Peer review of the genotoxicity, toxicokinetics, and carcinogenicity of a pesticide

Ni more comment.

No more comment.

We are determining a KMD, but the MOE depends of the KMD. Thus, either we decrease or increase the KMD (e.g., 10 or 100 ppm); this will have not real impact since the MOE will still be high then safe. In other words, the human exposure is very low compared to any KMD we may derive from the white paper. .
No more comments.

No more comment. I agree.

No more comment. I agree.

Therefore a lower KMD or higher uncertainty factor should be used for more precaution. Something closer to 10-30 ppm compared to 60 ppm.

First, I agree that hazard needs to be put in context of exposure, and risk is more informative. Also, I did not take issue with the WOE classifications provided in Q5.8 because a WoE needs to come to a judgement of whether there is sufficient evidence to make a projection of human risk (ie “Likely” or “not likely”) or there is not a sufficient basis to support a reliable projection of human risk given the nature of the responses/data (ie suggestive, inadequate).

The question is if we can estimate a human KMD. I don't understand why MOE are brought up in this discussion. Are we assessing risk or are we determining a kinetically derived value ?

answering to users 37600 [Expert 2] and 750802 [Expert 9] , if to consider portal of entry for respiratory system, the nasal cavity would be the selected one, and not the lungs.

I think we should not include the 2 NTP studies in our discussions, because of the contamination with epichlorohydrin. I would like to know the composition of TeloneII used for the other studies, and also the composition of DD92 used for some studies.

There are contrary views throughout this survey as to the ability of a 1% epichlorohydrin impurity to explain away ALL the positive findings in the NTP bioassays.

"IF" the WOE is centered around "not likely." While it certainly appears that most of the panelists across the subgroups agree on "not likely," this is not a consensus view.

See also the discussion in Topic 7.4 below about the possible relationship between "large MOEs" and hazard classification.

User-125195 [Expert 4] : I agree with you that labelilng requirements are "serious"-- but all I'm saying is that, as a logical matter, if the current "very small" exposures are *illogically* used to conclude that the pesticide is not carcinogenic AT ALL (a classification matter), THEN it is possible that the EPA requirements could change-- after all, why not allow more exposure for a "non-carcinogenic" substance. The hazard classification and the risk assessment are two separate exercises and must not be conflated.

Yes and for pesticide assessments EPA controls for this possibility by specifying the amounts of the pesticide to be used under each type of application and user group that may be exposed, including personal protective equipment requirements, field entry delays, etc. so that the exposure levels don’t change. Pesticides. Must be used as specified on the label. Not doing so is a violation of Fededral law and it is taken very seriously

I agree with the points raised here and think that the key point is raised by user-292467 [Expert 12] who notes that that the current MOE are already large (safe).

I agree with others that a quantitative assessment of carcinogenic potential is not needed if the WOE is centered around "not likely" for humans. A point of departure (POD) approach can be used assess risk and/or MOE but it should be based on the most sensitive toxic endpoint (non-cancer). Other endpoints have not been fully discussed in the White Paper.

The question is confounded by the lack of context...mutagenic potential in vitro? in vivo? in humans? This may explain the difference in the answers in the original review. I agree that, based on the available data, it is most likely not mutagenic in vivo.

I appreciate much of the comment (user-125195 [Expert 4] ) above, which seems to be in response to my "suggestive evidence" classification in Round 2. For clarification, I would classify 1,3D as "possibly carcinogenic", NOT because of the arguments in the white paper that the human threshold dose is far below current exposure levels (I don't fully agree with that logic anyway...), but because I see the bladder carcinomas, lung adenomas, and perhaps other findings among the 8 bioassays as of concern and NOT sufficiently debunked by the arguments in the white paper.

So the interesting issues in the above comment (what to do about a substance that CAN cause an effect in humans but can't if "used safely"?), don't apply to my original response to this particular question. I will note, however (albeit with ZERO knowledge of the lawsuit #125195 [Expert 4] refers to), that there is nothing inconsistent about (1) saying a substance poses no human risk as long as exposures are low; and (2) compensating a plaintiff whose exposures were NOT low!

I don't disagree with user-292467 [Expert 12] 's point that given what we know about current human exposures, the MOEs are "large" and it is therefore not that important to actually consider what the human TK are and the inter-individual variability therein. However, I have seen too many cases where an advocate uses this "logic": (1) chemical X might be a human *risk* if exposures were not currently very low; (2) therefore, chemical X is not a human carcinogen.

There's nothing wrong with combining hazard classification and risk assessment, but it has to be understood that "no significant risk" is a function of exposure, and cannot *by itself* determine the hazard classification. This becomes very worrisome in cases where exposures change and are no longer small relative to the level of concern.

I have nothing to add

nothing to add.

just flagging that in Section 8.6 below, there is a more complete discussion of the (not-agreed-upon) issue of whether the lung adenomas are in fact "slightly outside" of historical control rates or are completely different from historical control rates.

the current available data does not allow make any estimation of a human KMD value.

There is no human information to estimate a KMD value for humans. If animal data are to be used, I agree that uncertainty factors will be needed, as mentioned by user-292467 [Expert 12] .

I have nothing to add. I stick with the different Km explanation.

Human data (in vitro or in vivo) are simply lacking and therefore it is impossible to make a quantitative estimation of the sort.

no more comments.

Because it is likely that others will share user-37600 [Expert 2] ’s concerns, including potential regulatory reviewers, the sponsor may wish to address each of the points raised by 37600 [Expert 2] directly in the white paper when it is revised

no more comment

I agree that the overall picture support a KMD of 20 ppm or below (figure 5 of white paper).

There is definitely a non-linear phenomenon occurring . This is clearly the results of a diminishing clearance at exposure levels above 10 ppm (first measured occurence at 20 ppm). It remains to be determined if GSH depletion is the sole explanation, or a enzymatic saturation.

Nothing to add here.

There is not much to debate on this question.

Or just used more robust scaling factors for (I) the GSH level or activity in rodent versus diverse human populations, (Ii) plasma protein binding differences (differences in active free moiety) in rodent versus human and (iii) for the dosing scenario (intermittent versus repetitive dosing versus recuperation of the GSH or not versus the toxicity of the metabolite or not) as said in my previous comments. Overall, because their current toxicological risk assessment is not robust enough, this open the way to several debates; however, a lot of information is still missing. Therefore, they should used more robust and mechanistic based scaling factors to cover the remaining uncertainty and variability.

intermittent exposures above the KMD are OK because GSH can "recuperate...Yes after depletion (but it depends of the dosing interval; we still not know since no intermittent dosing scenario has been made), the GSH may recuperate for a higher elimination effect of the compound. This may be on the safe side and hence a higher KMD could be considered. However, this is only be true when the metabolite is not toxic....if the metabolite is also toxic this should be of concern since more metabolite could be produced if the Glutathione is less saturated ; this should be considered with scaling factor to correct the dosing scenario. Information in that domain are missing. From my side the current scaling factors used in their risk assessment are not conservative enough. Overall, using more robust scaling factors may stop our discussions on the uncertainty for the extrapolation to human.

Yes, based on my previous comments you may have two choices: 1. used the lowest 10 ppm to cover some uncertainty and variability effect, or 2. used 30 ppm but adjust the interspecies and/or inter individual scaling factors for differences in the unbound fraction in plasma (amount of active form) and differences in the amount of Gluthatione. However, as said the current MOE are already large (safe). No additional correction is probably not needed for the current human exposure.

