1

DEPARTMENT OF HEALTH AND HUMAN SERVICES

FOOD AND DRUG ADMINISTRATION

CENTER FOR FOOD SAFETY AND APPLIED NUTRITION

FOOD ADVISORY COMMITTEE MEETING

Advice on CFSAN'S Draft Report:

Approaches to Establish Thresholds for Major Food Allergens and for Gluten in Food

Thursday, July 14, 2005

8:30 A.M. to 5:20 P.M.

Greenbelt Marriott

6400 Ivy Lane

Grand Ballroom

Greenbelt, Maryland 20770

P A R T I C I P A N T S

FOOD ADVISORY COMMITTEE STANDING MEMBERS:

Richard A. Durst, Ph.D. - Acting Chairman

Jeffrey A. Barach, Ph.D. (Industry Representative)

Patrick S. Callery, Ph.D.

Dennis Gonsalves, Ph.D., M.S.

Jean M. Halloran (Consumer Representative)

Douglas C. Heimburger, M.D., M.S.

Margaret C. McBride, M.D.

Mark Nelson, Ph.D. (Industry Representative)

Carol I. Waslien Ghazaii, Ph.D., R.D.

TEMPORARY VOTING MEMBERS:

Petr Bocek, M.D., Ph.D. (Absent 7/14/05 Session)

Margaret Briley, Ph.D., R.D.

Erica Brittain, Ph.D.

Ciaran P. Kelly, M.D.

Soheila June Maleki, Ph.D.

David O. Oryang

Marc D. Silverstein, M.D.

Suzanne Teuber, M.D.

FOOD AND DRUG ADMINISTRATION:

Robert E. Brackett, Ph.D. - Director

Food and Drug Administration, CFSAN

Catherine Copp, J.D. - Senior Policy Advisor

Food and Drug Administration, CFSAN

Steven M. Gendel, Ph.D. - Senior Scientist

Food and Drug Administration

National Center for Food Safety and Technology

Rhonda Kane, M.S., R.D. - Consumer Officer

Food and Drug Administration, CFSAN

Marcia Moore, Food Advisory Committee, Executive Secretary

P A R T I C I P A N T S (Continued)

FOOD AND DRUG ADMINISTRATION STAFF:

Michael M. Landa, J.D. - Deputy Director for Regulatory Affairs

Food and Drug Administration, CFSAN

Stafano Luccioli, M.D. - Senior Medical Advisor

Food and Drug Administration, CFSAN

GUEST SPEAKERS:

Pekka Collin, M.D., M.P.H. - Professor

University of Tampere, Medical School, Finland

Catherine L. Copp, J.D.

Policy Advisor, CFSAN, FDA

Alessio Fasano, M.D. - Professor of Pediatrics

Medicine & Physiology and Director, the Mucosal

Biology Research Center, Center for Celiac

Research, University of Maryland School of

Medicine

Steven M. Gendel, Ph.D. - Senior Scientist

National Center for Food Safety and Technology, FDA

Rhonda R. Kane, M.S., R.D. - Consumer Safety

Officer CFSAN, FDA

Donald Kasarda, Ph.D. - Consultant and Retired

Senior Scientist, Agriculture Research Service,

USDA

Cynthia Kupper, R.D., C.D. - Executive Director

Gluten Intolerance Group of North America

Joseph A. Murray, M.D. - Professor of Medicine

The Mayo Clinic of Rochester, Minnesota

C O N T E N T S

PAGE

Call to Order and Welcome and Introductions

Richard Durst, Ph.D., Acting Chairman 6

Use of Gluten Thresholds

Catherine L. Copp, J.D., CFSAN, FDA 9

Introduction to Celiac Disease

Joseph Murray, M.D. 11

Patient Perspectives on Celiac Disease

Cynthia Kupper, R.D., C.D. 69

Grains

Donald Kasarda, Ph.D., USDA 84

Question and Answer Session 108

Prospective Studies

Alessio Fasano, M.D. 114

Question and Answer Session 146

Retrospective Studies

Pekka Collin, M.D., M.P.H. 161

Question and Answer Session 182

International Perspectives on Gluten-Free

Rhonda S. Kane, M.S., R.D., CFSAN, FDA 186

Question and Answer Session 198

Public Comments:

Elaine Monarch 212

Alice Bast 220

Mary Schluckabeer 225

Tom P. Sullivan 232

Steve Taylor 240

C O N T E N T S

PAGE

Overview of Approaches to Establishing

Thresholds: Gluten

Steven M. Gendel, Ph.D. 253

Question and Answer Session 257

Committee Discussion:

Panel Discussion with Guest Speakers 259

Gluten and Celiac Disease - FDA Questions 287

Revisit Food Allergens 354

Adjournment

Richard Durst, Ph.D., Acting Chairman 363

P R O C E E D I N G S

CALL TO ORDER AND WELCOME AND INTRODUCTIONS

CHAIRMAN DURST: Good morning. I would

like to call the meeting to order. All right, I

would like to welcome everyone back and also

welcome new participants in our meeting this

morning.

For those of you who weren't here

yesterday, there is a "Conflict of Interest

Statement" over on the table, if you want to refer

to that at all, otherwise I would also ask again

that maybe our participants or our members of the

Food Advisory Committee would introduce themselves

again for the benefit of those who were not here

yesterday.

I am Dick Durst, professor [of]chemistry at

Food Science and Technology Department at Cornell

University.

Marc, would you start it off?

DR. SILVERSTEIN: Marc Silverstein, I'm a

general internist and geriatrician at Baylor Health

Care System in Dallas.

DR. TEUBER: Suzanne Teuber, I am an

allergist at UC-Davis.

MR. ORYANG: I am David Oryang. I am a

risk analyst and agricultural engineer at the

United States Department of Agriculture, Animal and

Plant Health Inspection Service.

DR. KELLY: Good morning. Ciaran Kelly, I

am a gastroenterologist at Harvard Medical School

in Boston.

DR. MALEKI: I am Soheila Maleki. I am a

scientist with the USDA.

DR. BRITTAIN: Erica Brittain, I am a

statistician at the National Institute of Allergy

and Infectious Disease.

DR. BRILEY: Margaret Briley, University

of Texas at Austin, nutritionist.

MRS. MOORE: Marcia Moore, I am the

executive secretary for the Food Advisory Committee

and the Food and Drug Administration.

DR. WASLIEN: I am Carol Waslien,

nutritional epidemiologist at the School of

Medicine, the University of Hawaii.

DR. BRILEY: I am Margaret McBride, child

neurologist at Akron Children's Hospital.

DR. CALLERY: Pat Callery, West Virginia

University, pharmaceutical scientist.

DR. GONSALVES: I am Dennis Gonsalves, a

scientist at USDA.

DR. HEIMBURGER: I am Doug Heimburger, a

physician and nutrition specialist at the

University of Alabama at Birmingham.

DR. BARACH: Jeff Barach with Food

Products Association here, in Washington, D.C., in

regulatory affairs.

DR. NELSON: Mark Nelson with the Grocery

Manufacturers Association here, in Washington,

D.C., and I am responsible for scientific and

regulatory policy.

MS. HALLORAN: Jean Halloran with

Consumers Union located in Yonkers, New York, and I

am director of food policy initiatives.

CHAIRMAN DURST: Thank you very much.

I would also like to remind everyone and

also for our new people here at the meeting that

the "Charge" of the Committee is written out on the

meeting document. The most important thing is that

we are focusing on the "Threshold Working Group

Draft Report on Approaches to Establish Thresholds

for Major Food Allergens and for Gluten in Food."

We are not here to set any kind of

thresholds or discuss the labeling of these foods

for allergens, but strictly to make comments on the

best approaches to use for setting these

thresholds.

Did I cover everything that we need to?

(No verbal response.)

CHAIRMAN DURST: In that case, let's begin

with our first presentation. This is

Catherine Copp, the policy advisor for CFSAN, FDA,

on the use of gluten thresholds.

USE OF GLUTEN THRESHOLDS

MS. COPP: I have been asked this morning

to proceed with discussion on gluten and threshold

levels for gluten or possible thresholds for gluten

framework. It is similar to yesterday. I simply

want to provide you with some context for the

evaluating the Draft Report portion that addresses

gluten in food. This is the hazard of being first.

(Slide.)

MS. COPP: Yesterday, I mentioned the Food

Allergen Labeling and Consumer Protection Act.

This is a new law that Congress passed last August.

Although it focuses primarily on allergens, food

allergens, Congress also directed FDA to address

the separate problem of gluten in food.

When I say directed, I mean that Congress

has mandated that the Agency consider and then

establish regulations according to a schedule to

define "gluten-free" for use on food labels and

also to identify the appropriate use of the term.

As with Allergens, for consumers with

celiac disease, strict avoidance of gluten at

levels that will elicit an adverse effect is the

only means to prevent potentially serious

reactions.

Accurate, complete and informative

labeling on foods can help these consumers achieve

their goal. We believe that understanding

thresholds for gluten and having a sound scientific

framework for evaluating the existence of such

thresholds will help FDA develop a definition of

gluten-free and identify appropriate use of the

term. That's it.

Thank you.

CHAIRMAN DURST: Does anyone have any

comments or question on Catherine's presentation?

(No verbal response.)

CHAIRMAN DURST: Okay. We will move on

then to the presentation from Dr. Joseph Murray,

professor of medicine at the Mayo Clinic of

Rochester, Minnesota, on the introduction to celiac

disease.

INTRODUCTION TO CELIAC DISEASE

DR. MURRAY: Good morning, Committee

Members. I will be providing a general overview to

celiac disease.

(Slide.)

DR. MURRAY: First of all, we will discuss

what is celiac disease. We will discuss, briefly,

the pathogenesis of the disease; who gets it; what

the treatment is, at least in a very relatively

superficial fashion. We will discuss some of the

complications and compliance issues of celiac

disease and a prognosis or future of celiac

disease.

(Slide.)

DR. MURRAY: Obviously, yesterday was

focused on food allergies. "Celiac disease" is one

of the food intolerances that is immune-mediated,

though it is not thought to be IgE-mediated; so, it

comes into the non-IgE-mediated food intolerances

that are mediated by an immune response.

(Slide.)

DR. MURRAY: Where does it happen? It

happens within the smaller intestine, predominantly

the proximal, smaller intestine is the workhorse of

the digestive system. It is this surface of the

intestinal lining that is maximally expanded by the

development of circular folds and on top of these

circular folds the so-called "villi," these villi,

shown here in a histological picture, which

maximize the digestive surface area.

It is on the surface entrocytes of these

villi that most of the enzymes and in the layer

immediately above that in the lumen where most

digestion of the macromolecules from nutrition are

broken down and then absorbed. This is just a

picture. It looks like one of those shag-ply

carpets from the 1970s. This is a normal

appearance.

(Slide.)

DR. MURRAY: However, celiac disease is an

inflammatory state of the small intestinal mucosa.

It occurs in those who are genetically predisposed,

and it resolves, the damage resolves, with

exclusion of dietary gluten.

Here, on the left, is a normal intestine

with a normal villus structure; and on the right,

fully evolved celiac disease with complete

destruction.

The villi are gone, not only are they gone

but this entire intestinal mucosa is greatly

thickened and filled with inflammatory cells. This

is where the primary site of injury occurs in

celiac disease.

I didn't mention it, but it is a permanent

condition. While it will heal most of the time

with exclusion of gluten, the intolerance to gluten

is permanent and will recur when the individual is

reexposed to gluten.

(Slide.)

DR. MURRAY: Now, what causes celiac

disease?

(Slide.)

DR. MURRAY: We know there are two major

components to this disease: the first is the

genetic background of predisposition.

Much of that predisposition revolves in

the HLA type, which is part of our human leucocyte

antigen-recognition system. It is how we determine

self- and non-self and generate an immune response

as appropriate and its interaction with

environmental factors, primarily the environmental

factor of gluten.

These two conspire together to produce an

immune response that becomes out of control

resulting in inflammation, which we just showed to

you, that occurs primarily in the proximal small

intestine and then subsequently the consequences of

this inflammation leading to malabsorption and

symptoms.

(Slide.)

DR. MURRAY: What do we know about the

genetics of the disease? For many years, we know

there is a strong, familial predisposition to the

disease.

If you are unlucky enough to be a

monozygous twin of somebody affected with celiac

disease, your concordance rate is 80 percent. It

is not 100 percent, but it is about 80 percent. If

you are a sibling of a celiac, your chance of

having it is 10 percent. If you are a child of,

about 5 to 10 percent.

There is a very strong association with

certain HLA molecules. These are Class II MHC

molecules but particularly two types. First, DQ2

is the predominant type that is required for celiac

disease, and in some populations DQ2 is also an

enabling type.

These genes, however, while they are

essentially required for the disease, are not

sufficient alone to the development of the disease.

Probably 30 to 40 percent of the Caucasian

population carry one or both of these molecules,

but most of them don't get celiac disease. There

are other HLA genes that are likely involved,

though they have not been well elucidated and

certainly not confirmed in many populations.

There are other chromosomal disorders --

Down's syndrome, Turner's syndrome, and Williams

syndrome -- that are associated with a greatly

increased risk of developing celiac disease for

reasons that are not entirely clear, but probably

are associated with the increase risk of disease in

those chromosomal disorders.

(Slide.)

DR. MURRAY: Looking at the primary

environmental trigger for the disease -- that is,

gluten -- it is basically the storage proteins that

come from these particular cultivated grasses:

wheat, barley, rye, and other similar grains from

within those families. Other grasses -- for

example, rice, items such as corn, sorghum, millet,

and probably not even oats -- are not involved in

triggering the disease.

(Slide.)

DR. MURRAY: It is the storage proteins

from the endosperm compartment of the wheat kernel

particularly, and those are gliadins oar glutenins

that are thought to contain the antigenic moieties

that trigger the disease.

(Slide.)

DR. MURRAY: What is it about these wheat

proteins? Well, if you take wheat, as an example

of the others, these storage proteins are uniquely

high in certain amino acids, especially glutamines

and prolines.

Over 30 percent of the amino acids in

gliadin are glutamines. The glutamines, of course,

can be cross-linked to give the grain its

resiliency. Really the cooking ability, the

ability to use wheat as such an effective way of

making things that stick together like bread, for

example, and maintain their shape, is largely a

property of these particular combinations of amino

acids.

The proline sequences that contain or

proline residues contained within the wheat

proteins also appear to form helices, and these

helices are resistant to digestion within the

mammalian gut.

(Slide.)

DR. MURRAY: This may be a key factor in

what results in the likelihood of these peptides

basically being maintained and becoming, then,

still available for the immune system to recognize

in a patient with celiac disease.

Now, gliadin molecules are presented by

these HLA types to T-cells in the intestine, and

T-cells that are specifically primed to respond to

gluten. There are certain gliadin molecules that

have a higher affinity than others for these

Class II molecules and then the T-cell receptor.

These peptides may be processed or altered

within the gut, perhaps, to make them more

antigenic. They may not start out very antigenic,

but then they undergo some change within the gut

that may make them more antigenic.

It turns out that some of these peptides

that are particular immunodominant, these are the

ones that are most likely to produce an immune

response, that those immunodominant peptides may be

digestion resistant because they contain those

proline sequences that perform helix, making them

relatively poorly digestible by peptidases within

the gut.

(Slide.)

DR. MURRAY: Now, it turns out that there

is a contribution to this antigenic nature from

within the intestine itself, and this may well be

because of this enzyme tissue transglutaminase.

This is an enzyme that is present within

the gut mucosa. It is released by cells,

especially fibroblasts when they become inflamed,

and it cross-links cystine residues.

It turns out that it will also act on

gliadin by deamidating some of those glutamine

residues, some specific glutamine residues, to

glutamates, making it more antigenic by deamidating

that gliadin peptide and making it fit more

perfectly or with a tighter affinity into the

binding groove of the DQ2-HLA molecule, and, hence,

producing a more vigorous immune response within

the gut.

(Slide.)

DR. MURRAY: This is a schema, a

relatively simplified schema, of what I have just

talked about. We start with wheat. You look at

particularly the ethanol-soluble fraction, and

gliadin is probably broken up into smaller

peptides, but still of a sufficient size to produce

an immune response.

It is taken up across the epithelium

presented by antigen presenting cells to the

T-cells. These are T-cells that will specifically

respond to gluten then producing two types of

responses: a cellular response, characterized by

lymphocytes producing interferon gamma and possibly

other cytokines. It is probably the cellular

response that leads to the "inflammatory cascade"

that produces the damaged epithelium characteristic

of celiac disease.

It also produces help to the B-cell side,

to produce plasma cells that produce antibodies.

These antibodies are directed both against the

exogenous antigen gliadin and also antibodies

against tissue transglutaminase or what was known

as an the endomysial antibody. It is not known

what the actual pathogenic role of this is, but it

is a very useful serologic or blood marker for the

disease.

I mentioned about the antigen getting

changed by tissue transglutaminase. This is a

little cartoon which shows the peptide derived from

gluten. If you change one specific glutamine to a

glutamic acid, which could be done by tissue

transglutaminase, this then binds much more

tightly. This is the HLA molecule here on the

surface of the antigen presenting cell, and it fits

more perfectly into the T-cell receptor, producing

a more potent T-cell response.

(Slide.)

DR. MURRAY: Now, there are other things

that happen in the setting of celiac disease, and I

am really touching just on the surface of many of

these, but there are other things that result in

this inflammation that damage the lining of the

intestine.

For example, there are metallic proteases

that damage the structural elements that maintain

the structure that maintain villus structure.

There is endothelial injury that occurs affecting

the blood vessels in the villus. There are

antibodies, autoantibodies, that are produced that

affect actin that are involved in the site

maintaining the structure of the entrocyte itself.

(Slide.)

DR. MURRAY: Recently, there has been work

suggesting that there is a molecule called

"zonulin" that may be released in the setting of

celiac disease.

This is important because it opens up

tight junctions between entrocytes which may allow

even more ready access of the antigen, the foreign

antigens, between the cells into lamina propria

where antigen-presenting cells can then present

those peptides to the gluten-responsive T-cells,

further accelerating the disease.

(Slide.)

DR. MURRAY: Now, I pointed out that many

people, 30 percent or more of the Caucasian

population, carry DQ2 or DQ8. Virtually, the

entire population are exposed to gluten, but most

people don't get the disease.

There must be triggers that produce the

disease. There is evidence that suggests that

gluten in the infant diet, specifically the age of

introduction of gluten into the infant's diet, may

be important in triggering or at least producing

autoantibody markers suggestive of celiac disease

early in life.

It is not clear, however, if that changes,

whether you delay introduction or not whether that

changes, the lifetime risk of celiac disease, but

it certainly seems to be important in triggering or

producing evidence in childhood at least of celiac

disease immune markers.

The amount of gluten in the child's diet

may be important. There are other events such as

pregnancy, infection, or surgeries that may bring

previously asymptomatic celiac disease to clinical

presentation.

(Slide.)

DR. MURRAY: One could speculate, and I

think this is based on some data, putting data

together, that one's risk for celiac disease starts

with your HLA type. Only those who carry HLA types

are at risk. You, then, are exposed to gluten.

Perhaps the timing of exposure is important,

developing in some individuals a sensitivity to

gluten.

Then, with the interaction of other

factors such as other genes other than HLA, other

things that may predispose one to autoimmunity

including gender and other events that may occur --

gastroenteritis, aging, postsurgical or postpartum

changes in the immune system that may occur -- may

lead to a loss of tolerance, inflammation, and

subsequent malabsorption.

(Slide.

DR. MURRAY: Don Kasarda, who is here once

used the term or suggested that celiac disease was

a collision between our evolution of our immune

system and our ability to recognize self and

non-self through the HLA system and our cultivation

of wheat and these other grasses. This collision

occurs in the intestine.

(Slide.)

DR. MURRAY: Now, when this collision

occurs and results in damage, how does it present?

And who gets the disease?

(Slide.)

DR. MURRAY: Well, this is classic celiac

disease, and this is the way that I certainly

learned about celiac disease. A severe

malnourished child with evidence of malnutrition

often associated with the large, swollen abdomen

but great muscle, terrific muscle, wasting and

protein-calorie malnutrition with symptoms that

would occur sometime after the onset of gluten

introduction into the diet, sometime between the

age of six months and seven years of age: with

failure to thrive; abdominal distention; anorexia;

diarrhea; steatorrhea, that is the passage of

malabsorptive stools laden with fat; anemia; growth

failure; and vitamin deficiencies. That was really

the picture that we had of celiac disease 30 years

ago.

(Slide.)

DR. MURRAY: However, we now see celiac

disease in adulthood. In fact, celiac disease can

present at any age. Symptoms can include things

such as abdominal pain, even upper-GI symptoms:

heartburn, nausea, vomiting, anemia, fatigue.

There are of course patients who have

symptoms of malabsorption, though not necessarily

the classic, fully evolved malabsorptive picture.

Steatorrhea as a presenting symptom is relatively

rare, even patients may have constipation.

(Slide.)

DR. MURRAY: It can mimic other disorders

such as lactose intolerance. Indeed, lactose

intolerance may be secondary to the damage caused

by celiac disease. It may mimic the symptoms of

irritable bowel syndrome or symptoms of

inflammatory bowel disease.

(Slide.)

DR. MURRAY: There are specific

deficiencies that can occur in celiac disease,

especially the fat-soluble vitamins -- D, E, A,

and K -- with their resultant syndromes from

deficiencies.

Iron deficiency is especially common in

celiac disease because iron is absorbed in the

proximal small intestine; folate deficiency, again,

because it is absorbed in the proximal small

intestine; and, interestingly, B12 deficiency may

be relatively common in celiac disease by a variety

of mechanisms. Other trace elements -- zinc, B6,

selenium, and others -- may also be deficient in

celiac disease.

While in the past we would look for

combinations of these, often a patient would

present with many of these deficiencies at the time

of diagnosis, now it is relatively uncommon to see

the entire spectrum of deficiencies. Indeed, you

usually see one or two deficiencies that are

clinically evident, and the others may not even be

present.

(Slide.)

DR. MURRAY: How about non-intestinal? I

have mentioned that the major site of injury is in

the gut, but there are patients who will present

with non-intestinal presentations which can involve

into things such as the musculoskeletal system,

joint pains and osteoporosis or osteomalacia;

infertility or reproductive issues, delayed

puberty, spontaneous recurrent abortions have been

described; hematologic, which is predominantly

anemia; hyposplenism is an unusual consequence but

can present; and then dentition, enamel defects,

can be a presenting feature.

(Slide.)

DR. MURRAY: To focus on iron deficiency

anemia, which is probably one of the most common

reasons that I see celiac disease, about 5 to 8

percent of adults who present with unexplained iron

deficiency in some studies have celiac disease.

It is especially common in those who are

resistant to the use of oral iron. If you look at

individuals who are coming to gastroenterologists

for endoscopy, it may be 5 to 15 percent of those

patients, depending on the study, who may have

celiac disease if biopsies are taken from the small

intestine.

However, many of those patients are even

missed because the biopsies are still not taken

during routine endoscopy in patients who have got

anemia in about a third to a half of patients.

(Slide.)

DR. MURRAY: Osteomalacia or bone disease,

this is an example of severe disease with

pseudofractures in the pelvis of an individual,

whose only presentation was osteomalacia with no GI

symptoms, caused by celiac disease.

Other non-intestinal presentations include

neurologic or even neuropsychiatric syndromes such

as neuropathy, ataxia, seizures, cognitive decline,

or dementias; fibromyalgia-like syndromes or

chronic fatigue syndrome-like presentations;

individuals with skin and mucous membranes, there

is a specific rash associated with celiac disease,

a recurrent aphthous ulceration of the mouth; the

dental enamel defects we mentioned.

(Slide.)

DR. MURRAY: And, then, dermatitis

herpetiformis was specifically mentioned, because

this very blistering, extremely itchy skin rash

that affects the extensor surfaces is a direct

manifestation of intestinal gluten sensitivity.

(Slide.)

DR. MURRAY: Now, what about other

associated conditions? Celiac disease is

associated with other autoimmune conditions. It

may be seen in 3 to 7 percent of Type I diabetics,

individuals with thyroid disease, individuals with

inflammatory arthritis, primary biliary cirrhosis,

as examples of others; and then the congenital

disorders, especially those associated with

chromosomal abnormalities and also selective IgA

deficiency. If you look at relatives of celiacs,

it is anywhere between 5 to 20 percent, depending

on how one is related to someone with celiac

disease.

(Slide.)

DR. MURRAY: Now, beyond the symptomatic

celiac disease, there are also individuals who have

no symptoms, who already have fully evolved damage

within their intestine, and there may be no symptom

or there may not be occurring in someone with an

associated disease.

This is frequent to find this in first-

and second-degree relatives of patients with celiac

disease.

(Slide.)

DR. MURRAY: There is also what is termed

"latent" -- well, whether it is latent celiac

disease or latent gluten sensitivity -- individuals

who have a positive serologic response but have a

negative small-bowel biopsy. Some of those

patients will go on to develop the full-blown

disease, if followed, on a normal diet.

(Slide.)

DR. MURRAY: What about the epidemiology?

To summarize, while it was first identified in

Europe, it occurs essentially in all populations,

which could be termed Caucasian. Its prevalence is

probably somewhere between 1 in 90 to 1 in 300;

however, the diagnosis rate is much lower, which

would suggest a prevalence of about 1 in 2,000, if

you just look at the diagnosed cases.

It is one of the most frequent genetically

based diseases. If you look at other countries --

Latin America, or other areas; Africa, especially

North Africa; if you look at Asian countries --

there is celiac disease present in those. The

worldwide average prevalence is somewhere very

close to 1 percent.

(Slide.)

DR. MURRAY: I'm coming a little closer to

home. This is the data from Olmsted County, that

we published a couple of years ago, which looked at

the new case identification or the incidence rate.

The solid yellow line is the new cases per

100,000 in the population, which is essentially

quite low and stable over many decades until the

1990s and into 2000 to 2001, showing a greatly

increased rate of detection of celiac disease.

If we looked at who were being diagnosed,

this is the age of diagnosis by age category. This

is the incidence per 100,000 of people in that age

category in the community. You can see that the

new cases being diagnosed are predominantly people

between the age of 45 and 64.

The solid line indicates females; so,

females are diagnosed at a rate about twice that of

males of all ages.

There are a significant portion of

individuals, almost a third, who were diagnosed for

the first time over the age of 65.

(Slide.)

DR. MURRAY: One can term or consider

celiac disease like an iceberg. These are a series

of icebergs where the tip of the iceberg is the

part that has been detected, and the part

underwater is the part that can be found if one

screens the population. Of note, these are numbers

per thousand not percent.

Obviously, there are some places like

Ireland, for instance, which is quite a big iceberg

with a lot above water but also a lot below water.

Finland, which this iceberg is probably even more

out of the water as they have got a very active

program for finding celiac disease.

This is circa 1996. The U.S.A. iceberg is

still close to a very low level of actually being

diagnosed; but this number, the part underwater,

has actually grown to be something very close to

what you would find in Finland or Ireland,

especially when one looks at, at least what we know

about is really from a Caucasian population.

(Slide.)

DR. MURRAY: One miniature study, this is

one we simply looked at, Natrona County in Wyoming.

This is a very isolated community. Anybody who has

been to Wyoming, there is lots of nothing for miles

and miles.

We were able to study about 4,000

individuals from a health fair, generally healthy,

and found the numbers above just under 1 percent of

people who had serologic evidence for celiac

disease.

Only half of them had GI symptoms. Most

of them did not have other risk factors for celiac

disease, just two having family members with celiac

disease.

However, these are numbers that would

confirm basically the rest of the world's data that

suggests that the prevalence of celiac disease, if

you look for it, is probably slightly under

1 percent.

(Slide.)

DR. MURRAY: Now, how does one make the

diagnosis? It starts with suspicion. Serologic

tests may be very effective at detecting the

disease. The intestinal biopsies are regarded as

the gold standard. Then, one ultimately gets a

response to a gluten-free diet to confirm the

diagnosis.

(Slide.)

DR. MURRAY: The pathology, as we have

mentioned already, is a pathology of chronic

inflammation within the intestine with features

such as intraepithelial lymphocytes on the surface,

villus atrophy or the loss of the villus surface,

great crypt hyperplasia, and then characterized by

being a lamina propria filled with lymphocytes,

macrophages, plasma cells, and even eosinophils.

(Slide.)

DR. MURRAY: There is however a spectrum

of damage that occurs, typified here by the Marsh

classification. This is classic disease, but there

are also milder forms of the disease that may be

asymptomatic in most individuals.

(Slide.)

DR. MURRAY: Our algorithm for finding

celiac disease, if we have a high clinical

suspicion, an individual with malabsorption, we

biopsy those individuals. If we have an individual

who is at moderately increased risk, serology is

probably the most effective way of finding it.

Though, we yet do not depend on the serology alone

to detect this, there are other circumstances for

alternate; serologic testing may be necessary.

(Slide.)

DR. MURRAY: Now, what about treatment for

celiac disease, or, as one patient with celiac

disease called it, "playing gluten-free roulette"?

(Slide.)

DR. MURRAY: The nuts and bolts of a

gluten-free diet, basically one needs to avoid

foods that contain the offending grains: wheat,

barley, rye, and the wheatlike grains of spelt and

kamut.

I put down at the bottom that many corn or

rice commercial cereals do not appear to be

gluten-free because of their incorporation of

particularly barley extract in their flavor

ingredients.

(Slide.)

DR. MURRAY: Now, one of the issues and of

course the issue for today is, How much gluten is

too much? We will be hearing a lot more when you

hear prospective and retrospective data later this

morning on this. I am not going to dwell on it.

One thing from clinical appreciation is

that symptoms are not a good indicator of gluten

ingestion. Many patients can have significant

damage to their intestine, despite the absence of

symptoms when they ingest gluten.

Most patients diagnosed clinically with

celiac disease have never suspected that wheat or

gluten products are what are precipitating their

symptoms. They may have or are often likely to

have blamed other foods that they weren't able to

digest because of the damage.

Antibodies such as tissue transglutaminase

antibodies are really only positive of the

substantial gluten contaminating the diet. At

least in my practice if somebody admits to cheating

more than once a month they will like continue to

have injury in their gut. However, there is a high

degree of variability in the sensitivity to gluten

ingestion, at least clinically.

(Slide.)

DR. MURRAY: We will hear a little bit

about Codex Alimentarius draft standards. This,

however, is still a draft, I think a Stage IV

draft. I want to put this up really to demonstrate

that there is a variance between countries and what

one allows.

We will hear more from our colleagues from

Europe about some of the incorporation of rendered

gluten-free foods which use the gluten-containing

grains as a base and then remove the proteins from

them.

(Slide.)

DR. MURRAY: What about non-responsive

celiac disease? This is relatively common. By

far, the most common reason is inadvertent gluten

contamination of the diet and lymphocytic colitis,

pancreatic insufficiency, bacterial overgrowth, and

then only a few patients have true refractory

disease that no longer responds to exclusion of

gluten from the diet.

(Slide.)

DR. MURRAY: There are many potential

sources of contamination of the diet with gluten,

which include of course commercial cereals, eating

out, communion wafers, lipstick, airborne flour or

starch in certain work situations, so-called "soy"

sauces made from wheat, mislabeled or unlisted

ingredients have been an issue, and at least

allegedly some medications.

(Slide.)

DR. MURRAY: Some ingredients that people

are concerned about: seasonings and spice blends,

modified food starch, malt and malt extract,

modified hop extract or yeast-malts, sprout

extract, dextrins, caramel color. There are a

whole bunch of things that might be derived from

gluten-containing grains.

(Slide.)

DR. MURRAY: What about complications

associated with untreated celiac disease? We know

the mortality of symptomatic celiacs who do not

comply with the diet has doubled. The mortality of

celiac disease even when it is diagnosed is also

double, even following out for several years after

the diagnosis.

The predominant excess in death comes from

GI malignancies. There are also morbidity

consequences such osteoporosis or osteomalacia,

stunted growth, infertility, chronic ill-health --

all of which could be prevented by early detection

and treatment.

(Slide.)

DR. MURRAY: What are the dangers of

non-compliance? The increased mortality we

discussed, the osteoporosis. Children who were

diagnosed and then did not remain on a gluten-free

often have osteoporosis or diminished bone density

when they get to adulthood, lymphoma, other

cancers, and then the psychological effects of

non-compliance.

(Slide.)

DR. MURRAY: One of the most feared

complications of this are the T-cell lymphoma.

This is celiac disease on this side, and the T-cell

lymphoma on the other side is one of the more

feared complications with a very high mortality.

(Slide.)

DR. MURRAY: However, as I pointed out,

most celiac disease is probably not symptomatic, at

least when we look at a cross-section of the

population. We do not know whether those

identified by screening are less sick than

clinically diagnosed celiac disease.

We don't know the benefit or negative

effects of a gluten-free diet in those who are

found by screening alone. We don't know if they

are any more or less likely to comply with a

gluten-free diet. We don't know whether

intervening in those patients will actually affect

their ultimate mortality.

(Slide.)

DR. MURRAY: George Dennison Prentice

said, "What come call health, if purchased by

perpetual anxiety about diet, isn't much better

than tedious disease."

(Slide.)

DR. MURRAY: This comes to the future.

There is, I think, a promising future in celiac

disease, a variety of approaches, which I have

listed, that individuals and research groups are

looking at as alternates to the gluten-free diet,

though none of them are really even close to

clinical use.

(Slide.)

DR. MURRAY: This is one that has been

tested in patients using a lactobacillus digestion

strategy, trying to reduce the potential, harmful

effects of gluten.

In summary, I would like to suggest that

gluten or celiac disease is common. It has been

largely unrecognized until recently. There are

many challenges that face patients and their

physicians in the treatment.

The gluten-free diet is not simple. There

is widespread use of grain proteins in good, and

that makes it challenging for individuals with

celiac disease. Food ingredient source

identification is of great concern to patients.

Dietitians and those who counsel patients

with celiac disease, we are here because of

regulation or potential regulations. Defining

acceptable thresholds and verification of those may

be very important to patients with celiac disease.

(Slide.)

DR. MURRAY: I finish with an aside on

another food safety issue, an invitation to come to

Minnesota and enjoy Joe's and have worms at the

same time.

Thank you.

CHAIRMAN DURST: Thank you very much,

Dr. Murray.

QUESTION-AND-ANSWER SESSION

CHAIRMAN DURST: I would like to ask the

Committee if they have any questions or discussion,

and also to point out that Dr. Murray will not be

able to stay around for discussion this afternoon.

If there are questions, now is the time to ask

them.

Doug.

DR. HEIMBURGER: Thank you very much for

that presentation. It is very helpful. I do care

for some patients with celiac disease. As you

already mentioned, one of the big questions that

they have is, "What really are my risks if I play

with it a little bit? If I knowingly introduce a

little bit of gluten in my diet, how absolutely

obsessive do I need to be about it?"

With the mortality being primarily

associated with the risk of lymphoma, what is the

quality of the evidence that compliance with gluten

restriction is really correlated with the risk of

lymphoma?

