NATIONAL ADVISORY COMMITTEE ON MICROBIOLOGICAL CRITERIA FOR FOODS

NATIONAL ADVISORY COMMITTEE
ON MICROBIOLOGICAL CRITERIA FOR FOODS
RISK ASSESSMENT ON THE PUBLIC HEALTH IMPACT
OF FOODBORNE LISTERIA MONOCYTOGENES
Thursday, May 27, 1999
8:10 a.m. to 4:10 p.m.
The Ambassador West Hotel
George I Conference Room
1300 North State Parkway
Chicago, Illinois
MEMBERS:
MICHAEL L. JAHNCKE
Virginia Polytechnic Institute & State University
Hampton, VA 23669
MORRIS E. POTTER
Food and Drug Administration
Center for Food Safety & Applied Nutrition
Washington, D.C. 20204
ROBERT L. BUCHANAN
U.S. Food and Drug Administration - CFSAN
Washington, D.C. 20204
DANE T. BERNARD
National Food Processors Association
Food Safety Program
Washington, D.C. 20005
MICHAEL P. DOYLE
Department of Food Science & Technology
Center for Food Safety & Quality Enhancement
University of Georgia
Griffin, GA 30223
AIM REPORTING SERVICE
(773) 549 - 6351
2
ANGELA D. RUPLE
National Marine Fisheries
Pascagoula, MS 39567
MEMBERS:
(Continued)
MARGUERITE A. NEILL
Memorial Hospital of Rhode Island
Infectious Disease Division
Pawtuckett, RI 02860
MICHAEL C. ROBACH
Continental Grain Company
Gainesville, GA 30501
CATHERINE W. DONNELLY
College of Agriculture & Life Sciences
University of Vermont
Burlington, VT 05405
DAVID W. K. ACHESON
New England Medical Center
Boston, MA 02111
MARGARET HARDIN
National Pork Producers Council
Des Moines, Iowa 50306
BRUCE TOMPKIN
Armour Swift-Eckrich
Downers Grove, IL 60515
LEON R. RUSSELL, JR.
Texas Veterinary Medical Center
Department of Veterinary Anatomy & Public Health
College Station, TX 77845
AIM REPORTING SERVICE
(773) 549 - 6351
3
ALISON D. O'BRIEN
Uniformed Services University
of the Health Sciences
Bethesda, MD 20814
TERRY C. TROXELL
Food and Drug Administration
Center for Food Safety & Applied Nutrition
Washington, D.C. 20204
AIM REPORTING SERVICE
(773) 549 - 6351
4
I N D E X
LIST OF SPEAKERS
Page
DR. WESLEY LONG
8
DR. RICHARD WHITING
14
DR. MARY BENDER
46
DR. PAT McCARTHY
108
DR. RICHARD RAYBOURNE
124
DR. RICHARD WHITING - SUMMARY
168
AIM REPORTING SERVICE
(773) 549 - 6351
5
P R O C E E D I N G S
1
2
MR. MORRIS POTTER:
Good morning.
3 Listeria Risk Assessment Public Hearing.
Welcome to the
The Food and Drug
4 Administration and Food Safety and Inspection Service, USDA,
5 are conducting a risk assessment to determine the prevalence
6 and extent of foodborne exposure to Listeria monocytogenes
7 and the public health consequences of that exposure.
8
This is a call for information.
The Risk
9 Assessment Task Force will present the framework that they're
10 developing.
And we hope that the subcommittee and the full
11 committee and the audience participants will be able to
12 provide the necessary data input to make this a high-quality
13 product.
14
The goal of these risk assessments -- yesterday's
15 on Vibrio parahaemolyticus, and today's on Listeria -- is
16 to provide FDA and FSIS with the information needed to review
17 current policies and to ensure that future programs provide
18 the maximum public health benefit.
19
I'd like to thank the Risk Assessment Task Force
20 for preparing today's presentation.
And I'd like to thank
AIM REPORTING SERVICE
(773) 549 - 6351
6
1 the Risk Assessment Subcommittee under Dr. Jahncke's care
2 for their attention and counsel to the Task Force.
And I'd
3 also like to thank the audience participants whose presence
4 and participation underscores the importance of this public
5 health effort.
6
With that, I'd like to turn the mike over to Dr.
7 Jahncke, who will manage today's proceeding.
8
MR. MICHAEL JAHNCKE:
Thank you, Dr. Potter.
I
9 would also like to thank the Risk Assessment group that has
10 put together this document.
11 presentations.
Looking forward to the
And we would like to welcome all of the
12 subcommittee members and all of our guests in the audience.
13
14
Just a couple of procedural pieces.
Please
15 remember, when you do speak, to use the microphones -- we
16 are being transcribed -- and to identify yourself and your
17 association.
Also, keep in mind that this is a discussion
18 on risk assessment.
Everyone should have a copy of the
19 document, the presentations that will be based upon the
20 document in our NAC folder.
It should be under Tab 10, I
AIM REPORTING SERVICE
(773) 549 - 6351
7
1 believe.
The title of it is, "Structure and Initial Data
2 Survey for the Risk Assessment of the Public Health Impact
3 of Foodborne Listeria Monocytogenes."
4
Everyone also ought to have a Draft Agenda.
5 will be following this Draft Agenda.
6 their presentation.
We
The speakers will make
There will be time after each of the
7 presentations for questions.
The way the questions will work
8 is that we will take questions from the Risk Assessment
9 Subcommittee members around the table.
If there are no more
10 questions from the Risk Assessment Subcommittee members, we
11 will then take some questions from the National Advisory
12 Committee people who are in the audience.
And we will go
13 through each of the presenters in that manner.
14
If you do look at the schedule, when all the
15 speakers are finished in the morning, there will be a general
16 committee discussion where we will invite up all of the
17 speakers to the table and also any of the National Advisory
18 Committee members in the audience to the table, for an open
19 and general discussion.
Following that, there will be an
20 opportunity for open public comments.
AIM REPORTING SERVICE
(773) 549 - 6351
8
1
I would like to start off this morning by each one
2 of us around the table introducing ourselves and our
3 affiliation.
My name is Michael Jahncke, and I am with
4 Virginia Tech.
5
MR. TERRY TROXELL:
Terry Troxell, FDA.
6
MR. BRUCE TOMPKIN:
Bruce Tompkin, Armour
7 Swift-Eckrich.
8
MR. MICHAEL DOYLE:
I'm Mike Doyle with the
9 University of Georgia.
10
MR. LEON RUSSELL:
Leon Russell, Texas A & M.
11
MR. MICHAEL ROBACH:
Mike Robach, Continental
12 Grain.
13
MR. DAVID ACHESON:
David Acheson, New England
14 Medical Center, Tuft University.
15
MR. DANE BERNARD:
Dane Bernard, National Food
16 Processors Association.
17
MS. ANGELA RUPLE:
Angela Ruple, National Marine
18 Fisheries Service.
19
MR. ROBERT BUCHANAN:
20
MS. MARGUERITE NEILL:
Bob Buchanan, FDA.
Peggy Neill, Brown
AIM REPORTING SERVICE
(773) 549 - 6351
9
1 University School of Medicine.
2
MS. MARGARET HARDIN:
Margaret Hardin, National
3 Pork Producers Council.
4
MS. CATHY DONNELLY:
Cathy Donnelly, University
5 of Vermont.
6
MR. MICHAEL JAHNCKE:
Thank you very much.
7 that, let us begin our session today.
With
Our first two speakers
8 are going to be -- Dr. Wesley Long, I believe, will be
9 starting.
And then Dr. Richard Whiting will also be part
10 of the presentation.
11
The title of the presentation is Introduction to
12 Listeria Monocytogenes, Risk Assessment.
13
DR. WESLEY LONG:
Dr. Long?
Good morning, everyone.
14 Welcome to our day-long presentation on the structure and
15 initial data survey for the risk assessment of the public
16 health impact of foodborne Listeria monocytogenes.
17
Before we get into the technical presentations,
18 I want to set the stage for the rest of the day that I hope
19 will help both the Committee and the public help us to take
20 this risk assessment on to the next stage and on to conclusion.
AIM REPORTING SERVICE
(773) 549 - 6351
10
1
Before we get started, I just want to let you know
2 that we did do a risk assessment activity yesterday all day
3 and that we are starting anew today.
So, those of you that
4 were here yesterday, there will be some repetition.
I can
5 see that there are a number of different people in the
6 audience, so I'm glad we decided to take this approach.
7
The stated purpose of the risk assessment is to
8 determine the prevalence and extent of consumer exposure to
9 foodborne Listeria monocytogenes and to assess the resulting
10 public impact of such exposure.
This risk assessment, how
11 will the results of this risk assessment be used?
12
The risk assessment is intended to provide both
13 FDA and FSIS with the scientific information that they need
14 to review their current programs relating to the regulation
15 of Listeria monocytogenes contamination in foods and to
16 ensure that those programs provide the maximum level of public
17 health protection.
18
Now, if you've ever heard me speak before, you've
19 ever asked me to speak, you're bound to see a slide very
20 similar to this because I think it's very important that we
AIM REPORTING SERVICE
(773) 549 - 6351
11
1 point out a little bit that risk assessment is only one
2 component of the risk analysis process, which includes risk
3 communication, risk assessment, as well as risk management.
Risk managers have a number of factors that they
4
5 need to consider when they make a decision.
Public values
6 are very important, and that's why this is a public forum
7 and we are interested, of course, in what the public has to
8 say today.
There are senior-level risk managers here from
9 both FSIS and FDA, and they'll be happy to listen to your
10 comments and to consider those comments when we do get to
11 the stage of evaluating the programs.
12
13
There will always be economic factors to consider.
And while we do not base our public health decisions on
14 economics, those are always considerations in terms of the
15 cost benefit analysis of any sort of actions.
16
There will always be political factors.
I guess
17 the point here is that even our risk managers have bosses.
18
And factors such as budget and priority always have to be
19 considered.
20
Technology, what we may be able to do may be limited
AIM REPORTING SERVICE
(773) 549 - 6351
12
1 by technology.
And it may drive us in one direction as
2 opposed to another.
Statute.
Both FSIS's are governed by
3 laws that set the framework for what actions we can and cannot
4 take.
Finally, there's the science.
5
6 is in terms of the risk assessment.
7 the organization of that science.
The science today
The risk assessment is
So, the point here is just
8 that the risk assessment is one of the considerations that
9 will go into considering the revisions of our current programs
10 in terms of the regulation of Listeria contamination in foods.
11
Now, the risk assessment is a collection of the
12
13 scientific facts that are structured to try to clearly tell
14 what it is that we know and what it is that we don't know.
15
And they should be descriptive to characterize how well we
16 know what we know.
In addition, we want to put out extra
17 efforts to be very transparent and to reveal any biases we
18 may have.
So, for example, if we decide that we're going
19 to use one data set as opposed to another, that needs to be
20 very well-explained; and the effect on the end analysis needs
AIM REPORTING SERVICE
(773) 549 - 6351
13
1 to be clear for the risk managers and the public.
2
What questions do we hope that this risk assessment
3 will answer?
Can the relationship between the consumption
4 of Listeria monocytogenes in foods and the risk of becoming
5 ill be quantified?
Can we establish a quantitative
6 relationship between the numbers of Listeria that are
7 consumed and the likelihood or the extent of illness?
What
8 data do we need that will help reduce the uncertainty in these
9 estimates of risk that we're going to come up with?
And
10 does the assessment focus further efforts on specific foods
11 or populations at risk?
And I'll come back to this in a
12 moment.
13
14 answer?
So, what questions will the risk assessment not
Again, coming back to risk management, we're not
15 going to establish the appropriate level of public health
16 protection today or through this risk assessment process,
17 although the risk assessment will be considered in making
18 those determinations.
Right now, we're not gonna look at
19 control measures that may be implemented for producers,
20 manufacturers and consumers.
And we're not gonna decide what
AIM REPORTING SERVICE
(773) 549 - 6351
14
1 levels of Listeria should be allowed in or on foods.
2
So, what are we hoping to get from NACMCF today?
The
3 purpose of my presentation is to help us all focus on risk
4 assessment.
Is their scientific approach sound?
In the
5 document, you see that we've laid out what we consider the
6 parameters to be and the flow of the risk assessment.
And
7 we're very interested in whether you have recommendations
8 on how we can revise that approach or modify that approach
9 to enhance it and make it more useful.
10
Do we have all of the right data?
I think the
11 document that you have in front of you today was the result
12 of casting a fairly wide net to try to capture all of the
13 data and information that we could.
14 initial data screening.
There has been some
The area that I am most familiar
15 with is the dose-response component.
And, for example, we
16 decided to limit our data to post-1990 in terms of
17 immunobiology and virulence characteristics because of
18 advances in that science.
19
So, we are looking for your help in determining
20 whether we have the right data.
And I have a couple of
AIM REPORTING SERVICE
(773) 549 - 6351
15
1 comments about the data.
We're looking for data.
And this
2 is gonna come up over and over during the day -- on the
3 frequency of Listeria monocytogenes isolations from
4 different foods; on the serotypes isolated from foods; and,
5 of course, on the levels of Listeria monocytogenes in various
6 foods.
7
This data request is being structured in such a
8 way that we're not intending to utilize any data that you
9 might submit to us to take any sort of enforcement action
10 against the submitter.
And we are accepting data that is
11 blinded to protect the confidentiality of those who might
12 have data but might have various reasons for not feeling
13 comfortable in submitting it.
And, of course, have we
14 overlooked anything?
15
Time line for the process:
You know that we came
16 to NACMCF in February and made an initial presentation of
17 where we thought we were headed.
18 course, in May.
And we're here today, of
There was a Federal Register document
19 published a few weeks ago which has a comment period that
20 closes July the 6th.
And we're hoping for you to submit both
AIM REPORTING SERVICE
(773) 549 - 6351
16
1 your comments and any information and data you might have
2 by that date.
3
Starting on that date, we will be, of course,
4 between now and then revising our plans based on the comments
5 that we hear today and wrapping up our data collection and
6 beginning the modelling phase of this process.
7
We hope to have a draft report by September or
8 October of this year.
We want to come back to NACMCF and
9 to the public with the results of those analyses.
And then
10 we're going to initiate additional risk assessment activities
11 starting, hopefully, in November of this year.
12
Those additional activities will be based on where
13 this phase of this initial risk assessment leads us.
And
14 that might include analysis of product-specific pathways.
15
What this risk assessment may help us do is target specific
16 foods or classes of foods that we need to study more closely.
17
We will also look at the effects of various interventions
18 on pathogen load.
And we may identify the focus of further
19 research and technology development that will help us reduce
20 the uncertainty to come up with better risk estimates of risk
AIM REPORTING SERVICE
(773) 549 - 6351
17
1 to help us better set policies.
2
I'm going to turn this over now to Dr. Whiting,
3 and we'll get rolling on the day.
4
DR. RICHARD WHITING:
Thank you.
Wes just gave you a little
5 bit of the outline of the risk assessment process and the
6 structure.
I want to just sort of begin an introduction of
7 the risk assessment itself.
8
In the 1980's, I think you're all aware that we
9 realized that Listeria monocytogenes was a foodborne
10 pathogen.
And when I mentioned the word, "Listeria," I think
11 for this meeting, we will be meaning Listeria monocytogenes
12 pretty much through this whole presentation.
13
And at that time, there was a very intensive five-
14 to six-year period of research in which we recognized the
15 widespread occurrence of Listeria, both in the natural
16 agricultural environment, as well as the food processing
17 environment, and this widespread occurrence in foods.
18
There was also a couple papers where they found
19 the presence of Listeria monocytogenes in the
20 gastrointestinal tract of apparently healthy people.
AIM REPORTING SERVICE
(773) 549 - 6351
And
18
1 we realize that despite this widespread occurrence of the
2 organism, that the disease occurs relatively rarely.
3 figures are about .5 cases per 100,000 people.
But when it
4 does occur, it is very likely to be a serious disease.
5 it's a very opportunistic organism.
The
And
It strikes the
6 immunocompromised, including the elderly and pregnant women.
7
8
As a result of that, the Agencies, both FSIS and
9 FDA, have had a zero tolerance policy for this organism in
10 foods, which means if they find it in a sample, the food is
11 considered adulterated.
The industry put a very intensive
12 effort in improving sanitation and process controls.
And
13 from the period from about 1989-1990 into about the
14 mid-1990's, we've seen a decrease of 40 to 50 percent in the
15 incidents of Listeriosis.
16
However, in the last couple of years, we've seen
17 a levelling off in this decrease in incidents.
This may be
18 that the preventative measures that have been taken have sort
19 of run their course.
It may also reflect more increased
20 surveillance efforts and better detection of the disease.
AIM REPORTING SERVICE
(773) 549 - 6351
19
1
2
It's not really clear.
There's also been quite a bit of thinking in the
3 scientific community on the dose-response relationship of
4 just what consumption of low-dose of this organism means.
5
And as a result of some of this thinking, several other
6 countries that we are actively trading with -- I think of
7 Canada and Denmark, in particular -- have regulatory policies
8 in which if Listeria is found in certain classes of food,
9 it is not automatically considered adulterated and pulled.
10
So, these various considerations are some of those that are
11 driving this re-looking at the Listeria question.
12
I also should mention that this is just part of
13 our little broader effort by the Federal agencies to look
14 at Listeria.
FSIS has an ongoing survey.
15 a bit of research on the organism.
There is quite
I particularly want to
16 mention the FDA graph per-dose response research.
This is
17 with the University of Georgia Primate Center where they are
18 specifically doing challenge studies with pregnant monkeys
19 to try to determine what the dose-response relationship is
20 for this organism.
AIM REPORTING SERVICE
(773) 549 - 6351
20
Also, within the CDC, I think, our food-net
1
2 program, which you're aware of, has been very instrumental
3 in getting better information on this organism.
And the CDC
4 is also in the final stages of planning a case-control study,
5 which also should give us a lot more information on the
6 incidences and various other information.
And then,
7 finally, the risk assessment.
I just want to say a little bit about that other
8
9 risk assessment that you heard of yesterday, the follow-up
10 on Wes' remarks.
11
We have quite a different purpose here.
The Vibrio risk assessment is focusing on one organism and
12 primarily on one food.
And you saw some pathway-type
13 modelling where they were looking at increases and decreases
14 in Vibrio.
This risk assessment is more of a risk-ranking
15 type, where we're looking at the very broad spectrum of foods
16 and are interested in saying which foods contain how much
17 Listeria.
Both risk assessments, of course, then address
18 the question of the dose-response relationship.
19
So, with that, then I'd like to just say a little
20 bit about the structure that you will have today.
AIM REPORTING SERVICE
(773) 549 - 6351
Two of
21
1 the data-collecting areas in a risk assessment have been
2 described as the exposure assessment and then the hazard
3 assessment.
4
And we follow that organization this morning.
You will hear on the exposure assessment side.
5 have two presentations.
And we will
Tony Hitchins will look at the
6 survey for Listeria monocytogenes presence in food.
And then
7 Mary Bender will look at the consumption patterns in food.
8
The idea here is that the exposure of Listeria to
9 the population is a result of how many Listeria organisms
10 are in the food, then times the amount of the particular food
11 that is consumed.
12
Next slide.
Okay.
This afternoon, then, we will
13 move on to the other part, the hazard assessment.
And there
14 will be a presentation by Pat McCarthy looking at the
15 epidemiological record.
16 overlap here.
And I will admit, there is some
The epidemiological record, of course, gives
17 us information on exposure, as well.
18
And then, secondly, a presentation by Richard
19 Raybourne on the dose-response experimentation.
This would
20 be any animal, human feeding studies or in-vitro experiments
AIM REPORTING SERVICE
(773) 549 - 6351
22
1 that would give us information on the dose and response.
2
After these two parts are done, we then move into
3 the risk characterization phase, which you can call the
4 modelling or the number-crunching, if you will.
5 begin after this meeting.
This will
And just to complete our
6 description of the team, Dr. Clark Carrington will head up
7 the modelling section.
But since we have not gotten to that
8 point, there will be no presentation on that today.
9
So, thank you, Mr. Chairman.
I turn the meeting
10 back over to you.
11
MR. MICHAEL JAHNCKE:
Yes.
Are there any
12 questions from the subcommittee members for either Dr. Long
13 or Dr. Whiting?
14
If not, thank you very much for your presentation
15 and excellent introduction to the subject.
16
We're now moving, as Dr. Whiting indicated, the
17 next section is exposure assessment.
18 time is Dr. Tony Hitchins.
And our presenter this
And his topic is presence of
19 Listeria monocytogenes in foods.
20
Dr. Hitchins?
AIM REPORTING SERVICE
(773) 549 - 6351
23
1
DR. TONY HITCHINS:
2 honor to be here.
Thank you very much.
If I could have the first slide.
It's an
Our work
3 is done by the contamination work group of the FDA Center
4 for Food Safety and Applied Nutrition.
5
Next slide, please.
6 are as follows:
The members of the work group
Mary Lynn Datoc from FDA; Eric Ebel and Wayne
7 Schlosser from the USDA; myself from FDA CFSAN; and Pauline
8 Lerner from FDA CFSAN.
Mary Lynn has been collecting data
9 on vegetables and cheeses.
Eric and Wayne have been
10 collecting data on the meats.
Pauline Lerner has been
11 collecting data on the seafoods and is also now just moving
12 into the milks.
Myself, I've been collecting data on some
13 of the larger studies that cover or contain all areas of foods,
14 so all food types.
15
Next slide, please.
Today, I'd like to overview
16 Listeria monocytogenes in relation to food safety.
17 asked to do this as I am the first speaker up.
I've been
And I hope
18 the members of the audience, of whom there are a lot who are
19 experts on monocytogenes, will bear with me at my simplistic
20 approach.
Then we'll move on to the actual meat of the talk,
AIM REPORTING SERVICE
(773) 549 - 6351
24
1 which is the food contamination data collection and then give
2 an interim report on the results so far and perhaps indicate
3 at the end some future work.
4
Next slide, please.
Our role in the process or
5 this module's role is to collect data on food contamination
6 by Listeria monocytogenes.
We want to get a data base or
7 are getting a data base.
And we'll collate the items into
8 various food categories.
And these have to be harmonic with
9 the Food Consumption Work Group's data base categories.
They
10 have a much more finely-resolved data base than we do.
They
11 have things like fish and fish with chips and so on,
12 finely-resolved meals and foods -- whereas, we in the
13 contamination area tend to have more grossly-resolved
14 components such as even seafood, whatever that means.
15 means everything.
16
It
So, we have to work that problem out.
Then we're going to determine the foodborne
17 exposure to viable strains of the pathogen in the U.S.A.
18 And this will then be used by the other people to relate
19 exposure to risk of human foodborne Listeriosis in the U.S.A.
20
Next slide, please.
Briefly looking at the
AIM REPORTING SERVICE
(773) 549 - 6351
25
1 Listeria, then, there are six species recognized today.
2 Monocytogenes is the one of most interest.
3 and animal pathogen.
Then we have innocua and seeligeri
4 which are not pathogenic.
5
It is a human
Welshimeri, ivanovii, and grayi.
Ivanovii is another pathogen but doesn't appear to affect
6 humans.
7
These are sort of grouped roughly into two groups
8 that reflect their occurrence in foods.
When one gets
9 Listeria contamination of foods, these are the ones most often
10 involved, I think it's fair to say.
Whereas, although these
11 can occur in foods, they occur less frequently.
12
Next slide, please.
The six species are typed once
13 one has a Listeria isolate by various simple tests.
14 not going to go into that today.
We don't need to.
And I'm
But I
15 will just mention that one of the tests that is used is the
16 test for hemolytic activity by the species.
17 the two pathogens are hemolytic.
And we see that
And seeligeri is also
18 hemolytic, though it's not really considered to be a pathogen.
19
I mentioned the hemolysis because you're going to
20 be hearing more about it, I think, this afternoon.
AIM REPORTING SERVICE
(773) 549 - 6351
So, I
26
1 thought I would introduce that.
2
Next slide, please.
In regard to food safety, some
3 important properties of Listeria and Listeria monocytogenes
4 are that it -- and the most important one is in terms of its
5 control -- is that it grows very slowly at refrigeration
6 temperatures.
It can also grow without air.
It's quite good
7 at surviving freezing.
8
Next slide, please.
And it's relatively resistant
9 to many of the preservation agents, whether chemical or
10 physical.
Looking at one of the physical agents, we can get
11 a 90 percent heat kill in about .2 minutes at 150 Fahrenheit
12 or 65 Centigrade, which is sort of a pasteurization-type
13 temperature.
It's relatively heat-resistant.
Of course,
14 that figure depends on what matrix one is looking at of food.
15
16
Some of the other organisms which are more
17 heat-resistant are recognized as the most heat-resistant,
18 vegetative forms of bacteria are the salmonella senftenberg
19 and coxiella burnetti and mycobacterium tuberculosis.
20
Another important property is that it's slightly
AIM REPORTING SERVICE
(773) 549 - 6351
27
1 more resistant than most bugs, leaving out staphaureus, to
2 low-moisture levels.
It can grow at about 10 percent salt,
3 which is equivalent to about 92 percent equilibrium relative
4 humidity or a water activity of .92.
So, all these factors
5 make it, are of great importance in considering food safety.
Next.
6
7 serotypes.
We've heard a little bit already about the
Monocytogenes and other Listeria species can be
8 sub-classified into various serotypes that depend on the
9 chemical composition of their outer layers, the flagella and
10 cell wall components.
11
And there are essentially seven types.
Five and six are missing, but they are covered by other
12 species or occur in other species.
