1                             MEETING

 2                              OF THE









11                   255 SOUTH AIRPORT BOULEVARD






                     WEDNESDAY, OCTOBER 6, 1999
                              9:00 A.M.





24   Janet H. Nicol
     Certified Shorthand Reporter
25   License Number 9764



 1                          APPEARANCES
    Dr. John Froines, Chairman
 3  Dr. Paul D. Blanc
    Dr. Craig Byus
 4  Dr. Gary Friedman
    Dr. Anthony Fucaloro
 5  Dr. Stanton Glantz
    Dr. Peter S. Kennedy


 8  Mr. Bill Lockett
    Mr. Peter Mathews

    Dr. George Alexeeff, Deputy Director for Scientific Affairs
12  Dr. James Collins, Staff Toxicologist
    Dr. Melanie Marty, Supervising Toxicologist
13  Dr. Andy Salmon
    Dr. Martha Sandy














 1                             INDEX

 2                                                        PAGE
    1  Continuation of consideration of draft report:       3
    2  Review of health values for the air exposure       106
 7     pathway for methyl tertiary butyl ether (MTBE)

 8  3  Further Panel deliberations regarding the          ---
       Panel's September 1999 "Pesticides in the Air
 9     Workshop"

10  Adjournment                                           150

11  Certificate of Reporter                               151

















 1                      P R O C E E D I N G S

 2             CHAIRMAN FROINES:  Let's go on the record and

 3   start the meeting.

 4             Good morning, everybody.  The meeting is now

 5   official.

 6             And the question I'm raising with the individual

 7   panel members is they have all seen the modified findings,

 8   SRP findings for methyl parathion.  I'm just asking if there

 9   were any other comments that we need to discuss, because

10   we'd like to finalize the document.

11             And unless I hear something else, I'll assume

12   we're okay.

13             DR. KENNEDY:  I had some probably minor and fairly

14   insignificant concerns in terms of the overall piece.  And

15   such issues as particularly in the area of biologic effects

16   there are discussions of depression of red cells, hemoglobin

17   and hematocrit.  Those are obviously all the same thing.

18             CHAIRMAN FROINES:  Peter, I think -- Bill, this is

19   not on the agenda, so it means we can't have a discussion?

20   It is not on the agenda, so we can't have a discussion.  So

21   we'll have to resolve this separately.

22             MR. LOCKETT:  Where it was last left at the last

23   meeting was that the panel was going to be given the

24   strikeout changes so they could review to determine that,

25   yes, the changes are within the ambit of the discussion of



 1   what the panel decided.

 2             So my understanding from Craig's comments earlier

 3   is that's the case in terms of the changes that were made as

 4   lead person.

 5             And we've gotten Peter Kennedy's comments, but

 6   we've not reviewed them from that standpoint.

 7             DR. BLANC:  I think the easy way to resolve this,

 8   as long as there's no substantive issue that is an important

 9   omission or something which is internally contradictory, we

10   should just let it stand even if there are areas which, you

11   know, could be further honed, such as your comment about

12   redundancy.

13             So I'd probably just let it pass.

14             DR. KENNEDY:  That's how I prefaced my remarks.

15             I have no concerns over the major revisions.  I

16   have no concerns over the main points.  These are minor

17   issues that I think we can deal with --

18             DR. BLANC:  Prospectively, let's say, and the next

19   time we have one of these, we should just be cautious that

20   we are not overly redundant.

21             CHAIRMAN FROINES:  If there are minor issues we

22   can just make them.  If they are relatively minor issues

23   like redundancies, we can make the changes and not

24   necessarily go back to the panel again.  Because

25   theoretically what we should be only talking about here is



 1   the cross-outs, because that's where the questions were from

 2   the last meeting.  We shouldn't be raising new issues.

 3             DR. KENNEDY:  That's fine.

 4             CHAIRMAN FROINES:  So we're okay.

 5             DR. GLANTZ:  So I think just to be clear, so we've

 6   now finished with it?

 7             CHAIRMAN FROINES:  Yes.

 8             DR. GLANTZ:  Subject to any little minor editorial

 9   adjustments that the chair makes.

10             CHAIRMAN FROINES:  Yes.  And we've already taken a

11   vote, so it's -- and we'll work out the final things with

12   Peter.

13             Peter Mathews, I don't have an agenda.  Do you

14   have an agenda I can have?

15             I think basically the first, unless I'm mistaken,

16   the first item on the agenda -- that's September 16th -- is

17   the continuation of the discussion on the RELs.

18             And unless I'm really mistaken, I don't think Paul

19   Blanc is finished.

20             DR. BLANC:  No.  But you may have to remind me

21   which ones we discussed and which ones we didn't.  I

22   remember doing formaldehyde.

23             DR. COLLINS:  And hydrogen chloride and chlorine.

24             DR. BLANC:  And chlorine.

25             DR. FUCALORO:  Ammonia, dioxane, hydrogen cyanide.



 1             DR. MARTY:  Dr. Froines, we had also put together

 2   some information.

 3             CHAIRMAN FROINES:  Oh, that's right, Melanie.  I'm

 4   sorry, I'm preempting you.

 5             Why don't we go ahead with what you were going to

 6   provide the panel and then we'll go back to Paul and in the

 7   meantime he can figure out which ones he didn't do.

 8             DR. MARTY:  Okay.

 9             CHAIRMAN FROINES:  The people who we have yet to

10   cover, though, today are Paul, Peter Kennedy, Stan Glantz

11   and myself.

12             DR. MARTY:  I'm Melanie Marty from Office of

13   Environmental Health Hazard Assessment, and I have Jim

14   Collins on my left.

15             And Dr. Witschi and Dr. Glantz at the last meeting

16   suggested that we take a look at the reference exposure

17   levels that were based on human studies and see what the

18   reference exposure levels for those chemicals would be if we

19   based it on an animal study that we had some amount of

20   confidence in.

21             So we have done that.  I have I think it's about

22   12 examples.

23             DR. GLANTZ:  13.

24             DR. MARTY:  13, thank you.  13 examples of that.

25             I think I need to preface it by saying that we are



 1   not suggesting that we would use that animal study.  It's

 2   more of this is what happens when you use the animal study

 3   for those chemicals.

 4             So this -- you should all have a handout so you

 5   don't have to strain your eyeballs looking at the overheads.

 6             So, Peter, did they get those handouts?

 7             MR. MATHEWS:  Yes.

 8             DR. KENNEDY:  Yes, twice.

 9             DR. MARTY:  We did fax a preliminary version of

10   these tables to the panel members and there have been

11   changes made, so it's best to follow the handout that we

12   have today.

13             CHAIRMAN FROINES:  I just want to make one

14   comment.  In my REL here for methylene chloride, you have

15   the REL as 300, and up there it's 400.

16             DR. MARTY:  Interesting.

17             CHAIRMAN FROINES:  What's in this document is 300.

18             But go ahead.

19             DR. MARTY:  It was probably a rounding error.

20             Okay.  As you can see by just perusing this table,

21   some of the numbers come in pretty close to each other.

22   Most of them are within a factor of three, but there are

23   some that are bigger than that.

24             Next slide, Andy.

25             CHAIRMAN FROINES:  Before you go on, there are, I



 1   counted four where there are quite substantial differences.

 2             DR. MARTY:  Yes.

 3             CHAIRMAN FROINES:  And I don't know if I said this

 4   last time, but I'm still concerned that what happens is we

 5   tend to pick a value from a particular study and in some

 6   ways I still think it would be useful to look at the values

 7   you would get from a series of studies and then make a

 8   judgment about what makes the most sense, rather than simply

 9   picking the most conservative value.

10             In other words, I think that the scientific

11   justification for the number that -- I mean, I think that,

12   yes, we should proceed in a public health protective

13   fashion, but I think we should also look at all the studies

14   and decide what makes the most sense as we proceed as well.

15   Where is the consistent finding relative to the lowest

16   finding.

17             DR. GLANTZ:  Go ahead.

18             DR. MARTY:  I would say that at the staff level we

19   did try to do that, so we did look at the studies that we

20   thought were useful.  In many cases there really aren't that

21   many chronic studies by the inhalation route.  That's part

22   of the problem.

23             And we did look at the numbers that we would get

24   using several different studies or however many studies

25   there were that were available.  And, you know, we had



 1   discussions among ourselves as to what the major

 2   uncertainties were.

 3             I think that part of the difficulty for the panel

 4   in reviewing this document is all that thinking isn't

 5   written into the document so it's very hard for you guys to

 6   tell what we have looked at and what we haven't looked at.

 7             So but I agree that if the numbers are really

 8   wacky relative to the rest of the information you have on

 9   that chemical, then there's a lot more uncertainty in using

10   that number.

11             So there may be cases where we probably should not

12   have developed a REL and we did.  I think that's certainly

13   an issue for discussion.

14             DR. GLANTZ:  Although, when I looked this stuff

15   over a couple days ago when you sent it, I was actually

16   impressed at how close the agreement was, given all these

17   problems.  And I think 10 of the 13 they were within one

18   order of magnitude, which given the uncertainties that are

19   implicit in the process, especially if you're dealing with

20   animal data, is I thought quite remarkable.

21             So there are three or four where there were big

22   differences, but for the most part the thing that impressed

23   me was how close they were.

24             This actually increased -- actually I was

25   expecting either a completely random results or a situation



 1   where there were huge differences.

 2             But I think, you know, when one recognizes that

 3   this risk assessment process is not something you can do to

 4   17 significant digits, this is quite reassuring to me about

 5   the process that we've been using.

 6             Not always, but 10 out of 13 were pretty good.

 7   And they're actually, in this -- I walked off this morning

 8   and left all my stuff at home, but I actually plotted this

 9   stuff.  If you do a log-log plot of these, they're pretty

10   well correlated.

11             DR. BLANC:  Stan, you better be careful.  Someone

12   might think you're an absent-minded professor.

13             DR. GLANTZ:  No.  I'm just a space-out.

14             CHAIRMAN FROINES:  What makes me more worried

15   about it is if you walked out without, are you going to be

16   able to do your chemicals?

17             DR. GLANTZ:  I'm going to try.  But Blanc is here.

18   I talked to him on the phone about all this and anything I

19   don't know he knows.

20             But, I mean, I just think that that to me was -- I

21   thought this was a useful exercise actually.

22             DR. COLLINS:  And we did pick the study we thought

23   was the best animal study we knew of, rather than what would

24   give us the closest.

25             DR. MARTY:  I should also add, for this table that



 1   we just have where the human study was from an EPA RfC, that

 2   is what is represented in this table.  In a couple of cases,

 3   we would have done a little bit differently with the same

 4   study, and I'll show you that as we go through the

 5   chemicals.

 6             CHAIRMAN FROINES:  I think it's important that I

 7   believe that, because I do experiments, that there are

 8   people do experiments very differently and there can be

 9   wildly different outcomes.

10             And so I think in terms of protecting everyone's

11   interests, pursuing the most consistent pattern of findings

12   really makes the most sense, rather than picking studies

13   that are outliers, because they're conservative in public

14   health finding.

15             So that it's just something that I think we need

16   to be -- we need to try and do anything that increases our

17   confidence in the findings, and so that as well as having

18   public health protection as a key watchword.  But anyway.

19             DR. COLLINS:  At least one example we did that.

20   It's not in this cite.  With creosols, we had a study from

21   the Russian literature which had very very low levels of

22   facts, and they just seemed to be such outliers that we

23   abandoned that and used another study.  So that has entered

24   into the thinking, at least in some of them.

25             CHAIRMAN FROINES:  I just wanted to make one other



 1   comment, because you had mentioned EPA RfCs.

 2             Two or three or four years ago, you lose track

 3   with all this, that when we were in the more controversial

 4   phase of this panel's existence, when there was different

 5   leadership at OEHHA, there was a significant press to have

 6   our values be similar to those, if not identical, to the EPA

 7   values.  And this panel has always taken the position that,

 8   yes, it would be good to -- what's the term that --

 9             DR. COLLINS:  Harmonize.

10             CHAIRMAN FROINES:  Harmonize.  That we harmonize

11   our risk assessment values with EPA.

12             This panel has always taken the position that

13   harmonization is a good idea, but only if it makes

14   scientific, good scientific sense, and it is a matter of

15   good judgment.  It is not an ironclad rule.  And I think I

16   just want to say that for the record.

17             And unless there's somebody in here that

18   disagrees, please disagree now, because I want to make sure

19   that this panel is on record saying that harmonization is

20   good, but it's the quality of the science that we're

21   interested in, not harmonization as the endpoint.

22             Go ahead.

23             DR. MARTY:  Next slide, Andy.

24             The first example is ammonia, which I think is one

25   of Dr. Blanc's chemicals.



 1             We had a human-derived reference exposure level of

 2   100 micrograms per cubic meter.

 3             When we took the best study that we had on hand,

 4   in guinea pigs the reference exposure level ends up being

 5   ten.

 6             The human study has a slightly different endpoint.

 7   We're looking at pulmonary function as opposed to pathology,

 8   pulmonary edema in the guinea pig.

 9             And the duration of exposure was only six weeks in

10   the animal studies, so we have a subchronic uncertainty

11   factor there right away.

12             Also the human study, we identified a NOAEL from

13   that study.

14             Now, the US EPA when they evaluated this also

15   talked about a rat study, which they got a LOAEL from.  It's

16   a different endpoint, but what they wanted to say was just

17   what Dr. Froines was just talking about is within the

18   confines of the data that you have available that 9.2 NOAEL

19   doesn't look too bad with respect to LOAEL from an animal

20   study.  And that's the Broderson et al in the LOAEL row.

21             Of course the endpoint was an exacerbation of

22   effects of a microplasm infection, so it's not quite the

23   same as looking at pulmonary function or eye, skin or

24   respiratory irritation.

25             At any rate, I think use of the animal study is a



 1   little complicated, the Anderson study, because of the LOAEL

 2   uncertainty factor.  You also have to have an interspecies

 3   uncertainty factor.  And we have a subchronic uncertainty

 4   factor, so you end up with a rather large cumulative

 5   uncertainty factor of 3,000, versus 30 if you use the

 6   Holness et al '88 human study.  '89, I'm sorry.

 7             So that's one example of the kind of stuff we're

 8   faced with for ammonia.

 9             And we were somewhat surprised that there's a lot

10   of acute studies on ammonia because it's an irritant, but

11   there really is not very much information on chronic

12   exposures, so that in and of itself is a problem.

13             We can do the next example if there's no

14   questions.

15             DR. BLANC:  I think I'm not going to make my

16   comments now, I'll wait.  I think we'll just confuse things.

17             DR. MARTY:  Okay.

18             DR. BLANC:  Only make comments as they relate to

19   the specific task at hand.

20             DR. MARTY:  Okay.  Benzene, here's another example

21   where OEHHA used a study of male refinery workers where they

22   had a diverse exposure duration, one to 21 years, with an

23   average of 7.4, and a third of those were more than ten-year

24   exposures.

25             In the study they did not observe any effects.



 1   They actually looked at a lot of different things, but in

 2   particular we were interested in the hemotological effects,

 3   which are noted for benzene exposure.

 4             So they came up with basically a freestanding no

 5   observed effect level of a half a ppm, which when we account

 6   for exposure continuity and use an intraspecies uncertainty

 7   factor of 10, we end up with 20 parts per billion basically

 8   as our reference exposure level.

 9             DR. BLANC:  What's interesting here I think is you

10   have a fairly recent animal exposure study, whereas in the

11   previous ammonia example, it was a very old animal study.

12             And I think that one of the things that is worth

13   taking into account, perhaps in an introductory section to

14   the whole document, which would state why you weighted

15   things that were taken into account and study quality.

16             And one would be the time frame of the study, that

17   if there were two studies where they were both animal

18   studies, but one was more recent, and would therefore have,

19   you know, more currently accepted standards, that you would

20   tend to weight that more heavily.

21             But a question with something like this, since

22   basically you're looking at hematologic effects that are

23   probably not unrelated to benzene's oncologic effect, even

24   though this document is noncancer outcomes, it surprises me

25   that there wouldn't have been, for example, an animal study



 1   of longer duration which was primarily a carcinogenesis

 2   study which nonetheless reported noncarcinogenic hematologic

 3   outcomes.  So it surprises me that all you have for an

 4   animal study is an eight-week study.  I mean, I know that

 5   there are a lot of multi-month and even primate studies of

 6   benzene.

 7             DR. MARTY:  Yeah, there are.

 8             I think we used this, because they were looking

 9   specifically at hematological effects, but that's a good

10   point.  We could go back and look at some of the other

11   chronic studies that were oncogenicity studies and see what

12   else they looked at and what they measured.  I think that's

13   a good point.

14             One of the problems is we had to use the

15   subchronic uncertainty factor of ten.

16             DR. BLANC:  Right.  Because it seems to me that

17   those data are probably in those studies.  I would be very

18   surprised if somebody didn't -- somebody reported out on a

19   leukemia study and particularly in primates and didn't talk

20   about other hematologic effects.

21             Peter, this wasn't your chemical, was it?

22             DR. KENNEDY:  No.

23             In man, of course, the exposure is acute in terms

24   of the induction of marrow aplasia malignancy.  It's not a

25   chronic phenomenon.



 1             DR. BLANC:  It depends on how you define chronic,

 2   but I mean people have been exposed for months and years.

 3   It's a relatively short latency in the big scheme of

 4   carcinogenesis.

 5             CHAIRMAN FROINES:  But Paul's point is

 6   particularly interesting, because in looking at your

 7   references there are no references to chronic ano bioassays,

 8   and we know benzene was the first chemical this committee

 9   ever took up in 1984, and at that point we had a Maltoni

10   study, and I don't know if MTB has done one since, but if

11   MTB had done one, there clearly would have been -- I can't

12   say the word.

13             DR. BLANC:  Hematologic.

14             CHAIRMAN FROINES:  Whatever.  So that there's no

15   reference to any chronic studies in this document.

16             DR. BLANC:  Well, the other thing, the reason I

17   think it's particularly critical in something like benzene,

18   which is an air pollutant, people are likely to encounter,

19   there could be a lot of public health implications

20   between -- in the difference between 20 parts per billion

21   and six parts per billion, might really be important from a

22   public health action point of view.  I mean, it might drive

23   different approaches.  I don't know.

24             CHAIRMAN FROINES:  What's the 10 to the minus

25   fifth dose?



 1             DR. COLLINS:  For benzene?  Seven micrograms per

 2   day.  That's the NSRL under Prop 65.

 3             CHAIRMAN FROINES:  How did that -- we need you to

 4   translate that for us.

 5             DR. MARTY:  3.35 cubic meters per day if you did

 6   it gross --

 7             DR. COLLINS:  .35 micrograms per cubic meter.

 8             DR. MARTY:  So we're at 16 and 20.

 9             I think that background levels and urban levels

10   have dropped from about four ppb to around one ppb now.

11             So but I think we do -- I do want to back up and

12   say that for this comparison we did this in a pretty quick

13   time frame, but I think that's a good point, though, is I

14   know there's chronic studies.  Everybody has looked at the

15   carcinogenicity and so forth, so we should dig one up and

16   see what it looks like compared to the Tsai et al male

17   refinery worker study.

18             You know, it's the age-old problem with an epi

19   study in general that exposure is always an issue,

20   quantifying that exposure is not an easy thing.

21             Okay.  Another example we have, which has all

22   sorts of different problems, is ethylene glycol.  And you'll

23   recall that we used in the document a human study on

24   volunteers who were exposed, and respiratory tract

25   irritation was one of the endpoints that they looked at.



 1             We have a subchronic uncertainty factor because

 2   the exposure was only for 30 days.

 3             And the total cumulative uncertainty factor is 100

 4   and it includes intraspecies uncertainty factor also.

 5             We discussed at the last meeting whether or not

 6   respiratory tract irritation was actually a good endpoint

 7   for a chronic study, particularly if it's a sensory irritant

 8   effect.  So that's an issue right there.

 9             When we look at Tyl et al 1995, it's a whole body

10   inhalation study in mice and they were looking at

11   reproductive and developmental effects.  So they defined a

12   NOAEL in the study based primarily on fetotoxicity.

13             We have an interspecies uncertainty factor of

14   three, because there was an HEC estimate, a human equivalent

15   concentration estimate.

16             And then we have the intraspecies uncertainty

17   factor of ten, for a total uncertainty factor of 30.  So

18   right there you can see a difference of the uncertainty

19   factors are different, human versus animal study.  And the

20   results are two and a half fold different, 400 micrograms

21   per cubic meter versus a thousand.

22             If we didn't use the subchronic uncertainty factor

23   with the respiratory irritation endpoint, then our number

24   would be 800 micrograms or so -- did I get that right?  Even

25   more than that.  There's another issue there with whether or



 1   not we need that subchronic uncertainty facto that was used

 2   with the human study.

 3             We talked about formaldehyde last time.  We had

 4   based the reference exposure level on a couple of human

 5   studies.

 6             DR. BLANC:  It's really one human study.

 7             DR. MARTY:  It's really one.  The other one is in

 8   there because it's I guess it's somewhat supportive of it.

 9             DR. BLANC:  It's the same study, they just

10   analyzed a different endpoint.  You need to be clear about

11   that.

12             DR. MARTY:  Okay.

13             DR. BLANC:  From what I can tell.  I didn't pull

14   the papers, but it's the same study and it's just a

15   different --

16             DR. MARTY:  Different order of authors.

17             This was the study where they had a group of

18   office workers as the control group, quote, control group,

19   and, of course everybody is exposed to some formaldehyde, so

20   they made measurements in the office and then they had a

21   groups of workers that were exposed to either formaldehyde

22   alone or formaldehyde-pressed wood desks.

23             And they did a series of measurements looking at

24   nasal and eye irritation, nasal obstruction, and also some

25   lower airway symptoms.



 1             So we ended up with an inhalation reference level

 2   of three micrograms per cubic meter with a cumulative

 3   uncertainty factor of ten for intraspecies variability in

 4   humans, with the human study.