DR. MURRAY: The evidence, the quality of

the evidence is largely observational, taking

cohorts and following them, and often their

self-assessment of their level of compliance. The

data is based on those cohorts. They are

predominantly referral cohorts.

The longest follow up comes from Britain.

In those, there is quite a clear increased risk not

just for lymphoma but other malignancies in those

who are considered not to be completely compliant

with the diet. I would say the level of evidence

while it is not a prospective study, in retrospect

in fact it appears to be reasonably strong.

DR. BRILEY: Margaret Briley. I would

like to know what you can tell us what might be the

reason for the increased detection of celiac

disease? Is the serological test the definitive

reason? What is going on with those? It seems to

be more prevalent, as you said, in our society.

DR. MURRAY: I think there are probably

two reasons, one is probably serologic detection

has made this an acceptable diagnosis to primary

care physicians, number one; and the second is

suspicion, that is, the awareness of celiac disease

as a possibility.

I do not believe, however, we can rule out

the possibility that celiac disease is actually

increasing in prevalence over time.

We don't know, but we know that there

certainly were substantial increases in other

inflammatory bowel diseases over the last 50 years,

which may not have been accounted for by detection

rates, for example.

I do agree with you it is probably the

sero-detection. It is a combination of those two

things, suspicion and ease of detection have made

the biggest difference. However, I would not rule

out the possibility of its actually increasing in

prevalence.

CHAIRMAN DURST: Soheila.

DR. MALEKI: Soheila Maleki. I was

wondering how long does a person that is

asymptomatic go before they find they are

symptomatic? How do they usually find out?

DR. MURRAY: First of all, we don't know

when it starts. There is one study I didn't talk

about from Denver that suggested that children, if

you follow children at genetic risk for celiac

disease based on HLA type, that they will convert,

sero convert, by about the age of seven, though

most of them are asymptomatic.

It reaches all but 1 percent of that

childhood cohort, the same prevalence you find if

you look at adults. The suggestion is that you

have developed that celiac autoimmunity probably by

the age of seven.

However, the age, the median age, of

diagnosis is 45 years of age. It is likely that

those individuals have some clinical disease for a

long time before they present.

In our population, the most common reason

for presentation is postmenopausal anemia. Women

who have anemia sometimes go a long time,

especially when they are menstruating. The doctor

says that menstruating is an excuse for the anemia.

Now that they are menopausal and are no longer

menstruating they no longer have an excuse for

their anemia, and they are referred for evaluation

for the cause of anemia.

Iron deficiency anemia at least is one of

the most common, but there are other things. For

example, the development of chronic GI symptoms.

Largely, as the earlier questioner said, probably

awareness of the possibility of this disease, with

primary care doctors using sero-diagnosis to find

celiac disease.

There a lot of atypical symptoms. There

are reservoirs of celiac disease, the Type I

diabetics, the family members of celiacs, those

with chromosomal disorders. As the doctors looking

after those people become more aware of celiac

disease, they are testing their patients more.

DR. BRILEY: Could you speak a little bit more about

the evidence that you have in regard to physicians

using the serological test? Is it a pretty common

thing to do it at the very beginning of a patient's

coming in with GI problems, or is it more likely it

is down the road a while?

DR. MURRAY: If you look at the data on

patients’ presentation time, from presentation to

diagnosis of celiac disease, it is somewhere

between 8 and 11 years from the time they present

to their physician with a complaint to the time

their diagnosis is made.

I think that time period is now beginning

to shorten, thankfully. The serology is done close

to the end of that period, so it is often only when

the suspicion is generated. Partly that is because

it is not considered at the early differential

diagnosis of celiac disease.

It is also difficult. It is difficult

because the symptoms are not very specific.

Fatigue and some bowel disturbance and a little

anemia, that is not a rare syndrome or not a rare

collection of symptoms you will see in individuals

in our community. It is hard to know where you

look for celiac disease and when you look for it.

It is often a very delayed diagnosis.

DR. BRILEY: Is the serological test one

that is pretty accurate? Could one count on it if

you did it early?

DR. MURRAY: If you use the more modern

autoimmune tissue transglutaminase test, for

example, it is reasonably good. It is not perfect,

but it is quite efficient at detecting celiac

disease at that level, in that period. It is not

perfect, though, but it is a pretty good test.

DR. BRILEY: It sounds like we need to do

some education, then.

DR. HEIMBURGER: A follow up to that --

Doug Heimburger -- what is currently understood to

be the sensitivity specificity of anti-tissue

transglutaminase antibodies?

DR. MURRAY: Depending on the study, most

studies suggest that the sensitivity of tissue

transglutaminase for celiac disease is in the high

90 percent, so the high 90s.

The specificity is probably also in the

mid- to high 90s, so quite effective when you are

looking at a population you are suspicious of the

disease.

DR. BRILEY: One more question.

Margaret Briley. Do we have any data that shows

the celiac patient also maybe has been identified

with a lactose intolerance? Is there anything that

combines those two that you would know or does

that--?

DR. MURRAY: The combination is probably

the damage that is caused by celiac disease, the

damage to the entrocytes affects your disaccaridase

including lactase throughout the surface of the

small bowel, and, hence, you get a secondary

lactose intolerance.

In general, if you think of the genetic

basis or the ethnic groups that are involved,

celiac disease tends to hit those of European

extraction more so. I know that population will

tend to retain their lactase activity for longer.

If you look at Subsaharan Africans, for

example, we don't know actually what the prevalence

of celiac disease is in Subsaharan Africans, but

they are individuals who tend to lose their

lactase.

Genetically, they are probably separate, but they

come together because of secondary lactase

deficiency caused by the damage in celiac disease.

DR. NELSON: Thank you, Dr. Murray. That

really was a very interesting presentation. I

think the last time I studied this was 30 years

ago.

CHAIRMAN DURST: Identify yourself,

please.

DR. NELSON: Mark Nelson. I wanted to

touch base on the couple of slides where you talked

about how much gluten is too much and the proposed

Codex definition for gluten-free foods.

You had a question mark after the

threshold for damage being 20 to 100 milligrams per

day, and I guess that is similar to the naturally

gluten-free foods the Codex proposed.

Then, it goes on to talk about cheating is

greater than once per month. Also, is there an

issue of cumulative exposure, 20 milligrams per

day, or is the cheating an excursion of whole wheat

crackers, for example?

DR. MURRAY: Well, first of all, the

comment on cheating that is what the patients tell

me, those patients who will admit to gross

cheating, that is, eating a piece of bread or a

cookie or cake, which is what I regard as an

obvious source of gluten.

However, that may reflect some other

background, but a lack of detail of care, for

example, attention to detail in the rest of their

diet, maybe what they are not telling us. That

only applies to what they have told me.

The issue of actual threshold, I have a

question mark after that because we will hear a lot

more detail, science, about threshold testing for

thresholds of gluten contamination.

The intermittent contamination, once a

month obvious contamination. Something that is, to

some degree, under the patient's control with

appropriate education and exposure to information

then essentially it is under their control.

It is the low-level contamination on a

regular basis from sources they are not aware of,

and those are the patients that I see that make up

the majority of the patients I see who have

difficulty. They are coming to me because they are

trying to be gluten-free, but they are having

contamination of their diet on a daily basis,

probably a relatively small amount. We will hear a

lot more about the threshold, the actual testing of

the threshold using that type of low-level daily

contamination.

DR. MALEKI: Soheila Maleki. Is there an

adult onset or spontaneous development of celiac

disease, or does this have to come with a genetic

component and hereditary?

DR. MURRAY: It probably only occurs in

people who have that HLA or genetic type. You have

to have it. However, that is 30-plus percent of

the Caucasian population.

Can it start first in adulthood? We have

very little data on that. The only data is the

stuff I have mentioned, looking at children and

showing about 1 percent by the age of seven or

eight, positive by the age of seven or eight, the

same prevalence if you look at adulthood.

There are a few cases of what we call

"latent celiac disease," individuals who have got

antibodies with apparently normal small intestinal

biopsies as adults, then go on over a space of

years to develop full-blown celiac disease within

years of that initial identification of a positive

serology. Those are very rare cases that have been

found.

Of course, if you find somebody who

suspects they have got a problem, they ought to

change their diet anyway, and that changes

everything. There is very little data, I think, to

be sure of whether it first occurs, starts, in

adulthood.

If you look at the age of diagnosis, your

eighth or ninth decade can be the first time that

you are diagnosed with celiac disease. A lot of

those patients have suspicious symptoms going back

many years.

DR. BRITTAIN: Erica Brittain. Can you

quantify the level of damage? Is that only

possible with the biopsy, and is that very

invasive?

DR. MURRAY: You can -- well, if you take

biopsies, you can quantify the degree of injury. I

showed that slide which shows a spectrum of injury.

It tends to be variable, even within the same

individual. It started in the first part of the

small intestine and extends a variable distance

down the small intestine.

There has really only been one study,

which was done in the early 1960s, of taking

multiple biopsies down the intestine -- a handful

of courageous volunteers.

It is not clear that there is a

correlation between the extent of injury and the

severity of symptoms. In fact, we don't know how

to predict the occurrence of symptoms in patients

with celiac disease, so that is a "black box."

DR. BRITTAIN: You are saying there is

really no simple way to quantify the extent of

damage? There is nothing that would correlate with

these multiple biopsies?

DR. MURRAY: Some people have suggested

that the level of the antibodies, the tighter the

antibodies might correlate: The higher, the

tighter, maybe the more severe the injury to the

intestinal biopsies.

When we take a biopsy of the intestine, we

are sampling a tiny fraction of a percentage of

1 percent, a fraction of 1 percent, of the

intestinal lining. We have tried to look using

other imaging techniques to assess the extent of

injury. Using some of those techniques seems to be

very variable between individuals, and it doesn't

seem to predict their symptoms.

Really right now it is a yes or no issue:

yes, they have celiac disease; or, no, they do not.

It is very hard to measure the severity of the

disease.

We can look at the severity of

consequences: have they developed osteoporosis,

what their bone density is, have they lost a lot of

weight, whether they have severe malabsorption

based on fat malabsorption in their stool. We have

got other measures to look at the consequences or

impact of the disease, those we can assess.

DR. BRITTAIN: Do you think that it would

correlate with the intestinal damage? Not

necessarily?

DR. MURRAY: We have tried. I would say

that the data is not very good on that. People

with very mild injury may be more likely to be

asymptomatic, but the data is not sound.

CHAIRMAN DURST: Suzanne.

DR. MALEKI: Suzanne Teuber. I had a

question about the neurologic presentations. I

have read some on screening of pediatric

populations, but with adults how often does it

present as dementia without it being diagnosed as a

GI problem? Is this something we should be adding

to dementia screening?

DR. MURRAY: It is probably relatively

rare. In fact, I think there has only been one

good study. Maybe Dr. Collin, who is here as a

speaker later, has done a study and has looked at

and reported on that.

We occasionally see cognitive decline at

the time of diagnosis, but there is really no good

epidemiologic study to address that. Some of the

other presentations, peripheral neuropathy, for

example, or ataxia, there is more data on it. Many

neurologists are beginning to include celiac

disease as part of their differential diagnosis for

those syndromes. Good epidemiology data is

relatively small, very little data.

CHAIRMAN DURST: Ciaran.

DR. KELLY: Ciaran Kelly. This is

actually more clarification than a question. Dr.

Murray is quite correct that there isn't a measure

of either severity of intestine abnormality or even

height of antibody levels that reliably reflects

the degree of injury or correlates closely with

symptomatology.

However, with treatment one can use a

decline in antibody levels as a crude indicator of

at least reduced exposure to gross amounts of

gluten. It is not a very sensitive indicator, but

it is useful. Of course, with repeat biopsy, if

the histology has revered to normal, that of course

can be used. However, the less invasive test of

following antibody levels is used clinically to

follow response.

DR. MURRAY: Quite right. I think you

will agree, Joe.

DR. KELLY: If the antibody levels aren't

dropping, that is used as an indication that the

patient is successfully on a gluten-free diet.

DR. BRILEY: Margaret Briley. Can you

give us any idea of any behavior data that you may

have received from your patients regarding their

willingness to try foods that are not gluten-free

labeled?

DR. MURRAY: Oh, well, there are many

different attitudes among patients with regard to

what they want to eat or what they are afraid to

eat.

I advise my patients to be prudent, that

they try to select things based on identification

of ingredients, source ingredients, not containing

things contained from gluten, the use of substitute

grains that are gluten-free.

Many patients are quite willing to do that

on their own. Many of them use support group

information where maybe a group has cooperatively

contacted manufacturers who in good faith provide

information on their source ingredients.

There are some patients who are entirely

paranoid about it, and want to obtain a kit to test

the food. I don't know that we've got a very

effective kit yet for testing food for gluten

contamination. There are many different attitudes.

Fear is a major concern among my patients.

I mean, fear of even the slightest potential, not

even actual but potential contamination. This can

verge on, "Do we avoid taking prescription

medications for things like hypertension?"

resulting in life-threatening changes to their

medication regimens because of fear of

contamination.

I would say fear is a major part or a

major influence on the quality of life. We will

hear more I think shortly on the impact of a

gluten-free diet on patients' lifestyles a little

later. Certainly that does affect a substantial

portion of my patients.

Patients go through a substantial grief

reaction and feel socially isolated because of

their difficulty of interacting with society,

because so much of our society activities or social

activities revolve around food. There is that

safety sense of insecurity, which I think pervades

or affects many patients with celiac disease.

DR. BRILEY: Thank you.

CHAIRMAN DURST: Soheila.

DR. MALEKI: Well, I just want to know if

there are any coordinated studies for a

determination of thresholds? I know you mentioned

the level of PPMs. Do you know of any studies?

DR. MURRAY: There are and you will hear

about them. There are both retrospective and

prospective studies, and you will be hearing some

data on those later this morning.

CHAIRMAN DURST: I have one question --

Dick Durst -- on the biopsy and histological

studies to see the morphology, morphological

changes, you showed from the shag rug. I am just

curious whether just one of these cameras you can

swallow would be able to detect those kind of

changes without having to go through a biopsy?

DR. MURRAY: Yes, you can detect them.

Nobody would suggest that it would replace the need

for biopsies to make the diagnosis. There is

really relatively little published data on it.

There has been a paper suggesting that you can see

those changes. With a magnified view, you can see

with a capsule.

Yes, I think you can identify those

changes in a lot of individuals with celiac

disease, maybe all of them. Although, I really

can't comment more on that, because it hasn't

really been studied in any great detail.

CHAIRMAN DURST: Erica.

DR. BRITTAIN: Erica Brittain. Do you

have any insight about the cumulative effect of

decades of low levels, very, very low levels of

gluten exposure? Certainly, we are going to have

to think about what chronic exposure could do when

we talk about the thresholds. Can you provide any

insight into that?

DR. MURRAY: Probably the best clinical

insight I can give are individuals who I see who

were diagnosed 20 years ago and have not come back

to medical attention in 20 years. I see those

patients maybe every week.

I would see somebody who is diagnosed 20

years ago, and they got instruction at that time

that allowed them to eat things like barley malt or

that people weren't really instructed about some of

those rice or corn cereals that may have contained

malt, for example.

Those patients come back with anemia,

chronic GI complaints, maybe not as severe as they

had initially, but they certainly have accumulated

some health morbidity over those 20 years. Some of

them will come back with frank lymphomas and will

end up with a mortal complication of their celiac

disease.

Yes, at least my clinical observation is

that I frequently see individuals with problems

that we get rid of, once they now move to a much

more strict gluten-free diet, by eliminating those

things -- largely, because in 20 years they didn't

go back and get more education and realize that you

had to exclude those minor ingredients. That is

one way of looking at the effects of decades'

accumulation of low-level contamination.

CHAIRMAN DURST: Dick Durst again. On

your slide that showed the various causes, the

different grains, and so on, I believe you

indicated oats was not one of the causes. Could

you expand on that?

DR. MURRAY: We will hear a little more, I

think, from Dr. Collin on that issue. While oats

had been thought to be one of the offending grains

in things done in the fifties and sixties, it turns

out from recent very well-done studies that it

doesn't appear to impair the healing of the

intestine in newly diagnosed celiac disease. It

doesn't seem to result in a significant worsening

of production of damage in patients who are already

diagnosed with celiac disease.

For the vast majority of celiacs, it is

probably safe in its native, pure form. However,

there are some sequences within oats that can

produce an immune response, at least in vitro, in

lymphocytes derived from a few celiacs.

It is not an absolute. There may be some

individuals with celiac disease that can respond to

oats, both in the laboratory test and possibly also

clinically there are a few.

There are probably a relatively small

minority of celiacs in which that occurs. A bigger

concern is the issue of contamination of oats with

other grains that are well recognized to cause

injury.

CHAIRMAN DURST: Marc.

DR. SILVERSTEIN: Marc Silverstein. I

would like to inquire about the potential

subsequent lifelong increased risk of GI cancers.

I presume that the risk is predominantly small

bowel, but I wonder if there is any increased risk

of colorectal cancer?

Then, what are your thoughts about whether

there is sufficient risk that patients with celiac

disease should be in some sort of surveillance

program for early detection of GI cancer?

DR. MURRAY: Clarifying the risk of

cancers, it is particularly visceral cancer but

also includes: esophageal cancer, non-Hodgkin's

lymphoma of any site not just the intestine, and

probably also B-cell lymphomas, as well as the

T-cell lymphomas, small-bowel carcinoma.

There is a greatly increased relative risk

of small-bowel carcinoma. Of course small-bowel

carcinomas is a very rare disease to begin with, so

the lifetime risk of dying of a small-bowel cancer,

even in a celiacs, is still relatively small. The

data on colon cancer is mixed. There is some that

suggests there is an association; and some, that

does not.

When you look at other causes of

mortality, even non-cancer causes of mortality such

as infections, neurologic disorders and chronic

lung infections, there are other excesses of

mortality that occur in patients with celiac

disease.

There are some reductions in cancer

mortality. It appears, at least there is a

suggestion, that breast cancer may be less common

in celiac disease than non-celiac disease. There

were a couple of suggestions that lung cancer might

be less common in celiac disease than in non-celiac

disease.

Now, whether there is some competing issue

like smoking may be less common in celiac disease

than non-celiac disease, so there may be some other

competing issues that are involved. Body size may

make a difference. It may be another confounding

issue that confounds or is a competing risk for

malignancies.

While small-bowel cancers and lymphomas

are the two that have the greatest relative risk,

it is a small, absolute risk because of the

relative rarity of those cancers.

CHAIRMAN DURST: Do we have any further

questions for Dr. Murray?

(No verbal response.)

CHAIRMAN DURST: Thank you very much.

Our next speaker is Cynthia Kupper, who is

the executive director of the Gluten Intolerance

Group of North America who will present on patient

perspectives on celiac disease.

PATIENT PERSPECTIVES ON CELIAC DISEASE

MS. KUPPER: Good morning. I am a

dietician, not a doctor. I appreciate the honorary

doctor status.

My job here today is to give a face for a

person with celiac disease. I have been tasked

with letting you know who they are, letting you

know how they get their information and education,

and then also providing you with some information

about labeling that they have.

(Slide.)

MS. KUPPER: First of all, living with

celiac disease is very difficult. It is a chronic,

lifelong condition, as you have heard, and people

find this to impact greatly their quality of life.

Forty-four percent of the patients in a

Canadian study say that the diet is very difficult

to follow. In fact, there are some studies that

suggest that the compliance with the diet can be as

low as 50 percent in teenagers but probably ranges

around 70 percent compliance.

Eighty-four percent of these patients in

Canada suggested they have a difficult time

determining what is gluten-free and what is not.

They don't travel, and they don't eat out. It

impacts their family life and their career.

If you have celiac disease and are an

X-ray technician, oftentimes you change careers

because sometimes the X-ray slides are dusted with

flour. Chronic exposure could impact your quality

of life.

(Slide.)

MS. KUPPER: I did a study online of 620

patients a few months ago. In response to that

study, 75 percent of them said that they can tell

the difference between a gluten reaction and other

intolerance or a food allergy.

When they discussed their reaction

symptoms, they ranged from anaphylaxis, which is

not a gluten reaction for celiac disease, to

delayed reactions which could impact any aspect of

their GI and other health systems, body systems.

The average time to reaction was somewhere

between four to eight hours, but some of them

complained of immediate, almost allergic type

responses, and many of them said that their

responses or the symptoms that they had would last

for several days.

Keep in mind, there is no medication we

can give them to make this go away, so they just

have to let it work its course. This is really

disturbing to me.

(Slide.)

MS. KUPPER: As a dietitian, patients do

not rely on medical communities and professionals

for their information. They rely primarily on

support groups.

Actually, the Internet should probably be

the first one, because they are Internet savvy.

They have been out there and they have gotten all

kinds of information before they ever see a

dietitian. Not only do they get information from

the Internet and the support groups, but there are

list serves and chat groups that they belong to.

These can be very useful tools for a

person with celiac disease. However, they also

provide some very frightening and unreliable

information that the patients will hold onto as if

it were gospel. Then, they work with self-help

books as well.

Unfortunately, doctors, and especially

dietitians, are seen as unreliable. It is sad for

me to say that as a dietitian my profession doesn't

get this disease. They also treat it like it is a

rare thing, and they don't know anything about it.

In the United States, I can tell you that there is

probably a handful of dietitians who would be

considered experts in celiac disease.

Doctors don't get much more respect,

primarily because it has taken so long for the

patient to get a diagnosis that the patient has

lost faith.

Then, they will go to research facilities

like the University of Maryland, Chicago, and

New York, or the Mayo Clinic. Lastly, they will go

to medical Web sites. The bulk of our information

is coming from potentially unreliable and

non-research-based sites.

(Slide.)

MS. KUPPER: The consumers perceive that

gluten exposure levels -- the question was asked to

me, "What do consumers believe about gluten

exposure? Are they concerned about the health

risks?" The answer is yes and no.

On the study that I did, it depended upon

their confidence of the labeling, and it depended

on whether they accepted testimony or accepted

research. There is a group of celiacs, as

Dr. Murray suggested, who really don't want to

listen to what research suggests.

As you move forward, with not only

establishing how you are going to determine the

threshold but what that threshold will be, you will

have a fight in the celiac community for a lack of

education and understanding of research.

Consumers oftentimes also have an

inability to correctly interpret research findings.

These are people who have just enough medical

knowledge to be dangerous, so they don't have a

full understanding of the terminology they are

talking about.

There is this constant perpetuation of

misinformation. I don't know how many times when

we try to bury something that is inaccurate it gets

dug up.

(Slide.)

MS. KUPPER: There are varying levels of

gluten sensitivity, as you heard, too. There is

the perception that gluten is poison. Not unlike

the allergy people that we heard from yesterday,

this is a huge issue to the celiac consumer.

They believe most of the time that when

their gut hurts it is from gluten not from

something else. Consequently, we are trying to

help the patient understand that not everything

that makes their gut hurt is gluten.

As Dr. Murray said, there is a huge fear

reaction. If I had to put a psychological label to

a group or at least a portion of the celiac

community, they are filled with fear and a little

bit paranoid about what they can and can't do.

How do we define "gluten-free" in the

U.S.? This is a really interesting question. Of

the consumers, only 19 percent realize that there

is no definition right now for gluten-free in the

U.S. Many of them define that the true definition

is zero. This is a problem -- a lot don't know.

(Slide.)

MS. KUPPER: When I ask the question, "Do

you trust gluten-free labeling?" It was

interesting, too, because most of the people say

they do trust it. However, when you ask them if

they ever had a reaction to a product labeled

gluten-free, you can see that up to 50 percent

suggested that they might have had a reaction to a

gluten-free product.

When I talk to manufacturers that

manufacture only gluten-free products and ask them,

"Do you test, and what do you test to," many of

them are using older testing methods not the newer

testing methods, the monoclonal tests that we

talked about yesterday.

Some of them tested 200 parts per million,

some of them tested 20 parts per million, some of

them tested no detectible. For the gluten-free

consumer today, the label "gluten-free" really has

no meaning.

(Slide.)

MS. KUPPER: Again, the gluten-free

consumer is compulsive about their medical needs.

This is their only treatment. It is often referred

to as our drug of choice. There is nothing else we

can do, except to follow a strict gluten-free diet.

They have very limited trust in the

manufacturing industry. They believe that labels

that say "may contain" and different things like

that need to be distrusted. When they call the

manufacturers, they are not quite sure that they

are getting the right answer all the time.

Also, they have a limited understanding of

what good manufacturing practices really mean, so

they are always questioning what the manufacturer

will say. Yet, at the same time they want

accountability and they want reliability.

They may translate information to the

extreme. Let me give you an example. A few months

ago on one of the list serves, someone put out a

message about bottled water being gluten-free.

That got taken in a week's time to the

point where consumers were calling asking why water

had gluten in it, and how dare the food industry do

that to them. The reality is it never did.

A company, out of the graciousness of

their heart, put it on a list of gluten-free

products, and from that the consumer decided that

every other bottled water had gluten in it. This

is the extreme that the consumer can go to. Again,

they don't find descriptive labeling helpful at

all.

The changes that can occur in ingredients

in manufacturing processes make it difficult for

this consumer group to know what they can have.

The term "modified food starch" usually means

cornstarch, modified cornstarch, in this country.

However, if the manufacturer determines

that wheat starch is cheaper in the fall and they

switch and the consumer has determined that this

product is gluten-free, now they are in trouble if

they don't recheck.

When you talk to the food industry, you

will find that their calls have dramatically

increased over the last 2 to 5 years of consumers

calling in, and 90 percent of the questions do not

have to do with other allergens but have to do with

gluten.

(Slide.)

MS. KUPPER: When the consumer asks the

question about gluten, the problem is that they are

asking the wrong question. The consumers believe

that if they don't have effective labeling how can

anybody possibly know that they are going to be

able to be healthy and protect themselves.

They want to know that if you call a

company they are really giving you the right

answer, and they are just never confident about

that. Oftentimes, when the company answers too

quickly, they get suspicious.

Oh, I've had that experience. I will call

on a product and I'll say, "I need to know the

source of the modified food starch."

"Oh, you're talking about gluten?"

"Yes. Tell me the source of your modified

food starch. Let me make that decision about

whether I'm talking about gluten." That makes a

consumer suspicious.

Finally, you know, if a person eats a

gluten-free food and they get sick, whether it is

related to gluten ingestion or not, they have

determined that they can't trust that company any

longer.

Again, these list serves and chat groups,

I have seen them take small companies out of

business because of the spreading of rumors --

which are probably unfounded.

(Slide.)

MS. KUPPER: In closing thoughts, I really

encourage that through this entire process related

to labeling thresholds, that we be talking a common

language.

Let me use the example of threshold.

Yesterday, as I listened to Anne Munoz talk about

thresholds for allergens, I realized that we have

three different definitions of thresholds -- or

tolerance, excuse me. I want to use the word

tolerance.

The consumer says, "Tolerance is zero."

What that means is they think there should be zero

gluten in their food.

The medical community says "zero

tolerance." For them that means, you should be on

a strict diet, and you should never cheat.

The manufacturing industry wants to know

where that is. Is it 20? Is it 200? They know it

is not zero.

We are not talking the same language. The

consumer needs to know that the manufacturer and

the industry or the legal ramifications around any

labeling are all using common terminology in a

language they can understand.

Education is a huge component. As much as

I am a supporter of this regulation and this law,

one of the things that is going to happen, as you

heard yesterday in discussions about soy lecithin

and other ingredients, it is going to become a bag

of worms. For the consumer, it is going to be very

confusing, and we need to have an education

component as part of the new labeling laws.

I encourage you, too, although you heard

it yesterday and you will probably hear it today,

too, we know that there is no testing kit available

that tests to zero. We know, as you will probably

hear later, that it is probably an impracticality

or unnecessary to even go there.

I implore that when you set a threshold

method and testing methods, when you set the

threshold level, that it be reasonable and

something that meets the health needs of the

consumer but also allows the industry to meet the

need.

(Slide.)

MS. KUPPER: As they found out in

Australia, when you set zero as the tolerance level

and as the magic number for food manufacturers, a

lot of gluten-free products that patients used no

longer can be labeled gluten-free. Now the

consumer is once again confused and outraged.

CHAIRMAN DURST: Thank you very much.

Do we have any questions for our speaker?

Yes, Jeff?

DR. BARACH: Hi. Thank you. Jeff Barach

with Food Products Association. If I interpret

what you said correctly, you were talking about the

consumer really doesn't find descriptive labeling

very helpful in the case of gluten-free.

I assume then the consumer would go to the

ingredient list or the 800 numbers or their

internal chat groups to find out whether the

product really is gluten-free or not. Am I

interpreting that right? Your constituency does

not want gluten-free labeling?

MS. KUPPER: I would say that is probably

right, that is the message I got from the survey.

In fact, they found that labels that say "may

contain" or "processed in a plant with" really is

frightening to them. They will look at a product

like that, and they will simply avoid it.

They do go to chat rooms and there are

lists of gluten-free products. However, when you

look at those lists and you ask how they were

developed, there are no standards for developing

those lists.

DR. HEIMBURGER: Doug Heimburger, a follow

up to that. That is not the same, is it, as saying

"gluten-free"? Do they not want a label except it

is gluten-free with a consistent and clear

definition of that?

MS. KUPPER: They do want a label that

says gluten-free with a clear and consistent

definition.

DR. HEIMBURGER: Yes.

MS. KUPPER: I believe that gluten-free is

not is not going to mean zero; it can't mean zero.

CHAIRMAN DURST: Any further questions or

comments?

(No verbal response.)

CHAIRMAN DURST: Thank you very much.

CHAIRMAN DURST: Our next speaker is

Dr. Donald Kasarda, who is a consultant and retired

senior scientist from the Agricultural Research

Service of the USDA. He will make a presentation

on grains.

GRAINS

DR. KASARDA: Good morning everyone. I am

a research chemist retired from the U.S. Department

of Agriculture, although I still maintain a

relationship with my old lab in Albany, California,

as a collaborator.

Now, Dr. Murray covered a lot of the

things I am going to talk about. Maybe I will be

able to add a little bit more detail to some of

them, but he did an excellent job of talking about

some of the grain topics.

(Slide.)

DR. KASARDA: Immunology textbooks often

classify hypersensitivities into these four types.

Celiac disease is a delayed type hypersensitivity

that involves T-cells in the primary mechanism. It

falls into Type IV. Allergy is Type I and is

mediated by IgE antibodies.

Now, in the case of celiac disease, it is

often suggested that there is a Th1 mechanism

involved in which T-cells are presented with

gliadin peptides, and, ultimately produce

cytokines, inflammatory cytokines such as

interferon gammas as an example.

Now, in the case of allergy, however, the

same molecules that can induce the symptoms of

celiac disease are also capable of producing

allergies. We do have a certain amount of

confusion sometimes between immediate

hypersensitivities and the delayed-type, celiac

disease.

(Slide.)

DR. KASARDA: Now, this is the same

diagram that Dr. Murray showed. I want to talk

about primarily the endosperm, which is this white

part here (indicating) in the cutaway diagram.

The starchy endosperm is made up of about

75 percent starch, but it also contains about 7 to

17 percent protein, depending on the use of the

wheat. Most of this protein, about 75 percent of

it, is gluten protein.

The proteins are storage proteins. They

are used by the developing plant that comes from

the germ here. The germ is separated from the

outer layers and the endosperm during the milling

process after crushing and sieving.

The storage proteins are broken down upon

germination of the seed to produce a new plant.

The resulting amino acids and nitrogen are used in

the synthesis of new molecules needed by the

developing plant. Now, as I mentioned, about 75

percent of the storage protein is, in fact, gluten

protein.

(Slide.)

DR. KASARDA: This is a picture of flour

particles, a scanning electron micrograph. These

round, spherical structures are starch granules.

These (indicating) are A type, there are some

B types which are small here. The surrounding

rough-edged material is the gluten protein or

storage protein.

(Slide.)

DR. KASARDA: If you mix together flour

and water, as most of you have had the experience,

you can form a cohesive elastic dough. If you need

a dough under water, say, in a large container of

water or under a stream of water, you can wash out

the starch granules; they pop right out of the

matrix. You are left with a cohesive, elastic mask

consisting mainly of the storage or gluten

proteins.

(Slide.)

DR. KASARDA: Now, this is the traditional

cereal chemist definition of gluten. You cannot

carry out this process with rye and barley.

Therefore, to the traditional cereal chemist, there

is no gluten in rye and barley. However, the

celiac disease community has adopted the term

"gluten" for any protein that is toxic or harmful

to a celiac patient.

This terminology problem sometimes is

confusing when patients go to a company where they

might be dealing with a traditional cereal chemist,

and there is a certain amount of confusion as to

what is gluten. As I said, this is the traditional

definition, but it has been expanded to include

other grains that are harmful to celiac patients.

Now, from time to time, you will hear

about these fractions of gluten. Going way back,

at least over a hundred years, it has been

traditional to divide gluten into two, roughly,

equal fractions based on their solubility. This is

not an exact separation. No solubility fraction is

ever perfect.

Traditionally, it was alcohol-water

solution and sometimes we used detergent solutions.

We divide it up into the soluble fraction, which we

call "gliadin."

This is made up of monomeric proteins of

the prolamin class. The prolamin terminology comes

from Osbourne back around 1900. It is derived from

the fact that there are two major amino acids found

in the composition of these proteins. Proline and

glutamine, hence, prolamin.

By structure, we have three types: the

alpha type, gamma type, or omega types. Sometimes

people speak of the alpha/beta. I will talk about

that in a little as we go along.

Now, the insoluble fraction is called

"glutenin" by the cereal chemists. In rye and

barley, there is an equivalent fraction that we

called just generically "glutelin."

Now, this polymeric fraction consists of

prolamin subunits. Again, large amounts of proline

and glutamine in the composition of the proteins.

These subunits are linked together by disulfide

bonds into a higher level of polymer.

Of course, a protein is a polymer in

itself. It is divided into two main types the

low-molecular weight and the high-molecular weight

glutenin subunits.

(Slide.)

DR. KASARDA: This just is a table showing

the percentages in the various types of proteins.

For example, you have the sum of glutamine and

proline ranging from about 40 percent up to about

80 percent in some of the omega gliadins.

This is pretty unusual to have such a high

percentage of glutamine and proline. This is key

to the toxicity, because the toxic sequences

involve glutamine and proline and usually an

aromatic as well, either tyrosine or phenylalanine.

(Slide.)

DR. KASARDA: Now, the terminologies that

we use really go back to early electrophoretic

studies in the late sixties and early seventies.

Again, if we follow this diagram here, this is an

acid gel in which the proteins are separated by an

electric field in a polyacrylamide gel.

The terminology actually came from a sort

of free-boundary electrophoresis that was carried

out at our Northern Regional Research Center back

in the sixties. When they developed the

polyacrylamide gel electrophoresis, it was found

that the fractions fit with the mobility in the

electrophoretic gel.