Groups 1 and 2 are
13 classified together, so sometimes we'll say, "1, 2a" or if
14 we are a bit more slangy, we'll say, "1/2a, 1/2b, 1/2c."
15
These are the ones, then, that subdivide
16 Monocytogenes.
But it's not a perfect thing because one or
17 two, three, four of them also occur in other Listeria species.
18
19
Their main use is from the point of view, epidemiology.
Next slide, please.
The serotypes, although sort
20 of indicative of the ones that are most commonly occurring
AIM REPORTING SERVICE
(773) 549 - 6351
28
1 in clinical cases, or the ones that most commonly occur in
2 foods, are not perfect indicators.
But I will just try and
3 summarize the trends that are seen.
4
Listeriosis of any kind, whether foodborne or
5 otherwise, is due to all types, all serotypes.
6 1/2b and 4b are the most common types.
But 1-2a or
When we come to
7 foodborne outbreaks of Listeriosis, they are often due to
8 larger ones, to 4b.
9 type come into play.
But more recently, we've seen the 1/2b
The sporadic cases of foodborne
10 Listeriosis are most often due also to 4b, 1/2a and 1/2b.
11
Next slide, please.
Looking at the serotype
12 occurrence in foods, in dairy foods, we quite commonly find
13 4b, and the one type, particularly the 1/2a serotype.
In
14 meat products, we have the 4b serotype and the 1/2a, b's and
15 c's.
16
Plus, occasionally, 4ab and 4d.
In poultry products, we have the types 1 and types
17 4 and types 3.
More commonly, we have the 1/2b, 1/2c.
18 3b comes into play.
And
We haven't seen that much before in the
19 other bullets that I've shown you.
20
On vegetables, we find 1/2a, 1/2c and 4a, b and
AIM REPORTING SERVICE
(773) 549 - 6351
29
1 4b.
Coming to seafoods, another major area, we have the
2 1/2ab's and c's and the type 4's, particularly type 4b.
I
3 hope I haven't confused you, but I think the point is that
4 there are trends in the occurrence of the serotypes in various
5 kinds of foods and also trends in their occurrence in the
6 various cases of Listeria.
7
Next slide, please.
Coming now to the data
8 collection, our general approach has been to collect all
9 possible data without prejudice.
I'm a very nervous person,
10 so I like to be sure that we're going to try and get all the
11 data.
I'm afraid there won't be enough data.
I have a
12 horrible feeling, though, that there may be an avalanche of
13 data descending upon us.
Hopefully, some persons in this
14 room will contribute data of their own.
15
I hope so.
But doing this enables us to be choosy about what
16 we finally use in the analysis, what we will give to the
17 statistician to use, or what we can pick out for him to use
18 as he requests.
So, the choice of data to use will depend
19 on what's available and how it can be appropriately applied.
20
Next slide, please.
What sources of data are we
AIM REPORTING SERVICE
(773) 549 - 6351
30
1 using?
2
Well, we have a preference for the primary sources.
We'd rather look at the original publication rather than
3 some reference to it in a book chapter or a review.
But,
4 of course, the book chapters and reviews are good ways to
5 find the primary sources.
We're particularly interested in
6 publications in the scientific literature or in published
7 government documents.
And as we've already intimated this
8 morning, we're willing to consider other kinds of data.
So,
9 we're setting our net quite wide.
Move onto the next slide, please.
10
11 chronology.
This refers to what age of data we're going to
12 be looking at.
13 data.
14
Data
Again, we're going to consider all ages of
This is from about 1980 up to the end of this decade.
Obviously, we prefer the most recent data if sufficient is
15 available.
16 current.
That is, we want our exposure estimate to be
But if we get enough over the total time period,
17 then perhaps we can do some temporal comparisons.
Did things
18 change from the early 90's into the late 90's, that kind of
19 thing.
20
Next slide, please.
Geographical origins of data.
AIM REPORTING SERVICE
(773) 549 - 6351
31
1
Again, we're collecting data from all countries -- North
2 and South America, Western Europe.
We have data from the
3 Far East, a little bit from the Mideast, a little bit from
4 North Africa and a little bit from Australasia.
We have a
5 lot from Western Europe and a lot from North America.
I guess
6 they're the major regional areas.
Obviously, we have a preference for U.S. data.
7
8 But data from other industrialized countries that are
9 somewhat similar to the U.S. -- I like to think of England
10 in that category -- you know, we'll consider that, too, if
11 we have to.
So far, we've noticed, looking at the regions, that
12
13 contamination is universal.
14
It's not a surprise, of course.
But occurrence rates are not -- at least, looking at it off
15 the top of your head kind of look -- not dramatically
16 different.
That is, there's not a hundred-percent
17 contamination of foods in some countries and zero in others.
18
They all have some contamination between zero and a hundred
19 percent.
20
Move on, please.
Next slide, please.
AIM REPORTING SERVICE
(773) 549 - 6351
We've
32
1 mentioned subtypes.
The serotypes, in particular, they're
2 not always reported in the studies.
We'll keep an eye on
3 them and see if we get enough to do anything with, but we're
4 not -- it's not top in our priority at the moment.
But maybe
5 if we have enough, we'll change our mind.
6
We have to recognize that most food isolates have
7 tested virulent when they've been laboratory-tested.
8 talking monocytogenes, of course.
I'm
And we have to assume in
9 this analysis right now that all food isolates are equally
10 virulent in nature.
Subtype analysis of any kind -- serotype
11 or DNA type or whatever -- is temporary, at least, not a high
12 priority.
13
Next slide, please.
14 are we finding?
What types of occurrence data
Well, we prefer quantitative data.
15 there's not an awful lot of it.
16 very gratefully accepted and used.
But
But what we have will be
Quantity of data, I mean
17 by the colony-forming units per gram or mil of food.
It's
18 important here to get the total number of samples examined
19 and then the number occurring in given density ranges of
20 contamination.
And from that, we can get a percent.
AIM REPORTING SERVICE
(773) 549 - 6351
33
Percent, per se, is useful for weighting the data
1
2 between different observations.
But we have to be careful
3 of that because, obviously, the statistics get better, the
4 greater of number of samples examined for a given food
5 category.
Most of the data, as I've intimated, is
6 qualitative data, presence or absence in a food.
Again, the
7 number of sample examined is important to know that from the
8 statistical point of view.
And then, from that, we get the
9 number positive.
10
11 study.
Often, one can get both types of data in the same
Perhaps I should put that another way.
Most of the
12 data is qualitative data, and sometimes it is within that
13 study.
Also, some quantitative data.
14 to put that.
I don't think we have any examples of studies
15 with just quantitative data alone.
16
That's the best way
Next slide, please.
Maybe one.
Some people have questioned
17 what's the use of presence and absence data.
18 go into that now.
But it is useful.
And I won't
I think it can be really
19 just equivalent and just as useful as the concentration data.
20
AIM REPORTING SERVICE
(773) 549 - 6351
34
Generally, the analytical portion used in these
1
2 studies is 25 grams from a non-composite sample of the food
3 type.
Obviously, if there's a variation from that, we're
4 going to have to correct our data or standardize it and correct
5 the analytical portion size used.
And we may have to think
6 about composite sample types of data, correcting that, too.
7
Next slide, please.
8 sensitivity.
Isolation method,
This is an important factor with the
9 quantitative data.
10 colony count data.
Generally, the quantitative data are
So, we're sort of in the area of 50 to
11 a hundred cfu's per gram is sort of the minimum level
12 detectable, depending on the sample volume one place.
But
13 with the MPN's, depends again on the number of tubes used.
14
But, typically, they would only be three or five.
And so,
15 we would be somewhere in this detection range, minimum
16 detection range.
17
The qualitative methods, generally people are
18 using well-known standard methods, which seem to have
19 comparable sensitivity.
And so, with those methods, we can
20 detect at least one colony-forming unit per 25 grams.
AIM REPORTING SERVICE
(773) 549 - 6351
And,
35
1 as I said before, sometimes both kinds of methods are used
2 together.
For instance, you screen the foods, look for the
3 samples that are present or have Listeria mono in them.
And
4 then you say -- go back to it and count it within a day or
5 so.
If there are under-estimates of occurrence,
6
7 clearly, they will tend to err on the side of safety. More
8 or less, they err in a safe way.
Next slide, please.
9
10
This is a point that came up.
We're assuming, really, in this analysis, that the analysis
11 time and the ingestion time differential does not
12 significantly effect the counts consumed.
13 on average, they will tend to agree.
That is because,
But, clearly, a
14 particular food might have been counted at one time and
15 perhaps held a much longer time at refrigeration temperature
16 before it was consumed.
But this will be -- this kind of
17 error will be less critical, I think, with the short-life
18 products but perhaps more critical with longer shelf-life
19 products.
So, we're assuming the count will represent a
20 potential ingested dose.
That is, at the time we count the
AIM REPORTING SERVICE
(773) 549 - 6351
36
1 food, perhaps someone has already bought it; and then the
2 day we analyze it, they're eating it.
3 situation.
4
That's the best
But, obviously, it's a big assumption.
In difficult cases, we may want to look at survival
5 studies for monocytogenes in critical products.
There's
6 plenty of data on survival of monocytogenes in various kinds
7 of foods.
8
Next slide, please.
9 themselves?
But what about the foods
Well, the main emphasis clearly is on
10 ready-to-eat foods.
These are not always clearly defined
11 in the contamination studies we're looking at.
We're also,
12 though, going to look, I think, a little bit at undercooked
13 foods -- that is, partially-cooked hamburger and that kind
14 of thing.
It is possible to make some ballpark estimates
15 of the levels of mono one might ingest in a partially-cooked
16 hamburger.
We're not really looking at rewarmed, cooked,
17 chilled foods or cooked leftovers.
We're not really
18 considering that so much, how well they were reheated and
19 that kind of thing.
20
We're collecting data on raw foods.
AIM REPORTING SERVICE
(773) 549 - 6351
But, again,
37
1 it will be difficult to use that in this analysis.
2 people are eating that much raw food, I believe.
Not many
And then,
3 of course, I think we've already talked about this.
But
4 there's the harmonization of the contamination of dietary
5 data.
We're going to have to have appropriate pooling of
6 the food-type data.
7
Next slide, please.
You're going to hear more
8 about the groups of foods that are being studied in regard
9 to the dietary or ingestion data.
But we've done some partial
10 harmonization looking at our contamination data, talking to
11 Mary Bender and the Dietary Intake Group.
And so, we have
12 a major category of dairy foods broken down into cheese, ice
13 cream, milks and something called miscellaneous, which could
14 include butter and so on.
And, again, some of these
15 categories are broken down into appropriate further
16 breakdowns such as soft cheese versus others, or raw milks
17 versus pasteurized milks.
18
Next slide, please.
Fresh produce.
We're
19 concerned about vegetables that are eaten raw here in salads
20 or sandwiches -- those that are grown in the air away from
AIM REPORTING SERVICE
(773) 549 - 6351
38
1 the soil and those that are grown in the soil. Clearly, these
2 might be more likely to harbor Listeria monocytogenes.
3 There's something called miscellaneous vegetables,
4 catch-all.
And then we have fruits that are eaten raw, those
5 that grow or are grown near the soil and those that are growing
6 distal to the soil.
Seems an appropriate breakdown, if
7 possible, in regard to potential for contamination.
Next slide, please.
8
Juices, we're looking at
9 fruit and vegetable juices, pasteurized and raw.
Next slide, please.
10
I should say that with all
11 these groups that we've come out with, this is a tentative
12 list.
I don't say we have much or even any data for all of
13 them at the moment.
Salads, vegetable, fruit and nut salads -- that
14
15 is, salads without protein added, animal protein items added.
16
And then your other kinds of salads with the animal protein
17 items.
18
And something called miscellaneous mixed salads.
Next slide, please.
Coming to the meats.
We
19 break each kind of meat down into raw, ground and cooked meats,
20 in general.
And we have here the beef, pork, lamb and poultry
AIM REPORTING SERVICE
(773) 549 - 6351
39
1 as the major groups.
Next slide, please.
2
Other meats and products are
3 important, of course, in the ready-to-eat area.
We have our
4 deli and luncheon meats, the bolognas, the hot dogs, fermented
5 meat products and other kinds of meat products.
6 sausages.
I presume this includes things like bologna and
7 -- well, bologna is up here.
8 jerky.
We have
But salami and so on.
I have something for exotic meats.
9 spreads and pates.
The meat
Meatloafs,
And then the egg products have been put
10 here.
11
Next slide, please.
Very important category in
12 the ready-to-eat area, of course, and the food preparation
13 area is sandwiches.
Broken down into burgers -- the
14 cheeseburgers, hamburgers.
And then deli items, the various
15 meats, eggs, seafood and veggies.
16
17 category.
Next slide.
I think this is the last major
We're considering seafoods of the ready-to-eat
18 and raw type.
In fish categories, the shellfish, smoked
19 seafoods, and then anything else.
20
Next slide, please.
We have a few other food
AIM REPORTING SERVICE
(773) 549 - 6351
40
1 categories that seem to be miscellaneous in character.
2 we have Mexican-style cheese and on-cheese dishes.
But
Some
3 kinds of salad dressings such as blue cheese and things like
4 pastries that are cream-filled.
Next slide, please.
5
Our results, we have over a
6 thousand lines of data, 400 kilobases.
Seems to vary whether
7 you move one line from the database or add it back in.
8 can change going low.
It
But, anyway, gives you some idea of
9 the collection so far.
They're in various separate data
10 bases by the members of the work group at the moment.
11 they're gonna have to be combined.
So,
And we've essentially
12 covered seafoods, vegetables, cheeses, meats, poultries and
13 sandwich.
14 items.
That is, we haven't covered all the data on these
But we have fair amounts of data on all of them.
15 And, as I mentioned before, the milks are just being started,
16 raw milks and pasteurized milks.
17
Next slide, please.
In the document that you
18 probably have or I hope you have got out front, some of the
19 kinds of data we have are in there.
But just to run over
20 briefly the kinds of data that we're collecting and data base
AIM REPORTING SERVICE
(773) 549 - 6351
41
1 so we have a reference, this happens to be an acronym for
2 the West North Yorkshire Joint Working Group that's been
3 published in '91.
Large survey in the UK.
So, we have the
4 country, and then we'd have the food types examined by the
5 work group.
And then the components, type of food and the
6 categories.
So, we'd have a deli item with meat in the
7 sandwich category.
And then we'd record whether it's
8 ready-to-eat or raw.
And then what the species is --
9 hopefully, monocytogenes.
Next slide, please.
10
11 base sample a little more.
12 of sandwich.
13
This is the continuous data
Here were the data for this kind
And they looked at 47 and found 7
positive.
One of them, they didn't get any quantitative data.
The
14 other six, they got various kinds of quantitative data for
15 the six, the six positive ones.
16 than 20 cfu.
So, 1 out of 46 had less
They were plating half a mil, so pushing the
17 plating technique.
18
2 out of 46 were in this range.
19 in this range.
20 gram.
1 out of 46 were
And 2 were greater than a thousand cfu per
And they used the 25 gram sample for analytical portion
AIM REPORTING SERVICE
(773) 549 - 6351
42
1 size.
2
Obviously, we'd like, really, to get a lot of data
3 like this because it gives us some kind of distribution of
4 the various concentration levels, how frequently they occur.
5
In this particular example, there seem to be quite a lot
6 of high proportion of high-level contamination.
7 not always true when you look at other foods.
But that's
So, that gives
8 you some idea of the
9 kinds of data we're collecting.
10
Next slide, please.
11 complete our data gathering.
Obviously, we have to
And, hopefully, we will get
12 some more from volunteers outside of the Government.
Finish
13 the milks, in particular.
We've got to combine the data bases
14 and make them consistent.
We've got to edit it and pool the
15 data into the categories that I've sort of mentioned.
Then
16 we can sort the data into all Listeria species or just
17 monocytogenes species data.
And we can sort it into density
18 or presence and absence data.
19 important, I've decided.
Though that isn't quite so
We have to select amongst the data
20 for the ready-to-eat versus the raw.
And we have to collate
AIM REPORTING SERVICE
(773) 549 - 6351
43
1 it with the dietary data.
2
So, as I said, we're being quite Catholic in our
3 collection of data.
4 hold of.
We're grabbing everything we can get
I think that's gonna be important in terms of
5 determining at least an overall frequency distribution of
6 the data because, obviously, we're interested in any kind
7 of contamination level.
But we're particularly interested
8 from the point of view of possible disease in the:
how
9 frequent are the higher levels of contamination in the various
10 kinds of foods; how frequently does a food type have a count
11 of, let's say, ten to the three to ten to the four?
And,
12 hopefully, we'll be able to correlate that with frequency
13 of disease.
14
Get a match-up, if you like, a titration.
Next slide, please.
So, the results will be
15 estimates of the amounts of viable Listeria monocytogenes
16 in U.S. food and food subgroups.
And the estimates will be
17 in the form of pathogen and cell density frequency
18 distributions.
19
Next slide, please.
I don't have to introduce the
20 next speaker because I understand the committee is going to
AIM REPORTING SERVICE
(773) 549 - 6351
44
1 answer any questions there might be on this.
But to review
2 what we've covered, we've looked now at the contamination
3 module, the contamination rate for monocytogenes in foods,
4 and that the data collection for that, that has to be
5 multiplied by the consumption rate to give us an exposure
6 rate.
And then that exposure rate has to be used to derive
7 some function which will give us the frequency of Listeriosis
8 and, in particular, foodborne Listeriosis, relate these to,
9 in a risk analysis by the statistician.
10
So, without anymore ado, I'll close and let the
11 committee introduce the next speaker.
12
MR. MICHAEL JAHNCKE:
Thank you.
Thank you, Dr. Hitchins.
13 Are there questions from the subcommittee for Dr. Hitchins,
14 please?
Dr. Hitchins, if you could just wait a minute.
15 There are some questions from the subcommittee.
Dane
16 Bernard, please.
17
MR. DANE BERNARD:
18 for your presentation, Tony.
Thanks.
Dane Bernard.
Thanks
Looking at the data that you've
19 already collected -- and I recognize that you're not very
20 far on the dairy portion of your data collection -- but what
AIM REPORTING SERVICE
(773) 549 - 6351
45
1 would you say would be your greatest need and what are those
2 product areas?
DR. TONY HITCHINS:
3
4 Dane.
Well, that's a good question,
It depends on what you mean by, "need."
But, for
5 instance, in the area of fruits and fruit juices, I wouldn't
6 say we had very much, if any, data.
But whether there's a
7 real need for it, I don't know because -- it would be nice
8 to know what's there, you know, that kind of thing.
I think dairy area is fairly well-covered.
9
10 Seafoods is pretty well-covered, but one doesn't always know
11 what they mean by, "seafood."
They tend to lump things
12 together, and one doesn't always know whether it's
13 ready-to-eat and raw together and so on.
So, that's going
14 to have to take careful sorting.
Meats, we have a lot -- I think that's the major
15
16 groups.
Have I missed one?
Sandwiches.
In this country,
17 I would say sandwiches -- approaching your question another
18 way, we have a lot of data from various categories worldwide.
19
But in any given country, we may not have much data.
And
20 we have a lot of data on sandwiches in Northern Ireland.
AIM REPORTING SERVICE
(773) 549 - 6351
46
1 But, you know, maybe we need more on sandwiches in the U.S.
2
Yeah.
Thank you.
3
Quantitative data is needed, too.
As I mentioned,
4 though, the presence and absence data is, if you think about
5 it very carefully, a set of quantitative data.
6 reflect a distribution.
It does
The shape of that distribution, one
7 can either make assumptions about it, or one can look at the
8 quantitative data we have and see whether the distribution
9 is something we might expect, such as a log normal
10 distribution or whether it differs from a log normal
11 distribution.
12
So, I believe the presence and absence data will
13 tell us a lot more than just presence and absence.
14
MR. MICHAEL JAHNCKE:
15
MR. MIKE DOYLE:
16 questions.
Mike?
Mike Doyle.
Tony, a couple
Would you be asking for the methods that are used
17 to generate these data?
And, if so, will you take the methods
18 into consideration as to whether the data are acceptable or
19 not?
20
DR. TONY HITCHINS:
Yeah.
Thank you, Mike.
AIM REPORTING SERVICE
(773) 549 - 6351
Any
47
1 data we get, you know, we'd be glad to see the method.
I
2 think it should be -- if you can tell us the method, too,
3 that would be very helpful.
Particularly if it's a method
4 that's perhaps not one of the standard methods such as the
5 FDA method or the FSIS method or the Dutch method or the Nordic
6 group method.
You know, if it's not one of those, then we'd
7 like to know it, I think.
8 any good.
9
It doesn't mean to say it's not
But we would like to know, if possible.
MR. MIKE DOYLE:
10 classification.
You mentioned sandwiches as a
And I think that's great.
But what's in
11 the sandwich is probably more important because salami might
12 be different than chicken salad, for example.
I think you
13 may want to break those out into sandwiches and the various
14 ingredients within those sandwiches, rather than lumping it
15 into one group.
16
DR. TONY HITCHINS:
17
MR. MIKE DOYLE:
18 do with sprouts.
Right.
The other question I have has to
I didn't notice that up there.
And
19 certainly, we have a strong interest in sprouts today.
20 I wonder if there might be a focus in that area.
AIM REPORTING SERVICE
(773) 549 - 6351
And
48
DR. TONY HITCHINS:
1
Well, with regard to the
2 sandwiches, we have a crude resolution there into the meat
3 and non-meat types of sandwiches.
And I think the breakdown
4 is really going to be limited to what is available in the
5 data.
I mean, you know, if we have a lot of chicken sandwich
6 data and we have a lot of salami sandwich data, that's fine.
7
But if we don't have much, we're sort of reduced to pooling
8 into a category of meat sandwiches or, perhaps, poultry
9 sandwiches versus meat-type sandwiches and seeing what we
10 can get out of it.
But you're quite right.
11
12 those to be separated.
13 mentioned.
Ideally, we would want
Sprouts are not specifically
Perhaps they should have been.
And I would put
14 them in there somewhere, I think, under fresh vegetables.
15
Do you have a disagreement about that, or would you rather
16 have that kind of product separated out?
17
MR. MIKE DOYLE:
I just want to make sure we don't
18 overlook sprouts.
19
DR. TONY HITCHINS:
20
MR. MICHAEL JAHNCKE:
All right.
Peggy?
AIM REPORTING SERVICE
(773) 549 - 6351
49
MS. PEGGY NEILL:
1
Peggy Neill.
2 go back and just ask you to clarify.
3 it.
A couple of things.
One is:
I just wanted to
Maybe I just missed
In order for the data to
4 be acceptable to be included in the data base, it will need
5 to be a isolation as opposed to a molecular detection method?
6
Is that correct?
7
DR. TONY HITCHINS:
Well, speaking as a regulator,
8 we always like the data to be for the organism that is being
9 isolated and we have it in our hand.
But I think for the
10 purposes of this exercise, I think that molecular type
11 presence/absence or quantitation data would be useful.
12 Yeah.
13
I mean, would be acceptable to me, anyway.
MS. PEGGY NEILL:
At least to look at -- although,
14 then, hard to know whether at this point it would necessarily
15 go into the data base?
16
DR. TONY HITCHINS:
It can go into the data base
17 certainly, yeah.
18
MS. PEGGY NEILL:
19
DR. TONY HITCHINS:
Okay.
But whether we use it -- any
20 data in the data base may not be used in the analysis, okay.
AIM REPORTING SERVICE
(773) 549 - 6351
50
1
You know, that's going to be our statistician and other
2 people's choice what is actually used.
I hope I've given
3 the impression that we'll accept any data, and then we'll
4 see what we've got, what we can do with it.
I mean, obviously,
5 you have some ideas already what we can do with it.
6 no, that data will be very acceptable, Peggy.
MS. PEGGY NEILL:
7
But,
Yeah.
The other point that I wanted
8 to make sure that we all understood was that you had fairly
9 early on a slide in which you were addressing how not all
10 of the studies may have sub-typed or serotyped isolates.
11 But then you also made a point that you will be includ -12 the assumption is basically that all L.M. are virulent
13 regardless of whether additional virulent --
DR. TONY HITCHINS:
14
15 personal assumption.
16
(interrupting)
Other people may not agree with me.
I mean, obviously, not all L.M. are virulent.
17 certain types that are not virulent.
18 terribly frequent.
19
20
Yeah.
That's my
There are
But they're not
So, one is isolating those rather rarely.
Did I answer your question?
MS. PEGGY NEILL:
So, the assumption for inclusion
AIM REPORTING SERVICE
(773) 549 - 6351
51
1 in this module is basically if L.M. -2
DR. TONY HITCHINS:
(interrupting)
Any L.M.,
3 yeah, you know, we found L.M. in X samples of food product
4 by this method.
5
6 question.
MR. MICHAEL JAHNCKE:
Then we'll save it.
7 discussion a little bit later.
8
MR. ROBERT BUCHANAN:
We'll have one last
We will have a committee
Bob?
Bob Buchanan, FDA.
Tony,
9 the scientific literature has basically two primary sources
10 of information about Listeria in foods.
One of them is the
11 survey data that you have indicated you're going to
12 incorporate in your data base.
The other is a rather
13 extensive literature on inoculated pack studies in
14 determining what foods will and will not grow Listeria, under
15 what conditions or growth rates, et cetera.
I didn't see
16 any indication at all that you planned to use that one whole
17 group of data.
18
How are you going to handle the research that has
19 been provided in terms of inoculated pack studies,
20 experimental growth studies, et cetera, which probably has
AIM REPORTING SERVICE
(773) 549 - 6351
52
1 the best quantitative data that is available because it's
2 been done under usually fairly controlled conditions?
