>I was wondering what the general consensus is on the new EPA cancer
>guidelines Margin of Exposure non-linear default approach? Do you find it
>lacking in detail and rather uninformative?
Is the document below what you are discussing? If so, then yes it is pretty
nonspecific about lots of things. But then again, such is the state of the
art in risk assessment. It is a tall order to predict adverse effects of
chemicals from sparse, erroneous, or nonexistent data. Perhaps you could
give us your views on nonlinear dose-response assessment and margin of
This document was obtained from http://www.epa.gov/science1/ehc0297.htm
Environmental Health Committee
and the FIFRA Scientific Advisory Panel (SAP)
U.S. ENVIRONMENTAL PROTECTION AGENCY
SCIENCE ADVISORY BOARD (SAB)
MINUTES OF MEETING: February 13-14, 1997
PURPOSE: The main purpose of the meeting was to discuss and review EPA's
Proposed Guidelines for Carcinogen Risk
LOCATION: The meeting took place at the Holiday Inn Georgetown Hotel,
Washington DC 20007 and was announced in
the Federal Register for January 23, 1997 (Vol. 62, No. 15, page 3510)
PARTICIPANTS: SAB Members, Consultants, Federal Experts, and Staff: Dr.
Emil Pfitzer, Chair, Drs. Charles Capen,
Adolfo Correa (Day 2 only), Kenny Crump, Ernest McConnell, Frederica Perera
(Day 1 only), Henry Pitot (Day 1 only),
James Swenberg, Mark Utell (Day 1 only), and Lauren Zeise; Mr. Samuel
Rondberg, Designated Federal Official (DFO) and
Executive Secretary to the EHC, and Ms. Mary Winston, Staff Support
Secretary to the Committee. The Committee Roster
(incorporated as Attachment B) lists the EHC attendees and their
affiliations. A "Sign In" sheet listing other attendees is
incorporated as Attachment C.
Six members of the public had registered to make statements to the Meeting
per the invitation in the Federal Register and are
identified in the Agenda (Attachment D). During the course of the meeting,
several other members of the public did take the
opportunity to make brief statements or comments as part of the informal
discussions involving Committee Members and/or
Dr. Ellen Sibergeld, who had planned to attend the meeting as a member of
the Committee (and who appears on the
previously printed Agenda) could not attend; Dr. Silbergeld did address the
group briefly during the first day of the meeting
the issue of departures from default values -- the issue to which she had
earlier been assigned as lead discussant.
SUMMARY: The meeting followed the Agenda with some minor rearrangements and
deviations in timing to accommodate
The meeting opened at 9:00 AM with brief administrative comments by the
Chair and by the DFO. Each Member of the
Committee then presented statements as to their affiliations and any
previous involvement in issues related to topics to be
discussed or to any parties with an interest in the outcome of the review.
No one reported any associations or activities which
were, or which appeared to be, conflicts under current policies in so far
as the current meeting was concerned.
Dr. Arnold Kuzmack, EPA/OW briefed the Committee on the proposed
Carcinogenicity Risk Assessment Guidelines,
providing a brief history of their development, and addressing the
principle issues arising in the public comment process: Dr.
Kuzmack's slides are incorporated as Attachment E.
Following brief dialogue with Dr. Kuzmack, six members of the public
addressed the Committee: Dr. James Wilson,
Resources for the Future (slides incorporated as Attachment F); Dr. Matthew
Bogdanffy, Chemical Manufacturers
Association (slides incorporated as Attachment G); Mr. Steve Milloy,
Environmental Policy Analysis Network (slides
incorporated as Attachment H); Dr. Michael Gough, Cato Institute (slides
incorporated as Attachment I); Mr. William G.
Kelly, Jr., Federal Focus, Inc. (slides incorporated as Attachment J); and
Dr. M. Jane Teta, American Industrial Health
Council (slides incorporated as Attachment K).
In addition, the Committee was addressed by an invited speaker, Dr. Mel
Anderson. Dr. Anderson reported on the findings
an an Expert Panel convened by the International Life Sciences Institute to
review the draft Guidelines. His slides are
incorporated as Attachment L.
