[CONVENE 8:14 A.M.]
R.
HARRISON: Good morning, everyone. I’m Bob Harrison. Welcome to the meeting of the Ranch Hand Advisory Committee. As usual, we’ll start off with
introductions. And first of all, we’ll
go around the table and then we’ll start over on that end and just go down each
row just zigzagging back and forth. And
I guess first, Joel will be first. So
Joel, why don’t you introduce yourself?
J.
MICHALEK: Joel Michalek, Principal
Investigator of the Air Force Health Study.
L.
SCHECHTMAN: Excuse me. If everyone at the table would please use
the microphones; simply push the talk button so that your red light lights up
and then we — everybody in the room will be able to hear you. Thank you.
J.
MICHALEK: All right. I’m Joel Michalek, Principal Investigator of
the Ranch Hand Study or the Air Force Health Study.
M.
STOTO: I’m Mike Stoto, a member of
the Committee. I work at the RAND
Corporation.
P.
CAMACHO: Paul — I’m Paul Camacho, a
member of the Committee. I’m at
UMS-Boston.
L.
SCHECHTMAN: I’m Leonard
Schechtman. I’m with FDA’s National
Center for Toxicological Research, Associate Deputy Director of Washington
Operations and the Executive Secretary of the Ranch Hand Advisory Committee.
R.
HARRISON: I’m Robert Harrison. I’m Professor of Medicine Emeritus,
University of Rochester in Rochester, New York.
M.
GOUGH: I’m Michael Gough. I’m a member of the Committee. I’m retired from various places.
S.
LEFFINGWELL: Sandy Leffingwell,
physician, new kid on the block.
N.
RIVERA: Use your mike, sir.
S.
LEFFINGWELL: Sandy Leffingwell, a
new kid on the block. I’m a physician.
J.
ROBINSON: I’m Julie Robinson. I’m the Branch Chief for the Air Force
Health Study.
R.
TREWYN: Ron Trewyn, Kansas State
University, a principal troublemaker.
K.
OSEI: Kwami Osei, Professor of
Medicine, Director of Division of Endocrinology at Ohio State — Buckeyes.
R.
HARRISON: Kwami, hit the button
again and turn it off.
L.
SCHECHTMAN: Okay. Just a little bit of housekeeping; there
will ...
R.
HARRISON: Are we going to introduce
the ...
L.
SCHECHTMAN: Sorry?
R.
HARRISON: Are we going to introduce
the people in the audience?
L.
SCHECHTMAN: Oh yes, let’s finish
the introductions.
D.
BUTLER: My name is David
Butler. I’m a Senior Program Officer in
the National Academy of Sciences, Institute of Medicine. I direct environmental health studies, including
studies related to the Air Force Health Study.
D.
JOHNSON: I’m Dave Johnson. I’m the Executive Medical Director for the
Division of Environmental Health at the Florida Department of Health.
R.
HARRISON: At what?
D.
JOHNSON: Florida Department of
Health.
R.
HARRISON: Florida Department of
Health. Thank you.
K.
FOX: Colonel Karen Fox with the Air
Force.
J.
MINER: I’m Jay Miner, Operational
Technologies contractor. I work in
Program Management for the Air Force.
In my former life, I had Colonel Robinson’s job.
E.
HASSOUN: Ezdihar Hassoun. I’m the Associate Professor of Toxicology
for the University of Toledo College of Pharmacy.
M.
PAVUK: I’m Marian Pavuk; I’m with
SpecPro. I work with Joel Michalek on
the Ranch Hand Study.
R.
HARRISON: And we all know who you
are.
K.
CAMPBELL: I’m Kim Campbell. I’m the Management Specialist for the Ranch
Hand Advisory Committee.
R.
TREWYN: That’s who we make our
complaints to.
M.
MONTGOMERY: Lieutenant Margaret
Montgomery. I’m the Operations Manager
for the Ranch Hand Study.
M.
BLANCAS: Manny Blancas with
UDTech. I’m a contractor working for
Program Management on the Ranch Hand Study.
W.
MURRAY: Bill Murray. I’m with ANSER, representing the Air Force.
E.
SKILLEN: I’m Elizabeth Skillen. I’m with HHS, home base ATSDR.
R.
HARRISON: Let’s go ahead to the
rear — right rear there.
[INAUDIBLE]: I’m a pharmacy student.
[INAUDIBLE]: I’m a pharmacy student.
M.
OWENS: I’m Maurice Owens. I’m with Science Applications International
Corporation. I’m a Program Manager for
the Ranch Hand Study.
W.
GRUBBS: I’m Bill Grubbs and I’m
also with SAIC.
M.
YEAGER: I’m Meghan Yeager. I’m also with SAIC.
R.
HARRISON: Thank you very much.
L.
SCHECHTMAN: Thank you all and
welcome. A couple of things: we have an attendance sheet, sign-up sheet
for your — to acknowledge your presence that’ll be — that’s being passed around
the room. We would like you to please
fill out the information and sign that for us.
Most of you are used to seeing Barbara Jewell as part of these and every
meeting historically ever held relative to the Ranch Hand Group. Her good fortune and our misfortune is that
she has just retired and we will surely and sorely miss her.
Replacing
Barbara is our new Management Specialist who just introduced herself; that's
Kim Campbell. So you will see her name
coming through on lots and lots of e-mail messages that are both informative
and are designed to hound you so that you can get back to us the information
that we need from each and every one of you on a regular basis. So we welcome Kim to this effort and we know
that although no one could ever fill Barbara Jewell's shoes, Kim will do us
good service and she will be an excellent, excellent replacement. Go ahead.
R.
HARRISON: I’d like to make one
suggestion, alteration to the agenda.
As some of you may have noticed in your e-mail, the National Academy of
Sciences has been asked to look into the — certain issues concerning the Ranch
Hand Study. And Dr. Butler, who
introduced himself earlier, is here to discuss the status of that request and
the status of the National Academy of Sciences decision-making process. And I thought that it would be better to put
him before Joel begins his report. And
if it’s agreeable to the Committee, we’ll let Mr. Ogershok do the Program
Management update and then immediately after that, ask Dr. Butler to give his
presentation. Okay? So for right now then, Mr. Ogershok.
M.
STOTO: Are we going to do the
minutes?
R.
HARRISON: What?
M.
STOTO: Are we going to discuss the
minutes?
R.
HARRISON: Sorry. Hold on just a minute, Jay. Mike has caught me in another procedural
issue.
J.
MINER: Mine is going to be really
short.
L.
SCHECHTMAN: Okay.
Just to inform the Committee, the meeting minutes of November of last
year that were reviewed, edited and ultimately have been approved by Dr.
Harrison by his signature had gone through the Committee; changes and edits
were recommended. Those minutes were taken
back with the recommendations by Dr. Harrison, updated, and signed off and
approved — just to let you know that that took place; that was the November
2001 meeting minutes. Now as far as the
meeting minutes from the previous meeting, those have been circulated.
M.
STOTO: Is it really — is it really
2001 or is it 2002?
L.
SCHECHTMAN: 2002, sorry.
M.
STOTO: Okay.
L.
SCHECHTMAN: The — well, wait a
second.
M.
STOTO: It says “2001" in the
draft meeting of March 2003.
L.
SCHECHTMAN: We will confirm
that. We’ll confirm that. I’m pretty sure since it was the end of the
2001 year.
M.
STOTO: Okay.
L.
SCHECHTMAN: I think that's correct
actually — 2001. Yeah.
R.
HARRISON: I think we had a fallow
year there.
L.
SCHECHTMAN: And then we had the
March 2003 meeting. Those minutes were
generated, circulated to the Ranch Hand Committee members for their comments,
edits and suggestions. And we’re at a
point now where we’re welcoming that input and seeking approval for those
minutes.
R.
HARRISON: Any other questions,
Mike?
M.
STOTO: Yeah. I mean, I tried to look through those
minutes for the March 2003 meeting and found them very difficult to
understand. And because so much time
has passed since the meeting, I found them difficult to correct. I mean, at one point it quotes me saying
something that doesn’t make any sense.
And I suppose it’s possible I did say something that didn’t make any
sense, but it’s more likely that the minutes didn’t reflect what I said and I
can't remember what I said.
So
I guess rather than talk about this particular draft, I’d like to raise the
issue of how can we do a better job of getting these minutes done
accurate? And I think that one thing is
that we need to do them on a more timely basis because I really can't do my job
as a Committee member of correcting these minutes because ten months have
passed and I — and my memory is not that good.
So I think that one thing I'd like to discuss is try to find a way that
we can get something to look at in a more timely way. I’m not sure.
R.
HARRISON: And I think the start to
the answer to that would be the explanation for the time interval between the
meeting and getting the draft minutes to us.
Len, can you ...
L.
SCHECHTMAN: Well, there was
turnover in our office and the people who are responsible for the minutes early
on have since left Committee activities.
So we were doing what we could to get those minutes generated as quickly
as we could — unfortunately not quick enough — prior to this particular
meeting. What we were left with was
working from the verbatim transcript of the previous meeting and any
information that appears in the current minutes were generated from the
verbatim transcript as it was written into the transcript per the discussion
around the table.
Any
inaccuracies in those minutes may or may not be attributable to the content of
the verbatim transcript. And we would
be willing to sit down with any individuals, any members of the Committee who
had problems, such as those described by Dr. Stoto, and go over those matters
along with the public record, the verbatim transcript, and make sure that
despite the fact that the verbatim transcript may not have given or — the
minutes may not have transmitted the information accurately as to what the
thoughts were of the speaker that the verbatim transcript was, as I said, what
was used as the basis for the minutes so that we can make those corrections.
R.
HARRISON: I guess, so essentially
what you’re saying is that this was an unusual delay? But I wonder if you could give us an estimate of when we — for
instance, when will we see the minutes?
When will you all see the minutes for this meeting? And because what you’re kind of saying that
whoever did the minutes wasn't there and so the — to make any kind of judgment
about the verbatim transcript. The
other question I wondered is, is the — are the — are parts of the minutes key
to parts of the verbatim transcript?
So
if Mike has a question about something that he purportedly said, he shouldn’t
have to go through reams of pages of transcript to get to it. He ought to know that it’s on page 123 at
least. So those would be the two
questions I would have. What do you —
what is the time interval going to be and is there a way to key the minutes to
the transcript so that — so that when a question arises, the Committee member
can get directly to that point?
L.
SCHECHTMAN: Okay. I guess what I can do is answer the
questions in reverse order. We can work
with the transcriber to see if indeed there is a way to do what you request
about keying the minutes to the verbatim transcript; that’s never been done in
the past. We would look into that
possibility. But as far as the minutes
as they are generated in the order in which the information appears in the
minutes, that is the exact sequence of events that are transcribed in the
verbatim transcript. So that even if we
weren’t able to tie it page by page to the minutes, which I think is probably
highly unlikely, that we could go back in order with the verbatim transcript
and the minutes side by side and be able to hone in on a given area of the
verbatim transcript to match it against the minutes.
M.
STOTO: I think we should separate
two issues: one is how to fix the
minutes for the March 2003 meeting versus how to improve the process so that
we’re not in this situation in the future.
I mean, I — one of — one of the things that I don’t like about the way —
about this form of minutes is that it is chronological. And it really doesn’t do a good job of
describing what happened at the meeting to say “and then we broke for lunch” or
“and then somebody came in and did this” in the middle of a scientific
discussion, although that may be the way it transpired.
I
mean, I would like to see minutes that summarize the essence of what happened
at the meeting and are not just a condensed version of the transcription, which
I think that this is now. So I think
that I would like — obviously, this is open for discussion — but I personally
would like to see minutes that were more oriented, organized by subject, and
key decisions and so on rather than a summary of the — second thing, I think
that this is — it’s really important to get these minutes right.
And
it probably is worth an investment of someone like a writer with a background
in science who can listen to the discussion, prepare a draft on a timely basis;
someone, you know, whose job it is to do that and to get that draft back to the
Committee and to the Ranch Hand staff in a timely basis so that corrections can
be made while everyone remembers what was — what the issues were.
R.
HARRISON: Okay. Any other comments? Mike?
M.
GOUGH: Well, I want to echo what
Mike just said because we’ve had this discussion about minutes before. And the difficulty of waiting so long is
that people's memories fade and people's memories fade about different
things. So then instead of getting
five, or six or seven people agreeing on what was said, we often get one person
who remembers it — saying what was said.
Now when I do that, it’s perfectly right. But I mean, I think everybody does it as honestly as they
can. But I agree, it’s got to be done
more timely and also the structure of the minutes is just really clumsy.
R.
HARRISON: Anyone else? Yes?
S.
LEFFINGWELL: Leffingwell. Is it reasonable to assume that we’re supplied
the transcripts and the minutes in an electronic format as well as paper?
R.
HARRISON: That’s very unreasonable
to assume.
S.
LEFFINGWELL: Oh dear. Okay.
If we had the electronic version of the verbatim transcript, then we can
search on key terms and find things like Dr. Stoto referred to more easily.
R.
HARRISON: There are a number of
programs that allow you to search by key word that would take the thousands of
words that we’re going to generate today and make any particular sequence of
three or four of them findable. So ...
M.
STOTO: Well, we could ...
R.
HARRISON: I'm sorry, Mike. I interrupted Paul and he was going to say
something next.
P.
CAMACHO: It was just about the
technology. It’s the — do we have the
voice technology that can just make text out it? I doubt it; I don’t think we have that.
R.
HARRISON: I know for sure that’s
not feasible. Mike?
M.
STOTO: I don’t think it’s a
technology issue. I mean, honest. I mean, in fact, we did get the draft
minutes by e-mail and I don’t know whether it’s Word or pdf. But that's easy enough to do and, you know,
you can search there.
P.
CAMACHO: It doesn’t sound easy.
M.
STOTO: And I would search for
“Stoto” to see what was attributed to me.
But I think that the fundamental thing is that the person who drafts the
minutes needs to understand more about what the issues are.
P.
CAMACHO: Yeah.
M.
STOTO: And that's not going to
happen by fancy text ...
P.
CAMACHO: Well, this ...
M.
STOTO: ... transcribing and so on.
P.
CAMACHO: But this way — but this
way you have both. If you get a — if we
get a — if you had a technology
transfer, then you wouldn’t have an abbreviated version and you wouldn’t be
going through ...
R.
HARRISON: Let me — but I think the
point ...
M.
STOTO: I mean, I don’t — I want an
abbreviated version; I don’t want a transcript. I want something that’s — you know, minutes are supposed to
summarize the key points, not everything that was said.
R.
HARRISON: The point though, Mike,
that I think you’re implying but not saying is that the problem here is a
matter of resources; that instead of a sort of unknowledgeable person
transcribing this — not transcribing it, but converting it in the minutes —
that the Committee leadership needs to invest in a little more money and get a
little more pop for the — for its nickel.
M.
STOTO: Yes.
R.
HARRISON: Okay. Anyone else on the subject of minutes?
L.
SCHECHTMAN: Okay. As far as some of the information that was
exchanged just now, we did hire a new transcriber. I don’t know that you introduced yourself?
N.
RIVERA: No.
L.
SCHECHTMAN: Would you, please?
N.
RIVERA: Sorry. I’m Nadine Rivera. And Dr. Stoto, I am a writer/editor, a scientific writer/editor,
so I know exactly what you’re talking about.
L.
SCHECHTMAN: And which is ...
R.
HARRISON: Can you — can you — can
you tell us exactly where you live because Mike was ...
L.
SCHECHTMAN: ... which is exactly
the reason why we hired Nadine to do this job because of her background as a
scientific writer as well as a transcriber so that we would hopefully, and
having predicted your comments, we’ll be able to fulfill those requests.
M.
STOTO: Okay. Thank you.
R.
HARRISON: Dr. Stoto is beaming,
which means all is well and we can proceed with — and I could tell when he
started walking up that this wasn’t Mr. Ogershok who was going to do this
management thing. Go ahead, Jay.
J.
MINER: Mr. Ogershok was not able to be here at the
meeting and — but does send his greetings.
The Human Assistance Program Office pretty much — on Program Management,
there’s not a whole lot going on right now with contractual actions. I think the only one that we have pending of
interest is an action to put a contract out to UC-Davis to continue some
adipose tissue studies and glucose transport mechanisms. We did receive our budget authority and
unlike some of the rest of the federal government, we’re not under a continuing
resolution, so that’s a good thing.
And
as far as contract deliverables for our current contract, Science Applications
International Corporation is doing a great job: delivering chapters on time; obtaining Air Force comments;
incorporating, resolving and then getting the draft chapters back out to us. And then hopefully, that process will then
flow to you. You will receive initial
draft chapters, Air Force comments, comments as resolved, and then the final
draft chapter, which Lieutenant Colonel Robinson will be sending to the
Committee for distribution. Other than
that, that’s it.
N.
RIVERA: What’s your name, sir?
J.
MINER: Miner, M-I-N-E-R; first name
Judson.
R.
HARRISON: Thank you very much,
Jay. Len has just thought of something
else that he needs to tell us about before Dr. Butler gives his talk.
L.
SCHECHTMAN: Okay.
Actually, we may need the extra few minutes because one of the agenda
items was a committee management report by myself. And I wanted to give the Committee an update on just what had
transpired between oh, the fall of 2003 and the current situation in terms of
Committee member appointments. And some
of this information I’d like to read into the record just so that everyone will
be aware as to the steps that took place both from our end and the feedback we
got back from the Department of Health and Human Services.
So
in October of 2003 — October 14th — we sent a waiver request to the
Office of the Secretary of HHS requesting a special exception to Department
policy on Committee reappointment. And
it read, “This is to request a special, but essential exception to Department
policy limiting the term of service of advisory committee members. The committee in question is the Advisory
Committee on Special Studies Relating to the Possible Long-Term Health Effects
of Phenoxy Herbicides and Contaminants, a/k/a the Ranch Hand Advisory
Committee, which has been operating since 1981.
The
Committee advises the Secretary and the Assistant Secretary for Health
concerning its oversight of the conduct of the Ranch Hand Study by the Air
Force and other studies in which the Secretary or the Assistant Secretary of
Health believes involvement by the Committee is desirable. When the original charter was established on
January 19th, 1981, the function statement specifically stated that
‘members shall be invited to serve for the duration of the Committee.’ Thus, granting an exception to allow members
of the Ranch Hand Committee to continue to serve clearly meets the intent of
this element of the original Ranch Hand Committee charter under which the Committee
was established.
This
Ranch Hand Study is currently scheduled to conclude on September 30th,
2006. The fifth and final round of
physical examinations of the study participants has just been completed and the
medical data are now being compiled. A
review of those study results by the Ranch Hand Committee is currently
scheduled to begin in March of 2004.
Initially, there will be six chapters plus a glossary that will be ready
for review at that time. An additional
twelve chapters plus a discussion of future directions are currently projected
to be completed by September 2004.
The
draft final Ranch Hand Study report is also projected to be completed in the
last quarter of fiscal year 2004. These
will all subsequently be reviewed by the Ranch Hand Committee over the
following one to two years. Since the
Ranch Hand Study itself is anticipated to conclude September 30th,
2006, the Ranch Hand Advisory Committee activities are expected to conclude by
the end of 2006. Despite the fact that
the study is reaching completion, the Ranch Hand Advisory Committee is faced
with the possibility of having to replace four existing Committee members.
Of
the nine positions constituting the Ranch Hand Advisory Committee, one former
vacancy has been filled and two are in the process of being filled. As new members, none will have previously
participated in a Ranch Hand Committee meeting, four in the Ranch Hand Study
review process. Five of the remaining
seven experienced Ranch Hand Advisory Committee members are to be rotated off
the Committee in accordance with current Departmental policies unless the
necessary exceptions can be imposed.
Thus,
of the total of nine members of the Ranch Hand Advisory Committee, seven of
them, 80 percent, would be new replacements charged with the responsibility of
seeing the study through to its completion.
Considering the expertise and long-standing involvement of the current
Ranch Hand Advisory Committee members with the Ranch Hand study itself, the
study participants and the study director, and their familiarity with the
biological specimens, types of data generated, medical histories, data
interpretation, reporting process, et cetera, replacement of existing
Committee members with new individuals unfamiliar with the majority of these
elements would only serve to impose an extreme hardship on the work of the
current Ranch Hand Advisory Committee members and disrupt the efficiency of the
Ranch Hand Advisory Committee activities.
As
stated previously, it is also noteworthy that the function statement of the
original Ranch Hand Advisory Committee charter states ‘members shall be invited
to serve for the duration of the Committee.’
Thus, maintaining the existing membership of the Ranch Hand Advisory
Committee in accordance with the language of its original charter is in compliance
with that Committee charter. Obviously,
the removal of the existing knowledgeable and experienced Ranch Hand Committee
members whose terms are about to expire at a point in time when the study is
coming to an end would introduce an avoidable burden on the activities and
productivity of the Ranch Hand Advisory Committee.
In
fact, the entire life span of the Ranch Hand Study has led up to this
culmination point for the final analysis of study results. It is for these reasons that it will be of
critical importance to maintain the current makeup of the Ranch Hand Advisory
Committee. Retention of the existing
knowledge and experience of the present Ranch Hand Advisory Committee members
who have followed the study and analyzed the data over the years would preserve
the continuity of the Committee efforts and thereby avoid any retardation or
interruption of the Ranch Hand Advisory Committee productivity.
With
the present Ranch Hand Advisory Committee members, it would take new
inexperienced replacements the remaining years of the study to achieve a
comparable level of familiarity with the medical data and Committee processes
in order to provide the necessary expertise so essential to the completion of
the Ranch Hand Study analysis. It is
even conceivable that were 80 percent — that is, seven of nine of the existing
membership of the Ranch Hand Advisory Committee — to be replaced at this time,
the work of the Committee may not be completed by the time the study concludes
and resources are no longer available.
In
view of these factors, the National Center for Toxicological Research wishes to
reappoint” and we listed Drs. Camacho, Osei, Sills, Gough and Harrison. “Reappointment of these members would
require a waiver from Department policies which prohibit service of more than
four continuous years on an advisory committee without a one-year break in
service. We are requesting that they be
reappointed for an additional two-year term, which will take them through the end
of the study. The additional service
would commence February 1st, 2004 and end on January 31, 2006.
All
of these members have expressed a willingness to continue their service on the
Committee. In addition” — there is no
addition. Sorry. And that was signed by the Director of the
National Center for Toxicological Research as generated by our Washington
Office. So that letter to the Office of
the Secretary went out on the 14th of October 2003 as I
mentioned. On the 20th of
that month, we received word that permission to revert back to the original
charter and to put the language in the original charter that states that
members may remain with the Committee until the termination of 2006 had been
approved.
Two
days later, October 22nd, we received word from the Department that
they would not change the charter. As
such, we were forced to go with the language in the original charter and
nominations were forwarded to HHS on November 20th after all
attempts had failed in obtaining the waiver approval. I then sent an e-mail dated November 19th to all of
the standing Committee members, which I won’t take the time to read into the
record unless I’m requested to do so, informing all of the members of the Ranch
Hand Advisory Committee as to the state of affairs and who — and those members
who would no longer be allowed to serve on the Committee as a result of the
Department’s rules and their unwillingness to accept the waiver request. Are there any questions? Yes?
M.
