ENISS Comments on
ICRP draft Radiation Weighting for Reference Animals and Plants
General remarks
ICRP has developed radiation weighting factors for the use in human radiation protection (Publication 60 and updated in Publication 103). Although based on experimental RBE-values, the radiation weighting factors are biased for end-points relevant to human radiation protection i.e. cancer at low doses and dose rates. There are considerable elements of judgement involved and the same set of radiation weighting factors are used for all organs/tissues which is practical but not in line with well-founded radiobiological knowledge. Apparently, RBE is used differently in radiobiology compared to radiation protection which also has to be recognized in the case of environmental radiation protection.
In environmental protection the focus is on “fitness” of the population inferring that the critical health effects may differ to humans. Furthermore relevant exposure regimes refer to planning and existing situations while emergency exposure situations are questionable (see ICRP 124). This means that low dose and low dose rate regions prevail, implying that acute tissue effects (acute mortality) might be beyond the scope of this report. Inversely, lethal tissue damages occurring late after exposures could be of some relevance to the stated aims in the report.
In the current draft clear objectives are lacking (1) why it is important of applying radiation weighting to the exposures of biota in case of tritium and α particles and (2) why it is important to adopt a generic radiation weighting factor for α-particles in spite of all the listed “complexities”. For risk assessment, for example in an existing exposure situation, an arbitrary value of radiation weighting is not applicable (see ICRP 92) as considerations on physical dosimetry as well as radiobiological and ecological parameters need to be taken. Notably, the ICRP system of protection of the environment means performing risk assessments in different exposure situations, and comparing the final results with some reference levels (DCRLs).
Concomitantly , the emphasis on “tritium” has to be clearly motivated since other radionuclides appear in the environment that upon decay are releasing low energy electrons (e.g. Auger).
The literature surveys on RBE might be valuable but ought to be accompanied by profound discussions (not only numbers) on their relevance to the RAPs, to the exposure conditions and for DCRLs comparisons. Should any factor for biota be related to types of radonuclides rather than particle energy?
Out of the 41 studies reviewed for tritium in Tables B1 (early mortality), B2 (reproductive success) and B3 (morbidity), 21 studies refer mainly to experiments on mouse and only five on RAPs (rat). The majority of endpoints assessed in Table B2 apply to gametogenetic effects (spermatogenesis and oocyte survival) in the mouse. The high radiosensitivity of primary oocytes to β-rays from HTO is well-known (3 mGy/day kill 50 %). However the final effect on reproduction is less due to an overproduction of oocytes and may only cause shortening of the reproductive phase. This may indicate alternative mechanistic pathways in the system and should be commented.
There may be difficulties to draw conclusions on late effects from the 17 studies presented in Table B4 for mutation/chromosomal aberrations. Especially studies on the radiosensitive L5178Y and human peripheral lymphocytes are not representative for the aims of this report. Furthermore molecular damages (DNA damages, chromosomal aberrations) may either be lethal or non-lethal to the cells and, although there may be arguments for the use of RBE values obtained from studies on the induction of chromosome aberrations in human radiation protection, it seems more appropriate to base decisions on radiation weighting factors for any population in the environment on studies upon animals and plants.
The radiation weighting factor of 1 for tritium (or is it low energy electrons?) seems adequate but requires some better justification in view of the surveys presented in Tables B1-B4.
The determination of the nominal value of a weighting factor for α-particles is intricate because of complex dosimetry due to the heterogeneous energy depositions and the spatial distributions (short ranges). The molecular damages induced by high LET radiation have also shown differences compared to damages induced by low LET radiation. To address this complexity together with many other considerations and uncertainties might be beyond the scope of this report but in the review of studies and their relevance for the stated weighting factor some notations on the dosimetry for the goal of the environmental protection ought to be addressed in the main text (only mentioned briefly in the Annex).
