B2)
Is the rationale for not using hypertension as the critical effect justified and
objectively and transparently presented? Is this rationale correct?
B3)
Is the rationale for not using increased kidney weight justified and objectively
and transparently presented? Is this rationale correct?
C. Method of Analysis
Benchmark dose modeling has been used to derive the point of departure for determining
the proposed RfD.
C1)
Is there a suitable chemical-related dose-response relationship to justify
benchmark dose modeling of nephropathy? Is discussion of this effect
objectively and transparently presented?
C2)
Is the explanation for the choice of 5% extra risk as the benchmark response for
increased nephropathy transparently presented? Is the choice of 5% extra risk
scientifically justifiable?
D. Uncertainty Factors
A total uncertainty factor of 300 was applied to the point of departure: 10 for interspecies
differences, 10 for intraspecies variation, and 3 for deficiencies in the data base.
D1)
Are the choices of uncertainty factors transparently and objectively described?
D2)
Do the data support the use of different values than those proposed?
Scientific Comments from External Peer Review
A1) Is the NTP (1994) chronic animal study the most appropriate and scientifically
justifiable principal study for deriving the RfD? If not, what other study (or studies)
should be chosen and why?
A-5
Comment: All five reviewers agreed that the NTP (1994) animal study was the most appropriate
and scientifically justifiable principal study for deriving the RfD. Several reviewers commented
that, although the NTP study was the best available study, its ability to predict the effects of
chronic barium ingestion in humans was limited. A reviewer commented that there is some
uncertainty about whether the mouse is the most appropriate species for predicting the human
response to barium ingestion. Another reviewer noted that, absent any mechanistic data that
would indicate that the rat is a more appropriate model, the most sensitive species should be
used. This reviewer also remarked on the numerous strengths of the study, including the
excellent study design and methods, a chronic exposure duration, and quality dose-response data.
Another reviewer expressed concern that the NTP (1994) study did not report the dietary intake
of barium or account for it in the exposure estimates.
Response: EPA chose the NTP (1994) chronic animal study as the critical study for deriving the
barium RfD because the observed incidence of nephropathy in mice provided the best available
dose-response data. Recognizing there may be potential differences in the toxicodynamics and
toxicokinetics of barium between mice and humans, EPA has utilized a 10-fold interspecies
uncertainty factor in the derivation of the RfD. As one of the reviewers noted, mice were more
sensitive than rats to the nephrotoxic effects of barium. Treatment-related nephropathy was
observed in rats exposed for 13-weeks, but in the 2-year study the high incidence rate of
spontaneous lesions masked any treatment-related effect.
Barium was not listed as a contaminant of the NIH-07 rat and mouse ration in Appendix
L of the NTP Technical Report (1994), but it is not clear if NTP analyzed the feed for this
element. Barium serum levels in both rats and mice provided a biological measure of their
relative exposures. The dose-dependent increases in barium serum concentrations of treated
animals, and the significant difference between the treated and control groups, supports the
assumption that drinking water was the primary source of barium exposure.
A2) Is the explanation for why the human studies were not used as coprincipal
studies transparent and scientifically objective?
In the development of the existing RfD, hypertension was selected as a co-critical
effect. Evidence of hypertension was not observed in any of the principal studies. The
existing NOAEL is based on the highest exposure level in the human studies where no
hypertension was observed. In this case, was the selection of hypertension as the critical
effect and the derivation of the NOAEL scientifically objective and appropriate?
A-6
Comment: The reviewers unanimously agreed that the human studies (Brenniman and Levy,
1984; Wones et al., 1990) should not be used as coprincipal studies in the derivation of the RfD.
Two reviewers thought the rationale for not using the human studies was logical, transparent,
and objective. Two other reviewers suggested that EPA should elaborate on the limitations of
the human studies. It was not clear to one reviewer that the NOAEL from the human studies was
no longer a key data point used in the derivation of the RfD.
Two reviewers noted that it would be inappropriate to interpret the negative findings of
the human studies as evidence that barium has no effect on the cardiovascular system. One of
these reviewers recommended explicitly stating that the studies were inappropriate for evaluating
the effect of barium on blood pressure because of methodological limitations.
Only one reviewer addressed the second part of the question, which asked whether the
selection of hypertension as the co-critical effect for the existing (1998) RfD was scientifically
objective and appropriate. This reviewer stated that the authors and reviewers of the 1998
assessment were cognizant that an effect level for cardiovascular effects had not been defined.
Moreover, the decision at the time to base the RfD on a NOAEL was supported by a logical
rationale. However, he also stated that it was appropriate and scientifically justifiable for EPA to
refine the oral assessment and he endorsed the selection of nephropathy in mice as the critical
effect.
Response: There are methodological and design limitations associated with both the Brenniman
et al. (1981) and Wones et al. (1990) studies that limit the utility of their data. More importantly,
neither study provided sufficient data to support, nor refute, the hypothesis that chronic barium
exposure causes hypertension. It was not considered scientifically justifiable to base the RfD on
hypertension in the absence of dose-response data that support an association between chronic
barium exposure and this effect. As several of the reviewers noted, this does not mean that an
association between chronic barium exposure and hypertension has been ruled out, only that
there are insufficient data to draw a conclusion at this point in time. Additional text concerning
the limitations of the human studies and the rationale for selecting the animal data were added to
Sections 4.1 and 5.1 of this Toxicological Review.
A3) Are you aware of any other studies that may be relevant to the derivation of the
RfD?
A-7