FOOD AND DRUG ADMINISTRATION















                     Advice on CFSAN'S Draft Report:


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
















                         Wednesday, July 13, 2005


                          8:30 A.M. to 5:50 P.M.





                            Greenbelt Marriott

                              6400 Ivy Lane

                              Grand Ballroom

                        Greenbelt, Maryland 20770



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




       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.




       Petr Bocek, M.D., Ph.D.

       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.



       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


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

 Food and Drug Administration, CFSAN


       Stefano Luccioli, M.D. - Senior Medical Advisor

       Food and Drug Administration, CFSAN


 Marcia Moore, Food Advisory Committee, Executive Secretary


       Jenny Slaughter - Director

       Food and Drug Administration Integrity and Ethics Staff



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




       Rene Crevel, Ph.D. - Senior Scientist

 Unilever, Safety and Environmental Assurance Centre, United Kingdom


       Susan Hefle, Ph.D. - Associate Professor and Co-Director

 Food Allergy Research and Resource Program, University of Nebraska

       Stefano Luccioli, M.D., - Senior Medical Advisor

 CFSAN, FDA Assistant Professor, Georgetown University


       Anne Munoz-Furlong

 Director, Food Allergy & Anaphylaxis Network


       Steve Taylor, Ph.D. - Maxcy Distinguished Professor & Director

 Food Allergy Research and Resource Program, University of Nebraska


       Robert Wood, M.D. - Professor

 Johns Hopkins University School of Medicine



                             C O N T E N T S




       Call to Order, Welcome and Introductions

       Charge to the Food Advisory Committee

                 Richard Durst, Ph.D., Acting Chairman            6


       Conflict of Interest Statement & Other


                 Jenny Slaughter, FDA                            10


       Welcome and Opening Statement

                 Michael M. Landa, Esq., CSAN, FDA               14


       Use of Food allergen Thresholds

                 Catherine L. Copp, Esq., CFSAN, FDA             18


       Introduction to Food Allergens

                 Robert Wood, M.D.                               23


       Patient Perspectives on Food Allergies

                 Anne Munoz-Furlong                              72


       Allergenicity:  Analytical Methods

                 Susan Hefle, Ph.D.                              99


       Question and Answer Session                              126


       Oral Challenge Studies:  Purpose, Design,

       and Evaluation

                 Stefano Luccioli, M.D.

                 Senior Medical Advisor, CFSAN, FDA             133


       Question and Answer Session                              161


       Threshold Modeling Approach

                 Rene Crevel, Ph.D.                             170


       Food Allergen Thresholds

                 Steve Taylor, Ph.D.                            189


       Overview of Approaches to Establishing

       Thresholds:  Allergens

                 Steven M. Gendel, Ph.D., FDA                   218



                       C O N T E N T S (Continued)




       Public Comments:


                 Tracy Atagi                                    228


                 John Carroll                                   232


                 Diane Castiglione                              242



                 Will Duensing                                  247


                 Martin J. Hahn, Esq.                           250


                 Peggy Mockett                                  254


                 Kim Mudd                                       257


                 Kim Mulherin                                   260


                 Linda Webb                                     264


                 Jupiter Yeung                                  268


       Committee Discussion:


                 Panel Discussion with Guest Speakers

                 Thresholds for Food Allergens - Questions      272




                 Richard Durst, Ph.D., Acting Chairman          414



                          P R O C E E D I N G S






                 CHAIRMAN DURST:  I would like to call the


       meeting to order.


                 Good morning.  I am Dick Durst, professor


       of chemistry in the Food Science and Technology


       Department at Cornell University.  I was asked to


       chair this meeting over the next two and a half


       days.  I would like to make a few announcements


       before we begin our meeting this morning.


                 I would appreciate it if everyone would


       turn off their cell phones, unless they are


       expecting a call of a super emergency nature.  I


       would also like to ask if the guest speakers could


       make themselves available for the discussion this


       afternoon, I would really appreciate it.  We may


       have some additional questions.


                 We have received a charge from the FDA to


       give our evaluation of the draft report prepared by


       the Threshold Working Group.  I assume all of the


       members have read that thoroughly.  In my opinion,



       I it was fascinating.


                 It was an excellent article and I commend


       the Committee for coming up with it.  It was very


       educational.  Not being an expert on food allergens


       myself, it was extremely educational, and I was


       able to follow it quite clearly.


                 Our charge is to evaluate this report to


       determine whether the approaches that are presented


       in there are the only ones or the better ones,


       which of the ones that are in there might be the


       most appropriate.  This is the focus of our meeting


       today, both on the food allergens and on gluten.



                 Let me also begin by asking the committee


       members to introduce themselves.  We will start


       with Dr. Silverstein.


                 Marc, would you start it off?


                 DR. SILVERSTEIN:  Good morning.  My name


       is Marc Silverstein, and I'm a general internist


       and geriatrician at Baylor Health Care System in




                 DR. TEUBER:  Good morning.  My name is


       Suzanne Teuber, I am an allergist at UC-Davis.



                 MR. ORYANG:  Good morning.  I am


       David Oryang.  I am a risk analyst and agricultural


       engineer at the United States Department of


       Agriculture, Animal and Plant Health Inspection




                 DR. KELLY:  I am Ciaran Kelly, and I am a


       gastroenterologist at the Harvard Medical School in




                 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.


                 DR. BOCEK:  Good morning.  I am


       Petr Bocek, medical officer in NIH's National


       Institute of Allergy and Infectious Diseases.


                 MRS. MOORE:  I am Marcia Moore.  I am with


       the FDA as the executive secretary of the Food


       Advisory Committee.


                 DR. WASLIEN:  I am Carol Waslien.  I am a



       nutritional epidemiologist at the University of




                 DR. McBRIDE:  I am Margaret McBride.  I am


       a child neurologist at Akron Children's Hospital.


                 DR. CALLERY:  I am Patrick Callery, a


       pharmaceutical scientist from West Virginia




                 DR. GONSALVES:  I am Dennis Gonsalves, a


       scientist with USDA in Hawaii.


                 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, vice president for special


       projects and regulatory affairs.


                 DR. NELSON:  Mark Nelson with the Grocery


       Manufacturers Association responsible for


       regulatory and scientific policy.


                 MS. HALLORAN:  Jean Halloran from the


       Consumers Union where I am director of food policy




                 CHAIRMAN DURST:  Thank you very much.



                 One other item is that we may have some of


       our members leave early on Friday, depending on the


       amount of time we can spend.  What I propose is


       that today and tomorrow that we anticipate having


       to go perhaps till 6 o'clock so that we can be sure


       that we have enough time for all of our




                 Okay.  Let me introduce our first speaker.


       This will be Jenny Slaughter, director of Ethics


       and Integrity Staff at the FDA, to describe the


       "Conflict of Interest Statement" and other






                          AND OTHER INSTRUCTIONS


                 MS. SLAUGHTER:  Well, good morning and


       welcome.  The Food and Drug Administration is


       convening today's meeting of the Food Advisory


       Committee under the authority of the Federal


       Advisory Committee Act of 1972.


                 With the exception of the industry


       representatives, all members of the Committee are


       special government employees or regular Federal



       employees from other agencies subject to Federal


       conflict of interest laws and regulations.


                 FDA has determined that members of this


       Advisory Committee are in compliance with Federal


       ethics and conflict of interest laws including, but


       not limited to, 18 U.S.C. 208 and 21 U.S.C. 355 and




                 Under 18 U.S.C., Section 208, applicable


       to all government agencies, and 21 U.S.C. 355,


       applicable to only FDA, Congress has authorized FDA


       to grant waivers to special government employees


       who have financial conflicts when it is determined


       that the Agency's need for particular


       interventional services outweighs the potential


       conflict of interest.


                 Members who are special government


       employees at today's meeting including special


       government employees appointed as temporary voting


       members, have been screened for potential financial


       conflicts of interest of their own as well as those


       of their spouse, minor child, and employer, which


       are related to the discussions of today's and



       tomorrow's and Friday's meeting regarding the "FDA


       Draft Report: Approaches to Establish Thresholds


       for Major Food Allergens and for Gluten in Foods."


                 These interests may include investments,


       consulting, expert witness testimony, contracts,


       grants, research and development agreements, public


       speaking, writing, patents, royalties, and primary




                 In accordance with 18 U.S.C. 208(b)(3),


       full waivers have been granted to the following


       participants, Dr. Suzanne Teuber and Dr. Soheila


       Maleki, please note that all of the interests in


       the firms that could potentially be affected by the


       Committee's decisions.


                 A copy of the written waiver statements


       may be obtained by submitting a written request to


       the Agency's Freedom of Information Office, Room


       12A-30 of the Parklawn Building.


                 In addition, the following individuals are


       participating as FDA's invited guest speakers,


       July 13th: Dr. Rene Crevel, Dr. Susan Hefle,


       Anne Mun[MLM1] oz-Furlong, Dr. Steve Taylor, and



       Dr. Robert Wood.


                 The following individuals will be


       participating as FDA invited guest speakers


       tomorrow, July 14th: Dr. Pekka Collin,


       Dr. Alessio Fasano, Dr. Donald Kasarda,


       Dr. Cynthia Kupper, and Dr. Joseph Murray.


                 Lastly, I would like to report that


       Dr. Jeffrey Barach and Dr. Mark Nelson are serving


       as the industry representatives on the Committee at


       today's meeting.  They are acting on behalf of all


       regulated industry.


                 Dr. Jeffrey Barach is employed by the


       National Food Processors Association and


       Dr. Mark Nelson is employed by the Grocery


       Manufacturers of America.


                 A copy of this document will be placed on


       the back table, if anybody wishes to take a look at


       it.  I thank you.


                 CHAIRMAN DURST:  Thank you very much.


       We will now go on to the welcome and opening


       statement by Dr. Michael Landa, the deputy director


       for Regulatory Affairs at CFSAN, the FDA.







                 MR. LANDA:  Thank you, Dr. Durst.  You


       will be pleased to learn that I don't have a


       doctorate or an M.D.  I'm just a plain, old J.D.


                 (General laughter.)


                 MR. LANDA:  Thanks again.  Good morning to


       everyone.  Welcome to the members of the committee,


       to the guest speakers, to members of the public who


       have joined us today, and to my fellow FDA




                 I would like to give a special thanks to


       the Committee members for your willingness to take


       time from busy schedules to help us with your


       expertise for a meeting that will be several days


       long.  We are all here today, tomorrow and a fair


       chunk of Friday.


                 Let me just add that Dr. Brackett had


       hoped to be here this morning, but he wasn't able


       to make it.  I am hopeful that he will be here for


       some portion of the meeting.  He was called


       downtown for a meeting this morning.



                 I am going to refer to a couple of points


       on the food allergens, but the points I'm making


       apply to celiac disease as well.  It is just less


       cumbersome to start with the food allergens.  The


       agenda has been making, I think, an opening


       statement, of course I'm really not going to do




                 There are just a few points I want to make


       as you go into your inquiry today.  The first is


       virtually every FDA speaker makes at this kind of


       proceeding which is what we do really is based on




                 We talk about being a science-based


       agency.  It is the bedrock; it is the foundation.



       In that context, I am going to paraphrase what may


       be a rather obscure 19th century Senator, Karl


       Shrews from Pennsylvania.


                 The paraphrase essentially is, Our science


       correct or incorrect, when it is correct, help us


       keep it correct; when it is incorrect, help us to


       correct it.  That is as much as anything else what


       we want from you here in terms of your expertise in



       the science.


                 If with respect to the threshold in the


       Draft Report, we have gotten it right, we want to


       know from you that we have gotten it right.  We


       want your help in keeping it right.  If we have


       gotten it wrong, we want your help in getting it


       right.  That includes, as you will hear, if we have


       not considered an approach that we should have


       considered, we want to know that from  you.


                 The third point I will make is that


       Americans suffer from food allergies, particularly


       children.  There is some evidence that the number


       is increasing.  If you add to that family members,


       you really have tens of millions of folks who are


       involved.  At the moment their principle means of


       protection really is exquisite attention to the


       food label.  That is their pathway to safety I




                 We are hoping that eventually thresholds


       will provide another path to safety.  This is the


       beginning of the inquiry into thresholds, that is,


       the approaches that are outlined in the report.  It



       is the first step in a very important process.


                 The last point I will make is just that


       this is as much as anything else for members of the


       public, the docket is going to remain open until


       about the middle of August.


                 If people have comments, based on what


       they have heard today, for example, they should


       feel free to submit those comments to the docket.


       Again, it is until about, I don't remember the


       precise date, but it is the middle of August.


                 In that connection, I should say we are


       especially interested, as I think is always the


       case, in data.  In this case, data of the type


       outlined in the report.


                 Thank you.


                 CHAIRMAN DURST:  Thank you, Mike.  Since


       Mr. Landa didn't want me conferring a doctorate


       degree on him, I will not do it with Catherine


       Copp, who is the policy advisor at CFSAN, also the


       FDA, who will discuss the use of food allergens







                 MS. COPP:  I was hoping.  Oh, well.


                 (General laughter.)


                 MS. COPP:  Thank you, Dr. Durst.


                 Good morning.  As you know, the focus of


       this meeting today and tomorrow and the discussion


       on Friday is the Draft Report of CFSAN's Threshold


       Working Group:  Approaches to Establish Thresholds


       for Major Food Allergens and For Gluten in Food.


                 I have been asked to provide a context for


       the Draft Report in terms of CFSAN's programmatic


       efforts.  This is one thing that if I were a real


       doctor I could do.  Lawyer's don't do this.




                 MS. COPP:  Last August, Congress enacted


       the Food Allergen Labeling and Consumer Protection


       Act, which we refer to in-house by the somewhat


       awkward acronym "FALCPA."


                 This new law amends the Federal Food, Drug


       and Cosmetic Act, the principle statute


       administered by FDA by requiring that the label of


       a food product that is or contains an ingredient


       that bears or contains a major food allergen



       declare the presence of the allergen as specified


       in the law.  In shorthand, the declaration is to be


       in "consumer friendly" terms.


                 FALCPA defines a "major food allergen" as


       one of the eight foods or food groups or a food


       ingredient that contains protein derived from one


       of these foods.  Those are listed on the bottom of


       this slide.  By "food groups," I mean fish, tree


       nuts and crustacean shellfish, which were


       identified by Congress in the law.




                 MS. COPP:  The possible existence of


       threshold levels for food allergens is an important


       scientific issue, as Mr. Landa has pointed out,


       associated with our implementation of FALCPA.


                 Although the law does not require FDA to


       establish thresholds for any food allergen, there


       are three possible ways, which are listed on this


       slide, that such thresholds could be used to


       implement the new law, these are: administering the


       petition process provided for in FALCPA,


       administering its notification process, and



       addressing the issue or the occurrence of






                 MS. COPP:  FALCPA provides two processes


       by which an ingredient may be exempt from the


       FALCPA labeling requirements, a petition process


       and a notification process.  I'm trying to read my


       own slides (laughter).  No, okay.


                 Under the petition process, an ingredient


       may be exempt, if the petitioner demonstrates that


       the ingredient does not cause an allergenic


       response that poses a risk to human health.


                 Given this language for the petition


       exemption standard, we believe it will be very



       important for us to both understand food allergen


       thresholds and to have a sound scientific framework


       for evaluating the existence of such thresholds.


                 Under the notification process, an


       ingredient may be exempt, if the notification


       contains scientific evidence that demonstrates that


       the ingredient does not contain allergenic protein,


       or, if FDA has previously determined under the food



       additive approval process that the food ingredient


       does not cause an allergenic response that poses a


       risk to human health.






                 MS. COPP:  Given this language for the


       notification exemption standard, we also believe


       that it will be very important for us to understand


       food allergen thresholds and to have a sound


       scientific framework for evaluating the existence


       of such thresholds.




                 Finally, the FALCPA directs FDA to prepare


       and submit a report to Congress.  This report will


       focus principally on the issue of cross-contact of


       foods with food allergens and is to describe the


       types, current use of, and consumer preferences


       with respect to so-called "advisory labeling."


       Processed in a facility that also processes tree


       nuts is an example of such labeling.


                 Cross-contact may occur during food


       production when residues of an allergenic food are



       present in the manufacturing environment and are


       unintentionally incorporated into a food.  Because


       the food is not intended to contain the allergen,


       it is not declared as an ingredient on the food's


       label.  In some cases, however, the potential


       presence of the food allergen is declared by a


       voluntary advisory statement.


                 We also believe that understanding food


       allergen thresholds and developing a sound


       scientific framework for evaluating the existence


       of such thresholds may also be useful to us in


       evaluating and addressing food allergen


       cross-contact and the use of advisory labeling.


                 Thank you.


                 CHAIRMAN DURST:  Thank you very much.


                 Does the Committee have any questions or


       discussion of this presentation?


                 (No verbal response.)


                 CHAIRMAN DURST:  If not, I think we will




                 The next speaker is Dr. Robert Wood,


       professor at Johns Hopkins University School of



       Medicine, who will give us an introduction to food






                 DR. WOOD:  Thank you very much.  It is a


       pleasure to be here.  What I was asked to do is to


       provide an overview of food allergens and food


       allergy leading into the discussion that is going


       to go on over these next couple of days.




                 DR. WOOD:  The beginning of this, any talk


       about food allergy really requires that we have


       some common definition that we can all agree on.


       This is something that is not as easy as it might


       sound and often generates a lot of confusion.  The


       reality is that a lot of what is called food


       allergy is really not food allergy and may fall


       under more of a food intolerance category.


                 When we are talking about food allergy,


       there are a couple of key ingredients.  One of them


       is that there is an immunologic component to the


       reaction.  The reaction is typically to the protein


       component of the food as opposed to a food



       intolerance that is more often related to the


       carbohydrate component of the food.  Importantly to


       this meeting, exquisitely small amounts may cause a


       reaction and that these reactions can be severe and


       even life threatening.




                 DR. WOOD:  The pathophysiology of the


       allergic response is sort of very schematically


       diagramed here.  What we are thinking about is a


       process that begins with exposure and with most


       allergy, probably all allergy, you have to have


       some prior exposure to develop your sensitivity.




                 DR. WOOD:  There is a genetic


       predisposition that makes some people particularly


       more prone to develop allergy in general, whether



       it be food allergy or respiratory allergy, than


       others.   There are some people who no


       matter what, how, when and where they are exposed


       they will never develop an allergy, and others who


       with very trivial exposure may develop a severe





                 If you are in this group who is


       genetically predisposed, your immune system then


       goes through a process we will refer to as


       sensitization.  Sensitization is most often


       involving the production of IgE antibodies.  We


       will talk about this in a little bit more detail


       about some different food allergy syndromes.


                 However, it is also important to note that


       not every food allergy involves IgE and that there


       may be differences in the types of reactions and


       the doses of food required to induce a reaction in


       those patients that have IgE versus


       non-IgE-mediated food allergy.


                 Once you have become sensitized, then


       reexposure to this food will lead to symptoms.


       These symptoms may be abrupt, they may occur within


       seconds of eating the food, or they may be very


       low-grade and chronic.  This is another concept


       that we will come back and talk to a little bit.


                 With some patients it will be very easy to


       determine a threshold, and in some patients it will


       be virtually impossible to determine a threshold



       because their symptoms will not appear in a


       challenge test.  They may take days or weeks of


       chronic exposure and then develop very significant


       disease based on that chronic exposure.




                 DR. WOODS:  The prevalence of food allergy


       is substantial.  The numbers that we would be most


       comfortable with would be 5 to 7 percent of young


       children; 2 to 3 percent of adolescents and adults;


       at least 10 or 11 million Americans affected.


                 We do believe that the prevalence is


       rising.  We don't believe that this is specific to


       food allergy.  There has been a substantial rise in


       asthma and other allergic diseases as well as food




                 Now, the reason that these numbers change


       between childhood and adolescence and adulthood is


       because a large proportion of food allergy is


       outgrown over the first five to seven years of






                 DR. WOOD:  There is a long list of



       potential food allergens out there.  At least 200


       foods have been identified and characterized as


       truly food allergens, but there is a relatively


       shorter list that are focused upon because they are


       responsible for the vast majority of food allergy


       that occurs.


                 The list on the left-hand side


       representing what is most common in young children:


       milk, egg, peanut, soy, wheat, and tree nuts.


       Then, the list shifts a little bit as you get into


       older children, adolescents and adults and is


       dominated by peanuts, tree nuts, fish, and





                 The reason that this list changes from


       childhood to adulthood is because four of these


       most common food allergens in your children --


       milk, egg, soy, and wheat -- are typically




                 Eighty to 90 percent of children will


       outgrow those food allergens and not carry them


       into adolescence or adulthood, whereas the peanuts,


       tree nuts, fish and shellfish are significantly



       more difficult to outgrow, less commonly outgrown,


       and tend to persist into adulthood and actually


       through the patient's entire lifespan.




