Rat Liver Tumor Mode of Action

The word threshold can be problematic and perhaps should be clear the the MOA based on current literature support the use of standard BMD approach and NOT a linear no threshold default as used by EPA in the absence of data. The ROS MOA is supported by several studies and the pleiotropic effects of ROS on biological systems will enable a range of BMD calculation and an assessment of the most sensitive endpoint that can be used in a risk assessment.As suggested by Young et al and Expert 3 high-dose human occupational exposure scenarios are very unlikely and in my opinion, BMD based on the high doses required to produce a biological effect in rodents will likely yield order of magnitudes human equivalent exposures.

yes

I agree with expert 4, MIE have been defined as part of the AOP process for toxicology outcomes. The AOP for Dioxane includes this and the citation below support this hypothesis. Again, ROS is produced (Totsuka et al., 2020) at high dose associated with metabolic saturation. ROS based MOA is a threshold based MOA (AOP 296) and will enable a risk assessment based BMD and a POD to generate human equivalent concentrations and assess margin of exposures. 

Totsuka Y, Maesako Y, Ono H, Nagai M, Kato M, Gi M, Wanibuchi H, Fukushima S, Shiizaki K, Nakagama H. Comprehensive analysis of DNA adducts (DNA adductome analysis) in the liver of rats treated with 1,4-dioxane. Proc Jpn Acad Ser B Phys Biol Sci. 2020;96(5):180-187. doi: 10.2183/pjab.96.015. PMID: 32389918; PMCID: PMC7248212.

No
Although high dose DX exposure at metabolic saturation is associated with liver tumors in female mice, but this alone does not provide sufficient cause for the occurrence of the biological events (cytotoxicity, cell proliferation genotoxicity, epigenetic changes, clonal expansion of an altered cells) known to occur to produce a tumor response.
Metabolic saturation does not appear in the literature as a KCC (Key Characteristics of human Carcinogens), and is not regarded as a cancer key event.

As stated repeatedly, all OECD compliant are conducted at high doses, up to 1000 mg/kg, that likely produces metabolic saturation for most compounds and many of these do not cause liver tumors.

In vitro studies using human hepatocytes at low doses and doses that cause metabolic saturation combined with IVIVE may yield human equivalent concentration that can provide a wide margin of safety. Going to human hepatocytes can also provide MOA  bypass issues in rodent with human-relevant data

Expert 5 design will answer several questions in a single study and will also provide BMD data for several endpoints (ROS lesions, enzyme induction, histopathology assessment, cell proliferation) that can be used to identify PODs. A 28 design will not only allow an assessment of genotoxicity in bone marrow, but also in a target tissues (e.g. liver); an assessment of DNA damage and role of ROS in contribution to that damage by using ROS-lesion specific enzymes can be done. Secondly, a 28 day study will also allow an assessment of mutations by error-corrected DNA sequencing in wild type mice at doses that are known to cause tumors; this will put to rest whether or not there is a genotoxic/mutagenic MOA using time points and doses that are OECD compliant and useful for regulatory submissions 
Although the ETAP was made for "low information" chemicals it was based on several very well known prototypic  compounds using 5, 28, 90 day studies that cause a host of biological events that captured several MOAs.  Secondly the cost of a 28 day study and more importantly the "turn around time" to get useful data from a lab that has run these types of studies is around 6 - 9 months. I do disagree on ETAP philosophy to use the first dose that causes any change in gene expression. In the 28 day study design the investigators can use apical and MOA based transcriptomic changes to calculate BMD- these usually align quite nicely. 
With respect to dose saturation as a key event , as I stated early in my mind this happens in all OECD based testing based on the language written - the top dose is based on the fact that the next highest dose would be lethal or cause obvious adverse effects. Secondly, the current use of key characteristic of carcinogens is not chemical specific. I do agree that the high doses used that saturate metabolism results in the events e.g., ROS from Cyp2e1 leading to indirect genotoxicity or cell proliferation.
If designed properly a 28 day study and also include levels that approach worst case scenario human exposure levels to nail down NOELs,  support nonlinear threshold, and to demonstrate that the dose levels required to elicit biological effects as stated by Rev. 3 have no quantitative relationship to levels of exposures experienced by humans.

