P R O C E E D I N G S
Greeting and Housekeeping
CHEESEMAN: Good morning. I'm Mitchell Cheeseman. I am the Director of the Division of Food Contact Substance Notification Review at FDA. I just want to start by giving you a little bit of housekeeping information.
We have question cards, if you have not already received some of those. They are multicolored question cards and you can get another set down front. There is a box to place the completed cards in, if you have any questions. They are color-coded. Administrative is white. Environmental is yellow. Chemistry is pink or red and toxicology is blue. You will help us sort through the questions at lunch if you follow that scheme.
We will try to get to those questions in the afternoon. There will be questions and answers for each of the speakers speaking about guidance, also, so there should be plenty of time for discussion and feedback.
Just to go through the agenda. The morning we will spend talking about mostly administrative procedures and processes of the program followed by discussion of the guidance on each of the technical areas of chemistry, toxicology, and environmental.
There will be a morning break at 10:30, an afternoon break at 2:45 and a lunch break from 12:15 to 1:30.
In the afternoon session, we will have three workshops, one on environmental, one on chemistry matters and one on toxicology matters that will involve discussion of some more detailed topics including exposure estimation, development of environmental assessments and some new information on structure activity analysis in our toxicology review. Then we will end sometime before 5:00 with wrap-up questions.
I believe that is all I wanted to say at the beginning here so I will turn the podium over to Dr. Alan Rulis, the Director of the Office of Food Additive Safety for a few introductory remarks.
Thank you.
Statement from the Director
RULIS: Thank you, Mitch. Good morning. I am Alan Rulis. I am the Director of the Office of Food Additive Safety in the Center for Food Safety and Applied Nutrition at FDA. I would like to extend a warm welcome to everyone present including notifiers, possible notifiers, representatives of notifiers, consumer group reps and others in the room as well as FDAers who are here.
When we look back to 1997, when Congress passed the Food and Drug Administration Modernization Act, that was really a landmark period of time for us. It established, in statute, a new process for dealing with food-contact substances. In fact, it defined them as an entity and put in place a procedure for premarket notification of the use of those materials.
Of course, that statutory provision did not become active until Congress appropriated funds, which it did for Fiscal Year 2000. So, on October 1 of 1999, we had available earmarked funds for executing the program. As a result of that, we were able to start in January of 2000, January 18 of 2000, to receive, for the first time, food-contact notifications. Since then, we have been processing them.
Most recently, of course, you are aware that we have published a final rule--it became effective in June of this year--of governing the processes, the food-contact notification process, the submission-and-review process, and, as well, as final administrative guidance. Prior to that, at toxicology and chemistry guidance has been available.
So there is a lot of documentation out there. Of course, we all know that, in its pristine form, those documents can be helpful but not necessarily answer all of your questions. So this workshop is intended to provide additional guidance and help to everyone.
Over the past couple of years, since January 18 of 2000, we have converted a number of pending food-additive petitions and accepted a number of notifications and seen them go to closure such that now we have almost 300 in almost three years. Actually, when you think about that, from the perspective of somebody who saw food-additive petitions for food-contact substances going through the process years ago, and how arduous that was in many cases, that is a remarkable change and a lot of credit goes to both the industry and the government sides on this because there were advocates on both sides who worked very hard to bring into place the statutory provisions that enabled this work to go on.
As a matter of fact, I think it might not be a bad idea--there are many FDAers in this room who are, at the moment, contributing to the success of this process. We often think of our government regulators with some cynicism over the years. You hear cynical comments, but this is a group of people who are making the process work and work very efficiently for the benefit of all.
It is working efficiently for industry and it is protecting public health at the same time. I think it would be, hopefully not embarrassing too much to the FDAers in the room--to ask you all to who are in here to just raise your hands so that folks in the room can see who is available.
You see the number of FDAers in the room. These are folks in the Division of Food Contact Substance Notification Review and elsewhere in the Office of Food Additive Safety who are contributing to the success of this process. So thank you all for the work you have done and the work you will continue to do.
Even though this process has been successful, this workshop is being held in recognition of the fact that there is always improvement that can be made. We, on the FDA side, want to make the process work as efficiently as possible. We know there are gains to be made. There are ways to improve the way we do our job.
At the same time, we are aware that industry submitters of notifications can improve those notifications. There are ways to avoid mistakes, but they have to be identified and you have to be provided guidance. So the notion of this workshop, then, was born to try to bring both sides together to make the process work better.
In addition to that, I think we are open to the idea that we need to push back the boundaries of science and administration such that new ways of doing business can be seriously considered. So you are going to hear discussions today, in addition to the traditional kinds of submission of information and review of information, scientific information that we have always done over the years, you are going to hear some discussions about some of the scientific barriers that we are trying to overcome.
We will be discussing a little bit about structure activity analysis so that we can do a more efficient job of evaluating these notifications. We are open, obviously, to examining modern methods of migration testing and modeling migration in food, and even applying new and sort of barrier-breaking toxicological evaluation methodologies to this process. So, with all of that, we hope that, as time goes on, the process will become ever more efficient.
One other thing I just want to mention is that, since September, we have had an assignment--as you know, over the last year, we have reorganized the Office of Food Additive Safety. Just in the past month, we have assigned permanent directors to two of the divisions in that office and, as it turns out, we haven't made it publicly very noteworthy. We haven't made a splash in the press about it but it turns out that Mitch Cheeseman, who was up here a moment ago, is now the permanent Director of the Division of Food Contact Substance Notification Review and Dr. Francis Lin is the Deputy Director of that Division.
Those of you who follow this area and who work in this area know those two have been champions of this process and are very well suited to be senior managers in the Division. They will continue to carry the torch, so to speak, and I commend you to them and their senior management skills.
So, with that, I wish you good luck today in the workshop. It is my sincere wish that everyone finds this session helpful and, ultimately, will bring in even better notifications for us to evaluate and tune our process up so that we can more efficiently evaluate them.
With that, I will turn the microphone back to Mitch Cheeseman and send you forward. I will probably have to leave about mid-morning, but if you have any questions of me, I would be happy to try to answer them either now or before I leave.
Thank you.
CHEESEMAN: Maybe I should yield for any questions for Dr. Rulis. If not, I will go on to the purpose of the meeting.
[Presentation slides for Dr. Cheeseman are not available]Purpose and Question/Answer
CHEESEMAN: Good morning.
[Slide.]
Welcome to FDA's 2002 Food Contact Notification Stakeholder's Workshop. In honor of the season, my talk is entitled Food Contact Notification Review Process, Trick or Treat.
[Slide.]
The first thing I want to do is to give credit for a lot of the work that went into this workshop where it is due. As many of you know, I just returned a month ago to the Office of Food Additive Safety from a temporary assignment so most of the hard work of organizing this meeting was done in my absence.
In addition to all the excellent speakers that you are going to hear today, the organizational work has been done by a few individuals that I would like to recognize specifically. Among those, are Bill Trotter, Anna Shanklin and Carolyn Young who are here today. Also, I would like to recognize Sandra Varner who is the past Acting Division Director of the Division. Under her leadership, this meeting was developed and planned for the most part.
Last, but not least, I would like to recognize two of the Division support staff, Sylvia Dodson and Julie Mayer, who were helping out with registration this morning.
[Slide.]
Next, following up on Alan's mention of the reorganization and assignment of new folks, I wanted to give you a snapshot of what the Division looks like today. Francis and I are respectively the Deputy Director and Director. We have a small support staff and a vacant, at the moment, senior special assistant position under the Division Director's staff.
The remainder of the Division is broken into six teams. There are two each of regulatory teams which do primarily CSO duty, chemistry review teams that do chemistry review, and toxicology review teams. The supervisors of those teams are Arthur Lipman for the CSO team, and we do have one vacant supervisory position in that section. Some of you may have dealt with Andrew Zajak who was a supervisor in this Division up until September 22. Andy transferred, at that point, over to the Division of Petition Review and so we have listed and hope to fill shortly his previous position.
Dr. Michael Adams and Alan Bailey are the supervisors in chemistry review teams. Dr. Chingju Sheu is the current supervisor of one of the toxicology teams. Dr. Lin's promotion created a vacancy of a second supervisory position. We have, again, listed that position and hope to fill it shortly.
[Slide.]
I am going to talk a little bit about our goal today, and our goal overall. The mission of this Division is to protect the public health by effective regulation of food-contact materials. We expect to accomplish that through a couple of avenues, through informed regulation, which requires us to stay engaged with you, the industry that we regulate, as well as the public whom we serve.
Through the current and complete knowledge of the food-contact materials in the industry that we regulate, we are able to accomplish the most efficient use of the limited resources that we have.
I have mentioned before in talks, and I will mention again, that, in the absence of complete knowledge, regulatory agencies tend to make conservative assumptions or conservative decisions. So the process of staying in touch with you, the regulators, and all our other customers, ultimately pays benefits for all of us by permitting us to make more realistic estimates of consumer risk and better decisions regarding consumer protection.
Moreover, knowledgeable regulators breed confidence amongst the public in the decisions those regulators make. In addition to informed regulation, we also seek to achieve efficient review of food-contact submissions, whether they be food-contact notifications, inquiries, or prenotification consultations. We hope to make our review processes more efficient by producing better guidance, by striving to make presubmission interactions more fruitful, by continuing interactions such as this one today and by focusing our review efforts at a level that is commensurate with that likely risks of the consumer.
[Slide.]
With that goal in mind, our theme today is one of the importance of telling a complete story. Much of what we will discuss today are simple but important aspects of the food-contact notification process which, if ignored, create much more work for both you and us.
Implementation of our advice for you to tell a complete story requires each of us to try and think like the other. In many cases, where information is lacking in your submissions, we can formulate the most logical answer to our questions. However, we cannot act or make final decisions on such educated guesses so we must ask for additional information.
Being a bureaucracy, documenting and making those requests can be time-consuming, as you well know. So, in this way, providing complete information helps us to help you because we don't ask the questions that we can easily find the answers to. However, I don't want to suggest that we are offering to relieve you of your responsibility to document your safety decisions regarding notified food-contact substances. This requires a complete consideration of all migrants. Such consideration should be proportional to the likely risk and need not be exhaustive. However, it must be complete. If a substance is expected to result in an immeasurably low exposure, then often a brief, reasonable explanation of why is all we are looking for.
If the toxicity data recommendations are to supply a literature search, then the discussion of the parameters of that search, in addition to the report of no hits, is useful in determining that your safety assessment is adequate.
These are just a couple of the examples and you will hear more about that during the more specific discussions later on today.
[Slide.]
Following the advice of telling a complete story helps us get to end product that we all want and that is FDA helps you comply with the law and notify the substances that need notification and also not submit unnecessary submissions to the agency.
One of the end products is greater consumer confidence because we have a more knowledgeable regulatory agency and greater public-health assurance and, also, ultimately what you, the industry, want, greater ease in bringing innovative products to the marketplace.
[Slide.]
Today's task will be to present our guidance in an updated manner, to discuss some of the common shortcomings of food-contact notifications in relation to that guidance, to explain how we deal with different types of data such as genetic toxicity data, end-test data, things of that nature, but also, just as important, we are here to listen to your questions and to your feedback.
We held a stakeholder meeting prior to the beginning of the FCN program in March of 1999. Those of you who participated in that meeting, are aware that we listened at that meeting, that we made some changes to the program and that meeting resulted in a better program and I have great hopes for this meeting resulting in a continuing improvement of the program.
It is our hope, of course, to leave today with new ideas to improve the FCN process for both of us.
[Slide.]
Finally, I want to answer a few of your questions that have already been submitted. In our notice announcing the meeting published last month, we did offer the possibility of submitting questions in advance. We will attempt to answer those throughout the presentations this morning and in the early afternoon as appropriate.
I want to address a few of them that are purely administrative and policy-related this morning. First of all, let me assure you that the input that we have received regarding suggested revisions to Form 3480 and the administrative guidance will be seriously considered and we will act on those in conformance with FDA's good guidance practices and in conformance with the Paperwork Reduction Act.
One comment inquired into the reliability of information given by the Division during the prenotification consultation process. In particular, one question suggests that FDA has a history of, perhaps, reversing its decisions or changing its mind.
I have to say, without specifics, it is difficult respond to such questions. However, I can say that acting on specific information of this nature is an important part of my job, and of Dr. Lin's job, to insure that the guidance we give and our review practices are consistent.
I can also say from experience that, on many occasions, seemingly insignificant changes in identity, manufacturing details, and use are made by you after consultation with us. In these cases, we may have to modify our previous guidance so, again, let me go back to our theme of "tell a complete story" and add that you also need to tell us the same story each time.
Some of you have registered complaints regarding late-acknowledgment letters. I know that we have, in fact, sent out late-acknowledgment letters. You have also complained about short deadlines for submitting information to FDA.
It is difficult to deal with a general complaint like this again. I can only offer that, if you have specific problems, I hope that you will bring it to me or bring it to Francis Lin as quickly as possible so that we can act to correct the situation.
There may be some mitigating circumstances in many of these situations that drive the issuance of these letters to a later date or that require us to ask for information quicker than we might normally ask for it.
One of the comments has suggested that we incorporate into our administrative guidance a time frame of ten working days for response to requests for information in FCNs. We will be considering that change, among others, that were suggested for the administrative guidance but I believe that we can make the suggested change and I will say so today.
There was one final comment, and I am going to ask your leave to actually read it because I don't want to get it wrong, and I didn't really want to type it up into a slide for you.
"During FDA's review of FCNs, the agency frequently requests additional information. FDA sometimes accepts supplemental information and considers the date of the submission to be the date the initial information was received; i.e., the start of the 120-day review period.
"In other situations, however, FDA considers the date of receipt to be when the supplemental information was received as per 21 CFR 170.104(b)(1). Both the final rule and the administrative guidance specify that FDA will refuse to accept an FCN only if a component or element required by Section 170.101 is missing.
"We recommend that submissions of additional information in response to an FDA request should establish a new date of receipt only if the information is so significant as to render a required element of the FCN under Section 170.101 missing for purposes of Section 170.104. An element should be considered missing only when the information requested by FDA is sufficiently significant that its absence would support an objection to the FCN becoming effective or would preclude evaluation of environmental impact."
We certainly agree that, if environmental information is missing or insufficient, then that is grounds for not accepting the food-contact notification. In fact, I have spoken before in discussions of the program and stated pretty clearly that the reason for the issue of acceptance/non-acceptance was primarily the environmental portion of the food-contact notification and the fact that the environmental information was not a basis for objection. That still stands.
With regard to the interpretation of what is missing and when FDA issues a nonacceptance letter, in general, we are concerned about the date of receipt of the food-contact notification when the absent information hinders the process of the review. So, even a small deficiency in one part of the notification, if left unrectified for a period of time, will ultimately hinder the review process at some point.
What we are trying to correct in regard to resetting the receipt date is to provide us with a fair amount of time to review the food-contact notification as provided under the statute.
So, we won't totally agree with this particular interpretation. However, with regard to missing information being equivalent to information that FDA would object to, we will stick to the practice of only resetting the receipt date when the information that we are missing tends to hinder the review process from progressing.
Of course, our goal, once you tell a complete story and we give complete guidance, should be that the FCN review is a treat and not a trick.
With that, I will introduce the next speaker on our agenda, Dr. Anna Shanklin who is a consumer safety officer in the Division. She will be speaking on administrative guidance.
[See presentation slides for Dr. Shanklin]Administrative Guidance and Regulations
(FCN Background, Common Deficiencies,
Form 3480 and SOPs)
SHANKLIN: Good morning. Thanks, Mitch, for that introduction.
[Slide.]
I am Anna Shanklin. Today I will be speaking on FDA's Food Contact Notification Program, administrative guidance and regulations.
[Slide.]
In today's talk, we will have a brief introduction and look at the FCN Program from an historical and current point. Mitch and Dr. Rulis have already covered several of my points, so I will sort of run through this. We will also talk about the scope of the FCN Program. We will talk about the FCN vocabulary and review the standard operating procedures for FCNs.
We will cover a few of the administrative deficiencies that we have seen and, finally, a conclusion.
[Slide.]
Since we are part of the information age, I would like to begin by noting that OFAS has a new look on the website. This is our new web address. On this site, we have useful links and general guidance for our approval and notification programs of food ingredients and packaging.
Now, many of us wouldn't be here today if we didn't have a genuine fondness of interest in food additives, but let's begin by reviewing what a food additive is. According to Section 201(s) of the Food and Drug Cosmetic Act, the Act, the term food additive means any substance the intended use of which results or may reasonably be expected to result directly or indirectly in its becoming a component or otherwise affecting the characteristics of any food.
This includes any substance intended for use in producing, manufacturing, packing, processing, preparing, treating, packaging, transporting or holding of foods if such substance is not generally recognized as safe.
As you can note from the definition, we have two informal classes of food additives, direct or indirect, and a lot of our indirect additives will be our subject for today.
[Slide.]
Since FDA's mission, pure and simple, is to promote and protect the public health by bringing safe and effective products to the market and monitoring those products while they are still on the market, let's take a closer look at how FDA regulates food additives.
[Slide.]
The 1958 Amendment to the Federal Food and Drug Cosmetic Act, in addition to providing us with a definition for food additives also requires FDA approval of food additives prior to their inclusion in food, thus the term "premarket."
It also establishes the standard of review, the standard of safety and also gave us formal rulemaking procedures. As Dr. Rulis has already mentioned, in 1997, the Food and Drug Administration Modernization Act also defined the food-contact substance and established the Premarket Notification Program for food-contact substance and this is the primary method for authorizing new uses of food additives that are food-contact substances.
[Slide.]
What is a food-contact substance? Just briefly, it is any substance intended for us as a component of materials used in manufacturing, packing, packaging, transporting or holding food if such use is not intended to have any technical effect in the food.
[Slide.]
So here are examples of several types of food packaging and food-contact substances in direct food additives: coatings, and those are paper and metal; plastics, including both polymers as well as their monomeric components; paper; adhesives; and ingredients used in packaging such as colorants, antioxidants, antimicrobials.
[Slide.]
Here are some more specific examples of food-contact substances: they are substances that are used in the manufacturing of foods such as boiler-water additives for steam or ion-exchange resins commonly known as your secondary directs; constituents of food additives are also considered to be food-contact substances such as the monomers of the polymeric material; and also GRAS and prior-sanctioned substances that meet the statutory definition of a food-contact substance such as sodium sulfate and pulp used in making paper.
[Slide.]
Dr. Rulis has already covered but I will briefly run through some key events in the FCN Program. On October 22, 1999, the FCN started and we had several food-additive petitions and threshold regulations that were converted.
On November 12, 1999, FDA announced the availability of draft chemistry and Draft Chemistry and Toxicology Guidance Documents. On January 18, 2000, FDA began celebrating my birthday, as I told you, as SPI nine months prior to my arrival and that, on this day, we began accepting new notifications.
[Slide.]
Also, on July 13, of 2000, FDA published this Proposed Rule. On May 21 of this year, three documents were published in the Federal Register. They were the Final Rule, the Notice of Availability and Advanced Notice of Proposed Rulemaking.
[Slide.]
Now, as of October 7, 2002, we have approximately 291 FCNs to date. Eighty-three were converted for petitions and threshold regulations. We have 206 FCNs that are effective and 206 are also posted on our website. We have only had 48 that were withdrawn after Phase 1 review and, really only three that were withdrawn after Phase 2 review. Only one FCN was not accepted and we are currently reviewing 36 FCNs. So we are busy.
[Slide.]
Throughout this morning's talk, you will hear from all of the technical disciplines. That will give general requirements. They will also discuss some of the concerns we have, our lack of information that we have seen in our FCNs. We would like to keep in mind that guidance is available on the OFAS website.
[Slide.]
So, since our mission is not only to bring beneficial products but safe products, let's look at what a notification must contain. A notification must contain the identity of the food-contact substance, the manufacturing process and intended conditions of use, data and information that forms the basis of the notifier safety determination, environmental considerations and Form 3480.
All these criteria make for, hopefully, an acceptable FCN.
[Slide.]
This is Form 3480. The address is located on the front of the form and this is the mechanism of submission. I will briefly run through the parts of Form 3480.
[Slide.]
Part I contains general information where the person submitting as well as their representative is listed. We would like to pay special attention, if you have had a prenotification consultation, if you could list that number, we would appreciate it. Also, if you have previously submitted and FCN for this food-contact substance, we would like to see that also indicated.
Also, if there is a place on our website in the listing of effective notifications where this substance is listed maybe for another intended use, we would like to see this. Again, this is a part of our search that we do regardless, but it helps if this information is readily available.
[Slide.]
Part II is the Chemistry Section.
[Slide.]
Part III is the Toxicology Information where your safety narrative and comprehensive toxicology profile is mentioned.
[Slide.]
With that, I would like to say that one of the most essential components of the FCN is the toxicology part. The burden for demonstrating safety lies with the notifier. It is the notifier's determination that is essential here.
Actually, there was some sound that was supposed to go "boom" so that you could really feel that this is an essential component for the notifier.
[Slide.]
Also, the National Environmental Policy Act, which you will hear about later today, requires that FDA considers the environmental impact of all its actions, so Part IV contains our Environmental Section.
[Slide.]
Part V is the Certification Section. It just says that all the statements are accurate and complete that are placed in the notification. So we would like to see that sign.
[Slide.]
Attachments are in Part VI. One of the points I would like to bring out is that the attachment name, along with the page number, is clearly indicated. That also helps facilitate our review process.
[Slide.]
So, now that we have looked at Form 3480 and some of the historical aspects, let's take a closer look at the scope of the FCN Program, the whos, the whats and the whys.
[Slide.]
When should a notification be submitted? FCNs are required only for new uses of food-contact substances that are indirect food additives or substances used in the manufacture of food such as the boiler-water additives we talked about earlier.
FCNs can be used to notify FDA of new uses of food-contact substances that are not considered food additives in the definition such as constituents of food additives, GRAS and prior sanctioned substances, that these are not a requirement whereas these are a requirement.
[Slide.]
The same safety standard applies in evaluating food additives under the food-additive petition process as with the food-contact notification process. There are two cases in which FDA would warrant a food-additive petition and that is when the new use of the food-contact substance will increase the cumulative estimated daily intake of the food-contact substance to greater than or equal to 1 part per million or 3 milligrams per person per day or, in the case of a biocide, if the level is 200 parts per billion or 0.6 milligrams per person per day.
I would like to pause at this moment and stress, as is stressed in our administrative guidance, that prior to the submission of a food-additive petition, even if you meet these two cases, to please contact FDA because an FCN may be warranted under those cases, also.
[Slide.]
Who should notify? A manufacturer or a supplier of a food-contact substance may submit to FDA a notification for a new use of a food-contact substance. Now, what does supplier mean? The supplier means any person supplying the food-contact substance including the company supplying the food-contact substance to themselves for manufacture of the food-contact material.
[Slide.]
Who may rely on effective notifications? The manufacturer of the food-contact substance and anyone that can demonstrate that the food-contact substance being marketed has been manufactured or supplied by the manufacturer identified in the FCN and that it is being used under those exact intended conditions of use that are the subject of the FCN. This is important.
[Slide.]
We will briefly review some of the FCN vocabulary, receipt dates, starting the clock, withdrawal, nonacceptance and objection. The receipt date is the date indicating receipt of a complete notification. That is very important. Starting the clock. This is when the receipt date is set which indicates when the 120-day FCN review period begins. Then the clock starts. So, thus, a started clock equals complete submission received by FDA. We have had some language, resetting, starting the clock. But the clock does not officially start until we have accepted the notification.
Withdrawal; a notifier may withdraw without prejudice to a future submission at any time prior to our completion of review of their FCN.
[Slide.]
Nonacceptance; as Mitch mentioned, nonacceptance is when components are missing under 21 CFR 170.101 and, thus, it gives us incomplete submission and the notifier chooses not to withdraw the submission, then it is a nonacceptance status.
Objection; FDA recommends, due to noncompliance with general criteria for the FCN, failure to demonstrate safety under the intended conditions of use, or when 120 days of the receipt of the FCN occurs within a fiscal year when the program is inoperative.
[Slide.]
Who are our key players in the FCN process? We have the notifiers who prepare our notifications. We have our notification control assistant who is busy at work here. [Slide.]
Our review team is very essential. It consists of a toxicologist, and environmental review scientist, a chemist and a consumer-safety officer.
[Slide.]
What happens when you submit your FCN to FDA? As a notifier, when you submit your FCN, the notification control assistant receives it and she set a receipt date, the date that we received it and also 120-day target date is set. Then the notification is distributed to all the review teams, a CSO environmentalist, chemist and toxicologist.
Within three weeks, we hold what we call a Phase 1 review meeting to note whether the FCN is considered acceptable, which gives us Case A, or not acceptable, which gives us Case B. Let's review Case A first since those are the types of FCNs we like to see the most.
[Slide.]
If your notification is accepted, you will receive an acknowledgment letter and Phase 2 review will continue. If all goes well, your FCN will become effective. A final letter will be issued and a listing will be published on our website in the Inventory of Effective Notification.
However, in Phase 2 review, if we have safety concerns, the notifier has the option to withdraw or receive an objection letter. So those are the case for Case A.
[Slide.]
In Case B, if we don't accept the notification or find it unacceptable in Phase 1, the notifier receives a deficiency letter. The notifier has two choices at this point. They can either respond to the deficiencies or they can choose to withdraw. If they choose to withdraw to the deficiencies, they have up to two weeks or ten working days for us to receive the information.
