Registration Dossier

Diss Factsheets

Administrative data

Endpoint:
extended one-generation reproductive toxicity - basic test design (Cohorts 1A, and 1B without extension)
Type of information:
experimental study planned (based on read-across)
Justification for type of information:
Hypothesis: Hypothesized MoA predicts local effects in the lungs and no significant systemic exposure to unreacted NCO since it reacts with biological nucleophiles before being absorbed as GHS/protein adducts. With no systemic exposure to unreacted isocyanate or toxic metabolite, no effects on fertility are predicted.

Justification: The substance is part of a category read-across based on the hypothesized MoA that predicts local effects in the lungs and no significant systemic exposure to unreacted NCO since it reacts with biological nucleophiles before being absorbed as GHS/protein adducts. With no systemic exposure to unreacted isocyanate or toxic metabolite, no effects on fertility are predicted. This is justified by weight of evidence is based on three independent sources of information including a) Absence of effects on reproduction and fertility in a guideline two-generation reproductive toxicity study on an analogue aromatic diisocyanate TDI, b) Evidence for a lack of systemic availability of toxicologically active parent or metabolite, and c) A lack of systemic toxicity and effects on reproductive organs in repeated dose toxicity of representative MDI substances and TDI. As all substances of the MDI category as well as the analogous substance TDI contain significant quantities of bioaccessible NCO groups required for the hypothesized MoA.
However, as no fertility data from the required test guideline for Annex X is available for any category test substance, the current study is proposed on the boundary substances (4,4’-MDI and 4,4’-MDI/DPG/HMWP) to support the common mechanism. Additional screening studies (OECD 422) will be performed on selected category members from each sub-group to act as bridging studies. These additional studies will confirm that hypothesis that additional structural features to not contribute to the systemic and reproductive toxicity and that 4,4’-MDI would still be considered the worst-case substance. If these additional studies suggest additional toxic potential, additional OECD 443 studies will be proposed.

The final report on the definitive study is expected to be available 48 months following receipt of the final decision. However, due to limited global capacity for this study type, result could be delayed.

Data source

Materials and methods

Test guideline
Qualifier:
according to guideline
Guideline:
OECD Guideline 443 (Extended One-Generation Reproductive Toxicity Study)
Justification for study design:
SPECIFICATION OF STUDY DESIGN FOR EXTENDED ONE-GENERATION REPRODUCTION TOXICITY STUDY WITH JUSTIFICATIONS:

- Premating exposure duration for parental (P0) animals:
The toxicity of MDI substances is driven by the bio-accessible NCO value which is responsible for chemical and physiological reactivity and subsequent toxicological profile. The hypothesized MoA of toxicity is the reaction of the electrophilic NCO group with extracellular biological nucleophiles such as cellular proteins and is limited to point of contact effects without subsequent systemic toxicity. Reliable inhalation data is available for 4,4’-MDI, pMDI and 4,4’-MDI/DPG/HMWP). Results from these studies are consistent with the hypothesized MoA and assertion that MDI substance toxicity after repeated inhalation exposure in animals is limited to local effects caused by inflammation and irritation to the respiratory tract without subsequent systemic toxicity. In regard to these local effects after inhalation, the lungs are the primary organ affected since olfactory lesions were only observed at higher concentrations/doses. As an MDI substance enters the lung, NCO groups react with biological nucleophiles at the MDI/lung fluid interface to form MDI-conjugates. Formation of these MDI-adducts depletes protective nucleophiles in the lung and results in pulmonary irritation and inflammatory cell influx. MDI only enters the systemic circulation in the form of MDI-GSH or protein adducts. Consequently, there is no systemic exposure to the toxic NCO functional group which is consistent with the lack of distal toxicity in any study conducted. Thus, a premating exposure period of 2 weeks is considered sufficient.
- Basis for dose level selection: An OECD 422 study will be performed prior to main study to serve as a dose-range finder and basis for comparing bridging studies for multiple category members. The preliminary OECD 422 study will include an extension of dosing in the F1 generation to ensure toxic effects on juveniles is characterized. In addition, study data from existing sub-chronic and chronic study data will be used for dose selection (Hoymann et al. 1994 for 4,4’-MDI, Reuzel et al. 1994 for pMDI). These supporting studies demonstrate that even with a chronic/lifetime exposure duration with group sizes equivalent to or greater than guideline reproductive studies, effects from pMDI and 4,4’-MDI aerosols are confined to the lungs. Effects on systemic organs including reproductive organs were not observed at exposure concentrations associated with marked respiratory tract toxicity.
- Inclusion/exclusion of extension of Cohort 1B: not planned, based on reliable repeated dose toxicity study data there is no indication of distal toxicity induced by MDI, see above. However, if preliminary OECD 422 studies indicate concern, additional cohorts will be added to the definitive OECD 443 study.
- Termination time for F2: not applicable, the available data indicate no specific concern regard to neuronic see above
- Inclusion/exclusion of developmental neurotoxicity Cohorts 2A and 2B: not applicable, the available data indicate no specific concern regard to neurotoxicity in adult animals or to developmental neurotoxicity, see above. However, if preliminary OECD 422 studies indicate concern, additional cohorts will be added to the definitive OECD 443 study.

