Consultation publique concernant le projet de l’agence nationale de sécurité sanitaire de l’alimentation, de l’environnement et du travail (Anses) d’analyse de la meilleure option de gestion réglementaire (RMOA) du disulfure de carbone (CS2)
Suite à l’évaluation de la substance disulfure de carbone (CS2) par la France dans le cadre du processus d’évaluation prévu par le règlement REACH (https://echa.europa.eu/documents/10162/80f86071-5648-67e8-1205-d5ee7fe174e1), une analyse de la meilleure option de gestion réglementaire (RMOA) a été menée par l’Anses pour clarifier l’existence d’un risque lié à l’utilisation de la substance par les travailleurs. Cette analyse a aussi été l’occasion de questionner si des mesures complémentaires de gestion des risques doivent être mises en œuvre.
Le projet de RMOA s’est tout d’abord focalisé sur la pertinence de la VLEP européenne. Il est conclu que la valeur actuelle ne nécessite pas d’être modifiée, en particulier au vu des études épidémiologiques et de la nouvelle étude étendue de toxicité pour la reproduction sur une génération.
Cependant, le projet de RMOA fait état d’un manque de données d’exposition ne permettant pas de conclure solidement sur l’évaluation des risques, et donc sur la meilleure option de gestion réglementaire. Les connaissances sur les utilisations du disulfure de carbone sont limitées en particulier concernant les utilisations professionnelles en laboratoire. Le projet de RMOA mentionne également un manque d’informations sur les données d’exposition et sur les alternatives potentielles. La présente consultation est une opportunité pour transmettre ces données sur la substance et ainsi permettre de compléter l’analyse proposée par l’Anses.
En l’état des connaissances, le projet de RMOA recommande de mettre à jour la classification harmonisée du CS2 (précision des organes cibles pour la classification quant à sa toxicité spécifique en cas d’exposition répétée et ajout d’une classification pour la toxicité aiguë par inhalation).
Le RMOA pourrait également préconiser de proposer une restriction des utilisations du disulfure de carbone pour limiter les expositions par inhalation et voie cutanée mais les données actuelles ne permettent pas d’identifier un risque avéré pour les professionnels même si celui-ci apparait par modélisation dans quelques situations. Des données d’exposition affinées collectées dans le cadre de cette consultation publique permettront de reconsidérer si cette mesure de gestion est nécessaire.
L’identification SVHC est discutée et représenterait un signal fort pour communiquer sur les propriétés de danger de la substance. Cependant, le processus d’autorisation ne couvre pas les usages de fabrication, d’intermédiaire et de recherche et développement, or ces usages sont les plus représentatifs pour le CS2.
Enfin, au vu des différentes VLEP utilisées en Europe, le projet de RMOA indique qu’une autre option de gestion réglementaire serait de fixer une VLEP réglementairement contraignante et non plus indicative dans le cadre de la directive européenne sur les agents chimiques pour permettre de limiter l’exposition des travailleurs et d’harmoniser les pratiques au niveau européen. L’Anses indique que la fixation d’une valeur limite biologique en complément de cette VLEP permettrait également d’assurer une plus grande sécurité pour la santé des travailleurs.
La consultation publique se termine le 27 février 2023.
The DE CA agrees in most parts to the RMOA prepared by the FR CA. According to the RMOA, in 2008 the SCOEL recommended an 8-hour TWA OEL of 5 ppm or 15 mg/m3 based on 10 ppm as point of departure and applying an uncertainty factor of 2 based on the severity of the effect (cardiac ischemia). Additionally, as stated in the RMOA, “the Health Council of the Netherlands (HCNL) concluded that 15 mg/m3 (5 ppm) is a LOAEC for cardiovascular effects based on the induction of minor cardiac ischemic findings in human, observed in Takebayashi et al., 2004 study”. Furthermore, it is stated that “Due to the large amount of human data on workers, an uncertainty factor of 1 for intraspecies differences seems also appropriate.” The DE CA agrees with using the results of the study by Takebayashi et al. (2004) to derive a DNEL for workers. However, we would like to express our concerns at using an uncertainty factor of 1 for intraspecies differences instead of the standard intraspecies assessment factor of 5 for workers. According to the ECHA Guidance on Derivation of DNEL/DMEL from Human Data “use of AFs lower than the standard assessment factors is appropriate when it can be shown that some of the factors that cause the intraspecies variation in the target population, such as gender, age, nutritional status, health, susceptibility and genetic polymorphism have been covered in the study population”. However, the study by Takebayashi et al. (2004) included only male workers of similar age with no history of cardiovascular or cerebrovascular disease as described in Annex 4, Section 1.1.1 of this RMOA. Therefore, the study cohort cannot be regarded as representative in terms of gender, age, and health, and an intraspecies assessment factor above 1 should be chosen. Additionally, with regard to the severity of effects, the LOAEC of 15 mg/m3 (5 ppm) derived by HCNL may be used as point of departure for DNEL derivation together with an extrapolation factor of 3 from LOAEC to NOAEC and an intraspecies assessment factor above 1.
As far as we know, the Institute for Occupational Safety and Health of the German Social Accident Insurance (IFA) has collected some measurement data for carbon disulfide. These data are not yet published in the MEGA-database, but it might be worthwhile to contact the IfA : https://www.dguv.de/ifa/gestis/expositionsdatenbank-mega/index-2.jsp
NL comments on the RMO paper for carbon disulfide
EC number : 200-843-6
CAS number : 75-15-0
We would like to thank ANSES/ the French CA for finalizing the substance evaluation (from which it was concluded that carbon disulfide possesses no endocrine disrupting properties), drafting and sharing the RMOA for carbon disulfide from the perspective of establishing and addressing the concern for workers.
