequency. Therefore, the few emerging aberrant cells are cleared out of the blood. However, our results do not exclude any connection between chromosome aberrations and longterm clinical results of LDR BT. In summary, there is a relatively loose link between the chromosome aberrations and ISV-s in case of LDR BT.
In case of HDR BT, volumes receiving the smallest 1% dose cannot be calculated by the TPS because of the limited size of the calculation matrix. On the other hand, we found that most ISVs correlated positively with radiation specific aberrations (dicentrics + rings) three months after the therapeutic intervention (Table 5). All examined volumes turned out to be significant predictors in the univariate regression analysis after three months and most ISVs were predictors at twelve month (Table 6). HDR therapy uses higher photon energy compared to LDR therapy, which should cause less DNA damage [29]. The ISVs of absolute doses are also smaller in case of HDR, however, the higher dose rate increases the aberration frequency. These effects cumulatively caused less dicentric + ring frequency in our cohort (Fig. 1). Despite of the lower level of CAs, in HDR therapy, there is still a close relationship between aberration frequency and irradiated volumes. The effect of the volumes was also long-lasting (one year at least), the durability of the connection was not limited to teletherapy.
In EBRT, most correlations were found immediately after completing the irradiation, but they were observable at the third-, sixth- and at the twelfth-month follow-up time too (Table 5). Twenty-seven of the variables were observed to be predictors in regression analysis (Table 7). Besides, the highest predictive values (R2 = 49.4) were found in EBRT, where the highest irradiated volumes can be found, which may cause a bigger impact on chromosomal aberrations. The strongest significant predictor was found in the case of V10%. Although fractionation may eliminate the link between the ISV and biodosimetric values, it is clearly detectable in our results.
It may be of interest to note, that in HDR BT mostly dicentrics + rings showed correlations at the third month, which aberrations considered to be radiation specific. However, after LDR BT and EBRT and HDR at the twelfth month, not just the radiation specific aberrations were in relation with the volumetric features. We think that these other aberrations also show radiation dose dependence, but not all of them follow linear quadratic dose-dependence, as we published before [30]. Their frequency can also have diverse time dependence as our data measured right after EBRT suggest.
We also observed, that volumetric features have long-lasting impact and its disappearance lasts for at least a year. The aberrations measured at later time points are possible to be caused by the long-lived lymphocytes, which were stopping in the lymph nodes and later got back into the circulation.