Out of approx. 86 studies reviewed for “α-emitting radionuclides” in Tables C1-C4, 56 studies (about 65 %) refer to experiments on cells (primarily mouse and hamster cells). Arguments for the inclusion of studies of DNA damages , of chromosomal damage and of neoplastic transformation might be relevant for human radiation protection but is debatable in case of non-human biota. Stochastic effects such as cancer have been deemed less important here, and the survey as well as the background provided in Annex A ought to be more focused on the aims of this report (see comment below). Notwithstanding these above comments , studies on α-exposures are relatively rare which may justify the inclusion of some cellular/molecular studies but not without comments.
A majority of studies are performed with internal exposures to 238-239Pu,210Po and 241Am (is that external exposures?) but there also appear exposures to protons, cyclotron and tandem accelerated particles (14N, 4He) with very short ranges and 125IUdR. Thus it seems necessary to revise the tables and give emphasis to the most important studies for protection of the environment in the case of α-emitting radionuclides, by also considering the limitations of the studies. A suggestion is to move Chapters 1.5 (expanded) and 2 in Annex C to the main text since the wide spread of studies (and the decision) must be followed by some explanations and clarifications. Since there appear an almost endless list of RBE determinations in the literature , the decision on the selection of studies has to be clarified, as data from a few research groups seems to have been much utilized.
There is an immense variety of biota, exposure conditions , response to radiation, in interaction with other environmental parameters etcetera which may call for a prudent and simplistic approach to the protection of the environment. But the question is whether a single conservative value based on a limited number of experimental studies is too simplistic and too generic and thus of no relevance for its purpose? Table C6 in Annex C is indicative of the problem of finding a nominal value for weighting factors for -particles in the environment (for the RAPs) and a more valid recommendation of the ICRP would be to recommend range of tentative values for application in risk assessment.
Accordingly,
- the Introduction chapters need to be revised in order to include more explanatory text and a thorough justifications to the suggested main recommendations! The proposed amended Annexes (see comments below) might be helpful in this case,
- the survey of studies in the Annexes ought to be more condensed and focus on studies of most relevance for recommending weighting as such, and the range of values for weighting factors. As stated, the RBE-surveys for some endpoints seems inadequate,
- due to the inconsistencies in the report (see above and detailed remarks below) and current recommendations from the ICRP in human radiation protection it seems illogical to apply a single radiation weighting factor for α-particles for every exposure situation and for a number of different end-points.
- Correspondingly, the recommendation should be that (see point 26), whenever there is a need in a specific exposure situation to consider the efficiency of high LET radiations in the response to radiation, possible weighting of the assessed absorbed dose rates might be warranted. The range of RBE values in Table C6 could be indicative but not without clear justification. A value of 10 does not seem appropriate in view of (among others) the uncertainties in the surveys, the multifaceted environment and the intricate dosimetry and radiobiology.
Detailed remarks
Chapter 1.1
Most of what is stated is repetition of previous ICRP publications. The subject of this draft report is intended to advice the ICRP on assessing radiological impact on biota in planned and existing exposure situations. Thus, it is not the relevance of RBE for the RAPs that needs to be discussed but the relevance of the radiation quality for assessing the impact on populations of non-human biota. In that aspect parts of Annex A-C should be moved to the front (main report) but revised (see below).
The Derived Consideration Reference Levels are set in absorbed dose per day but it is not clearly indicated if the estimated radiation weighted dose rate is to be compared with a no weighted DCRL or a weighted DCRL? How different radiation qualities have been regarded when it comes to the derivation of the DCRLs ought to be discussed and clarified. Seemingly this contradicts ICRP current recommendation regarding humans and use of equivalent dose?