                 DR. WOOD:  Now, the signs and symptoms of


       food allergy are highly varied.  They may be


       chronic and low grade as I mentioned, they may be


       acute and life threatening.  What I want to run


       through in the next couple of minutes are just some


       examples of allergic reactions that will point out


       a number of things about not only the kinds of


       reactions, but the exquisitely small amounts of


       food that induce these reactions we are going to


       show you, and the sort of day-to-day issues that


       patients with food allergy are facing.




                 DR. WOOD:  The first couple of patients I


       am going to show you have urticaria or hives.  This


       is a total body hive reaction that this boy is


       experiencing, a patient I have known since he was


       an infant.


                 He is school age at this point.  This



       reaction occurred when he was in the grade school


       cafeteria, was being teased about this food


       allergy, another child blew a straw full of milk


       across the table into his face, and he had this


       really significant reaction.




                 DR. WOOD:  This baby here was identified


       with milk allergy in the first few weeks of life.


       There are some children who don't show up with food


       allergy until they are two or three or four years


       old, while there are others who are really


       demonstrating food allergy in the first days of




                 This was a baby who was so allergic that


       he would react very acutely if his mother, who was


       breast feeding him, ingested any milk protein.  She


       was on a very strict avoidance diet after we


       identified his milk allergy, but on the occasion of


       her birthday ate a piece of cheesecake, breastfed


       him an hour and a half later, and he had this acute


       hive reaction.





                 DR. WOOD:  Now, when we are thinking about


       urticaria or hives, there are patients that may


       have chronic urticaria.  Food allergy is rarely a


       cause of chronic urticaria.


                 However, when someone shows up with an


       acute episode of hives, the chance that it is food


       allergy becomes higher.  Again, we are looking a


       relatively short list of foods that are most


       commonly implicated: peanut, nuts, eggs, milk,


       fish, and shellfish.


                 Importantly, these reactions are usually


       very quick in their onset.  Ninety percent of them


       or thereabouts will have an onset within 30


       minutes; at least half of them, within 5 minutes;


       and virtually all of them, within 2 hours.


                 When a patient has this type of reaction,


       it is often very easy to identify the culprit food


       because of the abrupt association of the ingestion


       of that food with the onset of these hives.


                 Then, in more severe episodes, there may


       be swelling or angioedema or associated


       gastrointestinal or respiratory symptoms.  That is



       moving into more of a systemic reaction that we


       would refer to as "anaphylaxis."




                 DR. WOOD:  Now, this is a patient here who


       is having an anaphylactic reaction.  When you look


       at her back here, it looks just like hives.  When


       you see her front, though, she is having swelling


       and breathing difficulty.






                 DR. WOOD:  This is a patient who was


       having a reaction in the midst of a food challenge


       -- not in the midst of it, after her first tiny


       dose of egg protein, she went into this very


       severe, anaphylactic reaction.




                 DR. WOOD:  This boy here is someone who is


       having a dramatic episode of swelling.  His


       reaction occurred.  Most patients, we should say,


       who are having severe reactions know about their


       food allergy and are making efforts to avoid it.


                 He was shellfish allergic -- he is



       shellfish allergic.  He was making efforts to avoid


       shellfish, and he had been reaction-free for


       several years.


                 Then, on another birthday occasion, he ate


       chicken in a restaurant and the chicken had been


       fried in the same oil as shrimp had been fried.


       With that cross-contact, this severe reaction.




                 DR. WOOD:  Anaphylactic reactions are


       defined as a systemic allergic reaction,


       involvement of multiple organ systems.  These have


       an abrupt onset typically.  They are related to IgE




                 You can identify these by doing a skin


       test or a blood test looking for IgE.  The


       manifestations are not always severe.  There is an


       impression that all anaphylaxis is


       life-threatening.  Some episodes are relatively


       mild, but others progress rapidly to


       life-threatening or fatal reactions.


                 We think that there are at least 150


       deaths in the United States each year due to fatal



       food-induced anaphylaxis.  That number is probably


       a substantial underestimation, but we would be very


       comfortable saying that it is well identified of


       100 to 150 deaths per year.


                 There are different types of reactions:


       some are single phase and some have two phases,


       where a patient may look better and then two or


       three or four hours later have an even more severe


       reaction than they had initially, some of those


       lead to the worst outcomes.




                 DR. WOOD:  This is a patient with one of


       the more chronic forms of food allergies, the


       patient with severe itching due to his eczema.  In


       Eczema, a food allergy is often underappreciated


       because there is not an obvious cause and effect.


                 This is one where it is more of a


       low-grade, chronic reaction.  Hence, this is much


       harder for a patient or a family member to identify


       that, yes, he ate this food and he is more itchy


       now, rather it is really more of a low-grade


       reaction where you don't see these direct



       relationships between ingestion of the food and the


       outcome being their eczema or atopic dermatitis.


                 It is also a condition where food allergy


       is underappreciated by physicians and where


       patients may be treated with a variety of different


       creams and lotions and only later on find out that


       it was really a food allergy that was driving the




                 Overall, 40 to 50 percent of patients with


       severe atopic dermatitis and 20 or 25 percent with


       less severe cases have an underlying food allergy.


                 The same list of foods: egg allergy being


       most common, followed by milk, peanuts, soy, wheat,


       and fish.  These six foods account for the vast


       majority of food sensitivities seen in atopic




                 From our standpoint, it makes it


       relatively easy to screen patients and find which


       of them are allergic by testing for a relatively


       short list of foods.




                 DR. WOOD:  Now, the last category that I



       want to mention is something that we will lump


       together as gastrointestinal food hypersensitivity.


       There are a variety of conditions that fall under


       this umbrella.


                 There are some that are in the immediate


       hypersensitivity category.  This would be part,


       say, of an anaphylactic reaction where someone ate


       food, broke out in hives, had vomiting, diarrhea,


       abdominal pain, or other gastrointestinal symptoms.


                 There is another condition called "oral


       allergy syndrome" where patients have reactions


       that are confined to their mouth or throat or lips,


       particularly related to fresh fruits and




                 There is another group of conditions that


       are lumped under a category of eosinophilic


       disorders of the GI tract.  There is a specific


       condition, eosinophilic esophagitis, where only the


       esophagus is involved.  As most people in the


       audience know, the eosinophil is a type of white


       blood cell that is most affiliated with allergic





                 If you take someone who is having a bad


       hay fever day outside today and look at their nasal


       secretions, their nasal secretions will be loaded


       with eosinophils.  If you take someone that is


       having difficult asthma, their bronchial mucosa


       will be loaded with eosinophils.


                 By the same token, if you have allergic


       eosinophilic esophagitis, the lining of your


       esophagus is loaded with eosinophils.  It may be


       isolated to the stomach, it may be more diffuse


       where we would call it "allergic eosinophilic


       gastroenteritis."  This is somebody who may have


       disease anywhere in their GI tract, and oftentimes


       very diffusely.


                 There are some other conditions,


       enterocolitis syndrome and dietary protein


       proctitis, that are much more common in very young




                 The importance of presenting these


       different syndromes here is that some of these


       syndromes are IgE mediated and some of them are not


       IgE mediated, some of them are very acute and some



       of them are very chronic.


                 It turns out that those syndromes that are


       more chronic and low-grade that don't present with


       any acute symptoms, don't present with any clear


       cause and effect of eating the food and having


       increased gastrointestinal symptoms are going to


       be, potentially, the most difficult for this


       Committee to grasp.  That is because these patients


       are often reacting to remarkably small exposures.


                 I will come back at the end to sort of


       give a couple of examples of the dilemma that kind


       of patient is going to present to us as we really


       try to figure out what is safe and what is not




                 It also turns out in the same vein that


       the non-IgE conditions in general are probably


       going to be most difficult to deal with, both


       because they often don't have the acute IgE-type


       symptoms, and because they are predominantly


       mediated by a different part of your immune system


       that can recognize even smaller degrees of these


       food proteins that identifying thresholds are going



       to be much more difficult.




                 DR. WOOD:  Now, when we are trying to


       approach a patient with a food allergy, one of the


       real difficulties is making an accurate diagnosis.


       The diagnosis, as in most everything we do, begins


       with a history, talking about the foods they


       suspect are causing problems, whether we think the


       symptoms are consistent with food allergy, whether


       this is something that may not be food allergy at


       all, or whether it may be a food intolerance rather


       than an allergy.  We are going to be interested in


       the timing of the symptoms and the reproducibility


       of reactions.


                 It turns out that when you do a very


       careful history, most of the time it is wrong.  It


       will be correct in the acute reactions, where you


       have a patient who comes in and says, "I fed him


       scrambled eggs for the first time last week, and he


       had hives all over."


                 "She took her first bite of peanut butter,


       and developed hives within 2 minutes."



                 It is very likely that the history will be


       born out when you do further testing.  However,


       when you look at the bulk of patients with food


       allergies, many of them will have these more


       chronic conditions like eczema or the


       gastrointestinal disorders.  When you are looking


       at those patients, you will only verify the history


       when you do further testing about a third of the






                 DR. WOOD:  The next set of tests we do


       after taking a history would typically either be


       skin testing or serologic testing.  A RAST test,


       "radioallergosorbent test," is the most common


       serologic test that is used.


                 These tests have some value and they also


       have some problems.  The problems they have is that


       there is a relatively high rate of false-positive


       tests.  They do not have a terribly good positive


       predictive accuracy.


                 They are generally accurate when they are


       negative.  Although, they will only be active when



       they are negative when you are convinced this


       patient has an IgE-mediated condition, because both


       of these tests rely on the presence of IgE


       antibodies to identify the specific food allergy.


                 An example would be if a patient develops


       hives or anaphylaxis, which typically are


       IgE-mediated, and they suspect that it is a certain


       food.  If you get a positive test back, it is very


       likely that they have that allergy.  If you get a


       negative test back, then you need to keep looking.


       It was not likely that food that caused that




                 However, if you have a patient with


       something like the allergic eosinophilic


       gastroenteritis where there may not always be IgE


       antibodies, you cannot stop with a negative test


       and say, "We've proven you don't have food


       allergy."  That is something that happens all the


       time, but it is often going to lead to a


       misdiagnosis and mismanagement of that patient.


                 The bottom line is that we need to


       carefully interpret our tests in the context of the



       overall clinical picture, and that we need to rely


       on oral challenge tests as the more accurate tests,


       so that we will say that they are not completely


       definitive.  They are more definitive but not


       completely definitive.


                 Again, they are going to be less


       definitive in the patients that have more delayed


       type reactions or more chronic conditions where


       they won't react in that four-hour observation


       period of your food challenge.




                 DR. WOOD:  You are going to hear more


       about food challenges this afternoon, but I will


       just mention a couple of issues here in terms of


       the way that they can be done.  They can be broken


       down as open challenges where both the patient and


       the person administering the challenge knows what


       is being given.


                 A single-blind challenge is where the


       patient is blinded but the person administering the


       challenge knows the food that is being


       administered, whereas a double-blind,



       placebo-controlled challenge is regarded as the


       most accurate test because it eliminates the bias


       that may occur on the part of both the patient, who


       may be feeling a great deal of anxiety about this


       food challenge, or on the part of the observer, who


       may have their own biases about this patient's


       allergy and might overinterpret or underinterpret




                 We would say that these are going to be


       the most accurate tests for the diagnosis of food


       allergy.  We would use them, if the history and lab


       results don't provide a clear diagnosis.  That is


       often the case, again, when we have both a history


       that may not be accurate and laboratory tests that


       may not be completely accurate.


                 Then, we also do them very commonly to


       determine when an allergy has been outgrown.  This


       would be a patient who has been known to be


       allergic to a food, and we would be monitoring them


       with some regularity in determining at some point


       that it is worth trying to retry that food.


                 We would typically do it in a controlled



       setting, just because even in some patients you


       don't expect to react at all there may be


       significant reaction.  Consequently, we have to do


       these with considerable caution.




                 DR. WOOD:  I think I pretty much mentioned






                 DR. WOOD:  Now, they asked me to mention,


       briefly, a study that we published last year


       looking at the risk of oral food challenges.  What


       we have presented in this paper were results on


       almost 600 challenges, 253 of which were failed


       challenges.  The patients reacted in the challenge,


       so that is where we can look at the risk.  The


       other 57 percent, the patients had no symptoms, so


       it was a risk-free challenge once they might have


       gotten over the anxiety of being there.


                 We collected a lot of information on


       demographics, other atopic disease, symptoms during


       challenges, treatment needed, doses at which


       reactions occurred.  Even though there is a lot



       said about safety of food challenges, there has


       been very little published before this paper on


       what really occurs.


                 Now, I'm going to say this again a couple


       of times looking at the data, but I will say it up


       front here, that these results are not


       representative of the general population of food




                 These patients that are being challenged


       in this either had an unclear diagnosis, so it


       wasn't a dramatic kind of situation, or they were


       thought to have potentially outgrown their allergy


       and were being challenged to potentially prove that


       their allergy was gone.


                 We are really looking at very low-risk


       population, and it is not representative of the


       whole population of food allergy patients that are


       out there.  Again, I will say this a couple more


       times looking at the specific data.




                 DR. WOOD:  Now, whenever we are doing this


       sort of analysis, we try to break things into



       categories.  One of the tough categories to decide


       is how do you rate reactions.  You will see in the


       literature some different definitions that have


       been used.


                 We chose to create our own for a series of


       studies that we were doing, and talked about mild


       reactions that were skin and/or oral symptoms only.


       Oral symptoms is just at itching or they will often


       have an obvious hive-like reaction in their mouth


       or pharynx when they are having one of these


       localized reactions.


                 A "moderate reaction" was described as


       upper respiratory and or GI symptoms only or any


       three systems.  When we are talking about systems,


       we broke that into: skin, GI, upper respiratory,


       lower respiratory and cardiovascular.


                 Then, severe reactions were those that


       were that were potentially life threatening, where


       they have lower respiratory and/or cardiovascular


       symptoms or any four systems were involved.




                 DR. WOOD:  When we broke things down into



       these different systems which were involved in


       which challenges, you will see here that when we


       look at this column on the right here, which is the


       total in this paper we reported on milk, egg,


       peanut, soy and wheat.


                 The greatest number of failed challenges


       was to milk, 90; 56 to egg; 71 to peanut; 21 to


       soy; 15 to wheat; for a total of 253.  You will see


       that skin manifestations were most common, 78




                 This is actually similar to what we have


       seen and what is in the literature in terms of


       reactions that happen out in the real world.


       Eighty percent of food reactions, 80 percent of


       anaphylactic reactions involve the skin, but about


       20 percent do not.


                 Oral symptoms occurred in about a quarter,


       upper respiratory in a quarter, lower respiratory


       in about a third, GI in 43 percent.  We,


       thankfully, had no cardiovascular reactions in this




                 Now, why would that be the case?  It would



       be for two reasons.  The biggest reason is that


       cardiovascular reactions are not that common in




                 The cardiovascular system of a child is


       really sturdy enough to put up with the insult of


       an allergic reaction without necessarily becoming


       involved.  Cardiovascular reactions are much more


       common in adults, and this population was entirely




                 The other reason that we might have seen


       the absence of cardiovascular reactions would be


       that we were dealing with a relatively low-risk




                 When we break it down into those three


       severity classifications -- mild, moderate and


       severe -- you will see that the numbers are


       relatively similar for each food.  When we look at


       the total category, they broke pretty close to a


       third in mild, a third in moderate, and a third in




                 When you look across the specific foods,


       the most important point that came out of this is



       that you can't say that one type of food allergy in


       this kind of setting is more dangerous than




                 It turned out that the greatest number of


       severe reactions occurred with egg challenges.


       This was important information we thought to get


       out to get out to people doing challenges.


                 A lot of allergists will say, "I'm going


       refer you, Dr. Wood, all of my peanut challenges.


       I'm not touching a peanut challenge because they


       are really dangerous.  However, I will do egg and


       milk challenges out in my office any time."


                 The message there is that really all of


       these foods have a potential to have severe


       reactions and need to be done in a setting where


       you are really equipped to deal with that potential


       for a severe reaction.




                 DR. WOOD:  When we looked at the RAST test


       score or the median IgE level for these different


       challenge results, we found that there was really


       no strong association between their IgE level and



       the reaction severity.


                 Now, this is an example of where this


       population is not a good one to look at for this


       data.  The reason is that we were essentially only


       challenging people that had relatively or very low




                 We were not challenging people with very


       high levels where they were extremely likely to


       fail the challenge.  There is no reason in most


       instances to prove that they are allergic.  When


       you know with, say, 99 percent certainty that they


       are allergic, we would not put that patient through


       a challenge.


                 Consequently, if you went out in the real


       world where the RAST test levels range anywhere


       from zero to 100, you would typically see


       escalating reaction severity with levels that are


       higher.  We have that data for peanut allergy where


       the group of patients that had levels at 100 did


       have more severe reactions when they had accidental







                 DR. WOOD:  Then, I think the last thing to


       present from this study is whether reaction


       severity was correlated or related to the percent


       of food ingested in these challenges.  It turns


       out, if anything, it is inversely correlated.  The


       more severe reactions, and none of these were


       statistically significantly, but if you look at the


       general trends, you will see here that the more


       severe reactions occurred with milk and eggs.


                 As you can see, the severe reaction for


       milk is 15 percent and 30 percent for eggs.  When


       you look at the total group here, 50 percent, 45


       percent and 30 percent.




                 DR. WOOD:  What is the reason this


       happens?  Does this make any sense at all?  Do you


       have your more severe reactions with smaller


       exposures?  The reason we think it happens is


       because it is just identifying the more reactive




                 It is picking out those that even though


       our test scores said that they are not so allergic



       that they should do this, it is picking out those


       that react more abruptly and have more severe


       symptoms early in the challenge just because they


       were higher risk patients.


                 Now, we have come up in our studies about


       some decision making about when we would do food


       challenges.  This is purely for clinical purposes.


       These are for those reasons of when we are trying


       to decide if they are truly allergic or when we


       think that the food allergy might have been




                 What we would say is that we would do food


       challenges based on their history of reactions.  If


       they have reacted recently, we wouldn't feel the


       need to do a food challenge.


                 We would base it on their laboratory


       testing, the skin testing and the RAST testing.


       Then he would base it on the importance of the food


       to the diet.  There are some foods that are


       obviously much more important to the diet.


                 A family may never care whether that child


       ever eats a pea again the rest of their life.  They



       may elect to never have a pea challenge done, but


       they may be jumping to do a milk or what challenge


       at the first opportunity, because milk or wheat


       back in the diet would make such a dramatic


       difference in their day-to-day life.


                 Then, we have come up with some


       recommendations based on RAST testing of when we


       would recommend doing challenges.  These cutoffs


       for milk, egg and peanut are all where we found a


       greater than 50 percent chance of passing the


       challenge, if you have levels below that range.


       For other foods, it has been harder to determine


       cutoffs, and we would challenge at higher levels


       for things like wheat and soy.




                 DR. WOOD:  Just to go through an algorithm


       of how we approach diagnosis, then, because it does


       impact on the discussions that are going to happen


       here, we would first take our history.


                 Based on the history, we would make some


       distinction whether we think this is consistent


       with an IgE type reaction or whether we think that



       it is consistent with a non-IgE type reaction.


                 If it is IgE-mediated in all likelihood,


       then a skin test or a RAST test will help identify


       whether that food that was suspected to cause a


       reaction probably did or probably didn't.


                 If the test is negative, because the


       negative predictive accuracy is so high, we would


       feel that you could stop worrying about that food


       at that time.  If the skin test is positive,


       because there are false-positive tests that occur,


       we need to do something more.


                 We might do a trial on an elimination


       diet; we might do a food challenge in one order or


       the other; and based on all of that information, we


       would arrive on the specific elimination diet


       recommended for that patient.


                 If it falls into a non-IgE category, the


       situation is much more difficult because we can't


       rely on a simple screening test to weed out those




                 They are going to need some combination of


       challenges -- endoscopy, if it is a



       gastrointestinal symptom; elimination diets,


       rechallenges, maybe a reendoscopy -- so there is a


       much more difficult plan on this side of the screen


       to sort out those patients.




                 DR. WOOD:  Now, I'm going to finish here


       with a couple of conclusions and present a couple


       of dilemmas.  The conclusions are that food allergy


       is very common.  This is a remarkably worthwhile


       initiative that is going on here, and that right


       now avoidance is the only treatment plan.