 5             Then with the animal study that we used to make a

 6   comparison it was a monkey study, and they actually looked

 7   at squamous metaplasia in the nasal turbinates.  They had a

 8   no observed effect level of one ppm.  It's a subchronic

 9   study, so we have a subchronic uncertainty factor of ten.

10   We have interspecies uncertainty factor of ten and an

11   intraspecies uncertainty factor of a ten, for a total

12   uncertainty factor of a thousand.  And that ends up giving

13   you a reference exposure level of ten micrograms per cubic

14   meter.

15             So the uncertainties are actually different.  The

16   issues are different.  We have good exposure measurements in

17   the monkey study, whereas you may not have such good

18   exposure measurements in the human study, but the

19   uncertainties are shorter duration of exposure interspecies,

20   any potential interspecies differences.  Those are the key

21   issues there.

22             DR. BLANC:  Can I ask a couple questions just to

23   again it would be related to global approaches.

24             The interspecies uncertainty factor, primate

25   studies don't get a break on that?



 1             DR. MARTY:  Well --

 2             DR. BLANC:  It seems to me just from a biological

 3   point of view that using an interspecies factor of ten for

 4   primate studies and ten for rat studies is not the way, you

 5   know, most -- well, I don't know.  Just from a biological

 6   point of view, it doesn't seem --

 7             DR. MARTY:  Yes, I would agree with that.  It is

 8   not really taken into account in the general methods by

 9   either EPA or by us.

10             DR. BLANC:  I know that we don't have the luxury

11   of very many primate studies, but it seems to me that if we

12   are going to use primate studies, maybe there should be a

13   decision, at least for this document, given the great

14   difficulties in these particular evaluations.  This

15   particular set of evaluations seemed to be much more problem

16   ridden than either the acute or the cancer effects.  And

17   maybe we should have some approaches that take into account

18   the limitations of data overall, and one of those, I think,

19   should be -- I'm curious to hear -- or more animal testing

20   oriented.

21             CHAIRMAN FROINES:  Yeah.  If it were -- I'm not a

22   wild enthusiast for primate studies, period, but leaving the

23   ethical issue aside, it would seem that I wouldn't use more

24   than a factor of three.

25             DR. BLANC:  Which is what you use when you have



 1   human equivalencies, so it would be consistent.

 2             DR. MARTY:  You can do that.

 3             DR. BLANC:  You could assume that you have gotten

 4   the --

 5             DR. MARTY:  Okay.

 6             DR. BLANC:  The other thing that I think this

 7   side-by-side comparison underscores and comes back to your

 8   allusion to the issue of respiratory nasal irritant symptoms

 9   as compared to either a condition, such as chronic

10   bronchitis or asthma or a pathologic finding or a finding

11   such as airway nonspecific hyperresponsiveness or something

12   that one would assume was a more integrated measure of a

13   chronic effect rather than simply a repeated acute effect or

14   subacute effect, and that is that I think there needs to be

15   some weighting given to a pathologic finding such as

16   squamous metaplasia as opposed to an indeterminate issue

17   such as nasal and eye irritation.

18             As you remember from our discussion last time, I

19   did point out that there was another paper from the same

20   group where they did do, apparently, based on the title,

21   some kind of nasal biopsy or something.  Did you pull that?

22             DR. MARTY:  We didn't pull that paper yet, but,

23   yeah, that's an important paper to look at.

24             The other thing is I'm a little squeamish about

25   using metaplasia.  It's a little worse than hyperplasia, so



 1   that's, to me, an issue also with at least using this

 2   particular monkey study.

 3             DR. BLANC:  Although it's only 26 weeks, so.

 4             DR. MARTY:  Yeah.  I know -- I don't know how

 5   readily reversible metaplasia would be in these animals.

 6             DR. FRIEDMAN:  Isn't something like nasal and eye

 7   irritation an acute effect, an effect of acute exposure,

 8   rather than an effect of chronic exposure?

 9             DR. MARTY:  It is, but if you're exposed every

10   day, I mean, you know, have to -- you wouldn't want somebody

11   to be chronically exposed to a concentration that would

12   produce nasal and eye irritation, so that's why we consider

13   using those for chronic endpoints.  But it is an issue.

14             CHAIRMAN FROINES:  Does formaldehyde increase IGE?

15   Is it an adjuvant for asthma?

16             DR. BLANC:  The whole literature on formaldehyde

17   and true allergic responses is very very murky.  There are a

18   handful of case reports of formaldehyde-related asthma, and

19   it's really very murky.  It's a low molecular weight

20   substance, so nobody is clear if it's acting as a half tone,

21   or something else is going on that it's a non IGE-effect or

22   as you were asking whether or not it's somehow increased the

23   risk of sensitization to ambient air or allergens or

24   something.  It's a very murky literature.

25             But again as I said last time, so I'm not going to



 1   be making comments on the formaldehyde when you get back to

 2   me, but I don't think that the formaldehyde thing can stand

 3   as it was written.

 4             DR. MARTY:  Okay.

 5             DR. BLANC:  Otherwise we'd be telling the public

 6   health policy makers that they'd have to evacuate this room

 7   because of the REL has been exceeded for chronic health

 8   effects.  It just doesn't make sense.

 9             DR. MARTY:  Okay.  Then we have another example

10   with one of Dr. Froines' chemicals, hexane.

11             CHAIRMAN FROINES:  Why don't we just -- I thought

12   you agreed that this one, you agreed with my criticisms, and

13   so it was my assumption you were going to go back and redo

14   it.  Remember, I mentioned that.  You quickly said, yes, we

15   agree.  And we -- so I didn't go through my criticisms

16   because you so quickly agreed.

17             DR. MARTY:  Okay.

18             CHAIRMAN FROINES:  It doesn't seem to make sense

19   to use something that last time you said you agreed with my

20   criticisms and then go through and show how a flawed study

21   compares to an animal study.

22             DR. MARTY:  Well, the use of this is that there's

23   a flawed human study and a flawed animal study, and that's

24   basically what we have to work with, so that's why I thought

25   I would throw that up there, and also, you know, I think I



 1   was agreeing that I'm a little nervous just accepting US EPA

 2   RfCs on face value or even after we've looked at it, and we

 3   actually would have done something different with Sanagi et

 4   al, so I had that up there.

 5             Also there's a couple of issues that came up in

 6   your comments last time, I believe you were concerned about

 7   that in occupational settings there's also frequently dermal

 8   exposure and these authors specifically said that these

 9   folks were not exposed dermally so it was an inhalation

10   exposure.

11             And I think the bigger issue was the concurrent

12   exposures with acetone and potential potentiation.  We need

13   to look at that a little further.  I know that other C6

14   ketones that can also be metabolized to the hexanedione

15   would certainly potentiate the hexane exposure, because

16   there the ultimate toxicant is the same in their metabolism.

17             Acetone, I'm not familiar with in terms of the

18   interaction with hexane.  It may be that it's another

19   mechanism by which it's potentiating.

20             So I did want to look at that a little more.

21             CHAIRMAN FROINES:  We've got a paper that we are

22   publishing in Environmental Health Perspectives in which we

23   have done toxicokinetic modeling on 2,5-hexadiene's

24   potentiation of ketones, and it's in fact the acetone is

25   metabolized to the di-ketone and it's in the -- it's in the



 1   detoxification phase of the toxicokinetics where the

 2   interaction occurs, so that I think that I'll send you the

 3   paper and you can read it.

 4             But I feel pretty strongly that where you have

 5   exposure to these ketones where you can get toxicokinetic

 6   interaction either in the formation of 2,5-hexadiene or in

 7   the detoxification of it, that you just don't want to use

 8   those studies because you really can't -- you have no

 9   measure of the magnitude of the interactions, so you're not

10   really necessarily measuring the pure effect, you're

11   measuring -- it's a potentiation, so you're measuring -- so

12   it's your dose estimate is not accurate.

13             DR. MARTY:  Right.  Okay.

14             We did look at Miyagaki 1967.  The study itself is

15   in Japanese with an English summary, so we need to find

16   somebody who speaks Japanese to make sure that the English

17   summary is really reflective of what's in the paper.

18             And also we need to search a little more to find

19   more animal studies for this.

20             The end result is that from the animal study you

21   have quite a bit higher REL than from that human study, and

22   that could actually reflect the potentiation that the

23   workers were experiencing by having concurrent exposures.

24             It did -- a couple issues came to my mind that the

25   Miyagaki study used commercial grade, which is only about 68



 1   percent n-hexane, from what I can gather, just looking at

 2   other sources.  The rest of it is branch chain hexanes,

 3   which don't get metabolized to the hexanedione, so that may

 4   not be an issue for this study.

 5             It was a one-year study in mice, so we did not

 6   apply a subchronic uncertainty factor, because we felt for

 7   mice that that was sufficiently chronic.

 8             We did a HEC calculation so we only have an

 9   interspecies uncertainty factor of three, and a total

10   cumulative uncertainty factor of 30.

11             If we had used Sanagi ourselves and we wouldn't

12   have used EPA's modifying factor, which we can't ever really

13   figure out what they're using it for and they're not

14   consistent, so we would have actually had a slightly higher

15   number.  It would have been 700 versus 7,000.

16             But we are going back and looking for more

17   information on hexane.

18             DR. BLANC:  Here's another generic question for

19   you.  You have a lot of rounding that you do in this

20   document and I'm not sure where the rounding -- what stage

21   the rounding happens, since it's really a cumulative

22   uncertainty of 90 and not a hundred.

23             DR. MARTY:  You know what, that's because the

24   rounding actually, three represents really 3.16 rather than

25   three, so when we put that -- whenever it's three, it's



 1   really 3.16.

 2             DR. BLANC:  For which one?

 3             DR. MARTY:  For the intermediate, the quote,

 4   intermediate uncertainty factors.

 5             DR. BLANC:  For both of them?  For both the LOAEL

 6   uncertainty factor and the subchronic uncertainty factor,

 7   for both of them it's 3.16?

 8             DR. MARTY:  Right.

 9             DR. FUCALORO:  Remind me for a minute why it's

10   3.16.

11             DR. FRIEDMAN:  Square root of ten.

12             DR. FUCALORO:  Square root of ten.

13             DR. MARTY:  Right.

14             DR. BLANC:  Can you tell me why you would put a

15   subchronic uncertainty factor for a human exposure duration

16   of six years?  I understand it's a part of -- it's because

17   it has to be more than 20 years for you not to put in the

18   factors.

19             DR. MARTY:  It has to be more than 12 percent of

20   the lifetime, so 12 percent of 70, so if it's greater than

21   eight years, we don't use a subchronic uncertainty factor.

22   If it's between five and eight years, we use a subchronic

23   uncertainty factor of three and if it's less than five

24   years, we use a subchronic uncertainty factor of ten.

25             So that's why we would have applied a subchronic



 1   uncertainty factor in that case.

 2             Actually, when you go back and read US EPA RfC,

 3   they say we have applied an uncertainty factor of ten for

 4   use of a LOAEL and for less than chronic.  So they actually

 5   don't spell it out, but they're doing three and three also

 6   in that case.

 7             I guess one of the reasons we do a lot of rounding

 8   too is Stan's point earlier that there's so much uncertainty

 9   it's kind of useless to have two or three or four or five

10   significant figures.

11             CHAIRMAN FROINES:  We have done a review of all

12   the hexane literature, which we'll send you.

13             DR. MARTY:  Okay.  That's great.

14             DR. BLANC:  I mean, hexane is a good example of a

15   chemical for which you really wouldn't want this kind of

16   analysis for, because really with hexane what you really

17   care about is a noncancer chronic exposure effect.  So it

18   certainly is an example for which something this exercise

19   makes sense, wouldn't you say?

20             CHAIRMAN FROINES:  Yeah.

21             I was going to ask you, I wonder if ARB knows

22   anything about the ambient hexane levels, since we have a

23   lot of -- down where I live we have a lot of petroleum

24   refining, and so there's a lot of hexane in the air from

25   gasoline, and we have a lot of automobiles too that are



 1   going to release it.

 2             DR. BLANC:  There's a lot of over-the-counter

 3   rubber cement for which this is still the solvent.

 4             DR. MARTY:  Yeah.  We can find that out, I'm

 5   pretty sure.

 6             DR. SALMON:  The study that we are involved in

 7   about in terms of the fuels projects, their predictive model

 8   and atmosphere model, we're coming up with hexane levels

 9   which were something between one and two orders of magnitude

10   below the proposed chronic REL for the South Coast air

11   basin.

12             DR. MARTY:  Thanks, Andy.  I had forgotten about

13   that.  That's the ARB's project on --

14             DR. SALMON:  So it varies quite a bit according to

15   the scenario, but basically it's what we learned.

16             DR. MARTY:  It's interesting too that Miyagaki in

17   the hexane makes the statement in his summary that mice are

18   much less sensitive to the neurotoxic effects than humans

19   and I want to know if it's because he's looking at Sanagi,

20   which has the potentiation problem or not.  So that makes me

21   want to look more.

22             CHAIRMAN FROINES:  One of the most difficult

23   things about -- hexane is a very good example of one of the

24   problems we have insofar as you get people who do

25   essentially short-term animal studies at high dose and we



 1   really have very little information on chronic low dose for

 2   neurotoxicity of this kind.  And it's because toxicologists

 3   do what's easiest, you know.  They give a dose to the animal

 4   over a period of couple weeks and they watch and see what

 5   happens.

 6             And even here you have a year, but for the most

 7   part it's just not the way people do these studies.  So that

 8   the issue of the chronic low dose effects is still, from a

 9   standpoint of neurotoxicity, it still seems to me to be a

10   problematic area.

11             It's pretty clear we have a lot of axonal

12   degeneration in people in Southern California, so you never

13   know where it comes from.

14             DR. MARTY:  I can keep going if this is useful,

15   but I don't want to bog the whole meeting down with our

16   examples.

17             CHAIRMAN FROINES:  How does the panel -- do you

18   want to go through all of them or is this --

19             DR. GLANTZ:  I don't think that's necessary.  I

20   mean, I think the points have been made.  We're not really

21   reviewing this for setting levels.  It's really for the

22   purposes of illustration.

23             I mean, again, despite all of the issues that are

24   being discussed, I was still impressed at how close the

25   numbers came out most of the time, which was the original



 1   issue.

 2             But I think we should move on to the hot spots.

 3             CHAIRMAN FROINES:  We're coming to Craig.

 4             DR. BYUS:  I did mine.

 5             CHAIRMAN FROINES:  You're right.

 6             The one I don't think we necessarily need to go

 7   through it, but one of the ones that has the very large

 8   difference is styrene, and styrene is an important chemical

 9   and maybe it would be worth taking a minute just to relook

10   at that for a second, because styrene is such a widely used

11   chemicals and the numbers are so --

12             DR. BLANC:  Discordant.

13             CHAIRMAN FROINES:  Discordant.

14             DR. MARTY:  For the styrene REL we used the

15   proposed US EPA RfC, which, incidentally if we had done it

16   would have been the same because we would have used an

17   interspecies uncertainty factor of ten rather than three,

18   and no modifying factors.

19             So but at any rate, the critical effects were

20   basically reflections of CNS toxicity in workers.

21             We have a NOAEL, so but the exposures, I guess the

22   8.6 years, but not too many more than that, so EPA had used

23   three and we just decided to go along with what they had

24   decided to use.

25             And the cumulative uncertainty factor is 30, so we



 1   get 1,000 micrograms per cubic meter as the REL.

 2             There was a '98 study by George Cruzan, et al, in

 3   Toxicological Sciences, which was essentially a chronic tox

 4   oncogenicity study in CD rats and they did not observe a

 5   NOAEL, but they have a LOAEL of 50 ppm.  It's a chronic

 6   study, so you don't need a subchronic uncertainty factor.

 7             But we did look at two ways to look at it.  If you

 8   have the LOAEL factor of ten or because the pathology was

 9   relatively low grade and low incidence, particularly in the

10   females, you could opt to use a LOAEL uncertainty factor of

11   three in that case.

12             So if you do that, you either end up with a

13   cumulative uncertainty factor of 300 or 100, and a REL of

14   either six parts per billion or 20 parts per billion,

15   depending on what you decide about the LOAEL uncertainty

16   factor.

17             So we have the difference between a cumulative

18   uncertainty factor of 30 and a cumulative uncertainty factor

19   of 300.

20             The Mutti study focused on central nervous system

21   effects.  The Cruzan study was looking at degenerative

22   changes in the olfactory epithelium, so you actually have

23   different endpoints.

24             DR. BLANC:  What is the central nervous system

25   effect we're looking at, just to refresh our memories.



 1             DR. MARTY:  It was a battery of neuropsychological

 2   tests, and they had a number of endpoints that were -- they

 3   had higher incidence in the exposed versus control.

 4             DR. COLLINS:  Verbal learning skills were

 5   significantly impaired in workers.  Logical memory and

 6   visual constructive ability were shown to be significantly

 7   affected in workers.

 8             DR. BLANC:  These were using the computerized

 9   battery of --

10             DR. MARTY:  Yeah.  There were a number.  There was

11   vocabulary block design, the digit symbol, 30-minute recall

12   of ten words and 30-minute recall of --

13             DR. BLANC:  The reference group was what?

14             DR. MARTY:  I believe it was workers in the same

15   factory, but not exposed.

16             DR. BLANC:  Well --

17             DR. MARTY:  Am I remembering that correctly?

18             DR. BLANC:  I would say that that's a difficult

19   endpoint, those batteries, and how they're interpreted, the

20   neuropsychiatric test batteries.  So that would be an

21   example I think of what John was referring to of taking

22   studies in the context of other studies.

23             So it's a very very -- those are endpoints which

24   are first of all they're multiple tests that are done so you

25   have one issue with multiple test comparisons.  Sometimes



 1   some of those batteries involve 30 or 40 different measures,

 2   so you have to read the study very carefully and see whether

 3   or not they have adjusted for multiple comparisons, which

 4   frequently is not done.

 5             Secondly, they are very very subject to

 6   confounding effects, which may have been controlled for if

 7   the reference truly was workers from the same facility or

 8   the same socioeconomic background, but it's very prone to

 9   confounding.

10             So this, quite a different weighting that one

11   would give to the degenerative changes of the olfactory

12   epithelium, which is again the kind of chronic effect which

13   makes a lot of sense to look at since some of the chemicals

14   that you're interested in do have -- probably do have

15   chronic olfactory effects.

16             DR. MARTY:  I think when you look at the database

17   as a whole, you do -- there are other studies besides Mutti

18   that talk about central nervous system effects of styrene at

19   relatively low exposures.  I think that was something that

20   was taken into consideration by EPA when they did their RfC.

21             And in this study also they looked at urinary

22   metabolites, so they had a fairly good exposure estimate and

23   they put the workers into two groups, based on their urinary

24   metabolites the following day.

25             DR. FRIEDMAN:  If the central nervous system



 1   effects are based on retesting the same individuals, are

 2   they before and after exposure?

 3             DR. COLLINS:  The 50 controls versus 50.

 4             DR. FRIEDMAN:  I was curious about what

 5   confounding there would be in that case, but apparently that

 6   wasn't the way the study was done.  I would think that's the

 7   way the study should have been done.

 8             DR. COLLINS:  Then you would have to get them

 9   before they were employed or to get --

10             DR. MARTY:  Workday --

11             DR. BLANC:  Acute effect, not a chronic effect.

12             DR. MARTY:  At any rate, there are a handful of

13   studies that have looked at reaction time and other things

14   that are also hard to measure and hard to interpret, but

15   that's the reason that EPA went with the human study.

16             DR. BYUS:  So, Paul, in a occupational setting are

17   those sorts of -- I had some concerns about it last time

18   just because I didn't understand what the tests were.  It

19   wasn't described very well.  But is that a valid way to

20   measure neurological effects?

21             DR. BLANC:  I mean, there is a discipline and it

22   is used and it's just something that you have to interpret

23   with caution.  I think when a lot --

24             DR. BYUS:  It's complex.

25             DR. BLANC:  It became sort of vogue in, I would



 1   say around 1980 or so, the Baker battery and all that stuff.

 2   It was popular, there was some batteries of computer

 3   administrable neuropsychiatric tests, but they're really

 4   challenged, I think, from a biostatistical point of view and

 5   an epidemiologic point of view because you're talking about

 6   batteries of 20 or 30 tests.  So from that, from the

 7   analytic point of view, people haven't really used, by and

 8   large, don't use sophisticated analyses where they look to

 9   see if the entire pattern is different, and then look at

10   individual subtests, and then from an epidemiologic point of

11   view it's very prone to a series of potential complicating

12   factors, including alcohol intake and educational intake.

13             When taken serially, which is not the issue here,

14   is a learning factor, it's just methodologically difficulty.

15             But in any event you come out with a much lower

16   level when you use the animal study in this particular

17   thing.

18             DR. COLLINS:  It's also a very given endpoint with

19   the nasal problem, and then you have this huge RGDR

20   calculation, which keeps driving things down.

21             DR. BLANC:  By a factor of four, though, right?

22             DR. COLLINS:  Five for the RGDR thing.

23             DR. BLANC:  That's because it's something in the

24   nose?  That's what we went through before.  Whereas if it

25   was truly an extrapulmonary factor --



 1             DR. COLLINS:  Extrathoracic.

 2             DR. BLANC:  This is extrathoracic, but if it was

 3   some kind of systemic absorption, then you don't use this

 4   factor.

 5             DR. COLLINS:  That's right.

 6             DR. BLANC:  So this only has to do with nose

 7   things.  Did I understand that from the last time?

 8             DR. FRIEDMAN:  I'm sorry I wasn't here at the

 9   beginning.

10             I'm just curious what is the bottom line of this

11   exercise?  Are you just trying to show us how different the

12   RELs are when you use animal versus human or you suggesting

13   that we always use the lowest or always use the human or

14   what?

