1,1'-methylenebis(4-isocyanatobenzene) and its oligomeric reaction products with butane-1,3-diol, 2,2'-oxydiethanol and propane-1,2-diol

Basic toxicokinetics

Administrative data

Endpoint:

basic toxicokinetics, other

Remarks:

In silico modeling

Type of information:

calculation (if not (Q)SAR)

Adequacy of study:

supporting study

Study period:

2021

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

accepted calculation method

Data source

Reference

Reference Type:

study report

Title:

Unnamed

Year:

2021

Materials and methods

Objective of study:

absorption

bioaccessibility (or bioavailability)

Principles of method if other than guideline:

The potential absorption and systemic bioavailability of monomeric MDI isomers and modified MDI substances in human via inhalation was predicted with the mechanistically-based pharmacokinetic software, GastroPlus®.

GLP compliance:

no

Test material

Specific details on test material used for the study:

GastroPlus models the behavior of individual chemical entities, not UVCB-substances . In order to cover the range of substances included in the EU-REACH MDI category, a representative sample of 39 expected/possible MDI constituents was selected for modeling. In addition to the monomeric MDI isomers, variation of the following parameters was considered:
• Type of terminal NCO groups and diol modifiers
o 4,4’- and 2,4’-terminated first glycol adduct for different diols
o A limited selection of the much less prevalent 2,2’-terminated first glycol adducts
• Molecular weight
o DPG/TPG constituents of higher molecular weight, including some representing constituents that may be present <0.1%
o Combinations of diols (see below)
• Kow values
o Specific DEG /TPG constituents to cover a range of log Kow values
• NCO functionality
o Multifunctional oligomers, e.g., homopolymer or isocyanurate and modifiers
• Combinations of diols in the higher glycol adducts
o A selection of “mixed” glycol adducts to represent some of the potential options when using mixtures of diols

Test animals

Species:

other: human

Administration / exposure

Route of administration:

inhalation: aerosol

Details on study design:

All predictions of inhalation systemic absorption were conducted with GastroPlus, v9.8. Initial modeling was conducted with 4,4’-MDI in rat to optimize model parameters and evaluate the accuracy of GastroPlus inhalation absorption predictions vs. prior in vivo data. Presystemic metabolic clearance values for 4,4’-MDI in rat were derived by fitting metabolic clearance terms in five model pulmonary tissues (nose, extra-thoracic, thoracic, bronchiolar, alveolar-interstitial) to 0.0001 L/hr (based on the in vivo rat bioavailability results of Gledhill (2005)) to represent presystemic loss of test materials via reaction with glutathione or other nucleophiles such as proteins. Absorption via the pulmonary route was defined as intact test material that becomes systemically bioavailable through the pulmonary tissues. Presystemic chemical degradation in the GI tract was also included, as per observations of Gledhill (2005) describing substantial polyurea formation from 4,4’-MDI, and optimized to a half-life of 10 min. Absorption via the oral route is defined as intact (undegraded) test material that absorbed from the GI tract. Other parameters included in MDI inhalation modeling of this in vivo study were: test material exposure via 6 hr pulmonary infusion of 0.0609 mg test material as a powder (scaled from 0.078 mg dose and mean body weight of 320 g in prior in vivo study to 0.25 kg default rat body weight used in GastroPlus) and measured dimension for test material particles (1.39 µm diameter). The pattern of test material deposition in the rat pulmonary compartments was calculated with Multiple-path Particle Dosimetry Model (MPPD v.3.04, Applied Research Associates, Arlington VA, USA). Further definition of the distribution between the combined Thoracic-Bronchiolar MPPD compartment was derived from GastroPlus-ICRP66 calculations of relative distribution between these two compartments in humans exposed to particulate material of 1.39 µm diameter.
The prior in vivo study results of Gledhill et al. (2005) described nose-only inhalation administration of 4,4’-MDI in the rat as a combined inhalation/oral exposure. Therefore, uptake via inhalation was assumed to represent 38.5% of the received inhalation exposure, with the remainder as two bolus oral doses at the start and end of the 6-hr exposure. The value of 38.5% of the received dose deposited in the lung was an average the minimum and maximum possible for this intake route based on data from the prior in vivo study: a minimum of 13% of the dose recovered in lung at the end of the 6 hr exposure up to a maximum of 64% representing the portion of the dose not found on the rat skin (36%) at the end of the 6 hr exposure. The two oral dose amounts were then set based on the 36% of the dose found on the rat skin immediately post-exposure (oral dose 2) and the remainder of the dose not attributed to the inhalation route or post-exposure oral, assumed to occur at time zero (25.5%; oral dose 1). Based on conclusions of Gledhill et al. (2005) that the majority of systemically absorbed 4,4’-MDI occurred from the oral route, the fractions of systemically bioavailable parent 4,4’-MDI via the pulmonary tissue and GI tract were modeled as 10% and 90% of total, respectively.
For in silico predictions of absorption of the different MDI constituents in human, the pulmonary metabolic clearance rate in rat was scaled to species differences in total lung surface area (human=average of 201 fold higher than rat; Frohlich 1996) affording a rate of 0.0201 L/hr, while chemical degradation in the GI tract was kept at the rat value of a 10 min half-life. Other settings were pulmonary 6-hr infusion of 1 mg of a test material powder (dose based on 0.01 ppm x 6 hr exposure to monomeric MDI, using a light intensity minute volume of 0.0273 m3/min for adults (US EPA, 2011), test material particle size of 2.5 µm diameter, and deposition in pulmonary tissue based on the ICRP66-Nose model incorporated into GastroPlus. Test materials were assumed to be non-volatile so the diffusion loss from blood to lung air parameter was set to zero in the GastroPlus nasal-pulmonary module. Physical-chemical properties of water solubility and Log Kow used in the GastroPlus modeling were derived within the GastroPlus software. All other parameters were set to model defaults.
To evaluate the impact of test material particle size on absorption in humans, 6 hr exposures to 0.01 ppm (1 mg dose) of the different MDI constituents were modeled, with theoretical particles sizes of 1.0, 2,5 10, 25 and 100 µm. Fractional deposition patterns in the pulmonary compartments were derived with the ICRP66-Nose model (listed in Text Table 1). All other model parameters were as described above.