STOTO: Did they formally turn the
waiver request down or did they just not respond? Sorry.
L.
SCHECHTMAN: No, it was a formal rejection of the waiver.
R.
TREWYN: I thought when you were
describing it though, you said that the — that you were — then had to go with
the original language and the original language was the term. And I assume it meant that it was the
existing regulations that you had to go with.
L.
SCHECHTMAN: I’m sorry; that’s correct.
R.
HARRISON: So are we here or not
here?
R.
TREWYN: You’re not here.
R.
HARRISON: We’re not here; this is
it. Okay. Free at last, free at last, thank God Almighty.
L.
SCHECHTMAN: Basically what it boils down to, if I can
just for purposes of clarification, Dr. Harrison and Dr. Gough, who have served
on this Committee forever essentially ...
M.
GOUGH: No, I had a break.
L.
SCHECHTMAN: You did?
You had one break? And the
duration of your service, as I recall, was about twelve years? I’m sorry.
M.
GOUGH: Ten years and six months.
L.
SCHECHTMAN: Ten years and six months — I stand
corrected. And Dr. Harrison’s term of
service was fourteen years and four months with no break in service. Okay.
So because of their extended service, despite the quality of their
service, the Department decided that they had served the purposes of the Ranch
Hand Advisory Committee for as long as they possibly could be allowed to.
M.
STOTO: And there were — and there
were others though that you mentioned.
L.
SCHECHTMAN: Right.
Everyone else ...
M.
STOTO: Everyone else was
reappointed?
L.
SCHECHTMAN: ... was reappointed because the duration of
their service hadn’t exceeded the approvable limit by the Department.
R.
TREWYN: So was the limits placed by
the Department — I’m just curious — were those just sort of ignored over the
years? I mean, obviously for fourteen
years-four months, ten years-six months, that's a fairly long period of time.
L.
SCHECHTMAN: No, they were not ignored. Requests were put in to have them — their
memberships renewed. Those were
approved over the years and it just reached a point in time where they said “enough”
unfortunately.
R.
HARRISON: So we’re — so we’re
finished then. Well, this is — this is
basically just an informational item just telling you what’s happened. I don’t think that there’s any reason to
spend a lot of time on it. I have a
plane to catch in just about an hour.
Do you have something else to say?
P.
CAMACHO: Well, now it’s on? Well, doesn’t this — how do we — who’s
taking your place in a bottom line?
L.
SCHECHTMAN: Okay.
That was taken care of by the Department. Dr. Stoto was requested because his long-standing — but not quite
as long as them and experience on the Committee — he was requested to serve as
the new chair of the Committee and he has accepted that position.
M.
STOTO: Thank you. Has it been approved by the Secretary yet or
...
L.
SCHECHTMAN: To the best of my knowledge, it has.
M.
STOTO: Oh, okay.
L.
SCHECHTMAN: If it hasn’t, we'll let you know.
M.
STOTO: Okay. Well, thank you. I didn’t know that and I want to just, at this point, make sure
we give tribute to the two members who are — who are leaving now for their
service.
L.
SCHECHTMAN: Absolutely; I agree. Dr. Harrison will serve on today’s — at
today’s meeting as will Dr. Gough as a member and as chair: Dr. Harrison as chair and Dr. Gough as
member of the Committee. Their terms
don’t expire until the end of the month, which is why we — one of the primary
reasons for having this meeting prior to their leaving, departing from the
Committee because we indeed wanted to give him — give them the recognition that
they so truly deserve for all of their efforts over the years.
I
will say this: we will thank you both
verbally at this time. We’ve been
pressing the Department for letters of recognition and plaques for each of you
for your service; we know those will come.
But sometimes things, as you know in the government, do lag behind the
desired schedule. So verbally though,
we want to thank the both of you for exceptional efforts on the Ranch Hand
Advisory Committee. Your terms of
office, despite the length of office, were efforts that the Committee could not
have done without.
And
I know I speak for the rest of the members of the Committee as well as the rest
of the individuals who participate in our meetings in acknowledging your
valuable input, your efforts, and the time and energy that you’ve invested in
the Ranch Hand Advisory Committee over the years. And I would welcome anyone around the table to offer any comments
in addition to those at this time.
R.
TREWYN: Wing of the building going
to be named after them or something here?
Let’s make this appropriate, folks.
R.
HARRISON: Well, I would just first
of all like to — like to say that, and I think that most of you all know this,
that Mike Stoto came to this Committee after serving on — in the National
Academy of Sciences and brings a store of knowledge about this area that I
suspect rivals anyone in this room. So
it’s — number one, it’s comfortable to — comforting to see him become chair and
particularly because I would also like to say that as someone who views most
things governmental and the Armed Forces with a high degree of cynicism, that
I’ve been impressed over the years that I’ve served on this Committee about the
concern that the Ranch Hand Study personnel have had on the Air Force side for
making this a good study and actually finding out what’s going on. And it’s nice to know that there are at
least some parts of our government that actually work the way you expect them
to and that’s probably the part that I’ll miss most of all out of this. Okay, schmaltzy stuff aside. Let’s — no.
M.
GOUGH: I've been called upon to say
something.
N.
RIVERA: Your light needs to be on.
R.
TREWYN: Your light is very dim.
M.
GOUGH: It’s a reflection of other
things. Well, you know, I — yesterday
when I got the e-mail with the attachment of Chapter 4 ...
N.
RIVERA: It’s not on, sir.
M.
GOUGH: I’m sorry. Yesterday, when I got the e-mail with
Chapter 4 attached, and I looked at it and I thought, “Well, I’m off the
Committee, so I don’t need to print this one out and read it” — all 1,100 —
1,700 pages. Is that what it’s going to
be?
J.
MICHALEK: Approximately.
M.
GOUGH: So for those remaining on
the Committee, enjoy yourselves. I have
a thought: I think that — I think that
this has been a really good relationship between a research group that’s doing
a very difficult and very important study and an advisory panel that’s
generally been a good advisory panel.
So I think we did well. I’m sure
you’ll continue to do well; you’re going to need a lot of manpower. When I'm corrected about political
correctness by Bob Harrison, indeed. I
already have a great big plaque from the last time I went off the Committee.
R.
HARRISON: Two plaques should
provide some balance and everything.
Okay. Any other comments, or questions or statements? Very well.
I think we’re finally to Dr. Butler and the Vietnam Veterans Health
Studies at the National Academies. Dr.
Butler?
D.
BUTLER: Thank you.
I’m going to take just a moment here to call up my particular set of
slides if people will indulge me until then.
Dr. Michalek was kind enough to let me borrow his system here. My name is David Butler and I am a Senior
Program Officer with the Institute of Medicine of the National Academies. I wanted to start out for just a moment and
explain what our institution is and what it isn't. Importantly, what it isn’t is a part of the federal
government. We were chartered by the Congress
back during President Lincoln's administration to be an independent non-governmental
institution whose job it was to advise the government on various science
subjects.
Our
first bit of advice actually was to the military who asked the original members
of the National Academy of Sciences to advise them on how to build compasses
that would work on iron clads — a very important issue of the day. The Institute of Medicine itself is a
relatively new part of the National Academies.
We were chartered in 1970 as a part of the institution whose specific
mission it was to look at medicine and public health issues. National Academies studies follow a common
outline.
Importantly,
for this — for what I’m about to talk to you about — all studies must start
with the Governing Board’s Executive Council reviewing and approving study
conduct. The research that I’m going to
be speaking with you about later has been submitted to our Governing Board, but
the Governing Board hasn’t yet acted on it.
So what I’ll be talking to you about is speculative at this point,
although I have every expectation that the Governing Board will approve this
study to go forward.
Presuming
it does, what we will do is to form a committee of volunteer experts who will
then research the topics that we've been asked to look at and write a report
based on literature reviews, public hearings, and workshops, and written
testimony and other information. In
other words, we are an information gathering body. And we will need a lot of input on this study and we’ll certainly
be looking for it. When we’ve completed
the report process, our final reports are subject to an external peer review
before they are made public.
The
National Academies and the Institute of Medicine were asked to do a number of
controversial pieces of research over the past couple of years. I’ve picked out a few hot button topics just
to cite that we have a rather broad mandate for looking into tough science
questions in advising the federal government.
Our studies of Vietnam veterans health, of which there are several, were
mandated originally under the Agent Orange Act of 1991. That Act instructed the Secretary of the
Department of Veterans Affairs to enter into a contract with us to conduct
research.
This
set of studies evaluate both the statistical association between wartime
exposure to herbicides and health impacts; that’s a task for which we use
information from the Air Force Health Study, and numerous other research
studies, and also a separate project where we looked into the means and methods
of retrospectively evaluating herbicide exposure of veterans. There have been a number of reports that
have been issued from us over the years.
There’s a complete listing of them on the back page of the slides
handout that I gave you.
The
two most recent are the Veterans and Agent Orange: Update 2002 report and a report that closed out the second
section of work that I just mentioned on the Characterization of Exposure of
Veterans to Agent Orange. All of
this information and information on our other studies, including those of
Persian Gulf veterans and veterans of other wars, are on a web site that just
went live in December of this year and now serves as sort of a central point
for getting information on all of our veteran studies.
More
specifically, the reason that I was invited to come here today is the Veterans
Benefits Act of 2003. That Act was
signed about a month ago into law and contains a section requesting that the
Secretary of Veterans Affairs enter into a contract with the National Academies
to conduct a study on the disposition of the Ranch Hand Study. That legislation contains five separate
elements that were to be addressed in the study. The first of these is to evaluate the scientific merit of
retaining and maintaining the medical records and all of the other pieces of
information that Dr. Michalek and his staff have gathered over the years past
the scheduled termination date of the study.
The
second primary element is to evaluate all of the issues related to retaining
and maintaining those medical records, including importantly privacy concerns
of the veterans themselves, and of course, the issues like informed
consent. The third is the advisability
of providing some independent oversight of these records and other data, and if
independent oversight is deemed appropriate, a mechanism for providing such
oversight. So the first three elements
of this are all related to retention and maintenance of samples and records,
the advisability and the mechanism that would underlie it.
Element
four comes to a separate topic: the
advisability of extending the study and some evaluation of its potential value
and relevance, its cost and the entity that's best suited to continue the study
if it is extended. And finally, the
advisability of making the laboratory specimens available for independent research;
that is, as I understand it, research outside that's strictly specified by the
Air Force Health Study, and the value and relevance of such research and its
cost.
So
in brief what we’ve been asked to do is take a comprehensive look at this huge
longitudinal study, the information that’s been gathered in it over the years,
and to offer advice on how best to manage that information, and whether to
extend the study as it presently stands.
Just briefly, in the course of this study, we will be posting information
on our progress. Information on that
study and on any other study that the National Academies conducts can be found
on our current projects web site, which can be reached through the web page
that I've listed here.
And
as I indicated before, all of our reports are up on the web for open access of
anyone who might like to read them through the National Academies Press. Because we haven’t yet signed a contract for
this study, it hasn’t yet been approved by our Governing Board, there isn't a
whole lot I can tell you about the details of the conduct of the study. I have every expectation that the study will
be approved by our Governing Board; a contract will be signed.
Given
that that's the case, we would expect that we would hold public meetings of
this committee; that we would gather information from numerous sources,
including this Committee, including veterans, the Department of Veterans
Affairs, veteran service organizations, other individuals who wanted to have
input on the questions that have been outlined. And at this point, I'm in an information gathering mode. I’m looking to learn from the Committee
today, looking to simply inform people that we are looking for information and
for input on this study. And I’m
supposed to mention that I’m looking forward to this because it sounds like a
fascinating study and I’m eager to see how it turns out. Dr. Gough?
R.
HARRISON: Thank you. Any questions?
M.
GOUGH: Am I on?
R.
HARRISON: You’re on.
M.
GOUGH: Dave, do you have a — once
it’s approved, it looks like a relatively tight timetable. And what is the nature of this report? Is it going to be advice to the Secretary of
the VA that this is what we would consider doing with study — with the Ranch
Hand Study?
D.
BUTLER: The Secretary has been
advised, has been told to enter into a contract with us to do this research,
but the advice is not specifically to the Secretary of Veterans Affairs. It is general advice on what should be done
with the study. The timetable is still
being worked out. You’re right; that
legislation has specified a rather tight timetable for the conduct of the
study.
M.
GOUGH: David and I used to work
there. We know the VA doesn’t give
enough time for projects. What do you
think — what do you think realistically, once it’s — once you get approval from
the Governing Council to do it, how long do you think it’ll take?
D.
BUTLER: That’s up to the contract
that we make with the Department of Veterans Affairs and we haven’t finalized
that yet. Clearly in the best of all
worlds, Mike, we would like to have a little bit more time to think about
this. There are some pretty heavy
issues ...
M.
GOUGH: Oh yeah.
D.
BUTLER: ... that have been outlined
here and we want to make sure that a thorough and thoughtful job is done of
it. We'll be talking with the
Department of Veterans Affairs on that and we’ll see, once we sign a contract
with them, what timetable they feel comfortable with given the Congressional
mandate.
M.
GOUGH: And I have one other — one
specific question. Is the National
Institute of Aging aware of this and do they know about the study? I mean, this is the biggest, longest
longitudinal study of guys ever. And it
has a — and it has a disadvantage there are no women in it, but on the other
hand, it does have a racial mix, which is unusual for a study like this. And it seems to me that they — there might
be interest in other agencies in the government, NIA for instance, in
participating in the discussion of what’s going to happen with these samples and
whether additional information is going to be collected.
D.
BUTLER: I think you make an
excellent point. In talking about who
we would want to have on the committee, we thought that certainly
representation from a gerontologist would be appropriate given the makeup of
this population. I haven’t specifically
thought about the National Institute of Aging, but now that you mention it, I
certainly will bring it up with them and see if we can ask them for some advice
on this because I agree; this is — this is a fairly large population. They've been studied, followed for such a
long time.
In
the Update 2000 report — I don’t have it in front of me — I know that
the committee who wrote that report specifically pointed out that
notwithstanding the value of the study from a Vietnam veteran’s health point of
view, that that committee thought that there might be other valid reasons for
following — continuing to follow this population given the information that's
already been collected.
R.
HARRISON: Yes sir?
R.
TREWYN: Since I know that the
committee will reach out looking for members, I would just suggest that, you
know, there are a couple of geriatric members of this Committee that will be
stepping down soon and would have a wealth of knowledge that could be, I think,
invaluable. And at least looking at
them as a resource, if not as members, might be something to think about.
D.
BUTLER: We’re certainly happy to
take recommendations for individuals who should be considered for membership
and whether or not people make it on to the committee, as I said, we’ll be
holding public workshops for gathering information.
R.
HARRISON: Mike?
M.
STOTO: First of all, I want to say
I think this is great news that the Congress asked for this study and the
Academies seems willing to do it. I
mean, this is something that’s really very important and I’m glad that we —
that David was able to come talk about it and even happier that they — looks
like they're going to do it. I also
want to second the point that Mike made about the value — the potential value
of this to beyond Ranch Hand and even veterans health specifically. And so I’m glad to see that item number five
is in the — is in the charge and I think that NIA is a good one.
I
suspect that there are other parts of NIH that might be of interest as well; I
mean, the environmental side and potentially others as well. I mean, given the findings on diabetes, it
suggests that there would be some interest conceivably there. And the point that I think we've tried to
make in the past bears repeating here; is that if the scientific community, and
particularly a funding agency like NIH knows about the resource of the
materials that are available, they’re in a position to provide funding for
studies through normal NIH funding channels; that a combination of their
funding and the Air Force’s resources are a potential gold mine. And I think that point — that connection
needs to be made.
Now
there are issues about whether the data can be used for purposes like that,
which will also have to be addressed as well.
But, you know, it’s not that we’re asking the Air Force to put more
money into — or the VA to put more money into doing more studies on this. It’s that the resource that’s available
would be useful for scientists funded through other means and that point needs
to be explored and clarified.
D.
BUTLER: Assuming the interest of
the study participants in allowing this information to be used in other forms,
I agree that there is tremendous potential to learn from this information.
R.
HARRISON: Before you say something,
Kwame, can I — can I make what I think is a point? And it’s a human subjects issue that I’m not sure I quite
understand. But I know that the
subjects that we spoke with, the Ranch Hand participants that we spoke with out
in California, would be quite willing to agree to allow their samples and so
forth to be used in other studies. It’s
just not clear to me that that’s — that you can do it post-hoc. I think the biggest sticky wicket in this
whole thing is how to deal with the protection of human subjects issues that
comes up from trying to retrospectively obtain agreement to utilize materials
and samples that you already have which has a coercive value that’s very
difficult to understand. That’s —
okay. Mike and then Kwame.
M.
STOTO: On that point, first of all,
I agree with you; that is, I think, the most difficult issue that needs to be
addressed. And the other point though
is that I believe that — and I'll ask Joel now — is that in the last round, the
study participants were asked about their willingness for the data to be used
in the future for Ranch Hand purposes — or I guess Agent Orange purposes — and
for other health purposes and it’ll be interesting to know what the results
were. Can you tell us?
J.
MICHALEK: I brought those; those
slides are on my machine. I think I
showed them at the March meeting and I can show them again when I stand up, the
percent who complied, so I have those numbers with me. They were a very high percentage — over 90 percent.
K.
OSEI: Yeah, David, is this going to
be open to other private institutions to compete because, you know, the
National Institute may not be able to do all these? The ideas may have to come from outside and the question comes
in: would you, you know, is the
National Institute going to allow, say Ohio State, or any other institutions to
apply to participate rather than individual, you know, groups?
D.
BUTLER: Well, in the conduct of
this particular study ...
K.
OSEI: Okay.
D.
BUTLER: ... we will be — we will be
doing this particular study. As to the
disposition of the materials in the future, one possibility — just talking
hypothetically — is that there could be a competitive process for determining management
of the — of the samples and data, but that simply exists as a possible
option. Our present concern with this
study is this is something that where we will be conducting this work and
generating the report. The National
Academies does not in general contract — subcontract out pieces of studies that
they’ve been requested to conduct by the Congress.
K.
OSEI: But there may be other — the
reason I'm bringing this up, for example, if you look at the Physician Health
Study where we have all these males who are in their 70s who are comparable in
age, so that may be, if you look at the health of, you know, people who have
never been exposed to, you know, Agent Orange, asbestos Agent Orange, and all
the variables that can come in. And
that will have to come from outside for you to be able to actually look at a
control group that is comparable to what we are looking at.
R.
HARRISON: Kwame, it sounds to me
like you're talking about two different things. David and the National Academies is being asked by Congress — I
mean, as things presently stand, in another year or so, Joel and his crew are
going to take all these Revcos, and computers and optical drives, put them in a
big hole somewhere and pour concrete on top of them, and the whole thing is
going to be done. And Congress is
asking is there something better we can do than that?
And
what you’re asking is once the — you’re assuming that the decision is going to
say “Yes, let’s use it for other stuff.”
And then you’re talking about what gets done after that; those are two
separate items. And that second item,
the one you’re asking about, is something I don’t think the National Academies
is going to do anything about other than say that, you know, maybe the NIH
should provide some funds or someone else should provide some funds. Am I — am I understanding you correctly,
Kwame?
K.
OSEI: I think what I’m talking
about ...
R.
HARRISON: I think what happens
after the samples are available ...
K.
OSEI: Right.
R.
HARRISON: ... for other use and
that’s not what David is here to talk about.
D.
BUTLER: Right. In general, the National Academies does not
conduct primary research.
R.
HARRISON: Correct.
D.
BUTLER: And we are not a funding
organization and don’t have any money to support others’ primary research.
P.
CAMACHO: So your recommendations
would be confined to the value of keeping the data, not particularly managing
and the management structure of the data:
the whole piece, access? Because
he’s talking about who’s going to get access to the data ...
K.
OSEI: Exactly.
P.
CAMACHO: ... assuming it was around
somewhere. So ...
D.
BUTLER: The committee might choose
to offer advice on the process by which such decisions would be made, but not
the details.
P.
CAMACHO: There’s no end result, but
rather approach limits on what they ought to be doing?
D.
BUTLER: That’s correct.
M.
STOTO: And the — and the Academies
hasn’t been asked to actually manage it.
P.
CAMACHO: That’s — yeah; that’s what
I’m getting at.
M.
STOTO: We have a distinction
between recommendations about and managing.
R.
HARRISON: Precisely. What we’ve talked about over the last
several meetings is there’s this treasure trove of data and biological samples
that if we follow the road map as it’s presently drawn out, there’s this —
there’s this end right at the edge of this very deep crevice, and everything
just drives over the edge and drops out.
And NAS is being asked, “Is this the right thing to do?” That’s all they’re being — my understanding
of it is that’s all that they’re being asked.
D.
BUTLER: That’s correct.
R.
HARRISON: And I know you’re aware
of this, but I just feel like it ought to be said anyway; and that is, there
isn’t a Revco good enough, or Forma or whichever to keep biological samples
indefinitely. And I sure hope that the
committee emphasizes that not only if the committee does decide to recommend
that other things be done with this material, that it’s not only important, but
that as far as the biological samples are concerned, it’s desirable that it be
done quickly rather than at a snail’s pace.
D.
BUTLER: Well, we — I — we would
seek to include, either on the committee or in the workshops, expertise on the
long-term management and maintenance of biologic samples. And to get back to your earlier point, also
to include individuals who have experience in bioethics, treatment of study
subjects and materials from study subjects.
R.
HARRISON: Kwame?
K.
OSEI: So from what you’re saying,
this decision has to be made before September when the — September 30th
of 2006, right, because that's the end of this, you know, whole Ranch Hand
Study? So you have to move very quickly
because this is 2004.
D.
BUTLER: Yes.
K.
OSEI: And so, you know ...
D.
BUTLER: No. We're very aware of the time-lines here and
...
K.
OSEI: Okay. All right.
M.
STOTO: I mean, and that the
Congress has asked that it be done in 120 days. I think that we should thank
Dr. Butler for coming and, you know, express our willingness as a Committee to
interact with the committee that the Academies sets up and bring the expertise
that we’ve developed to bear on that.
R.
HARRISON: I agree.
Thank you very much. I wish that
I had the power to award you a plaque as an honor. But it — I think I speak for everyone who's participated in this
process and participated in the discussion about the disposition of the Ranch
Hand materials; that we’ve all felt that this was a truly important set of
materials to — even if — even if you all decide that nothing should be done
with it, I’d be much more comfortable that there’s a sound scientific
discussion and a reasoned decision made rather than just things dropping
out. And I’m happy to see that and I
hope that the decision is positive, but we very much appreciate that it’s being
looked into.
D.
BUTLER: Well, the National Academies
Update 2000 — Veterans and Agent Orange: Update 2000 — report pretty much echoes what you’re
saying. That report does not reach any
conclusions on this, but simply identified a need for there to be a reasoned
discussion and evaluation of what ought to be done on it. And, you know, speaking from the standpoint
of the committees, I’m sure they’re pleased that that recommendation is now
being followed up.
R.
