• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services

Animal & Veterinary

  • Print
  • Share
  • E-mail

Surveillance in Food Animals by Paula Fedorka-Cray, Ph.D., USDA

DR. FEDORKA-CRAY: I appreciate the opportunity to address you. You will notice in your packets that you have many, many, many more slides that I’m going to show today, and part of that is we just don’t have enough time to show everything, particularly when we talk about four different animals species. So it’s going to be a condensed overview and I would direct you to look at your slides for any other information.

I will release those to Joanne so that they can be put up on the web, too. And with the pitch that Beth Karp gave us for the interactives, I will also direct you to the web sites for the BEAR units so that you can look at additional interactive graphs at your leisure.


I just want to acknowledge a large number of people it takes to run this program. And many of them are here in your audience today. And to them I’m eternally grateful.


There is one change in the collaborators now, Doctor Nora Wineland has moved on to another position in AFIS, and Doctor Bruce Wagner has now taken her place. So that is a change in -- and in just the person that we’re collaborating with, not certainly in the operation of the system.


Something a little different, I know that you will see a lot of these as Ezra pointed out over time, but I wanted to pointed out was the total number of isolates that in fact we have tested since 1997, 57,000 when the slide was made, there will be many more. These include isolates from the slaughter program which I’ll show you in a slide next. In addition to some diagnostic and sentinel sites and on farm isolates. So we have about 30,996 because I counted them last night, in slaughter and the rest of those are going to be diagnostic and other isolates.
You will notice that we have 242 serotypes that we represent. This is quite interesting since there are over 2,500 different serotypes of Salmonella. So you know you find -- I find this to be fascinating in that we only capture about 10 percent of the total number of serotypes of Salmonella that have been reported over time.

One other interesting thing to note is that we have 180 different antigenic formulas that are not identified by serotypes but by antigenic formula. And this almost equals the total number of serotypes that we have. So, I think that this points out some of the limitations that we still have with Salmonella serotyping and what this means as far as disease and transmission and all of the other harbridge of variant factors is really unknown at this point. But it’s something that I think we need to take a look at.

The total number of Campylobacter isolates we’ve tested is at 6,300. We have about 22,000 E. coli isolates and just a little under 11,000 Enterococci isolates.

Of these 57,000 Salmonella isolates, since 2004 when VetNet was launched, we have about 19,000 of those with pulsive types representing 150 of the 242 serotypes and 110 of the 180 antigenic formula. So that gives you an overview of exactly what we have available.

And, you know, as always I’ll make the offer that if there is something in particular that you want to see, some serotype that we need to go back and do retrospectively, we can certainly do that.


These are the total, these are the 30,996, if I punch the calculator right, slaughter isolates, primary isolates that were focused on for this program, by far the number of isolates that we have are from chickens. As part of our work with AGISAR now and my involvement with them, we know that there’s a minimum number of isolates that we would like to have, which is 170, particular per commodity, and you’ll notice that for the last couple of years, with turkeys and swine as FSIS has changed their sampling program, that we now have fewer isolates represented for those particular commodities. And that’s something that we’ll keep a look out for.


Like Ezra, I’m just going to go over some of the data until I get to the crux of what I want to talk about, which is the revamp sampling that we hope to have in the next couple of years.

Interestingly about 80 percent of the non-typhi isolates from humans are pan-susceptible and you will notice that the isolates that are pan-susceptible amongst chickens and cattle are certainly -- there are more of those then we see in turkey and swine. Turkeys by far have the fewest pan-susceptible isolates.

I would point out that as we’ve looked at these over time that again the numbers are going to shift. You will see that there has been a decrease and this could be due to the sampling that changed in 2006 from FSIS.


It’s important when we talk about Salmonella that we look at the serotypes in particular because the serotypes are going to be what drive the resistance that we see. So when we talk about Salmonella on a macro level, include everything together, we’re going to have one representation of resistance within that population. When you start to look by serotype you’ll see very different things emerge.

And I think that what’s important, just to point out here is that the wide variety of serotypes that we have, although Typhimurium shows up amongst three of the four animals versus you’ll notice that it doesn’t show up for turkeys, which I find quite interesting. And you will find that there are some very different serotypes among turkeys then we find for the other species over time.

And again, a lot of the resistance that we see driven per commodity is going to be driven by the serotypes that we see within that particular commodity.


This is one of our newest interactive graphs. And I just want to point out that it works very much the same way as before. The drugs are over on the left-hand side. Serotypes are on the top. Unfortunately we can’t put all of the commodities on one, so you will have to go by each commodity and look at the four different ones. They don’t show up very well on your slides, but they will be posted to the website over time.