                 We really hope in the next 5 or 10 years


       that there are going to be other treatments for


       food allergy.  It may be enough so that even if


       they don't cure the disease, that they will elevate


       the threshold to a point that we don't even need to


       have these meetings, that small exposures won't


       even be relevant.  We are not even close to their


       yet, so avoidance is the only option.


                 Strict avoidance is essential to prevent


       reactions obviously, but we also think that in many


       patients it also helps to promote the outgrowing





                 Here is where we may have very different


       thresholds.  We may have a threshold that this


       child, say, with milk allergy -- they know for a


       fact that they can eat this bread that has whey as


       the tenth ingredient and never have a symptom.


       They are perfectly fine with it.


                 What we have found that getting that bread


       on a regular basis may keep their immune system


       more revved up to maintain the allergy so this


       thing that is way below their threshold for


       reacting acutely may still drive the immune system


       to maintain the allergy and prevent them from


       outgrowing the allergy.


                 The next conclusion is that food


       challenges are a useful means to diagnose food


       allergy and a useful means to determine threshold


       doses.  There are going to be some limitations of


       challenges, and one of them is that as opposed to


       the study that I presented that Dr. Perry did with


       me, you have to include in a threshold type study


       the most allergic patients.



                 Doing the kind of patients that we are


       studying on the lower end of the spectrum has


       nothing to do with thresholds.  It is irrelevant


       data.  You can't go to my study and say, "This


       looks like a threshold because we are not including


       in those kinds of studies those highly allergic




                 The greater dilemma, and this one is


       solvable, there are plenty of real allergic


       patients out there.  They won't necessarily want to


       undergo these studies, because it is not a pleasant


       thing to have allergic reactions, but that part is


       potentially solvable.


                 The more difficult thing is a


       determination of the threshold doses that I


       mentioned for the chronic allergic conditions,


       especially those that are not IgE mediated probably


       isn't possible.


                 To give a couple of examples, if we take,


       say, milk allergy, the most common food allergy of


       all, and we are talking about an infant who is on a


       formula, there are a bunch of different options we



       could have.  Some of them can have soy, but some of


       them are also allergic to soy.


                 Some would go on to a formula like


       Alimentum or Nutramigen, which is a formula where


       the milk protein has hydrolyzed to a small enough


       fragment that in 98 or 99 percent of kids with milk


       allergy.  It completely solves the problem.  They


       don't react at all to that level or that type of


       protein that remains in that formula.


                 That other 2 percent, though, may react


       severely to that.  They are typically the patients


       with the gastrointestinal disease.  They are


       typically very sick; they are typically not


       growing; they are typically malnourished.


                 They are a group of patients who aren't at


       risk for the acute dangerous reactions, but they


       may be at very high risk for chronic disease from


       their food allergy.


                 Those patients will typically respond


       dramatically to a formula that is based in a single


       amino acids as a protein source, and that is a


       formula like Neocate and Elecare.



                 Now, when you take that population, and


       this is what I deal with every day, there is going


       to be a group of them -- and that is probably even


       less than 1 or 2 percent, it is probably only 1 out


       of 500 -- who still react to the Neocate.  They can


       react severely to it.


                 We know that because of their


       gastrointestinal biopsies, their biopsies that are


       taken from their esophagus or stomach or intestinal


       tract still show evidence of severe allergy.


                 What we think those patients are reacting


       to would be either the absolutely trivial amounts


       of, say, soy protein that is in the soy lecithin,


       that is the eighteenth ingredient in Neocate, or


       the trivial, trivial amounts of protein that may be


       left in the safflower oil that is used as a fat


       component of Neocate.


                 When we switched those patients off of


       Neocate we can prove, and we have 15 patients now


       who we have proven, that taking them off Neocate


       resolved their food allergy.  In this supposedly


       non-allergenic formula, they were still reacting.



                 Now, whether the direction this Committee


       needs to focus on is this very unusual patient or


       not is sort of a separate debate all together, but


       it is safe to say that there are going to be


       patients out there who break all rules.  No matter


       what rules are established, there will be patients


       who completely break them and make all of our lives


       difficult from that standpoint.


                 I would be delighted to take any questions


       from the Committee or otherwise.  Thank you for


       your attention.


                 CHAIRMAN DURST:  Thank you, Dr. Wood.


                 Are there questions for discussion?






                 DR. TEUBER:  This is Suzanne Teuber.  I


       had a question about your patients with the Neocate


       sensitivity in terms of what the company reported


       for the soy lecithin, did they have any values that


       you could report back as to a chronic ingestion




                 DR. WOOD:  No.  I mean, most of these kids



       it is most likely the soy lecithin.  SHS doesn't


       have that data on the protein content of their soy


       lecithin.  They say it is zero.  These kids when


       they were switched to Neocate One Plus, which has


       no soy lecithin, their disease went away.  We have


       to assume that there was enough there to drive that




                 CHAIRMAN DURST:  Yes.


                 MS. HALLORAN:  Jean Halloran.  Could you


       say something about the process about growing


       allergies?  How does that work?  What actually




                 DR. WOOD:  Well, that is a very good


       question.  There are a number of things that we


       don't understand too well.  However, what we think


       is that in the majority of patients we think that


       outgrowing is most related to the immune system


       gradually forgetting about that concern that it


       earlier had.


                 That is where we think that strict


       avoidance is likely to promote the outgrowing


       process, and with a prolonged period of strict



       avoidance for many of these foods, the immune


       system has a memory that isn't long enough to


       maintain the allergy and that it will gradually


       wane and then full tolerance will be accomplished.


       There are probably lots of other mechanisms going


       on immunologically that are not well understood.


                 The other question with this that we have


       no great explanations for, lots of theories but no


       great explanations, is why you can take a food


       allergy like milk, which in early infancy can be


       every bit as severe as a peanut allergy, and have


       most kids outgrow that allergy, while very few kids


       outgrow the peanut allergies.  There is something


       very different about the immunologic memory of one


       food allergen versus another.


                 CHAIRMAN DURST:  Yes.


                 DR. KELLY:  Ciaran Kelly.  I wanted to


       come back to the issue of challenging individuals


       with severe allergies as a method for determining a


       threshold.  I would like to hear your comments as


       regards the feasibility and safety and whether that


       would be ethical to perform?  I guess my concern is



       that once the threshold is crossed, whatever that


       threshold might be, isn't there a potential for


       severe allergic reaction?


                 DR. KELLY:  Yes.  Absolutely.  There have


       been threshold studies done for the biggie, peanut,


       with very allergic people so it is doable.  Now,


       what we can say about this is that these studies


       won't be done in children.  It is not going to




                 That automatically limits your population


       of people, because when you go out and try to find


       your group of milk-allergic adults to do these


       studies on, you are limited.


                 Now, they do tend to be more severe


       reactors.  From that standpoint, you have some


       patients out there, but there is no IRB that is


       going to let us do this in children.  There has to


       be demonstrated benefit to do a study with risk.


                 The safety element is one that we are


       comfortable with, recognizing that you need to have


       emergency management available to you because there


       will be people that have bad reactions.



                 The safety that is built into that is


       starting with exquisitely small doses and working


       up very gradually and aborting the challenge


       whenever you see your first symptom.


                 That may lead you to end some challenges


       prematurely.  You may end up with a false


       threshold, but you are obligated to stop when you


       have objective signs that patient is reacting.


                 The ethics beyond that to me is that if it


       is an adult patient who is willing to consent to


       that process, I have no problem with the ethics of


       doing it and have no fear that I will ever lose a


       patient to a food challenge.


                 CHAIRMAN DURST:  Yes.


                 DR. BRITTAIN:  This is Erica Brittain.


       Since you can't study children in that way, do you


       know how this threshold might be different in


       children, if you've got the threshold for adults?


                 CHAIRMAN DURST:  No, we don't know that.


       That data is, to my knowledge, not available in a


       large enough sample to have any validity


       whatsoever.  It is a superb question.  The argument



       is going to be and will always be these children


       are much more reactive than the adults for most of


       these foods.


                 For peanut allergy it is going to be the


       simplest, because allergy tends to persist.  We


       think that people usually hit their peak level of


       severity as an adolescent or young adult, so that


       would be fairly easy to solve.


                 However, when you look at the others like


       milk and egg and soy and wheat, you are by and


       large going to have the highest level of reactivity


       in your first couple of years of life.


                 When we think about those allergies, we


       usually think of growing into the allergy for one


       or two or three years where they are becoming more


       and more allergic, and then they are becoming less



       and less allergic over the next one or two or three


       or four or five years as they outgrow the allergy.


       It is a moving target at all points, but the most


       severe reactivity is likely to be early on.


                 CHAIRMAN DURST:  Dr. Wood, I have a


       question -- this is Dick Durst -- just points of



       clarification.  On your slides where you indicated


       "wheat," now this is the IgE-mediated type allergy


       as opposed to our discussion tomorrow on celiac




                 DR. WOOD:  Yes, these results are entirely




                 CHAIRMAN DURST:  Okay.  Do other grains


       cause the IgE type reaction as the wheat?


                 DR. WOOD:  Yes, our study there, about 600


       challenges, came out of about 3,000 food challenges


       that we have done.  There were five most common


       foods that I had enough data to make some


       conclusions that we were comfortable with.  All of


       the grains cause allergic reactions.


                 It turns out that wheat and rye are very


       cross reactive from an IgE-mediated allergy


       standpoint, and that most patients allergic to


       wheat are also allergic to rye; it turns out that


       about half are allergic to barley; and 10 to 20


       percent are allergic to oat.  Beyond those grains,


       all of the other grains and grain substitutes are


       clearly capable of causing allergy in select





                 CHAIRMAN DURST:  Thank you.  One other


       question as far as clarification at least for my


       mind.  One of your slides with the food challenge


       decision making had the units in caps "KU/L."  I


       don't know if you defined that?  I was curious.


                 DR. WOOD:  Yes.  It stands for "kilo unit"


       of IgE in a specific assay that Pharmacia has


       developed called an immunoCAP RAST.  It all goes


       back to this one technology that is thought to be


       the most accurate quantitative measure of specific


       IgE, and the results are represented in that kilo


       unit of IgE, the specific IgE antibody per liter of




                 CHAIRMAN DURST:  Thank you.


                 There is another question?


                 DR. KELLY:  I have one other question.


       Dr. Wood, you made a very important comment about


       the potential for continued subclinical exposure to


       allergens perpetuating an allergic response.  How


       well accepted and how well documented is that, or


       is that largely a clinical impression?



                 DR. WOOD:  Very well accepted, very poorly


       documented.  It is widely accepted.  There is very


       poor information to support it.  There are only a


       couple of studies.  The problem we have is we tried


       to do the study, and we were turned down because it


       is so widely accepted that to go to the IRB and


       propose to them that we are going to take this


       group of kids with milk allergy and keep them on


       low-dose milk and take this group and have them


       strictly avoid it was turned down.


                 Now, there is some work being done that


       has identified instead of looking at the IgE


       against milk globally, it has turned out that if


       you have IgE against certain portions of the milk


       molecule it may be more predictive of a longer-term


       allergy, and if you have it toward others, other


       epitopes, it may be more predictive of an allergy


       that is easier to lose.


                 We think that it may be feasible to focus


       on that population that has a very good chance of


       losing their allergy, even if we make a mistake, to


       be able to do this study.  It is doable, but the



       outcome is about 10 years down.


                 CHAIRMAN DURST:  Marc.


                 DR. SILVERSTEIN:  I have had some


       experience --


                 CHAIRMAN DURST:  Identify yourself.


                 DR. SILVERSTEIN:  Marc Silverstein, Baylor


       Health Care System in Dallas.  I have had some


       experience in studying the epidemiology of asthma


       and anaphylaxis.  In both of those conditions, your


       findings are very much dependent upon your


       diagnostic criteria.


                 In clinical medicine, we have diagnostic


       criteria.  You have described the criteria for food


       allergy, which would involve components of:


       history, physical exam, laboratory tests, food


       challenge, and response to clinical management with


       elimination diets.


                 Are there standardized criteria that you


       would see moving the diagnostic criteria that you


       would use from clinical practice to investigation


       and publication in peer review literature and/or


       perhaps the policy in making regulatory decisions?



                 I am interested in, Is there a set of


       standardized criteria that professional


       organizations or clinicians would use for


       investigation or for recommending policy?  I


       understand there is some recent work on definitions


       and standards for anaphylaxis?


                 DR. WOOD:  The definitions for


       IgE-mediated food allergy are pretty clear and it


       is pretty well accepted that it is if you have a


       history that is consistent, you have a positive


       allergy test, and you either fail a challenge test


       or pass a challenge with a dose that is generally


       accepted to indicate full tolerance.  It is fairly


       straightforward and well accepted in the peer


       review literature.


                 It is much more difficult on the group of


       patients with, say, eosinophilic gastroenteritis


       where they don't necessarily have IgE.  You require


       a histologic diagnosis to identify the condition,


       and then figuring out whether they have food


       allergy driving the process exclusively, partially


       or not at all is a much more difficult process.



                 It is doable, but you have to eliminate



       foods, rebiopsy, reintroduce foods, and rebiopsy.


       There are studies that have done that, but it is so


       much more difficult to do that there is much less


       of an acceptance of an absolute diagnostic


       criteria, much, much less.


                 It is being looked at.  This is a form of


       allergy that is clearly either happening much more


       often or being identified much more often or both,


       so that the potential is there, but it is much


       further away from a definition that is well agreed




                 CHAIRMAN DURST:  Yes.


                 DR. BRITTAIN:  This is Erica Brittain.  I


       have a clarification question on the food


       challenge.  How is the placebo control implemented?


                 DR. WOOD:  I think you are going to hear a


       lot more about food challenges this afternoon, but


       the idea, and it is going to vary depending on the


       age of the patient and what they can do, but the


       idea that it needs to be well disguised and


       obviously safe from the perspective of that



       patient's allergen --


                 (Simultaneous discussion.)


                 DR. BRITTAIN:  But --


                 DR. WOOD:  Go ahead.


                 DR. BRITTAIN:  I'm sorry.  Is it by a


       dose?  Is a particular dose placebo, or does a


       patient get all placebo?


                 DR. WOOD:  Yes.  I'm sorry I


       misunderstood.  The normal way the challenge is


       done is to have a separate challenge for the


       placebo and for the actual food being studied.  The


       usual way it is done is that the patient would come


       in and have a day doing a placebo challenge and


       come in and have a day doing the food challenge.


                 Challenges can be done in a matter of a


       couple of hours in some situations, but to do


       highly allergic people in a placebo-controlled


       manner would usually take 8 or 10 hours for each




                 CHAIRMAN DURST:  All right.  Seeing no


       further hands in the air, I think we will thank


       Dr. Wood.  We are right on schedule.  Thanks again.



                 Our next speaker will be


       Anne Munoz-Furlong, who is director of the


       Food Allergy and Anaphylaxis Network, who will


       discuss patient perspectives on food allergies.




                 MS. MUNOZ-FURLONG:  Thank you.  I would


       like to thank the organizers of the meeting for the


       opportunity to be here.




                 MS. MUNOZ-FURLONG:  What I would like to


       do is in that time that I have been allotted is


       give you a sense of who this food allergic consumer


       is; the food allergen labeling from their


       perspective; and then, most importantly, their way


       of looking at threshold levels for food allergens.




                 MS. MUNOZ-FURLONG:  By way of background,


       the Food Allergy & Anaphylaxis Network or "FAAN" is


       a non-profit organization.  We were established in


       1991 and have 27,000 members, almost 28,000


       members.  Eighty percent of these people come to us


       from physician referrals, so we know we are talking



       about IgE-mediated responses when we are looking at


       our membership.


                 Our mission has four points: to increase


       public awareness, provide advocacy and education,


       and advance research on behalf of those with food






                 MS. MUNOZ-FURLONG:  Now, as Dr. Wood said,


       food allergy is believed to affect about 11 million


       Americans or 4 percent of the population; fish and


       shellfish allergy, 2.3 percent or 6.5 million;


       individuals in peanut and tree nut, 3 million.


                 Consequently, between these four foods we


       are talking about almost 10 million Americans.


       These are the four foods, as was presented earlier,


       that are lifetime allergies and also are believed


       to cause the majority of the severe or fatal


       reactions in this country.


                 The other point I want to make here is


       that although we are talking about 11 million


       patients, our data shows us over and over again


       that most of these patients have families who



       follow their restricted diet.  The impact is


       actually many times greater than the number of






                 MS. MUNOZ-FURLONG:  When we look at


       shellfish allergy, this is looking at data that we


       published about a year ago now.  Te prevalence of


       shellfish, we found about 2 percent of the


       population or 6 million Americans.


                 The key foods responsible for the majority


       of these reactions in rank order are: shrimp, crab,


       lobster, and clam.  For fish allergy, .4 percent of


       the population: salmon, tuna, catfish, and cod


       being the primary fish that cause reactions.


                 However, if you look at these a different


       way, these foods, especially shrimp or salmon, are


       available on almost every menu that you are going


       to look at in a restaurant or food service


       establishment.  Therefore, the risk for these


       individuals is constant.




                 MS. MUNOZ-FURLONG:  Talking about tree



       nuts, and these most of you already know, are not


       peanuts; they are different.  Most people with a


       peanut allergy avoid tree nuts as a precaution but


       not because they are allergic to them.  About


       20 percent of the 20 peanut allergic population is


       allergic to tree nuts as well.


                 When we are talking about tree nuts, it


       affects about 1.5 million Americans.  Again,


       looking at data from our patient registry of 5,000


       patients, we find that walnut, cashew, almond and


       pecan are the leading cause of tree-nut-allergic


       reactions in this country.




                 MS. MUNOZ-FURLONG:  What does it mean to


       have food allergies?  It is vigilant label reading.


       You have got to read labels not just for food


       ingredients but anything coming into the home.


       Bath products can have tree nuts, milk or eggs in


       them, for example.


                 Pet food, if you have ever looked at the


       ingredient statement on a pet food, it can have


       almost every single one of the major eight





                 That is something you have to worry about,


       especially if you have a toddler who will pick up


       food from the floor or anyplace else they can get


       it.  Also, medications have been known to have


       allergens in them, particularly milk.


                 It is not just a question of label reading


       for food; it is for anything.  Trace amounts can


       cause an allergic reaction, and that has been



       proven over and over again.


                 Just one bite can cause a reaction.


       Therefore, we can't tell by looking at someone how


       allergic they are going to be or what their


       tolerance will be to that food.


                 Currently, as Dr. Woods said, the only


       cure now is a dose of epinephrine, if the patient


       has a history of severe reaction.  The onus is on


       the patient or the family to read the label and


       avoid the allergen and then be quickly prepared to


       handle an allergic reaction, if they have made a


       mistake or accidentally ingested the food to which


       they are allergic.





                 MS. MUNOZ-FURLONG:  Because there is no


       cure, decisions about any part of the person's life


       are centered around food allergy.  This is what


       makes food allergy so stressful on the family and


       on the patients.


                 Whereas with other allergies you have


       seasonal components and you might have an easy


       spring but fall is the bad season or if you are


       allergic to cats or dogs you can avoid those, with


       a food allergy every decision every single day is


       affected by your food allergy.


                 Food shopping can take two to three to


       hours just from reading labels.  Cooking, if the


       family is bringing the allergen into the home, they


       then have to prepare two meals, the


       non-allergen-containing meal and then the


       allergen-containing meal, and take precautions to


       avoid cross-contact.


                 Decisions about dining out and socializing


       are made based on not a food preference, but is the


       food safe.



                 "Can the manager be trusted to give us


       accurate information?"


                 "Can the person we are visiting be trusted


       not to slip some of the allergen into the food?"


                 Then, the decision is made to move forward


       based on the answers to those questions.


                 Even what school or childcare the


       individual will be sending their food allergic


       child to are going to first be centered on food


       safety from a food allergy perspective.


                 Vacation and travel where you and I might


       decide whether we want to go someplace warm or go


       skiing in the winter, these families have to think


       first about food.


                 "Can we ship food there?"


                 "Is there a safe place?"


                 "Can we rent a room with a kitchenette and


       make some of the meals so that we can maintain some


       level of safety?"


                 Even family relationships, there is always


       somebody in the family that does not believe the


       food allergy is real, and so decisions are made



       about whether they can visit that individual or






                 MS. MUNOZ-FURLONG:  As a result of all of


       this, it has a tremendous impact on quality of


       life.  We published a study several years ago


       looking at the impact of food allergy on quality of




                 What we found is that families who have a


       food-allergic child score lower on their perception


       of whether their child has good health or not, the


       emotional health and family activities than the


       general population.