15             DR. MARTY:  No.  What we're doing is responding to

16   Dr. Witschi's and Dr. Glantz at the last meeting suggested

17   that for the human, the RELs that were based on a human

18   study, that we look at a REL based on a animal study and

19   make the comparison.

20             DR. FRIEDMAN:  With what purpose?

21             DR. MARTY:  Just sort of an educational purpose to

22   see what we go through and the types of uncertainties and

23   how they're different and where they're different and

24   whether or not the numbers come even close.

25             DR. FRIEDMAN:  If they don't, do you have an idea



 1   of which one you would pick?

 2             DR. MARTY:  We have tended to bias towards the

 3   human studies, just because of the uncertainty with

 4   extrapolating from an animal to a person.  So that's -- and,

 5   you know, we've tried to go with human studies that we

 6   thought were fairly well conducted.  And they all have some

 7   flaws.  There's just no way around that.

 8             But the general thinking is it's better to use

 9   humans if you're trying to figure out what happens in a

10   human, than to use the rat.

11             And, you know, I was actually somewhat surprised

12   at the result that the styrene, that it came out so much

13   lower from the rat study.

14             CHAIRMAN FROINES:  What do you get if you use

15   Guillemin?

16             DR. MARTY:  I don't know.  That's, how did you say

17   his name?

18             CHAIRMAN FROINES:  It's Swiss, and so I can't ever

19   get it right.

20             DR. MARTY:  I don't know.  We could whip that out

21   and see what happens.  I know --

22             DR. FRIEDMAN:  Do you plan to modify the report to

23   include these comparisons or not?  Just for our education.

24             DR. MARTY:  Right.  We hadn't planned on doing

25   that.



 1             And also we did it with a fairly fast turnaround,

 2   so it's not like we went and hunted through all the

 3   literature that we may have missed or --

 4             CHAIRMAN FROINES:  Whose study is this, the

 5   Cruzan?

 6             DR. MARTY:  Cruzan.

 7             DR. COLLINS:  It's from several industries.  One

 8   woman is at Chevron and he's at Tox Works in New Jersey.

 9   Somebody else from Arco.  Someone else from Hunting Life

10   Sciences.

11             CHAIRMAN FROINES:  It's an industry chronic animal

12   bioassay?

13             DR. COLLINS:  That's correct.  Mainly looking

14   at -- they were mainly interested in not finding

15   carcinogenicity and that's what they did.

16             CHAIRMAN FROINES:  They're interested because of

17   the IARC changes in the classification for styrene.

18             DR. COLLINS:  It was funded by the SIRC, the

19   Styrene Industry Research --

20             DR. MARTY:  But it was primarily focused on

21   oncogenicity, rather than other endpoints.

22             CHAIRMAN FROINES:  Any other comments?

23             I think Paul's points are very well taken.  These

24   studies that were so popular in the late '70s and early '80s

25   are really troubled by confounding historically.  And it's



 1   worth trying to look at that in detail.  I mean, there were

 2   lots of styrene, but the old studies where people were

 3   looking at JP 4 jet fuels and they weren't taking into

 4   account alcohol intake.  I mean it was ridiculous.  And in a

 5   heavily drinking population too.  So it was these things

 6   come up over and over again.

 7             DR. MARTY:  Yeah.  If I'm not mistaken, Mutti they

 8   did in that study look at alcohol intake as a huge problem.

 9             DR. BLANC:  It's a bit less an issue for you with

10   the styrene, because we'd be concerned about all those

11   confounders to the extent that they would have somehow given

12   you an oddball, you know, positive finding.

13             But since your real policy issue is going to be

14   whether or not you used this more recent study, which US EPA

15   did not have available to it when they did their analysis,

16   and whether or not your cumulative uncertainty factors

17   should be 300 or 100, depending on the LOAEL uncertainty

18   factor.

19             I think those are your methodologic issues, not so

20   much because the animal study would give you a more

21   conservative value in this particular case.

22             DR. MARTY:  Yes.

23             DR. BLANC:  So since we've been going for almost

24   an hour and a half, we should take a break for your

25   reporter.



 1             CHAIRMAN FROINES:  I should say just before

 2   leaving this that the Swiss investigator is an extremely

 3   fine investigator, and that's one of the factors that we

 4   haven't really talked about was the quality of the

 5   scientists, as well as the science, so it's always worth

 6   noticing when you see somebody you know that does good work.

 7             Let's take a break for ten minutes and then we'll

 8   go on.

 9             (Thereupon a short recess was taken.)

10             CHAIRMAN FROINES:  Melanie, I had one question.

11             Are you going to make a presentation?

12             DR. MARTY:  No.

13             CHAIRMAN FROINES:  On methyl bromide.

14             DR. MARTY:  No, not on methyl bromide.  We have

15   for the second agenda item.

16             DR. COLLINS:  We deferred methyl bromide.

17             CHAIRMAN FROINES:  You deferred methyl bromide.

18             DR. MARTY:  In the meantime I've spoken to

19   Dr. Glantz, who the lead on methyl bromide, and also the

20   panel got a response to comments sent to them a couple weeks

21   ago from -- it was our response to the methyl bromide

22   industry panel comments, but we don't have a presentation,

23   but when Stan comes to -- when he's at bat, we're going to

24   talk about what the issue was with the methyl bromide.

25             CHAIRMAN FROINES:  I think that's important to be



 1   as thorough as possible on that particular substance.

 2             We're back on the record.

 3             Because there is obviously sufficient interest,

 4   significant interest.

 5             So do you have more before we turn it back to the

 6   panel?

 7             DR. MARTY:  No.

 8             CHAIRMAN FROINES:  Then let's turn it back to Paul

 9   Blanc.

10             We're turning the heat up, literally.  The

11   temperature in the room is going to increase, we hope.

12             DR. BLANC:  So, Melanie, I'm going to start with

13   ammonia.

14             DR. MARTY:  Okay.

15             DR. BLANC:  Because I think we didn't do ammonia.

16             DR. MARTY:  Let's go with ammonia.

17             DR. BLANC:  Then I'll go through in alphabetical

18   order maybe.

19             Here's a generic comment on the document.

20             When you do physical properties, I think it would

21   be more consistent, and I think this came up in a previous,

22   and maybe in the acute, if you would give us physical

23   properties at some kind of standard temperature and

24   pressure.  You know, anhydrous ammonia may be a liquid on

25   the planet Jupiter, but not on the surface of Earth.



 1             DR. MARTY:  Dr. Fucaloro has already pointed that

 2   out to me.

 3             DR. GLANTZ:  But, Paul, we should take a broad

 4   view.

 5             DR. BLANC:  Secondly, I would like to see for each

 6   inhalation reference exposure level where it's done at the

 7   beginning, to also include the parts per million, that

 8   conversion, so that somebody doesn't have to do that.

 9             DR. MARTY:  Okay.  Right up in the front.

10             DR. BLANC:  I know you have it other places.  You

11   have the conversion factor, but when you say inhalation its

12   reference level, for example, for ammonia, is hundred

13   micrograms per meter.  I think in parentheses it would be

14   helpful to have the parts per million.

15             Another generic comment is that there's a lot of

16   heterogeneity in the document in major uses and sources in

17   that section.

18             Now, I don't think each of these things needs to

19   be a NIOSH criteria document, but it does seem to be overly

20   telegraphic in places and this is one where it does.  For

21   example, a major use of ammonia is as a refrigerant.  That's

22   not listed.

23             Secondly, the whole section on ammonia disregards

24   that the major public exposure to ammonia is in ammonia in

25   solution.  So unless you know something about ammonia, it's



 1   really hard from this document to understand the difference

 2   between anhydrous ammonia and aqueous ammonia solutions,

 3   which are going to off gas them, ammonia or aerosolized

 4   ammonia.  So I think it would be useful -- and the place to

 5   do that I think would be in the major uses or sources and

 6   maybe to say something about commercial over-the-counter

 7   ammonia containing products as compared to industrial

 8   ammonia products which tend to be 25 percent ammonia or

 9   whatever.

10             I actually don't know what the max you can go to

11   in an industrial solution, but you can get pretty

12   concentrated aqueous ammonia.

13             So that would be a comment there.

14             This is a fairly brief section on what is a fairly

15   major industrial chemical, and that may simply be driven by

16   the extremely limited nature of any kind of chronic exposure

17   data, so I recognize that.

18             And there may need to be some generic approach to

19   the document which explicitly acknowledges that inconsistent

20   way.

21             And it may be that what you need in addition as

22   one -- an additional fixed section, you know how you have

23   one, two, three, four, five, six, et cetera.  Maybe, you

24   know, a section for each one needs to be data limitations,

25   which has animal data, you know, limited modern excellent



 1   human data, limited modern excellent, you know, overall

 2   quality of data including recent publication methods, study

 3   size, some kind of comment like that for each section.

 4             CHAIRMAN FROINES:  Can I ask a question.

 5             Melanie, when you have document like this and say

 6   let's just use ammonia that has, following up on what Paul

 7   said, has basically four pages, but that most of which is in

 8   references, when if there were a person who submitted

 9   comments, they are kept in some kind of record that's

10   accessible to the public.  If somebody wanted to read the

11   comments as well as this short text, because one of the

12   things that does bother me about we're so used to the big

13   fat documents that are this thick for diesel, and we end up

14   with files like this and these turn out to be the other

15   extreme.  So in a way you don't have as much as you might

16   like.  On the other hand you may not need it.  But if people

17   do have comments you do need to have that in a place that

18   becomes accessible, it seems to me.

19             DR. MARTY:  They're on the Web.  What we do is we

20   pretty much take the comments that are sent in to us, we

21   scan them in and we respond.  We respond -- we try to keep

22   the comments as verbatim as possible to avoid us

23   paraphrasing and misinterpreting the comment, but that is

24   the material that the panel gets and that is available on

25   the Web.  And if people are interested in actually looking



 1   at the letter they came in, they can always do that also.

 2   So we've got that in our files.

 3             DR. GLANTZ:  Yeah.  That's another document we got

 4   was the comments documents.

 5             CHAIRMAN FROINES:  I know we get it.  I was just

 6   thinking about the public.

 7             DR. BLANC:  Just another example where you have a

 8   mixture of critical effects.  The one where you're on more

 9   solid ground I think is pulmonary function and then the eye,

10   skin and respiratory symptoms, irritation are more problem

11   ridden for all the reasons that we said, and I think that

12   there does need to be some kind of generic approach in the

13   document when you're dealing with that.

14             The corollary to pulmonary function would be, you

15   know, studies of olfactory deficits.

16             And I forget whether some of this human exposure

17   data had that.  That's something that for ammonia, because

18   of the high water solubility is particularly the area in

19   which I would give a clear look, like your case, even more

20   so, since there's good evidence that acute overexposure of

21   formaldehyde can cause an acute olfactory deficit.

22             But you have some pulmonary functions.

23             Okay.  So should I move on?

24             One other thing.

25             The number that you come out with at the end is .1



 1   part per million which you then translate as 100 micrograms

 2   per cubic meter, but your conversion factor would say that

 3   it's 70 micrograms per meter, 71 micrograms per meter, and

 4   there's an area where I think there's too much rounding, if

 5   that's what you're doing.

 6             DR. MARTY:  Okay.

 7             DR. BLANC:  I mean, just as a reader it was

 8   confusing to me.

 9             Chlorine we did, so now we can go to dioxane.

10             CHAIRMAN FROINES:  Can I make a generic comment.

11             Melanie, you know what I'd like to see you do with

12   these, since we have a hundred million of these chemicals --

13   no, I know.

14             DR. COLLINS:  It seems like that way.

15             CHAIRMAN FROINES:  I'd like to have a little

16   notation with each chemical where ARB has done airborne

17   monitoring and there's actual exposure or airborne

18   monitoring data.  So that I don't know whether you've done

19   it for styrene, I don't know whether you've done it for

20   methyl bromide, I don't know whether they've done it for

21   ammonia, but in the document, right in the beginning, if you

22   just had a little thing, air monitoring, yes, air

23   monitoring, no, then the reader would know that they could

24   go to, presumably go to the Web and look for any air

25   monitoring data that may exist, so the reader at least knows



 1   that somebody has -- that there may be some comparable air

 2   monitoring information that they can then compare the REL to

 3   something that somebody is actually taking a look at it.  I

 4   think it would be helpful to everybody who might be

 5   interested in these issues.

 6             DR. MARTY:  Okay.  Also I'll see if they have any

 7   modeling information based on some of their modeling

 8   efforts.

 9             DR. BLANC:  Dioxane if I -- there seems to be an

10   arithmetic error, again this can't even be a rounding error,

11   the inhalation reference exposure levels is listed as 3,000

12   micrograms per meter, but it is at the end of the document

13   3.8 milligrams per meter, which should be 3800 micrograms.

14             DR. COLLINS:  That should be 2.8, and actually it

15   came out to be 2.88 milligrams per cubic meter.  The three

16   was an error and it rounded to 3,000.  Sorry.

17             DR. BLANC:  At least I knew something wasn't

18   right.

19             DR. COLLINS:  I like to say, we did look up that

20   review you gave us and they basically said the animal study

21   available was this study.

22             DR. BLANC:  Okay.  Good.

23             And I think that in the major uses section here

24   there is one thing I would add, and that is, unless it's

25   changed, dioxane had a major use as a stabilizing additive



 1   to chlorinated hydrocarbon solvents, such as

 2   1,1,1-trichloroethane, and I think probably that's the major

 3   source of exposure would be my guess.

 4             DR. MARTY:  Yeah.  It's my understanding that that

 5   use has declined, but I can --

 6             DR. BLANC:  Can you double check that?

 7             DR. MARTY:  I can check with ARB folks and ask

 8   about that.

 9             DR. BLANC:  Because historically that was

10   certainly --.

11             DR. MARTY:  Yes, correct.

12             DR. BLANC:  Since you're basing in part your

13   reference exposure level on a study by Thiess et al, if I

14   understand the section correctly, which is described as in

15   German described in NIOSH 1977 and not otherwise listed in

16   the reference, meaning that you got everything you got about

17   it from the NIOSH summary and you never pulled it yourself,

18   that study.

19             DR. MARTY:  I'm not sure why it's not listed.

20             DR. BLANC:  Well, that's what I would suspect when

21   I read it that way.  That's nicht gut.

22             DR. COLLINS:  Which one?

23             DR. MARTY:  It's not in here.  That's what he's

24   saying.

25             DR. BLANC:  So please pull that original.



 1             I think what I think would be more precise when

 2   you talk about the enzyme abnormalities, transaminase

 3   elevations, particularly, I think the language that I would

 4   suggest you use in this section and elsewhere, if there are

 5   other cases where that's your endpoint is that you talk

 6   about abnormal values, rather than increased activities,

 7   because it's not really increased activities.  What it is is

 8   liver cells leaking out the enzymes into the circulation, so

 9   it's not really an increased activity in that sense.

10             DR. MARTY:  Right.  Okay.

11             DR. BLANC:  It would be more precise, I think, to

12   say that.  And you may want to use along with the older

13   terminology, which is SGOT and SGPT, the preferred current

14   is AST and ALT.

15             DR. MARTY:  Okay.

16             DR. BLANC:  Even though I'm sure when they

17   published it -- and these are really measures of liver

18   dysfunction.  I know they're called liver function tests,

19   but just to clarify.

20             And there's also in the next paragraph where you

21   say dioxane levels ranging from less to 25 to greater than

22   75 parts per million.  Do you mean ranging from 25 to 75

23   parts per million?  It's page A-50.

24             DR. MARTY:  Let's check.  I see what you mean.  It

25   doesn't make sense.  I think --



 1             DR. BLANC:  You mean between 25 and 75.  Maybe the

 2   arrows got reversed.

 3             DR. MARTY:  Exactly.

 4             DR. COLLINS:  These are the others.

 5             DR. MARTY:  We will check that, because Dave Lewis

 6   just informed me that it could be that there was groups

 7   exposed to less than 25 ppm and groups exposed to greater

 8   than 75.

 9             DR. BLANC:  However it is, you should just

10   clarify.  It's not very clear.

11             And then I guess in the next page there was a

12   whole discussion about a rat study by Torkelson from 1974

13   where they had increased red blood cells and decreased white

14   blood cells.  I think you should say concentrations, if

15   that's what you mean, by the way.

16             DR. MARTY:  Counts or concentrations.

17             DR. BLANC:  Or something.

18             And then there's this sort of this comment a

19   little bit later, although Pilipyuk et al, 1978 considered

20   the changes to be a reflection of adverse effects due to

21   exposure to dioxane, Torkelson et al in 1974 do not consider

22   the hematological and clinic changes of toxicologic

23   importance.

24             And could you tell me what --

25             DR. MARTY:  Purpose of that statement is?



 1             DR. BLANC:  Yeah.  Because you go on, you use in

 2   part the Torkelson study --

 3             DR. MARTY:  I think we're just pointing out --

 4             DR. BLANC:  As a no effect level.  But I assume

 5   that these effects were seen, right?

 6             DR. MARTY:  Yes.

 7             DR. BLANC:  You used it as a NOAEL and not a

 8   LOAEL, because you're discounting the fact that there were

 9   increased concentrations of red blood cells and white blood

10   cells and decreased alpha and phosphatase, which doesn't --

11   I think it doesn't have a lot of meaning whether it's

12   decreased.

13             DR. MARTY:  We're going to look at the paper.

14   It's not clear.

15             DR. BLANC:  And then such dysfunctions, however,

16   were observed in rats exposed to dioxane by ingestion, and

17   in humans.  So I really found this whole thing rather

18   confusing.  That's sort of the bottom line.

19             DR. MARTY:  Looks like we were trying to figure

20   out why Pilipyuk et al '78 and Torkelson et al '74 seem to

21   be not arriving at the same conclusion.

22             DR. BLANC:  Because basically isn't what's going

23   to happen if you accept that those do matter, then you're

24   going to have to use 111 as a lowest effect level and not as

25   a no effect level and then that's going to drive everything



 1   down by a factor of --

 2             DR. COLLINS:  Three or ten.

 3             DR. BLANC:  Three or ten, depending.

 4             DR. MARTY:  Torkelson is saying that the minor

 5   differences between a control and exposed rats with respect

 6   to the hematological values were within normal physiological

 7   limits and were not considered of any toxic importance.

 8             DR. BLANC:  But they were statistically unlikely

 9   to be due to chance.

10             DR. MARTY:  Correct.

11             DR. BLANC:  And then but on the other hand the

12   study in German, which you haven't pulled, in humans, which

13   is described by NIOSH, did observe such changes, maybe, if I

14   understand what you mean.  But I don't know what levels.

15   So, you know, it's really badly put together.  It's put

16   together in a way which I can't figure out whether it's --

17             CHAIRMAN FROINES:  I think what I'm going to do is

18   take the prerogative of the chair and say this one should be

19   redone.

20             And I think that you really should avoid

21   statements like that.  Torkelson et al do not consider the

22   hematologic and clinical changes of toxicologic importance.

23   Quite frankly, that is a highly self-serving statement.  And

24   if you go back and look at that paper, it's not entirely

25   surprising.



 1             So I think that one should look at that with some

 2   care.

 3             I think Paul is right about the --

 4             DR. BLANC:  I'm not going to revisit the

 5   formaldehyde, hydrogen chloride and cyanide, because --

 6   actually I never did cyanide.  That's the one I need to do

 7   also, is that right?

 8             DR. MARTY:  We talked about formaldehyde.

 9             DR. BLANC:  Formaldehyde I did.

10             DR. MARTY:  I'm not sure you finished.

11             DR. BLANC:  Chlorine we did.

12             DR. GLANTZ:  All the ones that were discussed you

13   finished.

14             DR. BLANC:  Let me do hydrogen cyanide as the last

15   one.  That's under H.

16             Here's an example where we really have lousy

17   chronic human data.  And I know that the EPA used, and I

18   know because I've contributed some of the lousy chronic

19   hydrogen cyanide data.  Although, I don't think my study was

20   any worse than El Ghawabi.

21             And it would seem to me that here's an effect for

22   which animal data of some sort must be more informative, and

23   particularly I don't see why you couldn't use animal dietary

24   values, because for inhalation we have good data to tell you

25   how much the delivered systemic dose would be from



 1   inhalation, and there's nothing particular about the

 2   inhalation other than it being one route of exposure.

 3             So since there's a lot probably of dietary chronic

 4   cyanide feeding data in animals, I would assume -- it seems

 5   to me it would make much more sense to extrapolate from

 6   feeding studies in animals than to use the El Ghawabi study

 7   of metalplaters in Egypt.

 8             DR. COLLINS:  We do have this chronic animal

 9   study, which is in the summary in which we used two studies

10   from Scandinavia about 1980, which ended with inhalation

11   reference level of two micrograms per cubic meter, so in

12   that case the animal data went along well with the lousy

13   human data.

14             DR. BLANC:  They were given fairly short shrift in

15   summary, though.

16             DR. COLLINS:  There's EPA -- deference to the EPA

17   at that time.

18             DR. BLANC:  Yeah.

19             DR. MARTY:  Yeah.  We could -- that was one of the

20   other examples I had where we compared the human versus the

21   animal.

22             DR. BLANC:  Right.

23             DR. MARTY:  And in human we have a cumulative

24   uncertainty factor of a thousand and the animal we have a

25   cumulative uncertainty factor of 300, so that tells you



 1   right there that there's issues with both studies.  But we

 2   do end up with in one case three micrograms per cubic meter

 3   and the other case two micrograms per cubic meter.

 4             Also I do want to comment on the route-to-route

 5   exposure extrapolations.  We do them as a last resort.  The

 6   issues are primarily pharmacokinetic differences which may

 7   or may not be well characterized.

 8             DR. BLANC:  I'm just saying for cyanide it's one

 9   of the ones where you have less trepidation of all the

10   things I could think about.  We know it's absorbed very well

11   systemically from diet or from inhalation, rapidly lethal in

12   the acute model.  So, you know, reason to think it's pretty

13   well distributed to where it matters.