HARRISON: Anyone else? Thank you again. Not
included in the agenda for today is time for public comments. And what I would like to suggest, and would
hear disagreements if there are any, is that we’re going to still try and
finish the morning session — everything in the morning session. We’re about twenty minutes behind now, but
Joel is an expert at compressing his — at data compression and I’d like to
suggest that we do public comments immediately after lunch — 1:00. And I would like to ask anyone who plans to
make a public comment to give your name and affiliation to Ms. Campbell, who
was introduced earlier, so that I’ll be able to introduce you. But if someone has not introduced — has not
given Ms. Campbell the information, but wishes to make a public comment, you
will certainly be able to do so.
P.
CAMACHO: You may have people show
up at the — before then. Why don’t you
leave it that they can make the public comment to her; that they — for the
record or do you want to ...
R.
HARRISON: No, public comment is public. I mean, they get to say ...
P.
CAMACHO: Well, I’m not saying that
they don’t have to — well, if you — they ...
R.
HARRISON: We have one more item
after ...
P.
CAMACHO: What if they come in at
3:00?
R.
HARRISON: We’ll be gone.
P.
CAMACHO: What if they come in at
2:00?
R.
HARRISON: If they come in at 2:00
and it turns out that they wish to make a comment ...
P.
CAMACHO: Fine.
R.
HARRISON: ... we’ll discuss
it. But I’m assuming that most of the
people who might be interested in making public comments are already here or
will be here before lunchtime; that’s the way it usually works. But no, I’m not going to shut anyone off if
— I’m not going to shut anybody off period.
I’m just trying to organize it so it flows a little better.
P.
CAMACHO: Thank you.
R.
HARRISON: Any other questions or
comments? Okay. So let’s get on now to ...
M.
STOTO: Could we maybe take a break
for a minute before we get started?
R.
HARRISON: Okay, Dr.; the bathroom
is around that corner. Five minutes.
[BREAK 9:31 A.M. - 9:46 A.M.]
R.
TREWYN: At your advanced age, this is five minutes? Is that right, Bob?
R.
HARRISON: Okay.
Well, just wanted to make sure Mike had plenty of time to accomplish his
mission. Okay. So we now have recent findings: Dr. Michalek and Committee discussion. Joel?
R.
TREWYN: You have twelve minutes.
Serum Dioxin and Four Biochemical
Parameters of Adipose Tissue
J.
MICHALEK: All right.
This is the first of several presentations. The activity is to understand at the cellular level whether
dioxin is implicated in the metabolic pathways leading to diabetes mellitus
type 2. This study was launched several
years ago. It’s based on data collected
at the cycle 5 or 1997 physical examination.
The primary scientists involved:
Phillip Fujiyoshi and Fumio Matsumura are the toxicologists; I am the
statistician. Haejo Hwang was a — one of the toxicologists involved with
actually performing the measurements, but he was replaced by Phillip.
Remember
that I am a statistician; I am not a biologist. I am not going to pretend to talk about mechanisms. I do not have the mechanisms memorized. I will just display the data for you from my
point of view and you can interpret as you please. To cite some information at the top, right off the top before we
start any slides, you should know that this is a completely new molecular
epidemiologic approach that has never been done before — ever. It is unprecedented.
This
is the first time the method of preliminary — polymerase chain reaction
measurements have been made in human tissues for the public health issue of importance
that we have here today in front of us.
Measuring the expression of metabolic intermediaries in the diabetic
pathway through gene expression is — was considered the best way to determine
whether there really is a causal pathway between dioxin and diabetes. The criteria to establish exactly what
should be measured were written by Dr. Matsumura in his initial proposal to the
Air Force.
The
first is that the measurements must be — must have been proven in animal
experiments to have been related to TCDD and diabetes. Every parameter in the study must be
expressed in enough quantities in human adipose tissues to be measured in the
first place. Every parameter must be
clearly recognizable from other parameters.
They must be amenable to standardization. They should be quantitatable and the laboratory methods should be
reliable enough to be carried out by reasonably trained people. All of those criteria are met by this study.
I
hope that in future presentations, either Phillip or Fumio will be here to tell
you what all of these data mean. We are
— we are attempting to determine whether dioxin is adversely associated with
one or more of the components of the biochemical pathway leading to diabetes. The study was launched prior to cycle 5,
which was the 1997 physical, based on data collected at cycle 4, which was the
1992 physical. We envisioned that about
50 percent — well, first of all, we did not have the resources to collect
adipose tissue specimens on all 2,000 study subjects.
Not
only did we have — not have the resources, it was not even logistically
possible to have done that because you would have to have a surgical procedure
done, namely liposuction, on each of thirty people — or sixty people every week
and there weren’t enough resources at Scripps Clinics to handle that. So based on the logistical and fiscal
limitations of the study, we — and based on power calculations that I — that I
constructed along with Fumio Matsumura, based on animal studies, we concluded
that approximately 300 study subjects would be sufficient to establish
statistical power that was reasonable.
We
expected 50 percent compliance due to a variety of reasons: refusals, medical deferments and other such
things. So we prepared a roster by a
random — a random selection process and a factorial design from the cycle 4
data of 650 subjects who were then entered onto a roster and were flagged at
the cycle 5 exam to be invited to provide twelve grams of adipose tissue by
liposuction. The selection factors in
the factorial design were grouped:
obesity, age and diabetic status.
And all of these components were prescribed before-hand in a study
protocol that was reviewed by the Committee and by us, and was approved and
funded with the University of California at Davis.
Of
the 650, 313 actually completed — volunteered and completed the adipose tissue
liposuction very close to our expected 50 percent compliance. The four biochemical parameters that were
actually measured are glucose transporter 4, C-SRC, NFκβ and
C/EBPα. And these parameters were
selected through an iterative procedure by Dr. Matsumura based on feasibility
and state-of-the-art measurement techniques in current research. And there you see a description of what
these are and why they’re important.
Here’s
a breakdown of what happened to those 650 individuals: 10 percent — 10.2 percent actually refused;
16 percent roughly were deemed too lean by the surgeon to draw any fat at all
and so that's good news for some of those people. I had enough fat to have been drawn on the practice run of this
procedure and it was an interesting process.
So there you see what happened.
The final bottom line was that 313 or 48.2 percent of the original
roster completed the procedure.
And
here’s a final sample size layout of the factorial. In the columns, you see the obesity determination — lean and
obese — and on the far right is the total.
And there you see the breakout by age.
The cut point for age was the median in the — at the baseline exam — 42
years at baseline. And diabetic status
was based on our old diabetic determination before the new ADA criteria and
I’ll talk about that on the next slide.
The older criteria was simply that the individual had to have an
elevated two-hour post-random glucose or a physician's diagnosis of diabetes.
M.
GOUGH: Joel?
J.
MICHALEK: Yes?
M.
GOUGH: When you made this cut at
1942 for birth date ...
J.
MICHALEK: No, age. Let’s go back up. Let me see what that cut was; I forgot it already.
M.
GOUGH: Yeah, it was — it was the
...
R.
HARRISON: It said “1942.”
J.
MICHALEK: It did? Okay.
That was the median birth year of individuals at baseline.
M.
GOUGH: Did that result — does that
result in there being more officers and enlisted men in the old group?
J.
MICHALEK: Yes, of course, because
officers are older than enlisted; that’s true.
M.
GOUGH: So that’s a — that’s a real
complication then because of socioeconomic?
J.
MICHALEK: Occupation — military
occupation is in the analysis as a covariate.
So ...
M.
GOUGH: Okay.
J.
MICHALEK: ... it was deemed
important to adjust for age. Doctors in level, of course, too is in the
analysis.
M.
GOUGH: I was more concerned about
...
J.
MICHALEK: Yeah, socioeconomic. Right.
Yeah.
R.
HARRISON: Hold on, Joel. Mike has a question.
M.
STOTO: How about the exclusion of
the people who were too lean? I mean,
presumably, they have less body fat.
J.
MICHALEK: “Too lean” means they
have not even enough body fat to be extracted by liposuction; that’s true.
M.
STOTO: So ...
J.
MICHALEK: Yeah, you’re right. I mean, we’re dealing here with — these are
people and some of them don’t even want to have it done in the first place,
like those 10 percent. Some are taking
medications that they were — confessed to the surgeon; that was the first
question he asked. He asked me. And so the surgeon decided in certain cases,
“No, you're not going to have this done.”
M.
STOTO: You know, I mean, I
understand that. I’m just thinking
about what biases may have ...
J.
MICHALEK: Biases, right.
M.
STOTO: ... been the cause of that
that are unavoidable ...
J.
MICHALEK: Yes.
M.
STOTO: ... but need to be ...
J.
MICHALEK: They’ll be discussed ...
M.
STOTO: ... acknowledged, yeah.
J.
MICHALEK: ... right up front. And this table will appear in the subsequent
journal article that is in preparation.
All right. Diabetic status, as
you know, has changed. Our definition
of diabetes has changed from 1995 to 2002.
And this slide doesn’t even represent the latest change, which was to
the ADA definition. This just
represents changes to our own understanding of diabetic status from cycle —
from 1995 to 2002. The roster for the
selection of the individuals for the adipose tissue study was developed in 1995
based on information we had at that time.
Since then — since then, our information has changed.
For
example, individuals that we thought were diabetic in 1995, eight of those were
subsequently found to be not diabetic because those eight — there was something
in the doctor’s diagnosis that was qualified:
“He’s not diabetic; he’s borderline diabetic.” And so we backed off and redefined them as non-diabetic in
2002. And of some of those that we had
originally decided were non-diabetic in 1995 were subsequently classified as
diabetic because they produced a very high or abnormally high glucose
measurement subsequent to 1995. And so
this definition is a little bit in flux, but you see on the diagonal elements
there, the large number down here and the 39 people up here, those are people
that have been consistently designated; these are diabetic or
non-diabetic. And this visit ...
R.
HARRISON: Excuse me for a minute, Joel.
J.
MICHALEK: Yes sir?
R.
HARRISON: Kwame, would you like to
comment on the borderline diabetes? Or
if you don’t, I would.
K.
OSEI: No, Bob, the problem with all
these — we talked about it before. The
change from the WHO criteria to ADA changes a little bit, but not dramatically
if you use the two-hour, you know, post-glucose challenge. And I think, if I remember, that’s what you
guys did. Okay. So that actually helps a lot, but whether
you were a diabetic in 1995, you are not diabetic in 2002 depends on what you
did to your life. If you’re — if you’re
a diabetic and you lose weight or you exercise more, you can eventually become
normal. So that means that they were
not diabetic; what they said is that they’ve done something, you know, to
themselves to make them “un-diabetic,” if there’s any word, you know, like
that.
The
most important thing is the list of medications. I think if you were diabetic, somebody might be taking care of
them, giving them medication, they’re seeing a doctor for intervention and that
will influence your GLUT4, you know, transporters. And they, you know, change the things that you are measuring so
it will be very important for us to know precisely what were they taking; they
were taking current medications, and so on and so forth. And I hope you’re going to show us that, you
know, the list of those medications.
J.
MICHALEK: That will part of the
next presentation. The detail is not
here today, but I have given those list of medications to Drs. Fumio and
Fujiyoshi. And they — I must admit —
they are a bit overwhelmed by the amount of data available because they’re used
to doing animal studies on rats, mice and guinea pigs. And so this is — we are working in an
unprecedented area here, an unprecedented complication.
K.
OSEI: But the ...
J.
MICHALEK: The medications are part
of the complication, yes.
K.
OSEI: Right. The pile of data in animals, are you — can
we have access to that? Do you have the
...
J.
MICHALEK: The animal data?
K.
OSEI: Their data in animals.
R.
HARRISON: I don't think, Kwame, that he’s saying that
there’s pilot data ...
J.
MICHALEK: No, this is published data.
R.
HARRISON: ... generated by these
investigators. He was just making the
general comment that these guys were more used to working with animals than
people.
J.
MICHALEK: It was based on their
published data that this study was launched.
K.
OSEI: Okay.
J.
MICHALEK: So I can tell you that
the diabetic status does influence the value of their measurements. And so you’re right; this is an important
factor, however, whether we use the ‘95 or the 2002 definition is fairly
irrelevant. As I understand it, the DNA
resides in the aqueous portion of the cell and these are adipose tissue
cells. There’s a fat droplet in the
cell. The larger that fat droplet, the
less DNA is available.
And
so they expected to see a negative correlation with body fat and they did
between their measurements and the body fat measured in 1997, which was the
body fat exhibited by the men as they were having their adipose tissue
drawn. There was a significant
correlation between these measurements, which they view as artifactual because
of the physiology of the cells.
Together with the four measurements that are prescribed by the protocol
...
R.
HARRISON: Oh wait, whoa. I'm sorry Joel. There’s
something that got missed in terms of what was being done. I mean, these people are doing, are
measuring these levels. They’re
measuring levels of messenger RNA ...
J.
MICHALEK: Messenger RNA; I’m sorry.
R.
HARRISON: ... using a reverse
transcriptase ...
J.
MICHALEK: Yes.
R.
HARRISON: ... polymerase chain
reaction assay they have to normalize.
So what they do is to extract RNA from these cells and probably use
primers that find a very short, unique segment of the RNA. In order to do that assay, they have to know
how much RNA they’re putting into each test tube. So the fact that each cell by volume has less RNA or DNA in it
than a non-adiposite cell doesn’t enter into the equation because they should
be using the same amount of RNA in each test tube. So it can't — I can't imagine that this correlation with percent
body fat is a simple matter of some cells having a lot of fat in them and other
cells not.
J.
MICHALEK: Well, this is where we
run into our problem of me presenting this material. Let me read to you from the script, which was written by
them. “The log transform values were
negatively correlated with percent body fat at the time of adipose tissue
sampling, suggesting an effect of fat droplet size on messenger RNA yield. The transform in GAPDH level was not
significantly different between diabetics and non-diabetics — a p value of 0.02
— suggesting that if this gene were used to normalize the others, it would not
introduce artifacts related to diabetes.”
So what they were trying to do was find a way to standardize these
values in such a way that they would not introduce an artifact. And they believed that these correlations
were related to the fat droplet size.
R.
HARRISON: Okay.
So it sounds like they didn’t try and assess — okay.
J.
MICHALEK: And in fact, I’m going to
stay with the script now. I’m not going
to ad lib at all.
K.
OSEI: Joel, one — before you move
forward, when you talk of “percent body fat,” how do you measure that?
J.
MICHALEK: To do that, we measured
body mass index, which was weight over height squared in metric units, and we
multiplied ...
K.
OSEI: But that’s not body fat.
J.
MICHALEK: No. We measured — the percent body fat
calculation has been used in all of our big reports. It’s something like — Bill Grubbs, you can correct me on this —
1.24 times BMI minus 13 roughly. And
that little linear equation was derived from a study done twenty years ago on
Army veterans to find a correlation between BMI and PBF. And that’s the — that’s the little linear
equation we use. And of course
statistically, it makes no difference whether you use BMI or PBF; the
correlations are the same because one is simply a linear function of the other.
K.
OSEI: Which body segment did you
get the fat — the biopsies? Was it the
abdomen?
J.
MICHALEK: Oh, it was the — it was
right here.
K.
OSEI: On the front?
J.
MICHALEK: The “love handles” so to
speak, right in here. So they decided
to use GAPDH as the — as the gene to standardize the others. GAPDH is considered a housekeeping gene
expressed at a constant level regardless of external factors and is thus often
used to normalize the PCR-determined values of other genes for comparison among
samples. The literature's ambiguous
about whether total RNA or GAPDH is a better measure of normalization. But as our GAPDH ranges — values range over
four orders of magnitude in one microgram total RNA, it is clear that we must
discard one of these measurements for normalization.
As
obesity confounds values normalized by total RNA, GAPDH is a better
candidate. GAPDH, let’s see, GAPDH
eliminated — that’s this slide — GAPDH normalization eliminated the correlation
of NFκβ and C/EBPα with percent body fat. As you see on this slide, the significance
is gone, demonstrating that GAPDH is a suitable housekeeping gene for
normalization to remove the effects of obesity on messenger RNA. Interestingly, GAPDH normalized C-SRC and
GLUT4 still correlate with body fat after normalization. However, the r squared is small.
That’s
the point they’re making; that minus 0.23, if you square that, you get about
0.05. So in other words, only about 5
percent of a variation in GLUT4 was explained by body fat. So what we’re talking about here, our
findings are now right on the threshold of knowledge and there’s little they
can say except that it’s interesting.
They’ve never seen this before.
In fact, that’s the — that’s the theme.
I myself as a statistician, I’m going faster than they can keep up with
me in the — in the results. I mean,
they’re — I produced a couple, two-three hundred pages of statistical analyses
that they’re still trying to understand.
R.
TREWYN: As we all are.
J.
MICHALEK: Yes. The genes were normalized by dividing by
GAPDH values and mean values were computed by group diabetic status and percent
body fat. I can tell you right now that
there are no — let’s back up one slide; I’m not here — okay; there are no
overall Ranch Hand to comparison differences on these four measurements. However, dioxin interacts very strongly with
these measurements as you’ll see. And
in particular in this slide, there’s an interaction between diabetic status,
comparison exposure group and GLUT4 — glucose transporter 4 — in that this
value right here, this 0 minus 0.61, is significantly different from the other
values in the comparison group, but not so in the Ranch Hand group because
there’s something different about the Ranch Hand group among lean non-diabetic
individuals, and correspondingly, lean non-diabetic individuals in the
comparison group as regard to glucose transporter 4. And that interaction reaches significance — less than 0.05.
There’s
something — the statistical analyses which these are detecting a significant
change in the relationship between diabetic status and glucose transporter 4
between Ranch Hand and comparison groups among lean non-diabetic individuals;
namely that this — there’s a significant increase in the glucose transporter 4
expression among lean non-diabetics in the comparison group, but not so in the
Ranch Hand group.
P.
CAMACHO: Not so?
J.
MICHALEK: Not so. This is, in other words, in the Ranch Hand
group, there is no such, but there are other correlations in the Ranch Hand
group that are significant. But I’m
just pointing out one; I know that there are many.
P.
CAMACHO: You talked about the lean
people.
J.
MICHALEK: These particular — I’m
just isolating a particular example of what we would call a three-factor
interaction between group diabetic status and GLUT4.
P.
CAMACHO: All right. For us sociologists, does this take us back
to the issue of the two lean people that we left out?
J.
MICHALEK: We’re talking right now
only about those individuals that are left in and these are the
individuals. These data are summarizing
the 313 people who are actually in this study.
P.
CAMACHO: But if we had — but the
lean issue is what I’m ...
J.
MICHALEK: If there was some —
perhaps some way to have included those that were deemed too lean to be in the
study, perhaps the results would’ve changed.
But then it may be true about other people who weren’t included in the
study either. I'm just telling you ...
P.
CAMACHO: Yes.
J.
MICHALEK: ... what we find among
313 that are in the study.
K.
OSEI: Were the diabetics in the
Ranch Hand Study different with respect to BMI than in the comparison group —
the comparison diabetic group? Could
that make a difference?
J.
MICHALEK: I’m sorry I don’t have
that demographic information; all I have for you is this. These are the sample sizes. To the extent — I can answer your question
to the extent of the design; that these individuals who are diabetic — sorry;
let’s go down here — the older Ranch Hand diabetics, five of them were deemed
to be obese. And the old — of the older
diabetic comparisons, nine of them were deemed to be obese meaning they — these
nine and those five both have percent body fat greater than 25 percent. You want to know whether the average body
fat of those nine was similar to the average body fat of those five? I don’t have that slide handy, but I can
maybe get it over lunch for you guys. I
have the data on my laptop.
K.
OSEI: Given the small sample size
you have right now ...
J.
MICHALEK: Small numbers.
K.
OSEI: ... do you have power to look
at this?
J.
MICHALEK: For individual sample
size — little cells like this one of three people — obviously we do not have a
lot of power there; that’s true. But we
do have power as you’ll see for combinations of these — of these factors.
M.
STOTO: Can I make a comment? I’m clearly not familiar with the biology
here, but it strikes me that when you’re looking at third-order interactions,
that’s — you’re on pretty thin ice in statistical terms and it’s hard to
interpret those in the best of situations.
But when you have sample selection issues like we talked about earlier,
I could easily imagine those causing third-order interactions and us not really
understanding how that would be the case.
J.
MICHALEK: That’s true. You can easily produce many, many
third-order interactions there beyond the ability of them to interpret and the
problem is to sort your way through all that.
And there are — there are ways and I’m leading to that on the last few
slides. So you can go off and produce
all kinds of things that are not interpretable.
R.
HARRISON: Why don’t you lead on, Joel, and ...
J.
MICHALEK: Okay.
R.
HARRISON: ... and maybe what we’ll
...
J.
MICHALEK: Your points are well
taken.
R.
HARRISON: ... maybe what we’ll do
is let's let Joel ...
J.
MICHALEK: Finish.
R.
HARRISON: ... dig a — dig a little
bit deeper ...
J.
MICHALEK: Hole.
R.
HARRISON: ... before we start shoveling
on him.
J.
MICHALEK: Thank you. Well, here’s another example of an
interaction which now involves dioxin.
I have to tell you that the normalized — the values for GLUT4 or the raw
values show similar trends with dioxin as when they’re normalized by GAPDH or
even when they’re normalized by other genes further validating the GAPDH
normalization. GLUT4 is down-regulated
in lab animals exposed to dioxin. This
is one of the results that led to this particular study. However, among lean diabetic Ranch Hands,
GLUT4 increases with dioxin — with dioxin load.
And
there’s dioxin across the horizontal axis up here in this cell and you see a
significant trend — a p value of 0.02 on that trend — whereas, you see no
corresponding pattern in the control group.
You see it in the Ranch Hand group, but not in the control group and
that change from this slide to this one is significant. That’s called an interaction. They speculate that perhaps GLUT4 expression
has become uncoupled from glucose regulation in the Ranch Hand group as
influenced by dioxin. And there I’ve
brought you almost to the threshold of knowledge to the point where I cannot
push these two guys any further on interpretation for another couple of months
until I visit them again.
Now
how do you make sense of all this data?
Well, one way is to use existing biological pathways that they believe
up until now had been established in animals.
And this is accomplished by a means of path analysis, another
statistical procedure which is now heavily dependent on the subject matter
specialist to provide beforehand, before even seeing the data, to provide a
pathway. And then the job of the
statistician is to ask whether the data is consistent with the pathway. And this is a new field which has been applied
a lot in sociology and psychology, but only now for the first time to this area
of toxicology.
What
you have here is a path diagram which was postulated by Dr. Matsumura before
ever having seen the data. This is a
pathway that he had derived from his animal experimentation: that he believed the dioxin should influence
C-SRC, and body fat influence NFκβ, and that those two would be
upstream from the outcome, which was diabetes measured here by fasting
glucose. The job of the — statistical
job is to invoke a SAS procedure called “PROC CALIS” to describe this pathway,
which is called “recursive” in this case because we’re — well, this is called —
this is an example of recursive pathway, which there’s a causative pattern from
left to right.
The
arrows mean that dioxin is supposed to cause changes in C-SRC; C-SRC causes
changes in C/EBPα and so on to GLUT4, and finally producing a diabetic
state. In fact, of course, they will
point out very quickly that there are feedback loops involved in the
process. And to a statistician, those
are called “non-recursive” paths. The
procedure or the SAS procedure that estimates these parameters and p values is
fully capable of incorporating feedback loops and I’ve invoked those. However, there’s a complication in the
method itself to produce a path that is not only biologically sensible, but
which is in fact numerically solvable.
