Registration Dossier

Environmental fate & pathways

Bioaccumulation: aquatic / sediment

Currently viewing:

Administrative data

Link to relevant study record(s)

Description of key information

The test item is not bioaccumulative according to PBT criteria. 

Key value for chemical safety assessment

Additional information

The bioaccumulation potential of the substance was assessed in a weight of evidence approach using several QSAR calculations as well as one experimental study.

 

The only available experimental result is a BCF study conducted in 1984 (Gakashuin University 1984). The study was conducted according to "Bioconcentration test for chemical substances in fishery products in accordance with Notification of Kanpogyo No. 5, Yakuhatsu No. 615, Kikyoku No. 49-392". For test substance preparation an emulsifier was used. The tested concentrations of the substance were 0.1 and 1 mg/L, respectively, and it resulted in BCF values of 13.6 - 13.8 (1 mg/L test concentration) and 14 - 19 (0.1 mg/L test concentration). Because the tested concentrations are far above the actual water solubility of the substance (< 0.0017 mg/L) and the additional use of a solubilizer in substance preparation the study cannot be regarded as valid. However, in a weight of evidence approach the results give additional information about the degree of bioaccumulation.

 

Several QSAR models have been used to properly assess the bioaccumulation potential in a weight-of-evidence approach. The single models and their results are summarized in the table below:

 

 

Model

 

BCF

Log BCF

Remarks

Catalogic v5.11.2

 

209.41

2.321

all mitigating factors applied; 96 % in domain

T.E.S.T. v4.01

 

1418.78

3.15

EPISuite v4.10

Regression-based estimate

3321

3.52

Arnot-Gobas upper trophic level

452.9

2.656

Including biotransformation rate estimates

Arnot-Gobas mid trophic level

637.1

2.804

Including biotransformation rate estimates

Arnot-Gobas lower trophic level

706.9

2.849

Including biotransformation rate estimates

 VEGA CAESAR v1.0.8   90   1.96  could be out of model Applicability Domain

 

 

In general, the QSAR models only taking into account the logKow of the substance are hardly applicable (e.g. EPISuite regression-based estimate)because the partition coefficient of the substance is in a range were the linearity between logKow and log BCF is decreasing and therefore the obtained results are of low value.

 

Regarding the results of the model calculation, the calculated BCF values range from 209.41 (Catalogic) to 3321 (EPISuite v4.10, regression-based estimate). Of these models, Catalogic, VEGA CAESAR and the Arnot-Gobas model from EPISuite v4.10 take into account mitigating factors, e.g. metabolism, water solubility and/or size. Catalogic revealed a corrected BCF of 209.41 and the compound is 96% within the model’s applicability domain. . The Arnot-Gobas model from the EPISuite takes into account the biotransformation rate of the compound and calculates BCF values for the upper, mid and lower trophic levels. The values for the present compound range from 452.9 (upper trophic level) to 706.9 (lower trophic level). The model assumes default lipid contents of 10.7%, 6.85% and 5.98% for the upper, middle and lower trophic levels, respectively. Usually, in the context of REACH a default lipid value of 5% is assumed which represents the average lipid content of the small fish used in OECD 305 studies. Thus, the higher lipid values of the Arnot-Gobas model can be regarded as reasonable worst-case scenarios as higher lipid contents are usually associated with a higher potential for bioaccumulation.

The regression-based estimate from the EPISuite revealed a BCF value of 3321 based on a calculated logKow of 8.23. This model correlates the logBCF to the logKow. According to ECHA’s guidance on information requirements and chemical safety assessment, chapter R.11: PBT assessment, logBCF increases linearly with logKow which is the basis for the “B” screening criterion of logKow > 4.5. However, at very high logKow (> 6), a decreasing relationship between the two parameters is observed which makes a solid assessment difficult. The present compound has a calculated logKow of 8.23. Additionally, ACD labs was used to calculate the logKow. It resulted in a logKow of 8.7. Due to these values, the correlation of the logKow and the logBCF is uncertain and the QSAR calculation of the regression-based estimation from EPISuite can only be handled with caution.

The T.E.S.T. package from the US EPA estimates BCF values using several different advanced QSAR methodologies. The recommended model of the T.E.S.T. package is the consensus method since it provides the most accurate prediction. This model estimates the BCF by taking an average of the predicted BCF values from the other applicable QSAR methods of the package. For the present substance the consensus method averaged the results from (1) the hierarchical clustering method, (2) the single model method, (3) the group contribution method, (4) the FDA method and (5) the nearest neighbor method. The resulting BCF value is 1418.78.

The VEGA model was developed with several descriptors and is based on a dataset of 473 compounds. It offers detailed information on the applicability domain. In the present case, the calculation gave a BCF value of 90. Since the Global AD Index was calculated to be 0.85 the substance could be out of the Applicability Domain of the model.

 

In summary, in a weight-of-evidence approach balancing the different QSAR estimations and the experimental results the test substance is not bioaccumulative according PBT-criteria.