14             DR. MARTY:  We can certainly look at it.  We can

15   look at the Hugod studies of '79 and '81 and also some

16   dietary studies and see what --

17             DR. BLANC:  Or put in your final table, you know,

18   make the animal study appear more prominently in your -- I

19   just, that's just my take on it.  Since you have animal

20   data, which is very supportive, I would probably put a

21   little more prominently, it just makes your argument a lot

22   more strong, a lot stronger, because the El Ghawabi study is

23   both in terms of what, you know, how they measured the

24   thyroid stuff and --

25             DR. MARTY:  Okay.  We'll actually in the document



 1   then discuss the REL derivation from the animal studies and

 2   the fact that it's close and then we'll also go ahead and

 3   take a look at the oral studies.

 4             DR. BLANC:  And actually cyanide, despite there

 5   aren't good chronic human inhalation data because, thank

 6   goodness, there aren't occupational settings where people

 7   are chronically exposed to levels of cyanide that are that

 8   high, by and large.

 9             DR. COLLINS:  The only comment we had was from the

10   mining industry, which is because of the large amount of

11   cyanide used there.

12             DR. BLANC:  But we do have really large

13   epidemiologic studies of dietary cyanide, so it is an

14   epidemiologic, chronic epidemiologic issue and there

15   probably are really important and chronic health effects

16   from chronic cyanide exposure.  Where at risk people are --

17             CHAIRMAN FROINES:  I spent quite a while looking

18   at chronic neurologic effects of cyanide.

19             DR. BLANC:  Okay.  That's it.  I'm done.

20             CHAIRMAN FROINES:  You're done with everything.

21             So we can shift over to -- Craig's finished.

22             So let's shift over to Stan.

23             DR. GLANTZ:  Okay.  Well, I have to say that I

24   read all the stuff getting ready for the last meeting and

25   thought that -- and there were a lot of public comments on



 1   several of the chemicals that you assigned me, and I thought

 2   OEHHA actually did a pretty reasonable job.

 3             And then I listened to all of Paul's detailed

 4   comments, and I felt inadequate, so I went back and read it

 5   all again and I still thought --

 6             DR. BLANC:  I told him it was criticism envy.

 7             DR. GLANTZ:  Yes, it was.  It was criticism envy.

 8             I still thought OEHHA did a pretty reasonable good

 9   job.  I don't have a whole lot to say, actually.

10             I just want to clarify a couple of points and that

11   is on the isopropanol, that was one place where the

12   commenters made a lot of very specific suggestions for

13   changes, and OEHHA, seemed to me, accepted almost all of the

14   changes.  And I just wanted to make sure that AL is reading

15   that right.  I thought it was appropriate to accept it, by

16   the way, and that the document that we have before us

17   reflects all those changes.

18             DR. MARTY:  Yes.

19             DR. GLANTZ:  I don't know, I don't think it's

20   worth going through them, but that was one place.

21             And then the other one that I thought was worth

22   some discussion is the issue of methyl bromide, because

23   that's one where there were a bunch of comments received

24   fairly late before the last meeting, and issues relating to

25   the pathology that underlied the study that they used, and



 1   whether the number, the exposure in the study should be

 2   considered NOAEL or a LOAEL.  And I just think it would be

 3   worthwhile for the panel letting OEHHA kind of review that

 4   correspondence and the comments and how it all worked out.

 5             Again, I think the resolution was fine, but I just

 6   think, given the back and forth, it would be worth a brief

 7   discussion of what happened.

 8             Is that okay?

 9             DR. MARTY:  That's fine.

10             DR. GLANTZ:  Do you want copies of the letters?

11             CHAIRMAN FROINES:  Could I ask a question?

12             DR. GLANTZ:  Yeah.

13             CHAIRMAN FROINES:  Why did you use EPA RfC?

14             DR. GLANTZ:  For which one?

15             CHAIRMAN FROINES:  Methyl bromide.

16             DR. MARTY:  We did.

17             Why did we?

18             DR. GLANTZ:  Can we just go through this other

19   thing first and then if --

20             DR. MARTY:  I think I can sum it up pretty

21   quickly, and that is that the methyl bromide industry panel,

22   CMA, sent in a letter saying that they had requested another

23   pathologist to review the slides of the histologic lesions

24   in the olfactory epithelium.

25             DR. GLANTZ:  It was a study they used to get the



 1   REL.

 2             DR. MARTY:  Right.  And Dr. Hardesty, who was the

 3   pathologist, reviewed the slides and was in general

 4   agreement with the higher dose that with the initial study

 5   pathologist with regards to the impacts on the olfactory

 6   epithelium at the higher doses, but at the lower doses there

 7   was disagreement as to whether or not there was in fact any

 8   histopathology.

 9             The methyl bromide industry panel concluded that

10   therefore that that dose, rather than treating it as a

11   LOAEL, as was done, should be a NOAEL.

12             So I contacted the methyl bromide industry panel

13   and asked them to have the original study pathologist send a

14   letter stating that she agreed with the reevaluation.

15             And I got a letter and she agrees with it to an

16   extent, but if you look at where she disagrees, that low

17   observed effect level is still a low observed effect level.

18   She thinks that there were effects in one sex that are

19   statistically significant.

20             However, yesterday or the day before, they sent me

21   another letter from the original study pathologist, in which

22   she implies that it's such a minor difference that in her

23   opinion we shouldn't use more than an uncertainty factor of

24   two, and we have used an uncertainty factor of three.

25             So I am suggesting that we stay with the



 1   uncertainty factor of three for the extrapolation from the

 2   LOAEL to the NOAEL.

 3             So that basically sums up the issue.

 4             DR. GLANTZ:  So the bottom line of all this

 5   correspondence is nothing changed.

 6             And I think that the decision by OEHHA to leave

 7   things as they were is justified.

 8             So that was basically all that I had to say on

 9   these compounds.  I mean, if you guys, I'm sure Paul will

10   have incredibly insightful comments or something, but I read

11   through all of the stuff several times and I'm satisfied

12   with OEHHA's responses to the public comments.  And in

13   reading the report itself it all seemed reasonable to me,

14   subject to the changes that they made in response to the


16             DR. BYUS:  I read with great care all of the

17   correspondence with the pathologists and all that you sent

18   forward, and I find it a little disturbing when they

19   reanalyze the pathology and don't do it in a blind way

20   fielding a panel.

21             And then where they assess the quality of the

22   panel by giving them known slides in a blind way, and then

23   they -- then you look at the slides in a blind way by a

24   panel, recording votes and then do it.

25             The way this methyl bromide was not done that way.



 1   The pathologist, from my reading of it, knew what he was

 2   looking at when he looked at it.  And I find that not

 3   necessarily a good way to do it.

 4             And I think my reading of your interpretations of

 5   all this is very good.  So I mean I think you did exactly

 6   the right thing.

 7             And then the original letter it appeared that I

 8   had the impression that the original pathologist agreed

 9   completely with the re-evaluation, and then when you

10   actually read her letter, which was nice for you to get, it

11   was some letter, that was not the case.

12             And so I think there is a lot of times when

13   reanalysis of the pathology is a good idea, but it has to be

14   done under, say, good laboratory practices guidelines or

15   definitely blind, and preferably a panel of pathologists,

16   rather than one.  So.

17             DR. GLANTZ:  I guess just for the record I had

18   very similar feelings and I'm really glad that you went back

19   and contacted the pathologist, rather than taking a

20   secondhand assessment of what -- I think it was a woman,

21   wasn't it, what she said?

22             DR. MARTY:  Yes.

23             DR. GLANTZ:  But as I say, I thought these were

24   relatively well done and the responses to the public

25   comments were good and that when the commenters had made



 1   good points there were appropriate changes made to the

 2   report.

 3             So I think we can, unless somebody else has

 4   something, we can just go on.

 5             CHAIRMAN FROINES:  Why did you not, going back to

 6   my question, why did you not do -- why is this listed as an

 7   US EPA RfC?  Because what you're -- because the way you're

 8   talking about the study, it's as though you did an

 9   evaluation such that the number that you're using is in fact

10   a number that you would have come up with had you described

11   it yourself, and if that's the case, I'd rather you said

12   it's your finding, not an EPA finding.

13             DR. MARTY:  Okay.  I can say that when we looked

14   at the EPA RfCs, we did go back and look at all of their

15   documentation and pulled the original studies and so on.

16             If we really had heartburn over what we did, then

17   we decided we weren't going to forward it on as a proposal.

18             If we agreed that the study chosen was appropriate

19   and the endpoints were appropriate, there may have been some

20   slight disagreement over the modifying factor, which I

21   pointed out earlier, but we in general we did look at the

22   information that was used as the basis of the EPA RfC and in

23   this case we do agree with that choice of study and

24   et cetera.

25             CHAIRMAN FROINES:  The reason I say that is more a



 1   policy than a science statement in the sense that methyl

 2   bromide is a very important compound in the State of

 3   California, and it seems to me that it's worth saying that

 4   OEHHA, these are OEHHA's findings and this is your

 5   conclusions, rather than saying this is EPA's finding and we

 6   sort of agree with them.

 7             I would think it would be stronger if you said

 8   this is the OEHHA position on the chronic noncancer effects

 9   of methyl bromide.

10             And methyl bromide obviously is a, you know, we

11   spend a fair amount of time on dioxane, which does have

12   virtually no exposure by relative basis, by comparison, so

13   that this is important.

14             Do we know -- I was hoping Lyn Baker would be

15   here, but there's no reason why he should have been here, so

16   do we have some sense of how this value compares to what we

17   found in air monitoring from methyl bromide?

18             DR. MARTY:  We don't.  We will follow up with Lyn

19   and find out what he knows.

20             CHAIRMAN FROINES:  If you take the NTP study --

21   I'm glad I asked.  It's nice to see George get out of his

22   chair in this meeting.

23             DR. ALEXEEFF:  George Alexeeff with OEHHA.

24             Just to let you know about methyl bromide, it's

25   currently DPR has prepared a risk characterization document



 1   on methyl bromide.  For this particular chronic effect they

 2   have the same value.  They have developed the same value as

 3   this number.  It's a slight different way they calculate it,

 4   but it ends up being the same number.

 5             And their -- since the document is a risk

 6   characterization document and not yet a TAC document, okay,

 7   they are having it right now peer reviewed by the National

 8   Academy of Sciences Committee, which is supposed to be done

 9   fairly shortly.  I believe at some point it will come also

10   to the panel when it's further completed, but it's a

11   slightly different format, for some reason, that they chose

12   that.

13             CHAIRMAN FROINES:  Does the National Academy

14   Committee do methyl bromide, per se?

15             DR. ALEXEEFF:  No.  They're just paying the

16   National Academy of Sciences to peer review it on some sort

17   of expedited basis.

18             So but there is some -- so that is undergoing peer

19   review.  I think it will be done sometime in the next few

20   months.

21             So they probably do have levels that they

22   measured.  They have had levels.  I've seen a draft of the

23   document that's available.  But it's not a TAC document that

24   they have.  It's what they call -- it's prior to the TAC,

25   but they felt because of the findings they had in that risk



 1   characterization document, they felt to be peer reviewed

 2   before they even developed a TAC document.

 3             So that's the status of that.

 4             CHAIRMAN FROINES:  Well, we'll put on the record

 5   that this was a matter that Dr. Froines and Bill Lockett and

 6   Jim Behrmann will follow up on, because we'd like this to

 7   be -- we'd like something -- the panel to be informed as to

 8   what's happening on methyl bromide because it is such an

 9   important compound.

10             DR. ALEXEEFF:  They could possibly give a status

11   next meeting.

12             CHAIRMAN FROINES:  If you, back to science, if you

13   look at the NTP bioassay and you had a dose level for mice

14   that was down to as low as ten, do you remember what the

15   findings were with the mice in terms of the various

16   parameters?

17             In other words, did they find effects in at ten

18   parts per million, because your LOAEL is three parts per

19   million in rats.  Do you remember what the NTP findings were

20   in mice at ten?

21             DR. MARTY:  I don't.  I need to go back and look

22   at that.

23             CHAIRMAN FROINES:  Again, it goes to what I was

24   raising at the very beginning, which is looking for

25   consistency across findings, and that's obviously is a



 1   well-done study and it's obviously a dose range that has

 2   relevance to your Reuzel study.

 3             DR. MARTY:  I'm looking at our document on page

 4   153, it does say a low incidence of sternal dysplasia and a

 5   significant decrease in locomotor activity were noted in the

 6   ten ppm group.  This is --

 7             CHAIRMAN FROINES:  I don't think I have that.

 8             DR. MARTY:  Fourth paragraph down.

 9             CHAIRMAN FROINES:  I don't have that document.  My

10   methyl bromide is A-470.

11             It doesn't matter.  Let's just --

12             DR. COLLINS:  Are you looking at the original

13   October 1997 draft?

14             CHAIRMAN FROINES:  Yeah.

15             DR. COLLINS:  I'm sorry.  May 1999.

16             DR. BLANC:  That's all right.  He's always a

17   couple years behind.

18             DR. COLLINS:  I don't think this changed much.

19             CHAIRMAN FROINES:  I know.

20             DR. COLLINS:  An exposure of mice --

21             CHAIRMAN FROINES:  I left the other one.

22             DR. COLLINS:  An exposure of mice, 86 animals per

23   group, to 0, 10, 33, a hundred ppm methyl bromide for six

24   hours a day, five days a week for 103 weeks, a low incidence

25   of sternal dysplasia and a significant decrease in locomotor



 1   activity where noted in the ten ppm group.

 2             Would you like to know what -- we can look up what

 3   the incidence is.

 4             CHAIRMAN FROINES:  No, that's okay.

 5             DR. BLANC:  Stan.

 6             DR. GLANTZ:  If you'd like.

 7             DR. BLANC:  Can we go to inorganic mercury.

 8             For consistency, first of all, I think it should

 9   be mercury, comma, inorganic, since you're going in

10   alphabetical order and you have it under M.

11             DR. FUCALORO:  I was looking it up under I.

12             DR. BLANC:  Okay.  Let's go to the chemical

13   property summary.

14             First of all, clearly the critical thing about

15   inorganic elemental mercury is that although it's a heavy

16   liquid, at room temperature it's easily vaporized.

17             And is the vapor pressure correct?  The vapor

18   pressure that's listed there.

19             DR. MARTY:  I can check it against another source.

20   We used HSDB, a secondary source, and we found it's full of

21   mistakes.

22             DR. BLANC:  Doesn't it have a high vapor pressure?

23             DR. FUCALORO:  I don't know.

24             DR. BLANC:  I mean, it's weird because it is

25   mercury and it's vapor pressure and mercury --



 1             DR. FUCALORO:  You know, the modern symbology for

 2   millimeters mercury is torr, t-o-r-r.  And I only mention

 3   that because it was named after that famous Italian

 4   scientist Torracelli.

 5             DR. COLLINS:  You mean the guy from New Jersey?

 6             DR. BLANC:  There's a tendency to shorten last

 7   names, so it's a good thing there's nothing in your name.

 8             DR. FUCALORO:  There's plenty of them.

 9             DR. MARTY:  We will double check the vapor

10   pressure.

11             DR. BLANC:  You seem to be dealing in this section

12   with both inorganic elemental mercury and then with mercuric

13   chloride?  Are those the two things, or have I got --

14             DR. MARTY:  Mercuric.

15             DR. BLANC:  That the same thing as sublimate, you

16   know, what they used to call --

17             DR. COLLINS:  It may be.  Corrosive sublimate.

18             DR. BLANC:  Corrosive sublimate, I'm sorry.  Is

19   that the same thing?

20             DR. MARTY:  I think so.

21             DR. BLANC:  Those are the only two substances

22   you're dealing with here, because those are the only two

23   things that would become airborne with any likelihood; is

24   that right?

25             DR. COLLINS:  Maybe.



 1             DR. BLANC:  As opposed to a whole bunch -- I mean

 2   there are a whole bunch of mercuric salts, which wouldn't

 3   make sense for you to deal with here.

 4             DR. COLLINS:  I think those are the ones that tend

 5   to be the ones with the most information.

 6             DR. MARTY:  ARB may have to comment on that,

 7   because the emissions are not speciated that we get hold of.

 8   We just get total mercury emissions.

 9             I would assume that the majority are elemental,

10   particularly from a combustion source, but there may be

11   salts of mercury that end up being emitted into the air

12   under high temperatures or high pressure.

13             DR. BLANC:  I wouldn't think -- I would think it

14   would be very minor and only situation where you would get

15   airborne.  And actually I'm not sure with corrosive

16   sublimate, what the ultimate exposure is, whether it is in

17   fact -- but since you have a boiling point for it, which is

18   actually lower than the boiling point for mercury, that was

19   the other thing that I was surprised at.  I thought mercury

20   boiled a little bit lower than that temperature, elemental

21   mercury.

22             DR. FUCALORO:  Can't help you.

23             DR. BLANC:  Just double check that.

24             Then if you could in the text of this kind of

25   break out or deal with separately what would be the exposure



 1   routes and uses for corrosive sublimate if there are still

 2   any, if that's what you're really dealing with, because it

 3   would be, I think fairly obscure to any reader otherwise.

 4             DR. MARTY:  Okay.

 5             DR. BLANC:  And it would also perhaps affect when

 6   you say inorganic mercury, if you're talking about both

 7   elemental and this one other compound and there's some other

 8   common names for them, it's going to make the whole thing

 9   more complicated, since you're not talking about simply

10   elemental mercury.

11             But I can see why you need to treat it as one

12   thing, but I think it needs to be clarified.

13             And I think that if your RfC, which is derived

14   from the EPA, and you cite that the study is the Piikivi

15   study.

16             DR. MARTY:  Andy, can you turn on that slide.

17             Yeah, I am going to apologize for what is in the

18   text in that it did not really explain well enough all of

19   the studies that ended up being the basis.

20             DR. BLANC:  Right.  They were all at 25

21   micrograms, so it is actually a much -- this is one of the

22   ones where you really have very strong human data that's

23   consistent.

24             DR. MARTY:  Exactly.  Interestingly enough, the

25   rat study came up with same REL.



 1             DR. BLANC:  I think you're really on firm ground.

 2             DR. MARTY:  I have notes to myself to flesh that

 3   out and make it clearer.

 4             DR. BLANC:  And then can you tell me why there's

 5   this whole lengthy appendage at the end about the --

 6             DR. MARTY:  Oral reference dose?

 7             DR. BLANC:  Yeah.  What is -- why does that have

 8   to be in there?

 9             DR. MARTY:  In the air toxics hot spots site

10   specific risk assessments certain compounds are looked at by

11   multiple routes of exposure.  So things that may end up in

12   particulate for metals, semivolatile organics like PAHs and

13   dioxanes and furans, we evaluate not just from inhalation

14   route of exposure, but we also have an exposure paradigm

15   looking at deposition onto food that you might be growing in

16   your backyard onto soil where you may have dermal contact

17   and some ingestion from hand to mouth.

18             DR. BLANC:  Is there anything else in this

19   document --

20             DR. MARTY:  I'd have to look through, but

21   certainly dioxanes, which I believe is in this set, and any

22   of the metals.

23             DR. BLANC:  So nickle has this too?

24             DR. MARTY:  Nickle should have this too.

25             DR. FUCALORO:  If I can get back to the mercury



 1   just for a second, the numbers you point out, it's not

 2   hard -- it's hard to get a good sense of this thing just

 3   from looking at the numbers without doing the calculation.

 4             At a boiling point of 360 degrees centigrade, that

 5   means that the atmospheric pressure, vapor pressure, is one

 6   atmosphere or 760 torr.  Vapor pressure at 25 degrees is two

 7   times ten to the minus three torr, .002.

 8             You can put that in something called a Celsius

 9   clapper on the equation to calculate, estimate the heat of

10   vaporization, and, by so doing, you come up with about 60

11   kilojoules per mole, which is not unreasonable.  I mean, I

12   don't have the numbers with me.

13             Water, for example, is about 44 kilojoules per

14   mole.

15             So maybe it's those numbers are correct, even

16   though it does look -- I mean I have to say, it's a .002

17   millimeters mercury looks a little low, but the calculation

18   seems to indicate that it's ballpark.  That's about -- I can

19   check it.  It's easy enough.

20             DR. BLANC:  Yeah.  Well, for those few places

21   could you have a brief one sentence that says because this

22   is, the following section is included.

23             DR. MARTY:  Okay.

24             DR. BLANC:  And I forgot to say something on

25   cyanide.  Can I go back to that for a just a second?



 1             And it's not unrelated to the mercury, which made

 2   me come back to it.

 3             In cyanide, which you describe as a colorless

 4   liquid/gas, again I think it would be useful to simply have

 5   a sentence or two that says that hydrogen cyanide is a gas,

 6   but that there are multiple cyanide salts, which, if they

 7   come in contact with an acid environment, will release

 8   cyanide gas.  It's the parallel to the whole discussion

 9   about hypochlorite containing products, because that's

10   really the route that most people are exposed or the

11   releases occur, not because somebody has got compressed

12   hydrogen cyanide gas in tanks that's released, but because

13   they've got metal cyanide salts that --

14             DR. MARTY:  Okay.

15             DR. BLANC:  And so the colorless liquid also,

16   cyanide gas is not a liquid in any kind of normal condition,

17   but a lot of these cyanide salt solutions are liquids.  So

18   plating operations or cyanide salts.

19             And I also think it would be useful to emphasize

20   that there are chemicals which can be airborne, which are

21   metabolized to cyanide in the body such as a acrylonitrile

22   and acetonitrile.  And I think that needs to be alluded to

23   in that section.