These are the 15 top serotypes that we see and we can just change that over time to look at what’s going on. Unlike the increase in SIP resistance that they’ve seen in the human arm we have not seen any SIP or now-resistance for S. enterididis which is the yellow line here over the years, except for a couple of isolates back here in the early 2000's.

One of the things that Beth didn’t point out that I would like to draw your attention to is, is that it’s important to look down here at the total number of isolates, those -- this just scrolls down. When you get to isolates that have N=1 or 5, or 8, or 7, you really have to take all of that with a grain of salt because we really don’t have a lot of numbers there. And that’s important so when you look at some of these graphs and all of a sudden you’ll see something shoot up to 100 percent, it’s really important to go and look at the numbers that we’re talking about.
I think it’s quite interesting when we look at some of the serotypes you see some either the top 15 and we see that Istanbul which is really and odd serotype that one doesn’t think about very often, as far as human illness goes, now it starts to show up in the chicken population.
We saw this with S. bendoca with several other sources early in the 2000's and that has increased over time to have major significance among some of the animal sources.

So, it’s these kinds of things that we’ll be on the lookout for and will be interacting with CDC and others to see how that will change.


A series of graphs that you will now see. And then you will have greater and equal to two, and then greater than equal to five. And so I’m just going to show you some of the greater than equal to five. And again point out that you may see some of these types of increases, particular for like for Typhimunium and Typhimunium-5, but you have to come down here and you have to look at the small numbers. And the numbers except for Montevideo have really dropped off over a period of time.

So you can see amongst the major serotypes and a serotype particularly Montevideo which is the most prevalent amongst cattle isolates that we really don’t see a big change in multi-drug resistance. You will see a change in multi-drug resistance for some of the other serotypes, and again we’re talking about very small numbers.


Interestingly, though, even though there is a lot of concern about the, what goes on in the poultry area, Typhimunium and Typhimunium-5 minus really don’t have large numbers in the chicken arena. What we are taking a look at here is, is that by far S. Kentucky is the most prevalent serotype. It is in green here. And interestingly what we start to see is this change over time. S. Heidelberg was certainly the most prevalent serotype back here, and now we’re starting to see Heidelberg go down along with the multi-resistance associated with it. The S. Kentucky appears to be coming up.

Whether this is, since 2006, a reflection of what’s going on with the over sampling based on the risk-based sampling that FSIS has undertaken or not is something that we need to take a closer look at. And we’re looking at those clones, too. We hope to have some idea of what’s going on by the end of the year.


There are a series of these types of tables. This shows up in the Executive Report. What it does is it breaks down interesting resistance patterns that are of concern particularly in human medicine and you will see that for chickens, this Acc -- I don’t know what happened there. It just kind of went off scale -- the ACCSuT really hasn’t changed very much particularly amongst the chicken population and neither has the increased pattern where we’re looking at the MCAmp phenotype with augmentin and ceftiofur. So we really have not seen a lot of change there.


We’ve seen more of a change -- well, that behaved right -- more of a change with what’s happening from swine, and I’ll show you in the next couple of slides, even don’t we don’t seem to see or hear much going on with the Typhimunium and DT104, it seems to be increasing in the swine population, which I find to be somewhat interesting and something that we’ll be working with pork producers to try to figure out.
You’ll notice that while we see an increase in the ACCSuT we don’t see as big a jump in the AmpC phenotype.


Two serotypes of interest are S. Newport and you’ll notice that the data we have from animals it continues to decline over time. And the of isolates we’re receiving decreases, it does have the extended AmpC pattern, this is in contrast to two percent that we see as it contrasts to at least the 10 percent that they continue to see to see in the humans through 2008, really from Ezra’s last graph.


One of the things with DT104 is, is that as the numbers go down we have seen in 2009 this increase here in ACCSuT patterns. These are the confirmed DT104. One of the things that I’ll point out is that the humans only identify the ACCSuT, they do not identify how many are actually DT104 over time.

And so you have to temper this increase here with the decrease in total numbers. So what it appears is, is that the population within these numbers appear to be expressing more of the ACCSuT. Again whether this is something that is a phenomenon based on the risk based sampling or not is something that we’re exploring with FSIS.


Quite interesting, we’ve always seen it in cattle, cattle has been the main reservoir. We have seen it in turkeys in the blue and some in chickens, in past years, but what’s really noticeable for us is that since 2007 it’s really cropped up in the swine population as I mentioned in the previous slide. And that’s something that we’ll be taking a closer look at.


So, now, I’ll just finish up the last few slides with the crux of the whole system is, is that we’ve long been criticized or it’s been encouraged that we look at changing our system. This is something that requires more than just our will. We always want to have different things occur, there are a lot of reasons that you look at why you may change things over time, and there are collective reasons why it just doesn’t happen over night.