Chapter 1.2
From biophysical and radiobiological knowledge it is apparent that RBE is a very complex quantity depending on many factors such as LET, dose, dose rate, biological system, biological impact and reference radiation. Although based on RBE, the suggestion of weighting factors in view of the environment must involve a considerable amount of practicality and simplifications. Hence “a table of presumptive RBE values “ has to be interpreted for the purpose of this report. As an example, in a planned exposure situation and tritium, the expected doses due to HTO are low or very low, but many exposures reported in in the tables in Annex B are very high. The corollary of this discrepancy for the recommendations ought to be judged in the main text (although briefly mentioned in the Annex).
Furthermore, is tritium important for an existing exposure situation?
The aim of the ICRP to have a prudent approach to the protection of the environment is acknowledged and it may be important to consider the efficiency of different types of radiation for that purpose. Nonetheless the consequences of effects on populations seem to be more than complicated and the statement (lines 253-254), much research remains to properly address populations effects, should not be an excuse for excluding an important feature of environmental protection from the analysis. The reported RBE data are derived from evaluation of laboratory experiments i.e. effects on individuals. How relevant are these studies for giving recommendation in radiation weighting of absorbed dose rates in view of protection of the population? Also, studies on transformed cells – how reliable are these data for late effects in RAPs?
(lines 213-215)- Thus, for completeness and to allow comparisons to be made, less relevant stochastic data on cancer and chromosome damage are included together with directly relevant data on tissue reactions – This may add confusion, consider to delete data of no relevance.
(lines 241-244) - DCRLs, however, are set at absorbed dose rates where deleterious effects may occur; the selection of an appropriate weighting factor thus has direct relevance for our understanding of likelihood of effects and need for protective measures. – Is this really true? DCRLs apply to individuals and there are elements of expert judgement involved in the interpretations so far. Consider to delete!
(lines 247-249) - ….there is a possibility of effects in the population. To that end it may be important to take into account the RBE, when the radiations of concern warrant. – RBE may be dependent on many things but not populations. Consider to delete!
(lines 251-253) Although, the decision to ignore if population effects are deterministic or stochastic is natural, still, in the context of this report, it is remarkable. RBEs show major differences for different end-points and to adopt a single nominal value to conceal the difference doesn´t seem appropriate. The majority of reviewed studies applies to either tissue level effects or stochastic effects (cancer) which is obvious due to their purpose of supporting human radiation protection.
Concerning tissue reactions in humans after exposures, ICRP stated in the earlier draft report on the use of effective dose, that there is limited information available to compare the effectiveness of different radiation qualities in causing tissue reactions. It was further concluded that adding specific radiation weighting factors for this type of reactions was not warranted in most cases. It is also well known that for high LET radiation, tissue effects due to cell killing have much lower RBEs compared to stochastic effects. These differences must also apply to the data presented in this report and to combine the reported RBEs for deterministic and stochastic effects to a single value for α-radiation is rather simplistic (illogical) even in view of pragmatism.
Chapter 2
Fig 2.1 – 2.4. Information about the dosimetry is incomprehensible and must be sorted out. The dose rates also seem to be rather high and the extrapolation to doses of relevance in planning exposure situations has to be revealed.
Figure 2.1 and table B.2 Please clarify the relationship between them and the basis for the recommended radiation weighting factor for tritium.
The several-fold uncertainty in dosimetry may justify the selection of a weighting factor of 1 for tritium (low energy electrons)? But what about Auger electrons emitted by radionuclides in the environment? High ionization density in close proximity of DNA is crucial for the radiation impact (see conclusion point 30).
Chapter 3
The complex dosimetry of exposures to internal α-particles needs some explanations, especially in view of the data presented in Tables C1-C4. This is addressed in the Annexes but should after amendment be inserted here. For better usefulness when risk assessments are undertaken, the implication of a non-uniform energy deposition and the spatial distribution of the particles in regard to the review of studies and their relevance for the doses and the goal of the environment ought to be better addressed.
(lines 390-392) Of these, 58 were reviewed in detail; the remainder were considered to have inadequate precision with regard to dosimetry, or had other limitations- Recommendation to revise the tables and only include relevant studies. Also consider to include more recent studies and studies from different research groups.