If it takes more than two weeks, FDA considers the information not forthcoming. It is missing. The notification is incomplete. Thus, you receive an nonacceptance letter. However, if we do receive the information within the time frame and we don't accept it, you still have an option to withdraw or receive an nonacceptance letter.
However, if, within the two weeks, we do accept it, a new 120-day date is set if the information is substantive and it remains the same if it is not substantive and we have Case A, and you follow your notification through to listing on the website.
[Slide.]
These are just a few administrative concerns that we have seen to date. We have seen some incomplete Form 3480s, so we have to sort of look for the information which runs over or ties into organizations of the FCN. Some of the information is there, but it is not as well organized so we have to spend time looking, and someone may say, "Hey; I actually saw that information on Page 210 when maybe it should have been on Page 20."
We have seen illegible and faint copies of documents, especially or chromatograms. Timeliness of submission of additional information; we would like to see it within the ten working days. That would really help us out a lot because we don't if there has been a mailing error or if you are planning not to submit the additional information, so that would help.
Also, we have seen where we don't get sufficient copies of additional information. We get five copies of the FCN but, when you send the additional information, it is also distributed to our review team so we need extra copies for those, also.
[Slide.]
So how do we avoid some of these deficiencies? FDA wants you to come in and talk to us. We have prenotification consultations. They are interactions with you, the industry, and our office prior to your submission of an FCN. We would like to see those and they can arrive via e-mail, fax or letter.
[Slide.]
What types of prenotification consultations have we seen? We have meetings to discuss the requirements that will make for an acceptable FCN submission. Sometimes, we receive prenotification consultation information concerning the CEDI and ADI for your food-contact substance if it is not already available on our web.
I would just like to say that this website is updated every couple of months. Also, information concerning the eligibility of your substance prior it being considered for an FCN. We are here and we are willing and able to help.
[Slide.]
Now, I would like to say that we are accepting electronic submissions. However, they must meet the requirements in 21 CFR Part 11 and we also need a paper copy. They must be submitted on CD ROM or IBM-compatible disc and we are currently developing guidelines for these electronic submissions that will better aid you. But, for now, Part 21 CFR Part 11 is your bible.
[Slide.]
The Freedom of Information Act requires that all federal agencies make their information publicly available. But, for FCNs, let's take a look at what is releasable and what is not releasable. We will not release any information until the 120-day effective date has been reached. However, the notifier must remember that FDA's conclusion is made publicly available on our website in the Inventory of Effective Notifications.
What is releasable? General information, safety and functionality data and also information incorporated by reference. What is not releasable? Trade-secret and confidential commercial information is not releasable and information in withdrawn FCNs we do not release. So your withdrawn FCNs are safe with us.
[Slide.]
If you are looking for us on a Wednesday morning and we are in our Phase 1 review and you need some FCN guidance, how do you find us on the web? If you go to www.cfsan.fda.gov under Program Areas, Food Ingredients and Packaging--
[Slide.]
--if you click to us, you will come to our Approval and Notification Page. Also, if you look under Notification Programs for Food Ingredients and Packaging, you will see Food Contact Substances.
[Slide.]
This is our website. Our Inventory of Effective Notifications, you can get to there. Also, we have guidance that is available here. The CEDI/ADI database is available from this website and also some of the documents that were published in the Federal Register are also available on our website.
[Slide.]
So Form 3480 is also available and can be retrieved on our website. We are currently revising Form 3480. However, if you would like to print out the form and fill it in via typing, we recommend that you print out the PDF format. Also, we have received several of your suggestions concerning modifications that can be made to Form 3480. As Mitch stated, we are considering those. Just keep in mind that all forms have to go through OMB for approval so we are working on your suggestions.
[Slide.]
In conclusion, I would just like to briefly run over what is needed in the FCN. The comprehensive summary, which deals with our safety. The chemical identity, intended conditions of use, the intended technical effects. [Slide.]
We need to have the estimation of intake, exposure to the food-contact substance, toxicity information, environmental information and FDA form 3480 completed and signs. These are points to remember.
[Slide.]
We would also like to have five copies of the FCN and all FCNs should be submitted to the notifications control assistant at this address. Again, the address is available on the front of Form 3480.
[Slide.]
So, take-home points. Notifications for food-contact substance must contain sufficient scientific information to demonstrate that the substance of the notification is safe for the intended use for both human and environmental safety and also FDA encourages you to please come in for prenotification consultations to facilitate development of a complete FCN.
Are there any questions at this point?
CAMENISCH: Steve Camenisch with Englehard. One question around the ten-day. One of the problems that I have run into is actually getting the notice to you through Federal Express or mail. Are you all addressing that at all?
SHANKLIN: As far as safety or security, someone may want to comment, but what I have done with you in the past, and others, if I know that a Fed Ex copy is coming, I will let that serve as my hard copy, but I have accepted a fax. But, again, Mitch or Dr. Rulis may want to comment on that.
CHEESEMAN: I think accepting a fax is fine. If you are sending it via a carrier like Federal Express or UPS, you can also give us a tracking number and we can track it over the web, I think, also, if I am not mistaken.
SHANKLIN: Yes; you can track on the web. That's fine.
Yes?
BOWER: Does the ten days start when you send it or when I receive it?
SHANKLIN: It starts on the date that I send it which, again, I am not in management to make that call but usually the day that I send it is when the ten days start, because usually we send it via e-mail or fax where you readily get it. We never really send it through mail. We may send a copy of the deficiency letter through the mail, but I always make sure that, on that date, someone receives it via fax or e-mail.
CHEESEMAN: I just wanted to mention that folks should identify themselves and their affiliation when they ask a question for purposes of the transcript.
BOWER: I am David Bower from R.T. Vanderbilt Company.
SHANKLIN: Did we answer your question? The ten days start the day that I send it which usually is the day that you receive it because I am sending it through electronic media.
BOWER: Do you attach a receipt to your electronic mail so that you can document that I got it?
SHANKLIN: The notifiers, again, that I have worked with, I have always followed up to make sure that they have received the deficiencies. So we probably have to make that a part of making sure that the notifier has received the deficiencies. I do. I'm sure more CSOs do also.
Yes, sir?
SIMMONS: Ralph Simmons from Keller and Heckman. If we are filing electronically, how many paper copies do you want?
SHANKLIN: One.
SIMMONS: Just one?
SHANKLIN: One paper copy.
SIMMONS: The other; just a comment. We have, on occasion, received requests for additional information by mail only which can reach us pretty close to the ten-day period. So I guess we would just ask that there be a consistent policy of sending those by e-mail or fax.
SHANKLIN: Okay. We will make sure of that.
SIMMONS: Thanks.
SHANKLIN: Yes; Naeem?
MADY: Naeem Mady from Ciba Specialty Chemicals. In the electronic submissions, the toxicology will accept tox studies without signature electronically, if we are submitting the whole submission electronically, not just the form?
SHANKLIN: Tox doesn't really like to receive electronically. So, in those cases, I have seen where we let the ten days vary a little. Usually, if you Fed Ex it, then we sort of let the ten days vary. Whatever comes on fax, we would like to receive--we will receive it, but we would still like the hard copy especially for toxicology information.
MADY: I am just talking about not the ten days. I am talking in the original submission, if you are submitting, accepting electronically, the form, we can submit the whole submission electronically.
SHANKLIN: Yes.
MADY: Including the tox studies.
SHANKLIN: Yes.
MADY: However, there are electronic signatures because these are GLP studies. I am not sure if you would accept those electronically--what is the signature? Is it important to the agency at all?
CHEESEMAN: I am not sure we have dealt with that specific question.
SHANKLIN: JoAnn is back there for electronic signature.
CHEESEMAN: I will let JoAnn add something to this but I would suspect that Part 11 would apply to any sort of required signatures in electronic submissions.
MADY: Can I ask another question?
ZIYAD: In terms of electronic signatures, we are working on that module and we hope to have it ready by the end of October, 2003--hopefully, before that time. But, as soon as we are ready, we will let you know.
MADY: Thank you.
SHANKLIN: Just one paper copy, also, with that electronic copy, just to reiterate that.
MADY: I just have another question. I think Dr. Cheeseman mentioned that before. When we receive the final notification letter which it is not really that important as long as we know that the time came in and we did not receive a letter--it usually comes in, like, a month later, or twenty days later. However, Dr. Cheeseman mentioned that before, but I am not sure what the answer was.
When we eventually receive that letter, if there are changes from the original acceptance letter, then it causes a problem. So what I am saying is we don't mind receiving the letter on time, the final notification letter. However, it should not change, if it doesn't come on time, from the original acceptance letter.
SHANKLIN: Do you mean the final letter that you receive at the end? We are not bound to send those out, as you know, but we sort of do that as a courtesy.
MADY: I understand. But if I receive a letter twenty days after, or a month after, I submit the final-version submission and they accepted everything there, then I am already marketing the product after the 120 days because I didn't hear from the agency. Now, I am marketing the product for a month and then I get another letter saying, "We are changing the name and we are changing the CS number. That is not correct." It is an issue.
SHANKLIN: Okay. We will definitely take that under consideration. I hope that doesn't happen often, but we will take that under consideration.
CHEESEMAN: Again, I am going to have to say that bring specific problems like this to our attention. There are certainly situations where information does get changed after the acknowledgment letter. That is, in fact, the purpose of the acknowledgement letter, to work out any changes that are necessary.
But there shouldn't be any surprises in the final letter. If there are surprises in the final letter, then please bring it to our attention immediately and we will work to correct that problem.
SHANKLIN: Yes, sir?
DOWNES: My name is Jim Downs with Solutia. I have a follow-up question on your slide, when should you provide a food-contact notification? In there, you stated that, for example, if you have a GRAS substance, do you still require a food-contact notification for a new use for that?
SHANKLIN: It is not a requirement. It is just that some notifiers wish to notify us that they are using that food-contact notification that is GRAS, they are using that for that use. But it is not a requirement. You don't have to do that.
DOWNES: Thank you.
SHANKLIN: Are there any more questions?
ETTINGER: David Ettinger of Keller and Heckman. One of your slides addresses the issue of information that is disclosed by the agency. Typically, notifiers will sometimes mark portions of the notification that they deem to be confidential. If the agency disagrees with that, is it the agency's practice to notify the notifier and give them an opportunity to respond before the information is disclosed?
SHANKLIN: I have only been here two years and I have seen both cases. Sometimes, I do receive FCNs and they also have what we call redacted versions with the FCNs. Sometimes I have and sometimes I have not contacted. Some people make it a standard practice. I am quite sure we will have an administrative meeting and decide if we will make that a part of our practice.
We are not, under law, bound to do that.
CHEESEMAN: I believe that there are some exceptions to the procedure you are referring to and one of those exceptions is where there are some specific regulations describing the releasable information. That is the case, actually, for food-additive petitions, for color-additive petitions and for food-contact notifications.
So we are actually not required to discuss that in advance although, if we think that there is likely to be an issue, if we are talking about information that may be in a grey area, we sometimes do seek to discuss it.
SHANKLIN: Do you want to continue?
ETTINGER: I guess it is just a concern for most people. I am just wondering if there could be something that could be taken maybe a little more consistently and more seriously, that a phone call or a fax be sent to the notifier if that information is deemed to be disposable and the notifier claims it to be confidential.
SHANKLIN: Okay. We will definitely take that into consideration. That is one of our points and it has come up in our hallway meetings.
Are there any more questions?
[Slide.]
If not, I would like to close with some acknowledgements. I would like to acknowledge the Division of Food Contact Substance and Notification Review; also Blondell Anderson who is one of our consumer-safety officers; Art Lipman, who is my current supervisor who keeps me busy with interesting projects; Sandra Varner, who is also a great repository of information; and also Mitchell Cheeseman, who was my first supervisor when I came to FDA who actually had my first notification assigned to me even before I reached yje post.
So I would like to thank all of these.
[Slide.]
I have the pleasure of introducing our next member on the review team. If you remember our slide, we have toxicologists, environmentalists, consumer-safety officers, and we have our chemists.
[Slide.]
Our chemist will give us some guidance and take us on a little journey. Who is one of our premier chemists that will be taking us on this journey?
[Slide.]
None other than Mr. Kirk Arvidson. So, Kirk, take it away.
[See presentation slides for Dr. Arvidson]Chemistry Guidance
ARVIDSON: Good morning. Thank you for the introduction.
[Slide.]
As Anna indicated, I will be going over one of our guidance documents, the chemistry recommendations.
[Slide.]
Basically, what I am going to try to provide today is a brief overview of our chemistry recommendations. I will be covering these topics including what we recommend providing for the identity of a food-contact substance, its use, use level, and also some information on migration testing and analytical methods. I will also touch briefly on how to calculate consumer exposure to food-contact substances.
At this time, I would like to remind everybody that this afternoon we will be conducting a workshop where Dr. VanDerveer, one of my colleagues in the Chemistry Section, will be able to answer any very detailed questions about calculating exposures and any other scenarios that are of interest to you.
[Slide.]
As Anna mentioned, our guidance documents can be found by accessing the CFSAN web page. I have provided a screen shot of the CFSAN web page here. This is what you will actually see when you get to our web page.
[Slide.]
You can find our guidance documents by clicking on Food Ingredients and Packaging hyperlink, which will bring you to our OFAS web page. Down here, on the right-hand side, are our guidance documents. All three guidance documents, chemistry, administrative and toxicology are located here.
[Slide.]
As I indicated, the first topic I want to discuss today is the identity of your food-contact substance. I have arranged my talk to mirror the information provided in our guidance document and I have also included on each slide the actual section number where you will find this information in the guidance document.
As we discussed previously, you need to provide a name for your food-contact substance and we recommend providing the CAS name or a systematic name for your material. You can also provide trade names and common names for the material and other synonyms, but we do recommend that these names are not the only forms of identification. The CAS name is the best option.
We also recommend providing a CAS registry number for the food-contact substance. This serves as a verification of not only the name of the food-contact substance but also the structure that you provided in your notification. Not only do the chemists use the CAS registry number but the toxicologists also use this number to perform a variety of searches for additional information on your food-contact substance.
[Slide.]
Continuing with identity, we recommend providing the identities of all potential migrants contained in your food-contact substance and also their amounts found in your food-contact substance. These potential migrants may include any residual monomers, impurities and breakdown products that you may have from, say, quenching a catalyst in your reaction.
We also recommend providing CAS registry numbers for all of these materials. Again, we can use these numbers to verify structures, names and also search for additional information in our toxicology databases and any additional chemistry information that we may need.
We recommend providing a full description of the manufacturing information. As Anna mentioned, we do keep this material confidential. We like to see the identity of all your reagents, including catalysts. We realize these might be trade secrets. But we do like to see the identity of the catalyst and the solvents and the amounts of these materials that are charged to the reactors.
Include times and temperatures for your reactions and also any purification steps or purification aids that you use in the production of your materials.
All of this information is important. For example, we can use this information to verify your list of potential migrants. We can also use the information to estimate consumer exposure or to refine an exposure estimate that you have provided to us in your notification.
Finally, I would also like to mention that we recommend providing chemical equations for the production of your food-contact substance and any known byproducts that are in your food-contact substance.
[Slide.]
We also recommend providing some sort of spectroscopic data to identify your food-contact substance. Typically, people incorporate infrared spectra but we will accept other information such as proton, carbon-13 NMR spectra, other NMR active nuclei, and mass spectra. I believe we even accepted some Mossbauer data in a notification. But I would like to emphasize that you should make sure, when you send these spectra in, that you assign the peaks in your spectra and label them identifying what type of a spectra it is. I'll mention briefly here, please provide the structure of the food-contact substance and any of the impurities that are in your material.
[Slide.]
We also recommend providing the physical and chemical specifications for your food-contact substance. For example, if your food contact notification is for an adjuvant, you may want to include the density, boiling point, melting point, and the solubility of the adjuvant in the food simulants that you will be using.
This is not an exhaustive list, obviously, but this is just giving you the idea of the information that you may want to include in your notification.
For polymeric materials we recommend including residual levels of the monomers that are present in your food-contact substance and also any information on the molecular weight, number-average molecular weight, weight-average molecular weight. If you don't have that information, you can provide some viscosity data, inherent or relative viscosity information.
Also you should identify whether the food-contact substance is a random or a block copolymer, the tacticity of your material, crystallinity, if it is crystalline, and the percentage. Also, you should provide the glass-transition temperature for the polymer.
You may also want to consider providing a thermal gravimetric analysis of your material. There is a section in Form 3480 that you have to attest to the stability of your material under the conditions of use and TGA could be used to support your determination.
Also, we would like to see the analysis methods used to determine impurity levels in your food-contact substance and also the analysis method used to determine the concentration of the food-contact substance in a currently authorized material, if applicable. Remember to provide supporting data for these analyses
[Slide.]
Now that we have thoroughly identified your material, what is the material going to be used in, its use and use level. We recommend providing a full description of the use and the typical and maximum use levels of your food-contact substance. Also try to include any limitations that you may have for your material.
Some examples I have here are: an antioxidant that is intended to be used at 0.1 weight percent in a very specific resin or maybe you have a polymeric material that is going to have a specific use. For example, the polymer could be a coating on the inside of a can or just specifically used for production of soft-drink bottles.
We also recommend providing the maximum thickness of the food-contact article that will contain your food-contact substance and also information on whether this material is for single or repeat use. All of this information can be use to either estimate a consumer exposure or refine an exposure estimate that you have provided.
Also, we recommend providing the types of foods that your food-contact substance will come into contact with. Typically, people reference Table 1 in 21 CFR 176.170. This table gives some generic raw and processed-food types. In our guidance document, in Section II.D.1.c, we do provide a cross-reference table, which correlates those particular food categories with our recommended food simulants.
[Slide.]
We also recommend providing the conditions of use of your material. Normally, people reference Table 2 of the same regulation here. These are our standard conditions of use A. through H. I have reproduced all of them here for your reading enjoyment.
[Slide.]
Now that we know what your material is going to be used for, we recommend providing data to support that the food-contact substance actually achieves the intended technical effect that you stated the material is going to exert. We also recommend providing data to support that the use level of your food-contact substance is actually the minimum amount required to achieve your intended technical effect. This is a very important point.
[Slide.]
The next topic I would like to talk about is migration testing and analytical methods. I'll briefly describe the different types of migration cells that you could use. These descriptions are located in Section II.D.1.a of our guidance document. If you have a specific use, perhaps your resin is going to be used in soft-drink bottles you can actually formulate your soft-drink bottle and actually do the migration tests with the soft-drink bottle (i.e. place the appropriate type and amount of food simulant in the bottles and subject the bottles to the appropriate migration protocol).
Alternatively, if you have broader uses for your material, you can use a two-sided migration cell. We recommend one by Snyder and Breder. This cell is described in a reference listed in Appendix V of our guidance document.
Although this cell may not be universally applicable to every scenario, we do recommend that if you use a cell of your own design that you keep a couple of points in mind. We do recommend that the solvent should always flow freely around the plaques in your test cell. We also recommend that you minimize any headspace and also maintain an air-tight and solvent-tight seal. You don't want to lose your solvent over the extended test periods. Also, we recommend that you agitate your cell periodically.
Finally, sometimes you may need to use a single-sided migration cell. This is most often used when you are dealing with multilayer laminants. A reference for these migration cells can be found in Appendix V of the guidance document.
[Slide.]
The next thing you will want to consider is what kind of a test sample do you want to test in your migration cells. I am going to briefly talk about the formulation of the test sample and what types of information about the test sample we recommend providing in a notification. If your notification is for an adjuvant, you will want to formulate your test sample using the maximum-use level of your food-contact substance.
We recommend providing the characteristics of your test sample. For example, provide the concentrations of other adjuvants in the test sample such as plasticizers. These can affect migration. Provide the chemical composition and physical properties of the test sample. Is the test sample a homopolymer or is it a copolymer? If it is a copolymer, what is the comonomer content of the material? Again, provide molecular-weight data such as number-average and weight-average molecular weight and the density of the material that you are working with. If you are working with a paper sample, please provide the basis weight of your paper sample.
[Slide.]
We recommend providing the dimensions of the material; not only thickness but also length and width so we can calculate or verify the surface area of your test plaques.
I just want to make a note here; if you use a sample that is 20 mils thick and also not more than 25 percent of your food-contact substance migrates out of the test plaque, you can actually consider the migration of these materials to be independent from either side of the plaque and you can use both sides of the test plaque in your consumer exposure calculations.
Just a quick look at some of the polymer properties that you may want to consider when you formulate your test sample. If you are putting an adjuvant into an already notified or regulated material, we recommend testing a polymer with the lowest molecular weight and the characteristics of the currently regulated material as described in 21 CFR 177.
If it is a new polymer, you would want to test the polymer that is going to give you the highest migration results. This will tend to be the copolymers, also the materials with the lowest molecular weight, crystallinity and cross-linking density.
[Slide.]
As I briefly mentioned before, we do have a variety of recommended food simulants. The food simulants that we recommend are shown here. For aqueous and acidic foods, we do recommend 10 percent ethanol. If 10 percent ethanol is causing a problem, either with your analysis or if it is reacting with your migrant, you can do separate extractions in water and 3 percent acidic acid.
Alcoholic foods; if it is below 15 percent ethyl alcohol, such as beer and wine, 10 percent ethanol is recommended. If it is above 15 percent ethanol, we recommend 50 percent ethanol as a food simulant.
Fatty food simulants; we recommend the use of food oils such as corn oil. Alternatively, you can use Miglyol 812 or HB307.
I did want to stop here a moment. We have talked about the food simulants in our sample cells, but how much food simulant do you actually want to put into your sample cell and have in contact with your test sample. Typically, we recommend maintaining a realistic ratio of food-simulant volume to surface area of your test plaque.
A lot of times, we will accept 10 mL per square inch. If you us use less than 10 mL per square inch, we do recommend providing a justification for why you went below this particular level and also show that you haven't reached the partition limit for your system.
[Slide.]
We do have alternative fatty food simulants. Obviously, corn oil is not the most easily dealt with in an analytical situation. For polyolefins (complying with 21 CFR 177.1520) and poly(vinylacetate)copolymers (complying with 21 CFR 177.1350), we recommend 95 percent ethanol or absolute ethanol. Although, this solvent can exaggerate migration from other polymers so it is not universally applicable.
We recommend for PVC, polystyrene, rubber-modified polystyrene and PET that you use 50 percent ethanol as a fatty food simulant. You may, in your own experience, have come up with an acceptable food simulant but we do ask that you contact us first before you go ahead and do full-blown migration studies and submit a notification with the alternative food simulant.
[Slide.]
Times and temperatures of migration tests are discussed in Section II.D.1.d of the guidance document. The generic migration protocols that we recommend are located in Appendix II of our guidance document. These protocols correspond to the Conditions of Use A through H that I mentioned a few slides back. Again, you should give careful consideration to all potential uses that your food-contact substance may have, to ensure that you test under the conditions that are most severe.
Typically, our migration protocols recommend testing for ten days and that you sample your migration solutions at four different time intervals.
I want to make a couple of notes here. Polymers that are used below their glass-transition temperature, such as polystyrene and PET, we do recommend that you extrapolate your ten-day studies to thirty days. You can, to avoid any uncertainties in your extrapolation, actually test your materials for thirty days.
Another point I would like to make at this time, and there have been a number of incidences of this, concerns the use of end tests in place of migration tests. End tests are often found in the CFR for a variety of different polymeric materials. These end tests are meant for producers to use to make sure the materials that they are producing meet the specifications listed in the Code of Federal Regulations. These tests cannot be used as a substitute for migration testing.
[Slide.]
As mentioned, the recommended migration protocols are located in Appendix II of the guidance document. Regarding these general protocols, which correspond to the conditions of us A through H, I just want to provide some caveats for these particular recommendations. For example, Test Condition C, hot-filled or pasteurized above 66 degrees Celsius. This protocol supports Conditions of Use C through G only. It does not support Condition of Use H.
I would also like to point out that the Conditions of Use A through H do not cover cooking above 121 degrees Celsius. We do have a protocol that we recommend for food-contact substances that will be used in a cooking scenario above 121 degrees Celsius.
[Slide.]
Along with our general migration protocols, we do have a variety of specific protocols. I believe, if you look in our guidance document, there are roughly fourteen different scenarios. I have just listed a few of them out here. Unfortunately, I don't have time to go into depth on some of these.
If you are interested in submitting a food contact notification or testing adjuvants for polyolefins, we do have guidance on how to test these materials. For example, we have protocols that cover the use of an adjuvant in just one particular polyolefin or its use in all polyolefins. We also have protocols that would cover the use of the adjuvant in all polymers.
As I mentioned, we have protocols for high-temperature uses such as dual-ovenable trays, and microwave susceptors. Dry foods with no surface fats or oils, at this point in time; we don't recommend any testing for this scenario. We have a scenario for articles that are intended for repeat use. These are conveyor belts, rubber gaskets, O-rings that are used in the production of food. We also have a scenario for additives used in the manufacture of paper, and paperboard.
Finally, I want to bring to your attention a new addition to our guidance document. If you are interested in submitting a food-contact notification for materials, either a new polymer or an additive for a polymer, that will be used during the irradiation of prepackaged foods we can supply you with specific guidance tailored to your needs through the PNC process. We don't have any guidance document typed up at this point in time, but if you come to us, we will be more than happy to sit down with you and go over your ideas and provide you with some guidance on the types of studies that we recommend.
[Slide.]
Now that we have discussed your migration studies, what about characterization and data reporting. We do recommend triplicate migration studies. As I mentioned previously, we do recommend analyzing your samples at four different time intervals. We also recommend that you test blanks or control samples. The control samples serve as a baseline for your migration tests and they can also be used in the validation of your analytical method.