Some of the new things that will be happening, we’re in the process of redoing our CHRIS projects in the USDA and the NARMS is now going to be recognized with the separate CRIS project within the unit. That gives it heightened awareness not only within the surveillance arenas, but within ARS.

One of the areas that we’ve been mandated to address in addition to the FDA’s request and the Science Board’s request, isn’t also our own program staff request, to look at redesigning the sampling scheme.

So, in some ways we have to add or move sampling closer to the farm. We can look at sampling in plant prior to interventions as one way of accomplishing this.

One of the other things that we will add to the program because of the letters that have been sent through Congress most recently, means that we’ll be looking at adding MRSA and C. difficile to our repertoire. There is controversy over both of those. We recognize but that’s something that -- that’s the political will of the time.

We also will be looking at expanding USDA VetNet to the best of our ability. And until all of these changes take place we’re certainly going to continue testing so that we have a body of data that allows us to continue looking at trends over time, good, bad or ugly.


So, where is it best to test. A lot of people will say it’s best to test on the farm. There are sampling issues and there are monetary issues that have to be addressed when we go back there.

We can also look at trucks and we can look at --- and we can look at something prior to intervention in the plant before we get to the actual carcass itself.


So, these are the discussions that I would like to have. And of the things that -- it struck me as I have been sitting in the hotel room for probably the last two weeks in various places in the world is to -- it takes more than just myself and a few others to put how we’re going to have all this accomplished.

So I think that what I’m going to do is put -- not I think, I’m going to because I’ve already set it up this morning, is, is that by the break there is going to be a sheet of paper on the table in the back and I would like to collect any names who would like to be involved in these types of discussions, either by e-mail or conference call over time, just because I would like to hear everybody’s ideas. I would like to see if there is something that we haven’t thought about in how we can accomplish this.


One of the things is regardless whether we’re looking at in-plant prior to the interventions or on-farm, is who collects these samples and who has to provide the buy-in. I know my place in the food chain is one step above Sponge Bob, I’ve always said that, I’m well aware of that, and I’m really happy to be there, Patty and I have a relationship going on. And so there are people above me who have to make other decisions. And there are people within the industry who have to make decisions to make all of this work.

We understand how -- that we have to maintain confidentiality. Some of the questions that we have to address is, is that do we have the funding and other resources available. Do we have to do every commodity every year. Is there some other sampling scheme we can think of. NARMS does every five years. Maybe that’s too long. And how many isolates and that’s a discussion that we continue to have within AGISAR now.

And then is there some consideration we have to give for serotype differences, because if you talk about getting 170 isolates by serotype then you ramp your numbers up considerably when you’re looking for those. And cost becomes a real factor.

So, in order to meet this goal I would challenge particularly everybody from industry because that’s in fact where we’re going to have to get the samples from. At some place is, is that we have to have industry buy-in, in some way, shape or form and how do we get that.

We have to have a willingness for samples to be collected. We can provide all of the supplies and boxes and shipping labels and we can tell you we’re going to send everything back, but we still have to get them in the first place. So how best can we do this. Really we don’t have a need for farm I.D. for -- in order for us to maintain confidentiality, but then it talks about who maintains that and who keeps track of things over time. Well, we have some ideas about that and I would like to further discuss that.

The swine industry tends to be a very progressive industry. We’ve had a lot of discussions with them and we’re really encouraged about where we can go particularly within that commodity.

One of the other things is, is that we would very much like to include a short questionnaire with our sampling, whether it’s in-plant or whether it’s on-farm, to look at management inflow and the class of antimicrobial or disinfectants used in either place. And that would help address the well, what’s happening, can you associate resistance with what’s going on, with antimicrobial use.


And with that I’ll end. I’ll talk about Campylobacter and E. coli and Enterococci tomorrow a little bit. I just wanted to give you that quick overview on Salmonella and to introduce the fact that we’re serious about changing our sampling scheme and it’s going to take more than a couple of heads to get this done. Thank you very much.


DR. GREEN: Thank you very much, Paula. Our next speaker will be talking about surveillance of antimicrobial resistance in retail meats. And that is Doctor Shaohua Zhao. She’s a Senior Microbiologist at the Division of Animal and Food Microbiology in the Office of Research at FDA Center for Veterinary Medicine. And her primary responsibilities at CVM including coordinating NARMS with the Centers for Disease Control and Prevention, maintaining molecular genotype and databases, and advancing genotype technologies for foodborne pathogens.

I also wanted to point out that there is a list of biographies for the speakers and moderators out front so you can grab that during the break.