Registration Dossier

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

420 mg/m³

Most sensitive endpoint:

repeated dose toxicity

Overall assessment factor (AF):

4

Modified dose descriptor starting point:

NOAEC

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

595 000 mg/kg bw/day

Most sensitive endpoint:

repeated dose toxicity

Overall assessment factor (AF):

17

Modified dose descriptor starting point:

NOAEL

Correction of dose descriptors (route-to-route extrapolation), application of assessment factors (AFs) and derivation of the endpoint specific DNELs

Since there is no dose descriptor for every exposure route, dose descriptors were converted into a correct starting point by route-to-route extrapolation based on the ECHA guidance document "Guidance on information requirements and chemical safety assessment. Chapter R.8: Characterisation of dose [concentration]-response for human health", May 2008.

Within the ECHA Guidance a factor of 2 is suggested for the extrapolation from oral to inhalation absorption. On the contrary, the Technical guidance document on risk assessment in support of Commission directive 93/67/EEC, 2003 appendix IV A and B gives a number of physico-chemical properties that normally determine oral, inhalation and dermal absorption. These parameters include molecular weight, log Kow, pKa values and for inhalation also particle size distribution, vapour pressure etc.

Molecules with a molecular weight < 500 and a log Kow between 0 and 4 can be assumed to be well absorbed equivalently by the oral and inhalation route. Oral absorption may be reduced for acids and bases depending on their pKa value and their possibility of absorption in the GI tract. More lipophilic substances may be better absorbed in the GI tract due to the solubilisation with bile acids and thus oral absorption may be higher than inhalation absorption. The consideration of physico-chemical parameters should be performed before using default assumptions. Unless valid data suggest that inhalation leads to higher absorption than oral ingestion, equal absorption will be assumed when extrapolating from oral to inhalation route.

To convert an oral NOAEL into a dermal NAEL, the differences in absorption between routes as well as differences in dermal absorption between rats and humans have to be accounted for. The percutaneous absorption of capryl glucoside was measured in a human in vitro study (Zingerly, 2009). Considering that the in vitro test was conducted with a 10% dilution and skin levels (viable epidermis and dermis) were included as absorbed, assuming a dermal absorption of 0.01% displays a sufficient conservative approach.

Since short-term exposure scenarios will not be assessed, only long-term DNELs for workers and the general population are derived. The oral route is not relevant for workers. In addition it is assumed that only workers will come in contact with the neat substances. Due to the known irritating potential of undiluted APGs it is common to use personal protective equipment like gloves to avoid dermal contact therewith considering local DNELs as obsolete.

It is believed that use of default AFs recommended by ECHA will in most cases lead to DNELs that are not scientifically justified and consequently lead to unrealistically low levels for establishing safe use. Moreover, ECHA provides limited scientific justification for the default AFs; it is supposed they have relied on ‘standard approaches’ used by other organisations. In contrast, ECETOC based its AFs on a critical review of the literature consequently they should be used preferentially.

Furthermore, the overall analysis of comparable datasets on expert level distributions indicates that toxicologically supported factors for the time extrapolation of repeated dose studies might be lower than those proposed so far. Batke et al. (submitted) could show that chemical bias, dosage spacing and interspecies bias influence the outcome of a safety factor assessment and if these effects are removed by appropriate inclusion/exclusion criteria ("expert level") the resulting extrapolation factors decrease. They concluded that the accepted level of safety can be reached with lower assessment factors than previously thought (e.g. a factor of 1.5 for extrapolation from subacute to subchronic, 3.1 for subacute to chronic and 1.4 for subchronic to chronic study duration).

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

124 mg/m³

Most sensitive endpoint:

repeated dose toxicity

Overall assessment factor (AF):

7

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

357 000 mg/kg bw/day

Most sensitive endpoint:

repeated dose toxicity

Overall assessment factor (AF):

28

Modified dose descriptor starting point:

NOAEL

Hazard assessment conclusion:

DNEL (Derived No Effect Level)

Value:

35.7 mg/kg bw/day

Most sensitive endpoint:

repeated dose toxicity

Overall assessment factor (AF):

28

Modified dose descriptor starting point:

NOAEL

Correction of dose descriptors (route-to-route extrapolation), application of assessment factors (AFs) and derivation of the endpoint specific DNELs

Since there is no dose descriptor for every exposure route, dose descriptors were converted into a correct starting point by route-to-route extrapolation based on the ECHA guidance document "Guidance on information requirements and chemical safety assessment. Chapter R.8: Characterisation of dose [concentration]-response for human health", May 2008.

Within the ECHA Guidance a factor of 2 is suggested for the extrapolation from oral to inhalation absorption. On the contrary, the Technical guidance document on risk assessment in support of Commission directive 93/67/EEC, 2003 appendix IV A and B gives a number of physico-chemical properties that normally determine oral, inhalation and dermal absorption. These parameters include molecular weight, log Kow, pKa values and for inhalation also particle size distribution, vapour pressure etc.

Molecules with a molecular weight < 500 and a log Kow between 0 and 4 can be assumed to be well absorbed equivalently by the oral and inhalation route. Oral absorption may be reduced for acids and bases depending on their pKa value and their possibility of absorption in the GI tract. More lipophilic substances may be better absorbed in the GI tract due to the solubilisation with bile acids and thus oral absorption may be higher than inhalation absorption. The consideration of physico-chemical parameters should be performed before using default assumptions. Unless valid data suggest that inhalation leads to higher absorption than oral ingestion, equal absorption will be assumed when extrapolating from oral to inhalation route.

To convert an oral NOAEL into a dermal NAEL, the differences in absorption between routes as well as differences in dermal absorption between rats and humans have to be accounted for.The percutaneous absorption of capryl glucoside was measured in a humanin vitrostudy (Zingerly, 2009). Considering that thein vitrotest was conducted with a 10% dilution and skin levels (viable epidermis and dermis) were included as absorbed, assuming a dermal absorption of 0.01% displays a sufficient conservative approach.

Since short-term exposure scenarios will not be assessed, only long-term DNELs for workers and the general population are derived. The oral route is not relevant for workers. In addition it is assumed that only workers will come in contact with the neat substances. Due to the known irritating potential of undiluted APGs it is common to use personal protective equipment like gloves to avoid dermal contact therewith considering local DNELs as obsolete.

It is believed that use of default AFs recommended by ECHA will in most cases lead to DNELs that are not scientifically justified and consequently lead to unrealistically low levels for establishing safe use. Moreover, ECHA provides limited scientific justification for the default AFs; it is supposed they have relied on ‘standard approaches’ used by other organisations. In contrast, ECETOC based its AFs on a critical review of the literature consequently they should be used preferentially.

Furthermore, the overall analysis of comparable datasets on expert level distributions indicates that toxicologically supported factors for the time extrapolation of repeated dose studies might be lower than those proposed so far. Batkeet al.(submitted) could show that chemical bias, dosage spacing and interspecies bias influence the outcome of a safety factor assessment and if these effects are removed by appropriate inclusion/exclusion criteria ("expert level") the resulting extrapolation factors decrease. They concluded that the accepted level of safety can be reached with lower assessment factors than previously thought (e.g. a factor of 1.5 for extrapolation from subacute to subchronic, 3.1 for subacute to chronic and 1.4 for subchronic to chronic study duration).