And you can easily go off and produce a path which is not numerically
solvable. This path is; this path
diagram is solvable.
The
diagnostics in the software will tell us whether or not the path diagram is
explaining the data. And in this case,
this path diagram is — according to the diagnostics — is fitting in the Ranch
Hand cohort, but not in the control cohort because the null hypothesis is that
the path is explaining the data. The
alternative is that it’s not and so our rejection, a large Chi-square as you
see in the comparison group, indicates that this path is not holding in the
comparison group. But the p value above
0.05 in the Ranch Hand group indicates that the path is explaining the data in
the Ranch Hand group.
Other
diagnostics in the software will tell us whether paths exist that we had not
taken into account, whether there’s a statistical basis for paths that were not
in the original diagram. The path from
dioxin to GLUT4 was not in the original diagram. The original diagrams provided by Dr. Matsumura did not include
paths from dioxin directly to GLUT4 or from PBF directly to GLUT4. Invoking the SAS procedure on the path
diagram without those paths produces a diagnostic that tells us that literally,
SAS wants to put a path from dioxin to GLUT4.
And if you don't put a path from dioxin to GLUT4, you’re going to
degrade your fit. That's the procedure.
The
method is supposed to be conducted in close coordination with a biologist to
determine every step of the way whether the paths make sense biologically. Following the diagnostic from the SAS
routine and the input from Drs. Matsumura and Fujiyoshi, I drew a path from
dioxin to GLUT4. And when you do that,
you find a significant link from dioxin to glucose transporter 4 further
upstream than they had thought. They
didn’t expect to see this. And not only
did they not expect to see it, they expected to see a negative correlation, and
in fact, it’s positive, which is consistent with the scatter diagram I showed
you on the previous slide, suggesting to them that there is something being
disrupted by dioxin and its connection with glucose transporter 4 and diabetes.
There,
I have brought you to the final interpretation. There’s nothing more that they can say at this point; I expect to
visit them again in a couple of months.
The analysis is very difficult and can only be done together in front of
a computer with interactions with them.
So what we’ve got is a data set that is, to put it mildly,
complicated. Interactions, to put in
terms — statistical terms — prevent simple interpretations. Interactions prevent the analysis of main
effects, to put it in statistical terms.
There are no simple main effects, but there are interactions. And as you saw on the previous slide, some
of those interactions may be interpretable.
And
as you already realize, small sample sizes prevent more detailed path
analyses. Path analyses have been
applied to a very large data set; it’s over 1,000 subjects. We have 300. And the limitations of the path analysis is that they must
include normally distributed data. So I
can’t include a single path for all — both cohorts combined because I can’t
include a group flag Ranch Hand control.
The data must be normally distributed.
So I have to make a separate path and apply it to Ranch Hand and a
separate path and apply it to comparison.
So when I do that, I have to cut the sample size in half. So now we’re down to about 120 subjects,
which is pushing it to the — to the limit of believability.
And
of course, the power of the drill path analysis is to — is to invoke a model to
fit it to a data set and then replicate, conduct the study over again somewhere
else on a different cohort. And of
course, that’s not possible. And
finally, there may be other factors that we haven’t measured. In the path diagram associated with each of
these components is an error term — a statistical term which is there to
measure error. Those error terms are
supposed to be uncorrelated; in fact, in the data set they are. And that’s a flag to indicate there are
probably other factors involved common to these which we have not measured and
that’s causing correlations between the error terms. And so that will be mentioned in the final discussion section of
the paper.
So
what we’ve got is that this is the first study of dioxin and gene expression
ever done. It has been done under
strict quality control. We have
complete data on every subject. We have
good follow-up and many risk factors were measured that includes medications
which haven’t even been dealt with yet in this presentation. And of course, the limitations are — which
you’ve already pointed out — sample sizes, interactions, replication is not
possible, and unmeasured factors. So
the bottom line at this point today is that the data suggest dioxin may be
disrupting the normal diabetic pathway through interaction with GLUT4. However, they will be quick to say that more
study is needed, including more thinking about the complex regulatory mechanism
of diabetes and other pathways. Thank
you very much.
R.
HARRISON: Who shall cast the first stone?
M.
GOUGH: I have two questions: one’s about the biology. What is the — I don’t know about animal
studies they’re depending on. But
typically, animal studies of biochemical interaction or biochemical studies
with dioxin are done at doses that are thousands of times higher than those
experienced in human beings. What is
the ...
J.
MICHALEK: Well ...
M.
GOUGH: What is the ratio here
between the animal studies and what people have — these people have?
J.
MICHALEK: Good point. In fact, those are — that was the reason why
we focused on Dr. Matsumura’s research in the first place. He was working in a — in a — in a dose range
which is compatible with what the Ranch Handers received in Vietnam. In other words, he was working with less
than one-tenth of a microgram per kilogram body weight exposure to the
animals. The animal — other animal
studies you’re talking about were very
heavy exposures to the point of toxicity.
These are very light ...
M.
GOUGH: Yeah. Yeah.
Okay.
J.
MICHALEK: ... very, very small
doses. And I think they can defend that
point and it will be defended in the — in the introduction to the — to the
paper.
M.
GOUGH: Okay. Well, that’s good. The second thing is on the — you have a slide entitled “mean raw
gene expression.”
J.
MICHALEK: Yes.
M.
GOUGH: Those numbers were the — I
mean, those numbers between the comparison to the Ranch Hands were the same so
far as I ...
J.
MICHALEK: That’s a display of the
fact that there are no — what we would call “main effects” in the data set.
M.
GOUGH: Yeah. So why shouldn’t I — given reservations
about sample selection, and exclusion of people and then what Mike said — and I
don’t — I’m taking his word that a statistician said that these third factors
...
M.
STOTO: Third-order interactions.
M.
GOUGH: ... third-order interactions
being weak, why shouldn’t I walk away from this saying there’s no difference at
least between the Ranch Hands and comparison?
J.
MICHALEK: You can easily do
that. In fact, you can do that ...
M.
GOUGH: Good.
J.
MICHALEK: ... at any time you want
to.
M.
GOUGH: But in fact, you’re getting
into this morass. What I — I’m not a
statistician; I’m also not a diabetologist.
J.
MICHALEK: I know what you’re
saying. Go ahead; I’m sorry.
M.
GOUGH: But you’re just getting into
this morass of analyses and I'm not — it — I’m not sure if it's
informative. It’s not informative to
me; I don’t know if it’s informative to anybody. And I’m wondering about what the endpoint in your mind is?
J.
MICHALEK: To me, the endpoint is
about this slide right here: it’s to
reach a consensus by the biologist that this path — a pathway like this or
similar to this — is either supported or not supported by the data.
M.
GOUGH: Can you do that with the
qualifications about who was excluded and who was included and ...
J.
MICHALEK: Let’s — actually, all I
can tell you about that ...
M.
STOTO: Can I respond to that?
J.
MICHALEK: Yes, go ahead.
M.
STOTO: First of all, I want to say
that I think it’s important that this kind of study be done. I mean, this is — this is an example of how
the Ranch Hand Study is providing information that’s potentially very valuable
for understanding the causes of diabetes, the natural history of diabetes. Now that doesn't mean it’s easy to sort out
what these — what this all means and — but, you know, when you do observational
studies in humans, this is always the case; that you have to sort of do the
best you can and then when you write up the interpretation, say, you know, is
it possible that what you found is due to the limitations in the study? And you just have to look at that carefully
and make a judgment.
And
I think that the — that the key issues here that need to be looked at — and I
have every confidence that they will — are, is whether issues in the sample
selection — I mean, you don’t have complete data on everybody. You have complete data on people who weren’t
lean and who weren’t on medication. And
both of those things seem to me to have the potential for putting bias into
this study and there needs to be careful thought about whether that is the
case. And that’s what a discussion
section of the — of the paper ought to do and I presume you will do.
The
other issue, potential I think that needs to be looked at is that when you look
at the two slides that have to do with the — with the scattergrams — it’s just
the one above this and the one before that — one thing that you see and you’ve
— we’re not surprised is that the Ranch Hands have more dioxin body burden than
the controls. That’s the very nature of
the study. And I mean, it strikes me
that, you know, the relationship that we see — I mean, I think the key thing
here is that in the — on the right — the two graphs on the right goes up on the
top and it goes down on the bottom when if you go to the comparison slide, who
knows whether that would be the case if you had people with more dioxin?
I
mean, it’s simply not enough, particularly in the — there’s simply not enough
people with high levels of dioxin to see whether the slope goes up that
way. So I mean, I think that you have
to look at this and you have to consider that, but that's an important thing that
needs to be ...
R.
HARRISON: Kwame?
K.
OSEI: Yeah. Joel, I’m so very confused about whether
dioxin increased or decreased messenger RNA of GLUT4. So far, I haven’t seen it.
And as you know — maybe Phillip might’ve talked to you about the, you
know, what GLUT4 does. GLUT4 is just a
plain glucose transporter that pushes glucose into the cell. And the two components over it is the number
of GLUT4 and the activity of GLUT4.
So
you can have any of these variables being changed that will impact on, you
know, diabetes. What we know in diabetes
in humans is that GLUT4 is either the same or down-regulated to give you an
inability to transport glucose. That’s
when you become diabetic. So far, I
don’t see any — the connection at all from the evidence that you have; either
you have true diabetes. And so, it’s
not explainable to me by what you’ve provided so far, you know. So probably Phillip needs to sit down and
look at this very carefully.
R.
HARRISON: Kwame, what you're saying is that if dioxin
produced diabetes, you’d expect GLUT4 to go down, but the data looks like it
goes up?
K.
OSEI: Right.
R.
HARRISON: Is that what you’re
saying?
K.
OSEI: That’s what it is.
R.
HARRISON: Okay.
K.
OSEI: The data is saying that ...
R.
HARRISON: You just have to explain
this for a country boy from Mississippi.
K.
OSEI: Yes. You have to increase your — actually, if you
exercise, you increase your GLUT4 and that prevents you from becoming
diabetic. So if I take dioxin, which
I’m not going to take, and to prevent — I hope that it will prevent diabetes if
what you’re saying is true; that I should increase my GLUT4 numbers with proper
activity. If I don’t do that, then
something else is wrong. It means
either the GLUT4 is not representative or reflective of dioxin effect and there
has to be some other variables, you know, that you need to go back and look at.
That’s
where the whole idea of exercise becomes very important, and the physical
fitness becomes important and other medications that these people were on. But we know a lot of these medications may,
you know, influence GLUT4 levels, such as TZD, and so on and so forth. So it’s so important that you tease out all
that before you even you start doing all these pathways because I think you
muddy the water by excluding, you know, not including all those variables.
R.
HARRISON: For the transcription, that's “TZD” —
thiazolidinediones.
M.
STOTO: I think that what these
graphs here on that slide are showing is that there’s a positive relationship
between dioxin in GLUT4 in lean people and it’s a negative in obese
people. And then the other set of
slides shows that there’s — that it’s no such relationship to either one for
the non — for the control group. Now what
that means, I don’t know, but that’s — I think that’s what they’ve — I think that's
the key thing they found, right?
J.
MICHALEK: So far. This is all I’m willing to talk about at
this point. They have a lot more on
their table than this, but ...
R.
HARRISON: Just to — just to — a couple of other
points: and that is that there’s
intracellular feedback mechanisms that require one at some point in time — I’m
not suggesting that it needs to be done now — but it requires you at some point
in time to establish that the changes in messenger RNA levels that you’re
looking at here are actually reflective of changes in protein transcription or
changes in protein levels and not some other, for instance, a reduction in
GLUT4 degradation, which then essentially feeds back on the GLUT4 message and
reduces its activity. So at some point,
you’re going to have to make that kind of correlation.
J.
MICHALEK: Yeah. Exactly.
Yes.
R.
HARRISON: The ...
J.
MICHALEK: I wrote code to run a
feedback loop from GLUT4 to PBF and from C/EBPα and of NFκβ.
R.
HARRISON: Those are all — you’re all looking at, with
the polymerase chain reaction, you’re looking at MRNA levels. And I’m saying that actually, the assumption
implicit in that is that MRNA reflects the amount of protein being synthesized. It’s the protein that does the work, but
that’s not always the case. And
someone's going to ask you at some point to show that protein levels are
actually changed either by immunohistochemistry, which is kind of qualitative,
or by some other means of actually measuring the amount of GLUT4 in the cell.
R.
TREWYN: The activity part of it.
R.
HARRISON: Yeah.
Yeah, even as far as activity is concerned. The — almost the point I should’ve made first is that I view all
this as phenomenology that says that this phenomenon occurs. And until you actually have the mechanism or
mechanistic explanation of what’s occurring, I think we all understand that
what we’re looking at here are phenomena.
And what we see in this study is not a mechanistic explanation, but is
an investigation of phenomenology at the cellular level as opposed to the
investigation in large population groups.
The
last thing I want to suggest, Joel, is that there are a number of projects
around now that are doing cell simulations, computer simulations of cells. And the one that comes particular to mind is
a Japanese project that I believe is called “E-CELL” — E hyphen CELL. And I just remember that they’ve done a lot
of work with glucose metabolism in this simulation and there’s some non — there
are non-intuitive things to be found when you start actually putting these
simulations together. And I think that
this plays right to your strengths in terms of how to analyze data and would be
a different way of looking at this than a straight statistical — these are —
these are actual — they’re actually cranking in all the different parts and
then tweaking little parts at a time.
J.
MICHALEK: Thank you.
R.
HARRISON: Oh, and I — and I would like a copy of the
manuscript when you — when you all ...
J.
MICHALEK: As soon as they’re ready
to let go of it, yes.
R.
HARRISON: Yeah.
J.
MICHALEK: Just another note: these asterisks mean that the coefficient is
positive and significant — statistically significant — connecting C-SRC, and
C/EBPα and NFκβ, C/EBPα, which is completely consistent
with their — with their preconceived ideas regarding what the pathway really
is. So they see that as a strength. That’s all I had on that study.
R.
HARRISON: Now we have Syndrome X?
J.
MICHALEK: Yes, I think. No.
We have cancer in the comparison group and Syndrome X comes — if you
want, we can do Syndrome X next.
R.
HARRISON: I’m sorry, Joel. The agenda says Syndrome X is next.
J.
MICHALEK: Syndrome X, all right.
R.
HARRISON: I — go ahead. I’m not — no; no, seriously. I think the papers are arranged in the
cancer order.
Cancer Prevalence in Comparisons
J.
MICHALEK: Okay.
During our analysis of the data which will soon be published in the Journal
of Occupational and Environmental Medicine, we discovered trends in the
comparison group that were unexpected and the purpose of this presentation is
to show those to you. First off, Marian
Pavuk is with us here today. Fatema
Akhtar no longer is with us; she moved to the University of Texas. This is the bottom line: cancer prevalence is adversely related to
years of service in Southeast Asia in the comparison group. And as you will see, this has apparently
nothing to do with dioxin or Agent Orange.
And
like I just said, we discovered this pattern while analyzing Ranch Hand
data. We're using the same materials
and methods that were used in the Ranch Hand cancer analysis — the same data
set. Cancer prevalence is determined
through 31 December 1999; all cases confirmed by record review and this
includes all individuals who were either partially or fully compliant to any
physical exam through the 19 — to the fifth one, which is up to 1997. Currently, we are updating our databases and
should have the new data up through the current physical ready within a couple
of weeks.
This
analysis includes people who were what we call “partially compliant,” which
meant they simply came, and were interviewed and then refused to take a
physical. And most of those partial
compliants occurred in the baseline exam.
And of course, you know, adds an sample size, but no physical exam data
for those people. Time to onset is
measured and this is different; we’d never done this before: measured from the end of the last Southeast
Asia experience to the earliest occurrence of cancer, death or end of
follow-up.
In
all of our previous reports, we had measured time to onset from what we would
call the “qualifying tour,” which was the tour that was used to identify an
individual to be in the comparison group in the first place. But because we found this association
between years in Southeast Asia and cancer, we’ve decided to extend the cut
point to the last Southeast Asia experience in an attempt to imitate exposures
that would occur in industry, for example, where time to onset is measured from
the last exposure period.
We
assumed a fifteen-year latency, which meant cancers that occurred within
fifteen years of the last Southeast Asia experience were not considered. They were intentionally excluded. And anyone who didn’t have fifteen years of
follow-up from the last Southeast Asia experience was excluded. So people, to get into this study, had to
have fifteen years of follow-up after the last Southeast Asia experience to be
in this study in the first place and that didn’t exclude very many people
because of the good follow-up we’ve had.
Did you want to ask a question?
R.
TREWYN: Yeah.
Southeast Asia, because the comparison group consists of individuals who
were stationed outside of Vietnam and those who were stationed in Vietnam, has
that been segregated out in this study?
J.
MICHALEK: Not yet. I’m showing preliminary information.
R.
TREWYN: Okay.
J.
MICHALEK: What you will see is that
— and it may surprise you — that the Southeast Asia tours occurred between 1942
and 1982. There’s a spread of Southeast
Asia tours that coincide with World War II and move all the way up to the
present time. There are tours from
those early years all the way to the present.
And the maximum amount of time spent in the Southeast Asia region among
comparisons is about fifteen years.
We’re using the same cancer definitions that we’ve used in the other
papers we’re working on — all of them prescribed by the National Cancer
Institute Surveillance Epidemiology End Results section, the SEER cancer
category definitions.
In
particular, the all-site SEER category includes all cancers except basal cell
and squamous cell carcinoma, which we call “non-melanotic skin cancer.” And throughout the entire analyses, we used
proportional hazards modeling, which includes not just a flag as to whether or
not they had cancer, but includes the actual time from exposure to onset. All of that is taken into account along with
other risk factors. Other risk factors
for all-site cancer that were accommodated in the model include year of birth,
race, military occupation, skin reaction to sunlight, eye color and pack-years
of smoking.
Pack-years
of smoking is measured from baseline because to use otherwise, to include
pack-years of smoking up until the present time would induce an artifact
because then we would have a — what’s called “time-dependent covariate,” which
would continue to accumulate value as the years passed and it's not a good
idea. Respiratory cancer is adjusted
for, of course, pack-years, and age, race and military occupation. Prostate cancer — the same covariates we
used, in fact, in the Ranch Hand paper which is soon to be published. And digestive system cancer is year of
birth, military occupation.
These
covariates are derived from our knowledge of the literature and are subject to
change depending on new information about covariates. And there’s a set of covariates for the skin cancers; they were
restricted to non-black veterans which include about 92 percent of the cohort
adjusted for those variables: skin
exposure, and eye color and military occupation, year of birth. We began with 1,785 comparisons who were
ever partially compliant or fully compliant to any exam between 1982 and 1997.
For
the purpose of dioxin summaries, we excluded people with no dioxin, of course,
but those were not used. This dioxin
reduction was not used in the main analysis.
It’s only used in one place and that’s only to show you that the data we
have are unrelated to dioxin level.
It’s the sample size 1,785 that’s important for cancer analysis. Using standard methods to display data, we
broke the range of years spent in the Southeast Asia region by up into
quartiles and that produces a balanced design of approximately equal numbers of
subjects by interval of time.
It
does not produce nice intervals, but it does produce a balanced design which
gives us better statistical properties.
So what you've got is that about 50 percent of the cohort spent less
than 2.1 years in Southeast Asia, for example.
So 2.1 is the approximate median of years spent in the Southeast Asia
region. Demographically, we have a
situation where there’s a strong age effect.
Those individuals who were there between 3.6 and fifteen years were born
in approximately 1934 and those that were there for a short time between 0.1
and 1.3 years were born in 1945.
There’s about a ten-year difference in age there, which is accommodated
by the analysis.
And
their age as represented on the second line — laser pointer just quit; oh well
— represents their age during the qualifying tour. Thank you. Qualifying
tours were primarily during the Vietnam war period. Baseline pack-years of smoking, you see a strong trend of
increased smoking with increased years in SEA; the analysis accommodates that
too. Let’s see; which way do I have to
— got it; there we go. Smoking — and of
course, it’s flat with regard to dioxin.
These men are all at background levels and what you see there, perhaps a
slight age effect. We know that dioxin
levels increase with age, so what we might be seeing there is an artifact
simply due to the age differences.
And
the tour years coincide, like I said, with the — is the qualifying tours with
the war itself and the body fat is the body fat measured during their
qualifying tour. They were all
relatively lean then. Days in SEA, of
course, they go up with years spent in SEA — up to 1,500 days in — that’s the
median in the high category. And days
in Vietnam, there’s a trend too down to 75 days, median in the lowest
category. And all the other variables
are fairly stable. Enlisted ground
individuals are less represented in the longer time periods in the region and
officers are slightly more represented.
All
of these variables are statistically accommodated by the proportional hazards
model. These are data on the reported
exposures that we gleaned from the National Opinion Research Center
questionnaires that were given during the physical exams. This — the question was, “Have you ever been
exposed to asbestos — yes or no?” This
is very rough data. “Were you ever
exposed to insecticides in your whole life?”
And so you see these numbers and they’re fairly stable across the years.
Well,
here’s the bottom line. On the highest
years in SEA category, we have a significant increase in the risk of all-site
SEER cancer in the comparison group:
relative risk 2.9, p value less than 0.001. And there’s a significant trend of increased risk with increased
years in SEA: a p value less than 0.001
in the comparison group. That’s after
adjustment for age, smoking, skin color and eye color, military occupation and
all the covariates I've listed before.
A
significant increase in the risk of cancer of the digestive system — smaller
counts — prostate cancer in the high category.
In other words, high years of exposure is a relative risk of 8.7, p
value 0.003. We have done supplementary
analyses to probe this data to make sure that the proportional hazards model is
not leading us astray. And we have
demonstrated that that is the case; that the SAS proportional hazards procedure
is in fact not leading us astray and that we are not — we are not simply
displaying an artifact due to age.
Cancer
of the respiratory system is there’s a weak trend overall, but no significant
increase in relative risk in the high Southeast Asia category. The numbers in melanoma were too small to
analyze and there was no significant trend or increase in relative risk in any
Southeast Asia category with regard to basal or squamous cell carcinoma. And here are the other sites: and you see lots of zeroes, and small
numbers and which is the reason why these were not analyzed. There’s nothing here that stands out as a —
as a cause for alarm in other words.
And
just to show you what happens if we give up the fifteen-year latency, what
happens then is the findings go away and so it seems to be important that there
be a fifteen-year latency. In fact, you
can — you can change your latency definition and find a continued statistical
significance if you push it out to even twenty years. And here is one of our checks on the proportional hazards model
to make sure we weren’t simply seeing an artifact; and that is, we stratified
by year of birth, and by Southeast Asia years and we considered prostate
cancer. And you find that if you — if
you combine the first two years of birth categories and assess trend using a
Chi-square method called the “Mantel-Haenszel,” a Chi-square test for trend,
you find a significant trend of increased risk among individuals born prior to
1944.
So
in other words, this supports the idea that the proportional hazards model is
giving us a non-artifactual outcome.
And here’s the histogram of those tours running all the way from World
War II up to the present time — the bulk of them, of course, occurring during
the Vietnam war. So the issue, of
course, is why are we seeing this? And
all we can say at this point is that it appears that whatever it is, appears to
be unrelated to dioxin or any exposures to dioxin-contaminated herbicides that
may have occurred somehow during the war.
