Opinion of the Scientific Committee on Toxicity, Ecotoxicity and the Environment (CSTEE) on the results of the Environmental Risk Assessment of Pentabromodiphenyl ether [CAS N° 32534-81-9], carried out in the framework of Council Regulation (EEC) 793/93 on the evaluation and control of the risks of existing substances - Opinion expressed at the 13th CSTEE plenary meeting, Brussels, 4 February 2000.
Terms of
reference
The CSTEE
has been invited to examine the Risk
Assessment Report for Pentabromodiphenyl ether
and address the following issues:
1. Does the CSTEE agree with the
conclusions of the Risk Assessment Report?
2. If the CSTEE disagrees with such
conclusions, the CSTEE is invited to elaborate
the reasons for this divergence of opinion.
Introduction
Polybrominated diphenyl ethers are used as an
additive flame retardant mainly in polymers
and textiles. There are three groups of
products with an average of five, eight or ten
bromine atoms in the molecules. The products
containing compounds with an average of five
bromine atoms are called pentabromodiphenyl
ethers, even if there also are considerable
amounts of tetrabromo- and hexabromodiphenyl
ethers in these products. These substances
have been widely found in environmental
samples and a risk assessment was performed
under Regulation 793/93. The environmental
part of that is ready [CSTEE/99/13, Risk
assessment of diphenyl ether, pentabromo
derivative, (pentabromodiphenyl ether), Final
draft of August 1999] and has been examined by
the CSTEE. The outcome of this examination may
later be influenced by the content in the
report for human risk assessment when that is
available.
1.
General substance information
The
problem with the complexity of the product is
well handled and efforts have been made to
compare results for the product with the total
results of the components. In most cases these
results are in good agreement. The acronym
used in the document, PeBDPE, is long and the
more common PeBDE is easier to use. There is
no reference to the numbering system for the
congeners (the same as for the PCBs), which is
often used. This would make the text easier to
read.
2.
General information on exposure
PeBDE is
not produced within the EU (although in
section 3.1.4.4 samples from downstream an
PeBDE production plant in the UK are
mentioned) and the import data supplied by
industry indicate that <300 tonnes/year of the
product is imported. The major problem is to
estimate the amount imported in finished
articles. The total amount of PeBDE present in
articles is estimated to 1100 tonnes/year,
which seems to be rather low.
The major
use of PeBDE is in polyurethane foam and the
assessment assumes this to be the only use.
Some of the data on environmental
concentrations indicate that there may also
be, or at least have been, other uses, such as
in the textile industry. Emissions from
diffuse sources, including imported finished
articles, may also be of greater importance
than is obvious from the present report,
although there are presently very few data on
these emissions.
The
possible formation of polybrominated dibenzo-
p-dioxins
and dibenzofurans is discussed thoroughly in
Appendix 1 of the assessment. It is focussed
on thermal reactions and the conclusion is
that the major problems may occur in
production and recycling of polymers
containing PeBDE. In waste incineration it is
expected that problems with polychlorinated
dibenzo-
p-dioxins and dibenzofurans are
much more severe and that measures taken to
decrease those compounds will also reduce the
brominated counterparts. It is surprising that
there is no reference to the WHO/IPCS
Environmental Health Criteria 205 (Polybrominated
Dibenzo-
p-dioxins and Dibenzofurans) in
the report. That document contains most of the
information in Appendix 1 plus a lot more
(including effects assessments).
3.1.
Exposure assessment
In the
report the major emissions of PeBDE are
expected from production (not applicable in
the EU) and polyurethane foam production. The
emission of the substance from the finished
products is calculated with a formula derived
for migration of plasticisers from polymers.
The emission from the very porous polyurethane
foam may be higher, especially if it is
frequently squeezed in furniture. In addition
other sources may be important, especially as
PeBDE has been included in other materials
that are still in use.