If you are submitting a food-contact notification for a new polymer, it is beneficial to you to actually analyze the migration solution for residual monomers instead of relying on their residual level in the polymer to estimate exposure. The magnitude of the exposure that you calculate in your notification will determine the amount of toxicology information that you will need to supply for your food-contact notification.
We recommend identifying and quantifying the total non-volatile extractives (TNE). These are all of the materials that migrate out of your test plaques including monomers, oligomers, antioxidants, whatever else might come out.
We recommend identifying those materials for the same reason that we recommend analyzing for the monomers. Again, you need to supply a certain quantity of toxicology information for these materials.
You don't, necessarily, have to identify every single material in the TNEs. You can classify them into different classes. Perhaps, you have fatty acids in the TNEs. You can classify the fatty acids as C8 through C18 fatty acids. You don't have to identify each specific one in there.
For polymer adjuvants, you will want to analyze the migration solutions for the adjuvant itself. Again, you will want to test for impurities and decomposition products of your adjuvant. For example, antioxidants decompose into different compounds when they exert their intended technical effect. You will want to test for these compounds for the same reason that you wanted to analyze for your residual monomers. Again, keep in mind that the amount of toxicology data that you will have to supply for this material is tied to exposure.
We do recommend reporting your results in somewhat strange units, milligrams per square inch. This facilitates conversion to the concentration of the additive in food.
[Slide.]
The next topic I would like to talk about is your analytical methods. This slide is somewhat deceiving since this is the bulk of your chemistry section in your notification. But we do recommend supplying a full description of your analytical method. Not only how you handled your samples, from concentrations to the different dilution factors that you may have used, but also providing the analysis method, and the type of instrument you used. Did you use GC; if so, what kind of a column did you use, what was the temperature program for your GC, and what type of detector was used?
We recommend providing detailed information on how you prepared your standards; what did you use as the standard, concentrations, et cetera. We recommend including calibration curves and also the raw data used to generate the curves. A lot of times, we send out deficiencies for lack of raw data. What I mean by raw data is the actual instrument response that you get with each of your analyses.
For example, for GC it will be area. If it is a UV/Vis, it will be an absorbance. But we do recommend providing all raw data for each of your analyses and also for the analysis of your standard solutions.
We recommend providing example spectra or chromatograms for your analyses. You don't have to submit every single one of them. We recommend including example calculations showing how you converted all of your raw data to the actual concentration of your food-contact substance in food.
Keep in mind when you are putting this notification together that the chemist's job is to verify all of your data and results. We are going to verify your calibration curves. We are going to verify the equations that you used and all of your calculations that you performed in support of your food-contact notification. So the more information you can give us, the more easily we can verify your data and results.
[Slide.]
As I mentioned a little while ago, you need to validate your analytical method. Typically, this is done by the fortification of your migration solutions; in other words, spike and recovery analysis. You are going to essentially report your precision and the accuracy of your analytical method.
We recommend performing triplicate analyses of migration solutions spiked at levels equivalent to the 0.5, 1.0 and 2.0 times the level of the food-contact substance found in the food simulant. If you don't detect your material in the food simulant, you should determine the detection limit and, again, for all of this information, you want to provide your supporting data, all your raw data, example spectra, example calculations need to verify this information.
[Slide.]
Finally, leading up to our consumer exposure. There are a variety of ways that you can calculate consumer exposure. One of them is 100 percent migration. As I mentioned in the beginning of my talk we recommend that you provide residual levels of the monomers in your food-contact substance, if it is a polymer. You can use these residual levels to estimate exposure to the monomers by assume that 100 percent of the residual monomers migrate to food.
For example, you have provided the residual level of the monomer in the food package, the package thickness, and the density of material, so we can calculate the quantity of your food-contact substance (residual monomer) per given volume of the food package. We than assume 100 percent of that material migrates into food.
Again, keep in mind the magnitude of exposure is going to determine the amount of toxicology data that you need. So you may want to try 100% migration calculations first but then you may want to actually refine this down by doing migration studies.
As I mentioned just a few slides ago, your migration results can be used to calculate exposure by combining the concentration of the migrant in the various food simulants (Mx) with food-type distribution factors and consumption factors. The food-type distribution factors and consumption factors are located in Appendix IV of our guidance document.
I have placed the equations that we use to calculate the concentration of a migrant in food <M>, dietary concentration (DC) and estimated daily intake (EDI) on this slide. The concentration of a migrant in food, <M>, is calculated using the first equation. In the second equation, the concentration <M>, usually provided in micrograms or milligrams per kilogram of food, and a consumption factor (CF) are used to calculate the dietary concentration of the migrant.
The DC is then used to calculate an estimated daily intake (EDI)or micrograms of your food-contact substance consumed per person per day.
[Slide.]
There are other methods of estimating consumer exposure such as migration modeling. This is based on Fickian diffusion of a migrant through your polymeric test material or your food-contact packaging. I am not the expert in this. If you have any questions on this, Dr. VanDerveer would be a better source of information concerning migration modeling.
We do have a variety of specialty scenarios. For example, repeat-use items. Perhaps, you have a new additive that you want to put into some O-rings. This is essentially a variation of the 100 percent migration calculation. We just divide the result by the total quantity of food that the O-ring, seal or conveyor belt will touch during its service life.
We also have a scenario for lubricants and their additives. We even have a scenario for calculating consumer exposure to rinse aids on plates and silverware.
[Slide.]
Finally, as Anna mentioned before, we have the cumulative estimated daily intake (CEDI). This is essentially the sum of all the different estimated daily intakes for the different uses of your food-contact substance. We have a database of these numbers located, again, at our OFAS web page.
If you can't find the CEDI in our database, you can request a CEDI through the PNC process and we can calculate one for you.
[Slide.]
As I mentioned, you can get to the CEDI/ADI database by going to our OFAS web page, which is shown here. At the bottom left corner of the OFAS web page you will see the hyperlink to the CEDI/ADI database.
[Slide.]
I have talked about a majority of the appendices that are located in our guidance document. The only one that I haven't really discussed is Appendix III, which contains examples of how to validate your analytical methods. I recommend that everybody takes a look at that appendix. It does give a couple of good examples of how to validate your data.
[Slide.]
In summary, I would like to remind everybody, to provide all of the details for your manufacturing process. Again, we can use this information to refine exposure estimates. Please provide the details of your analytical methods and your validation experiments. This will allow us to verify the results of your migration studies and the consumer exposure that you have calculated.
In addition, not only do we recommend providing a dietary concentration and EDI for your food-contact substance but we also recommend providing them for any of your impurities such as residual monomers or catalysts in your food-contact substance.
That's it. If you have any questions, I will try to answer them now. I would just like to remind everybody, we do have a workshop this afternoon that you can ask more detailed questions.
Dr. Lickly?
LICKLY: Tim Lickly with Dow Chemical. You ask for all the raw data.
ARVIDSON: Yes.
LICKLY: I guess my problem with supplying you with all the raw data is, as you know, we do different food simulants. We do multiple-day analyses. There is a different calibration curve for each one and modern analytical equipment nowadays processes that and takes that down to micrograms per ml of solvent or something like that.
Isn't it more appropriate to just give you an example of raw data, again, a consistent package where you give a linearity curve, give a response. You give an area for that response so you can see the typical analysis of a given then and then give you a process data, because the raw data would be meaningless without all the other data. That would get voluminous to review.
ARVIDSON: Well, yes. You can give us a typical chromatogram, a typical calibration curve. Typically, people--Anna and I are fairly new people here; typically, I have seen the majority of the data, or all of the data, provided for the analyses for your food-contact substance. There will be a stack of information. We like to get I guess what I would call the raw data, micrograms per square inch or micrograms per milliliter and peak areas so we can go through your equations and see how you calculate everything and double check.
Granted, we may not be able to double-check all of your calculations, but we do need to go through and make sure there are no error in your equations and calculations. We do run into mistakes and things like that. But, if you can provide as much information as possible--the more information, the better.
HECKMAN: My name is Jerry Heckman. I am with Keller and Heckman. This is not really a question. You may well want to say that this should be taken up with the General Counsel's Office or something, in which case we will be glad to do that.
But, in your paper, on Page 4 and in the Form 3480, both, you tend to indicate that people have an obligation to supply you with extensive data about intended technical effect and the amounts necessary to accomplish the effect, et cetera, et cetera.
The statute doesn't support that. Under the statute, you are not required to prove efficacy with food additives, not even direct additives much less indirect. That fiction has been carried forward in FDA for reasons that are not clear to me, other than that maybe some people got transferred from the drug side, for a long time.
So anytime I see it, I have sort of an obsession about trying to straighten that out. You are entitled to information to indicate that the substance will accomplish some effect and you have a right to deny the application if it will work a fraud or deception on consumers, but you don't have a right to inquire any further than that with regard to intended technical effect, in my opinion.
We have written a couple of papers on that, one of which is on our website, a copy of which I can give you today if you would like to have it. But I think it is time for FDA to stop perpetuating that myth.
In 1958, we fought long and hard after it was proposed that you be required to prove efficacy to convince Congress that that was neither necessary nor proper. So, every time I see it, I am going to say something about it, as you may have just gathered.
CHEESEMAN: Let me thank Jerry for that input and also you don't have to send us any copies of the paper. I actually have several copies in my office that you gave us the last time we had you at a meeting. So we appreciate that input.
ARVIDSON: Are there any other questions?
McCOURT: Chuck McCourt from Dow Corning. You mentioned that there are special calculations for repeated use and for lubricants.
ARVIDSON: Yes.
McCOURT: In the guidance documents, I haven't seen a whole lot of special calculations. Are those somewhere else?
ARVIDSON: Like I mentioned, the rinse aids, we have come up with scenarios. Often, they are not included in the guidance document but we have scenarios that we have on hand.
McCOURT: So, if I contact you, I could get that information?
ARVIDSON: Yes; you can contact us through a PNC and we can write something up for you and get you a copy of those particular recommendations that you are interested in.
McCOURT: Okay.
ARVIDSON: Yes, sir; in the back?
PACE: Gregory Pace, Sun Chemical Corporation. On Page 7, you are talking about the analytical method and chromatograms and raw data. You say to label all spectra and chromatograms. If we have groups, functional groups, like you said, food acids, can they be labeled that way or does each individual peak need to be labeled?
ARVIDSON: I would think you would be able to provide us with some general indication like peaks, with retention times 5 through 17 are fatty acids between this molecular weight and that molecular weight or C8 through C16. I don't think, in that scenario, you would have to identify every single one as long as we know that it is a generic category and which peaks are assigned to that category.
There could be something in the middle that maybe some other substance that you would want to bring to our attention, "This is some other compound." But you could, I believe, do that; yes.
REED: Peter Reed of Ondewo Nalco. I would like some clarification on the confidential aspect of the information we give you; for example, it may be nice for us to send our manufacturing procedure, our catalyst information. You indicated that would be confidential. However, in the previous question and answer session, there was a discussion about the requirement for you do that. Apparently, there is nothing preventing you from releasing that publicly.
I would like some clarification on that if I misunderstood anything because, obviously, we don't want to send you information that would compromise our ability to make a profit.
ARVIDSON: That is completely understandable. Mitch, I am sure, can address that better than I.
CHEESEMAN: Since I created the confusion, I should try to clear it up. What I was speaking to of not having an obligation was not necessarily having an obligation of coming back to notifiers about information about information that is marked confidential that we do not necessarily agree is confidential.
There has never been any question, I don't believe, that manufacturing, specific manufacturing information, catalyst information, is confidential, commercial or trade-secret information and is not subject to release under FOI.
SIMMONS: Ralph Simmons from Keller and Heckman. It is not really a question. It is a comment but it relates to that last comment from Dr. Cheeseman. Our position is that you do have an obligation to contact notifiers before releasing confidential information. We do not agree that you are exempt from that regulation.
We can get into the details of that at another time, but, beyond that, we have run into at least one occasion recently where the failure to make that contact resulted in the release of confidential manufacturing information, specifically impurity information.
Fortunately, we caught that before it went someplace other than our office. But, nevertheless, that did happen and it resulted from the failure to contact us before the information was made public.
ARVIDSON: Any other questions?
SHANKLIN: I would just like to say that there is an administrative workshop set up for this afternoon, so I will be available along with other CSOs for questions that you have targeted for administration and maybe clarify some of the things regarding confidentiality of information.
ARVIDSON: I guess we are scheduled for a break at this point in time? Is that right, Mitch?
CHEESEMAN: Yes. Why don't we go ahead and start that.
[Break.]
[See presentation slides for Dr. Young]
Toxicology Guidance
YOUNG: Welcome back. Thank you for being here.
[Slide.]
I am Carolyn Young. I am a toxicologist and I will be presenting information this morning on toxicology and the Toxicology Guidance Document that became final in April of 2002. I will also present information on the deficiencies that we have seen.
[Slide.]
This afternoon, we will have presentations on the Form 3480, the Tox portion of that form. We will have speakers talk about genetic toxicology, structure-activity relationships and the integration of that genetic toxicity information in SAR and risk assessment.
[Slide.]
So, welcome to my portion which is Toxicology Guidance 101.
[Slide.]
As you have heard before, we have webpages to assist you in locating these documents. The first address up there, as you know, is pretty long. FDA has a new webpage which you may find helpful. This is a little bit easier, at least for me, to remember so, in the next few slides, I will take you through the steps to get down to the guidance document itself.
[Slide.]
This new page has so much information, I just want to make you aware that it exists.
[Slide.]
You have seen some of those pages before.
[Slide.]
When you get to this page, you do have to scroll down considerably--
[Slide.]
--to get down to the individual guidance documents. Again, there is one for administrative, chemistry and for toxicology.
[Slide.]
The Toxicology Guidance Document is divided into twelve separate sections. I am going to be focussing on these six sections. The first one is an introductory section.
[Slide.]
Moving right along into the topic here, exposure estimates for the food-contact substance and constituents. I am going to bring the point up right here that we do need information on the constituents. This is a little bit different than what has been the practice with indirect-additive petitions. So you will see repeatedly, in these slides, reference to constituents as well as the food-contact substance.
[Slide.]
As mentioned here, the exposure estimates are important because we use these as a guideline for the level of safety testing that we recommend that you submit to us.
[Slide.]
We have already heard the definition of a food-contact substance, so I don't need to go through this again.
[Slide.]
But, for a constituent, it is a nonfunctional and unavoidable chemical entity associated with a food-contact substance. Examples of this are impurities, unreactive monomers or ligomers, processing aids, side-reaction products during synthesis and decomposition products. These do need to be considered in developing your tox information that you will submit in with your notification.
[Slide.]
We do ask that the test substance that you use in toxicity studies represent what you expect to migrate into the food. So this would be either the food-contact substance, itself. When the food-contact substance has been a polymer, sometimes, then, the best material would be low-molecular-weight oligomers. Again, the constituents. We do recommend that you use a single batch when you are doing your toxicity studies instead of using multiple batches or to use similar batches that are comparable.
[Slide.]
We also want to know exactly what the test compound is, the name, the structural formula, some of the components and constituents of the test material and, as mentioned earlier, the CAS number is very helpful for us in being able to do our literature searches and carry out our review.
[Slide.]
The safety-testing recommendations that we have in our guidance document are a tiered structure. That means that, based on the exposure in the daily diet, the lower the exposure, the less testing that we recommend.
There are basically three separate levels in this scheme. The first group is the lowest exposure which is anything that is less than or equal to 0.5 parts per billion. For this level of exposure, we are not recommending that you provide or conduct toxicity studies. However, we do need to make sure that any potential carcinogenicity issues regarding those compounds is addressed.
One way to do that is to make sure that you conduct a thorough literature search of published and unpublished studies and then provide to us summaries of the these studies. If it is a pivotal decision that you are making, sometimes we will even ask for the actual studies, themselves. But, at this level of exposure, you can see that we are not asking for you to go out and conduct studies.
I have listed this information on the bottom half of the screen in a lighter yellow color. The reason I have done that is that this same information in this lighter yellow color will show up on the next several slides as we go through these three different levels. So it is easier for you to know that this information repeats at every level.
[Slide.]
As I mentioned before, we need to know about any carcinogenic potential that is brought about by migration of the material into the diet.
[Slide.]
The second level is for exposures that are greater than 0.5 ppb but less than 50 ppb. At this level, we have genetic toxicity studies that are recommended. We recommend gene mutations in bacteria and we also recommend an in vitro study. Chingju Sheu will be talking about these studies in more detail during the workshop section this afternoon.
[Slide.]
As I mentioned before, this is that lighter yellow color. It is the same information requested for the lower exposure is, again, requested at this exposure level.
[Slide.]
Again, the carcinogenic constituents need to be addressed.
[Slide.]
The third level is anything that is greater than 50 ppb but less than 1 ppm. At this level of exposure, the recommendation is for two subchronic oral toxicity studies, one in rodents and one in nonrodents. These studies should be able to provide an adequate basis for determining the ADI, the acceptable daily intake.
[Slide.]
Once again, you see genetic toxicity studies. This time, in addition to the first two where you have a choice, either one of the in vitro studies, a gene-mutation study in bacteria. This time, the third study is an in vivo test for chromosomal damage using rodent hematopoietic cells. So that is something new that gets added at that third level.
[Slide.]
Once again, the potential carcinogenicity needs to be discussed and this information carries through including the carcinogenic constituents.
[Slide.]
Anything greater in exposure than 1 ppm, the testing significantly increases. At that point, we ask for comprehensive studies including bioassays. But, as mentioned earlier, I would like to remind you that the recommendation to either go with an FCN or a petition, FDA would really appreciate it if you would contact us before submitting a notification or a petition so that we can help you make that appropriate decision as to where that notification or petition should be; should it be a petition or a notification.
[Slide.]
We also have provided, in the guidance document, a statement that when the test material is a biocide that the previously discussed testing recommendations, those three levels that I previously mentioned, are the same, however, at one-fifth the value of CEDI. So, in your handout, you will see that this information is provided for you. I just took one-fifth of the values that I previously mentioned.
[Slide.]
Some of the information about the safety-testing protocols; I would like to mention that, originally, in 1982, the toxicological principles for the safety assessment of direct food additives and color additives, better known as the Redbook, was published. In 1992, there was a second draft of this. Beginning in 2000, we have started putting the final edited chapters of the Redbook available on the web. The address is provided here.
Not all of the chapters are available yet. We work on those on a continual basis. So, if you don't find the chapter that you are interested in right now up on the web, then refer back to Redbook II or back to the original Redbook I.
Of course, whenever you are planning to conduct toxicity studies, if the recommendations that we have provided in the Redbook are different from what you plan on doing, please contact us so that we can offer our opinion and our suggestions so that when the study is eventually submitted to us, we don't have to go back to you and say, "Oh; if you had only come and talked to us, we could have told you that we needed it done this way or we needed an extra piece of information."
[Slide.]
For the organizational structure of the information that we would like to have in any food-contact notification, the safety summary is the first part and then there is a second part, the safety-data package.
The primary component of the safety summary, from my point of view, is a safety narrative. After that, there is a listing of toxicity studies for the food-contact substance and the constituents and a description of other relevant studies. All of this part of the safety summary is found in Form 3480, Part III.
The Safety Data Package forms part of the rest of your submission. The major component of that, in my opinion, is the comprehensive tox profile. The comprehensive tox profile is a summary of the different studies. I will provide more information in just a few minutes on that.
We would also like to have you submit the original study reports, a complete published study. If it is a published paper, don't send us just the top page and say, "Here it is." Please provide us with an entire study.
Appendices; underneath the appendices, I have included literature searches. The documentation for searches is included in this section.
[Slide.]
The safety narrative is a concise summary of the scientific basis for a decision of safety for the food-contact substance and its constituents. So the safety narrative is a combination of all the information that you have put together to form that basis of safety.
It should include an estimate of human consumption, the EDI that Kirk previously mentioned. It should discuss all the potential toxicity and to include any information and decisions about mutagenicity and carcinogenicity including the worst-case upper-bound lifetime risk levels for carcinogenic constituents.
[Slide.]
In the guidance document, it mentions the ADI, the acceptable daily intake and a way to arrive at the ADI using the no-effect level. We remind you to use the most relevant studies and the most relevant endpoints when you are doing your calculation to come up with the ADI.
The document also provides the general safety factors that are used in the calculation with the NOELs. I have provided this table for your use. Once again, we do have our database, the CEDI/ADI database. As you have heard before, we do appreciate when you contact us ahead of time and request information on the CEDI or the ADI so that we can assist you before you actually submit your notification and have all that information in the notification when it comes in.
[Slide.]
So the take-home message is, for the safety narrative, have one for notification. It is part of the safety summary and it is located in Part III of the Form 3480.
[Slide.]
As you heard Mitch mention earlier, we like to have a story--not that we really need to have all story stories, but we do like a complete picture of what you are trying to present to us. Make sure that the safety narrative and the information includes a concise summary for decisions of safety for the food-contact substance and the constituents. Include in that, your ADI, your NOELS, your ADIs and discuss the main points that lead you to make that decision of safety.
[Slide.]
The comprehensive tox profile is a little bit different. It is a summary and critical evaluation of all the available toxicological information pertinent to the safety evaluation of the food-contact substance and its constituents. So, in a notification where you have one safety narrative, you may have five, six, two CTPs. You should have one CTP for each compound, whether it is a food-contact substance or constituents that are migrating in at levels that would cause us to think about how much is there, what toxicity testing do we need to substantiate that decision of safety for each one of those.
As I have indicated here, we ask that you do literature searches so that the information and your discussion to us, your story to us, is up to date. Based on the information that you have found out about the food-contact substance and/or the constituents, we ask that you provide us with summaries of that information.
Sometimes, that summary is going to be really short. "We conducted a literature search and there wasn't any information available." Other times, there will be a lot of studies and you will be submitting the actual study reports to us and we ask, in the comprehensive tox profile, to give us a synopsis of the study and then, as indicated earlier in the safety data package slide, to make sure that you give us a complete study submitted in the food-contact notification.
Use the information from the available studies in your determination of your NOEL and ADI. We recognize that the NOEL, usually we have studies that are a minimum of 90 days in duration that we use in order to use that NOEL and calculate an ADI. Again, risk assessments for carcinogenic constituents. Then there is a bibliography.
[Slide.]
So the take-home message here is that there is one CTP for the food-contact substance and one CTP for each migrating constituent. This information is located in the safety data package which is part of the body of the notification.
[Slide.]
Once again, tell us a story with the CTP. Again, you can have multiple CTPs in a notification. Tell us a story. Link all the pieces together. It is like anything you would like to read where all the pieces fit. We like to see that as well.
[Slide.]
This concludes the portion on the Tox Guidance document.
[Slide.]
You have heard a lot about the stories you want you to tell us. This is my opportunity to tell you the story of food-contact notification, some of the things that we have seen, some of the deficiencies that we have experienced and how we have dealt with them.
[Slide.]
These are two of the characters in the stories. As you can see here, "We sure have seen a lot of variety of deficiencies in the food-contact notifications that we have received." As, is my hope, that the stakeholders meeting that we are having today helps you in the future when you are putting together a notification and I hope it helps everybody so that these notifications are submitted without any deficiencies and that everybody can take care of what they need to easily and efficiently.
[Slide.]
Deficiencies can be grouped by the prenotification consultations, exposure estimates, the safety narrative, safety assessment of the food-contact substance and of the constituent. A special case; a low-molecular-weight oligomers and the organization of that safety information.
[Slide.]
All too often, we have seen this scenario where we look at the notification when it comes in and we say, "That is not what they asked when they came in. They came in and they talked to us. We thought we knew what was going on". Sometimes, it is a different contact substance. Sometimes, the use is different and I am sure it is frustrating for all of you in the industry as well as for us.
[Slide.]
Our recommendation is to provide sufficient background and use information so that when you submit the request for information that we can provide to you exactly what will serve best in the long run.
[Slide.]
For exposure estimates to the food-contact substance and constituents, we have noticed that some notifications do not take into consideration the cumulative exposure. They have not actually calculated exposures. Oftentimes, I have seen this in relationship to constituents. We need all of this information to be addressed or it becomes a deficient submission and then you hear about it from us.
Our recommendation is to submit appropriate chemical-identity information and exposure estimates for the food-contact substance and the constituents.
[Slide.]
For the exposure estimates, a deficiency is the notification does not include the scientific basis for a decision of safety for the food-contact substance, and, again, the constituents. Sometimes, we see the narratives and they have really done a great job talking about the food-contact substance but there is absolutely nothing about any of the constituents.
So, please, when you submit the notification, go back and read your safety narrative and make sure that it is a complete story for us.
[Slide.]
We need adequate information to support that decision of safety. Sometimes, we encounter inadequate or missing toxicity studies. Sometimes, this means that, in our literature search, we have identified studies that have been conducted but you have not mentioned them. Sometimes, we have only gotten parts of toxicity studies. Sometimes, there are pages missing.
So we ask for a complete submission of toxicity studies, especially when they are the pivotal studies for that decision of safety.
[Slide.]
Another big thing is the missing or inadequate literature search. I have presented here what a literature search should include; the names of the databases that are searched, the period of years that are searched, the specific search terms that are used, documentation of the search, justification and relevance of the search, and review and the summary of the individual references and how they are pivotal to that safety decision.
It is very frustrating, from my point of view, to open up an FCN and see 150 pages of literature search but you can't tell what the search terms were, how it relates to the FCN. You just sit there and look at it, and "How does this fit?" So, my personal preference is, "Tell us a full story when you submit your literature searches." It would make our job a lot easier and it would avoid that call back to you saying, "I'm sorry, but your FCN is deficient."
[Slide.]
