Boron trifluoride

Endpoint summary

Additional information

Three studies were conducted to assess the genotoxic potential of boron trifluoride. For this, boron trifluoride dihydrate, was tested in an Ames test (May K., 2010), a mouse lymphoma assay (Woods I., 2010a) and an in vitro aberration chromosomal test (Woods I., 2010b).The three tests were conducted according to OECD guidelines and were GLP.

In the Ames test, no evidence of mutagenic activity was seen at any concentration of Boron Trifluoride Dihydrate in either mutation test (concentrations up to 5000 µg/plate were tested, vehicle: water). In the mouse lymphoma assay, Boron Trifluoride Dihydrate did not demonstrate mutagenic potential, up to concentrations that did not cause a fluctuation in pH of more than 1.0 unit and that did not cause toxicity.

The in vitro chromosomal aberration test gave inconclusive results: on the basis of pH that did not cause fluctuation in pH of more that 1.0 unit, the following concentrations were selected for metaphase analysis: first test (3 hour treatment, 18 hour recovery): in the absence of S9 mix: 162, 270 and 450 µg/mL; in the presence of S9 mix - 162, 270 and 450 µg/mL; second test: in the absence of S9 mix - 21 hour continuous treatment: 97.2, 162 and 270 µg/mL; in the presence of S9 mix (5% v/v) - 3 hour treatment, 21 hour recovery: 162, 270 and 450 µg/mL. In the absence of S9 mix, Boron Trifluoride Dihydrate caused statistically significant increases in the proportion of metaphase figures containing chromosomal aberrations at the highest concentration 450 µg/mL after 3 hour treatment (including gaps; but no significant increase if excluding gaps) and at 162 µg/mL after 21 hour treatment (including gaps, but no significant if excluding gaps) and 270 µg/mL (21 hour treatment, excluding and including gaps), when compared with the concurrent solvent control.

In the presence of S9 mix, Boron Trifluoride Dihydrate caused no statistically significant increases in the proportion of metaphase figures containing chromosomal aberrations, at any concentration, when compared with the solvent control, in either test.

In conclusion, Boron Trifluoride Dihydrate has shown evidence of causing an increase in the frequency of structural chromosome aberrations, in the absence of S9 mix, including and excluding gaps, but only after 21 -hour exposure period.

This leads to inconclusive results: Ames test and in vitro gene mutation assay were negative. It should be noted that the mouse lymphoma assay is able to detect certain clastogenic assay, but this was not observed in the current test. The in vitro chromosomal aberration test was positive under certain conditions only.

On Juli 5th, 2012 the European Chemicals Agency (ECHA) has taken the decision (based on a proposal from the Registrants) that an in vivo micronucleus test via inhalation route should be performed with boron trifluoride dihydrate by the registrant in accordance with the procedure set out in Articles 50 and 51 of Regulation (EC) No. 1907/2006. The investigation has been started with a technical trial, however, the pilot experiments and recalculations (reported below) let the Registrants ask ECHA for a  reconsideration of the decision for animal welfare reasons. Indeed, it has been shown that even at low atmospheric concentrations the substance will be an aerosol liquid whose corrosive properties would generate severe focal damage. In addition, it wil not be possible to achieve the concentration leading in vitro to the increase in the frequency of structural chromosome aberrations (See genetic toxicity in vivo-waiver).

The current German occupational exposure limits (OELs) are set at 1 mg/m3 for boron trifluoride (being 600 fold below the needed concentration to receive an effective dose in the blood). The high concentrations that need to be tested within the in vivo MNT will according to our calculations not lead to observable effects and will not decrease the DNEL below these OELs. For the reasons explained above, no in vivo mutagenicity tests including the in vivo MNT as well as other in vivo tests such as the Comet Assay should be conducted.

By letter on the 28th January 2014, Arkema as Lead Registrant was informed that a Statement Of Non Compliance (SONC) was issued against them. Indeed, pursuant to Article 41(3) of Rgulation (EC) 1907/2006 (REACH Regulation) the European Chemicals Agency (ECHA) has performed a testing proposal examination on the dossier boron trifluoride, CAS No 7637 -07 -2, EC No. 231 -569 -5 and taken the decision TPE-D-0000002344 -80 -04/F. In that decision, the Registrant was requested to submit by 5th July 2013 information on the following issue: Mammalian erythrocyte micronucleus test, inhalation route, (Annex IX, 8.4., test method : EU B.12/OECD 474) according to certain conditions.

ECHA has examined the information submitted in the update dossier on the 27th June 2013 and considered that the updated registration dossier did not contain the information requested by the ECHA decision as a waiving justification was provided instead of experimental data. ECHA did not consider the rationale for this waiving justification was provided instead of experimental data. ECHA did not consider the rationale for this waiving as being sufficient;

Please find below our answers regarding the following ECHA comments:

-" First, the Registrant refers to pilot experiments and calculations performed which he states are attached. However, the attachements could not be found in the technical dossier. Therefore, it is not clear on what arguments the registrant has based his adaptation".

Answer of the Registrant:

Please find attached the report of this experiment (12I048 -SR-LM-4) which demonstrated that the exposure atmosphere could have been almost exclusively composed of BF3 -2H2O aerosol. Under the current study conditions, a high aerosol fraction was detected at a dosing rate as low as 1.8 mL/h. In the range of LC50 value, BF3 is present as liquid aerosol in atmosphere. In addition, the sampling method for gas and vapours failed to provide reliable data for the atmosphere concentration due to a high absorption onto the chamber walls. Since the substance is corrosive, and given that the test atmosphere will be almost exclusively composed of BF3 -2H2O aerosol, it is evident that local corrosive effects would be induced at each impaction area as the substance would be available almost undiluted at each contact site.