Where
were they when they were over there?
Well, that question itself is obvious, but difficult to answer. We’re still working to create a database of
exactly where they were during all of those tours from World War II to 1982. That’s a very labor-intensive goal, task
because these men have many tours: TDYs
all the way up to PCSs which are all recorded and require line-by-line study
and interpretation because unit designators change with time. An individual that may be with the 130th
Air Tactical Squadron may have been stationed in Taiwan in 1965, but may have
been stationed in Alaska in 1972, you know.
So
you have to know you can’t simply use the designators as your information. You have to study these and we’ve been
working to do that; we’re not done yet.
We’re still creating a database of location — exact locations and
dates. And we’ll have that, but we
don’t have it today. And of course, we
don’t have a detailed exposure history of information from these men while they
were there. That was not collected,
unfortunately. We only measured one
chemical and that is dioxin in these men.
We didn’t ask them about what they did over there, about their diet and
about their exposures. We never did
that.
The
strengths, of course, are the same strengths applied to the whole study: everything is verified up to three times by
record review; we have rigorous quality control; and dates of onset are all
determined by record review, not by the individual’s opinion. So what could be the reason? We believe intuitively that what we’re
seeing here is that the Southeast Asia — time in Southeast Asia is a surrogate
for something else and we don’t know what that something else is.
A
candidate — speculation is something in the environment, something in the
water, or the food or perhaps insecticides.
During the Vietnam war, the World Health Organization implemented an
extensive campaign to eradicate malaria in the entire Southeast Asia region and
that's been documented in the literature.
And of course, there are the tropics themselves and everything that goes
along with that.
So
what we’ve got is a strong effect in the comparison group which has confounded
our analysis of cancer in the Ranch Hand group and will continue to do so as
long as we’re with this data because this isn’t going to go away. We’re talking about trends that were
produced years ago and are not going to change. Whatever we’re seeing has nothing to do statistically with dioxin
and we don’t know what it is. Thank
you.
R.
HARRISON: Mike?
M.
STOTO: Well, I think this is very
interesting and this is another good example of the kind of work that can be
done with these data and nicely done as well.
The thing that troubles me is the big difference in smoking between the
people who were there a long time versus a short time. And I understand that you've done your best
to control for that statistically, but you know, that’s never — there’s always
a possibility that you can’t control all of it statistically.
It
is reassuring that the things that you’ve — that the — that the cancer sites
that you found to be significant, digestive and prostate, are the ones that are
not generally thought to be associated with smoking. So that suggests that smoking may not be the problem. The fact that there is a trend though in
respiratory cancer suggests that maybe there might be something there, but the
sample size may not have been big enough.
So I guess one thing I would suggest is to maybe group together, you
know, the five or ten cancer sites that are most thought to be associated with
smoking in some sense and maybe see whether you find anything there just as a
check.
R.
HARRISON: If memory serves me correctly, there’s more
liver cancer in Asians than there is in Americans, which would come under the
heading of digestive system. And it led
me to wonder if this difference between the occurrence of digestive and
prostate versus respiratory and skin might be reflected in people who have
always lived in Asia versus people who just passed through there during the
military; that what we’re reflecting is something environmental, but something
that’s — if it’s environmental, it should be reflected in the indigenous
population as well as the visitors.
M.
STOTO: Isn’t it thought to be
associated with hepatitis?
R.
HARRISON: Well, my guess would be
that it’s associated with viruses that like to live in hepatic cells; I’d be a
little more general than that. It may
even in some cases in some particular areas be related to parasitic
infestations that involve the liver as well.
Yes sir?
S.
LEFFINGWELL: Leffingwell. I think I have heard about an association
between human papillomavirus and prostatic cancer, which would lead to the
possibility of sexually transmitted diseases and prostate cancer. And I wondered if we had either a
serological method of looking for HPV antibodies, or as a surrogate, looking at
treatment for sexually transmitted diseases during their tours in Vietnam. Similarly if it should turn out that the
liver is in fact the site in the digestive system that’s high, then certainly
looking at hepatitis B and C, which also might correlate with treatment for
gonorrhea — it certainly does in the civilian population — might be
helpful. Now why that would or would
not correlate with dioxin exposure isn't at all clear, but there may have been
some unit factors, morale sorts of things that would affect it.
J.
MICHALEK: Let’s see. Dr. Pavuk has done a literature review on
this.
R.
HARRISON: Dr., if you could come to
the microphone somewhere, it’s going to help us. Otherwise, Mike Stoto’s going to get on us about transcription
and everything.
N.
RIVERA: State your name, sir.
M.
PAVUK: Actually, I would’ve ...
R.
HARRISON: State your name, please.
M.
PAVUK: Marian Pavuk. I’m an epidemiologist; I work with SpecPro
on the Air Force Health Study as a contractor with Dr. Michalek. I worked previously on the study when I was
still at the University of Texas with Dr. Schecter also. Concerning the digestive system, there have
been 26 cancers of the digestive system.
And of those, there were — there was only one liver cancer identified in
those cancers. There were six cancers
of colon, nine cancers of rectum-to-stomach cancers, and three pancreatic
cancers.
So
except for a slightly different proportion of colon and rectal cancer which
usually is a little higher — more colon than rectal cancers in the general U.S.
population — we didn’t — we couldn’t see, because of the small numbers,
couldn’t really say, you know, with just one liver cancer whether there was
something unusual about the distribution of the cancers — digestive cancers in
this group. You correctly stated that
infectious diseases or parasitic diseases in the — in the region are really
related to liver cancer. For example,
Thailand has the highest rate of liver cancers in the world reported related to
those other diseases.
R.
HARRISON: Yes sir?
R.
TREWYN: Any correlation to head and
neck tumors? That nasopharyngeal
carcinoma is very ...
M.
PAVUK: Is another ...
R.
TREWYN: ... characteristic ...
M.
PAVUK: Is another characteristic
group of cancers in this area. I don’t
think we had any numbers — really, very small numbers in those areas. I don’t think really, there was only two
brain and central nervous system tumors all in all at this point and no
nasopharyngeal. But that’s not what the
doctor is suggesting.
R.
HARRISON: Mike Gough?
M.
GOUGH: Just two things: one is when the members of the Committee and
you guys review these slides, I — you — I would appreciate, I think other
readers would appreciate if you would pay attention to the labeling of the
slides because the “all-sites” — you have a slide called “all-sites cancer” and
that’s really the slide where you eliminate the latency period.
J.
MICHALEK: That’s true.
M.
GOUGH: And it ...
J.
MICHALEK: That slide needs a better
title.
M.
GOUGH: And it — I’ve had that
feeling in the — in the earlier discussion too. I read the title of the slide and it wasn’t very informative; it
didn’t focus me.
J.
MICHALEK: Yeah.
M.
GOUGH: The second thing is I think
you’re biting off much more than anybody should chew than try and thinking
you’re going to be able to find an exposure or exposures to associate with
these excesses; 25 percent of us get cancer and very, very, very rarely can
anybody point to what caused it. And I
just — I mean, I think looking at the unit locations may be a reasonable thing,
but beyond that, I'm not — I’m sure you have more valuable things to do.
R.
HARRISON: Any other — yes?
M.
PAVUK: Yes, if I may, that’s, you
know, the stored samples which include blood samples for the Ranch Hands and
the comparison veterans would still be analyzed for the whole battery of other
organochlorine compounds and investigate, for example, if DDT, or DDE or
pesticides used in Southeast Asia has any association with this cancer trend
observed.
R.
HARRISON: Thank you.
Anyone else?
R.
TREWYN: One more point: I also think chlordane was used a lot, which
has been banned in that part of the world at least during that period of time,
and might be another candidate if you feel compelled to search for possible
associations.
R.
HARRISON: Thank you very much. Let's see.
We’re running almost — well, we’re running at least a half hour behind. We were supposed to have a break at
10:30. I feel the urgent need for — to
have — to have — to have a cigarette.
Does anyone else object to us taking a break now and for let’s say ten minutes
and resuming? We can perhaps run just a
little late into the lunch hour, and try to catch back up and maybe even change
the format by holding our questions until the end of the presentation rather
than sprinkling them through the presentation.
R.
TREWYN: Is there any reason we
can’t work through lunch?
R.
HARRISON: Yes.
L.
SCHECHTMAN: If I can comment there,
we’re having lunch brought in so that we can — we can make that judgment as we
proceed through the morning depending on how much time Joel needs. So I’ll take the executive secretary’s
prerogative, and if need be, overrule the chairperson.
R.
HARRISON: Fine with me; last
time. Ten minutes.
[BREAK 11:02 A.M. - 11:20 A.M.]
R.
HARRISON: Okay.
Dr. Camacho is back; we can get started.
J.
MICHALEK: Let me make a correction;
I made a mistake. I’m going to back up
right here; that’s not it. Right
there. I said that this slide described
the all-site cancer without regard to latency; that’s not true. This is — yeah — this slide is still with
the fifteen-year latency, however, we are now using literally all sites and
we’re throwing in basal cell and squamous cell together with all the other SEER
categories. And now we’re expanding, in
other words, to literally all sites.
In
other words, this is the definition that we had used in older published
research from the study. And when you
do that, all the significance goes away.
So in other words, there’s something happening. There’s something special about basal and
squamous cell carcinoma. By excluding
those, we see the overall trend with years in Southeast Asia. Yes sir?
R.
HARRISON: I don’t understand why
you do that, Joel.
J.
MICHALEK: Why did I make this
slide?
R.
HARRISON: Because I don’t think anybody thinks that
all cancers have the same bases. And so
when you do something like this where you throw in all sites, it’s like taking
the average temperature of the world irrespective of the seasons and saying
that that describes something that happens, you know. People are going to be really surprised when they wind up in
Rochester, New York in the winter expecting it to be whatever — sixty —
whatever the average temperature for the year is — sixty degrees.
J.
MICHALEK: I understand that we
always have that problem when we combine or we lump together different disease
categories and your point is well taken.
R.
HARRISON: It makes sense to go by systems.
J.
MICHALEK: Yes.
M.
STOTO: And that’s also an issue
with the all SEER sites where they found the significant relationship.
J.
MICHALEK: That’s right.
M.
STOTO: So ...
J.
MICHALEK: All SEER sites ...
M.
STOTO: ... I mean, normally you
would expect ...
J.
MICHALEK: ... are right here.
M.
STOTO: ... that by lumping ...
J.
MICHALEK: Yeah.
M.
STOTO: ... lumping things together,
you wouldn’t find anything for exactly the reasons you said. But when you do, at least you have something
that you’ve got to scratch your head and think about. And then I think that what they did was then — was appropriate;
is that they broke it down by looking at specific sites.
R.
HARRISON: I guess the question I'd have is, I wouldn't
even try and — I wouldn’t even try and show the all SEER sites.
M.
STOTO: There’s certainly — I mean,
I think I would’ve agreed with you from the outset; that that’s not the way I
would’ve analyzed the data. But having
done it and they found this, then you have to sort of figure out what it
means. And they’ve gone the next step
beyond that.
R.
HARRISON: Let's go on, Joel, and we’ll ask questions —
everybody take notes. We’ll ask
questions at the end of the presentation.
J.
MICHALEK: Well, there are two more
cancer presentations. This one is
really a repeat of what you saw in March 2003.
I mean, it’s only here because the paper describing these slides will be
published on the 11th of February of this year in the Journal of
Occupational and Environmental Medicine.
I’m putting this here asking you whether you would like to see these
slides again or should I just skip them and move on?
R.
HARRISON: All in favor of Joel ...
M.
GOUGH: Can I ask one more
question? It’s not a — it’s — let’s —
Joel, I — one question: I finally
figured out these tiny little numbers down here. On your slide 81, which is “cancer incidence and mortality,” does
that have a latency period in it?
J.
MICHALEK: I can't see the numbers
on the slide.
M.
GOUGH: It's called “cancer
incidence and mortality.”
J.
MICHALEK: Okay. No.
Those slides — these slides do not consider latency.
M.
GOUGH: Okay.
J.
MICHALEK: But they could.
M.
GOUGH: Well ...
J.
MICHALEK: And in fact, subsequent
...
M.
GOUGH: ... there should be some ...
J.
MICHALEK: Yeah.
M.
GOUGH: Again ...
J.
MICHALEK: Yeah.
M.
GOUGH: ... you know, the people who
are going to read these things ...
J.
MICHALEK: Yeah.
M.
GOUGH: ... and they’re going to
say, “Well, how does that relate?” And
in fact, it’s all right to present it two different ways so far as I’m
concerned, but you just have ...
J.
MICHALEK: To be careful on the
title of the slides. Yeah. I should tell you that in the prostate
presentation, we do consider latency of twenty years and that’s in the title of
the — that’s in the methods section.
Regarding prostate cancer, I’m suggesting that we move the prostate
presentation adjacent to this one so we will have looked at all the cancer data
together before moving on to memory loss because right now, the prostate talk
is at the very end.
R.
HARRISON: So unless there's an objection, we’re going
to skip what’s been presented in a previous meeting.
M.
GOUGH: Well, I just want to
register, I sent Joel an e-mail in October.
I do not understand why their analysis was restricted to men who were —
among the Ranch Hands, why the analysis was restricted to men who had served
less than two years in Vietnam.
J.
MICHALEK: Okay. That — I have a slide show on that too and
that’s coming.
M.
GOUGH: Okay. I just ...
R.
HARRISON: So do you want this?
M.
GOUGH: No. No, I just want — I just want those two
things. I — one — I’m finished.
R.
HARRISON: Okay.
M.
GOUGH: Let me turn off my mike.
R.
TREWYN: Can I make one quick
point? Because in answer to your
question, there is one aspect in here with the 100 percent of Southeast Asia
service being in Vietnam for the Ranch Hand and 0 percent of that being in
Vietnam that I think is an important discriminator and that I would hope would
be fairly easy to apply to other groupings that you have so one could quickly
note does that pull out any differences.
So I think that is an important component that I only see in this — in
this particular section and I would hope that this is something that we could
review in the future.
R.
HARRISON: Well, for the sake of lucidity then, why
don’t we go ahead, and do the presentation and discuss it because there are
points to be made? And otherwise,
they’re going to be sitting here isolated.
And when the minutes come out and Mike has to review them, it won’t make
sense. So Joel, let's go ahead, but I
agree with moving the prostate so that all the cancers are together. Okay.
Cancer in Air Force Veterans of the
Vietnam War
J.
MICHALEK: All right.
Thank you. So we’re now going to
talk about the paper.
M.
GOUGH: What's the journal again,
please?
J.
MICHALEK: The Journal of
Occupational and Environmental Medicine.
The paper is a summary of these slides.
The paper was submitted in the summer of 2003. It was peer reviewed up to October of 2003; it was accepted for
publication. It was edited to — in our
response to the referees. It was — has
been typeset and we have edited the typeset version. And that’s been sent to the publisher and it will be published on
the 11th of February. The
paper has been selected by the journal for CME training, which means that the
front end of the paper has CME learning goals printed right into the first
page. I believe that means that any
physician who’s taking CME training will be possibly assigned this paper to be
studied as an example of good epidemiology as remarked by the — by the editor.
M.
STOTO: Isn’t this the one that was
rejected by Epidemiology?
J.
MICHALEK: Yes sir, it is.
R.
HARRISON: All comments after the presentation.
J.
MICHALEK: Good point, yes. So these are the authors you’ve seen
before: David Garabrant is at the
University of Michigan. Since then,
we’ve added a new author — Norma Ketchum — on our staff. By the way, probably I should tell you this;
that Fatema Akhtar left the study last year and is now at the University of
Texas Health Science Center at San Antonio.
And when she left, I asked Norma Ketchum, our on-staff statistician, to
check everything that Fatema did — everything.
And she made an independent check of all SAS code and all analyses and
found no mistakes.
I
think that’s an important remark to make because subsequent to that, we
submitted this to the Journal.
It’s important that Norma on staff can reproduce any of these analyses
at any time because she now has all of the SAS code that Fatema wrote under
control. And Fatema wrote a lot of code
in SAS macro language that now Norma is now the on-staff expert in this — in
this paper. So that’s — it well states
the depth of our knowledge on our staff.
We have four statisticians or epidemiologists, which you’ve met already
one — Dr. Pavuk.
And
David Garabrant himself did an independent check. Garabrant is not just the — what you would call an administrative
co-author. He is a co-author; he’s an
M.D. epidemiologist at the University of Michigan-Ann Arbor. And when he co-authors a paper, he wants the
SAS code and the data sets and he makes his own independent checks of
everything we do. And by the way, a
final remark is that I never show anything to you unless it’s been checked and
rechecked many, many times.
The
methods are similar to what you just saw.
It’s the same data set: end of
follow-up is 31 December ‘99; records were confirmed by — outcomes were confirmed
by record review, ICD codes, category definitions. The paper has two parts:
it has contrast with external national rates for U.S. males and internal
analyses by dioxin category stratified by different ways to isolate issues of
interest to veterans: namely, presence
or absence in Vietnam, amount of time spent in Southeast Asia region. Those factors were considered here and the
years in which you were actually in Vietnam were considered. Those were never considered before in our
analyses. To isolate those years where
Agent Orange was most heavily sprayed, those analyses were included in the
study.
The
two-year cut point is important and you’ll learn more about that when I show
the slides responding to Dr. Gough. But
you already have an idea from the previous presentation that there is a trend
in the control group and that two years is about the median years in the
Southeast Asia region among controls.
And I’ll have more to tell you about that. And in a little while, I’ll also give you more information to
tell you why we went to the 100 percent, 0 percent adjustments.
And
here are the covariates; you’ve seen this before for all-site SEER cancer, for
melanoma, prostate cancer. There are
many ways to approach the data to arrive at a two-year cut point. One way is to simply graph the percentage of
time spent in Vietnam on the vertical and the number of years in Southeast
Asia, and you start to see patterns.
And then look at the Ranch Hand group.
What you see if less than two years is that you’re — when you stratify
less than two years, that you’re pulling off a lot of Ranch Handers who were
there 100 percent of the time. The very
upper part of the graph there, there’s a lot of people piled up right there.
And
on the comparison side, it’s sort of the other way around. It’s not as pretty, but there are a lot of
people down here with only a short amount of time in Vietnam — less than two
years in Southeast Asia. And I’ll give
you more justification for that two-year cut-point in a few minutes. So beginning with all Ranch Handers who ever
were compliant, either partially or fully compliant to any of the first five
exams — and those same 1,785 comparisons with the same entrance criteria —
threw away or excluded people who had cancer before or during their tours of
duty and that led to 2,965 people who were available for contrast with external
rates. And then when considering
internal contrast based on dioxin exposure category, we excluded anybody who
had a missing dioxin level, which brought the sample size down to 2,438.
And
this is the same dioxin category definition you’ve seen in many of our
papers. And I won’t dwell on that very
much, except that the data shown here with respect to whites only. And we had to do this racial restriction
because of the way the national rate tables are made. They’re made by racial category — whites, non-whites — and that’s
just the way it is. And so we had to do
these and so we decided to be consistent throughout. In all these slides, we're talking about white people. But we did consider blacks, but many of the
slides are whites only.
And
here is one of the main slides against national rates. We’re finding some increases and some
decreases. In cancer of the genitalia,
it’s almost identical to cancer of the prostate. There’s only a three-case difference and we’re seeing a
significant increase in cancer of the prostate against national rates and a
significant increase in melanoma.
As
brought out in the latest and final version of the article, one of the referees
suggested that we may be seeing these increases due to the very close scrutiny
that we’re giving these men at the Scripps Clinic in subsequent follow-ups
where they’re being examined by a dermatologist. They’re being examined by a doctor of internal medicine. They’re being given a rectal exam and
PSAs. And they’ve been done — we’ve
been doing this since 1982, although PSAs didn’t come into play until 1992. The first physical occurred in 1982.
And
so they asked us to — the referees asked us to delineate which cancers were a
direct result of the Scripps physicals.
And it turns out that of the melanomas, four of the 17 among the
comparison group were a direct result of the Scripps physical exams and a
certain number that I forgot on prostate were a direct result. And that's specified in the paper now. So the — there’s a possibility that at least
part of the increase we’re seeing here could be due to the close follow-up that
these men received. Just jumping to the
bottom end of the paper in the corresponding picture in the comparison group,
is that we see a significant increase in cancer of the prostate, but no
significant increase in melanoma and decreases elsewhere.
For
example, backing up one slide, we see a significant decrease in cancer of the
digestive system against national rates, but no significant decrease in the
comparison group against national rates.
The paper cuts the data many ways following ideas we received from you
over two years ago. And that's an
important point too; that we just didn’t make this up. We got this by talking to you, and
therefore, departed from our previous analyses which was published in the American
Journal of Epidemiology where none of this was done. And we simply combined all cancers together
against dioxin category.
In
particular, in year of tour, the period — in the period of heaviest spraying,
we see an increase in prostate cancer among Ranch Handers and control were —
controls who were there during that period.
But melanoma, we see an increase in Ranch Hands only during the Agent
Orange period and that’s brought out in the article. And here you have a summary table against national rates done two
ways: once less than two years in SEA
and the other using the 100 percent, 0 percent cut.
And
what you see on the 100 percent, 0 percent is the SIR; it’s the standardized
incidence ratios are increased even though we have decreased the sample size
because of our cut. And, you know, this
melanoma that we saw before is now a relative risk of — or SIR of 3.05. And the same thing occurs with prostate
cancer, although the significance isn’t as great. The — numerically, the SIRs are increased. And here's the dioxin category analysis,
which is displayed in the paper, which is among those who were there less than two
years. And the analysis on two years
was done because — less than two years, greater than two years — because we
found that years in SEA was statistically a confounder and we adjust for
confounding by stratification.
And
by stratifying here on the median years spent in SEA among controls, we have
this significant pattern in the Ranch Hand group with increased risk with
increased dioxin body burden. The
relative risk of 2.0 in the high category reached statistical significance as
did the relative risk in the low category on all-site SEER cancers. And on the 100 percent, 0 percent analysis,
we see increased numerically larger relative risk, smaller sample sizes that
reached statistical significance on all-site SEER cancers.
So
in conclusion — you've seen this summary before; I won't read it to you — that
we see increases against national rates on prostate and melanoma. And in internal analyses after adjustment
for this confounder, we see increases in all-site SEER cancer, prostate and
melanoma. And then we see the same
thing on the 100 percent, 0 percent, but the significance is restricted to
all-site cancer. Now I’d like to move
to — well, unless you have questions. I
know there’s a series of questions I answered for Dr. Gough and they come on
another slide show that I can show in a few minutes.
R.
HARRISON: Go ahead, Mike.
M.
GOUGH: I just have some specific
questions. Yeah, I guess the first one
is on the “melanoma by exposure opportunity category,” which is your ...
J.
MICHALEK: The 100 percent, 0
percent in the less than two years?
Yes?
M.
GOUGH: That — no, you’ve gone too
far.
J.
MICHALEK: It was melanoma.
M.
GOUGH: Oh, that’s it. Yeah.
In the 100 percent of SEA versus Vietnam — I mean, 100 percent of SEA,
is that restricted to men who were there for less than two years?
J.
