Methyl methacrylate

Endpoint summary

Description of key information

Skin sensitisation

Human data: contact sensitiser in humans with indication of a weak potency (based on skin patch data with low prevalence in relevant cohort studies)   

Animal data - LLNA: inconsistent results from negative to weakly positive in guideline LLNAs according to OECD 429, 442A or 442B; e.g. weakly skin sensitising (OECD 429; Betts et al. 2006)

Animal data - Guinea pig: inconsistent results from various test methods with indication of a weak potency 

In chemico/ in vitro according OECD Adverse Outcome Pathway (2014)

Key events 1-3: consistently moderate positive in DPRA/ OECD 442C, KeratinoSens/ OECD 442D and h-CLAT/ OECD 442E

 

Respiratory sensitisation

The outcome of a detailed weight-of-evidence assessment of all the available information in July 2022 indicates that there is insufficient evidence to classify MMA as a respiratory sensitiser. Evaluated lines of evidence were human data (differentiated towards worker health studies, case studies, exposure, health surveillance databases and SIC cases) and, with lower relevance for the assessment, metabolism, in vitro and in silico data.

Key value for chemical safety assessment

Skin sensitisation

Link to relevant study records

Reference

Endpoint:

skin sensitisation: in vivo (LLNA)

Type of information:

experimental study

Adequacy of study:

key study

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

comparable to guideline study

Qualifier:

equivalent or similar to guideline

Guideline:

OECD Guideline 429 (Skin Sensitisation: Local Lymph Node Assay)

Version / remarks:

As described in: Kimber I, Basketter D A. The murine local lymph node assay: a commentary on collaborative studies and new directions. Food Chem Toxicol 1992: 30: 165–169.

GLP compliance:

not specified

Type of study:

mouse local lymph node assay (LLNA)

Specific details on test material used for the study:

MMA (99.96% pure) was supplied by Lucite International UK Ltd via Cassel Works, Cleveland, UK.

Species:

mouse

Strain:

CBA/Ca

Sex:

not specified

Details on test animals and environmental conditions:

- Young adult CBA/Ca mice (Harlan, Bicester, Oxfordshire, UK)
- Age: 8–12 weeks
- Housing: metal cages
- Food: SDS PCD pelleted diet; Special Diets Services Ltd, Witham, UK)
- Water: ad libitum.
-Temperature: 22 +/- 3°C
- Relative humidity: 50 +/-20%
- Lightning: 12 hr light/dark cycle.

Vehicle:

other: acetone or acetone/olive oil (4:1 v/v)

Concentration:

0, 10, 30, 50 75, and 100% (neat)

No. of animals per dose:

4 per group

Details on study design:

Groups of mice were exposed topically on the dorsum of both ears to 25 µl of various concentrations of chemical, or to the same volume of vehicle alone, daily for 3 consecutive days. Five days after the initiation of exposure, all mice were injected intravenously via the tail vein with 20 µCi of [3H] methyl thymidine in 250 µl of phosphate buffered saline. Five hrs later, mice were killed and the draining auricular lymph nodes were excised and pooled for each experimental group. A single-cell suspension of lymph node cells (LNC) was prepared; approximately 12 hr later, incorporation of [3H]TdR was measured and a stimulation index relative to the concurrent vehicle-treated control value was derived.

Positive control substance(s):

other: 2,4-Dinitrochlorobenzene

Statistics:

The estimated concentration of chemical required inducing an SI of 3 relative to concurrent vehicle treated controls, or EC3 value, was derived by linear interpolation of dose–response data.
The EC3 value was calculated by interpolating between 2 points on the SI axis, one immediately above, and the other immediately below, the SI value of 3.

Positive control results:

Positive control: 2,4-Dinitrochlorobenzene (DNCB), Vehicle: Acetone/Olive Oil):
Conc (% w/v) SI
0 1
0.01 1.4
0.025 2.2
0.05 4.0
0.1 9.8
0.25 16.2

Parameter:

SI

Value:

1.4

Test group / Remarks:

10 % (w/v)

Remarks on result:

other: Vehicle: acetone

Parameter:

SI

Value:

2.3

Test group / Remarks:

30 % (w/v)

Remarks on result:

other: Vehicle: acetone

Parameter:

SI

Value:

2

Test group / Remarks:

50 % (w/v)

Remarks on result:

other: Vehicle: acetone

Key result

Parameter:

SI

Value:

4.4

Test group / Remarks:

75 % (w/v)

Remarks on result:

other: Vehicle: acetone

Key result

Parameter:

SI

Value:

7.3

Test group / Remarks:

100 % (w/v)

Remarks on result:

other: Vehicle: acetone

Parameter:

SI

Value:

1.5

Test group / Remarks:

10 % (w/v)

Remarks on result:

other: Vehicle: acetone/olive oil

Parameter:

SI

Value:

1.5

Test group / Remarks:

30 % (w/v)

Remarks on result:

other: Vehicle: acetone/olive oil

Parameter:

SI

Value:

1.5

Test group / Remarks:

50 % (w/v)

Remarks on result:

other: Vehicle: acetone/olive oil

Key result

Parameter:

SI

Value:

4.4

Test group / Remarks:

75 % (w/v)

Remarks on result:

other: Vehicle: acetone/olive oil

Key result

Parameter:

SI

Value:

3.6

Test group / Remarks:

100 % (w/v)

Remarks on result:

other: Vehicle: acetone/olive oil

Key result

Parameter:

EC3

Value:

ca. 60

Test group / Remarks:

Vehicle: acetone

Key result

Parameter:

EC3

Value:

ca. 90

Test group / Remarks:

Vehicle: acetone/olive oil

Cellular proliferation data / Observations:

see table below

Desintegrations per minute (dpm)
Conc. (% w/v)	MMA, vehicle acetone	MMA, vehicle acetone/olive oil	DNCB/ pos. Ctrl., vehicle acetone/olive oil
0	137	243	609
0.01	ND	ND	839
0.025	ND	ND	1352
0.05	ND	ND	2455
0.1	ND	ND	5971
0.25	ND	ND	9843
10	205	341	ND
30	312	363	ND
50	274	369	ND
75	605	513	ND
100 (neat)	999	874	ND
Stimulation Index (SI)
Conc. (% w/v)	MMA, vehicle acetone	MMA, vehicle acetone/olive oil	DNCB/ pos. Ctrl., vehicle acetone/olive oil
0	1.0	1.0	1.0
0.01	ND	ND	1.4
0.025	ND	ND	2.2
0.05	ND	ND	4.0
0.1	ND	ND	9.8
0.25	ND	ND	16.2
10	1.5	1.4	ND
30	2.3	1.5	ND
50	2.0	1.5	ND
75	4.4	2.1	ND
100 (neat)	7.3	3.6	ND

Interpretation of results:

Category 1B (indication of skin sensitising potential) based on GHS criteria

Conclusions:

In this local lymphnode assay methyl methacrylate attained the criteria of Skin sens Category 1 B

Executive summary:

.In a mouse local lymphnode assay (Betts 2006) acc. to OECD 429 methyl methacrylate was tested at concentrations of 10; 30; 50; 75 and 100 % in two different vehicles (acetone and acetone/olive oil). EC50 values were 50 % (acetone) and 90 % (acetone/olive oil). In this study methyl methacrylate is a weak skin sensitizer and attained in this study the criteria of skin sens category 1 B.

