Document OXNYwVeLkJx42kwkMmEzBqgp
.nul Institu e ot Environmental Ite m t Vaencoa
More Pieces in the
sensitive to dioxin at low doses. That obser vation, Lucier concedes, is "not inconsistent" with a threshold. But the bottom line, he and
Dioxin Puzzle
Portier argue, is that they found "lots of different dose-response curves for receptor-
mediated events"--which should put an end
Ten months ago it seemed there might he a "threshold"for
to the notion that just because dioxin binds to a receptor, there is ipso facto a threshold.
dioxin's effects; now comes evidence to the contrary
"That is their interpretation," retorts
Gallo, who has taken the lead in developing
EPA's new receptor-based model. Though
Almost halfway t h r o u g h a major re- receptor complex is "activated."
he calls Lucier and Fortier's
assessment of dioxin's toxicity, the infamous N ot so, say Lucier and
new study "super," he thinks
chemical is proving to be even more slippery Portier, who about a year ago
they go too far in their inter
than many scientists initially suspected.
sec out to answer the dioxin
pretation* At even lower
Less than a year ago, some of the natron's threshold question once and for
doses, he suspects, they would
leading dioxin experts, including Michael all. Their strategy, which they
have seen a threshold. Says
Gallo, associate dean tor research at the described in late September at
Gallo: "I am still convinced
Robert Wood Johnson Medical School in the Dioxin '91 meeting in Re
that a fixed number of recep
New Jersey, and Linda Bimbaum of the search Triangle Park, was to
tors have to be occupied be
Environmental Protection Agency's (EPA) focus on specific biochemical
fore we see a response in ani
Health Effects Research Laboratory in changes believed to occur along
mals." But Lucier and Po-
North Carolina, urged EPA to reassess the pathway to carcinogenesis
rtier's new data have made a
dioxin's risk. The scientists were spurred on and see how each one responds
convert out o f EPA's Bim
by new evidence about how the chemical to dioxin dose. Lucier concen
baum, who is playing a key
acts at the molecular level, which suggested trated on the biology; Portier
role in the dioxin rvalua
there might be a practical "threshold" or on the modeling.
tion. "We arc not sure a cer
"safe" dose below which no toxic effects Lucier's group began by
tain number of receptors have
would occur (Science, 8 February 1991, p. feeding rats varying but ex
to be activated anymore," she
624). If so, they said, then the linear model tremely low doses of dioxin ev
concedes, adding that the is
EPA uses to predict risk is probably off, ery 2 weeks. (The lowest was
sue "is turning out to be even
perhaps by a large margin.
100 picograms per kilogram per
In response, last May EPA administrator day.) At the end of 30 weeks,
George Lucier
more complicated than peo ple thought earlier."
William K. Reilly launched ah effort to build the researchers measured several simple, di The key task now, say Lucier and Portier,
a new, biologically based model
oxin-induced changes, such as is to determine which of the effects they have
for assessing dioxin's risk. Once
the activity of the cytochrome analyzed is the most representative of can
complete, this will be one o f the
p450 enzyme system, which cer--and thus most useful in reevaluating
first times EPA has replaced its
goes up in response to dioxin, dioxin's cancer risk. In the end, says Lucier,
generic cancer risk model--
as well as alterations in the "Mike and I may wind up in full agreement.
which assumes that risk in
binding capacity of several re It may be that the receptor-mediated events
creases linearly, starting with the
ceptors, including the epider important to cancer do exhibit a threshold."
slightest exposure--with a
mal growth factor receptor, Portier is now incorporating Lucier's ani
mechanistic one. But now, just
which signals cells to divide. mal findings, along with data from other
5 months into this rvaluation,
They also looked at more com groups studying dioxin's mechanisms, into
comes new evidence from
plex responses, such as cell pro a biologically based mathematical model
George Lucier and Chris Por
liferation and the number and similar to the one Gallo is working on with
tier of the National Institute of
size of precancerous lesions in EPA scientists. The goal of both exercises is
Environmental Health Sciences
the animals' livers.
to reconstruct each step in the pathway that
in North Carolina suggesting
The result? Contrary to the leads from dioxin exposure to cancer, using
that there is no threshold, at
predictions of just 10 months all existing data and filling in the holes with
least for some of dioxin's effects.
ago, there was "absolutely no their best guesses. Right now the two groups
The old linear model might
deviation from linearity" for are working independently but exchanging
not be that far offafter all. They
the simple responses they mea information. Birnbaum hopes to bring them
emphasize, however, that the
sured, says Lucier. In other together to see which model is best or
final verdict is not in yet. The linear model was first
Chris Portier
words, the effects increased in whether a "fusion approach" is needed. proportion to the dioxin doses Once the new model is complete--Reilly
called into question by evidence from a num measured in the rats' livers.
has called for it in May 1992--it should be
ber of labs that dioxin, unlike many carcino But this simple linear relationship may possible to get a more accurate fix on just
gens, must first bind to and activate a recep not hold for more complicated responses how risky dioxin really is. At this stage, no
tor before it can cause cancer or any of its (ike cell proliferation or cancer, Lucier and one is placing any bets, not even Gallo, who
other effects. The notion o f a practical Portier caution. Indeed, they are still ana just 10 months ago predicted that it would
"threshold" came in, explains Gallo, because lyzing their data for cell proliferation, but so turn out ro be far less risky than EPA now
it appeared that a minimum number of recep far it appears to have not a linear response but estimates. "I have no idea how it will come
tors must be occupied by dioxin before the a sublinear one--in other words, it is less out," he says now. L eslie Roberts
IS OCTOBER 1991
000400-
RESEARCH NEWS 377
O'JiLMb