The 28 day design can provide not only traditional MOA-based endpoints but also transcriptomic data that can support the MOA for the observed biological effects. As importantly, these endpoints can be combined to derive BMD, POD and calculate a human equivalent dose with margins of safety for quantitative risk assessments.

In my view a MOA based 28 day study using several doses known to be NOEL and doses known to cause tumors is the best option. Several doses means around 8 doses to clearly define the shape of the dose response curves particularly at low doses and to leave no doubt what the NOEL and the BMD10 are. 

One note, a preliminary assessment of transcriptomic MOA can be done using existing FFPE tissues from an archived 28-day study. This can be very reliable source of RNA for gene expression assessment and at a fraction of the cost of doing a de novo study.

I agree with Expert 3 concerning the need for additional toxicokinetic studies to better understand metabolism saturating conditions for 1,4-DX.  The unpublished study in cyp 2E1 knockout mice referenced in the commentary of Wang et al. (2022) also provides important supporting information on the role of cyp 2E1 induction in the hypothesized MOA. Though referenced as a "manuscript in preparation" in the 2022 paper, I've not been able to identify a subsequent publication.

The existing MoA framework analyses provide for addressing an initial question that if the hypothesized MoAs could plausibly occur in humans regardless of dose conditions required to elicit the MoAs, the tumors are regarded as potentially (qualitatively) human relevant.  However, and importantly, the MoA then allow for consideration of the whether or not the tumor are quantitatively relevant to human, e.g., humans have quantitatively lower metabolism to hypothesized toxic metabolite(s) driving the MoA.  Saturation of metabolism, resulting in high-dose specific tissue accumulation of parent DX driving the hypothesized MoA(s), is a parallel scenario potentially lacking quantitative human relevance if the doses sufficient to saturate DX metabolism are demonstrably quantitatively implausible in humans.  Human DX exposure scenarios support such a conclusion that, e.g., no humans would tolerate a lifetime of likely very high concentration nasal irritating exposures to DX, and lifetime drinking water scenarios in the high 100s to 1000s of ppm necessary to elicit metabolic saturation are a completely unrealistic human exposure scenario.  Thus, although the biology associated with liver tumors in rodents theoretically exists in humans, such biology would never be stimulated given that no plausible human exposure scenarios would quantitatively sufficient to result in doses required to initiate even the first biological events hypothesized for the MoA

The nasal tumors resulting from drinking water treatment have been reasonably attributed, using inferences from dye-treated water exposures to nasal cavities, to site-of-contact irritation associated with rodent nasal contact with nasal-irritating concentrations of DX in drinking water and thus lack quantitative dose-relevance to humans (i.e., humans remove themselves from highly-irritating nasal exposures).

I agree with expert 6 observation.  The "threshold" for all(?) high-dose specific tumor MOA-hypothesized events appears to be plausibly dictated by confronting the animals with an initial dose regimen for which the organism is unable to mount relevant detoxification MoAs over the lifespan of the chemcal treatment.  Importantly, human toxicokinetic evidence (Young et al) suggest likely high-dose human occupational exposure scenarios are unlikely to trigger the responsive stressor event (cyp2e1 induction associated with excessive DX dosing) and thus not lead to any of the ensuing postulated/evaluated subsequent "toxicity-inducing" effects. 