Endpoint conclusion

Endpoint conclusion:

adverse effect observed (sensitising)

Additional information:

General introduction to sensitisation 

The capacity of MMA to induce sensitization is directly related to its metabolism and reactive chemistry. The general metabolic pathway for MMA consists of ester hydrolysis to methacrylic acid and methanol, a process catalyzed by carboxylesterase enzymes, which are further metabolized to carbon dioxide and water. Hydrolysis and further metabolism of MMA occur rapidly following exposure as discussed in the chapter “Toxicokinetics”, based on animal in vivo and human in vitro data.   

Another potentially important metabolic pathway for MMA and related esters involves reaction with tissue nucleophiles via Michael addition on Cβ of the α,β-unsaturated carboxyl group. Increased urinary excretion of thioethers and depletion of hepatocyte GSH have been documented after exposure to MMA as indication of the conjugation with glutathione (GSH) prototype for such reactions. However, the electrophilic activity of methacrylate esters, as reflected by GSH conjugation, is less than that of corresponding acrylate esters, as shown in Borak, et al. 2011 (therein Table 1, see attachment). The electrophilic reactivity of low-molecular-weight molecules is an important aspect of their ability to act as sensitizers. In skin sensitization studies, a key early step in the process leading to sensitization is the formation of covalent adducts with a carrier protein, thereby forming an antigenic hapten-protein complex. Accordingly, electrophilic reactivity also serves as a predictor of sensitization potential. Michael acceptor electrophiles, such as MMA and related esters, are generally predicted to be strong sensitizers, but there is a broad spectrum of reactivity across esters (see again Borak et al. 2011).   

The hydrophobicity (i.e., Log P) of individual molecules is a second factor that impacts their ability to react covalently with the nucleophile groups of carrier proteins. It has been estimated that sensitization potential is about twice as dependent on electrophilic reactivity as hydrophobicity (Roberts et al., 2008). In general, the Log P of methacrylate and acrylate esters increases as the length of their alkyl side chains increase. Because of its relatively low electrophilic reactivity and relatively low Log P as contrasted with other methacrylate and acrylate esters, one would expect that MMA has comparatively low sensitizing potential.   

Hydrolysis of MMA reduces its sensitization potential because, under physiological conditions, methacrylic acid is not electrophilic or protein reactive as shown in vivo (as discussed by Borak et al. 2011). The same is true for methanol as other primary metabolite. These findings suggest that hydrolysis is the principal pathway of MMA metabolism, that the hydrolysis is a detoxification process for the sensitisation potential of MMA, and that electrophilic reaction via Michael addition plays only a minor role and seems to occur only at high tissue concentrations.   

 

Skin sensitisation 

The skin sensitization potential was assessed following the considerations of the OECD guidance “The Adverse Outcome Pathway for Skin Sensitisation Initiated by Covalent Binding to Proteins” (2012*). The adverse outcome itself can be assessed with the help of human data or animal data with guinea pigs, last in well-established but disputable test systems from an animal welfare perspective. Thus, test systems were developed following the 3R principle to investigate the key events for skin sensitization and have been already validated under OECD or are in progress. There is common understanding especially for the in chemico and in vitro test methods that these are individually not adequate for use as stand-alone methods for the hazard assessment, despite various achievements, so that various defined weight-of-evidence approaches are currently in discussion. 

For all key aspects of skin sensitisation, reliable data are available. 

 

Human data 

Numerous case reports of skin sensitisation exist from certain occupational environments, where frequent and prolonged unprotected skin contact with monomer containing mixtures was common practice. Repeated exposure to undiluted MMA may lead to skin sensitisation in susceptible persons. The available data revealed extensive variation between the reports cited with respect to the number of subjects in whom a positive response was observed as a function of those tested. In fact, the prevalence was found to vary between studies from 0% to 16% (see table below). According to Kimber & Pemberton (2014), “the higher prevalence rates were reported in those studies where relatively small numbers of patients with dermatitis had been referred to clinic. The corollary was that much lower prevalence rates were observed among larger cohorts, but even in those cases it could be argued that the figures obtained are an over-estimation of the frequency with which skin sensitisation develops among populations exposed to MMA. The data of Pflug (Pflug, 1995, 2000) were considered to perhaps provide the most accurate reflection of the prevalence of sensitisation to MMA in the dental industry (values of between 0.25% and 0.378%), although even in these reports there is possibly some over-estimation of true rates of sensitisation to MMA (Betts et al., 2006; Pflug, 1995, 2000). The conclusion drawn at that time was that, although there was no doubt that MMA is a contact allergen, the available evidence from predictive test methods indicated relatively weak skin sensitising potential. The data available from reports of skin sensitisation to MMA among exposed human population was considered to be consistent with that conclusion (Betts et al., 2006). It is, of course, important to consider whether there are any other data that have become available regarding human skin sensitisation to MMA since publication of the Betts et al. article (Betts et al., 2006). Little new information has emerged that is relevant to the prevalence of sensitisation to MMA, or to its sensitising potency. ... Based on all available data it is concluded that MMA is a contact allergen, but displays weak skin sensitising potential. The relatively low frequency of sensitisation among those exposed to MMA is consistent with that conclusion.“ (cited references only partially contained in this dossier)

In the absence of any other data that informs understanding of human skin sensitisation to MMA there is no reason to revise the conclusions published in Kimber & Pemberton (2014).

 

Animal data in mice (LLNA)

The LLNA is generally considered as appropriate test system for a more precise view on the potential for skin sensitisation due to the studied dose relationships, so that in a weigh-of-evidence approach this line of evidence is considered with a rather high impact. Six reliable LLNAs are available following the respective OECD guidelines OECD 429, 442A or 442B. The results are relative consistently showing borderline effects close to the criteria threshold of the respective test method, with four being negative and two being weakly positive. This assessment is in line with OECD (2018; Report of the validation study of the Local Lymph Node Assay BrdU-FCM (LLNA: BrdU-FCM) test method. Series on Testing and Assessment No. 283, ENV/JM/MONO(2018)1) declaring MMA as weak or borderline skin sensitizer. Thus, MMA is considered as having a weak potential in these fully reliable test systems. 

 

Animal data in guinea pigs 

There is a great number of reliable studies available to assess the skin sensitising potential of MMA. The variety of used test methods is large, providing positive and negative results in almost equal proportions. Nevertheless, in the most common test systems at least one positive test result can be found. 

 

In chemico/ in vitro data

The OECD Adverse Outcome Pathway (AOP, 2012*) was developed to assess the potential to cause skin sensitization with the help of non-animal test methods that address specific key events in the acquisition of contact allergy/ skin sensitization. While these test methods and AOP were developed to avoid further animal testing during regulatory programs like REACH, for MMA extensive human and animal data were already available before the validation of the in vitro methods. Thus the following in chemico/ in vitro data are of subordinated impact for the hazard assessment. For the substance group of lower methacrylate esters including MMA, Kimber & Pemberton (2021) found out that results from DPRA and Keratinosens assays match well with animal data while h-Clat data are less reliable; thus h-Clat results should be even less considered in the hazard assessment.

 

Protein binding/ molecular initiating event (KE1) 

MMA was consistently moderately positive in a few reliable DPRA studies according to OECD guideline 442C, with cysteine depletion being stronger than lysine depletion. Thus, MMA is considered as having a moderate potential in these test systems. 

Keratinocyte inflammation/ cellular response (KE2) 

MMA was weakly positive in a KeratinoSens assay according to OECD 442D, while other test methods showed variable results with MMA in pre-GL studies (LuSens positive; HaCaT positive and negative). Together with heterogeneous pre-GL data with other methods, MMA is considered as predominantly weak positive in these test systems. 