I agree with the two other reviewers. We knows already lot things but some additional uncertainty factors could be needed. First, a greater unbound fraction (most active form) of the compound in human plasma versus rodent would not be protective for human. Since we do not have this information, the interspecies scaling factor could be adjusted. And I also agree that the toxicological risk assessment should be refined since it is too simple from now.; at least a scaling factor based on species and/or interindividual differences in the amount of Glutathione would be relevant. However, as said, the current MOE are already large (safe).

Yes I agree, the kinetics of both isomers could be different due to different Km values but the toxicity could also be similar.

Yes, this is why I asked to make at least a plasma protein binding study in rodent and human to define the amount of free drug in human plasma versus rodent plasma. Considering that only the free drug moiety is active, the KMD in rodent would have to be corrected by an additional interspecies scaling factor for an extrapolation to human while the unbound fraction in plasma is not the same in rodent and human.

Should be an "add" before "much" in the above cooment.

Yes, not add the NTP studies due to the presence of epichlorohydrin in Telone II used for that study.

Hard to find a good surrogate. 1;3-D is such a small molcule that almost any change will alter it's reactivity. Substitution at the CH2 would likely be the most conservative. I found a couple of refs to 1,3,3-TCpropene. Inactive
in mn assay, in vivo and in vitro, but produces strand breaks in vitro. Doesn't really much to discussion.

The direct answer to this question is yes, but this does not mean it is completely effective in all organs.

agreed

I agree with the comment from user-2441 that the existing bioassays provide sufficient data to support a hazard WOE, and that a more cogent and logical explication of the most probable MOAs would likely shed more light on the issue

I agree that at this time application of Tox21 is at present of limited support to better understand the potential for human genotoxic risk with respect to carcinogenicity. However, it seems that it will be useful for risk assessment in the future as networks of AOPs are worked out and their interactions characterized.

re: answer 6, regulatory agencies focus on risk assessment which is a careful and detailed evaluation of both exposure and hazard to estimate risk and not simply hazard characterization or the inherent property of the substance to cause disease at some level of exposure that may greatly exceed that ever seen by humans. Regulatory Agencies use conservative assumptions and safety factors to ensure that allowable use levels have large margins of exposures that protect public health. They also recognize that at very high exposure levels, xenobiotics can overwhelm detoxification and repair systems causing diseases, such as cancer, that simply would never happen at the levels encountered in the environment. IARC and NTP can be very helpful to EPA, FDA and other regulators around the world by conducting tests to identify chemicals that can cause tumors at any exposure level in a hazard identification manner. Their use of terms like "Possibly (or Probably) carcinogenic to humans" means that the substance so identified has the potential to cause cancer in humans, without regard to the exposure levels required to produce tumors and whether such levels might ever be encountered by humans, to alert regulatory authorities that cancer is a possibility for that substance (i.e. there is a potential hazard). The regulatory agencies then take the results and conditions of the study into account and carefully evaluate potential exposures to the substance in context with the levels that humans will encounter to judge the substance's risk. In addition to providing this benefit the IARC and NTP approach can also create unintended negative consequences, such as categorizing a chemical substance that has been tested many times with negative results as a possible human carcinogen because of one positive study employing very high doses that greatly exceed that maximum dose levels required to be tested. For instance, glyphosate has been studied in nearly two score cancer bioassays almost all of which were negative. Every regulatory authority in countries around the world had concluded that if used according the the pesticide label requirements, which is the law, glyphosate could be used safely. However, based on one lifetime study using something like 5000 mg/kg/d cancers were found in the high dose level, far above the 1000 mg/kg/d maximum dose level the EPA requires to be used, and thus according to their categorization scheme IARC categorized it as a possible human carcinogen. This was of use to EPA, and every other regulatory authority around the world (some 30 odd countries) because they now knew it could cause cancer at least under very high dose exposures. They then evaluated the data sets within the requirements of their own countries laws and regulations and unanimously concluded, that glyphosate is safe to use under the conditions and restrictions on its the label (i.e., while glyphosate may pose a hazard at very high dose levels the risk assessments conducted in some 30 odd countries all said its safe to use as directed on the label because the exposures encountered by humans will be orders of magnitude less ). On the other hand that same IARC hazard identification classification of possibly carcinogenic was also successfully used in a lawsuit to claim that a worker got cancer from its use despite unanimous agreement among regulatory agencies worldwide that it does not and will not do so. Given the dichotomy between hazard only categorizations vs. hazard x exposure = risk categorizations at best one compares apples to oranges when applying the IARC, NTP classifications with those of organizations such as EPA. While those who feel that it is irresponsible not to use hazard identification alone to restrict exposure to substances that might cause cancer under any circumstance, because we just don't know everything about the initiation and progression of the disease there is another side to the argument. One can also argue that it is irresponsible to regulate a substance without consideration of the exposure levels required to cause the tumors , because of the societal benefits the substance provides (e.g., the public health benefits from disease control).

I agree with user-232578 [Expert 7] that while clarification of the issue of clastogenicity might provide increased confidence about the carcinogenicity of 1,3-D it falls into what I might term "nice but not necessary" studies as it would not likely change the bottom line conclusion of "not likely to be carcinogenic"

Nothing more to add

I agree with the points raised by reviewer 292467 [Expert 12] , but to me the overall picture supports a KMD of 20 or below

I have nothing to add

my thinking is aligned with answer 1 above and I do not think that consideration of structural surrogates will contribute significantly to the WOE analysis during risk assessment

I agree with user-553126 [Expert 5] that while 1,3-D may pose a hazard that at realistic conditions the risk is negligible.

I agree with user-553126 [Expert 5]

I too do not think it poses a mutagenic threat in vivo

I agree with both answers 1 and 2

Nothing to add

Nothing to add

I agree with #24419 [Expert 8] that risk (probability of harm) must consider both potency (dose-response) and exposure; I disagree, however, that the White Paper attempts to assess risk via a probable MOA. Rather, it asserts what the MOA is without fairly considering alternative modes, and (more importantly) then concludes that the human risk conditional on this MOA is zero, *without* actually undertaking any kind of quantiative risk assessment of that claim.

To user 717898 [Expert 1] : I was of course referring to compounds positive in the in vitro cytogenetic assays (e.g. chromosome aberration assay; micronucleus assay) for which negative findings in the in vivo bone marrow micronucleus test with no clear target tissue exposure are observed. Based also on my direct experience, many of these in vitro genotoxins do not reach at a sufficient level the bone marrow. In these cases, the in vivo comet assay does represent an obligate in vivo follow-up study since alternative methods (e.g. induction of micronuclei in the rat/mouse colon and intestinal epithelium and also in liver of young rats) are not yet validated. Aneugenicity would not be covered in these cases and additional ad hoc in vitro evaluations would clarifying the MoA. I would also note that the Transgenic Rodent Somatic and Germ Cell Gene Mutation Assays is only used as in vivo follow-up study in case of positive outcome for gene mutation in vitro (Ames test, mammalian cell gene mutation assay).