24             DR. FUCALORO:  Can I say something about --

25   something that was said earlier regarding the conversion



 1   factor.  We saw parts per million into so many milligrams

 2   per meter cube.  I'm assuming that's done the same way each

 3   time, in other words you take the molecular weight and

 4   divide it by 24.4, which is so many liters per mole.  It's

 5   the standard --

 6             DR. COLLINS:  25.

 7             DR. FUCALORO:  298 kelvin.

 8             So why don't you just put that down in the

 9   beginning and people can do the conversion themselves.  Just

10   a suggestion.  I mean, you just take the molecular weight

11   divided by 24.4, and you come out with equivalency of one

12   part million to so many milligrams per meter cubed.

13             DR. BLANC:  I just think they have that in the

14   general introduction, but most people who deal with, you

15   know, OSHA literature and NIOSH literature will be looking

16   for conversion factor, because it's in most standard

17   industrial hygiene related so I think it's useful to have it

18   for each one too.

19             DR. MARTY:  We can also put it in the introduction

20   what the formula is.

21             DR. FUCALORO:  Is it in the introduction?

22             DR. MARTY:  You know, I don't think it is.

23             DR. FUCALORO:  I didn't notice it, but then I

24   didn't look.

25             CHAIRMAN FROINES:  Can we finish, Paul?



 1             DR. BLANC:  Yeah.  That was the one I wanted to

 2   comment on.

 3             CHAIRMAN FROINES:  I have a quick question.  It's

 4   just of interest.

 5             I don't know if there are things like mercury

 6   thermometer plants in the State of California, but are there

 7   hot spot -- this is a relatively low REL, four hundredths of

 8   a part per billion, so it seems like there's a potential for

 9   hot spots.

10             Are there, as far as you know, under -- does the

11   2588 have any indication that there are?

12             DR. MARTY:  What we have available to us is the

13   total statewide emissions from the facilities that report in

14   the hot spots.

15             CHAIRMAN FROINES:  Oh, so you wouldn't find most

16   of them.

17             DR. MARTY:  Right.  If it's evenly divided over

18   thousands of facilities, then you may not have a hot spot.

19             DR. COLLINS:  I think a lot of it's fuel

20   combustion too.

21             DR. FUCALORO:  I bet you there's a hot spot in

22   your laboratory, just as in my laboratory and his

23   laboratory, that there's a lot of mercury vapor there.

24             CHAIRMAN FROINES:  We never spilled any mercury.

25             DR. MARTY:  I can go back and look at our



 1   inventory risk assessments and see if mercury ever came up

 2   as an issue in any of the risk assessments, but I don't

 3   remember that it has.

 4             CHAIRMAN FROINES:  Well, it's the other way around

 5   in a sense, that the trouble with that kind of risk

 6   assessment is there's only certain companies that have to

 7   comply with 2588, but if you did it logically and said who

 8   releases mercury, then you would say thermometer plants,

 9   electroplating plants, et cetera, et cetera, and you could

10   then --

11             DR. GLANTZ:  Toxicology professors.

12             CHAIRMAN FROINES:  College professors who don't

13   put sulfur all over their floors to prevent it.

14             But one could ask the question if you wanted to

15   identify problems, you got a number like this, then there

16   may be some places that you'd want to take a look at, see if

17   there is.

18             Mercury is obviously not inconsequential problem.

19   Okay.

20             I think we're at Peter Kennedy.

21             DR. KENNEDY:  My list begins with ethyl chloride.

22             I have the same generic comment regarding the

23   consistency of enumeration of the description of the

24   conversion factor.

25             I don't deal with this stuff at all, and would



 1   like to see it all the same way, so I can be more

 2   comfortable with it.

 3             It appears that the total sum of human exposure is

 4   some nutty lady who spilled two to three hundred cc's on her

 5   sleeve daily for four months.  Hard to know how to make any

 6   sense of that.

 7             This is, I have a question really informationally

 8   for me as much as anything.  In terms of developmental

 9   toxicity, delayed ossification, particularly in the skull,

10   is apparently a fairly common parameter that's used as an

11   endpoint in these evaluations.

12             It is also seen with developmental abnormalities

13   in ribs and particularly cervical ribs.  At least at an

14   intuitive level that suggests to me not one, but maybe two,

15   mechanisms of action is -- do you all have any information

16   on that or are these just observations that are made?

17             DR. MARTY:  Primarily observations.  In speaking

18   with our developmental and repro tox group at OEHHA, the

19   delayed ossification is rather indicative of general

20   fetotoxicity and general chemical stress.

21             DR. KENNEDY:  Okay.

22             DR. MARTY:  I'm not sure that anybody has really

23   outlined a actual mechanism, but I can certainly talk to

24   them about that.

25             DR. KENNEDY:  I can accept that.



 1             But if you assume that, I have trouble linking

 2   that and cervical ribs, which is really probably a

 3   mutational or developmental malady of the other direction.

 4   You've got bone being hypertrophic, rather than delayed in

 5   its development.  It's a small point, but it keeps coming

 6   up.

 7             DR. MARTY:  The supernumerary ribs?

 8             DR. KENNEDY:  Yeah.

 9             Hey, that's better.

10             DR. MARTY:  Yeah.  We can look into that further

11   and add some information if there's information with regard

12   to that.

13             DR. KENNEDY:  Since that's all you have to work

14   with in this, I mean that's basically what you have.  It

15   might be interesting.

16             The remaining animal studies with inhalation for a

17   hundred odd weeks with extensive histopathologic review

18   shows nothing.

19             I think some interest is this issue of cardiac

20   sensitization when used in anesthetic concentrations.  But

21   there's, I guess, no more data.  It's very very old and

22   there's nothing else that's presented.

23             So you really have only the development effect,

24   developmental effect as your observed endpoint.

25             I'm not sure that the strengths of the inhalation



 1   REL include multiple species, multiple concentrations and

 2   multiple histopathologic analysis, since all those are

 3   negative.  I mean, it's true enough, but it doesn't really

 4   contribute to anything.

 5             DR. MARTY:  We can strike that.

 6             DR. BLANC:  Can I ask a question about this

 7   compound?

 8             Did you double check the Med Line or Tox Line

 9   since you did this, just to double check, since -- I'm not

10   saying that you have to systematically reiterate and never

11   be finished with the document, but for something which has

12   so little information, I would just do a double check of

13   your own.

14             DR. MARTY:  We have been doing that, but Jim says

15   he hasn't done it in a while.  We can do it again.

16             DR. BLANC:  I'm not saying --

17             DR. KENNEDY:  I thought I had missed some lead

18   article in the New England Journal or something when you

19   brought it up, but it's a good point.

20             DR. MARTY:  Pretty surprised there's not more on

21   ethyl chloride myself.

22             CHAIRMAN FROINES:  Could I ask a question that's

23   related to Paul and Peter's question?

24             Under three major uses and/or sources you have

25   ethyl chloride is used as a starting point in the production



 1   of tetraethyl lead.  Is there a lot of tetraethyl lead made

 2   in the United States these days?

 3             DR. BLANC:  Yes, there is.

 4             DR. MARTY:  No.

 5             DR. BLANC:  It's exported.

 6             DR. FUCALORO:  It would be for an export industry.

 7             DR. BLANC:  Ethyl Corporation it's still their

 8   major source of income.  Unless they make it off shore.

 9             CHAIRMAN FROINES:  There is?

10             DR. BLANC:  I don't -- I don't know that for a

11   fact that they make it in the United States.  I do know that

12   Ethyl Corporation it's still their major cash cow, as far as

13   I understand.

14             CHAIRMAN FROINES:  I bet it's in CNA news.

15             DR. BLANC:  The reason why Ethyl Corporation has

16   pushed organic manganese is because they have a patent on

17   that.  For their future they think that's going to be --

18             CHAIRMAN FROINES:  We're coming to that.

19             DR. BLANC:  I think that the phrase it is also

20   used as a topical anesthetic.  Based on Patty's Industrial

21   Hygiene from 1994, perhaps it would be more conservative to

22   say it has been used.  I doubt that it's being used very

23   much.

24             DR. KENNEDY:  It's still a major component of

25   walk-in clinics.



 1             DR. BLANC:  Is it?

 2             DR. KENNEDY:  Yeah.

 3             DR. BLANC:  So it's on the pharmacopeia, because

 4   it evaporates so quickly?

 5             DR. KENNEDY:  Yeah.

 6             DR. BLANC:  I stand corrected.

 7             DR. GLANTZ:  You have to pop a stitch in a little

 8   kid, it's right there.  Less traumatic than infiltrating.

 9             DR. BLANC:  There's got to be data out there, I'm

10   sorry, on exposure, it seems to me.  There's got to be.

11   Somebody has got to.

12             And when you say it's an alkylating agent.

13             DR. KENNEDY:  I had an question about that too.

14   At least not in the sense of how I think of an alkylating

15   agent.

16             DR. MARTY:  In chemical production it's used as

17   the source --

18             DR. KENNEDY:  It might be worth stating that

19   specifically.

20             DR. MARTY:  Not as like nitrogen mustard, which is

21   what you guys are thinking about.

22             DR. FUCALORO:  Craig, you must have a lot of

23   organic chemistry.  Is it in fact a alkylating agent?

24             DR. BYUS:  I think it is.  I think you're right.

25   That's exactly my interpretation of it, in inorganic



 1   chemistry it's alkylating, but it's not a DNA alkylating

 2   agent.

 3             DR. KENNEDY:  You might --

 4             CHAIRMAN FROINES:  Wait a second.  Wait a second.

 5   I am an organic chemist.  I have a Ph.D in that field.

 6             And so it is an alkylating agent.  In fact when

 7   you take undergraduate inorganic chemistry you know that

 8   ethyl chloride will react with sulfate to form ethyl

 9   sulfate.  And you study it, it undergoes SN 2 reaction, not

10   SN 1 reaction, so you don't get pure carbonium, but still it

11   will obviously interact with nuclear files.

12             DR. KENNEDY:  I'm sure it would.

13             CHAIRMAN FROINES:  The related question is has

14   anybody ever looked at it as a carcinogen, since it is an

15   alkylating agent?

16             DR. MARTY:  There is an NTP study, apparently.

17             DR. BLANC:  Well, then I would like --

18             DR. MARTY:  It's an older one.

19             DR. BLANC:  Appropre of the discussion on whatever

20   that other chemical is on benzene where I said look for the

21   noncarcinogenic hematologic effects, then I would pull --

22   seemed like this is one where you just relied too much on

23   the EPA and maybe they were not as aggressive.

24             CHAIRMAN FROINES:  Must have been tested for

25   carcinogenicity at some point by somebody.



 1             DR. COLLINS:  The NTP --

 2             CHAIRMAN FROINES:  They have.  Well, then,

 3   that's -- I would have had that referenced in here.

 4             DR. KENNEDY:  The low little compound gets more

 5   attention.

 6             DR. MARTY:  It doesn't have a lot of play.  We

 7   talked about the study.  It does say --

 8             CHAIRMAN FROINES:  I got it.

 9             DR. MARTY:  Yeah.  It's at necropsy a complete

10   histopath examination failed to identify evidence of

11   toxicity up to 15,000 ppm.

12             CHAIRMAN FROINES:  All that proves is that it's

13   not a very good alkylator.

14             DR. BLANC:  Sorry, Peter.

15             DR. KENNEDY:  No problem.

16             I guess my one last comment with regard to that is

17   that ethyl chloride you comment that it produced

18   vacuolization in hepatocytes.  There are a whole lot of

19   things that do that.  I'm not sure of what significance that

20   is.

21             The next compound is ethylene glycol monoethyl

22   ether.

23             Limited human study.  And sounds as if it was

24   contaminated significantly, since sperm counts were lower

25   both in exposed and control subjects.  May be important to



 1   find out where these people worked.  We may not want to go

 2   there.

 3             Animal studies, I guess my thoughts on this relate

 4   to the comments that Dr. Froines had made previously.

 5             Obviously, the major point of reference that you

 6   use in developing the REL is the study by Barbee in rabbits

 7   and rats.

 8             Again, as suggested by the human study, the main

 9   issue that you were concerned with or that was identified

10   was degenerative changes in seminiferous tubules, and

11   reduced testes weight.

12             The hematologic effects that this keeps coming

13   around again and again and bothers me that we -- the

14   hemoglobin, hematocrit and red count are all the same thing,

15   just different measures of expression, and we never really

16   find out whether this is a suppressive effect or a

17   destructive phenomenon.

18             And this is the study that you used to develop the

19   LOAEL.  Some of your other studies, however, suggest that

20   you, even in the absence of a poor dose response curve,

21   which you don't really see even in the Barbee study, that we

22   see there are CNS effects seen in the offspring at 100 ppm

23   per meter squared, bone alterations at 175.

24             I guess the hundred is close to the 103 that you

25   recorded.



 1             But it may be valuable to -- again it brings up

 2   the issue of do you take a single study to develop our

 3   NOAEL, or can you collect the information from multiple

 4   studies, which are certainly suggested by the teratologic

 5   effects.

 6             I'm not sure how you respond to that, but it seems

 7   that this is a compound where that's of substantial

 8   importance.

 9             Again, your final statement, your final paragraph

10   is repetition of the last compound.  Availability of

11   subchronic inhalation, well-conducted histopathologic study

12   and observation of a NOAEL.  I would suggest to you that the

13   strongest study at least in the information that we have

14   here is really the neurodevelopmental studies from the

15   pregnant animals.

16             DR. MARTY:  Nelson et al, '81.  I think that's a

17   good point.  We should revisit that.  Your point is that

18   Nelson et al, if I'm reading this correctly, has effects at

19   100 ppm.

20             DR. KENNEDY:  Right.

21             DR. MARTY:  And that we adopted the EPA RfC, which

22   used 100 ppm basically as a no observed adverse effect level

23   for testicular toxicity.

24             CHAIRMAN FROINES:  I'm sorry.  I'm missing

25   something.  Where are you at?



 1             DR. KENNEDY:  Where am I at?

 2             CHAIRMAN FROINES:  What paragraph, what page?

 3             DR. KENNEDY:  In the, I think this is the May '99

 4   version.  Yeah.

 5             DR. COLLINS:  The effects of animal exposure

 6   paragraph, behavioral teratogenic effects were examined in

 7   pregnant --

 8             DR. KENNEDY:  Two basic --

 9             CHAIRMAN FROINES:  Why are we in the May '99

10   version?

11             DR. KENNEDY:  The first study done relates to

12   degenerative changes in the testes with a NOAEL at 103.

13             The second study suggests that there are

14   neurodevelopmental changes in offspring of pregnant animals

15   at that dose level.

16             That starts on the bottom of page 70, Nelson's

17   work.

18             DR. MARTY:  Okay.  I think we should go back and

19   revisit that.

20             DR. KENNEDY:  Okay.

21             CHAIRMAN FROINES:  I was on the monomethyl of

22   rats, so that's why I was off.

23             But you're going back to this monoethyl, Nelson's

24   work is very good.  He's with NIOSH and his work is quite

25   solid.  So I would look at it pretty carefully if I were



 1   you.

 2             DR. KENNEDY:  That's all I've got for that.

 3             DR. MARTY:  George just pointed out to me that

 4   when using repro studies we don't have a subchronic

 5   uncertainty factor because the effect is reproductive, so if

 6   we --

 7             DR. KENNEDY:  We're back to that again.  We went

 8   through that in February.

 9             DR. MARTY:  If we used Nelson et al we'd get about

10   the same number, because we wouldn't have this subchronic

11   uncertainty factor of 10 but we would have a low observed

12   adverse effect level uncertainty factor of 10, so the

13   uncertainty -- you exchange the type of uncertainty, but you

14   end up with the same number, but I think we need to put that

15   in there.

16             DR. KENNEDY:  At least express that --

17             CHAIRMAN FROINES:  It's also the issue of trying

18   to put a round peg in a square hole, which is, you know, we

19   have these categories, subchronic, chronic, carcinogenesis.

20   We always seem to have problems of trying to figure out

21   whether reproductive problem is a chronic or acute.  Why

22   don't we have a category called reproductive and make a

23   document that has that and has all the chemicals which you

24   can put numbers in based on those endpoints, and not sort of

25   always trying to justify whether it's chronic and acute.



 1             And I know Paul is rolling his eyes because he

 2   knows it's a lot of work but --

 3             DR. BLANC:  Drop it.  Don't even go there.

 4             CHAIRMAN FROINES:  It's worth considering.  Let's

 5   leave it at that.

 6             DR. BLANC:  At some future date.

 7             CHAIRMAN FROINES:  At some future date.

 8             DR. KENNEDY:  Monoethyl ether acetate, same issue

 9   of reproductive effects.  I guess I don't know how you'll

10   deal with that.

11             But you do in this case obviously develop your

12   references using that as the endpoint, although this is

13   direct fetotoxicity.  All those poor little animals die.

14             Does anybody ever do any real analysis on any of

15   these hematologic effects except to make basic observations?

16   You've got decreased red counts and now you've got little

17   red cells as well suggesting a possibility of a hemolytic

18   response, which would be destructive, rather than just

19   marrow suppressive effect.  If these small cells are

20   spherocytes --

21             DR. BLANC:  You know in that regard, which is the

22   one of this family of glycol ether or glycol ether acetates

23   that was associated with bone marrow effects in humans and

24   this series that Cullen did of printers where they actually

25   did bone marrow aspirates.  I looked quickly through this



 1   group to see where he was cited, because I don't remember

 2   which --

 3             DR. COLLINS:  Spell his name.

 4             DR. BLANC:  C-u-l-l-e-n.

 5             DR. COLLINS:  C-u-l-e-n?

 6             DR. BLANC:  C-u-l-l-e-n.  Mark.  I think he was

 7   first author.

 8             CHAIRMAN FROINES:  I think it's EGME --

 9             DR. COLLINS:  It will show up in Med Line.

10             CHAIRMAN FROINES:  I think it's one of those two.

11             DR. BLANC:  I think, Peter, the thing is that

12   there is in this group it's one of these is the only one, is

13   the only chemical really, industrial chemical, outside of

14   benzene, for which there is really quite convincing human

15   data in the workplace of bone marrow suppression with bone

16   marrow aspirates, quite a good clinical case series.

17             DR. MARTY:  We should look at those studies and

18   see what --

19             DR. KENNEDY:  Run that down.

20             DR. MARTY:  How that relates to the dose levels

21   we're calling NOAELs.

22             CHAIRMAN FROINES:  Paul is right, because we were

23   doing a study on glycol ethers in cleaning solvents in

24   janitors at one point, and we found, for example, the most

25   widely used was the butyl compound, and it was contrary to



 1   the rat data.  There is a species difference.  And I think

 2   it's relatively specific that there's -- that there may be

 3   only one or two for which there was actually hematologic

 4   data, although there's a much more rich database with the

 5   animals.

 6             DR. MARTY:  The butyl is, if I recall, is the one

 7   that causes hemolysis in laboratory rodents, EDBE.  But that

 8   effect doesn't appear to happen in people, although the data

 9   isn't that strong to say that it absolutely doesn't happen.

10   It just may be not a very strong effect.

11             EGME has the bone marrow effect, so I guess we

12   cited an EG monoethyl ether in that tox summary.

13             DR. BLANC:  I'm sorry.  It's the monomethyl ether?

14             DR. COLLINS:  Page A-81, Larese, et al, appearing

15   in the British Journal of Industrial Medicine.

16             CHAIRMAN FROINES:  Let's go ahead.

17             Let's Peter finish and then we'll go on.

18             DR. KENNEDY:  I think I would only say in

19   finishing up EGEEA that I think your comment at the end

20   indicating the difference, that the difficulties in using

21   reproductive effects as an endpoint is clear and reasonable.

22             And monoethyl ether, don't have a whole lot to

23   say.  Interesting human observation of transient, presumed

24   granulopenia during exposure, which is reversable.  They

25   were also exposed to acetone, for whatever reason.  It



 1   popped up in ancient literature as well.

 2             No major comments otherwise.

 3             CHAIRMAN FROINES:  You know, this compound, these,

 4   the EGME, EGEE and their respective acetates, they were very

 5   important, because when people shifted, for example, from

 6   acetone in 1973 when they had the gasoline shortage and

 7   people started to use a lot of these compounds in place of

 8   acetone, there was a heck of a lot of case studies on

 9   neurobehavioral and neurologic effects, and I'm surprised

10   that either you didn't quote it or there isn't any, but I'm

11   surprised there is so little data on neurotoxicity on these

12   compounds.

13             I know people have been looking at the

14   reproductive effects for years, because that's the endpoint

15   of most concern, because the effects are closest to the

16   existing occupational standard.

17             But that it does seem to me that, you know, these

18   older data that Peter just cited on neurologic effects, I

19   think there must be some animal data that -- don't go back

20   and change it, but I mean it's just worth being aware of.

21             The other comment I was going to make is that for

22   these ethyl glycol, monomethyl and monoethyl ether is,

23   remember, that where this whole controversy about these two

24   compounds got started was when they were used as a solvent

25   in photoresistance and semiconductor manufacturing.  That's



 1   where the interest all emerged about glycol ethers.  It was

 2   in the spontaneous abortions associated and at teratogenesis

 3   associated with photoresistance in the semiconductor

 4   industry.

 5             Now I think it's pretty much all gone from that

 6   industry.  I think they've gotten rid of it essentially

 7   entirely, but you might want to have at least a reference to

 8   it in your uses, since it is at least a historical use.

 9             DR. KENNEDY:  You certainly do for NMEA, sort of

10   scary defects in a woman not only exposed acutely who had a

11   child with terrible deformities, but then three years later

12   she had another one.  Presumably she didn't keep -- she

13   broke off her exposure.

14             I think NMEA basically looking at the same issues

15   again.  Would be interested personally in some details about

16   the changes in blood counts, what the mechanism is here,

17   whether it's a suppressive or not.  It seems to be specific

18   to at least one cell line.

19             DR. MARTY:  We can look at the studies on

20   hemotologic and bone marrow effects and flesh that

21   information out in this section.

22             DR. KENNEDY:  It looks to be important as in the

23   group as a whole.  It might be worth doing that.