The Committee then turned to discussion of the specific issues embodied in
the Charge for the meeting. The entire
proceedings of the Committee are recorded in a transcript, incorporated as
Attachment M. A brief summary of the
Committee's major findings on the Charge issues follows.
a) General: There was general support within the Committee for the proposed
guidelines. The EHC found that the
Guidelines were a significant step forward for the Agency, but will
recommend some rewording in several areas.
b) Hazard Classification: EPA had proposed 3 broad
descriptors--"known/likely", "cannot be determined" and "not
likely." The EHC supported the importance of having a narrative to provide
the evidence for carcinogenicity. However, it was
not possible to reach a consensus in the meeting for a list of descriptor
terms that would provide a heading for the narrative. It
was clear that the "known/likely" descriptor was meant to be two different
descriptors, but several members felt that the
descriptor "likely" was not a good term. There was some support for the 8
descriptors proposed by Ashby, et al., which is
basically an expanded IARC classification. Concern was expressed that the
EPA's list of approximately 11
descriptors/sub-descriptors was too detailed. The discussion ranged from an
expression of using only a narrative without
descriptors to an expression of narrative with a list of 3 to 8
descriptors. There was no consensus at the meeting on what the
descriptors should be.
c) Defaults/Mode of Action: The EHC was in agreement that deviations from
default assumptions must be based on
strong scientific evidence. There was extensive discussion about the
problem of defining the "strength" of scientific evidence,
with agreement that additional guidance was important. Specifically, there
were expressions of concern that, without
additional guidance, mode-of-action as a basis for deviation from default
would be misused. It was requested that the
guidelines re-emphasize that when there is uncertainty the decision should
favor the protection of the public's health. There
were expressions of concern about the use of the term "default." Suggested
alternatives were "in the absence of information
use the most conservative option" or "inference guideline"; a term
suggested back in 1983.
d) Dose-Response: EPA suggested dividing the dose-response curve into an
observed range where data exist, and a range
of extrapolation where there were no data -- using the lower bound on the
10% cancer incidence as a point of departure for
extrapolation to lower doses. Cancer precursor data (e.g., hyperplasia
incidence, hormone levels) might also be used. EPA
also proposed biologically based dose-response as the preferred
extrapolation technique, with linear and nonlinear defaults.
The Committee supported the division of the dose-response range and the
proposal to use the lower bound on the 10% tumor
incidence. There was not uniform endorsement of the use of precursor
information for quantitative assessment however,
although it could be used qualitatively.
e) Margin of Exposure Analysis: The EHC discussed the Agency's proposal to
use a margin of exposure analysis for
nonlinear dose response assessment. It was felt that this topic needed the
addition of case studies to better illustrate its use,
and there was no consensus on the advice to be given to a risk assessor on
the magnitude of an appropriate MoE. The EHC
also requested further wording on how the Agency would use peer review in
the process of deciding alternative approaches.
f) Human Data: The EHC understood that dose-response data will often be
deficient in studies with humans. It was agreed
that wherever possible information on dose-response should be used, but
that decisions would necessarily be on a
case-by-case basis. It was noted that it will often be important to have
access to raw data for the best utilization of human data
for risk assessment, and that EPA should request this. The EHC agreed that,
because of apparent misunderstandings, it was
important to clarify that statistical significance would be used in judging
human data. However, it was considered that
statistical significance should not be a strict criterion for causality,
but that it should be coupled with reasoned judgment. The
Bradford-Hill recommendations for judging chance should used, but as
guidelines not as criteria.
g) NRC Recommendations for Tumor Data Analysis: The EHC considered that
neither the NRC approach nor the
Agency's proposed approach should be used without modifications. It was
agreed that the summing of tumor types was not
scientifically appropriate. However, the occurrence of tumors at different
sites should be recognized in the weight of evidence
approach. The decision to sum potencies for independent tumor types should
be limited to a case-by-case decision. Since
summing of tumor types or summing of animals with tumors can often "swamp"
the critical factors for risk assessment, it
was considered that this should not be a routine practice.
h) The EHC agreed that the Agency should take susceptibility factors into
account when they are known. In particular
concern for children and individuals with polymorphisms should be evaluated
as potential populations at risk. There was
considerable discussion regarding a potential generalized uncertainty
factor for linear extrapolations to human populations;
there was no consensus for a specific number.
Following a brief discussion of report preparation, Dr. Pfitzer adjourned
the meeting at 1:06 pm.
Mr. Samuel Rondberg
Designated Federal Official
Dr. Emil Pfitzer