The losses
of PeBDE from polyurethane foam in a landfill
are also addressed. It is considered unlikely
that significant amounts will leach, as the
substances would be expected to adsorb
strongly onto soils, and it is thus not
included in the PEC estimates. The
polyurethane foam may, however, be fragmented
to small particles that can follow the
leachate out from the landfill and contaminate
the recipient.
In
Appendix E different values of some
physical/chemical parameters have been used in
the EUSES model to investigate the influence
on the predicted results. This interesting
exercise is done for decabromodiphenyl ether,
but the same type of interesting outcome could
be expected for PeBDE. The change of water
solubility over three orders of magnitude, for
example, does hardly influence the local water
concentrations at all and the regional water
concentration only by a factor of 3. The logK
ow,
on the other hand, has a strong influence on
the predicted human intake.
The
determination of BCF has been scrutinised in
the report. A general problem with these very
lipophilic compounds is to get them in aqueous
solution and the assessors have been observant
on this and identified an experiment where
concentrations higher than the solubility have
been reported. An attempt to recalculate the
BCF values is done but, as is pointed out, the
new values will probably not represent the
true values either. The extremely low BCF for
2,2',4,4',5-PeBDE (BDE99) reported in the
original article is adjusted to a more
realistic value in the recalculation.
The SAR
data from the EPI program (Syracuse Research
Corporation) is reported on an isomer specific
basis in several places of the report. This is
not relevant, as these programs give identical
results for different isomers.
In the
calculation of PECs for the terrestrial
compartment the outcome (Table 3.19) indicates
a higher level for natural soil than for
agricultural soil. This must be rather
unlikely as the agricultural soil often is
treated with sewage sludge containing
considerable amounts of PeBDE. There are a few
unpublished results from an experimental field
study in southern Sweden where 0.3 - 0.4
microg PeBDE/kg soil ww were found in a field
treated with sewage sludge.
In
addition to the data presented on PeBDE in
sewage sludge samples there are some data
published from WWTPs in Stockholm (Sellström
U, Kierkegaard A, Alsberg T, Jonsson P,
Wahlberg C and de Wit C, Organohalogen
Compounds, 40 (1999) 383). The levels there
seem to be at least an order of magnitude
above the data presented in the PeBDE report
but this will not influence the conclusions in
it.
The
predicted levels of PeBDE in human food (Table
3.24) indicate 70 - 95 % of the intake to come
from root crops. The EUSES model is probably
not reliable in this prediction. This part of
the model was developed for substances with
logK
ow < 4.2 and only meant to
predict the concentrations in thin roots. In
an unpublished Swedish study the
concentrations of PeBDE in sugar beets were
compared with the concentration in the
surrounding soil. The results are comparable
with the corresponding data for PCBs.
PBDE congener |
Conc. in beet (ww)/conc.
in soil (ww) |
2,2',4,4'-TeBDE (BDE47) |
0.13 |
2,2',4,4',5-PeBDE (BDE99) |
0.06 |
2,2',4,4',6-PeBDE
(BDE100) |
0.07 |
These
findings will probably have more relevance for
the human risk assessment.
3.2.
Effects assessment
Toxicity
data available for PeBDE are scarce. For the
aquatic compartment, the minimum data
requirements for calculating a PNEC for water
(short term data on fish,
Daphnia and
algae and long term on
Daphnia) are
satisfied, even if data on algae are
controversial due to the possible adsorption
of the chemicals onto the algae.
No
information is reported on the toxicity of
PeBDE to sediment and soil dwelling organisms,
and therefore the equilibrium partitioning
method suggested in the TGD is used. The CSTEE
considers that for these chemicals, with very
low aquatic solubility and high persistence in
soil, the equilibrium partitioning method is
not adequate, even assuming the additional
uncertainty factor of 10 for the PEC/PNEC
ratio, suggested by the TGD. Only when enough
information on the toxicokinetics,
environmental fate and mechanisms of action of
the compound is available, can the equilibrium
partitioning method be useful when assessing
poorly soluble chemicals, and this is not the
case for PeBDE.