                 Certainly, they scored lower or worse than


       families who are looking at or dealing with other


       chronic diseases such as diabetes, juvenile,


       rheumatoid arthritis and attention deficit



       disorder, for example.


                 We also looked at some of the other


       influences.  If the individual has a food allergy


       and asthma or atopic dermatitis, that further


       lowers their score for the quality of life.



                 If a family has a child with two or more


       food allergies, that group scored much lower in 9


       out of 12 scales compared to those who only have


       one or two food allergies that they are dealing




                 When we look at our patient population at


       FAAN, we see that it is not uncommon for our


       members to report a child with a milk, egg and


       peanut allergy simultaneously.  You can imagine


       eliminating those three foods and how it compares


       to the impact on the quality of life for the entire






                 MS. MUNOZ-FURLONG:  This is how, again


       looking at the same data, you can see here in blue


       is "General health" perception.  Food allergy lower


       than the normal for asthma, attention deficit


       disorder and some of these other symptom scores.


                 Now, in talking about label reading, which


       is really the cornerstone of managing a food


       allergy.  Here is what goes on.





                 MS. MUNOZ-FURLONG:  The person with a food


       allergy is told by the physician, as you heard


       earlier from Dr. Wood, "You have an allergy, avoid


       the food."  Zero tolerance.  They must live in a


       black-and-white world.  If you are allergic, you


       don't eat that product.


                 If the allergen is listed on the label or


       the label says "Contains allergen," they are not


       going to eat that product because they are trying


       to avoid a reaction.  As a result, they expect


       ingredient labels to be consistent and, most of


       all, reliable because this is what they are basing


       the decision about food on.  It will affect their


       health and safety.


                 When they see the same product with


       different ingredient statements, it makes them very


       confused and frustrated and sometimes very nervous


       because they, again, are looking for consistency in





                 What we are already seeing with some of


       the companies complying with FALCPA regulations is


       that there are products on the market that are



       pre-FALCPA and FALCPA compliant with different


       ingredient information regarding allergens.


       Already we are getting calls from our members.


                 "Which one of these labels is correct?"


                 "What if I hadn't picked up that second


       label?  How would I have known?"


                 This is what we are heading into as we


       start to change these labels.




                 MS. MUNOZ-FURLONG:  The challenge for


       food-allergic individuals is that the patients are


       told to strictly avoid the allergen, there is zero


       tolerance or be prepared to handle an allergic


       reaction.  Once a reaction begins, we don't know


       how severe that is going to be.


                 They are not aware that there are


       scientific names to foods when they are newly


       diagnosed.  This is something FAAN spends a lot of


       time doing.  It will get better as FALCPA is


       implemented because labels will have simple


       ingredient terms on them.


                 We have to remember it is not just the



       patient or the patient's family reading the label,


       but it is the teacher, the scout leader, the


       friends and family members.  The impact for any


       labeling decisions are going to be quite broad.




                 MS. MUNOZ-FURLONG:  Allergens can appear


       in unexpected places.  This is just one slide of a


       number of examples that we have for "Common Foods


       in Unexpected Places."  Every one of these examples


       have caused an allergic reaction to one of our


       members, because they were not expecting to find


       the allergen.


                 Just to give you an example, if you have a


       milk allergy, you would not have expected that


       barbecue-flavored potato crisps might have milk in


       them, and you might not have read that label, or


       that canned tuna might have soy in it.  Therefore,


       it is not as easy as avoid the food, you've got to


       be looking for unexpected sources.




                 MS. MUN[MLM2] OZ-FURLONG:  We can see this


       reflected in a study that was published in 2002 by



       Joshi, et al.  They took some food-allergic


       individuals, gave them products that were on the


       market, and asked them to read the label for the


       food they were trying to avoid.


                 You can see here that families avoiding


       milk, only 7 percent were able to accurately


       identify milk on the labels that were presented to


       them; for soy, they did a little better at 22


       percent; but peanut, only 54 percent got the label


       reading correct, and most of this was because of


       confusion about allergen labeling information.




                 MS. MUNOZ-FURLONG:  The problem with


       allergen labeling information, there are no


       guidelines or standards for use.  This is


       completely voluntary.  As a result, every company


       has their own decision tree and algorithm and


       wording for what terms they will use and under what




                 This makes it very difficult for us to


       educate consumers and the others who are reading


       labels on their behalf and telling them what to do



       and what these mean.


                 The proliferation of "may contain"


       labeling has really caused us some problems.  Just


       to give you a sense of what is going on, we had one


       volunteer go out in the Northern Virginia area to


       one grocery store and look at products from


       cookies, crackers, candy and bakery.  We were


       trying to follow the model of a previous FDA study.


                 She came back with 28 different versions


       of "may contain" statements.  From the consumer's


       perspective, what does that mean?  Can they be


       trusted, or should we ignore them?




                 MS. MUNOZ-FURLONG:  The current


       environment because of this, there are some


       physicians that advise their patients to ignore


       precautionary labeling, because it is everywhere


       and there wouldn't be any food for them to eat.


                 There are others who tell them, "Heed the


       warning and avoid those foods."


                 Then, there are some companies who tell


       the consumers, "It is on the package only because



       our legal counsel has advised us to put this on




                 Then, there are others that say, "You have


       to trust that wording and not go near the product."


                 How does a consumer determine which is




                 We are also seeing advisory statements for


       peanut allergy only.  The way the consumer


       interprets these statements is that they are


       shortcuts to label reading.


                 If they see "contains peanuts" or "may


       contain peanut," they may not read the rest of the


       ingredient declaration if they are looking for milk


       or soy, because they think that the company


       understands food allergy and would have listed all


       of the allergens on there.


                 As a result of all of this, consumers are


       confused and frustrated.  Particularly what is


       going on as their food choices are further


       minimized is that there is risk taking behavior by


       parents of kids with food allergies who decide,


       seemingly randomly to us, that some companies can



       be trusted and others not, so they will ignore "may


       contain" on the companies they trust.


                 Then, the teenagers, our highest-risk


       population for a severe reaction, want to be like


       everyone else are reporting that they are ignoring


       "may contain" statements, because it is on so many


       foods they have eaten the food and not had a


       reaction, so they don't really believe that these


       are true.




                 MS. MUNOZ-FURLONG:  This is one of the


       labeling studies that we conducted with our FAAN


       members during a spring meeting a year or two ago.


       We asked a question.  They were supposed to answer,


       "I would never purchase a product that says it


       contains" whatever the "allergen" is.  You can see



       that almost 100 percent of them would avoid a


       contain statement.


                 However, as you go from very specific to


       black-and-white to vague "packaged in a facility


       that also produces," say, peanuts or nuts or


       whatever the allergen might be, only 74 percent



       would avoid purchasing that product.


                 Consequently, 25 percent of the allergic


       consumers are going to purchase products where they


       don't really understand the precautionary labeling.


       If the company is putting this on here because of


       some risk, we've got a miscommunication or a


       communication gap going on.




                 MS. MUNOZ-FURLONG: Let's talk about


       thresholds, then.  Again, from the consumer's


       perspective, their physicians advise, as you heard


       from Dr. Wood, is strict avoidance or a reaction


       may occur and you will not outgrow this allergen.


       They are very motivated to try to strictly avoid


       that food.


                 When we talk about thresholds to our


       members, and these tend to be the most motivated


       and well-educated of the food allergy population,


       this is what we consistently get back.  They


       believe that threshold levels may put their


       children at risk because their child is so





                 They also wonder whether the threshold


       levels, the whole discussion is based on the


       industry or the government trying to figure out a


       way not to have to clean or label for allergens.


       Again, they are wary that this might be a loophole


       that is trying to be directed at them.




                 MS. MUNOZ-FURLONG:  The catch 22 here,


       from where we are at FAAN, is that we understand


       that if we label for all allergens at all levels it


       will further restrict diets.  If we further


       restrict the diet, we are going to increase


       frustration which will yield risk taking.


                 It is going to undermine the integrity of


       the ingredient label.  As I showed already with


       "may contain," we are already seeing that.  They


       believe "contains."  However, if we put "contains"


       on everything and they eat it and don't have a


       reaction, we are going to diminish the validity of


       that statement.


                 If we undermine the integrity of the


       ingredient label this will potentially lead to more



       allergic reactions as they take more risk, which is


       going to increase the number of doctor visits;


       hospital visits; and, potentially, fatalities.




                 MS. MUNOZ-FURLONG:  Here is an example of


       what can go on and what we see as what we may all


       be facing.  This is a report that came to us from


       one of our members who had a soy-allergic child who


       had safely eaten soy lecithin in the past.  Most of


       our members, although we tell them to read the


       ingredient declaration on products every time they


       purchase them, become brand dependent and stop


       reading the ingredient label.  That is exactly what


       happened here.


                 This was a product that the child had


       safely eaten in the past.  The mother did not read


       the label, gave it to the child, he started eating



       it.  She then started reading the label and saw


       that it now says "contains soy."  She got very


       nervous and screamed that it contained soy and


       asked the child to spit the food out.


                 Immediately, he started having itching,



       leading to hives, and a feeling of impending doom.


       The mother gave him medication and thought she was


       having a full-blown reaction.


                 The question we have to ask ourselves, Was


       this a reaction, or was it a panic attack?  She


       called the manufacturer and was told that the


       "contains soy" is because it contains soy lecithin.


       Therefore, the ingredients hadn't really changed


       from the product that they had safely eaten before.


                 From our perspective, we do not want to


       see consumers or their families subjected to this


       kind of fear.  Because what you don't realize is


       that once this reaction is taken care of, it takes


       a long time for the family to trust again.  We do


       have reports of children developing eating


       disorders and just being very cautious about being


       around other people once they have had a reaction.




                 MS. MUNOZ-FURLONG:  From the consumer's


       perspective, if we are looking at developing a


       threshold level, and as I said there are pros and


       cons to both sides of this issue, the key here is



       we have got to do a good job of education.  We have


       got to educate physicians and registered dieticians


       so that they can counsel patients accurately.


                 As you saw, we have done no training for


       "may contain."  We have got some doctors that say,


       "Just ignore it."  We can't afford to do that with


       threshold levels.


                 We also have to educate patients and their


       families and assure them that the food is still


       safe and that they can trust the information on the


       label.  We also have to do outreach to the food


       industry so that they can answer the queries from


       food-allergic consumers in a way that will give


       them confidence instead of make them nervous or


       suspicious about whether they can trust the


       information on the label.




                 MS. MUNOZ-FURLONG:  In summary,


       food-allergic consumers want as many food choices


       as safely possible.  This is really why we are here


       and why we are seeing some of this behavior with


       advisory statements.



                 They want to open the diet.  The children


       want to be like everyone else, and they want the


       least amount of restrictions, but they need to be




                 The consumer needs to understand the


       information on the ingredient statement.  They need


       most of all to trust that that information is


       reliable and it is going to be consistent from one


       product to the other.  They also need a minimal


       number of precautionary allergen statements and a


       guideline so that they understand what these


       statements mean and what they should do as a result


       when they see these on products.




                 MS. MUNOZ-FURLONG:  In conclusion, the


       current labeling and manufacturing practices


       present enormous challenges to food-allergic


       consumers.  As Dr. Wood said, the number of these


       patients is increasing.


                 To give you an example, we conducted a


       prevalence study of peanut and tree nut allergy in


       1997, repeated that same study in 2002, and found



       that in that five-year period the number of


       children with peanut allergy had doubled.  We don't


       know how it is continuing to trend, but reports are


       that it is still increasing.




                 MS. MUNOZ-FURLONG:  The bottom line is


       above all we must protect the integrity of the


       ingredient information.  Because from the


       food-allergic consumer's perspective, they depend


       on this information to avoid an allergic reaction


       and, most of all, to maintain their health and


       safety.  We already have data showing that food


       allergy impacts the quality of life.  We don't want


       to further diminish their quality of life.


                 With that, I will end here and open for




                 CHAIRMAN DURST:  Thank you.


                 Does the Committee have any questions?




                 MS. HALLORAN:   I mean, obviously a person


       can survive without ever having to buy any packaged


       food.  I am wondering in terms of the kinds of



       things you were talking about -- teenager's


       preferences, the needs of a busy mother, et cetera


       -- are there particular categories of food that are


       prepared and packaged that are most sort of


       important and essential in our modern life?  I


       mean, would it be bread or breakfast cereal or--?


                 MS. MUNOZ-FURLONG:  If they ate


       vegetables, they would be fine.  How many kids want



       to eat vegetables?


                 (General laughter.)


                 MS. MUNOZ-FURLONG:  I think it really goes


       back to quality of life.  Children want to be like


       everyone else, and they will do everything they can


       to fit that mold.


                 I have a daughter that was diagnosed with


       milk allergy and egg allergy when she was an


       infant.  I will tell you that I did everything I


       could to make sure that she felt like her friends.


                 It is not just the patient or the child,


       it is also the family wanting to not have their


       child isolated or feel stigmatized because of the





                 If everyone else is having breakfast in a


       box, that is what these kids want.  What we want is


       to make sure that those labels are accurate, if the


       family makes that decision.


                 Granted, there are some families that are


       very cautious and will only make food from home,


       make it from scratch.  However, as the child gets


       older and is out with friends, that is just not




                 MS. HALLORAN:  Are there any particular


       categories of foods?


                 MS. MUNOZ-FURLONG:  No.  As you saw in


       that slide, "Common Foods In Unexpected Places," we


       are seeing allergens everywhere.  We have just got


       to make sure that all of the labels are correct and


       can be trusted.


                 CHAIRMAN DURST:  Yes.


                 DR. KELLY:  Ciaran Kelly.  A question for


       you from your perspective and the perspective of



       the people you represent, the patients with food




                 I understand that you are frustrated and



       find it very difficult to work with the current


       system of many different types of wording.  Would


       it be better for you to have a two-level system,


       "does not contain" and "may contain traces of" --


       or even three levels, "contains" and "may contain


       traces of" and "does not contain"?  Would that be




                 MS. MUNOZ-FURLONG:  Well, I will start


       from the back end of your question.  If you poll


       our members or just the general consumers, they all


       want "does not contain" labeling.


                 I would caution to you because of the


       reports I've seen.  This is very widely used in the


       U.K., our colleagues in the U.K. have reported,


       recalls to products that say "does not contain


       peanuts" when they do contain peanuts undeclared.


                 From the way the consumer is going to


       behave if they see "does not contain," they may not


       read that ingredient declaration because that is


       the guarantee they have been waiting for.


                 I am not in favor of "does not contain."


       I am in favor of let's have them read the



       ingredient declaration and know that they can trust


       if it doesn't have peanuts in that ingredient


       statement, the product should be safe for them.


                 When we start to see different allergen


       statements, we want to make sure that those can be


       trusted.  When we are talking about "does not


       contain," that is an implied endorsement or


       guarantee, which makes me very worried.  If the


       company makes a mistake and that is on the label in


       error, we could have someone pay for it by having a




                 Now, if we have two levels, "contains" and


       "may contain," as along as we know what that means


       and that all companies are following this


       guideline, that makes it much easier.  Right now,


       you can go poll 12 companies and they each do


       different things.


                 CHAIRMAN DURST:  I think we need to move




                 Thank you.


                 Our next speaker will be Susan Hefle,


       associate professor and co-director of the Food



       Allergy Research and Resource Program at the


       University of Nebraska, who will be speaking on


       "Allergenicity:  Analytical Methods."


                 Dr. Hefle?




                 DR. HEFLE:  Thank you, Chairman Durst.


                 Good morning.  I am going to discuss the


       basic analytical methods for allergens.  The model


       used is the ELISA-based model which has lateral


       flow.  This model has been used for several years


       now.  We will discuss this more later.


                 Our second bullet, the most successful


       kids do use polyclonal antibodies but occasionally


       a kit uses monoclonal antibodies directed against a


       single protein.  Usually, the antibodies are


       directed against a crude extract of an allergenic


       food not the specific proteins themselves.  It is


       not necessary to really measure the allergen.


                 The industry just cares if any peanut is


       there, not if one particular protein from a peanut


       is there.  "Ara h 1" is a particular peanut


       allergen.  The industry just wants to know if any



       peanut or whichever peanut is there.


                 A lot of times a lot of the successful


       kids use a much more kind of crude approach to


       detecting peanut rather than specifically horning


       on the allergens themselves.


                 There is a challenge, though, in that


       different standards are used in the different kids,


       depending on the manufacturer, and also different


       antibodies are used in the different kids depending


       on the manufacturer.  It is not like a standardized


       approach across the board, necessarily.




                 DR. HEFLE:  The detection limits range


       from around 0.1 to 2.5 parts per million for the


       quantitative methods.  There are also quality


       methods; however, if we are talking about threshold


       levels, we need to talk about quantitation here.


                 Using a method that has a very low


       detection limit has certain challenges.  Every kit



       has the ability to have a low detection limit.  Ten


       years ago, when I started developing kits,


       Steve Taylor and I sat around and thought about



       what the detection limit should be based on our


       years of experience in dealing with consumer


       reactions and things like that.


                 We set a certain level with the kits we


       developed; we picked 2.5.  It seems to have gone


       very well over the last seven or eight years since


       these kits have been on the market.  Some of the


       other companies have a little bit lower range of


       detection limit, and that seems to work okay, too.


                 However, if you go way too late, I mean,


       they can all push these kits really, really, really


       low.  The problem is, Is there clinical relevance


       at that point?


                 If there is no clinical relevance,


       companies may be chasing molecules around their


       processing plant.  They will have all of this


       positive data at a low level, and they won't know


       what it means.  We like to call this "paralysis by




                 We want the data to be relevant.  We want


       the data to be useful.  If the industry goes back


       in and says, "I want to fix this, but what if I get



       all of these positive results at a low level?"


       Detection limits have to be kept in mind.  They


       should be tied to threshold levels, whatever we


       decide the threshold levels should be.


                 It adversely affects the quality of life


       for food-allergic consumers, if you use detection


       limits that are really low or push those detection


       limits without a good clinical basis.  Because of


       the industry reaction in the form of increased use


       of "may contain" label.


                 When they did paralysis by analysis and


       they get positive results, maybe they throw a lot


       of "may contain" labeling on that product that they


       are worried about and so they are going to put that


       on there.  That decreases the number of foods that


       allergic individuals can eat.


                 The current detection limits that are set


       that the industry uses right now have worked very


       well for seven years in protecting the


       food-allergic consumer.


                 I don't think at this point there is any


       need to change them right now.  But, again, as



       science comes in and we know more about threshold


       levels, there might be an adjustment here or there.


                 We just finished an egg threshold study.


       Contrary to what Robert Wood said, you can do


       threshold studies in kids, because we did this in


       30 egg-allergic children.  That is the only kind of


       people we could find to have egg allergy are kids.


                 When we crunch those numbers and look at


       that data, if the threshold is low enough that we


       need to adjust the kids for egg out there, the


       manufacturers have all said they would be willing


       to do that based on the science.




                 DR. HEFLE:  Many companies are testing for


       allergen residues.  What they are primarily testing


       is not-finished product, but they are using it to


       verify sanitation procedures.  They have been using


       them for as long as they have been on the market.


                 Certainly with the new law coming up,


       there are a lot more using them than used to use


       them.  In general in the U.S., companies are


       incorporating testing using these test kits.  As



       the test kits get faster and easier to use, it is


       easier for them to use them.


                 Again, the ELISA or lateral flow, which is


       kind of like a dipstick method are the preferred


       methods.  Some do the test in house.  They really


       like it if they can do that because they can fix


       things right away.


                 However, if you don't have in-house


       capabilities, they will send it out to a contractor


       lab or if they want third-party verification, they


       will do that.


                 Most companies, as I said, are not testing


       finished product.  They are testing to validate



       sanitation methods or doing environmental swabbing


       to try to find where the problem is before they get


       to the final product.  They want to fix the problem


       before they get there and figure out if their


       sanitation is accurate before they get to the final




                 Some testing of finished product on


       certain occasions though is done, especially when


       you can have the product under full control.  They



       don't usually want to release something that they


       have tested and they find out there is a problem


       and they have to call it back from the marketplace


       later and perhaps put consumers at risk.


                 There are tests that are based on DNA


       detection, and they are called "PCR."  We don't


       advocate these for allergenic residue detection


       because it doesn't prove the presence of the


       protein.  You need the protein to have the allergic


       reaction.  It just says that there is DNA from that


       particular allergenic food there.


                 It is not practical at all for in-plant


       use.  You can't put one of these machines next to a


       processing line.  It is very expensive and requires


       a lot of segregation and things.  It is meant more


       for a regulatory agency or a big corporate lab who


       has this ability.  It does not prove the absence or


       presence of the protein or the allergen.  It is


       just an indirect kind of a marker.