I concur with comments of expert 4 and responding debate of experts 2 and 4.  I view the inability of the organism to metabolically "handle" the high doses presented to it through its normal biological status as representing evidence of a chemically-induced stressor event (which can be readily detected by a dose-disproportionate shift in plasma toxicokinetics) , i.e., the organism senses the dose challenge as being outside of its existing homeostatic capabilities and responds to restore that homeostasis by altering the capacity of its detoxification resources.  If the responding and biologically-limiting homeostatic defense mechanisms remain overwhelmed by a very high dose stressor (as is for DX), it moves the organism from a condition of adaptive "eustress" to toxicologically important "distress" (Sies  http://dx.doi.org/10.1016/j.redox.2016.12.035).  Although Sies developed the eustress/distress concept specifically for oxidative stress, its fundamental underlying principles reasonably extend to metabolic saturation

The Lafranconi etal (2021) 90-day drinking water toxicokinetic data provide valuable insight that metabolic saturation of cyp2e1 metabolism is present even post-induction at 90 days (i.e, the substantial dose-disproportionate increase in plasma DX at 6000 ppm).  Importantly, these toxicokinetic data are collected using a dosing regimen, drinking water, producing liver tumors in a chronic bioassay. Thus, it is reasonable to assume that DX metabolism is saturated very early on (if not immediately for inhalation) following initiation of DX treatment.

I believe refined toxicokinetic studies would not be cost or time consuming to conduct (other than repeated dose inhalation and drinking water administration) in that validated and appropriately sensitive DX and HEAA analytical methods are readily available.  However, as noted by expert 6, such studies would go a long way in supporting a conclusion that whatever MoAs might be driving tumor outcomes, they are demonstrably specific only to dose regimens sufficient to challenge sufficiency of DX first-order toxicokinetics.  If such can be more robustly confirmed, any further MoA refinements are unlikely to better inform human tumor relevance, i.e., whatever is going on at such metabolically saturating conditions, it very likely involves a multiplicity of highly complex toxicological interactions for which a further characterization of which events are primary or associative will be challenging if not impossible to reliably identify.  However, what can be confidently established, however, is that those potentially complex events are restricted to saturating dose conditions that have no quantitative relevance to realistic human exposure scenarios.  

I agree with general consensus that the overall data support a non-genotoxic threshold-based (high-dose specific) tumorigenic MoA for liver tumors.  In addition, there is reasonable debate on whether metabolic induction most likely and primarily of cyp2e1 (which data indicates as an enzyme stabilization, not new protein synthesis), represents a plausible molecular initiating event to the overall MoA hypothesis.  However, it also seems reasonable to posit that the true initial key event is not induction per se, but the immediately earlier step that infers induction is only initiated when the initial systemic doses are sufficiently high enough to overwhelm the initial capacity of cyp2e1 to adequately maintain rapid first-order metabolic clearance of DX.  The data also indicate that even post onset of induction occurring after repeated dosing, the delivered systemic inhalation or oral doses remains sufficient to saturate even the induced state of cyp2E1, resulting in a sustained increased in systemic DX concentrations.  The high-dose specific and immediate saturation of metabolism appears could also be reasonably viewed as the key driving force responsible for activation of subsequent and likely complex set of high-dose specific events as thoroughly outlined/debated by other experts (recognizing, of course, that such complexities can nonetheless be distilled down to the identified higher-level and well-established key events plausibly driving high-dose and threshold specific tumor outcomes).  

Consideration of initial and continuing repeated dosing sufficient to overwhelm metabolic capacity represents perhaps the most important molecular initiating event in that the overall data indicate that at dose levels insufficient to overwhelm metabolic DX clearance, many if not most of the following postulated KEs critical to tumor outcomes do not occur.  Such is seemingly clearly apparent with DX, i.e., tumors are consistently restricted to doses that are greater than both initial and post-induction metabolic capacities of the animals. Picking up on this consideration, international guidance on the design of animal rodent toxicity bioassays strongly and consistently encourages that a priori consideration of dose-dependent toxicokinetics should play a pivotal role in top dose selection for animal toxicology studies (and logically conversely, facilitate interpretation of quantitative relevance of rodent tumor human risk relevance when such a DX-like toxicokinetic picture is developed post-hoc to bioassay conduct).  This advice rests on long-established and extensive evidence that saturation of ADME can set in motion a spectrum of potentially interactive and complex MoAs that are unique to the resulting dose-disproportionately high plasma/tissue concentrations. 