24             There's one of these that where you go and it goes

25   stir crazy over the so-called immunosuppressive effects and



 1   the methodology is pretty crude.  Depression in splenocytes,

 2   give me a break.  Is there anything that's more contemporary

 3   than that?

 4             This is in a monomethyl ether.

 5             I've lost it.

 6             The immune compromising effect of these compounds

 7   is referred to a couple of times, but the methodology is

 8   pretty archaic.  I'm not sure what significance it has.

 9             DR. MARTY:  The decreased --

10             DR. KENNEDY:  How relevant.

11             DR. MARTY:  Decreased spleen cell numbers.  That's

12   the Exxon et al, '91.

13             DR. KENNEDY:  It would be of value, I think, to

14   see whether that's -- unless there's a -- unless it exists

15   as a model that's well accepted in the field, it borders on

16   being not real relevant.

17             And then finally for propylene glycol, which looks

18   to save the day in terms of being the gentler, kinder

19   compound, I think you've done what you can.

20             There's not really chronic exposure information,

21   but apparently you've gotten some comments from the Chemical

22   Manufacturers' Association where they're doing a study and

23   will keep you appraised of that.  So I think that would be

24   important when it comes down the pike.

25             DR. MARTY:  Yes.  We did get a new chronic study



 1   to look at for PGME, and as far as we can tell it's a

 2   well-conducted study, and so we are going to recalculate the

 3   REL.  It indicates a chronic REL of 7,000 micrograms per

 4   cubic meter, rather than 2,000.  So the EPA did not have the

 5   benefit of that study when they developed their EPA RfC,

 6   which is based on a subchronic study.

 7             So that change is coming down the pike.  We got

 8   that comment in the latest public comment period, and

 9   received the report in July.

10             CHAIRMAN FROINES:  All right.

11             DR. KENNEDY:  That's it.

12             CHAIRMAN FROINES:  I think we can finish before we

13   break for lunch.

14             I'm sorry, Gene, if we're going to hold you over

15   after lunch.

16             So I can be very quick, I think.

17             Melanie, if you're okay.

18             DR. MARTY:  Okay.

19             CHAIRMAN FROINES:  Hexane, I won't say anything

20   more about.  I've made my comments about that.

21             I'll come back to manganese.

22             We're going to deal with MTBE this afternoon.

23             I only wanted to make one comment about MTBE,

24   wherever the devil I put it, but that I only had one comment

25   about the document on MTBE, and that is that if I remember



 1   correctly the problem I noticed that the NOAEL is 403 ppm,

 2   and your LOAEL is 3,023 ppm.  Seems to me that one could use

 3   this as a textbook case for students as to why the safety

 4   factor approach isn't worth the paper it's printed on

 5   sometimes.

 6             Sometimes, I don't want the record to reflect some

 7   rejection of the safety factor approach, but I am -- I do

 8   want to say that when you have a study where somebody looked

 9   at 403 parts per million and they jump up to the next dose

10   level of 3,000 and then we call the 403 as the no effect

11   level, that's strange credibility, I think.  And it is

12   reflective of the kind of problems that we get into when we

13   have to force ourselves to have NOAELs that are based on

14   experimental dose levels that we've all read about over and

15   over again and the criticism of this kind of approach, that

16   we end up with these values that are ultimately defined by

17   the dose levels that people picked in their studies.

18             So that's the only comment I make.

19             Whether or not that's a correct, 403 is in fact a

20   NOEL, is a reasonable question, I think.  Any rate.

21             Naphthalene we were going to -- let me say one

22   thing about methylene chloride.

23             Stan, what's the blood hemoglobin levels of people

24   who are exposed to environmental tobacco smoke?

25             DR. GLANTZ:  Blood?



 1             CHAIRMAN FROINES:  Carboxy hemoglobin.

 2             DR. GLANTZ:  Oh, boy, I don't remember those

 3   numbers.

 4             DR. BLANC:  Low.

 5             DR. GLANTZ:  They're low.

 6             DR. BLANC:  An active smoker gets up to, unless

 7   they're really really heavy, only gets up to about four or

 8   five percent.

 9             DR. GLANTZ:  Yeah.

10             DR. BLANC:  So, I would bet it would be --

11             DR. GLANTZ:  Well under a percent, I think.

12   Seven-tenths of a percent is what I remember, but I'd have

13   to look those up.

14             CHAIRMAN FROINES:  Just a second.  So where people

15   who are exposed to 50 parts per million of methylene

16   chloride, and they had peak blood carboxy hemoglobin levels

17   of 1.9 percent, do we assume that that is --

18             DR. GLANTZ:  That could be enough to be

19   significant in some people, because what happens is you run

20   very close to a hundred percent or to high levels of oxygen

21   saturation and if you get -- in some people if you

22   compromise that a little bit the curve is not linear.

23             CHAIRMAN FROINES:  So we're running here about,

24   what?  What's my blood carboxy hemoglobin right now?

25             DR. BLANC:  Less than one percent.



 1             CHAIRMAN FROINES:  Less than one percent.

 2             DR. MARTY:  We also note in the document that the

 3   ambient air quality standard for carbon monoxide is based on

 4   a carboxy hemoglobin -- or avoiding a carboxy hemoglobin

 5   level of two percent.

 6             So there was lots of discussion at some point

 7   about what a adverse carboxy hemoglobin level would be.

 8             CHAIRMAN FROINES:  I had one question and it's

 9   just a very minor matter.

10             You wrote down a LOEL is 33 parts per million, but

11   when I look at your document it says that the 19 workers

12   were exposed to 40 parts per million, so I don't understand

13   where the 40 gets to 33.

14             It's not a big deal.  Let's go on.

15             DR. MARTY:  It's actually underneath, the

16   paragraph underneath that.  Workers were exposed to average

17   measured concentrations of 40 ppm during the workday and the

18   personal monitors on three of the subjects indicated an

19   eight-hour time-weighted average of 33 ppm over a two-week

20   period.

21             CHAIRMAN FROINES:  And you have their specific

22   carboxy hemoglobin levels?

23             DR. MARTY:  It's not clear from looking at this

24   whether that average carboxy hemoglobin level of 3.9 percent

25   at the end of the work shift refers -- I think that refers



 1   to all of the 19 workers, not just the three that were used

 2   to calculate the time-weighted average.

 3             CHAIRMAN FROINES:  That's the point.  If it's 40,

 4   then use 40.  If you are saying 3.9 relates to 40, and then

 5   you use 40.  If it means three people related to 33, then

 6   use 33.  But you shouldn't -- you should be internally

 7   consistent.  And it's not a big enough issue to spend any

 8   time on.  It's almost the kind of thing you could talk about

 9   outside of the --

10             Melanie, I made a number of notes on manganese and

11   now that we're doing manganese, I've lost track of what my

12   notes mean.  I may talk with you about it separately.

13             I think manganese is an important compound if it's

14   going to be used as a gasoline additive at some point, and

15   that's an organic manganese compound, I would assume.

16             So do we know that that's happening?  Is manganese

17   currently now being used --

18             DR. COLLINS:  Not in California.

19             CHAIRMAN FROINES:  -- as a gasoline additive or is

20   it just proposed to be used?

21             DR. MARTY:  I asked the Air Resources Board, who

22   called back home, and they said that it's illegal to use

23   manganese in California as a gasoline additive.

24             And that the US EPA has rejected the use of

25   manganese as a gasoline additive.



 1             I think there are some legal issues remaining with

 2   the US EPA's decision.

 3             It is used in Canada.

 4             CHAIRMAN FROINES:  It is used?

 5             DR. BLANC:  I think it's used by the military, and

 6   I'm not sure whether or not they're prohibited under

 7   California law.  It's used in certain, I think, military

 8   uses.

 9             CHAIRMAN FROINES:  I had a question about the LOEL

10   based on the Roels study.  And if I understand it, correct

11   me if I'm wrong, that we're looking at what is essentially a

12   systemic effect.  That is neurobehavioral changes.

13             It looks to me as though in your LOEL it's .15

14   milligrams of respirable manganese dust per meter cubed.

15             That's, to me, means that that's an

16   underestimation of the actual dose, because the workers were

17   breathing manganese that included both respirable and

18   nonrespirable, and so they were getting some manganese being

19   absorbed into their systemic circulation through their GI

20   tract, as well as through the lung.

21             Therefore, the amount reaching the brain is a

22   combination of what goes in through the lung and what goes

23   into the GI tract, and so this is not an accurate

24   representation of their dose, of their internal dose.

25             So I don't think you can use -- either you have to



 1   adjust for that or you need to do something, but looking at

 2   a respirable fraction -- now, if these people were only -- I

 3   didn't look at the paper, so if it is a situation where they

 4   were only exposed to fume, then one might --

 5             DR. MARTY:  It's dust.  It's manganese oxide dust.

 6             CHAIRMAN FROINES:  If it's manganese oxide dust,

 7   then most of what they breathe was not respirable, so that

 8   the estimate of using respirable dust is an absolute and

 9   it's especially since the mass changes by the cube root,

10   that the amount that they got into their GI tract was

11   obviously the vast majority of the material.

12             I don't know what the actual efficiency is

13   absorption to the GI tract is, but it's not zero.

14             So that you can't use this study or at least you

15   can't use this dose level for this determination.  It's

16   incorrect.

17             DR. BLANC:  It's actually more complicated and

18   interesting than that.

19             And there's, Melanie, there's a recent issue of

20   one of the toxicology journals devoted to this, which was --

21   or the proceedings of a symposium that just came out, but

22   the symposium was two years ago, and you really need to take

23   a look at that.  I have the reference back at the office, if

24   you can't find it.

25             But there's a lot of stuff happening with



 1   manganese and you may have better animal and human data.

 2             It turns out that there's animal data suggesting

 3   that part of the toxic mechanism in terms of route of

 4   delivery for manganese may be uptake of manganese by

 5   olfactory nerves and direct translocation to the CNS and

 6   that that's why manganese has some of the targeted effects

 7   it has in the, you know --

 8             CHAIRMAN FROINES:  Brain.

 9             DR. BLANC:  In that area of the brain.  I think

10   it's in the putamen, but I'm not absolutely sure.

11             So it may not only matter what the nonrespirable

12   part is, but that may be more important.

13             But it actually is very poorly absorbed and very

14   tightly regulated by the GI tract.  So GI absorption is

15   probably not particularly an issue with manganese.

16             CHAIRMAN FROINES:  I don't agree.  Because you

17   should read the papers by Froines on this subject.

18             You find, for example, that lead is deposited, is

19   absorbed with an efficiency of about 40 percent in the lung

20   and it's absorbed with about the efficiency of five percent

21   in the GI tract.  But when you calculate the dose to the GI

22   tract based on the amount, the percentage of the size

23   distribution of the particulate, then the amount that is

24   going to the GI tract is way over --

25             DR. BLANC:  I believe that's true for lead.  I



 1   just don't think it happens to be true for manganese.

 2             CHAIRMAN FROINES:  I don't think it's going to be

 3   zero.

 4             DR. BLANC:  It's really really low, because there

 5   are a lot of dietary additives with manganese, in fact.

 6   It's a really important issue, but I think all the data

 7   suggests that it just so happens that for manganese that's

 8   not particularly an issue.

 9             CHAIRMAN FROINES:  Okay.  Well, I think that the

10   key question is, Melanie, is the size distribution of the

11   particulate, the potential for olfactory nerve uptake and GI

12   tract absorption efficiency.

13             And we need GI tract absorption efficiency in

14   fasting and non-fasting.  We need -- it's obviously

15   complicated.

16             Lead goes up to about 50 percent in fasting.  So

17   it varies depending on what's going on.

18             DR. MARTY:  Okay.  We can look at into that.

19             We can also look at the Roels paper.  They have

20   some discussion of manganese measured in blood, in urine and

21   we may be able to relate that --

22             CHAIRMAN FROINES:  But the point is --

23             DR. MARTY:  -- back to using that, rather than the

24   concentrations somehow relate that back to a total dose.

25             CHAIRMAN FROINES:  Depending on the issue of the



 1   efficiency of uptake, I would bet that the vast majority of

 2   the exposure is to nonrespirable manganese.  What then

 3   happens, we can argue about it.

 4             DR. BLANC:  I would say that for consistency you

 5   should title this section manganese, comma, inorganic,

 6   because you really are not dealing with organic manganese

 7   compounds here.  Not just the gasoline additive, but, you

 8   know, man ebb as well, which is going to be coming up to us

 9   separately from our friends at Food and Ag.

10             CHAIRMAN FROINES:  And can Paul send us that

11   reference.  I think that would be interesting to look at.

12             DR. MARTY:  Yeah.

13             CHAIRMAN FROINES:  Do you have it?  Do you know

14   that?

15             DR. BLANC:  I don't know it by heart, but I was at

16   the conference and it took them a year and a half, but they

17   did finally send it.

18             DR. COLLINS:  It's a whole series of articles.

19             DR. BLANC:  Yeah.

20             CHAIRMAN FROINES:  Good.

21             DR. BLANC:  It's driven in part by the concern

22   over gasoline additive, obviously.

23             CHAIRMAN FROINES:  This issue of nonrespirable and

24   respirable is very important.  There are lots of people who

25   go out and use cyclones to sample for various metals,



 1   thinking that you want to look at the respirable fraction

 2   when in fact there's -- it's as though people believe this

 3   stuff is silica or something and it's pulmonary effects

 4   they're looking at and in fact most of the time it's not,

 5   especially where it's systemic effects.

 6             Anyway, let's take a break for lunch and come

 7   back.

 8             DR. BLANC:  What's left afterwards?

 9             CHAIRMAN FROINES:  We're going to do MTBE.

10             DR. BLANC:  And that's it?

11             CHAIRMAN FROINES:  That's it.

12             DR. FUCALORO:  When do you anticipate us breaking?

13             CHAIRMAN FROINES:  So if we get back here in say

14   45 minutes, and then we'll break about, I would guess 2:00.

15             Let me say this, MTBE is an extremely important

16   compound, so that we shouldn't short-circuit it.  We

17   shouldn't try and not have a lunch and short-circuit that

18   discussion.

19             (Thereupon the lunch recess was taken.)









 1                A F T E R N O O N    S E S S I O N

 2             CHAIRMAN FROINES:  Melanie, why don't you go

 3   ahead.

 4             DR. MARTY:  George.

 5             DR. ALEXEEFF:  George Alexeeff with OEHHA.

 6             I thought I'd just give a little bit of a

 7   introduction to this document, because it's a little bit

 8   different from some of the other documents.

 9             First of all, MTBE is already a toxic air

10   contaminant, so we're not necessarily identifying it as a

11   toxic air contaminant.

12             In this case the question was if we considered it

13   carcinogenic or do we consider it carcinogenic and should we

14   have a potency for it, something more along that manner.

15             And another thing that's unusual about this

16   document is most of the information has been based upon a

17   previous OEHHA document that we developed for our water

18   program, it's called a Public Health Goal Program.  And that

19   particular document was developed last year.

20             It went out for public comment.  We had two peer

21   reviewers, quote, review it, two members of the faculty in

22   the UC system, and then we also had a public workshop.  We

23   had public comments.  We revised it in response to public

24   comments.  So there's a lot of information with regards to

25   development of the document.



 1             And that process all went through last year and

 2   then the document was finalized earlier this year.

 3             So what we thought what we would do is since that

 4   document was fairly recent, up-to-date in the literature, it

 5   made sense to not write a whole new MTBE document, but just

 6   to base as much of that as possible on the PHG document.

 7             And finally we're also, as part of the Governor's

 8   Executive Order to phase out MTBE, we're required to write

 9   our report, which we're doing right now, and Melanie and

10   Andy are writing that report on replacement fuels for MTBE

11   and what are the risks of replacement fuels of MTBE, or

12   alternative fuels from the current fuel that's in use.

13             And in that process of comparing alternative fuel

14   risks, we would be using the MTBE potency number to help

15   evaluate what are the risks of using one fuel versus another

16   fuel.

17             I wanted to give you that background as to why --

18   I know that you've probably heard it's being phased out so

19   why are we concerned about it or why are we using it.

20             And there's two reasons.  One is we have a recent

21   literature review and the other is we still need it to

22   evaluate some relative risk alternatives.

23             And Dr. Andy Salmon will be sort of presenting the

24   document and then we have Dr. Martha Sandy, who prepared the

25   document or the characterization portion of the cancer



 1   portion for our water program, and she'll be providing some

 2   information too.

 3             CHAIRMAN FROINES:  Wait a second.

 4             George, we have three people who have to leave at

 5   2:30.  And I think actually it's four.  I think

 6   Dr. Friedman, Dr. Fucaloro, Dr. Kennedy and Dr. Byus, so

 7   that will leave Paul and me, which doesn't constitute a

 8   quorum.  So we can only go to 2:30.  Point one.

 9             Point two is, just to reiterate, this compound has

10   already been designated a toxic air contaminant.  There is

11   no Part A document on exposure.  There is no Part C

12   document, although the formal Part C document, although

13   you've had comments.

14             So in essence the question is what are you asking

15   the panel to do besides take notice of it.

16             And so I'm assuming for the sake of argument that

17   the risk assessment that you are going to present this

18   afternoon is in the context of in a sense the risk

19   assessments we've been doing today where the panel isn't

20   identifying the TAC and then sending findings or rather a

21   submittal of findings, but that we are rather simply

22   reviewing what you say and then basically approving what you

23   say.

24             DR. ALEXEEFF:  That will be fine, since we're not

25   identifying a toxic contaminant.  That's fine to review it



 1   in that context.

 2             DR. GLANTZ:  But does that mean, George, that then

 3   we will at some point vote to recommend or to make a

 4   conclusion that we agree with the document?

 5             DR. ALEXEEFF:  Right.

 6             DR. GLANTZ:  In particular with the unit risk

 7   number that you come up with, that comes up at the end of

 8   this process.

 9             DR. ALEXEEFF:  Yes.  As we did with the previous,

10   with the acute documents, let's say, or the cancer potency

11   documents in the past.

12             DR. GLANTZ:  Okay.

13             CHAIRMAN FROINES:  So we're not doing a full --

14   this is not a formal TAC procedure then in that context, as

15   I understand it.

16             DR. ALEXEEFF:  It's not following all the formal

17   aspects of the TAC program where findings are then presented

18   to the board and there's a presentation to the Air Board.

19   That is not -- that portion is not happening.

20             CHAIRMAN FROINES:  Okay.  Is everybody happy with

21   that?  Okay.

22             Andy, go ahead.

23             DR. SALMON:  I've got the structure and a few

24   details of the compound on the first slide for your

25   information.



 1             What I'm going to do in fact is I'm just going to

 2   very briefly describe the studies which are the basis of the

 3   number which we calculated, and describe to you how we

 4   reached that number.

 5             If I can have the next slide, please.

 6             The effects on which we based the number,

 7   obviously, is the suspected carcinogenicity of MTBE.  There

 8   are no data on human cancer risk.  However, there are

 9   several studies that are available in animals.

10             One study in rats by gavage and a study in rats

11   and study in mice by inhalation.

12             If I could have the next slide, please, Jim.

13             The first study I want to describe is the gavage

14   study conducted by Belpoggi and colleagues.  This is from

15   Dr. Maltoni's laboratory.

16             The original report appeared in 1995.  In response

17   to some comments about the original report, in fact they

18   undertook a pathology review which was published in 1998.

19   The results of that review are also presented in the slide.

20             We have used the revised data from the pathology

21   review in our numerical analysis.

22             The principal findings are, firstly, in female

23   rats an increase in the incidence of lymphomas and leukemias

24   and in male rats an increase in the incidence of

25   interstitial cell tumors in the testes.



 1             If I can have the next slide, please.

 2             DR. BYUS:  Excuse me.  Do you want us to ask

 3   questions sort of now or what do you want us to do?

 4             DR. SALMON:  I'm happy to answer them now.

 5             DR. BYUS:  Since you have that up there, put the

 6   previous one back.

 7             DR. SALMON:  Sure.

 8             DR. BYUS:  As I read the report, I have no

 9   problems with grouping the lymphomas and leukemias together.

10   I realize that was a concern.

11             DR. SALMON:  There's been considerable amounts of

12   debate about that and one of the purposes of the pathology

13   review was to in fact clarify that and get a second opinion

14   on that.

15             DR. BYUS:  I think one of the documents, I think

16   from peer reviews, did a very nice job of reviewing that and

17   putting it in the right context.

18             My concern with this study has to do with this --

19   well, there's two concerns.

20             Number one, in regards to lymphomas and leukemias,

21   why did only the female mice show up with increased

22   incidence?  I see no explanation for it.  It's never

23   commented on.  Some of these, a number of these studies are

24   sex specific, other than the Leydig cells, which is sex

25   specific.



 1             DR. SALMON:  Easy to understand that, certainly.

 2             DR. BYUS:  I just find it somewhat unusual.

 3             And also the fact that it wasn't commented on.

 4             DR. SALMON:  I think it's one of the -- as far as

 5   I am prepared to say on this occasion, it's an observation.

 6   It's an observation which is not, by any means,

 7   unprecedented.

 8             DR. BYUS:  Okay.  That was one question.

 9             My next question was about doing the tumor

10   incidences on the surviving rats.

11             DR. SALMON:  Yeah.  In fact, because of the

12   duration of the starting, there's a complexity which is

13   explained in the PHG document, which gives the full

14   analysis.  We actually didn't do the survivor analysis for

15   the potency calculation.  We present it here, because it's

16   important for the significance calculation to show there was

17   real effects.

18             But the actual potency calculation was not using

19   this survival adjustment.

20             One of the things which we tried to do on the

21   inhalation study was to use a difference survival

22   adjustment.

23             DR. BYUS:  It's very confusing.

24             DR. SALMON:  Yes.

25             DR. BYUS:  It's extremely confusing.



 1             DR. SALMON:  I apologize for the complexity of the

 2   issues.