3
DR. TONY HITCHINS:
Well, I sort of agree with you
4 that we, perhaps, should make use of that kind of data.
But
5 I can't quite see how putting ten to the three Listeria into
6 a food and seeing what happens to it has usefulness as telling
7 us what is out there and what is consumed by the public.
8 It tells us things about how likely it is a contamination
9 with mono will develop into a larger population of mono in
10 the food or whether it will decline or stay constant.
And
11 that would be an important breakdown in terms of the same,
12 "Well, these are the foods in which mono will grow."
13
Is that what you mean?
14
MR. ROBERT BUCHANAN:
15 level.
Not only grow but to what
It just seems to me that you're missing an entire
16 source of data that needs to be capped or at least examined
17 in some way.
18
DR. TONY HITCHINS:
19 in our mandate, I guess.
Well, you know, that wasn't
It was to see what the contamination
20 of foods was in nature, not in the laboratory in terms of
AIM REPORTING SERVICE
(773) 549 - 6351
53
1 inoculating them.
But we can certainly incorporate those
2 kinds of data if people care to send them to us.
MR. MICHAEL JAHNCKE:
3
Excuse me.
To stay on
4 schedule, we will have, after Dr. Bender is in a break, a
5 general committee discussion.
6 discussion.
We can continue on with this
But to stay on schedule, I would like to thank
7 you, Dr. Hitchins, for a very well-organized and very
8 excellent presentation.
Thank you very much.
Our next speaker is Dr. Mary Bender.
9
And she will
10 be talking about food consumption patterns.
DR. MARY BENDER:
11
Thank you.
Yes, I would like
12 to bring you up to date with the progress that we've made
13 so far with our food consumption module.
If we had an
14 unlimited amount of time, we would probably not really get
15 into it heavily until Tony has finished his first module.
16
But there was no way to do this, so we've kind of plunged
17 in headfirst and are really in progress of still addressing
18 the issues.
19
20 team.
First, I would like to bring your attention to our
The team has evolved over the last three-and-a-half
AIM REPORTING SERVICE
(773) 549 - 6351
54
1 months that we've been doing this.
2 Food Labelling at FDA CFSAN.
I work in the Office of
I'm not sure if anyone has had
3 microbiology.
I have, and I'm a research methodologist
4 statistician.
But we do work with the food consumption
5 composition sales, labelling, whatever data bases, and do
6 use data provided by other agencies and also do data
7 collection, have it available to the world.
Also, the team has worked with consultants from
8
9 ARS and from CDC on their specific data collection tools.
10
And it's been very, very helpful.
11
The purpose of this model is to model a consumption
12 of foods that have a high potential for contamination by
13 Listeria monocytogenes, which brings us to the first question
14 of what foods are at greatest risk for contamination.
And
15 not having a background in Listeria monocytogenes -- I've
16 heard about a few outbreaks related to cheese or hot dogs
17 -- and I thought, "Oh, this will be a piece of cake.
Put
18 in a little bit of computer code and come out with what's
19 there."
But looking at the literature, it was very apparent
20 very quickly that Listeria is in many, many foods, as Tony
AIM REPORTING SERVICE
(773) 549 - 6351
55
1 and others have mentioned.
And so, trying to look at the
2 case data for Listeriosis, the outbreaks, the sporadic cases.
Also, recalls by the U.S. government and Canadian
3
4 government, as well as some of the analytical testing data,
5 although I didn't really get into that too thoroughly.
As Tony mentioned, Listeria is in many raw,
6
7 unprocessed foods and at grocery refrigeration temperatures.
8
And freezing doesn't necessarily kill it.
9 heat-resistant.
10 Listeria.
11
And it can be
Cross-contamination in the home can spread
And it may be on foods that are ready-to-eat.
Next.
Which brings us to the next question:
12 First reaction was just to make all these foods more
13 manageable.
We should look at different groupings.
Also,
14 it's critical that we do look at food categories in order
15 to allow the merger of the data from the contamination module
16 and the consumption data.
17 easily.
And they don't necessarily merge
So, it is a challenge that we're continuing to
18 address.
19
Our categories are evolving.
Tony listed the
20 categories, and I'll go into a little bit of the information.
AIM REPORTING SERVICE
(773) 549 - 6351
56
1
We have several meat categories.
I know there's a lot of
2 analytical testing data on various meats, recalls also.
3
Next.
Poultry, I know there have been some
4 outbreaks related to poultry, sporadic cases that have been
5 cooked.
Deli-luncheon meats is a very important area,
6 especially with the latest outbreaks and also a number of
7 recalls in the last, I guess, six months from the Midwest,
8 but recalls from a number of places.
9
We're still trying to get the sausage categories
10 straightened out.
Luckily, the hot dog/sausage website did
11 have some kind of explanation of the various sausages.
But
12 we're thinking in terms of like salami and pastrami or
13 whatever going in the fermented area.
But I don't know.
14 Tony and I are still working on this.
15
And then the deli meats listed there, as well as
16 miscellaneous bulk and link sausages.
And we had thought
17 that that was primarily like the breakfast sausages, because
18 there have been recalls related to those areas.
19
Next.
Listeria has been identified on jerky.
And
20 as far as exotic meats, we not only mean the game meats like
AIM REPORTING SERVICE
(773) 549 - 6351
57
1 venison or buffalo or rabbit, but also exotic concoctions.
2
There was an outbreak in Europe related to pates, and I know
3 ham roulettes.
And there was pork tongue in jelly -- and
4 I don't know who would eat that anyway, but maybe it's really
5 good.
But we do want to cover whatever we can find.
As far as fruits, we've been able to find very,
6
7 very small amount of information in the literature.
I did
8 find one article on AIDS patients and Listeriosis and
9 contacted Dr. Mescall (phonetic) at Los Angeles County Health
10 Department.
She was thrilled to death that we were doing
11 this study and said that they had unpublished data on unwashed
12 grapes.
And she also mentioned a vegetable I had never heard
13 of called jicama, that apparently it was like a potato.
14 you eat it raw and slice it.
But
And they did find data that
15 they didn't publish.
In one of the really good textbooks, it's a Riser
16
17 and Marth [phonetic].
18
And the update came out this year.
They cited an outbreak related to strawberries, blueberries
19 and nectarines.
20 read it.
And so, I ordered Dr. Schlech's article and
And none of the people on the team contacted him.
AIM REPORTING SERVICE
(773) 549 - 6351
58
1
And he referred us to a Dr. Lin at CDC, I believe, who is
2 in Viet Nam -- and we haven't heard from him yet -- to find
3 out exactly what this outbreak was.
I know Tony did find analytical data on plums and
4
5 peaches and raisins.
But we need to identify fruits that
6 could be a problem, and I'm not really sure what they are.
7
I know there was a recall of frozen blueberries within the
8 last year, was another one.
But I don't know what to do with
9 that yet.
Okay.
10
Next.
As far as vegetables, we have it
11 broken into those vegetables eaten raw and miscellaneous
12 vegetables.
And the raw, with the raw vegetables being
13 broken into those grown above the ground and below the ground.
14
I know there was one outbreak in the U.S. that was -- I guess
15 there was some epidemiological link to lettuce, tomatoes and
16 celery.
Well-known outbreak in Canada related to coleslaw,
17 which is the cabbage.
Sprouts.
We did get sprouts in there.
18
19
Within the last year, there were a number of recalls
20 of various kinds of sprouts.
They looked like they were
AIM REPORTING SERVICE
(773) 549 - 6351
59
1 processed, though, because they had crunchy sprouts and dill
2 sprouts and whatever.
3 screen today.
But that's how sprouts made it on the
Grown below the ground.
And radishes.
I
4 know that there's analytical data on a number of vegetables
5 with not a lot of results.
But the radishes did show
6 Listeria, as well as some potatoes.
But I'm not sure who
7 eats raw potatoes, so I didn't put it up there.
8
Next.
Listeria is a very big area for Food and
9 Drug Administration.
For cheese, it isn't going to end up
10 being as simple as soft and other.
11 where we're going to go with that.
But I'm not sure exactly
Again, it rests a lot
12 on what Tony finds with the contamination data.
But there
13 have been outbreaks -- a well-known one in the United States
14 with Mexican cheese, and various outbreaks in Europe.
15
As far as ice cream, I did see that there was an
16 outbreak related to ice cream.
And there have been recalls
17 of ice cream and ice cream products with Listeria.
18
Fluid milk, we're still working with this.
And
19 there will be some slides later where I go into it a little
20 bit more deeply.
But there have been outbreaks in the United
AIM REPORTING SERVICE
(773) 549 - 6351
60
1 States linked to pasteurized chocolate milk.
And another
2 with -- I think it was Holland -- maybe 2 percent lowfat milk.
3
And outbreaks in other countries from the raw milk.
We have a category for miscellaneous dairy products
4
5 and know there was an outbreak related to cream, butter.
6 There have been recalls related to the other products listed
7 up there and probably some additional ones.
Seafood is an important area.
8
I immediately
9 jumped down to smoked seafood because that's most of the
10 literature that I have found that has linked Listeria with
11 outbreak due to smoked mussels in New Zealand and Australia.
12
Another outbreak related to smoked rainbow trout.
That was
13 someplace in Europe and sporadic cases also with smoked salmon
14 and smoked cod roe.
15
I know that there have been -- I'm not sure what
16 -- I guess recalls related to ready-to-eat seafood.
17 this really is an important area.
And so,
And I'm sure we'll get
18 into these categories and look at the relationship with
19 Listeria.
20
This one broke my heart.
Tony found some
AIM REPORTING SERVICE
(773) 549 - 6351
61
1 analytical data for cream-filled pastries, I believe, in the
2 United Kingdom.
3 here."
And my first thought was, "Good.
It wasn't
But I know that there have been recalls related to
4 whipping cream and other dairy products, so the possibility
5 does exist.
I found very little information on juices.
6
There
7 was one article that came out of FDA in Seattle where they
8 tested fruit juices and found Listeria in unpasteurized apple
9 and apple-raspberry juices.
And I did contact the author
10 who said that it was, I guess, in a jar.
11
It was unpasteurized.
12 Listeria.
13
It was jarred juice.
But they did not test for levels of
It was strictly presence and absence.
Salads are also important.
I know there have been
14 recalls related -- I'm kind of jumping around -- to different
15 types of meat, fish, salads.
I know that there have been
16 several outbreaks related to potato salad in this country
17 and others.
I don't have details on that.
Hummus, there
18 have been recalls related to hummus, various types of hummus,
19 and that analytical testing, I think, primarily -- I believe
20 Tony said in Ireland -- I know it was in the United Kingdom
AIM REPORTING SERVICE
(773) 549 - 6351
62
1 someplace -- where they did look to see what ready-cut salads
2 had Listeria.
And I know the International Cut Produce
3 Association is really interested in this area and is doing
4 everything they can to try to provide us with safe cut salads.
Again, we've isolated or we've put down burgers
5
6 and deli.
I'm wondering if we'll end up including hot dogs
7 also as another category.
I know the USDA regulates meats;
8 but if you put it in a bun or between bread, it falls into
9 FDA's jurisdiction.
So, we are interested in the sandwiches
10 that have been implicated with Listeria.
I know there have
11 been a couple recalls of frozen cheeseburgers.
And I have
12 no clue to what level you'd have to heat the products in order
13 to eat them.
But they have had Listeria, and they have been
14 in recalls.
15
Other foods.
This is still evolving.
I'm not
16 sure where Mexican-style foods will -- whether it will
17 continue in here.
There was a recall related to chicken
18 burritos, a recall related to blue-cheese salad dressing.
19
And one book had raw eggs implicated an outbreak in the U.S.
20
But I haven't been able to get any further to find out if
AIM REPORTING SERVICE
(773) 549 - 6351
63
1 that was really true.
Okay.
2
3
Which brings us to the next question as to:
What are the best sources of food consumption data? There
4 are two large-scale U.S. food consumption surveys.
And those
5 are the tools that we're going to use to answer our questions.
6
First is an ARS survey that has been going on -- I don't
7 remember -- 20 years.
8 remember.
Maybe more than that.
I can't
It's known as CSFII, the continuing survey of food
9 intakes by individuals.
And this is the most current food
10 consumption data that are available.
The survey collects two, 24-hour recalls of foods
11
12 eaten.
It is a probability sample of respondents.
And the
13 survey does have weights whereby you can look to find out
14 -- you weight the sample size and also weight the amount eaten.
15
And it will give you a national probability sample.
But
16 it is food that is eaten, non-institutionalized people.
17
So, when you get down to the bottom and see the
18 number of respondents, it would not account for adults 65
19 and older who are living in nursing homes or other types of
20 assisted living.
Also, the sample of pregnant or lactating
AIM REPORTING SERVICE
(773) 549 - 6351
64
1 women is not actually large enough to be nationally
2 generalizable.
But this is where we are with this.
CSFII has collected data under an EPA contract for
3
4 children, and there will be a lot of data available to EPA
5 at the end of the year and to FDA and others right after the
6 first of the year.
7 in those data.
But it won't be ready in time for this survey.
The second survey is not strictly food consumption.
8
9
But maybe later we'll go back and pull
It's the National Health and Examination in Nutrition
10 Surveys.
And these have been conducted by CDC by National
11 Center for Health Statistics in D.C. area for a long time.
12
I don't remember exactly.
These data are older than the
13 CSFII, and they do have one 24-hour recall of foods eaten.
14
They do also have test measurements, body measurements of
15 the respondents.
But to our chagrin, none of the
16 measurements could really be used to determine
17 immunocompromized conditions.
18
It's a probability sample.
19 to reflect the U.S. population.
They provide weights
Many respondents -- probably
20 a large sample of pregnant or lactating women, but probably
AIM REPORTING SERVICE
(773) 549 - 6351
65
1 not large enough to really be truly representative, even
2 though it's weighted.
3
I believe OMB of the government is the source that
4 asked CDC and ARS to combine their surveys.
So, starting
5 in 2000, there will be separate data collection, but they'll
6 be using the same sampling design.
7 of other changes.
There will be a number
But future data will be -- they'll be able
8 to combine the different groups.
And I know that they're
9 going to attempt to collect data on more pregnant or lactating
10 women.
Last year, we looked into seeing if we could give
11 them funding to double the sample.
And that was -- I think
12 it was something like 500,000 for each agency.
And that
13 didn't include the testing.
So, there are a lot of
14 consumption data available.
And it just didn't look like
15 the best place to invest the funds.
16
The next question:
17 were 7,500 food codes.
I saw somewhere where there
And we are looking at all of these
18 food codes to figure out exactly what should go into the pot
19 because, obviously, all foods are not implicated.
Until I
20 hear differently from somebody who knows a lot more than I
AIM REPORTING SERVICE
(773) 549 - 6351
66
1 do, we aren't going to put bread or cookies or a number of
2 foods.
This is just one example.
3
I know there were over
4 a hundred food codes linked to cheese.
And there will be
5 some later slides that explain what we've tried to do.
Some
6 of these cheeses would be at greater risk and others wouldn't.
The next question:
7
What measure of food
8 consumption will best represent exposure?
Both surveys
9 provide data as the amount eaten in grams per eating occasion.
10
And I know CFSAN issues those data a lot in order to figure
11 out serving size because Congress said in the early 90's,
12 "You will figure out the serving size based on the amount
13 of food that's customarily consumed."
14 in eating occasions.
So, we will have data
But speaking with the modeler, Clark
15 Carrington, he said he would like the amount eaten per person
16 per day.
And it will be in grams.
And we can also figure
17 the proportion of the population who are eaters, and that's
18 really important.
19
Our steps.
We will select the appropriate Food
20 Codes and for each Code determine the amount of the food eaten
AIM REPORTING SERVICE
(773) 549 - 6351
67
1 per person per day in grams.
2 population.
Weight the data to reflect
Sort the data in to the groups that we've
3 mentioned earlier.
And then, in some instances, merge data
4 from CSFII and NHANES.
And I'm not sure how often that's
5 going to happen because the more we look at it, the more we
6 see that that is not totally a good idea.
But we will just
7 see.
8
Okay.
There are limitations, under-reporting as
9 well as over-reporting.
10 been a problem.
This is a problem.
It always has
It always will be a problem.
Partially
11 because people don't remember what they eat, and partially
12 because a number of people might not want to say that they
13 had twelve doughnuts.
But both of the data collection
14 agencies realize that it is a problem.
And when I mentioned
15 this when I met with them, they said, "Yes."
So, we're gonna
16 consider this as a limitation and with no known correction
17 at this point.
18
Different weighting factors for each survey
19 mentioned this.
The first example that the programmers
20 pulled out sort of blew me away when we looked at the eating
AIM REPORTING SERVICE
(773) 549 - 6351
68
1 occasions for raw, smoked and pickled seafood and came out
2 with 79 from CSFII and 87 from NHANES.
And then I looked
3 to see, "Boy, you look at the weights."
And you say, "Boy,
4 1.6 million.
And the 1.5 million.
This is great."
But then
5 when you start to look at the sample descriptives, there's
6 a problem.
7
There's another problem here.
Now, if you look at the -- for CSFII, as a rule,
8 medians are really the best estimate for food consumption
9 data because the data tend to be skewed.
But you look here
10 and you say, "Wow, these are fairly close."
11 you weight, the data comes out fairly close.
And even when
And if I go
12 back to intermediate statistics, I think, "Oh, well, maybe
13 the distributions are normal."
But they're not.
So, then
14 when you look at NHANES and see a median that's higher and
15 a mean that's way off the board, it's one of these where you
16 go back to the raw data and see what was going on.
17
And there was a 19-year-old Hispanic male who ate
18 a ton of raw oysters.
And when I talked to the people at
19 NCHS, they said, "Oh, yeah.
We know that guy."
So, and even
20 if you weight the data, you come out with parameters that
AIM REPORTING SERVICE
(773) 549 - 6351
69
1 are not close.
2
Next slide.
This is not a perfect slide, but I
3 wanted to come up with some idea of what the distributions
4 look like.
So, if you look at the XX's and see the amount
5 eaten grams, and then the number of individuals in thousands
6 -- this is weighted data -- you can see that from NHANES,
7 which is the yellow bars, there were more eaters of small
8 amounts.
Now, I'm not sure -- I guess 28 grams and whatever
9 would be an ounce.
So, it looks like there were a lot of
10 eaters of small amounts.
And then the guy -- or I'm not sure
11 who all, up at the top.
But these are different
12 distributions.
And there's no way you would combine these
13 two weighted distributions, even though you go back to the
14 original sample sizes and see that there are not many eaters
15 in those surveys.
16
Another limitation is individual ingredients from
17 mixed dishes.
And I know this comment was mentioned earlier
18 that -- we can look at beef, or we can look at cheeseburgers
19 and hamburgers.
20 burgers.
And I know there were 43 Food Codes of
But if you want to pull the beef off, which makes
AIM REPORTING SERVICE
(773) 549 - 6351
70
1 sense, it means picking out the Food Codes and going into
2 the amount eaten per person per day in grams, and then looking
3 at a proportion of the overall, which would be the beef.
Varying sample sizes of food groups.
4
Not only is
5 there a challenge when you have a low sample, but if you take
6 consumption of fluid milk and come out with -- who knows how
7 many data points -- it would not even probably fit in a
8 computer.
It isn't gonna work.
And so, we've been talking
9 to the modeler and figure for some of these foods -- and then,
10 of course, if you take and look at the entire group, you're
11 gonna have many, many, many of the foods.
We will probably
12 look at percentiles to figure out what the data are at the
13 first percentile and the second and the third.
So, at least
14 there would be a hundred data points for him to work with.
15
I don't know what difference this will make to overall
16 exposure estimates, but this is one of the challenges.
17
A limitation is merging the data from the earlier
18 module and this module.
And, again, it really will be
19 important to see what Tony and his team is able to find.
20 And then we need to adjust according to them.
AIM REPORTING SERVICE
(773) 549 - 6351
They don't
71
1 adjust according to us.
2
Then we went over this last week one day.
And
3 someone sent me an E-mail that said, "Don't you think it's
4 a limitation to have one or two days of eating?"
5
Absolutely.
But when you work with data all the time and you know it's
6 the best that's supposed to be out there, sometimes the
7 obvious isn't always clear.
8 a limitation.
But, yes, this is definitely
And when these two surveys attempt to
9 integrate, there may be data collection over the telephone
10 instead of in-person.
And there may be one day of eating,
11 or there may be only a subgroup where they can get to the
12 people in person and talk to them.
But both agencies are
13 doing a lot of highlighting to try to find out what works.
14
15
Okay.
Risk assessment always has to include
16 uncertainty, and one source that is very implicit in our
17 module is that we want to have a reasonable proportion of
18 the food consumed that would model the consumption.
19
One example is with fluid milk.
We looked at CDC's
20 behavioral risk factor surveillance systems survey; and their
AIM REPORTING SERVICE
(773) 549 - 6351
72
1 latest data -- I believe that's the latest data -- indicated
2 that 1.4 percent of the respondents said that they drink raw
3 milk.
And so, we're right now working under the assumption
4 that if we look at fluid milk consumption, that 1.4 percent
5 would be unpasteurized.
We're still trying to figure out
6 whether to put most of the risk in unpasteurized or to look
7 at the pasteurized.
Pasteurized milk will be included, but
8 I'm just not sure where that will go.
And also, there was some data from CDC that fall
9
10 into assumption to -- where they said 5 percent of
11 unpasteurized milk contains Listeria.
So, possibly this is
12 going to limit the amount of milk that's really at risk.
Okay.
13
14 projects.
You'll learn a lot when you do these new
And I was very surprised to find that over half
15 the states -- actually, 28 allow intrastate, the intrastate
16 sale of milk.
17
These are the 27 that allow the sale from farms.
Luckily, the BRFSS I just mentioned did include Missouri
18 and New York.
For milk, they did not have a question for
19 South Dakota.
So, some of their states were in that survey.
20
And I know for one of them, the proportion of people who
AIM REPORTING SERVICE
(773) 549 - 6351
73
1 said that they drink raw milk was a little bit higher than
2 the other ones.
3
Two separate columns here.
4 grocery stores to sell raw milk.
5 restaurants.
Eleven states allow
Six states allow
And then the list is getting smaller.
But some
6 states even allow the sale of raw milk in schools and in
7 hospitals.
8
Surprise.
Next.
I was surprised.
You may not be.
Listeria is not on this slide, but we did
9 just find one article that was published last year that linked
10 outbreaks to raw milk.
11 includes Listeria.
12
Now, I'm not sure if "unknown"
But there are problems.
We go back to our burgers again.
And the BRFSS
13 reported that just under 20 percent of the respondents
14 reported that they eat pink hamburger.
Now, I don't know
15 what proportion of the burger would be pink -- probably not
16 the outside -- but at this point, our assumption is that 19.7
17 percent of the ground beef consumed is undercooked and at
18 greater risk.
19
You could spend a couple years on cheese.
20 really fascinating.
It's
But we've tried to look at the type of
AIM REPORTING SERVICE
(773) 549 - 6351
74
1 cheese, the category, the pasteurizations required, the
2 implication in cases and recalls.
3 an overall risk designation.
And then our team is given
I'm not sure where that's gonna
4 go.
5
Type of cheese.
6 four categories.
They appear to fall into these
You just can't say, "soft cheese" because,
7 I mean, there have been cases linked to feta cheese and
8 outbreak -- that one Mexican-style cheese is listed.
9 cheeses, there have been recalls.
Cream
I don't know that there's
10 been that much a problem with cottage cheese.
11 fair just to say soft versus other.
So, it isn't
There have
12 been outbreaks linked to soft, ripened cheeses, both in this
13 country -- I know they've been epidemiologically linked and
14 also in Europe they have been linked.
15
Okay.
Semi-soft cheese.
I believe everything up
16 here has been linked to either some sort of case or to a recall.
17
18
Hard cheeses.
I haven't found much problem.
A
19 lot of the literature that Bob Buchanan mentioned where you
20 inoculate and see what happens, I've seen literature on these
AIM REPORTING SERVICE
(773) 549 - 6351
75
1 cheeses, but not too many problems that I'm qualified to
2 address, anyway.
3
You would think the processed cheeses wouldn't be
4 a problem, but there have been recalls linked to cheese
5 spreads and various types of cheese pack foods.
6
We receive literature from our Land Foods.
That's
7 not the total office title -- but the fellow who's a cheese
8 expert -- and he brought out the literature and said this
9 is how cheese production in the U.S. is broken down, which
10 would at first glance indicate that two-thirds of the cheeses
11 that we produce here are at less risk.
But the other little
12 over a third doesn't always-- it isn't very clear-cut.
13
I was surprised to find out that some cheeses are
14 pasteurized, or else the milk is pasteurized first, but that
15 there are also heat treatments and the temperature is not
16 as high as pasteurization.
So, like the sharp cheddar, there
17 still could be some kind of risk.
We've contacted Dr.
18 Johnson, Wisconsin -- I'm not sure where.
And he is going
19 to have his people look at later data to see if it still falls
20 in this proportion.
AIM REPORTING SERVICE
(773) 549 - 6351
76
Now, the earlier slide was U.S. production.
1
2 that doesn't include what we import.
And
And I don't think that
3 it's fair to say that what we import is at greater risk.
4 But these are some data from the National Cheese Institute
5 where you can see that Camembert and Brie, which is part of
6 the cheeses that have been linked to outbreaks in different
7 parts of the world.
50 percent we import, as well as Gouda
8 and Edam, and in a smaller proportion.