- Inclusion/exclusion of developmental immunotoxicity Cohort 3: not applicable, the available data indicate no specific concern with regard to immunotoxicity in adult animals or developmental immunotoxicity. However, if preliminary OECD 422 studies indicate concern, additional cohorts will be added to the definitive OECD 443 study.
- Route of administration: inhalation route
A need for an oral hazard assessment for MDI is not indicated by its use or by toxicokinetic profile of exposure routes. Furthermore, oral toxicity data for MDI cannot be extrapolated for risk assessment of inhaled aerosols as the relevant route of exposure for human risk assessment. Therefore, there is no apparent benefit from any oral toxicity data for MDI. Toxicokinetic data for the inhalation route of exposure is sufficient, and the performance of an additional oral animal toxicity study would not create data that would influence the risk management measures and therefore would be in conflict with the principles of animal use and welfare. These differences are described below:

Route of exposure specific differences in MDI metabolism

4,4’-MDI (MDI) contains two highly reactive NCO-groups, which are responsible for the distinct portal of entry toxicity described by the available toxicological data. The NCO-group reacts readily with nucleophilic biomolecules and depending on the chemical and physico-chemical composition of the interphase at the site of primary contact, distinct differences in primary reaction products can be described. Therefore, significant differences in bioavailability and metabolic fate can be described for the oral, dermal and inhalation route of exposure (see end point summary on toxicokinetics\):
• in the pH neutral medium of the lung inhaled respirable MDI aerosols react with the proteins and peptides (mainly glutathione) of the bronchioalveolar fluid, partly representing bioavailable adducts,
• direct intubation of large MDI doses into the stomach is an artificial exposure route and can only simulate accidental swallowing. In the acidic pH of the stomach MDI polymerizes with the stomach content and forms solid and inert polyureas. Information from analogous diisocyanates and US reports on accidental ingestion of MDI based glues in domestic animals describe formation of high molecular primary reaction products with CO2 liberation, without apparent systemic chemical toxicity. Polymeric reaction products are of low bioavailability.
• Following oral swallowing of traces of MDI, reactions will commence at once with biological macromolecules in the buccal region and will continue along the oesophagus prior to reaching the stomach. Reaction products will be a variety of polyureas and macromolecular conjugates with for example mucus, proteins and cell components.
• at the interface of the skin, reactions with nucleophilic groups of the skin matrix and polymerization to a solid polyurea crust occurs, significantly reducing bioavailability. (Based on this knowledge it should only be speculated on the effects of an agglutination of the reactive MDI with blood proteins following i.v. application.)