You have the alpha, fastest moving; beta;

gamma; and omega. Structurally, the alphas and

betas are pretty similar. Some people will talk

about alpha/beta types. I just lump them together

as alpha types.

Now, the alpha type and gamma type are

about the same size. If you carry out SDS page or

polyacrylamide gel electrophoresis in detergent,

sodium dodecyl sulfate, which is a very good

dissociating agent for proteins.

Reduced or unreduced the gliadins give a

pattern somewhat similar to this. It is not quite

as good at resolving alpha, beta and gammas as you

find in the acid gels where aluminum lactate was

one of the favorite buffers.

If we go over to the glutenin fraction,

and these are the subunits linked together by

disulfide bonds, if you try to take a purified

glutenin fraction and run it into an acid gel or

into a detergent gel, mostly you've just got a

little bit of streaking around the origin because

the polymers are too large to migrate into the gel.

(Slide.)

DR. KASARDA: Upon reduction, however, you

begin to see this type of pattern here in which

there is a group of high-molecular-weight subunits

and a group of low-molecular-weight subunits. This

only occurs for the glutenin fraction when you

reduce the disulfide bond.

(Slide.)

DR. KASARDA: This is a two-dimensional

pattern, electrophoretic pattern, of the gluten

proteins. All you have to recognize is that each

spot here represents a separated protein. There

are quite a few different gluten proteins, and we

can count easily 50, 60, 70 spots in such a

pattern. Therefore, there are at least 50, 60, 70

gluten protein components.

We know from genomic studies that, in

fact, there are probably at least a hundred genes

coding for these proteins, and probably several

hundred genes coding for the proteins. The loci in

the genome that code for these proteins are spread

out over about nine different loci in the genome.

Now, as far as we know, all of these

gluten proteins are toxic in celiac disease. This

group here (indicating) are the omega gliadins, and

they seem to be particularly active.

However, all of the gluten proteins have

been tested by Paul Ciclitira's group and

Peter Howdle's group in the U.K., and they all

indicated by direct installation into the small

intestine that these proteins, all of these

different classes are toxic.

These omega gliadins are noted for being

strong allergen in exercise-induced anaphylaxis.

They are one of the really strong antigens involved

in that particular allergy.

(Slide.)

DR. KASARDA: This is a schematic diagram

that illustrates the fact that all of these

proteins are noted for a repetitive domain in which

certain amino acid sequences are repeated over and

over again.

They are somewhat degenerate, but we can

derive consensus sequences. These are glutenin

subunits, gamma-type gliadins. These red,

staplelike lines indicate intramolecular disulfide

bonds.

In a glutenin subunit, we also have free

cysteines which can link up to another molecule to

form these higher-level polymers. For example,

here is an alpha-gliadin -- I'm going to talk about

this a little bit more -- the end terminal, or the

first half of the molecule, is made up of these

repeating sequences.

The second half is not repetitive and

contains most of the disulfide bonds. Toxicity

seems pretty likely to be limited to the repeat

regions. These are the high-glutamine, the

high-proline regions.

Now, the omega-type gliadin seems to have

lost -- well, we are not entirely sure whether they

lost this type of domain or not, but in any case

they are made up almost entirely of repeating

sequences.

(Slide.)

DR. KASARDA: This is a hypothetical model

of the gluten polymer or glutenin in which the

subunits are joined by intermolecular disulfide

bonds, there are also these intramolecular

disulfide bonds, to form a higher-level polymer

that provides elasticity to a dough.

The gliadins and the glutenins are

cohesive with one another, but the gliadins

contribute more to the extensibility of the dough,

and the elasticity comes primarily from the

glutenin fraction.

(Slide.)

DR. KASARDA: Now, here I show some of the

types of sequences that you find in the repeats.

Now, they look pretty similar, a lot of glutamine,

which is represented by "Q"; a lot of proline

represented by "P"; and usually an aromatic

residue, either phenylalanine, "F," or tyrosine,

"Y."

Somewhere, and these are often degenerate.

They are not exactly according to the consensus

that I show here. Somehow along the line these

sequences have acquired toxicity in celiac disease.

(Slide.)

DR. KASARDA: Now, this is the complete

sequence of an alpha gliadin. This is the

end-terminal region up here (pointing). It starts

at one, and there are 263 amino acid residues.

Here, note the predominance of the blue Q's and the

red P's for the proline and glutamine residues.

This half of the molecule here is the

repeat region. There is also this interesting set

of glutamines, which really hasn't been studied in

celiac disease. It is probably not toxic, but, as

I say, there has been almost no study of this

polyglutamine stretch here.

Note also these vertical lines here which

I show. Those are sites that we have observed

where cleavages occur with gastric enzyme, pepsin

and pancreatic enzymes, trypsin and chymotripsin.

Now, most proteins would be broken down by

the digestive enzymes into single amino acids or

very small peptides: diatride, tetrapeptides that

are easily absorbed, which are probably not toxic.

(Slide.)

DR. KASARDA: In the case, as Dr. Murray

mentioned, because we have a lot of proline which

interferes with the breakdown by the proteolytic

enzymes, we can get some pretty large stretches.

This stretch here from 31 to 55 right here is

something that we have tested as toxic.

Other people have dealt with sequences

from this stretch and found them also to be at

given toxic. The fact that this gliadin and

glutenin proteins are difficult to digest by the

digestive enzymes allows these toxic stretches to

exist longer than you would find for other

proteins.

Now, this half of the molecule has a fair

amount of glutamine and proline, but as far as we

know toxicity does not reside in this C-terminal

half or sort of the end of the molecule. It is in

sort of the forward end of the molecule.

(Slide.)

DR. KASARDA: Now, this is what I would

call my string of beads model in which I have just

taken that sequence, 1 to 263, and shown it as

beads on a string.

Each bead represents a different amino

acid. I tried to assign the different amino acids

different code words to distinguish them. This is

the end terminal region of repeats. This

C-terminal where we have the disulfide bonds here.

Toxicity resides in this part here.

This sequence here from 31 to 43 was

synthesized by Mike Marsh in the U.K. first, and he

instilled the synthetic peptide directly into the

small intestine of several celiac patients and

found changes in the mucosa that were indicative of

celiac disease. So this does seem to be a toxic

sequence.

Here I show a computer molecular model of

what that sequence would look like in the

polyproline II left-handed helical confirmation

that Dr. Murray mentioned. We think that these

peptides do have a strong tendency to assume this

polyproline II confirmation.

Here, I show the sequence in three-letter

code as a string of beads model and here just a

single-letter code. I know most people are not

used to dealing with these codes. I apologize for

using them in some of the slides, but often I am

just trying to make a general point, and you don't

really have to follow the sequences according to

their exact correlations with the amino acids.

(Slide.)

DR. KASARDA: Now, this is a list of some

of the either toxic or immunoactive peptides that

have been described in the literature. This is

from Sean, et al, from Chaitan Khosla's lab at

Stanford. The 33-Mer appears to be a very active

sequence. I have indicated some sort of homology

100

here by the yellow boxes here.

All of these sequences, with the exception

of this one, have been found to be toxic by direct

installation into the small intestine or they have

been found to stimulate T-cells, T-cell clones,

derived from biopsies of celiac patients.

These are just some of the toxic

sequences. We don't know all of the toxic

sequences at this point. There are certainly

others to be found, so it is a pretty complicated

situation in trying to sort out exactly what it is

about the sequences that produces toxicity in

celiac disease.

(Slide.)

DR. KASARDA: Now I want to move on and

talk a little bit about the other grains. If we

start with the class flowering plants, which is one

of the major divisions or plants in terms of

taxonomy.

We go down to the major two subclasses,

monocotyledones plants and dicotyledones plants.

101

We can follow the monocots down to the Gramineae or

the grass family.

Here, we have only wheat, rye and barley

that are toxic. Triticale is a cross between wheat

and rye, and so would be expected to be toxic.

Now, oats I have had to put in both

columns, and I will explain why. There are many

other grasses in which the grains do not have toxic

proteins as far as we know.

There are only two grains that have been

studied with modern methods and modern approaches

to understanding their relationship to celiac

disease, and that is wheat and oats.

These others have often been studied very

minimally including rye and barley, but rye and

barley do contain proteins that have sequences

quite close to those in wheat.

We assume that rye and barley are probably

toxic grains according to the early results of

Dicke in the Netherlands back around 1950, where

they considered rye, barley and oats as part of the

toxic group.

102

Now about 15 years ago, I suggested that

if there are only a few grasses that contain the

toxic sequences, and they are closely related as

you will see, and then there are many other grasses

that do not contain the toxic sequences.

I suggested that if you get into the dicot

group -- the buckwheat, quinoa, amaranth, and these

other grains -- it would not be toxic, simply

because of their distant taxonomic relationship to

wheat.

I was a little bit apprehensive about

suggesting this, but over the past 15 years since I

suggested this, people have been eating these other

grains. As far as I know, there hasn't been any

serious indication that these do, in fact, have

toxicity for celiac patients.

There have been some very fine studies

from Finland and throughout the world, indicating

that oats were safe for celiac patients. But

towards the end of last year the Oslo, Norway,

group under Knut Lundin and Ludvig Sollid. They

have found -- well, I'm getting a little bit ahead

103

of myself. Let's deal with this slide first.

(Slide.)

DR. KASARDA: If we take a subfamily,

festucoidiae, of the grass family and we look at

the tribal level, and I made this slide before the

results from Oslo were published. The hordeae --

which includes wheat, rye, and barley -- were one

tribe. I thought that oats were probably

non-toxic, so I put them in -- well, they belong in

a separate tribe. It was only this one tribe that

had the toxic sequences.

Now, it is pretty certain that the oats

are toxic to a few probably rare individuals, but

we don't really know how this works out. They

found three celiac patients who definitely reacted

to oats by the same mechanism that they reacted to

gliadin peptides. This, I think, was pretty well

demonstrated by the work from Norway.

Now, here I show the proteins that you

find in wheat, gamma-type gliadins and related

low-molecular-weight glutenin subunits, they are

found in rye, barley, not in oats. Alpha-type

104

gliadins are found in wheat, but you don't find

them in rye and barley or in oats, and so on down

the line.

Now, the avenins are a small fraction of

the total proteins in oats. These make up only

about 10 percent of the protein. Most of the

protein is oat globulin, which as far as we know if

not harmful or toxic to celiac patients. There are

low-molecular weight proteins related to the

avenins in rye, barley, and wheat.

(Slide.)

DR. KASARDA: Now, this is another one of

those sequence slides, but let me just try to make

a few points here. This top sequence is a gliadin

sequence. It starts here (indicating) and runs

down to here.

The bottom sequence is an avenin, which

shows a lot of homology with the C-terminal half of

this gamma-gliadin molecule. This is also true for

the alpha gliadins. Where the amino acids are the

same, I have colored them in blue.

There is a lot of homology here in the

105

C-terminal half, but that is not where the toxicity

lies. The Oslo group has shown that this

particular sequence, which I have underlined, does

have certain characteristics including glutamic

acid at key positions that are important, as Dr.

Murray pointed out, for binding to MHC proteins.

Consequently, most of the repeat region is

absent from the avenins. There is just this sort

of residual section here, which does have a lot of

glutamine and a lot of proline and some key

glutamic acid sequences or amino acids that seem to

be responsible in these few patients that have been

studied for the toxicity. This sequence is

certainly capable of stimulating T-cell clones from

these patients.

I think that the evidence is pretty good

that there are at least a few -- it seems as though

there are probably rare individuals who respond to

oats and probably most celiac patients do not

respond to oats.

This is a rather puzzling situation,

because I have always thought of the proteins as

106

being pretty definitive for celiac disease, that

all celiac patients reacted to wheat and probably

to rye and barley, and that this was part of the

definition of celiac disease.

Now we have a situation here where it

appears that some patients react to oats and some

don't. This is some ongoing research that needs

some elucidation.

(Slide.)

DR. KASARDA: Just to jump back into the

classification, this is another subfamily,

panacoideae. Here we have maize and sorghum and

millet. We have actually done a little bit of

end-terminal sequencing on sorghum and millet

proteins, and they do seem fairly close to the maze

protein. This would explain to some degree why

they are not toxic in celiac disease. As far as we

know, it is still wheat, rye, and barley that are

harmful.

(Slide.)

DR. KASARDA: Now, this is my last slide.

The currently favored method for determination of

107

gluten in food is the R5 monoclonal antibody ELISA

test developed by Mendez in Spain. This seems to

be a pretty good test; it is not perfect.

The antibody reacts to monomeric wheat,

rye and barley prolamins, but not the oat avenins,

and it reacts weakly or not at all with the

glutenin or glutelin, pretty good sensitivity,

recognizes these particular motifs more strongly,

although some others that are similar are

recognized weakly.

The Codex Committee on Methods agreed in

2004 to endorse temporarily the R5 ELISA for the

determination of gluten. Now, there are some

possible problems. There is the failure to detect

the glutenin proteins.

In some preliminary work from our

laboratory, for example, when we look at wheat

starch that is intended for use by celiac patients,

we find that the gliadins are pretty well washed

out, but we do find evidence of

high-molecular-weight glutenin subunits attached to

the starch surface. These would not be picked up

108

by the R5 ELISA.

Then, there is the question about small

peptides from hydrolases. There has been some work

described using a competitive assay, which might

possibly solve the problem of the small peptides

from hydrolases.

Also, there is a certain amount of data

indicating differences in the results from

different labs on the same sample. On the whole,

it seems like a pretty good test that can be used.

As I say, it is not perfect, but it is a pretty

impressive test on the whole, and it may be as

close as we are going to get.

Although, I certainly can think of some

ideas for maybe improving it. At any case I think

I will end my talk here and I thank you very much

for your attention.

CHAIRMAN DURST: Thank you.

QUESTION AND ANSWER SESSION

CHAIRMAN DURST: Are there questions?

Okay. Margaret?

DR. BRILEY: Margaret Briley. Can you

109

give us any idea of the use of this ELISA test by

industry in terms of the frequency and the

acceptance and willingness to do it? Do you have

any feel for that?

DR. KASARDA: Well, I think it is being

used, and Dr. Collin can comment on this,

extensively used in Europe and it is becoming used

in the U.S. Susan Hefle, who spoke yesterday, I

think that they have done a certain amount of work

using this test for industry.

My impression is that not much testing is

done in the U.S. Maybe Cynthia could comment on

that. I think it is developing, but we are quite a

bit behind the Europeans in terms of a willingness

to test and desire to test products and make sure

that they are as close to gluten-free as they

possibly get. I can't give you a definitive

comment on that. I haven't made any surveys.

CHAIRMAN DURST: Any further questions or

discussion?

Yes?

DR. CALLERY: Pat Callery. Thank you for

110

the review of the relevant biochemistry. Could you

relate the transglutaminase substrate specificity

to these various glutamine-containing --

(Simultaneous discussion.)

DR. KASARDA: Not personally, but other

people have. There are certain sequences that are

susceptible to deamidation and probably

transamidation as well. These have been described

in some recent publications. There are many sites

in the gliadins that are susceptible to

transglutaminase.

DR. CALLERY: The transglutaminase I

understood was an important feature in binding

these proteins and causing the --

(Simultaneous discussion.)

DR. KASARDA: Well, you know, in the MHC

proteins there is a positive charge in the binding

pocket that binds well to a negatively charged

glutamic acid. This does enhance the strength of

the binding to the binding site of DQ2/DQ8.

Now, I didn't get into it, although

possibly someone else will, there seems to be two

111

legs to the celiac disease situation. There is the

adaptive immune system, which has been worked on

quite a bit in terms of this presentation of

gliadin peptides to T-cells, to the T-cell

receptor, and stimulation along that leg.

However, one of the peptides I described

that Mike Marsh had studied, that particular

peptide is not immunoactive. It does not stimulate

the T-cells, yet when instilled directly into the

small intestine it produced changes that were

characteristic of celiac disease.

In the last couple of years, there has

been an interest in the role of the innate immune

system in possibly triggering the first leg of

celiac disease, which then progresses on to involve

the adaptive immune system and the CD4 T-cells of

the lamina propria.

I think that part is becoming pretty well

understood. Ludvig Sollid and his co-workers,

while he was on sabbatical at Stanford, they

actually crystalized a DQ2 with the gliadin peptide

in the binding site. They have defined a number of

112

characteristics of the peptides that are important

for binding.

However, I think we still don't understand

a good part of what is active or toxic about these

peptides, and it may have to do with this

triggering and innate immune response. That is

research that is really developing right now.

CHAIRMAN DURST: Soheila.

DR. MALEKI: Soheila Maleki, USDA.

Essentially, the substrate for the transglutaminase

is the same peptide that is presented by the

antigen-presenting cells?

DR. KASARDA: Well, after you deaminate a

particular glutamine, then the binding strate goes

up for the receptor site on the MHC protein.

DR. MALEKI: Essentially, it is the same

substrate, just after deamination --

(Simultaneous discussion.)

DR. KASARDA: Well, there is also the

question of, Why do you have antibodies to the

transglutaminase? This may involve transamination

reactions where you get a binding of gliadin in the

113

transglutaminase, and then this triggers the

apparent autoimmune antibodies to transglutaminase.

DR. MALEKI: I see. Well, I just find it

amazing that when you show the lineup of the

peptides, the homology, that you had a two amino

acid difference and went from immunoreactive to

non-toxic. I'm sure by now they probably can

explain that?

DR. KASARDA: No, they can't.

DR. MALEKI: Well, even in the fitting up

to the transglutaminase or to the processing by

antigen-presenting cells?

DR. KASARDA: Well, it is very puzzling,

very interesting. I really can't answer your

question.

DR. MALEKI: Thank you.

CHAIRMAN DURST: Any further discussion?

(No verbal response.)

CHAIRMAN DURST: All right. Thank you,

Dr. Kasarda.

We are now scheduled for a break. We are

just about on schedule, so we will take a brief

114

break and reconvene at 10:45.

(Thereupon, from 10:25 a.m. to 10:45 a.m.,

there was a pause in the proceedings.)

CHAIRMAN DURST: Our first speaker after

the break is Dr. Alessio Fasano, professor of

pediatrics, medicine and physiology and director of

the Mucosal Biology Research Center, Center for

Celiac Research, University of Maryland School of

Medicine.

PROSPECTIVE STUDIES

DR. FASANO: Thanks so much. I've got to

do this. I need really to thank the FDA, who has

been so kind to invite me, but also to be so

sensitive to use Italian candies.

(Laughter.)

DR. FASANO: This is very nice of you

guys, and we appreciate that. I also want to tell

you guys that because of the other speakers, I

decided to reduce a little bit my talk, so a few

slides have been taken out from the handouts to go

straight to the point.

There has been a general perception that

115

this is a quite young disease, in other words,

something that we are dealing with kind of

recently. I want to put this in the right

perspective and give you some of the background to

justify this prospective study to decide what is

threshold of tolerable gluten.

First of all, believe it or not the first

trace of a description of this disease goes back to

the Roman Empire. This is not something that has

happened in the last few years.

(Slide.)

DR. FASANO: Who really put the disease on

the map is this fellow here. Samuel Gee, at the

end of the past century, around 1890, gave an

historical lecture to a place where I had the

privilege to study for a little while at Saint

Bart's Hospital in London.

He really put celiac disease on the

"scientific map." I took little sentences here and

there from his lectures to give you the sense of

how this guy got the story right more than 120

years ago.

116

He described these as a chronic

indigestion that is met in every single age.

Again, our misconception in the past was celiac

disease was confined to a specific age group. He

knew already that was not the case; it can affect

at any age.

Of course, it is particularly more

frequent in all kids between one and five years

old, and that was the observation at the time. He

spent time and effort to clarify the fact that

everybody can be affected.

Now, symbiotics, hands-on, was the way to

do a disease diagnosis at that time. We didn't

have a lot of sophisticated tools, so it was really

hands-on.

For a gastroenterologist, dealing with a

problem like that means describing feces, stools,

and that's what it is. He introduced with this

description a very important concept about celiac

disease in terms of classical GI presentation,

malabsorption.

In other words, right before they would

117

know about the genetics, right before they would

know about the grains, the eyes are telling us that

the feces are loose, malformed, but not watery,

definitely more bulky than the food taken seems to

account for, i.e., malabsorption.

What is remarkable is this part here. He

ventured also to understand what was the

pathogenesis of the disease and introduced two

concepts: the genetics and environmental trigger.

He said kids that suffer from it are not

all weak in constitution, errors in diet. I want

to clarify that the first time that the link

between celiac disease and grains was made was soon

after World War II. Until then we had no clue

whatsoever what was the trigger leading to celiac

disease.

The link was made during World War II

because there was a higher rate of mortality among

kids in Middle Europe that was not explained.

During World War II, grains were not available

anymore.

They were fed with potato starch, potato

118

flour, and the mortality dropped dramatically, to

reappear after the end of the war when flour was

again available. That is when the link was made.

This guy is already there. Errors in diet

may be perhaps a cause, but whatever. Why, out of

a family of kids all brought up in a much similar

way, should only one suffer?

Again, he is trying to understand what is

the genetic component, what is the environmental

component, why some people have got it and some not

from the same family eating the same stuff.

Then, he finished up by saying, "Okay, I

think that I have a way to get to the bottom line

in treatment. The treatment has to be regulating

food in the main part of the treatment. It is

amazing if you come already with this conclusion.

The allowance of farinaceous food must be small.

Again, I find this remarkable. Highly starchy

food, rice, sorghum, corn-flour are unfit.

Now he is losing himself a little bit when

he says malted food is better. Also, rusks or

bread, provided it is cut thin and well-toasted on

119

both sides, will be all right.

Grant him the benefit of that. I believe

that, again, in 1890 making this kind of statement,

even if he [made] this little boo-boo here, I think

that it is absolutely remarkable.

(Slide.)

DR. FASANO: Now, fast forward that 120

years later, and that is what we understand about

celiac disease. You heard from Dr. Kasarda and

Dr. Murray already that this is an immune-mediated

reaction.

It is not an allergic reaction, but rather

right now we really truly believe that this is an

autoimmune condition. In other words, we are in

the same kind of range as multiple sclerosis, type

1 diabetes, and so on and so forth.

Therefore, as such there are two key

elements to develop the disease: You have to be

genetically susceptible. I'm not going to spend

more time about this DQ2/DQ8, but they are the

docking station, the "eyes," of the autoimmune

system to see the trigger from the environment

120

coming in.

It is unique because the only other

autoimmune disease for which we know everything

specifically is the only autoimmune disease for

which we know the trigger, that is, gluten.

I wish that we had that kind of

information for other autoimmune diseases, for

which we will have a solution. Theoretically, we

have on hand the possibility of treatment of this

disease.

However, I will argue that unless we have

a clear rule of engagement, i.e., a food labeling

bill that will really clearly define what is

"gluten-free," this is a theoretical solution but

very difficult to put in practice.

(Slide.)

DR. FASANO: Again, it is pretty obvious

that you have to have these two ingredients, you

have the genes and you have to have the grains.

When they interplay, you may end up developing

celiac disease.

We heard already that variability in terms

121

of the timing, how long it is going to take, the

outcome in terms of symptoms, and so on and so

forth, is unbearable. However, they are all under

the same kind of umbrella of celiac disease.

What are our treatment options at the

moment? If these two elements are absolutely

necessary to developing the disease, I believe it

is a no-brainer, it is pretty simple, there are

only two solutions.

First, we can remove the genes, and I

don't think that we can do that. We are not quite

there yet anyhow. As Dr. Murray explained, we know

some of them but we don't know all of them. Or,

secondly, we eliminate the grains. Those are the

options that we have available. There is no other

way to turn from this.

Don Kasarda went extensively into this. I

didn't know that he was invited, by the way.

However, the bottom line is the only treatment

right now is strict, lifelong -- as you heard, you

don't grow out of this, so you have got to endure

it for the rest of your life -- avoidance of wheat,

122

rye and barley. The oats story, again, I am not

going to go back because you heard about that.

(Slide.)

DR. FASANO: It is pretty obvious what are

the major sources of gluten. This is the easy part

when you have to deal with the patients freshly

diagnosed. It is easy to say, "You know what? No

bread, pasta, pizza, beer, cookies, muffins, and so

on and so forth.

(Slide.)

DR. FASANO: This is a little bit more

complicated, and that is where I believe a food

labeling bill will help. Of course, it is not

necessary to go and say, "You know what? This

muffin that you buy at the bakery needs a label."

We know that already whether it is gluten-free.

However, this stuff here (showing "Sources

of Gluten" slide) definitely needs a label, some of

them, because it is not clear if they have gluten

or not because they can be processed with or

without gluten.

Gluten is a formidable, extremely cheap

123

biological glue. Don told you the physical,

chemical characteristics of the molecule. The

reason why manufacturers use that is because when

you have two elements of a processed food that does

not stick together, the cheapest way to keep them

together over time is to use gluten. Right now,

the label can see just the nature of flavor but not

gluten, not necessarily so.

Then, there are really the tough ones in

which, this is not even food really, a source of

gluten needs to be considered. I can't

conceptualize enough how many times we've gotten

E-mails of people asking, "Is my husband, who has

celiac disease, going to be sick or whatever," or

the Playdough for the kids in kindergarten, and so

on and so forth. These are elements to keep in

mind that we deal with all the time.

Of course, the big deal is right here --

medications, prescriptions. As for foods,

processed foods, also medication they enjoy gluten

as an additive to keep elements together.

Now, while I was saying adhere to a diet

124

is a pure theoretical no-brainer, but in a

practical sense it is extremely complicated. It is

a chronic intervention that you have to do, and you

have to stick with it with full commitment for the

rest of your life.

Every single individual in this room I am

pretty sure that you have made some commitments

here and there to go on a certain diet or to

exercise or to decide to change your lifestyle. To

keep that constantly for the rest of your life, it

takes a lot of stamina. That is the reality of the

story.

That is true particularly in the American

society in which any chronic illness will require

chronic treatment, whether it be diet or exercise

or medication or whatever, will pose a problem of

compliance. Definitely among different

interventions, a diet compliance can be really a

difficult aspect of treatment.

In my book, food is one of the few joys in

life. How many times do we leave home and go to

work, we drive, we don't think about it, and we

125

find ourselves at work without having to pay

attention to directions, streets, and so on and so

forth? We are used to it.

That is the same with food, we are used to

it. However, that is not the case for celiacs

because they have to think about this over and over

and over again. It will become not a natural,

spontaneous activity in life, but it will become a

very, very demanding operation.

(Slide.)

DR. FASANO: Why don't people stick with

diets? This is a survey that was done in

Upstate New York. This statement, and this is just

to paraphrase something that Cynthia was telling

us:

"If I eat less gluten, I will have less

intestinal damage."

Half of the people say, "You know what? I

really don't have to stay a hundred percent gluten

free. As far as I decrease this, I will have less

problems. I will be all right."

"I've lived this long eating gluten, how

126

much will a gluten-free diet really help me now? I

mean, you know, if it's not been a big deal so far,

why should I just dramatically change my lifestyle?

I've survived so far, I'm not going to die from

it."

"It's not me, that I have to do this.

It's my doctor who should tell me when I need

follow-up testing or whether I need to stick with a

diet, and so on and so forth." One-fourth of the

people say that.

Again, you heard Dr. Murray, that

unfortunately some of the confusion is generated by

the professionals, the healthcare professionals.

They don't know the rule of the game, and,

therefore, they cannot transmit how to play the

game.

It is pretty much the sense that you go to

the doctor as an individual that has to teach you

how to play chess, and this fellow has no clue

whatsoever how to move the pieces. Patients have

to learn how to play chess while playing against a

professional player. How fair is that? It is an

127

ongoing process.

This is the one that disturbed me the

most: "Scientists and doctors still haven't proven

that gluten really hurts them." You know, there is

no clear information that gluten is dangerous to

celiacs, and that is quite disturbing.

(Slide.)

DR. FASANO: What are the current barriers

in compliance? Again, you heard about the emotion

of the person, anxiety. There is a tremendous

reaction when you are diagnosed with a chronic

illness, no matter how you want to put it. Now,

grief and fear and denial are part of the story.

The ability to resist temptation and to be

disciplined on a gluten-free diet is tough. There

are feelings of deprivation. A few years ago I was

with one of the patients, and he got the chance to

drink a gluten-free beer. Soon after he started to

drink the beer, I saw tears coming down his cheeks.

His simple statement was, "I've waited 25 years for

this." Imagine, 25 years to drink again beer.

He was disciplined, and he didn't touch

128

it. However, there are many others, particularly

adolescents, in which that kind of discipline is

really hard to obtain.

This is very much the heart of the

problem, fear generated by inaccurate information.

If we do not have clear ideas, we, as

professionals, and one says black and the other one

says white, and the other one says up and the other

one says down, that creates a lot of confusion and

a lack of trust.

(Slide.)

DR. FASANO: Other barriers to compliance

are of course we live in a society that drives 150

miles an hour, and we don't have the time to seek

to prepare our food to enjoy. My kids consider

that the stove is the microwave. The stove does

not exist.

Cynthia teaches us the fact that the new

generation believes that cooking is just powder

mixed with water, stick it in the microwave, and

the only thing that you've got to do is read how

long should that go on and that's it. What

129

sophistication.

Here, assessing gluten content in food and

label reading is the most compelling change in

lifestyle that these people go through. Right now,

I don't know about you guys, but I don't enjoy food

shopping. I really do not. I tend to go at

midnight when nobody is there, because I want in

and I want out.

That is not an option for celiacs. One

thing that will take you, I don't know, half an

hour will take four or five hours for celiacs

because you've got to read every single label to

the nitty-gritty and make decisions.

Many times now I see people with cell

phones calling an 800-number right there on the

spot saying, "I have your Box XYZ, is this

gluten-free or not?" It is cumbersome.

(Slide.)

DR. FASANO: All of this to come to the

heart of what I'm going to share with you guys.

How much is too much? Unfortunately, I can't

conceptualize and stress enough what Cynthia

130

already said. In biology, the absolute zero does

not exist. If you really do believe that we can

achieve zero as gluten-free, this is a pure

theoretical concept that nobody will ever be able

to achieve.

Assume, just for a moment, that we will

have a sophisticated, super-duper sophisticated,

monoclonal ELISA to really go down to zero. To

manufacture food in that way, people in that

particular factory should be dressed with spray

suits, all antiseptic. A piece of bread will cost

$250, because that is what that level of

sophistication and controlled environments will

take. Consequently, it is impossible.

At the same time we need to give industry,

manufacturers, a parameter of what is tolerable and

what is not. There have been many retrospective

studies that Dr. Collin is going to tell us about,

very few prospective studies because they are

extremely challenging to do right.

This study that I am going to show you the

data of has really been coordinated by

131

Dr. Carlo Catassi, who has been involved in this

kind of topic for the past 15 years. He is a

member of our center, and we have been doing this

in coordination for the past four years.

Why do we need to do this? Because again

this is a long-term, strict gluten-free diet. If

we do a prospective study design, we can answer

questions that a retrospective study was not able

to answer.

How we did this? We did it in a way that

the gluten-free diet, people that come in are

already diagnosed on a gluten-free diet. We are

monitoring this gluten-free diet in a blind fashion

where a given amount of gluten is added to the

diet, then, the clinical, serological and biopsy

evaluation before and after the microchallenge.

The background noise, this is very

important, is caused by possible contamination of

the food was minimized by using a control group, in

other words, to really do this by the book.

(Slide.)

DR. FASANO: Studies done in the past, for

132

example, from Dr. Catassi almost a decade ago,

showed a linear relationship between the amount of

gliadin -- that is the toxic part of the story here

-- a daily dose, and it causes damage between 100

and 1,000 milligrams a day.

The intraepithelial lymphocytes -- and we

are going to go back to what these intraepithelial

lymphocytes are all about, the meaning -- was the

most sensitive index, not the serology and not the

symptoms.

What you heard already from Dr. Murray is

that after all these red flags the antibodies may

not be sensitive enough to uncover exposure to

gluten. Indeed, even 10 years ago this was very

clear.

(Slide.)

DR. FASANO: Why do this again? If it was

done 10 years ago, why revisit this if we have

already the information? Several reasons. The

need, first of all, to investigate the effects of

lower gluten doses. Because at that time they were

using large doses, because that was the level of

133

sensitivity of the tests for the foodstuff.

There is a need for prolonging the

duration of the microchallenge. In the past, the

longest that we went was a month, and people would

ask, "How about two months or three months?"

How about if the period, the lag period,

between the exposure to gluten and when you react

is longer? You believe it to be safe for one

month, but you keep going, and eventually you

react.

There is a need of a control group that

was never used before, and, most importantly, you

heard that gliadin is part of the story. They are

the glutamines.

If you do the study just as done in the

past, you may really not uncover what is really the

story; in other words, what you leave out there is

not pure gliadin but rather this mixture of

proteins that Don Kasarda was telling us about.

(Slide.)

DR. FASANO: I don't want to spend too

much time on this, but for a matter of

134

quantification, to give a sense of what we are

talking about. In 200 grams of wheat-based

products -- bread, pasta, so on and so forth -- you

heard that the main proteic fraction in wheat is

gluten. For 8 to 14 percent of the overall amount

is wheat. Gluten is 75 percent of all the protein.

Between gluten and glutamine, we can say that all

of this 8 to 14 percent are these toxic proteins

for celiacs. This 8 to 14 percent translates into

15 grams.

The real toxicity, the main toxic, is due

to the gliadins. Again, glutamines contributed to

toxicity. Of the 200 grams, 8 to 14 percent is

equal to 15 grams. Half of it is gliadin. Gliadin

has more than 50 toxic fragments, and so on and so

forth.

If you go on a gluten-free diet, an adult

that is on a gluten-free diet, roughly, consumes --

I mean, in a normal diet, roughly, the amount that

you consume is this, 15 grams. Roughly, you

consume 200 grams of wheat-based products.

If you are on a gluten-free diet, a

135

typical gluten-free diet, the subject consumes

gluten-free flour-based, that is roughly 80 grams.

The key element is how much of this 80 grams of

gluten-free products can be contaminated with the

toxic element, gluten? How much is the amount that

you can tolerate? That is the heart of the problem

here.

(Slide.)

DR. FASANO: That prompted the design of

the study. It is a quite complicated study. The

aim was to evaluate the consequences of the

protracting just minimal intake, either 10 of 50

milligrams, a very small intake.

In a group of adult celiacs on long-term

treatment with the gluten-free diet, why this

amount? Because, again, 100 milligrams was already

tested and proved to be dangerous 10 years ago.

How the study was designed was as a

multicenter, prospective randomized,

placebo-controlled, double-blind and was a

three-year study. It was entirely sponsored by the

Italian Celiac Society.

136

The reason why we did it in Italy, as I

was mentioning before, is mainly because economical

support of such a complicated and expensive study

could be executed at this time only in a place

other than the United States where we don't have

that kind of resources.

(Slide.)

DR. FASANO: Who was eligible? Patients

with biopsy-proven celiac disease had to be on a

gluten-free diet for at least two years. These

people that had been diagnosed with all of the

criteria are accepted and have to be complying with

the diet for at least two years.