They did not have
9 data for the amount of Hispanic cheese that we import.
So, we have come up with our own little risk
10
11 designations with the lower; and then if there's been a
12 recall, we move the lower to the higher.
And then if the
13 cheese has been associated with an outbreak or spreaded case,
14 then we move them up to the highest risk.
One example would be blue cheese.
15
16 cheese.
17
Pasteurization is optional.
It is not required.
It has been implicated in recall outbreak.
18 outbreak might have been in Denmark.
It's semi-soft
I believe the
I'm not for sure.
And
19 so, blue cheese is one that will be coded at highest risk.
20
Juice.
FDA is currently working on important
AIM REPORTING SERVICE
(773) 549 - 6351
77
1 juice HACCP regulation.
But the one last year, the economic
2 people put together a lot of data sources.
And we're able
3 to estimate that 1.7 percent of the apple and orange juices
4 consumed is unpasteurized.
5 greater risk.
And, therefore, it would be at
That's how the assumption, how we're wording
6 the assumption.
7
I did find one article that -- again, I don't
8 understand the microbiological aspects.
But it did explain
9 that some products at high acid, like orange juice, could
10 maintain Listeria.
11 going to go.
So, I don't know exactly where that's
A slide that I don't have -- because Dick and
12 I just talked about it on the way here--is:
13 with frozen produce?
What do you do
I know that there was one outbreak
14 related to frozen broccoli and cauliflower.
And so, went
15 back to Texas -- ordered the article and called Texas State
16 Health Department.
And they said that there were, indeed,
17 people who -- well, they considered it an outbreak along the
18 Texas and Mexican border.
And they were able to go back to
19 the stores and find Listeria in their frozen product.
20
They didn't have a clue what the people did, whether
AIM REPORTING SERVICE
(773) 549 - 6351
78
1 they ate the frozen product out of the bag or maybe they didn't
2 cook it high enough.
But there was a problem here.
And then
3 I mentioned earlier with Listeria in a well-known brand of
4 frozen blueberries.
And so, I figured that there might be
5 some way to go to -- we have sales data from A. C. Nielsen
6 and from Information Resources -- and figure out how much
7 of the packaged frozen product is sold and then possibly go
8 to commodity groups or produce marketing association, figure
9 out what is sold raw.
And then maybe come up with some way
10 to consider the frozen vegetables.
Because there's no way
11 you would go to food consumption data bases and see that
12 anybody says they've eaten frozen broccoli.
You know, it's
13 either raw or else they've cooked it or put it into a mixed
14 dish.
15
I'd like to leave you with the thought that we are
16 using the best U.S. food consumption data available.
We are
17 considering limitations -- data bases are not perfect -- and
18 attempting to reduce uncertainty as much as we are able within
19 our tight time constraints.
20
Thank you.
MR. MICHAEL JAHNCKE:
Thank you, Dr. Bender.
AIM REPORTING SERVICE
(773) 549 - 6351
We
79
1 have about five minutes for general questions from the
2 subcommittee.
And we will go to a break and then have a full
3 discussion with the entire NAC members with all the
4 presenters.
Are there questions from the subcommittee?
5
Yes,
6 Catherine.
MS. CATHERINE DONNELLY:
7
8 I really enjoyed your presentation.
Cathy Donnelly.
Mary,
I'm wondering if you've
9 given any thought to breaking out of, especially the
10 continuing survey of food intake data, maybe regional
11 differences or socioeconomic trends.
DR. MARY BENDER:
12
13 variables.
14
They do have some of those
So, in addition to the age groups, it's possible.
It's just that when you get down to the cells within the
15 overall survey, it just -- you know, hopefully they'll be
16 a large enough sample to make sense.
17
MS. CATHERINE DONNELLY:
But I'm thinking with
18 certain food consumption trends, there really are regional
19 differences.
20
DR. MARY BENDER:
Right.
AIM REPORTING SERVICE
(773) 549 - 6351
80
1
MS. CATHERINE DONNELLY:
And that might be useful
2 in the data.
3
DR. MARY BENDER:
4
MR. MICHAEL JAHNCKE:
5
MR. ROBERT BUCHANAN:
6
Mary, again, let me echo.
Right.
Thank you.
Bob?
Bob Buchanan, Food and Drug.
A very nice, interesting
7 presentation.
8
I guess one of the questions I have is:
Since
9 Listeria, Listeriosis primarily affects the very young and
10 the very old or people that in some way have suppressed immune
11 systems, your working assumption is the dietary patterns of
12 these individuals are not in any way different from the
13 patterns you're seeing in the rest of the population; or I
14 didn't pick up anything in your presentation.
15
Do you anticipate any kind of a problem in making
16 that working assumption?
17
DR. MARY BENDER:
This is a problem and something
18 that we're definitely still considering.
I know that there
19 was one article -- maybe a couple -- that were in the United
20 Kingdom where they looked at pregnant women and then women
AIM REPORTING SERVICE
(773) 549 - 6351
81
1 as a proxy, women in child-bearing age.
And they did look
2 at their consumption and found out that there was very little
3 difference.
So, I've tried to reach some of the
4 nutritionists in our center to see if they have a comment.
5
But nobody has gotten back to me about that.
6
It's a problem.
I mean, we can come up with
7 aggregate estimates for the population, and I know that that
8 isn't going to be adequate.
But I hope that we can do
9 something within the time frame.
But last year, I spoke a
10 number of times with people from CDC; and they were willing
11 to go out and collect data from pregnant women from their
12 sites.
And also, CDC and ARS were willing to double the
13 sample pregnant women.
But it didn't work out that we had
14 the funds that would go toward this data collection.
And
15 it wouldn't be ready anyway.
16
As far as the immune-compromised people, I know
17 there are some variables on some of the data bases.
Like,
18 we could pick out people who either say they have diabetes
19 or who are being treated for diabetes, possibly some other
20 conditions.
But when I spoke with CDC about this, the
AIM REPORTING SERVICE
(773) 549 - 6351
82
1 National Centers for Health Statistics say that NHANES is
2 the wrong survey.
You're not going to be able to determine
3 who was immunocompromised.
4
And I know people in our Center have spoken with
5 those folks about including data, some other measures in the
6 future.
But I don't know where that has gone.
7 have a problem.
So, we do
We could find the children, and we could
8 find older people who are not in institutions.
9
One thing that I did learn is that over time, a
10 lot of people have moved out of nursing homes into home
11 situations where the -- I guess Medicare will now -- Medicare,
12 Medicaid, I can't remember -- I'm not quite there yet -- it's
13 getting closer -- where they will provide support in the home
14 for the people.
And so, because of that, the likelihood would
15 be that there would be better data on older folks.
16
Something else that I did find in the literature
17 was that a lot of people in nursing homes have a very
18 restricted diet.
And many people don't eat anything compared
19 to what we eat, anyway.
20 problems.
And so, we know that these are
And it's actually a major limitation and should
AIM REPORTING SERVICE
(773) 549 - 6351
83
1 be included.
So I can't really answer your question, but
2 we know it's a problem and we're gonna work on it.
3
MR. ROBERT BUCHANAN:
Is there any data available
4 through programs such as Meals on Wheels in terms of the
5 patterns, consumption patterns?
6
DR. MARY BENDER:
7 on Wheels data.
8
CSFII captures some of the Meals
But I don't know.
MR. MICHAEL JAHNCKE:
Thank you, Dr. Bender, for
9 a very thorough presentation of a very complex issue.
10 you very much.
It is time for the break.
11 promptly at 10:15.
Thank
We will reassemble
At that point we'll have a general
12 committee discussion with all the presenters.
13 questions and there will be plenty of time.
So, hold your
Thank you very
14 much.
15
(Whereupon, a recess was had in this matter.)
16
MR. MICHAEL JAHNCKE:
Welcome back, everybody.
17 We're now at a little bit after the break for the committee
18 discussion.
I'd like to open this up for all the people
19 around this table, the National Advisory Committee people
20 and the presenters, for questions and comments.
AIM REPORTING SERVICE
(773) 549 - 6351
84
Yes, Dane.
1
You had a question right before the
2 break.
MR. DANE BERNARD:
3
Thank you.
4 Yes, I did have a question and a comment.
5
Dane Bernard.
The comment, first.
We've used in presentations so far the word, "risk" rather
6 liberally.
And in my humble opinion, maybe not in the
7 appropriate context.
I know it's tempting to talk about risk
8 in context of probability of contamination when we actually
9 mean the probability of contamination.
So, I would caution
10 as we go forward with the project, with the risk assessment,
11 when we show slides that already categorize things by risk,
12 it would tend to give the impression that a decision has
13 already been made.
And my impression is that's supposed to
14 be one of the outputs of the risk assessment.
And unless
15 I've missed a whole bunch of history, we're not to the output
16 stage yet.
17
So, I would caution that as we go forward, we
18 consider the impact of those kind of statements and how we,
19 in fact, are using the word, "risk."
20
The question:
In neither the first two
AIM REPORTING SERVICE
(773) 549 - 6351
85
1 presentations did I see reference to the impact of food
2 preparation steps on the actual amount of Listeria
3 monocytogenes ingested.
There were several of the foods
4 listed -- for example, hot dogs -- that, while they are sold
5 and legally defined as a ready-to-eat product in the package,
6 they are customarily further prepared before consumption.
7
We all know that there are occasions when that may not happen.
8
But certainly, if one is to calculate a good estimate of
9 exposure, I think you have to consider how the products are
10 going to be prepared before consumption.
So, we had a list
11 of products where we're collecting data on incidents in the
12 marketplace and then how much of that particular product is
13 consumed.
14
But I think in order to get a good estimate on how
15 much Listeria monocytogenes is consumed, you're going to have
16 to consider the impact of further preparation on the actual
17 population.
18
Thank you.
MR. MORRIS POTTER:
I wonder, Dane, if that doesn't
19 go back to something that Bob brought up in terms of modelling
20 the survival and growth of Listeria.
Given the presence,
AIM REPORTING SERVICE
(773) 549 - 6351
86
1 perhaps, at some point in the chain, given the amount that
2 are then given -- how it's prepared, how it's used and how
3 often it's consumed and by whom, then would all come together
4 in characterization.
5
MR. DANE BERNARD:
I think it does.
I think you're
6 exactly right, that if there is a place within the risk
7 assessment for considering that kind of information, that's
8 exactly where it should be considered.
9
Now, as you know, we're conducting our own study
10 that's very similar to some of this.
And I think I would
11 agree a bit with Tony's comment earlier that considering the
12 dynamics of the marketplace, many of these things will sort
13 of null out.
But you do need to think about what the
14 population or what the quantity of L.M. would be at time of
15 consumption and what factors, including either growth or
16 decline in a product, might affect that.
17
MR. MICHAEL JAHNCKE:
Bob?
18
MR. ROBERT BUCHANAN:
Bob Buchanan, FDA.
19 comments were gonna sort of echo some of yours, Dane.
My
What
20 I wasn't sure in the data base -- and Tony or Dick, maybe
AIM REPORTING SERVICE
(773) 549 - 6351
87
1 you can give me a hint -- are you going to be determining
2 or attempting to estimate at what point in a product's shelf
3 life the sample was actually taken or in some way
4 differentiate in your data base whether the sample was taken
5 at the time of manufacture, versus it was sampled in a retail
6 market, versus it was sampled in someone's refrigerator?
7 And, certainly, that could have a very large impact.
I know
8 it's an extremely difficult problem, particular when you're
9 dealing with what appear to be about 10,000 different kinds
10 of foods that you are considering.
11 to do with this?
But any plans on what
Regretfully, much of the data is collected
12 at the point of manufacture and doesn't take into account
13 that whole distribution potential for temperature abuse, the
14 effect of preparation practices, et cetera.
15
MR. MICHAEL JAHNCKE:
Please identify yourself,
16 please.
17
DR. TONY HITCHINS:
Tony Hitchins, FDA.
Well,
18 it's a complex problem, allowing for the differential between
19 analysis time and how the actual food might be treated by
20 a given consumer.
AIM REPORTING SERVICE
(773) 549 - 6351
88
1
I think as Bob was already trying to tell us when
2 he questioned me earlier, one can take into account data from
3 survival studies, impact inoculation studies. One can do
4 that.
I guess the way I would do it is:
I would say what
5 is the frequency of contamination of franks?
6 total of franks consumed?
7 mono consumed?
What is the
Therefore, what is the total of
And then I would apply corrections to that
8 based on some feeling for survival curves.
9 it puts a lot of wobble in the final answer.
I mean, you know,
But that's what
10 risk analysis is about, I think, that one has to say, "This
11 is what we would consume if so-and-so applies.
12 be less if it doesn't apply."
And it will
That kind of thing, I think.
13
I don't know if that helps.
14
DR. RICHARD WHITING:
Richard Whiting, FDA.
15 Following that, would you say your data bases that you're
16 working with, Tony, generally identify where the sample is
17 taken?
So, I mean, you could take a series of a luncheon
18 meat, for example, and you might have a certain data set was
19 at manufacture, and then another data set was taken in the
20 deli.
And that would become part of the way you would work
AIM REPORTING SERVICE
(773) 549 - 6351
89
1 up.
2
DR. TONY HITCHINS:
Thank you, Dick.
I should
3 have said that we can certainly classify our different pieces
4 of data into whether it was taken from someone's fridge or
5 whether it came out of retail or whatever, from the factory
6 or whatever.
We can do that to a large degree.
And that
7 may help us, too.
8
MR. MICHAEL JAHNCKE:
9
MR. BRUCE TOMPKIN:
Bruce?
Bruce Tompkin.
This question
10 about where the products are sampled, I assume that FDA/USDA
11 samples are from point of manufacture.
12 basically are.
Certainly, USDA
I'm less familiar with FDA.
One of the
13 outcomes of the risk assessment eventually will be to address
14 the question of which foods are at higher risk, at least in
15 terms of consumer, from a consumer perspective.
So, whether
16 growth can or cannot occur in a product is important, as is
17 another important aspect in this thing.
18
There is often confusion whether it's with CDC or
19 here in this study as to what foods are.
20 a fermented sausage?
I mean, what is
What does "cured" mean?
AIM REPORTING SERVICE
(773) 549 - 6351
As the
90
1 consumers are polled by telephone and the questions are asked,
2 do they really know what Lebanon bologna is, for example?
3
As you go down through the various categories of foods and
4 their recall, all this has impact on that outcome.
5
As you consider your different food categories,
6 I think you can get help in terms of identifying these foods,
7 whether they be cheeses, which you already have a pretty good
8 fix on, or on the meat and poultry products.
There's a number
9 of us who could help you with a better understanding of what
10 the different classes of meat and poultry products are.
We
11 can give you references; we can sit down and talk with you
12 -- however best that could be done.
I'm sure that a number
13 of us would be very willing to help you get that clarification.
14
And then when it comes to the data, would it be
15 helpful to then at least group the products for which you're
16 painting data into perhaps three groupings -- One, those where
17 growth can occur; those where growth cannot occur; and those
18 where you're uncertain.
19
In terms of growth, products in which growth can
20 occur, some of us have data on that.
There are published
AIM REPORTING SERVICE
(773) 549 - 6351
91
1 data such as what Mike did at Wisconsin.
2 help you with.
So, that we could
And that also would have some impact on your
3 interpretation of the significance of the results.
4
MR. MICHAEL JAHNCKE:
Bob?
5
MR. ROBERT BUCHANAN:
Bob Buchanan, FDA.
6 I like that idea.
7 contacting you.
Bruce,
And certainly, FDA won't be bashful about
I wonder if it would also be good to include
8 in that subdivision of food products what was the type of
9 consumer preparation.
And we can get that into the mix also,
10 because certainly that's going to have a very large impact.
11
MR. MICHAEL JAHNCKE:
12
MS. CATHY DONNELLY:
Yes, Cathy?
Cathy Donnelly.
13 Doyle and Bob Buchanan asked about methodology.
Both Mike
And I think
14 it's really going to be important to take the data on presence,
15 whether it's qualitative or quantitative, to focus in on the
16 methods used to arrive at an estimate of degree of
17 contamination because increasingly, as we look at injured
18 Listeria, both regulatory methods in use now do significantly
19 underestimate Listeria better injured.
20
For instance -- and I'd be happy to furnish some
AIM REPORTING SERVICE
(773) 549 - 6351
92
1 of these data because I think they will be helpful to the
2 risk assessment.
But products like salsa, for instance, if
3 you use a method that considers recovery of injured organisms,
4 you go from 3 out of 30 samples being contaminated to about
5 23 out of 30.
And so, I think that it gets to Bruce's point
6 of data from point of manufacture, using highly-selective
7 methods is really underestimating what's there.
And that's
8 why I think inclusion of data that had been stored under
9 refrigeration conditions, for instance, gets that injury
10 issue backwards kind of way and I think would be very
11 instructive.
12
MR. MICHAEL JAHNCKE:
13
MR. BRUCE TOMPKIN:
Bruce?
Bruce Tompkin.
14 mentioned something about a study of some sort.
15 quite clear what that meant.
16
So, Dane, you
And I'm not
And maybe I misinterpreted it.
But is anyone actually going to undertake a market basket
17 survey to determine what is available at retail?
I note
18 there's some issues associated with that kind of a study.
19
But is this being pursued in any manner?
20 quantitative?
AIM REPORTING SERVICE
(773) 549 - 6351
And will it be
93
1
MR. MICHAEL JAHNCKE:
Richard?
2
DR. RICHARD WHITING:
Richard Whiting.
Well,
3 FSIS has an ongoing survey of meat products, although at this
4 point maybe somebody can clarify it.
5 a presence/absence study.
I think it is basically
I don't think within FDA right
6 now we have any ongoing survey-type for foods with Listeria.
7
Our field office does do samples as part of the regulatory
8 role.
And we have data which we have been collecting from
9 our field offices on the presence of Listeria that they find
10 in certain foods.
But we also realize that that is somewhat
11 biased data and that samples are collected when the inspector
12 often sees a need to take the sample.
And I think, again
13 -- and somebody can correct me if I'm wrong.
I think this
14 is basically presence/absence data that we collect.
15
So, I think there is a great shortage of ongoing
16 data collection right now in this country as to just what
17 the quantitative levels of Listeria are in our foods.
18 Unfortunately, we've done some thinking within the house of
19 what this takes.
And when you have a situation like Listeria
20 where we're often talking about 1, 2 or 3 percent of the
AIM REPORTING SERVICE
(773) 549 - 6351
94
1 samples being positive, and then you say of those 1 percent
2 that's positive, how many do we need to then quantify so we
3 have reasonable idea of what the average and distribution
4 of positive samples are?
Our statistician came back and said
5 we need to take something like 2,000 samples for each
6 particular food in order to come up with good data.
And so, you start talking about 2,000 samples for
7
8 everything.
And then, you know, to be reasonable, now we've
9 got to start lumping food categories together.
10 put in all raw meats and pool that or what?
And do we
And it becomes
11 a very daunting analytical problem to come up with this data.
12
MR. MICHAEL JAHNCKE:
13
MR. BRUCE TOMPKIN:
Yes, Bruce?
Bruce Tompkin.
Perhaps this
14 is where data from the UK and Germany -- I think those two
15 countries in particular would be helpful because they sample
16 at retail.
And I don't know how you're able to -- what your
17 connections are.
18
I'm sure if you can't get it, nobody could.
But those two countries do sample at retail.
And it's
19 primarily by the regional health districts that are doing
20 the sampling.
And it's just a matter of collecting that
AIM REPORTING SERVICE
(773) 549 - 6351
95
1 information.
2 quantitate.
And at least in Germany, I believe, they also
So, that information would be a good source for
3 not only presence/absence but the numbers associated with
4 foods that are available for purchase.
5
MR. MICHAEL JAHNCKE:
6
DR. TONY HITCHINS:
Other questions, comments?
Tony Hitchins, FDA.
7 I'd like to address Bruce's points.
Yeah.
We do have data from
8 the literature from the UK Public Health Laboratory survey
9 and the Yorkshire survey.
And we have data from Germany from
10 the Toyful (phonetic) and Benzulla (phonetic) survey.
I
11 mean, your statement seems to imply, though, that there's
12 a lot more data than that, even, that is current.
13
MR. BRUCE TOMPKIN:
The published information is
14 summaries of that kind of information.
But I believe they're
15 ongoing as part of the responsibility for the regional health
16 authorities.
So, it's just a matter of what's available and
17 ongoing.
18
DR. TONY HITCHINS:
Yeah.
I'll just have to write
19 to Dr. McLaughlin and so on in the UK and try and ask them.
20
MR. MICHAEL JAHNCKE:
Yes?
AIM REPORTING SERVICE
(773) 549 - 6351
96
1
DR. WESLEY LONG:
This is Wes Long with FDA.
2 have a further point to make on that.
I
I think we need to
3 -- I think those are good sources of data, but I think we
4 need to be careful because they may be under a different
5 regulatory construct, and the measures that they have in
6 place, be they regulatory, HACCP, whatever, may result in
7 different levels of those contaminations of those foods in
8 those countries.
So, we have to take that into account when
9 we consider their data.
10
11
12
MR. MICHAEL JAHNCKE:
Other comments, questions?
Yes, Michael.
MR. MICHAEL DOYLE:
Mike Doyle.
Last week at a
13 meeting in Georgetown addressing Listeria, a point was raised
14 about missed opportunities.
And we ought to be thinking
15 about in the future when there are recalls, to see if we can
16 relate those data as to the number of Listeria that are present
17 and pounds of that type of food that was consumed.
And that
18 would fit very nicely into the risk assessments.
19
MR. MICHAEL JAHNCKE:
Yes, Bob?
20
MR. ROBERT BUCHANAN:
I do want to sort of take
AIM REPORTING SERVICE
(773) 549 - 6351
97
1 off my Advisory Committee hat and put on my FDA hat for a
2 second and remind everyone that this information, there's
3 sort of a bright, shiny line drawn in the sand about when
4 data will be available.
And while future work is pertinent
5 in terms of validating whatever the current team is putting
6 together or to be data for future risk assessments, at some
7 point we have to take whatever we have and do the risk
8 assessments.
9
And that date is July 6th.
So, as we talk about future programs, please
10 understand that they're not really directly pertinent to the
11 questions at hand before the working group.
12
MR. MICHAEL JAHNCKE:
13 comments, questions?
14
Thank you, Bob.
Other
Yes, Tony?
DR. TONY HITCHINS:
15 that -- in keeping my thoughts.
I agree with Bob, of course,
But, no, seriously, you know,
16 we do have to go with what we've got.
And, really, we can
17 go a long way with presence and absence data.
That can be
18 converted into means and distributions if one makes certain
19 assumptions.
So that for the time being, we can get by
20 without further collection of data that is more enumerated
AIM REPORTING SERVICE
(773) 549 - 6351
98
1 directly.
2
MR. MICHAEL JAHNCKE:
3
DR. WESLEY LONG:
Yes, Wes?
Wes Long, FDA.
I want to go back
4 to something that Dane Bernard raised earlier before we opened
5 things up for additional comment.
I would hate for the sound
6 bite from this morning to be that certain soft cheeses are
7 at highest risk.
And I just want to clarify that what Dr.
8 Bender was referring to was this probability of contamination
9 and that it was important for her to categorize these
10 different cheeses differently because when she has to match
11 that up with Dr. Hitchins' data that's not as specific, we've
12 got to figure out where do we put his data, into which
13 categories do we put his data.
14
So, she was just referring to a probability of
15 contamination and not referring to the high-risk,
16 medium-risk, low-risk cheeses.
17 final output of this process.
18 now.
Certainly, that may be a
But we are not at that stage
We're not ready to make any statement to that effect.
19
MR. MICHAEL JAHNCKE:
20
MR. DANE BERNARD:
Dane?
Thank you.
AIM REPORTING SERVICE
(773) 549 - 6351
Dane Bernard.
99
1 Thanks, Wes, for that clarification.
Before I forget it --
2 because at my age, I do forget things -- Tony, I went through
3 your references on the seafood list.
And there were some
4 additional references that both John Glenburg (phonetic) and
5 I were made available to us at the NFAO consultation last
6 week.
And I think you might find quite interesting some very
7 recent studies from the Nordic countries, some populations
8 of L.M. in seafood products.
So, before I forget to mention
9 that, we'll get that to you.
10
DR. TONY HITCHINS:
11
MR. MICHAEL JAHNCKE:
Thank you, Dane.
Michael Jahncke.
Along the
12 same lines, I know that Mel Eklund has additional information
13 also that if he does not remember to come up to you, please
14 keep that in mind.
15
DR. TONY HITCHINS:
He alerted me to that.
Thank
16 you very much.
17
MR. MICHAEL JAHNCKE:
18
DR. RICHARD WHITING:
Yes, Richard?
Richard Whiting.
Yeah.
On
19 that line, the purpose of the document that we've given out
20 today is exactly for that reason.
You will notice about half
AIM REPORTING SERVICE
(773) 549 - 6351
100
1 of it is just lists of references.
2 straightforward, dry reading.
It is rather
But the purpose of it is to
3 put it out there and show people what we are looking at.
4 And if you are knowledgeable in an area, skim through those
5 references.
And if you see something there that we have not
6 listed, bring that to our attention.
That's one of the
7 purposes of this document.
MR. MICHAEL JAHNCKE:
8
9 from the committee members?
Other comments and questions
Yes, Bruce?
MR. BRUCE TOMPKIN:
10
Are we allowed to talk about
11 the documents at this point, too?
MR. MICHAEL JAHNCKE:
12
If you can keep it focused
13 on the presentations this morning, it will tie in nicely.