The report
also recommends conducting specific studies on
soil dwelling organisms instead of a
refinement of the risk for soil organisms
after the submission of additional information
on aquatic organisms, and this recommendation
is strongly supported by the CSTEE. The
proposed bioassays include three taxonomic
groups but no recommendation on the time frame
is presented. The CSTEE recommends clarifying
that the bioassays should cover both the acute
and chronic toxicity of the chemical (i.e.
chronic reproduction studies for soil
invertebrates, long-term relevant exposure
times and endpoints for plants and
micro-organisms).
The risk
for vertebrates (secondary poisoning) is also
addressed using the TGD protocol, which
includes a single food-chain step
(water->fish->fish-eating vertebrate or
soil->earthworm->earthworm-eating vertebrate).
The potential for food chain biomagnification
is supported by experimental data (i.e.
bioaccumulation from food exposures) and
monitoring data (higher concentrations in
species representing higher levels in the food
chain). The report presents these issues in a
clear and sound scientific way; however, this
information is not used in the final risk
characterisation, probably because these
aspects are not clearly identified in the TGD.
Even excluding this potential, an unacceptable
risk for vertebrates due to secondary
poisoning has been identified. Therefore, the
CSTEE supports the conclusion of the report
and considers that the risk for top predators
could be even higher than that estimated for
single secondary poisoning. This risk should
be taken into account either in risk
refinements or when proposing risk reduction
measures.
For
secondary poisoning, a PNEC was calculated by
applying a factor of 10 to the NOAEL of a 30
day repeated dose study on mammals. This
factor (instead of a factor of 100 as
suggested by the TGD) is justified by the fact
that the end point considered is the most
sensitive in a range of repeated dose studies.
Anyway, even with this low application factor,
a PEC/PNEC>1 has been calculated. It must be
underlined that mammalian data are not
reported in the document provided so far to
the CSTEE.
In the
environmental assessment report of PeBDE the
information on potential endocrine disrupting
potency (e.g. effects on the thyroid system)
and dioxin like activity (e.g. effects on EROD
induction) is limited. For a full risk
assessment of these compounds further
information on these activities are needed.
Receptor activity tests should therefore be
performed. A multigeneration test in rodents
is also important.
5.
Results
PEC/PNEC >
1 was obtained for water and sediment from
local sources and for both local and regional
sources for the terrestrial compartment. It is
mentioned that further toxicity testing is
needed to refine the PNEC for these
compartments. A refinement of the PEC is
probably more relevant, as the models used
often overestimate the exposure in lack of
measured data.
CSTEE
Conclusion
The risk
assessment of pentabromodiphenyl ether is well
performed and the CSTEE generally agrees with
the conclusions drawn, but has the following
comments:
-
There may
be several sources other than just the use
of PeBDE in the production of polyurethane
foam, especially from products in use and
from disposed products.
-
The
predicted concentrations in roots are
probably overestimations.
-
To verify
the PEC/PNEC > 1 both PEC and PNEC have to
be investigated.
-
In the risk
reduction, emissions from imported goods
should also be considered.
-
The
equilibrium partitioning method is not
adequate for persistent lipophilic
chemicals.
-
The
bioassays with soil organisms should cover
both acute and chronic toxicity.
-
Given the
high levels of PeBDE in the aquatic food
chain there is a need for toxicity testing
on (aquatic) birds, especially as there are
no data on toxicity in birds available.
-
The
potential for endocrine disruption and/or
dioxin-like effects should be further
investigated.
-
A
multigeneration test in rodents should be
performed.
-
The risk
for top predators may be even higher than
anticipated in the RAR as there is a
possibility for biomagnification in several
steps.
-
Some of the repetitions in the report could
be substituted with cross-references to
decrease the volume.