If you are younger than 18 years old, poor

compliance, abnormal results at the baseline

evaluation or you have IgE deficiency, that will be

an exclusion criteria.

(Slide.)

DR. FASANO: Now, how we did this? Well,

again, these people were heroes to accept such a

study, but this was the only way to do it. These

people would come in to be scrutinized to see if

137

they were eligible.

If they were eligible, a consent form was

obtained and there was an intense, strict

monitoring of their gluten-free diets for a month

before the beginning of the study was obtained.

Baseline clinical serological and a biopsy was

obtained. In other words, they underwent a

endoscopy with a biopsy to show that they were

fine.

They were blindly randomized in three

groups, either no gluten, 10 milligrams of gluten

or 50 milligrams of gluten. They were followed for

three months. At a monthly interval there was a

check with the serologist for symptoms.

At the end of the study, at the end of the

three months, once again there was a clinical

evaluation and a serological evaluation and a

second intestinal biopsy under endoscopy.

This was the kind of study that this was

the only way, given the fact that the we know

symptoms and serology tests cannot be sensitive

enough to do this right.

138

(Slide.)

DR. FASANO: The purified gluten was used

for the challenge. Gluten -- or lactose-containing

placebo -- capsules were randomly prepared. The

lab tests were centralized. There was monthly

monitoring of adherence to the protocol; it was

checked by a nutritionist.

Measurement of gluten contamination in

commercially available gluten-free food that they

had during the challenge was checked by ELISA. The

serum AGA and anti-tTG antibodies were checked; a

biopsy was performed with morphometry; there was an

intraepithelial lymphocytes count; and control

biopsies from non-celiac patients were used.

(Slide.)

DR. FASANO: These are the foods that they

had a gluten-free foods. You keep in mind that in

Italy right now the food labeling policy is to be

labeled as gluten-free you have to have 20 parts

per million or less.

Indeed, with this simple exception, the

vast majority of the foods that these people there

139

are eating was gluten-free, by definition of the 20

parts per million.

Consequently, the only gluten that these

people were seeing was actually the ones that were

dealing with the challenge, if they were in the

group of gluten exposure.

(Slide.)

DR. FASANO: We were able to recruit 39

people, who were divided equally into three groups.

There were a couple of things that were interesting

to us.

Of all the parameters that we measure, two

are extremely important to establish the health of

the intestine and the exposure to gluten damage,

one was the villous height/crypt depth ratio. It

is very typical use of morphometric analysis that

we do in clinical practice.

Typically, we want to see this: roughly, a

ratio of 3:1. In other words, the height of the

line has to be 3 times the depth of the crypt.

That is what typically we consider to be normal.

Despite the fact that they were on a

140

gluten-free diet, despite that, they fulfilled the

criteria. They were gluten-free, symptom-free,

immunologically negative, and all the 9 yards.

They went on a one-month controlled diet.

When we did the starting biopsy, there was

a slight decrease of the villus-crypt ratio,

meaning, the villi were a little bit shorter. That

is what happens when you have an insult, the villi

become short and the crypts go deeper.

The other parameter is the number of

CD3-positive cells, the intraepithelial lymphocytes

if you wish, was again 20 per hundred entrocytes

and controls and 30 in the celiacs on a gluten-free

diet.

Therefore, at baseline already something

was going on. It is like there is a status of

inflammation in which this is like a very

well-trained athlete, ready to react to anything if

it smells gluten coming through. It is really at

the edge, ready to jump.

There was a strong correlation between the

number of intraepithelial lymphocytes and the

141

villus/crypt ratio, meaning, that the more healthy

is tissue, the less intraepithelial lymphocytes.

The healthier the tissue -- when the crypts are

elongated and the villa get short, the more

intraepithelial lymphocytes are there. The

intraepithelial lymphocytes are really soldiers

that the immune system sensed at the forefront and

ready to fight the battle. That is what it is.

(Slide.)

DR. FASANO: Now, what kind of symptoms

after the three months these people experienced in

the three groups? There were not really

significant differences: abdominal distention,

anemia, iron deficiency, loss of appetite,

bloating, and so on and so forth.

There were equally distributed in all

groups including the placebo, but two really stand

out -- all in the 50 milligrams. This stomatitis

and the mouth, there are the typical signs of

mucosal involvement of the oral cavity in celiacs,

well-described, it was present only in the

50 milligrams. Weight loss was experienced only in

142

the 50 milligrams. For the rest, we didn't see any

major differences.

To revisit the concept that the antibodies

were useless -- these are the antibodies, IgA and

anti-tTG and IgG anti-gliadin antibodies -- before

and after the challenge in placebo 10 and 50

milligrams, there was no difference. Pretty much

there was no difference among the groups.

What we saw as the difference was the

villus/crypt ratio, that all in the 50 milligrams

started to decrease to a level of significance.

After three months, we saw the crypts become a

little bit deeper and the villi to become a little

bit shorter. This translates in the fact that

there was damage that started to occur, or possibly

damage that started to occur.

The intraepithelial lymphocytes, there are

these spots here (indicating). Again, these are

lymphocytes under normal circumstances you see in a

smaller quantity in between epithelial cells.

It came to be of a very increased number

in people with 50 but not in 10, not reaching

143

statistical significance, but these are trends that

I have the obligation to report. It is not

significant that there are more of these cells in

the 50 milligrams compared to the starting point,

but it is a trend there.

(Slide.)

DR. FASANO: I believe the heart of all of

this is this table. I believe this really cuts to

the chase. It is extremely confusing, particularly

to patients, when you talk about milligrams and

parts per million. What the heck are you talking

about? Why do we use this parameter of parts per

million and not just straight milligrams?

Because, by the way, say, you do the study

and you show that 50 milligrams could be dangerous,

so how can it be 10 milligrams? How much is 10

milligrams? How much of a pizza is 10 milligrams?

You say, "Well, let me give you the bad news. It's

less than a fraction of a crumb of a piece of

bread. That is what we're talking about.

Still, it doesn't give you clearly what is

the magnitude of the stuff that we are talking

144

about. The reason why we prefer to express in part

per million rather than in milligrams is because

the amount that is tolerable really depends on how

much you eat.

(Slide.)

DR. FASANO: As you see here, this is the

daily intake of gluten-free flour or whatever

products are based on gluten-free. If you eat

50 milligrams, of course you end up to ingest much

less than 300 milligrams of the substance that you

are eating.

Let's say that, for example, we set the

parameter at 200 parts per million. If we want to

accept the outcome of this study as something to

keep in mind, 10 milligrams is safe for everybody,

50 milligrams start to be questionable.

If you set the threshold at 200 parts per

million, if you eat a relatively small amount of

the stuff a day, you are okay. If you eat a little

bit more, you are in an area that we don't quite

know, because again it is between 10 to 50

milligrams. You can argue, "Is 40 okay? Is 30

145

okay?" We don't know.

Definitely, if you eat 300 grams a day --

in other words, you eat large amounts of

gluten-free products that is contaminated to the

level of 200 parts per million -- you start to go

into the red zone. That is dangerous.

If you go down this table, you see that if

you set 20 parts per million, no matter how much

your Italian lifestyle of eating like crazy food

that is gluten-free based, no matter how far you

go, you still are well below the threshold.

Therefore, at least based on this study,

that I believe has been done really the way that it

is supposed to be done, long enough, because three

months is definitely a long period, a threshold of

20 parts per million should be safe for the vast

majority of the people because it will keep you way

below the cutoff that seems to be dangerous, i.e.,

the 10 milligrams.

(Slide.)

DR. FASANO: Now, this litany of names is

just to explain that this was not a trivial study.

146

It was a multicenter study that involved a

tremendous amount of work and a tremendous amount

of dedication of people that have no business to

undergo this, particularly two endoscopies with two

biopsies. However, it is only because of the

dedication and the commitment of these people that

we have an answer and we have a chance to come to

you today with something that is a little bit less

foggy than so far we have had in terms of

prospective studies.

I will stop there, and I will take any

questions that you have.

CHAIRMAN DURST: Thank you very much.

QUESTION AND ANSWER SESSION

CHAIRMAN DURST: Questions?

Margaret.

DR. BRILEY: Margaret Briley. Can you

tell me, I didn't understand, how often did they do

the biopsies? Every month? Every three months?

DR. FASANO: No. No, no, no, that would

kill us if we do it every month. No, the biopsy

was done at the beginning of the study, at the

147

entrance, and at the end of the study, three months

after, the idea being how much insult did you get

in two months.

What was done at intermediate intervals

was a survey of the diet, to make sure that they

were complying with a gluten-free diet, survey

compliance of taking the pill, and the serological

tests for the antibodies. Those were done on a

monthly basis.

DR. BRILEY: On a monthly basis?

DR. FASANO: That's right.

DR. BRILEY: Thank you. That was good.

CHAIRMAN DURST: Erica.

DR. BRITTAIN: Erica Brittain. If I'm

understanding correctly, the conclusion of the

study is that 10 milligrams daily would be safe,

was shown at least to be fairly similar to your

placebo group in this four-month exposure. How

would you know how that would translate to four

decades of exposure?

DR. FASANO: Only with decades of

prospective study. You are a statistician, and you

148

know better than I do that you've got to start

somewhere.

There is no question in my mind that the

only way to do that is like when you put a new drug

on the market, and you go to Phase I and you do 10

people. When you do Phase II, you do 100 people.

You do Phase III, and 10,000 people. Everything is

fine. Ten years later, because millions of people

took it, it may be that something wrong will come

up, a classical example.

I don't have an answer for you. How do I

know in 10 years what's going to happen? But, you

know, we have to have some way to start. I believe

that this study is giving us a parameter, a

justification, a scientific rationale to say,

"Let's start here."

CHAIRMAN DURST: Okay. Soheila.

DR. MALEKI: Soheila Maleki. I was just

wondering, this is probably not directly related to

your topic, I heard earlier mention of wheat flour

and exposure. How much is inhaled exposure

involved in some of these reactions?

149

DR. FASANO: I don't think that anybody

can answer with scientific confidence that inhaling

is or is not a possible port of entry of gluten for

people with celiac disease to react to.

What we know as a fact, an undisputable

fact, is that the intestine is the port of entry,

the key port of entry. I can tell you anecdotally

that we have patients that react to inhalation of

gluten leading to asthma as an allergic reaction to

gluten rather than to celiac disease.

How confident am I that this could be an

alternative to the other route? I'm not really

confident, because I don't think that we have the

scientific proof beyond any reasonable doubt, as we

do with the other route, that it could be a

possibility.

DR. MALEKI: Thank you.

CHAIRMAN DURST: Ciaran.

DR. KELLY: Yes.

Thank you, Alessio, I agree. Thank you

for sharing the data with us, and I agree that at

least it is a basis that we can begin to work from

150

and make some rational approaches to what is best

for our patients with celiac disease.

A couple of questions: The first relates

to the earlier question about the 40-year

experiment. There is one that there is a

20 part-per-million threshold set already in Italy.

Could you comment on how well that is tolerated by

every or the vast majority of celiac patients in

Italy?

DR. FASANO: Actually, it is much more

than that. There are interesting natural

experiments being done. Italy for many years now

reinforced the 20 parts per million. England has

this 20 to a hundred, and so on and so forth.

As far as I can tell you, this is

something that in Italy the food labeling

legislation setting it at 20 parts per million has

been there for 7 years. It has been considered to

be absolutely safe with very sporadical reports of

reactions.

Now, I think it was telling you when you

have a stomachache and you are a celiac, you tend

151

to go that way to the extreme that some people say,

"Today's ache is because I had gluten."

I mean, this is the reality of the story.

But if you want to, statistically speaking, work on

the large numbers, I would say that 20 parts per

million has been proved to be safe.

DR. KELLY: In your study, then, and this

is something that we discussed a lot yesterday in

the context of food allergy and challenge studies,

is there the potential for bias in selection; in

other words, individuals who are highly sensitive,

in terms of symptomatically highly sensitive, to

low levels of gluten would either be afraid or not

choose to enter the study?

DR. FASANO: Absolutely. Absolutely, no

question about it. The reality of the story is

that if you are extremely sensitive to gluten, you

would be less willing to expose yourself to

something that you know is going to harm you.

The point is, What percentage of the

population does that represent? Is it 10 percent,

20 percent, 50 percent, or a fraction of 1 percent

152

of the celiac population?

You know, I'm pretty sure everybody that

is involved in the clinical care of people with

celiac disease has run into people who are

extremely, extremely sensitive to gluten out there.

The exception are people where actually the problem

is the opposite; these are people who can eat

dangerous amounts of gluten and they do not react.

That is a problem.

CHAIRMAN DURST: Dick Durst. Just to

follow up on that, How did you recruit the people

for these studies?

DR. FASANO: The method of recruitment is

a major advantage of the Italian setting is that

there is a single Celiac Society, and they are

extremely committed. What we did was very simple.

They have a national bulletin, both electronically

and on paper, that is read pretty much by the vast

majority of the members of the celiac community.

I believe that we originally asked for 45

volunteers. That is the number that the

biostatistician told us to go for to have a

153

meaningful outcome. We got 470 volunteers, so we

had to turn people down.

CHAIRMAN DURST: Did you at that point

know which ones were the hypersensitive or the more

sensitive versus other and select on that basis at

all?

DR. FASANO: No. The way that these

people were selected was completely random. In

other words, the least that we had every "X," three

or four up -- I don't know, to make the number --

were called to make that unbiased. We really

wanted a representative portion of the population.

This was done by also sex and age.

Yes?

DR. KELLY: Ciaran Kelly again. I do have

one other question, and it has to do with the

interpretation of the data on villus/crypt ratio

and IEL counts in the controls versus the

well-controlled celiacs.

You showed that there was a small

difference at baseline, even though those

individuals were doing well on a gluten-free diet.

154

Your interpretation is that there is an underlying

immune activation. My question is, Is it possible

or likely or relevant that the 20 parts per million

that they are taking is perpetuating that?

DR. FASANO: What I am trying to convey is

the difference is that the recovery -- even if you

are completely, religiously gluten-free -- is not

100 percent. That is what I meant.

I don't know if this is due to an ongoing

immune response. I believe that to not probably be

the case. Because after all, after all with all of

the machinery in the community, these people have

been proved not to go back to normal. Whereas,

again, the fact is that no matter how you push it,

you can't really go back to normal.

I think that the fact that for three

months, even if you were really "touched," so to

speak, you did not react to 10 milligrams. For me

it was a great level of confidence that this is the

way to go -- together again with data with a

retrospective study, that we are going to hear

about in a moment, and on-the-field exercise in

155

Italy.

CHAIRMAN DURST: Jean?

MS. HALLORAN: Another question about the

sample group. When you did the baseline study, how

much variability did you find in the members of

that group?

DR. FASANO: Let me see if I can go back

on this.

MS. HALLORAN: You had two factors that

you looked at, the villus height --

DR. FASANO: Can you put on the slide show

for a second?

(Ms. Sylvia Smith complies.)

DR. FASANO: You will see that there was

MS. HALLORAN: Slide 32.

DR. FASANO: Can you bring me over there,

please?

(Slide.)

DR. FASANO: There is a fair amount of

variability. You see that, and there is some

overlapping at baseline.

156

If you go down -- keep going -- now, if

you can go up to 28, please?

(Slide.)

MS. HALLORAN: It is 32, I think.

DR. FASANO: You want 32? I thought that

you were talking about the variability of the

villus/crypt ratio. Is that what you are talking

about?

CHAIRMAN DURST: Yes.

MS. HALLORAN: Yes.

DR. FASANO: It is a little bit higher

than that.

Can you go higher?

MS. HALLORAN: Ah.

DR. FASANO: Stop here. I need 26.

(Slide.)

DR. FASANO: All right. You see here,

this is the variability. You see here that this is

the variability. These are the single points. If

there was somebody that was high right here

(indicating), and someone like here, these are the

celiacs. There was a continuum, so it is not that

157

there are people here, people there; it is a

continuum.

This is the standard deviation, and this

is the mean. Again, there is some variability but

not huge. There is much more variability in the

intraepithelial lymphocytes -- you can see this

scatter -- that are being monitored.

CHAIRMAN DURST: Suzanne.

DR. TEUBER: Suzanne Teuber. I would

assume, and this may be a completely incorrect

assumption, that in the population that is

following a gluten-free diet strictly, as those you

indicated you recruited, would actually be a subset

of patients who perceive themselves to be very

sensitive, and thus would have a higher motivation

to follow such a diet.

This would bring up in Italy what percent

of patients do comply with the gluten-free diet?

We heard about the extreme difficulties here and

the poor compliance rate. Is it better in Italy?

Would this mean that, perhaps, this population that

you recruited from really might be a good sensitive

158

population?

DR. FASANO: I think, and I'm paraphrasing

Joe Murray on this, that the compliance with the

diet is the results of many factors, some of them

diet there. Education I believe is at the top of

all.

It is not that you feel it to be more

sensitive or less sensitive. If you understand the

facts, if you understand the rules of the game, no

matter how you are perceived as being sensitive or

not sensitive, you know that you can't cheat. You

know that you need to start with that.

If you go to 10 doctors and they say all

the same things, "I'm sorry, you don't have an

alternative," then the level of confidence

increases. We don't have that here. We don't have

it, honestly.

Let's be honest. We have people, doctors,

that will tell you, "You know, you need to go on a

gluten-free diet." These are the teaching sheets

that were printed 20 years ago. "After three

months, go back on a regular diet. You're going to

159

grow out of it." What level of confidence do you

have?

Definitely, a study like here, like this

done here, will have a tremendous amount of bias.

Because who is going to do that? It will be only

the ones that are extremely compliant. The

population in Italy that is compliant -- in Italy?

I should not say in Italy, in Europe --

because they are like 10 or 15 years ahead of us in

this, because the level of awareness has been there

for quite a long time -- is pretty high.

They understand exactly what is at risk.

That is the reality of the story. It is more than

to be the people with high cholesterol, high blood

pressure and to be on medication because there is

much more flexibility there.

These people they understand that if they

don't comply the pay a price, and they do. The

level of frustration, particularly here, is that

they want to do that, the ones that understand the

game, but they can't because there is no way in

their current situation they can comply.

160

CHAIRMAN DURST: Doug.

DR. HEIMBURGER: Doug Heimburger. Would

you go to the next slide, please, after this one?

DR. FASANO: Sure.

(Slide.)

DR. HEIMBURGER: Does this graph include

the controls or only celiac patients?

DR. FASANO: These are only the celiacs.

DR. HEIMBURGER: Just out of interest, did

you test for this correlation in the controls?

DR. FASANO: Yes, it is the same. We put

it all together, yes. There is a strong

correlation. Again, if you conceptualize this

intraepithelial lymphocytes as, again, the first

folks to go there -- just two weeks ago, for

example, there was a paper in science in which they

claimed that the lymphocytes, they are called

gamma/delta, they are able to present antigens.

They can see gluten and they can start the

entire reaction, at least to the adaptive immunity

Th2 response to interferon-gamma, that will

translate in damage, i.e., to make the villi short

161

and the crypt deeper. That makes a lot of sense.

The more you have, the more cytokines you can use,

the more damage you have.

CHAIRMAN DURST: Dr. Fasano, will you be

around for the discussion this afternoon?

DR. FASANO: Yes. Yes, I will.

CHAIRMAN DURST: Because I think maybe we

will stop the questions.

DR. FASANO: I have my candy so I can't

leave you.

(General laughter.)

CHAIRMAN DURST: Okay. We will probably

move on so we don't go too far into the lunch hour.

Our next speaker is Dr. Pekka Collin. He

is a professor at the University of Tampere Medical

School in Finland. He will discuss retrospective

studies.

RETROSPECTIVE STUDIES

DR. COLLIN: Yes, good morning everyone.

I come from Tampere. You probably know where

Finland is and Tampere is a hundred miles north of

our capitol, Helsinki.

162

(Slide.)

DR. COLLIN: We at least in Tampere think

that is the celiac center of Finland, but maybe

somebody disagrees with that. We have a half a

million people around our hospital and now our

clinical prevalence of celiac disease is

approaching 1 percent. I think it is .7 at the

moment, so we have 1,000 patients with celiac

disease. Consequently, we have tried to examine

both the symptoms and the diet.

(Slide.)

DR. COLLIN: I had some specific issues

which I should address at this meeting, and they

are here. I should explain why we carried out our

retrospective analysis of the gluten content in our

gluten-free products; then, also, calculate what is

the significance of daily gluten exposure in this

small amount of gluten; and then, also, to discuss

is there some variability in the sensitivity of

people with gluten intolerance which has been

discussed here already many times. That should

also include patients who are taking with

163

starch-based, gluten-free products and who are

taking oats where we have a lot of experience.

(Slide.)

DR. COLLIN: I think that celiac disease

has been described very well by previous speakers,

so I will go straight into the point. However, I

will emphasize that now we are talking about parts

per million or 10 milligrams or 20 milligrams of

gluten intake.

(Slide.)

DR. COLLIN: In real life, if you have 100

patients with celiac disease, I think 90 percent of

them are taking 15 grams of gluten a day because

they do not know that they have celiac disease.

Only 10 out of 100, for instance, in the

U.S.A. I think know that they suffer from celiac

disease. Of the remainder 10, maybe 3 or 4 do not

follow a gluten-free diet strictly because they

don't care, or it is more likely because there are

not enough products when they are eating out or

eating in restaurants, and so on.

I think that is very important, that we

164

have a good choice of products. That is more

important than some parts per million in order to

achieve a good percentage of compliance.

The amount of threshold, I think it

started more than 10 years ago in Europe. The

celiac societies were very, very active in these

respects. From southern countries, some people say

that we are in northern countries poisoning our

people because they know that we are giving them

wheat-starch-based, gluten-free products.

On the other hand, our society, I think

they are very -- I don't find the right word -- but

I admire them because they said, "Please make the

study. Look at what we are now eating. Celiac

patients are the last who will have some extra,

unnecessary dietary restrictions, so please make a

study where you show whether we are now eating

safely or not." I think that was the background

for our so-called "retrospective study."

At that time we were quite relaxed. We

were not afraid that we are poisoning our people,

because we published a study where we showed that

165

in our patients we did not have, in treating

patients we did not have, any extra mortality and

even we did not have any extra risk of malignant

conditions at that time.

Then, we looked at what the Finnish

celiacs are eating. As expected, the majority of

them took wheat-starch-based, gluten-free products.

(Slide.)

We can also see that compliance was very

good. These patients they were invited, after 5 or

10 years on a gluten-free diet they were invited, a

cohort of those patients, both so-called

"sensitive" and not sensitive, and we can see that

only a small percentage of patients had dietary

transgressions. Although there were a few who

daily or twice a week or once a month had dietary

lapses, most people preferred to follow a naturally

gluten-free diet.

We also show that for these patients their

quality of life is good, and they did not have any

additional symptoms compared to the population. As

has been mentioned many times earlier, symptom is

166

not a very reliable objective sign of gluten

intolerance.

(Slide.)

DR. COLLIN: This is an example how

symptoms can be misleading. This is maybe a little

bit out of the topic, but I think this is very

interesting.

We ask family doctors to send us all such

patients who spontaneously reported that they get

symptoms after taking wheat or rye. The majority

of them had also on their own account tried to

avoid or withdraw these products from their diets,

and they experienced clear improvement in symptoms.

We thought that many of them had latent or

overt celiac disease, but to our surprise only 10

percent of people with a clear history of

intolerance to gluten had really celiac disease.

Then, there are some which we thought that they

maybe had wheat allergy.

When I was here yesterday I heard about

that. Yes, the diagnosis is so difficult, so I

hope that I don't have to discuss this in more

167

detail.

However, the majority of them, even with

sophisticated methods, they did not have any signs

of celiac disease and probably they have irritable

bowel syndrome. Hence, we cannot trust symptoms

even in the diagnosis of celiac disease.

(Slide.)

DR. COLLIN: Then, of course we have to go

to small bowel biopsy as they did also earlier. We

took a control biopsy after 5 to 10 years from

these patients who had been diagnosed with celiac

disease and who were asked to come to our hospital.

(Slide.)

DR. COLLIN: What we can see here is that

this is the same villous height/crypt depth ratio

which has been measured by, for instance,

Alessio Fasano. Here is our reference value for

people who have no suspicion of celiac disease.

They have come to endoscopy because of suspected

some gastrointestinal disorder, reflux symptoms or

dyspepsia.

We can see that in our long-term treated

168

patients, there is a 95 confidence interval, so it

was exactly the same as in our non-celiac people.

I could show also a similar slide of

intraepithelial lymphocytes, very similar. They

did not have extra intraepithelial lymphocytes.

We did not either have any so-called

"highly sensitive" patients with celiac disease.

We had some here who had not a complete recovery in

the mucosa.

After dietary inspection, it turned out

that all of these people are taking occasionally

gluten. Even once a month, I think that was in the

data, the histological recovery was not complete

Then, we had also here are the celiac patients

where the ratio was of course low.

Then, we had some short-term treated

patients, that means from half year to one year.

We show that the healing was not complete at that

time. From this slide we had two questions.

First, when we have a complete recovery,

are those patients still taking some small amounts

of gluten or are their products complete

169

gluten-free?

The second question was, When we have this

incomplete recovery, does it depend on wheat starch

or gluten contamination or is it normal life in

celiac disease? In other words, would the healing

be better if instead of wheat starch used, the use

of naturally gluten-free products?

To the first issue, Are those products

contaminated which have shown that our people are

doing well and their mucosal is healthy? It was

not surprising that most of naturally gluten-free

products had less than 10 ppm gluten.

However, I think it is very important to

realize that some of the so-called naturally

gluten-free products, they may be contaminated with

gluten, even quite high. All of these were

fulfilling the current European Codex standard.

If we go to the wheat-starch-based,

gluten-free flours, there were two with zero

gluten, and as expected most of them contained

trace amounts of gluten. Two had more than 100,

but the majority has less than 100. That was our

170

idea that maybe we can set the limit to 100 ppm.

When I had this slide and my conclusion in

Europe, one of the representatives of industry said

that he was disappointed because I am talking about

100 ppm, and I should have talked about the limit

of 200 ppm because it is much easier for them.

However, I said that we had too few

products here to assert that 200 ppm would be

recommendable. I think I will remind you that

90 percent of our celiacs have used this product

for 40 years or even more, and we have

biopsy-proven results from that so-called challenge

from 5 to 10 years. The mucosal recovery, as I

said, was perfect.

(Slide.)

DR. COLLIN: We also looked at how much

they did use those flours. Maybe somebody who has

taken gluten-free products can know that they are

not necessarily as good as wheat, baking with

wheat.

Nevertheless, here are how the patients

used these products. There was no difference

171

between wheat-starch-based products or a naturally

gluten-free diet. The average was 80 grams, and

the majority took less than 150 grams as you can

see here. There was no correlation between the

villus damage and the amount of data used of loss.

(Slide.)

DR. COLLIN: From here we come to this

conclusion, which maybe you have seen this kind of

table in Alessio Fasano's presentation. Provided

that we set the limit to 100 ppm, and provided that

each of these products also contained the maximum

amount allowed, when patients are taking 100 grams

of those or 200 grams of those the gluten

contamination is from 10 to 20 milligrams.

If you look at Fasano's results and if you

look at some earlier, small studies -- even the

Catassi study, which was referred to, and some

smaller studies made by Sturgis and so on -- I

think we are very, very safe here at the 100 ppm.

I think also that our clinical experience will show

that the same.

Of course, this is not a prospective

172

study, and we did not have any control group, and,

unfortunately, we did not have many patients who

have clear dietary restrictions, so we cannot make

any statistical analogies between those who are --

what is the word -- cheating with their diet and

who are not. However, I think with this kind of

system, we can treat our patients and have good

compliance.

(Slide.)

DR. COLLIN: If I can, go to the issue

whether patients are more sensitive or

hypersensitive patients with celiac disease. When

we look at those patients, we can see that their

mucosal recovery takes place in a different way in

different people. That has been very well shown in

some challenge studies. Where earlier it was

customary to accept diagnosis, we have once again

to challenge the patients to gluten-free diets and

look at if there will emerge new villus atrophy.

We show that in some cases it took two

months or one month to see a mucosal relapse, but

in some cases it took two or three years. Our

173

record is 15 years. Fifteen years with normal diet

and earlier diagnosed celiac disease, after 15

years a mucosal relapse occurred.

Here we can see that in the short-term

some people do not respond, and you could think

that these might be so-called "hypersensitive." If

we give enough time and the patients are truly

following the gluten-free diet, which means that we

must be really accurate that they do not take wheat

at the same time, I think in the long-term we have

almost complete recovery. We did not have any

so-called "hypersensitive."

I think patients with refractory sprue

they can be very sensitive because they do not

respond at all to celiac disease, but that is a

different issue. It is probable that even zero

gluten would not help them. There is something

wrong in their gut. Probably the diagnosis has

been made too late, and it does not recover any

more. I think that refractory sprue is outside of

the topic of this day.

Also, we were discussing with Peter Chen,

174

when he wrote to "Gastrointestinal Endoscopy" that

complete mucosal recovery is not possible, and we

had a very friendly, friendly discussion in the

pages of that journal. However, we said that it is

possible when we have a good choice of products and

people also outside the home know what celiac

disease is what this means for the patient with a

gluten-free diet.

The second issue in my slide was that

could it be that the mucosal healing would be more

rapid in those who are on a naturally gluten-free

diet than in those who are maintaining

wheat-starch-based, gluten-free products?

Here, we carried out a randomized

prospective study of one year in newly detected

celiac disease patients. If we look at the villus

healing here and here, villous height/crypt depth

ratio, there were no differences between these two

groups. We can also see that in one year, you

cannot achieve the limit of three, which is

considered normal.

Similarly, when we look at intraepithelial

175

lymphocytes, they decreased in a similar way in

both patients. At that time, unfortunately, we

could not measure what was the exact amount of

gluten these patients were taking; we did not have

methods. We can assume they took those same

products which were mentioned in my last slide

which contained trace amounts of gluten but not

more than 100 ppm.

(Slide.)

DR. COLLIN: If I may say some words about

oats. It was in Finland, the first publication.

After that, very soon it was accepted for celiacs

in Finland that they may use oats. At the

beginning we were very careful. We followed up

with them each month and looked at what to do, but

now we do not do it anymore.

We made a question out, too. We sent a

question out to members of the Celiac Society, how

do they appreciate oats. As you can see, they like

about the permission to eat oats.

Almost all said that it is a very

significant part of every day gluten-free diet in

176

terms of tasty and low lost. They even thought

that it is healthy, diversifies the diet, and we

have a good availability in Finland of oat

products. I understand that maybe in some

countries oat is not so important.

Some might say that in Finland they are

not eating good, so maybe people in Italy do not

operate yet in the same manner as in Finland, but

we can discuss it.

(Slide.)

DR. COLLIN: Here are how our people have

now used oats, the majority of patients -- not

great amounts, it is only 20 grams, 15 or 20 grams.

There, most of the studies are about approximately

50 grams, so less than in those randomized studies.

Some people do not prefer oats, and that

is the same thing in people in general not only in

celiac patients. Some of them had stopped, and the

reason is that they had developed symptoms. Some

even got a rash, basically dermatitis

herpetiformis. We do not have any proof that the

reason for stopping would be that they

177

simultaneously had mucosal damage. Usually, the

mucosa is good even though the patient has stopped

the diet.

The rest, in dermatitis herpetiformis, we

also saw that even in patients with no oat diet, so

even they may have a temporary rash. There are

some clinical relapses in patients with dermatitis

herpetiformis also.

It is excellent to study these questions,

because we can change the subjective symptoms quite

rapidly to objective science, count the number of

blisters, for instance.

(Slide.)

DR. COLLIN: We also looked at the quality

of life in patients with oats. Actually, there was

no change, difference, compared to patients with no

oats. This also was a prospective, randomized

study in treating celiac disease.

Interestingly, those patients who were

taking oats, they reported more symptoms of

diarrhea, which was statistically significant.

They also reported more constipation, which was not

178

significant. Even in these patients, we did not

have any mucosal deterioration.

From this we learned that if we start on a

gluten-free diet with oats, we must inform the

patient that "You may have symptoms after this. If

you have symptoms, why continue. But it is

improbable that we have done any harm to your

small-bowel mucosa.

We also saw that those who were taking

oats had a little bit more intraepithelial

lymphocytes, not CD3 lymphocytes, which we have

discussed today, but gamma/delta lymphocytes.

The gamma/delta lymphocytes were a little

bit increased in the oat group. I cannot explain

the reason for that, and that has not been

published elsewhere -- but that is the fact.

(Slide.)

DR. COLLIN: Here are my conclusions to

the questions which I was asked to answer. Maybe I

also specific questions which you have, specific

issues which you have to address in the final

report.

179

If I may say something about the

subpopulation, the most highly sensitive people, I

think such people of course may be, but eventually

they have good mucosal recovery, provided that they

follow a gluten-free diet. The majority of these

highly sensitive patients are probably such people

who have advertent or inadvertent gluten intake.

We can also remember that even if it

happens, the consequences are not disastrous,

because they do not develop an anaphylaxis aspect

as do people with peanut allergy as we heard today.

We can quite easily detect these highly

sensitive, if we after the diagnosis, one year

after the diagnosis, take a small-bowel biopsy and

look at whether there is an improvement in the

mucosal architecture. If there is not, we must

consider that they may be very sensitive, but

usually they do not follow the gluten-free diet.

About the risk of malignant diseases, I

think the whole literature tells that those people

who are at an increase risk of malignant lymphoma,

their diagnosis has been made too late. They

180

already have lymphoma when the symptoms of celiac

disease appear and when they get the diagnosis of

celiac disease, or they have had dietary

transgressions for a prolonged time.

Of over 1,000 patients I have seen during

the 15 years, I have seen one patient who has

developed lymphoma after being 5 or 10 years on an

apparently gluten-free diet. The risk of these

severe complications in those small daily intake is

probably very low. Even our new data show the

same, which is now published only in abstract.

Similarly, the mortality, it depends on

those patients who come to the hospital together

with the diagnosis of celiac disease and later,

usually within six months, we can see that they

also have lymphoma.

(Slide.)

DR. COLLIN: What about the oats? Here I

summarized some studies. Those with plus signs

they are those who have shown that oats have no

adverse effect on the mucosa. I think nearly

almost all of these studies are randomized,

181

open-randomized. They have a control group with

non-oat. We have hundred of patients who seem to

tolerate oats.

But I think I would be stupid if I did not

see also those two papers and patients who are

sensitive. I cannot close my eyes from the

results, because Don Kasarda told the data very

convincingly.

I don't know who they are. Maybe there

are some who really develop villus atrophy after

taking oats, but that must be an extremely rare

condition. Because, as you see, we have so many,

many patients who are taking oats, and we have not

seen this phenomenon.

Still, we must be careful, and we must be

careful because patients with oats may develop

symptoms. If everything does not go well, of

course we stop the use of oats. However, we must

be aware of that, that maybe there are some rare

patients where it acts the same as gliadin for most

people with celiac disease.