14
Because there will be a chance this afternoon also to go
15 over the -- as all the presenters will be addressing this,
16 too.
17
MR. BRUCE TOMPKIN:
What I would discuss would be
18 off, not what we've just heard.
19
MR. MICHAEL JAHNCKE:
Something else.
Yes, any other questions and
20 comments from the group?
AIM REPORTING SERVICE
(773) 549 - 6351
101
1
Yes, Bob?
2
MR. ROBERT BUCHANAN:
Yeah, I would like to make
3 a point and get some additional clarification from Tony on
4 one comment he made earlier this morning.
5
The traditional taxonomy of Listeria monocytogenes
6 really divides Listeria monocytogenes/innocua into
7 pathogenic and nonpathogenic isolates based in hemolysin
8 production.
9
Tony, you indicated that there are monocytogenes
10 species that are not virulent.
11 that designation?
On what evidence did you make
As far as I know, there's nothing in the
12 literature that identifies other than
13 genetically-manipulated strains or strains that have in some
14 way lost a virulence characteristic due to a deletion
15 mutation, any monocytogenes that is truly a monocytogenes
16 that has not been considered pathogenic in an appropriate
17 animal model.
18
19 Bob.
DR. TONY HITCHINS:
Tony Hitchins, FDA.
Yeah,
I only refer there, really, to hemolysin negative
20 strains that do crop up occasionally when one is isolating
AIM REPORTING SERVICE
(773) 549 - 6351
102
1 monocytogenes from foods -- very rarely, in fact, that kind
2 of strain.
And by inference from the deletion-type studies,
3 one assumes that their virulence is less than the normal
4 isolates.
I didn't say there -- Well, if I implied they're
5 totally non-virulent, only in the sense that probably a
6 greater dose of them would be necessary to produce some kind
7 of symptoms.
8
MR. ROBERT BUCHANAN:
It might be helpful to the
9 Committee members to refresh our memories on what is the
10 distinction between innocua and monocytogenes.
11
DR. TONY HITCHINS:
12 difficult, really.
13
Well, it's very -- it's quite
I mean, it's not a hundred-percent clear.
But basically, you know, the taxonomists would say
14 monocytogenes are this set of properties.
And it's
15 hemolytic, basically.
16
And if you've got a non-hemolytic strain, you would
17 be in trouble in terms of normal taxonomic methods.
But by
18 other methods, you would say it's a monocytogenic.
19
MR. ROBERT BUCHANAN:
Yeah. I guess that was my
20 point, is that on anything except very fine genetic analysis,
AIM REPORTING SERVICE
(773) 549 - 6351
103
1 innocua and monocytogenes are identical except for one
2 virulence-associated determinant.
And so, it's almost by
3 the classical taxonomy; all the pathogens wind up in
4 monocytogenes, and all the non-pathogens wind up in innocua.
5
6
DR. TONY HITCHINS:
We really don't know that all
7 monocytogenes strains are virulent, quite frankly, do we?
8
We just don't know that.
9 monocytogenes strains.
We can isolate a lot of
But unless we give them some test
10 -- I mean, we can argue about what the test should be.
11 don't know they're all virulent, really.
12
MR. ROBERT BUCHANAN:
We
Do we?
That was my question.
Is
13 there any evidence at all that when we've tested a
14 monocytogenes, regardless of its serotype, as long as it has
15 all of the appropriate virulence markers, it is pathogenic?
16
DR. TONY HITCHINS:
Yes, but I think we'll have
17 to wait until this afternoon until Dr. Raybourne discusses
18 the virulence factors.
19
MR. ROBERT BUCHANAN:
20
DR. TONY HITCHINS:
Okay.
I don't think they've really
AIM REPORTING SERVICE
(773) 549 - 6351
104
1 been thoroughly defined.
2 that kind of thing.
3 know about.
I mean, we know the hemolysis and
But there may be other factors we don't
We just don't know that if a hemolytic
4 monocytogenes is isolated from a food and it doesn't
5 correspond to any strain that had been isolated from a case
6 of Listeriosis, exactly, exactly correspond.
We just don't
7 know it's virulent unless we then do some tests.
Again, we
8 might not agree on what those tests should be, apart from
9 human trials or something that comes close to human trials
10 like primate trials.
11
MR. MICHAEL JAHNCKE:
Cathy, yes?
12
MS. CATHERINE DONNELLY:
Cathy Donnelly.
Will
13 there be any attempts in building the risk assessment model
14 to be proactive and contact some of the companies involved
15 in rapid methods, whether they be typing or -- any type of
16 DNA or life-based technology?
Because to validate these
17 methods, there's been a large amount of data collection -18 Bob is sitting over there smiling -- but with the proviso
19 that the purpose of the data isn't to engage in regulatory
20 enforcement.
I think those data bases will reveal a lot of
AIM REPORTING SERVICE
(773) 549 - 6351
105
1 interesting information for this analysis.
2
MR. MICHAEL JAHNCKE:
Wes?
3
DR. RICHARD WHITING:
Richard Whiting.
4 was just going to say:
5 afternoon?
Richard?
Well, I
Why don't we leave that one for this
And we'll put our two speakers who will get into
6 more of the hazards and so on of the organism and let them
7 deal with that.
8
MR. MICHAEL JAHNCKE:
Other questions and comments
9 from the group?
10
If not, I'll pass this over to Dr. Potter.
11
DR. MORRIS POTTER:
At this point on the schedule,
12 there's time for public comment.
Since this is a public
13 meeting in addition to being a meeting of the National
14 Advisory Committee, we would like to give non-committee
15 participants in today's proceedings an opportunity to talk.
16
For those people in the audience who would like
17 to speak, it would perhaps be most appropriate this morning
18 to talk about those aspects of Listeriosis that relate to
19 the presence of Listeria in foods and human consumption.
20 But if there are folks here who would like to make comments
AIM REPORTING SERVICE
(773) 549 - 6351
106
1 who will not be able to stay for this afternoon and the
2 comments are off-point, please feel free.
We understand that no one has signed up outside
3
4 to make a formal presentation.
But if there are comments
5 based on what you've heard this morning or other comments,
6 please step up to the mike and identify yourself.
I know some of you aren't this polite.
7
8
All right.
Good.
9
MR. WALLY SCHLECH:
Just to get the ball rolling,
10 Wally Schlech from Delhausen (phonetic) University.
11 a long interest in Listeriosis.
I have
I listened this morning with
12 great interest in some of the regulatory aspects of what's
13 being attempted to do.
I think July 6th or whatever it is
14 is a pretty short time line considering the lack of data that
15 you have.
16
What I've seen expressed today is a lot of large
17 but really anecdotal collections of data from around the
18 country that is being pooled to determine what types of food
19 products may be risky.
And I think what I would encourage
20 the group -- and this is obviously something you can't do
AIM REPORTING SERVICE
(773) 549 - 6351
107
1 before July 6th -- but that in terms of -- I think a risk
2 assessment is a project in progress.
In other words, even
3 if you produce something July 6th, you'll still have to
4 continue to refine it.
The idea of doing some sort
5 of retail market sampling similar to some in the UK, I think,
6 is critical.
The numbers are large.
But because of the
7 variability and how the consumer, who basically we're trying
8 to protect here, handles food, I think that at the retail
9 level is the time to do some sampling.
And the sampling has
10 to be done in such a way that there can be cross-comparisons
11 of various food products.
I could be a little controversial
12 and say if products are meant to be cooked before eating,
13 don't bother sampling that group of products.
That leaves
14 out things like hot dogs, which obviously would be politically
15 incorrect to leave out of any sampling procedure.
16
But, theoretically, if the public's not gonna take
17 care of itself by cooking these things properly, I'm not sure
18 that we should spend a lot of money looking for Listeria in
19 those products.
20
I'm more concerned about the deli meats and the
AIM REPORTING SERVICE
(773) 549 - 6351
108
1 others, salads, that are in fact meant to be consumed as
2 they've come out of the plant in appropriate packaging.
And
3 there, I think, we do have a role in protecting them from
4 that.
5
I'm sure there will be more this afternoon about
6 the issue of virulence.
My own bias would be, for example,
7 that this E-strain phage-type that was present in this most
8 recent problem is intrinsically different in some way than
9 the sort of standard, run-of-the-mill serotype 4b.
10 question -- We just don't know.
And the
And maybe it will come up
11 this afternoon in discussions of virulence.
But I think that
12 that's something that needs to be critically looked at.
And
13 only science can answer those kinds of questions.
14
But, hopefully, it would inform the regulatory
15 stance once that kind of data is available.
I don't think
16 -- it sounds like you're not going to go there for this
17 meeting.
Obviously, you're not planning to with the decision
18 about zero tolerance.
And I don't think there's anything
19 in the virulence area or in the identification of the
20 organisms as monocytogenes that allows anyone to change the
AIM REPORTING SERVICE
(773) 549 - 6351
109
1 stance based on that some may be less virulent than others.
2
So, with those comments, I appreciate the
3
4 opportunity.
MR. MICHAEL JAHNCKE:
5
6
Thanks.
Other comments?
Dane, did you have something?
MR. DANE BERNARD:
7
8 NFPA.
9 there.
Thank you very much, Wally.
Thank you.
Dane Bernard from
Just a follow-up to something that Wally had said
He mentioned sample products intended to be cooked.
10
I don't think you're talking about a sampling program here.
11
You're talking about what data do you consider and how do
12 you consider it.
13
He also mentioned that it's probably not
14 politically correct to do so.
15 to weigh that.
The Agency is going to have
But if the purpose of the risk assessment
16 is not to look at specific products and do a risk assessment
17 on products, it may be appropriate to follow Dr. Schlech's
18 advice here.
Look at where your data is good, look at how
19 you can utilize that data to make an easier projection, a
20 more accurate projection of what is actually consumed.
AIM REPORTING SERVICE
(773) 549 - 6351
And
110
1 maybe you don't use products in certain categories.
And
2 maybe it is that category where, for example, with hot dogs
3 it would be very difficult to factor in what is actually
4 consumed -- the basis, the further preparation of those
5 products.
I mean, it's a challenge.
I don't think we need
6 to worry too much about the politics of whether to include
7 it in the data base or not.
What you need is a data base
8 that you can work with simply and minimize your uncertainty
9 predictions but still have a solid prediction of what is being
10 consumed.
So, I think it's not a comment that should be taken
11 lightly.
I think it should be given due consideration.
12
MR. MORRIS POTTER:
13
MR. ROBERT BUCHANAN:
Bob?
Bob Buchanan, FDA.
Dane,
14 I guess I have to disagree in terms of passing on advice to
15 this group.
If the primary purpose of this risk assessment
16 is to evaluate the public health impact of Listeria, foodborne
17 Listeriosis, and you have documented outbreaks associated
18 with this class of products, then how are you going to get
19 an estimate of the risk face by the consumer regardless of
20 contributing factors without considering all products and
AIM REPORTING SERVICE
(773) 549 - 6351
111
1 including consideration of the likelihood that a product that
2 will be abused, mishandled, inappropriately handled, or
3 handled absolutely correctly and still be associated with
4 outbreaks?
5
I mean, you know, the example that was provided
6 is one that I can confirm, based on the CDC data of the most
7 recent outbreak and the reports I've heard of it, is that
8 all those hot dogs that were consumed were cooked.
So, we
9 may have a representative from CDC that may want to follow
10 that up.
But I would be very cautious about eliminating
11 products when you're attempting to get a risk assessment
12 that's looking at the overall impact of an organism on public
13 health.
14
MR. MORRIS POTTER:
Thank you, Bob.
Remember that
15 today we're looking at the prevalence and extent of exposure,
16 and then the public health impact about exposure.
And for
17 the risk assessment team, a principal take-home from today's
18 meeting is advice on the model and help with their data
19 collection.
20
Some of the comments that have been made this
AIM REPORTING SERVICE
(773) 549 - 6351
112
1 morning would imply that perhaps some of the classification
2 of foods into categories needs some help.
Certainly, that
3 we need some help and data on presence, absence and numbers
4 of Listeria in those categories of foods and perhaps
5 information on post-purchasing handling of foods where those
6 data exist.
For those who may not be ready for oral comments
7 today during the meeting, remember that there is an
8 opportunity to make written comments and to share information
9 with the risk assessment team after this meeting up to the
10 drop-dead date that Bob gave us.
11
Another comment?
Could you identify yourself,
12 please?
13
14 Services.
MS. PETRA BOYSEN:
Petra Boysen from Fresh Check
I have a question concerning the data collection
15 for consumption data.
And I was wondering:
In response to
16 the question of regional information, has any of the sales
17 of certain products been taken into account, assuming that
18 the sales, that these products that are sold are being
19 consumed?
20
DR. MARY BENDER:
Mary Bender, FDA.
AIM REPORTING SERVICE
(773) 549 - 6351
No.
Only
113
1 probably looking at some market share.
But even though I
2 am not a nutritionist, I work with a lot of nutritionists
3 who bristle at the idea of looking at sales data or production
4 data as consumption because you don't know who eats what.
5
And you might have a -- you know, it's very important data.
6
I mean, it's critical.
But as far as consumption, the
7 philosophy at FDA is to stick with consumption data if you
8 have it.
That's another can of worms there.
9
MR. MORRIS POTTER:
MR. PAUL HALL:
10
11 Foods.
Thank you.
Paul?
Good morning.
Paul Hall, Kraft
First of all, I want to compliment this morning's
12 speakers for their presentations and treatment of this
13 difficult subject, to say the least.
14 comments.
15 point.
A couple questions and
First of all, I want to reiterate Bruce Tompkin's
This issue of probability of contamination that we're
16 talking about when we're doing a risk assessment.
And I just
17 want to reiterate the point that Bruce made that I think is
18 extremely important to have some measure of the ability of
19 these products to support growth to high levels of Listeria.
20
AIM REPORTING SERVICE
(773) 549 - 6351
114
1
I know Dr. Bender talked about the cheese category
2 and how difficult it is in classification of cheeses.
And
3 that's a category, of course, near and dear to our heart.
4
And cheese is not cheese is not cheese.
And we all know
5 that you have soft cheeses where we had large outbreaks linked
6 to Listeriosis.
And then you have, say, processed cheese
7 category in which that product, some of those products are
8 hot-packed at a temperature that is lethal to Listeria and
9 there's no opportunity for recontamination, versus a
10 cold-pack type of processed cheese in which it receives no
11 thermal treatment and there is opportunity for
12 post-processing contamination.
But most cold-pack cheeses
13 won't support the growth of Listeria to high levels.
And
14 even though there may have been recalls of those products,
15 it really speaks to whether, indeed, those products represent
16 a present, imminent danger to public health.
17
So, I think it is very critical that we have some
18 kind of measure on whether these products can support growth.
19
And I think Bruce's point is well-taken.
I think there are
20 industry folks that can help out in that assessment.
AIM REPORTING SERVICE
(773) 549 - 6351
115
And Bob Buchanan talked about the role of challenge
1
2 studies, inoculated pack studies.
And there's plenty of
3 those as well that can be factored in on that particular point.
So, to me, it goes just beyond probability of
4
5 contamination, you know, whether it's contaminated there or
6 not.
But it's also level of contamination at the time of
7 consumption, I think.
I know that's a difficult thing to
8 model.
9
The other thing, the other piece I guess I would
10 ask about is the quantitative data.
I know the UK data that
11 Dr. Hitchins presented had as a upper limit greater than a
12 thousand per gram, I believe.
And given the scientific
13 nature of the risk assessment that we're trying to do, given
14 the work at the University of Georgia and the Emory Primate
15 Center on the L.M. Monkey Study, as I call it, trying to get
16 at infective dose, my question is that we try and measure
17 up the levels of Listeria in these products that we're
18 consuming versus anything that would come out at infective
19 dose study, is greater than a thousand per gram or a thousand
20 per gram sufficient enough?
Or should we be trying to go
AIM REPORTING SERVICE
(773) 549 - 6351
116
1 higher in terms of quantitating levels?
And then, I guess my other question is the issue
2
3 of, really, how do we harmonize ready-to-eat foods,
4 definitions of ready-to-eat foods in this whole process
5 versus, say, frozen foods, for example?
6 that in as well?
Thank you.
MR. MORRIS POTTER:
7
8 from -- Okay.
Thanks, Paul.
Other comments
Tony?
DR. TONY HITCHINS:
9
And how do we factor
10 comment on Paul's comments.
Tony Hitchins, FDA.
Just a
There are data in the collection
11 already that have, you know, numbers greater than a thousand
12 per gram.
It's just that in that particular study or that
13 piece of that study, it wasn't apparent; it wasn't done.
MR. MORRIS POTTER:
14
You got a little far from the
15 mike there, Tony.
DR. TONY HITCHINS:
16
Sorry.
There is data in the
17 data base, at least from some other studies, that gives
18 numbers for rare cases where the counts are greater than a
19 thousand per gram.
20 do.
And not all studies have that, but some
Yeah.
AIM REPORTING SERVICE
(773) 549 - 6351
117
1
MR. MORRIS POTTER:
Wally?
2
MR. WALLY SCHLECH:
Wally Schlech again.
3 wanted to comment about the quantitation.
I just
I think that you
4 also, particularly if you're looking at levels of
5 quantitation, need to look again at the host.
There is clear
6 data in the Boston outbreak in the late 70's that antacids
7 were a risk factor.
So, you may decide to, say, allow ten
8 to the two Listeria per gram to get out into the market.
9 But that may not be sufficient to protect one of these
10 immunocompromised individuals.
And if you look at all the
11 Pepcid AC ads on TV, it seems like the entire American
12 population is swallowing them.
Then maybe that would argue
13 against -- and presumably the monkey studies might give some
14 additional information.
But we certainly have studies in
15 a gastric model in rats that is a real phenomenon.
16
MR. MORRIS POTTER:
Thanks, Wally.
The BRFSS
17 surveys have looked at antacid and H2 blocker consumption,
18 at least in some of them, so there are some data there.
But
19 that may be a bit difficult to model.
20
MS. CARY FRYE:
Cary Frye, International Dairy
AIM REPORTING SERVICE
(773) 549 - 6351
118
1 Foods Association.
Also, the National Cheese Institute.
2 And we really appreciate the comments here today, and we're
3 very supportive of the risk assessment.
I did speak to Mary
4 Bender about some of the data she presented with the food
5 consumption, specifically in the cheese category.
And the
6 slide that you showed about mandatory pasteurization of 33
7 percent is certainly accurate.
8
I don't disagree with that.
However, I think commercial practices of cheese
9 manufacturing, specifically cheese manufacturing that could
10 have a higher probability of contamination, are showing that
11 pasteurized milk is used.
I know commercially,
12 Mexican-style cheese by one of our members, all of their milk
13 is pasteurized.
So, it appears there could be a data gap
14 here that might need additional information that we could
15 assist with, rather than just looking at the regulations,
16 but maybe providing actual practices for cheese manufacture.
17
So, I realize that, and we hope that we can provide that
18 because many soft cheeses are made with pasteurized milk for
19 that very reason.
20
Secondly, I had a question related to the risk
AIM REPORTING SERVICE
(773) 549 - 6351
119
1 assessment similar to this same line of thinking.
If you
2 look at the literature, you're looking at it worldwide,
3 cheeses that may show levels of Listeria that were made from
4 raw milk because there's different regulations in different
5 countries.
And how will you account for that in the risk
6 assessment?
Will there be any accounting for the different
7 practices of how cheeses are produced?
Because it's my
8 understanding the risk assessment will be looking at the risk
9 of the U.S. population.
Will you look at the imported cheeses
10 such as the data we have at NCI and weight that, or will you
11 look at all cheeses?
12
Thank you.
DR. MARY BENDER:
Mary Bender, FDA.
There's
13 somebody back there right now who's trying to get data, as
14 you're discussing.
Our Regulatory Affairs Office at FDA does
15 collect some data on imports.
And they're really excited
16 that they have a data base going, but they've warned us not
17 to take everything as is because this is a developing data
18 base.
But we have been able to look at some of the imports
19 of the lots of cheeses.
And a certain proportion that's been
20 tested or held back for Listeria, and then some where there
AIM REPORTING SERVICE
(773) 549 - 6351
120
1 have been positive results.
But it's been a challenge to
2 try to put this all together to come out with something that
3 makes sense and is accurate.
There was one slide that I had
4 that -- we do want to look at this further to try to figure
5 out.
And I really do appreciate any help.
6
Now, Cary and two others did come to me at the break
7 and said that there really has not been an outbreak related
8 to ice cream.
And I looked back at my file, and there was
9 an epidemiological link -- I don't know -- it was from a CDC
10 article.
11 I've read.
And you all are the experts.
This is something
So, I really appreciate the input.
12
MR. MORRIS POTTER:
13
MR. LARRY BORCHERT:
14 American Meat Institute.
Other comments?
Larry Borchert with the
And it really is following
16 up on points that have already been made.
18 as a cheese is.
Yes.
My comments also deal with data
15 acquisition and consideration.
17 as an example.
Thanks.
And I'll use that
A hot dog is not a hot dog is not a hot dog
If we are considering international data,
19 for example, the hot dogs that are made in Germany, for
20 example, have probably twice the brine concentration, traces
AIM REPORTING SERVICE
(773) 549 - 6351
121
1 of salt and water concentration, that they do in this country.
2
So, it warrants us to be very careful of the use of
3 international data.
4
Likewise, acquisition of data, I think we do need
5 to be cognizant of sales data.
For example, two major
6 companies in the United States produce 40 percent of the hot
7 dogs in the United States.
So, looking at broad-based
8 consumption data might distort the overall picture,
9 particularly if one or both of these companies are using some
10 intervention technique that might decrease the prevalence
11 of Listeria in their products.
12
So, I think the point I'm trying to make is that
13 we must be very, very careful in acquiring the data and using
14 the data that we are applying that to the specific products
15 that we're talking about, not just a generic family of those
16 particular products.
17
18 comments?
Thank you.
MR. MORRIS POTTER:
Thanks, Larry.
Other
Seeing none, the schedule calls for us to be back
19 in session at 1:00.
Since we're a little ahead of schedule,
20 I hope folks will be prompt.
We will start again at 1:00.
AIM REPORTING SERVICE
(773) 549 - 6351
122
1
Be here.
2
(Whereupon, a lunch recess was had in this matter.)
3
MR. MICHAEL JAHNCKE:
Welcome back, everybody. I
4 hope everyone had a nice lunch.
We're going to get started.
5
We have two more presentations this afternoon-- three more,
6 with the summary.
I'm just waiting for a slide.
Here we
7 go.
8
As I mentioned, we're going to have two more
9 presentations.
Then Dr. Whiting later will do a summary of
10 what has been presented to this day.
11 of Hazard Assessment.
We're in the session
And the two presenters will be Dr.
12 Pat McCarthy looking at some epidemiologic records.
And the
13 second speaker will be Dr. Richard Raybourne on dose-response
14 experimentation.
15
16 McCarthy.
Let me introduce our first speaker, Dr. Pat
And he will be speaking on epidemiology of
17 Listeria monocytogenes outbreaks.
18
DR. PATRICK McCARTHY:
Good afternoon.
I'm going
19 to talk about the epidemiology of Listeriosis.
20
Next slide, please.
Listeria was first described
AIM REPORTING SERVICE
(773) 549 - 6351
123
1 in 1926.
And a few years later, the organism was recognized
2 as a human pathogen.
The suggestion that Listeria,
3 Listeriosis could be transmitted to humans in food dates back
4 to the 1930's.
But it was not until the 1980's that evidence
5 was obtained that Listeriosis is a foodborne disease.
6
Since the 1980's, foodborne outbreaks in sporadic
7 cases have been reported in many countries throughout the
8 world.
And in 1986, the Council of State and Territorial
9 Epidemiologists recommended that Listeriosis be a reportable
10 disease.
11
Next slide.
Listeria is the name of a group of
12 disorders caused by Listeria.
Listeriosis is the name of
13 a group of disorders caused by the organism, Listeria
14 monocytogenes.
Listeriosis is clinically defined when
15 Listeria is isolated from blood cultures, spinal fluid or
16 an otherwise normally-sterile site like a placenta or a fetus.
17
Cases of Listeriosis are usually divided into
18 perinatal and nonperinatal groups.
The perinatal group
19 includes pregnant women and their fetus or newborn.
Women
20 may get Listeriosis at any time during pregnancy, but most
AIM REPORTING SERVICE
(773) 549 - 6351
124
1 cases are reported in the third trimester.
Often, pregnant women will present with an
2
3 influenza-like illness which includes fever, chills and
4 headache.
This prodromal illness occurs in about two-thirds
5 of women with pregnancy-associated Listeriosis.
About three
6 to seven days after the onset of prodromal symptoms, women
7 will abort the fetus or will have premature labor.
In the first trimester, Listeriosis results in
8
9 spontaneous abortions.
In later stages of pregnancy, the
10 result can be a stillbirth or a critically-ill newborn.
11 Sepsis occurs in about 30 percent of pregnant women with
12 Listeriosis, and there are a few reports of meningitis in
13 pregnant women.
14
The fetus can suffer abortion, stillbirth.
And the newborn can present with sepsis, meningitis or can
15 die.
16
The nonperinatal group includes all non-pregnant
17 persons over the age of 28 days.
Nonperinatal cases
18 primarily include persons that are taking immunosuppressive
19 medications, persons with chronic debilitating diseases like
20 cancer, diabetes or alcoholism, and persons over the age of
AIM REPORTING SERVICE
(773) 549 - 6351
125
1 60.
Healthy children and adults have a relatively low risk
2 of infection from Listeria.
When infection does occur in children and adults,
3
4 Listeriosis is usually superimposed upon some other illness.
5
Nonperinatal cases often present with meningitis or sepsis.