For DX, evidence of metabolic saturation was observed only shortly after the earliest DX chronic bioassays, e.g., the Young et al late 1970s-era studies.  Given the above contention/perspective, however, the current status of dose-dependent toxicokinetic studies could and should be supplemented to better characterize at what doses and by what human-relevant routes and modes of administration (inhalation and drinking water) likely present metabolism-saturating conditions.  At present there is a particular absence of appropriately designed single- and repeat-dose inhalation toxicokinetic studies that would more clearly define the range of doses at which DX metabolism transitions to clear evidence of parent DX dose disproportionality.  Such studies would ideally characterize plasma and/or liver DX and HEAA, urine HEAA and exhaled DX, and include sufficient time-course data to identify Cmax and AUC determinations. Although similar studies could considered for drinking water, the Lafranconi et al. (2021, Fig.4) drinking water toxicokinetic study is likely adequate to demonstrate that metabolic saturation is present at 2000 ppm and greater.  This study's conclusions might have been more robustly supported if a differing timepoint sampling regimen had been used for each of the selected doses and sampling on study days 7, 28 and 90.  Instead of collecting just a single plasma sample from each animal (possibly reflecting a Cmax value depending on when the samples were collected), recent toxicokinetic studies reported by Saghir et al. have shown that for chemicals with short parent rodent plasma half-lives (as is case for DX), sampling of animals as few as 3 times during the course of a 24 hr period while remaining on diet or drinking water treatments can reasonably provide an estimate of 24 hr AUCs of parent and possibly key metabolite(s).  Conduct of such toxicokinetic refinement studies, reflecting dose/exposure conditions reflecting dose-response conditions used for DX chronic bioassays, would be (in this expert's opinion) an extremely value-added dataset in demonstrating whether or not tumors findings are confidently restricted to doses only sufficient to saturate DX metabolism (both initially and after onset of metabolic induction).  Given the dose-specific tumor profile of DX, such information would substantially support a conclusion that, absent doses sufficient to saturate metabolism, subsequent postulated key events would likely be sufficiently attenuated or even absent such that tumorigenicity does not result.  The existing DX MoA data highlighted by other experts certainly support this, and particularly the state-of-art toxicogenomic studies indicating that DX-induced toxicological/biological signals are restricted largely and/or only to doses saturating DX metabolism (further informed by refined toxicokinetics).  Also consistent with the metabolic saturation hypothesis is the commentary of Wang et al (2022; a member of the Charkoftaki and Chen research team) in which they describe an unpublished12 week 5000 ppm drinking water study in cyp2e1 knockout mice that reduced plasma HEAA by approximately 85% and attenuated both hepatic oxidative stress and the Nrf2 response. However, they also noted "unexpected" increased in liver 8OHdG and suppression of DNA repair and postulated that such events might be due to presumed DX-parent mediated events independent of cyp2e1 induction key event.  Such speculation is not implausible in that, unless an unidentified supplemental clearance mechanism beyond cyp2e1 metabolism was available in these KO mice (e.g., triggering of CAR activation), the plasma/liver concentration of DX in would reasonably have otherwise been extremely high with possibility of initiating any number of unidentified MoAs.

The point of offering this additional perspective is that if it can be more confidently shown that DX tumorigenicity hinges entirely on doses saturating DX metabolism and under dosing regimens associated with cancers, the existing DX MoA database should be viewed as complementary sufficient to support a high-dose specific MoA that is quantitatively not relevant to human risk given the wide disparity between the animal tumor-producing and metabolically saturating DX doses and those realistically encountered by any human chronic exposure scenario.