MICHALEK: No.
M.
GOUGH: So that’s all years?
J.
MICHALEK: This is all years. The only restriction on the Ranch Handers
was they had to have 100 percent of their SEA service in Vietnam.
M.
GOUGH: Okay.
J.
MICHALEK: And this is all
comparisons who spent no time whatsoever in Vietnam. And notice among those comparisons, we observe two and we expect
two ...
M.
GOUGH: Yeah.
J.
MICHALEK: ... a relative risk of
almost one. In the Ranch Hand group,
it’s three times.
M.
GOUGH: Right, I see that. The other — then in your conclusion in the
internal contrast ...
J.
MICHALEK: Well, that’s
different. Internal contrast, right
here.
M.
GOUGH: Yeah. Yeah.
I’m just going to ask you this question. Internal — the first thing it says is “less than or equal to two
years in SEA” and then “high category.”
That means that’s where you compared just within the Ranch Hands?
J.
MICHALEK: Now we’re asking — we’re
giving the restriction that they were there less than two years.
M.
GOUGH: Yeah.
J.
MICHALEK: Is there a difference
between the high category and comparisons?
And the answer is statistically yes ...
M.
GOUGH: Okay.
J.
MICHALEK: ... by a relative risk of
2.0.
M.
GOUGH: Okay. Thanks.
J.
MICHALEK: Okay.
R.
HARRISON: Anyone else?
M.
STOTO: On that one there, is there
a p value or something?
J.
MICHALEK: There is in the
paper. I didn’t put it on here.
M.
STOTO: Oh, so red — the red — the
red ...
J.
MICHALEK: Red means significant.
M.
STOTO: Okay.
J.
MICHALEK: And the p value and
confidence interval are in the paper.
M.
STOTO: Okay.
J.
MICHALEK: I have the paper actually
right here.
M.
STOTO: No, that’s okay. I just — I was looking at this and my copy
doesn’t have red.
J.
MICHALEK: Right. Sorry.
R.
HARRISON: Moving right along.
TCDD and Prostate Cancer in U.S. Air
Force Veterans of the Vietnam War
J.
MICHALEK: Okay.
Connected with that, we have a separate paper strictly on prostate
cancer, which is in submission to a journal, in fact, and which has been — has
been presented. I wanted to show you
those slides right here. First author
is Marian Pavuk, myself and then Arnold Schecter. This presentation considers prostate cancer in more detail.
M.
GOUGH: Where is it?
J.
MICHALEK: This is the very — near
the very end.
R.
TREWYN: Page 18.
M.
GOUGH: Okay. Thanks.
J.
MICHALEK: Okay. Just a little background — these slides were
made by Dr. Marian Pavuk. In the first
few bullets there, you’ll see prostate cancer is common in older men, second
only to lung cancer. TCDD has been
classified as carcinogenic in animals and humans. And the National Academy has indicated “suggestive/insufficient”
evidence to declare TCDD associated with prostate cancer. The study goal here is to investigate in
more detail an association to TCDD and dioxin in this study.
Now
here we have some data up through the year 2002. Dioxin is measured through 1997, which is all — which is true for
the whole study. And Marian, was this
on the first five physicals only, fully compliant? Yeah. So this is the same
entrance criteria we had on the previous talk and diagnoses up to 31 December
1999. And now we have introduced a
latency requirement that they had to have at least twenty years follow-up after
the last — after their qualifying tour.
And so cancers that occurred earlier than twenty years after a tour are
not counted and only individuals who had at least twenty years of follow-up are
counted. This is now more stringent
than what you saw in the previous talk.
In
addition, we did not simply use the measured dioxin level. We computed area under the curve, which is
sort of a favorite metric in animal studies, which is a metric which is
proportional to the bioavailable dose.
And we did that using the latest information on dioxin half-life — 7.6
years. And so now he’s breaking that
measure out into tertiles on the Ranch Hand group into low, medium, high, and
comparing each of those with the comparisons, and once again, excluding
individuals with less than twenty years of follow-up. And all of the analyses are done using the time to onset flag — a
time to onset as well as a flag for cancer using a Cox or proportional hazards
models.
And
so here you see the total sample size and the exclusions, which are necessarily
tied to dioxin and less than twenty years of follow-up. We lost twenty comparisons and eight Ranch
Handers who did not have twenty years of follow-up in other words. And we restricted to white individuals to
remain consistent with previous papers.
And here’s a display of the demographics by Ranch control and you see
the good concordance on age and other factors that were matched — that are
matching factors for this study.
BMI
is similar; time in Vietnam is not similar.
The Ranch, of course, the comparisons spent less time in Vietnam than
the Ranch Handers did. Time in SEA, the
other way around; the comparisons spent more time in the region than the Ranch
Handers did on the average — on the median.
Pack-years of smoking — no remarkable differences there or on alcohol
use. And this is more demographics on
the dioxin levels. You see, of course,
increased dioxin levels among the high area under the curve category: 59 ppt on the mean; dioxin in the high area
under the curve category, which is expected because that’s how you compute.
And
down here, you see the actual area under the curve — 2,499; ppt-years is the
unit — ppt-years. And the range — and
the full range down here after 24,569 ppt-years, which is about — it’s a
fraction of what the chemical workers received in the NIOSH study and I think
it’s like less than 50 percent — maybe less than 20 percent than what the
chemical workers received in the United States who made herbicides. And here’s a picture of the cumulative TCDD level
by area under the curve category and, of course, they’re going up. Covariate associations with prostate cancer
— nothing remarkable here except time in SEA is a clue that there’s something
happening with years in Southeast Asia, which you already know about because I
showed you the other talk.
It
turns out that, in fact, yes; there is a correlation in the control group
between years in Southeast Asia and prostate cancer and that’s what this
analysis is pulling out — this 1.5. And
so, and here you see another representation than what you saw on the previous
talk. In the comparisons and Ranch
Handers, you see in the comparison group a significant increased trend of
increased risk with years in the region, whereas, not so in the Ranch Hand
group.
This
is the point — this fairly subtle point — of the published paper; that the less
than two-year cut point has to do with trends we see in the comparison group
and not in the Ranch Hand group. It
appears counter-intuitive to the — to the — to the reader that we should see
increased risk in the less than two-year stratum because people would think,
“Gosh, more than two years is where you should see the effect, right?” Well, the reason we’re seeing that is that
the trend in the control group is driving that cut point — not the Ranch Hand
group.
And
here you see that the area under the curve categories are unrelated to dioxin
level. I’m sorry; the area under the
curve in the comparison group is unrelated to years in — years in Vietnam,
whereas, in the Ranch Hand group, they are, which is consistent with the idea
of an exposure experience in the Ranch Hand group, but not in the comparison
group in Vietnam.
M.
GOUGH: Joel, that is days in
Vietnam for the comparison, not days in SEA?
J.
MICHALEK: No, it’s days in
Vietnam. Separately, if you — if we
were to do a slide on days in Southeast Asia, you would see roughly the same
picture. This is just another look at
the data. And so here it is: the overall analyses without adjustment for
that confounder and what you see is pretty much nothing. You see a flat picture of relative risk that
are all near 1.0. So without regard to
the confounding, you are led to conclude that there’s nothing happening in this
data.
However,
if you adjust for the confounder by carrying out the stratification, you find
that among those that were there less than two years, you see a significant
increase of the risk of prostate cancer in the high area under the curve
category relative to comparisons and you see a significant trend right
here. This AUC column is a — indicates
the change in risk with a unit change in area under the curve and that reaches
significance too. But you don’t see
that on the greater than two years. And
intuitively the reason for that is that once you go to the greater than
two-year category, now you're bringing in those comparisons that are already
experiencing increased risk associated with years in Southeast Asia, which I
don’t know why that is, but that’s what’s happening. And so as soon as you do that, you lose your adjustment and you
lose your effect.
So
what we’ve got is that without accommodating this confounder, we see no
association. But after accounting for
it, we see an association and that’s what’s new about it because previous
publications — the article published in the American Journal of Epidemiology
1998 — does not include adjustment for this confounder. And so you are led to conclude based on that
paper there is no association, but now we do see one because we have discovered
this confounder. And the reason we
discovered it is through our discussions with you.
And
this shows the — more characterizations of the prostate cancer cases looking
only at those 47 comparisons who did have prostate cancer more than twenty
years after SEA in the 33 Ranch Handers by age and surgical procedures that
were used to treat their prostate cancer.
Finally, we considered the Gleason scores that were extracted from
medical records and found no connection between Gleason scores and the — is
there a cite for that? No; between
Gleason scores and the area under the curve categories.
R.
TREWYN: What are Gleason scores?
J.
MICHALEK: Gleason scores are a
medical determination by the — Marian, can you tell us briefly about Gleason
scores? Microphone. I’m sorry it’s not here.
M.
PAVUK: A Gleason score is a — is a
measure. It’s made from a pathological
examination of prostate tissue obtained either on biopsy or surgery, preferably
on biopsy at the time of making a diagnosis.
And it’s a measure of a — of a — of a stage of prostate cancer, one of —
one of several and it’s a combined score.
There are at least — it combines — at least two areas have to be
evaluated on a pathological slide and maximal — it’s a scale from 0 to 5 and
the minimal level is prostate cancers that have scored about 7. You can have two scores from 1 to 5 and then
it’s combined for a total Gleason score.
And those with a level above 7 have a worst prognosis for survival, but
we didn't see in this data an association with dioxin levels.
J.
MICHALEK: That ends this particular
presentation, however, I still have slides to respond to Dr. Gough.
M.
STOTO: Three quick comments: one is that I think this is great detective
work separating out, you know, that — those different groups and seeing what
was going on. Second, in the background
slide, the term that the IOM uses is “limited/suggestive,” not the words that
you have here to describe prostate cancer.
The third one is a question; it’s about using the area under the
curve. If you use — if you assume
everybody has the same half-life, does area under the curve just become
proportional to their ...
J.
MICHALEK: Yes.
M.
STOTO: ... their measured level?
J.
MICHALEK: If everyone had the same
length of follow-up.
M.
STOTO: So but length of follow-up,
does ...
J.
MICHALEK: Yeah. Right.
M.
STOTO: ... it differ with length of
follow-up? Okay.
J.
MICHALEK: Right. But we — but we — what we did was we
accommodated this twenty-year. Explain
the area under the curve calculation for me, please.
M.
PAVUK: Well, just to answer your
question maybe: first, no; it’s not
totally proportional to their dioxin level because dioxin level as measured is
just a number at a certain point of time — their actual level. Area under the curve reflects how long — how
long they’ve been followed up and how long from the exposure this also
happened. So it approximates their body
burden a little better than ...
J.
MICHALEK: Cumulative body burden.
M.
PAVUK: ... cumulative body burden
instead of just one measurement at one point of time, but it’s an estimate
still.
M.
STOTO: If they all had the same
amount of follow-up, then the two would be proportional, but they do have
different levels.
M.
PAVUK: Correct. I mean, at this point, there are developments
— ongoing ones that would allow us, you know, to use not just the average
half-life and accommodate other functions as, you know, hepatic metabolism or
other excretions of dioxin to make better estimations of area under the curve
or body burden, or even better, the, you know, the initial levels at the — at
the end of the exposure in Southeast Asia.
M.
STOTO: Well, one of the — one of
the ongoing themes that we've had here is whether to assume that the — that the
half-life is constant.
M.
PAVUK: Well, I think it becomes
very clear that it is not constant over all — over the whole period of time;
that the half-life is shorter at the very beginning of the exposure, especially
of high exposure, and then the half-life becomes longer. However, still using some half-life or
having an estimate of half-life approximates a little better the estimated
levels than not having any half-life at all.
M.
STOTO: Well, what I — what I — what
I had in mind is that variation across an individual’s half-life and in particular,
for something, you know, that the guys with more body fat are likely to have a
longer half-life than I think we’re seeing.
M.
PAVUK: Yes, you’re ...
M.
STOTO: If that were — if that were
a risk factor for prostate cancer, then it would be important, but I think that
...
J.
MICHALEK: Right, it's not.
M.
STOTO: ... it’s not.
J.
MICHALEK: Actually, you’re alluding
to and you’re opening up the idea of giving up these assumptions and we’re on a
— right on the threshold of doing that.
We are co-authoring an effort with EPA to introduce
physiologically-based pharmacokinetic modeling to accommodate changes in
half-life with time, and with lipid content and body fat. And that’s under the works right now. There's a paper in submission to the Lancet
on our new pharmacokinetic modeling of half-life in the Ranch Hand cohort. And subsequent to that, we’ll develop a new
initial dose estimate based on the very latest modeling, which will then
preempt whatever you’re seeing here today.
R.
HARRISON: Just one comment: I've never understood how you are likely to extrapolate back to
the initial dose without having data from the initial period because all of the
measurements that you’re basing the — your new half-life study on comes long
after that fast initial slope is gone.
J.
MICHALEK: True. In fact, that data was presented to the
Committee perhaps a year ago when we combined Ranch Hand and Seveso data into a
single analysis. We had something like
forty Seveso males, adults who were exposed during the Seveso accident in
1976. We combined that with the 300 or
so Ranch Hand veterans in the — in this study who had repeated dioxin
measurements. And when you put the data
together, you find in log units a linear pattern way back to within three
months of the initial exposure.
And
before that, you see a non-linear decrease, a very rapid decrease which Marian
also already alluded to, which is highly significant. In other words, you do see a non-linear pattern real close to the
initial dose. And so our analysis —
this analysis you’re seeing here today does not accommodate that non-linearity;
however, new analyses that are coming in the next few months will.
R.
HARRISON: If I — if I recall things correctly, what
you see is in the initial curve is a — what happens with a lipid soluble
compound when you essentially inject it intravenously; and that is, a lot of it
gets cleared real quickly because it's not water soluble and then a lot of it
gets stored ...
J.
MICHALEK: In the — in the liver.
R.
HARRISON: ... mostly in fat and in cell membrane.
J.
MICHALEK: In the liver too.
R.
HARRISON: So then the — that second
half-life or that longer half-life then is really more dependent on body shape
almost than it is anything else and ...
J.
MICHALEK: The seven-year half-life
...
R.
HARRISON: ... I’ve still never
understood how you — I guess the point I'm making is that if you were — if you
— if you had enough fat to have a dioxin level of 100, the peak dioxin level
could be 100 or anything above 100 and anything after the three-month mark will
look the same. So you — it's hard to
use that data to go back and decide what the peak level was, which is ...
J.
MICHALEK: I think I'll have
something to show you later this year perhaps on that in more detail.
R.
HARRISON: Not me, but that’s
okay. All right. Yes, Dr. Leffingwell?
S.
LEFFINGWELL: I had a couple of
questions for orientation. In your
twenty-year exclusion, was that from first or last exposure?
J.
MICHALEK: That was from their
qualifying tour, which is the tour that was used to decide eligibility for the
study.
S.
LEFFINGWELL: And do you remember
about how many were excluded on that basis?
J.
MICHALEK: Very few; in fact, I have
that slide right here: twenty
comparisons. And of those five cases in
eight Ranch Handers, of those were two cases of prostate cancer were excluded
based on the twenty-year follow-up. So
I’m near the bottom of the slide right here — these numbers.
R.
HARRISON: Let's do — we're, I believe, one, two, three
presentations short of where we're scheduled to be before we break for
lunch. I think at least one of us has a
fairly early plane this afternoon. And
so what I would suggest is that we do one more presentation, take approximately
ten minutes to gather our lunch material, and then munch as Joel continues his
presentation. He looks well fed and can
probably last until that time without any undue harm. Is there any objection from the Committee or from any of the
participants at our doing that? One
more presentation, ten minutes to gather food and then we just keep chugging on
through. Okay, Joel.
Adjustments for Ethnicity and Cut
Points
J.
MICHALEK: A few months ago, Dr. Gough wrote us a
letter with some questions about the slides you just saw and this — these slides
respond to two points that he made that I thought were really very good and
instructive. In particular, why — what
happens to the melanoma analysis when we adjust for ethnicity? The idea being that Scandinavians — the
blue-eyed, fair-skinned Danish or Swedish individuals — have a much higher risk
of melanoma than individuals with darker skin.
And what happens there if you tried to accommodate this ethnicity?
So
we have one of our physical exams, we asked people what they thought of
themselves regarding their ethnicity.
And this is a quagmire of issues about the mother and the father, and
what are they and what are you, you know?
And you can’t probe this very far if we start running into some serious
issues. But we simply took at face
value what these people said about themselves and so we produced a single
covariate. It was, “Are you
Scandinavian or are you not — yes or no?”
And to approach this data, we are going to again, use the same methods
and data sets you just saw: the 2,965
people and we're going to be considering the same melanoma analysis one more
time.
And
so here are the results of ethnicity on the 2,965 broken out by Ranch Hand by
these — by these categories:
comparison, background, low, high.
And here, I’ve highlighted Scandinavian — people who declare themselves
as Scandinavian; about 5 percent across the board think of themselves as
Scandinavian. So I created a new
covariate, which was a 1 if you're Scandinavian; it’s a 0 otherwise. And we introduced that covariate into the melanoma
model.
And
there — well, you might as well stare at this slide. I mean, you can see that quite a large percentage for some reason
are Irish; I don’t know why. But 15
percent to 20 percent of this cohort are Irish. They think of themselves as Irish. And there are the rest of the ethnicities here on the next slide,
small percents, and a fairly large percent consider themselves French; and 10
percent or so American Indian; and no response, 12 percent of the comparisons
didn’t want to answer the question at all; and all the Ranch Handers
answered. I don’t know why that is
either.
So
and here is, again, I just again reproduced the same table that’s in the
article just for your reference on melanoma — less than two years in SEA. So what we’re going to do is take this same
model and without any changes, just introduce that single covariate on less
than two years. And when you do, you
find it’s a continued significance. In
other words, the modeling is not disrupted and the conclusions are not changed
by introducing this additional covariate.
Now I have to say as a statistician, I really don't like pushing the
data this far because these are small counts.
They’re small counts and so I only answered your question, which is what
happens if you introduce ethnicity and that’s what happens.
M.
GOUGH: Actually, I asked
specifically about Kelts because John Nichols — who was the Director of
IARC? John ...
J.
MICHALEK: Okay.
M.
GOUGH: I can’t remember his
name. Anyway, he published a paper
showing ...
J.
MICHALEK: On Kelts? I’m sorry.
I thought you meant ...
M.
GOUGH: ... showing Keltics ...
J.
MICHALEK: Okay.
M.
GOUGH: ... people from the British
Isles. If you go to Australia and South
Africa where those people went, skin cancer rates are — or melanoma rates are
very high. So ...
J.
MICHALEK: Well, maybe then Irish
would be closer to what you were talking about because we don’t have Keltics.
M.
GOUGH: Well, I think English and
Welsh. I don’t — no; I said I don’t
know how you define them either.
J.
MICHALEK: I don’t either.
M.
GOUGH: But yeah.
J.
MICHALEK: Okay.
M.
GOUGH: Well, I'm glad you did this
though. I think that you’ve provided a
great service for somebody who wants to look at it because I — because there is
such a tendency. I mean — I mean, my —
I'm a dermatologist's dream as are several other people around this room and I
suspect you are too, close to it. So
anyway, I — but I appreciate you doing that and I appreciate you digging that
out. And I know what a — that you can’t
rely with a great deal of confidence on what people think they are.
J.
MICHALEK: Yes sir, that’s true.
M.
STOTO: Two quick things: one is that this is the way it’s done in the
census and almost every other way about ethnicity.
M.
GOUGH: Unless you’re Tiger Woods
and you’re mixed.
M.
STOTO: Right, and then — so this is
as good as you can do.
M.
GOUGH: Yeah.
M.
STOTO: The other thought is that if
you want to pursue this, one way to — you might be able to do this is to look
at national rates and try to find the countries that had the relative high
rates of ...
J.
MICHALEK: Melanoma.
M.
STOTO: ... melanoma. And yeah; I would imagine they would be
English, Scottish, Irish and Scandinavian out of this list and look at them as
a group rather than just the Scandinavians.
J.
MICHALEK: Okay.
R.
HARRISON: Yes sir?
S.
LEFFINGWELL: The problem with
looking at country data would be there’d be a latitude variant, which is why
their skin color differs too.
M.
GOUGH: Well, that's true.
M.
STOTO: Yes. I’m just — I’m just trying to find a
somewhat objective way of dividing the ethnicities into ones that are prone to
melanomas and others not.
P.
CAMACHO: Well, you were the one who
brought up the eye color and all of that stuff. I mean, isn’t that more ability to — of those eye color and all
that, are they kind of indices? All
right, then so divide people by ...
J.
MICHALEK: We’ve done that already.
P.
CAMACHO: All of — all of us with
blue eyes get over in a corner.
J.
MICHALEK: Yeah.
P.
CAMACHO: Isn’t that pretty
objective?
J.
MICHALEK: The analyses ...
R.
HARRISON: I'd prefer you move ...
J.
MICHALEK: ... of the blue eyes ...
R.
HARRISON: I’d prefer you move a
little further.
J.
MICHALEK: Yeah. The analyses in the paper already adjusted
for eye color and skin reaction to sunlight.
M.
GOUGH: You know, I wasn't really —
perhaps I didn’t read the slides in the last presentation correctly, but I
wasn’t aware that you did adjust for skin color. So I think that’s ...
J.
MICHALEK: Skin reaction to
sunlight. It’s not ...
M.
GOUGH: Whatever — I think that’s
fine.
J.
MICHALEK: Okay.
R.
HARRISON: The one possible explanation, by the way,
for the high Irish count might be that the largest St. Patrick’s Day
celebration in the United States is in Savannah, Georgia. The South was populated by English that
lived at the English-Scottish Border, the most violent part of England back
during the 1700s. This is in a book
called Albion Sea. And when they
left England, they want to two places:
they want to Northern Ireland and the American South. That's why the — that’s why the North
would've lost if the South had had just a little more — little more capital
behind it. It’s, and still to this day,
it’s probably the most violent part of the country and just — it explains a lot
about Northern Ireland.
J.
MICHALEK: All right. The other part of our talk regarding Dr.
Gough’s comments is this issue of the two-year cut point. And so we want to see what happens when we
play with that cut point — change it.
So here are the demographics according to a quartile of years in SEA —
and you already saw these quartiles in the earlier presentation — to show you
that there is a trend on year of birth in the comparison group and a slight
trend on pack-years. And in the Ranch
Hand group, there’s of course the change too.
And the individuals who spent a long time there are relatively older.
And
I guess this is the slide that is probably our best indicator of why the
two-year cut point is important as regarding adjusting for this
confounder. If you look at up to two
years — up to 2.1 years in SEA, you’re talking about 6 percent of the — of the
comparisons are experiencing cancer; whereas, up here, it’s 12 to 16
percent. If you look down here at the
Ranch Hand group, it’s fairly flat as compared to years in SEA as relative to
the Ranch — comparison group. So what’s
going on is there’s a dynamic going on in the comparison group that we don’t
understand and no corresponding dynamic in the Ranch Hand group.