In the safety assessment of the food-contact substance, we have an inadequate assessment of toxicity. Sometimes the studies that were used for the basis of safety or for determination of an ADI was an old study and there could be newer studies that are out there.
Again, make sure you do a comprehensive literature search and that you are aware of the information and then look at the studies carefully, make that you are understanding the basis for the NOEL, that it is the most relevant endpoint and that the study that you are using has the most relevant endpoints and that is most applicable.
[Slide.]
So a reminder to use the lowest ADI, the NOEL from the most relevant study and from the most sensitive endpoint, use the appropriate safety factors and also use the CEDI/ADI database as a reference.
[Slide.]
"Well, they submitted a pretty good FCN except for constituent information." Her response is, "I agree, but it is still a deficient FCN."
I, unfortunately, deleted my gold star. I really want to compliment a lot of people who have submitted FCNs and have done a fantastic job, not only on that safety narrative but they have been very good at including that constituent information.
[Slide.]
Some special notes about constituents. Exposure is generally low relative to that of the food-contact substance. Safety evaluation should usually focus on the potential mutagenicity and carcinogenicity and the quantitative risk assessment required for carcinogenic constituents. We need all of that information.
To make sure that when there is a carcinogen of concern, that concern is based on more than one piece of information. It is based on the weight of evidence that we have from toxicological studies, the exposure estimate and we oftentimes will rely on the structure-activity relationship information.
[Slide.]
As I have indicated before, the safety assessment also applies to the constituents. We want to make sure that we have adequate information for that decision of safety and, once again, what we have experienced as inadequate or missing toxicity studies and missing literature searches. So, please make sure you include that information.
[Slide.]
As we have seen with the food-contact substance, itself, we want to make sure that you are using the most up-to-date information possible and the most relevant studies for the NOELs and the calculation of ADIs.
[Slide.]
We have also noticed, with the constituents, that there has oftentimes been a lack of quantitative risk assessment or an inaccurate calculation of the extreme case upper-bound lifetime risk to humans from exposure to carcinogenic constituents. So our recommendation is make sure you present convincing scientific evidence to justify any alternative approach to estimating risk.
So, if you are planning to develop an argument, look at the Tox Guidance. If your plan is different from what ours is for justifying and supporting food safety, talk to us about it. Lay it out in your story and your narrative so that we know that you have really considered it and we can also evaluate it and look at it through your eyes, and then make a reasonable decision.
[Slide.]
We have a special case, low-molecular-weight oligomers. We have seen where notifications did not include exposure estimates, assessment of safety including toxicity studies. I have an asterisk here as a reminder that sometimes the oligomers may be chemically synthesized for animal tests. We don't expect you to go out and spend exorbitant amounts of money and time to get enough test material when it is just not feasible to do that.
This is where the chemically synthesized oligomers may be the way to go. Again, we want you to discuss these low-molecular-weight oligomers in the safety narrative and also in a separate CTP.
[Slide.]
Generally speaking, the organization of the safety information; we really appreciate it and we give gold stars to all of you who have taken the time and effort to organize this information in a way that is easy for us, when we pick up that notification, that we can easily find the information.
Again, the safety narrative, the comprehensive tox profile and the literature-search information are things that I look for right off the bat when I get a notification. It is very nice when we have tables and it says, "The literature-search information is located here in this notification," and that there is that little table of contents that many of you provide for us. That is extremely helpful.
[Slide.]
The grand take-home message; make sure that your notification provides adequate qualitative and quantitative information regarding what migrates into the diet; food-contact substance, low-molecular-weight oligomers and the constituents
[Slide.]
Tell me a story. Make it comprehensive, concise and make it a summary of all the information that you use to form a decision of safety. Also, make sure in your safety data package that you include these comprehensive tox profiles and that you include literature searches, the documentation of those searches, for the food-contact substance and the constituents.
[Slide.]
Provide a detailed review and fair interpretation of the pivotal toxicity data and determination of the NOEL, or NOELs, as the case may be, and the ADI or ADIs, as the case may be.
[Slide.]
Be sure, when there are carcinogenic constituents to calculate the acceptable lifetime cancer risk.
[Slide.]
As our characters in our twisted tales remind us, "I think that we have covered most of the main points related to the tox portion of the food-contact notification," and, "I guess this wraps it up for us." I hope everybody remembers that--
[Slide.] --when you are standing there and you are pulling your hair out, this could be for those submitting the notifications. Keep your hair attached to your head. Turn around. Consult the Tox Guidance Document. Consult with FDA by phone or by e-mail. Let's keep your hair attached to your head.
It also keeps us from looking like that as well.
[Slide.]
So, if we look into the future a little bit, this is my hope for the future, at this time next year that we can say, or your supervisors or your bosses can say, "Congratulations. During this year, all of your FCN submissions have been complete and no deficiencies."
As we have had our little characters throughout the second portion of my presentation, here we can see them in the Year 2020. "It has been great working with you on all these FCNs all these years." Her response is, "It seems like it was just yesterday that we started with FCNs and had that stakeholder's meeting."
[Slide.]
His response is, "That meeting sure did make a big difference in the FCNs that we received." Hers is, "As I recall, the percentage of deficient FCNs significantly decreased immediately after that meeting."
So that is my hope and my intention was to present the information and hopefully keep everybody's hair attached to their head rather than pulling their hair out.
So I appreciate your attention. I will take any questions that you have and hope that I can answer at this time.
DOWNES: My name is Jim Downes with Solutia. Has the agency had any experience at accepting HPV dossiers such as the Euclid format for comprehensive tox profiles?
YOUNG: I didn't understand. Let me say back to you what I did hear, that there is some other information that you want to know whether we would accept as being equivalent to the information that I presented?
DOWNES: Yes. In the High Production Volume Chemicals Program that USEPA has along with the OECD, industry is doing a lot of pretty comprehensive tox profiles under a Euclid format which is a European format. My question is are those profiles acceptable in food-contact notifications to FDA?
YOUNG: I am not personally familiar with those documents so I can't answer the question. But we can find out an answer and get back to you and let you know.
DOWNES: It may be a unique subset of chemicals that are going to be notified because they are chemicals which are in commerce at greater than 1 million pounds a year. But those documents are generally available and would be a much smoother operation than starting from scratch to prepare a new one.
YOUNG: Thank you.
Any other questions? Yes?
CAMENISCH: Steve Camenisch with Englehard. What information would you require for constituents that are previously FDA-approved? If you have a constituent that is FDA-approved, are there any differences?
YOUNG: My understanding is if the constituent is already previously approved and there is no increase in exposure, remembering that our recommendation is based, to a large extent, on exposure. So, if you have an increase in exposure to a constituent and it is not covered by a regulation that is already there, then we would recommend that you provide us with the additional studies.
Does that answer the question for you?
Yes?
ITZKOFF: Mark Itzkoff, Olsson, Frank and Weeda. For constituents that are present at less than half a part per billion, how much information do you need in terms of identifying that constituent if it is a minor constituent that is hard to identify? Is it sufficient to say that it is present at, say, 0.05 ppb, that you can't identify it to do the tox profile?
YOUNG: When we have an exposure that is very, very low, we do ask that if there is a potential that it is carcinogenic, that you have done a literature search, sometimes there is not going go be any information there. Sometimes, we recognize that it is very difficult to positively identify exactly what that constituent may be because of limit of detections.
In those cases, you can always submit a PNC and ask ahead of time whether there is anything else. But I think that, if in your summary, in your CTP and the safety narrative, that you mention exactly the parameters that you do know and the exposure, and you discuss it, that that, a lot of times, will be adequate. But make sure that if there is any likelihood of mutagenicity, carcinogenicity, that that is adequately discussed.
ITZKOFF: Thank you.
YOUNG: Yes?
LICKLY: Tim Lickly with Dow Chemical. In the case of oligomeric material where we come up with, say, a new monomer or a new comonomer in a polymer, and we are really replacing an existing material such as, say, a new alphaolefin in a polyethylene or something, it is very hard, again, to do the mutagenicity studies and stuff for low-molecular-weight materials and where it is a replacement for a similar constituent, again, methyl acrylate versus ethyl acrylate.
Maybe methyl-acrylate polymers are already regulated and you want to use ethyl acrylate. Would that be acceptable to just identify it as a replacement and identify the amount of oligomeric fraction in the diet rather than trying to generate the mutagenicity studies?
YOUNG: We recognize that there are those times when this type of situation occurs. But we also know that a different substance can be a different substance and then the tox profile would be significantly different.
We do look for information that would correlate the information, supportive documentation of similarity, similarity in structure, what might be known about the compound that you have previously brought to our attention and this new compound.
This would be another time when that prenotification consultation could be very helpful. Based on the similarity and what information is known and what the exposure is, then we would make some sort of determination that would tell you specifically whether or not you would need to go the route of actually conducting the studies.
If there are any other questions, I encourage you to use our question box and our three-by-five cards over here--they are color-coded--to help us in sorting through the questions and getting them to the right discipline during lunchtime. We hope to answer those questions for you during the afternoon portion of our presentation.
This concludes my portion. I introduce Paul DeLeo.
[See presentation slides for Dr. DeLeo]
Environmental Guidance
DeLEO: Thank you, Carolyn.
Good morning.
[Slide.]
I'm Paul DeLeo. I am an environmental scientist in the Division of Chemistry, Research and Environmental Review. That is not to be confused with the Division of Food Contact Substance Notification Review. Our environmental review team supports activities in the Office of Food Additive Safety both in that division and the Division of Petition Review and we also support the Center for Food Safety and Applied Nutrition in other areas where they need to meet their environmental requirements. So we are in a separate division.
This morning, I am going to be discussing environmental requirements for food-contact notifications.
[Slide.]
As part of my talk, I will discuss some of the background of the National Environmental Policy Act, what we call NEPA. I will talk about why food-contact substances and notifications are the subject of NEPA. Then I will go into the specific requirements that you need to be aware of in your notifications including environmental assessments or categorical exclusions from the need to prepare an environmental assessment and how you can fill out the Form 3480.
[Slide.]
The National Environmental Policy Act was passed by Congress in 1969. It requires each federal agency, including FDA, to take environmental considerations into account in their final agency actions. The requirements of NEPA supplement FDA's authority under the Federal Food Drug and Cosmetic Act and other public-health statutes.
NEPA does not supersede the authority of these statutes. In fact, NEPA does not require the agency to favor environmental protection over other considerations such as human health in its decision-making, just that environmental impacts be considered during the decision-making process.
FDA's NEPA regulations are found in 21 CFR Part 25.
[Slide.]
In addition, NEPA is a full-disclosure act. It requires that interested parties in the affected public are informed of the environmental analysis including the decision-making process for major agency actions. The complete environmental record is open to the public and public participation is encouraged.
NEPA is also a broad statute. It requires consideration of all aspects of actions which may have an impact on man's environment including impacts on natural resources and energy use. In addition, it applies to actions which may have impacts abroad.
[Slide.]
NEPA states that all agencies of the federal government shall include an environmental impact statement with every major federal action significantly affecting the quality of the human environment. Environmental impact statements are extremely resource-intensive documents to prepare. Therefore, environmental assessments are prepared to help determine if an action will significantly affect the quality of the human environment and thus require an EIS.
[Slide.]
At this time, I am going to move on to why FCNs are subject to the National Environmental Policy Act.
[Slide.]
I am going to be summarizing a discussion of this analysis that was published in the Federal Register, the May 11, 2000 edition of the Federal Register. This was part of FDA's Direct Final Rule expanding NEPA regulations to cover food-contact notifications. So, again, this is a summary of a legal argument or a policy decision. I just want you to be aware of that because there is some detailed legal analysis in here.
[Slide.]
NEPA requires agencies to consider environmental impacts for actions that are major and final. So I am going to go into these issues of finality and major actions.
[Slide.]
Regulations implementing NEPA define major actions as those which include circumstances where the responsible official fails to act and that failure is reviewable by the courts under the Administrative Procedure Act. This definition is in the NEPA regulations which are under EPA's regulations at 1508.18.
The obvious example here is that failure to object to a notification falls under this definition of failure to act.
[Slide.]
The agency's decision not to object also meets the finality requirement under the Administrative Procedure Act as was recently articulated by the Supreme Court. It marks the consummation of the agency's decision-making process; that is, by not objecting, the agency can be considered to have reached a conclusion about the safety of a food-contact substance and rights or obligations are determined; that is, authorization for marketing is a direct and appreciable legal consequence of the agency's decision not to object.
[Slide.]
Subsequently, the agency's decision not to object is a major and final action and therefore the requirements of NEPA apply to the food-contact-notification process.
[Slide.]
I am going to go into those requirements now.
[Slide.]
Every food-contact notification must contain either an environmental assessment or a claim of categorical exclusion from the need to prepare an environmental assessment. If an FCN does not contain an EA or claim of categorical exclusion, the agency does not consider the notification to be complete and will not accept it for review.
Similarly, if an EA or categorical exclusion is submitted with an FCN but a component of that is deficient, your FCN may not be accepted for review. That is a possibility.
[Slide.]
I am going to talk now about environmental assessments and categorical exclusions but I am going to actually do them in reverse. I am going to talk about the categorical exclusions first.
To begin with, what is a categorical exclusion? Well, a categorical exclusion is a category of action that the agency has determined do not individually or cumulative affect the quality of human environment. Ordinarily, when an action meets the criteria of a categorical exclusion, neither an environmental assessment nor an environmental impact statement is required.
You will recall that EAs are prepared to determine whether an action has significant impacts to the environment and would require an environmental impact statement.
Also, you should be aware that there is a provision in NEPA that requires an EA to be prepared even when an action meets the criteria of a categorical exclusion but when extraordinary circumstances exist. I am going to be speaking about extraordinary circumstances a little bit later.
[Slide.]
FDA expects most food-contact notifications will be the subject of a categorical exclusion. In fact, to date, about 80 percent of effective notifications have a warranted claim of categorical exclusion. Anna Shanklin previously provided some statistical information about the notification program. The categorical exclusions have only been able to be used later in the process so that is why only 112 effective notifications are considered here. Of that number, 89 were excluded from the need to prepare an environmental assessment.
You also need to keep in mind that you make the claim of categorical exclusion but it is the agency that determines whether that exclusion is warranted.
[Slide.]
If FDA determines that, for a given claim of categorical exclusion is not warranted or extraordinary circumstances exist, then you would have to prepare an EA. The categorical exclusions which pertain to FDA's Center for Food Safety and Applied Nutrition in general are located in the CFR under 21 CFR 25.32.
[Slide.]
The exclusions which we feel are specific to food-contact notifications are 25.32(i), (j), (k), (q) and (r.) But, in reality, we have seen notifications using mostly 25.32(i) and (j) and there is one instance where 25.34 (q) was used and there is an effective notification using that. So I am going to just go into what the criteria for those three exclusions are, (i), (j) and (q), very briefly.
[Slide.]
If you look in the regulations, 21 CFR 25.32(i) is a categorical exclusion for a food-contact substance that is present in finished food-packaging material at not greater than 5 percent by weight. That food-contact substance is expected to remain with the finished food-packaging material through use by the consumer or that food-contact substance is a component of a coating of a finished food-packaging material.
So this exclusion basically applies for food-contact substances that are a component of a coating or that are in a packaging material at low levels.
[Slide.]
The circumstance for the exclusion under 25.32(j) is for a food-contact substance that is used as a component of a food-contact surface of permanent or semi-permanent equipment or of another food-contact article intended for repeat use. So this is essentially for food-contact substances used in food-processing equipment.
[Slide.]
The categorical exclusion under 21 CFR 28.32(q) if for a food-contact substance which is registered with the EPA under FIFRA for the same requested use. So this would be for an antimicrobial substance, for an example, that has a label that indicates a food-contact use identical for the use that is being notified for.
[Slide.]
In order to include your claim of categorical exclusion into your notification, there are three simple steps. You need to cite the section of the CFR for the exclusion you are claiming--for example, 21 CFR 25.32(i). You need to include a statement of compliance with the exclusion criteria and you need to state that no extraordinary circumstances exist that would require the submission of an EA.
[Slide.]
I am going to turn now to Part IV of Form 3480. The three simple steps are shown here. If you are making a claim of categorical exclusion under 21 CFR 25.38(i), you would just indicate that there. You would check, yes, affirm that what you are planning to do, your intended use under the notification meets the criteria and that, to the best of your knowledge, there are no extraordinary circumstances.
If you are making a claim of categorical exclusion, that is all you have to do as part of your notification, assuming it is true. But we will still evaluate it to see if you meet the criteria. Dr. Layla Batarseh is going to be speaking immediately after lunch. I think she is going to go into a little more detail about this aspect.
So, with regard to categorical exclusions, it can be as simple as that. Now I am going to go into a case where you would be submitting an environmental assessment.
[Slide.]
With regard to Form 3480, it is pretty simple. We just want you to include a statement that says, "We have attached an environmental assessment." That's it. Then, your appendix or attachment would include your environmental assessment.
[Slide.]
I want to discuss, first of all, what is an environmental assessment. It is defined under the NEPA regulations as a concise public document that serves to provide sufficient evidence and analysis for determining whether the agency should prepare a finding of no significant impact or an environmental impact statement. This, again, is defined in the NEPA regulations under 1508.9.
Again, going back to we don't want to prepare an environmental impact statement for every little thing so we can prepare a much shorter document, a environmental assessment. This will help us determine whether we need to prepare an environmental impact statement. If we don't, we just prepare a finding of no significant impact.
[Slide.]
The environmental assessments are prepared by the notifier, you, but often with recommendations for us, from us. We provide these recommendations because it is the agency that prepares the finding of no significant impact or it is the agency that prepares the environmental impact statement.
[Slide.]
I have a list of four general situations where there are not categorical exclusions which can be used and an environmental assessment is generally required. The first example is really--it is a petition type of situation, not a food-contact-substance situation, a macronutrient replacement that is expected to remain with the food. So it is not really relevant to this discussion here.
The other three examples are for a substance used in food production that is not intended to remain with the food, the general classification of chemicals here being the secondary direct-food additive. An example would be a boiler-water additive.
So, if you are submitting a food-contact notification for a boiler-water additive, chances are you are going to need an environmental assessment.
The second category with regard to food-contact notifications is processing agents used in food packaging that do not remain with the packaging. An example here might be a defoamer used in the manufacture of paper where it is not incorporated into the paper but the substance might be released to the environment.
A third example is for a component that is present in food packaging at greater than 5 percent. An example here might be a plasticizer that is used in a film at greater than 5 percent.
[Slide.]
As I mentioned before, an environmental assessment will be required in situations where a food-contact substance meets the exclusion criteria but where extraordinary circumstances exist.
[Slide.]
I just have identified some extraordinary circumstances as examples. These include actions for which there is potential for serious harm to the environment, actions that adversely affect endangered species, actions that threaten a violation of law, a unique emission circumstance or an action that may have significant effects on solid-waste management or resource and energy use.
An example here might be, say, for a plant extract, you have a plant extract that you want to notify for that isn't GRAS and you want to use it at a low level in your finished food-packaging for some reason at less than 5 percent.
You might think this qualifies for an exclusion under 35.32(i), but if the plant that you are extracting it from is on the endangered-species list, we would consider that an extraordinary circumstance and we might require an environmental assessment in that case.
[Slide.]
Getting back to actually preparing an environmental assessment, there are two general areas of analysis that we typically see included in environmental assessments for food-contact notifications. The first is with regard to chemical releases to the environment and the second is with regard to impacts of the food-contact substance on new food packaging material and resource and energy use, how the food-contact substance might affect the recyclability, other solid-waste management issues, land-fill and incineration and those sorts of things. So we typically see analyses that address these two areas.
[Slide.]
There are five general areas, topics, that an environmental assessment should address. We typically provide detailed guidance or recommendation for any notifier who needs an environmental assessment. But the general areas cover a discussion of a need for the action, a description of the introduction of the substance into the environment, information on the fate and effects of the substance or impacts to solid-waste management, alternatives to the proposed action and impacts of the proposed action such as how does the release of the chemical to the environment affect environmental receptors.
[Slide.]
There are no specific data requirements in the NEPA regulations but it is usually necessary that the notifier provide actual data from environmental testing or results from a prediction model to demonstrate that there is no significant impact. We recommend that you contact us at FDA before you consider any environmental testing. We don't want to see anyone do any unnecessary testing and there may be alternative means of getting the data you need to satisfy the requirements of an EA.
[Slide.]
The fate and effects portion of an environmental assessment is usually the most data-intensive portion. Some of the typical data we would see in the fate-and-effects portion would include chemical-property information about a chemical compound, information on depletion mechanisms for a specific chemical in the fate section, and then environmental effects data. Typically aquatic toxicology data is what we might look at.
[Slide.]
There are other ways to collect data. There might be alternative means to generate data for environmental assessment rather than through environmental testing. Specifically, EPA's Office of Pollution Prevention and Toxics is involved in the development of environmental-prediction software and they have several websites for which information is provided on a variety of those kinds of computer programs.
[Slide.]
EPA has also sponsored the development of estimate software for environmental fate and effects of chemicals which use structure-activity relationships. Again, EPA, in its contract with Syracuse Research, have websites that further describe the software.
[Slide.]
EPA has made available for free its suite of environmental fate modules referred to as EpiWin. The software will calculate a variety of environmental-fate parameters including hydrolysis half-life, octonol water-partition coefficient, Henry's law constant, soil-absorption coefficient and it will give a prediction of biodegradability.
The suite of software also includes a module for the prediction of ecotoxicological parameters and this module is called EcoSAR. This software will calculate acute and chronic toxicity concentrations for specific organisms. Again, this software was developed by EPA and a more in-depth discussion can be found at EPA's specific website about EcoSAR.
[Slide.]
I talked a little bit about the fate-and-effects data for the chemical releases portion of an environmental assessment. The second portion, the issue of data for dealing with resource and energy use, is going to be discussed this afternoon by Dr. Batarseh.
[Slide.]
So I am going to conclude with some summary comments about environmental requirements. Food-contact notifications are required to have an environmental assessment or a claim of categorical exclusion. If you do not include one or the other, your food-contact notification will not be accepted. The categorical exclusions which have been implied most frequently to food-contact notifications are the exclusion under 21 CFR 25.32(i) for additives to food packaging at low levels and 21 CFR 25.32(j) for substances used in the manufacturer of food-processing equipment.
An action which meets categorical-exclusion criteria may still require and environmental assessment if there are extraordinary circumstances associated with the action. When you are not sure about your environmental requirement for your food-contact notification, we recommend that you contact us and we will be happy to provide you with an environmental assessment or even to see whether one of the exclusions, one of the categorical exclusions, applies and what kind of information you need to supply to use the categorical exclusions.
That concludes my presentation. I would be happy to answer questions right now if there are questions about the environmental requirements for food-contact notifications.
LICKLY: Lori Lickly from Dow Chemical. Do we need to address constituents in the environmental assessment or just the food-contact substance, itself?
DeLEO: The constituents may need to be addressed. It sort of depends on a case-by-case basis. I believe, if you are submitting an exclusion, a claim of categorical exclusion, that would be for the food-contact substance. But, again, we might look at the constituents with regard to an extraordinary circumstance. If that constituent is particularly toxic or there is some issue related to it, we may have an interest in that.
Then, if you are submitting an environmental assessment, it may be necessary to submit data about the individual constituents. But, typically, I think the impurities or whatever, if it is such a low concentration compared to the food-contact substance, itself, we don't often see data regarding that.
Layla, would you like to respond?
BATARSEH: Yes. Layla Batarseh, the Environmental Review Group. As Paul explained, the focus of the environmental assessment is to make an assessment of the impact on organisms in the environment. So I would rather look at this issue in a different way. Rather than talk about constituents, talk about substances that are introduced into the environment.
The word "constituents" probably makes more sense in relation to toxicology and chemistry. So you would have to look at the substances that are introduced into the environment, and these are the chemicals that you would need to look at regarding how much and what impact.
DeLEO: Do you have another question, Lori?
LICKLY: Yes.
DeLEO: Go ahead.
LICKLY: It is sort of related. There were a couple of places in here where you said that the environmental assessment was a public document.
DeLEO: That's right.
LICKLY: So, if we have to discuss our manufacturing process in order to discuss environmental impacts, how do we go about doing that without disclosing it?
DeLEO: I guess I will say, in general, we have worked with notifiers and petition submitters so that we can accommodate both the issue of providing information that the public needs but keeping information that a petitioner, or submitter needs confidential confidential. We have done that on a case-by-case basis and I think we have been successful in meeting both areas.
Generally, at this point in time, with regard to how environmental assessments work, we don't request information on manufacturer of individual compounds. We look primarily at issues related to the use and disposal at the consumer end or the user end, as it were.
Are there other questions? Layla, would you like to make a comment?
BATARSEH: Yes. If you feel that there is some information that is important for the environmental assessment to help us know what the impact on the environment is, but you think that it is protected from disclosure, you can always include it in a confidential section and reference it in the EA.
When you submit that appendix, which is a confidential appendix, don't call it an environmental assessment. Just say, "confidential environmental information," so it will not be part of the environmental assessment.
DeLEO: I will remind you that we have our questions box if you want to include environmental questions for us. We are scheduled to reconvene at 1:30, so, at this time, if it is okay, I will declare a break for lunch.
[Whereupon, at 12:00 p.m., the proceedings were recessed to be resumed at 1:30 p.m.]
A F T E R N O O N P R O C E E D I N G S
[1:30 p.m.]
CHEESEMAN: Our first speaker this afternoon is Dr. Layla Batarseh who is the Supervisor over our Environmental Review Group in the Division of Chemistry Research and Environmental Review. Layla will be expanding on Paul DeLeo's talk on the environmental requirements of the Food Contact Notification Program.