6
7
In the next few minutes, I'll discuss the early
8 foodborne outbreaks and surveillance for Listeriosis; and
9 I'll provide some examples of recent outbreaks and sporadic
10 reports.
11
Listeriosis is known to cause severe illness, but
12 there have been events in which the majority of cases
13 developed mild symptoms.
I'll identify a few events where
14 mild symptoms were primarily reported.
15
I have a slide on the incubation period for
16 Listeriosis and another slide on fecal carriage studies.
17 I'll show you the incident trend for Listeriosis in the United
18 States between 1989 and 1993.
And I have some recent data
19 from FoodNet, the ongoing active surveillance program for
20 foodborne diseases.
AIM REPORTING SERVICE
(773) 549 - 6351
126
1
Next slide.
The earliest evidence that
2 Listeriosis is a foodborne illness was obtained from
3 outbreaks that occurred in Nova Scotia, Massachusetts, Los
4 Angeles, and Switzerland between 1981 and 1987.
Other
5 outbreaks occurred before 1981, but the vehicle of infection
6 was not identified.
These outbreaks during the 80's lasted
7 for several months each but involved relatively few cases.
8
On the other hand, there were several deaths associated with
9 these outbreaks.
10
Next slide.
Both nonperinatal and perinatal cases
11 were identified in each outbreak.
The age range for the
12 nonperinatal cases was between age 21 and 100.
The median
13 age in the nonperinatal cases was about 60 years.
In these
14 outbreaks, the majority of the nonperinatal cases were taking
15 immunosuppressive medications, had a debilitating disease
16 or were over age 60.
About one-third of the nonperinatal
17 cases died.
18
In the perinatal group, the mother and fetus or
19 newborn was considered as a single case.
The fatality rate
20 in the perinatal group was about one-third.
AIM REPORTING SERVICE
(773) 549 - 6351
127
Matched case-control studies implicated a
1
2 particular food in each outbreak.
3 was implicated.
4 other outbreaks.
In Nova Scotia, coleslaw
And dairy products were implicated in the
The odds ratios that implicated the food
5 were all significant at the 0.05 level or below the 0.05 level.
6
Listeria monocytogenes 4b was isolated from cases in each
7 of the outbreaks and from the implicated food in all outbreaks
8 except from Massachusetts.
9
The incident rates that I show here are for the
10 populations in which the outbreaks occurred.
I don't have
11 the background incident rates for all these outbreaks.
But
12 in Switzerland in the years preceding the outbreak, the
13 background rate was approximately .5 per hundred thousand
14 cases.
At the end of the outbreak, the incident rate was
15 about 5 cases per 100,000.
Low-incident rates make the
16 outbreaks very difficult to detect.
These outbreaks were
17 only detected because all the cases occurred in a single
18 hospital or were reported to a single laboratory.
For
19 example, in Los Angeles, a hospital infectious control nurse
20 noticed the increase in cases; and her observation led to
AIM REPORTING SERVICE
(773) 549 - 6351
128
1 the investigation which implicated the Mexican-style cheese.
2
The likely source of Listeria in the Nova Scotia
3 outbreak was the raw manure used to fertilize the cabbage
4 which was made into coleslaw.
The sources for all these
5 outbreaks suggest that Listeriosis was linked to the farm
6 or to food production facilities.
7
These early outbreaks showed that Listeriosis, the
8 foodborne Listeriosis can cause abortion, stillbirth,
9 sepsis, meningitis and death.
Matched case-control
10 investigations showed that significantly more cases than
11 controls ate the implicated food.
The L. monocytogenes 4b
12 was identified in most of the infections occurring during
13 the epidemic period.
And the epidemic strain of Listeria
14 monocytogenes was isolated from opened and unopened samples
15 of food implicated in 3 of the 4 outbreaks.
16
Following the Los Angeles outbreak in 1985, CDC
17 started Listeria surveillance.
I show here data from two
18 surveillance populations, but there were other reports in
19 the literature of surveillance that took place between 1985
20 and 1993.
There were 34 million people in the 1986
AIM REPORTING SERVICE
(773) 549 - 6351
129
1 surveillance.
And between 1989 and 1993, in that
2 surveillance, there was 19 million people.
Both
3 surveillance periods included people from Oklahoma,
4 Tennessee and Los Angeles County.
The 1986 surveillance
5 population was larger because health departments in Missouri,
6 New Jersey and Washington were included.
7
Before the surveillance was started, hospitals,
8 laboratories and physicians in the surveillance area were
9 contacted and asked to report cases of Listeriosis.
At the
10 end of the surveillance period, facilities that reported
11 cases were audited to determine the sensitivity of the
12 surveillance.
The case ascertainment for the 1986
13 surveillance was 93 percent, and case ascertainment in 1993
14 was shown to be 97 percent.
15
246 cases were reported in 1986.
16 and 1993, about 400 cases were reported.
And between 1989
Now, I'm going to
17 show additional data from the 1986 surveillance.
And in a
18 few minutes, I'm going to show the incident trend that was
19 developed for Listeriosis between 1989 and 1993.
20
Overall, in 1986 there were .7 culture positive
AIM REPORTING SERVICE
(773) 549 - 6351
130
1 cases of Listeriosis per 100,000 population.
The rate was
2 slightly less in the nonperinatal group but was much higher,
3 7.8, in the perinatal group.
If Los Angeles County was included, the cases per
4
5 100,000 would be approximately 24.
But Los Angeles County
6 experienced an outbreak during 1985, and this heightened
7 awareness could have been the reason for the increase in
8 cases.
So, I have excluded it in what I'm reporting to you.
Listeria monocytogenes has 13 serobars.
9
But 3
10 serotypes accounted for approximately 96 percent of the
11 cases.
1/2a accounted for 30 percent; 1/2b for 33 percent;
12 and 4b accounted for 33 percent of the isolates during the
13 1986 surveillance.
Based on surveillance data, it was
14 projected that about 1700 cases and 450 deaths due to
15 Listeriosis occurred in the United States in 1986.
16
Next slide.
Listeria monocytogenes can cause
17 illness if it penetrates the lining of the GI tract.
Once
18 the organism penetrates the tissue, it can protect itself
19 from phagocytosis, grow and then migrate throughout the host.
20
The chance of tissue invasion is thought to depend upon the
AIM REPORTING SERVICE
(773) 549 - 6351
131
1 number of organisms consumed, host susceptibility and
2 virulence of the organism.
3
In the 1986 surveillance, there were 179
4 nonperinatal cases.
There was a 2-month-old and a
5 3-year-old, but the other 177 cases were all age 16 or over.
6
56 percent of the cases occurred in males; 66 percent of
7 the cases had sepsis; 19 had sepsis and meningitis; and 12
8 percent had meningitis only.
About 3 percent of the cases
9 had a focal infection caused by Listeria.
10 Listeriosis increased with age.
The incidence of
84 percent of the cases were
11 over age 50, and 40 percent of the cases were over age 70.
12
In adults, fatalities also increased with age.
13 there was a 35 percent fatality rate.
Overall,
In cases over age 60,
14 the fatality rate was 41 percent.
15
There were 67 affected pregnancies.
80 percent
16 of the pregnancies resulted in live birth, and one of the
17 neonates died.
Of the live births, 75 percent were culture
18 positive, so transmission of Listeria to the fetus does not
19 always occur.
80 percent of the culture positive babies had
20 an early onset Listeriosis.
AIM REPORTING SERVICE
(773) 549 - 6351
132
1
Early onset is defined as a case of Listeriosis
2 in a neonate between birth and seven days of age.
Early onset
3 is often characterized by a premature birth, respiratory
4 distress and circulatory failure.
In 1986, 80 percent of
5 the early onset neonates had sepsis, and 20 percent had
6 meningitis.
7
20 percent of the culture positive babies had late
8 onset Listeriosis.
Late onset is defined as Listeriosis in
9 a neonate between 8 days and 28 days of life.
10 onset neonates are born healthy and at fullterm.
11 is more common in the late onset babies.
Usually late
Meningitis
The mothers of late
12 onset babies usually had an unaffected pregnancy and no
13 prodromal illness.
Listeria is rarely isolated from the
14 mother, and the source of Listeriosis is often not identified
15 in late onset cases.
Data was available for 31
16 maternal cases in the 1986 surveillance.
58 percent of the
17 mothers experienced premature labor or premature membrane
18 rupture; 32 percent of the mothers had sepsis or fever; and
19 10 percent aborted their fetus.
There was no meningitis and
20 no deaths reported in the maternal cases.
AIM REPORTING SERVICE
(773) 549 - 6351
Listeriosis is
133
1 rarely life-threatening to the mother.
Other studies in the
2 literature suggest that Listeria does not cause repeated
3 abortions in the same women.
4
Next slide.
This slide shows a few examples of
5 outbreaks and sporadic reports of Listeriosis that have
6 occurred since 1988.
Listeriosis has been reported in
7 several countries, and a variety of foods have been implicated
8 as the vehicle of infection, including turkey franks, cheese,
9 mushrooms, pate, fish and hot dogs.
10
This slide shows some of the milder symptoms that
11 have been associated with Listeria infection.
It's been
12 estimated that 33 percent of all cases give mild symptoms
13 and that most cases occur sporadically.
Mild symptoms
14 include chills, diarrhea, nausea, vomiting, fatigue,
15 abdominal cramps.
Reports of mild symptoms suggest the
16 possibility that many illnesses caused by Listeria may go
17 unreported.
18
This slide shows events where most of the cases
19 reported mild symptoms -- not all the cases, but most of the
20 cases.
Again, mild symptoms associated with Listeria
AIM REPORTING SERVICE
(773) 549 - 6351
134
1 infection have been reported in several countries, and a
2 variety of foods have been implicated as the vehicle of
3 infection.
I'd just like to speak a little bit about the cases
4
5 in Denmark.
These cases involved babies at a daycare center.
6
There was a 2-year-old that got fever and was hospitalized.
7
After the fever subsided, he got diarrhea.
8 cultures were obtained.
Blood and stool
The child was treated for his
9 symptoms and released after two days in good clinical
10 condition.
After discharge, blood cultures grew Listeria
11 monocytogenes.
12 symptoms.
The baby was readmitted but no longer had
Two other babies that attended the same daycare
13 were also admitted to the hospital, released in good condition
14 and then readmitted when the blood culture came back positive.
15
After the second admission, blood cultures from all three
16 babies were negative, but stool cultures grew Listeria
17 monocytogenes 4b.
18 established.
The source of the outbreak was not
But this example shows that mild symptoms can
19 occur even if a blood culture is positive.
20
The peer reviewed literature shows that the
AIM REPORTING SERVICE
(773) 549 - 6351
135
1 incubation period associated with Listeria infection can
2 range from less than 24 hours to approximately 3 months.
3 Incubation associated with severe illness, like sepsis and
4 meningitis, can range between several days to a few months.
5
The incubation period associated with gastrointestinal
6 symptoms can range between several hours and a few days.
7
The large bowel is the principal reservoir for
8 Listeria in humans.
Several studies have looked at fecal
9 carriage to gain insight into how the disease is transmitted,
10 especially in sporadic cases.
I show here two examples of
11 fecal carriage studies.
12
In Germany, less than 1 percent of persons with
13 diarrhea and healthy food workers were fecal carriers.
In
14 Scotland, approximately 2 percent of pregnant women and 3
15 percent of nonpregnant women were fecal carriers.
In the
16 literature, estimates of fecal carriage ranges between less
17 than 1 percent to 21 percent.
18
It's not known how fecal carriage relates to the
19 length of incubation or to the occurrence of Listeriosis,
20 although it's been suggested that in fecal carriers, stress
AIM REPORTING SERVICE
(773) 549 - 6351
136
1 can undermine resistance; and then carriers can get the
2 disease.
3
This is the Listeriosis incident trend from the
4 1989 to 1993 surveillance.
The bar chart shows cases per
5 million on the y-axis and year on the x-axis.
About 1990,
6 as more information became available, the regulatory agencies
7 and private industry developed plans to reduce the incidence
8 of Listeriosis.
9
Industry initiated HACCP programs and increased
10 sanitation to eliminate contamination.
The regulatory
11 agencies expanded programs to remove contaminated foods
12 before retail sale.
There was also a consumer education
13 campaign that focused on food safety.
14
Shortly after these efforts were initiated,
15 Listeriosis declined from about 7.9 cases per million in 1989
16 to about 4.4 cases per million in 1993.
The decline occurred
17 in diverse geographic areas of the United States.
And also,
18 about the same time, Listeriosis declined in the United
19 Kingdom after the government issued a health warning.
20
This data is from FoodNet.
FoodNet is an active
AIM REPORTING SERVICE
(773) 549 - 6351
137
1 surveillance program.
The purpose of FoodNet is to determine
2 the frequency and severity of foodborne illness.
To identify
3 all cases of confirmed disease, FoodNet personnel contact
4 each clinical laboratory in each surveillance area in each
5 catchment area, either weekly or monthly.
6
This slide shows Listeriosis compared to other
7 pathogens that are tracked by FoodNet.
There were
8 approximately .5 cases per 100,000 population in 1998.
Data
9 for 1996 and 1997 also showed that there was approximately
10 .5 cases per 100,000 population in those years.
11
This chart shows FoodNet data from 1997.
The
12 y-axis shows cases per 100,000, and the x-axis shows ages
13 in years.
From this graph, you can see that most cases occur
14 in the very young and in the very old.
When this same data
15 was broken down by sex, the ratio of males to females was
16 approximately equal.
This is approximately the same picture
17 that you would see from the 1986 surveillance.
18
A seasonal trend of Listeriosis has been referred
19 to in literature for many years.
This slide shows combined
20 FoodNet data from 1986 and 1997.
The y-axis shows cases per
AIM REPORTING SERVICE
(773) 549 - 6351
138
1 month per million population.
2 of the year.
And the x-axis shows month
There's an apparent increase in cases between
3 late spring to autumn, but the reason for this apparent
4 increase is not known.
5
This graphic shows some of the pathogens that are
6 being tracked by FoodNet on the y-axis.
On the x-axis, it
7 shows the percent of isolates from hospitalized individuals.
8
Listeria had the highest hospitalization rate in 1998.
9 Compared with other pathogens like Salmonella and Shigella,
10 which occurred more often, Listeria put more people into the
11 hospital on a percent basis.
12
Listeriosis also had the highest hospitalization
13 rate and the highest case fatality rate in 1997, 1998.
14
In conclusion, I found by reviewing the literature
15 that Listeriosis is a deadly foodborne illness that can be
16 transmitted in many foods, but it is not product specific.
17
Of the FoodNet pathogens, Listeria has the highest
18 hospitalization rate and the highest case fatality rate.
19 Listeriosis cases could possibly increase in the future due
20 to our aging population and to the use of immunosuppressive
AIM REPORTING SERVICE
(773) 549 - 6351
139
1 medications in surgery and due to the AIDS epidemic.
And
2 intervention may decrease cases of Listeriosis in the future.
3
That's the end of my presentation.
4
MR. MICHAEL JAHNCKE:
Thank you, Dr. McCarthy.
5 Are there questions from the subcommittee?
6
MR. BRUCE TOMPKIN:
Bruce?
This is Bruce Tompkin.
On the
7 conclusion, it states that Listeriosis is not product
8 specific.
And in a general sense that may be true; however,
9 it is product-specific in terms of those foods in which
10 multiplication can occur.
11
DR. PATRICK McCARTHY:
What I tried to point out
12 there is that it's in hot dogs; it's in vegetables; it's in
13 a variety of foods.
And in that sense, it's not
14 product-specific.
15
MR. BRUCE TOMPKIN:
So, within each of those
16 commodities, it is product-specific is what I was saying.
17
18 questions?
19
MR. MICHAEL JAHNCKE:
Thank you.
Other
Yes, Mike.
MR. MICHAEL DOYLE:
This is Mike Doyle.
Could you
20 elaborate on this outbreak in Finland that was associated
AIM REPORTING SERVICE
(773) 549 - 6351
140
1 with butter?
DR. PATRICK McCARTHY:
2
3 to at this time.
I don't think I'm prepared
I'd need some more time before I could talk
4 about that.
5
MR. MICHAEL JAHNCKE:
6
MR. MORRIS POTTER:
Other questions?
Morris Potter.
Yes.
I'd just like
7 to point out for the committee that three of the areas covered
8 by surveillance in the last case-control study fall into the
9 FoodNet catchment area, so while all of the studies on
10 Listeriosis aren't the same, there is some overlap that allows
11 one to look for general trends.
MR. MICHAEL JAHNCKE:
12
Thank you.
Any other
13 questions?
14
15
Thank you very much for an excellent presentation.
Thank you.
16
DR. PATRICK McCARTHY:
17
MR. MICHAEL JAHNCKE:
18 Richard Raybourne.
Thank you.
Our next speaker is Dr.
He will be addressing characteristics
19 of Listeria monocytogenes, dose-response.
20
DR. RICHARD RAYBOURNE:
I'd like to thank the
AIM REPORTING SERVICE
(773) 549 - 6351
141
1 committee for the opportunity to make this presentation and
2 also to thank the collaborators in the dose-response effort
3 whose names are listed there and two of whom are in attendance
4 today.
Next slide, please.
5
There are probably many ways
6 to define -- or at least several ways to define dose-response
7 and the concept of the dose-response model.
I've chosen one
8 that was in one of the other Listeria risk assessments by
9 Farber, et al., and that is the dose-response model provides
10 a functional relationship between the probability that an
11 individual will contract Listeriosis and a specific dose or
12 level of exposure to a virulent strain of Listeria
13 monocytogenes.
And I thought that was a reasonable
14 definition, and I didn't think I could improve on it very
15 much.
16
So, I just lifted it from the paper.
In looking at the possible sources for information
17 on dose-response, there are four listed here.
The first
18 we've heard something about in Dr. McCarthy's previous talk
19 -- that is, the epidemiology and case report information.
20
In addition to that, other possible sources include animal
AIM REPORTING SERVICE
(773) 549 - 6351
142
1 studies and in-vitro studies of various sorts which have
2 addressed questions which are also related to dose-response.
3
Go on to the next slide, please.
Some of the
4 parameters that might go into calculating or developing a
5 dose-response model are, obviously, the number of organisms;
6 the food matrix or the food in which the organisms are existing
7 at the time that they are consumed; the virulence of the
8 particular Listeria strain; and the host susceptibility -9 that is, the resistance or susceptibility of the host to
10 infection.
11
By combining these various factors, you would
12 develop several types of outcomes ranging all the way from
13 asymptomatic carriage of Listeria through more mild
14 diarrheal-type illness to invasive disease to the ultimate
15 end point of death in some individuals and also the fetal
16 abortions, as well.
17
The first issue I'm going to touch on is the issue
18 of the food matrix.
And this goes to the point that was made
19 earlier in regard to the data initially on survival of
20 Listeria in various foods, except the way that I'm presenting
AIM REPORTING SERVICE
(773) 549 - 6351
143
1 or thinking of it here is in the more qualitative sense of
2 the effects of the types of treatment as opposed to the
3 quantitative or number of things -- that is, to raise the
4 question of whether adaptation of Listeria to a acidic or
5 a high-salt environment can actually alter or result in the
6 selection or adaptation of a functionally more virulent
7 population of Listeria such as improving its ability to
8 survive the stomach acid barrier or within some host
9 phagocytic cells, as well as a result of adaptation to a harsh
10 environment.
Whether the specific environment, the specific
11 stress in the food environment is actually the same stress
12 may not actually be relevant due to the sort of global stress
13 responses in some of these organisms resulting in the
14 phenomenon that's sometimes referred to as cross tolerance
15 among these pathogens.
16
In addition, another area that might well be
17 considered is the issue of the fat content in foods,
18 specifically again the question of whether a high-fat content
19 and the sort of relationship between Listeria and the
20 structure of the food and the fat mice cells, for example,
AIM REPORTING SERVICE
(773) 549 - 6351
144
1 could actually protect Listeria from gastric acid or even
2 modulate its interaction with some host cells, perhaps.
3
I have not directly found a tremendous amount of
4 evidence on this area.
But I did find one reference in --
5 I think it was in the Massachusetts outbreak where there was
6 actually a protective effect of skim milk versus whole or
7 2 percent milk on one outbreak.
I think this is an area where
8 additional data would also be needed.
9
Moving on from the food matrix issue to the area
10 of numbers of organisms associated with illness, this is a
11 collection of basically case report and epidemiological data
12 which contains some dose information in it in which an effort
13 was made to quantify the level of Listeria.
And in some
14 cases, an effort was also made to determine what the
15 consumption was to actually get to a dose.
So, in these cases
16 where it just says, "The dose was a given CFU," that means
17 that it was normalized for food intake.
And in those where
18 it says, "CFU per gram," it means that the intake of the food
19 was uncertain.
So, we don't actually know how much was
20 consumed.
AIM REPORTING SERVICE
(773) 549 - 6351
145
Again, there may be other cases that I don't know
1
2 about or that our group doesn't know about.
And we would
3 definitely appreciate information related to dose from any
4 other sources that the audience may know of.
What you can say about this is that there's
5
6 certainly a wide range of doses, and they're basically all
7 over the place in terms of the level of Listeria implicated
8 in illness.
This type of data and various other subsets of
9 data like this have been used in three other Listeria risk
10 assessments to produce dose-response models.
The next slide, please.
11
In the dose-response of
12 studies in the Farber, et al. risk assessment, they developed
13 the dose-response curves for both high -- normal populations
14 and high-risk populations based on a Weibull-Gamma model.
15
In this particular graph, it plots the total number of
16 Listeria monocytogenes cells versus the probability of
17 illness.
This was based on approximate ID-10 and ID-90 doses
18 which were extrapolated from case report information.
19
In another risk assessment, Buchanan, et al.
20 developed a conservative model using consumption data for
AIM REPORTING SERVICE
(773) 549 - 6351
146
1 a single food source and Listeria incidence data.
In this
2 dose curve, the plot is again the log of Listeria
3 monocytogenes cells versus the probability of illness.
4
Finally, more recently, another risk assessment
5 was done for Listeriosis derived from soft cheese
6 consumption.
7
Again, this used the same mathematical model.
This is a little bit harder to sort of access what the
8 cystograms represent.
But I will explain that the plot here,
9 the risk of illness from one serving of cheese versus the
10 probability of illness.
The upper curve represents the curve
11 for the high-risk population, and the lower curve represents
12 the low-risk population.
13
The point here is not to particularly dwell on these
14 models but to make the point that there are some limitations
15 to the approach used in these studies.
And, clearly, these
16 are all based on epidemiologic data which -- in addition to
17 this, in these studies, the virulence is basically assumed
18 in the sense that virulence would be considered a more or
19 less absolute characteristic, either virulent or avirulent,
20 and that the host susceptibility in both of these studies
AIM REPORTING SERVICE
(773) 549 - 6351
147
1 -- in all three of these risk assessments -- was identified
2 as an important variable.
However, in terms of developing
3 ways to address the issue of relative susceptibility, this
4 was essentially based on, to use the term quoted from one
5 of the studies, a "rough approximation of the relative
6 susceptibility."
7
So, for the rest of the time, I'm going to try to
8 present some approaches by which we could use some other data
9 sources other than the epi-data and case report data to try
10 to improve the level of -- or decrease the level of uncertainty
11 in these dose-response models, particularly dwelling on the
12 issues of pathogen virulence and host susceptibility.
13
And so, I'm going to present some animal and various
14 other kinds of -- and other kinds of data, in-vitro data,
15 which have been developed extensively in Listeria since
16 Listeria is a favored organism for both microbiologists and
17 immunologists alike.
18
This is a brief overview of the types of studies
19 that have been done and is not intended to be an exhaustive
20 review of Listeria virulence or immunological mechanisms
AIM REPORTING SERVICE
(773) 549 - 6351
148
1 associated with Listeria.
But the focus is on what kind of
2 data in these studies can be used to help us in development
3 of models.
First, dealing with the issue of pathogen
4
5 virulence.
We might pose the question:
Can experimental
6 virulence studies be used to identify a range of relative
7 Listeria virulence?
If you'll look at our -- going back to
8 our data sources, in looking at human studies, as we've heard,
9 the outbreaks are focused on a small number of predominant
10 serotypes: the 1/2a, 1/2b and 4b.
Although, if you noticed
11 in the slide on the outbreaks, the butter outbreak was
12 mentioned in there.
13 3a.
14
And I believe it was actually a serotype
So, an exception to every rule, I guess.
And it's important here, I think, to remember when
15 talking about these serotypes -- and also, the phagetypes
16 and ribotypes -- that these data are essentially valuable
17 epidemiologic tools but are not necessarily mechanistically
18 related to the virulence of the organism as well, which I'm
19 sure you're all aware of.
20
Next, please.
One virulence factor that's been
AIM REPORTING SERVICE
(773) 549 - 6351
149
1 studied extensively in in-vitro studies is Listeriolysin O,
2 which we've already heard discussed today.
Essentially,
3 it's produced by all clinical isolates of Listeria.
And
4 in-vitro studies have revealed that it's required for
5 survival within macrophage cell lines, which are an important
6 line of defense against Listeria.
But this is also not an
7 absolute in that the survival of even Listeriolysin O positive
8 Listeria is actually limited in in-vitro studies to a small
9 percentage of the bacteria, indicating that there is some
10 selection or adaptation that goes on in this system, as well.
11
Listeriolysin O negative strains, however, do not survive
12 at all in these in-vitro macrophage survival models.