Text Table 1. Fractional Deposition of Test Material in Nasal-pulmonary Compartments of Human vs. Particle Size (as calculated with the ICRP-66 module of GastroPlus).
Fractional Deposition in Compartment
Particle size (µm diameter) Nose Extra-Thoracic Thoracic Bronchiolar Alveolar-Interstitial Total
1.0 0.0694 0.0697 0.0041 0.0225 0.1472 0.3129
2.5 0.2099 0.2834 0.0158 0.0487 0.2214 0.7792
10 0.3639 0.4025 0.0173 0.0107 0.0077 0.8021
25 0.2728 0.279 0.0031 0 0 0.5549
100 0.2504 0.2508 0.0001 0 0 0.5013

To evaluate the impact of dose amount, metabolic/chemical clearance rates and exposure duration on test material absorption in humans, 3 or 6-hr exposures to 0.01 ppm of the different MDI constituents were modeled, with a theoretical particle size of 2.5 µm. Dose amounts were varied from 0.1x to 10x the default 1 mg value. Metabolic clearance/chemical hydrolysis rates were varied from 0.1x to 10x the default values of 0.0201 L/hr and 10 min half-life, respectively. All other model parameters were as described above.
In silico predictions of absorption from inhalation exposures were also made for the three monomeric MDI isomers, vs. mono-glutathione (S-linked and cyclic) and bis-glutathione adducts of the three isomers. Exposure parameters were as described above, except that the glutathione adducts were administered as a 6-hr pulmonary infusion of a test material solution (vs. as a powder for the monomers). No metabolic clearance in pulmonary tissues or chemical degradation in the GI tract were used for the glutathione adducts as these adducts represent products of initial metabolism in the lung compartment. All other model parameters were as described above.
A complete list of default model parameters used for all 39 compounds are presented in Appendix Table 8. Definitions of the inhalation model parameter terms are listed in Appendix Table 4.

Results and discussion

Bioaccessibility (or Bioavailability)

Bioaccessibility (or Bioavailability) testing results:

GastroPlus modeling of 4,4’-MDI inhalation exposure to rat, as a combined inhalation/oral dose, afforded predicted absorption of 26.1% of the received dose as parent compound, with 24.5% via the GI tract and 1.6% through pulmonary tissues. These in silico predictions were quite comparable to the estimated total in vivo absorption of 28.5% (25.7% via oral and 2.8% via inhalation; as parent compound and/or metabolite). Predicted inhalation exposure of MDI isomers, and non-monomeric constituents of modified MDI substances show the highest pulmonary absorption for the three MDI isomers (38-54%), with lower levels (3-27%) for constituents 4-19 (MW 381-751) and less than 0.1% absorption for the remaining, higher-molecular weight (MW) constituents. Predicted absorption via the oral route, representing mucociliary transported dose, ranged from 5-10% for the MDI isomers, higher levels (10-25%) for Test Substances 4-19, and ≤ 3% for compounds with MW >900. Increased mucociliary transport for Test Substances 4-19 vs. the three MDI isomers is consistent with the lower water solubility and corresponding lower pulmonary absorption for these higher molecular weight analogs. The impact of particle size was shown to have a substantial impact on calculated inhalation and oral uptake of the monomeric MDI isomers and modified MDI substances, with predicted uptake through pulmonary tissues highest for the three MDI isomers for all particle sizes from 1-100 µm. Predicted absorption by this route was less than 0.1% for all compounds with MW > 900, for all particle sizes. For any given MDI constituent, pulmonary absorption is predicted to be highest at 2.5 µm, with decreasing values at 1, 10, 25, and 100, respectively. Patterns of oral absorption for the 39 MDI constituents are complementary to inhalation uptake, with a consistent drop in appreciable absorption above MW 900. Decreasing or increasing metabolic clearance and chemical hydrolysis rates had a modest impact on predicted inhalation absorption (1-7 fold) but more substantial impact on predicted oral absorption (2-50 fold). Altering theoretical dose levels also afforded up to 10 fold changes in predicted inhalation, vs. less than 2 fold change in oral absorption for any test substance. A shorter exposure time (3 hr vs. 6 hr default) resulted in less than 10% changes in predicted fractional uptake for MDI constituents 1-19 by either route. In all cases, in silico-predicted pulmonary absorption of the three MDI isomers was higher than the remaining 36 possible/expected MDI constituents. The relative absorption of the three MDI isomers and the corresponding mono- and bis-glutathione conjugates through the pulmonary tissues was predicted to be quite comparable, at 38-54% for the MDI isomers vs. 44% for all 9 of the glutathione conjugate analogs.

Applicant's summary and conclusion

Conclusions:

Pulmonary absorption is predicted to be highest for the three MDI isomers, with uptake by this route ~ 2 fold lower for analogs of MW 381-751. No pulmonary absorption is predicted for constituents > 900 MW. Mucociliary transport is substantial, based on predicted oral absorption of up to 25% for constituents with MW 300-751. Oral absorption of MDI isomers is predicted to be slightly lower than MW 300-751 analogs, probably due to higher pulmonary uptake. Higher molecular weight analogs > 751 may also undergo substantial mucociliary clearance, as per Kuehl et al. (Kuehl et al., 2016), but were not predicted to be appreciably absorbed via the GI tract.
Numerous studies in the literature support the concept of decreased dermal or inhalation absorption for compounds with high molecular weight and/or lipophilicity. Several groups have summarized dermal penetration studies for a variety of hydrocarbons and have shown decreased uptake across the Log Kow range of 3-8 (Figure 12) (Jakasa et al., 2015; Kezic and Kruse, 2010). Other investigators have evaluated the distribution of highly lipophilic compounds within the skin and have shown that this class of chemicals are sequestered in the stratum corneum, with little penetration into the epidermis or subsequent layers (Chuberre et al., 2019; Petry et al., 2017). Bos et al. have proposed a cutoff for substantial dermal uptake of chemicals at 500 molecular weight (Bos and Meinardi, 2000). Based on these datasets and others, several regulatory groups have listed physical-chemical properties that limit dermal absorption as being 1) molecular weight > 500, 2) Log Kow ≥ 4, or 3) water solubility below 0.001 mg/mL (ECHA, 2017; SCCS, 2015). The pulmonary absorption of lipophilic compounds such as corticosteroids is also compromised by substantial (40-90% of dose) mucociliary clearance of an inhaled or intranasal dose (Winkler et al., 2004). This decreased absorption may be due to lower water solubility and/or sequestration in the pulmonary mucus (Olsson et al., 2011; Sigurdsson et al., 2013).
Finally, the pulmonary absorption of the glutathione conjugates of MDI isomers is predicted to be comparable to, or greater than the parent diisocyanates. These results are consistent with numerous other studies showing pulmonary uptake of peptides and proteins to be rapid and substantial (Folkesson et al., 1998; Liang et al., 2020).
The results of these in silico evaluations should be useful in category-based, Read-Across Assessments for monomeric MDI isomers and modified MDI substances.

Executive summary:

Finally, the pulmonary absorption of the glutathione conjugates of MDI isomers is predicted to be comparable to, or greater than the parent diisocyanates. These results are consistent with numerous other studies showing pulmonary uptake of peptides and proteins to be rapid and substantial (Folkesson et al., 1998; Liang et al., 2020).

The results of these in silico evaluations should be useful in category-based, Read-Across Assessments for monomeric MDI isomers and modified MDI substances.