Layla has also asked me to let you know that she will be departing, I'm sure to go back to work at the office, after her talk and so you should take the opportunity after her talk that she provides to ask any questions that you have.
Thank you.
[See presentation slides for Dr. Batarseh]
Environmental Workshop
BATARSEH: Good afternoon, everybody.
[Slide.]
I am going to be focusing on the categorical-exclusion requirements and I also will be focusing on the inadequacies that we see both in environmental assessments and in claims of categorical exclusion.
[Slide.]
As Paul said this morning, an FCS must contain a warranted claim of categorical exclusion or an environmental assessment. Therefore, we will not accept a food-contact notification for review if the environmental component is missing or is deficient.
[Slide.]
The categorical-exclusion applicability is-usually facially determined without requesting any additional information; you can focus on the use and find out whether the categorical exclusion is applicable or not. The CEQ's view is that information submitted in a request of categorical exclusion is usually enough for us to determine that the categorical exclusion is applicable.
Therefore, our categorical exclusion, or the categorical exclusions, were formulated to include specific criteria so that, in most cases, a categorical exclusion can be determined either facially or by reviewing other sections of the food-contact notification.
[Slide.]
However, and some of you might have FCNs in our shop where we ask you for additional information even though you submitted a claim and you said that the categorical exclusion is warranted. The reason is that, in limited instances, it may be necessary that FDA request additional information to establish, to the agency's satisfaction, that the criteria are met.
This is most common when we talk about categorical exclusions (i), (q) and (r). The information we request will help us decide if your requested use complies with the exclusion criteria.
[Slide.]
I am going to focus on two categorical exclusions. I am going to talk about (i) because that is the exclusion that is mostly requested by you. And then I am going to talk about (q) because of the misunderstanding of the applicability of (q). So these two present specific cases.
Exclusion (i) may apply when, one, the processing aid is present in finished food packaging at no greater than 5 percent by weight and, two, it is expected to remain with the finished food-packaging material through use by consumers. These two points are, I believe, written in the regulations. However, many notifiers miss the last point which says that the percentage of the processing aid that is incorporated into the finished food-packaging material should be greater than 95 percent.
What happens sometimes is that some food-contact substances will remain in the finished food-packaging material as probably contaminants or they would remain at very low percentage. Some notifiers would come to us and say, "Well, it is less than 5 percent in the finished food-packaging material."
However, you should always think of the percentage that remains with the finished food-packaging material in light of the total amount of the total market volume and, therefore, most of the market volume should remain with the food-packaging material such that, at the end, it will be not greater than 5 percent by weight of the food-packaging material.
[Slide.]
Now, the requirements might be different for the categorical exclusion (i) depending on the function of the food-contact substance. There is a difference whether the food-contact substance is functional or nonfunctional. So, when claiming this categorical exclusion--that is, (i)--no additional information will usually be needed when the substance remains with and functions in the finished food-packaging material because, obviously, if it is functioning in the finished food-packaging material, then most of it should remain with the food-packaging material.
However, for substances that have no function in the finished food-packaging material, such as processing aids, but that remain with the food-packaging material probably, or most probably, because of their chemistry, then we need you to submit an estimate of the percentage of the amount of the substance used that is incorporated in the finished food-packaging material.
[Slide.]
Categorical exclusion (q) is for substances that are registered by the EPA under FIFRA, mostly antimicrobials. They should be registered for the same use requested in the FCN. As I said, antimicrobials.
[Slide.]
Many people would ask what do you mean by "same use." We figured out an answer for that. Same use means that when comparing the food-additive use, when you come to us, FDA, and you compare that to the pesticide use, the following should be substantially identical.
The purpose of the use is the same, any components used with the substance are the same, the amount of the substance--that is the use level--should be the same, and the amount of any components used with the substance, all of these should be substantially identical.
[Slide.]
When you have categorical exclusion (q), you should submit to us two things. One is the current FIFRA registration which is to show us that, indeed, the substance is registered under FIFRA. Then you have to submit a proposed draft FIFRA registration that will reflect the changes that will take place in the label to reflect that this use is now regulated by the Food and Drug Administration.
[Slide.]
So, offhand, as Paul showed you this morning, the form is very simple for the environmental requirements, especially for categorical exclusion. All you need is to write the Section 21 CFR 25(i) and then you would say no extraordinary circumstances exist and that you meet the criteria.
But let me share with you the long list of inadequacies that we see in the claim of categorical exclusion.
[Slide.]
We receive some FCNs that have no environmental submission at all. I mean, they are totally silent on the environmental aspect. Second, the citation is for thewrong exclusion. They put (i), and the use is (j), or vice versa.
Incomplete claim; the three components are not all present in the claim of the categorical exclusion. And the claim lacks an explicit statement that the exclusion complies with exclusion criteria and there are no extraordinary circumstances.
This is your responsibility so you have to write that explicitly in the claim of categorical exclusion.
[Slide.]
Next, the exclusion is applicable only to one part of the use, or only to some uses, and not all. We see this often with categorical exclusion (i) and (j). Sometimes, you claim the exclusion under (i) and then, when we look somewhere else in the notification, it says, "For all kinds of polymer, for all food-contact uses." That, therefore, opens the possibility that some of the uses might be repeat food-contact uses. In that case, you would have to say (i) and (j).
You always have to look at the use and say, "How many claims do I need?" Usually, you need one claim. Then you would have to say (i) or (j). But there are times when both sections should be cited in the claim of categorical exclusion.
The use requested in the claim is different than the use requested in other sections of FCN. We see this less frequently now. We used to see this, in the past, when the notifiers did not have to use the form. When you get to that page of categorical exclusion, you would say what the use is, and you would state the section. Then, when we compared that use with the use requested at the beginning of the submission, we noticed that they were different. Now that you have to use the form, you don't really cite the use in the claim of categorical exclusion so this is not a problem, I should say, at the present time.
Last, there is insufficient evidence in the notification for us to determine if the exclusion criteria are met and, in this case, we have to ask for additional information.
[Slide.]
I am going to move to the second part of the talk which is the frequent inadequacies in the environmental assessment.
[Slide.]
Again, no environmental submission whatsoever. The use requested in the environmental assessment is different than the use requested elsewhere in the notification. Also, it lacks an explicit statement that there are no extraordinary circumstances that apply to the site of production.
As Paul said in his talk, we no longer ask for information about the production site. However, you would have to state that there are no extraordinary circumstances that would require the submission of information about the production site. Also, sometimes the EA is not independent of the submission.
In other words, you would say, under format item 5, which is identity of the food-contact substance, "See Chemistry, Section IV." That should not be the case because the environmental assessment should be a complete and independent document that should be available to the public to understand the whole story and understand the basis for the agency's decision.
[Slide.]
Next is that it does not include an estimate of environmental introduction concentrations. Again, the introductions we are talking about are not limited to the introduction of the food-contact substance itself only. The whole process of using the substance might lead to other introductions into the environment and these should also be looked at.
Next, it lacks a discussion of potential impact on solid-waste strategies. I want to stress here that when we talk about solid-waste strategies, we are talking about environmental impact resulting from the management of solid waste. So we are talking about land filling. We are talking about incineration. And we are talking about recycling.
Recycling is an issue that, I believe, over the years, the agency and industry have discussed and talked about. But the point I want to highlight here is that the concern is not the impact on the recycling, per se. Rather, it is the impact on the environment such as increased emissions from incinerators and such as increased landfill volume that would result from an impact on the recycling process. I will talk more about this in a couple of slides
The last part is, it contains confidential information. It should not contain confidential information, as Paul said this morning.
I think this is a good time for me to answer a question that I found in our question box, which says, "We submitted a FOIA request for all information for an FCN. We received no EA. It was all considered confidential. We know the EA was quite extensive. How can this be?" It should not be. It is incorrect."
What happens is you need to know that environmental assessments for food-additive petitions are made publicly available at the time of filing the petition. They are sent to the dockets. There is a comment period and the whole environmental record, is available to the public.
With the FCN, because even the idea that we have an FCN should be kept within the agency as confidential, we cannot make the environmental assessment available to the public at the time we received and FCN and up to the time that the FCN becomes effective.
After that time, the decision was made not to send the environmental assessment to the docket, rather to say that it is available if you want it. We should be able to provide you with the environmental assessment.
So, whoever wrote this question, please do contact me and I will look at the environmental assessment and, if you still need it, we have to make sure that you get a copy of the environmental assessment because it is a publicly available document.
[Slide.]
Now, I thought that this might be a place to talk just a little bit, just two or three slides, about resource and energy use. We have four kinds, as Paul mentioned this morning, of uses where you would need to submit an environmental assessment. One of these is when your proposed use is for a polymer that will be--or for a food-contact substance that will be greater than 5 percent by weight in the finished food-packaging material.
One of the questions I thought someone would ask me, but let me just touch on that question, is where is our guidance. I am sure many of you are wondering where is the environmental guidance. We have been working on the environmental guidance for all kinds of uses including the polymeric packaging material. Because of issues related to polymeric packaging material, we decided to focus on the other three guidance documents for processing secondary food additives and additives that are not macronutrient replacement.
These are out of our office. I cannot say when because, as you know, it has to go in a queue for people upper in the chain, or higher in the chain, to look at it. But, meanwhile, we have been providing guidance for you if you come and ask us for guidance.
Actually, we will tailor the guidance, or rather call them recommendations--we tailor them to fit your situation. So please do not hesitate to call us if you are at all in doubt what kind of information we request from you.
The other guidance, we call it the fourth guidance, is for polymeric packaging material. That will take some time to be published. Therefore, I encourage you to talk to us when you have a new polymeric material because if you look at previous EAs, as we said, they could be very extensive, you might not need to do all of that.
So, please, before you go ahead, get previous environmental assessments and try to do everything that is in those assessments. Please come and talk to us and then we will tell you what you need to do, what kind of information you need to submit.
But the basis for all of this, the basis for our--I should refrain from saying our authority to look at the polymeric material and its environment impact-- is actually stated in NEPA Section 102(b)(6). It states that one goal of NEPA is to enhance the quality of renewable resources and approach the maximum attainable recycling of depletable resources.
So we understand this to mean that we need to look at the environmental impact of new polymeric material since it might affect the resources and the energy.
For polymeric materials in food packaging, they might impact the environment as a result of disposal by affecting solid-waste-management practices. As I mentioned earlier, we are talking about recycling, land-filling and incineration.
[Slide.]
Therefore, if you have a new polymer and you want to use an FCN in the making of a new polymer where the concentration is going to be greater than 5 percent by weight, you need to prepare an environmental assessment. There are times--well, it hasn't happened yet, but there might be cases where, even if you believe that your submission or your requested use qualifies for categorical exclusion because the concentration in the finished food-packaging material is less than 5 percent by weight, if we have a basis to show that there are extraordinary circumstances, then we may ask you to submit an environmental assessment.
As I mentioned earlier, this hasn't happened yet but it is a possibility.
[Slide.]
Information needed to assess impact on the environment as a result of packaging disposal. Usually, we look closely at food packaging that is free-standing, rigid or semi-rigid containers, especially bottles. Information needed to make an assessment of the impact of a new polymeric material is the technical effect of the proposed additive, the estimated yearly market volume, and I stress here we consider it confidential and, therefore, usually you have to provide it elsewhere in the notification.
[Slide.]
Also, you need to provide the types of packaging you want to use this food-contact substance in; the way we look at films is different from the way we look at bottles. Size of the container is important, especially if you are making bottles because we use the size in our calculation of the energy consumption for that proposed use.
Then, five, intended food application. Is it a beverage? Is it alcoholic beverages? Is it soft-drink packages? Then, six, currently using packaging material that the proposed additive may compete with and replace.
[Slide.]
Now, I put two slides at the end. In your handout, there are two slides. I noticed this morning that one more slide is missing. So what I put here is my e-mail address. Please send me a note if you are interested in getting that last slide and I will certainly make sure I send it to you.
So that, in a sense, sums my presentation. If you have any questions, I will try to answer them.
Yes?
MADY: When we have our additive in plastic lower than 5 percent, we take advantage of the exclusion. However, we have a polymer that will be 30 percent of an ink material. But the ink will not be more than 5 percent of the actual package. Do we have to submit an environmental impact statement?
BATARSEH: If you have an FCN, you always have to submit an environmental component to it. One thing you need to look at is what is the substance that is the subject of the notification? In other words, once your notification becomes effective, what is the substance that is going to be the FCS in the listing of effective notifications?
You look at that and say, "Okay; this is my substance. What is the concentration of this substance in the finished food-packaging material?" So if the food-contact substance is not greater than 5 percent, then you qualify for categorical exclusion (i). But, if it is greater than 5 percent, then you need to submit an environmental assessment.
MADY: Let me ask the question differently. Is the 5 percent calculated based on where the material is to be sold into, like the ink, or the 5 percent based on the whole package of plastic?
BATARSEH: Finished food-packaging material. Definitely, yes.
MADY: Thank you.
BATARSEH: You're welcome. Anything more? I am going to leave now.
BINDRA: I am Amrit Bindra from Englhard. I have this question. You mentioned about this explicit statement, about the claim lacks explicit statement that the exclusion complies with the applicable criteria. Is sticking on that Form 3080--is that enough or do we have to--
BATARSEH: Yes, because you are saying yes to what the form is saying, that there are no extraordinary circumstances.
BINDRA: So we don't have to literally add that statement.
BATARSEH: No; you don't have to repeat that. You are agreeing, or your answer to that is explicit enough, that you meet the criteria.
BINDRA: If you don't mind explaining again. I am a little bit confused about that if it is less than 5 percent, then it is excluded and if it is more than 95 percent, it is excluded. I kind of missed that part.
BATARSEH: The way the exclusion is written in the CFR, it says, "It should be not greater than 5 percent by weight of the finished food-packaging material."
If you look at the preamble, it goes in more details what we mean by that--there is another point that we should have, actually, explicitly said in the CFR and that is most of the market volume should end up with the food-packaging material because there are situations where you use a processing aid, you add it at the head box when you are making paper and paperboard.
Then the finished food-packaging material, when it is time to submit your analysis and to meet the chemistry requirements, you would say, "Very little, if any at all, remains with the food-packaging material. It is less than 5 percent.
Let's say it is probably 1 percent. Having 1 percent in the finished food-packaging material is not the whole story. You have to make sure that most of the market volume that you added actually remains with the food-packaging material meaning that you didn't lose most of the market volume between adding the chemical to the head box and the time you have your paper ready because, for example, in the paper manufacturing, you go through the process of removing water and the drying.
So, if the market volume is 1,000 kilogram and then, as you add it, you lose most of it when you remove the water from the paper and very little remains with the food-packaging material, that amount that was left with the food-packaging material definitely will remain with food-packaging material through use by the consumer. But, remember that that amount that remained with the food-packaging material is a tiny fraction of the total market volume that you use.
So you have to look not only at what is left in the food-packaging material but compare that with what you started with. If what you started with, of total market volume, all of it, stayed with the packaging through the whole paper manufacturing process and ended up with the paper, then (i) qualifies.
But if you use 1,000 kilograms and 995 were lost during the process and only 5 kilograms remained with the finished food-packaging material, then (i) doesn't qualify. So you have to look at what remains with the packaging as a percentage of what you started with, and 95 percent or more should stay with the food packaging.
BINDRA: Let me just restate what I understand?
BATARSEH: Yes, please.
BINDRA: If we are making a product which goes into food packaging, we are using a processing additive in the manufacturing process, if more than 5 percent of that processing additive goes into the waste water, and less than 95 percent stays with the FCN substance, is that what it implies, that then we need this assessment?
BATARSEH: Yes; 95 percent remains with the packaging, then (i) qualifies.
BINDRA: Then we don't.
BATARSEH: Yes; then it qualifies.
BINDRA: Thank you.
BATARSEH: There was someone else?
McCOURT: Chuck McCourt from Dow Corning. A similar type of question, actually, but it has to do with if the processing aid ends up in, like, say the waste water but yet the amount of change based on the food-contact notification versus the other industry application for that use is small, can we state that in the environmental assessment, that the actual amount is going to be very small?
BATARSEH: You should first make an assessment of the introductions as a result of the proposed use. So you would say, "This is how much we expect to be introduced into the environment." Actually, it does make sense that you compare that with the total amount in the environment and say something to the effect that, "This change is not significant."
I know this is, like, a loose term, "not significant." How much should it be to become significant? But you have, I suppose, to use your best judgment and then we would work with you to see how we can best make a decision whether there is an impact or not.
But remember that, under NEPA, we have the responsibility to also look at cumulative impact. So the increase might be a small increase but when you add it to previous uses or current uses, then that might make the increase significant.
REED: I have a very similar question. Peter Reed from Ondewo Nalco. There are cases where you use a processing aid. Maybe 95 percent of it doesn't go, say, with the paper. But, in the waste water, that could go to, say, a cooling tower. It could go to evaporative train. In that case, they may use that particular polymer additive as a processing aid to that, as maybe a scale inhibitor.
So, in that case, it doesn't seem reasonable that there would be a significant environmental assessment.
BATARSEH: Exactly. The whole idea of an EA is to make an assessment of the impact of substances introduced into the environment--not only introduced; but that stay in the environment, that get the chance to get to the environment. There are times, of course, when industry would discharge waste into a waste-water treatment plant. So, even though you discharge that substance, it doesn't really end up in the environment because of the depletion factors. Therefore, that is exactly what you need to do is to look at other factors that remove the substance from the discharge or remove it from the environment and say that, "Even though it is in the discharge, it is not ending up in the environment."
If it is not ending up in the environment, then there is no environmental impact, or very little, if any at all. So all you need is just to follow the substance and see if it ends up in the environment.
But, also, I might add, it might end up in the environment but it might not have an effect on organisms in the environment because of dilution factors in the environment. You dump something into the waste-water treatment plant and then the waste-water treatment plant discharges into a river.
By that time, you have 10, 50, 100 times dilution factor. So if you start with 1 part per million, you might end up with 50 parts per billion. Therefore, you have to, in a sense, look at all the factors that would reduce, one, the introduction levels and, two, will reduce the exposure. Therefore, that will definitely reduce the toxicity to organisms.
But I really want to stress on the fact that you should call us. We are there to help you get us a complete notification. I am sure that my colleagues have stressed this this morning. You will hear it again in the afternoon. Do call us. Come to us. Say, "I'm not sure about this. How do you think I can handle this issue?" We will be very happy to help you.
Yes?
McCOURT: Chuck McCourt from Dow Corning again. Another point that you made was that you don't need any manufacturing information if we can state that there are no extraordinary circumstances?
BATARSEH: Correct.
McCOURT: That is based on that it is already approved by TSCA?
BATARSEH: Yes.
McCOURT: Okay.
BATARSEH: What we have looked at, obviously hundreds of petitions that resulted in hundreds of finding of no significant impact. We found that, almost in every case, there were enough environmental laws and regulations, either the Environmental Protection Agency or at the state level, that are set such that they would control the emissions and the discharge into the environment.
So we felt confident enough to say, "Well, let's depend on EPA or the state regulations unless you or we find extraordinary circumstances." In that case, we say, "Well, we need to look at this more closely." So tell us about the production.
I'm lucky. I am the first speaker after lunch. With that, I thank you very much for your attention.
We will next hear from Dr. Michael VanDerveer with the Chemistry Review Group.
[See presentation slides for Dr. VanDerveer]
Chemistry Workshop
VanDERVEER: I am Mike VanDerveer. I am a member of the Division of Food Contact Substance Notification Review and I was asked to assist in this probably because I am one of the senior reviewers and, thus, I have seen several years of notifications and have previously been part of the petition process.
[Slide.]
When I was first asked about this, it was said that I should talk about common chemistry deficiencies at our stakeholders' meeting. I was puzzled by exactly what a stakeholder is. I found out, through process of elimination, that, since we are not a business, we don't sell anything, we don't have customers. So, a stakeholder, to us, would be a customer to a business.
But when I first encountered it, I didn't know what a stakeholder was, so I went to the dictionary and looked it up. Do you know what a stakeholder is? According to Webster, it is, "A disinterested third-party who holds the stake in a wager and pays off the winner of the wager." I am not sure exactly how that applies here, but that is what I think of when I think of a stakeholder.
[Slide.]
One thing we have noticed and, since I am a veteran of the petition process, it seems as though the notifications we are getting assume that, since we have to review it quickly, we don't need as much information. This is an impression that we have.
Unfortunately, that is not the case because, for those of you new to the notification process, who have not used petitions, the petition process was quite iterative. We would review it. If there were deficiencies, we would send out a letter. Then the petitioner would come back with more information.
If it was okay, then it would get regulated. But then, sometimes, we would find more deficiencies, send out another letter, then go back and do more studies and then, a few months or a few years later, we might get around to regulation.
Apparently some of our stakeholders were not satisfied with that so we now have the notification process. We do the same thing but we do it faster so it is more important that we have all the information. If we don't have the information, we will ask that the material be withdrawn and resubmitted at a later date.
[Slide.]
In the notification process, we have these wonderful prenotification consultations. We have had comments received saying that sometimes what we say at the prenotification consultation doesn't agree with what comes out once the notification is submitted.
PNCs are a two-way process. You need to tell us everything before we can tell you what we think you should be doing. Two specific instances I can think of. One was someone didn't tell us about the catalyst used in the process and this catalyst decomposed into a potential carcinogen. They didn't mention that at the PNC. It didn't come out until the notification was submitted.
The solution to that was pretty easy. They just wound up changing to a different catalyst.
Or, say you are submitting something for a stabilizer and you are concerned about the decomposition product but you forget to tell us about a byproduct in the manufacture that has tox concerns. That is the type of thing. We need to know everything about the notification in the prenotification consultation before we can give you a good answer.
[Slide.]
This is from Form 3480, the chemistry information and the important thing is right here in the middle. "Include data to enable FDA to confirm your estimated daily intake." We do check your calculations. We don't necessarily check all of them, but we do check sample calculations.
We need enough data so that we can verify that the number you get agrees with what we think it should be. Now, this morning Dr. Arvidson went through the guidance documents emphasizing what we need and mentioning briefly why we need it. I am going to go through Form 3480, mentioning briefly what we need and talking more about why we need it.
[Slide.]
Starting then with identification; CAS name, CAS number, trade common name, other names. Can there be problems with this? Yes. We have had chemical names given to us that did not agree with the structure that was given. One thing I remember particularly was propanoic acid versus propenoic acid. Just a simple typo, most likely. This is something that, in the notification process, can be taken care of with a telephone call.
We don't necessarily need CAS names. We need some sort of systematic name so that we can take the name and come up with a structure from the name. Common names, also very useful. The X through the trade names is not because you shouldn't use trade names, but a trade name should not be the only identification of material.
[Slide.]
CAS numbers are useful for us to have. Apparently, we can't require them but much of our databases, most of our databases, are based on CAS numbers. It is useful if we have that. Make sure you have the right CAS number. It is very easy to transpose numbers. It is very easy to switch numbers. It is not uncommon for the CAS numbers to not agree with the chemical that is supplied.
Code names; this is a particular concern because you are doing studies, migration studies. You are doing toxicology studies. You don't want to use the name of the material for everything, so you are using code names. Say, you have a study from 1996. You are a different company now. You have a different owner now. They might have used a different code name.
If you use code names to identify the studies, migration studies, toxicology studies, please, right up front in the identity section, say, "The following code names have been used in studies on this substance," because if you have a study for something that is a code name that we don't know what it is and it doesn't match some of the other code names, you are going to get a call from us and we are going to have to ask about it.
Last on the list, we continue to get claims for proprietary information. "We can't tell you because it is proprietary." Well, we need to know. If you can't tell us, we will not be able to give you the notification. We will deny it or ask you to withdraw it.
How do you get around this. One of the questions that was submitted pertains to this. "If you are a supplier and there is a manufacturer who makes the material, how do you get the manufacturer information?" One of the best ways is the Food Additive Master File. The manufacturer would submit a master file identifying their manufacturing process and they would get a number for that master file. Then you would refer to that number of the master file.
You, as a supplier, will never see this. The manufacturer is protected from FOI. Getting information out of a master file is much more difficult than from a petitioner notification. I am a reviewer and not a regulator. But the master file is the way to go to solve many of those problems.
[Slide.]
Finishing up on the identification. It is nice to have a description and characterization, structure, that sort of thing. Make sure the structure matches the name. Describe, say what it is. If it is a polymer, come up with molecular-weight distributions.
As mentioned before, clearly label all spectra and chromatograms. There is nothing more frustrating than to have a nice chromatogram and have no idea what it is for. Or, if you submit NMR spectra, and we don't know what it is, it makes it a little tough.
One specific example of this came in not too many weeks ago. It was obviously an NMR spectrum, and there were two peaks on there about -105 and -110 parts per million. Now, I looked at this and I had never seen anything like this. I have been a chemist for a few years now. I went to the other chemists in the group and they had no idea what this was.
I looked at the molecule and I said, "Well, it is obviously not proton. There is no phosphorous in there. Is it silicon, an NMR-active nucleus. So I went to my old Silverstein and Bassler from the qualitative organic analysis days--the book is older than many people in this room, I'm sure--and found that, indeed, silicon 29 is active, spin one-half.
So I went to the web, did a little web search on silicon 29 NMR, found out it is a relatively new technique, found a sample database that told me what I needed to know; yes, this NMR is consistent with the structure that was proposed. But it would have saved a lot of time if that was, number one, labeled as silicon 29 NMR and, number two, some reference to some place I could find out about that.