13
Functionally, the Listeriolysin O enables the
14 organism to escape from the phagolysosome of the macrophage
15 and mediate the next phase of its virulence cascade or
16 mechanisms which would be the cell-to-cell spread.
That is,
17 Listeria can also invade nonphagocytic cells -- such as liver
18 cells, for example, and move within epithelial cells -- and
19 move within the cytoplasm and spread from cell to cell by
20 means of actin polymerization.
The molecule or the virulence
AIM REPORTING SERVICE
(773) 549 - 6351
150
1 determinate responsible for this is a surface protein Act A
2 which mediates actin polymerization.
3
In addition to this, there are also a series of
4 proteins involved in getting the organism into the cell in
5 the first place.
One of these is the Internalin protein InLA
6 which facilitates adherence to and invasion of phagocytic
7 cells.
8
Next, please.
Looking at how these studies based
9 on essentially salt culture models pan out in animal studies,
10 it's observable that Listeriolysin strains are all -11 Listeriolysin O negative strains are avirulent in mice in
12 parenteral and oral inoculation studies.
13
In addition to this, Act A negative strains also
14 show reduced infectivity in mice.
And, finally, another
15 group of virulence determinates, the phospholipases, play
16 an important role in the ability of Listeria to evade the
17 early host neutrophil-mediated defense mechanism in the mouse
18 liver, which has been shown in in-vitro studies.
19
So, we can look at what some of this data tells
20 us in terms of dose-response in the next slide.
AIM REPORTING SERVICE
(773) 549 - 6351
In this
151
1 study, this is a study based on oral inoculation and shows
2 a reduction in the number of colony-forming units in the mouse
3 spleen and liver comparing hemolysin positive and hemolysin
4 negative Listeria strains.
So, this gives us a kind of
5 quantitative data based on the presence or absence of
6 hemolysin in an oral inoculation model.
7
The next example shows the fact -- basically, the
8 take-home message from this is that the Listeriolycin is not
9 the whole story in terms of in-vivo virulence in the animal
10 models in that strains which have the Listeriolycin but lack
11 the phospholipase C are reduced in virulence.
12
Putting all the sort of animal virulence factor
13 studies together into a model of what happens in the oral
14 infection in the mouse model in Listeria, you could summarize
15 it by saying that Listeria can attach via the attachment
16 virulence factors to either M-cells in the gut or gut
17 epithelial cells, become internalized, then move through the
18 cell via means of actin polymerization and emerge on the other
19 side of the gut barrier to be taken up by macrophages, which
20 they are capable of survival in, and from there they're
AIM REPORTING SERVICE
(773) 549 - 6351
152
1 capable of then disseminating to various tissues and causing
2 various pathologies in the animal.
3
Next, please.
Looking at the last component of
4 the dose-response parameters, host susceptibility, the
5 question that we're posing here is:
Can animal models of
6 immunocompromised states provide us with any useful
7 quantitative data on relative susceptibility in humans?
8 This is a fairly ambitious question.
However, I think that
9 as we progress through there, you can see that there may be
10 some relationships that are possible to exploit in this
11 question.
12
We know from looking at human studies that healthy
13 adults are usually asymptomatic carriers.
Nonperinatal
14 disease usually occurs in individuals as various predisposing
15 conditions.
For example, pregnancy, very young, infants,
16 individuals with AIDS -- although, this is actually kind of
17 an interesting case because parenthetically, when the AIDS
18 epidemic first developed, it was initially thought that
19 Listeria would be a common opportunistic infection.
And,
20 in fact, it turned out to be actually a sort of unusual
AIM REPORTING SERVICE
(773) 549 - 6351
153
1 opportunistic infection in AIDS, relatively speaking.
And
2 there's a reason for that which will emerge later on in the
3 discussion.
Cancer, immunosuppressive therapies of various
4 kinds and, finally, old age are other predisposing
5 conditions.
6
What you can say about this is that all of these
7 predisposing conditions are likely to involve different types
8 of immunosuppression mechanisms.
That is to say, the factors
9 that predispose in pregnancy are probably different than the
10 factors that may predispose in cancer or in infancy or in
11 old age on a mechanistic level.
And this is more or less
12 what the mouse animal model of Listeria infection tells us.
13
In fact, one of the most useful of these models
14 and instructive has been the use of the severe combined
15 immunodeficiency mouse model.
And it was this model that
16 led to the realization that there was an extremely important
17 interaction of innate and adaptive immune systems in the
18 mouse.
That is that the SCIDS mice, the immune-deficient
19 mice which lack either both T-cells and B-cells, do not clear
20 an infection but also, at the same time, do not succumb readily
AIM REPORTING SERVICE
(773) 549 - 6351
154
1 to the infection.
In fact, they remain chronically infected,
2 which was kind of a surprise at the time of the initial
3 observation, I would think.
4
The neutralization, however, of the Cytokine
5 Interleukin 12 or tumor necrosis factor L for either one of
6 those results in an increase in the lethality of the infection
7 in SCIDS mice and an increase in CFUs to quantify it again,
8 thinking always of what quantitative data we can get from
9 this, by between 1 and 3 logs.
10
The take-home message from the SCIDS mouse model
11 is that in the absence of T-cells, the infection is controlled
12 but not eliminated.
Various studies have demonstrated that
13 this effect is mediated by the polymorphic nuclear
14 leukocytes, neutrophils -- primarily, monocytes, which are
15 producing Interleukin 12 -- and NK-cells, which are present
16 in these animals which produce NK or natural killer cells,
17 which produce gamma interferon, which is one of the most
18 important host-resistance mediators in the mouse model of
19 Listeria.
20
On the next slide, this model, the SCIDS model is
AIM REPORTING SERVICE
(773) 549 - 6351
155
1 summarized by showing on the top, "SCIDS Mice," which remain
2 heavily infected, chronically infected with Listeria.
But
3 the Listeria is held in check by the innate immune system
4 mechanisms -- that is, the NK cells and the neutrophil
5 populations.
6
In the normal mice, these things are operating
7 early on in the infection until such time as the T-cell
8 mediated mechanisms kick in, resulting eventually in sterile
9 immunity in this model.
10
Looking at the next slide, you can see that this
11 has a direct impact on the dose-response to Listeria in a
12 system where neutrophils are depleted by a monoclonal
13 antibody against the neutrophil determinant.
The
14 dose-response effect is really quite remarkable.
That is,
15 the infective -- the lethal dose in this system essentially
16 drops from four times ten to the eighth to four times ten
17 to the fourth or a four-log increase in susceptibility, you
18 might put it, in this particular mouse model in that zero
19 of five of these -- it may even go lower than this -- zero
20 of five of the controls are killed, whereas three of five
AIM REPORTING SERVICE
(773) 549 - 6351
156
1 of the neutrophil-depleted animals are killed.
2
Next, please.
The purpose of this slide is not
3 to have you figure out one single thing that's on this.
This
4 is the pathway of the -- and I put it up here for the point
5 of showing that extensive studies have been done to show,
6 to elucidate the various pathways involved in resistance.
7
The point is that within these various mechanistic
8 studies are embedded information on dose-response to Listeria
9 that is linked specifically to certain kinds of immune
10 mechanisms.
11 slide.
These I have tried to summarize on the next
Looking at various types of ways to manipulate this
12 system, you can see that recombinant Interleukin-1
13 administered to the mouse results in a 250-fold decrease in
14 the level of infection in the spleen.
15
Looking at the Interleukin 6 knockout, there's a
16 300-fold effect.
That is a knockout animal.
But this animal
17 lacks Interleukin-6; therefore, in the absence of that
18 component of the immune system, there's a 300-fold increase
19 in CFUs.
20
Using, again, a monoclonal antibody to deplete
AIM REPORTING SERVICE
(773) 549 - 6351
157
1 Interleukin-12, there's a 500-fold effect.
2 is a thousand-fold effect.
Gamma interferon
TNF alpha, also a thousand-fold
3 effect in the mouse model.
4
I wanted to also mention at this point, while we're
5 on the topic of Cytokines, what is happening and some of the
6 events that go on in the pregnancy model as well because they
7 fit in nicely to what we know from the mouse studies.
And
8 that is that there's studies in both human and animal systems
9 that show there's actually an inhibition of NK cell function
10 during pregnancy.
And we know from the animal studies that
11 NK cells are extremely important in the resistance to Listeria
12 infection.
13
In addition to this, there's a shifting of the
14 T-cell responses during pregnancy towards what's called a
15 Th2 or T-helper-2 type Cytokine secretion pattern.
That is,
16 Interleukin-1, Interleukin-5 and Interleukin-10 are
17 produced.
It's also been shown in other -- in studies in
18 the mouse model that the inhibition of Interleukin-4 actually
19 has a beneficial effect on survival of mice infected with
20 Listeria monocytogenes so that those things which tend to
AIM REPORTING SERVICE
(773) 549 - 6351
158
1 favor production of IL-2 are actually detrimental in -- of
2 Interleukin-4 are actually detrimental in terms of the
3 infection.
And this is one of the events that's going on
4 during pregnancy.
5
In addition to this, it's also been reported that
6 spontaneous abortions in humans are associated with an
7 increase in the sort of yin-to-the-yang here, the Th1
8 Cytokine.
When this type of response gains predominance,
9 it essentially begins to recognize the fetus as a foreign
10 body and reject it.
And it's worth noting that Listeria is
11 one of the prime ways to attempt to drive this kind of
12 response.
So, there may be a link there in the human system
13 that's doing what we can see in the animals.
14
Finally, of course, in terms of these animal
15 studies, there are some serious questions that need to be
16 asked about the use of these various animal models.
17 of all, would be:
First
Does the use of gene knockout or monoclonal
18 antibody-based deletion have any relevance in humans?
19
Secondly, do the mechanisms defined in the mouse
20 model operate in human infections?
There's very, very little
AIM REPORTING SERVICE
(773) 549 - 6351
159
1 information on what is happening mechanistically in human
2 Listeriosis, at least that I've found.
Maybe, again, some
3 of the committee members know more information that I'm not
4 aware of.
5
And finally, a kind of subset to this:
Can the
6 host-resistance mechanisms identified in the animal studies
7 be connected with human biomarkers of exposure and
8 susceptibility?
That is, can we use what we know are
9 important biomarkers in animals -- gamma interferon, TNF
10 alpha, for example -- and use them to answer questions about
11 human exposure and susceptibility to Listeria?
12
In the next slide, this is kind of a bit of a
13 tongue-in-cheek slide in a sense, coming from the Washington
14 Post just this past May 13th.
Not to give anyone the
15 impression that the Centers and FDA might be working at
16 cross-purposes in some instances.
But the recently-approved
17 drug, Enbrel, which has produced spectacular results in
18 treatment of rheumatoid arthritis, may have caused serious
19 infections in some patients, six of who have died.
20
Enbrel is a biological response modifier,
AIM REPORTING SERVICE
(773) 549 - 6351
160
1 chemically engineered to attract and neutralize tumor
2 necrosis factor alpha.
Therefore, there is some
3 relationship in terms of what we know, at least about Listeria
4 infection in mice and these kinds of drugs.
5
In addition to that, one could only anticipate that
6 as more of these mechanisms are investigated and the drug
7 design becomes more sophisticated, there will be more and
8 more therapies like this that are not just general
9 immunosuppressive therapies, but very specifically targeted
10 to certain immune mechanisms.
So that there may be more and
11 more instances where sort of designer drugs can knock out
12 specific components of the immune system to a good effect
13 in the treatment of inflammatory disease, but to a possible
14 detrimental effect in terms of susceptibility to illness.
15
Secondly, as has been mentioned previously, we're
16 in the process of developing in conjunction with the
17 University of Georgia the Rhesus-pregnancy model.
And in
18 addition to the dose-response data -- which will undoubtedly
19 not be available for the July 6th deadline, but hopefully
20 sometime in the future, the absolute numbers in dose-response
AIM REPORTING SERVICE
(773) 549 - 6351
161
1 -- we're also trying to develop some biomarker data in
2 conjunction with that study so that we can then if not look
3 at -- if we can then verify what's happening in the mouse
4 model and this sort of closely-related non-human primate
5 model, it may go a long way to validating the use of the animal
6 data in terms of modelling the relative susceptibility.
7
Next, please.
Going back to the first slide and
8 sort of summing it up and restating or stating maybe clearly
9 for the first time, how we're going to use these various pieces
10 of data or how we're proposing to use these pieces of data,
11 in terms of the issue of numbers of organisms and food matrix,
12 we're proposing to develop distributions for probability of
13 illness based on the human data.
14
Ultimately, we hope in the future to be able to
15 incorporate information from the dose-response studies
16 ongoing now when they become available.
We also will,
17 hopefully, as more information from epi-studies comes
18 available, that will also be incorporated.
But at the
19 present, we're essentially operating from the same data set
20 that other risk assessment efforts have operated from in terms
AIM REPORTING SERVICE
(773) 549 - 6351
162
1 of human data.
2
Next, please.
In terms of organism virulence,
3 we're proposing the concept of using the in-vitro and animal
4 data to model a range of virulence for Listeria monocytogenes
5 to determine -- rather than a sort of a plus-minus virulence
6 situation, to see if that could be-- help refine the model.
7
And, finally, in terms of host susceptibility,
8 we're hoping to explore the use of the animal, primarily mouse
9 data, to model relative susceptibility in various
10 immune-compromised states.
Ultimately, we would like to
11 correlate the mouse biomarkers with the primate model as
12 surrogates for human infection.
13
And that's pretty much the status of the
14 dose-response effort and data forces.
15
MR. MICHAEL JAHNCKE:
Thank you.
Thank you, Dr. Raybourne,
16 for an excellent presentation.
17
We're going to just break from regular procedure
18 a little bit.
It's warm in this room, and our audience is
19 probably wilting.
20
We're going to take a 20-minute break.
What that will allow people to do is to break down this wall
AIM REPORTING SERVICE
(773) 549 - 6351
163
1 and open up the two rooms to air this out a little bit.
And
2 then the next one will be our committee discussion with all
3 the speakers and our National Advisory people.
4
So, 2:25, come on back.
5
(Whereupon, a recess was had in this matter.
6
MR. MICHAEL JAHNCKE:
7 afternoon session.
Let us get started on the
Before we do, there's one little
8 housekeeping point.
Committee members need to turn in their
9 -- they've got a calendar for August through December as far
10 as availability for meetings.
Fill that out and leave it
11 with the staff in the hallway.
We're going to have our committee discussion with
12
13 all the committee members plus the presenters for today.
14 Keep in mind, if there are any questions that any of you have
15 about the document itself, now is the time to bring those
16 up.
And also, keep in mind the three questions that were
17 first presented this morning.
18
Question one:
19
The second question was:
20
And the third one is:
Is the scientific approach sound?
Do they have all the right data?
Have they overlooked anything?
AIM REPORTING SERVICE
(773) 549 - 6351
With
164
1 that, we'll -- Yes?
2
DR. ALISON O'BRIEN:
3 this committee.
4
I'm not a regular member of
May I ask a question?
MR. MICHAEL JAHNCKE:
Absolutely.
Identify
5 yourself.
6
DR. ALISON O'BRIEN:
7 Safety Committee.
8
I'm a member of the Food
It's Dr. Alison O'Brien.
I wanted to ask a question of the last speaker,
9 Dr. Raybourne, who is right next to me.
Dr. Raybourne, you
10 were talking about using animal model data, pili mouse model
11 data as part of your dose-response assessment, guesstimates,
12 estimates.
13
Are you aware of the older data from Christina
14 Cheers (phonetic) looking at innate susceptibility of
15 different mouse strains to Listeria?
Because there was
16 nothing about the basic genetic host background in your
17 discussions today.
18 modulated.
You talked about Cytokine response being
And I can't remember, unfortunately -- I'm gonna
19 say what I think she found.
And she found there was a gene
20 in mice that controlled early response to infection which
AIM REPORTING SERVICE
(773) 549 - 6351
165
1 allowed certain strains of mice to be several logs more
2 susceptible to Listeria than others.
I think it was a
3 complement, actually, complement-medicated factor on mouse
4 chromosome 5.
And you never -- I could be wrong about that,
5 and I don't want to mislead.
But there's a whole set of data
6 on that.
DR. RICHARD RAYBOURNE:
7
8 Yes, I'm aware of that data.
Rich Raybourne, FDA.
I think it's kind of, as I
9 recall, almost a mirror image of the salmonella ITY data;
10 is it not?
11
DR. ALISON O'BRIEN:
It is not.
12
DR. RICHARD RAYBOURNE:
The strains are different,
13 though.
DR. ALISON O'BRIEN:
14
It is not the same gene; and
15 it doesn't have exactly the same mouse profile in the product,
16 no.
17
But the product of the gene is not ITY, IEN RAM now.
It's a different gene, and I think it affected complement,
18 C-5 component of complement.
I believe the AJ strain of mice,
19 which is low in that complement component, was particularly
20 susceptible to Listeria.
AIM REPORTING SERVICE
(773) 549 - 6351
166
So, since you're using mouse models, I thought you
1
2 might go back and check that.
My data may be wrong, but I
3 know it isn't the same profile as salmonella exactly.
DR. RICHARD RAYBOURNE:
4
5 agreeing with you.
Yeah, that's -- I'm
I'm saying it's not the same.
6
DR. ALISON O'BRIEN:
Oh, it's not the same.
7
DR. RICHARD RAYBOURNE:
I think it's -- in the C-57
8 is relatively more resistant in Listeria and it's more
9 susceptible in salmonella.
10
MR. MICHAEL JAHNCKE:
11
DR. RICHARD RAYBOURNE:
12
Other questions?
But that's a good point.
Thank you.
13
MR. MICHAEL JAHNCKE:
14
MR. BRUCE TOMPKIN:
Bruce?
Bruce Tompkin.
I just had one
15 question.
Both of you mentioned carriage, asymptomatic
16 carriage.
Another one was the phrase where healthy adults
17 are usually asymptomatic carriers.
Is this a reality?
Are
18 there carriers whereby normal, healthy individuals may have
19 an indigenous population of Listeria monocytogenes in the
20 GI tract?
Or is it a transient, just as a result of consuming
AIM REPORTING SERVICE
(773) 549 - 6351
167
1 food; and when stool surveys are conducted, they merely show
2 up as a positive because of whatever exposure?
3
DR. PATRICK McCARTHY:
In the studies that I
4 referred to, they were point prevalence.
5 were there at the time.
And so, they simply
In the German studies, several
6 thousand people were involved.
And they found it in those
7 individuals.
8
They found higher rates when they tested the same
9 person over a period of time.
It's my understanding that
10 there are people that are carrying the organism but do not
11 show symptoms.
How long they carry the organism, I don't
12 know.
13
MR. MICHAEL JAHNCKE:
14
MR. DAVID ACHESON:
Yes, David?
David Acheson.
That, to me,
15 raises of the question of any data out there on
16 person-to-person transmission.
17
DR. PATRICK McCARTHY:
This is Pat McCarthy:
I
18 did see one study -- and, of course, I can't remember exactly
19 the name of the study at this time -- but there was a suggestion
20 that individuals living in the same household may have --
AIM REPORTING SERVICE
(773) 549 - 6351
168
1 there may have been transmission person-to-person.
But, for
2 the most part, in all the studies I looked at, that was not
3 an issue.
Person-to-person transmission was not an issue.
MR. MICHAEL JAHNCKE:
4
5 committee?
Yes, Michael?
MR. MICHAEL DOYLE:
6
Other questions from the
This is Mike Doyle.
Richard,
7 I think I noticed on your slide, you had a estimated dose
8 of ten to the ninth for the butter-associated outbreak.
Did
9 I read that right?
DR. RICHARD RAYBOURNE:
10
Rich Raybourne.
It
11 should not -- if that's what it said, it shouldn't have said
12 that.
I think the dose was, as I recall, ranging between
13 a hundred and about ten to the fourth.
14
MR. MICHAEL DOYLE:
15 the butter.
16
17
Yeah.
That was the count from
But above that, I think I saw ten to the ninth.
And I was curious to know how you arrived at that number.
DR. RICHARD RAYBOURNE:
No.
I think the number
18 is much lower than that.
19
MR. MICHAEL JAHNCKE:
20
MR. BRUCE TOMPKIN:
Other questions?
Bruce?
We haven't really discussed
AIM REPORTING SERVICE
(773) 549 - 6351
169
1 the document.
And I only have two questions.
2 is figure one.
3
The simplest
I've tried to understand it, and I don't.
And there's no sense spending time on it now.
But I couldn't
4 figure it out -- the top portion, in particular.
5
But my other comment really was relating to Page 6.
6
And as part of the background information where this is just
7 all background and introduction, Pages 4, 5 and 6, and it's
8 not in here -- and I'd just like to suggest perhaps you may
9 wish to do this -- but it is to actually compare the policies.
10
I know the intent of this risk assessment is not to address
11 policy at this point in time.
But as a matter of comparison,
12 I thought it would be helpful to compare the policies in other
13 comparable countries, industrialized countries, in terms of
14 their Listeria policies, the numbers of cases per hundred
15 thousand -- and I know CDC will wince at that thought because
16 no one has as good a system as the United States what the
17 data are saying.
18
Anyway, the number of cases per hundred thousand
19 and also any information on percent of positive food samples
20 with the intent to see whether or not there's any relationship
AIM REPORTING SERVICE
(773) 549 - 6351
170
1 between the policy, the actual exposure in terms of percent
2 positive foods that are reported in those countries, and then
3 the public health impact.
And that would just be a matter
4 of background information at this point.
5 be.
That's all it would
It would not be anything actionable, as I understand,
6 from this risk assessment.
7
MR. MICHAEL JAHNCKE:
8
MS. CATHY DONNELLY:
Cathy?
Cathy Donnelly.
I'd just
9 like to follow up on Bruce's point and put in a plug for a
10 comment that was made earlier today in the public comments
11 section.
And that being a focus on production practices that
12 leads to production of a food.
And the case that was being
13 discussed was cheeses, and the focus of the risk assessment
14 was on food type or cheese type.
And I'd like to put in an
15 appeal for production practices, i.e. farmstead cheese versus
16 cheeses made in the manufacturing plant.
And I think you'll
17 see a big difference in incidents.
18
19
20
MR. MICHAEL JAHNCKE:
Other comments, questions?
Yes, Dane?
MR. DANE BERNARD:
Thank you.
AIM REPORTING SERVICE
(773) 549 - 6351
Dane Bernard, not
171
1 an immunologist.
2
So, take your question for what it's worth.
The fine report on how we're gonna model immune response,
3 how do you plan to take what you've got and translate that
4 into what I think most of us would accept as a population
5 who distributes a wide range of immunological conditions
6 which vary.
I guess I'm just curious because we've got models
7 that show different parts of how the immune response can be
8 activated or not activated against this particular challenge.
9
But how do you go from where we're at now to what
10 you, what I think will need to do, which is look at the human
11 condition and the whole host of immunological conditions from
12 whatever you call normal or whatever we rank as normal down
13 to those who are very, very severely immunocompromised?
14
DR. RICHARD RAYBOURNE:
Rich Raybourne, FDA.
I
15 think that the issue that you're raising is one that we're
16 at the moment struggling with as well.
I think that clearly
17 there's a spectrum of -- going to be a spectrum of
18 immunocompromised individuals.
I don't think at the moment
19 we have a good handle on ways that we can realistically measure
20 that in the population as a whole to even get it, to get at
AIM REPORTING SERVICE
(773) 549 - 6351
172
1 what proportion of the population is, quote, unquote,
2 "immunocompromised" and to what degree they're
3 immunocompromised.
It's kind of a technically daunting
4 task.
5
I think that the positive side of using the data
6 that I -- of sort -- of the type that I presented is that
7 it's at least a quantifiable measure as opposed to kind of
8 a rough approximation.
I think we need to try to also in
9 as many ways as we can make sure that what we learn from the
10 animal models, particularly the mouse models, is translatable
11 into the human situation.
This is particularly difficult
12 in Listeria because there's essentially no prospect for doing
13 any kind of human clinical trials in Listeriosis.
And so,
14 the best approach that we have right now is to try to develop
15 a surrogate model, which we're trying to do in a primate,
16 in a primate system.
17
It might also be possible, for example, to develop
18 some of this kind of correlative human data in outbreaks or
19 in following up on patients involved in outbreaks.
But it
20 just hasn't really been done to any extent at the present
AIM REPORTING SERVICE
(773) 549 - 6351
173
1 time.
So, it's a good issue, but I'm sorry we don't have
2 a better answer at the moment for you.
3
MR. DANE BERNARD:
Follow-up, if I might?
We've
4 got data on those populations which seem to be more at-risk
5 -- this is outbreak data -- who gets Listeriosis predominantly
6 and who doesn't.
7 immunologist.
We know enough, I think-- not an
We know enough basis, what you've presented,
8 I think, to theorize what some of the mechanisms of
9 susceptibility might be in those categories.
10
Have you thought into that scenario to see if
11 there's any mileage there?
12 than one-year-old group.
I mean, for example, the less
We know the immune system is still
13 developing, immature, unchallenged, da, da, da, da.
Based
14 on the mouse models that you've got, is there anything that
15 applies there?
16
At the other end of the spectrum, same thing.
DR. RICHARD RAYBOURNE:
Rich Raybourne again.
I
17 think in terms of doing those kinds of studies, we should
18 look at, for example, in levels of quantifiable types of
19 markers, like the Cytokines I mentioned, in these populous
20 -- it's theoretically possible to do that.
AIM REPORTING SERVICE
(773) 549 - 6351
The problem with
174
1 doing that -- at least my understanding of it -- is in the
2 absence of an ongoing infection measuring levels of
3 circulating Cytokines is not going to be very worthwhile.