In conclusion the toxicokinetic behaviour of MDI needs to be considered with respect to the specific physico-chemical and chemical exposure conditions at the site of first contact. For MDI, significant differences in primary reaction products and by this in the subsequent bioavailability and metabolic fate imply a high level of uncertainty for route to route extrapolation of toxicological data. In accordance to REACh Annex I (0.3.), the chemical safety assessment of a substance shall be based on a “…known or reasonably foreseeable exposure…”, accidental exposures are not considered. Therefore, and in accordance to ECHA Guidance Chapter R.7a and the REACh Annexes on information requirements the potential hazard of MDI should be determined on the most relevant route of exposure for risk assessment which is inhalation.

Relevance of oral route of exposure for risk and hazard assessment:
ECHA Guidance Chapter R.14 (Occupational exposure estimation, R.14.2 Types and routes of exposure) indicates that, exposure through ingestion is “…generally not considered further in the assessment of workplace exposure“. For proof of concept, working processes for professional uses of MDI include handling and spray application of foams in concentrations up to max 30 % pbw. Other uses include handling and application of coatings, adhesives and paints in which MDI is not contained as such but as pre-reacted high viscous polymer with low monomer content. For spray applications in which aerosols are generated respiratory protection by full mask, and for all application’s effective dermal protection by full body protection and gloves is prescribed. These risk reduction measures effectively prevent from any oral exposure, e.g. via contaminated skin or clothing or inhalation cross-contamination. Cured PU applications, e.g. spray foam, do not contain residual unreacted MDI (see Topic 6 below). In conclusion, no significant uncertainties regarding a potential oral exposure can be anticipated in the holistic exposure assessment for professionals.

ECHA Guidance Chapter R.15 (Consumer exposure estimation, R.15.2.2 Reasonable worst-case situations) indicates that, “…the consumer exposure estimation should normally address the intended uses of the products that contain the substances under investigation. However, since consumers may not accurately follow instructions for use of products, an estimation of other reasonably foreseeable uses should be made. Consideration of deliberate abuse is not part of the exposure estimation process.” Consumer uses include one component rigid foam available in cans as well as coatings, adhesives & sealants, and paintings. These uses contain pre-reacted and polymerized MDI derivatives (prepolymers and higher oligomers originating form polymeric MDI (PMDI)) with low amounts of residual monomer. Reaction is readily with humidity in the air, resulting in entirely cured product, free of residual MDI.

Consumer uses of MDI are covered by national regulations, which e.g. restrictively prescribe selling in correspondingly equipped DIY stores, if the consumer agrees to information provided by the staff addressing the risk. In addition, the appropriate storage and handling is explicitly prescribed, e.g. by the recommendation to use gloves which are delivered with every can as part of an existing restriction under REACH Annex XVII as follow up of the 2005 Risk Assessment. Therefore, there is a very low possibility for oral exposure by the intended use, since e.g. skin contact should be prevented by the use of gloves, and by this dermal to oral cross-contamination is minimized. Since cured PU applications, e.g. spray foam, do not contain residual unreacted MDI (see Appendix 1 on Toxicological information, DNEL justification provided as a separate document) migration from articles through sucking, chewing or licking can likewise be excluded. In addition, no foreseeable misuse other than abuse has to reasonably be anticipated due to specific regulations of the consumer application.

- Other considerations, e.g. on choice of species, strain, vehicle and number of animals [if applicable]: Wistar rats will be used as the species as they have been extensively used in MDI substances and provide an appropriate basis for comparison between studies.