Are there any questions on just basic identity of the material? Yes?
REED: Peter Reed, Ondewo Nalco. We often work with water-soluble polymers that may be registered as CAS numbers under a different form; in other words, the counter ion is different. Maybe it is the sodium form or the acid form. So if we don't have a CAS number, say, for a sodium salt but a CAS number exists for the acid form, how should we handle that?
VanDERVEER: I think just say that the CAS number is thus for this form, but we are using that form. One thing we have discovered is CAS numbers are not as unique as we might like to think, polymers especially. Often, there are two distinct names. We have two distinct CAS numbers and CAS names for them. One is named based on structure and one is named based on how it was manufactured.
So, now that we know that, it is not a problem.
[Slide.]
Manufacturing information; what we like to have with this is the complete recipe, the whole story, amounts, times, temperatures, equations, adjuvants, details of the purification steps and drying and finishing.
This is so that we can see that, yes, indeed, you had an excess of this reagent and probably all of this other stuff that we might be more concerned will be completely consumed. We can also see such things as you dried it long enough so that we don't have to worry about any benzine left over in the material.
If you don't provide full information, we will have to ask you about it. One case that came up recently was specifically tetrafluoroethylene was used to make the polymer and tetrafluoroethylene apparently is a suspected carcinogen. They did not do any studies for residual tetrafluoroethylene and they didn't have any details of the manufacturing process other than the fact that these things were polymerized.
So we asked, just a telephone call. They faxed in, quite quickly, a description. It turns out, in the final step of the process, there is an annealing stage where they take and heat stuff at 200 Celsius for six hours. I looked at that and I said, "Oh; we don't have to worry about tetrafluoroethylene. It is not going to be there anymore."
So the more information you have about the manufacturer, the better we are at being able to decide whether or not we have to worry about residuals.
Once again, master files. If you can't get manufacturing information from the manufacturer, have them submit a master file or, if you are a manufacturer, yourself, you can submit that information as a master file and refer to it.
[Slide.]
Impurities; this is important. To the best of your abilities, list the impurities such as residual solvents and monomers. Any side reactions or byproducts, the decomposition products and low-molecular-weight oligomers.
We have another question here about low-molecular-weight oligomers. "Oligomers under 1000 Daltons are of toxicology concern. How can you characterize something like this if you don't really know what it is?" To the best of your abilities, try to figure out what it is. We need to know what it is. We need to know how much of it can get into the diet with the low-molecular-weight oligomers, especially if you have a new polymer or a modification of an existing polymer.
Say you had a new diole or a new diacid for a polyester. What you should specifically do is try to look for low-molecular-weight oligomers unique to that new monomer, to the best of your abilities. This can be very though. I understand. But that is what we are interested in. If you have a basically regulated material, you are putting in a new chunk. Some of those new chunks are going to wind up in the oligomers. How much might they be, to the best of your abilities?
[Slide.]
Physical specifications. This reminds me of something else that came up in one of the questions that was submitted before the meeting. Form 3480 probably will not be adequate for all the of chemistry data unless you have a very simple antioxidant used in repeat-use articles or something like that because, generally, you are going to wind up with some sort of migration data and you probably will have to have attachments to it.
Form 3480 is evolving so maybe things will change in the future. But, for now, we are using this form.
[Slide.]
Polymers; the specifications. A molecular-weight profile if it is a polymer or melting point, boiling point, other types of things, molecular weight of normal materials.
If you have specifications, either as a polymer or specifications for an antioxidant, whatever, specify things like purity. We have had notifications come in and say it is 99 percent pure. That's nice. What is the other 1 percent? It is nice that it is 99 percent pure but we need to know everything. A telephone call and then they come back, "Oh, it's water." Fine. But, if you tell us that, it saves a few days or a few weeks.
The residual levels; are there residual levels of solvent in there? Are there residual levels of solvent in there? Are there analytical methods to specify the residual levels of solvents and such? Describe the analytical methods in detail because we need to be able to replicate the studies if need be. And include the data.
Now, we are not talking migration studies here. We are still just on specifications and purity of the material. Include your data for any measurements.
[Slide.]
So, if you are specifying the characteristics of your substance, are there any questions on that?
Okay.
[Slide.]
The intended use; the intended use should include a use level. You want to use an antioxidant at up to 3 percent by weight of polyethylene polymers. Fine. If you don't include a use level, we will ask about it and it will add a few days to the review.
What types of foods do you intend to use this with and under what conditions? We have had several that just do not have that information.
[Slide.]
I meant to mention this earlier. This is pictures of Form 3480 edited for our use here. This is a chunk beginning the migration studies. Indeed, migration studies are the biggest chunk of chemistry data and the most common source of deficiencies.
Here are some examples that are used here. We have had a suggestion that these examples are not really appropriate. I agree with that. There are just too many variables for the food-contact substances. You can't have one example that can cover all of the possibilities, but if you are unsure how to handle your calculations based on your use, prenotification consultation, telephone call, e-mail, whatever.
[Slide.]
Further down the list here; for repeat-use articles provide typical-use scenarios. We have had one particular question about repeat-use materials, ion-exchange resins. It doesn't say who asked it. I think I know who asked the question. We do not have separate protocols for ion-exchange resins other than in the repeat-use suggestions there is something about the maximum temperature, anticipated temperature, of use and maximum time of contact. So it depends on what your use is.
If you are running it through an ion-exchange column and it is only spending 15 minutes on the column at 50 degrees Celsius, then, presumably, your migration studies would be done for 15 minutes at 50 degrees Celsius.
Back in the manufacturing process; I knew there was another question. "The manufacturing process is constantly changed because of the need to optimize and improve. In theory, a new FCN should be done for each process change, but this is not practical." I agree; it is not practical. But I believe the way the law is written right now, if you change your manufacturing process, you will need to submit a new notification.
Am I correct on that, Mitch?
CHEESEMAN: Only if it results in substantive changes.
VanDERVEER: What is a substantive change?
CHEESEMAN: Usually, we encourage them to ask in a prenotification consultation.
[Slide.]
Intended technical effect. I believe we have been down this road before this morning. Is Jerry still here? We like to have information on technical effects primarily to make sure that the stuff you are putting into the polymer, or whatever, does what it is supposed to do because we are wasting a lot of time trying to figure out if the use is safe. We would like to make sure it is actually doing some good.
So you should include minimum amounts, a description of the effect and some sort of testing. I was hoping Jerry was still here because I wanted to tell him something from back a few years ago. This was back in the petition days. Someone submitted information on an antioxidant. They wanted to use up to 3 percent by weight of polyolefin polymers. I don't remember the specifics.
The exposure was a little on the high side so there were some problems with the toxicology concerns. But they had not submitted any technical-effect data. So we had them go back and do studies showing the minimum amount required, et cetera.
They found out that 1 percent by weight of the polymer, they got 94 percent of the efficacy they did at 3 percent by weight of the polymer. So, they just knocked it back up to 1 percent by weight, knocked the exposure down by a factor of 3. Everybody was happy except they couldn't sell quite as much of the stuff. So I suppose that was the downside on that.
[Slide.]
Stability; we have already heard mentioned that, if it is an antioxidant, you know it is going to decompose in the process. What are the breakdown products? Even if it is not an antioxidant, we like to have some information on how stable is the material, does it break down under the intended-use conditions.
[Slide.]
Now, the heart of the matter, the migration studies. The migration testing is the biggest part of most of these. These are just pictures of the sections on migration. Summarize the results of the test. For each test specimen, give individual and average migration values. The data should be presented--it is in there somewhere. We need to have the data to do the calculations. The number-one deficiency here that we have had with notifications, not so much back in the petition days, the migration studies submitted to not match the intended-use conditions.
You can't submit room-temperature migration studies for Condition of Use A. This has happened several times, or that type of thing has happened several times. Another problem we have been having is provide the composition of the specimen. You have a plaque of polymer with your antioxidant in it. Is there something else in there?
One specific example of this was a material we didn't know--it was PVC. Was it plasticized or not? Those of you in the polymer business know that plasticized PVC migration is much higher, orders of magnitude higher, than nonplasticized. They had done it in nonplasticized PVC so we had to restrict their use to nonplasticized PVC--or they agreed. We didn't have to--let me rephrase that. They agreed to restrict their use as we suggested they do.
Otherwise, we would have denied that particular notification.
Provide details of the analytical workup. This is complete details, all sorts of dilution factors and samples and that sort of thing. We just had one--I was talking to one of my colleagues on the telephone, when I had asked for deficiencies, they had notice.
If you take a 2 milliliter aliquot and dilute that to 10 milliliters, tell us about it because when we do the calculations, our number is going to come up five times bigger than your number and we are going to wonder why. And we are going to have to ask you and you are going to have to figure it out and it is going to take time.
So, give us the compete details of your analytical process.
[Slide.]
Migration testing continued; raw data. We have already discussed this once this morning. When we say raw data, we mean peak area for GCs, absorbances for other spectra and, indeed, it can be extensive. Tables of raw data are what I am used to seeing. A page-full of nothing but data for your migration studies.
We don't need notebook pages. That is not necessary unless we have any questions. We might ask for them later but, initially, just tables of the data.
Calibrations; now this is something that also came up in this morning's discussion. We like to see nice calibration curves but with some new instrumentation, we are getting into something that one of my old analytical professors talked about, the black box analytical chemist. You have a black box. You take your sample. You push a button and a number comes out and you have no idea what is going on in between.
This is part of something that we are seeing more and more of these days. We had a prenotification consultation where there were some atomic absorption studies and the calibration consisted of one point. I questioned this. It turns out, you buy the instrument from this company, you buy the calibration solutions from this company, you put the calibration solution into the instrument, you poke the button, it is calibrated.
You then run your samples. I am a little uncomfortable with that because what if the technician who was using this calibration solution purchased from the company spit into the bottle or something. It is no longer what the company said it was. So how do you validate the calibration solution that the company sent you? It bothers me a little bit.
What we like to see is nice lines, linear plots. By far the biggest problem with migration studies, is the validation, spiking and validation. Make sure you spike the extracts after you have removed the samples from the extracts. The most common problem here is people spike the extracts with sample plaques still in there.
You can get migration from the spiked solution back into the plaque. This goes back to one of the questions I had fairly frequently early on. People would do migration studies for Condition of Use A, high temperature, and they would be shocked to find that they get a relatively high number and then it would decrease at the lower temperature. They thought there was something wrong and they would keep redoing their studies, perfectly normal behavior.
At the high temperature, it migrates out. You lower the temperature, it migrates back. Don't worry about that. But the validation needs to be done, spiked appropriate and appropriate recoveries. The appropriate recoveries are discussed in the guidance documents.
[Slide.]
This is a typical calibration plot taken from a notification I was just reviewing. We like to see things like this. You have got five or six points there. It's great because we tend to take these numbers not for every calibration but we look at it and we do a least squares to get the correlation coefficient on it, make sure it matches what you got, make sure we get the same equation you got.
Now, one time there was a curve a lot like this except there was a point about there. I did the correlation coefficient and I came up with like a 0.92 or something like that. It is not really very great, but they reported 0.997 or something like that.
I said, "Wait a minute." This was back in the petition days. We had to write a letter and found out from them that, oh, they had ignored that point up there. Okay. Fine. Tell us that. That is all we need to know.
[Slide.]
Here is a typical sample chromatogram. We like to see chromatograms like this. Relatively flat baseline. Someone has even drawn in the baseline there and there is a little bit of a shoulder. I guess they have adjusted for the shoulder, but this is the type of thing we mean when we say sample chromatograms.
[Slide.]
Migration calculation options. What are the options on this? 100 percent migration. This should be your first choice if you are doing something like a repeat-use article or if you are doing something like an additive, a stabilizer in an antioxidant. We have seen things like this.
A tenth of a percent stabililizer in an antioxidant used at 1 percent by weight in can coatings that are very thin coatings. Maybe you can just do 100 percent migration, assume it all migrates out. You don't even have to bother with migration studies.
Or you can do migration modeling. Assuming Fickian diffusion and using the Peringer models, we have done quite well with this recently. It was just a short while ago we had a case where someone had very migration into the fatty-food simulants. It happens fairly often. So they had only done migration studies in fatty food. So we had to come up with migration into aqueous and acidic foods.
How did we do it? Modeling. Knocked the number down by at least an order of magnitude. I don't remember exactly what it was. But then I said, "Now, wait a minute. What would we get if we took this number and did migration modeling for the fatty-food simulant?
They had reported 27 parts per billion. I cranked it through and came up with 30 parts per billion. I was very pleased to see that. That is typical of what it does. The migration modeling typically overestimates migration usually by a little more than that. It is usually by a factor of about 2. But migration modeling is quite useful for studying, for doing, for determining exposures that are very difficult to do with standard modeling.
[Slide.]
End tests are not appropriate for safety studies. End tests, as Dr. Arvidson mentioned this morning, are only quality-controlled checks to make sure that the material you have meets some minimum specifications.
I will have to admit it has been a long time since we have had a notification with end tests instead of migration studies but we do get prenotification consultations, we do get inquiries, where people say, "I have this new polymer. I have done the end tests according to here. Is it okay to use this stuff?"
If it is a new polymer, no. If it is an existing regulated polymer, and the end tests show that you are in compliance with the regulations, then, yes. But if it is a new polymer, you need to do migration studies to come up with a safety assessment.
[Slide.]
I am realizing at this point there is one thing I did forget to mention. I am sure there are several things I forgot to mention, but sample calculations. We ask that you provide sample calculations. There is a reason for that. It is so we can easily get from your peak area to your milligrams per square inch. Also, it is very useful for you.
We had one case a few years back where I went through--I did my exposure estimate and I got a number that was about 200 times higher than the number the petitioner in this case had reported. After my many years trying to teach freshman chemistry, I prided myself in trying to figure out where the freshmen were going wrong in these calculations.
I could not figure out why this factor of about 200.
So, one of our deficiency letters went out and we said, "We get this number. You get that number. Could you explain this to us?" A few months later, we got back a fairly sheepish letter saying, "Oh; we mistakenly divided twice by the molecular weight, which was 189."
If they had included a sample calculation, they probably would have noticed that at that point.
[Slide.]
Final slide here. The bottom line. What does the Chemistry Review do? We come up with estimated daily intakes for the food-contact substance. Surprisingly, we have very few deficiencies at this point. If you get this far, you generally can crank the numbers through.
The only points I want to make on this aspect are if you have a very specific use and, indeed, many of the notifications are coming in with very specific uses, the general consumption-factor, food-type distribution factor, may not be the best way to come up with an exposure.
We have available to us USDA food-consumption data so if you are making a food packaging for a particular food type, we will come up with concentration of your migrant into the food simulant, assume it is that concentration in that food type and knowing how much of that food is consumed by the consumer in the diet, we can come up with an exposure that way.
The only recent deficiency I have had is someone had an antimicrobial for use in all polymers and they used a consumption factor of 0.05. 0.05 is appropriate for colorants in all polymers because of some strange history, but for antimicrobials, if it is all polymers, the consumption factor should have been 0.8. But that is a small point.
For repeat-use articles, I will mention that again. With repeat-use articles, if you can give us a reasonable story about how much food contacts that particular article over its given useful lifetime, we will generally accept your numbers.
That's all I have for this. Are there any questions?
MADY: When we conduct migration studies, our objective is to run one migration study that we will be able to submit to FDA and EU. Recently, EU has been accepting the FDA migration protocol as a worst-case migration number. The only part that the EU does not accept is the validation FDA protocol.
Basically, the simple difference is that FDA requires to spike at the end of the study where the EU requires to spike at the beginning of the study with no polymer, as you indicated in the effect of the polymer on the actual food-contact material.
Is it possible for the FDA--since the EU already took into consider accepting the migration protocol of FDA, is it possible FDA will reconsider validation based on the EU. It simply will really address one of the issues that is not clear in a lot of the submissions that we look at, is that the stability of the additive is not really addressed.
If you spike during the migration, that will address the stability of the additive through the migration, so you achieve both validation and stability in one step.
VanDERVEER: I should let you know that, for some reason, it seems like there is a speaker way in the back there that is broadcasting these voices. There are a lot of echos going on. I got the basic thrust of the question.
We are trying to harmonize the European and American things. We have made a lot of progress. We now have the same food simulants except they tend to use olive oil where we tend to suggest corn oil. We have the same time and temperature protocols.
We are working toward harmonization in that respect. The way things are here, I think it would be very tough to accept validation by spiking before the workup like that. I don't know.
But, to answer your question, yes; we are working toward harmonizing with the European Union but the validation is not yet one of those things.
Yes, sir?
BINDRA: Amrit Bindra from Engelhard. Do I understand right that we may not need to do the migration studies, we can just use the modeling?
VanDERVEER: That is a possibility; yes.
BINDRA: Would you recommend some kind of program that you would feel comfortable with?
VanDERVEER: I believe our latest guidance documents on the web there have links or have mention of the studies, the literature references, to the modeling. We also have in-house two experts on that in the lab that spend a lot of their time working on these migration models.
BINDRA: So if we have a new potential substance, then in the prenotification meeting, could we discuss that and how to go about it?
VanDERVEER: If you know the molecular weight and you know which polymers you are talking about, there are very good models that you can model them--
BINDRA: This is only for polymers or it is for all substances?
VanDERVEER: The migration is typically from a polymer. It is an additive in a polymer or the polymer, itself.
BINDRA: Right. So any substance, the modeling will kind of show how much it will migrate from a given polymer.
VanDERVEER: There are good models that are apparently very capable of calculating this completely. They are based on empirical studies. These are mathematical equations designed to experimental data.
One more question? Yes.
OLSON: Bill Olson, CFR Services. I have a question regarding exposure assessment for dry food contact, as was said this morning by Dr. Arvidson, there is no migration testing for dry food contact. It is my understanding the default concentration for dry food contact migration is 50 ppb which is a level higher than you often find in doing extraction studies into aqueous or fatty media.
Do you have any comments as to my understanding? Is that correct, and what we can do about it with respect to refining the migration potential to dry food.
VanDERVEER: Your impression is correct. There is a lot of history involved there. It goes back to the virtually-nil concept that we had, if you check forty years ago. Eventually, we needed a number to go with virtually nil. At the time, 50 parts per billion was chosen because that was an unimaginably low concentration.
Times have changed. But still, for dry foods, the assumption, I believe, still is 50 parts per billion.
I should mention that there is cake at the break. So I think one more question and then we better have our break. Yes?
KINNE: Patty Kinne, Ciba Specialty Chemicals. Is it still possible for pigments to use the solubility or insolubility of pigments?
VanDERVEER: To the best of my knowledge, we still take solubility data but I haven't seen any for many years now.
KINNE: If you wanted to end up with compliance of, for example, B through H, would you need to look at the solubility at 100 degrees C, then?
VanDERVEER: The highest temperature; yes.
KINNE: In the food-simulating solvents.
VanDERVEER: Yes.
KINNE: Thank you.
VanDERVEER: There is a break scheduled now. After the break, we will have a toxicology team up here.
[Break.]
CHEESEMAN: Our first speaker this afternoon is Dr. Mary Shackelford who is a toxicology in the Division of Food Contact Substance Notification Review.
Mary?
[See presentation slides for Dr. Shackelford]
Toxicology Workshop
SHACKELFORD: Good afternoon.
[Slide.]
In this Toxicology Workshop, I will discuss FDA Form 3480 Part III.
[Slide.]
I will first review some information which Dr. Young presented in the morning session. A complete notification should contain toxicology information. We recommend that you organize the toxicology information into a safety summary and a safety data package. FDA Form 3480 provides a structured format for communicating the safety summary.
In this form, there are four sections. There is one section for the Safety Narrative, two sections for listing of toxicology studies for the food-contact substance and constituents and a fourth section for description of other relevant information.
The safety data package includes: comprehensive toxicologic profile, original study reports of toxicity studies, published literature and appendices.
[Slide.]
The Safety Narrative is a concise summary of the scientific basis of the safety decision for the food-contact substance and its constituents. The following information on the food-contact substance and its constituents should be included in the Safety Narrative: the estimate of human consumption, the discussion of potential toxicity, conclusions regarding mutagenic and carcinogenic potential, and the calculation of the worst-case upper-bound lifetime risk levels for carcinogenic constituents.
[Slide.]
There is one Safety narrative per notification. It is located in Part III of FDA Form 3480 and is part of the Safety Summary.
[Slide.]
The Comprehensive Toxicological Profile is a summary and critical evaluation of all available toxicological information pertinent to the safety evaluation of the food-contact substance or its constituents. The CTP may include: the literature search, study summaries of published and unpublished data, a determination of the no-observed-effect level and ADI, risk assessment for carcinogenic constituents, and a bibliography.
[Slide.]
There is one Comprehensive Toxicology Profile for the food-contact substance and one for each migrating constituent. It is located in the Safety Data Package.
[Slide.]
Let us look at FDA Form 3480 Part III. There are four sections in Part III. I will just briefly review what is in this written area. The toxicology information requested in this part includes: the Safety Narrative which is in Section A, a tabulation of relevant safety studies on the food-contact substance which is in Section B, information about the potential carcinogenicity and toxicity of the constituents which is Section C, and Section D which is a brief description of any other relevant information not included in the other sections.
If an acceptable daily intake has been calculated by the Agency for the food-contact substance, do not complete the table for other relevant studies in Section B unless new data are available.
In the case the database is extensive for a constituent, contact FDA for guidance on preparing Section C.
Do not provide detailed study summaries in any of the sections below. Such summaries should be included in the Comprehensive Toxicology Profile of your notification.
[Slide.]
Now let's look at the four individual sections of Part III of FDA Form 3480. In Section A, provide your safety narrative. We have already described what we recommend should be in the safety narrative.
[Slide.]
Let us next look at Section B. Tabulate all of the relevant safety studies on the food-contact substance or food additive using the tables provided in Section B. Relevant safety studies include both unpublished and published genetic toxicity tests, carcinogenicity studies and other relevant safety studies.
Other relevant safety studies include: subchronic studies, chronic toxicity studies, reproduction and developmental toxicity studies and other specialized studies such as immunotoxicity and neurotoxicity studies. Metabolism, pharmacokinetic and epidemiological studies should also be included if relevant.
[Slide]
Let's look first at some blank tables in Section B and then some tables with information entered in the tables. This is the table for genetic toxicity studies for the food-contact substance or food additive.
[Slide.]
This is the table for other relevant studies on the food-contact substance or food additive. I would like to point out that there is an error in the table. This should read, "List the lowest no-observable effect level from each study." That will be changed in the form.
[Slide.]
This is the table for the carcinogenicity studies on the food-contact substance or food additive.
[Slide.]
Let's look at some examples of how information may be entered in the tables in Section B starting with the table on genetic toxicity studies. You will notice that the test substance may be entered as the CAS number or trade name. A summary of the study should be provided in the comprehensive toxicology profile and a full study report or published article should be provided in the notification. In the last column of the table, identify the locations in the notification of the corresponding summary and of the full study report or published article.
The name of the test system may be abbreviated. Please include individual tester strains for the Ames test.
In the example in this table, the concentrations tested were the same for each of the Salmonella tester strains with and without metabolic activation for the initial and confirmatory assays. In the mouse-lymphoma assay, all the dose levels used were the same under all conditions of the assay. Thus, all the dose levels utilized in the assay were entered in the table. If the concentrations were different for the different conditions of a genetic toxicity assay and the information does not fit in the box, then enter the range of concentrations that were tested.
In this table, I have also shown an entry for the in vivo chromosomal aberration study using rat bone marrow.
If the test substance were a solid compound, then the concentrations might be expressed as microgram per plate, microgram per ml or milligram per kilogram body weight.
In the example in the table, the test substance was a liquid and thus the reason that the concentrations were, for example, expressed as microliters per plate.
In this table, the results may be described as positive, negative or equivocal. The Ames test assay and the mouse-lymphoma assay are performed with and without metabolic activation. Therefore, I have entered this study detail as plus or minus S9.
[Slide.]
Let's next look at the table for other relevant safety studies on the food-contact substance or food additive. Notice again that the test substance needs to be identified. You also need to identify the location in the notification of the corresponding summary and of the full study report or published article.
In this table, I have listed a subchronic toxicity study in rats, a chronic toxicity study in monkeys, a two-year chronic study in rats, a two-generation reproduction study in rats. If there is a combined chronic toxicity and carcinogenicity study, then the chronic phase should be listed in this table.
Enter details of the study design such as duration of study, route of administration, species and sex, number of animals per group and dose levels. Notice that I have used some standard abbreviations when filling in the table. Male and female are abbreviated as M and F. A two-generation reproduction study may be abbreviated as two-gen-repro.
In this table, LOEL refers to the lowest observable effect level and NOEL refers to the no observable effect level. The dose levels for the LOEL and the NOEL should be expressed as milligram per kilogram body weight per day. If the NOEL cannot be established, write, "None."
[Slide]
Let us next look at the table for carcinogenicity studies on the food-contact substance or food additive. Notice that the test substance needs to be identified. You also need to identify the location in the notification of the corresponding summary and of the full study report or published article.
As I told you, there was a combined chronic toxicity and carcinogenicity study. In this table, I have listed the carcinogenicity study in rats. The chronic phase of the study was listed in the table for other relevant safety studies. Enter details of the study design such as duration of study, route of administration, species and sex, number of animals per group and dose levels. Abbreviations may be used.
Generally, only oral carcinogenicity studies should be listed, but carcinogenicity studies by routes other than oral administration should be included if systemic toxicity is observed.
Identify treatment-related neoplastic and/or non-neoplastic lesions. If no treatment-related lesions are identified in the table, indicate no in the table.