Dendritic cell activation/ cellular response (KE3) 

MMA was tested consistently positive in a h-CLAT assay according to OECD 442E and also in pre-GL studies with various test methods (MUSST, USENS or h-CLAT). Thus, MMA is considered as positive in these test systems.

A discrimination by the potency of substances is currently not possible for these methods. 

   

in silico data 

Two commercial QSAR systems were used to evaluate the sensitising potential of MMA and provided inconsistent results (Kimber 2019*): DEREKTM predicted MMA to have a ‘plausible’ skin sensitisation alert whereas ‘plausible’ was defined as ‘the weight of evidence supports the proposition’ (of skin sensitising activity). The skin sensitising potential of MMA was derived from the presence of an alpha/beta-unsaturated ester. With TIMES-SS however, MMA was negative. Accordingly, Kimber (2019)** concluded that “Data obtained using 2 QSAR methods (DEREKTMand TIMES-SS) were less encouraging and failed to reflect results recorded with in vivo or in vitro test methods, or with available human data.” Irrespective of the inconclusive results for MMA, the role of SAR analyses for the identification of sensitizers remains limited.  

 

Defined Approaches

In the OECD Guidance 497 of 2021***, three Defined Approachs (DA) are presented. These DAs are based on fixed data interpretations of different combinations of in chemico, in vitro and in silico data and come to a rules-based conclusion on the potential dermal sensitisation hazard of a substance. These DAs were validated against available LLNA/ OECD 429 data with a set of training substances including MMA. The DA for hazard identification (i.e. sensitiser versus non-sensitiser, using DPRA & h-CLAT) resulted in a prediction for MMA as skin sensitizer with a high level of confidence. The two of the DAs that provide information for the sensitisation potency categorisation (using as third component data from either DEREKTM or the OECD Toolbox) resulted in two predictions for MMA as skin sensitizer Cat 1B with a high level of confidence. This outcome on the potency is consistent with the other lines of evidences (most relevant, human & animal data) but the reliability is considered as somewhat limited due to the integration of in silico data (see above).

  

Conclusion 

The overall picture from all lines of evidence (physico-chemical properties, toxicokinetic information, human data, in vivo animal data on guinea pigs and mice/ LLNA, in vitro data, in silico data, defined approaches combining in vitro and in silico data) shows consistently that MMA is a skin sensitizer of weak potency.

Regarding the weighing of different data types/ lines of evidences, human data are generally considered as the most relevant and most adequate type of data for basic hazard assessment of this endpoint (i.e. differentiation of sensitisers from non-sensitisers). Animal data, specifically guideline LLNA data, are considered as data type with the highest adequacy for potency characterization towards Cat 1A/ 1B. The data types/ lines of evidences with lower weight in the assessment support the outcome of the human and animal data by closely consistent outcomes.

 

References    

 * OECD (2012). The Adverse Outcome Pathway for Skin Sensitisation Initiated by Covalent Binding to Proteins, Part 1: Scientific Evidence. Series on Testing and Assessment, No.168, May 2012 

 ** Kimber I. (2019). The activity of methacrylate esters in skin sensitisation test methods: A review. Reg Tox Pharm 104, June 2019, 14-20 

*** OECD (2021). Guideline No. 497: Defined Approaches on Skin Sensitisation. Published on June 22, 2021/ OECD Series on Testing and Assessment No. 336, Annex 2 (2021); https://www.oecd.org/chemicalsafety/testing/Annex-2-in-vitro-in-silico-in-vivo-defined-approaches.xlsx

  --

  Prevalence of MMA Sensitisation: Case Reports and Other Data 

Test Group (selection criteria)	Total no. or no. tested	No. tested positive	Incidence (%)	Reference	Potential bias comments
dental technicians and students (with and without exposure to MMA, no details reported)	175	0	0	Marx et al., 1982	Bias-negative: possibly no history of exposure to MMA Rem.: small cohort
dental technicians inwho are members of the Employer’s Liability Insurances Association BGFE, Berufsgenossenschaft Feinmechanik & Elektrotechnik, Germany (~ 95 % of the workforce in that field in) No. of applications due to recurrent, severe skin disease (BK 5101)	83191 85238 86102 73871 66696 65419 66412	195 206 200 209 214 247 194	0.23 0.24 0.23 0.28 0.32 0.38 0.29	BGFE, 1995 BGFE, 1996 BGFE, 1997 BGFE, 1998 BGFE, 1999 BGFE, 2000 BGFE, 2001 (all: personal communication)	Bias-small (Two effects balance each other - not everyone is expected to apply (negative bias) but cases with allergies against metals or other chemicals in the field are also included (positive bias)) Rem.: Probably the most reliable estimate
patients with dermatitis with previous contact with (meth)acrylates (dental products, adhesives, ptinting agents, photopolymers, plastics) ((meth,)acrylate allergy suspected, no further details)	82	1	0.8	Guerra et al. (1993)**	Bias-positive: patients of dermatologists or dermal clinics are a cohort pre-selected for having a problem with the skin
patients with dermatitis (suspected of (meth)acrylate allergy; no further details)	1161	9	0.8	Schnuch (1997)	Bias-positive: patients of dermatologists or dermal clinics are a cohort pre-selected for having a problem with the skin
patients with dermatitis (suspected of (meth)acrylate allergy; no further details)	4221	51	1.2	Schnuch (1996)	Bias-positive: patients of dermatologists or dermal clinics are a cohort pre-selected for having a problem with the skin
patients with dermatitis (suspected of (meth)acrylate allergy; no further details)	4900 3080 4099 5812 4427		1.4 1.2 1.6 1.4 1.0	Pratt et al. (2004) Marks et al.(1995, 1998 and 2003) Zug (2009)	Bias-positive: patients of dermatologists or dermal clinics are a cohort pre-selected for having a problem with the skin
patients with dermatitis with previous contact with (meth)acrylate (no further details)	352	17	4.8	Tucker and Beck (1999)	Bias-positive: patients of dermatologists or dermal clinics are a cohort pre-selected for having a problem with the skin
patients with dermatitis with previous contact with (meth)acrylate (no further details)	271	20	7.4	Kanerva et al. (1997)	Bias-positive: patients of dermatologists or dermal clinics are a cohort pre-selected for having a problem with the skin
dental technicians with dermatitis, (details re. Exposure to MMA not reported)	72	9	13	Schnuch and Geier (1994)	Bias-positive: patients of dermatologists or dermal clinics are a cohort pre-selected for having a problem with the skin
1619 patients with contact dermatitis (23 (meth)acrylate-exposed persons were tested, no further details)	23	3	13	Kiec-Swierczynska (1996)	Bias-positive: patients of dermatologists or dermal clinics are a cohort pre-selected for having a problem with the skin
dental technicians (with dermatitis, details re. Exposure to MMA not reported)	93	17	16	Peiler et al., 1996	Bias-positive: patients of dermatologists or dermal clinics are a cohort pre-selected for having a problem with the skin

 * This reference and all other ones below look at patients, i.e. cohorts which are more or less selective towards the occurrence of dermatitis, all persons without skin problems have no reason to go to a dermatologist and are so inadvertently excluded from the study 

** Betts et. al. (2006) cited wrongly: Actually a printer was sensitized against methacyl methacrylate instead of patients working with dental products or adhesives. 

Respiratory sensitisation

Endpoint conclusion

Endpoint conclusion:

no adverse effect observed (not sensitising)

Additional information:

A general introduction to the sensitisation potential of MMA is given above in the chapter “Skin Sensitisation”.