3.1-3.3 and Table 3.1 See comments above! Please clarify!
(Lines 421-422) It should be noted however, that these effects are stochastic in nature and at present, it is uncertain how to extrapolate such effects to relevant population endpoints. - This is an important statement and does indicate that these types of studies may only be valuable from qualitative viewpoints and not quantitative. Consider revision of the tables to exclude data that are not appropriate for the conclusions.
(Lines 460-461)The values obtained are in a wide range but centre around values of the order of 10. – Some further explanations are needed for this conclusion since not all the data presented are appropriate. It also seems doubtful to make some kind of “averaging” in view of differences in fundamental mechanisms.
(Lines 459-461); A value of 5 seems to be more justified recognizing uncertainty in dosimetry, selection of end-points and overly conservative extrapolations (ICRP 92).
Chapter 4
(Lines 469) The validity of statement, there is some consistency in numerical values and (line 471) similarity across organisms, is based on the presented surveys and not generic . How reliable is the statement on similarity with regard to the purpose of the report? Only one type of RAPs appear in the tables in Annex B and C, the rat. The affirmation may be true but some vagueness in the statement must be considered.
Lines (478-480/Annex A) The extrapolation of effects to doses below thresholds of an observed effect involves a high degree of uncertainty and may be overly conservative. The assumption that cell killing is the only cause of tissue effects (ICRP 58) has been discussed more recently by the ICRP (ICRP 92). As late tissue effects also may be relevant for the goals in environmental radiation protection it is recommended to consider the conclusions from the ICRP 92 report.
(Lines 480-484) For the purposes of this report, therefore, it is considered reasonable to base proposals for radiation weighting factors for biota on the observed RBE data without further adjustment to obtain RBEm values for tissue reactions and RBEM values for stochastic effects, although RBEm and RBEM values were calculated for some studies with alpha particle emitting radionuclides (Annex C) – Please clarify this statement! Is that justified?
(lines 488-489)- … all chromosome damage/mutation studies relate to stochastic effects and their relevance in the context of this report is more questionable. – It seems necessary to clarify whether the suggested weighting factor comprises these data.
Lines (490-493). The division of tentative end-points of biota in line with human radiation protection has severe implications. Annex A is in principle a review of the background for selecting weighting factors for humans. Nonetheless, the recommendation is to move larger part of this text to the front after an extensive revision. The focus must be on non-human biota, their exposures and the most relevant end-points, and ignoring other end-points of less or no concern..
Lines 490-493;No clear differences in the ranges of RBE values for low LET radiations may justify some deeper explanations (ICRP 92). Anyhow the recommendation of a weighting factor of 1 for low LET radiations is adequate but the justification of a similar value for tritium is missing.
Similarly, a weighting factor of 10 for α-radiation has to be justified since the underlying radiobiological and biophysical basis for such a high value is disputable.
Lines 497-499; Consider changing this recommendation to : If internal exposures to tritium beta or α-emitting particles are within or close to the DCRL, possible weighting of the assessed absorbed dose rates might be warranted. Let that be the main recommendation from this report!
Annex B-C
Part of this information is valuable for the objective of this report. However the data is unbalanced and ought to be revised with a higher focus on environmental radiation protection rather than human protection. The impact on the environment varies and to identify end-points of concern is therefore not straightforward but has to be done for justifying the chosen RBE-values/weighting factors. Furthermore it must be realized that choosing an arbitrary radiation weighting factor for α-emitting radionuclides has some disadvantages. Any assessment involves consideration of a diverse environment and complex dosimetry, and pragmatism must apply not only to the exposure but also to other environmental parameters. The evaluations of a few studies presented in the tables might be helpful but a specific number is not warranted. Overall, a more comprehensive discussion about the selection process is preferable.
In the end, costs due to unwarranted conservative values in radiation protection are not justified especially in comparison with other environmental impact.