I agree that the transcriptomics assessment can be helpful for the early consequences of toxicity; however, this may not be predictive for the multistep process of carcinogenesis since those additional steps may profoundly alter the cellular landscape contributing to or emanating beyond the early transcriptomic consequences. The divergences in transcriptomics, proteomics (including phosphoproteins or other active forms), and additional effectors bearing upon cellular changes, is another matter altogether that could be confounding.   

I also support the recommendation of Expert 5 concerning the nature of study that may provide valuable additional information on the sequence and dose-response/temporal concordance of key events. 

While the EPA Transcriptomic Assessment Product recommended by Expert 1 may provide some additional information on dose-response relevant to hypothesized early key events, it’s been designed for a different purpose, namely identifying likely most sensitive non cancer endpoints for the development of transcriptomic based reference values (TRVs) for data poor substances.  EPA defines data poor substances as those with “no existing or publicly accessible repeat dose toxicity studies or suitable human evidence”.  The TRVs are considered relevant for “effects other than cancer or related to cancer if a necessary key precursor event does not occur below a specific exposure level”. A more focused design to consider dose-response for hypothesized key events at several levels of biological organization such as that proposed by Expert 5 based on the available mechanistic data on 1,4-DX (debate on Question 1.2) seems likely to provide information more relevant to the necessary prerequisite to apply a threshold based approach.

The available literature provides good evidence for the critical metabolic steps in carbon tetrachloride hepatotoxicity. For 1,4-D we know less and there seems to be no further literature which may help at this point to understand better what the molecular initiating events leading to cell death are. Also the fact that a few classical inducers (acetone, isopropanol) of CYP2E1 are not typical hepatotoxicants raises some doubts about  the hypothesis that induction per se is the source of damage. This statement does not mean, that I would not envisage oxidative stress to play a role in 1,4-D toxicity. 

I agree with expert 5. The fact that human exposure levels are or have been lower than in the top doses in the rodent studies most probably explains why the human epidemiology is not showing an increase in liver tumors . This observations supports the notion that there is no evidence for a fundamentally different MoA in humans.

I agree with the colleagues that there is some indication that tumors in the nasal mucosa may arise from a similar MoA (cytotoxicity, oxidative stress, regenerative proliferation..). 

The rat mesothelioma of the processes vag. peritonei have been used in risk assessment of chemical carcinigens although this tumor type and that localization is very rare in humans. I can not tell anything substantial with respect to the mode of action.

 

The existence of a threshold in this case is more than likely (a strong evidence!). Clearly, the size or range of the threshold may vary with species, strain etc., but this is obvious for thresholded effects. Nevertheless, the range of onset of toxicity seems to be similar in rat and mouse. I do not see a reason to believe that this assumption was not true for humans.

A molecular initiating event (MIE) sounds like a type of biochemical effect or interaction. The saturation of metabolic clearance most probably results in such an event, but should not be called an MIE.

First of all, I support the suggestion made by expert 5 although the experiment sounds like a big (and expensive) effort. Nevertheless, measuring several key parameters in parallel is crucial to understand better both the dose-response and the MoA. Secondly, I suggest to explore further possible modes of action by targeted experiments including cell culture assays, e.g., in hepatocytes. It would be very helpful to understand the molecular events leading to cytotoxicity.

There are several uncertainties surrounding what ROS species since there are limited analytical techniques beyond 8-oxoG and Comet assay using enzymes that cut at ROS-derived DNA adducts. High doses to needed to elicit these responses , there are limited/no data showing dose-response for DNA breaks (a common ROS DNA lesion) as measured by H2AX. 
The studies measuring ROS only had these types of data (H2AX, Cyp2e1 induction) or high dose dioxane. My bias would be to use the EPA's design for a 5 day transcriptome study that includes liver histopathology - I think this approach will provide a RfD that EPA has endorsed (this approach is ongoing for 18 PFAS compounds). I served as an ad hoc reviewer of the EPA ETAP and the data appear to be very much in line with 2 yr bioassay data.
Secondly not sure it "matters" on source of ROS or specific ROS species since there could be many and could take yrs chasing high dose artefacts. 