I don't know whether these, my results and

182

recommendations, can be applied in the United

States but that is how we are doing now. Our

celiac society is very happy because we said that

you can continue with starch-based, gluten-free

products.

Thank you very much.

CHAIRMAN DURST: Thank you.

QUESTION AND ANSWER SESSION

CHAIRMAN DURST: Do we have questions?

Suzanne.

DR. TEUBER: Suzanne Teuber. My question

relates to the applicability of the diet parameters

to the United States dealing with how much

gluten-free flour do people in different parts of

the world ingest, if they were to have the option

of knowing that something was truly, truly

gluten-free.

You talk about 100 parts per million. It

was your data that came up with the 80 grams a day

that people ingest. I'm wondering -- you know, we

are not setting any level here today -- in terms of

United States' folks, I have no idea how that would

183

apply. Would this be a safe level for them? Or,

here, would people be preferring to adjust much

more? Do you have any input on that?

DR. COLLIN: I think there is not much

data on that, how much people really in different

parts of the world are really using wheat or other

flours which may be harmful to patients with celiac

disease.

I think that this is a subject for further

studies. Maybe somebody here knows how much celiac

patients are here using gluten-free flours, but I

don't know. I have not seen any publications about

this issue.

CHAIRMAN DURST: Any other questions?

DR. McBRIDE: Margaret McBride. Did I

understand correctly that gluten-free in Finland

means 100 part per million?

I guess for me, as I'm thinking about it,

maybe part of the difference between the two

studies, aside from the obvious

retrospective/prospective, et cetera, is that in

Italy the gluten-free diet did contain some,

184

although very little at 20 parts per million

gluten, in addition to what was administered.

I don't know if there is an estimation of

how much that would be. I'm also thinking maybe

there is more interest in pasta in Southern Europe

than in Northern Europe.

DR. COLLIN: I think that today we have

given the formal Codex standard which says that

200 ppm is okay, but of course we need to

reconsider that.

I think that in the whole of Europe there

will be two limits, that is the 20 milligram which

can be used in the highly sensitive people, but in

the majority of people it is 100 ppm.

Of course, there is a problem with

labeling, how we should label that. We cannot say

that it is "low gluten," because then people will

use that. That is our problem.

What our recommendation is, is that maybe

the majority of people with celiac disease can

tolerate products which are under the limit of

100 ppm.

185

DR. KELLY: Ciaran Kelly. I wonder in

terms of compliance with the diet and acceptance of

the diet, is there a big difference between 20

parts per million or 100 parts per million from the

perspective of the palatability of the food?

DR. COLLIN: I think the important thing

is, at least the industry in Europe says, that, if

we go to very low level, there are not so many

alternatives for gluten-free products, which again

may result in that general compliance will be worse

than I have shown now.

How the products, how they--? I think

that those wheat-starch products, I think they are

very tasteful. Does it depend on the small

milligrams of gluten or not? I don't know. But,

as can be seen, most of the people are preferring

those products instead of naturally gluten-free.

CHAIRMAN DURST: Anyone else?

(No verbal response.)

CHAIRMAN DURST: If not, thank you,

Dr. Collin.

Our final speaker for this morning is

186

Rhonda R. Kane from the Consumer Safety Office of

CFSAN, FDA, on international perspectives on

gluten-free.

INTERNATIONAL PERSPECTIVES ON GLUTEN-FREE

MS. KANE: Good afternoon. My name is

Rhonda Kane. I am with the Food and Drug

Administration, and I was asked to present

information to the Food Advisory Committee about

how the term "gluten-free" is defined in other

countries and the basis for those definitions.

(Slide.)

MS. KANE: My presentation today will

focus on four examples of international or national

definitions of the term "gluten-free" that apply to

labeled packaged foods.

The first two examples I will be

discussing pertain to Codex Alimentarius and they

include, the first one, Codex Standard 118-1981,

which pertains to the Codex standard for

gluten-free foods that was established in 1981, was

amended in 1983 and is in effect today; and, two,

the Proposed Draft Revised Standard for Gluten-Free

187

Foods at Step 7 that is now under consideration by

the Codex Committee on Nutrition and Foods for

Special Dietary Uses as a replacement for the

current standard.

For ease in my presentation, I am going to

refer the Codex Committee on Nutrition and Foods

for Special Dietary Uses simply as the "Codex

Nutrition Committee."

In the early 1990s, members of the Codex

Nutrition Committee agreed that developments in the

characterization of gluten on studies on gluten

tolerance warranted a revisiting of the current

standard and an updating of it.

The current proposed standard has

undergone several revisions and is now at Step 7 of

an 8-step process pending resolution of certain

issues including what method of detection is going

to be used for gluten and the results of gluten

threshold studies in celiac patients. The Codex

Nutrition Committee will be meeting in November

2005, and will be discussing the proposed standard.

The third example of gluten-free that I

188

will be discussing is found in Canada's Food and

Drug Regulations at Section B.24.018. It became

effective on May 1, 1996.

Lastly, I will review the definitions of

both "gluten-free" and "low-gluten" that are found

in Clause 16 of Standard 1.2.8 of the Australia

New Zealand Food Standards Code, and I will also

discuss the definition of gluten found in Clause 1

of that same standard.

(Slide.)

MS. KANE: The current Codex standard that

is in effect today defines "gluten" as "Those

proteins commonly found in wheat, triticale, rye,

barley or oats to which some persons are

intolerant."

The current standard further defined the

term "gluten-free" to mean that "The total nitrogen

content of gluten-containing cereal grains used in

the product does not exceed 0.5 gram nitrogen per

100 grams of the cereal grains on a dry weight

basis."

(Slide.)

189

MS. KANE: The current standard states

that it does not apply to foods which in their

normal form do not contain gluten. Gluten-free

foods are defined according to two categories,

those that contain the cereal ingredients -- wheat,

triticale, rye, barley or oats or their

constituents, which have been rendered gluten-free

-- or those foods in which any ingredients normally

present that contain gluten have been substituted

by other ingredients that do not contain gluten.

(Slide.)

MS. KANE: In comparison, the Codex

Proposed Draft Revised Standard for Gluten-Free

Foods at Step 7 defines "gluten" to be "The protein

fraction from wheat, rye, barley oats or their

crossbred varieties and derivatives to which some

persons are intolerant and that is insoluble in

water and 0.5 molar solution of sodium chloride."

You will see that in this definition and

in others that are occurring in the proposed

standard information within brackets is intended to

indicate that that information is pending

190

additional discussion at the Codex Nutrition

Committee. Their next session meets in November

2005.

The Proposed Standard also defines the

term "Prolamin" to mean "The fraction from gluten

that can be extracted by 40 to 70 percent aqueous

ethanol." This definition specifically identifies

the prolamins: gliadin from wheat, secalin from

rye, hordein from barley, and avenin from oats.

(Slide.)

MS. KANE: The Proposed Standard also

states that it applies to those foodstuffs and

ingredients which have been especially processed or

prepared to meet the dietary needs of persons

intolerant to gluten.

Therefore, this parameter is similar to

the one for the current standard in that neither of

the two standards, the current and the proposed,

would include foods that are naturally or

inherently free of gluten.

The proposed standard also identifies

three categories of gluten-free foods where their

191

definitions specify certain limits on their gluten

content.

(Slide.)

MS. KANE: In the first proposed category,

gluten-free foods consisting of ingredients which

do not contain any prolamins from wheat or all

Triticum species -- rye, barley, oats -- or their

crossbred varieties cannot have a gluten level that

exceeds 20 parts per million. Again, you will see

that "20 parts per million" is within brackets,

therefore, this number is pending.

(Slide.)

MS. KANE: This proposed definition also

specifically cites three examples of grains within

different species of Triticum, they are: spelt,

kamut, and durum wheat.

Although triticale is not one of the

grains that is identified within the definition by

its name, it is included because it is a crossbred

hybrid of wheat and rye.

In the second proposed category of

gluten-free foods, they are those consisting of

192

ingredients from wheat, rye, barley, oats, spelt or

their crossbred varieties that have been rendered

gluten-free and cannot have a gluten level that

exceeds 200 parts per million. Again, "200 parts

per million" is cited in brackets, and it is

therefore pending.

(Slide.)

MS. KANE: In the third proposed category,

gluten-free foods consisting of any mixture of the

ingredients as described in the previous two

categories, cannot have a gluten level that exceeds

200 parts per million. Again, "200 parts per

million" is cited in brackets and it is pending.

(Slide.)

MS. KANE: Based upon my reading of the

session reports for the Codex Nutrition Committee

and related documents, it appears that the

rationale for including two levels, the 20 and 200

parts per million, in the definition of gluten-free

foods was to accommodate different points of view

of the Codex member countries that thought there

should be a different level of gluten based upon

193

their experience with their populations, what would

be adequately protective.

There were some countries that believed

either the lowest limits of detection or 20 parts

per million would be most protective of those that

are very sensitive to gluten.

Twenty parts per million was considered a

practical limit to make it more feasible for

industry to produce gluten-free foods in that

category.

Other countries believed that the higher

level of 200 parts per million would be

appropriate, because they had experiences with

citizens in their country that had celiac disease

where they had been consuming wheat-starch-based

products for years without harm, and they enjoyed

them.

The 20 parts per million level would

essentially prohibit the inclusion of those

wheat-starch-based products. Therefore, it was a

compromise, the low limit and the high limit, and

they realized they could create some confusion on

194

the part of the consumer.

I also want to point out that the proposed

definition of gluten-free foods specifically cites

that whatever detection method is used it should

have a detection limit of at least 10 parts per

million gluten in the product on a dry weight

basis.

(Slide.)

MS. KANE: The next definition I will

discuss is that found in Canada's Food and Drug

Regulations at Section B.24.018, and it states:

"No person shall label, package, sell or advertise

a food in a manner likely to create an impression

that is a gluten-free food unless the food does not

contain wheat, including spelt and kamut, or oats,

barley, rye, triticale or any part thereof."

(Slide.)

MS. KANE: Canada's definition of

gluten-free prohibits the use of derivatives or

constituents of any of the cited grains.

Therefore, wheat starch would not be allowed in a

product that was labeled gluten-free.

195

It is my understanding based upon

communication with staff who work with Health

Canada and the Canadian Food Inspection Agency,

that the definition that Canada is using was

developed using a rule-making process, but they

closely coordinated with the Canadian Celiac

Association in the parameters for this definition.

Canada underwent a rule-making process

similar to the one that we use in the United States

where they reviewed the relevant scientific

literature, they published a proposed rule,

considered comments before it went final, and they

determined that back in the mid-1990s that there

was insignificant or insufficient I should say

scientific evidence to support establishing a level

that would be safe for all celiac patients.

(Slide.)

MS. KANE: In the last definitions of

gluten-free that I will be discussing, in the

Australia New Zealand Food Standards Code, first,

in Clause 1 of Standard 1.2.8 of the "Code," which

I will refer to simply as that rather than

196

repeating that long name, it defines "gluten" as

"'The main protein in wheat, oats, barley,

triticale and spelt relevant to the medical

conditions, Coealic disease and dermatitis

herpetiformis.'"

It also defines in Clause 16 of that same

standard the terms "gluten-free" and "low gluten."

(Slide.)

MS. KANE: "Gluten-free" is defined as

those foods that contain no detectable amount of

gluten. They also cannot contain any oats or their

products or any cereals containing gluten that had

been malted or their products. It has to meet all

of those three criteria not just one.

In addition, their Code defines the term

"low-gluten foods" to mean those that contain no

more than 20 milligrams of gluten per 100 grams of

food. Now, although not stated in the Code as

such, this level of gluten is equivalent to 200

parts per million.

(Slide.)

MS. KANE: It is my understanding based

197

upon communication with Food Standards Australia

New Zealand's staff that they also underwent a

rule-making process where they proposed these

definitions for "gluten-free" and "low-gluten"

before they went final.

They did a review of the relevant

scientific literature. They considered public

comment, and they also consulted with experts in

the appropriate fields to develop the definitions

that are in effect today.

In addition, the fair trading laws in both

Australia and New Zealand were interpreted as

prohibiting the term "gluten-free" from being used

with any foods that contained any detectible amount

of gluten.

(Slide.)

MS. KANE: Further, the definition of

gluten-free was influenced by a lack of reliable

analytical methods to detect gluten in oats and

malted cereals. Essentially, their definition says

not only no detectible amount of gluten, but no

oats or other products, no malted cereals

198

containing gluten and their products because of

this limitation of analytical methods.

(Slide.)

MS. KANE: The Code includes two

definitions, "gluten-free" and "low gluten," to

provide citizens who have celiac disease a choice

between which level of gluten-containing foods they

want to consume based upon their individual gluten

tolerance level and the advice of their healthcare

provider.

In closing, I would like to sincerely

thank the staff that I consulted with at Health

Canada, and the Canadian Food Inspection Agency, as

well as Food Standards Australia and New Zealand.

With that, I will take any questions.

CHAIRMAN DURST: Thank you very much.

QUESTION AND ANSWER SESSION

CHAIRMAN DURST: Does the Committee have a

question or comment?

Erica.

DR. BRITTAIN: Erica Brittain. I guess I

find it appealing the idea of the two levels, just

199

as a comment, in the last one you cited. This

might be applicable to the allergy situation as

well.

CHAIRMAN DURST: Okay. Anything else?

Mark.

DR. NELSON: Mark Nelson. Did your

contacts in Australia, New Zealand and Canada give

any indication that they might change their

definitions or their categorizations if there were

more work done on thresholds, if that data based

changed?

MS. KANE: That sort of conversation

didn't occur between me and them, but I would think

because they are government agencies, just like FDA

is, if there were newer information on the horizon,

they would probably consider it. Whether they

would go through the rule-making process and change

it, I guess they would base it on the needs of

their own populations.

DR. NELSON: I guess the opportunity for

-- this is Mark Nelson again -- two categories does

have some attractiveness. I guess at Codex it is

200

going to be gluten-free and really gluten-free.

(General laughter.)

CHAIRMAN DURST: Suzanne.

DR. TEUBER: Suzanne Teuber. To your

knowledge in talking with these folks, have their

been any consumer-preference studies or behavior

studies completed or underway with how the celiac

disease patient is using these standards in terms

of their overall intake?

MS. KANE: I don't have personal knowledge

of that. However, the Canadian Celiac Association

is very supportive of Canada's definition of

gluten-free. Because they were instrumental in

helping develop it, so they were very supportive of

it.

CHAIRMAN DURST: Yes.

DR. McBRIDE: Margaret McBride. Do I

understand from your slides about the Codex

proposed changes that the term "gluten-free" would

be applied both to those foods that contained none

of the products in question and are lower than 20

parts per million, and to those foods that are

201

wheat-based and gluten has been removed but would

contain up to 200 parts per million? I realize the

numbers are in question.

MS. KANE: Right. They do not apply to

the term. They don't want the term "gluten-free"

to apply to naturally gluten-free foods but those

that have been specially processed or prepared

where the formulation has been controlled.

There is a substitution of ingredients or

a removal of gluten from ingredients. It would

cover categories that are wheat-starch-based. That

is where the 200 parts per million definition is

coming into play.

Member countries did not want

wheat-starch-based products to be excluded from

being called gluten-free, if there was only one

definition of 20 parts per million. That is why

they compromised and had the two levels that would

apply.

DR. McBRIDE: A follow-up. Would I assume

that they would then be called something different,

or would we be expecting the consumer --

202

(Simultaneous discussion.)

MS. KANE: No. Right now, as it stands,

they are saying one definition "gluten-free" to

apply to three categories of gluten-free. However,

that could change.

Now, keep in mind all of this

is pending. It is at Step 7 of an 8-step process.

I know there is a Working Group, the Prolamin

Analysis and Toxicity Group. That information will

come into play. These levels are not definite and

they could change.

If both of those situations or all three

were called gluten-free, then we would have to

expect that the consumer who felt that they were

very sensitive and wanted truly a very low level,

below 20 parts per million, would have to read and

understand the names for the various grains,

et cetera, that would be on the ones where in fact

products that at least one time had contained

gluten were used.

I understand that, and the report I cited

on my second slide, the "ALINORM Report" is the

203

latest one, to my knowledge, that contains the

language of the current proposed standard at

Step 7. It doesn't go into those details about how

it might be labeled alternatively or what

additional information it would include. You're

right, it does create confusion. How would you

know if it is 20 parts? How would you know if it

is 200 parts?

That issue was brought up in some related

documents, but it is not found in the latest

session report. However, you're absolutely right.

DR. NELSON: This is Mark Nelson. I just

want to address that question about the Codex

label. There is a separate committee, Codex

Committee on Food Labeling, and these definitions I

would expect would ultimately be referred to the

Codex Committee on Labeling to address the issue

you have just raised about the potential confusion.

CHAIRMAN DURST: Suzanne.

DR. TEUBER: Suzanne Teuber. I also see

an issue about cross-contamination problems with

foods that you wouldn't expect to contain gluten

204

and yet might contain contaminants because some of

these, the rules that you are talking about, really

don't address that.

Do you have any information on that, like

say, corn that may be processed in a place that

also has processed wheat? It really would be

beneficial to the consumer if it were to undergo

testing and have a specific label, and yet these

other definitions in other countries don't seem to

cover that all. It would probably just come out

with no statement. Is that a correct

interpretation?

Or, actually maybe, Dr. Nelson--?

DR. NELSON: I think in Europe and Codex

also has a standard for good manufacturing

practices; the Europeans have the equivalent. I

think the issue there would be the responsibility

of the manufacturer to maintain good manufacturing

practices and prevent as much possible that cross

contact.

CHAIRMAN DURST: Marc.

DR. SILVERSTEIN: Marc Silverstein. Would

205

you clarify the categories of foods to which this

would apply? I would like you to, because I'm not

sure I understood the criteria exactly. If a food

has multiple ingredients, does this apply to all of

the ingredients in the food?

This is packaged and labeled food. One or

the major ingredient may be a food which in its

normal form does not contain gluten, yet there

might be other ingredients perhaps mixed in with it

that would.

Would it be that it applies to a labeled

package food which any of the ingredients contain

gluten, or would it be just the major ingredient

does not contain gluten and there might be some

additive or some other component ingredient?

MS. KANE: It is my understanding it would

apply to all ingredients. It would be selectively.

If a packaged food that is labeled gluten-free, it

would have to conform to the proposed. Of course,

again, it is proposed so it is not a done deal.

However, there are categories going back.

Can we go back? Can you reverse it back.

206

It is probably more towards the front. Okay, that

one right there.

(Slide.)

MS. KANE: That is the first category

consisting of ingredients. It doesn't say primary

ingredients. It means ingredients. That is how I

understand it. Keep in mind I've never been a

member of the U.S. delegation to a Codex Committee

meeting. I do not have firsthand knowledge of the

discussions. It is only based on my reading of

their session reports and related documents. The

way that is written I would interpret that to mean

all ingredients. Maybe someone who has attended

the Codex could speak to that?

CHAIRMAN DURST: Mark.

DR. NELSON: Mark Nelson. I think

everybody would interpret that as all ingredients

not just the main ingredients but including the

minor ingredients, flavors, spices, and so on.

I can just talk a little bit about my

experience in the food industry. I have worked

both for packaged goods companies but also

207

suppliers to packaged goods companies.

They look at it very carefully to find out

what the subingredients might be in, say, flavors

or an additive or carriers or something like that.

I can assure you, being a supplier to companies

like Nestle or Kellogg's or Kraft, we have to

provide a fairly substantial dossier to them for

every ingredient we supply them to deal with issues

like allergens and gluten levels as well. The food

industry itself does take this very seriously.

CHAIRMAN DURST: Soheila.

DR. MALEKI: Soheila Maleki. I guess this

is more a question. It seems to me that based on

what we have seen on some of the slides you've

shown today that there really isn't good analytical

method to be able to determine.

For example, the nitrogen content, you

could measure every protein in there and you could

weigh overestimate the amount of gluten. Measuring

gluten in the insoluble water fraction, that seems

to be, again, if you can solubilize it. If you

can't really detect it, okay.

208

DR. NELSON: I'm sorry, you may have to

start over. Sorry about that.

(General laughter.)

DR. MALEKI: It is kind of a question.

Based on this, I don't think there is really an

analytical method that can make you comply to this,

so how does this work? How are they going to

enforce it?

MS. KANE: Keep in mind that the nitrogen

definition of gluten is the current one. They are

proposing it be defined as the protein fraction for

wheat, rye, barley, et cetera, to which persons are

intolerant and it is insoluble in water and a

0.5 molar solution to sodium chloride.

However, there is an analytical method

component of a standard, and that is pending

because they were talking about the R5 Mendez

method, ELISA. They knew that they would have to

have a method that was sensitive enough, reliable,

accurate and would detect the types of proteins

that they are talking about in their definition.

That is going to be, I'm assuming, part of

209

the discussion at the next Codex meeting is to

bring that information about the methodology into

play, because those were the two components, the

methodology and threshold levels. Those are the

two areas needed to be worked out, and so I think

that is going to be the crux of the discussion at

the next Codex meeting.

DR. MALEKI: I just wanted to make a

comment as a follow-up.

CHAIRMAN DURST: Oh, okay.

DR. MALEKI: I'm Soheila Maleki. It seems

like the antibodies, the R5 kit again doesn't

detect gluten it detects gliadin. Maybe Steve can

help with that somewhere along the line.

All right, go ahead.

DR. CALLERY: Pat Callery. If the

analytical part can be worked out, which I think it

can. I wonder if there is an analogy here with

caffeine where we have caffeine-free sodas and

such, which we expect to have no caffeine, and

coffee that is decaffeinated that does have

caffeine in it. The word is not very pretty,

210

"deglutinated."

There may be an analogy that says when it

is gluten-free it is truly gluten-free and when it

is deglutinated, then there is a perhaps 20 parts

per million or something, whatever the standard

would be. That might be easier to understand.

CHAIRMAN DURST: Dick Durst. You

mentioned that the next meeting is in November of

this year. Do you get the sense that they will

finalize the document at that point?

MS. KANE: Oh, I wouldn't venture to say

that at all. I don't know, and I don't know how

close. Again, I've never been involved in their

meetings, and there is an eight-step process. They

could go back and revisit the issues; they could go

forward, and then it could advance. However, I

don't have a clue.

DR. NELSON: This is Mark Nelson. Even if

they did adopt it at the committee meeting, it

would then have to be forwarded to the overarching

body, which is the Codex Commission for them to

adopt it, and that will be next July.

211

CHAIRMAN DURST: Thank you.

Any further discussion?

Jean.

MS. HALLORAN: I think everyone should

realize that Codex standards are not biding on

anybody.

CHAIRMAN DURST: Okay. Thank you, Rhonda.

MS. KANE: You're welcome.

CHAIRMAN DURST: We will take our lunch

break. We are about 15 minutes over, but I think

we have sufficient time to reconvene at 2 o'clock.

Marcia, do you have anything?

MRS. MOORE: No.

(Luncheon recess.)

212

A F T E R N O O N S E S S I O N

CHAIRMAN DURST: We will reconvene for our

afternoon session.

It turns out that we haven't been able to

locate our first speaker, Steve Gendel, but we will

go on then to the public comments portion.

PUBLIC COMMENTS

CHAIRMAN DURST: Since today we have only

five signed-up speakers, we are going to give them

5 minutes instead of the 3 minutes that we used

yesterday. Hopefully, all of our speakers are

here. The first one is Alice Bast from the

National Foundation for Celiac Awareness.

(No verbal response.)

CHAIRMAN DURST: She is not here; okay.

Our second speaker is Elaine Monarch from Celiac

Disease Foundation.

MS. MONARCH: Good afternoon. I was

slightly unprepared to make a statement until I was

called on earlier today, and I am more than pleased

to do so.

My name is Elaine Monarch. I am the

213

founder and director of the Celiac Disease

Foundation, a national organization for individuals

with celiac disease and dermatitis herpetiformis.

Our offices are in Los Angeles California.

I am pleased to thank several of my

medical advisory board for making their appropriate

presentations today. I want to thank this

Committee for the opportunity to say a few words,

and for the hard work that you are doing on behalf

of all celiacs.

On behalf of the Celiac Foundation, I am

an active participant in creating more awareness of

this disease. As a member of the NIH Planning

Committee for the 2004 Consensus Conference, I was

hands on in the awareness process, and I am still

involved in getting the message out to the medical

community. I am also a member of the DDNC, the

NDDIC, and the American Celiac Disease Alliance.

It sounds like alphabet soup.

Oh, by the way, I am a celiac. I am a

typical celiac. I was not diagnosed as a child. I

was told that I was a banana baby, that I would

214

outgrow whatever stomach distress my parents said I

had. I was diagnosed when I was 40. I fit right

into everybody's statistics for not being diagnosed

appropriately.

As validated by the 2004 NIH Celiac

Disease Conference, celiac disease affects 1

percent of the total population in the United

States. We have heard today that celiac disease is

the only digestive disease that we know the trigger

for, and we call that trigger "gluten."

It is also the only digestive disease that

doesn't require pharmaceutical intervention. It

can totally be controlled by the strict adherence

to a gluten-free diet.

Adhering to this diet or lifestyle is not

as easy at sounds as you have heard here today and

yesterday. For example, there are limited choices

that I will have later today as I wait an hour and

a half at the airport for my plane. I could

probably find drinks, possibly a banana at

Starbucks, and very few other food choices.

I feel very fortunate that all I have to

215

eliminate from my life is gluten, yet there is no

standard for how much is too much, and that is what

I am hopeful will be the outcome of this meeting.

The simple casual snacking, something that

most of the population take for granted, is not so

for me. We need to examine everything that we

ingest. There is a wheat protein in everything

from Campbell's soup to licorice.

In today's busy society, fastfoods have

become a way of life for most people, convenience

foods. We talk to people on a daily basis in our

office, they are in a quandary of what to eat.

Fastfoods, sticking something simply in your mouth

at a cocktail party at somebody else's home is not

an option for a celiac.

There are as many stories in the celiac

community as there are diagnosed celiacs and those

yet to be diagnosed, and a broad range of

sensitivity. We are relying on this Committee to

supply our community with reliable, evidence-based

guidelines so that the majority of us can live the

gluten-free lifestyle to its fullest.

216

This past summer I am very pleased to say

-- or this summer our organization sent 12 celiac

children to camps across the United States where

food wasn't an issue.

We put the word out through the Internet

through our newsletters and our fellow celiac

organizations that we had the opportunity to

provide this camping experience for these children.

We asked them to please supply us with essays.

Twelve essays came in. We were never

going to turn anybody down. Twelve essays came in

from 12 children. Their ages were between 8 and

14. Each essay focused on food.

They were afraid to eat at camp. Their

parents would send for them for their other camping

or overnight experiences. They were afraid that

whoever was in the kitchen was going to serve them

improperly.

When you take a gluten-free waffle out of

a package, you have no idea if it is gluten-free or

not gluten-free. If you took two, square waffles

out of a package, is one gluten-free and one not?

217

You would have no way of knowing.

Every one of these 12 essays was based

around the fact that food was an issue for these

children. They didn't want to be different. They

didn't want the camp to run out of food. They

didn't want the camp to say, "Oh, Joey, this is

your meal."

We sent these 12 children to camp. We are

now just starting to get replies from the camps.

The smiles on the photographs go from ear to ear.

They had the best experience, because they could

experience the camping experience to its fullest

without the fear factor of food or being sick.

They weren't different; they were just campers.

All celiacs are totally dependent on the

food industry's manufacturing processes, practices,

and the accuracy of labels. Diligent label reading

is what we do. Yes, it does take us a little

longer to go through the food store. Yes, you do

have to read the label every time you buy a

product.

Warnings that foods are made in facilities

218

that also manufacture foods that may be toxic to

us, like the inclusion of wheat on a food label, is

going to be extremely helpful. We see the word

"wheat" and we know that we don't have to read any

further.

Patient compliance will improve when there

is a reliability on a food label. I think

compliance is low now because people aren't sure,

and they might as well cheat, because they are not

too sure if what they're eating is safe or not.

Food is truly our drug of choice. The

decision of this Committee will impact the quality

of life of 1 percent of our total population. That

is close to 2 million people.

Please decide on a standard that is

healthy, and that is doable by the food industry.

Thank you. Please help us to make more informed

decisions so we can take care of ourselves.

Thank you.

CHAIRMAN DURST: Thank you.

Does the Committee have any questions?

Yes?

219

DR. KELLY: Ciaran Kelly. I do have one

question. I think we are all in agreement about

the importance of clear, reliable labeling of

gluten-free foods. As we approach that,

approaching the question of thresholds, what we are

also struggling with are what the preferences of

individuals with celiac disease might be.

We all know that it is going to be

impossible to have zero gluten in food. The

question is, How rigorous a standard of gluten-free

do you think most individuals would like to see?

Do they want to see a highly rigorous or a less

rigorous standard?

MS. MONARCH: Well, we think that based on

the information that was provided here today, 20

parts per million to 100 parts per million, I think

each of us following the gluten-free diet would be

safe. I think that is probably a good industry

standard that the industry could comply with.

I think listening to some of the comments

yesterday from the allergy people I think catering

to the small fraction of people that have the most

220

severe sensitivities would do the entire population

a disservice.

CHAIRMAN DURST: Any further questions?

(No verbal response.)

CHAIRMAN DURST: Thank you. We will go

back now to our first speaker, Alice Bast, from the

National Foundation for Celiac Awareness.

MS. BAST: Hello, my name is Alice Bast,

and I am the executive director of the National

Foundation for Celiac Awareness. I am co-chair of

the Greater Philadelphia Celiac Sprue Support

Group. I am also a celiac sufferer.

Thank you for the opportunity to speak

with you today about the importance of clear,

unambiguous labeling of food so that the estimated

3 million Americans with celiac disease can

confidently choose food that is safe for us to eat.

We agree with the consensus statement

published after the conference of experts convened

by the National Institutes of Health, which noted

that the strict definition of a gluten-free diet

remains controversial due to the lack of accurate

221

method to detect gluten in food products and the

lack of scientific evidence for what constitutes a

safe amount of gluten ingestion.

These experts concluded that additional

research is needed to define the minimum, safe

exposure threshold of gluten in a diet relative to

celiac disease.

Celiac disease is underdiagnosed, in part,

because it has many forms. Patients suffering from

classical celiac disease exhibit digestive tract

reactions to gluten in the form of diarrhea,

bloating and constipation, but many more of us have

atypical or silent or latent celiac disease, and

many others are genetically predisposed waiting for

the disease to strike.

Unlike people suffering from food

allergies, addressed elsewhere in the draft report,

many celiac patients do not exhibit acute reactions

to food containing gluten.

Celiac disease must be confirmed through

blood antibody tests followed by an endoscopic

examination of the villi of the small intestine.

222

The result of continual ingestion of gluten is

chronic suffering in the form of: anemia,

osteoporosis, diabetes, thyroid disease,

infertility, stillbirths, and cancer.

With the level of complexity, it is

understandable that there is not yet the consensus

regarding a threshold level for gluten in the diet

of a celiac sufferer. How can a no-observable or

lowest-observable effect level be set when many

celiac patients exhibit no obvious symptoms?

We are encouraged by the research that is

underway to set a threshold, but we believe it is

premature for the Working Group to recommend an

approach to setting the threshold without more

data.

We encourage the FDA to consider including

its report to Congress on this subject a request

for an appropriation to be made to the National

Institute of Health to fund further research in

this important area with the goal of defining an

appropriate and healthful threshold level.

Gluten is not one but a family of proteins

223

that separately and together can trigger reactions

in celiac patients. These proteins are present in

wheat including durum, spelt, kamut, barley, malt

and rye, and the cross-hybrids and related proteins

are present in oats causing reactions in some

people with celiac disease.

Flour milling and food manufacturing

processes are ripe with opportunities for

cross-contamination, putting celiac patients at

risk of ingesting gluten from apparently safe

sources.

Again, we suggest that funding be made

available to develop and refine analytical methods

that will enable food processors to determine the

level of gluten present. We believe this is the

first critical step not only in the rational food

labeling program, but making food safe to eat for

celiac sufferers.

Cross-contamination represents a risk that

we can manage through proper equipment clean out

and product isolation procedures that are routinely

practiced by other industries. Providing standard

224

analytical methods to the food processing industry

will enable manufacturers to label their food

products properly, engendering the trust of celiac

patients throughout America.

Thank you for the opportunity to have me

speak to you today.

CHAIRMAN DURST: Thank you.

Are there any questions?

DR. MALEKI: Yes.

CHAIRMAN DURST: Okay. Soheila.

DR. MALEKI: Soheila Maleki. This could

have been asked for either one of the previous

speaker or you, but how does the consumer feel

about the labeling of two, like a double-scale

labeling, "low-gluten" versus "gluten-free"?

MS. BAST: I would have to speak on behalf

of myself. I would say that we have had one

incidence. There is a wafer, a communion wafer,

that has been labeled as low-gluten. There are a

number of people that are very hesitant in taking

that, because it is low-gluten.

I think that at least they have an idea or

225

an understanding that there would be no gluten

versus low gluten. That might be a good

compromise, because they know that there are

potential risks. If they are feeling that they

don't want to take those risks, then they have a

choice.

CHAIRMAN DURST: Ciaran, did you have a--?

DR. KELLY: (Shaking head.)

CHAIRMAN DURST: All right. Thank you

very much.

MS. BAST: Thank you.

CHAIRMAN DURST: Our third speaker is

Mary Schluckebeer from the Celiac Sprue

Association.

MS. SCHLUCKEBEER: I want to than you all

for having listened to so many different parts and

pieces of this rather complex problem. You see how

many questions there are? That is what we get in

our office every day as we reach people who are

newly diagnosed in our Celiac Sprue Association.

We get about 80 calls a day. We get about

the same number of E-mails and over 2 million hits

226

to our Web site every month. This is one where

people are looking for answers.

Like the researchers have shown you today,

answers aren't real easy to come by on this. We

really don't know the entire scope of the program.

This is probably because after diagnosis the

doctors are very pleased. They have figured out

what all of these strange little symptoms finally

came to, and it is diagnosed.

They are thrilled and they say, "All you

have to do is just go home and just eliminate all

of those amino acid sequences that you find in

wheat, barley, rye, oats, and their derivatives."

And you say, "I'm not going to die."

Then, you go home and you try to figure out what to

do.

Well, I am one of those people who is a

celiac. I was the daughter of a celiac. While I

was going to the University of Nebraska as a home

economist. "Oh, dad, a little bit won't hurt you."

Every time he got into a little bit he suffered a

lot.

227

At the time he was diagnosed, around 1959,

the smaller tube was introduced. The doctor said,

"I just read about this, and I think maybe we

should check you out." It was almost a fluke that

you got diagnosed around 1959, 1960.

At that time food was not labeled. Dairy

products had to have their recipe on file at the

state. You didn't know as a consumer exactly what

was in that.