[Slide.]
Let us now focus on Section C for listing information about the potential carcinogenicity and toxicity of the constituents.
[Slide]
Tabulate all of the unpublished and published genetic toxicity tests and carcinogenicity tests for each of the constituents. This is the table for the genetic toxicity studies on the constituents.
[Slide]
This is the table for the carcinogenicity studies on the constituents.
[Slide]
Let's look at some examples of how information may be entered in the tables in Section C starting with the table on genetic toxicity studies. In this table, I have listed genetic toxicity studies on eight constituents. Please remember that you need to provide a literature search on the constituents as well as any new studies as indicated by the exposure level as described in the Toxicology Guidance.
You must also include a review and evaluation of the toxicity studies in the Comprehensive Toxicology Profile and tabulate the genetic toxicity and carcinogenicity studies in the tables in Section C. You will notice that the constituent which is tested in the study may be entered as the CAS number or acronym.
A summary of the study should be provided in the Comprehensive Toxicology Profile and a full study report or published article should be provided in the notification. In the last column of the table, identify the locations in the notification of the corresponding summary and of the full study report or published article.
The name of the test system may be abbreviated. Please include individual tester strains for the Ames test. Under concentrations or dose levels, specify the concentrations that were tested in the genetic toxicity assay. If the concentrations were different for various concentrations of the assay, then enter the range of concentrations that were tested.
In this table, the results may be described as positive, negative or equivocal.
[Slide]
Let's next look at the table for carcinogenicity studies on the constituents. Notice that the test substance may, again, be identified as CAS number or acronym. You also need to identify the location in the notification of the corresponding summary and of the full study report or published article. Enter details of the study design such as duration of study, route of administration, species and sex, number of animals per group and dose levels. Abbreviations may be used.
Generally, only oral carcinogenicity studies should be listed but carcinogenicity studies by routes other than oral administration should be included if systemic toxicity is observed.
Identify treatment-related neoplastic and/or non-neoplastic lesions. Notice that I have identified the neoplastic lesions by the organ in which they occurred.
[Slide]
In Section C, you will notice that there is a space provided after the table on the carcinogenicity studies for constituents. In this space, please provide information about whether any constituent is known to be a potent toxicant. If so, indicate the nature of the toxicity, the test system, and the dose level at which the toxicity was observed.
[Slide.]
Section D in Part III of FDA Form 3480 provides a space where other relevant toxicology data or information may be included. Provide a brief description of other data and information which are not mentioned above that are relevant to the safety evaluation. A description of the number and types of supportive studies and evaluation of the structural similarities of the food-contact substance or constituents to known toxicants are examples of the kind of information that might be appropriate to include in this section.
It would also be appropriate to mention in this section pertinent hazard or risk analyses from other regulatory agencies and/or international bodies. These may include IARC monographs, JECFA monographs, EPA, IRIS documents, etc.
[Slide.]
Part VI of FDA Form 3480 has a table for identifying attachment sheets. If there is not sufficient space provided in Sections A though D, then use an attachment for continuation sheets. If there is not enough space in the tables provided in Sections B and C, then create a table with the same information and specify this as an attachment.
[Slide.]
In summary, the toxicology information for a complete notification should include a safety summary and a safety data package. We have described how to complete Part III of FDA Form 3480. A safety data package should also be submitted for the food-contact substance and the constituents.
We will have questions at the end of all the presentations for this afternoon. The next speaker is Dr. Chingju Sheu.
SHEU: Thank you, Mary.
[See presentation slides for Dr. Sheu]
[Slide]
Let's turn our attention to genotoxicity testing.
[Slide]
As we know, genotoxicity testing is an essential component of safety assessment. Based on our FCN Guidance of April, 2002 it is based on exposure to the chemicals. For a cumulative exposure greater than 0.5 ppb, we recommend genotoxicity testing. At a level below or at 0.5 ppb, we do not ask for new testing, but we do need a literature search for any genotoxicity information available.
That means, we are still concerned even at that level, concerned with the genetox information.
[Slide]
In our FDA Redbook 2000, four tests are listed. As we know, there are many genotoxicity tests available using different organisms, testing different endpoints, like DNA damage is getting quite popular, and gene mutations or chromosome aberrations, both in vitro and in vivo.
The four tests recommended here are based on national and international expert consensus that they are, at present, the most reliable, useful tests.
As I mentioned, it is based on the exposure. At the lower exposure, at 50 or below ppb, we usually recommend two tests. The most commonly used test is the Ames test. That, number one, is a test for gene mutation in bacteria. The second test is a choice. Some prefer to test for chromosomal aberrations. Some test for gene mutations. Either one is acceptable.
At a higher exposure above 50 ppb, in case of a biocide, it is 10 ppb, we recommend an addition, the third test. Usually it is the in vivo test, a micronucleus test. But any other chromosomal aberration test is all right.
[Slide]
How do we do the genotoxicity testing? In our FDA Redbook 2000, we have a detailed description of the one we prefer. But, actually, it is very similar or almost identical to the one published by EPA in their guidelines. It is also very similar to OECD guideline or ICH guideline. So they are all acceptable.
[Slide]
This is the final question. What do we use the genotoxicity test for, or the data? How do we use it? Especially in the absence of bioassay data, genetox tests have become very important. In the presence of a bioassay, the genetox data will serve as the supporting evidence.
[Slide.]
In the absence of a bioassay, we would have to see if the compound is genotoxic. As people who are familiar with the genotoxicity assessment know that it is not an easy plus or minus because we are testing different endpoints. We could come up with positive or negative or equivocal. It is not easy to just come up with plus or minus.
So, in our assessment, we will look at all the available information. That means, the studies provided by the notifiers. We also look at all the information we have in-house in our databases. We also search the literature for any published reports on genotoxicity. We will look at all the information and come up with an assessment.
If the assessment turns out to be negative, then no further action is taken. But if the assessment comes up equivocal or positive, then further action is required. This subject will be covered by the next three speakers.
But, at this point, I would like to answer one question raised earlier. This one says, "If a known constituent is a suspect mutagen but it is not detectable at 0.50 ppb and a lower detection is not possible, what is the position of the FDA?"
There are two key points in this question. The first point is "it is not detectable at 0.5 ppb" I would take it as meaning that the detection limit is 0.5 ppb. That means new testing is not required. But we do require a literature search. That means we are still concerned with the genetox assays. That is key point one.
The second point, "it is a suspect mutagen." That means you must have seen some genetox tests or reports indicating it is a potential mutagen. So, based on these two points, I will say this compound could be equivocal or positive. That means we will require additional evaluation. So that is my response.
Any comment on that?
This is the end of my talk. The next speaker is Dr. Chanderbhan. He will discuss the question raised with equivocal or positive genetox response.
[See presentation slides for Dr. Chanderbahn]
CHANDERBHAN: It is terrible coming after lunch but it is even worse coming after cake and cookies. I sympathize. I will try to make it interesting or lively.
[Slide.]
My name is Ronald Chanderbhan. I am with one of the tox groups in the Division of Food Contact Substance Notification Review.
To expand on Dr. Sheu's comment about need for further evaluation, we can ask the question, when faced with borderline, ambiguous or troublesome findings, what can we do? We can kick it under the rug, forget about it and go on to another test. We could flip a coin. We could, perish the thought, ask for more information from the notifier, more data, more testing.
[Slide.]
Or we can utilize quantitative structure activity relationships to help us answer these troublesome questions or to help us to reach a safety decision.
What I would like to do in the next few minutes is to present to you a couple of the QSAR systems that we use currently and to discuss very generally their advantages and disadvantages. The speakers that will follow after me will delve into a little more detail on the use of SAR.
[Slide.]
Having said that, I have got to throw this at you, and that is that QSAR analysis results are currently not used as the sole criteria for a safety decision. I have to emphasize, it is part of the whole package of safety information.
I also say currently, because I don't want to discount that, perhaps, some time in the future, we can have decisions made just on QSAR. As QSAR gets better, perhaps the day will come when maybe, for some types of exposures low enough, QSAR alone may be good enough. But, currently, it is used as part of the whole package.
[Slide.]
Here are a couple of hypothetical cases. One, Case 1; exposure estimate greater than 0.5 ppb less than 50 ppb. There are four genetox tests submitted; two are negative, two tests are equivocal. Case 2 could be an exposure estimate that is approximately 0.3 ppb. The notifier submitted an extensive literature search and, within that literature search, we found that a few publications indicate some potential for the substance being a mutagen. So this is where QSAR could be useful.
[Slide.]
Here are a couple of the QSAR systems. First, OncoLogic and, second, Multicase Expert System.
[Slide.]
First, I would like to just, again as I said before, give you generally what are the advantages and disadvantages of these two systems.
[Slide.]
The first is OncoLogic. It is an expert rule-based system developed by LogicChem in collaboration with the EPA's Office of Pollution Prevention and Toxics. It is used mainly to predict carcinogenicity. It is based on the chemical conduction of cancer monographs. You have a choice of four modules; namely, fibers, metals or metal-containing compounds, polymers and organics.
[Slide.]
The advantages of OncoLogic; it is relatively easy to use. The computer resources needed for running OncoLogic are not extensive. We also have in-house expertise. In saying that it is easy to use, I have to add to that that sometimes it can be difficult in terms of getting structures in, how to classify a structure chemically. We have in-house expertise that can help us with that, people in the Chemistry Section, et cetera.
It also is a decision-tree process presented in the justification as well as you get the reference chemicals when you get that decision.
[Slide.]
Some of the disadvantages. The concern level is qualitative. It is concrete. The modules cannot be user modified. The structures are not always modifiable. If you cannot enter your structure, it is not very clear why it won't accept your structure. Finally, carcinogenicity is the only endpoint.
[Slide.]
MCASE is a computer-automated structure-evaluation technology developed by Dr. Klopman at Case Western. It is a VAX-based problem and, again, requires extensive--it is one of the--I will get to that. It uses a training set to break molecules into all possible fragments to identify structural alerts.
[Slide.]
It uses statistical methods to identify fragments as biophores. It presents data in scaler units with a range from 10 to 90, 10 being the least alerting. There are several modules in MCASE that include mutagenicity, carcinogenicity and teratology.
[Slide.]
Some of the advantages; it is easy to draw the structures that you need to enter into it. The database or the training set can be expanded and/or contracted. The modules can also be expanded.
[Slide.]
Some of the disadvantages and limitations. The computer-requirements are extensive. The end values do not lend themselves to quantitative risk assessment. It doesn't give you the unit cancer risk or a TD50 value.
[Slide.]
Continuing with the disadvantages. It gives unitless case units which are difficult to interpret in a quantitative manner and there is no mechanistic significant to the fragments, although, retrospectively, the user can go back and, from the identified biophores, come up with some kind of mechanism of action.
The molecular weights have to be less than 5000 Daltons and organometallic molecules cannot be tested in MCASE.
[Slide.]
With this in mind, some of these disadvantages; MCASE-ES was developed to try and address some of those disadvantages. It was developed at FDA at CBER, which is the drug section. It contains expanded training sets. It offers broader coverage than the original MCASE and it is designed to filter out single-sex species carcinogens. It focuses more on trans-species, trans-gender, carcinogens.
[Slide.]
So, a validation of this system was done and the constraints used are a virtually safe dose that is equivalent to one in a million. A test set of 643 compounds was used and that was taken from the carcinogenicity potency database or more commonly called the Gold database. Oral administration was one of the criteria, a TD50 with a significance level less than or equal to 0.01, and it had to have a definable structure.
Modules were run for predicting carcinogenicity in male and female rats and mice and a cutoff was defined as a combined score of 100 case units.
[Slide.]
Here is a graph. It is kind of busy. The take-home message basically that I want to get across here is that even at the least-potent compounds, we still have about an 80 percent correct prediction rate. Of course, at the far left end of the scale, at the more potent compounds, you have a better predictivity.
[Slide.]
Future directions? We will continue to upgrade MCASE training sets and modules. We will continue to evaluate other QSAR systems as they become available and that we think are applicable to our needs.
Basically, that is end of my talk but, before I leave, I would like to leave you with a couple of words of wisdom or aphorisms, as you might call it, for those times when you take a break from tearing your hair out while you are looking at all that positive genetox data.
I particularly enjoy this one. "The great tragedy of science is the slaying of a beautiful hypothesis by an ugly fact."
The next, which I think is applicable, is that "Science has promised us truth. It has never promised us either peace or happiness."
Thank you for your attention. At this time, I would like to introduce Dr. Nancy Braier who will give you a little more detail about the chemistry aspects of SAR.
[See presentation slides for Dr. Braier]
BRAIER: Thanks, Ron. Good afternoon.
[Slide.]
I just first start by saying that I am the adopted child of the Toxicology Group. I am a chemist, a review chemist, in Food Contact Notification.
[Slide.]
We have put this talk in here because, in order for us to do our structure-activity relationships, there is one thing that always comes around; we need to do structures. We need a chemical structure. A structure-activity begins with a representative chemical structure.
Why do we need that structure? We need that structure to determine the structural analogues. We need the structure to search for information. You are asked to conduct searches. Most of the searches that are nowadays conducted are text-based searches. We want to expand and add a new dimension to those searches and that dimension is structure.
When you add structure, then you have a name. But, sometimes, as we know, there are many ways to name a compound. There is a structure and there is a similar structure that is really a very good identifier for that compound and similar compounds.
We also like to start mining our in-house databases, so we have started a project to try to put structure connected to the information we have to be able to search that information in yet another, more efficient, way.
[Slide.]
As you have seen from the two programs that Ron has showed you, you need that information also at the input of the structure-activity software programs. Dr. Chanderbhan discussed already two of the programs we have available, MCASE, MCASE-ES, and OncoLogic.
We have many other programs available or that we are testing. We don't use all of them, but we do keep our eye out for everything that comes out, and we ask questions about it. When we decide to use a program, and these are some examples, and there are many numerous examples out there, we talk to the people who make those programs because we want to know what the coverage is.
That is a point that is very important for us. As you see, we do use two expert systems. The reason that we use two expert systems is because the coverage in MCASE is different from the coverage in OncoLogic. So, when you put the two systems together, we get a broader picture of the kind of molecules that we can actually study and do SAR with.
[Slide.]
By now, I hope you are familiar with Phase 1. Phase 1 is the process prior to us sending you, or talking to you, about your deficiencies of the acknowledgment letter. At this stage, SAR is used to complement the available data.
As I said, the kind of software that is used in Phase 1 analysis is the expert systems and it is also structural databases. The expert systems, we have already talked about, MCASE and OncoLogic. We also use a program called TOXSYS which is a database, a structural database, search program.
What are the results? What do we give in SAR? The SAR resource for Phase 1 provides a prediction. In this case, you heard already, it is not a number. I am a theoretician. I have always been asking for, "Give me a number." Here, all we can give you is maybe yes, maybe no.
A point to the wise. Maybe yes usually has a little bit more weight, and I will explain myself with that.
If you actually look at the way the results come, and I will probably quote from OncoLogic, what it says. "No biophores found is presumed to be inactive." You have to make sure; presumed to be inactive doesn't mean that it is inactive.
So, as we have been telling you all day, if you do any kind of SAR analysis, give us the complete story. Because, if this compound is presumed to be inactive but there is a line there that says, "We found all these biophores in the compound," and all these biophores are actually found in compounds that are not active, that gives a lot more weight.
So, when you are using this information, if you decide to send some of the SAR information, make sure that you give us as much information as possible.
What do we do with the results of the SAR analysis? As Dr. Chanderbhan said, they are not the sole tool that we use to make decisions. We, the SAR team, give these results to the FCN Review Team and they, in conjunction with all the other information they have, make the decision.
[Slide.]
Now, as I said, aside from the expert system, we do have some database-based systems. TOXSYS is a comprehensive toxicological information system. It actually contains information on over 230,000 compounds. It has literature references so, if we find something in TOXSYS, we can give it back to the team and they can look for that information.
It has chemical structure, which is the way we search, although we can search by many other endpoints or CAS number or whatsoever. It is a compilation of RTECS records and it also includes food-additive information from the PAFA database.
The databases in TOXSYS, all this information is actually divided into different modules. These are examples of the modules, the modules that are underlined, the mutagenicity, the carcinogenicity and the food-additive modules are the modules that I usually use to perform my searches.
[Slide.]
What about Phase 2? Now we are really into trying to figure out what is happening here. Now, we are into, "Give me a number." So this is what we do. We try to do, in Phase 2, a theoretical risk analysis. The software we use, we use in-house structural databases. What do we use? To be able to do a Phase 2 analysis, we want good-quality data. So we are going to use what we consider good-quality data.
What is this software used for? It is used for the analogue search. It is used for selecting the analogue or the analogues that are going to be used to do the prediction. The SAR results, at this time, are an estimate of the cancer risk associated with the requested substance. Here I will say we will give an estimate if and only if and when we find that the data that is available on that compound is adequate.
At this time, we communicate our results in a written memorandum to the review team.
[Slide.]
From now on, I am just going to into the chemistry of how do we--I am going to give you a working example of how do we determine what a suitable analogue is.
There are two kinds of analogues that we use when we are doing a Phase 2 analysis. One is a structure analogue. As the name says, in a structural analogue, compounds have overall similar structures.
We also do searches on functional analogues. This is when we know what the target is. We know that there are parts of this molecule that have similar functionality and are the ones who are probably most--we believe that those are the parts that are actually giving the properties to the molecule, so we search for other molecules that have that functionality.
[Slide.]
I am going to take you now to a working example of how do we do a prediction on an molecule. The example I am going to use is 2-ethylhexyl acrylate. I will not give you the tox aspects of it. That will be the next talk. I will just guide you through the selection of the analogues for this compound.
In order to do our search, we use a structurable, searchable version of the carcinogenic potency database, or CPDB, and the reference is below. You can always find the data in the book. The search criterion for this structure; we use structure and we use similarity.
Structure means that we are looking for identical parts of the molecule. Similarity means that we are looking for molecules that, on the average, look like this molecule but are not exactly the same molecule.
So, what do we do? We use the molecule, the target molecule, 2-EHA, and we also use some substructures to make the search. We will require, in our search for this example, at least 80 percent similarity. You may say, "You know, that may not be so similar." You are absolutely right. What we are trying to find is caught with a broader coverage. We want to get the big picture.
We will go through the elimination process but let us eliminate the molecules that we are seeing are not reasonable.
[Slide.]
So, the initial set for this search included structure and substructure searches. We went through the CPDB database and we found 91 hits. What did we do? The SAR team went through the compounds and did some obvious elimination, looked at the size of this molecule. Larger molecules were eliminated. The molecule that we are searching for is nonaromatic, so most aromatics were eliminated.
Then we looked for mechanistic reasons. The joy of working in the office that we work at is that we are composed of many different people with many different backgrounds. We have found very useful that there is always an expert or somebody who works on X or Y compound and knows the history that can help us and we do use every resource available to be able to make an educated decision. So our final set was reduced to only six molecules.
What do they look like? There were phthalate-like molecules and they were acrylate-like molecules. This is how we did the search. We not only used the entire molecule. We did this substructure search for a part of the molecule and we also searched for small molecules. What we are trying to do here is to cover as much of the molecule as we can, really narrow the focus but also, keeping in the big picture, we also used the entire molecule.
[Slide.]
So what did the finalists look like? This, in the center, is our molecule. The first molecule you can see--I am going to guide you to the four molecules in the phthalate-like family, and you will see the parts of the molecule that actually resemble the target molecule.
These phthalates actually mechanistically behave--we believe that, potentially, can behave as the 2-EHA, so that is what--there is an aromatic molecule here but there is also a part of the molecule that can be found in each compound. Again, the allyl compound will give us even a broader coverage. It will start getting into the chain of the 2-EHA.
Most of the molecules, you can actually find in the DEHP compound and in the DEHA compound. So what we end up is with enough compounds to actually cover, or try to do our best to cover, the entire structure of the molecule. If you actually look at it, all these circles have just totally gone and covered the entire molecule, so we have a good coverage.
So, now we have found the molecules for this family. Let's go to the other family.
[Slide.]
This is the acrylate part of the molecule. You can see, again, and I am going to go quickly through this, that we have found, in these compounds, parts of the molecule that resemble our target molecule.
[Slide.]
So what do we do now? We have found the analogues. We have our database and we are going to use that do to a risk analysis. That is the subject of the next talk by Dr. Michelle Twaroski.
[See presentation slides for Dr. Twaroski]
TWAROSKI: Thank you, Nancy. Before I begin, my talk, I wanted to go ahead and I would like to introduce myself.
[Slide.]
I have been with the FDA for two years after completing following the completion of my doctorate degree in toxicology. I would be remiss to stand up here and take all the credit for all the institutional knowledge that comes with giving this talk, so I wanted want to make sure to and thank Dr. Lin and Dr. Sheu, Dr. Layla Batarseh, for an opportunity to learn SAR from EPA, Dr. Cheeseman, as well as my colleagues in the SAR group who have spoken today, Drs. Braier and Chanderbhan.
[Slide.]
I want to give you an overview of what I am going to talk about. First, I'll review the pertinent aspects of food and drug law to try to give some guidelines for what kinds of constraints constraints we are under when we evaluate potential carcinogens.
Second, I will go into carcinogenic risk assessment in a concrete, abstract analysis with data.
I will then go on to talk about some scenarios where we have mutagenicity data and structure-activity data, more along the lines of problems that Drs. Sheu, Chanderbhan and Braier have mentioned. Within that, hopefully, that will cover some issues of problem resolution.
[Slide.]
So what kinds of constraints are we under? If you look at the Federal Food, Drug and Cosmetic Act, the Delaney Clause or the Food Additives Anticancer Clause, provides that no additive shall be deemed safe if it is found to induce cancer when ingested by man or animal. This is a pretty black-and-white situation. [Slide.]
The courts have interpreted this law to mean that it applies to the additive, itself, and not to constituents of the additive. This is where we have a little leeway to work with you, the notifier.
The way the courts have interpreted the law is that constituents of the additive can be evaluated under the general safety standard using risk assessment procedures.
[Slide.]
I'll now go into a little bit of the history of how we evaluated constituents as well as food additives for carcinogenicity and then how we are doing that today. If we look at the committees that have been set up to handle this, there are two that come to mind that were established in the early to mid '80s: the Cancer Assessment Committee and the Quantitative Risk Assessment Committee. These committees feed into each other, the first committee deciding whether or not something is or is not a carcinogen and the second committee giving the quantitative answer as to the virtually safe dose or the unit cancer risks that can be derived from that virtually safe dose calculation.
The Cancer Assessment Committee that was established is made up of numerous experts including mathematicians, pathologists, statisticians, chemists as well as toxicologists.
Before the Food Contact Notification Program came into inception in the late '90s, OFAS--at that time, it was the Office of Premarket Approval--was trying to determine how to come to the answer of making a safety decision and a regulatory decision without actually making a decision of carcinogenicity which was set in the responsibilities is historically the responsibility of the CAC.
That is where the type of analysis that I will be talking about was conceived, and that is, can you make a decision as to whether something is safe, taking the worst-case situation without actually making a decision that, yes, something is a carcinogen or no, it is not? .
[Slide.]
So this is the question that we are really trying to answer, does the chemical have the potential for carcinogenic activity? As Drs. Sheu, and Drs. Chanderbhan and Braier have mentioned, if there is no evidence of mutagenicity or structural alerts, then the answer to this question is simply, no. Although you might have other toxicological issues, this makes for a very easy FCN as far as carcinogenicity goes.
Alternatively, the answer could be yes. Here, just for simplicity, I have listed three scenarios that could lead to that answer. One is you have neoplastic data that demonstrates that the chemical can cause tumors. The other is that you have non-negative mutagenicity data. Since we are asking for several tests, and they do not all test the same parameter, you could have conflicting reports and that is why I have listed this as non-negative. You could have positive results in one strain of the Ames assay that would lead you to think it is possibly a mutagen.
The other case is where SAR comes into play, and that is structural alerts. As I said, this is a more simplistic model to try to capture risk assessment for this broad audience, but I wanted to point out that you may also have functional alerts in which I am talking about alters in the manner of a 90-day study that shows increased relative liver rate, increased peroxisome number, increased enzyme activity that might lead you to believe you have a peroxisome proliferator which would be a non genotoxic carcinogen, for example.
But, keep that in mind when you consider structural alerts as well, that you could have structural or functional alerts that you might be able to resolve through SAR pathways.
[Slide.]
Going on to the question of carcinogenicity in the most concrete manner would be neoplastic data that is available on the chemical. The reason I have this separate is we all know that there are both nongenotoxic as well as genotoxic carcinogens, so these things (boxes presented) don't always cross.
So, if you have neoplastic data available, you can calculate a unit cancer risk and an upper-bound lifetime cancer risk.
[Slide.]
I want to go into a little of depth as to what is required there and then an example of how you would do this. When you submit data, and you need to calculate a carcinogenic risk assessment using neoplastic data--and here I am talking again about constituents and not food additive themselves, which are covered under Delaney. When you submit, you need to review the data and submit the data in full.
What I mean by "in full" is include all the appendices. These are important for reviewing the full submission and we do look at all the appendices to review the data. So make sure that this is included so that it doesn't come up thirty days into the submission at Phase 1 that that is missing and then, as you know, the ten days, and it goes on and on. So it is best to go ahead and submit this at the beginning.
You need to conduct a risk assessment using the linearized multistage model. There are other models out there, but this is the model that we are currently using. Calculate the upper-bound lifetime cancer risk based on your exposure.