                 There are ATP tests out there.  This is a


       test that is commonly used in the industry for


       sanitation assessment.  Some companies would like



       to use this to detect allergens, specifically the


       ATP does not detect protein also.  Right now, it is


       not knowing that these correlate well with the more


       specific protein-based tests.




                 DR. HEFLE:  Three peanut, like ELISA test


       kits, have been performance tested by FDA through


       AOAC-RI.  Those companies with those tests are


       Neogen, R-Biopharm, Tepnel.


                 Five peanut ELISA kits have been studied


       in one JRC interlab trial.  This is the European


       Union's group in Belgium that does these sorts of


       things, and they put these three tests plus two


       more through a validation trial.  They are


       currently doing another validation trial on the two


       peanut lateral flow devices.  They are not finished


       with that yet, but they are only doing one matrix


       not several matrices.  They are just testing it in


       cookies right now.




                 DR. HEFLE:  FDA works with AOAC and has


       said they plan more validation studies with other



       test kits, and that has been the case for more than


       a couple of years now with no apparent progress on


       this front, though.


                 The U.S. food industry and other


       regulatory agencies -- for example, the Canadian


       regulatory agency, the JRC -- has moved way ahead


       of FDA/AOAC at this point.  The industry is not


       running validation trials themselves, but they run


       in-house validation things like that.


                 However, there are regulatory agencies who


       have said, "Well, we're going to move ahead.  We


       can't wait for AOAC anymore.  We have to get these


       things done in validated interlab trials."  There


       are several trials that are planned right now


       internationally to, hopefully, get some of these


       things "validated" in the next few years.


                 The U.S. industry has been testing for


       about seven years now, since the first peanut tests


       came out.  They have increased the amount of


       testing each year and, I've got to say, have spent


       millions of dollars once they've gotten test


       results to change equipment, to make modifications,



       for allergens specifically.


                 Before about 10 years ago, we didn't have


       any tests at all to do this.  Since the tests have


       become implemented, they have used them to make


       changes in how they manufacture food.


                 Health Canada/CFIA has a Compendium of


       Food Allergen Methodologies.  They crunch through a


       lot more of these kind of in-house validations that


       they do so that they can use them for their


       purposes.  There is a Web site for that.  They use


       both commercial and their own in-house methods.




                 DR. HEFLE:  Validation of kits, there are


       more JRC trials coming out of the EU more likely.


       We know of several that are planned, and other


       groups have them planned, too.  Other groups are


       planning more interlab trials, some with kind of


       "modeled" foods.


                 A lot of these tests are done where you


       spike peanut into something else.  It is not really


       like a manufactured product, so it doesn't really


       mimic the manufacturing process.



                 A "model food" is actually where the


       allergen is manufactured into the matrix, so that


       it more appropriately represents what would happen


       in the food industry.


                 Therefore, those are king of challenging


       to make.  You can't just make them in your back


       yard or in your home kitchen.  You need to make it


       on an industrial level, so it can be quite an






                 DR. HEFLE:  Kit companies do much more


       extensive validation than ever will be done by any


       regulatory agency or academic center.  It is


       usually that the are in the process of selling


       kits, and they don't necessarily share the data


       like they should.  I have been encouraging all of


       them to go ahead and publish all of this great data


       they have, and it would be a lot easier for all of


       us to evaluate how good their kits really are.  So


       far, they still want to sell kits and not spend


       time writing papers.


                 However, they do have liability issues. 



       Their kits have to work.  They have liability


       issues.  They have reputation issues if the kits


       don't work, so it behooves them to do their own


       validations before they put a product on the






                 DR. HEFLE:  Reference materials are solely


       lacking for allergens.  It would be really nice if


       we had a bunch of reference materials we could do


       all of these interlab validations with.


                 However, we are having a problem finding


       the appropriate reference materials.  There are not


       many available, and they are really needed.  NIST


       is one source of reference materials.


                 Unfortunately, the NIST standards that are


       available were not made for allergen testing, were


       not designed for that and often do not represent


       the type of allergenic materials used in the food




                 A case in point was the standard that was


       used in the AOAC-RI-FDA study.  It was peanut


       butter made by a major manufacturer.  It is fine



       for things like aphlatoxin determination and other


       things.  Unfortunately, the varieties are not known


       with certainty, because the manufacturer wouldn't


       tell FDA about every little peanut that might be in


       there.  They wouldn't divulge it.  I'm referencing


       NIST not FDA, I'm sorry.


                 Different peanut varieties have different


       responses in the kits.  It is imperative to know


       exactly what is one of these standards.


       Unfortunately, there aren't a whole lot of other


       standards around the world around to do that.


                 There are other sources of materials that


       could be used as reference materials, but we have


       to come to a worldwide decision on what is the


       appropriate criteria for considering something in


       the reference material.  Is something the JRC makes


       in Europe representative of something we use in the


       United States?


                 There are several of these materials


       available, and we could begin to talk about going


       through some interlab trials with some of these, if


       they met certain criteria.





                 DR. HEFLE:  Processing can have a huge


       effect on extraction and kit performance.  Most


       kits are not validated using these model foods, so


       we have to do some more of this stuff;


       international call for more use of modeled foods.


                 The old method of spiking, which is where


       you put a peanut extract into some of the matrix


       and mix it together and see how it performs.  This,


       again, does not truly represent what happens in the


       food industry.


                 However, the spiking does provide some


       useful information, but the manufacturing of these


       model foods gives the best information about how a


       kit will work.


                 Model foods have to be made on a pilot,


       plant or industrial size scale.  If you make this


       in your backyard or your kitchen, then it doesn't


       really appropriate what a model food is in the


       industry, either.


                 If you make many cookies in a home-size


       oven or a Suzy Homemaker or Easy-Bake Oven, it is



       not going to be the same thing as what Keebler or


       what Pepperidge do on a huge scale.


                 The results of these are not practical or


       useful for the food industry.  Let's make some real


       model foods.  They are involving for assessing how


       a kit is going to work with a specific commodity,


       how efficient the extraction method is under


       industrial conditions.


                 It is becoming more and more important to


       use these types of standards in assessing the kit's


       performance for certain commodities and processing.


       I think spiking is pass.


                 I get yelled at in my professional


       society, AOAC, because spiking is the way of the


       food chemists.  However, we have to do some spiking


       and look at things, but we have to make these model


       foods and do that sort of assessment also.




                 DR. HEFLE:  The extraction method, is it


       sufficient?  We've got to think about it.  Is it


       sufficient?  Is the recovery good?  Can we trust


       the results?



                 Some foods are challenging.  There are


       tannins and polyphenols in dark chocolate that bind


       protein.  It is a famous matrix, one of the most


       difficult matrices to do with allergen




                 High fat levels can hide the allergen in


       other types of ingredients.  If the product is


       hydrolyzed, you cannot analyze hydrolyzed or


       fermented ingredients in these test kits.  They


       were never designed for this.  When you start


       chopping up the proteins, the ELISA signals go


       away.  The methods are meant to detect intact


       proteins and not peptides.


                 Processing, if you burn stuff, it is going


       to be less detectible; it is less soluble.  That is


       a factor.  Now, most companies don't burn their


       food, but sometimes they want to detect burned


       foods on band ovens or something they can't readily


       clean.  These are challenges to kit performance,






                 DR. HEFLE:  Most kits for most allergens



       have good reactivity with processed forms of the


       allergenic food in my experiences over the last 15


       years, and that is just my experience.


                 The use of polyclonal antibodies and crude


       extracts and making antibodies against processed


       forms are recipes for successful kits.  There are


       several on the market today that do very well.


                 Monoclonals are okay if they use a


       heat-resistant epitope in making the monoclonals.


       They can accommodate the processing changes that




                 Some of the egg residue kits have some


       issues in this regard.  The industry has been able


       to adjust and adapt.  Many survey the raw material


       instead.  Instead of worrying about the processed


       egg, they will just do the raw egg and handle it


       that way, or use a kit that has antibodies against


       raw and processed egg, to get around that


       particular issue.


                 Matrix effects, my lab has used all of the


       ELISA-based test kits available on the market in


       our own validations and tests.  It is kind of my



       hobby so I like to do this.  The matrix effects are


       usually not a problem for most of the test kits out


       there, for the vast majority.


                 Kit companies have added extraction


       additives to their extraction buffers to assist.


       When it was recognized dark chocolate was a


       problem, they added some secret extraction


       additives to help you pull the protein out of dark


       chocolate easier.


                 Model foods, though, again are going to be


       of great use in assessing the true extraction


       performance of a kit.  Again, I can't stress enough


       we need to make more of these.


                 In cross-reactivity issues, even though


       most methods do use polyclonal antibodies, which


       those of you who know something about polyclonal


       antibodies could say, "Boy, there could be a lot of


       cross-reactivity problems with them."


                 We really don't see this happening.  The


       kit companies really couldn't sell any kits if


       their peanut kit cross reacted with everything


       else, too.  Therefore, we don't usually see these



       problems in that they have looked at that before


       they have launched it, so we don't see the


       cross-reactivity.  I am not saying that there isn't


       one that is going to crop up sometime.




                 DR. HEFLE:  Again, we've got a problem


       with hydrolyzed proteins, hydrolyzed vegetable


       proteins, hydrolyzed soy proteins.  You can't


       really detect them.


                 The industry would love to do this, to


       chase them through the facility and see if they


       have cleaned up afterwards because we know there


       can be some residual allergenicity in hydrolyzed


       protein preparations.


                 However, the ELISAs are pretty much


       rendered useless when trying to analyze for


       hydrolyzed protein.  It is not what they are


       designed to do.  The company has had to make a


       decision, "What is most of our market?"  It is not


       chasing hydrolyzed proteins, but it is chasing the


       intact proteins.  We have to balance the kits to go


       towards that, so you can't use it for this.



                 Unfortunately, a negative result in an


       ELISA in this case does not mean that there is no


       allergenic residue left.  You have to ascertain


       residual allergenicity via a different method using


       human allergic IgE in something like a Western blot


       or a RAST analysis.


                 Another related area is the analysis of


       fermented ingredients: gums, Lactobacillus


       cultures, starter cultures.  Once they start eating


       at the substrate, the proteins are partially


       hydrolyzed and the ELISAs won't detect them


       anymore.  You need to use an IgE-based method to


       just ascertain the true allergenicity.


                 Companies don't tell contract labs the


       nature of their samples.  They just say, "Here is


       Sample X."  They are not going to tell them it is


       hydrolyzed, so we have some challenges.


                 I try to communicate with the contract


       labs and say, "Be sure you ask the question.  Just


       don't give them a negative result, because it


       couldn't be truly negative maybe from an allergenic


       standpoint."  I think this is the minority of the



       samples out there.




                 DR. HEFLE:  My lab performs testing for


       food-allergic consumers, their physicians, their


       lawyers when they call for free when they report a


       reaction to a food.  We work with some members of


       the Food Allergy & Anaphylaxis Network when they


       have a problem.


                 If there is an analysis I can do, I will


       try to help a food-allergic consumer identify what


       happened with that particular food, if they have


       managed to keep it in the height of the moment.


                 In 10 years of doing this, we have only


       seen "large" -- now notice I say "large" with a


       quotation around it, I don't want to make a lot of


       judgments on that right now -- amounts of


       undeclared allergenic food causing reactions.





                 DR. HEFLE:  We cannot currently do


       immediate monitoring in the food industry, though.


       The technology doesn't exist.  It is getting



       better.  These lateral flow devices can sometimes


       get down to 5 minutes now.  I think in the future


       they will be able to make a more immediate




                 Right now, a lot of them are 30 minutes


       long.  If you are swabbing things and waiting


       around for 30 minutes to see if the result is


       positive and then having to go back and clean


       again, it is pretty impractical for the food


       industry to do.


                 Sanitation and verification is the most


       practical, not the test and release kind of thing.


       My dad is a fisherman, so I like to the catch and


       release and test and release kind of analogy.


                 We do not have tests for some of the


       allergens, and fish is a notable example.  You


       cannot test for the hydrolyzed or the fermented


       allergen sources using these types of methods.


                 Some types of cross-contact are not


       homogenous or 100 percent cleaning is not possible


       due to the nature of the product.  Food equipment


       was never historically designed for allergen clean.



                 Sometimes these facilities are quite old,


       and there is no room.  There is no room to bring in


       different equipment.  They have to try to redesign


       as they can, but they can't get completely rid of


       hangup areas.


                 You cannot take enough samples to


       practically test, to be a hundred percent sure all


       of the time.  That is impossible.  If I get a


       statistician in to tell me how many samples I would


       need, the industry would just spend the whole day


       testing rather than trying to make food product.


                 In some of these cases, precautionary


       labeling is justified due to the nature of the


       product and the process in FARRP's opinion.  For


       example, dark chocolate and milk chocolate on the


       same line is one example where we think


       precautionary labeling is justified.  That doesn't


       mean we think precautionary labeling is justified


       in every case.




                 DR. HEFLE:  This is a study that we


       recently completed and published in 2004 of some



       incidents from milk allergic consumer complaints.


       These were the casein levels we found in those


       particular products.  They range from 5,000 on up


       to 44,000 parts per million in things that were


       supposed to be free of milk or labeled even


       "dairy-free" or "kosher," quite high numbers of


       parts per million.


                 They also asked me to talk a little bit


       about highly refined oils.  What does HRO mean?  In


       FARRP's opinion, "highly refined oil" means


       neutralized, bleached and deodorized or refined


       bleached and deodorized.


                 The definition of what "refined oil" is,


       is kind of debated a lot right now in terms of


       FALCPA, opinions based on scientific review of oil


       challenges with oils in the literature and what we


       feel refined oil should be.


                 The available quantitative methods, there


       are methods used in the literature including ELISA


       and other methods that reports the levels of


       protein in highly refined oils.  None of these,


       though, have been validated in interlab trials or



       other types of validation for protein and oil


       determination to date.


                 Somebody will run something and they will


       report it, and they will do a certain number of


       samples, but no one has looked at whether that is


       an appropriate method across the board for


       detecting this.


                 There is a question as to whether a small


       amount of protein in the HRO is completely


       extracted in aqueous buffer.  "Aqueous buffer" is


       something that people often use to do these sorts


       of biochemical tests.  It means trying to partition


       the proteins from the oil into an aqueous buffer.


                 If they really like oil, they might not


       all come over.  They might want to stay in the oil.


       The question is, Does this capture the true protein


       content of the oil or whether some of the more


       hydrophobic proteins stay in the oil fraction, and,


       therefore, do not get extracted and therefore




                 My lab uses an amino acid determination


       based on Edman degradation, but we also use aqueous



       extraction.  We try to maximize that aqueous




                 We use heat; we use a large amount of


       buffer; and we concentrate the sample.  However, I


       cannot guarantee that I'm pulling all of the


       protein out of that highly refined oil when I


       measure that.


                 We report the results as relative and not


       a complete picture of the possible protein count


       out of HR oil.  I still think you are capturing


       most of the protein that is there, but I just can't


       sit up here and say we are covering a hundred


       percent of it.




                 DR. HEFLE:  The protein levels of HRO are


       reported in the literature, and there are lots of


       different reports and levels.  The caveats again:


       The use of aqueous buffers in the determination,


       how good if they use an immuno-chemical-based


       method is the epitope recognition of the antibody?


       Does it really recognize those soy proteins at that


       level of processing?



                 Relating "total" nitrogen, sometimes they


       use the total nitrogen amount to what the protein


       is.  Well, total nitrogen can be free and running


       around in the protein and not associated with --


       free and running around in the oil and not


       associated with the protein.  Consequently, it may


       be an overestimate actually of the protein amount.


                 Limitations of certain types of methods


       like dye binding.  "Dye binding" is a method that


       will bind to certain proteins preferentially and


       not bind to others as well.  When you use a


       dye-binding method, is it really representative of


       everything that is in there?  You can't absolutely




                 The protein levels reported in the


       literature are usually a few milligrams per


       kilogram, which are a few parts per million.  You


       will see some widely ranging estimates, though,


       from different investigators.  A lot of times I


       question their methods sometimes or their ability


       to reproduce that particular result.


                 I think that is the end of my



       presentation, and I thank you very much for your




                 CHAIRMAN DURST:  Thank you.


                 Committee, do you have any questions or




                       QUESTION-AND-ANSWER SESSION


                 DR. MALEKI:  Soheila.


                 DR. HEFLE:  Soheila.


                 DR. MALEKI:  Yes, Soheila Maleki.  I was


       wondering, just based on your experience and you


       have been around a lot of industry, if there is any


       kind of correlation or if there are any standards


       between what the companies use to label "may


       contain" versus "contains"?  Do they use the same


       2.5 parts per million that the kits provide as a


       may contain or a not contain and so forth?


                 DR. HEFLE:  They don't really use the


       analytical results to make a definite decision


       about that.  Usually, the companies make a decision


       to put precautionary labeling on through a certain


       stringent set of criteria.  It is something they


       have tried to clean up, and they are still having





                 They have intermittent contamination.


       They would never allow something that consistently


       had a significant amount of allergen in it to be


       called a "may contain."  They would try to clean up


       more, if it is not supposed to be there.


                 They don't set a level like that.  They


       use the analytical results to help them determine


       whether that is justified or not.  It has to be


       potentially hazardous, intermittent, hard to clean.


       Those sorts of things are taken into consideration


       much more than just the simple analytical result.


                 DR. MALEKI:  Thank you.


                 CHAIRMAN DURST:  Yes.


                 DR. NELSON:  Mark Nelson.  I just wanted


       to follow up to that in response to Soheila.  In


       2001, the food industry, a group of associations



       representing their members did put together


       guidelines on labeling.


                 The preference is obviously and clearly


       the requirement is to label the ingredient in the


       presence of an allergen when it is directly added



       to the food.  In the situation where there is a


       potential for cross contact, we did establish some


       guidelines before companies should use "may



       contain" labeling because of the concerns we have


       heard about before.


                 One of those key guidelines was to make


       sure that we could not avoid it even after applying


       good manufacturing practices: appropriate cleaning,


       appropriate separation, and so on and so forth.


                 DR. MALEKI:  I see.  Depending on how much


       you detected, it didn't matter, if you detected, it


       went to "may contain," if it was on the line or --


       well, if it contains it was directly added to the


       product?  I'm trying to make sure I understand that




                 DR. NELSON:   Yes, I think it is more to


       Sue's point that we aren't necessarily measuring


       the finished food so much.  It is not a catch and


       release situation.


                 DR. MALEKI:  I see.


                 DR. NELSON:  It is understanding your


       system; what ingredients are going into the food;



       what other products might be made on that line;


       validating your cleaning processes between


       products; scheduling products, depending on the


       ingredients that they contain; the sequence in


       which you might make the product and so on.  There


       are a lot of things that go into it.


                 CHAIRMAN DURST:  Anything else?




                 DR. CALLERY:  Pat Callery.  It appears


       that the allergens themselves are not that well


       defined, especially when you can find in actuality


       generated new allergens by treating food in a


       certain way.  I am wondering how you address the


       analytical problem of false negatives and false




                 DR. HEFLE:  For a lot of foods the


       allergens are indeed known, and there are very rare


       cases where you make new allergens through


       processing.  That is an extreme case in the


       literature, I think.


                 However, false positives and false


       negatives are evaluated at the company level first



       by testing tens of thousands of food commodities


       and looking for potential issues.  Also, I kind of


       poke around myself and see if there is anything


       that I can challenge the kits with.


                 In my experience, the false positive/false


       negative rate for most of these methods is very


       low.  I can't give you a number.  I can't tell you


       how good that is, because I haven't done a


       systematic study.


                 However, I think that the use of these


       interlab trials with model foods will help us look


       at some of those issues a little bit more, but I


       don't have a good sense of how much false positive


       and negative is out there.


                 I just know in our experience, and we use


       these every day, we don't have a lot of issues.


       When the occasional issue crops up, and we call the


       manufacturer.  We usually work through it pretty




                 They do tell the manufacturers to validate


       or run their own in-house validations before they


       truly test the results.  The manufacturers do tell



       the manufacturers to do that, so, theoretically,


       they should hopefully find some of these things.


       However, every method has a chance of a false


       positive or a false negative.


                 DR. CALLERY:  I'm not sure how you do that


       without standard materials.


                 DR. HEFLE:  I'm sorry?


                 DR. CALLERY:  I don't know how you do any


       of that without standard materials to validate




                 DR. HEFLE:  Some of the manufacturers will


       give you a standard to work with, either the


       standards from the kit or a recognized standard or


       perhaps one of the NIST standards, which is what we


       are all defaulting to because we have nothing else.