Test material

Constituent 1
Reference substance name:
4,4'-Methylenediphenyl diisocyanate, oligomeric reaction products with butane-1,3-diol, 2,4'-diisocyanatodiphenylmethane, [(methylethylene)bis(oxy)]dipropanol and propane-1,2-diol
EC Number:
500-312-1
EC Name:
4,4'-Methylenediphenyl diisocyanate, oligomeric reaction products with butane-1,3-diol, 2,4'-diisocyanatodiphenylmethane, [(methylethylene)bis(oxy)]dipropanol and propane-1,2-diol
Cas Number:
123714-19-2
Test material form:
liquid: viscous
Details on test material:
Batch # 290832

Results and discussion

Applicant's summary and conclusion

Conclusions:
No reproductive toxicity studies exist on the target substance 44MDI/1,3-BD/TPG/PG. This endpoint will be satisfied by weight of evidence and read across from valid OECD 443 Extended One-Generation Reproduction Studies for the inhalation route on two category substances that are considered the boundary substances (4,4,’-MDI and 4,4,’-MDI/DPG/HMWP). Additional reproductive and developmental screens (OECD 422) will be performed on 9 substances representing all sub-groups and key structural/chemical characteristics (see Test Proposal Summary attached in Chapter 13). These studies will confirm the proposed MoA for effects on reproductive endpoints or identify substances that may require additional testing.
The basis of read-across for all MDI substance categories is via a common toxicological mode of action of point of contact reactivity of the NCO group on the most bioaccessible constituents of the substance. The hypothesized MoA predicts local effects in the lungs exclusively and no significant systemic exposure to unreacted NCO since it reacts with biological nucleophiles before being absorbed as GHS/protein adducts. In the case of MDI substances, no systemic effects have ever been noted in acute, subacute and chronic bioassays. This is consistent with the toxicokinetic evidence for a lack of systemic bioavailability of any toxicologically parent molecule (i.e. active NCO) or metabolite.
All MDI substances (including the target substance and proposed boundary source substances) share the common chemical characteristic of two NCO functional groups per molecule with each NCO group bound to an aromatic ring and this ring connected to a second aromatic ring by the methylene group . The proposed boundary source substance 4,4’-MDI has a NCO content of up to 33.6 % (with 99% monomeric MDI) while the other proposed boundary substance 4,4,’-MDI/DPG/HMWP has 20-23% NCO (with 45-50% monomeric MDI). Since it is the monomeric NCO group which drives the reactivity and thus toxicity of MDI substances, 4,4’-MDI represents the worst case with the highest levels of bioaccessible NCO groups. Conversely, 4,4,’-MDI/DPG/HMWP represents the least toxic category substance due to the least amount of bioaccessible NCO. All other category members (including the target substance) lie between these 2 boundary substances with respect to the amount of bioaccessible NCO. This evidence supports that the potential hazard for MDI Category substances to elicit toxic effects lies between the 2 boundary substances and can be interpolated between them.
Further supporting evidence of lack of systemic and reproductive toxicity is provided by a negative OECD 416 (Two-generation inhalation study) on an analogous diisocyanate TDI (Tyl, et al., 1999). While the presence of the two NCO groups on the same aromatic ring in TDI may affect reactivity, exposure, and deposition, once it becomes bioaccessible in the lung lavage fluid, it behaves similarly (i.e. adduct formation with biological nucleophiles) before absorption. Therefore, like MDI, TDI has the same lack of systemic exposure and lack of toxicity in any systemic organs including effects on reproduction and fertility. Thus, with respect to systemic toxicity, TDI is appropriate for to support the hypothesized MoA for fertility. It should also be noted that TDI, with its higher volatility (vapour at room temperature; (Pauluhn, 2015)) and water solubility, has a higher respiratory exposure potential than MDI substances. Hence, the TDI substance represents worst-case with which to draw analogy, thereby providing sufficient confidence in the assessment.