For number 2, I am uncertain if specific xenobiotic metabolic saturation can be considered a key event. Regulatory toxicology OECD testing requires testing up to 1000 mg/kg day, so in my biased view this is likely true for all regulatory toxicology studies; OECD 408 (90 day repeat dose study states "the highest dose level should be chosen with the aim to induce toxicity but not death or severe suffering").  This is why careful dose-response study capturing dose levels used in the cancer bioassay can be used to define that shape of the dose-response curve at low doses, determine an accurate NOEL for use in risk assessment. 

Given the high dose needed to elicit the biological effects of dioxane in rats and an ROS based MOA as a key event a transcriptome based BMD will likely reveal a large margin of exposure to humans.

I agree with expert 2. there is insufficent data to define a MOA for these other tumor types.      

In considering the possible mode of action for 1,4 DX  several data gaps should be addressed.  Specifically linking the treatment of 1,4 DX to liver cell proliferation, oxidative stress , and cytotoxicity in a dose response manner.  A simple short term in vivo study has been used for other compounds to address the liver mode of action and could be used to fill some of the data gaps.  Suggestion: 

Treat mice or rats (use the chronic bioassay strains) with 1,4 DX  (control and at least 4 doses including the tumorigenic dose and one dose higher and two doses lower) for 7 and 28 days continuous. At 7 and 28 day sampling times measure liver histopath and also cell proliferation (either via Brdu pumps or Ki67), measure liver serum enzymes, measure oxidative damage (lipid peroxidation,8OHdg), measure Cyp2E1 activity, secure liver RNA for future pathway analysis. These results will show a dose response correlation between cell proliferation, possible necrosis, oxidative stress, and metabolic saturation. These data may help resolve the ox stress versus metabolic saturation debate as the initial key event.   If necrosis, serum enzymes and cell proliferation occur at doses that induce oxidative stress but not metabolic saturation the oxidative stress should be considered the key event.   

In reviewing the excellent and thorough comments from my colleagues, it appears that two potential initial pathways have been discussed. One the potential for metabolic saturation of 1,4 DX  that in turn will result in cytotoxicity subsequently leading to hepatic compensatory (regenerative) hyperplasia and liver tumor formation.  The second is the induction of oxidative stress/damage that appears to involve CYP2E1 which culminates in a cytotoxic response  and/or an indirect genotoxic response again leading to liver tumors. These two pathways may not be exclusive with regard to rodent liver tumor mode of action if both eventually lead to a cytotoxicity , hepatocellular necrosis, and hyperplasia. 

Before defining the MIE or first key event we may want to step back and review the extensive literature base for rodent liver tumor modes of action.  Besides the first key event(s) the pathway involved is similar for “nongenotoxic” carcinogens.  That is after the initial insult the next key event (lets label as key event 2) is the induction of hepatocyte growth (usually measured by cell proliferation although inhibition of apoptosis has also been suggested as a mechanism for increase liver growth).  The increase in cell proliferation results in the selective clonal expansion of preneoplastic cells (Key event 3) resulting in the clonal expansion of focal lesions.  Selective focal lesion can progress into liver adenomas and carcinomas  (Key event 4).  This process seems to hold true for receptor mediated and cytotoxic modes of action.   

Important considerations for the rodent liver tumor mode of action are that each key event exhibits dose response characteristics and that the series of key events are temporal, requiring completion of the previous key event to progress to the next key event.  Therefore, although a key event may be “activated” the insult must be enough to allow further progression.  

Another important point when considering rodent liver carcinogenesis is that both in rats and mice the formation of preneoplastic cells appears to occur spontaneously. This is supported by the appearance of preneoplastic foci in untreated rodents as well as the observation of liver tumors in untreated rodents. Therefore, in key event 3 the proliferation stimulus may be targeting already present preneoplastic cells or alternately the compound may be inducing new preneoplastic cells.  