 3             DR. BYUS:  And I'm not -- let me ask another

 4   question.  My other question would be why are the rats in

 5   this study, there's a significant number of rats dying

 6   before the end of the experiment even in the control groups,

 7   I believe.

 8             DR. SALMON:  This experiment was in fact run for a

 9   rather longer period than the standard 104 weeks.  The study

10   design they used actually continued, although dosing

11   terminated 104 weeks, their actual observations --

12             DR. BYUS:  Was a lifetime study.

13             DR. SALMON:  Was literally, yes.  I think it went

14   out to about 132 weeks for some groups.  So some of that

15   reflects the difference in study design.

16             DR. BYUS:  I got the impression that wasn't the

17   main reason.  I got the impression from reading that there

18   was just a lot of animals dying.  And why they died, that

19   was my next question, is why they were dying.  Were they

20   dying due to the tumors or were they dying from the

21   leukemias or were they dying due to something else?  You

22   know what I'm saying?

23             DR. SALMON:  I think I must admit I don't have the

24   full mortality analysis at hand, but certainly the --

25             DR. SANDY:  If I can answer that.



 1             For the females in the study, there was a dose

 2   related decrease in survival.

 3             DR. BYUS:  There's a dose decrease in survival?

 4             DR. SANDY:  Right.  For the females.  They did not

 5   see that effect for the males.

 6             DR. BYUS:  Another one that is very unusual in a

 7   sense.

 8             But were they -- was the death due to the tumors

 9   or was it due to infection or renal toxicity, which is the

10   inhalation study which is what kills everything off.

11             DR. SANDY:  It's not reported by the investigators

12   that there was any infection.

13             DR. SALMON:  I think if you look at the lymphomas

14   and leukemias are in fact a regular cause of death in aged

15   rats.  So if we assume that the tumors here are lethal

16   tumors, then the incidence reported in the high dose group

17   would have a significant impact on the survival.

18             DR. BYUS:  Right.  If that is in fact the case.

19   And then you can do the aged -- you can somewhat adjust your

20   incidence values if that's the cause of death, but if it's

21   not the cause of death, then it becomes much more

22   complicated to do that.

23             DR. SALMON:  One of the problems with this

24   particular study, as far as doing this type of analysis is

25   concerned, is that whereas with the inhalation studies we do



 1   have the full individual animal data, actually Dr. Maltoni's

 2   group have never published the individual animal typed tumor

 3   and mortality data.  We don't have that.  So we're basically

 4   left with the analysis of the --

 5             DR. BYUS:  And you did do the incidence aged --

 6   the death rate adjusted without that data?

 7             DR. SALMON:  We used the data as reported, which

 8   for the statistical analysis, which you see on the slide

 9   here, includes the information on the number of animals

10   alive at the point in time of appearance of the first

11   characteristic tumor, but that information is reported.

12             But for the potency analysis, we felt that that

13   would actually distort the number, because of the

14   considerably longer observation period, which is included.

15   So for the potency analysis we actually used the number at

16   risk as the number started on the start date.

17             DR. FRIEDMAN:  I'd like to follow up with another

18   question that relates to other studies of this type.  I mean

19   I am impressed with the trend as you increase the dose going

20   from a 3.4 percent up to 25.5 percent, but I'm wondering

21   could there be a multiple comparisons problem here?  In

22   other words, was it any special reason to pick those cancer

23   sites where they hypothesized in advance?  In other words,

24   there's many potential cancer sites.  Could others have gone

25   the opposite way and this just be a chance finding?



 1             DR. SALMON:  In fact, if there are any significant

 2   changes in, as you say, in the opposite direction, those

 3   would have been identified by the pathology analysis, and

 4   should have been reported had they occurred.  So and

 5   certainly if you have the full report of the study like

 6   this, you are in a position to examine the possibility of

 7   trends like that.

 8             This particular analysis is not in any way

 9   different from the analyses which are done on studies of

10   this type generally, and in fact there has been quite a

11   debate of some years ago on whether or not it was

12   appropriate to apply a continuity correction to significance

13   tests and things of that sort.

14             And so the statistical tests which we used and the

15   ones which Dr. Maltoni's group used here to some extent

16   reflects the fact that those types of issues have been

17   hashed over in the analysis of study designs of this type

18   over a number of years.

19             Is that -- am I missing anything?

20             DR. FRIEDMAN:  Is there an overall -- was the

21   incidence of all cancers combined higher in the treated

22   rats?  I mean, because you might think of that as a first

23   screen --

24             DR. SALMON:  To be honest, I don't have those

25   numbers in front of me, but I'm pretty confident that that's



 1   the case.

 2             DR. FRIEDMAN:  Is that something that's routinely

 3   looked at and sort of as a first screen before you look at

 4   the individual sites?

 5             DR. SALMON:  The way one of these studies would be

 6   conducted, the animals would be subject to gross

 7   pathological examination which would identify all tumors,

 8   including of course the control groups.  And then a

 9   histopathological analysis of all observed tumors and of

10   tissues from either all exposed animals or a selection of

11   tissues.

12             But normally there's a comprehensive analysis

13   which looks specifically not just at things which jump out

14   at you, but what are the backgrounds in various tissues.

15   And a well-designed study typically will include a fairly

16   thorough histopathologic analysis of any tissues in which

17   effects are either seen or anticipated.

18             DR. FRIEDMAN:  Thank you.  That's reassuring.

19             DR. SANDY:  If I can add a few things.

20             Dr. Sandy, OEHHA.

21             Back to the number of animals in the study.  The

22   females, there were 60 animals per group and males 60

23   animals.  So in the females you see there was not much high

24   mortality down to 58 in the controls, the 56 weeks.

25             And just another point that the statistical



 1   analysis that was reported in the Belpoggi study was

 2   performed by Dr. David Hoel and his group, and that is

 3   indicated in the paper.

 4             DR. BYUS:  My last question.  Sorry.

 5             This was a two-tailed analysis or one-tailed, for

 6   the significance.

 7             DR. SANDY:  I believe it's a one tail.

 8             DR. BYUS:  One tail?  Is that right?

 9             DR. SANDY:  I believe that --

10             DR. BYUS:  That's bad.

11             DR. SANDY:  -- is often common, commonly used when

12   we --

13             DR. BYUS:  It was only done one tail.  That

14   increases the significance significantly.  I just went

15   through this with a journal.

16             DR. SANDY:  I'll have to check.

17             DR. BYUS:  Check, please.

18             DR. SANDY:  The report --

19             DR. BYUS:  It doesn't say anywhere in any of these

20   documents.

21             DR. SANDY:  What we were able to do in OEHHA was

22   we did a one-tail T test or Fisher --

23             DR. SALMON:  Fisher Exact.

24             DR. SANDY:  It was analyzed in the paper and we've

25   indicated that here as significant by a log-ranked test and



 1   I'm --

 2             DR. BYUS:  But the P values you report are one

 3   tailed or two tailed here in this document?  That's all I

 4   need to know, because if they're one tailed it's usually

 5   much higher degree of significance than a two tailed.

 6             DR. BLANC:  All I was going say is it's only an

 7   issue for your values there that are presented as P less

 8   than .05 and presumably are between .05 and .01.  If that's

 9   your concern, you can just divide the ones by .01 by one and

10   a half and --

11             DR. BYUS:  The one is .05 and the other one is --

12   nothing is above --

13             DR. SALMON:  It's just the way it's always done

14   for this kind of analysis is the -- I mean, certainly if

15   what you're saying is that is the gold standard of

16   significance the .O1 and the .05 not quite so good, then,

17   yes, I would agree with you.

18             The methodology here is exactly what you've seen

19   for the last, however many other hundred --

20             DR. BYUS:  Sure.  Is it one tail?

21             DR. SALMON:  It's one tail, because it's whether

22   or not it's significantly different from the control group.

23             DR. FRIEDMAN:  It sounds like you're making the

24   assumption that it can only be harmful.  You aren't looking

25   at the possibility that it could be beneficial, if you don't



 1   do a two-tail test.

 2             DR. COLLINS:  The controls are two at 58.  If you

 3   went down to zero, 58 you couldn't see any significance.

 4             DR. SALMON:  I think the expectation -- there are

 5   very very few cases in which the effects which you're seeing

 6   is anything other than harmful.  There are a small number of

 7   instances where you see reduction of incidents relative to

 8   the control group, due to the variety of interests, many of

 9   which actually are reflective of defects in the execution of

10   the study or something like that, differential mortality,

11   things of that sort, but usually that isn't an issue.

12             DR. FRIEDMAN:  Craig, if you agree with that

13   assumption that these --

14             DR. BYUS:  I don't agree.

15             DR. FRIEDMAN:  -- can only be harmful, then the

16   one tail --

17             DR. BYUS:  I just went through this long

18   discussion of this with a journal editor and we did the

19   analysis both ways.

20             Statistically, if you design your experiment to

21   look for only an increase in incidents, and for example if

22   you set up so that your control groups have a very very low

23   tumor incidence and you're only looking for an increase,

24   there's only any room to see an increase, a one-tailed

25   analysis, at least by my estimation, is appropriate,



 1   although the journal -- and this is the way we did the

 2   experiment to look for an increase only, because it was not

 3   set up to look for a decrease.  If you want to look for a

 4   decrease you want to pick an incidence, we can control tumor

 5   incidence, we would pick an incidence that was higher, so

 6   you can see it go both ways.

 7             So we argued that a one tail was appropriate, and

 8   the journal refused this and said that the standard is a

 9   two-tailed test.

10             And in this case since you weren't -- you were in

11   a sense looking for whatever in the beginning, either an

12   increase or a decrease, you're not -- the experiment wasn't

13   biased one way or the other by design ahead of time.

14             DR. SALMON:  Yes.  It's biased in the sense that

15   the control incidence of any these tumors is relatively

16   extremely small.  So your chances of seeing --

17             DR. BYUS:  In this experiment it is.

18             DR. SALMON:  That's --

19             DR. BYUS:  The other Leydig cell it's very very

20   high.

21             DR. SALMON:  I think where the control incidence

22   is high, that certainly seems a defect of the experimental

23   data.

24             DR. BYUS:  Most the literature will report the

25   two-tailed test, which is usually lower significance.



 1             DR. SALMON:  I can't --

 2             CHAIRMAN FROINES:  Can we hold on.  Stan's been

 3   trying to enter in.

 4             DR. GLANTZ:  I actually was just talking about

 5   this in my statistics class yesterday, and I think --

 6             DR. BLANC:  Glad you're taking a class.

 7             DR. GLANTZ:  I am.  Thank you for sharing that.

 8             Actually, I think you were right and the journal

 9   was wrong.

10             DR. BYUS:  Thank you, Stan.

11             DR. GLANTZ:  Okay.  And it really does come down

12   to -- and Gary made this point earlier.  It comes down to

13   what the question is.  If the null hypothesis is that

14   there's no increase in tumor incidence, then a one-tail test

15   is completely appropriate, but then you are also basically

16   saying you don't care if this is good for you to be drinking

17   MTBE.  And I mean if that's the case, then it's appropriate

18   to use a one-tail test.

19             I mean the thing that I'm a little confused by in

20   looking at the slide here and that I was confused by when I

21   read the report was when you're talking about a log-rank,

22   it's not quite clear what you did here, because when you

23   talk about a log-rank test, that's usually done as a

24   two-tailed test and then you're saying you're doing a

25   one-tail Fisher Exact test and it seems that those are



 1   inconsistent.  I mean, you need to be consistent in the

 2   approach that you're taking.

 3             DR. SANDY:  Maybe I didn't explain it correctly.

 4             We're reporting what was reported by the

 5   investigators.  They have the original data, the

 6   time-to-tumor data.  We could not obtain that.  So we're

 7   reporting their statistical result.

 8             DR. SALMON:  Fisher Exact.

 9             DR. SANDY:  Their result, the investigator's

10   result, was the logged-ranked test.

11             Then we have done, based on just the data that's

12   here presented in front of you in the table, without

13   time-to-tumor data, the Fisher Exact test, which was a

14   one-tailed test.

15             DR. SALMON:  If I can append to that, certainly

16   Dr. Glantz has hit the nail on the head when he points out

17   that the expectation in the design of a standard

18   carcinogenesis bioassay, the default assumption is that

19   there is no tumor increasing effects and the study design is

20   predicated on that as a null hypothesis.

21             Now, that's the standard approach which is used in

22   conducting bioassays and that is the model which we used in

23   choosing the analysis which we reported that we were able to

24   do, but it clearly -- Dr. Hoel's analysis is different and

25   somewhat more complicated, but unfortunately we're actually



 1   not privy to many of the details of that.

 2             DR. GLANTZ:  At the very least, I think you need

 3   to clarify this in the report.

 4             The other thing is it's called the log-rank test,

 5   not the log-ranked test.  No e-d.

 6             DR. SALMON:  We can amend that.

 7             CHAIRMAN FROINES:  Let's go ahead.

 8             DR. GLANTZ:  That's a major criticism.

 9             DR. SALMON:  Actually, I don't know if we actually

10   cite that in the report, which we have for the -- but

11   anyway.  We will note that for future reference.

12             The next study is an inhalation study which was

13   conducted in rats.  One of the things to note is that in

14   fact this was a different strain of rat.  This was the

15   Fisher 344 as opposed to the Sprague-Dawley, which was used

16   by the Belpoggi study.

17             The observations in this case were in male rats

18   only, the renal tubular adenoma and carcinoma and again and

19   interstitial cell tumor of the testes.

20             And many of the same comments about the

21   statistical analysis apply here, except that in this case we

22   are reporting the usual analysis by Fishers Exact test.

23             And one of the problems with this study is

24   relatively poor survival.  Mid and high dose animals did

25   suffer quite significant mortality during the study.  In



 1   fact the exposed groups were terminated somewhat earlier

 2   than the projected 104-week duration of the study.

 3             DR. BYUS:  I've got a comment, question.

 4             Now, the incident of the Leydig cell tumors here

 5   is much higher in this strain of rats than the other strain.

 6   I mean, clearly there's this huge strain dependence on this.

 7   And people commented that even in incidence in the controls

 8   of 64 percent is on the low side of the historical range for

 9   Leydig cell tumors in Fisher rats.  They're even up to 70, I

10   forget what the number is, the range is quite high.

11             DR. SALMON:  Yes.

12             DR. BYUS:  So my question to you is what this

13   looks likes to me for this Leydig cell tumor data, which is

14   used as the best evidence for a dose response, somewhere in

15   the document, it's one of the documents --

16             DR. SALMON:  It's not that result actually.

17             DR. BYUS:  It is not that result?

18             DR. SALMON:  No.

19             DR. BYUS:  But in any --

20             DR. SALMON:  I'll describe that in a little while.

21             DR. BYUS:  I'll show you where it's in here.  Not

22   your document.

23             Anyway, what this looks like to me for the Leydig

24   cell tumors is really an effect on lag time, not on

25   incidence.  In other words, all these animals basically



 1   ultimately are going to get Leydig cell tumors, and what

 2   you're seeing here is an increase or decrease in the time

 3   that it takes for them to show up.

 4             DR. SALMON:  I think you could describe any

 5   carcinogenic effects which involves a naturally occurring

 6   tumor as an effect on the time to incidence of the tumor.

 7             DR. BYUS:  You can, but it becomes much more a

 8   relevant argument when the incidence approaches 100 percent

 9   in the controls.  Especially historical, but even here is 60

10   something percent.

11             In my opinion it's much more likely.

12             And the reason I'm making this point is if it is

13   an effect on lag time it's more of a promotional stimulus,

14   and that means it's likely to have some kind of threshold,

15   rather than being a nonthreshold type of response.  That's

16   why I think it's important at least to consider it.

17             So in other words if you wait long enough, they're

18   all going to have the same incidence, which is not the case

19   in the leukemia study that you showed before or in the

20   Leydig cell study you showed before.

21             CHAIRMAN FROINES:  Remember, that you notice again

22   your mid and high dose animals are terminated at 97 and 82

23   weeks, so that in fact your high dose animals are being

24   terminated earlier in the study, rather than later in the

25   study.



 1             DR. BYUS:  That's true.  But the problem is

 2   because over on top of this problem with this study is this

 3   horrible renal toxicity that is killing off the animals in

 4   the high doses.

 5             So I just think it's just not very good from these

 6   points of view.  You have a very high toxicity, renal

 7   toxicity.  You have certainly in the Leydig cell, you have a

 8   very high incidence, which is increased statistically here.

 9   But to me it looks like an effect on the lag time only and

10   not necessarily any real effect on an incidence mechanism.

11             DR. SALMON:  A theoretical analysis of tumor

12   incidence actually does treat the time to tumor appearance

13   being reduced by the exposures to the carcinogen as a

14   carcinogenic effect.  So that isn't conceptually -- there's

15   not conceptually saying anything different than what the

16   underlying assumption of the model is in the first place.

17             However, I would entirely agree with you that

18   there are significant problems with this study, one of which

19   is the high incidence of tumors in the testes and another of

20   which is the relatively severe toxicity and early mortality

21   in the study.

22             I would say that although clearly this study, in

23   our opinion, has some very significant defects, and contrary

24   to some of the comments of which we've heard from elsewhere,

25   we inclined to the view that the Bellpoggi study is actually



 1   a much better one in terms of both its conduct and it

 2   interpretability.

 3             Nonetheless, yes, there are certainly some

 4   significant defects to this study.

 5             However, I think it's a moot point as to whether

 6   those are any worse than the defects of studies which are,

 7   you know, often used in other contexts.

 8             But I would also say that, you know, one of the

 9   ways that we can address whether or not these defects are

10   fatal is to actually look at the dose response calculations,

11   which I will describe in a minute, in a few minutes, to see

12   to what extent those results are consistent with other

13   studies, because I think that where you have a study which

14   has defects of this sort that does become an important

15   question.

16             CHAIRMAN FROINES:  I think at this point we should

17   go through and let you finish, because we're going to have

18   to stop and I'd like to have a complete presentation on the

19   material before the panel.

20             I know Craig wants to jump in, but --

21             DR. BYUS:  It's all right.

22             CHAIRMAN FROINES:  I think it be better if we got

23   through the entire discussion.

24             DR. SALMON:  Okay.  Well, the mouse study

25   inhalation study in fact reports incidence of hepatocellular



 1   adenoma and carcinoma, again with significant mortality

 2   issues with this study.

 3             Obviously, there are other things which can be

 4   said about all these studies.

 5             I'd like to proceed now to just describing what we

 6   did with these data, which basically we were trying to make

 7   the best of what we had available in coming up with a

 8   number.

 9             And one of the things we did, because we had

10   studies both by inhalation and by gavage, we developed a

11   pharmacokinetic model, which was designed to allow us to

12   compare the results of the studies by the two routes.

13             We did not have sufficient pharmacokinetic data in

14   mice to be able to apply the model to the mice, so we were

15   only able to consider the rat data for the pharmacokinetic

16   models.

17             If I can have the next slide, please.

18             This is one of the sort of one of those

19   isn't-that-neat type pictures.  This is a diagram of the

20   model.  Obviously, if you require more details, then we can

21   point you to those.

22             But the idea here was, as I said, was to allow us

23   to compare the results of the inhalation studies and -- or

24   the rat inhalation study and the mouse.

25             DR. GLANTZ:  Can I ask a quick question here.



 1             When you look at this model and you describe it in

 2   the report, there's a zillion parameters.

 3             DR. SALMON:  Yes.

 4             DR. GLANTZ:  That go into all these

 5   pharmacokinetic models.  How good are those parameters?

 6             DR. SALMON:  As many of them as are possible are

 7   externally determined.  Things like blood flow rates and

 8   breathing rates and things of that sort are obtained

 9   independently from other sources.

10             The partition coefficients are things which in

11   general can be at least estimated on the base of independent

12   experimental observations.

13             However, I think that that's a legitimate question

14   as to how good the individual detail predictions of the

15   model are.

16             As far as the use that we are putting the model to

17   is concerned, I think the primary thing is does it make a

18   reasonable job of fitting the observed data and is it

19   therefore appropriate for making a fairly crude level of

20   estimate.

21             And what we were looking for basically was the

22   overall amounts of material metabolized at the various doses

23   and by the two routes.

24             So I think what I'm saying is that some of this

25   detail, obviously the individual parameters, particularly



 1   those which are not externally measured and valued against

 2   data obtained from other sources, are somewhat uncertain,

 3   but that the overall conclusion of the model, if you like,

 4   is probably not too far from the truth, especially provided

 5   you don't extrapolate too far away from the starting point.

 6             DR. BLANC:  Stan, we better move on or we're not

 7   going to be able to hear from our other speaker.

 8             DR. GLANTZ:  If I can say one quick thing.

 9             I think that -- and this is something I've talked

10   about and we've done before -- I think it would be useful to

11   do a little sensitivity analysis on the parameters and make

12   sure that the results are not highly dependent on the

13   specific values.

14             Let's just go on.

15             DR. SALMON:  Yes, I agree with you.  And in

16   general we do do things like that.

17             DR. GLANTZ:  Go on.  I'm sorry.

18             DR. SALMON:  I also wanted to just include a slide

19   explaining what we were doing here was using the LED 10

20   methodology as described in the proposed risk assessment

21   guidelines from US EPA, rather than the traditional

22   linearized multistage model.

23             We actually did both calculations, and they don't

24   produce a very substantially different result.  But we tend

25   to prefer this methodology, especially in cases where the



 1   underlying mechanism is uncertain and we aren't therefore

 2   necessarily entitled to assume that the linearized

 3   multistage model is a good description of the dose response

 4   across all levels of dose right away from the experimental

 5   doses down to low levels of exposure in the environment.

 6             However, one of the things which received quite a

 7   bit of discussion in the PHG reports was the fact that the

 8   mechanism of the effect here by MTBE is uncertain.

 9             We therefore chose the default methodology, which

10   does assume a linear dependence of risk on dose, low doses.