By considering the animal welfare, the Study Director has taken the decision not to perform the study (see Boron trifluoride statement document) which is in line with the REACH and OECD guidelines. Specifically, Article 13(4) of Reach requires that toxicological and ecotoxicological tests shall be carried out in compliance with EU Directive 2010 -63 -EU on animal protection which sets out the basic requirements for the care and accomodation of Laboratory animals, and stipulates that experiments shall be designed to avoid distress and unnecessary pain and suffering to the animal. In addition, according to the general part of Annexes VII-X, in vivo testing with corrosive substances at concentration/dose levels causing corrosivity should be avoided to the extent possible.

-" Secondly, the Registant states that it is not possible to obtain concentration in vivo that would be similar to the concentrations used in the in vitro study that caused an increase in structural chromosomal aberrations due to the corrosive nature of the substance. The OECD/EU test guideline specifies that test concentrations should be selected to "cover a range the maximum to little or no toxicity observed, and further"... the highest dose may also be specified as a dose that produces some indications of toxicity of bone marrow..." This means that toxicity of bonne marrow is not a pre-requisite for the dose selection. The dossier contains studies that have been performed in vivo via inhalation route, where toxicity was observed already at low doses e.g. 6 mg/m3 (repeated dose toxicity study, kidney effects). Respiratory distress was noted in all doses (2 to 180 mg/m3), however with no abnormal histological findings. Moreover, the direct comparison between concentrations used in in vitro studies and doses to be used in in vivo study cannot be made, due to the differences in test systems. Therefore, ECHA considers that the Registrant's reasoning according to which " high enough dose levels cannot be obtained" is not supported by evidence.

Answer of the Registrant:

In the final decision TPE-D-0000002344 -80 -04/F, ECHA emphasizes that the Registrant needs to demonstrate that the test method is applicable for the test substances taking into account specifically paragraph 7 of the OECD TG 474 ("if there is evidence that the test substance, or a reactive metabolite, will not reach the target tissue, it is not apppropriate to use this test"). In view of the results of the experimental trial and the in vivo data indicating significant toxicity even at low doses, it became evident that the selection of dose-levels will be significantly low as well the low probability that the test substance reaches the bonne marrow. Thus, taking into account the animal welfare considérations and the pertinence of the test, a waiving statement was written by the Study Director.

Now, it appears that the comments made by ECHA in the SONC are in contradiction with the previous comments. ECHA seems now to consider that there is no need to prove the reaching of the target tissue (i.e. the bone marrow) and that the observation of clinical or systemic effects (kidney findings) which could indicate that the bone marrow has been reached were observed at high dose-levels in only few animals.

- Lastly, the Registrant refers to German OEL of 1 mg/m3 and the potential 600 time higher concentration (not clear if the Registrant's calculation is based on the in vitro study) that would need to be used to obtain an "effective dose in blood". The Registrant's conclusion is that the results obtained by using these high concentrations would not change the DNEL below the OEL. However, this statement has two drawbacks. First, hazard data from either animal studies or human evidence is needed before DNELs can be calculated. Therefore, before the data is available, it is impossible to know if DNELS or OELs would need to be changed. Moreover, structural chromosomal aberrations are considered to be following non threshold mechanisms; hence a dose without potential effects (DNEL) cannot be derived from such data. Consequently, the Registrant's arguments concerning the potential concentrations used and reference to DNELs is not accepted by ECHA.

Answer of the Registrant:

We agree with the fact that we cannot anticipate the results of a study, however, we firmly believe that whatever the result of the micronucleus test, this will not impact the systemic inhalation long term exposure DNEL. Firstly, all the available data indicate that the critical short- or long-term effect is the highly corrosive potential observed since low exposure and the existing DNEL is based on this effect. Secondly, in the absence of reliable chronic study data and/or elements that could suggest a carcinogen popential, no extrapolation to humans can be made.

Lastly, as the toxicity of the substance requires Risk Management Measures and low occupational exposure limits, there is no concern regarding the carcinogen potential of the substance.

In conclusion, we firmly believe that the in vivo micronucleus study in the rat by inhalation with boron trifluoride should not be conducted for the above-specified reasons. However, in view of the new elements provided by ECHA, we would be willing to perform the test and the update of the dossier by July 18th 2015 (binding deadline defined by the French Authorities in the Arrêté préfectoral N° 2014 -120 dated June 17th 2014) at dose-levels that are not expected to cause severe pain and/or distress in animals in the case where ECHA still considers our justification and additional information as insufficient and agrees on the absence of proof for reaching the target tissue i.e. the bone marrow.

Short description of key information:
Three tests are available. All were conducted according to OECD guidelines and are GLP. Ames test and mouse lymphoma assay are negative. The in vitro mammalian chromosome aberration test in human lymphocytes is positive but only without S9 mix and for a 21-hour exposure period (including and excluding gaps). The high damaging aerosol concentrations that would be needed to reproduce the same exposure in an in vivo test (erythrocyte micronucleus test) contra-vene animal welfare regulations and are unnecessary in regard to the low OELs.

Endpoint Conclusion: No adverse effect observed (negative)

Justification for classification or non-classification

According to EU directive 67/548/EEC and EU Regulation (EC) N0. 1272/2008 (CLP), BF3 is not classified for genotoxicity.