 

For this human health hazard, a well-defined test strategy does not exist in the REACh context for the time being. Moreover, no appropriately documented guideline studies with humans (specific inhalation challenge tests, SIC tests) exist for MMA. Thus, a scientifically valid weight-of-evidence (WoE) approach has to be applied. For MMA and this hazard, such a comprehensive WoE investigation on predominantly human data has been performed by Pemberton & Kimber (2022) that represents the most reliable hazard assessment for the time being. This publication (attached to this E.S.) thus forms the basis of the following information.

Additional information is related to 6 SIC tests claimed to be related to MMA that became available after the literature search of the Pemberton & Kimber paper. Although sparsely documented, these cases are in the focus of a currently ongoing classification process (2019-2022) so that relevant information from these cases is integrated in the respective line of evidence of the Pemberton & Kimber evaluation.

The considered data set is completed by further, non-human data of less relevant lines of evidences. Those have been assessed by Borak et al. (2011) in a slightly different WoE approach, and a continuative assessment of 2020 covering data between the Borak publication and 2020. These less relevant lines of evidences are summarized below; the continuative assessment and the Borak publication are attached as well.

For overview and volume reasons, the cited references from the Kimber & Pemberton paper are reduced to a minimum in the following text: complete references are available in the attached documents.

 

Human Data

Human data are in general the most relevant type of data for classification purposes under CLP.

In their WoE assessment, Pemberton & Kimber categorized all available information into six lines of evidence and evaluated individually their quality (using a Klimisch scoring system adopted to clinical data; K) and strength for causal relationship (by modified Bradford-Hill criteria; BH). Then, the level of confidence and the uncertainty levels were defined against the tested hypothesis (“MMA can cause the development of occupational asthma (OA) in subjects that were not previously asthmatic”) for each information before integration of all lines of evidences. The content of this assessment is summarized as follows.

 

Line 1 - Worker health studies

Eight out of 13 worker health records were rated Klimisch category (K2) and 6 of these were cross-sectional worker health studies in the cast acrylic sheet manufacturing industry by independent physicians with Della Torre, Monroe, Pausch, and Pickering studies being more comprehensive with high inclusion rates compared with those of Marez and Mizunuma. No increased risk of asthma was reported in the K2 rated studies.

The only indication of an increased risk of asthma came from two K3 rated studies. Jaakkola et al. (2007) reported an increased risk of adult-onset asthma in dental assistants with claimed exposure to methacrylates. Low weight was attributed to this study, as no details of exposure levels or chemicals involved were reported, and no individual case reports of asthma were included. Moreover, the incidence of asthma and lower respiratory tract symptoms were reported to be comparable between former Swedish dental technician students and general population controls.

Jedrychowski and Fonte (1984) also reported a higher incidence of asthma and obstructive syndrome, but not chronic bronchitis, in styrene, benzene and MMA workers. Low weight was assigned to this study as mixed exposures were reported and any possible effect attributable to MMA alone could not be discerned. Overall, there is no convincing evidence for an increased risk of asthma in worker-health studies involving high exposures that were predominantly to MMA.

Irritation of the eyes and the upper respiratory tract, mucosa and throat was reported consistent with the Harmonized classification of MMA within the EU as irritating to the eyes (H319) and respiratory system (STOT SE 3, H335) (ECHA, 2021c).

The information was evaluated against the adapted BH criteria as follows:

Adapted BH Criteria	Confidence score	Justification
Strength	high	There is no increased risk of asthma in the K2 rated worker-health studies with no consistent contradictory evidence from the K3 rated studies
Consistency	high	The health effects being observed are in workforces with high exposures to predominantly MMA and that the only indication for an increased risk of asthma come from 2 studies with low reliability and mixed exposures
Specificity	high	These associations are likely to be due to the known irritant property of MMA vapour
Temporality	high	The health effects being observed follow, in the majority of cases, many years of employment in the same industry and thereby discount any healthy worker affect
Biological gradient	moderate	The health effects reported are being observed in workers with the highest exposures to MMA
Plausibility	moderate	The health effects observed are typically associated with irritation, a known property of MMA
Coherence	moderate	Primary chemical irritation is known to cause these health effects
Experiment	N/A	N/A
Analogy	N/A	N/A

This led to an overall high level of confidence in the line of evidence for worker health studies as some inconsistencies from the K3 studies were included in the BH considerations above. 

 

Line 2 - Case study records

All 29 case studies (in 19 publications) were rated K3 (27 cases) or K4 (2) as they lacked essential detail on exposure within the workplace to meet the criteria according to ECHA R.7a guidance to a sufficient extend (ECHA, 2017a), thereby limiting the maximum level of confidence that could be achieved.  

Six clinical cases involved products that apparently did not contain liquid MMA so did not add to the body of evidence for this chemical. Five clinical cases involved mixing and use of orthopaedic cements. These cements were described as containing, or likely contained, gentamicin. Gentamicin is an antibiotic that has been linked to development of OA. The temporal alignment between starting work with the cement containing gentamicin and the development of OA (Pickering et al., 1986), and the negative SIC when challenged with cement containing MMA, but not gentamicin (Reynaud-Gaubert et al., 1991), points strongly towards development of OA due to gentamicin rather than MMA. However, the elicitation of LAR when mixing cement with MMA, but not when this was replaced with water (Pickering et al., 1986), perhaps points to MMA being able to trigger an asthmatic response in a subject that has already developed asthma due to prior exposure to another substance. In terms of consistency and the prevalence of OA in this sector, there are 22,288 orthopaedic surgeons with presumably a corresponding number of orthopaedic nurses within the EU conducting 750,000 hip replacements per year. So, irrespective of the causative agent, the prevalence rate of OA in this sector is extremely low.

Six case studies were from the dental sector in which use of MMA is commonly described for technicians. However, the authors apparently attribute the development of OA in these cases to use of MMA without providing any evidence as to why other chemicals present in the wide range of restorative materials used in this sector can be discounted. In terms of consistency, and the prevalence of OA cases claimed to be due to MMA, there are 40,000 dental laboratories and 210,000 dental technicians and 837,000 dental assistants within the EU indicating that the prevalence rate, in any event, is extremely low.

Eleven case studies contained insufficient detail on workplace exposure to indicate with any confidence a causal inference for MMA. For example, nail/beauty technicians are recognised as using a number of artificial nail technologies including gel-, wrap- and cyanoacrylate-nails none of which contain MMA. Exposure to hydroxyethyl methacrylate (HEMA), hydroxypropyl methacrylate (HPMA), ethylene glycol dimethacrylate (EGDMA) and cyanoacrylate used in these technologies has been reported (Moshe et al., 2019 and Sauni et al., 2008) thereby weakening the strength of attributing the development of OA solely to MMA in these cases.

In terms of analogous evidence, there have been 7 clinical cases reports of OA due to other methacrylates in the literature, primarily again related to dental sector but dentists and dental nurses. Here, complex mixtures are used as adhesive and primer products with several methacrylates and acrylates as typical components but rarely, if at all, MMA as component. Those products, combined with the use of other dental products, renders the attribution of the development of OA to MMA subject to some uncertainty.