                 DR. CALLERY:  I think you mentioned that


       one kit, they have some secret materials that they


       put into the kit to help extract protein.  This


       seems inconsistent with being able to validate a


       method if you don't even know what the test


       material, how it was made and what the scope of the


       antibodies are that are made.




                 DR. HEFLE:  Well, the extraction additive


       is not a reference material.  The extraction


       additive is just an aid in extraction.  Usually,


       the companies will tell you what it is.  It is


       usually non-fat dry milk or soy protein.  It is


       secret, but it is not that secret.


                 It is just an additional protein in the


       mix that helps pull the proteins out of oily


       matrices or hard to extract matrices.  The


       companies know this, and they share that with


       customers.  However, these sorts of extraction


       additives aren't really the reference materials or


       the standards used in the kit.


                 CHAIRMAN DURST:  Sue, will you be around


       for discussion this afternoon?


                 DR. HEFLE:  Yes, I will.


                 CHAIRMAN DURST:  I think we will hold


       further questions until that time because we are


       running a little bit late.


                 DR. HEFLE:  Okay.


                 CHAIRMAN DURST:  I would like to take the


       recess now.  We will take a 10-minute break and



       reconvene at 10:45.


                 Thank you.


                 (Thereupon, from 10:30 a.m. to 10:40 a.m.,


       there was a pause in the proceedings.)


                 CHAIRMAN DURST:  We will start with our


       next speaker, who is Dr. Stefano Luccioli, who is a


       senior medical advisor to CFSAN, FDA.  He is also


       assistant professor at Georgetown University.  He


       will be speaking on "Oral Challenge Studies:


       Purpose, Design and Evaluation."


                         ORAL CHALLENGE STUDIES:


                      PURPOSE, DESIGN AND EVALUATION


                 DR. LUCCIOLI:  Thank you, Dr. Durst.


                 Good morning.  Today, I really want to not


       talk to you as an FDA medical officer, but as an


       allergist who has experience in performing and


       evaluating oral challenge studies.




                 DR. LUCCIOLI:  The goals of my talk today


       are basically just to give you a basic overview of


       oral challenge studies, the purpose, why they are


       done, the design and conduct, and also spend a



       little time on evaluation and interpretation of


       data, especially with regard to sensitivity of


       subjects as well as clinical response and severity


       and maybe present some data gaps that may be of


       interest while you deliberate on thresholds




                 DR. LUCCIOLI:  The purpose of challenge


       studies are manifold, but the primary reason is to


       diagnose allergy, food allergy.  The gold standard,


       as we have already heard, is the double-blind,


       placebo-controlled, food challenge.


                 As we have heard, also people outgrow


       their allergies.  They are done also to evaluate


       tolerance where those individuals have outgrown


       their allergies.  They have also been done to


       evaluate specific ingredients that are allergens in


       specific populations.  For instance, there have


       been some studies on highly refined oils in


       peanut-allergic populations.


                 However, in recent years, there has been a


       lot of emphasis on using oral challenge studies to


       determine minimal eliciting doses.  This has



       important implications potentially to determine


       sensitivities of individuals within a population,


       but also potentially some therapeutic opportunities


       in that, as Dr. Wood had mentioned, there is a


       feeling that maybe if we can't cure food allergy,


       maybe we can raise people's sensitivity levels so


       that they may not react to very low trace amounts


       of food.


                 For reasons that you are all here today,


       also for establishing threshold challenges, they


       may be able to provide you data on low-effect


       levels and no-effect levels.


                 A problem in this field is that there are


       insufficient animal models which are commonly used


       to evaluate toxicologic ingredients and also


       scattered data about case reports where there is


       not a lot of information about exact doses that


       cause reactions.


                 Very few studies are done or have been


       done.  One study was reported by Dr. Wood on


       evaluating reaction severity, and we don't have any


       current biomarkers to predict severity.  This is an



       important, I think, factor when we are looking at


       evaluating minimal eliciting doses.




                 DR. LUCCIOLI:  I'm just presenting this


       slide, but I'm not really going to go into it, to


       just give you an overview that oral challenge


       studies are somewhat different to traditional tox


       models that are used to determine potential


       thresholds or acceptable doses.  I will, hopefully,


       be able to highlight some of these issues in my


       talk and present, as I said, some data gaps.




                 DR. LUCCIOLI:  When you are designing oral


       challenge studies, obviously the selection of


       subjects is an important factor.  Usually, you have


       populations of adults, children or infants just to


       keep the statistics in order.  Most studies involve


       both men and women as well as are from foreign


       countries and most high ethnicities.


                 The selection of subjects is basically


       geared to what the purpose of the study is for,


       whether you want to diagnose individuals with an



       equivocal IgE or clinical history; evaluate


       evidence of outgrowth of tolerance, as we have


       mentioned; and also potentially to evaluate


       co-existent allergies,  for instance, milk-allergic


       individuals who may have soy, especially in the


       infant population and also for evaluating specific


       ingredients, in this case how to evaluate infant




                 Obviously, for specific ingredients, you


       may want to pick particular populations for that.


       In fact, most infant studies are done to evaluate


       infant formulas, and the majority of studies are in




                 Another important factor is that there is


       a notable exclusion of individuals from these


       studies.  As Dr. Wood had alluded to, there are


       individuals who have a cutoff level of their IgE


       where above this level they have a 95 percent or


       more risk of already failing the challenge.  The


       challenge is basically useless.  You already have


       the information, and you tell those individuals to


       avoid the food.



                 However, these individuals may represent a


       fairly sensitive population.  Now with IRBs as they


       currently stand, it is very difficult to get these


       individuals tested in studies.


                 Also, classically individuals who have had


       anaphylaxis or very severe reactions which were


       fairly convincing for the actual food are excluded


       from the studies, because another rule of thumb is


       do no harm.


                 Consequently, you don't really want to


       test people who could have potentially severe


       reactions when you have already had a high clinical


       index that they are allergic.


                 Of course, there are a lot of people who


       self-exclude themselves from studies who may be


       part of a sensitive population.  I also mentioned


       here unstable asthma because in any study you don't


       want to test individuals who are unstable to begin




                 Individuals with asthma tend to have more


       severe reactions and are probably the group most


       representative of fatal reactions.  By not



       including these individuals, you may be missing not


       only sensitive individuals but individuals who are


       potentially very severe responders.




                 DR. LUCCIOLI:  With regards to test


       materials, there is a variety of test materials


       that can be used.  Various preparations, if you


       just look at peanut, you can have peanut flour,


       ground peanut, peanut butter.


                 There is evidence that the processing


       method of these various preparations may affect the


       allergenicity profile of proteins within these




                 You may have some individuals who are more


       sensitive to peanut flour versus peanut butter.


       The importance, too, with choosing the material is


       that for logistic purposes you want to have it for


       an increased time, if you are going to be doing


       challenges over multiple months or time points.


                 A preferred method for these types of


       ingredients are dried ingredients.  You get into a


       problem where dried milk or spray-dried egg are not



       very commonly ingested ingredients in the


       population.  It is more common, I mean, the raw or


       cooked egg or milk, liquid milk.  Therefore, these


       are factors that need to be assessed.


                 Also, fresh versus processed foods, some


       individuals are more likely to react to the fresh


       food versus the processed as well as raw versus


       cooked.  These are issues that need to be


       considered when you choose a food for a particular




                 Then, the dose units are different within



       these challenges.  Some studies report milligram


       for food; others milligram for protein of food;


       and, very rarely, milligram per kilogram which


       would be fairly ideal if we wanted to evaluate


       potential differences between adults and smaller


       adults, infants.




                 DR. LUCCIOLI:  Obviously, people who


       partake in these studies are people who think they


       have an allergy; may have had a fairly significant


       reaction; and are, understandably, under a lot of



       stress and are afraid.


                 Blinding is an important fact, since there


       is unfortunately a high incidence of the "nocebo


       effect," which is actually the opposite of placebo,


       people reacting to a substance that they think is


       going to harm them.


                 In blinding it is important to mask the


       food, because you don't want the subject to know


       what they are eating.  Factors that are used are


       called "vehicles" in one sense, and they are


       basically other types of foods that are thick that


       can hide the taste and smell and texture and that


       are also pleasant tasting, you hope.


                 However, when you are thinking about doing


       a challenge study over a few time limits, obviously


       you don't want to give some of these vehicles too


       much of this, too many milkshakes -- you have to


       make sure that the individual is not milk allergic


       -- but also they may cause some GI effects or other


       things independent of what the actual food that you


       are studying would have.


                 In some cases, they don't always mask the



       taste.  Therefore, some researchers have preferred


       to use capsules, since this basically bypasses the


       taste issue.


                 However, using a capsule is difficult,


       especially if you are going in higher doses of


       food, it is hard to put a serving of some food into


       a capsule.  I think people would know when they see


       a big capsule that there is more food in that.


                 Also, an important factor is that you may


       delay the absorption of that food putting it in a


       capsule, and also you bypass the oral cavity which


       may be a primary target organ for the initial


       allergic response.  You may have not only a delayed


       response but potentially a less severe response.


                 I won't talk about the protocol, I think


       that was basically well-mentioned by Dr. Wood, but


       also a question about placebos.  There are some


       studies that use placebos within the challenge.


       They use a dose and then the next is a placebo.


                 You know, it is a very complicated process


       where you usually need some other people that blind


       those to both the researcher and the subject, but



       they are used as well.  However, I think the


       preferred method and the easier method is to do a


       separate placebo day.




                 DR. LUCCIOLI:  Now this is just a


       schematic of an example of a dose protocol.  I


       think the important factor is this is an escalation


       study of divided doses.  One of the important


       things, too, is you don't want to be there all day,


       and you don't want the patient, too, to be there


       all day.


                 To be able to determine a dose of food and


       get up to the final dose, which is usually a


       serving of the food, which is like 10 grams of


       solid or 60 grams of wet food is what you want to




                 If there is no response at that dose,


       there is a good likelihood that the challenge is


       negative.  However, in many cases you still want to


       have the patient come back and do an open




                 Now, with choosing the starting dose, this



       varies among studies.  In many diagnostic studies,


       because of this issue about not wanting to be


       there, you choose a dose that is roughly half of


       the dose that caused the reaction.


                 Now, I don't know how a lot of people


       figure that out, but that is what has classically


       been used as the starting dose.  Even within a


       study, these doses shifts.  This dose usually comes


       out to be in the milligram range.


                 Now, more recent studies that have


       actually been targeted to study minimal eliciting


       doses, have started in doses in the even microgram


       range.  However, there are a variety of studies


       when you are looking at evaluating studies for


       eliciting doses.


                 Also, in this protocol, it is important to


       know the time interval differences.  Usually, also


       that is tailored to the patient when their symptoms


       first occurred.  Most allergic reactions occur


       within 15 to 30 minutes, so that is usually the


       time gap, but some other reactions may be a little


       bit more delayed.



                 As Dr. Wood discussed, there are


       individuals who have delayed reactions as well.


       Unfortunately, it is just not logistical to do a


       study and wait for these people's reactions to


       occur, because they might not occur that day; they


       may occur on a separate day.


                 In this model that I use, I just use a


       twofold dose incrementation, but also this could


       vary.  Some studies go up to even tenfold, so this


       could affect also the starting dose and


       interpretation of doses in the dose response.


                 Now, you go and you do the challenge.  If


       it is negative, it is negative, or you stop it


       after the first objective symptom occurs.  Some


       studies will also record the subjective symptoms,


       but that is not always the case, because the


       objective symptom is the symptom that denotes a


       positive allergic response.


                 When you record the dose, you can either


       record it as the 4X, which is the discrete dose


       recorded or the 7X, which would be the cumulative


       dose adding the X, 2X or 4X.



                 Just to put this also into some


       perspective in terms of safety assessment, when we


       are talking about LOAELs and NOAELs, the 4X would


       be the low-effect level for this study.  If there


       are doses before that, at least for this individual


       you can say that this dose did not cause a response


       and could be considered a no-effect level.




                 DR. LUCCIOLI:  Some other issues are don't


       do this at home.  People can have a very severe


       reaction.  These studies are done in a clinic or an


       office where there is emergency equipment and


       personnel.  It is not a challenge that is done out


       in the open.  It is in an experimental setting, so


       that can also affect the interpretation or results.


                 Medications, too, most studies now have


       people stop the medicines, but with some earlier


       studies this was not a factor.  Antihistamines and


       other things, if people are on these drugs, may


       block the early responses so that can factor in.


                 Fasting, too, most people fast before the


       study, but in some studies this was not necessarily



       explained.  If you have a full meal right before


       the challenge, this could affect, potentially


       affect, absorption of the allergen and therefore


       affect the interpretation of the study.


                 The clinical history or reactivity, too,


       is important.  Dr. Wood talked about oral allergy


       syndrome, but he did not mention about exercise.


       There are some individuals who eat a food and have


       no problem.  However, if they eat the food and


       exercise, they have a problem.


                 Some studies actually test the individual


       and then put them on a treadmill and have them


       exercise to see if you can elicit the reaction.  I


       mean, this is very rare, but that is something that


       also can be done in terms of the oral challenge






                 DR. LUCCIOLI:  Statistical endpoints, I


       think these are fairly straightforward for most


       challenge studies.  You want to just know what


       percentage of individuals will react or not react


       to the challenge, or in cases where you are



       studying reaction severity which ones will have a


       mild versus a severe reaction.


                 If you assume that all of these


       individuals in the study are part of the sensitive


       population or general population, you can maybe


       make some assumptions about that and decide a


       percentage that will or will not react to a


       specific food concentration.


                 Also, there an importance in this is also


       when you are designing a study, you may want to try



       to achieve a certain number of individuals to give


       you confidence levels for the incidence of allergic




                 In this example, this is a table that


       shows over here (pointing) the number of


       individuals that need to be tested to give you a


       confidence level that the incidence will be less


       than this.


                 For instance, if we were to design a study


       with 66 people, that would give us 99 percent


       confidence that 1 in 10 would potentially react, so


       90 percent would not react.  Also, you could use if



       66 is more than 59, you could also say, well, 95


       percent confidence that 95 percent, 1 in 20, will


       not react.


                 Twenty-nine has been usually seen as a


       magic number for infant formulas.  If 29 patients


       do not react, if the infant with milk allergy does


       not react to a cow's milk infant formula, that is a


       basis for hypoallergenicity.




                 DR. LUCCIOLI:  I will spend the rest of my


       talk on evaluation and interpretation of challenge



       study data.  Basically, a general interpretation as


       we just talked about the statistics, many of these


       studies are done in a very small population of


       patients, therefore you cannot make a very general


       assumption for the general population.




                 DR. LUCCIOLI:  Because some of these


       studies do test the same food, there is a tendency


       to group these studies together to try to get the


       power higher and then potentially make some





                 The problem with this is that I think it


       is important to note that all of the studies that


       are currently available are not standardized.  I


       think that was a question asked just a little




                 This is not standardized data.  They are


       not standardized to dose.  Starting dose or


       blinding or testing could also be a factor and also


       interpretation of clinical symptoms, which I will


       address a little later.


                 Another issue here is that all sensitive


       populations, are they included.  If you have


       information only on adults, is that going to


       predict what harm it will be to infants.


                 Again, in terms of statistical power, if


       you get individuals who are not reactive, if you


       are looking at total numbers to say "This is how


       many people did not react to this dose," well, what


       about people who didn't react to the challenge at


       all?  Should they be included in the final analysis


       of individuals, or should only the ones who react


       to the challenge be part of that analysis?



                 What about foreign study data.  For


       instance, China has a very low prevalence of peanut


       allergy, presumably because peanuts there are


       boiled or fried versus in this country they are dry


       roasted.  If you have all of this data in the


       United States about peanut allergy, could that be


       transferred to data in China?




                 DR. LUCCIOLI:  I just should mention, too,


       that with standardization it is important to note


       that there have been some very nice reviews on


       actually proposed protocols, standardized


       protocols, for food challenges which have been


       published in the last year or so.  However, to my


       knowledge, there have been no studies that have


       used this protocol at least for a major food


       allergen for evaluation.


                 Another general interpretation is that


       this is an experimental exposure.  It is not real


       life.  There could be false negatives.  Individuals


       who have had a negative food challenge go out and


       have an open challenge and react.  It is not always



       a definitive assessment of allergy.  Also, I think


       it is difficult to predict reactions to future


       exposures.  I will try to talk about that as we


       come up.




                 DR. LUCCIOLI:  Subject sensitivity, this


       is I think an important issue to consider when


       looking at evaluating food ingredients.  The


       genetic heterogeneity of individuals, there are


       multiple allergens in food.


                 People can be sensitized specifically to


       certain allergens within that food.  If you cook


       the food in a certain way or process it, you may


       affect their allergenicity positively or


       negatively.  This may be what is apparent when they


       do studies and you see this enormous gap in




                 You have almost a millionfold gap between


       the high responders or I should say the least


       sensitive who respond to low doses and the most


       sensitive to who respond to high doses.


                 There is also this potential link with



       severity, as Dr. Wood study has suggested and some


       others, that some studies suggest that the


       individuals most sensitive to low doses appear to


       have the most severe reactions.  Are we talking


       about a specific subpopulation of individuals here


       who are not only sensitive but severe?  Also, there


       is a sensitivity issue between foods and between


       food products.


                 Another important aspect is that the


       individual sensitivities may vary over time.


       Allergies can progress and individuals with food


       allergies develop asthma later in life.  This


       asthma, therefore, makes their reactions a little


       bit more severe.


                 Telling somebody right now that they


       reacted at a certain dose and that it is okay to


       ingest doses before that may not be relevant a year


       or five years from now.




                 DR. LUCCIOLI:  This just is a hypothetical


       dose curve adapted from Jonathan Hourihane, who has


       done some nice research in this area, basically



       just to show you how severity and sensitivity may


       factor in.  I don't really want to spend time on






                 DR. LUCCIOLI:  Evaluation of clinical


       responses, this is where interpretation of


       eliciting doses is important with regards to


       subjective versus objective symptoms as well as


       reaction severity in the dose response.


                 This table summarizes some of the


       reactions that you can see from an allergic


       response.  Basically, they are divided into


       subjective versus objective.  "Subjective" means


       that they are reported by the individual or the


       subject, and "objective" are responses that are


       actually visible or observed by the observer.


                 These reactions are reported in this


       manner.  As I said, it is when objective symptoms


       occur, that is when the study is felt to represent


       a positive reaction and stopped.




                 DR. LUCCIOLI:  To just show you some of



       these reactions, not only is there a wide range in


       reactions, but there are some fairly milder


       reactions, hives.  You down here to shock and this


       is anaphylaxis.  Wheezing and syncope are very


       close to systemic reaction and potential


       anaphylaxis.  Consequently, even within an


       objective response, you may have a severe


       anaphylactic response.


                 There are also some subjective reactions


       that may be somewhat severe: throat tightness,


       dizziness, sense of impending doom.  I haven't had


       the pleasure, fortunately, to experience a patient


       with this, but I hear it is fairly dramatic.  They


       have this sense of impending doom and go rapidly


       into anaphylaxis.  It is very, very serious.  It


       doesn't take much for a subjective reaction to go


       to something severe.


                 Also, there are some reactions that kind


       of are in between the line of what is subjective,


       what is objective: fussiness behavior, abdominal


       pain.  In adults, that could be suggestive of a


       nocebo effect.  However, in infants, infants don't



       mess around.  This is their symptom, so these could


       be positive responses for infants.


                 At the same time, you could have skin


       flushing or shortness of breath leading to


       increased respiratory rate, which could be an


       objective sign.  However, many times this could be


       due to also a nocebo effect.  Whether these are


       actual positive reactions is hard to determine.


       There is some clinical interpretation differences


       that can occur here.




                 DR. LUCCIOLI:  Subjective versus objective


       symptoms -- as I told you, the measurable indicator


       of allergic response is the objective symptom.  It


       has got many different endpoints, and the


       interpretation may vary.  This could also be true


       for the subjective reactions.


                 Many times, subjective reactions do occur


       as part of a nocebo effect.  However, there are


       some that are potentially indicative of an allergic




                 How should these be factored into the



       assessment?  Many times they are not recorded in


       the study, so we don't know if there are earlier


       reactions to the objective dose, which may


       represent an earlier adverse event level.




                 DR. LUCCIOLI:  Some other eliciting dose


       considerations, the starting dose is important.  If


       the response occurs at this dose, you cannot


       determine the no-effect level.  Obviously, there is


       no dose below that that doesn't cause an effect,


       but is this starting dose the low-effect level?