An important component in the liver tumor mode of action is Key event 2 (cell proliferation).  The results shown are not convincing to me that there are strong data to support a linkage between metabolic saturation or oxidative stress/damage and cell proliferation in the 1,4 DX mode of action.

Questions that appear to remain are: 

1.       How is the oxidative stress/damage produced by 1,4 DX?  Is it through futile cycling? Which oxygen species are involved?   Is there a dose response pattern for oxidiative stress induction that correlates with cell proliferation and cytotoixcity (necrosis) 

2.       In the case of  metabolic saturation, is the parent compound responsible for the cytotoxicity or is there another metabolite?  Does the metabolic saturation occur at doses that induce measurable necrosis and induction of cell proliferation?  

As indicated in my comments on round 1, evidence for direct genotoxicity in target tissues is limited. The high doses required to induce MN in the bone marrow do not meet 2016 criteria for a MN assay - doses too high based on length of exposures (OECD 474). The origin of genotoxicity in bone marrow is uncertain but is high dose irrelevant to human exposure levels. The induction of DNA strand breaks in liver as measured by H2AX only occurs again at the top dose of 5,000 mg/L in drinking water. 
In my view irrespective of origin oxidative stress caused by high doses of dioxane this is a threshold based MOA and BMD should apply. 
Secondly if more work is to be done I suggested to use the design put forth by US EPA in the EPA Transcriptomic Assessment Product (see: https://www.epa.gov/etap): a 5 Day transcriptomic study in liver will fulfill 2 criteria - insight in to MOA to identify tipping points for key events (e.g., NrF2 transcription regulator of liver oxidative stress response) and as importantly develop a traditional threshold based  RfD for use in risk assessment - EPA has "endorsed" this approach based on several studies showing that a short term (5-28 day) transcriptomic based BMD is similar to a 2 yr cancer bioassay based BMD. Combined with PBPK modeling and reverse dosimetry a margin of exposure and can be calculated that will reflect the high dose induced biological effects not relevant to levels of exposure experienced in the environment.

Oxidative stress induced cytotoxicity by high dose dioxane causing cell liver regeneration and proliferation can "fix" mutations caused by either ROS-DNA adducts or lipid peroxides as liver cells undergo necrosis/apoptosis. This MOA is threshold based.

This seems an important area of further investigation, from a regulatory perspective.  The bar for acceptance of an hypothesized mode of action leading to a threshold approach is often much higher for multi-site carcinogens (See, for example, rationale of the Michigan Department of Environmental Quality at: https://www.michigan.gov/-/media/Project/Websites/egle/Documents/Groups/TSG/Subcommittee-and-Workgroup-Reports/presentation-2013-10-08-dioxane.pdf?rev=5f6e0fd625584610a0e0a912e034d255)

The molecular initiating event is defined as "a specialised type of key event that represents the initial point of chemical/stressor interaction at the molecular level within the organism that results in a perturbation that starts the AOP/MOA" (i.e., the first key event).  It was introduced in the context of AOPs to incorporate evolving information on molecular level interactions.  I don't believe that there was any intent to confer special significance to the MIE as a more or most important determinant within a described pathway though it may be being interpreted this way.  As far as I'm aware, the terminology was developed in full recognition of the need to trigger subsequent key events (each with relevant dose-response relationships).

The effects of dioxin on mitochondrial respiration leading to sustained increases in reactive oxygen species are established in excellent studies (Senft et al. Toxicol Appl Pharm 2002; 178:15-21). This will be significant for early or continued mitochondrial and nuclear genome damage.   

I recommend the literature that pertains to hepatic carcinogenesis involving oxidative stress and continued liver injury, e.g., chronic HBV and HCV, alcohol-associated liver disease, and MASH. 

I suggest careful analysis will be helpful of somatic mutations including in the setting of pre-existing conditions, e.g., hepatic steatosis.   

I agree with the nuances articulated in the above responses from multiple Experts. The hepatotoxic aspects should be common to animals and humans. 