The information was evaluated against the adapted BH criteria as follows:

Adapted BH Criteria	Confidence score	Justification
Strength	very low	There were no K1 or K2 studies identified. All studies lacked essential details on workplace exposure to assert with confidence that MMA and not other chemicals used was the causative agent.
Consistency	very low	The prevalence rate in orthopaedic and dental sectors is extremely low and inconsistent with other recognised respiratory sensitisers.
Specificity	low	These associations are more likely to be due to the known irritant property of MMA vapour triggering asthma-like symptoms in an asthmatic.
Temporality	low	It is claimed that the OA developed after exposure to MMA but this is not substantiated with evidence. Insufficient data upon which to score.
Biological gradient	very low	Insufficient data upon which to score.
Plausibility	very low	The causal link between MMA and development of OA is not established neither is an immunological MoA. Observed effects could equally be due to irritation.
Coherence	very low	Irritation is known to trigger asthma in asthmatics and could be equally responsible for the pattern of disease.
Experiment	N/A	N/A
Analogy	low	There is some evidence that other methacrylates or acrylates may be linked with OA in dentists. However, the numbers are small and exposures are mixed so it is uncertain whether this is causally related or coincidental elicitation due to irritation.

The strength of the clinical case study evidence is hampered by the understandable difficulties in recording an accurate and historical profile of workplace exposure to the suspected causative agent, as well as other agents that could be responsible. Typically, this is only achievable in the industrial setting, with the worker health studies comprising the first line of evidence. 

The main inconsistency in the line of evidence is the reporting in some studies of exposure to other chemicals that might have equally been responsible for the development of OA, for example in the orthopaedic and dental sectors. Importantly in other case studies from the same sector, no such co-exposure is reported. This, combined with the low prevalence of effect and lack of dose response, results in an overall very low level of confidence.

 

Line 3 - Quantitative Structure-Activity Relationship (QSAR) analyses

Three records describing QSARs for asthma and mentioning MMA have been identified. In summary, QSARs have been developed based upon learning set of substances that have been cited in the literature as being the cause of OA (asthmagens) as opposed to chemicals that have been proven to be respiratory allergens. These QSARs have subsequently identified the presence of the acryl (R-C=O) functional group capable of Michael addition as a structural alert for asthma thereby forming a rationale for associating acrylics (acrylates, methacrylates and cyanoacrylates) with asthma. While the claimed predictive value against all LMW substances based upon the learning set has been optimised, their performance with substances outside the learning set is of lower confidence. Overall, these QSARs are rated as K3 as their predictive value for MMA is low.

 

Line 4 - Exposure data

Methyl methacrylate has a vapor pressure of 37 hPa at 20 °C and vapour density quoted as between 3.45 (Air = 1) or 4.16 meaning that the vapour will settle at floor level and motions in the air created by wind and other pressure differences make the vapour rise.

Ten records on exposure mentioning MMA have been identified in the literature search and references therein. Three records addressed dental technicians, 2 were in dental clinics, 3 were in nail salons and 2 in operating theatres. A further 6 studies on the health of workers in the cast acrylic sheet manufacturing industry reported exposure data on MMA.

In the published studies in dental technicians MMA exposures (8hr TWA) reportedly varied between 2.94 – 6.7 mg/m³ and 327.28 ± 79.42 mg/m³. Peaks were reported up to 40.74 mg/m³ in the first study and only slightly higher than the TWA, in the second, while in the third study MMA concentrations (likely peaks) “up to 2.4 times higher” than the local MAK value of 410 mg/m³ were reported, i.e. up to 900 to 1,000 mg/m³. Dust exposure of technicians was reported in one study as 2.35 ± 2.70 mg/m³ and in second study, pure silica dust was measured to be 3.6 times and 2.6 times higher than the respective local MAC of 0.1 mg/m³ during sandblasting of metals and ceramics grinding, respectively.

In a study in dental clinics, levels for MMA were not reported. Levels of another methacrylate, HEMA, were reported also as being low (0.003 mg/m³) in the breathing zone of the nurse with a maximum concentration of 0.033 mg/m³) with peaks (maximum concentration) 3-5 times higher. In a study on HEMA and MMA exposures in five public dental clinics, 8hr TWA levels were low while short-term exposure levels were higher (0.079 mg/m³ for HEMA and 0.015 m/m³ for MMA. Similar findings were also reported during dental filling treatments with maximum concentrations of 0.4 mg/m3 for MMA, 45 µg/m3 for HEMA, 13 µg/m3 for EGDMA, and 45 µg/m3 for TREGDMA being reported.

In a study in nail worker stations low levels of ethyl methacrylate were measured but no exposure to MMA was detected. In another study, 8 h TWA and peak exposures to MMA during fingernail application in manicurists’ shops were 22 mg/m³ and 84 mg/m³, respectively. In a third study, mean TWA concentrations of ethyl methacrylate in sculptured nail salons was 18.5 mg/m³ (mean TWA) and 64.3 mg/m³ (peak) and levels of PMMA dust in 16 personal samples were 0.9 mg/m³ and 1.4 mg/m³ for respirable dust and total dust, respectively.

Reported orthopaedic workplace concentrations were between 210 and 420 mg/m³ of MMA during hip and knee replacement operations under conventional operating conditions without laminar airflow. In another study on exposure of hospital operating personnel during operations where MMA was used in surgery only in 4 of 27 cases MMA concentrations above the detection limit (1.2 mg/m³) were found (3.7; 4.0; 4.0; 55.3 mg/m³).

In a further study, acute exposure levels of up to 1570 mg/m³ were measured during mixing bone cement without local exhaust ventilation. Further the review by Borak et al., 2011 cites previously unpublished data in which MMA levels measured by photo-acoustic spectrophotometry continue to increase for 10–20 minutes after mixing bone cement, reaching levels of greater than 4000 mg/m³ above the mixing bowl if emission controls were not employed.

One study reported levels of MMA in workers involved in the recycling of PMMA and manufacture of acrylic sheets with exposures of up to 736 mg/m3 (8hr-TWA). A second study reported mean 8hr TWA levels in four cast acrylic sheet manufacturing plants as being 2.94 mg/m³ (range “trace” to 161.7 mg/m³) with 1 hr TWAs of 488.8 mg/m³ to 1680 mg/m³. The same authors reported MMA exposure in the same plants 2 years later, as having mean 8hr TWA levels of 77.7 and 90.7 mg/m³ with ranges of 38-260 and 50-162 mg/m³. Further studies show the same pattern, with historical values for peak exposures up to 2856 mg/m³ and 8-h TWA as high as 420 mg/m³.

 

Further unpublished exposure data for these identified industries are available and cited within the EU Existing Substance Risk Assessment for MMA (EU RAR) (EC, 2002). Mean TWA exposures of 146 mg/m³, with peaks (15min TWA) up to 749 mg/m³ were recorded during manufacture of cast sheet. The EU RAR (EC, 2002) also reports exposures in dental laboratories and surgeries at workplaces with local exhaust ventilation (LEV) being usually between 3 and 6 mg/m³. According to the Federal Monitoring Authority of Hamburg, Germany, the short-term values at workplaces with suitable LEV are below 42 mg/m³. The German Federal Monitoring Authority reported significant differences depending on whether LEV is used. Short-term exposures (30 min, n = 4) up to 144 mg/m³ without LEV and 600 mg/m³ under unsuitable ventilation conditions have been measured. Short-term values (5 min) of between 420 – 840 mg/m³ were measured during specific tasks (alternately humidifying orthodontic components with liquid MMA and strewing with powdery pre-polymerised PMMA) in six laboratory workplaces that didn’t use LEV. Sometimes even higher levels were recorded when Flame Ionization Detectors (FIDs) capable of measuring peaks of even shorter duration were used. Real-time measurements in dental laboratories with a Fourier Transform Infrared (FTIR) detector reveal that short-term concentrations up to 650 mg/m³ occur for up to one minute (BG ETEM, 2021a).