11   So we are using a linear extrapolation here, but it's based

12   on the LED 10 benchmark, rather than basing it on the

13   traditional linearized multistage model.

14             DR. FRIEDMAN:  What do P and D stand for in that

15   equation?

16             DR. SALMON:  Probability of tumor at dose D.

17   Okay.

18             Now, this slide reports the route-to-route

19   extrapolation for the rat oral inhalation studies.

20             And what I want to point out to you, I'll try and

21   do this briefly here, is that with the exception of the

22   testicular tumors in the inhalation rat study, the Chun et

23   al study, the other numbers are in fact consistent within a

24   fairly narrow range, and we do believe that the divergence

25   of the number for the testicular tumors in the Chun et al



 1   rat does in fact reflect the problems with the high

 2   background incidence and such things, which is exactly what

 3   Dr. Byus was just talking about just now.

 4             So in a sense I think what I'm saying is that I

 5   agree that this defect does cause a problem for the

 6   analysis, but it isn't perhaps disastrous, but we can see

 7   how it is influencing the number.

 8             Apart from that, the Leydig cell tumors in the

 9   leukemia lymphomas in the Belpoggi study and the renal

10   tubular cells in the male rats of Chun et al are fairly

11   concordant, at least very much so within the sort of bounds

12   of confidence that we have with these kinds of estimates.

13             The other thing which I haven't actually got on

14   this slide, which is described in the PHG report, is the

15   fact that although we can't compare the mouse results on the

16   slide, because we don't have a mouse pharmacokinetic model,

17   in fact the mouse liver tumor results are also consistent

18   with the Chun et al rat renal tubular cell adenoma and

19   carcinoma data.

20             So by and large we're seeing that although there's

21   a diversity of sites, a diversity of species, a diversity of

22   strains, and a number of questions which have been raised

23   about the individual studies, by and large we are able to

24   obtain a reasonably consistent estimate of carcinogenic

25   potency.



 1             This slide also incidentally illustrates the point

 2   I made just now, that in this particular data set, the LED

 3   10 methodology does not produce a highly different result

 4   from what you see with the traditional linearized multistage

 5   model, which is what is reported here as Q 1 star in the

 6   table.

 7             If I can have the next slide, please.

 8             DR. GLANTZ:  Can I ask one other quick question?

 9             DR. SALMON:  Sure.

10             DR. GLANTZ:  One other thing that I had a hard

11   time figuring out, which this slide I also maybe I wasn't

12   understanding, is how the model predictions from the PBPK

13   model compared to the actual observed data.  I mean, that's

14   not in this slide, is it?

15             DR. SALMON:  No.  The numbers from the -- well,

16   from the -- with the actual pharmacokinetic data --

17             DR. GLANTZ:  Well, no.  You had --

18             DR. SALMON:  You talking about the potencies

19   calculated on the basis of the exposed concentrations?

20             DR. GLANTZ:  Again, I don't want to derail the

21   conversation, but if need be we can come back to this.

22             DR. SALMON:  Those data are presented in the PHG

23   report, so we can discuss those later if you would like to

24   do that.

25             DR. GLANTZ:  I don't want to derail.  We can put



 1   this on the agenda for later.

 2             But when I was looking through the report, like I

 3   think it was table 11, 13, yeah, 11 and 13, where it looked

 4   like you were coming up with predictions the PBPK model and

 5   then presenting actual data, and I was having a very hard

 6   time making the comparison and deciding if the predictions

 7   were good or bad.

 8             I think I don't want to derail the discussion.

 9             DR. SALMON:  The data are in the report.

10             DR. GLANTZ:  Maybe next time we discuss this you

11   can address that specifically.

12             Go on.  I'm sorry.

13             DR. SALMON:  So in order to obtain a best estimate

14   for the cancer potency, we actually took a geometric mean,

15   and the data we used were those data sets which were

16   reasonably consistent, which was the kidney tumors in the

17   rat inhalation study, and both the leukemia lymphoma in

18   females and the Leydig cell tumors in the males from the

19   Bellpoggi study.

20             The individual tumor endpoints varied over what by

21   the standards of these sorts of estimates is considered to

22   be a relatively small range.

23             The other thing which we considered was we did

24   look at a time-to-tumor analysis, which we could do from the

25   inhalation study, because we have those data, but we didn't



 1   particularly have much confidence in that, because we felt

 2   that one of the problems with the time-to-tumor analysis is

 3   that you're actually extrapolating to a theoretical end of

 4   lifetime tumor incidence, which in this case unfortunately

 5   is well outside the range of any observations in the study.

 6             So we felt this estimate, which was rather higher

 7   than the one we quoted, it had a rather low degree of

 8   confidence.  We didn't use that estimate.

 9             If I can have the next slide.

10             We used the potency estimate calculated using the

11   pharmacokinetic model to derive an inhalation potency for

12   humans at low dose, and we relied on the recently published

13   data by Nihlen and colleagues suggesting that approximately

14   50 percent of the inhaled MTBE at low doses would be

15   metabolized.

16             And making the usual assumptions we come up with

17   an inhalation potency estimate as a unit risk factor of 9.3

18   times 10 to the minus 7 part per billion, or 2.6 times 10 to

19   the minus 7 per microgram per cubic meter.

20             I'm just going to briefly summarize.

21             We're relying on the identification of this

22   substance as a toxic air contaminant with other hazardous

23   air pollutants as Dr. Alexeeff described at the beginning.

24             We are depending for the further analysis of the

25   carcinogenesis data, firstly on the University of California



 1   report, which Dr. Froines has, I think, already drawn to

 2   your attention, and also on the hazard identification and

 3   quantitative analysis by OEHHA, which is in the PHG report.

 4             Cancer slope factors based on oral and inhalation

 5   experiments in two different strains of rats at three sites.

 6             Pharmacokinetic model is used for interroute

 7   extrapolation in animals.

 8             That's the end of what I have to say.

 9             DR. MARTY:  We had a public comment period on the

10   derivation of the inhalation unit risk factor at the

11   beginning of the summer, and we got comments from the

12   Oxygenated Fuels Association, the Western States Petroleum

13   Association, and from Dr. Richard Wilson for Lyondell

14   Chemical Company.

15             Basically, most of the comments concerned what I

16   call hazard identification issues, is it or is it not a

17   carcinogen.

18             And also the choice of the model used to develop

19   our public health goal number, and also our cancer potency

20   factor.

21             I think it's easiest just to briefly go over the

22   hazard ID issues.  The concerns raised were that it's

23   probably not genotoxic, therefore you should not use a

24   linear, type of a linear model to estimate the cancer

25   potency factor.



 1             That carcinogenicity may be a result of toxicity

 2   resulting from high doses that were used.

 3             And that the types of tumors present have

 4   questionable relevancy to people, namely the male kidney

 5   tumors, the mouse liver tumors and the Leydig cell tumors,

 6   for a variety of reasons.

 7             I thought it would be most appropriate to let

 8   Martha Sandy address those issues.  The same issues came up

 9   when they sent the public health goal out for a public

10   comment and peer review.

11             And then finally there was one comment from

12   Dr. Richard Wilson.  What he did was went back and he

13   calculated his own estimate of what the cancer potency

14   factor would look like and it's about he estimated a PHG

15   based on carcinogenicity that's about six fold higher than

16   the PHG that was eventually adopted by OEHHA.  He based it

17   on his own estimates of potency and exposure assumptions.

18             So given that there's a lot of uncertainty and

19   there's all these issues that come to bear, my personal

20   opinion is that that's pretty good agreement and doesn't

21   really represent a major disagreement.

22             CHAIRMAN FROINES:  Six fold meaning six fold less

23   health conservative?

24             DR. MARTY:  Right.  Correct.

25             CHAIRMAN FROINES:  We have a cab for the people



 1   who are leaving and so they should leave here about 25 to

 2   3:00.

 3             Is that right, Peter?

 4             He says ten minutes, the cab will be here in ten

 5   minutes.

 6             DR. MARTY:  Martha has six slides.

 7             DR. SANDY:  You have copies of them.  I don't know

 8   if you'd like me to go through them or not.

 9             CHAIRMAN FROINES:  Just go through what you want.

10             DR. SANDY:  Very quickly.  You've seen this

11   before, this is just a summary of the data we have, where

12   this data is coming from.

13             And again the tumor sites, lymphomas and

14   leukemias, which are, as far as I know there is no question

15   as to whether they're relevant to humans.

16             Leydig cell tumors are still considered for risk

17   assessment purposes relevant to human risk.

18             Renal tubular adenomas and carcinomas, the

19   question here was whether alpha 2 U globulin nephropathy was

20   involved.

21             And very quickly it appears that MTBE may be

22   binding weakly to that male rat specific protein.  So there

23   may be some interaction there, but it's very very weak and

24   it doesn't seem to explain the entire picture or explain

25   entirely the concerns of those tumors in the bioassay.



 1             And then in mouse we have the hepatocellular

 2   adenomas and carcinomas.

 3             MTBE is not nonpositive in a number of

 4   genotoxicity assays.  There are a couple where we have seen

 5   positive results.  It's positive in the activated mouse

 6   lymphoma forward mutation assay and the rat lymphocyte comet

 7   assay for DNA strand breaks.

 8             The metabolites, I guess I should back up.  We

 9   don't know if it's MTBE that's the active carcinogen, or if

10   there are metabolites of MTBE that are contributing to the

11   tumors we see on the animal studies.

12             The tumors -- the metabolites that have been

13   tested in genotoxicity assays, there's only two.  One is

14   tertiary butyl alcohol, it's nonpositive.  Formaldehyde has

15   been tested and is positive.

16             And further studies of the possible involvement of

17   formaldehyde in inducing the mouse liver cell tumors, the

18   studies that have been performed to date do not suggest a

19   role for formaldehyde at that tumor site, but we haven't

20   looked at other tissues.

21             Next slide.

22             MTBE has induced renal tubular cell adenomas and

23   carcinomas --

24             CHAIRMAN FROINES:  I should say that, remember,

25   that isobutane is also a combustion product and it is



 1   mutagenic.

 2             DR. SANDY:  That's right.

 3             For tertiary butyl alcohol, we have observed the

 4   same type of kidney tumors in Fisher rats exposed via

 5   drinking water.

 6             The inhalation studies were in Fisher rats as

 7   well.

 8             Formaldehyde is induced, leukemias in

 9   Sprague-Dawley rats exposed via drinking water, so there's a

10   commonality of tumor sites between these metabolites and

11   MTBE.

12             And the next slide.

13             There are a number of concerns, as Andy has

14   alluded to, with the data.  All of the studies have their

15   problems.  High doses, et cetera, for study design problems.

16   Early mortality.

17             Study reporting, we would like to have more

18   complete reporting of all the studies, not just the

19   Bellpoggi study.

20             Relevance of tumor site and type.  I've just

21   described.

22             Next slide, please.

23             Then I have slides on the alpha 2 U globulin,

24   nephropathy, which I don't know if I need to go into detail

25   here.



 1             Dr. Froines, or the committee, would you like me

 2   to elaborate a little bit on this?

 3             CHAIRMAN FROINES:  We're not going to really have

 4   any time to discuss any of this so you might as well go fill

 5   out the time and then we'll take it up next time.

 6             DR. SANDY:  Let me back up a bit then.

 7             There are no mechanistic hypotheses that are out

 8   there to explain the lymphomas or leukemias that have been

 9   seen in the female rats in the gavage study.

10             It has been hypothesized that the liver tumors

11   seen in the mouse, and specifically the female mouse, which

12   had an increase of adenomas and carcinomas.  The male mice

13   only had a an increase in the carcinoma.

14             DR. BYUS:  This thing is very unusual, that

15   effect.  Go ahead.

16             DR. SANDY:  People have looked at whether looking,

17   focusing just on the liver tumors in the female mice, if

18   there could be some sort of hormonal interaction going on or

19   promotional effect of MTBE.  And the data available to date

20   does not show that out.  There is no hypothesis that has

21   been supported by the data, so we do not know what the

22   mechanism of action in the mouse liver is.

23             To go to the rat kidney tumors, the alpha 2 U

24   globulin nephropathy hypothesis has been pursued to a large

25   extent, and there's a lot of data.



 1             There are criteria that have been developed by

 2   both US EPA and more recent by IARC, which are there to help

 3   evaluate a study to decide if in fact an agent is working

 4   through this mechanism.  We have looked at that using both

 5   the US EPA criteria and the IARC criteria and we felt that

 6   the data available for MTBE just do not indicate that this

 7   is an agent working solely through this mechanism.

 8             The data supporting this mechanism include

 9   observed of mild to moderate number of {SKPAOEUZ} drop

10   {HREUTS} in {PHAOEPBL} approximate mall tubular cells which

11   stains weakly for alpha 2 U globulin.

12             There's a slight dose dependent increase in alpha

13   2 U globulin in renal reactivity in the rat kidney cytosol

14   by the Elisa assay and here we're reporting results that

15   came out in '97 by inhalation where they found an increase

16   of up to 150 micrograms of the alpha 2 U globulin per mig

17   total protein was the control level and the level in the

18   MTBE treated at the high dose by inhalation was 200.  So

19   there wasn't much of an increase.

20             There's a recent paper out in Toxicological

21   Sciences where by an oral dose of MTBE they now have they're

22   observing the bindings up to 250 micrograms.  So it's still

23   very weak binding.  They have statistical significance, but

24   it's not great level binding.

25             And MTBE binds weakly to alpha 2 U globulin in



 1   vitro.

 2             There's a dose dependent increase in cell

 3   proliferation in the renal cortex.

 4             And MTBE has little or no genotoxicity activity.

 5             Next slide, please.

 6             Now, there's on the other side data indicating

 7   that there's lack of a significant role of this nephropathy

 8   in the development of tumors.  It does not appear to be

 9   specific to the male rat for the nephropathy or the renal

10   tumorigenicity.  There was an observation of one renal cell

11   adenoma in the female and the females had chronic

12   progressive nephropathy dose dependent with MTBE.

13             No clear exposure related increase in staining for

14   alpha 2 U globulin was observed in the MTBE treated male

15   rats.

16             And there's no alpha 2 U globulin positive

17   proteinaceous casts at the junction of the proximal tubules

18   in the thin loop of Henle, and these casts are something

19   that's a hallmark of alpha 2 U globulin nephropathy.

20             No linear mineralization of papillary tubules was

21   observed.  Again, that's another hallmark of this condition.

22             No binding of MTBE to alpha 2 U globulin or male

23   rat renal proteins was observed in vivo.

24             And a very weak interaction between MTBE and the

25   male rat renal proteins hasn't been observed in vitro using



 1   a kidney homogenate system and the interaction did not

 2   survive dialysis or anion exchange chromatography.  Again,

 3   this has been looked at in this more recent paper and again

 4   confirming it's a very weak interaction.  It doesn't survive

 5   chromatography.

 6             In summary, we do see tumors, Leydig cell tumors

 7   in the testes, leukemias and lymphomas in both sexes of the

 8   rat in a lifetime gavage study.

 9             We have Leydig cell tumors in the testes and renal

10   cell tumors in male rats of a different strain in a 24-month

11   inhalation study.

12             We have hepatocellular tumors of both sexes of the

13   mouse in an 18-month inhalation study.

14             There's little or no genotoxicity observed.

15             The mechanism is unknown for induction of tumors

16   at all these sites.

17             There's supporting evidence provided by the

18   carcinogenic activity of formaldehyde and TBA, which are

19   primary metabolites of MTBE.

20             Thank you.

21             CHAIRMAN FROINES:  Well, Peter, what do we have

22   time wise?

23             Five minutes.  Five minutes of comments.

24             Who would like to start?

25             DR. BLANC:  Well, I'm very, again, I'm, despite



 1   what you said at the beginning, I'm a little unclear as to

 2   how this panel can be most useful in regard to this single

 3   point estimate that I assume you are required to promulgate

 4   as a supplement to previous analyses.  So how can we be of

 5   service?

 6             DR. ALEXEEFF:  I think here's the issue.

 7             Like a previous compound that the panel dealt

 8   with, DEF, MTBE has not been identified as a carcinogen,

 9   either in the state or by IARC or nationally, so it's not a

10   identified carcinogen.  But there is data, as you've seen,

11   and now you've seen the good, the bad and the ugly of that

12   data.

13             But under the statute that we're working under

14   where in the face of uncertainty we're supposed to estimate

15   a range of risk, and we're supposed to -- that is the sort

16   of the area that we're working in, so the question we're

17   asking you whether or not our use or our development of a

18   potency cancer potency estimate for this compound is

19   appropriate and whether the actual number we develop is

20   reasonable, scientifically speaking.  That's really what

21   we're asking you for.

22             How you actually agree with it or review it or

23   adopt it --

24             DR. BLANC:  And then you would be transmitting it

25   or having us transmit to the Air Resources Board?



 1             DR. ALEXEEFF:  No.  There's no transmittal to the

 2   Air Resources Board.  We will simply be -- we would record

 3   the conclusions similar as we did with the other chemicals

 4   we're working on, these multiple chemicals, and we would use

 5   that number and probably the Air Resources Board use that

 6   number of any calculations they need to make.

 7             DR. FUCALORO:  So the number you're thinking of

 8   recording is the one that's 2.3 times 10 to the minus 7

 9   meters cubed per microgram, is that right?

10             DR. ALEXEEFF:  Right.

11             DR. BLANC:  Is there some reason why you wouldn't

12   want to take, given the level of uncertainty, wouldn't want

13   to take the approach that you took with diesel exhaust of

14   having a range of estimates or reporting a range of

15   estimates, rather than coming down with a single number?

16             DR. ALEXEEFF:  We have a range of estimates in the

17   document already presented.

18             Previously in our PHG, in our water standard, we

19   pretty much had to come up with one level.  We had to sort

20   of make a decision for our water standard.  So part of it is

21   simply we just made -- decided to make the same decision for

22   the air.

23             But definitely we do have a range, and you can see

24   the uncertainty.

25             But once you throw out that study that Craig was



 1   referring to, that Leydig tumors where the high incidence,

 2   once you throw that out, the other numbers are fairly

 3   consistent, so the range would be fairly narrow.  I mean,

 4   it's not like diesel where you have a range of some, you

 5   know, at least an order of magnitude, let's say.

 6             DR. BLANC:  I mean, then I would say it would make

 7   your document even stronger if you would say, and

 8   furthermore if we take a range excluding the most

 9   problem-ridden study, the range is from 1.7 to 2.3, or

10   whatever the hell it is.

11             CHAIRMAN FROINES:  Let me interject, because we

12   have to stop.

13             I have a letter here from Mike Kenny from the Air

14   Resources Board and he says, and I quote, this letter is to

15   formally request that the Scientific Review Panel review the

16   Office of, well, OEHHA, enclosed documentation on the

17   carcinogenic potency of methyl tertiary butyl ether in

18   accordance with the usual procedures for peer review of the

19   health values for toxic air contaminants.

20             So we have a request basically to follow, quote,

21   usual procedures.  And within the context of the 1807 law,

22   that means that we have to determine whether or not your

23   document is, what's the words?

24             DR. GLANTZ:  Scientifically horrible.

25             CHAIRMAN FROINES:  No, no.  What is the word,



 1   Bill?  What's the expression?

 2             DR. ALEXEEFF:  It's scientifically sound.

 3             CHAIRMAN FROINES:  Not seriously deficient.

 4             DR. ALEXEEFF:  That language has been struck from

 5   ours.  It's now scientifically sound.

 6             CHAIRMAN FROINES:  This is from the Air Resources

 7   Board.  Therefore we have to meet the criteria within the

 8   context of the law that establishes this panel, I think.

 9             I don't know about what law you operate under, but

10   I know the law we operate under, and so we have to determine

11   whether or not this document is seriously deficient.

12             Is that correct or not correct, Bill?

13             This is from Mike Kenny, not from Joan Denton, and

14   he says the usual procedures for peer review.  And so by

15   that definition, as far as I know, we have to determine

16   whether it's seriously flawed.

17             MR. LOCKETT:  I think there is the statute allows

18   you to find the document acceptable in terms of current

19   science.

20             There is also under 1807 if you were going to do a

21   finding that it's not seriously deficient.

22             But I think in this case, review and approval of

23   the document will satisfy the statute.

24             CHAIRMAN FROINES:  Okay.

25             DR. ALEXEEFF:  The statute says you have to find



 1   that it's based upon sound scientific knowledge, methods or

 2   practices.  So it's a slight revision.

 3             DR. BYUS:  It's in the statute, which is not

 4   mentioned, I mentioned to you, it's whether or not a

 5   threshold exists for this compound.  We usually answer that

 6   question for each compound that we think is a carcinogen,

 7   does a threshold exist or not.

 8             That's another -- I mean, as I told you, it's

 9   another concern I have, especially on this compound.

10             CHAIRMAN FROINES:  Unfortunately, there's no

11   evidence whatsoever about a threshold.  It's almost moot.

12             DR. ALEXEEFF:  The question is can we identify a

13   threshold, is really what's asked, not does it exist or not,

14   but can we identify.

15             DR. BYUS:  Can we identify it.  Okay.

16             DR. BLANC:  I'd like to move that we adjourn the

17   meeting.

18             DR. KENNEDY:  Second.

19             (Thereupon the meeting was adjourned

20             at 2:43 p.m.)










 3             I, JANET H. NICOL, a Certified Shorthand Reporter

 4   of the State of California, do hereby certify that I am a

 5   disinterested person herein; that I reported the foregoing

 6   meeting in shorthand writing; that I thereafter caused my

 7   shorthand writing to be transcribed into typewriting.

 8             I further certify that I am not of counsel or

 9   attorney for any of the parties to said meeting, or in any

10   way interested in the outcome of said meeting.

11             IN WITNESS WHEREOF, I have hereunto set my hand

12   this 12th day of October 1999.




                                     Janet H. Nicol
17                                   Certified Shorthand Reporter
                                     License Number 9764