4
And at the very least, what you would want to do to get into
5 sort of a more technical way of approaching this, if I could,
6 what you would want to do is to somehow collect materials
7 from these individuals, stimulate them in-vitro and look at
8 the ability of the cells to respond.
I mean, it would be
9 a huge and expensive task to do this kind of thing.
10
There may be other simpler ways you can measure
11 this, looking at -- and non-invasive ways, too.
And we're
12 currently trying to think of approaches to this in terms of
13 even to the point of doing serological-type surveys, although
14 this is problematic in Listeria because there's not a lot
15 of evidence that I'm aware of that serum antibody responses
16 are important in resistance to Listeria.
So, I mean, it's
17 a great question.
I wish we could answer it and come up with
18 an approach to it.
And we've certainly thought about it but
19 have not done that at this point.
20
MR. MORRIS POTTER:
Morris Potter.
AIM REPORTING SERVICE
(773) 549 - 6351
Rich, I think
175
1 what Dane is suggesting is that say, for instance, in the
2 geriatric literature, it's known more or less which subsets
3 go first.
And, therefore, if we can look at susceptibilities
4 of various mouse strains that are absent, those things that
5 go in 50-year-olds and then the things that start to go when
6 we hit 60 and so forth, that we might be able to then model
7 the human population for those age groups and suggest when
8 people are going to become more susceptible to infection,
9 when they're going to become more susceptible to serious
10 invasive disease and that sort of thing.
11
MR. WILLIAM JAHNCKE:
Bob?
12
MR. ROBERT BUCHANAN:
Bob Buchanan, FDA.
13 I think I'd like to echo on what Morrie says.
Yeah,
I'm wondering
14 if you may be making this more complicated than is warranted
15 considering the huge range of -- and certainly, you're going
16 to face with the rest of your risk assessment.
Morrie and
17 Jim Smith and I did a presentation a bunch of years ago on
18 trying to get some estimates of increased risks associated
19 with aging.
And while certainly you're gonna have to come
20 up with some kind of fudge factor to relate the increased
AIM REPORTING SERVICE
(773) 549 - 6351
176
1 susceptibility, it was not very difficult to find some
2 age-related decreases in, for example, T-cell proliferation.
3
It was not difficult to come up with age-related equations
4 that we could develop for achlorhydria in the aged.
So, I'm
5 wondering if we couldn't just start off with trying a couple
6 of fairly simple relationships that have been gleaned from
7 these broad population studies, start simple.
And if it
8 didn't work, then get more sophisticated.
9
DR. RICHARD RAYBOURNE:
Rich Raybourne again.
I
10 think that's a good approach, yes.
11
MR. MICHAEL JAHNCKE:
12
DR. WESLEY LONG:
Yes?
I do want to make one point,
13 though.
14
MR. MICHAEL JAHNCKE:
Identify yourself, please,
15 Wes.
16
DR. WESLEY LONG:
Wes Long, FDA.
That it's
17 consistent with some of our conversations yesterday that what
18 we're doing is, you know, we don't have all the data now
19 certainly, clearly.
But what we're doing is laying a
20 framework at this stage and using that to figure out what
AIM REPORTING SERVICE
(773) 549 - 6351
177
1 to do next.
And we talked about how we can modify the risk
2 assessments as more information becomes available. So, I
3 think this sort of thinking is important.
4
Rich sort of mixed up the data we'd like to get
5 from outbreaks, that sort of thing, which is future, which
6 we don't have now.
But by doing this sort of thinking now,
7 I'm hoping that we will sort of lay the groundwork, even though
8 we may not be able to utilize some of the things that he's
9 talking about immediately.
10
MR. MICHAEL JAHNCKE:
11
DR. ALISON O'BRIEN:
Yes?
This is Alison O'Brien.
12 Following up on what Bob Buchanan said about T-cell
13 proliferation, some kind of marker that suggests you might
14 be more susceptible to Listeria.
The question goes back to
15 something Dr. Raybourne said during his talk.
Why aren't
16 a lot of AIDS patients infected with Listeria, or are there?
17
I mean, I know that I saw that as a subcategory.
But to
18 me, it seems a surprisingly small portion, given that if we
19 accept the paradigm that this is an organism that uses
20 protective immunity as related to cell-mediated immunity,
AIM REPORTING SERVICE
(773) 549 - 6351
178
1 not pneumo-immunity.
DR. RICHARD RAYBOURNE:
2
Rich Raybourne again.
In
3 terms of the AIDS question, I think part of the answer -4 and you're right.
It's not as common as you would think it
5 would be among AIDS patients.
And this was one of the sort
6 of statements in their first papers that came out when there
7 were finally some Listeria AIDS cases.
And I think part of
8 the reason for that may relate to the observations with the
9 effects of, for example, the Interleukin 4 and the fact that
10 it acts -- which in CD-4 deficient patients, is going to be
11 lower.
And in the mouse model, when you neutralize -- and
12 this is not a complete answer, but it's sort of a clue -13 that if you neutralize IL-4, you actually ameliorate the
14 Listeria infection in the mouse model.
So, it has kind of
15 a detrimental effect.
16
MR. MICHAEL JAHNCKE:
Yes, go ahead.
17
DR. PATRICK McCARTHY:
This is Pat McCarthy.
It's
18 true that in AIDS patients, in some of the earlier literature,
19 researchers were saying that it's not very common in the AIDS
20 patients.
But then about '86, '87, '88, there was an estimate
AIM REPORTING SERVICE
(773) 549 - 6351
179
1 that AIDS patients have Listeriosis about 150 times more
2 often.
Then, more recent, I believe there was another
3 estimate that AIDS patients may have Listeriosis about 280
4 times as often.
5
So, it's true that in the beginning, the
6 researchers were wondering why Listeriosis wasn't showing
7 up in AIDS patients.
But as more information became
8 available, estimates started to increase.
9
MR. MICHAEL JAHNCKE:
Yes, Bob?
10
MR. ROBERT BUCHANAN:
Bob Buchanan, FDA.
11 I just wanted to affirm that.
Yeah,
My recollection was that the
12 approximate increase in risk associated with Listeriosis and
13 AIDS was about 300-fold.
So, I think there is a very
14 substantial increase in risk.
15
MR. MICHAEL JAHNCKE:
16
MR. DANE BERNARD:
17 an immunologist.
18
Still not.
Dane?
Thank you.
Dane Bernard, not
But we're not on immunology.
Another factor, I think, when you look at the data
19 on incidence of Listeriosis in people with AIDS, you've got
20 to look at the interventions that go on there as well.
AIM REPORTING SERVICE
(773) 549 - 6351
180
1 Prophylactic use of antibiotics, extensive dietary advice,
2 is all provided once a person is diagnosed in that category.
3
So, there's a risk mitigation or risk management strategy
4 there, I think, that has a strong intervention that you're
5 seeing showing up in health statistics.
MR. MICHAEL JAHNCKE:
6
Other comments and questions
7 from the committee?
8
Yes, Bruce?
9
MR. BRUCE TOMPKIN:
This is Bruce Tompkin.
I'd
10 like to have a little clearer understanding as to how
11 information is to be given to the risk assessment team.
If
12 someone has data, do they just say, "Here, Dick Whiting.
13 Here it is"?
Or is there a mechanism -- I know that you went
14 through with a published announcement in the Federal
15 Register, and so there's probably a formal mechanism.
But
16 how do we know that information provided will be used or
17 considered and so on?
And once provided, to what degree --
18 I know this process is one of the processes we're going through
19 now.
20
This is a public process.
This is an open process.
So, I assume data that's provided will become public or
AIM REPORTING SERVICE
(773) 549 - 6351
181
1 available to the public.
2
Could you help me with that a little bit?
3
DR. RICHARD WHITING:
Richard Whiting.
Yeah.
4 There's a paragraph or two in that Federal Register Notice
5 as a result of some of the discussions with us and our lawyers
6 and so on, that I think we're probably breaking some new ground
7 for FDA here, as well.
8
The information that would be submitted, I think
9 you would have to expect that it would become public
10 information.
But we did say in there that we would accept
11 information that has been summarized or blinded and various
12 terminologies like this.
So, if, say, the meat industry,
13 for example, through one of your trade associations wanted
14 to do a quick survey of whatever Listeria your members might
15 have, and the trade association would just present a summary
16 to us, that would be the information that we would have.
17
And as risk assessors, we would then try to evaluate
18 that information as best we can.
The more information, the
19 more details you could provide, the more useful the
20 information would be to us.
But we'll accept what is offered.
AIM REPORTING SERVICE
(773) 549 - 6351
182
1
I would like to think that if some data came in, we might
2 have an indication of what methods were used, what sensitivity
3 -- if it was presence/absence data, what sensitivity might
4 be there.
5
But, again, we will just accept to try to use
6 whatever people are willing to submit to us.
And we recognize
7 this is sort of a new situation, I think, for all of us.
8 And we're gonna try to use this as a scientific process and
9 not a regulatory process, and I guess we'll have to see how
10 it goes.
11
MR. MORRIS POTTER:
This is Morris Potter.
12 could amplify on that a little bit.
If I
Part of the rationale
13 for using risk assessment is that it's a transparent process
14 and that people who look at the risk assessment ought to be
15 able to repeat it using different assumptions.
And that does
16 create a need to make the data sources available.
If there
17 are data that could be useful for the risk assessment but
18 that might be felt inappropriate to become public, they may
19 still be useful in terms of trying to validate things
20 internally.
But I think that our preference is to use risk
AIM REPORTING SERVICE
(773) 549 - 6351
183
1 assessment to help in making our decisions on risk more
2 transparent, more understandable to the broader audience.
3
We wouldn't want to turn our backs on data that
4 could be useful.
And if you have things that you'd like to
5 discuss, we can chat with you.
6
Wes, did you have any clarification on that?
7
DR. WESLEY LONG:
8 Wes Long, FDA.
I do.
First of all -- This is
I have the answer to this because it was
9 reported in Food Chemical News on the back cover page there
10 about six weeks ago that the Risk Assessment Clearing House
11 set up through the Food Safety Initiative was going to be
12 collecting the Listeria data.
13
14 was.
I'm not sure what the source of that information
But there is a Risk Assessment Clearinghouse that is
15 at the University of Maryland that's a part of FDA's new Joint
16 Institute for Food Safety and Applied Nutrition.
And we are
17 in the process of -- This clearinghouse is intended to be
18 a repository for data methods, models, anything to do with
19 risk assessment, initially focusing on microbial risk
20 assessment needs.
AIM REPORTING SERVICE
(773) 549 - 6351
184
Dave Lineback (phonetic), who is the -- I'm not
1
2 sure of his title, the chair or the head -3
MR. MICHAEL JAHNCKE:
4
DR. WESLEY LONG:
Director.
The Director, thank you.
The
5 Director of the Joint Institute for Food Safety and Applied
6 Nutrition will take responsibility -- at this point, we're
7 not really set up, but we could be set up very soon to take
8 data.
He will take the responsibility to do sort of a
9 secondary cleansing of data.
I think you would want to --
10 if you had to blind the data, you would probably want to do
11 that first.
12
But he would be a second mechanism to do that.
And he would provide the assurance, again, that -- of course,
13 all of the information that goes in the Clearinghouse does
14 become public.
But that FDA would never -- this is an
15 opportunity to blind the data again and further assure
16 submitters of the confidentiality of that information.
17
So, Dave Lineback -- I could give you information
18 about how to get in contact with him.
That might be another
19 way to get that information.
20
MR. MICHAEL JAHNCKE:
Morrie?
AIM REPORTING SERVICE
(773) 549 - 6351
185
1
MR. MORRIS POTTER:
I guess if I could recap where
2 we are, Richard has suggested that there is information in
3 the Federal Register where you can send it.
4 that you could send it to Dave.
Wes has suggested
And, in fact, you could also
5 send it to Richard or to me or to Joe Levitt (phonetic) or
6 to anybody else in the building, and we will see that it gets
7 to the right place.
8
MR. MICHAEL JAHNCKE:
9
MR. ROBERT BUCHANAN:
Yes, Bob?
And to answer the second half
10 of your question, Bruce, on how to tell whether or not the
11 data is being used similar to the document that you have in
12 front of you that outlines the data sources that are being
13 considered, the risk assessment itself will detail the data
14 that was selected for use and the criteria for using it.
15 So, if your data didn't get used, you would know it by reading
16 the final document.
17
18 questions?
19
MR. MICHAEL JAHNCKE:
Other comments and
Yes, Dane?
MR. DANE BERNARD:
Thank you.
Dane Bernard.
20 Pat, you covered a number of things in your review of the
AIM REPORTING SERVICE
(773) 549 - 6351
186
1 epidemiological information.
Some of us, I think, who have
2 watched outbreak information and epidemiological studies for
3 some years, occasionally run across things that we don't
4 necessarily agree with, that maybe they weren't in fact quite
5 as well-established as we thought they might have been.
6
Is there any need to or will you be reviewing any
7 of the source information on past studies to see whether it
8 meets some kind of criteria of acceptability in terms of
9 whether we, in fact, have targeted all the right foods or
10 maybe have targeted one or two too many as being implicated
11 in outbreaks?
12
MR. MICHAEL JAHNCKE:
13
DR. PATRICK McCARTHY:
Please identify yourself.
Pat McCarthy.
Before I
14 refer some data over to Clark Carrington who's going to be
15 doing the modelling, I am going to look at it to make sure
16 that it seems reasonable to me that the cases are
17 well-described and that it has the basic information in there,
18 including the rationale for implicating the particular food.
19
20
I'm going to try to summarize the data a little
AIM REPORTING SERVICE
(773) 549 - 6351
187
1 bit, in addition to giving them the raw data.
But summarize
2 the data a little bit to give them an indication of how often
3 a particular food or type of food is being referred to in
4 the studies that I refer to.
So, yes, I'm going to try to
5 be critical in terms of which studies are referred.
6
MR. MICHAEL JAHNCKE:
Bob?
7
MR. ROBERT BUCHANAN:
Bob Buchanan, FDA.
In
8 yesterday's session on vibrio parahaemolyticus, we spent
9 quite a long time on the dose-response area talking about
10 multiple biological end points and what would be appropriate
11 to model in the case of vibrio.
12 enteric pathogen.
And that's a fairly classic
You deal with colonization of the
13 intestinal tract as one biological end point.
14
Sepsis is a second biological end point.
15 Gastroenteritis is an intermediate biological end point.
16 In your presentation, Pat, you gave several different
17 potential biological end points.
And, Rich, you provided
18 a model for infection that would not be too dissimilar from
19 what we were discussing yesterday on vibrio.
20
Have you decided yet on what will be your biological
AIM REPORTING SERVICE
(773) 549 - 6351
188
1 end points that you're going to be modelling or considering?
2
Are you going to do multiple ones?
Is it going to be one
3 for sepsis, one for meningitis, one for neonates?
DR. PATRICK McCARTHY:
4
This is Pat McCarthy.
I
5 had planned to take a look at the studies and, again, to put
6 them together in terms of -- in a lot of different ways.
7 Organize the data for the model in a lot of different ways.
8
And, certainly, sepsis and meningitis are two big end points.
9
I was going to try to give the modeler an estimate of how
10 often those particular end points come up in the literature
11 that I reviewed.
And also, since the more mild symptoms,
12 there seems to be several studies that refer to mild symptoms,
13 I was also going to give them an estimate of how often that
14 seems to show up.
15
In terms of headache or chills or abdominal pain,
16 I'm not there yet in terms of how I'm going to group that
17 data; but it might be -- I do have estimates in different
18 papers of how often subjects or cases had diarrhea or had
19 chills.
And so, I haven't really decided how I'm going to
20 give it to them, but I'm going to try to be open when I do
AIM REPORTING SERVICE
(773) 549 - 6351
189
1 and give it to him the way that's most productive for him.
2
MR. MICHAEL JAHNCKE:
3
MR. ROBERT BUCHANAN:
Yes, Bob?
Sort of a follow-up on this,
4 now directed towards Rich.
5
Rich, do we have any estimates on the probability
6 of colonization or attachment?
In presenting your model of
7 the infection, you just sort of said "attachment," and you
8 didn't really deal with that.
9
Do we have any estimates on what it takes to get
10 attachment, or are there different known attachment
11 mechanisms?
Can you come up with any kind of probability
12 of attachment?
13
MR. MICHAEL JAHNCKE:
Identify yourself, please.
14
DR. RICHARD RAYBOURNE:
Yes.
Rich Raybourne,
15 FDA.
16
In the model I presented, I mentioned the
17 Internalin, which is essentially an attachment-type
18 virulence determinant.
In terms of numbers associated with
19 colonization, I'm not aware -- I don't have that information
20 right now.
There may, in fact, be some because a number of
AIM REPORTING SERVICE
(773) 549 - 6351
190
1 oral infectivity studies have been done.
And whether they
2 used attachment as an end point or not, I'm aware of one study
3 where they looked at invasion in the intestinal wall and
4 quantified organisms invading the intestine of the mouse.
5
But beyond that, no, I don't know of any.
MR. MICHAEL JAHNCKE:
6
Other questions and
7 comments?
Our next presenter, before we have it, on behalf
8
9 of the subcommittee, we'd certainly like to express our
10 appreciation to all the presenters today and all the hard
11 work.
The product is coming along quite nicely.
Thank you
12 very much.
13
Our next presenter is Dr. Richard Whiting.
And
14 he is going to be giving a summary of what has been presented
15 and discussed, presented today.
16
DR. RICHARD WHITING:
I'm just going to be very
17 brief in light of this good discussion we had.
But maybe
18 we can start with Bruce's question on this disputed figure
19 here.
This is the one Bruce is referring to.
We did have
20 some comments on the draft phase, but I guess we didn't get
AIM REPORTING SERVICE
(773) 549 - 6351
191
1 around to revising it.
But it's just trying to say here at
2 the top, "Food consumption, food contamination."
3 go together to form your exposure.
These two
And then this middle part,
4 "Food Vehicle Virulence and Susceptibility" is the disease
5 triangle idea, the hazard assessment, and that leading on
6 to illness.
7 it.
It's trying to give a little sense of flow to
I guess I can see we can do a little bit of editing and
8 reworking of that to make it a little bit more clear.
9
There's kind of an old saying that risk assessment people
10 have had, "Let the data speak."
And I guess that's largely
11 where we are at this point in the risk assessment.
12 see we've accumulated a lot of information.
You can
We've given you
13 some ideas of where we would like to go with it.
14
The next stage for the risk assessors is to try
15 to take all of this information and, really, just see what
16 we can do with it, see what the information can be summarized
17 as.
And I think you've seen the problems with trying to
18 combine the information on the presence of Listeria in foods
19 with the consumption.
20
That data base that Mary has -- I forgot.
AIM REPORTING SERVICE
(773) 549 - 6351
She had
192
1 something like eight or ten different categories of
2 hamburgers.
And then when you get to data like the cheeses,
3 we have some good consumption information on some of the
4 Hispanic cheeses, for example.
But then these data bases
5 don't say anything about whether this was pasteurized or
6 unpasteurized cheese.
And this is the kind of information
7 that we have to try now to pool together and bring out of
8 it the conclusions that we feel are justified in bringing
9 out.
And we may find there will be quite a few areas where
10 we will just say there is not information available that we
11 can go further.
And part of the exercise of doing a risk
12 assessment like this is to highlight the data gaps.
13
I also think that this will be an iterative-type
14 of process.
I think it will be occurring both within the
15 next few months, and I can see us doing an initial summary
16 of the data, which will perhaps highlight certain areas that
17 we will then go back out and try to find more detailed
18 information on.
19
And I can see sort of this second round as a point
20 where we will probably try to get in contact with various
AIM REPORTING SERVICE
(773) 549 - 6351
193
1 people in the industry who might have information on specific
2 consumptions, you know, my question of pasteurized versus
3 unpasteurized in certain groups of cheeses.
Or perhaps the
4 industry might have some information on consumption patterns
5 or food preparation habits.
Various questions like this
6 which we don't have yet but might be very relevant for
7 particular classes of food.
8
So, I see this as an iterative-type process, and
9 it will probably continue beyond the September, October date
10 that we have set as a target for completion of the first part
11 of the risk assessment.
12
And we've also made quite a few references here
13 today to various research projects that are underway, the
14 primate pregnant monkey primate study, and various studies
15 like this -- which obviously won't be ready by September and
16 October but yet, obviously, we want to take that and look
17 at the risk assessment again as soon as that data becomes
18 available.
19
I have been quite heartened today by, I think, the
20 sense of participation here by the industry.
AIM REPORTING SERVICE
(773) 549 - 6351
I was on the
194
1 Salmonella enteritis and egg risk assessment team that the
2 USDA did a year or so ago.
3 approached the industry.
And at that point, we sort of
And I would say they approached
4 us back with quite a bit of trepidation.
And we really did
5 not get very much back that was helpful to it.
I think there
6 was a lot of apprehension about what the whole risk assessment
7 process is about.
I hope everyone is becoming a little bit
8 more aware of just what a risk assessment is and what it does
9 and that people will be a little more willing to participate
10 in this.
I think for all of us, our goals are increased food
11 safety.
And I really do put a plug out there and echo again
12 the conversation we did just have about the submission of
13 data and blinded data.
And I really do put an invitation
14 out to industry and everyone to become active and follow it.
15
And if you have specific information that you can bring to
16 us, that you do that.
17
So, with that, Mr. Chairman, I thank you very much.
18
MR. MICHAEL JAHNCKE:
19 thanks to your entire group.
Thank you very much.
Again,
Certainly appreciate it and
20 thank you.
AIM REPORTING SERVICE
(773) 549 - 6351
195
1
Morrie, I'll turn this over to you.
2
MR. MORRIS POTTER:
I'd like to add
3 congratulations to the risk assessment team.
4 nicely done.
That was very
And on behalf of FDA and FSIS, I would like
5 once more to invite participants today who are not members
6 of the committee to come to the mike, identify yourself and
7 make whatever statements you'd like to about the risk
8 assessment model that was presented today, the direction the
9 team is taking or other comments on the risk of foodborne
10 Listeriosis.
11
Perhaps while people are thinking about that, I'd
12 like to direct a question to Wally.
13 earlier about colonization.
The question arose
And I wondered if you had any
14 information on either transient or long-term colonization
15 from your clinical experience.
16
MR. WALLY SCHLECH:
17 Schlech from Delhausen.
Thanks, Morrie.
Wally
I don't have any information other
18 than to say we did some carriage studies during the
19 now-ancient maritime outbreak in family members, primarily,
20 and did find some carriage.
We assume that's probably
AIM REPORTING SERVICE
(773) 549 - 6351
196
1 because they were eating the same items in the menu and during
2 the time we were sampling which, in fact, was often 30 -3 probably several months after the case, were able to find
4 some Listeria.
But whether these were new items or leftover
5 from the previous, I don't know.
I think there are some Dutch studies -- I believe
6
7 in Europe, some old studies of longitudinal carriage, as I
8 recall.
I think, although most people would suggest that
9 carriage is transient, that it may remain in the bowel flora
10 for a period of time.
11
So, I'm not certain.
I think I wanted to raise a question about the
12 biological endpoints.
I think the febrile gastroenteritis
13 syndrome is very much a distinct entity.
And the thing that
14 wasn't talked about today is the extraordinarily high attack
15 rates within the exposed group for that particular syndrome,
16 whereas mostly the Listeriosis you see, even in the outbreak
17 situation, the attack rates in the overall population are
18 quite low.
19
I think there is some evidence that that may be
20 more related to a huge dose and possibly even local.
AIM REPORTING SERVICE
(773) 549 - 6351
So,
197
1 maybe the hemolysin acts as a local cytotoxin in the gut for
2 a period or something.
3
I don't know.
But I think there is some data.
I've done some
4 work in gastrointestinal carriage in mice and rats.
And we
5 can see carriage persist for a couple of weeks in the
6 droppings.
But we haven't really gone beyond that at this
7 point in time.
8
And there certainly doesn't seem to be any specific
9 attachment factors.
The Internalin protein, I think, is an
10 interesting protein.
But in terms of the things we think
11 about, pili and other sort of typical attachment factors,
12 Listeria doesn't exhibit them.
And we really don't have any
13 information there.
14
15 comments?
MR. MORRIS POTTER:
Thanks, Wally.
Other
In that case, I think we can wrap this up.
16 Tomorrow morning we start again at 8:00 for the plenary
17 session of the National Food Advisory Committee.
18
(Whereupon, the hearing in this matter was
concluded at 19
3:20 p.m.)
20
AIM REPORTING SERVICE
(773) 549 - 6351
198
1
STATE OF ILLINOIS
COUNTY OF C O O K
)
)
)
I, ANNE I. MAZIORKA, CERT, a Notary Public within
and for the County of Cook and State of Illinois do hereby
certify:
That the foregoing transcript was reported to me
by electronic recording, was thereafter reduced to
typewriting under my personal direction and constitutes a
true record of the testimony given;
That the said hearing was taken before me at the
time and place specified;
That the hearing was adjourned as stated herein;
That I am not a relative or employee or attorney
or counsel, not a relative or employee of such attorney or
counsel for any of the parties hereto, not interested directly
or indirectly in the outcome of this action.
IN WITNESS WHEREOF, I do hereunto set my hand and
affix my seal of office at Chicago, Illinois, this
of
, 19
.
AIM REPORTING SERVICE
(773) 549 - 6351
day
199
ANNE I. MAZIORKA, CERT
Notary Public, Cook County, IL
CERT NO. 00140
AIM REPORTING SERVICE
(773) 549 - 6351