I have mentioned specifically upper-bound lifetime cancer risk because, a lot of times we have been seeing the virtually safe dose. Although this does give you an indication of how far away your exposure is from giving a the virtually safe dose, it is not the parameter that we are necessarily looking for.
We are looking more for the actual risk for the requested use. So keep that in mind, that we would prefer to see the upper-bound lifetime cancer risk calculated from your exposure.
[Slide.]
As I said, instead of saying that something is or is not a carcinogen, we are, instead, operating under a conservative conclusion that the chemical, if it shows positive neoplastic data, is a potential human carcinogen and an assessment to human risk is essential for the safety review.
[Slide.]
When you submit your bioassay data and your summary of that data, I wanted to point out a couple of pivotal findings that you should be sure to include. You want to include all nonneoplastic and neoplastic lesions. Be sure to emphasize progressive lesions. For example, if you see hyperplasia of the liver, then you see benign tumors of the liver, then you malignant tumors of the liver.
You would want to mention that there is clearly a liver insult because, although, as I mentioned, we do not necessarily go to a formal CAC, we do still consult with pathologists on occasion when necessary as well as statisticians to get to review the data and we will be looking at whether or not there are progressive lesions present in your study.
You want to specify all neoplastic findings and you want to tabulate those by sex and by site with details of the statistical significance.
I have an example here. When I say details of the statistical significance, you just want to indicate which findings were statistically significant, what the p-value cutoff might have been for the overall study and what type of tests were done, whether that is logistic regression or pairwise comparisons, you would want to indicate that early on in the tabular format for ease of doing a rough calculation at the Phase 1.
[Slide.]
How we are defining the unit cancer risk at OFAS currently is we are defining it as the sum of the slopes of lines drawn from the lowest apparent effective dose of the chemical through zero for each tumor site.
This definition incorporates a lot of assumptions and I have listed at least two of them below, and that is that tumors arising from multiple sites are assumed to be independent and are added to obtain the overall unit cancer risk. What this means is, although you might have a very compound that has attacked the liver, you have clear neoplasms of the liver and you also see lesions of the lung, we are not going to assume that the tumor has metasticized to the lung. Instead, we are going to assume that the chemical has attacked both the lung and the liver independently and that we need to add the slopes of these lines together to get the unit cancer risk.
We are also not going to assume that, because a dose response isn't present that the tumors being formed are not related to chemical treatment.
The second point here is that the lowest dose at which you see findings are reported is used to calculate the UCR. For example, if you did a study and you used three doses and only the lowest produced significant findings of tumors, we are going to use that lowest dose to calculate the UCR even though the two upper doses do not show significant findings. That is important to keep in mind.
[Slide.]
Here I have an example of how to calculate a unit cancer risk. This is from a mouse study. First, I am going to present data for male mice and then I will present it for female mice.
If you look to the left, you have the lesion column. And then we have the chemical reported in milligrams per kilogram body weight per day, and that is how we would like to see it as opposed to parts per million. Include the doses used as well as the incidence and the percentage incidence.
This example is clearly a liver insult. You have adenomas, carcinomas and hepatoblastomas. As is done here, it is appropriate at times to add adenomas and carcinomas together and then compare them individually as well as combined. We also have renal tubule adenomas that are detailed here. I have included just these two because these were the statistically significant findings of this study. Obviously, when you submit your neoplastic-lesion data, you are going to have all your organs present so you would have a full table.
How do we perform this calculation based on the data shown here? Down at the bottom of the slide is how to calculate the unit cancer risk. This is for the males and this is this data. So, what you would want to do is, first of all, identify what is the lowest dose with significant findings.
Here, for the liver, it was 16 mgs per kg per day. For the kidney, it was 32. So what you do is you subtract the control incidence from the treated incidence and then divide by the dose administered. Since we have two different organs with two significant findings of neoplastic data, as I mentioned, these would be additive. So you would perform this calculation twice and add the two slopes together.
At this point, we have the calculation resulting in 0.01 for mgs per kg body weight per day to the minus 1.
[Slide.]
If we look at female mice from the same study using the same doses, setting up the table in the same format. Again, this chemical, is very toxic to the liver in producing adenomas, carcinomas and blastomas of the liver.
The significant findings identified here, clearly 50 out of 50 animals had this response at the lowest dose tested. We could perform the same calculations subtracting out the control value, which was 33 out of 50 incidence, from the incidence of the treated lowest effect dose which is 50 out of 50. Dividing by the dose administered, we get 0.0212. This is a pretty straightforward mathematical operation.
Sometimes, you may only have one study. That makes for a very easy--you do the calculation and come up with your unit cancer risk. If you are looking at an NTP study or if you are fortunate enough to be going through the archives and finding lots of old studies on some chemical that has been tested, you may find numerous UCRs that you need to calculate, in which case, you want to make sure that when you calculate your upper-bound lifetime cancer risk, you choose the unit cancer risk from the study and the sex and the species that the chemical demonstrated the most potency.
[Slide.]
For the example that I have just shown, for the males, the result was 0.0144 whereas, for the females, it was 0.0212. I have highlighted here the females because this is a larger number and that is the number we are going to use to calculate our upper-bound lifetime cancer risk.
[Slide.]
We would prefer to see an upper-bound lifetime cancer risk calculation as opposed to a virtually safe dose. When you do that, you need to multiply the unit cancer risk that you have defined by the estimated daily intake.
When you do so, please make sure you convert that to milligrams as well as make sure that you convert that to kilogram body weight per day, here using the default value of 60 kilograms body weight.
Multiply the EDI from the unit cancer risk gives you the upper-bound lifetime cancer risk.
[Slide.]
That concludes the section that was on a straightforward example where you have concrete data, neoplastic data, available to you to do a calculation. Situations arise when we do not have bioassays on a chemical but have data supporting a concern for carcinogenicity. In this case, what we have is mutagenicity data and/or structure-activity relationship data that indicates there is a risk to the chemical.
[Slide.]
Here, graphically, I have just shown that you have non-negative mutagenicity data. If you have neoplastic data, you can go through, as I just showed you, and calculate out that upper-bound lifetime cancer risk.
If you don't, but you may or may not have a structural alert or you may have a structural alert and no mutagenicity data, the question becomes what do you then? As Drs. Chanderbhan and Braier mentioned, you look for structural analogues. If those are available, then you can calculate either the unit cancer risk or you can use the TD50 to estimate upper-bound lifetime cancer risk.
You would want to consider both the exposure and the uncertainty as far as how good the analogue is in your risk-management decision.
If the answer to this is no, then this is a very good example of when you would want to have a prenotification consultation because you would want to determine, based on the exposure, whether or not it is necessary to do a bioassay.
[Slide.]
So how do we estimate risk using structure-activity relationships? I will just briefly go through these couple slides here that lead into the examples because the previous speakers very nicely detailed them.
Non-negative mutagenicity data and/or a chemical class, which either doesn't have data on or we know, is structurally alerting, trigger SAR. The first phase, Phase 1, is a qualitative analysis. It answers the question, "Does the chemical contain structural alerts?"
[Slide.]
If the answer to this question is yes, we move on to what we are referring to as a Phase 2 analysis. Phase 2 is the quantitative approach. It is used to estimate the risk to chemicals with structural concerns or non-negative data. Currently, as Nancy mentioned, we are using the CDPB which is Louis Gold's compilation of TD50 values from numerous studies as well as we have our own database which contains chemicals that we have already calculated unit cancer risks for.
We can use either the TD50 or the unit cancer risk to estimate the upper-bound lifetime cancer risk.
[Slide.]
I am going to just briefly go over two examples of how SAR and risk assessment can be used to make a risk assessment to give to Risk Management to make a decision. If we look at the SAR approach, and this is Example 1, "Chemical X is possibly mutagenic." In this real example, the chemical was weekly mutagenic in the Ames assay. Weakly mutagenic, meaning that it was positive in one Ames strain and was negative in the other, and was also negative in another assay for mutagenicity.
The Phase 1 SAR analysis indicated that there was concern for some structural alerts present in the molecule. When we looked using an analogue approach, we found that there were neoplastic data not just on isomers of X. So a series of compounds, 3-5, 2-4, et cetera, that we could look at that had been tested, reputably tested--and I think that is important to point out--so that we could use that data, then, to estimate the risk.
We looked at the relevant bioassay data and calculated the unit cancer risk and upper-bound lifetime cancer risk of the analogue and made a safety conclusion that no additional data were required.
[Slide.]
Moving on to Example 2, "Chemical X is possibly carcinogenic." This is the example Dr. Braier mentioned with 1-ethylhexyl acrylate. As Nancy mentioned, we picked out ~91 analogues that were chemically similar but then filtered that out to six analogues that were considered biologically relevant, biologically meaning that they either fit into the phthalate class, which the majority of phthalates have been shown--in this case, the DEHP, the DEHA--to produce liver tumors through peroxisome proliferation, or in a DEHP sense, it is also a teratogen, and also to look at the acrylate chemicals which have some positive mutagenicity data but, for the most part, it is still equivocal whether or not those are carcinogens and available bioassay data.
So we set those aside for further analysis. We looked at the relevant bioassay data. We looked at the actual study summary such as from the National Toxicology Program for these analogues as well the literature to decide what endpoints the authors clearly stated were the result of chemical insult, which tumors were statistically significant and biologically relevant as a result of chemical treatment.
Then we used these TD50s and converted them to unit cancer risks through a simple equation and multiplied that by the EDI to determine the upper-bound lifetime cancer risk.
Since we had six analogues of two classes, we then concluded that the most potent UCR from that whole global spectrum was the one to use for the unit cancer risk but we also reported the range that the analogues represent in terms of managing the risk to management.
In this case, a safety conclusion was made and additional data were not required.
[Slide.]
In conclusion, back to this little graphic. I think the most important question here, obviously, is does the chemical have the potential for carcinogenic activity. I mentioned three possible scenarios where the answer could be yes as well as a scenario not on here whichhere, which is functional alerts. I also mentioned ways that we are working in to learn how to deal with this in a non-black-and-white-type data- requirement; in other words, to use the data that is already out there available to try to answer these questions.
I hope that this talk provided some insight into that.
CHEESEMAN: Questions for any of the speakers in the Toxicology session?
WIEDOW: Al Wiedow, Ciba Specialty Chemicals. You seem to cover the gamut from genetox and then you jump into cancer bioassays. I didn't hear much about noncarcinogenic target tissues. Dr. Sheu mentioned about a 5 ppb range for genetox and a 50 ppb range for triage. I didn't hear anything about triage for noncarcinogenic materials.
I didn't hear anything about susceptibility, children's health, children's issues. Do you add that as a safety factor? I didn't hear anything about endocrine-disrupting chemicals as another subtarget tissue. Could you please comment on some of those areas?
TWAROSKI: I think your point is really well-taken. I have asked Dr. Lin maybe to comment since he is here, but I think the thing to keep in mind is you made some excellent points about issues that we deal with in toxicology, but there are also very specific issues and the reason they might not have been discussed is to keep this particular conference is limited to one day and is on a broad spectrum of topics.
But maybe Dr. Lin would like to comment specifically about any of those points.
LIN: If I remember correctly, during this last year or two, we did have the situation where we had to deal with nongenotoxic lesions which oftentimes have been used as a predictor for potential carcinogenicity. The endpoint I am referring to is the peroxisome proliferation. Actually, there were a couple cases that we had to deal with that.
I think our concept is a substance is not a carcinogen unless there are really bioassay to indicate so.
As far as the genotoxicity endpoint, I think we have to risk assessment because of the general belief that this is a one-hit theory. So, for the genotoxic endpoint, we really have to do risk assessment, do the cancer risk assessment.
But for other endpoints like peroxisome proliferation, the office has formulated a policy that that endpoint is treated just as any other endpoint for which we can apply the general safety standards which means if, in that particular study or bioasssay--not bioassay, in a shorter-term study, like a 90-day study, if you see peroxisome proliferation, we can apply this general safety standard as long as we can find a no-effect level in that particular study.
Then we apply appropriate an safety factor to derive an ADI. So I think that is the difference between the endpoint such as peroxisome proliferation versus a genotoxic endpoint. I don't know whether I answered your questions or not.
WIEDOW: You answered the question as part of the symposium while you are doing cancer-risk bioassay. But how many cancer bioassays do you receive? That is the endpoint of all the toxicity testing one would do. I was just trying to focus in on the 90-day studies, what happens to those, the non-cancer target tissues.
TWAROSKI: I think your question--this symposium, what we have talked about here toxicologically is the endpoint of carcinogenicity and genotoxicity where your point is what about the non-carcinogenic safety assessment; is that correct?
WIEDOW: Correct.
TWAROSKI: I think the reason that we have focused so much on that is because of some studies, if you look at threshold, looking at where the most potent toxicological endpoint is and then looking at what type of exposures we routinely deal with in this lower exposure of no data.
But Francis could elaborate more on the general safety evaluation of 90-day studies which is the noncarcinogenic safety assessment.
CHEESEMAN: Do you mind if I interject for just a second? I think it the reason we started the effort dealing with genotoxicity and dealing with carcinogenicity prediction is because, as Michelle said, it is the most sensitive endpoint. It is the one that we have, perhaps, the most difficulty dealing with at the lowest exposures for that reason.
I think it is fair to say that we have the same sort of goals in mind of what you are talking about. A significant part of our effort in this area right now is database development. That database development includes the entire gamut of toxicity data that we have in our files and is out there in the published literature.
But the carcinogenicity databases, the mutagenicity databases, that are used to make these predictions that we are talking about today are simply more robust and more useful right now than the other databases, not that there is not the possibility and we won't eventually get to some capability of being able to use these tools in other areas.
WIEDOW: Thank you.
TWAROSKI: Are there any other questions for general toxicology or is everyone ready to go home?
KEITHLINE: I'm Jeff Keithline from Keller and Heckman. It is my understanding that qualitative, at least, SAR is done on every FCN that is provided; is that correct?
TWAROSKI: When we first initiated doing SAR, that was the approach that we were going to take. That is an approach we would like to take once we build up our databases enough sufficiently based on what data we have in our files to make sure we are not looking at duplicative--obviously, for our own timesaving ability.
If we already have seen something, or something similar, it is not just to your advantage, but also to our advantage, to be aware of that. But we are also considering an exposure before we do SAR currently, just because of wanting to make sure we utilize our resources in the most effective manner.
So if we, for instance, go to a Phase 1 meeting and a notifier has mentioned a certain chemical as having exposure, but our chemists state that, due to the volatility and the temperatures used, it is not likely to be there, then that is a case where we probably would have done SAR initially but may or may not do SAR now.
Does that answer your question?
KEITHLINE: Yes.
ETTINGER: David Ettinger, Keller and Heckman. If the agency does an SAR analysis, would that analysis be disclosed under FOIA?
TWAROSKI: Yes. I think that is something to keep in mind, that SAR, since you mention that, is very fluid. You are constantly developing your databases. You are constantly trying out new expert systems. So the answer you might get today may or may not necessarily be a concrete solid answer the next time you run the system because you could have modified the database.
So I think it is best to keep in mind when you do FOI something that, if you feel that you have gotten conflicting results, you can bring that up, but it is a fluid process and you are adding to the training base, just like Mitch and Ron presented in multicase.
Back to your question, there are teratology modules, there are other endpoints in that system that need to be developed further. Again, part of the fluid process is getting that data, getting it into the system so that then we can validate it and have more use of it.
CHEESEMAN: Let me modify my answer, David, because of your earlier point. It would be releasable to the extent that we wouldn't reveal confidential commercial or trade-secret information. As you have pointed out, we could do that in talking impurities and we wouldn't do that, or we would certainly seek not to do it.
I would also emphasize what Michelle said. This area of investigation and these tools are more fluid than you can possibly imagine.
SOTOMAYOR: Renee Sotomayor, CFSAN. When you do your upper-bound lifetime cancer risk, do you take into consideration factors, for instance, like type of exposure? Normally, the cancer risk is not a continuous exposure but I doubt that we ever eat continuously one chemical. So, do you take into consideration factors like intermittent exposure and what do you do in cases like that?
TWAROSKI: This is a very excellent point when you consider something may not be a complete carcinogen. It may just be a promoter and be promoting age-induced mutations. So, if you take it away, you are not going to get cancer.
Like you mentioned, lifetime exposure is considered in the model. But, as far as I know, there are not really good models out there to simulate this risk that are being used in the regulatory framework. Instead, more along the lines that something being a complete carcinogen not whether or not it is initiator, a promoter but, instead, that it is a complete carcinogen and that it is going to be exposed in a lifetime human basis.
COLLINS: Tom Collins from ORSO or CFSAN. As far as the fetus, you are going to have teratology bases, right, or teratology databases. What do you do about the kids, what we call neonates? They are going to be more exposed because of their small size as they are exposed to some of these additives, et cetera.
TWAROSKI: So your question specifically is are there SAR modules being developed to look at adolescence and young children and their insult to chemical exposure?
COLLINS: Right. Some of these chemicals, you are being exposed to in utero whether you like it or not.
TWAROSKI: Definitely. Maybe Ron could comment if he is aware--I know most of our collaboration is with CDER in their development of modules for MCASE. But we are developing, like I said, our database in an analogue search type situation, in other words structure search more than expert system.
CHANDERBHAN: I am not aware currently of any specific move to develop modules in in utero exposure. But that is not to say that it couldn't come in the future. I know one area that they are working on right now in MCASE in developing is maximum tolerated doses using just human data not animal data. That, when it is developed, could be very useful because we are always using animal data and making the leap of faith to humans.
If we have solid human data in a good module that has been validated, tested, it could be very useful. I don't see that, in the future, we couldn't go to something like you are talking about, adolescent, in utero, type exposures. It is very possible. The sky is the limit. I don't know where it will go. It could very well be in the future.
TWAROSKI: If there are no more questions, then Carolyn Young would like to take the podium. Maybe there is a question or also closing remarks. So thank you very much.
Wrapup/Questions
YOUNG: I want to take this opportunity to respond to a question that we received on an index card in relationship to biocides and recommended toxicity testing at different levels.
[Slide.]
In my talk earlier, I presented this particular slide. According to what is written in the Toxicology Guidance Document, it mentioned to apply the testing recommendations on one-fifth of the CEDI for the nonbiocide compounds. The question that we received specifically relates to the first and the second level whereby it indicated an exposure of 0.1 ppb which would be one-fifth of the 0.5 level.
I appreciate the comment tremendously because it points out a place in the Tox Guidance where we need to be more specific. The comment that was made is very right, that we do not require or recommend testing at less than the 0.5 ppb level. So this slide, and in your handouts, in the Toxicology portion from this morning on Page 9, it is the top slide. If you would please go to that section and make the following correction.
That could be changed, then. At the first level, for incremental exposure and daily diet less than or equal to 0.5 ppb or the 1.5 microgram per person per day where we do not recommend toxicity testing and the second level, instead of starting at 0.1, it would, again, start at the 0.5 and go to below 10 ppb. That would be represented by the studies that we recommend for the second level.
Does anybody have any comments or questions about that? Is that clear?
SOTOMAYOR: Could you repeat, please?
YOUNG: Page 9.
SOTOMAYOR: I found it. Now what do I do?
YOUNG: What I would like for you to do is next to 1--on your copy, it says, 0.1 ppb. Cross out the 1 and put a 5. Where it says 0.5, change that to 1.5 micrograms per person per day. For No. 2, instead of greater than 0.1 ppb, please change that to 0.5 and change, again, the micrograms per person per day to 1.5.
SOTOMAYOR: The 10 parts per billion; that stays?
YOUNG: The 10 parts per billion stays. That doesn't change. So, as I have been told, our guidance document is never etched in stone and I look forward to being able to make some modifications to that over time.
That's the end of the questions that I have. If there are no other questions for Toxicology, I will pass the mike.
CHEESEMAN: Thank you, Carolyn.
We have, I think, two other written questions and then we have a few more minutes to take any additional final questions that you might have or we can adjourn if everybody is ready to go.
I have, actually, a recommendation that changes made into various guidance documents be highlighted in some way or italicized to denote the modifications. I will take that back to our webteam, but I can tell you that they don't it that way based on the decision that is made basically above the Office of Food Additive Safety.
I think it is a very reasonable suggestion, by the way, and I have made it, myself, in the past. I would point out that when we put guidance documents up in final on the website, it is our habit and our procedure to link to the previous version. So you do have both the Draft 1999 or 2000 version of the guidance that's available on the website, or you should. And you have the 2002 version available, also, so you can physically compare them.
We also have made it our habit to include highlights in each version, to highlight the significant changes.
The last question, written question, is Dr. Shanklin's to address.
SHANKLIN: I received a two-part question. It says, "If a supplier submits an FCN, how does he provide manufacturing process?" It is our hope that the suppliers will be working with the manufacturers. I also have seen cases where the manufacturer submits what we call a Food Master File and the supplier is allowed to reference the Food Master File. So, hopefully, that answers your question. But you do need to have manufacturing information as a supplier.
My second question is, "Can the manufacturer then sell to anyone, not just the supplier notifier for that particular end use?" The key point in that question is yes, but that for that particular end use. Also in our guidance it says that in addition to the manufacturer or supplier specified in the notification, the FCN is also effective for that manufacturer's or supplier's customers.
So, yes, it can be. But it has to be for that particular use. That is the key.
ITZKOFF If a manufacturer does their FCN, then their supplier can authorize his other customers to use the same material? Isn't it limited to the manufacturer's customers?
SHANKLIN: The notification is effective for that manufacturer or supplier, but that supplier and/or its customers, as long as they can prove that the food-contact substance is being used as it is intended in the notification and produced by that manufacturer, it is effective for them, also.
CHEESEMAN: I was mistaken. We have one more toxicology question that Dr. Lin will answer.
LIN: This is a two-part question. Actually, I am not too clear about the question here. So, I need whoever asked this question to help me out. "When evaluating subchronic toxicity data in the absence of chronic toxicity data, does the agency recommend a safety factor of 10 or a safety factor of 3?"
I think what it really means is that the additional factor of 10 or 3--I don't know who asked this question. When we extrapolate from subchronic to chronic data, traditionally, historically, the agency has been using an additional factor of 10 on the top of 100. Everybody knows about that. So the safety factor we use is 1000.
So, as far as the safety factor of 3, I understand that there have been publications and also maybe other agencies have, indeed, used additional factor of 3 to extrapolate from subchronic to chronic. But, traditionally, the agency has not really done that.
Now, the second part of the question is when calculating NOEL, I do understand, unless it meant to say that you extrapolate from the lowest data, a factor or LOEL to NOEL, and whether we apply additional 10-fold or 3-fold safety factor. But, since the person who asked the question evidently is not here, so maybe you are not interested in knowing what the answer is.
Actually, traditionally, we have not, until very recently, to extrapolate from LOEL to NOEL. In the past, if there is no no-effect level. The agency will say, "You may have to back to do additional tests to find a clean no-effect level. But this past two or three years, we have become, I would say, probably more flexible. So we have, in certain cases, if I remember correctly, at least a couple of cases where we do extrapolate from the lowest effect dose to the NOEL.
As far as with an additional 10, or an additional 3, the way we determine which one depends on the severity of the concern of the toxicity that we are talking about. If it is just some minor effect--for example, the effect on the body weight gain--and nothing else, at this lowest effect dose, we would agree to the use of an additional safety factor of 3 to extrapolate from the lowest effect dose to NOEL.
On the other hand, if the endpoint of concern is what we consider of a serious nature--for example, you have a liver necrosis or you have a fatty liver, things of this sort, and we would, in the absence of no-effect level, the true no-effect level, apply an additional 10 for safety factor for this extrapolation.
So are there any questions? Or any objection?
WIEDOW: Just one question. The difference between the NOEL and the NOAEL. EPA uses the NOAEL, no adverse effect level, to determine some of their ADIs. What is the position here between the NOEL and the NOAEL?
LIN: We really don't make that distinction. We use the two interchangeably. Of course, when we determine an effect level, we don't just say, "Well, you see a little bit increase in the relative liver weight but you don't have any other clinical chemistry or histopathology." And we say that is an effect. We don't do that so that is why we don't want to make a clear distinction between NOEL and NOAEL.
WIEDOW: But you could have orders of magnitude sometimes in the difference in the dose. You could have a very large discrepancy in the dose of what you set your, say, ADI to develop from. You still would use the NOEL, the no-effect level, as opposed to--just like the example you used, maybe a change in some liver weight gain. That is the only clinical sign you might even see?
LIN: Like I said, we try not to get into an argument whether one given effect is adverse or is not. I think the agency tries to use judgment. Just like I said, we are not determining the no-effect level without taking into account all the considerations. Really, is it fair or not to call something an effect if it only changes say a relative liver-weight change without anything else.
So I would use that as an example.
WIEDOW: Would you consider that the NOEL, then?
LIN: Yes; it is a NOEL.
WEIDOW: As opposed to not seeing any change.
LIN: Excuse me?
WEIDOW: As opposed to not seeing any change in the study, where you say at the lowest dose, you didn't see any change. At the next highest dose, you saw the slight increase in liver weight.
LIN: If you use the next higher dose, you see very--just minor. We are talking about the magnitude. If you have, say, a very small percentage of an increase in relative liver weight, even at the dose next to the NOEL, the true NOEL, I think that we would consider just using the lowest dose as the NOEL. Even though there is, at the next dose, there is no effect at all, including the relative liver-weight change.
I think the point that I am trying to get across is that we do use our judgment in calling something an effect or not. Here, effect, what we say effect, what actually we mean is adverse effect.
Anything else? Thanks.
CHEESEMAN: Any further questions for any of the speakers or anyone else? Have a good evening.
[Whereupon, at 5:00 p.m., the workshop was adjourned.]