With respect to potential toxic effects not linked to the hypothesized MoA, target constituents (e.g. non-monomeric) but it is recognized that the oligomeric portion of the target substance constituents while containing no functional groups capable of biological reactivity, contain the greatest variability in terms of chemical structure. As described above in previous endpoints and in the Category Justification Document, the non-monomeric MDI constituents lowers the solubility and reduces its availability to react with biological molecules. In vitro reactivity experiment by Zhang et al. (2021) demonstrated that the mMDI constituents in modified MDI substances followed a similar pattern of glutathione adduct formation seen previously (Wisnewski et al., 2019a). However, when pMDI was incubated with GSH, adducts with the higher molecular weight constituents (>3 ring oligomer) were not detected in subsequent GC/MS. Additional solubility experiments by Zhang et al. (2021) confirmed that these constituents are not soluble and thus unable to react. In vivo inhalation experiments by Pauluhn (2002b) with pMDI demonstrates that inhalation exposure with pMDI (50 % of mMDI, 34 % of three-ring MDI, ratio about 3/2) in rats subsequent adducts detected (measured as two- or three-core MDA after hydrolysis) were present in a ratio of approximately 10/1. In other words, the presence of three-core adducts was seven times lower than expected from the composition of the test substance. Thus, even for the most soluble constituent next to mMDI of the MDI category substances, the 3-ring oligomer showed a significant reduction in bioaccessibility. Further increasing the molecular weight (either by condensation, or glycol adduct formation) even further reduces this potential bioaccessibility of the molecule. Taken together, the increasing octanol-water partition coefficients of the non-monomeric MDI constituents quickly attenuates the availability of these constituents to react with nucleophiles and elicit inflammation and irritation. Therefore, these constituents do not need to be considered for the prediction of systemic exposure and toxicity and predictions for all endpoints can be based on bioaccessible NCO as described in the hypothesis.
If the OECD 422 bridging studies described above identify additional hazards, additional higher tier studies will be considered.
It should also be noted that a need for an oral hazard assessment for MDI is not indicated by its use or by toxicokinetic profile of exposure routes. Furthermore, oral toxicity data for MDI cannot be extrapolated for risk assessment of inhaled aerosols as the relevant route of exposure for human risk assessment. Therefore, there is no apparent benefit from any oral toxicity data for MDI Category substances. Toxicokinetic data for the inhalation route of exposure is sufficient, and the performance of an additional oral animal toxicity study would not create data that would influence the risk management measures and therefore would be in conflict with the principles of animal use and welfare.
In summary, a weight of evidence with proposed OECD 443 tests via inhalation on source boundary substances 4,4’-MDI and 4,4’-MDI/DPG/HMWP, will be used to close the data gap for the target substance. Supporting toxicokinetic data and a 2-generation study on an analogous diisocyanate for this endpoint are used to demonstrate the hypothesized MoA and read-across. Based on this weight of evidence all MDI substances are not classified as a reproductive toxicant according to Regulation (EC) No 1272/2008 (CLP). For MDI, there is evidence that the strong respiratory irritation restricts exposure and ensures that there will be no exposure at levels at which there is a realistic possibility of reproductive toxicity. Additional proposed testing is proposed to confirm this hypothesis and support read-across to all category members.
Executive summary:

The hypothesized MoA for the substances of the MDI category for reproductive (and other systemic toxicity endpoints) is based on the common mode of action for local and systemic toxicity and subsequent effects on fertility.  Reliable evidence from an two-generation reproduction study on an analogue diisocyanate (TDI) shows no effects on any reproductive parameter and this is consistent with findings from acute and repeated exposure studies on 4,4’-MDI and pMDI that show that MDI substances will not produce toxicity in any systemic organs, including those of the reproductive system, outside the site of contact (lungs).  This data will be used to support the basis of using Extended One-Generation Reproduction Studies (OECD 443) on the MDI substance boundary substances 4,4,’-MDI and 4,4,’-MDI/DPG/HMWP as the source for a category read-across to the target substance.   Additional testing will be considered based on OECD 422 bridging studies on multiple other key category substances.  


Since there is high confidence in the hypothesized MoA for MDI substances through the reactive NCO group and as systemic exposure is not anticipated for any substance of the MDI category to be supported by testing on the category boundary substances, then the available data is considered on a weight of evidence basis as being sufficient for risk and hazard assessment purposes.


Additional details on the hypothesized MoA are provided in the MDI Substance Category Justification Document.