To use 615591: I am not sure why OECD 489 is (or would be) the 'ultimate in vivo follow up study when positive findings are observed in the in vitro cytogenetic assays" - where is this stated in the guideline? I do agree that OECD 474 has more limitations regarding exposure but for BB there is plenty of other in vivo data available that show that there is systemic (and bone marrow) exposure. OECD 474 is a measuring mutations, the Comet assay is an indicator test. Don't get me wrong, I like the Comet assay and it has an important place in the testing strategy but in the given case here I would prefer the in vivo micronucleus test

I am still unclear of what is being asked here in 6.1 which makes it also difficult to comment on the above discussion.

also agree with my peers that the cancer data are supportive of a non-genotoxic mode of action, wich is why I had said 'no' in the above answer. No - it does not affect my conclusion since I, from the other overall data (although that dataset has its weaknesses), do not believe BB is a DNR reactive carcinogen.

To note, the overall confidence level of all of us with the dataset is similar between 6/10 and 8/10. I would assume that the sponsor of this study would want to increase this closer to '10' - and from what I have seen in the debate so far it seems that we all struggle with coming to a more clear answer for the lack of state-of-the-art data for clastogenicity/aneugenicity

This question 2.8 and the variability in views will likely resolve if there was a clean (presumed to be negative) recent OECD 474 study available (a negative would also negate aneugenic effects)

It seems like none of the reviewers is now providing clear support anymore for simply excluding studies for the presence of epichlorohydrin . While I agree with most of the statements from reviewer 617591 [Expert 14] regarding limitations of 'non-standard' studies I do not believe that this disqualifies them for any consideration in the risk assessment. If there was a state-of-the-art micronucleus/clastogenicity in vivo study available then that would be an easier call

nothing to add here, have given detailed feedback on every assay in my initial review

Don't think the acceptance of the Gallapudi 1985 study is all that relevant since there is a recent, guideline compliant TGR study available which is clearly negative. As stated earlier, for this study (Young 2018), the lack of through substantiation of the maximum dose used. Given the votes shown in the above table, it seems like there is good consensus on a lack of mutation by BB but less clarity on the 'clastogenicity' side

No more comments...

I agree with user 320359 [Expert 3] . I think the yes and no answers reflect different interpretations in what the question is asking.

In response to user 617591 [Expert 14] , more data would certainly be helpful and increase confidence. With respect to strand break assays, I don't necessarily think that a lack of cytotoxicity indicates the dose level was appropriate, as the mechanism of action may not reflect the mechanism that occurs at the lower exposures to which humans are more likely exposed. If the mechanism of action is not relevant, then risk characteristics in humans starting with that dose will not be relevant. A smooth increasing dose-response would address this question, but the data are not available. The closest we have is the kinetically-derived maximum dose, and the strand breakage assays were well above this value.

Important typo correction in the sentence, "The strand break assays used doses well above the kinetically-derived maximum dose, likely indicating their mechanisms of action were relevant to human exposure conditions". Should read, The strand break assays used doses well above the kinetically-derived maximum dose, likely indicating their mechanisms of action were NOT relevant to human exposure conditions.

agreed

The question suggests metabolic saturation above 20 ppm. Several colleagues suggest that the data support a level (slightly) below 20 ppm for the start of the non-linearity. I think that is a more accurate interpretation of the available data

It appears that the colleagues who answered Yes and No are saying the same thing. The pattern of response in the bioassays is not what might be expected with a mutagenic/genotoxic agent. While not definitive, these results can support our thinking regarding genotoxc potential when considering all of the evidence.

For hazard identification the “strand break assays” employed appropriate dose-levels for the individual assays and this is what is needed to establish whether a compound is genotoxic or not. The kinetically-derived maximum dose is a different issue in the context. For the strand break assays and more specifically the in vivo comet assay by Sasaki et al., (1998) where marked and statistically significant increases in DNA breakage in the stomach, liver, kidney, blood, lung, brain and bone marrow three hours after treatment, no signs of cytotoxicity or necrosis were observed by histopathological observation in the organs in which DNA damage was observed. This confirms that the dose-levels used were appropriate. Concerning the last sentence of the above comment, the statement that "1,3 -D is not genotoxic to humans" does not appear to me correct. In terms of hazard identification, genotoxicity of 1,3-D (Mainly clastogenicity) cannot be ruled out based on the available data. So, what is needed is to firmly establish if 1,3-D is genotoxic, as also discussed in previous comments by myself. Once this is achieved the risk characterization for humans at realistic exposure conditions can be performed.

While I agree with answer 1 above (yes), answer 2 (I cannot answer) provides the appropriate caveat. A detailed weighing of the protocols is almost always needed.

I am confident that 1,3-D has genotoxic potential in vitro, but this greatly reduced when glutathione is present (a realistic scenario). Other than strand breakage assays I see no evidence that 1,3-D is genotoxic in vivo. The strand break assays used doses well above the kinetically-derived maximum dose, likely indicating their mechanisms of action were relevant to human exposure conditions. However, full dose responses were not carried out, so results at exposure-relevant doses are unknown. In a weight of evidence approach, these considerations lead me to conclude with a fairly high degree of confidence the 1,3-D is not genotoxic to humans at realistic exposure conditions.

Concerning the Gallapudi Big Blue assay, I don't agree with the white paper. The OECD guideline specifies the use of 5 animals to detect a 2 fold change in mutant fraction (However, as 1,3-D is likely weakly mutagenic (if at all)) 5 animals would not be sufficient to detect a 1.5 fold increase . Yet this could be important biologically. Also, since 1,3-D is at best weakly mutagenic, and mutations in the BB assay simply integrate over time, the 28 day treatment period would have doubled the sensitivity.
The Young BB assay was much more thorough, with a larger number animals and a longer exposure period, but the fixation was only several days for the last exposures (but longer for the earlier exposures). The sensitivity of the assay might have been increased with a longer fixation period. Although the OECD guideline states oral exposure is acceptable, it also states that the anticipated route of exposure should e considered when designing the assay; and inhalation exposure seems more relevant for 1,3-D.
Nonetheless, although certain limitations can be pointed out, (others were identified by the above reviewers)I think the weight of evidence is still against in vivo genotoxicity for 1,3-D. The in vivo tests referred to in the question, do not show any clear evidence of genotoxicity, If unlimited resources were available, the sensitivity of the post-labelling assay and the BB assays could be increased, but at biologically relevant exposures it would likely not change risk estimation.

Agreed - base excision repair (most likely type for small adducts) generates a strand break.

JUST FOR CLARIFICATION:
• Clear evidence of carcinogenic activity is demonstrated by studies that are interpreted as showing a dose-related
(i) increase of malignant neoplasms, (ii) increase of a combination of malignant and benign neoplasms, or (iii) marked increase of benign
neoplasms if there is an indication from this or other studies of the ability of such tumors to progress to malignancy.
• Some evidence of carcinogenic activity is demonstrated by studies that are interpreted as showing a chemical-related increased incidence of neoplasms (malignant, benign, or combined) in which the strength of the response is less than that required for clear evidence.
• Equivocal evidence of carcinogenic activity is demonstrated by studies that are interpreted as showing a marginal increase of neoplasms
that may be chemical related.
• No evidence of carcinogenic activity is demonstrated by studies that are interpreted as showing no chemical-related increases in malignant or benign neoplasms.
• Inadequate study of carcinogenic activity is demonstrated by studies that, because of major qualitative or quantitative limitations, cannot be interpreted as valid for showing either the presence or absence of carcinogenic activity.