Well, at that time ice milk was almost

always thickened with wheat flour, to help get that

feel in your mouth. Since you take the cream out,

you've got to put something in.

Oh, I never wanted to have his disease.

Now, he ate bread that was this (indicating) high.

I mean, that is as high as it ever got. At that

time wheat starch was allowed in the diet and was

in the packages that were said appropriate for

people with celiac disease.

Elaine Hartsook of the Gluten Intolerance

Group of North America started working with one of

the companies and said, "You know, this is still

228

making people sick."

When they eliminated the wheat starch from

these packages, my father's final symptoms of some

of the rash disappeared. He assumed this was

something he could eat with confidence. He

figured, "It's got to be something else I'm getting

into." He just couldn't figure it out. It was

that little, tiny bit of wheat starch.

So I'm always a little hesitant about

saying, "Oh, let's put this in" or "Let's take this

out," because, again, symptoms are not specific.

You can't say "I chewed this piece of gum, and I

got symptoms."

You go around and you're trying to figure

out, "What all did I get into in my environment in

this last two or three days that may have created a

symptom?"

When a person is diagnosed and the doctor

says "Go home and be well and just eat," because

you don't die -- researchers aren't real interested

in us when there are other problems that people do

die and we haven't solved and haven't gotten cures

229

for.

A celiac is left to have a team usually of

other people who have celiac disease, or they come

to support groups like ours where we have almost

10,000 members right now.

We are the largest celiac support

member-based group in the Western Hemisphere, and

we are very pleased. Canada has around 6,000

members in their association.

What we find is, though, that people get

very comfortable after they stick around with the

supports for a while and then they go off on their

own, because "Oh, I'm very comfortable in my diet."

I have learned how to live the lifestyle, and I

really don't need the help of everyone else.

We do a survey once a year of our

membership. One of the things that we did this

time was ask people to self rate where they

consider their sensitivity. Another question was,

How risk averse are you? Because it is a very

risk-averse population.

No matter what the sensitivity level a

230

person classified some of the cells as -- over 90

percent of the people put themselves at -- they

will take no risk, no known risk, in their food

choices.

That is a pretty high level of at least intent that

is expressed, no matter what they say their

sensitivity level is.

Again, that is why something like money

for research to find this threshold -- you notice

this threshold is the problem in each of these

countries. There is not any real good basis for us

to come up with a threshold.

That is why the physician said, "Go home

and don't eat any."

When you are talking to grandma she says,

"Just have a little," that's kind of where zero

comes out.

It is that place-taker or a way of

communicating, "I can't have some. I have to have

none." I don't know, if I could have some, I have

no idea how much "some" is.

It may be different when you are under

231

stress like in a hospital situation, at a childcare

center. What kind of risk do you want to take at

the training table, athletic training table, when

somebody else is picking out the food for those at

the table?

Again, without a threshold, it really

makes it very difficult to make some of these

choices because it is all subjective. Right now,

it would be awfully nice to be able to say it is

not subjective. We have some concrete information.

This is what will work as a workable definition for

the celiac patient and for the manufacturer, and it

is easy to communicate all of that information to

each other.

Thank you.

CHAIRMAN DURST: Thank you.

Any questions, Committee?

(No verbal response.)

CHAIRMAN DURST: I think there are not.

Thank you.

(Sotto voce discussion.)

CHAIRMAN DURST: Our next speaker, also

232

from the Celiac Sprue Association, is

Tom P. Sullivan.

MR. SULLIVAN: Good afternoon and thank

you. My name is Tom Sullivan. I happen to be the

president this year for the Celiac Sprue

Association. I myself am not a celiac. However, I

have very, very good association, and that may be

one of the reasons the patients decided that I

should be their president this year.

I have a wife who is a celiac; I have a

son who is a celiac; and I have a great niece who

is a celiac. The horror stories that lead to the

11 years' average time for diagnosis I can

perfectly well relate to and admit to. Because I

sat in a gastroenterologist's office and shook my

head most emphatically no four times to force that

man to take a biopsy on my wife.

The man never spoke to me again, never

looked in my direction after coming out of the

biopsy room. It was flat out knowledgeable on his

face what she had. It took that kind of forcing

effort.

233

Education is still necessary. It is a

major factor. It is one of the reasons CSA exists.

When a patient is diagnosed, they are informed

basically "Go and sin no more, my child. Change

from a wheat-based lifestyle to a rice-based

lifestyle. Goodbye" (waving).

What the heck does that mean? I haven't a

clue. It turns out the only ones who have a clue,

who know what to buy, where to buy it, who sells

what, how to use it, what do I do in my kitchen,

how do I travel, who do I see for this or that

problem, what does this symptom mean are other

celiac patients.

That is why CSA came to be, that is why

its mission is to be celiacs helping celiacs, and

that is why its function is to be the voice of the

patient. The patients are very, very good. We go

out with surveys each year, and they tell us what

do we need, what don't we need.

With reference to this afternoon's

proceedings or this week's proceedings rather with

this draft, the draft is a very good working draft.

234

It agrees with the patients, and that is, the

risk-assessment method is the method of desire.

In fact, the patients themselves have

moved to a risk-assessment process. It has been

done intuitively, it has been done with

cross-communication among all of them, and it has

produced the capabilities that CSA currently has to

speak for the patient.

What the patient does is very simple.

They say, "I have to eliminate wheat, barley, rye

and oats." Let's not talk gluten. Let's get away

from the. Source ingredients of wheat, barley, rye

and oats; okay? So my target is zero. Now we all

know, scientifically, zero is unmeasurable.

That isn't the situation. I have a

problem. I want none of it. How do I do it? Now

we get practical. Now we start asking

manufacturers, "What levels are you at? What do

you do? Can we trust you? Are you consistent?"

We put together lists of products. This

year's product listing is approximately 70 pages

listing products that the manufacturers will stand

235

behind, because they have told us that they do not

use wheat, barley, rye or oats in their product, in

their packaging, or in their processing. It is a

great source to help people get started.

In fact, one of the fun things I have when

I get an E-mail or a telephone call -- and it comes

from all over the world by the way, both into the

office and personally -- my immediate reply is

"Relax, take a deep breath, and let it out very

slowly. There is life after diagnosis." Then, we

teach them how to do it.

From a practical standpoint, the patients

evaluate the products that are out there. They

evaluate them against their target of zero, and

they handle them as a result of their reactions to

the ingestion of that product.

If they have a problem, they go look in

the book and find another similar product, a

different brand name, or they go to another label

of the same product in the store.

However, they have a method and a

technique that they have instinctively gone to, to

236

say, "How do I protect me? I have a health

problem. How do I protect me."

A very interesting result of this is that

when the patient starts on the gluten-free diet

they very quickly become better. This is why you

end up with a very wide range of variability in

your responses and in the reactions because most of

them, by and large, don't ever want to go back

there again. They didn't like it; they don't like

it; and they don't want it repeated.

One of the things that has helped is the

labeling and the information available out of the

manufacturers, the fact that they will respond, the

fact that the patient community is getting much,

much better on their knowledge of the questions to

ask and who to go to.

For example, not too long ago it was very

common to just pick up the phone and call the

manufacturer and say, "Do you have gluten in your

product?"

"Yes."

Okay, forget that product. Now, the

237

question is because rice gluten and corn gluten are

no problems, most of the patients will now ask,

"What is the source of that ingredient?"

"Oh, it's corn."

"Thank you." Problem solved. They have

set the risk level at zero. They have evaluated

the products that are out there, and they have

communicated that among themselves. That is

celiacs helping celiacs. That is what keeps them

safe. That is the way they have done it.

I am very, very happy to see that that is

exactly the way you have chosen as the recommended

technique for doing it. I think in the long run it

is the only one that is going to do it, what is the

minimum level. Beyond that, then, I've got

problems I can go looking at. Right now, we have

nothing. I think it is a very good start. Thank

you kindly.

CHAIRMAN DURST: Thank you.

Do we have any questions?

CHAIRMAN DURST: Do we have any questions?

Soheila.

238

DR. MALEKI: Sure. I guess I'll pose the

same question as far as previously, How do you feel

about two-scale labeling such as low gluten versus

gluten free?

Sorry, Soheila Maleki.

MR. SULLIVAN: That is a question we have

not yet asked our members, so I can't answer for

the membership. That is a question we will ask on

this year's survey, however, and we will have the

answer for you probably sometime just after the

first of the year.

Personally and based upon the input I've

had from the other celiacs over the years, if a

definition is precise and they can depend upon it,

then I don't think they will have any problems.

Quite frankly, a celiac patient is one of

the smartest people you're ever going to meet. It

is their health, their body, and they and they

alone are completely responsible for it.

By the way, at the end of September of

this year and the first of October in Tyson's

Corner, CSA is having our National Annual Education

239

Conference. You are all invited to come and find

out what the patients think and why they think it.

They will ask you some of the toughest

questions. It is a shame I didn't have Dr. Murray

say that this morning, because he has admitted they

ask nasty questions.

(General laughter.)

MR. SULLIVAN: They want to know because

it is my (pointing) body, and it is my

responsibility solely and completely. You tell me,

and I'll make the decision for me. That is where

it is coming from. It is more information. More

information is always to our benefit as a patient.

CHAIRMAN DURST: Ciaran, did you have a

question?

DR. KELLY: That addressed it.

CHAIRMAN DURST: Okay.

Anything else from anyone?

(No verbal response.)

CHAIRMAN DURST: Thank you very much.

MR. SULLIVAN: Thank you.

CHAIRMAN DURST: Our final public comment

240

speaker is Steve Taylor from the University of

Nebraska.

MR. TAYLOR: Good afternoon. My name is

Steve Taylor, and I am a professor and co-director

of the Food Allergy Research and Resource Program

at the University of Nebraska.

In addition to what you all heard from me

yesterday, our group provides analytical services

to the food industry including gluten testing

services, so I thought perhaps I could get up here

and say a bit about testing methods.

I should also say that this is a

fee-for-service activity that we provide to the

industry, but we also provide services on a lesser

cost basis to the Celiac Sprue Association and to

the Food Allergy and Anaphylaxis Network.

I want to make several points. One is

about testing methods and frequency. Our

laboratory uses the R5 monocle antibody test that

you have heard about this morning. That test is

commercially available from a company called

R-Biopharm in Germany as an ELISA kit. There are

241

other equivalent test methods that are on the

market as well.

This test detects prolamins, the prolamins

gliadin from wheat, secalin from rye, and hordein

from barley. It does not detect oats but will

detect the presence of wheat, rye and barley

proteins in oats, which is perhaps somewhat of a

significant concern to celiac sufferers. Our

advice is that they continue to avoid oats in

North America because of the chance that oats could

be contaminated with wheat, rye or barley.

The test detects the prolamin proteins

more reliably than it detects the glutelin

proteins, the higher-molecular weight ones, but we

can very easily detect the gluten levels in wheat

starch and other ingredients that you have

discussed today.

I think this test is very reliable for the

food industry to use to determine whether the

products are gluten free. I can say that the food

industry in North America has been using this and

similar tests for a number of years now to help

242

assure that products that are labeled gluten-free

indeed fit that definition.

I can say that it is my experience that

the industry is doing a much finer job in that

regard than perhaps they were 6 or 7 or 10 years

ago. That is partly because the Government of

Canada established this regulatory framework that

you have heard about this morning.

In the countries where the legislation has

said "gluten-free" is "zero," it can't get to zero,

so operationally you still have to have some

definition of it. The Canadian Food Inspection

Agency uses less than 20 parts per million as their

operational definition of gluten-free.

When they established this regulation in

1996, they began to be very vigilant in the

analysis of U.S.-made, gluten-free products

crossing the border into Canada to be sure that

those products met the definition. Well, most of

them did not.

They met the previous definition that

you've heard about this morning, the Codex

243

Alimentarius Commission's definition of less than

200 parts per million, but did not meet the

definition of less than 20.

I can tell you that since 1996 till today

almost all of those companies have succeeded in

protecting their Canadian market by now adhering to

the less than 20-part-per-million standard.

If you establish a standard of zero, many

of these companies will not be able to produce

gluten-free products because zero is unattainable.

We have heard that from some of your speakers this

morning.

I also want to say a few words about

grain-add mixtures, because the adventitious

presence of one grain in another grain is allowed

by something called "USDA grain standards." Wheat

can be in oats, soybeans can be in corn, soybeans

can be in wheat. That is allowed by USDA grain

standards, which are recognized around the world.

Raw agriculture commodities are another

exemption that is in the FALCPA legislation. I

think this establishes another potential

244

consideration for the panel in terms of a

statutorily derived threshold.

Once you convert these raw agricultural

commodities that are exempt into milled wheat

flour, milled oat flour and milled cornmeal, then

they are not exempt anymore.

Yes, if you establish a threshold at zero,

then this contamination occurs on the farm, and

there is no way to completely prevent it. However,

it is quite possible to have safe and effective

gluten-free products meeting the strictest

definitions in the world, those of Italy and

Canada, with less than 20 parts per million gluten.

I was convinced by the data I saw this morning that

seems to protect the vast majority of celiac

suffers.

Thank you. Dick Durst. I have question

on what the Canadians use as far as their method of

detection? What is the limit of quantitation on

their immunoassay?

MR. TAYLOR: I think they use the same

test that our laboratory uses, which is the

245

R-Biopharm Test. R-Biopharm sells several

different tests. I wish Dr. Hefle were still here.

I think the limit of sensitivity of the tests that

we are currently using is in the neighborhood of 5

parts per million slightly lower than that.

CHAIRMAN DURST: Okay. The limit of

quantitation is right near the limit that is set,

this 20 parts per million?

MR. TAYLOR: Well, it is severalfold below

that. I mean, it is 5 parts per million and the

limit is 20.

CHAIRMAN DURST: Yes. Well, the limit of

quantitation, I would think, is what you would need

to use in order to really verify the amount of

gluten or whatever or prolamin that is in the

product. I'm not sure the limit of detection is

the kind of best characterization.

MR. TAYLOR: Yes. The limit of

quantitation with that test is in the neighborhood

of five parts per million. I don't know what the

lowest limit of sensitivity is. We know that we

can reliably test 5 parts per million with that

246

test. I know Dr. Hefle knows the answer to that

question, I just don't.

CHAIRMAN DURST: Marc had his hand up?

DR. SILVERSTEIN: Marc Silverstein. Do

manufacturers continually test during production,

or is it just in developing a new product for the

market?

MR. TAYLOR: It has been our experience

that many of the producers of products that are

labeled gluten-free test rather frequently. There

are several very noteworthy companies that make

gluten-free products that are rather popular among

celiac sufferers, and these companies test very

frequently.

One that I can cite as an example would be

Arrowhead Mills, which was one of our more frequent

clients for a number of years. They were doing so

many analysis that they built their own laboratory

at the plant in Texas, and they do the ELISA

testing on a regular basis in their own facility.

CHAIRMAN DURST: Soheila.

DR. MALEKI: Yes. I just want to say that

247

you brought up a really good point about the

farmers that grow in and have rotation crops. That

essentially brings up a good point. I don't think

you could ever reach zero, even just because of

that, because of the same trucks they use, the same

dirt it is grown in, and so forth.

MR. TAYLOR: Yes. I mean, it is the same

farms, the same farmers, the same harvesting

equipment, the same on-farm transportation, the

same elevators, the same off-farm transportation.

The system, the commercial system, for handling

grains in the United States and around the world

doesn't offer you the opportunity to get to zero.

CHAIRMAN DURST: Erica.

DR. BRITTAIN: Yes. Is it possible,

though, to drop somewhere between twenty versus

zero? Is that realistic at all? Or, do you really

think 20 is as far as you could go? Could you go

to 10?

MR. TAYLOR: Well, you can do anything you

want with the analytical testing capabilities. I

think I would defer to the clinical experts, that

248

we have heard from already here, about what the

threshold level for celiac sufferers ought to be

and the way that ought to be established.

We don't do any clinical research on

celiac disease. We avidly read their papers, but

we are just analysts with respect to celiac

disease, and we do not pretend to be clinical

experts on this difficult subject.

I mean, in terms of being able to make

products that would pass that standard, 10 versus

20, would that make a big difference in terms of

making products?

MR. TAYLOR: Well, to me does 10 versus 20

make a difference? It depends upon whether it

makes a difference to the celiac sufferers in terms

of their health status.

The industry struggled when we went from

200 to 20. Many of them already could probably

come close to meeting 10, if they don't already do

it. Some of them might struggle to get there.

Consistently? Consistency is another key point.

CHAIRMAN DURST: Jeff.

249

DR. BARACH: Yes. Jeff Barach. Could you

speak to the validation of the test that you

described, the monoclonal antibody test? Do you

know if it is validated?

MR. TAYLOR: Well, I don't know if it has

been "validated" by the procedures that FDA prefers

to use when it uses that term, but it has been

validated by the company that made the kit. The

Prolamin Working Group has done some

interlaboratory testing of that kit as well. I am

not so sure that there have been comparisons

between that test and tests by competing companies

that are largely similar, so there may be some

analytical work to do. I am not so familiar with

the Prolamin Working Group. Dr. Hefle follows that

group, but I don't.

CHAIRMAN DURST: Margaret.

DR. BRILEY: Margaret Briley. Could you

give us any kind of estimate of the cost factor for

industry in terms of how often they use this test

and what it would add to the cost of the product?

Once you start testing it, I would think

250

you would test everything that came through,

whether it was for celiac or not. Am I wrong? I

mean, if you're going to run a test, wouldn't you

just run it? You're putting it out as an industry.

You wouldn't do a separate run just for celiacs?

MR. TAYLOR: Well, that is a very

complicated question as to how frequently you test,

how you devise a credible sample plan, and whether

the results of your test are reliable in terms of

all of the product manufacturers. Obviously, you

can't do a test on ever package of product, because

then you wouldn't have anything left to sell.

(General laughter.)

DR. BRILEY: No.

MR. TAYLOR: The tests are not very

expensive in some terms. We charge $50 to $75 for

the test per sample. I mean, that is some cost and

companies are going to question whether they want

to do 100 tests, 1,000 tests or 10,000 tests,

because it I going to be a cost factor.

DR. BRILEY: Well, I guess I was thinking

that you would probably test a run. You wouldn't

251

test every package that came out.

MR. TAYLOR: Yes. You've got to design

your testing system very strategically depending

upon where you think your sources of contamination

are.

Companies typically test the source

materials that are coming into manufacturing. They

test the first product manufactured after

changeover, if they have shared equipment.

However, you've got to pay attention to

things like whether you think there are hangup

points in your manufacturing equipment. That

varies from facility to facility and line to line.

DR. BRILEY: From company to company.

MR. TAYLOR: I wouldn't give the same

advice to every company.

DR. BRILEY: Okay. Thank you.

CHAIRMAN DURST: Ciaran.

DR. KELLY: Dr. Taylor, thank you. Ciaran

here. I'm going to keep you on the podium for

another moment or two.

We heard about the line spots in the

252

currently available test, the inability using the

widely used test to detect gluten in oats. Are

there to your knowledge intrinsic, technical

challenges there, or is it simply that nobody has

bothered to try?

MR. TAYLOR: I don't think anyone has

tried to develop a test for oats. I am convinced

you could develop an ELISA test for any

protein-containing food known to man. Yes, you

could develop a specific oat test.

There is this debate about whether oats

are safe or unsafe, the companies that were

developing these tests for gluten-free products

targeted these peptide sequences in wheat, rye and

barley. You could argue that is what they should

have done. I would advise them to do the same

thing.

CHAIRMAN DURST: Anything further?

(No verbal response.)

CHAIRMAN DURST: Thank you.

Okay. Now we will jump back in time to

hearing Steve Gendel, who is now with us, to speak

253

on the overview of approaches to establishing

thresholds for gluten.

OVERVIEW OF APPROACHES TO ESTABLISHING

THRESHOLDS: GLUTEN

MR. GENDEL: I guess I can say that one

way of keeping people from going into an

after-lunch slump is to mess with the agenda. You

have to pay attention to know where we are. I'll

take credit for that.

(General laughter.)

(Slide.)

MR. GENDEL: What I'm going to do today is

going to be an abbreviated form of my shortened

talk from yesterday, again, just to serve as a

refresher for what is in the "Draft Report"; to set

the stage for your discussion; and, again, to

remind you that the purpose of the report is to

identify approaches that can be used to establish

thresholds, not to decide on which approach to use

and not to discuss specific threshold values. We

are interested in potential approaches, the

advantages, disadvantages and data needs of each.

254

(Slide.)

DR. GENDEL: The organization of the

report hasn't changed since yesterday. There is a

section where we review celiac disease and one

which we talk about the approaches we have

identified for setting thresholds for celiac or for

gluten.

(Slide.)

DR. GENDEL: In the section on celiac

disease, we reviewed the mechanism of pathogenesis,

information on prevalence, foods of concern, we

looked at the clinical challenge studies that were

available, and looked at issues related to

measuring gluten in food -- all of the things that

we have heard about this morning.

(Slide.)

DR. GENDEL: As with the allergens, we

identified four potential approaches, and really in

this case three: the analytical methods-based

approach, the safety assessment-based approach, and

a quantitative risk-assessment-based approach.

255

I mentioned the statutorily derived one

here, for the sake of consistency with what we

talked about yesterday, where we felt that there

was no language in FALCPA comparable to that for

allergens that could be applied in the case of

gluten.

(Slides.)

DR. GENDEL: I am not going to go through

these approaches again. I think you are familiar

with them. The analytical-methods-based approach,

which is based on the sensitivity and detection

methods available; the safety-assessment-based

approach relies on LOAELS and NOAELS from clinical

data and appropriate uncertainty factors based on

the gaps in those data; and the

risk-assessment-based approach; and the

quantitative approach, which takes all of the dose

response information available into account.

(Slide.)

DR. GENDEL: The findings of the Working

Group, there were again five, the first one again

to reiterate the fact that whatever approach -- if

256

a decision to set thresholds is made, whatever

approach is chosen at this time, that these

decisions should be reevaluated frequently as new

data became available.

We heard a lot of discussion this morning

about clinical studies here also that are in

progress, and new data will become available. We

recognize the fact that any decisions made in the

short term should be reevaluated periodically.

The Working Group found that the

analytical methods-based approach could be used for

gluten also. However, as we talked about

yesterday, if it is used, we feel that it should be

replaced by a risk- or public-health-based approach

as soon as that is feasible.

The safety-assessment-based approach, the

Working Group found that approach could be viable

also based on data from the literature and

appropriate safety factors, taking into account the

nature of the clinical studies available to use.

The risk-assessment-based approach we felt

was not feasible at this time due to the lack of

257

data to quantitate risk in a dose-response type

manner.

Finally, as I mentioned, the statutorily

derived approach is not viable due to the lack of

appropriate statutory language.

That is really all I have to say about the

report. Are there are any questions about the

report itself?

QUESTION AND ANSWER SESSION

CHAIRMAN DURST: Thank you, Steve.

Do we have any questions?

Ciaran.

DR. KELLY: Yes -- Ciaran Kelly -- just a

technical question. When we are talking about the

safety-assessment approach, does that include

population observations in addition to prospective

studies? We heard this morning about a prospective

study, retrospective studies, and clinical

experience with a population that have been using

particular standards for many years. Is that

information incorporated within a safety-assessment

approach?

258

DR. GENDEL: I would say that the

safety-assessment approach would be one where any

data that can be used to establish a LOAEL or a

NOAEL is used. Then, depending upon where those

numbers come from with that number, then you would

apply appropriate uncertainty factors, and the

nature of the data which goes into establishing

those numbers would then be taken into account as

uncertainty factors would apply.

DR. KELLY: Would it be true to say, then,

that if similar numbers were arrived at from

different sources in the data, if independent

studies using different methodologies all arrived

at a similar number, that would reduce the

uncertainty factor?

DR. GENDEL: I would say that is probably

fair. Anytime you can replicate data, the degree

of uncertainty associated with it is less.

CHAIRMAN DURST: Any other questions for

Steve.

(No verbal response.)

CHAIRMAN DURST: All right. Thanks,

259

Steve.

DR. GENDEL: You're welcome.

CHAIRMAN DURST: We are now scheduled for

a break. We are about 15 minutes ahead of

schedule, so we will take our 15-minute break and

reconvene at 3:15.

Thank you.

(Thereupon, from 2:55 p.m. to 3:15 p.m.,

there was a pause in the proceedings.)

COMMITTEE DISCUSSION

CHAIRMAN DURST: Would everyone take their

seats please, and we can continue the afternoon

session.

All right. At this point I guess Steve

just before gave a nice review of the charge and

the questions that we are supposed to address.

What I would propose is that we initially begin

with just open discussions of the general points on

celiac disease; then address some of the specific

questions; and then, finally, if there is time at

the end of the day, also open discussion again on

the allergens and perhaps any cross-reference to

260

the celiac disease. There are certainly similar

questions in both of those cases.

I would like to mention, just to expedite

tomorrow's discussions, I have asked three members,

Marc and Suzanne to deal with the allergens and

Ciaran to deal with the gluten, try to come up with

a summary or a consensus of what they felt our

discussions have been leading to in terms of how we

want to address these approaches for setting the

thresholds.

I think that would help us in the morning

to focus in on those particular aspects and, again,

have the discussion bring in any new points or

additional points that members may want to add to

those summaries. I think that is all I want to say

on this point. Let's open the discussion on the

gluten and celiac disease.

Does anyone want to start with any general

comments on that?

Soheila.

DR. MALEKI: Soheila Maleki. I actually

have questions. Is that appropriate at the time,

261

at this time, to ask the panel questions?

CHAIRMAN DURST: Yes.

DR. MALEKI: Well, I have some questions

for Dr. Fasano. Well, I've got multiple questions,

but I will try to go through them where you can

answer them. What is it the specificity of the

activated CD3 T-cells? Do you know if they are

gamma/delta, alpha/beta, CD8, CD4?

What is particularly their specificity as

far as are they transglutaminase-specific or the

PEQ-specific? Anyway, do you know of any studies

that have looked at gluten-specific T-cells that

actually are reacting to oats or one of the other

products? How about the antibody

cross-reactivities of gluten versus barley, wheat,

and then oats? I think that's it.

(General laughter.)

DR. FASANO: Let me tell you the facts the

way we know right now. The activation of the

intraepithelial lymphocytes, particularly through

CD3 and gamma/delta, are considered highly specific

for celiac disease.

262

As a matter of fact, in the early Marsh

classification, Marsh I, we don't have any damage

whatsoever but you have all the infiltration,

intraepithelial infiltration, into the lymphocytes.

If you want to know that is malignancy of

the disease, you do the specific CD3 staining. If

it is positive, then you can say, "Okay, this is

Stage 1 of a Marsh grade for celiac disease."

Yes, as far as we know, there are

gluten-specific T-cells epitopes. You can isolate

T-cells for gluten or a fraction of gluten in terms

of a reaction activation of T-cell and K-cells and

so on and so forth. Absolutely, that is the way to

do that.

Of course, the specificity of

transglutaminase is an issue that is out there.

The only thing that I can tell you, at least based

on serological data, i.e., how specific is tissue

transglutaminase or inflammation-related celiac

disease, I would say that it is fairly specific,

with a few exceptions.

If it is true the current theory that the

263

reaction to transglutaminase is due to an initial

insult of the cells, it leaks transglutaminase, and

therefore becomes not naive anymore to the immune

system, leading to the immune response.

If this starts at the intestine level,

there is some specificity with celiac disease as

compared, for example, to Crohn's disease in which

we don't see that. However, we see patients with

Type I diabetes, for example, and not co-morbidity

with celiac disease in which the insult translates

with increased antitissue transglutaminase

antibodies.

In terms of cross activity among grains, I

will not give rights to the arguments. I believe

that Don will be much better than I am to give you

that kind of response. I can give you an

unprofessional, amateur response. The general

wisdom is, yes, there is cross-reaction.

DR. MALEKI: At T-cell level also?

DR. FASANO: Say that again?

DR. MALEKI: At the T-cell level? Of

course, yes.

264

DR. FASANO: Yes.

DR. KASARDA: At the T-cell level? Is

that what you said?

DR. MALEKI: Both actually, antibody and

T-cell.

DR. KASARDA: That is difficult to answer.

The problem is that almost all of the studies focus

on wheat, and very little has been done with rye

and barley.

As far as immunological epitopes are

concerned and cross-reactivity with IgG and IgE and

probably IgA, yes, there is a lot of

cross-reactivity because there is sufficient

homologies. However, when it comes to the T-cell

and the T-cell receptors, that is a whole other

ball game. I can't answer that. Maybe Pekka

Collin can.

DR. MALEKI: Thank you.

DR. COLLIN: I would say about gamma/delta

T-cells that they are thought to be very specific

for celiac disease, but we are a little bit

disappointed. I think specificity and sensitivity

265

for increased density is about 90 percent. The

strongest evidence against that they would be

specific is that in many cases those with elevated

gamma/delta T-cells they do not share their DQ2 or

DQ8, so they probably are not celiacs.

DR. FASANO: That is gamma/delta. The

CD3, that is the one that I was talking about, that

seems to be much more specific.

DR. MALEKI: Thank you.

CHAIRMAN DURST: Erica.

DR. BRITTAIN: I have a totally different

question.

CHAIRMAN DURST: Okay.

DR. BRITTAIN: Yes, it is about the

Italian study again. As a statistician, it is my

job to be skeptical. I just wanted to ask, and I

know that you have demonstrated there was a

difference between the placebo group and the

50-milligram group. You didn't see a statistically

significant difference between the zero and ten.

In fact, the means were very, very similar.

Normally, when you want to show that two

266

groups are equivalent or similar, you would

construct a confidence interval to define that

difference and estimate that difference. Is that

something that has been done?

DR. FASANO: Yes. The analysis for the

villus/crypt ratio was done on a confidence

interval level and there was no overlapping.

However, I have to be super-duper skeptical and say

that a morphometric measurement is not as accurate

as any other biological readout that you can

consider.

In other words, it is operated dependently

of course. That is how you make the measurement.

It is not a machine, so there is some degree of

possible error in there that you have to consider.

Nevertheless, if you did this in the blind

fashion, as we did, if you have two operators and

there is a 100 percent concordance, as happened to

us, the level of confidence that that was right

increases.

A more objective measurement, i.e., the

intraepithelial lymphocytes for which you say, "I

267

want to know how many there are per hundred

entrocytes, per hundred T-cells."

Why are we saying that 50 milligrams may

create a problem is because, again, we want to be

extremely conservative -- if we say, "Well,

actually this data is suggested but not conclusive

for damage" -- the answer is yes. We don't want

any question marks. It could be suggested, but

that is not the way that we want to go.

Conversely, with 10 milligrams, no matter

how you go, if you look at the intraepithelial

lymphocytes, if you look at the gamma/delta, if you

look at the alpha/beta that we did -- I didn't have

the time to show all the data -- in other words, if

you look at all of the parameters that you can

conceptualize to look at for possible histological,

serological, clinical evidence of a reaction, we

consistently in all of these patients found zero

reaction.

Consequently, our level of confidence is

associated also to the many years of implementation

of that threshold; there have been hundreds. That

268

makes us to say with some level of confidence that

we feel comfortable with 10 milligrams while with

50 we do not.

CHAIRMAN DURST: Soheila.

DR. MALEKI: Just one quick follow up on

that. That is part of the reason I was asking this

question is you would have to have biopsies to get

T-cells. Is that like some type of valid

immunological method to try to test for dosage, and

so forth, although it is in vitro? Could that be

some method, too, you would look at exposure, and

so forth?

DR. FASANO: I hate to be the one to

answer all the questions here. Depending on the

scientist, you will find different answers. There

are people that strongly believe that if you take

blood and you isolate its lymphocytes and you do

this exercise in vitro, it reflects exactly what is

happening intestinally also.

For example, the group from Australia is

trying to develop a vaccine for celiac disease.

The types of screening that they are using to

269

establish the many peptides can be toxic or

immunogenic, which one they have to look at, and

you target it for a vaccine.

They use an immunoblock reaction in which

they take blood from patients with celiac disease

to see with these peptides which one will react and

which one will not by the reaction of

interferon-gamma.

The skeptics of the group will say, "Well,

not necessarily does this reflect what happens at

the mucosal level." Technically, that is not

necessarily the same lymphocytes.

You can make an argument that you can test

negative in vitro, because you don't have the right

cells to migrate from the gut into the systemic

circulation. I don't think that there is a final

argument either way to sustain that you can do this

in vitro versus the biopsy.

CHAIRMAN DURST: Ciaran.

DR. KELLY: Ciaran Kelly. I've got a

question for Drs. Collin and Fasano. It relates to

a safety-assessment-based approach again relating

270

to data, historical data.

Both of you mentioned, but I wonder if you

can expand a little, upon studies that have looked

at the outcomes of individuals on gluten-free diets

set at certain levels as regards morbidity and

mortality outcomes. You both mentioned it. I

wonder if you could tell us a little more maybe

about the methodologies and results of these

studies?

MRS. MOORE: Excuse me, I'm sorry. When

you reply, say your name.

DR. COLLIN: Pekka Collin. If you look at

the mortality figures and risk of malignancy, I

think the most quoted paper is from Dr. Holmes

where he showed that if the patient stays on a

gluten-free diet for five or six years the risk

virtually disappears.

At that time, I think it was from 1984, I

think that diet was not so strict as today. I

suppose that at that time also people in the

United Kingdom they used the wheat-starch products.

The difference was between those who are on

271

gluten-free diet and between those who are on a

normal diet, which is very seldom today that people

are on a totally normal diet.

Then, the group from Italy, Corrao, the

excess mortality, it was all due to those patients

who had dietary lapses or who did not maintain the

gluten-free diet.

In our own studies, the first had 300

patients. We did not find any extra malignancy and

mortality in patients who were not on a gluten-free

diet. Now, later we have our odds ratio for

lymphoma, which is about four. It is almost the

same as that Peter Green had in New York and what

was in Corrao's paper and in the latest papers also

from Sweden.

However, each case except one has been not

on a strict gluten-free diet, and the majority have

occurred immediately, as I told you, after the

diagnosis of celiac disease. Probably, they had

celiac disease and lymphoma simultaneously. Only

one patient with celiac disease, presumably, on a

gluten-free diet developed lymphoma.

272

I think that all of the evidence shows

that if we try to avoid lymphoma, we should detect

the cases early enough, then put them on a

gluten-free diet, take care that they are on a

gluten-free diet, there is some circumstantial

evidence that those patients who remain undiagnosed

and who are asymptomatic the risk of lymphoma there

in them is very low.

In the United States, I don't know, maybe

you have 200 million people or even more, and you

should have 2 million with celiac disease, the

majority is undiagnosed.

Still, I think that small-bowel lymphoma,

especially small intestinal cancer, they are very

rare even here. That would be a serious risk

factor, I think we should see a lot of lymphomas

here.

Our mortality risk, our odds ratio is now

1.2, so it is very little excess mortality, and it

depends on the appearance of lymphoma at the same

time as the celiac in the patient.