  

Line 5 –European national surveillance data

The fifth line of evidence comprises clinical case studies reported to EU national surveillance networks. With exception of the BG ETEM (2021b) investigation, there were no comprehensive clinical details or exposure data available for any case study as prescribed by ECHA R.7a guidance. In some cases, details of the age, sex, sector or job and basis of diagnosis were available, while for many details were even more limited or completely absent. The level of confidence in the diagnosis of MMA causing the development of OA in these cases is therefore low.

The information was evaluated against the adapted BH criteria as follows:

Adapted BH Criteria	Confidence score	Justification
Strength	low	The clinical and exposure evidence is limited or absent in all cases. Co-exposure and likely other causative agents cannot be excluded and in some cases are more likely explanations
Consistency	low	The cases come from a wide range of sectors indicating a low prevalence. Evidence for dental and orthopaedic sectors under Case Studies would suggest that a very low prevalence rate exists but no comprehensive analysis is available.
Specificity	low	No data on workplace exposure levels upon which to exclude irritation exists for any case
Temporality	low	No clinical history is given for any case study
Biological gradient	low	No data are available on workplace exposure levels upon which to make this determination. However, no case studies exist for sectors with high exposures such as cast sheet manufacture or reactive floor coating resins.
Plausibility	low	It is not discernible if exposure to MMA occurred and if so whether this caused the development of OA or acted as an irritant trigger of symptoms.
Coherence	low	Irritant provocation of pre-existing asthma cannot be excluded.
Experiment	N/A	N/A
Analogy	N/A	N/A

The lack of details relating to the cases prevents any further modification and results in an overall low level of confidence in this line of evidence.

 

Line 6 – Specific Inhalation Challenge (SIC) test data

This line of evidence initially comprised a total of 31 identified SIC tests in patients who claimed to have developed OA as a result of occupational exposure to MMA (22 in the dataset reported by Suojalehto during the CLH process in 2020/21, 8 in publications, 1 by BG ETEM reported in 2021). Thereof, 17 patients did not respond during SIC, so that these cases are not further investigated.

Three of the remaining cases were excluded as the effects were more likely to be due to widely recognised respiratory sensitisers, i.e. gentamicin in the orthopedic sector (Pickering 1986 and patient #6 of Walters et al 2017). Two further cases were excluded as in the respective work sector or product no liquid MMA is used and thus false reporting is concluded (administration – Wittczak et al. 2003; plumber using Tangit® - Uriarte et al. 2013).

The body of evidence for the remaining 9 cases with positive LAR (late asthmatic response) can be broken down into following sub-lines of evidence: a) that patients were claimed to have given positive SIC in response to the substance MMA, b) that late asthmatic responses (LAR) in these SIC responses are indicative for immunogenic mechanisms, and c) that exposure to MMA is so low that the responses could not have been the result of irritation. The last aspect is specifically relevant for 6 SIC tests that were vaguely described in a letter of some authors of Suojalehto et al. 2020 to RAC (Annex 5 of the RAC-56 Opinion of March 2021). The protocols of these tests, which were performed by FIOH and three other centers, were claimed to be consistently performed by mixing “2-part cement with predominantly (>90%) MMA”.

Firstly, there is little information on the identity and levels of the substances that these patients were exposed  in the workplace and during the SIC.

Three of the remaining cases, i.e. case #8 of Piirila et al 1998 and two cases of FIOH mentioned in Annex 5 of RAC-56 Opinion, are described as dentists/ dental nurses in which the published literature recognises that there is workplace exposure to a wide range of complex products with function as adhesives, primers, or restorative composite materials. These products contain other, functional methacrylates and acrylates, plus organic or inorganic dust exposure, while MMA is rarely used in such products. In this workplace Glutaraldehyde (CAS 111-30-8), a known respiratory sensitizer, is used frequently for cold sterilization and also as component in some primer products. In a fourth case, describing a beautician with pre-existing asthma that was studied in a clinical centre of Madrid and that was also mentioned in Annex 5 of RAC-56 Opinion, personal communication with the responsible clinician reveals that the patient was actually tested with a product containing an undefined cyanoacrylate and MMA. Based on knowledge of products tested with other beauticians in this centre, MMA exposure during SIC test is unlikely. In a fifth case, describing an orthopedic technician that was studied in a clinical centre of Strasbourg and that was also mentioned in Annex 5 of RAC-56 Opinion, the patient was grinding a “recently hardened prosthesis” during SIC and thus not exposed to MMA vapours other than those minimum levels from residual monomers. For all these 5 cases, a causal link to MMA on the basis of typical occupational exposure alone can only be made with low confidence. The Annex 5 letter information reveals irregularities in the reporting of the mentioned 6 SIC tests to RAC (e.g., on tested products, SIC protocols and health status of the patients) that practically invalidates the Annex 5 letter as key basis for classification in RAC-56.  

For the remaining 4 cases of dental technicians (each one case tested by FIOH, by an Belgian clinical centre in Yvoir – both mentioned in the Annex 5 letter – , patient #6 of Lozewicz et al. 1985, and a bronchial hyperreactive person reported by BG ETEM, 2021b) it appears plausible that the patients used products in the SIC test whose “products’ main component was MMA (>90%) in liquid component according to the SDS or ingredient list” and that the SIC protocol foresees “mixing MMA monomer liquid with a suitable amount of methacrylic polymer powder” as described in the Annex 5 letter. However, such PMMA powders typically contain Dibenzoyl peroxide (BPO, CAS 94-36-0) as initiator for the intended polymerization of the liquid, which is a potent skin sensitizer in the LLNA. For these 4 cases, a causal link to MMA on the basis of typical occupational exposure alone can only be made with maximum moderate confidence.

Secondly, it is recognized that SIC tests with the respiratory irritant bleach (chlorine, which lacks sensitising potential) elicited two isolated LAR and one early and late (dual or DAR) in 13 cleaning employees with occupational asthma (Sastre et al. 2011). This indicates that LAR is not definitive indication of the involvement of immune mechanisms. Moreover, in quantitative terms of the types of responses observed in the 14 positive SIC, isolated EAR was obtained in 4 cases, DAR in 5 cases, and isolated LAR in 5 cases. For the 4 cases with maximum moderate confidence based on test substance considerations above, the numbers are 2 or 3 EAR and 1 or 2 DAR, thus no clear pattern in SIC responses potentially related to MMA. Finally, the dominance of 10/13 negative SIC tests with bleach in the Sastre paper appears comparable to the SIC pattern seen with MMA showing 17 negative tests. Overall, there is very low evidence that the SIC response pattern related to MMA is indicative for immunogenic mechanisms.

Thirdly, in terms of the strength of evidence that exposures in these positive SIC were too low to cause irritation there are two aspects that require further consideration. 1) Where exposure measurements during SIC tests were reported, those were not made in the positive SIC with MMA, but instead in SIC with “similar products and SIC protocols”. Since no details of these SIC are available other than they were made with different products, on other occasions with different patients and eventually other SIC protocols established at FIOH, and gave negative a SIC responses. The extent of their comparability to the 6 positive SIC made with MMA mentioned in Annex 5 of RAC-56 Opinion remains uncertain. 2) The measurement of MMA levels in SIC tests made by FIOH used the ISO 16000-6 method that is stated as being unsuitable for measurement of shorter sample intervals of less than 5 minutes. This is significant since short peak exposures that would not be detected by this method have been shown to occur in dental workplaces and have been reported as the cause of respiratory irritation in workers. Taken together, these aspects lessen the level of confidence from low to very low in the assertion that MMA exposure employed in 6 of these positive SIC mentioned in Annex 5 of RAC-56 Opinion were so low that it could not have resulted in irritation.