(PS: the simplest formula for the uncertainty in the pooled historical control rate of 40/865 (4.6%) gives 95% confidence limits that extend from 3.2% to 6.0%, both very far below 18%. This formula assumes that there is not some systematic effect making the individual experiments non-random, so if in fact the control rate was changing from year to year not by chance, these limits would be too narrow. But the White Paper makes no attempt to posit any factor that would make some control rates relevant to the experiment being analyzed and others not, let alone to posit some model for pooling the rates in some non-uniform way).

In response to user-750802 [Expert 9] : I agree that historical controls are relevant; my problem is with the non-statistical handwaving inherent in Table 16 and the accompanying narrative. The historical control rate is 4.6%. This number has some uncertainty around it, but the right way to portray the uncertainty is not to choose the single highest experiment out of about 20 different experiments and call that the "upper bound" in order to say that 18% (four times the pooled historical control rate) is "within the range."

I agree with User-850922 [Expert 10] , it should be possible to use PBTK modeling to make extrapolations in humans.

No more comment from my side.

No more comment.

agree with user 477751 [Expert 11] and user 232578 [Expert 7]

This comment pertains to the comment by user-37600 [Expert 2] above questioning the use of historical control data as described in the White Paper with reference to Table 16. In my experience , historical control tumor data are often used to aid in the interpretation of tumor data from a specific study. The historical control incidences represent a range of incidences that may occur in a control group by chance alone. If the tumor incidence in a specific study falls within the range of historical control incidences, it is fair to conclude that this tumor incidence may have occurred by chance alone and may not be treatment-related. Application of HCD in this way is fairly standard and is not considered to be "cherry picking". At the same time, application of HCD in this way does not "prove" that a specific tumor incidence was not treatment-related. This is true for the liver adenoma incidences in the Scott, 1995 study. We should also note that whereas Table 16 lists HCD on liver tumors over a 21 year interval (1991-2012), the highest HCD incidence in males (16%) occurred in 1992 and the highest incidence in females (8%) occurred in 1997. Since the Scott study was completed in 1995, the highest HCD incidences occurred within a few years of the completion of the Scott study. The time frame of these HCD data in relation to the Scott study increases their relevance in terms of interpreting the tumor data from the Scott study.

I agree with 477751 [Expert 11] immediately above-- just flagging that my comment above was about historical controls GENERALLY, and I'm not sure if any of the pages on this site have a discusssion of the way the White Paper (mis?)uses historical control data to suggest that statistically significant increases are not what they seem. I'm happy to move this comment and my prior one to a different page if asked.

To user 553126 [Expert 5] : I agree that 32P-labelling assay is not really sensitive to detect small adducts, as potentially could be for 1,3-D. This does not mean that DNA adducts are not present. They simply cannot be detected because the limit of the assay. However, DNA breakage observed (e.g. comet assay) could be the result of intermediate steps in the repair of small DNA adducts (excision repair) and not necessarily via a secondary MoA.

Agree with user-232578 [Expert 7]

Although I found in general that the White paper to be clear in how the data were selected and intrepreted, I agree with 24419 [Expert 8] . Adding to this comment, the paper needs a better hazard/risk characterization toward the end that provides the reasoning that underlies the choices in the available evidence as the foundation for conclusions, and what is particularly lacking is a characterization of the uncertainties. There will always be uncertainties (some were pointed out in the peer review). I dont think they distract from the overall conclusion regarding risk, but nonetheless, these uncertainities need to acknowledge in put in context.

to user 47751 - That is my question too. Is the composition of Telone II (given in the diet to F344 rats) different from the composition of DD-92 (oral-gavage to Sprague Dawle rats)? In many other situations, when comparing gavage x dietary studies, the dietary studies produce less cancer. In this case it is the opposite.
Either the composition is different (information not given) or there is a different rat strain susceptibility.
In addition, the hepatocellular adenomas percentage was significantly higher only in the highest dose.
In fact, I agree with the comment on page 36 of the white paper: "The fact that lower or comparable systemic exposures led to more severe toxicity (tumors vs. nontumors) further supports that the tumorigenicity only observed in rats via dietary exposure is not consistent.".

I agree, simply because a response is port of entry, is not a basis for disregarding. My point is given the high variable background and susceptibility, and that this benign tumor response was slightly outside of HCs, it is not convincing evidence of carcinogenicity. Some of the statements in the White paper should be revised so as not to imply that port of entry is a basis for downplaying or dismissing.

As stated, it woud be helpful to clarify the issue of clastogenicity with additional information to increase confidence. But it is still my opinion, this deficiency does not distract from the "not likely to be carcinogenic" conclusion because of the totality of the evidence (lack of gene mutations in vivo, TK, cancer bioassay results, etc). Clarifying the issue of aneuploidy, in my opinion, is more important in the context of reproductive risk. However, in this case, it appears that 1,3-D does not pose reproductive/developmental risk from the results of traditional TGs (as discussed in other documents). I agree with the comment that the White paper needs to include discussion of these studies that may have bearing on this issue.

I am not in disagreement with 617591 [Expert 14] . My only point was, in general, effective DNA reactive mutagens typicall y induce point/gene mutations (including deletions) and chromosomal aberrations. But there are exceptions of chemicals that are only clastogenic and do not induce
point mutations. Again, if the sponsor wants to strengthen the conclusions regarding genotoxicity, they need to address clastogenicity.

To user 717898 [Expert 1] : The in vivo mammalian alkaline Comet Assay is regulated by OECD TG 489 and currently represents the ultimate in vivo follow up study when positive findings are observed in the in vitro cytogenetic assays (e.g. micronucleus test). Many of these in vitro genotoxins do not reach at a sufficient level the bone marrow (negative findings in the in vivo bone marrow micronucleus test with no clear target tissue exposure). In these cases the in vivo comet assay does represent, at present, the obligate in vivo follow-up study.
Furthermore, I do not consider a good idea to recommend a new OECD 474 study (bone marrow micronucleus test) for the reasons I indicated in round 1 (Result 2.1) where I reported for the bone marrow micronucleus test by Gollapudi (1985) as main shortcoming, the absence of a clear target tissue exposure. This is corroborated by the findings that IARC carcinogens aloalkanes and haloakenes including 1,3-D, are poor or no inducers of micronuclei in the bone marrow in vivo as reported by Morita et al., (1997).

To user 232578 [Expert 7] : Reactive compounds to DNA can be clastogenic but not necessarily mutagenic (induction of point mutation) because the specific MoA. If a gene is broken down by a clastogenic MoA (e.g. large deletion) this is not detected by a conventional assay for the induction of gene mutation (except by the in vitro mouse lymphoma TK assay).

To user 477751 [Expert 11]
Genotoxicity on the "negative side" is not supported by solid data and results. On the contrary, there are evidences that 1,3-D could be clastogenic as shown by the in vivo Comet assay in stomach, liver, kidney, bladder, lung , brain and bone marrow (Sasaki et al., 1998). In addition, aneugenicity , which I believe to be not a real issue, cannot be ruled out by the available studies.