DR. FASANO: I believe that what you are

273

looking for is if there are any systemic studies

that would compare 20 versus 200 versus 400, and of

course there is none. However, there are studies

that Pekka already outlined between people

complying and people admitting to being less than

compliance. The lymphoma probably is the least

proper variable outcome to look at, because it is

very rare to start with.

I am not a biostatistician, but I am

assuming that if you are dealing with a condition

that is one in a million, that will go to 1 in

890,000, it is hard to make the difference.

However, if you see co-morbidities,

autoimmune diseases like diabetes or Hashimoto, in

which you reach as high as 17 or 18 percent, then

you start to really look at the differences in

which you have an outcome such as osteoporosis, the

same story, short stature, and so on and so forth.

DR. KELLY: Ciaran Kelly. Yes, I know

there are no systematic studies. I was more

wanting you to elaborate on the experience, the

clinical experience, for many years at different

274

levels of definition of a gluten-free diet.

It seems as though there is a lot of

experience with 200 parts per million, 100 parts

per million, and 20 parts per million. I am trying

to get a sense for whether there have been any

studies to determine differences in outcomes with

those different levels.

DR. FASANO: Alessio Fasano here. Again,

in Italy the switch from the 200 parts per million

to 20 parts per million occurred, again, six or

seven years ago. There are no published studies to

show if this which translated into decreased

co-morbidity of that outcome.

The general wisdom for what is in there,

in terms of the co-morbidity reports within the

Celiac Society in Italy, seems to suggest that

indeed there has been a decrease of some of the

co-morbidities -- particularly, anemia,

osteoporosis, short stature, and miscarriages, the

fourth -- following the switch. However, these are

very anecdotical, and I don't think there is such a

strong scientific outcome to make that statement as

275

a scientifically acceptable one.

DR. COLLIN: We have experience in Finland

that the important deviation is not between

200 milligrams and 50 milligrams. I think I do not

have a very strong scientific evidence but only

experience in what you were asking. I think that

the risk of lymphoma in those patients is very,

very low.

The problem is that we do not know what

our wheat starch was 15 years ago when everyone

also was using them, but it is logical to assume

that it may have contained even more wheat or

gluten as today. Even at that time, we did not

have any increased risk of malignancies. It has

been similar all the time when we have had this

follow up since 1970 or 1975.

CHAIRMAN DURST: Marc.

DR. SILVERSTEIN: Marc Silverstein. One

of the questions that comes up in study conditions

such as allergic diseases and celiac disease is the

spectrum of disease and selection or selection bias

in patients who enter studies and about whom we

276

make inferences to the larger population.

As I'm understanding the available data,

it seems that clearly there would be a large number

of individuals who would have serologic evidence

and genetic susceptibility.

It seems there would be a large number of

individuals who probably have the disease, who have

manifestations that would be detectible were they

evaluated, but because they have very mild symptoms

don't come to attention.

What I'm asking is, then, would it be your

sense that the spectrum of patients who come to

clinical attention who have symptoms, who go to

medical centers, who ultimately get diagnosed, even

if it is 11 years on average later, that spectrum

of patients would be more severe?

Would they be more severe than the general

spectrum of patients who might be in the population

with a genetic predisposition, perhaps, with some

inflammatory changes in their bowels, yet are not

under medical care, so that the "selection bias,"

if you will, is to exclude the relatively mild

277

disease and the ones you see would be the more

severe disease in the general population? I guess

it is a question to Dr. Murray, who is not here.

Dr. Kelly might step in and comment.

Dr. Collin.

DR. COLLIN: Of course, if you detect more

and more patients with celiac disease, then the

figures become less biased. We have a lot of

people who are asymptomatic, patients with celiac

disease. The question is very complex.

On one end, on one side, we have a patient

who has very mild, mild inflammation and very mild

villous shortening in the mucosa, and they may have

some symptoms such as iron-deficiency anemia. On

the other end, we have patients who have totally

flat mucosa, and they are totally asymptomatic.

We have learned a lot of data from family

studies where we are actively screening all of the

celiac disease patients. We have seen truly

asymptomatic patients who will not be detected

without the serology screening. Still, we do not

want to extend the screening program to the whole

278

population.

I think our policy is that we apply

screening in risk groups and if they have even

minor symptoms and then we can achieve almost 1

percent of celiac disease, clinically diagnosed.

I'm not sure whether I answered your question. I

couldn't hear it very well.

DR. SILVERSTEIN: May I follow up? There

is a paper, also from Finland, the prevalence of

celiac disease in children in the New England

Journal paper a couple of years ago. Again, that

was serologic, serology was available, so you had

population-based samples.

Those who had abnormal serology, when they

were followed up, and then you found some spectrum

of undiagnosed disease in those children. It would

seem to me the children detected through that type

of mechanism, would have generally milder disease

than those who would have come forward because of

the clinical presentations.

DR. COLLIN: Yes. In that paper, I think

the prevalence was not 1 percent, more than

279

1 percent of people, children in Oulu, Finland.

You are right, I think many of them who were not

detected earlier for celiac disease, who were

detected by this serologic screening, they were

asymptomatic.

We have also carried out some quality of

life studies from those patients who have been

screened for celiac disease not due to symptoms but

because they belong to the risk groups.

We have seen that their quality of like

with these measurements that we have used are very

similar to that of the population in general, and

it is better than those who are symptomatic

patients.

With a gluten-free diet, it still

increases, and it becomes after one year even

better than in the population. We call it maybe a

honeymoon period. We have no long-term data on

that, but that is very interesting that many people

really are asymptomatic.

DR. FASANO: Alessio Fasano. Marc, you

are absolutely right. The people that come to our

280

clinic with symptoms definitely are the biased part

of the population with celiac disease because they

are the ones that have symptoms that seek

attention. It is undisputable that compared to the

overall picture of celiac disease, the one that we

see on the clinical grounds are biased in that

direction. No question about it.

However, for example, it is policy for us

right now that every single time you make the

diagnosis of an individual the entire household is

screened.

Epidemiology studies out there suggest

that up to 10 percent of first-degree relatives

they have the disease, irrespective of if they have

symptoms or not. Sometimes when we diagnose these

people that apparently are completely clinically

silent, you do a truly, you know, well-done workup,

they have osteoporosis or osteopenia.

How would you consider the otherwise

completely silent? If you make the diagnosis on

time, and according to the current literature "on

time" meaning two to five, you can fix and correct

281

the problem. If you are too late, you can't do

anything about that in these people.

I would consider that a great, great

danger, even if clinically they are absolutely

silent because these are people at risk for

fractures in their thirties. We have seen these

cases.

DR. SILVERSTEIN: If I could follow up --

in terms of the spectrum of disease, unlike the

situation where you have a patient with perhaps a

food allergy, who we heard about yesterday, whose

physician or family member or even the patient may

decide the risk of the food challenge test would be

too great and they would be excluded, in your

experience in caring for patients with celiac

disease, is there a similar phenomenon where the

patient who were more severe would be less likely

to undergo evaluation biopsy or participation in

studies?

DR. FASANO: In terms of a challenge, in

other words, I'll give you a practical scenario.

An individual comes to our clinic because they have

282

symptoms for many years. They have never been

diagnosed, but they spoke with a friend or a

colleague or a family member that raised the issue

of the possibility of celiac disease.

They go on a gluten-free diet without

being diagnosed, and they are feeling better. They

come to your clinic, and they want to know if this

could be a definite diagnosis or not. You say,

"The only way to do this is you have to do a

challenge."

If this individual had a really hard time

in his or her life -- in other words, the symptoms

are severe -- the likelihood that this individual

will accept the challenge is much lower than the

person that had the stomachache or the bloating

here and there with vague symptoms that now are

gone away.

However, they want to know for sure,

because now they realize that a gluten-free diet

for life is not a joke, that this is indeed the

kind of direction to go. That individual is more

likely amenable to a challenge.

283

However, an individual who has been

absolutely sick with tremendous symptoms that

affect their lifestyle, that individual will be

very, very unlikely to be open for a challenge.

That is my personal experience.

DR. KELLY: Ciaran Kelly. Just to expand

upon the question, that is certainly the case.

That is certainly my experience. However, we are

talking about clinical gluten challenges. Do you

think it is the same for a prospective study where

one would be performing a low-dose, a minimal dose,

gluten challenge? Do you think that highly

sensitive individuals would also be less likely to

participate?

DR. FASANO: I think that there is a

serious possibility. In other words, when you do a

prospective study like the one that was done in

Italy and say, "Look, there is a chance that we're

going to give you a placebo, i.e., water and you're

going to be all right, or you could get some amount

of gluten that we don't know if it's going to harm

you or not," if this individual has a really hard

284

time, that individual will probably be less likely

inclined to participate.

Saying that, though, of the people that

have participated in this study, there was the

entire variation of the spectrum, if you wish, of a

gravity of symptoms. I can't tell you if there

were people that claimed to be hypersensitive to

gluten, those that will react like two hours after

eating.

If we have a few of these people because

this was randomized or because it was blind, I

can't remember. Actually, I don't know yet if they

were included in the study or not. There were

people like that who volunteered to do the study.

If they end up to do the study or not, I don't

know.

DR. COLLIN: Some half a year ago, we

started a study where we looked at hydrolyzed

products derived from wheat starch and the outcome

of histology where we have also to take one biopsy

before and one after a half year's period.

I did not have the feeling that the most

285

sensitive would not come. I think that those

refuse to come who don't have a strict diet,

because they feel that maybe their small intestine

is not in a good condition, and the doctor will

blame him or her about that.

Again, I would like to say that I suspect

whether there are really highly sensitive people.

Usually, when we start a study, we take 100

patients, and another 100 call to ask us, "Why

can't I participate in this project?"

CHAIRMAN DURST: Okay. Marc.

DR. SILVERSTEIN: I have a question on a

different area, so if there are further questions

following up on this selection, then we should

pursue that--?

CHAIRMAN DURST: I don't think so, go

ahead.

DR. SILVERSTEIN: Could I ask

Cynthia Kupper a question, if I may?

MS. KUPPER: Certainly.

DR. SILVERSTEIN: I was interested in

understanding the extent to which a celiac

286

patient's likelihood of following up with a

healthcare provider, physician or dietitian as

opposed to a disease association or an informal

network with regard to dietary advice, how that

might be changed by more helpful information on

food labeling?

In other words, if the labeling were more

consistent, more trustworthy, more reliable, would

there be increasing reliance on the labeling or the

non-clinical advice, or would there be even more

likelihood that physicians, dieticians and others

would be able to be more effective in managing

their patients? How would that likely affect

patients' behavior, do you think?

MS. KUPPER: In many ways, it is two

different issues. First of all, patients when they

are diagnosed currently oftentimes they are

referred to a dietician, but in many states

dietician services are not paid for.

Consequently, many patients don't go, or,

if they do go, the dietician is inadequate in

preparing them with the information they need, so

287

they are very frustrated.

Referrals back to dieticians should

happen, as suggested by the NIH Consensus

Conference, because part of the problems with

compliance is that they don't have that consistent

follow up and they aren't monitored by a dietician,

and there are some nutritional concerns about a

gluten-free diet.

In a sense, it is a different issue. They

need to be seeing a dietician, but they are just

not referred or their insurance isn't going to pay,

so they don't go.

Would a patient rely on labeling more than

a medical professional like a dietician, if the

labeling were more accurate? I really don't think

so. I think they still need the dietician.

I think they will be happier being able to

find sound advice from the labels, but it is still

a matter of teaching them how to read a label,

learning what the terminology is, and understanding

that so they can make wise choices.

QUESTIONS FROM FDA: GLUTEN AND CELIAC DISEASE

288

CHAIRMAN DURST: All right. If there is

no more general discussion or immediate questions

for the speakers today, I would like to suggest

that we start addressing some of the specific

questions from the FDA, so that we don't run out of

time at the end -- unless someone feels there is

some other urgent question they want to bring up?

(No verbal response.)

CHAIRMAN DURST: Okay. The first one is

the question of whether there is a distinct

subpopulation of individuals with celiac disease

and then going into the uncertainty factors

involved in these measurements. Would anyone like

to make some comments on that?

DR. KELLY: Ciaran Kelly. I think you can

approach that question from two angles, one is easy

to answer and the other is more difficult to

answer.

From a clinical perspective, it is very

clear that there is a broad range of clinical

manifestations of the disease and that some

individuals with celiac disease are able to ingest

289

the same amount of gluten in their diet as

everybody else and don't demonstrate any clinical

or nutritional ill-effects, at least in the

short-term.

Whereas others, if they ingest a tiny

amount of gluten, a crumb of bread, will have in a

very short period of time a gluten reaction, a

reproducible reaction that lasts a predictable

length of time; so, clinically there are.

What is more difficult, however in mind is

the fact that those clinical reactions don't

predict the severity of the mucosal abnormality.

At one level yes.

At another level there is also a variation

we saw earlier, the Marsh classification, of the

histologic abnormalities. There is a variation in

that also, but they don't overlap neatly. You

won't find always low-level lesions in silent

patients. The answer is yes, I believe.

If you ask it either way, clinically and

presentation, there is a huge spectrum

histologically. Immunologically there is a

290

spectrum. It is just that they aren't always

parallel.

CHAIRMAN DURST: Erica.

DR. BRITTAIN: Just to follow up, are they

reasonably correlated, even if they aren't

perfectly concordant?

DR. KELLY: I don't believe so. I will

ask Alessio and --

DR. FASANO: I can't hear you.

DR. KELLY: You can't hear?

DR. FASANO: No. What was the question?

(General laughter.)

DR. BRITTAIN: We are talking about the

relationship between clinical manifestations,

immediate clinical manifestations, and I guess what

you can observe in a biopsy? The people who are

sensitive with respect to immediate reactions don't

look the worst on biopsies? Is that what you're

saying?

I'm asking are they fairly correlated?

Could it also have to do with the length of

disease? I would think the damage in the

291

intestines would be a function of a lengthy

disease, whereas the short-term reaction had

nothing to do with the length of disease.

DR. KELLY: Ciaran Kelly. As regards

duration of disease, well we seldom have the

opportunity to identify exactly when celiac disease

develops, except in children who manifest symptoms.

When an adult presents with celiac disease, it is

impossible to determine the duration of disease at

that stage. The other question is, Is there any

correlation -- can you hear me?

DR. FASANO: Yes, I can.

DR. KELLY: Is there a correlation between

histologic severity of disease and clinical

manifestation of disease in terms of

symptomatology?

DR. FASANO: The answer is no. It is

pretty much a straight no. Keep in mind that the

target organ of this autoimmune process is an organ

that has 200 square meters of surface, so it is

huge.

What do you define "severe" as? If you

292

define "severe" as 80 percent of the surface is

damaged, then it may be that we can have that kind

of correlation.

With our methodology right now, consider

that maybe there will be a change in that story,

but now with endoscopy we see the first few inches

of this 14 feet.

It can be absolutely destroyed what you

see. But there is absolutely no damage with many,

many times patch lesions where we go with the

endoscope, and these people are sick like dogs

because it is everywhere, all the way, to affect a

sizeable amount of the surface. Your processing

and absorption and digestion of foodstuffs is

dramatically affected.

That is the reason why there is no such

correlation on the clinical ground versus the

procedural ground, because the procedure cannot

give you the full breadth of the damage of the

intestines.

DR. COLLIN: May I comment?

CHAIRMAN DURST: That answer was given by

293

Dr. Fasano.

DR. FASANO: By now, because of the action

probably you know who we are, right?

(General laughter.)

DR. COLLIN: Pekka Collin. If I may

comment, I agree with Alessio that there is no

correlation because we have some patients with very

mild atrophy and severe osteoporosis, and then a

flat mucosa without any symptoms.

However, there is one correlation. Our

ultimate goal, if you look at who is sensitive and

who is not sensitive, if you look at how the mucosa

will recover, how is the mucosa recovery, if the

initial lesion is very severe and the patient has

remained undiagnosed for many, many decades, then

their recovery is very slow. Maybe in elderly

people it is seldom complete, but when the initial

mucosa is mild, I think we achieve full recovery

quite soon.

CHAIRMAN DURST: On that same question, I

think we have to address the uncertainty factors

and whether tenfold is sufficient using a safety

294

assessment-based approach. That is a reasonable

uncertainty factor?

DR. MALEKI: What is the starting point?

CHAIRMAN DURST: Soheila.

DR. MALEKI: Oh, Soheila Maleki. I assume

you would imagine what would be the starting point.

If you imagine it would be 200 or 20 and then

hitting the limits of detection for the methods at

this point and whether you can detect it, if you go

tenfold below 20, then I think you will surpass the

methods of detection, whereas if you are 100 or

200, then you may be able to say that would be

sufficient. I think will wait to see if the

statisticians differ.

CHAIRMAN DURST: Erica.

DR. BRITTAIN: This is Erica Brittain. It

also depends, I guess, we have been hinting or

talking about possibly having two levels,

gluten-free, which would probably be pretty close

to as low as you can go and maybe something that is

not so strict.

Obviously, you would use a different

295

uncertainty factor there for the two levels.

Again, it is the same discussion we had yesterday.

The 10 seems very arbitrary. It also depends on

which data set you start with. I mean, they all

have limitations.

DR. GONSALVES: This is Dennis Gonsalves.

It seems that a preponderance of data from all of

the different presentations suggest that 20 parts

per million for Canada and the various studies were

really more or less agreed upon. At that level you

don't get this reaction. If one looks at 20 and if

one looks a the uncertainty factor, it looks like

they licensed this at 20. If you have a tenfold

uncertainty factor, well, this was 18. I think

that there are data that suggest that --

DR. BRITTAIN: Two.

DR. GONSALVES: Yes. Ten percent up or

down?

DR. BRITTAIN: Down. It would be two.

DR. KELLY: Go down to two.

DR. GONSALVES: Two, yes. Well, so this

is five, so you can adjust that. Anyway, my

296

suggestion is that there really is pretty good

information that you are very close.

You can argue all of these different

exceptions, but at some point you have to decide

whether this uncertainty factor of 10 is

sufficient. I won't argue that based on what I

have heard it is pretty sufficient.

DR. KELLY: Ciaran Kelly. In fact, I

agree based on the data that is available, albeit

limited and albeit imperfect but scientific data is

always limited and imperfect, that it appears as

though there is agreement around the general range

that appears to be below a threshold for injury.

If there is broad agreement across the

data, perhaps an uncertainty factor of 10 might

even be considered excessive. I think there has

already been a sort of de facto uncertainty factor

enacted in going in other communities from 200 to

20, and that was largely based on concerns about

whether or not the 200 was low enough. I would

suggest it might be worth considering that.

Again, it depends on where you start. I

297

feel if you start at a conservative level below

which the scientific data that are available

suggest there is no evident injury, either by

symptoms or by histology, then that may be a

comfortable level without an uncertainty factor.

CHAIRMAN DURST: Okay. Marc and then

Erica.

DR. SILVERSTEIN: Marc Silverstein. I

would like to make a comment. It seems to me that

in medicine we have lots of uncertainty and

uncertainty from lots of sources. Some of the

uncertainty comes from bias and some of it comes

from confounding and some of it comes from

measurement error.

It seems to me that the rationale for

uncertainty factors that was applied to toxicology

for environmental exposures in our discussion

yesterday is we couldn't find a reasonable clinical

or biological reason to think that level of that

approach would be appropriate for IgE-mediated

immune reactions.

It seems to me that although we have

298

learned or heard about the non-IgE, cell-mediated

immune injury in celiac disease there is little

rationale, from what we understand about the

disease, to attach an uncertainty factor of tenfold

or whatever-fold.

I understand how a public safety

mechanism, it might be nice to have an uncertainty

factor, it doesn't seem to be consistent with our

understanding of either IgE-mediated immune injury

or cell-mediated immune injury for celiac disease.

It is kind of a comment.

For those who know more about celiac

disease than the biology of immune-mediated injury,

is there any reason to have a rationale for

thinking that you can measure the variation in the

response or the threshold for a response based on a

factor, whatever the factor might be?

I don't know what body of understanding,

whether from biology or medicine, would be

applicable so we are using toxicology here. I

would ask for some comment from those who study the

immune response.

299

DR. FASANO: If I may? This is Alessio

Fasano. We had a long discussion in '99 and 2000

when we were designing the study about what would

be the biology readout, because that is what it

really boils down to.

What would be a satisfactory readout to

make us comfortable in saying the immune system,

genetically skewed to react in not immune fashion,

will be turned on by "X" amount of gluten? A study

designed that should take three months took almost

two years to reach a consensus, because there were

different philosophies that were on the table.

There were people that say clinical, serological,

biochemical, histological, combination and

permutation, all of the above.

The reality of the story was that we

weren't on evidence-based on the retrospective

studies done where some of the people that

participated plus our other colleagues, and we

realized, one, clinical was absolutely not

reliable.

Two, the biochemical, the other antibodies

300

was also not reliable because we still don't know

the role of these other antibodies, the

pathogenesis.

We all agree, while there was some

disagreement about the statement I just said, we

all agree though that the final product of the

autoimmune process, i.e., the autoimmune biological

readout, is the damage of the intestine. That was

the only confidence parameter that everybody agreed

upon.

The reason why this was not a joke is

because unfortunately, based on that decision, the

only way they could make a statement in terms of

biological readout implied two endoscopies. That

in terms of study design was inartful.

I mean, not only do you have to go to

somebody that is healthy who goes on a gluten-free

diet and asks to have an endoscopy that he or she

has no business to do, but then we heard this after

three months. A wheat starch level is the only way

to do it.

DR. SILVERSTEIN: Could I follow up with

301

one question, then?

(No verbal response.)

DR. SILVERSTEIN: When you selected your

empirical doses of 10 milligrams and 50 milligrams,

was it based on a rationale that a ratio of a

5-to-1 dose, or was it sort of a sense that this

was a lower feasible dose and a higher feasible

dose of interest? I mean, how did you--?

DR. FASANO: As you can imagine, if we

were absolutely insane in how to design, also 6

months we discussed how much we should go. The

reality was you so package data in which the vast

majority of the North Europeans consumed, roughly,

150 grams of gluten-free-based grains.

If you take this European population and

extrapolate to the American one, because we want to

do a study as generalizable as possible, we consume

in general terms more than that.

Italians, it may be that they are at the

extreme of the spectrum, but definitely we want to

cover as much as we could. Dr. Catassi did a study

before in which he used 100 milligrams and clearly

302

showed the damage. Other studies were done as well

that clearly showed the damage.

We designed to take two doses -- because

we wished to do 10, 20, 30, 40 or 50, but that was

not doable -- in which we pretty much covered the

spectrum between the 20 and the 200 parts per

million because the Codex Alimentarius, you heard

Rhonda, is around number 7.

They have been discussing this for ages.

This 220 has been on the map there for quite a

while. The 50 and 10, based on a max consumption

of 300 grams a day, were chosen to cover the two

ends of the spectrum of the 20 and 200 parts per

million. That was the rationale.

DR. SILVERSTEIN: I would just comment

that it seems to me in looking hard for a rationale

for an uncertainty factor, we don't get an

uncertainty factor approach. We get a clinical

approach based on the knowledge of the exposures

that have caused injury.

CHAIRMAN DURST: Erica.

DR. BRITTAIN: I guess the questions, we

303

are going to have a second question about an

uncertainty factor for the short-term versus

long-term exposure. I know we were just asked the

first question, but sort of we are talking about

everything all together.

For me there is still that long-term

question at least from the clinical trial, which is

a wonderful study, but it only goes four months. I

am thinking that I don't know if cigarette smoking

and lung cancer are at all analogous.

Obviously, the longer you smoke, the

greater your risk is. If we were just studying

smoking for four months, we might not pick anything

up. That to me is my strongest concern where the

uncertainty factor would come in, because clearly

with respect to that study we have uncertainty

about long-term exposure.

I guess what is hard for me to evaluate is

the observational data that might support the

validity of the result of the prospective trial. I

don't really have enough details about it, and

perhaps strong enough to support it, in which case

304

there may not be all that much uncertainty.

DR. FASANO: May I comment on that? May

CHAIRMAN DURST: Yes.

DR. FASANO: The three months was also not

pulled out of a hat; it was part of a long

discussion how we come up with the three months.

You are absolutely right, if you asked me, "Are you

absolutely unequivocally sure that the fellow on

10 milligrams that do not react today will not

react in 10 years from now," the answer is of

course not, I'm not.

Why do we choose the three months?

Because if you are exposed to dangerous levels of

gluten after you have been on a gluten-free diet,

because your immune system is primed, 90 percent of

the people will react within three months. That

was the reason why we went to three months.

In other words, the vast majority of

people that have been on a gluten-free diet and

they are challenged because of diagnostic purposes

or because of cheating or because they said "to

305

heck with this diet, I want to go back and enjoy

myself," not only are we experienced in what you

read out there, but the vast majority will react,

not necessarily clinically, but they will react in

some shape or form within that period.

We are sure there are people that after 10

years they still don't react. These are the

extremes. Statistically, that is how we came up

with this three months. Does this for sure say

that in the long-run they are going to be all right

MS. KUPPER: The American Dietetic

Association --

CHAIRMAN DURST: Your name, please?

MS. KUPPER: My name is Cynthia Kupper.

The American Dietetic Association is in the process

of doing evidence analysis of the gluten-free diet.

Two of the questions they are looking at right now

are, Is the gluten-free diet useful in reversing or

stopping anemia and osteoporosis?

They are really struggling with these

questions right now, because a lot of the studies

306

come from Europe. They are actually thinking about

separating the American studies from the European

studies, which might help to address this question

because then we are talking wheat-starch-based,

gluten-free diets versus non-wheat-starch-based,

gluten-free diets.

CHAIRMAN DURST: Let's see, David first.

MR. ORYANG: Yes. David Oryang. Talking

about the uncertainty, you mentioned you could not

be absolutely certain that the person would not

react the next time after some time. There is

always uncertainty. Somehow we need to keep in

mind, I think, the fact that the study was done at

10 and 50.

The other thing is I were to ask suppose

the study was done at five, what would the outcome

have been? Could there have been reactors? If we

say that, yes, there could have been reactors, then

we definitely have uncertainty.

Since we do have uncertainty, then we need

to put some uncertainty factors around this

parameter, then the issue becomes what the

307

uncertainty factor should be.

I don't know whether the FDA

representatives maybe can say something about this,

but the whole issue of using a distribution for an

uncertainty factor, would that be considered

reasonable?

Maybe an expert panel could be put

together just to address that issue and at least

set bounds for what the uncertainty factor should

be and then look at the 95 percentile value of the

overall result when you divide by that uncertainty

factor, to determine the threshold value.

However, this is another issue of maybe

modifying the safety factor analysis methodology to

incorporate elasticity into some of those

parameters as opposed to just a plain value make it

a distribution. That is one alternative to deal

with uncertainty about uncertainty.

(General laughter.)

MRS. MOORE: Did you have a specific

question that you wanted the FDA to respond to?

MR. ORYANG: Yes, whether that has been

308

considered, whether there is anything in the

literature about that having been applied.

MRS. MOORE: Well, wait for a moment. Did

I see somebody raise their--?

CHAIRMAN DURST: Yes, Margaret.

DR. McBRIDE: Margaret McBride. It seems

like one of the --

CHAIRMAN DURST: Well, he is ready to

speak. Sorry.

DR. McBRIDE: Sorry.

CHAIRMAN DURST: He is coming out of the

huddle.

DR. GENDEL: Steve Gendel. From my

consultation with the experts, I have been told

that this is not necessarily a normal way that this

is done, but it is certainly something that can be

considered. One of the things that we are

interested in hearing from the panel are

suggestions about approaches such as that.

CHAIRMAN DURST: Okay.

Margaret.

DR. McBRIDE: One of the biggest

309

uncertainty factors is how much starch a person is

going to eat. We have to rely on some kind of

judgment and education from the consumer.

If we tell them how many parts per million

or set a level that is a certain amount of parts

per million, then how many pieces of bread with the

alternative flour they eat affects much more than

any, or possibly significantly more at least

equally with any uncertainty factor we try to put

into deciding a parts per million.

CHAIRMAN DURST: Marc.

DR. SILVERSTEIN: Marc Silverstein. It

seems to me that we have lots of precedents in

clinical medicine where we studied either the

effect of large doses for shorter duration, or we

set practical limits to the amount of resources

that can be placed in doing studies, and we make

decisions about medications and treatments based on

courses of therapy.

There might be one month, three months,

six months, to a year, yet many of these conditions

and exposures that patients will get, either their

310

treatments or medications or tests or radiation or

diet, are lifelong.

With regard to the issue of going from

short-term to long-term, short-term exposures and

effects to long-term outcomes, I think we should be

cognizant of, be aware of the fact that these are

individuals with a clinical condition, with a

disease, that have access to healthcare.

These individuals with the advice of their

physician are going to have a course of therapy

that could be periodic assessment of their response

to therapy, periodic assessment of their mucosa,

periodic assessment for the consequences of

long-term inflammatory disease.

I am less worried about the problem of

making inferences on long-term outcomes because, by

and large, these are patients with a clinical

condition who are having a dietary regimen under

the management of a physician.

It may be unfortunate that some patient's

experience has been, the physician has said, "Well,

now you should change your diet. Goodbye and good

311

luck." However, I think we should remember that is

not, indeed, the norm and probably is more the

exception I would hope.

In any case, I would think that we would

be able to make reasonable inferences based on the

short-term exposure, and three months seems to me

to be right now probably the upper limit of what

was feasible in a well-designed clinical study for

response to gluten.

CHAIRMAN DURST: Erica.

DR. BRITTAIN: I guess my question about

that would be I assume these patients don't get

routine biopsies or anything like that, or maybe

they do? Maybe that is my question. Without that,

I don't know how you would necessarily know if they

are doing badly.

DR. SILVERSTEIN: That is what medicine

and clinical research is all about. The physicians

-- those under their care, the patients -- will

come up with a course of therapy.

I am sure there are physicians based on

current and evolving data who will recommend either

312

no further surveillance or periodic surveillance or

surveillance at yearly or three yearly or monthly

or whatever it may be.

In the same way, for example, patients

with ulcerative colitis who have longstanding

disease and are at increased risk of colorectal

cancer are often in a periodic surveillance program

for that outcome. That would be separate from

saying how we should make recommendations about

pharmacotherapy for their disease.

Some people would say, okay, study a

reasonable period of time -- whether it is one

month, three months, or a year -- and let the

patients and the physicians together make their

best recommendations about long-term management.

I am just addressing this issue of

short-term and long-term. Clearly, you can't wait

20 years before you make any recommendations, and

experience will evolve as patients and physicians

together learn more about what works and what

doesn't work.

I do believe that we have many precedents

313

where we do have reasonable time periods -- whether

it is one month, three months or six months -- for

clinical studies and then make recommendations that

patients and their physicians can look at the

long-term outcome.

I guess I am repeating myself. However, I

do feel that there is a scientific basis and

precedence in many other areas for best information

for a short- or moderate-term studies to be the

basis of practice over the long-term.

CHAIRMAN DURST: Okay. Thank you.

Mark.

DR. NELSON: Mark Nelson. I don't know

how we want to factor in the natural experiment, if

you will, of labeling gluten-free products in

Italy. While it is not as elegant and as

controlled as the Italian study that we heard about

today, my understanding is we have seven years of

this labeling in Italy. We must have some

real-world experience about severity of symptoms,

change in symptoms, and change in prevalence in

Italy with that experience.

314

CHAIRMAN DURST: Dick Durst. I would just

like to make one comment on the uncertainty factor,

that is, we can certainly recommend that there has

to be a certain degree of uncertainty factors

associated with the recommendation, but it sounds

like from what I've been hearing that even a

tenfold uncertainty factor would be beyond what the

analytical methods can currently do. Therefore, it

is a luxury that we may not have the option at the

present time of setting.

Certainly technology in the future is,

hopefully, going to improve the point where we can

get down to a tenfold uncertainty below, say, that

20-part-per-million level.

DR. HEIMBURGER: Doug Heimburger. Related

to that, though, but the other problem is the

impossibility of creating products that are

significantly below that 20, even if you could

detect it. I'm not sure that new technologies will

necessarily change the true uncertainty factor with

regard to what is actually included.

CHAIRMAN DURST: Unless the processing

315

technology also improves.

(General laughter.)

DR. BRITTAIN: It sounds like maybe 10, I

mean, I don't know if I was hearing correctly, from

a practical point of view something like that might

be the absolute lowest you could go and still be

able to produce food. I don't know if that's

right.

DR. HEIMBURGER: Doug Heimburger again.

From what Steve Taylor was telling us it is not

just a matter of farming technologies, either. You

would have to revolutionize the entire agricultural

methods that are used.

CHAIRMAN DURST: The infrastructure.

DR. HEIMBURGER: That is just not going to

happen, particularly because they are built on

efficiency now probably being as efficient and

cost-effective as possible. Any move in the other

direction would have all kinds of forces against

it, including our pocketbooks.

CHAIRMAN DURST: Jeff.

DR. BARACH: Yes, Jeff Barach. I think

316

what we are really talking about here is risk

management and uncertainty factors, adding

uncertainty factors is one approach.

Another approach that we favor is really

to establish when the threshold is established, to

consider it as sort of an interim threshold, and

then with experience and more information perhaps

that interim threshold would change.

For us to assign uncertainty factors based

on numerical and environmental considerations at

this point does seem premature. With experience

and setting an interim threshold, we would have the

opportunity at some time to make an adjustment in

that, if we felt it was either too high or too low.

DR. KELLY: Ciaran Kelly. I wanted to

return to the question of timing and whether or not

three months is adequate to demonstrate a response

to gluten -- of course the longer, the better.

In reality, if you look at acute gluten

reactions, clinical symptoms, reproducible

symptoms, they tend to occur within a few hours.

If you look at the acute challenge studies,

317

morphology could be demonstrated to be abnormal

within hours of instilling toxic peptides into the

duodenum. Those very acute studies were able to

show abnormalities within a very short period of

time.

It would seem to me that understanding

that this response appears to occur within hours in

vivo, having individuals exposed for three months

it would seem to me to be more than adequate to

demonstrate any at least medium-term effects.

There is nothing, to my knowledge, to

suggest that an individual should be triggered to

respond at a later period in time who hadn't

responded earlier, albeit the fact that there are

always exceptions to that.

CHAIRMAN DURST: Erica.

DR. BRITTAIN: So are you saying that

someone, say, who is not very compliant and does

eat a lot of gluten, he would not get worse over

time if he had more exposure to gluten?

DR. KELLY: Yes, he would. I mean, if

somebody has severe celiac disease and they are not

318

diagnosed or don't go on a gluten-free diet, they

can become malnourished. Before a gluten-free diet

was available, the condition was often fatal. Yes,

it can accumulate.

What I'm saying is that in an individual

who is on a gluten-free diet and well treated, the

response to inadvertent and purposeful gluten

intake, if there is a reproducible, reliable

clinical response, it is very rapid.

In those individuals who respond in a

particular way, the response is quite rapid. As we

heard, it is a delayed Type 4 reaction that occurs

within a few hours of exposure, typically.