Although the Tier I modelling tool ECETOC TRA is normally conservative, the modelled exposure value obtained with by ECETOC TRA was certainly not worst case enough compared to the available air monitoring exposure measurements for manufacture. Given the fact that the air monitoring value proposed in the dossier was more than 50-fold higher than the modelling value obtained with ECETOC TRA Workers. This is even strengthened by the fact that carbon disulfide vapour may be readily absorbed via the skin, which may lead to underestimation of the total (both dermal and inhalation) exposure. It raises questions if the other exposure levels for industrial use sites estimated with ECETOC TRA can be considered reliable or that those in analogy with PROC 1 also are (heavily) underestimated. Given this information it is seems logic that extra biomonitoring or exposure data in the form of air measurements are desired to specify the concerns. (Why) has this not been taken into account during the substance evaluation? The NL-CA has no relevant data available. We would like to suggest the German MEGA-database that contains exposure data for many different substances, although we did not check for the availability of data regarding CS2.
In the RMOA several signals are noted with respect to the derived IOEL of 15 mg/m3/ 5 ppm. Study data seem to imply that the IOEL is sufficient or should slightly be lower. There is a large amount of human data available, the derivation of the OEL based on human data is considered more appropriate than based on animal data, being therefore less influenced by potential differences between species. The NL-CA agrees upon this statement. The FR-CA seems to conclude that the current IOEL is sufficient. The HCNL proposal (the Committee recommended a value of 5 mg/m3 (2 ppm) as a health-based occupational exposure limit applying a factor of 3 to the LOAEC of 15 mg/m3 (5 ppm), is not considered appropriate by the FR-CA. The marginal effects are not considered sufficiently critical to revise the current IOELV in the absence of confirmed clinical alteration at 5 ppm when rigorous “symptomatic” ECG criteria of ischemia were applied. We feel some sympathy for this point of view and agree that the effort should not be put in (slightly) lowering the IOEL. Looking at Table 3 in the RMOA we would like to remark that there are only two member states (Germany & Hungary), who set significantly (twofold) higher OEL’s. With this respect the effect of setting a BOEL seems to be rather limited and is to our point of view not an appropriate option.
Concern for workers
Based on the analysis provided by ANSES we share the conclusion that there is still a concern for workers in general (especially with respect to the RCR’s presented for the PROC’s in Table 8). The concern is mainly related to the mismatch between exposure estimates and measured data and the uncertainty in the exposure assessment. We share the conclusion that ECETOC TRA underestimates the exposure levels, which are already close to the IOEL. Consequently, the risks are also underestimated.
Regulatory management options
Requiring extra biomonitoring and air measurement data with respect to exposure seems necessary, however, the NL-CA does not (yet) have a clear picture how this will favor one or more regulatory management options.
As proposed by the lead registrant, additional classification of the substance as Acute Tox. 4 ; H332 is warranted. The specification of the nervous system and the cardiovascular system is also relevant for STOT RE 1 to better communicate on the hazards of the substance.
Overall, it is suggested by FR-CA to update the existing harmonised classification. Acute toxicity by inhalation is not a priority endpoint according to Art. 36 of CLP regulation for the need for Harmonised Classification and Labelling but the differences in self-classification may justify that action is needed at community level unless registrants align their self-classifications. Although this is true we do not see how this will address the concerns raised in the RMOA.
Carbon disulfide is a known neurotoxicant that can induce damages in the central nervous system (CNS) and peripheral nervous system (PNS) in experimental animals and in humans. Carbon disulfide induces also toxicity to the cardiovascular system. The substance is classified as STOT RE 1, in high potency class, under the CLP Regulation. So, identification under Article 57(f) based on equivalent level of concern (EloC) for STOT RE could be considered, with could be followed by listing on Annex XIV. The primary aim of authorisation under REACH is to substitute SVHCs and it has proven to be an effective driver for this substitution. Within this respect we would like to mention that to our point of view SVHC-identification followed by authorization would be a relevant option to stimulate substitution, but only for a few uses of the substance. In addition, it is unknown whether safer technically feasible alternatives are available for all uses regarding carbon disulfide. Authorisation does not cover the following uses of carbon disulfide : manufacturing, intermediate uses or research and development. If carbon disulfide is placed on Annex XIV, this would probably mostly cover the manufacturing of regenerated cellulose and the industrial uses when the substance is not used as an intermediate. Consequently, only the uses as a solvent and viscose manufacturing could be covered by REACH authorisation. For the widespread use of the substance as a laboratory reagent, it is not clear whether all the uses would be in the aim of research or development or if some uses would be in the scope of authorisation. Given the fact that authorization would only addresses a part of the concern, we do not favor this option.
The FR-CA states that a restriction of the uses which result in too high exposure levels to ensure safe use of carbon disulfide would be an option. This would be an and maybe also the most interesting option to consider to our point of view.
In addition, it is stated that adequate control of the substance could be ensured defining mandatory inhalation and dermal DNELs. We follow this reasoning with respect to dermal DNELs. However, for inhalation we can not agree, given the conclusion in the RMOA that the current IOEL (also used as DNEL) with respect to inhalation seems to be sufficient. Therefore, setting a binding DNEL for inhalation in a restriction would probably lead to a similar level and will probably not lead to any update of the exposure scenarios. The concern can mainly be traced back to compliance with legislation that is already in place and the enforcement of the OEL’s.
Centre for Safety of Substances and Products
Dutch National Institute for Public Health and the Environment
PO Box 1 | 3720 BA Bilthoven
Antonie van Leeuwenhoeklaan 9 | 3721 MA Bilthoven