Independently by the potential confounding role of epichlorohydrin, particularly for the outcome of the in vivo studies, I think that this is not a real issue. The studies with a positive outcome for genotoxicity in the presence of epichlorohydrin (e.g. bone marrow micronucleus test by Shelby,1993; Kevekordes, 1996; Alkaline elution by Kitchin, 1994) are not relevant for risk assessment due to significant limitations and shortcomings as also pointed out in “Result 2.8”.
For the bone marrow micronucleus test by Shelby (1993) considered by the author as positive at 150 mg/kg, the negative historical control data were not considered and the frequency of micronucleated PCE (Mn-PCE) observed at 150 mg/kg (0.31 %) is usually found in negative control animals. In addition, no evidence of target tissue exposure is present.
For the bone marrow micronucleus test by Kevekordes (1996), in addition to an inadequate study design (single animals in negative controls), it bears unrealistic values of PCE/NCE ratios which invalidate directly the study.
For the alkaline elution by Kitchin (1994) it should be noted that this test has never been fully validated.

Yes, and I am also pleased with answer 2.

Answer 3: I agree with the Colleague. Concerning the DL test , please see my comments in the previous result.

Answer 1: I agree with the Colleague that the Dominant lethal assay is not very sensitive. In addition, in the updated version of the relevant OECD TG 478 it is reported the following: "However due to its limitations and the use of a large number of animals this assay is not intended for use as a primary method, but rather as a supplemental test method which can only be used when there is no alternative for regulatory requirement". Overall, the DL should be considered "not relevant" or at the best "slightly relevant".

Answer 2: I agre with the Colleague. Concerning the dose-response for the observed resorptions (3 and 4) and also the increase from 1.7% to 7.7% for resorption 3 and its possible statistical significance I would not pay much attention. Statistical significance has not been reported in the white paper and I believe that this is due to the high variability of results obtained among the animals.

Answer 1: “Big blue assay for Gallapudi…..”: I cannot completely follow the Colleague in this respect. I think that this study, though performed before adoption of the relevant OECD Guideline (TG 488), it essentially adheres to the main requirements of the TG. I only noted (my comments in round 1) that the administration period was 14 days instead of 28 days requested for slow proliferating tissues/organs (e.g. liver). The background of mutation frequencies appears to regular (see also background mutation frequencies in the rat study by Young, 2018). Concerning the route of exposure in the study by Young (2018), I cannot see a real problem, since administration by diet is one of the recommended routes of exposure. In this respect, the only issue I can see, is the use of the maximum tolerated dose (MTD) as the top dose. From the white paper it is not possible to assess this, but I believe that MTD was used as top dose. Overall, the in vivo studies with transgenic animals relevant for induction of point mutations are the most reliable among the studies available. Based on this, I concluded that these results can overrule the concern for point mutation observed in some in vitro studies and in particular in the Ames test (Lawlor 2009) and in mammalian cells (Myhr and Caspary, 1991).

Somewhat agree with user-37600 [Expert 2] , but it is strange that the same dose, but a much lower Cmax causes tumors in one study. Same dose, but much higher Cmax does not do anything in a 2nd study ? Again emphasize that detailed information on the composition of the materials tested could help explain the complex dataset. It is not known what other impurities except epichlorohydrin were/are present and if variable impurities contribute to the inconsistent results.

The point is not ignoring portal of entry effect. I am questioning the arguments regarding relevance of the non-linear plasma kinetics to discuss away the lung effects. I do not think that this is logic since the conc. of 1,3-D in the respiratory tract is depending on the air conc. Plasma kinetics may indicate saturation of liver biotransformation. I would not rule out portal of entry effects in any assessment

My point is that with the variable incidences of lung tumors in mice, it will always be difficult to obtain reliable results no matter how large the groups are. I think that more information on the composition of the materials tested including impurities other then epichlorohydrin may help explain some of the contradictions. I also agree that presence of epichlorohydrin may only explain the forestomach tumors. My conclusion that the NTP study should be set aside is based on issues with study design and conduct. The bladder tumors are somewhat puzzling and there seems to be dose-response. I tend to agree with user 750802 [Expert 9] regarding consitency of the overall database on carcinogenicity

User-37600 [Expert 2] asks an important question. Should we rule out "portal of entry effects" as irrelevant for human risk assessment? Do we ignore UV-induced melanomas in humans? The answer is no. However, if a compound given by daily oral gavage causes forestomach tumors in a rodent model, should we conclude that the compound would be carcinogenic in humans exposed by inhalation? Not necessarily. Since inhalation is the primary route of human exposure to 1.3-D, I would not rule out the importance of lung tumors in an inhalation study.

It seems that we all agree that a primate study is not warranted.

I agree with the previous comments that AOP investigations are appropriate when we have clear evidence of an increase in tumor incidence.

The TK data indicate that the 1.3-D plasma exposures were higher than expected (linear) at the highest exposure of 60 ppm due to saturation of the clearance mechanism (glutathione conjugation). This fact does not rule out the possibility that the tumors were treatment-related. However, saturation of the clearance mechanism in the mice makes it very difficult to extrapolate this effect to humans because we don't know how to estimate the clearance of the compound in exposed humans. A linear extrapolation of the tumor incidence to lower human exposures is likely to yield erroneous results.

User-37600 [Expert 2] raises a valid point about the transitional cell carcinomas in female B6C3F1 mice in the NTP gavage study. The tumor incidences of 0, 16 and 42% in control, low and high-dose mice when viewed in isolation are impressive and suggest a carcinogenic effect in the bladder. However,
we need to look at these tumor incidences in the context of all the other relevant data. The presence of 1% epichlorohydrin in the test article is important. User-37600 [Expert 2] states that this 1% impurity is unlikely to have caused this effect based on EPA's data and on the absence of any prior information that epichlorohydrin could produce this effect. These are valid points. However, we have no information on the potential interaction between this impurity and 1,3-D. Is a synergistic effect possible? We don't know. Secondly, The same doses of 1,3-D were given to B6C3F1 mice by gavage in the Redmond, 1995 study with no suggestion of an effect on the bladder. Thirdly, Stott, 1987 reported no bladder tumors in their inhalation exposure study, a route that is more relevant to human exposure to 1,3-D. Finally, there is no evidence of bladder tumors in any of the other bioassays conducted in CD-1 mice mice and in F344 or CD rats. Therefore, looking at all the relevant data, the weight of evidence is that 1.3-D did not cause bladder tumors in multiple oral and inhalation bioassays with the exception of one study in which a potentially confounding impurity was present. Given this data base, I have to conclude that this one result was an outlier that could not be reproduced in multiple additional studies.

Perhaps one could estimate the contribution of say, 1% epichlorohydrin to the genotoxicity of purified 1,3-D by comparing the genotoxic effects of the pure compounds (i.e. if epichlorohydrin is 100x more potent than 1,3-D could it account all of the observed genotoxic effects?). In practice this might be difficult (but maybe not impossible) as conditions for assaying the pure compounds would have to be very similar. Without this information there is no way to know if contaminated samples should be discarded.