CHAIRMAN DURST: David and then Soheila.

MR. ORYANG: Yes. David Oryang. Yes, I

think in an individual we can say that we expect

pretty much the same thing. What about between

individuals? Are we going to expect the same thing

between individuals?

The safety factor that we are talking

about, the intraspecies safety factor, is looking

at, okay, how will he respond versus this person to

319

the same challenge. Can we say with a hundred

percent certainty -- in other words, a safety

factor of one -- that they are going to respond the

same?

CHAIRMAN DURST: Soheila.

DR. MALEKI: Soheila Maleki. One thing in

answer to yours, they are talking about being on a

gluten diet and they are talking about a certain

limit, not sitting around and eating a bunch of

known gluten. The study is about getting a limited

amount. It is showing within 45 people -- I think

that was the study group, right?

CHAIRMAN DURST: Thirty-nine.

DR. MALEKI: Oh. I'm sorry?

THE COMMITTEE: Thirty-nine.

CHAIRMAN DURST: It was 39.

DR. MALEKI: Thirty-nine people, close

enough. Anyway, within 39 people that they tested

with this particular limit over 3 months, given

that they would have a severe reaction and it would

show within 3 months, that they didn't. Just

clarifying that, because you were saying if they

320

eat a lot of gluten.

DR. BRITTAIN: Oh, no, what I'm getting at

is what is the effect of cumulative exposure. Just

like smoking cigarettes for many, many years,

you're not going to get immediate damage from

smoking -- you may get slight damage, but it might

not be incurable.

I can just quickly describe a study design

you might do. I know it would be hard to do, but

perhaps a study for a year and do biopsies every

four months or every six months, so that you could

see if there is a change over time.

If the change happens right away and it

doesn't go down any further, then you know, you

have some confidence that there is not going to be

a continued change over time, but if you see a time

relationship, then you would know.

DR. FASANO: Can I make a comment on that?

May I, Mr. Chairman?

CHAIRMAN DURST: Yes.

DR. FASANO: Okay. Alessio Fasano here.

The parallel between smoking and celiac disease I

321

don't think that is really pertinent, because

celiac disease is an autoimmune disease. It is a

step-by-step process. The variability is how long

it takes for Individual A versus Individual B to go

from point 1 to point 2.

The steps are well-known. You are exposed

to gluten, antigen presented in cells, and we see

them represented in intraepithelial lymphocytes,

there will be inflammation, cytokine production,

intracellular cells, and the damage.

People can do this journey in a few hours;

people can do the journey in a few days; people can

do the journey in a few months. The question is,

When do the vast majority of people go from point 1

to point 2? That is what I was alluding to.

When people genetically predisposed to

celiac disease are exposed to gluten because

challenged, a toxic amount of gluten, or challenged

because they decide to abandon the diet, the vast

majority of the people, they react within that time

limit. When I say "the vast majority," it is

because again there is still the possibility and

322

there are reports that you can take much longer.

It is the same story on clinical grounds.

There are people that are exposed to gluten and

they develop symptoms as kids after a few months or

a few weeks because and then the damage will become

clinically apparent in the next few weeks, but

there are people that it would take nine years.

How do you explain the variability? We

don't know. The first steps in the process of

damage to the intestines are at the very beginning

and this is likely to switch on and off. It goes

on and it goes on.

That is the reason why I am pretty

comfortable with the three-month business, because

it is a very comfortable interval in which you

should see the immune system react with parameters

that can be biologically looked at as we did with

histology and morphology.

CHAIRMAN DURST: Dick Durst. I would just

like to say that, presumably, the whole point of a

threshold amount of the gluten or whatever the

problem is, is that it is a level below which the

323

disease has not progressed. It doesn't get any

worse.

Using, even though it is a bad analogy,

the cigarette, if you can show that, okay, smoking

one cigarette a day is not hazardous, but smoking a

pack over time will really cause serious problems,

lung cancer or what have you, then the threshold is

that one cigarette a day.

Presumably, if you have done good studies,

it would show that, all right, that's safe to do.

However, I think if we set a threshold at a point

where the disease does not progress, even on a

short-term basis, then over the long term that

should still hold.

Marc.

DR. SILVERSTEIN: I think this is an

insightful discussion because I think we are

talking about immune-mediated injury, IgE, in the

case of the allergic diseases or cell-mediated for

celiac disease.

We are talking about carcinogenesis on

perhaps a multistage process with genetic mutations

324

due to a variety of exposures, or we may be talking

about toxicologic injuries due to environmental

factors.

I think we've got a conceptual model,

certainly based on environmental toxins, that we

are extrapolating to these immune-mediated

processes. I am cautious about that. Because I

can't see that rationale coming from science, in

terms of the application of the risk management for

environmental exposures, toxins, being applied to

these immune-mediated conditions. That is why I'm

being fairly skeptical here.

The discussion has helped me think about

what are the underlying mechanisms in

immune-mediated carcinogenesis with damage to cell

growth or other toxic mechanisms due to the

environmental accumulation of small molecules.

CHAIRMAN DURST: Soheila and then Suzanne.

DR. MALEKI: Soheila Maleki. Bottom line,

I think that we can talk about this, and all the

complications which of course exist such as they do

in the case of IgE-mediated allergy, but bottom

325

line based on the best data available, we have to

accept this as what we know.

This is the level within the three-month

study that has been done prospectively and the

retrospective study that they have actually

determined maybe seemingly short-term, but

clinically relevant according to Ciaran.

Of course, then, we have the limitation of

the methodological methods available to us. We can

discuss this for a long time and not really be able

to still answer that question as far as what fold

we would put that at.

I mean, I agree with you that it is very

complex and a multihit type of disease much the

same as cancer would be. With the data available

-- and, I mean, think it is good data -- that has

shown that these are some of the limits they react.

CHAIRMAN DURST: Suzanne and then

Margaret.

DR. TEUBER: Suzanne Teuber. It seems to

me that with the population data that Dr. Collin

presented you actually have a higher level already

326

seen in the population with some long-term data

that was very reassuring.

It seems that if you use that as your

starting point, then apply an uncertainty factor

that we don't really know to that, and then your

data comes in from Italy that is very reassuring,

that a factor that has been theoretically applied

to that upper limit is coming down even lower over

a three-month period with no immunologic reaction

seen, that is very, very reassuring.

I think that would imply that the levels

that should be looked at initially would be what

Dr. Collin was discussing with this additional

safety.

DR. COLLIN: May I have a comment?

CHAIRMAN DURST: Oh, okay.

DR. COLLIN: Pekka Collin. Another issue

is that, if one stay on the safe side, if there is

somebody very sensitive, some celiacs who are very

sensitive, they might react after three months; so,

the period would be too short.

All our retrospective studies show that

327

ultimately we can achieve a complete response. We

have methods to detect those patients who might be

very, very sensitive.

One mechanism is that we take usually, at

least in Europe or maybe also in the United States,

take one biopsy after one year of a gluten-free

diet.

If there is no clear improvement, then we

can concentrate on those patients who might be

truly sensitive, or, as I said, who might take some

extra gluten, excess gluten not PPMs but grams of

gluten.

Therefore, I think that it is very high,

that uncertainty factor. Maybe it is not very

relevant. Maybe you go to the conclusion of zero

level instead of a little bit higher level, which

is very well tolerated by the vast majority of

celiac disease patients.

CHAIRMAN DURST: Finally, Margaret?

DR. McBRIDE: Margaret McBride. I think

in a certain way that the data that we heard this

morning from both studies included some long-term

328

data. We know what the biopsy results were in

folks who had adhered to a diet, a gluten-free

diet, in both countries.

In one country, for seven years

"gluten-free" meant below 20 parts per million, and

we assumed that they were eating somewhere between

100 and 200 or, for real past lovers, 250 or

whatever.

We can calculate the amount of the number

of milligrams of gluten to which they may have been

exposed over a seven-year period. In fact, their

biopsies were like those of normal folks, at least

in regard to the height/crypt depth ratio.

Likewise, in Finland we have the same kind

of data. In other words, we almost really have a

NOAEL from both studies that is a long-term NOAEL.

In Finland, I think Peter and I, too, was under the

impression that there was a set limit, but, in

fact, I believe that is not true, that the limit is

in fact the current international group limit.

In fact, when the gluten-free foods were

tested, all but two were below the 100 part per

329

million; but if you look at that slide, a good many

of them were down closer to the 20 part per

million.

We have data without any uncertainty

factor really needed that suggests that if you

adhere to a gluten-free diet in those situations,

that you do quite well.

CHAIRMAN DURST: Okay. May I suggest we

move on to number three?

Do you have another--?

DR. WASLIEN: Well, no, this is still part

of this. I hope that we are counting as celiac

disease only people who have shown clinical

symptoms, right? We are not counting silent?

Because if you count silent gluten sensitivity,

you've got a much, much higher level of gluten that

is still acceptable, right?

DR. MALEKI: Soheila Maleki. Well, if

it's silent, then they are not avoiding gluten.

DR. WASLIEN: No.

DR. MALEKI: They aren't eating

gluten-free.

330

DR. WASLIEN: It said two separate groups

of patients, and I know we talked about the range

of latent to silent to acute cases. The range of

acceptable intake between acute cases or ongoing

cases and latent is very large. I don't think we

count that in the disease.

If we count only those who have "symptoms"

of celiac disease, we do have this 100 parts per

million, it looks like, level. If we are talking

about the silent, we are talking about any level;

or the latent level, we're talking about any level

because they haven't shown any symptoms yet.

I think it goes back to this. Yes, we

said yes, there were differences in the patient

groups. We need to go back and say yes, there are

differences in the patient group, and this is the

group we are looking at, the ones who have acute

conditions.

CHAIRMAN DURST: Ciaran.

DR. KELLY: Ciaran Kelly. To respond to

that, there is a lot of debate as regards whether

some, many or all individuals with silent celiac

331

disease should be or should not be on a gluten-free

diet. I think that is far beyond what we are going

to decide.

I think what I would suggest we should

think of this as in those individuals who decide to

go on to a gluten-free diet, whether they be silent

celiac disease or be individuals with severe

malabsorption, regardless of that, in those

individuals who decide to go on a gluten-free diet,

what appears to be a safe ingestion of gluten for

them.

CHAIRMAN DURST: David.

MR. ORYANG: Yes. Just going back to the

safety factor, assuming a safety factor of one, it

would indicate that we have a hundred percent

confidence that the parts per million, the 20 parts

per million, as an example, is an absolute value

below which we don't believe people would react, or

no one faced with a challenge below that would ever

come up with celiac disease.

I don't think that we can put ourselves in

a position to say that; there is uncertainty. We

332

do need to look at it scientifically and be able to

communicate our uncertainty to people when we do

these kinds of assessments.

The way that you do that is through

something like a safety factor. This specific

approach uses the safety factor to communicate the

degree of uncertainty we have, in other words, to

ensure that we say, "Well, maybe they could react

at five times less of a dose." Or, "Maybe that is

the reasonable dose we believe that 95 percent of

the people will react at."

Somehow we need to be able to communicate

that as opposed to just saying, "Well, this is the

value." I think somehow it needs to be

communicated and transparent. Otherwise, I wouldn't

have confidence in it if someone just told me,

"Well, 10 is it." I think it really does need to

be considered. We can keep talking about it later.

The approach is basically so that people

can have confidence in the fact that we have

clearly evaluated the data, and based on our

evaluation this is the degree of certainty we have

333

or the degree of confidence we have in the clinical

studies that were done. I think it does need to be

considered.

CHAIRMAN DURST: Dick Durst. I agree that

we can't be absolutely certain of the level and

there has to be some uncertainty associated with it

in the same way that we can't give a number for a

threshold, that is not our job here.

MR. ORYANG: That is what I thought.

CHAIRMAN DURST: We are looking at the

approaches. I think our suggestion to the FDA is

that they do attach some kind of uncertainty

factor, but that is something that, again, requires

more study to find out what level it really needs

to be.

Marc.

DR. SILVERSTEIN: Marc Silverstein. It

seems to me that if you have observational data or

you have clinical data from a randomized trial you

will have number of patients per person years of

observation and observational studies, or if you

have two different doses, you will have risk

334

ratios.

Obviously, in clinical trials you have a

metric, and so it seems to me that rather than

saying whatever we do in looking at the literature

we pull out a number, the number is going to come

from a clinical study.

There will be data available in the study,

either presented by the primary authors or maybe

original data may be available in addition from the

office -- and a statistician or an epidemiologist

can look at the reported data and then calculate

confidence intervals around the various metrics in

the data.

It seems to me that it is not -- yes,

there are reports of values, but we are not just

saying, "Oh, we take 20 and that's it." You walk

away, and there is no uncertainty.

Of course, there is uncertainty, but it is

retrievable, either the reexamination of the

original data from the author in the publication or

by a review of the data. It could be that these

data have some inherent uncertainty, as most rates

335

or rate ratios or effect differences or risk

differences do.

I am not worried about getting some

estimate, because I believe that we will be able to

derive confidence intervals or measures of

uncertainty from whatever studies that these

thresholds come from.

CHAIRMAN DURST: Yes. I would like to

move on now to number three, so that we can finish

up at a reasonable time. This has to do with the

susceptibility to oats. I guess we heard that it

is pretty inconclusive at this point.

Most of the studies indicate that there is

no problem with oats, whereas there were couple of

studies they were. From the question here, if

there is no certainty as far as the susceptibility

to oats, what additional data is needed to draw

such a conclusion?

Would someone like to--?

DR. MALEKI: I'll start.

CHAIRMAN DURST: Soheila.

DR. MALEKI: Soheila Maleki. Again, the

336

majority of the studies by far have shown that

there is no cross reactivity with oats. However,

there have been some individuals that have felt

like they had reactions to oats, and documented.

Bottom line, and according to what

Dr. Taylor was talking about, that if they are

severely sensitive and there is contamination in

the product that comes from the far, then they

could have had wheat exposure with an oat

contaminant, otherwise oat exposure with a slight

wheat contaminant that could have seemed like an

oat exposure, or they could have had a real

reaction to the oats. Either way you look at it, I

think it is going to be hard to determine until

there is more data.

Thank you.

CHAIRMAN DURST: Marc.

DR. SILVERSTEIN: I don't know the

original literature, but if these studies are a

series of observational studies or case control

studies or cohort studies, there are meta-analytic

techniques for looking for heterogeneity and

337

homogeneity in published studies, and then if there

is homogeneity, making overall assessment so you

have an increased ability to make inferences based

on multiple published studies.

It seems to me that this would be a

wonderful area. Maybe there is a published

meta-analysis of these exposures; but, if not, it

seems to me that the recommendation of the type of

data that could be helpful might be a

well-constructed meta-analysis of the published

data for the oat exposed. If we are thinking about

it today, I'm sure there is somebody already

working on it; and if not, somebody should be.

CHAIRMAN DURST: Okay. Anything else on

that one?

DR. KELLY: Ciaran Kelly. Well, just a

comment, and that is that the prospective,

published studies are all in agreement regarding

safety. The other studies that suggest that there

are maybe a small proportion who are sensitive,

essentially these case series studies are in very

limited groups of patients.

338

Basically, those individuals appear to be

outliers, and so it is not going to be possible to

devise a study, a prospective study, to identify

the very few patients who appear to be sensitive to

oats.

The problem is you can't just make a

blanket statement, "Nobody is sensitive to oats."

It is unfortunate, but it does appear that there

may be a small number of people.

DR. SILVERSTEIN: No, I fully understand.

What I'm suggesting is if you've got 10 studies

that show there is, essentially, no numerator

events, you are going to have a more precise

estimate about how close to zero the observed data

is when you do your meta-analysis and you

appropriately weight the estimate of zero and a

confidence interval around it by the sum total of

patients in all of the studies.

CHAIRMAN DURST: Dick Durst. While it is

inconclusive, should the FDA err on the side of

caution and include oats, then, or is the

preponderance of the evidence against oats

339

sufficient?

Mark.

DR. NELSON: Yes. Marc Nelson.

Basically, you asked the question I was going to

ask, which is considering that these individuals

seem to be very few and far between, should oats be

included, and does that limit further the broader

number of celiac patients enjoying other products?

As I understood it, these individuals if

they in fact were sensitive to oats, they would

obviously demonstrate symptoms, and then, as our

clinician colleagues mentioned, they would further

work with that patient to determine what the cause

actually was.

CHAIRMAN DURST: Jeff.

DR. BARACH: Jeff Barach. I think if we

go back to where we started with FALCPA and our

charge here of looking at major food allergens,

what we have found here is a subpopulation of

people who are allergic, to celiac disease. That

doesn't really fall under my thought about a major

food allergen. It is kind of a sideline issue that

340

maybe is interesting and important but not for this

group.

DR. KELLY: Ciaran Kelly. I wanted to say

something similar in a different way, and that is,

that the number of individuals who may be have an

immunologic reaction to oats to cause disease is

tiny.

A much larger issue is contamination of

oats by grains and proteins that are known to be

toxic. I think we shouldn't focus on a very, very

small subpopulation and forget the big picture.

DR. COLLIN: May I comment?

CHAIRMAN DURST: Okay.

DR. COLLIN: As to the studies in the oats

business, as you mentioned, there are several

studies and they almost all are randomized. There

are some studies which are five-year studies. That

was a continuum for randomized studies.

However, there are also some studies where

in Sweden they advised the patient to take 100

grams of oats per day. I can assure you that is a

huge amount. It was a study by Strsrud, and they

341

succeeded. I don't remember whether it was half a

year or one year. They did not see any adverse

effects, even when the patients took five times

more than our patients are willing to take.

As I mentioned, they took 20 grams, but in

Sweden they tested in a randomized manner with 100

grams. It seems that pure oats is very, very safe.

Of course, contamination is a problem. It is also

a problem in corn and rice. If we are talking

about contaminated foods to celiac people, then we

are talking about wheat, not about oats.

However, there is that small study by

Lundin. I would actually ask, Don Kasarda, as you

mentioned oats -- wheat has been studied most

thoroughly of all -- and celiacs, would it be

possible that also some people who eat corn or rice

would have similar reactions as those three

patients? What is your guess?

My name is Pekka Collin. Sorry I didn't

mention that.

DR. KASARDA: Don Kasarda here. I think

that is an excellent question, and one that I've

342

thought about a lot. We just don't have the

information. It might turn out that if you studied

corn that you would find, because there are many

celiac patients who say I can't eat corn, the same

thing as Knut Lundin with oats.

I would like to just make one comment on

something that Ciaran Kelly mentioned. Yes,

contamination is a big problem, but we could

possibly remove the avenin fraction, which only

makes up about 10 percent of the proteins, with RNA

interference.

We could possibly silence 90 to 95 percent

of the avenin genes or the expression of the avenin

genes to get rid of the avenin fraction. This way

at least, I mean, you still have to deal with the

contamination problem, but that is dealable with in

my mind.

However, if you don't deal with the avenin

question, if you accept the Oslo results, there is

always the question, "Well, will I react to oats or

not?" I think it would be a good thing to do, but

we don't have any money.

343

(General laughter.)

DR. BRILEY: Margaret Briley. I think it

is kind of interesting when we start talking about

contamination. When you have to think about going

back to the farm, wheat in most part of the

United States is not really grown on the same land,

generally, as corn and rice is not there, so the

contaminations are going to have to be at the silos

and at the mill area. It might be that would be an

easy way to solve it, other than having to do an

expensive kind of study.

CHAIRMAN DURST: Soheila.

DR. MALEKI: Soheila Maleki. One comment

to Margaret is that I virtually like impossible,

especially at the silo level.

The comment I wanted to make is that we

have to keep in mind here that these are maybe

three or four patients and outliers. We can't

forget that we are looking here at what is best for

the general and the majority of the population and

don't revert to going into two or three outliers,

even though they might have been true cases. That

344

is something that I don't think we are charged with

looking at as Dick Durst commented.

CHAIRMAN DURST: All right. Let's move on

again to conserve time here. The next question,

number four, has to do with the risk of developing

consequences such as cancer and increased

mortality. Would anyone like to address that one?

You're shaking your heads no.

(No verbal response.)

CHAIRMAN DURST: Okay. Let's skip that

question.

(General laughter.)

DR. BRITTAIN: We haven't heard any data

on this, have we?

DR. MALEKI: Exactly. Soheila Maleki. As

far as I know just physicians, and some of these

physicians like Dr. Kelly can comment, I know they

see patients that have complications as we all

heard based on the talks they gave today that, yes,

they can lead to cancers, they can lead to other

complications and mortality and morbidity in a lot

of different cases. Yes, clearly I would say the

345

answer is yes, because of clinical observations.

DR. BRITTAIN: Yes, that there are equally

at risk?

(Simultaneous discussion.)

DR. MALEKI: I don't think they are all

equally at risk.

DR. BRITTAIN: Then the answer is no.

(General laughter.)

DR. MALEKI: I mean, yes, that they are

actually these cases that are seen. Are they all

equally at risk? No. Clearly, if you can put them

on a gluten-free diet or an appropriate diet, then

they don't see as many of these risks. Like I

said, the clinicians can comment more about the

percentages and the people that they see.

DR. FASANO: May I?

CHAIRMAN DURST: Okay.

DR. FASANO: Alessio Fasano here. I think

that the way that the question is posed is kind of

deceptive. Focusing on the mortality part, yes, we

know that the mortality is twice as much in the

general population in untreated celiac. The

346

general population is not going to go on a

gluten-free diet.

I don't think this issue is as much the

mortality as the morbidity. For the mortality, of

course the answer is no, not everybody is at the

same risk.

However, if you talk about morbidity, that

also equals quality of life, I also believe that it

will be undisputable to everybody that not only do

science with celiac disease but see patients, the

answer is undisputably yes.

I mean, these people will pay a price. We

are talking about the symptomatic. Otherwise, why

do they come to you? There is definitely increased

morbidity.

This morbidity can be reversed if, for

example, we are talking about anemia. It may not

be reverted if you are talking about short stature

and you missed the diagnosis, because that person

remains short for his or her life. That is

undisputable.

I believe, again, this question should be

347

rephrased a little bit. Mortality? No, not

everybody is at the same risk maybe. The

morbidity? Definitely, everybody is at risk.

Morbidity depends on from individual to individual.

That is the way that I see it.

CHAIRMAN DURST: Pekka.

DR. COLLIN: I a little bit disagree with

you that the mortality is twofold in celiac

disease. We have to remember the iceberg

phenomenon. It was in 1950 or 1967 it was told

that 15 percent of patients with celiac disease

eventually developed lymphoma. Now we know that it

is 1 or 2 percent.

Still, we do not detect all patients with

celiac disease. I think in terms of mortality and

also in terms of morbidity there is a clear bias to

the most severe cases. Now, what is in clinical

practice? Only the most severe cases will be

detected.

Pekka Collin it was.

CHAIRMAN DURST: Marc.

DR. SILVERSTEIN: Marc Silverstein. If I

348

were going to make estimates or inferences about

subpopulations at risk, I would want to see a

long-term, cohort study in which I had strata of

exposures, knowing full well there is always some

misclassification in epidemiologic data, but then

want to know for each strata of exposure what was

an element of the relative risk of an outcome

whether it was death or some long-term

complications.

While I certainly believe clinically these

risks are not uniformly distributed across people,

but when you make inferences about subpopulations,

you have to be able to characterize your

populations.

You could characterize your populations

based on serology, you could base it on extent of

involvement in inflammation in the small bowel.

You could base it on other characteristics of the

patient, whether it was age or gender, duration of

clinical symptoms, whether you had specific markers

of genetic exposure -- a whole variety of types of

putative potential variables you might look at.

349

Unless a clinical study, a cohort study,

then shows me the risk and how the exposure groups

are characterized and what those risk ratios are, I

wouldn't have a basis to say that the risks are

unequal, although we all believe in the real world,

as we get more information, the risks are unequal.

I haven't heard yet any particular

markers, even gender-specific subpopulations, of

celiac disease patients who would be at increased

risk. Although I certainly believe there may yet

be discoverable subpopulations, I don't know that

there are any that we've heard about.

MRS. MOORE: In the interest of time and

for our code of conduct, we are going to have to

ask the guest speakers to only speak when they have

been given a question.

CHAIRMAN DURST: Erica.

DR. BRITTAIN: Yes, I guess I certainly

agree with what you just said. I guess I am

wondering what the rationale -- how does it relate

to the threshold question? Do you think the idea

is that if there were individuals at increased

350

risk, they might want to have a separate, a

different threshold, or you they might want to base

your threshold on those individuals? Is that the

motivation for the question? Otherwise, I am not

quite sure why we are addressing it.

CHAIRMAN DURST: Steve, can you address

that?

DR. GENDEL: [No microphone.] I don't

think it is necessary at this time.

CHAIRMAN DURST: Okay.

DR. BRITTAIN: He doesn't feel it is

necessary to.

CHAIRMAN DURST: Yes, he is probably

right, we're not dealing with thresholds.

Do you want to--?

DR. KELLY: Yes, just a comment.

Ciaran Kelly. I would agree. To sort of bring it

together, what I would say is clearly individuals

with celiac disease have different outcomes.

However, we don't have information as to what

specifically determines that. As far as we are

concerned, from a practical perspective, the answer

351

is yes, as far as we know all individuals are at

equal risk, insofar as we can't identify

individuals at greater or lesser risk. Therefore,

when it comes to terms of management, we give all

of the individuals the same advice.

DR. HEIMBURGER: Right. Doug Heimburger.

If we interpret the question to mean, Is there a

subpopulation that is at lesser risk or at lower

risk? We don't know that there is, so we need to

proceed, assuming that all patients with celiac

disease are at higher risk and none lower than any

other from any data that we have.

CHAIRMAN DURST: Good. All right, let's

move on to number five. "Is evidence of minimal

intestinal pathological change -- following a

gluten challenge, an appropriate symptom upon which

to base a LOAEL for long-term consequences?" Are

there other biomarkers that would be more accurate

predictors?

DR. HEIMBURGER: Doug Heimburger. The

clinicians here are all unanimous in saying that

the word "symptom" in there should be changed to

352

"signs."

(General laughter.)

DR. HEIMBURGER: It has to be changed

before we can even discuss the relevance.

(General laughter.)

CHAIRMAN DURST: Consider it done.

DR. HEIMBURGER: Okay. Consider that

done.

CHAIRMAN DURST: That was your comment?

DR. HEIMBURGER: No.

(General laughter.)

DR. HEIMBURGER: What was the term Ciaran

used yesterday? It burns a hole. There was a hole

burned in my mind, and that has to be patched.

CHAIRMAN DURST: Ciaran.

DR. KELLY: I think that in terms of a

sensitive marker of celiac disease activity and the

most widely accepted marker of celiac disease

activity is probably histologic change. It goes

back to the definition of celiac disease, which

"celiac disease" is immune-mediated enteritis.

The presence of intestinal inflammation

353

and the consequences of inflammation, which are the

architectural changes that you see in celiac

disease, define the disease. Therefore, I think we

are on fairly solid ground, if we use histologic

criteria and morphometric criteria to identify a

gluten reaction.

Certainly there are other criteria that

can be considered such as clinical criteria and

systemic, immunologic responses such as antibody

responses, as well as interferon-gamma responses in

peripheral blood. I am sure I'm forgetting

something. Oh, there are other intestinal

permeability studies.

There are actually quite a lot of

biological markers that could be considered, but I

think the field considers histology to still be not

perfect, but the closest thing we have to a gold

standard.

I don't know if Alessio or Dr. Collin

would like to comment on this? No?

CHAIRMAN DURST: I think that summed it up

very well.

354

(General laughter.)

CHAIRMAN DURST: Any other comments on

that one?

(No verbal response.)

CHAIRMAN DURST: Yes, I think that was

quite good.

REVISITING FOOD ALLERGENS

CHAIRMAN DURST: Okay. I think we have

gotten through these questions. I did mention that

we would like to address a few more points on the

food allergens. A couple of the members have

indicated they have some additional comments to

make.

Erica, do you have one?

DR. BRITTAIN: I would like to say, I

think I mentioned it earlier, more today but maybe

even more yesterday that it seems like if there is

a way for the information to be given, like, in

terms of the allergic patients at least, that there

are peanuts in it in some quantity," to have that

information.

However, then there should be a second

355

threshold which is the threshold that the typical

allergic patient has, "You are safe to eat it, if

you are a typical allergic patient." Kind of like

what we are hearing today about the gluten-free and

the low-gluten.

For those patients who want to take the

most conservative approach, they have the

information that there is an extraordinarily low

level of peanuts or whatever the allergen is.

There could be another standard so that a

more typical patient could say this is the level

that would be safe for them. To me that provides

more information than a single threshold, which may

be too strict for the typical patient and not

strict enough for the most extreme patient.

CHAIRMAN DURST: Okay. I think Soheila

has a comment.

DR. MALEKI: Soheila Maleki. I think it

is a lot of difference actually between the two.

In the case of peanut allergy or food allergies,

the amount that you react to or the type of

reaction you have can drastically change from one

356

exposure to the next. It is not even comparable.

Plus, the consequences can be immediate mortality.

A celiac patient can decide the level they

want to be exposed to or how much they can handle

based on experience of having gluten-free food

knowing that it is 20 parts per million or 200

parts per million. A peanut allergic person, if

you tell them it is low peanut, they won't touch

it.

You can pretty much ask anybody. I don't

know if Anne is still here? She is probably gone.

Any food-allergic person will tell you that if you

gave them a low-, two-level thing that it wouldn't

make a bit of difference to them at all, they would

definitely not touch it.

DR. BRITTAIN: Again, I'm worried that

whatever threshold, if it is a single threshold, it

depends on how you want to set it. Do you want to

set it for the most extreme person or those with

almost no reaction? Then you are limiting the food

choices of the more typical allergic patient.

DR. MALEKI: Soheila Maleki. I don't

357

think there is like a more "typical" -- a peanut

allergic person would not eat anything that said

"low levels of peanut."

CHAIRMAN DURST: Mark.

DR. NELSON: Mark Nelson. I'm thinking of

your concept in terms of working with labeling and

communicating with the consumers. My sense I that

would be more confusing and, consequently, much

less helpful than to have some threshold below

which you could effectively label it

"allergen-free" or there would be no consequences.

Now, that threshold may need to be

different for different allergens, because of the

severity of the response. I think that is really

something that would be helpful for a consumer.

Above that level, then, you would have to label the

presence of the allergen because the typical

consumer would respond.

DR. BRITTAIN: I mean, I know we heard

today someone speaking in terms of the gluten, but

she said, "I know you shouldn't label it,"

referring to the discussion yesterday, it would be

358

a disservice to most of the community if you set it

at a level that is for the most severe."

That is what I'm concerned about. There

are kind of two --

DR. NELSON: Yes. This is Mark Nelson.

You're right, I mean, that is FDA's responsibility

to set that number. I think that from a pragmatic

and a practical standpoint and a workable

standpoint, we might end up with something that was

described about the hypoallergenic formulas for

infants.

There is a confidence level that 90

percent of the population will not respond. That

is where population is taking all of their

nutrition from one food; their total food source is

that. It serves the great majority of that

allergic population.

In working with their physician, if they

are part of that 10 percent where they do respond,

then they just have to work harder to find a better

food supply that they can accept.

CHAIRMAN DURST: Okay.

359

David.

MR. ORYANG: Yes. David Oryang. Yes, not

being a biologist by background, just listening to

the speakers yesterday and today, I can see clearly

that in the case of gluten, it is easier to define

the NOAELs. Those that have celiac disease are

more willing to go for the challenge tests, because

the symptoms are not acute.

I mean, they don't have acute signs, or I

will say symptoms of disease, whereas with some of

the other food allergens -- peanuts, and so forth

-- I mean, if someone is going to suffer from

shock, you are not as likely to go for those

challenge tests. The data is not available in the

more acute allergens as it is in celiac disease.

As far as our approaches are concerned,

you see an analogy in that when you can't define

the NOAEL correctly, because of no data out there,

if you can't get enough subjects that are very

reactive to come for the challenge test, then you

can't get good data on the NOAELs.

Consequently, you can't set any safe

360

threshold directly using the safety approach, so

you have to resort to other methods. With gluten

and celiac disease, I think the safety approach

seems to be a reasonable approach. If we can work

on the issue of the uncertainty factor.

Because the challenge tests are there, I

think there is some kind of agreement as to the

parts per million. I mean, at least there are some

values that we can see have already been

established.

However, with the food allergens, I don't

think that is really there yet, because not enough

challenge tests have been done or can be done, so

we have to resort to other methods.

I am just trying to look at it from a

methodology point of view and just point out these

differences, so that we can maybe think about it

and see what other approaches can be used in

specifically the food allergens.

I think the gluten, I think all of us have

come to a consensus that the safety approach is

reasonable, at least so far I don't feel

361

uncomfortable with it. In the other case, I think

other approaches are better served.

CHAIRMAN DURST: Suzanne, did you have--?

DR. TEUBER: Suzanne Teuber. Again, in

terms of the individual response, I think there is

a greater uncertainty factor that applies to the

individual in this case because of day-to-day

variability. Again, those factors of exercise,

alcohol, illness, unstable asthma all play a role.

We have absolutely no data on repetitive

challenges for threshold in those threshold studies

that have been done on whether there are

differences in the circadian rhythm on the

threshold, on the NOAEL or the LOAEL, in the few

studies.

I think that food-allergic patients, as

Dr. Maleki said, if there were any detectable there

based on whatever limit the FDA chooses, those

consumers will choose to absolutely pay it safe and

avoid it.

They are not going to miss it knowing

that, "Oh, it only had 200 micrograms that I could

362

have tolerated versus 2 milligrams that I would

have reacted to."

They will be happy that it is labeled,

that it had something, and that they can avoid it

safely. It is a little different perspective of

the patient than the patient with celiac.

CHAIRMAN DURST: Soheila.

DR. MALEKI: Well, I was just going to

briefly say that essentially if it is not useful to

the consumer then -- I mean, as Mark commented, the

consumer, it wouldn't really help the consumer out

so it is not something --

CHAIRMAN DURST: Okay. Any other

comments?

(No verbal response.)

CHAIRMAN DURST: You mean we actually

might finish up early? Oh, that's right, there was

something left over from yesterday.

Steve, would you like to comment on

question number two?

DR. GENDEL: Thank you. After caucusing

with the rest of the Working Group, I guess what I

363

wanted to say in response to your question is we

recognize that the questions and issues are not

straightforward and they don't lend themselves to

simple yes/no answers.

We listened to your discussion that

occurred around that question and after the request

for clarification. The Working Group believes that

the Committee's discussion was responsive to the

question and provided useful information to us, and

we don't feel that there is any need for any other

clarification.

CHAIRMAN DURST: Good. Thank you.

In that case, I will adjourn the meeting

and remind you that we are beginning tomorrow at

8 o'clock instead of 8:30. See you at that time.

(Thereupon, at 5:20 p.m., the meeting was

adjourned, to reconvene Friday, July 15, 2005, at

8:00 a.m., this same place.)

- - -