The information was evaluated against the adapted BH criteria as follows:

Adapted BH Criteria	Confidence score	Justification
Strength	very low	There is insufficient detail as to workplace exposure to chemicals other than MMA (e.g., Glutaraldehyde, Dibenzoyl peroxide) and the products tested in the positive SIC to make causal inference. The lack of exposure measurements during the +ve SIC with the weak read across to other -ve SIC combined with the use of an insensitive method for measurements of peak exposures introduces a high degree of uncertainty.
Consistency	low	14 +ve SIC with MMA were reported. Other -ve SIC were performed with patients also claimed to have OA due to MMA, but were not reported. This incidence is extremely low compared with the high numbers involved in dentistry within the EU.
Specificity	very low	The evidence supporting low exposure during +ve SIC is insufficient and therefore an alternative hypothesis of irritation is more likely. Accordingly, the SIC pattern of MMA is comparable with that of the respiratory irritant bleach/ chlorine.
Temporality	low	The claimed development of OA after an unknown period of working in the dental industry is consistent. However, clinical histories are not available with a few exceptions and the causal inference to MMA is not established with sufficient confidence.
Biological gradient	very low	No data are available on workplace exposure levels upon which to make this determination. However, no case studies exist for sectors with high exposures such as cast sheet manufacture or reactive floor coating resins.
Plausibility	low	MMA is a weak Michael addition reactor capable of causing contact allergy so it is plausible that it could cause OA. However, it is also a respiratory irritant so could be provoking a non-specific response.
Coherence	very low	Irritant provocation of pre-existing asthma cannot be excluded.
Experiment	N/A	N/A
Analogy	low	QSARs have a low predictive value for MMA and methacrylates.

The lack of details for the case histories, workplace and SIC exposure levels and co-exposure results in an overall low level of confidence in this line of evidence.

 

 Integration of LoEs (from Pemberton & Kimber 2022)

In terms of the clinical evidence, there was very high confidence in the worker health studies. These consistently showed that MMA caused irritation of the respiratory tract, mild congestion, but no asthma in workers exposed to a combination of high shift average and very high peak levels of MMA over many years. It would be expected that this respiratory irritation observed in industrial workers would occur in asthmatics exposed to similarly high peak levels of MMA during SIC, where LEV would be absent or ineffective.

Despite a certain disparity in some of the exposure evidence, there remains a high confidence in this line of evidence. The more reliable evidence from the industrial sector as reviewed in the EU RAR (EC, 2002), combined with exposure modelling are consistent with MMA being a heavy vapour capable of producing very high peak exposures of short-duration. Several studies have reported similar very high peak exposures occurring when mixing bone cements, a task comparable to the mixing of 2-part cements by dental operatives as simulated in SIC with MMA.  A combination of underestimation of the impact of LEV, different work tasks with products of different composition, and use of inappropriate measurement methods that could not detect short peak exposure levels, are the likely explanation for why some studies in the dental sector and SIC reported much lower exposures.

There was low confidence in the other clinical evidence (case studies, national surveillance data and SIC tests) for a combination of overlapping reasons. One of the primary weaknesses in these data is the lack of a comprehensive worker exposure history on which to base identification of the causative agent(s). In many cases there is clear evidence of selection and reporting bias reflecting a priori conclusion of causation by MMA. The characteristic smell of MMA and other meth(acrylates) may play a significant role in this respect. In some cases, this was further compounded by imprecision in the description of the chemistry involved and the clustering of acrylates, methacrylates and cyanoacrylates, without regard to their very different physical and chemical properties, and biological reactivity. In this respect, ECHA R.7a and various clinicial authors highlight the importance of making accurate exposure assessments both in the workplace and during SIC. Failure to follow this guidance prohibits any level of confidence in the assertion of a work-related pattern of disease causally related to a specific substance(s) and exclusion of non-specific irritation from SIC tests.

In regard to the latter, it is a commonly held view that the appearance of, or greater tendency for, isolated LAR in SIC points to involvement of an immune mechanism. However, this is not scientifically established and in the case of MMA, and the relative small numbers of positive SIC potentially related to is not sufficiently strong evidence (Pemberton and Kimber, 2021). In any event, the relatively small numbers involved in comparison with the large number employed in this sector points to an extremely low prevalence that is inconsistent with an inherent property to cause respiratory sensitisation.

There was also low confidence in the available QSARs for respiratory sensitisation/asthma as the predictive performance is less than claimed.   

 

The overall level of confidence for the 6 lines of evidence is summarised in the table below.

Adapted BH Criteria	Worker health studies	Exposure	Case studies	National surveillance data	SIC test data	QSARs
Strength	high	moderate	very low	very low	very low	low
Consistency	high	moderate	very low	very low	low	very low
Specificity	high	high	low	very low	very low	low
Temporality	high	N/A	low	very low	low	N/A
Biological gradient	moderate	N/A	very low	very low	very low	N/A
Plausibility	moderate	N/A	very low	very low	low	low
Coherence	moderate	moderate	very low	very low	very low	very low
Experiment	N/A	N/A	N/A	N/A	N/A	N/A
Analogy	N/A	N/A	low	N/A	low	N/A
Overall level of confidence	high	moderate	very low	very low	very low	very low

 

Uncertainty analysis (from Pemberton & Kimber 2022)

There are no generally accepted or validated animal models for sensitisation of the respiratory tract to LMW chemicals, and currently available QSARs are not sufficiently predictive to be of any value. The majority of the available clinical evidence is of such low confidence that it inevitably is associated with a high level of uncertainty; the majority of the asthma cases reviewed in this context do not even allow a clear medical assignment to the different types of OA according to World Asthma Organisation.

A high level of uncertainty is also given for the evidence of exposure to MMA as causative agent for OA as long as co-exposure to material with the potential for relevant respiratory effects (e.g., gentamicin in the orthopaedic sector, glutaraldehyde and dibenzoyl peroxide plus a wide range of anorganic dusts including plaster, non-precious and precious metals, ceramics and rare metal oxides in the dental sector, PMMA dust in both sectors) cannot be excluded specifically in those sectors where OA cases of interest were observed. Also, with regards to sector-specific MMA exposure, uncertainty on causality arises from the observation that OA was not consistently found in sectors of high MMA exposure (e.g., not in cast sheet production).

 

Conclusions (from Pemberton & Kimber 2022)

It is recognised that the available clinical evidence on MMA and asthma is extremely weak and of low reliability rendering it unsuitable for assessment purposes. Nevertheless, the high number of claimed cases of OA in the literature and in national surveillance databases arguably justifies its consideration under a WoE framework. Overall, the clinical evidence suggesting that MMA is the cause of OA is incongruent with the finding of an absence of asthma in the industrial (cast sheet) sector with higher exposures over many years. 

1. There are no published cases of OA for which acquisition can be causally related to workplace exposure to MMA with moderate to high confidence. In those cases identified, the causative agent is more likely to be other known chemical respiratory sensitisers, or has not been adequately established. Any work-related pattern of disease corresponding with MMA exposure is more likely the result of irritant exacerbation of pre-existing disease.


2. The evidence in favour of the involvement of an immune-specific mechanism associated with allergic sensitisation is both weak and insufficient since no cases of “respiratory sensitisation” due to MMA have been documented. The only evidence suggestive of involvement of an immune mechanism comes in the form of a small number of OA cases with positive SIC from the dental sector. However, irritant provocation is the more likely explanation since a) exposure levels used were not sufficiently established to be below levels that cause irritation and, b) the appearance of LAR, combined with EAR, in a few cases does not indicate with any confidence immune involvement. Furthermore, considering workers in the dental sector are known to have mixed exposures to other chemicals, including dusts, both in their workplace and during the SIC tests, it is not possible to attribute causation of OA or, indeed, “specificity” of the asthma response in SIC, to MMA.