To user 553126 [Expert 5] : Among the studies present in the last 2 tables of the white paper and technically relevant for aneuploidy (bone marrow micronucleus test), none of them can establish whether 1,3-D is aneugenic or not for the following different reasons:
1) Micronucleus-bone marrow (Gollapudi 1985). The result is negative but the strength of the study is not sufficient to exclude either aneugenicity or clastogenicity since no evidence of target tissue exposure is present and the statistical sample (1000 PCE/animal) is limited. In addition, an unjustified difference in the percentage of PCE in the negative control group between male and female animals (56.8 and 68.8 respectively at the 24 hour sampling time) and within female animals at 24 and 48 hour sampling times (68.8 and 61.5, respectively) is present. Overall, this study cannot be considered adequate for risk assessment at present.
2) Micronucleus-bone marrow (Shelby 1993): Positive >150 mg/kg. In my opinion this study is not positive because the frequency of Mn-PCE observed at 150 mg/kg (0.31 %) is usually found in negative control animals. Here, the negative historical control data would have helped. In addition, as also noted before, no evidence of target tissue exposure is present. Overall not reliable for both aneugenicity and clastogenicity;
3) Micronucleus-bone marrow (Kevekordes 1996): Positive. In my opinion, this study, in addition to an inadequate study design (single animals in negative control), bears unrealistic values of PCE/NCE ratios which invalidate directly the study.
Overall, these studies cannot rule out the concern for both aneugenicity and clastogenicity.

I'm not sure where any further discussion of historical controls should go (here or on some other page), but I continue to be somewhat skeptical of Table 16 in the White Paper. As I wrote under topic 5.5 in the prior round, "Table 16 in the white paper represents extreme cherry-picking. The historical control rate for male F344 rats at Dow in the year of the study is 2% (pooled 5/250). The complete control rate in Table 16, for all years, is 4.6% (40/865). Only by picking the highest possible control rate for males (16% in one year) and adding an irrelevant value for females (8%) was the white paper able to inflate the control rate by an amount large enough to make the 18% incidence at 25 MKD seem POSSIBLY not significant."

So comments above to the effect that the liver tumor incidence in mid-dose males (12%) was within historical control rates is not really fair-- and of course that says nothing about the 18% incidence in the high-dose males. If these tumors are to be ignored for biological reasons, so be it, but I don't think it's appropriate to ignore them as "elevated in the experiment but not elevated if you happen to compare them to either of 2 hand-picked prior experiments out of the 17 that we have data for..."

Yes, everyone agrees that test in non human primates is not necessary.

I think that the 32P-labelling assay used by Zeiger to deny DNA adduct formation was too insensitive for small adducts at the time it was performed. The assay has become much more sensitive for smaller adducts in the meantime. This calls into question the conclusion that "whenever DNA or chromosome breaks is seen in the absence of adducts, a secondary mechanism of action always be considered".
However, the Young study offers strong support to the lack of genotoxicity in vivo,

To user-750802 [Expert 9] -- you raise a good question but I think one logical answer is "because positive findings have more weight than negative ones, unless/until the positive findings are EXPLAINED satisfactorily." If all we needed to dismiss, say, an epidemiology study with a relative risk of 5 is another one with a RR of 1 (or 0.5), then we would have failed to act on many "known" causes of human disease. You're correct that it's possible that all the positive findings in the NTP bioassays are explained by the impurity, but I'm just pointing out how weak that explanation seems to me-- especially w/respect to the female mouse bladder tumors.

I appreciate the comment by 37600 [Expert 2] to the effect that a 1% concentration of epichlorohydrin was too small to have resulted in the tumor incidences observed in the NTP studies. However, we don't know anything about the potential interaction between the 1% contaminant and the 1,3-D. In addition, we have other bioassays in which there was no possible interaction between these substances. Why would we not believe the results of these unconfounded bioassays as opposed to speculating about the possible contribution of the 1% contaminant in the NTP bioassays.

There are several suggestive positive results in the last 2 tables of the white paper, although they all have some questions. I do not think that the use of DMSO as a solvent necessarily disqualifies a test, as DMSO reacts with 1,3-D epoxide and not the 1,3-D and in some cases it is not known if the epoxide was present in the 1,3-D. As the question asks about potential I would say likely, but the risk under realistic conditions is negligible.

There is an error in the ref in the above comment - the paper referred to is R.E. Talcott, 1984. The Watson paper does show a clear mutagenic dose response for purified 1,3-D in the presence of metabolic activation, but this greatly reduced by glutathione.

To user 232578 [Expert 7] -- probably not our place to say, but if an important question (developmental tox) has been "answered" outside of the White Paper, then it would seem to be a good idea for the White Paper to reflect that (negative) evidence, with references.

To user 24419 [Expert 8] : I agree that nongenotoxicity should be "a consideration in the overall process," but I don't see any coherent MOA argumentation in the white paper, other than the TACIT argument that because 1,3D "is" nongenotoxic, it "must" have a threshold dose-response function. To the contrary, there are examples (and good theory) why some nongenotoxic MOAs could have linear, or piecewise linear, or sublinear, dose-response functions, without a population threshold at all, or with one that is so low as to be "not a consideration in the overall process" of risk assessment.

To users 477751 [Expert 11] and 915125 [Expert 6] : my (long) comment on Topic 5.5 (WOE subpanel) presents cancer slope factor estimates from the bioassays of epichlorohydrin alone, and in my opinion STRONGLY suggests that the 1% impurity was far too small an amount to account for the excesses in tumors in the NTP bioassays of 1,3D-- *especially* the female bladder carcinomas, since epichlorohydrin has never been found to increase this tumor type.

I am persuaded by the above comments that the results are equivocal.

I could not locate the Lawlor refs for the Ames test, so I can't evaluate them. I'd like to get the ref for the clean state of the art mutagenicity data referred to. The Watson paper reported a low, but apparently positive response for highly purified 1,3-D. However, no raw data were included and no dose response data were included. Metabolism of 1.3-D to its epoxide apparently occurs, but is minor pathway. However, the 1,3-D-epoxide is reactive with dG (Schneider, 1998) and based on reactivity of other alkyl epoxides would be expected to react with DNA. Thus, I believe a strict answer to the above question is yes. However, based on the low yield of the epoxide and its detoxification by glutathione, I do not think it poses a mutagenic threat in vivo.

Understanding of AOP (chemically agnostic toxicity pathway) and a chemical specific MOA can contribute to the evaluation of a response as treatment-related or not.

This is not to say that one must go through the long and resource intensive process of establishing an AOP on the OECD Wiki.

The WOE from reviewers suggests that the extant bioassays provide sufficient data to support a hazard WOE. Rather than require more cancer bioassays, I would recommend more cogent and logical explication of the most probable MOAs.

I support comments of 617591 [Expert 14] citing McGregor et al . One can observe NOGELs, or no effect levels for sensitive genotoxicity assays.

The genotoxicity or lack thereof is not an exoneration, but rather a consideration in the over all process of hazard characterization.

An honest question out of possible ignorance: how is it possible to ignore lung tumors **by inhalation** as a "portal of entry effect"? How is this different from ignoring melanoma among lifeguards as a "portal of entry effect"? Naively, perhaps, one would think that the first site of contact with a carcinogen would be more, rather than less, important than distal sites.

(separating this from my immediately prior comment, for emphasis)

I remain puzzled by the complete lack of attention, both in the White Paper and in this exercise to date, of one particular positive bioassay finding that seems quite important: the striking and dose-related increase (above a ZERO control rate) of transitional cell carcinomas of the bladder in female mice (NTP 1985-- 16% at 25 ppm and 42% at 50 ppm).

Can anyone explain why we should be completely ignoring these? I know-- "epichlorohydrin." EXCEPT that (1) epichlorohydrin has never to my knowledge been implicated in bladder carcinogenesis; and (2) if you believe the back-of-the-envelope risk assessment (see user-37600 [Expert 2] comment on the WOE topic 5.5 elsewhere on this site) using EPA's cancer potency (slope) factor for epichlorohydrin, it's virtually impossible for 1 PERCENT impurity to cause more than 1 or 2 tumors per 50 rodents at ANY site, let alone one where the impurity does not seem to be active.