       Could you have given a dose a little lower and they


       could have still reacted?


                 With dose increments, some are twofold and


       some are tenfold.  Using tenfold, you may miss some


       increment in between that there could have been a


       reaction, even maybe a fivefold difference.


                 Also, time intervals between doses, as


       Dr. Wood has explained, some doses are delayed.


       However, time intervals, if you don't give enough


       time, you might not know when a subjective response


       has become a subjective response or so forth.  This



       could also affect interpretation of these eliciting




                 Of course, discrete versus cumulative


       dose, some studies report just a discrete dose;


       some the cumulative; some both, which is better.


       However, how do these factor into a true exposure


       assessment or prediction?




                 DR. LUCCIOLI:  I just want to just show


       this, a few more slides, just to kind of put this


       into perspective here, give you a mechanistic view


       that allergy is a unique toxicologic response.


                 When you get food that gets challenged, it


       causes a massive release of mediators and


       cytokines.  This is an amplification system that


       the immune system uses to protect itself.


                 Now, in many cases, this response occurs


       locally and may not amount to very much, but in


       some cases this amplification can involve other


       organs and spread systemically very rapidly.




                 DR. LUCCIOLI:  What has been observed is



       that the severity of an allergic response is on a


       continuum.  You can have subjective responses at


       some point, objective anaphylaxis, and potentially


       death in worse cases.


                 A few points to note is that this is not a


       fixed response.  The early objective system may


       rapidly progress to something worse.  Also, the


       degree of amplification, this is not always


       predictable or reproducible, so symptoms may not


       always be reproducible on subsequent rechallenge.




                 DR. LUCCIOLI:  To end, with the reaction


       severity, most studies only report the actual


       symptom.  You don't know where this symptom is in


       the continuum of severity many times.  Those few


       that do report the symptoms, they report them as


       mild, moderate, and severe.


                 You have to interpret the researchers, I


       guess, response to this, how they interpret it; so,


       there is some interpretation.  Also, when you have


       severe response, like in Dr. Wood's study, in some


       cases a third of individuals reacted and had mild



       reactions, a third had moderate, and a third had


       severe.  How do you factor in those severe


       responses when you determine uncertainty or other




                 Also, potentiating mitigating factors are


       important: anxiety/stress, medications, and so


       forth.  These can either potentiate the reaction or


       stop it.


                 Then, the challenge stops after the first


       response.  A lot of times we don't have the luxury


       of knowing how far or how many more doses would


       have caused a more severe response.  Having that


       information is important when you are wanting to


       make some risk assessment decisions.  Again, it is


       a dose distribution, not a dose response.




                 DR. LUCCIOLI:  In conclusion, the oral


       food challenge does provide data on clinical


       sensitivity to minimal eliciting doses and also


       reaction severity to the initial dose.  However,


       challenge data currently available for


       interpretation is not standardized among studies.



                 The current data pool may not include


       extremely sensitive populations with regards to


       severity.  Challenges have a proven value as a


       diagnostic tool but less value in predicting


       reaction severity to future exposures.


                 Thank you, and I will be glad to answer


       some questions.


                 CHAIRMAN DURST:  Thank you very much.


                 Are there any questions from the




                 We will start here.


                       QUESTION-AND-ANSWER SESSION


                 DR. BRITTAIN:  Yes.  I have a comment on


       the sample size table that you showed us.  I am not


       sure that is incorporating the statistical power


       education we need to have more than these


       individuals.  Are you familiar with what I'm


       talking about like the 29 there?


                 DR. LUCCIOLI:  Yes.


                 DR. BRITTAIN:  I'm wondering if you get


       zero out of 29, then your confidence interval


       excludes --



                 DR. LUCCIOLI:  Well, what that 29 is, that


       is usually a number that is targeted to challenge a


       number of study subjects.  If you show that 29


       individuals with the specific allergy do not


       respond to that ingredient, that gives you 95


       percent confidence that 90 percent will not react.


                 DR. BRITTAIN:  I guess what I'm saying is


       that means if you observed 29, you get the desired


       confidence interval.  However, if you were planning


       a study and you wanted statistical power to be a


       certain amount, you would need to have a bigger




                 DR. LUCCIOLI:  Sure.


                 DR. BRITTAIN:  You couldn't assume that


       nobody would react.


                 DR. LUCCIOLI:  Yes.  Yes, I mean, you saw


       that to be totally assured you would have to test


       quite a few people.


                 DR. BRITTAIN:  I do have another question.


       You mentioned the placebos again, if someone does


       have a reaction with a placebo, how is that


       interpreted in terms of if they also have a





                 DR. LUCCIOLI:  Well, yes, many times you


       don't know, so then you unmask the study and then


       you find out that they reacted to the placebo.


       Now, technically, some studies will rechallenge


       that patient again.  They will have them come back


       just to say, "Well, maybe" -- sometimes people do




                 The difficulty is when they react to the


       active dose, to a real challenge, and to the


       placebo.  If the placebo is too close to the


       active, it may be that by the time you gave the


       placebo, they are still having the active reaction.


                 Basically, if they are rechallenged and


       show again, they are excluded from the analysis.


       Now, that is what should happen.  Unfortunately,


       you never get that information a lot of times from


       these challenges.


                 CHAIRMAN DURST:  Suzanne.


                 DR. TEUBER:  One of the aspects that we


       are all very concerned about is which threshold to


       use and when it may cause a subjective reaction. 



       Actually, oral itching is a very important


       subjective reaction that you didn't have on your


       table up there in this presentation.


                 DR. LUCCIOLI:  Okay.


                 DR. TEUBER:  However, if that is


       reproducible with two active challenges and not


       seen with two placebos, which I think Dr. Taylor


       may address a little bit later, but some of the


       studies that Dr. Wensing and Bindslev-Jensen and


       Dr. Hefle have been doing, they have been looking


       at that.  All of these have been followed by


       objective reactions at higher doses.


                 DR. LUCCIOLI:  Yes.


                 DR. TEUBER:  I would really like people to


       comment on that because this may be a much safer


       way to approach obtaining thresholds to get these


       extremely sensitive populations, if we can use


       reproducible subjective data knowing, too, that


       there are those other factors that may affect it.


                 DR. LUCCIOLI:  Sure.


                 DR. TEUBER:  For instance, in these


       threshold studies that are being designed



       specifically for thresholds, people with unstable


       asthma would still be excluded.  I am curious if


       anybody knows anything about how unstable asthma


       would affect the threshold for a LOAEL that is


       seen?  Is it a lawful difference?  I mean, is there


       any anecdotal experience with how that might


       change?  We want safety here.


                 DR. LUCCIOLI:  Obviously, a speaker that


       is coming after me would have some information on


       that, but some information from Jonathan Hourihane


       would suggest -- and I think some European studies


       actually do test some severe patients.  Now, I


       don't think that any of these patients are




                 I think that they are all excluding


       patients who have unstable asthma, but with asthma


       in general they haven't found that these


       individuals have a lower minimal eliciting dose


       than other individuals.  However, when they do get


       a reaction, they can have a much more severe




                 The assumption, though, is that because of



       the fatalities and other things that when their


       asthma becomes unstable their sensitivity could


       change and become more severe very quickly.


                 CHAIRMAN DURST:  Marc.


                 DR. SILVERSTEIN:  Marc Silverstein.  I


       have two comments that deal with sort of


       clarification of terminology and one comment that I


       think deals with a more difficult issue.  I thought


       this was a wonderfully helpful and concise summary


       of a variety of complex factors.


                 In terms of the two clarifications, in


       clinical medicine from the first days of medical


       school we are taught the difference between


       "symptoms," which are subjective, and "signs,"


       which are objective.


                 Some of us from the clinical side who will


       be reading the report will think that subjective


       symptom is redundant and objective symptom is an




                 To help the wide dissemination of the


       report and presentation, I would like to suggest


       that we in our thinking we may say "subjective



       finding such as symptoms of the disease" and


       "objective signs" is the sense that I use that.


                 DR. LUCCIOLI:  Sure.


                 DR. SILVERSTEIN:  I think it is helpful


       because there will be a variety of readers of the


       report who may not appreciate that on the clinical


       side there is a clarification about that.


                 The second clarification had to do with


       the incidence versus prevalence in the sample size


       table.  What we are talking about is the proportion


       of subjects being tested to the proportion of


       individuals in a population, so that sample size


       table or the table we have is the expected number


       of sample you would need.  You label it "incidence"


       but it is really a "prevalence" of a condition in a




                 DR. LUCCIOLI:  Okay.


                 DR. SILVERSTEIN:  I believe that what you


       are getting at is the sample size so that the lower


       confidence interval is that 10 percent or 1 percent


       rather than the sample size necessary to show that


       two populations differ in proportion or the sample



       size to show how tight you are around a rate of


       zero, which would be a different population.


                 DR. BRITTAIN:  Can I respond to that?


                 DR. SILVERSTEIN:  Sure.


                 DR. BRITTAIN:  I don't think when you are


       designing a study you want to think of it in terms


       of statistical power, which would be greater than


       the sample sizes.


                 DR. SILVERSTEIN:  The third comment I


       have, which is substantive and I think we may need


       to address this later in greater detail, has to do


       with sources of error.  There are two sort of


       classes of errors that we made in our inferences.


                 One types of error an epidemiologist or a


       clinical epidemiologist may say is biased, one of


       the most common sorts of types of errors we could


       make would be making inferences in the presence of


       certain biases.  The most common of which would be


       selection bias.


                 Of course, the selection of individuals


       who are referred to a physician, who are referred


       to an allergist, who are selected for an oral



       challenge, food challenge, study would potentially


       lead to erroneous inferences if there were


       non-representative selection.


                 That is something that in reading the


       literature and making decisions about inferences


       for studies or for policy I think we need to be


       aware of up front, so that is an important class of


       errors that we need to be alerted to.


                 The second would be an epidemiologist


       would talk about confounding.  In your example of a


       study subject who has asthma, whether it is stable


       or unstable and how that is defined, asthma might


       be considered an extraneous factor that affects the


       relationship between the allergen and the response


       to the test.  We could use the framework of


       thinking of it as a confounding benefit.


                 Factors such as bias and factors such as


       confounding, I think, are useful as we make


       decisions about the report and the evidence for


       that.  I would like for us to be alert to that as


       we think about the presentations.


                 DR. LUCCIOLI:  Thank you for the





                 CHAIRMAN DURST:  Are there any other


       Committee comments?


                 (No verbal response.)


                 CHAIRMAN DURST:  If not, thank you very




                 Our next speaker is Dr. Rene Crevel,


       senior scientist at Unilever, Safety and


       Environmental Assurance Centre in the United


       Kingdom.  He will be speaking on the "Threshold


       Modeling Approach."


                       THRESHOLD MODELING APPROACH


                 DR. CREVEL:  Well, first of all, thank you


       for inviting me to share some thoughts on the work


       we have been doing on modeling thresholds.




                 DR. CREVEL:  You have asked me to talk


       about the following, to look at different modeling


       approaches including what is named the


       "hyperallergen."  This is the model, and the


       Bindslev-Jensen, et al., allergen model; talk about


       the data requirements and underlying assumptions



       behind them; and then say something about


       interpreting the results of applying these models.




                 DR. CREVEL:  Now, just to take a step back


       and think about why we are doing this, we've got a


       challenge in allergen risk management as far as


       industry certainly is concerned.


                 We want to protect allergic consumers of


       course, but we also are aware that protecting them


       by certain measures of risk management such as we


       have heard about this morning like precautionary


       advisory labeling does actually affect their


       quality of life.


                 We want to minimize the effects on their


       quality of life, the adverse effects on their


       quality of life.  We ultimately also want to


       maintain economic operation of food manufacturing,


       because if that doesn't happen, then that will


       affect the quality of life of a considerable number


       of individuals and people throughout society as


       well.  It is an important point to bear in mind.





                 DR. CREVEL:  How can we meet the


       challenge?  Well, first of all, of course we could


       label where the allergen is present, and that is


       fine.  You have legislation over here now in the


       U.S. for that; we have legislation in Europe; and


       many other regions and nations also have






                 DR. CREVEL:  Or, we can ensure that the


       residual allergen content of product is low enough


       to be harmless.  I put in brackets here (to the


       vast majority of allergic consumers), because we


       have heard here this morning some instances of


       people reacting to extremely low amounts.


                 I think it is questionable, and I think


       the Committee must address that particular


       question, whether those people can be protected by


       whatever we can do in the food industry.  We need


       to think about what alternative measures may need


       to be done, whether they can ever eat the sort of


       foods we can produce.





                 DR. CREVEL:  How can we determine what is


       harmless to an allergic consumer?  Well, we have


       several sources of data, which I have listed on


       this particular table.


                 We can look at case reports from the


       literature, and those we have heard.  We have heard


       about people reacting to very low amounts.


       Unfortunately, the usefulness in risk assessment,


       actually an allergen risk assessment in my view is


       actually quite limited.


                 They do establish the hazard.  Yes, they


       tell you that a certain amount will affect some


       individuals, will provoke a reaction in some


       individuals.  They don't tell you in how many


       individuals that will happen.


                 We can use control challenge studies.  In


       fact, those of course provide the bedrock of what


       is needed, the information needed in allergen risk


       assessment because the population can actually be


       quite accurately describe.


                 You can describe them in terms of the



       symptoms they have experienced, the allergological



       history, or the medical history as appropriate --


       all of the demographics that you can think about.


                 Finally, dose distribution modeling, which


       I am going to spend obviously some time on, also is


       very useful in allergen risk assessment.  But of


       course it relies on the experimental clinical data


       which is generated in control challenge studies.


       It cannot operate in a vacuum.  We do not have


       enough of those sorts of data at the moment.




                 DR. CREVEL:  I have been asked to say


       something about the hypoallergenicity approach.  As


       I understand it, it is an unofficial standard for


       designating infant formula as hypoallergenic.


                 The original reference I found goes back


       to 1991, although I think the American Academy of


       Pediatrics has actually updated or at least issued


       the guidance more recently, I think, in 2000 or


       something of that sort.


                 The statistics of this approach are based


       on the binomial theorem, quite simply.  This shows


       that, for instance, if you have a study with 29



       participants, as we have already heard, and you


       observe no reactions, then you can be 95 percent


       confident that only 90 percent of the population


       from which these people have been taken will not




                 You can also extend that a bit, so if you


       observed one reaction and you added people to the


       study, then you would 46 for the same degree of




                 If you wanted 95 percent confidence then


       fewer than 99 percent would not react, then you


       have got those other numbers which already become


       very challenging, pardon the pun, for a challenge


       study both in terms of recruiting the people do it,


       to participate, and the economic cost of actually


       doing it.




                 DR. CREVEL:  However, those are very


       useful data when they are generated, but protecting


       90 or 95 percent or even 99 percent of the allergic


       population is not sufficient as far as we are



       concerned as an objective for the food industry.



                 What we asked ourselves is, How can we


       improve this?  There are several ways.  We could


       try to look at the conventional toxicological


       approach and apply a safety factor to the lowest


       observed adverse effect level or the no observed


       adverse effect level as the case may be, if you've


       got the no observed adverse effect level.


                 However, that particular safety factor, I


       would say, would be arbitrary because we don't


       actually know enough about interindividual


       differences within the allergic population to apply


       a science-based factor, I think.  The level of


       protection still is undefined.  You do not know how


       many people you are protecting by applying that


       particular safety factor.


                 What about modeling of those distribution


       of minimum eliciting doses?  Well, that can


       actually define the level of protection for


       individual allergen level.  You can actually use a


       safety factor there.  You can use something which


       is like a lower 95 percent confidence interval


       instead of using the figure itself.





                 DR. CREVEL:  Does modeling actually work?


       We asked ourselves could we fit a curve to the


       distribution of minimum eliciting doses that are


       generated by challenge studies, and could that


       curve be useful to predict the number of reactions


       likely to occur as a result of exposure to a


       specified amount of inadvertently present allergen


       in the food?


                 What I have to say, of course, is we are


       not so much concerned about "declared allergen,"


       people who are allergic can avoid that, but what


       our concern is about is that which is present by


       cross-contact, mainly inadvertently.




                 DR. CREVEL:  This just gives a very quick


       model curve.  It is just used to illustrate some of


       the points, right, okay.  From this particular


       curve, okay, we have got the data points


       schematically indicated like this.  The dose on


       this (indicating) particular axis an the proportion


       of the study population reacting here.  Obviously,



       it goes up to 100 percent.


                 Then, you have these particular points,


       which I have named "ED                                                  

                50" here.  This would be the


       dose expected to provoke a reaction in 50 percent


       of the study population or 10 percent. This


       particular one is an extrapolation, one way of


       extrapolating, which one could use.




                 DR. CREVEL:  What is the impact of the


       choice of model on the predicted minimum eliciting


       doses?  I should go back a bit actually and say


       something else.


                 We collaborated actually with


       Dr. Bindslev-Jensen of Denmark in the initial


       development of the model.  At that particular point


       we used the lognormal distribution.  Having the


       papers published and so on, after that we decided


       to go back and look at a few more parameters and


       try to refine this particular approach.


                 Good clinical data were available for egg,


       milk and peanut.  We fitted the data using the


       following statistical distributions and calculated



       ED10s, the dose which would be predicted to cause


       responses in 10 percent of the population; and


       ED1s, in 1 percent of the population for each of


       those.  We used the following linear extrapolation


       from lowest observed adverse effect level to zero


       dose, which I showed you that was the red line; the


       lognormal model, which was the original one in the


       2002 paper; the Weibull model; and the loglogistic






                 DR. CREVEL:  What I want to do is to


       illustrate how these variously fit using the


       different distributions.  This is using real data


       actually from the study by Wensing, et al., in 2002


       on roasted peanuts.  You have got the data points


       here.  That is still a normal fit, which is the


       original one we used.


                 You can fit loglogistic pretty well as


       well and the Weibull as well.  You can even fit a


       linear -- you can even correlate these points


       linearly as well.


                 Although I haven't got the parameter fits



       here, I can tell you that they are all pretty good


       for all of those.  Basically, the fit which you use


       doesn't actually tell you which is the most


       appropriate one for the particular distribution.




                 DR. CREVEL:  This is illustrated in the


       differences between ED10s and ED1s between studies


       and models.  Now, this is just using data on peanut


       actually from -- in this particular case, we are


       just comparing the number of studies on peanuts


       including studies performed by Bock and May in the


       1970s and the later study by Wensing.


                 What is quite clear actually is for ED10s,


       which are still within the experimental zone, what


       I call the "experimental zone" in one of the


       previous slides.


                 The data actually drives what the


       predictions are.  I mean, there is not a lot of


       difference between the ED10s, even though it is log


       scale, I know, even in this particular case.




                 DR. CREVEL:  When you move away and go



       outside of the experimental zone to the ED1 and


       even further actually, the case is even stronger


       beyond that, the actual choice of model starts to


       drive the predictive responses.  That is an


       important point to bear in mind.


                 This is summarized on this slide for the


       ED10s in the experimental zone, that the


       differences between studies is greater than between


       models.  In order to address that, the best way of


       doing that is to focus on standardizing protocols


       and having consistent patient selection criteria


       for the studies which you wish to undertake.


                 For the ED1 values, the differences


       between models are much larger as I showed and


       increase of course as we move further away from the


       experimental zone.


                 What this actually illustrates is that you


       need to validate the particular approach.  You need


       to validate whatever model you have chosen and


       adjust parameters in accordance with that.  What


       I'm going to talk about is actually how we might go


       about doing that, what sort of data is needed.





                 DR. CREVEL:  Now, there are a certain


       number of assumptions underlying the values


       generated by the model.  We are talking here about


       undeclared allergen, and that is quite important in


       relation to the one of these particular points.


                 First of all, we assume that the


       participants in a controlled challenge study are a


       representative sample of the whole allergic


       population.  That is a very important assumption,


       and one which actually is sometimes shall we say




                 We have heard a lot about whether people


       are included or not included in particular studies.


       I would tend to argue actually that the population


       used in challenge studies, because of the way they


       are selected, is actually shall we say at the more


       severe, more sensitive end of the allergic


       population, basically because there are people who


       actually normally are referred to tertiary care




                 There are people whose allergies are



       actually troubling them.  They are not people who


       might just get a small rash and just ignore it or


       ignore the particular food that caused it.  There


       are people who actually need to manage their


       allergy and they need some serious advice in doing




                 The second point is actually in terms of


       validating the model the allergic people actually


       eat the same foods as the non-allergics.  In this


       particular case, it is quite important because if


       they are already avoiding them, the number of


       reactions that you will be able to enumerate in


       epidemiologic studies will not be the correct one.