But
what you do have in those down here is a — is an increased risk of cancer in
the comparison group among those who were there less than two years. And that’s just a reflection of the — of the
data that is driven by a statistical analyses that I’ll tell about — talk to
you about in a few minutes. And here’s
a display of the percentage of time spent in Vietnam, a percentage of their SEA
service actually spent in Vietnam according to years in SEA. And what you’ve got here is an indicator of
why that would turn out to be an important confounder too because among those
Ranch Handers who were there less than two years, you tend to pull off people
who were there a lot.
They
were — there are — nearly 100 percent of their Southeast Asia tour was in Vietnam,
whereas, not so in the Ranch — in the comparison group. It’s pretty flat around 20 percent. And so the years spent in SEA may itself be
a surrogate for something else: namely,
how much time did you spend in Vietnam?
That's what this slide says and that’s one of the reasons why we used
the 100 percent, 0 percent approach in the paper. I have to remark that I expected the referees to tell us to dump
one of those two approaches, but they left both in the paper.
Now
here is the table and this is the summary of a so-called interaction model,
which is critical to our decision to cut the years in SEA. There is a, in the interaction model, a
significant years by dioxin by group interaction — p value of 0.05. And to a statistician, what this says is that
the relationship between dioxin and cancer is changing. It’s changing; it’s not constant. It varies with years in SEA and group. And that says, “You’d better do something
about this.” It says that the straight
main effects model, which is the basis for most of our paper, is not working
here; that there is a change in the relative risk. And the main effects model, which would be the basis for sort of
the naive approach to the analysis, is not appropriate. And so ...
M.
STOTO: Group is?
J.
MICHALEK: Ranch Hand and control.
R.
HARRISON: But Joel, didn’t a couple of slides ago,
looking at the Ranch Hands as being flat across time and the comparisons
climbing with time, what that suggests is that the dioxin exposure is producing
cancers earlier, which would mean that you would not want to do a cancer
comparison of people after — I'm not sure how to think about this. But it means that you should see a greater
difference between Ranch Handers and comparisons if you look after only five
years of exposure versus — five years of latency versus fifteen years of
latency; that the difference should be greatest at time 0 and less at fifteen
years.
M.
STOTO: I think you may be confusing
years of follow-up versus years of service.
R.
HARRISON: No, I was — I was
thinking of latency. I was thinking
that earlier — I mean, what he’s been saying is ...
M.
STOTO: No, I — but this year here
refers to years of service.
J.
MICHALEK: Years of service in
SEA. Latency is not ...
R.
HARRISON: No, all I’m responding to on this slide is
Joel’s statement that there’s a change with time; that’s all I was responding
to and it made me think — made me think about the previous slide. When I was looking at it, it said to me that
the change in time was that Ranch Handers get cancer right away and the
comparisons have to wait for two — not get cancer, but the years in South — it
...
P.
CAMACHO: The Ranch Handers did one
thing.
R.
HARRISON: One may be an effect of just aging like in
the comparisons. As you get older, you
get more cancer. In the Ranch Handers,
the aging one picks up there because if you’ve — but maybe I am confused.
R.
TREWYN: Do we know on this
one? I mean, it looks like we’re seeing
the same trend down there. I mean, I
assume death was considered. It isn’t
that out at the 3.6 to 14.5, they were just dead? I ...
J.
MICHALEK: No sir.
R.
TREWYN: Okay.
P.
CAMACHO: But isn’t one of the
presumptions the Ranch Handers constantly handled this material versus the
comparison group in SEA? I walk into a
certain sector in the DMZ where there was defoliant spray and I'm doing a
three-week operation there. I walk out
of that and now I’m down below Group 9, you know. I mean, I walk in ...
R.
HARRISON: Yeah.
It's almost like a dose-response.
P.
CAMACHO: Yeah. Well, I mean, I don’t think ...
R.
TREWYN: I noticed this same thing
and I was trying to figure out where that ...
P.
CAMACHO: It seems to me that ...
R.
TREWYN: Paul, I just want to
address you. I think most of it is, as
far as exposure is, anybody who was in Vietnam in a base camp, they sprayed the
“be-jeebers” out of every one of those base camps with herbicides. There wasn’t a blade of grass growing
anywhere and so if you were in there — now whether it was Agent Orange versus
one of the others that didn’t have dioxin in it, so I don't think it was
walking through the DMZ.
J.
MICHALEK: Correct.
R.
HARRISON: Well, while everybody’s running down on one
thing, Joel, there’s one other last thing.
A comment that was made earlier about self-reported — I mean, this is
not on this paper; this was on a previous paper. But I wrote it down, then I forgot to mention it; it was
self-reporting exposure to venereal diseases.
And at least when I was in the Navy, you got ...
P.
CAMACHO: Yeah, you got ...
R.
HARRISON: ... you got punished if
you got a venereal disease. So if you
were on good relations, if you had good relations with the medical staff, you
got treated, but you didn’t get reported.
So I don't think self-reporting is a reliable — that's one instance
where self-reporting is not a reliable tool.
P.
CAMACHO: Well, moving right along.
J.
MICHALEK: All right, moving along —
I just wanted to point out this slide has nothing to do with dioxin. It only relates to years in SEA with cancer
and group.
M.
GOUGH: That includes basal?
J.
MICHALEK: This is SEER cancer; this
does not include basal and squamous.
J.
MICHALEK: And this is simply a
reflection that the relation between dioxin and SEER cancer changes with years
in group. Now I’m coming down to Dr.
Gough’s comments. Actually what we’d
like to see, different year categories instead of the old quartiles, which are
these sort of odd-ball decimal years, lets you see a nice progression: 0 to 2, 2 to 4, 4 to 15. So we did that. And of course too, it approximates the same cut point you saw on
the previous slide and so you see the same pattern once again of reduced. Here you see reduced cancer in the
comparison group less than two years:
right there — 5.9 percent.
And
so we go to these nice round digital cut points like this. You’re going to see right here; this is the
old analysis published in the paper — less than two years and you see almost
identical results. We’re now using a
two-year cut point. If you go 2 to 4,
it goes away. So what’s going on is
that if you change your cut point, you give up the adjustment. The adjustment was the stratification. When you give up the adjustment, you lose
the effect and that’s the whole purpose of the adjustment. And 4 to 15, it further goes away.
Once
again to illustrate the effect or the attenuation of the adjustment, if you —
if you decide, “I don’t like two years; I’m going to use three years,” you
again attenuate the adjustment because now what you’re doing, you get 7.1
percent in the comparison group down here.
And the reason you get a higher percent is that now you’re pulling in
comparisons who were in that upper strata and you’re inflating your cancer rate
in the comparison group. And so that’s
going to make it harder to detect a treatment effect in the Ranch Hand group
and sure enough, that’s what happens.
When
you increase the cut point, you lose significance in this category, which is
still maintained. And if you go greater
than three years, you lose it altogether.
When it’s greater than three years, it’s totally gone. So the conclusions are that years in SEA is
a confounder; not only that, it induces an interaction. And that’s the second bullet. And we also see, as we’ve seen many times
now, the — in the comparison group, there’s an increased risk of cancer with
increased years in SEA. And following
the book, the textbooks, we adjust for confounding by stratification.
And
if we decide we’re going to change that cut point and increase it, what goes on
then is that we attenuate the adjustment.
And if you keep increasing it, there’s no adjustment anymore; you lost
it. And the bottom line is that years
in SEA is a surrogate for something else in the comparison group that we don’t
know what it is. But years in SEA is certainly
an obvious risk factor for cancer, but it’s a surrogate for something we don’t
understand in the comparison group.
Thank you very much. That ends
our cancer presentations.
R.
HARRISON: Any further questions
before we take our ten-minute break to raid the food bar? Ten minutes.
[LUNCH 12:20 P.M. - 12:34 P.M.]
Dioxin and Memory Loss
J.
MICHALEK: We’ve looked at this confounder — years in
SEA — on diabetes and on birth defects and we see no confounding. So what we’re seeing — and whatever is
happening with Southeast Asia is happening on cancer so far only and not on
diabetes or birth defects. Once we saw
this, of course, then we wondered, “Well, what happened with all the rest of
the things we’ve studied over the last twenty years?” And number one is diabetes and birth defects, and it was
fortunate that somehow, years in SEA seems to be specific to cancer and not to
endocrine-related conditions or diabetes — birth defects.
So
the study is unique in many ways. In
one way is that now we have paired Wechsler Memory Scale on 2,000 people spaced
twenty years apart. We measured their
short- and long-term memory at baseline in 1982 with the Wechsler Memory Scale
and we also measured it again in 2002.
And we have published that there is an adverse association between
memory loss and dioxin the Ranch Hand Group.
And that was published in Neurotoxicology last — in the year
2002, which was about a year and a half ago.
And
I guess I had previously presented this to the Committee, yeah. Drue Barrett is at CDC — she’s our
psychologist co-author — and John Cary, Scripps Clinic. And this talk is very preliminary because
Barrett and Cary haven’t even seen this yet.
And it’s also preliminary because I don’t have a full data set. At the time I made these slides, I didn’t
have a full data set of covariates, but I put it in because I thought you
should see the latest information.
And
there’s the hypothesis; that dioxin and memory loss are adversely associated
with — in the Ranch Hand group. And
that hypothesis is driven by information from animal and human studies. In particular, children exposed to PCBs from
mother’s breast milk show intellectual deficits against children who were not
exposed. And they also see those things
in animals; we see cognitive deficits in animals exposed to organochlorine
chemicals.
And
there you have the summary of the Wechsler Memory Scale as administered in both
1982 and 2002. And these are the
components: logical memory, immediate
and delayed; associate learning; visual reproduction, immediate and
delayed. Logical memory is derived
from, as I remember, a recall of a paragraph that’s read by the — by the
psychologist. And you’re asked to recall
components of the little story, one of which has to do with activities in New
York City, and the other one, I forgot.
Associate
learning is a — is a drill: opposites,
for example. She’ll say “high;” you say
“low.” And you have to do that in a
rapid sequence and she scores you. And
visual reproduction, you are given a pattern and very quickly, you have to
reproduce it on pencil and paper. And
your ability to do that is scored, whether you can get the components of this
pattern properly. And immediate and
delayed means in the stories — logical memory — you are asked to recount the
story immediately after it’s read and then again a half hour later.
And
the same thing is true with visual reproduction. They give you a — they flash a pattern in front of you; you’re
asked to reproduce it on pencil and paper.
And then they come back a half hour later and say, “Hey, let’s do that
again.” And then you get — but they
don’t show you the pattern again and you have to do it again. So what we’ve got in the published paper in Neurotoxicology
2002 was a significant deficit in the high exposure — dioxin exposure category
on immediate recall from the Wechsler Memory Scale.
These
scores — these negative indicate that the score is significantly lower among
these Ranch Hand veterans in the highest exposure category as compared to the
comparisons, about a half a point, and that reaches statistical
significance. And the covariates from
the journal article are listed here at the bottom. And those include exposures to psychotropic medications, and
diagnosed psychiatric problems and things like that, and marital status and
drinking. And this covariate data is —
was available to us for our published paper, which has only now been completed
for the current round. And I don’t have
that data yet to show you.
So
this is just a recount of what’s been published and this was already reviewed
by the National Academy of Sciences.
And here you have another significant decrease in the delayed immediate
— delayed logical memory: a deficit in
the highest exposure category, not as strong statistically as the previous
one. And in all other components of the
Wechsler Memory Scale at baseline were non-significant. There was no significant deficits in any of
the Ranch Hand categories on immediate visual reproduction, or on delayed
visual reproduction or on associate learning at baseline. And so that’s in the paper if you want to
see it; it’s out there.
So
now we have the new data on Wechsler Memory Scale given in the year 2002 at
cycle 6. And here I’ve displayed
missing data in the upper panel of the table.
So one person in the background category had missing data probably
because for some reason, he refused to cooperate during the Wechsler Memory
Scale administration. And three to four
comparisons had missing data. And so
the bottom line sample sizes are in the bottom panel and there you see about
1,170 comparisons and then the sample sizes for the background, low, high
category with complete data.
And
here’s what the group looked like at the cycle 6 physical of those who took the
Wechsler Memory Scale, which is nearly everybody. Mean age, about sixty to 65 years; dioxin levels, of course,
increased with dioxin exposure category; BMI is pretty stable around thirty —
around thirty; and high school education is an important indicator of ability
to do the test. And there you see 67
percent have only a high school education over here in the high category. And 34 percent have only a high school
education in the background category, reflecting the fact that many of the
people in the background category are officers and many of the people over here
are enlisted. That correlates highly
with military rank and there you see it down here again.
M.
GOUGH: “High school” means that’s
as far as they went?
J.
MICHALEK: I think Bill Grubbs can
answer that. I got that covariate from
one of Bill Grubbs’ data sets.
W.
GRUBBS: They completed high school.
M.
GOUGH: And went on to college?
W.
GRUBBS: If you’ve got a college
degree like an associate’s, that is acknowledged.
N.
RIVERA: What’s his name?
J.
MICHALEK: Bill Grubbs.
R.
HARRISON: So 34 percent of the background completed
high school or to put it another way ...
J.
MICHALEK: Had only a high school
education.
R.
HARRISON: ... 66 percent of the background did not
finish high school?
J.
MICHALEK: No. No; 34 percent of those in background had
only a high school education. In other
words, about 65 percent had a college education.
R.
HARRISON: You’re saying everybody
had a high school education?
J.
MICHALEK: No, the key word is
“only.” No. In the background category, 30 percent had at most a high school
education and 67 percent in the high category had at most a high school
education. So in other words, the
background category is less educated than the — than the background category.
M.
GOUGH: But the presentation's
incomplete because we don’t know about beyond high school.
R.
HARRISON: Let me ask you this question. How many, in any of those groups, did not
finish high school?
J.
MICHALEK: Oh, we can find that
out. I’m just — when I say “high school,”
I believe that means a high school diploma.
R.
HARRISON: Yeah, but how many did
not get a high school diploma?
J.
MICHALEK: We can find that
out. I don’t have that at my
fingertips, but I can certainly get that for you — meaning a GED degree or
something like that?
R.
HARRISON: No. I mean, just those who didn’t finish high
school.
M.
STOTO: Yeah, because this
essentially groups the people who didn’t finish high school with the people who
got a Ph.D. So this should be cut point
“high school or less” or something like that.
J.
MICHALEK: That’s right; it’s high
school or less versus more than high school — high school or less.
M.
STOTO: But it does mean that? It does?
It ...
W.
GRUBBS: What you just said is
correct.
M.
STOTO: Bill is saying that it does
mean that.
W.
GRUBBS: What you just said is
correct.
M.
STOTO: That it’s high school or
less. Okay. Thank you.
J.
MICHALEK: And by the way, we’ve
looked at this data in the past. It’s
very complicated. There are all — every
possible scenario is realized: all the
way from a little technical, you know, less than high school education all the
way to a Ph.D. It’s all in there. And so we could — we find this definition of
education without any — with some difficulty, but we can do it.
M.
STOTO: I mean, this is what you’d
expect; is that the guys — the technicians, the ground crew would be in the
high category and ...
J.
MICHALEK: Mostly high school.
M.
STOTO: ... less likely to have a —
beyond high school education.
J.
MICHALEK: So what happens if you
repeat the analysis at cycle 6? Now
here, I’m adjusting only for age and education because I don’t have all that
other data available yet. When I made
this slide, I didn’t have psychotropic medications; I didn’t have diagnoses; I
didn’t have any of that. All I had was
age and education, and so, I did the best I could. And in the subsequent weeks, this will all be done over again
using a full set of covariates which is now available, which became available
only a few days before this trip to come here.
And
so we see nothing. We see, actually, an
average score of 5.67 over here in the high category and 5.78; they’re doing
slightly worse, but not significant — a p value of 0.37. So there’s nothing happening at cycle 6 regarding
logical memory immediate as opposed to baseline where we did see
something. And the same thing is true,
logical memory delayed, no significant decrement at all. In fact, in the background category, they’re
doing a little better than the controls: almost significant — a p value of 0.08.
Associate
learning, cycle 6: no significant
findings there either. I did not
analyze visual reproduction in cycle 6 because I didn’t know how to resolve
that with baseline. That’s still
something that has to be worked out with Dr. Cary and Drue Barrett, so that is
yet to come — the analysis of visual reproduction at cycle 6. Now one of the strengths of this data, of
course, is the paired nature of the data.
We have Wechsler administered twice:
once in ‘82 and once in the year 2002.
And
so this final analysis is — the analysis takes both of those measurements into
account and exploits the paired nature of the data: two scores on most of these people spaced twenty years apart. And so this slide summarizes sample sizes
for those people that went to both baseline and cycle 6 and you see some of
them didn’t. And those are the people
up here who — or actually, these are the people with missing data for various
reasons.
The
final sample sizes are down here. There’s
200 people in the high category, for example, that had complete data on
associate learning at both cycle 1 and 6:
846 comparisons with complete data at both cycles. So now we see the demographics among those
people who were fully compliant at both cycle 1 and 6. Another way to put it, these are the
demographics at cycle 6 who were also compliant at baseline and you see almost
the same pattern we saw in the previous demographics. These people are not unusual as compared to everyone at cycle
6. For example, on high school
education, again, it’s around 68 percent of the high category had at most a
high school diploma and about 34 percent in the background had at most a high
school diploma.
So
now we’re going to analyze changes in the Wechsler Memory Scale, which are
three subjects computed as the baseline value minus the cycle 6 value. So positive values are detrimental. You expect people to do worse at cycle 6
because they’re twenty years older than when they took it the first time. And so you see the averages are positive
meaning yes; they are doing worse. For
example, the average is at 1.32 in the high category. That means they dropped 1.32 points on the average for this scale
over the twenty-year period.
And
the comparisons dropped 1.88 over the one-year — twenty-year period; however,
there’s no significant decrement — no significant differences between any of
these. Now however — actually, I’ll
take that back. There is one here on
the high category, but it’s in the — it’s in the positive — it’s in the good
direction. They didn’t — they didn’t
drop as much, so they’re doing better actually than the comparisons on their
change from cycle 1 to cycle 6.
M.
STOTO: Or actually, they — well, I
think what it means is they did worse the previous time.
J.
MICHALEK: Yeah. Right.
That’s a good point. They were
worse at baseline and so this could be a reflection simply of that; that they
were lower at baseline. And remember
that these are unadjusted for those other covariates that have already passed
peer review and weren’t available yet.
And so all of this is going to be done over again pretty soon. And all of the other results on the paired
analyses are negative. There are no
significant results here yet. Here once
again in the high category, you see significance, but it’s in the — it’s in —
not in the detrimental direction. And
associate learning, and here it was not significant.
And
finally just to show you changes from baseline, the Box and Whisker Plot shows
a flat pattern all the way across, which indicate statistically no
dose-response here, here or here. And
so here are the limitations I’ve already talked about. We’re still looking for confounders. We do spend a lot of time doing that on
every study. We have not resolved
visual reproduction data and we will do that.
And then the strengths are — you know all these already. So we have no data to support the hypothesis
that dioxin is detrimental to your memory at this point in time.
R.
TREWYN: Did you look at years in
Southeast Asia and all that stuff?
J.
MICHALEK: We have not checked years
in Southeast Asia on this data yet, but we will. Right now, we’re sensitive to that confounder and we have that
data set available to merge with this and do the check.
M.
STOTO: So, I mean, so that suggests
that the original result may just have been a false-positive?
J.
MICHALEK: Yeah, it may have been
artifactual.
R.
HARRISON: Anything else? Okay. Let's move on.
Dioxin and Syndrome X
J.
MICHALEK: Okay.
The last time I gave a talk on hypertension and that frustrated some of
you — Dr. Osei in particular — that we analyzed just a single endpoint: hypertension — not very interesting. So we expanded that paper to include
metabolic syndrome, which is also called “Syndrome X.” And that is defined in our recent
publication as having three or more of any of the following conditions: a big waist 102 centimeters or more, which
it comes out to about forty inches or more; high triglycerides; low HDL; high
blood pressure; or high glucose or a diagnosis of — by ADA criteria of
diabetes. All of these put together
comprise a pattern called “metabolic syndrome.” In other words, you’re getting basically old, and fat and sick.
R.
TREWYN: The opposite of Generation
X.
J.
MICHALEK: Just like us, yeah. And we’re going to analyze this data — or
these data and we do using the very latest data from cycle 6 at the 2002
physical and 2003 from Scripps. We’re
excluding people with missing dioxins.
You see very few exclusions because we have almost — we have so many
with complete dioxin data now in this study.
The sample size is above 1,000 in the comparison; it’s about 775 Ranch
Hand. Demographics, as you expect,
about 62 years on the average.
Everything there is similar to what you’ve seen on previous slides.
Family
history of diabetes, about 20 percent; family history of hypertension, which
are going to be important covariates, about 40 percent; and about 42 percent of
both groups have Syndrome X — 42 percent, which I thought was remarkable — not
remarkable to those who served as monitors at Scripps Clinic, I guess,
including Karen Fox and Dr. Julie — Colonel Julie Robinson. Here are the dioxin measurements according —
by dioxin exposure category and you’ve seen this — these data before too: dioxin is increased in the high category.
Finally,
here’s the bottom line. We have a
significant increase in the risk of Syndrome X in the high exposure category,
which is consistent with what we saw with hypertension and diabetes: 1.4 — a p value of 0.04. And we have a significant trend, but that’s
not printed. And this is adjusted for
all these variables you see listed at the bottom of this table of the
slide. This work is still in progress
and it’s a puzzle to us as with diabetes; that in the background category, we
have a significant deficit of metabolic syndrome, and at the same time, a
significant increase in the high category — the same pattern we saw with
diabetes.
M.
STOTO: And the same puzzle.
J.
MICHALEK: Same puzzle.
M.
GOUGH: Or the same puzzle is that
if you stratify the comparisons into three dioxin categories, or four or five
dioxin categories, you get the same dose — the same slope of the line. It’s just displaced down toward lower dioxin
levels, isn’t it?
J.
MICHALEK: That’s true. In the — I have those slides.
M.
GOUGH: Which is — which is — I
think it’s one of the oddest puzzles that’s come out of this whole thing.
J.
MICHALEK: Yeah; that we have trends
in the comparison group that are, to put it bluntly, no one wants to see. When we show dioxin data in the comparison
group, we have significant trends that are parallel to the Ranch Hand and only
on the range 0 to 10 ppt.
M.
GOUGH: I know a number of people
who would like to see this.
J.
MICHALEK: Except the journals we
submit those to don’t want to see it; our papers keep getting rejected. You like it and I like it, but anyway for
your reference, here’s the — what happens with hypertension by itself, a
similar pattern. The relative risk is
1.5 in the high category, a p value of 0.01.
Again, we have a deficit in the background category. Now when we further — by the way, I have to
tell you that Syndrome X is not adjusted for body fat index because body fat
index is part of the endpoint. In order
to have Syndrome X, you had to be overweight, hypertensive and diabetic or have
high lipids.
So
we didn’t put BMI in the model because BMI is intrinsically — already is a
dependent variable. And hypertension,
here this model is not adjusted for BMI.
And when you rerun the hypertension model with adjustment for BMI, the
finding is attenuated somewhat. The p
value on the high category is now 0.047, relative risk 1. — I mean, odds ratio
1.4. So putting in BMI attenuates the
result, but doesn't make it go away.