Various sections of the White Paper, and various commenters, have at least coherent theories as to why we should ignore the liver adenomas, the forestomach tumors, and the inhalation lung adenomas-- but I've seen no attempt-- other than silence-- to explain away the bladder tumors...

I would prefer to see more systematic data analysis in support of mode of action evaluation. I see little value gained in testing 1,3-D in pregnant animals or non-human primates. I agree thoroughly with comments noting that the purpose of these evaluations of pesticides is to judge the likelihood of potential harm to humans through their use, rather than establishing an enormous data base for the sake of filling "gaps".

Most of these responses (those in the prior round and some of the comments to date) sidestep the original question by focusing on genotoxicity. The original question proposed a larger-N bioassay (not necessarily a "mega-mouse" study by any means) so that we wouldn't have to make qualitative ("increases above controls aren't really increases because in some years the control rate would have been greater...") and quantitative ("at this dose there were no statistically significant tumors") pronouncements with low power to make them.

A larger study would allow for slightly lower doses to be used, and for the uncertainty in historical control rates to make far less potential difference.

I suggest that there is not sufficient value of information potentially gained to warrant further long term cancer bioassays.

I suggest that the White Paper consider comments of 232578 [Expert 7] and others in refining their discussion of the inhalation bioassay results for relevance to likely human exposure scenarios.

The comment from user-292467 [Expert 12] has a lot of good information in it, but the notion that 10 ppm is a lower bound on the 30 ppm has NOTHING to inform it with respect to humans-- all the "should be sufficient" conclusion comes from data (such as it is) on rodents. I also don't fully understand how at the same time this user can claim that (1) exposures below the KMD (whatever that value is) are OK because detoxification proceeds faster than damage; AND (2) intermittent exposures above the KMD are OK because GSH can "recuperate." It would seem to me that "recuperation" only happens after depletion, which is exactly why I (and others?) have concern about intermittent high doses.

Re 37600 [Expert 2] comments: I don't think most risk assessors consider a portal of entry effect to be artifactual, but rather a consideration in evaluating bioassay results for relevance to a human exposure scenario.

The debate under question 7.1 concerns part (4) of this question above-- and so far I don't see much on this page that addresses the interspecies (and the human intraspecies) concern here. User 432123 [Expert 13] states that humans and rodents have the same metabolic pathways and isoenzymes, but assuming that's true, this provides no basis for QUANTITATIVE interspecies (or intraspecies) extrapolation for risk assessment purposes.

I submit that the majority of the WOE panel feels that 1,3-D is not likely to be mutagenic in vivo at human relevant levels of exposure. I note 37600 [Expert 2] 's point that non-genotoxic is not equivalent to not risky. However, also note that for most analysts "Risk = Hazard x Exposure", and that hazard includes a measure of dose response as well as identification of a potential adverse outcome. The current analysis is structured toward an estimation of risk based on consideration of most probable modes of action; I think the White Paper is close to achieving this goal.

User-850922 [Expert 10] says it well: (to paraphrase) "we have no idea what the range of human KMDs are or might be. We would have to know a lot more than we do now about average rates of various human parameters and their interindividual range."

So why are we leaping from a qualitative finding (there appears to be a KMD somewhere) to ANY kind of quantitative conclusion about where the human KMD actually is? Frankly, I find the third comment above (user-292467 [Expert 12] ) to misstate the question, which asks about humans, rather than rodents. It's possible that a factor of 2-3 (from a starting point of 30 ppm) might provide a lower bound for the RODENT KMD, but this provides zero information on what interspecies and intraspecies adjustments would be necessary to declare a human KMD.

Agree wholeheartedly with 232578 [Expert 7] re this statement "Before presenting the genotoxicity results on 1,3-D, I think it would be useful for the final White paper to provide some text upfront as to the explicit weight-of-evidence approach used for judging a potential for in vivo mutagenesis and a DNA reactive mode of action for cancer (ie how information is weighed and integrated). It appears that the White paper is weighing evidence as I have outlined above A. 1-4. But again this could be laid out clearly at the beginning of Section 3. Genotoxicity." I found a lack of clarity in several descriptions in the White Paper, due to too terse descriptions of logic as well as scanty discussion of criteria for choices.

Although not summarize in the White paper, it appear from other documents that there is little indication that 1,3-D poses a reproductive toxicity or teratogenic risk. Thus, given the negative findings in the in vivo DNA adduct/mutation studies and the dominant lethal test (recognizing peer review comment on lack of sensitivity) as well as the lack of concern for reproductive/developmental toxicity, there is not a scientific justification to pursue this line of testing based on available evidence. (as stated above "unless a clear clastogen" and then issue would have to be re-visited )

The first comment above (user553126 [Expert 5] ) is right on target, and suggests that the original question could be broadened to ask "is it possible that 1,3D is a developmental toxin, whether via a genotoxic mechanism or any other mechanism?"

Agree with analysis by 125195 [Expert 4] , as well as others supporting the use of the TK data in overall weight of evidence. While I am not sure I followed all points made by 37600 [Expert 2] , I also found some of the information in the White Paper to be lacking/

I think the original question was unclear and is being misinterpreted. The question probably doesn't ask about additional exposure to 1,3D on top of previous exposures to 1,3D itself, but about the RISK of 1,3D outside of the laboratory where humans are exposed to many other carcinogens. To be more blunt, I'd ask it this way: why assume that a nongenotoxic carcinogen is any "better" (less risky) than a genotoxic one??

I think user-553126 [Expert 5] (first comment above at this date) says it well-- "it depends." So my question remains-- if we don't KNOW whether incremental exposures to 1,3D are non-harmful (as they might be if 1,3D only promoted tumors, but somehow all environmental initiators were removed), then why are we seemingly exonerating it for being "nongenotoxic"?

I am not a PK expert and I hope others contribute to the discussion on this point. But the PK issue would not change my view that the slight increase in lung ademonas via inhalation in a highly susceptible strain/sex does NOT provide a basis for a concern in humans. Also as stated in comments above, there was no evidence of carcinogenicity in female mice and in either sex of the rat in the inhalation bioassays. Slight increase likely due to growth of spontaneous lesions and ability to score as adenomas because of size.

Yes, not necessary. And agree with comments on practicalities as well as ethical considerations given current directions to reduce animal usage.

Guess that there is agreement here and no need for debate ?

I agree that information on possible differences in composition of the test material in the two oral rat studies is crucial

A point to discuss is the conclusion that the lung tumors occur at concentrations that induces non-linear plasma kinetics. I do not think that this argument can be made since plasma concentrations of 1,3-D may not reflect target tissue concentrations in the lung. This should be related to exposure concentration. Metabolism in the target cells could be saturated, but would this would only be reflected in plasma concentrations if there is a significant 1st-pass in the lung ?

Agree that there is no need to do AOP development if the tumor response is not treatment related. Increase in growth of spntaneous lesion may all explain the weak response in mouse lung as suggested by user-232578 [Expert 7]

All, any conclusion if the lung tumors are related to treatment or just an artifact due to variable background ?