There's insufficient information to link the mechanistic basis of tumors in various sites - e.g., the somatic mutations and their consequences have not been developed.  

The threshold can shift on the basis of multiple metabolic and other factors or alleles predisposing to alterations in mutagenesis or cell fate. Therefore, this aspect requires a context-driven approach. 

I very much agree with Expert 5. The term MIE can be greatly misleading for the multistep process of carcinogenesis and certainly for this case. 

I agree with Expert 5 that metabolic saturation represents only the first event. This process should continue over time for successive cell populations but this aspect has not been appropriately determined in the available studies.  

I would like to agree with and emphasize the DNA damage aspects pointed out by Expert 4. Moreover, the somatic mutation landscapes and nature of adaptive cell origins, including the interactive signals and events driving clonal origins of cell populations (healthy plus metabolically challenged livers) should be significant.  

The role of oxidative stress in carcinogenesis is certainly relevant for two major aspects: a) cytotoxicity leading to oxidative DNA lesions and the more serious DSBs that would eventually result in the depletion of damaged cells, as is evident from the studies cited above, including in the details provided by Expert 4; and b) the ongoing somatic mutations in hepatocytes, which can accumulate in the healthy liver with progression under chronic injury conditions (Brunner et al. Nature 2019; 574L538-42), and also during metabolic dysfunctions (Wong et al. Cell 2023; 186; 1968-84), leading to clonal selection pressures. The oncogenic clones won't be related to cells that can no longer cycle, but involve additional cell populations that may arise during the adaptive or secondary phases of the injuries with multistep mutagenesis giving rise to clonally derived malignant cells, including from stem/progenitor origins (e.g., Herms and Jones. Annu Rev Cancer Biol 2023; 7: 189-205). 

Might the metabolic basis of oxidative stress for 1,4-DX reside in mitochondrial dysfunction leading to damage in both mitochondrial and nuclear genomes to constitute the initiating event of growth arrest in damaged cells followed by continued somatic mutations in other cell populations that eventually will result in oncogenic cell clone? Despite the absence of a "direct" genotoxic mechanism for 1,4-DX, the mutagenic selection pressure and clonal origin of cancers are highly probable. This should require focused studies. Similarly, the clonal carcinogenesis aspect is likely to be amplified in the setting of ongoing metabolic or other injuries in the liver, which too should require further studies. 

The role of CYP2E1 for mutagenesis is neither unique since multiple other CYP isoforms and detoxification mechanisms are typically involved nor should this be an independent initiator of clonal carcinogenesis.    

Two open questions seem to be 

1. Is oxidative stress the key event in DX cytotoxicity, the latter  finally leading  to regenerative proliferation?

2. Is CYP2E1 induction the key event in oxidative stress elicited by DX?

With respect to the role of oxidative stress, I agree with a number of experts that there is some good indication for oxidative stress (e.g., Chen et al, 2022 in mice) at high dose levels. From my perspective,  it is not entirely clear, however, what the molecular cause of oxidative stress really is.
The role of induction of CYP2E1 in the chain of events 'cytotoxicity-regererative proliferation-tumour' reads well but there are open questions. Wang et al. (2022) state, e.g., that 'Unexpectedly, subchronic exposure to high dose DX induced a trend towards an increase in oxidative DNA damage (8-OHdG levels) and suppression of DNA damage repair (γH2AX/H2AX ratio) in the liver of Cyp2e1KO mice'.

It also needs to be kept in mind that classical inducers of CYP2E1 such as acetone or isopropanol are not carcinogenic in rodent liver and are not typical hepatotoxicants. Thus, particular circumstances need to be suggested to make the hypothesis plausible, that CYP2E1 induction per se is the key event for liver toxicity. In the case of ethanol (frequently cited!), the current view seems to be that CYP2E1 induction leading to enhanced metabolism towards acetaldehyde formed from ethanol is a key event and not the isolated enzyme induction in itself.