So
here are the usual limitations. There’s
always the possibility of other confounders we haven’t considered. And the same strengths as others and we have
a published case definition. And the
conclusion, you already saw. And here’s
the reference for our case definition of “metabolic syndrome.” It comes from the National Cholesterol
Education Program — NCEP — an NIH publication.
I think that ...
K.
OSEI: You may have one of the
highest rates of Syndrome X in the world because the national numbers look like
25 percent. So you’re talking about 42
percent and you’re talking about predominantly male. And I think probably the advantage you have is the gender and,
you know, it’s very difficult to get all the variables in females where it’s
more, you know, clear cut in men. So I
think having, you know, a group that’s predominantly male is driving your
numbers little higher, which is very good.
J.
MICHALEK: Yeah, which makes the
data set so interesting.
K.
OSEI: Yeah.
J.
ROBINSON: For the last briefing of
the day, it will be, as it says, brief.
As you know, the protocol ends 30 September ‘06. We’ve already heard from Program Management
in regards to the funding. And in the
previous meeting, we talked about the fact that the Air Force had developed a
variety of options ranging from the study once it’s completed in September ‘06
that that would be the end and what the associated cost would be up to doing
another physical exam.
Now
this is something that’s very difficult to read in your handout. And actually what I wanted to reinforce here
is the fact that we in the Air Force are working toward completion of our
protocol. So we have a time-line and
the entire protocol time-line is at the bottom. We’re required to give an IRB final report and we’ve programmed
that in July of ‘06. Up at the top, if
not in the middle, you’ll see that we talk about the Ranch Hand Advisory
Committee meeting again this year at three different months — April, September
and November as we had discussed in March — to do the chapter reviews.
And
so far, I believe you have received at least two chapters for review. In that attachment which you will find, I
will just reiterate is you’ll have a first draft. That was the draft that was sent from SAIC to the Air Force for
initial review. We then put together
another attachment that you’ll also have for each chapter, which is a
resolution comment. So we put in our
comments, and then with SAIC and the Air Force, came up with resolutions to
those comments. So you will have that
so that it — the third attachment is the second draft that incorporates our
comments and the changes as a result of those comments. So you will have all three documents as you
requested.
Additionally,
I provided kind of an instruction sheet, and an explanation of those very
attachments, and included my number that if you had any questions, you could
call me and I would help you with any issues that might arise from your
review. What we’re hoping is that by
the middle of February, you will have at least five of the chapters. We have to do a review of a couple of second
drafts and then I will get them off to you as soon as possible. So I’m hoping that under Dr. Stoto’s
leadership, the meetings will be able to come about as planned.
M.
STOTO: So the two that you sent us yesterday
are essentially samples of the kind of things we’ll be expecting to get?
J.
ROBINSON: Yes, and ...
M.
STOTO: We're not expected to review
them today?
J.
ROBINSON: No sir.
M.
STOTO: Okay.
J.
ROBINSON: No, that’s for April;
we’ll start the chapter review.
Hopefully, we’ll get them so that you have at least two months prior to
that meeting. That will give us — we
have completed seven — at least five second drafts. There’s twelve chapters, so we’re way ahead, and they can be
spread out as you see necessary over those three meetings.
M.
STOTO: Okay. So maybe we should — we should sit down, and
think about the order, and who we have to review them and what — which ones you
might need before other ones, and whether it makes any logical sense to do some
first and things like that.
J.
ROBINSON: The ...
R.
HARRISON: The other thing that you're going to have to
really take a look at is the Committee’s expertise and ...
M.
STOTO: Right.
R.
HARRISON: ... early on, identify
some consultants for review of the sections.
M.
STOTO: Yeah.
M.
GOUGH: Well, each chapter is
written by an expert, right?
J.
ROBINSON: The chapters are
written. There is literature research
that was done by an outside group that was contracted through SAIC. I’m looking toward Bill Grubbs for some
assistance with this. And then each of
the chapters had a lead writer, an expert to assist with that.
M.
STOTO: Do we know anything about
who those people are?
J.
ROBINSON: We could provide you a
list of those for each chapter ...
M.
STOTO: Yeah.
J.
ROBINSON: ... if you’d like to
know. Okay.
M.
STOTO: Yeah.
J.
ROBINSON: We can do that. Along with the chapters, you will have
various appendices that will come with them as well; not each chapter will have
an accompanied appendix for the review.
R.
HARRISON: Okay.
So one of the things that needs to be discussed is the actual — when the
Committee will meet, try to get some dates organized. We also have a few other Committee-type things to discuss as
well. Are you finished with your — do
you have ...
J.
ROBINSON: One more item.
R.
HARRISON: One more item.
J.
ROBINSON: I’d just like to say is
that we’re planning to submit articles to peer-reviewed journals up until about
1 January, and from there, then we will turn any subsequent reports, remaining
reports into technical — Air Force technical reports. And that’s it, sir.
M.
STOTO: 1 January of which year?
J.
ROBINSON: '06.
M.
STOTO: ‘06. Okay.
R.
HARRISON: Well, that would make sense; I mean, they
wouldn’t have time to deal with the critiques and stuff. Any other questions of Joel or the — what we
have remaining is a discussion of the approval process for the March '03
minutes; a discussion of the upcoming meeting schedule; and it is almost precisely
the time when we said we’d have the public comments if there were any. That’s what we have coming up. So what I’d like to do is do the public
comments first and then we’ll do the minutes — decide how we’re going to handle
the minutes — decide how we’re going to meet and then we’ll be done.
M.
GOUGH: Before we do that, Joel,
what happened to hepatitis?
J.
MICHALEK: We — I did not include a
discussion of hepatitis data because the database isn’t ready yet. And so next time we meet, I should or next
time we talk science separate from the reviews, I’ll have some hepatitis data
to show you.
R.
HARRISON: Yes, Mike? Oh, I saw your thing on — your, excuse me,
microphone.
R.
HARRISON: So do we have any public comment at this
time, anyone that — sure we don’t want the pharmacy students to make
spontaneous presentations? Okay. All right.
Sure? Okay. All right.
Okay. Okay.
R.
HARRISON: Let’s — well, actually,
let me — let me do this and I'm going to do it awkwardly because my memory is bad
and all the — everything that you all know.
But we have two people here who are attending as observers who are
awaiting confirmation to join the Committee.
And I’d like to, number one, recognize them; number two, ask me to tell
them — ask them to tell me their names again because otherwise, I’ll screw it
up. Dave ...
M.
STOTO: Would it be okay if they
joined us at the table for this discussion?
No?
R.
HARRISON: I think we certainly
would like to hear from them though in terms of the future meeting
schedule. Go ahead.
D.
JOHNSON: My name is Dave Johnson.
R.
HARRISON: And you’re with the
Department of ...
D.
JOHNSON: Yes, I’m the Executive
Medical Director for the Division of Environmental Health with the Florida
Department of Health and ...
R.
HARRISON: Wonderful.
Well, we look forward to — I will look forward to hearing of what you do
in the future. And a new Committee —
yes ma’am?
E.
HASSOUN: I am Ezdihar Hassoun from
the University of Toledo College of Pharmacy.
R.
HARRISON: And I understand that she
has knocked off innumerable rats over the last twenty years with dioxin, so she
should fit right in with the rest of this group. We’re going to discuss the minutes, which is going to be the Committee
discussion, then we’re going to discuss when schedules might allow another
meeting to be held. And that even
though these — they can’t sit at the table, you all — I hope you have your
calendars and will be able to work that through.
M.
STOTO: Can I — can I ask one
thing? There are seven of us here, soon
to be diminished by two and replaced by two.
Isn’t the ...
R.
HARRISON: Nine.
M.
STOTO: Nine is the limit, right?
L.
SCHECHTMAN: Nine is the upper
limit.
M.
STOTO: The upper — it’s the upper
limit? So in other words, we’re going
to be seven for the duration technically?
Are we — I guess what I’m ...
L.
SCHECHTMAN: Actually, there is a
possibility that there may be more appointments coming from the Department, but
right now, we’re sitting — we will be sitting at seven.
M.
STOTO: Okay.
R.
HARRISON: Well, our leader has suggested regarding
approval of the minutes that we use the process employed previously; that is
that you all will submit your suggested revisions and edits to the executive
secretary. And after the editing changes
are made, the revised minutes will be sent to the chair for review and
approval. And the executive secretary
says that since Dr. Harrison served as chair of that meeting, as such, he will
be the signatory on those minutes, which is fine. I have a consulting rate and I’ll be happy to. Is there — is that process acceptable? Everybody has received their e-mail copies
of the minutes, so that can be — that’s not a problem?
M.
STOTO: I don’t — it’s not a
problem. I have two things: one is did we get them in Word format so we
can do track changes and so on? That
probably would help if people could do it that way.
L.
SCHECHTMAN: Yeah. In fact, that was in the cover e-mail
message; the request was that the editing changes be made in track changes.
M.
STOTO: Okay. Okay.
The second thing is do we want to set a date that these are due? Oh, that’s in there too? Oh great.
R.
HARRISON: Does everyone know what track changes are or
have a secretary who knows what they are?
Because I don’t — I actually don’t — I don’t use Word, so I don’t — I
don’t know what you all are talking about.
I’m a WordPerfect guy.
M.
STOTO: It’s a way to indicate your
changes ...
R.
HARRISON: So I have to go ...
M.
STOTO: ... electronically.
R.
HARRISON: So I have to go and find
a copy of Word then, okay, and learn how to use it?
M.
GOUGH: You’re the chairman. You don’t have to.
S.
LEFFINGWELL: Use redline and
strikeout in WordPerfect and then transfer it to Word.
R.
HARRISON: Anything else? All right, on to meeting schedule. Mike, why don’t you conduct this piece of the business because
I’m out of here.
M.
STOTO: Okay. Thank you.
Thank you. The — I guess — I
guess the question is we’re asked about April, September and November;
everybody is here who actually has to schedule. Do we want to — do people have their calendars? Are they prepared to do this?
R.
TREWYN: No, but we can wing it.
M.
STOTO: No? Okay.
L.
SCHECHTMAN: Perhaps we could get
our best guesstimates at this point ...
M.
STOTO: And then ...
L.
SCHECHTMAN: ... and then we can
resurvey by e-mail. Once we’re back in
our offices, get a really quick response from everybody so that we can nail
these dates down because calendars fill up quickly.
M.
STOTO: Okay.
L.
SCHECHTMAN: Thank you.
M.
STOTO: Okay. Let’s look then. I’ll just look at — look at April to begin with. Any particular dates we need to avoid in
April?
R.
TREWYN: The 15th.
S.
LEFFINGWELL: The 16th.
M.
STOTO: Okay. Is there — is there from the — from the Air
Force side something that we need to pay attention to there?
J.
ROBINSON: No, we're open.
M.
STOTO: Okay, and day of the week?
P.
CAMACHO: Yeah, a Monday or a
Friday.
M.
STOTO: Monday and Friday are
better?
K.
OSEI: Yes, much better.
M.
STOTO: Okay. Yeah.
Wednesday like today?
P.
CAMACHO: A Monday or — a Monday or
a Friday is good. We can try a Friday
and enjoy Washington in April.
M.
STOTO: Yeah. In April, Mondays are bad for me.
P.
CAMACHO: How about Fridays?
L.
SCHECHTMAN: So toward the end of
April? Is that what we’re considering
after tax day?
M.
STOTO: Yes.
L.
SCHECHTMAN: Is that what we’re
looking — okay. So the first Friday
after tax day is May — April 16th and the following Friday is April
23rd.
M.
STOTO: Fridays are not — are out?
L.
SCHECHTMAN: Fridays are no good?
K.
OSEI: Wednesdays or Thursdays would
be fine.
L.
SCHECHTMAN: Wednesday or Thursday.
M.
STOTO: Are they — are they
impossible?
P.
CAMACHO: Well, then I have to get
another — I mean, I don’t know how many classes I can cut.
L.
SCHECHTMAN: How about the 3rd?
P.
CAMACHO: This is on the first day
of classes.
L.
SCHECHTMAN: How about the 3rd?
P.
CAMACHO: Yeah, classes started this
week, so I already got substitutes on the first day of class. I have that on my syllabus.
L.
SCHECHTMAN: Okay.
M.
STOTO: So Wednesday and Thursday
are bad days for you?
P.
CAMACHO: Yeah.
M.
STOTO: Friday is a bad day for
Kwame.
P.
CAMACHO: Tuesday, Wednesday,
Thursday, but Monday’s a nice day for me.
M.
STOTO: How about Monday? Is Monday ...
K.
OSEI: Monday and Friday are not
good.
P.
CAMACHO: See, his classes are what
— Monday, Wednesday and Friday? And
mine are Tuesday, Wednesday, Thursday.
So how about a Saturday? We make
a nice weekend out of it.
S.
LEFFINGWELL: I do have a hard
conflict on the 16th.
M.
STOTO: Okay. I’ve got that. Where are we at? The 16th
we want to avoid. That’s one particular
hard thing.
P.
CAMACHO: Yeah. Is Saturday possible?
L.
SCHECHTMAN: The government doesn’t
work on weekends.
P.
CAMACHO: The government doesn’t
work on weekends.
L.
SCHECHTMAN: The Sabbath begins at
sundown Friday and lasts until Monday at dawn.
P.
CAMACHO: I’m married; I don’t get
any days off.
J.
MICHALEK: All right; there's April.
M.
STOTO: Okay. Do you have the same problem for September —
the same days and the same of you?
P.
CAMACHO: Mine will change.
M.
STOTO: Okay. So yours will change? So suppose that we accommodate — how does
that work? We have to accommodate Paul
now because if we accommodate you later, he may be able to make that one. Is that — is that — is that fair?
K.
OSEI: You can accommodate me later.
M.
STOTO: Yeah. I mean, I want — I want to spread the burden
fairly. So the first time through,
we’ll do it on a Monday or a Friday.
N.
RIVERA: Turn your — turn your mike
on.
M.
STOTO: Turn your mike on.
K.
OSEI: Oh, excuse me. Since we’re going to — we have about twelve
chapters to review, you may not be reviewing endocrine on one meeting date.
M.
STOTO: Right.
K.
OSEI: So the days that I need to be
here, I’ll be glad to. So if you can
accommodate me and then ...
M.
STOTO: That's a good point.
K.
OSEI: ... we can work it out.
M.
STOTO: That’s a good point.
K.
OSEI: Yeah.
M.
STOTO: So we will — if there’s one
that you absolutely can’t make it, we’ll make sure we don’t schedule the
relevant chapters ...
K.
OSEI: Exactly.
M.
STOTO: ... on that day. Okay.
So that gets us to Monday and Friday in April and then Tuesday,
Wednesday or Thursday ...
K.
OSEI: Wednesday.
M.
STOTO: ... in September. Wednesday in September?
K.
OSEI: Yeah.
M.
STOTO: Okay. So we’re back to Friday. How about the 23rd of April? That’s a problem for you?
D.
JOHNSON: Only that one Friday.
M.
STOTO: How about the 30th?
P.
CAMACHO: The 30th is
fine with me.
M.
STOTO: Okay. That’s okay for everybody except ...
R.
HARRISON: Kwame.
M.
STOTO: ... except you, then we’ll
work — April 30th.
K.
OSEI: Okay.
M.
STOTO: Okay. For September, we want to look for a
Wednesday.
K.
OSEI: Right.
R.
TREWYN: Let’s get your calendar up
there, come on.
L.
SCHECHTMAN: Before we move on, that
would mean people flying in on Thursday afternoon or evening and flying
out. Presuming we have a single-day
meeting similar to today, they would fly out Friday evening, Friday
afternoon/evening.
R.
TREWYN: They can always stay
around.
L.
SCHECHTMAN: Or not, okay.
R.
TREWYN: Right.
L.
SCHECHTMAN: Okay.
R.
TREWYN: There’s always a problem
getting out of DC Friday night.
L.
SCHECHTMAN: Yeah.
K.
OSEI: Yeah; that’s why I don’t want
to be at the airport on Fridays. It’s
always a problem.
D.
JOHNSON: The third Wednesday of the
month I sit on an IRB committee which requires a quorum, so that's not good.
M.
STOTO: Okay. So not the third Wednesday? The first Wednesday is September 1st. Is that — no, Labor Day would be the
following.
R.
TREWYN: The 6th.
K.
OSEI: Labor Day would be the 6th,
right.
M.
STOTO: I’m clear every Wednesday as
far as in that month.
K.
OSEI: Yeah, and it’s good for me —
any of those Wednesdays.
M.
STOTO: Well, people — September 1st
is before Labor Day, so that’s often a bad week for people. Yeah, so how about if we say the 8th?
K.
OSEI: The 8th.
J.
ROBINSON: And I just need to
clarify; there's twelve clinical chapters.
There’s like an introduction ...
M.
STOTO: Yeah. Now my thought — I mean, I guess there’s two
ways to think about the different kinds of chapters: one is that we do all of the kind of background chapters like
that first and then do the substantive chapters.
R.
TREWYN: That's if they've got them
written.
J.
ROBINSON: Yeah, they’re done.
M.
STOTO: But they — okay; well, then
if that’s not — yeah. I mean, the other
thought is to kind of do some background and some substantive at every meeting.
J.
ROBINSON: You’re getting five
background chapters for the introduction and other sections.
M.
STOTO: Okay. How do people feel about doing — taking care
of those background chapters first?
P.
CAMACHO: It’s better for me; I
mean, I’m not an epidemiologist.
M.
STOTO: Yeah.
J.
ROBINSON: We’ll have them all
ready.
R.
TREWYN: That’s what you’re going to
get so ...
M.
STOTO: Okay. So now we need a date in November.
P.
CAMACHO: What’s the date in
September?
M.
STOTO: September 8th.
K.
OSEI: September 8th?
R.
TREWYN: How soon they forget.
M.
STOTO: Now since Paul doesn’t know
what his conflict is going to be in November, I think we’ll still try to go for
a Wednesday.
K.
OSEI: Right.
M.
STOTO: That’s — if that’s —
yeah. Veteran’s Day is in November — is
a Wednesday in November. That actually
would be a bad one — yeah; the 10th is a — is a — is a bad one for
me.
R.
TREWYN: The 17th.
K.
OSEI: The 17th would be
good.
M.
STOTO: Should we do 17th?
L.
SCHECHTMAN: No; that doesn't work
for me. I have a conflict.
M.
STOTO: Okay. How about the 3rd?
J.
ROBINSON: The 3rd of
November?
R.
TREWYN: Right after the election;
right after the Presidential election.
L.
SCHECHTMAN: Vote your hearts out.
R.
TREWYN: Everybody can come and bitch.
M.
STOTO: Okay, and they’ll all be in Washington? Okay.
That sounds like a plan. And
then we’ll — maybe Len, and I and somebody from the Air Force can talk about
the schedule of which chapters are which.
J.
ROBINSON: It’ll probably be me.
M.
STOTO: Okay. Thank you.
R.
HARRISON: We have a couple more things — at least one
more thing. I’m sorry; we have one
more: a list of the — a list of the
reviewers. Just a comment of my own, I
mean, I recognize a couple of names on here just because I know them. But if you're trying to figure out who's
doing this, don't you need to know where they’re from or something?
M.
STOTO: So these are the people who
are participating in writing the chapters?
R.
HARRISON: Right, and — yeah.
J.
ROBINSON: We can tell you right now that on “General Health,” it’s Dr. Ken
Pischel who’s ...
R.
HARRISON: Well, that’s what I mean.
J.
ROBINSON: ... he’s under the
alternate list. For “Immunology,” it’s
Dr. Ron Simon. For “Renal,” it’s Dr.
Andrew King. I can tell you that Dr.
David Williams is at Scripps Clinic in La Hoya, California. Dr. Roger Cornell is also a dermatologist at
Scripps Clinic. I don’t know Dr.
Dailey. They’re all — okay; they’re all
at Scripps.
M.
STOTO: Including the alternates?
J.
ROBINSON: Including the alternates.
M.
STOTO: Okay. I mean, I guess I’m not likely to make much
of this. I mean, if they’re — if
they’re people who are Board certified, as you said they would be, and they’re
at Scripps, which is certainly a decent place, are we going to make — what are
we going to do about that?
L.
SCHECHTMAN: There’s — there would
be no perceived conflicts of interest with these people in terms of them being
at Scripps and some of the activities are going on at Scripps as far as the
collection of data? We have to be very
careful about that. So are these people
totally disassociated?
M.
STOTO: Well, remember these are not
reviewers. These are ...
R.
HARRISON: The writers.
M.
STOTO: They would be the writers,
so they’re part of ...
L.
SCHECHTMAN: Okay. So that’s okay?
M.
STOTO: Yes.
L.
SCHECHTMAN: And the reviewers will
not be involved?
J.
ROBINSON: Historically, they
have. I know Dr. Dave Williams has
participated in writing chapters in previous reports.
M.
STOTO: Yeah.
J.
ROBINSON: And that’s not been an
issue in the past.
M.
STOTO: Well, I think — I think
that, if I recall, the work they’re doing is part of the overall contract that
— with the — with Scripps.
R.
HARRISON: Anything else or is our work done?
M.
STOTO: Do we have the opportunity
to identify consultants to this Committee to review chapters that we need?
R.
HARRISON: Well, that’s what I’m
suggesting; is that you and Len need to get together, look at what expertise is
just on the Committee or coming on the Committee ...
M.
STOTO: Yeah.
R.
HARRISON: ... look at the chapter
headings and decide do you have an immunologist. And if you — if you don’t, then you need to get one. And in the past, Committee members have been
asked to identify if they — if someone that could serve in that capacity.
M.
STOTO: You know, presumably, if we
could — especially if we’re down two people, and actually, there may be some
resources there to hire consultants.
R.
TREWYN: Can I ask if there would be
a restriction on taking — utilizing people in that capacity who have been on
the Committee in the past? And even
separate from the ones who are here now, would it be feasible to bring any of
those individuals in as the consultants since they would have some level of
experience with it?
L.
SCHECHTMAN: We can check with the
Committee Management Office to determine if that’s allowable under the rules by
which these committees abide by, so we can get back and discuss that as a possibility
as well.
R.
TREWYN: Okay.
R.
HARRISON: I know when Ron left as the executive
secretary, he came back as a consultant for a couple of rounds.
M.
STOTO: Right, but that’s somewhat
different than a Committee member coming back.
R.
HARRISON: Oh yeah. Yeah, it’s being a part of the Mafia.
J.
ROBINSON: Dr. Harrison, I do have
one thing I wanted to enter into the record.
R.
HARRISON: Oh yes; I’m sorry. Please go ahead.
J.
ROBINSON: Barbara Jewell has spent
at least fourteen years associated with the Ranch Hand Advisory Committee, and
thus, the Air Force Health Study. I
talked to her prior to her retirement and I think she was pulling my leg; she
said she’d be here for this meeting, which she’s not. But in recognition of her outstanding support in assisting us in
coordinating these meetings, the agenda and all, we have a certificate that
recognizes her now as an official member of the Air Force Health Study
Crew. And I’m going to give it to Dr. Schechtman
to ensure that it gets to Barbara, but we do thank her.
PARTICIPANTS: [Applause.]
R.
HARRISON: Okay.
Anything else? Going once, going
twice, finished. Thank you.
M.
STOTO: Thank you and Mike.
[ADJOURN 1:27 P.M.]