OFFICE
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON. O.C, 204iO
July 30 , 1985
E&norable lee M. -Thomas
Administrator ' •
U.S. Environmental Protection Agency THE ADMINISTRATOR
401 M Street", S.W.
Washington, D.G. 20460
tear Mr, Ihomas;
On July 24 1 1984, the Environmental Health committee of the Science
Advisory Board reviewed a document, "Major Issues Associated with Health
Effects of Asbestos in Drinking Water {Carcinogenesis of Digested Asbestos
Fibers)," prepared by the Criteria and Standards Division in the Office
of Drinking Water. Ihe Coraniittee provided advice on this document in a
letter of fetober 29, 1984, which noted a pending study of asbestos
ingestion by experimental animals conducted by the National Toxicology
Program {NTP). A subsequent report from the Office of Drinking Water,
titled "Risk fron Ingestion of Fibers in Drinking Water," evaluates this
NTP bioassay. On May 22, 1985, the Ccmmittee met in public session to
review this report* Based on the additional information, it finds no
reason to change the conclusion of the October 29 , 1984, letter i^iich
states as follows:
"Given the positive signal seen in some epidemiolcgic studies, plus
well documented evidence for the association between asbestos fiber
inhalation and lung cancer, it is hard for the .'Committee to feel
comfortable in dismissing the possibility of an increased risk of
gastrointestinal cancer in humans exposed to asbestos fibers from
drinking water. However, the Committee consensus is that current
peez^reviewed evidence for humans and animals does not support the
view that asbestos ingested in water -causes organ-specific cancers."
A description of the NTP bioassay and its results is included in our
attached technical comments. -. The Cotrtftittee also notes that additional,
toxieological bioassays 'or epidemiolcgidal surveys are unlikely fco'con-'"" "
tribute more information to our understanding of the toxicity of ingested
asbestos fibers. Should EPA desire to resolve the current uncertainty,
the Comdttee recoraaends that the Agency sponsor research into the
mechanism(s) of action of asbestos.
Sincerely,
Richard A, Griesemer, D.V.M.,
Chair, Environmental Health Committee
. ,
Norton Nelson, Ph.D.
Chair, Executive Committee
A. James Barnes [A-1Q1]
Henry L. longest II [WH-556]
Assistant Administrators
-------
TECHNICAL COMMENTS CF THE EWIROtWENXAL HEALTH COMMITTEE ON MAJOR
ISSUES ASSOCIATED WITH BEAEIH EFFECTS OF ASBESTOS IN IRINKING WATER
, • (CARCINQGENESLS Of INVESTED ASBESTOS FIBERS)
On May 22, 1985, the Environmental Health Committee of the Sci-
ence Advisory Board met in public session to review a report, "Risk frcm
Ingestion of Fibers in Drinking Water," prepared by the Criteria and
Standards Division in the Office of Drinking water (QEW). The primary
purpose of this report was to update the Division's previous evaluation
of the effects of asbestos in drinking water based on a National Toxi-
cology Program (NTP) report that was not available at the time when the
first report was prepared. Following its review of the most recent Orw
evaluation, the Coisaittee finds no reason to change its earlier conclusion
reached on October 29, 1984.
The NIP report (no. 295) on the bioassay in the rat of chrysotile
fibers was one of a number of NTP reports (Nos. 246, 249, 277, 280 and 295),
which collectively represent an NTP investigation of the carcincgenicity
of asbestiform fibers for animals when administered in the diet for a
lifetime. Chrysotile and arnosite were administered to both F334 rats
and Syrian hamsters, whereas crocidolite and nonfibrous tremolite were
administered only to F344 rats. The experimental design was similar in
all six studies. The test substance was mixed with food and formed
into pellets containing 1% of the test substance, tfiwever, there was a
lack of information about fiber size (length in particular) in the test
substance after the pelleting procedure. For five of-the experiments,
two sizes of fibers were studied (described as small and intermediate in
range of fiber sizes). The group sizes for each sex, each fiber size
range and controls ranged frora 88 to 250 animals and were selected on
the basis of estimated ability to detect changes in,the known spontaneous
incidences of gastrointestinal tumors in the two species.
The bioassays were conducted in two laboratories, one for hamsters
and one for rats, at about the same time. Staggered experimental starts
were required but were nearly contemporary, and the teams of investigators
remained intact, ihey conducted the experiments uniformly frcm beginning
to end. The bioassays were conducted for the lifetime of the animals,
starting with the mothers of the test animals. In seme experiments (in-
cluding chrysotile asbestos in the rat) separate groups of 100 animals
received the test substance by gavage during the preweaning period.
Thus, the bioassays were not standard but were of greater duration with
Jarger numbers of.'animals than .the .usual.protocol, T
One dose level, 1% in the diet, was used in all treated groups, and
control animals received the diet without added materials. At death, the
entire gastrointestinal tract was opened a«3 examined, and all grossly
visible lesions were examined microscopically along with samples of
grossly normal tissues. All the major organs were studied according to
NTP protocols.
Overall, this was a well-designed series of qualitative experiments
in which the probability of detecting carcinogenesis induced by the three
asbestiform fibers and nonfibrous tremolite was reasonably high. Con-
trols could have been added in which nonfibrous mineral, of similar chemi-
cal composition was fed to compare with fibrous mineral. Since the re-
results are not statistically significant, however, a control for fiber
alone was less important.
-------
-2-
only suggestive evidence of carcinogenicity was the finding of
3.6% (9/250) adenomatous polyps in the large intestines of male rats fed
1% chrysotile asbestos of intermediate size range. NO polyps were found
in 88 matched control animals, to rats given chrysotile by gavage in
the preweaning period, 2% (2/100) had polyps in the large intestine.
Similar lesions were found in 0.6% (3/S24) pooled control rats fron all
of the NTP asbestos^experiments. Multiple polyps tended to occur in the
affected rats. .Short fiber length chrysotile asbestos showed no evidence
of polyps. In experiments with other forms of asbestos, occasional
polyps also were found, but these lesions were of comparable incidence
to those in the control rats.
Ihe weight of the evidence for carcinogenicity of chrysotile by the
dietary route is slight at best, a marginal increase in the incidence of
one type of benign neoplasm, at one site, in one sex, of one species, in
one of two size ranges of test material, at one dose level. The Ctumittee
interprets this evidence as "equivocal." That is, **dle other known
effects of chrysotile asbestos may suggest carcinogenic potential by
this route of administration, the data do not support a cause and effect
relationship. A confirming experiment in both sexes would be needed.
Frcart the overall evidence available, however, the Committee does not be-
lieve that additional animal bioassays would resolve the issue. Instead,
equivalent resources devoted to the mechanisms of asbestos action are
more likely to provide crucial evidence bearing on the hazard.
------- |