For lung, the mouse is very susceptible to GSH-depletion, but the MoA for other chemicals causing lung tumors in mice require cytotoxicity and hyperplasia. Based on absence of toxicity in the mouse lung and no lung response in the rat inhalation study, my conclusion is that the lung tumors are not treatment related. Issue with Scott et al, are much lower Cmax as compared to Kelly. Still struggle a little with the indication of a dose-response in Scott ? Mechanisms could be GSH-depletion at high doses/concentrations ? Agree that it is highly unlikely that direct genotoxicity plays a role in the responses.

Agreement here, even in case of negative results in a new study, the results of Scott et al require interpretation. The white paper does a good job addressing the issues

I was giving the lowest confidence score, consider upgrading this based on my own WoE.

My conclusion was further strengthened by the data on genotoxicity and toxicokinetics and the significant reduction in bw gain in Scott. Is there an explanation for the reduced bw gain ?

Looks like agreement, I would not add the NTP studies to an evaluation due to presence of epichlorohydrin and potentially other contaminants ?

Agree based on tumor incidences and some indication of dose response. I have difficulties interpreting the response in Scott et al. considering toxicokinetics since Cmax was much lower as compared to Kelly at the same dose level. Either the observation is an artefact, the rat strain in Kelly is much more sensitive, or the two formulations have different compositions ? Any other suggestion ?

Information on composition of the different Telone II formulations including impurities would be helpful to interpret the inconsistent results in the newer studies (Scott et al., and Kelly et al.)

The white paper includes a reference on a study with epichlorohydrin. I am aware that epichlorohydrin was used as a stabilizer in other chemicals that were tested by NTP in the 1970s (eg. trichloroethene). It could help to check doses of epichlorohydrin in these studies and correlate to tumor response in the target liver.

What is the concentration of 1,3 D in Telone II and in DD-92? Are they the same?

I agree that it is necessary to have an agreement that there is a clear carcinogenic response to investigate.

to users-477751 [Expert 11] and 750802 [Expert 9] - agree with you -

I guess we all agree that the new formulation of Telone II is not genotoxic. Only 2 studies have shown some evidence of carcinogenesis, which are the Scott study (1995) (Increased incidences of hepatocellular adenoma were observed in both high dose males and females and in mid-dose males) and the inhalation study in B6C3F1 Mice, in which statistically identified increase in benign lung tumors (bronchioloalveolar adenomas) was observed in male mice only exposed to high-concentration (60 ppm )1,3-D (22/50 vs. 9/50 in controls).
Carcinogens nowadays must be characterized according to their mode and mechanism of action. What would be the mechanism of action of 1,3D for carcinogenesis? Any evidence?

Study one indeed provoked different answers, but with a tendence to consider it positive for hepatocellular adenomas in high dose male and female animals and in mid dose males.

We could also discuss about the potency of epichlorohydrin. Are you aware of any carcinogenesis study with epichlorohydrin alone?

Just a comment, both the old and new formulations of 1,3 D are called Telone II. It has not been informed if old Telone II and new Telone II formultions have the same concentration of 1,3D. It is also not known what is the concentration of 1,3 D in DD92, which has been used in rat experiments. Could anyone inform this, please?

As noted, in risk assessment, we are judging whether there is a concern for humans. Thus, the question we are addressing is 1,3-D’s potential to induce mutation in vivo. Given that “metabolic profiles seems to be sex, route and species invariant”, the totality of the in vivo evidence does NOT support a mutagenic concern for humans (nor does DNA reactivity/mutagensis support a cancer concern). And if I am reading the comments correctly, I think there is consensus that the induction of gene mutation in vivo is of no concern. But there is a remaining issue regarding clastogenicity as pointed out in the comments. Also, Although I can not rule this out, I do not think it is likely that 1,3-D would be a direct clastogen (vs secondary consequence) when it does NOT appear to interact with DNA and induce gene mutations (albeit there are exceptions). Also, I do not view this as a major issue in the context of a weight of evidence for carcinogenicity given that results of the relevant chronic bioassays do not provide convincing evidence of carcinogenicity. However, if the sponsor wants to strengthen the conclusions regarding genotoxicity, as pointed out in 6.5, “the cleanest way to resolve this” would be a new study (OECD TG 487 or 474).

Again concur with user # 850922 [Expert 10] . This question is almost the same as Q.1 and the later part is difficult to answer.

I agree with user # 850922 [Expert 10] that the information available is limited but based on literature review I am of the opinion that 1,3-D does not get accumulated and is rapidly excreted in urine resulting in a short biological half-life.

As stated in my earlier response that "cis -13D is more active as an alkylating agent than the trans-1,3D since the cis-form favors cation stabilization due to steric hindrances and neighboring effects of the choline atoms (Neudecker et al., 1980)." Therefore the relative contribution of each of the isomers to the toxicity is not only variable but also difficult to predict at higher levels (> 300 ppm) of exposure.

As I said to estimate the amount of free drug in mouse and human plasma to correct the KMD while the unbound fraction is not the same in the animal model and human. This only to estimate the KMD since it would not change the risk assessment considering the high MOE.
Performing intermittent versus repetitive dosing may indicate whether the GSH is involved since in intermittent dosing the CL of 1,3-D should increased due to the GSH recovering.

I don't think it is possible to make an assumption that that the KMD would probably be about 10 ppm based on limited animal exposure data.

I agree that the AOP approach can be useful in certain situations. But before attempting mechanistic research, there should be agreement that there is a clear carcinogenic response to investigate. In this case, the benign lung response is barely outside of HCs, and thus questionable. Also, the research effort would be very challenging given that the response occurs in a genetically susceptible strain/sex with a high and variable background. If considered treatment related, this slight increase in adenomas (ie size) would likely be explained by an ncrease in growth of the spontaneous lesions (promotion).

No more comment considering the already high margin of exposure.

The reviewer does not totally agree; based on figure 5 it could be that KMD=61 ppm, but based on figure 5 KMD could be lower than 20 ppm. One thing is for sure it is somewhere between 15 and 61 ppm

No more comment from my side since everybody agree about the possibility of the non-linearity above 20 ppm. However, 3 reviewers also suggested that non-linearity could be as low as 20 ppm or slightly lower to 20 ppm. Overall, all discussions suggested a non-linearity close to 20 ppm. The effect of DEM is an indication that the metabolic rate becomes lower (even if there is a depletion effect).

No more comment since everybody agreed.

No more comment from my side since there is nobody saying clearly that GSH cannot be depleted.

PLEASE DO NOT PROVIDE RESPONSES OR COMMENTS TO SUGGESTED FOLLOWUP QUESTIONS HERE. PLEASE SEE THE LAST 3 TABS WHICH CONTAIN THESE FOLLOWUP QUESTIONS ALONG WITH RESPONSES

PLEASE DO NOT PROVIDE RESPONSES OR COMMENTS TO SUGGESTED FOLLOWUP QUESTIONS HERE. PLEASE SEE THE LAST 3 TABS WHICH CONTAIN THESE FOLLOWUP QUESTIONS ALONG WITH RESPONSES

PLEASE DO NOT PROVIDE RESPONSES OR COMMENTS TO SUGGESTED FOLLOWUP QUESTIONS HERE. PLEASE SEE THE LAST 3 TABS WHICH CONTAIN THESE FOLLOWUP QUESTIONS ALONG WITH RESPONSES