                 The distribution of allergic reactivity


       study at the population level, now we've heard


       thresholds for individuals.  Minimal eliciting


       doses for individuals do vary.  However, what we


       are saying here actually is that overall it will be


       studied in these particular challenge studies.


                 Finally, the responses to a given dose of


       allergen are similar in the clinic to those


       experienced outside.  We are doing some work



       actually with Jonathan Hourihane, in fact, to try


       to quantitate the differences that may exist


       because, in fact, we are very aware that particular


       assumption probably does not hold entirely.




                 DR. CREVEL:  Okay.  What data do we


       require for validation and application of a


       modeling approach?  We want to arrive at the risk


       assessment.  We have the hazard characterization.


       I would put it to you that actually what we are


       doing by the modeling approach is actually


       characterizing the hazard.  We are establishing how


       many people are likely to respond to a particular


       amount, and we use all of these.  These particular


       factors all influence it.


                 However, we also need to know the number


       of allergic consumers.  That is quite important in


       terms of prioritizing allergen management and so


       on.  Effectively, what the legislation does is also


       to acknowledge that particular fact.


                 The legislation either here, in Europe or


       anywhere else does not protect everybody because of



       course it only specifies a certain number of major


       common allergies rather than all of the 200 or so


       foods that may provoke allergic reactions.


                 We also need to know what the exposure is.


       We need to know what residual allergen levels are


       in the foods, residual allergen levels that are not


       declared.  Finally, we also need to know what the


       number of reactions is overall in the community.


                 Taking all of that together, we can


       actually validate the model.  Using those sorts of


       data, we can also apply it properly.




                 DR. CREVEL:  To summarize, I think the


       modeling approach complements clinical studies and


       it certainly compliments clinical studies to


       establish minimal eliciting doses.  Of course, it


       relies on the data generated in those studies.


                 I think the advantage compared to just


       using the data as such is it actually permits more


       complete use of those data using the whole dose


       distribution rather than just one particular point,


       say, the lowest observed adverse effect level or



       the no observed adverse effect level.


                 It also, I think, makes the whole process


       of risk management more transparent, I guess you


       would say, allowing a more informed discussion of


       risk management objectives by all stakeholders.


       That is very important I think.


                 In order to agree on objectives, I think


       people need to know how or need to see the process


       by which they are reached.  However, and this is a


       big proviso, it does require validation before it


       can be fully operational.


                 We are doing work at the moment to see how


       we can address that.  Some of the data actually I


       should say will contribute to this particular


       assessment will be generated by some European


       projects which are currently running, but of course


       it will take a few years to get there.


                 That was my last slide.  Thank you.


                 CHAIRMAN DURST:  All right.  It is open


       for discussion.




                 DR. BRITTAIN:  That was a really



       interesting talk.  There was one aspect of it that


       I'm a little --


                 DR. CREVEL:  I'm sorry?  I can't hear you.


                 DR. BRITTAIN:  There is one aspect of it


       that I'm a little confused by, and that was in one


       of your last slides with all the graphics about the


       needing to know the number of allergic consumers.


                 If you are trying to find the dose at


       which the risk of a reaction, given you are


       allergic, which is what I thought we were trying to


       do, why do you need to know the number of allergic




                 DR. CREVEL:  Well, you need to know the


       prevalence of the condition within the population.


       In fact, perhaps the confusion is there isn't,


       because I mentioned validation as well as


       prediction in this particular context actually.


                 For validation, you certainly need to know


       how many reactions are occurring in order to see


       whether the model actually predicts the numbers of


       reaction which you are actually observing.


                 I mean, this is a big data gap at the



       moment.  I mean, I don't think either in the U.S.


       and certainly not in Europe do we have data on


       actually the number of reactions that do occur.


       Certainly, we do not have any information on the


       total number of severe reactions or less severe




                 DR. BRITTAIN:  You mean the number of


       reactions that occur across a population as opposed


       to your study?


                 DR. CREVEL:  Yes.  No, across a


       population, sorry.  Sorry, that was in the


       population, sorry, yes.


                 CHAIRMAN DURST:  Any further discussion or




                 (No verbal response.)


                 CHAIRMAN DURST:  If not, thank you very


       much, Dr. Crevel.


                 Our final speaker for this morning's


       session is Dr. Steve Taylor.  He is the Maxcy


       distinguished professor and director of the Food


       Allergy Research and Resource Program at the


       University of Nebraska, who will discuss Food



       Allergen Thresholds.


                         FOOD ALLERGEN THRESHOLDS


                 DR. TAYLOR:  Well, I would like to thank


       the Food and Drug Administration for giving me the


       opportunity to make a presentation to this panel.


       There are advantages and disadvantages to being the


       last speaker of the morning.  Much of what you are


       going to see on my slides may just be a reemphasis


       of some things that have already been said.


                 I think I got a rather difficult topic,


       also by being the last one on the agenda, because


       I'm supposed to talk about uncertainty factors,


       what are uncertainty factors and how are they


       derived and what is the underlying scientific


       rationale for such a factor.  I only wish I thought


       I knew the definitive answers to all of those






                 DR. TAYLOR:  I think the National Academy


       of Sciences outlined risk assessment approaches a


       number of years ago, and I always like to start


       with this slide, even though I'm not going to



       discuss all of these different points, because I


       think that the same assessment can be used for food


       allergens as is used for pesticide residues and


       food additives and other things.  This is a very


       robust risk assessment approach.





                 DR. TAYLOR:  I am only going to focus on a


       few things on this slide today, and one is


       dose/response evaluation.  I have been thinking


       about this issue for probably 30 years.


                 This is one of the earliest slides that I


       created.  At that point in time we didn't know very


       much, and I would argue we only know a little bit


       more now than we knew when I wrote this slide a


       long time ago.


                 Trace amounts can elicit reactions.  I


       would argue that the severity of the response is


       directly related to the dose.  The higher the dose,


       the more severe the response.


                 I would agree that individuals can have


       different responses on different days to the same



       dose.  However, I don't think those responses are


       as dramatically different, or at least I would say


       that is an unproven point regarding some of the


       things that have been said this morning.


                 There are a lot of assumptions that are


       made in this field, and I think as a panel you need


       to identify all of the assumptions and question




                 Stefano Luccioli made a good point, that


       individuals vary widely in their degree of


       sensitivity in these controlled challenge studies a


       millionfold.  I completely agree with that.  That


       is kind of amazing in itself.


                 The big question is, How much is too much?


       The food industry has been focusing on trying to


       get an answer to this question for a long time for


       some of the reasons that Dr. Crevel just pointed






                 DR. TAYLOR:  I think there is another part


       that we haven't heard quite enough about, and Rene


       kind of pointed it out in his presentation.  It is



       the exposure assessment piece of the equation.


                 How frequently are food products


       contaminated with potentially hazardous levels of


       unlabeled allergens, and how frequently do


       food-allergic consumers suffer reactions?  We


       really don't know that part very well.  Only


       recently, as Dr. Hefle pointed out, do we have the


       methodology necessary to determine with any degree


       of confidence how frequently food products might be


       contaminated and at what levels.




                 DR. TAYLOR:  Gil Houben from TNO [The

 Netherlands Organization] prepared


       this slide, and I always like to steal good slides


       from speakers that I invite to be on programs.  I


       think this kind of pictorially describes the


       situation that exists.


                 We have food products in the marketplace


       that contained for one reason or another some level


       of undeclared allergen.  This may be from


       cross-contact, this may be from use of ingredients


       derived from commonly allergenic foods that are


       processing aids and historically haven't been



       labeled in most countries.


                 Then, we have individual thresholds for


       clinical response that varied by a millionfold as


       Dr. Luccioli pointed out.  There is an intersection


       here between products that have enough undeclared


       allergens that at least the most sensitive


       individuals have some probability of reacting to




                 If I was going to draw this slide myself,


       I would lengthen the tail of this curve because we


       know from analytical studies that there are


       products in the marketplace that are quite


       hazardous for these individuals containing


       comparatively higher levels of allergens that


       provoke severe reactions.  Dr. Hefle showed some of


       those data today.




                 DR. TAYLOR:  I wanted to say just a little


       bit about the different kinds of clues that we can


       have for determining allergen thresholds.  Stefano


       already pointed this out, too.


                 Probably the best data we have comes from



       double-blind, placebo-controlled food challenges or


       clinical threshold experiments using double-blind,


       placebo-controlled food challenges and


       immunotherapy trials that also use challenge data.




                 DR. TAYLOR:  I actually don't think that


       allergen cross-contact episodes turn out to be very


       useful in determining thresholds, and I wanted to


       emphasize that point, because there is a lot of


       anecdotal material in the clinical literature about


       these cross-contact episodes.


                 A lot of them are deficient, because the


       analytical methods used to detect the residues in


       those studies were probably not as accurate as the


       methods that Dr. Hefle described in her


       presentation, the methods that we have had for the


       last few years.  There is often a lot of lacking


       information in the investigation of these studies.




                 DR. TAYLOR:  As I pointed out, this


       question of how much is too much has intrigued out


       group for a long time in the food allergy research



       and resource program.


                 I want to point out that we are funded by


       the food industry.  We have more than 40-member


       companies scattered around the world.  We began to


       focus on the threshold question in earnest in the


       mid-1990s and beyond.




                 DR. TAYLOR:  We have held a series of


       threshold conferences.  The first one was held in


       1999.  I was asked to say a little bit about these,


       and it is really hard to summarize it in 15 minutes


       or less.


                 I will point out the fact that the results


       of the First Threshold Conference have largely been


       published in the peer reviewed, scientific




                 The question we asked at the First


       Threshold Conference is we invited a number of


       clinicians from around the world to come to South


       Carolina, because we thought that perhaps they had


       information on low-dose challenge trials.


                 When you hear studies of the kind that



       Dr. Wood reported this morning, recognize that most


       diagnostic challenges start at 400 to 500




                 No wonder some people have severe


       reactions at those dose levels, because those are


       quite high in my opinion.  We were interested in


       clinicians who sometimes, because of the patient's


       history, started the challenge at a much lower






                 DR. TAYLOR:  What did we find out?  We


       found out that there was considerable data on


       low-dose challenges for peanut, egg and milk in


       particular and more scattered data for some of the


       other foods.


                 The data were really hard to evaluate


       because of the lack of standardized protocols.  I


       will come back to that in a little bit.  The lowest


       provoking dose -- we had 306 patients for peanut,


       281 patients for egg, and 299 for milk.  These


       physicians brought this data to this conference.





                 DR. TAYLOR:  The lowest provoking dose for


       peanut was about 1 milligram of peanut, which is


       .25 milligrams of peanut protein.  However, I have


       to tell you that Dr. Hefle and I spent an entire


       weekend in the conference room trying to figure out


       what the doses were in these challenge trials,


       because the physicians don't calculate that,


       particularly carefully in some cases.


                 Our personal favorite is the physician


       that used a drop of peanut butter as his lowest


       dose.  We had him send us his dropper bottle and we


       tried to figure out how much that actually was.


       These data look really finite when you show them


       this way, but there is a lot of glorified


       guesswork.  I just want you to understand that.




                 DR. TAYLOR:  We determined that minimal


       eliciting doses or threshold doses do exist for


       commonly allergenic foods, that the threshold doses


       are finite, measurable and above zero.


                 However, it was really difficult to reach


       consensus, and we didn't reach consensus.  We had



       about 20 clinicians at this conference, and we did


       not reach consensus on what threshold doses should




                 In fact, for most of them this was their


       first introduction to this concept.  We had to


       teach them what NOAELs and LOAELs were.  They make


       risk assessments every day but not these kind.




                 DR. TAYLOR:  We also found that reactions


       occur to hidden or undeclared allergens in foods.


       No big surprise there.  However, severe reactions


       to undeclared allergens tended to occur at higher


       dose levels.


                 We also determined that at least in these


       populations with these low-challenge doses that


       low- or very low-dose exposures, LOAELs, result in


       mild reversible symptoms.




                 DR. TAYLOR:  The Second Threshold


       Conference was held in 2002 and was geared to


       address the biggest concern we had from the first


       one, and that was a lack of a consensus protocol.





                 DR. TAYLOR:  I don't have time to describe


       the consensus protocol other than to indicate that


       it has been published; it does exist; and there are


       ongoing low-dose challenge trials underway around


       the world using this protocol or slight variations


       of it.


                 As Dr. Luccioli pointed out, most of those


       haven't been published yet because it takes a year


       to two years to do these studies to find the number


       of subjects to enroll in these studies.




                 DR. TAYLOR:  We did have the Third


       Threshold Conference where we tried to determine


       what you do with the data once you collect it.




                 DR. TAYLOR:  I won't go into that very


       much, because much of it relates around the


       modeling stuff that Rene Crevel already described.


       Because the binomial approaches are just plain


       difficult, because it is very difficult to identify


       even 29 soybean-allergic individuals in the world



       to do a challenge trial.  Believe me, we've been


       there, and we know how hard it is.


                 It is easier to do peanut trials than


       perhaps others.  It is hard to do milk and egg


       because young children outgrow their allergies, so


       you've got to be concerned that the child, the


       patient, still has the allergy that you are looking






                 DR. TAYLOR:  There were a number of


       advantages to modeling.  I think Rene pointed those


       out.  I will just make the point that the consensus


       of the group was that you could do modeling.  Of


       course, you've got to figure out which model you


       are going to use.


                 Maybe we haven't validated them yet so we


       don't exactly know; however, using this lower


       confidence interval as the threshold might be a


       reasonable approach to consider.




                 DR. TAYLOR:  Well, classical risk


       assessment involves determining the NOAEL for a



       food additive or a pesticide residue or something


       like that and dividing it by 100.


                 Classically, tenfold is for extrapolation


       from animals to humans, and tenfold is for


       intraindividual variation.  Consequently, what


       uncertainty factor should we use?




                 DR. TAYLOR:  For allergens, since you have


       human subjects that can be used, the ideal thing


       would be to determine the no observed adverse


       effect level for specific allergenic foods among a


       human population that is allergic to that food, and


       then apply an uncertainty factor to get your


       threshold dose.




                 DR. TAYLOR:  To do that with any degree of


       confidence, you have to challenge a fairly large


       number of allergic individuals.  You would have to


       identify the NOAEL for each patient.


                 You would probably also have to identify


       the LOAEL for each patient to prove the person is


       still allergic to the food that is under





                 It would be good to determine the


       variation between individuals in NOAELs because it


       is probably a millionfold.  A standardized protocol


       would be handy so that you didn't have uncertainty


       about the differences in protocols.




                 DR. TAYLOR:  There is no animal to human


       extrapolation needed for food allergy


       considerations because we have human data.  We have


       already selected a sensitive subpopulation of the


       human population.


                 The question arises, Did we include the


       most sensitive individual?  I think that is an


       important consideration for this panel.  We have


       heard several speakers say, "Well, maybe we have




                 My argument is that perhaps in terms of


       representing the whole allergic population to a


       particular food we have actually excluded the other


       end of the dose distribution curve, and we actually


       have included a number of people from the most



       sensitive subpopulation.




                 DR. TAYLOR:  I want to point to this study


       again.  People have interpreted this study as a


       publication involving the dose distribution for the


       whole food-allergic group allergic to peanuts,


       eggs, and milk.  It is not that; it is a study of


       the most sensitive individuals in clinical







                 DR. TAYLOR:  Well, how much data is out


       there?  Is there enough data to make your


       decisions?  I think there can be if you can wrestle


       with the uncertainty factors and the differences in


       protocols from one study to another.


                 I just went through what I think is the


       most relevant literature.  Some of these are in


       your "FDA Report," which contains a big table at


       the back that somebody very laboriously put


       together.  I think they actually found most of the


       relevant studies.


                 I congratulate them for that, because that



       is not particularly easy to do.  I went through


       those studies and added up the number of patients


       that are in each one of these studies that were


       subjected to double-blind, placebo-controlled food


       challenges and for which a published LOAEL exists.


                 Now, there are lots of differences in


       protocols, so there are uncertainty factors with


       how to plug this data into one of Rene's curves.


       What you can see is there are lots of subjects.


       This is for peanut.  Note, I put an asterisk by our


       2002 paper, and that is because that is not


       original data.  Some of those patients may also


       appear in some of the other studies.  We got


       concerned about whether to count them twice.




                 DR. TAYLOR:  This is for egg.




                 DR. TAYLOR:  This is for milk.




                 DR. TAYLOR:  We have got a lot of data


       points.  What are the uncertainty factors?  Well,


       you've got adults, adolescents, children, infants. 



       Many of the studies have been done on pediatric


       populations; fewer studies have been done on


       adults.  You can do challenge trials on both of


       those.  A lot of the diagnostic challenge trials


       are done on infants, but they are not done in


       threshold study types of experiments.




                 DR. TAYLOR:  You've got the problem with


       the nature of the challenge material and the


       allergen content of that challenge material.  This


       is again from our 2002 study from Threshold 1.


                 You can see the number of different


       materials: ground peanut, peanut flour, peanut


       butter, egg white, dried egg white, whole egg,


       dried whole egg, and raw versus cooked for most all


       of those.  Then, you've got whole milk, non-fat dry


       milk and even infant formula as the milk challenge




                 In many of these cases, the physicians


       didn't determine the protein content of the


       challenge materials, so you've got to make


       glorified guesses.  There are uncertainties about



       the challenge materials.




                 DR. TAYLOR:  I would argue that studies


       should be compared using protein content.  This


       failure to provide that data makes the evaluations


       really difficult.  If the protein content of the


       challenge material was not determined or cannot be


       determined using reliable data in the literature,


       then the study probably has to be rejected from


       consideration by groups like this.


                 There are well-characterized challenge


       materials like non-fat dry milk, dried egg white


       and soy flour that I think you can assume what the


       protein level is based on standardized industry


       data.  Thresholds should be established in terms


       that can be related to analytical methods like


       milligrams of food or milligrams of food protein.




                 DR. TAYLOR:  There are also issues related


       to blinding that Stefano already talked about.


       Some clinicians use labial challenges.  They put a


       drop of the food on the patient's tongue or lip. 



       That is often used for young infants.


                 I think that is particularly difficult to


       interpret what the dose was.  However,


       diagnostically it is good procedure, but otherwise


       it is kind of difficult to figure out what was


       going on.  Then, there is the choice of dosages


       used for the challenges.


                 Probably the biggest uncertainty is this


       issue of the fact that most of the publications


       were done for diagnostic purposes, and so when you


       look at the published literature you get the LOAEL


       and not the NOAEL.  I actually think a lot of the


       NOAELs are clinically available; they are just not




                 There is more uncertainty in using a LOAEL


       rather than a NOAEL to established threshold doses;


       there is patient selection criteria; exclusion of


       people on probably both ends of the curve; and


       there is variability in individual threshold doses.


                 Diagnostic challenges tend to report only


       the LOAELs; the NOAELs in some cases may not be


       recorded.  As Dr. Wood pointed out to us in his



       study, in many cases the patient reacted to the


       lowest dose administered.


                 However, if it was 400 or 500 milligrams,


       that is a pretty sizeable dose.  I think there is a


       lot of uncertainty if you use that as your LOAEL to


       try to try and figure out what the threshold dose


       might be.  You are much better off to focus on


       lower dose challenges where there is less


       uncertainty even if you have to use the LOAEL.


                 How far above the NOAEL is the LOAEL; and


       if using a LOAEL, how big should the UF be?  I


       think that is the question that they wanted me to


       try to answer.




                 DR. TAYLOR:  I got bold and I did try to


       answer it.  If the LOAEL is based on subjective


       symptoms, "My mouth itches," then I don't think we


       have to be very concerned about uncertainty


       factors, because in the limited experiences that


       exist in the literature there is a ten- to a


       hundredfold variations between the doses that begin


       to provoke subjective symptoms and the doses that



       tend to provoke objective signs.  I learned


       something today.  I'll try to say "signs."


                 Now then, if you have a LOAEL based on


       objective reactions, what uncertainty factors


       should you use?  Well, then I think you would need


       to have very careful expert analysis of the


       clinical study you are looking at.


                 I could argue that if you looked at one of


       these clinical threshold trials that have been done


       using microgram and low-milligram dose level, that


       you could use a very low uncertainty factor.


                 However, if you are going to rely on


       publications like Perry, et al., from 2004 where


       you only have data on the reactions that occur at


       the lowest diagnostic challenge dose and it is 400


       or 500 milligrams, then you've got a much bigger


       uncertainty factor because you could be a long ways


       above the NOAEL.


                 I don't know what you would do in this


       particular case, so I put a question mark by it.  I


       actually think those kinds of studies are not very