3. MMA is a recognised irritant chemical and has been reported to cause respiratory irritation in industrial workers exposed to high levels of MMA vapours. Strongly irritating substances are known to cause “irritant induced” asthma. However, MMA is only a weak irritant both in animals and humans and this is consistent with an absence of asthma in industries (cast sheet and floor resins) with much higher and dedicated MMA exposure levels compared with those in the dental sector.

Taken together, the available evidence does not support MMA as being the cause of OA. 

   

Further data (non-human)

Metabolism 

The aforementioned Michael addition is a widely accepted protein binding mechanism that can be used to identify skin sensitisers. However, for the time being, in recent publications no respiratory sensitizer has been identified with Michael addition as protein binding mechanism (Krutz et al. 2022; Sadekar et al. 2022) although Michael addition was postulated as potential mechanism in earlier QSAR publications (Enoch et al. 2012). This observation may serve as rather weak, mechanistic indication that MMA as Michael acceptor electrophile is not a respiratory sensitizer.

Specifically with regards to respiratory effects and potential sensitization, it is important to understand that high level exposure to MMA vapours are known to cause inflammatory reactions in the respiratory tract due the rapid, carboxylesterase-driven hydrolysis in these tissues to the corrosive and cytotoxic metabolite methacrylic acid (see chapters Toxikokinetics and Repeated dose - inhalation plus Chan et al. 2008). Thus enzymatic hydrolysis can be understood as weakly toxificating process for the irritative potential of MMA to the respiratory tract, eventually appearing with some delay, while these inflammatory reactions should not be misinterpreted as adverse outcome of a sensitizing process.

 

In chemico/ vitro 

An Adverse Outcome Pathway (AOP) for respiratory sensitisation to low molecular weight substances is under development at the OECD and the interim status has been published (Sullivan et al. 2017). From the proposed Key Events, only data for Key Event 1 (“covalent binding of substances to proteins”) are available for MMA.

DPRA tests of guideline and pre-guideline quality addressing show a consistent picture with a dominating depletion of cysteine generally being >30% and lysine depletion generally <10% (e.g. Dow, 2020, Wareing et al. 2017). According to ECHA Guidance R.7a (2017b; “there seems to be a greater selectivity of respiratory sensitisers for lysine reactivity than for cysteine, whereas skin sensitisers bind both to cysteine and lysine (Lalko et al., 2013a )”) this pattern appears contra-indicative for a respiratory sensitiser. Other research indicates that the protein reactivity (for either cysteine or lysine) is not a unifying, general pattern of respiratory sensitisation but associated primarily with chemical structure (Krutz et al. 2021) so there is a uncertainty about the interpretation of this pattern for the time being.

 

Animal data 

No data available for MMA specifically for respiratory sensitization. Moreover, as role of animal data for the identification of respiratory sensitizers is in general limited, there is no data gap identified which could improve the quality of the hazard assessment. 

Animal testing with MMA specific to sensitization is restricted to skin sensitization and is described above, indicating a maximum weak sensitizing potential. 

 

Overall summary 

In absence of clear key studies or key information, a weight-of-evidence assessment was performed to answer the question whether there is an additional risk for respiratory sensitisation of handling MMA which must be addressed with an additional classification, additional safety measures and/or modified occupational exposure levels. Details of the performed weight-of-evidence approaches are attached to this endpoint summary.  

In conclusion, a detailed evaluation of the available data indicates that it is inappropriate and unnecessary to classify MMA as a respiratory sensitiser. There is more than sufficient evidence that MMA has the potential to provoke respiratory irritation, especially in individuals with pre-existing asthma. Existing safety measures are sufficient for safe handling of MMA if they were fully implemented.

 

Reference list 

Borak J, Fields C, Andrews LS and Pemberton MA (2011). Methyl methacrylate and respiratory sensitization: A Critical review; Critical Reviews in Toxicology. 41(3): 230–268

Chan K, O’Brien O’Brien PJ. (2008). Structure-activity relationships for hepatocyte toxicity and electrophilic reactivity of alpha,beta unsaturated esters, acrylates and methacrylates. J Appl Toxicol. 28:1004–1015 

BG ETEM (2021a). Exposure data for Dental Technicians in Germany. Personal communication to CEFIC Methacrylate Sector Goup for submission to the CLH review on MMA. Berufsgenossenschaft Energie Textil Elektro Medienerzeugnisse. Köln 50968, Germany.

BG ETEM (2021b). Occupational Diseases by Acrylates in the Dental Technology in Germany 2015-2019 [Berufskrankheiten durch Acrylate im Bereich Zahntechnik]. Personal communication to CEFIC Methacrylate Sector Goup for submission to the CLH review on MMA. Berufsgenossenschaft Energie Textil Elektro Medienerzeugnisse. Köln 50968, Germany.

Pemberton MA & Kimber I (2022). Methyl methacrylate and respiratory sensitisation: a comprehensive review. Critical Reviews in Toxicology 52(2): 1-28

Pemberton MA, Kimber I. (2021). Classification of chemicals as respiratory allergens based on human data: requirements and practical considerations. Regul Toxicol Pharmacol. 123:104925

Roberts DW, Aptula AO, Patlewicz G, et al. (2008). Chemical reactivity indices and mechanism-based read-across for non-animal based assessment of skin sensitisation potential. J Appl Toxicol 28:443-454

Sastre, J et al. (2011). Airway Response to Chlorine inhalation (Bleach) Among Cleaning Workers With and Without Bronchial Hyperresponsiveness. American Journal of Industrial Medicine 54:293–299

Justification for classification or non-classification

Skin Sensitization  

The overall picture from all lines of evidence (physico-chemical properties, toxicokinetic information, human data, in vivo animal data on guinea pigs and mice/ LLNA, in vitro data, in silico data, defined approaches combining in vitro and in silico data) shows consistently that MMA is a skin sensitiser of weak potency.

Regarding the weighing of different data types/ lines of evidences, human data are generally considered as the most relevant and most adequate type of data for basic hazard assessment of this endpoint (i.e. differentiation of sensitisers from non-sensitisers). Animal data, specifically guideline LLNA data, are considered as data type with the highest adequacy for potency characterization towards Cat 1A/ 1B. The data types/ lines of evidences with lower weight in the assessment support the outcome of the human and animal data by closely consistent outcomes.

MMA is classified with Skin sens Cat. 1, H317 according to the requirements of Regulation (EC) 1272/2008, Annex VI in the EU (harmonized classification).

From UN GHS perspective, many studies from the key lines of evidences (human and animal data) consistently indicate a weak potency of MMA so that a subcategorisation into Cat 1B would be justified based on data. Valid data from other lines of evidences would support this subcategorisation.  

--

Respiratory Sensitization

The outcome of a detailed weight-of-evidence assessment of all the available information in July 2022 indicates that it is inappropriate and unnecessary to classify MMA as a respiratory sensitiser.

The WoE assessment covers information on 6 SIC cases that were discussed in RAC-56 as Annex 5 but this source of information was invalidated due to significant irregularities in the reporting to RAC (e.g., on tested products, SIC protocols and health status of the patients). There is more than sufficient evidence that MMA has the potential to provoke respiratory irritation, especially in individuals with pre-existing asthma. Existing safety measures are sufficient for safe handling of MMA if they were fully implemented.

Therefore, a classification for respiratory sensitisation is considered as not justified.