Blood Safety Transcripts

"Prioritizing Decisions in Transfusion Medicine: Transfusion Transmissible Diseases."

8:28 a.m.
Friday, January 24, 2002

Hyatt Regency Hotel on Capitol Hill
400 New Jersey Avenue, N.W.
Washington, D.C. 20001

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

Committee Members

• Larry Allen
• Mark Brecher, M.D., Chairman
• Celso Bianco, M.D.
• Richard Davey, M.D.
• Ronald Gilcher, M.D.
• Edward D. Gomperts, M.D.
• Paul F. Haas, Ph.D.
• W. Keith Hoots, M.D.
• Dana Kuhn, Ph.D.
• Jeanne Linden, M.D.
• Lola Lopes, Ph.D.
• Gargi Pahuja
• John Penner, M.D.
• Mark Skinner, J.D.
• John Walsh

Non-Voting Government Representatives

• Mary E. Chamberland, M.D.
• Jay Epstein, M.D.
• Colonel G. Michael Fitzpatrick

Consultants to the Committee

• Christopher Healey, J.D.
• Captain Lawrence McMurtry

C O N T E N T S

AGENDA ITEM PAGE

• Bacterial Contamination -- Mark Brecher, Chair 5
• Parasitic Contamination -- David Leiby 50
• Pathogen Reduction -- Stephen Wagner 102
• Break
• CJD Update -- Robert Rowher 132
• BPAC Update -- TBA 181
• Lunch
• Public Comments 196
• Committee Discussion 206

P R O C E E D I N G S

DR. BRECHER: Welcome to the second day of the Advisory Committee on Blood Safety and Availability. We have a couple of business items we wanted to take care of. One, I wanted to remind people that while we have presentations this morning, there will be a Committee discussion this afternoon, and we're discussing resolutions, and I don't think that we will take the entire afternoon. My guess is we'll be out by 4:00 at the latest, if not sooner, unless I hear any objections.

The second point is there was a question of us discussing smallpox and the ramifications for the blood supply. It was pointed out that we did not really have any speakers prepared to discuss this subject and that is a very important subject, and that really deserves the attention of this Committee, and it has been suggested to me that we postpone that discussion and actually reconvene the Committee on a one-day basis sometime in a couple of weeks for a one-day meeting to really address the issue of smallpox vaccination and its impact on the blood supply. And so if the Committee would agree to that, I would propose that that's what we do.

We can, perhaps, at the same time could also have one or two presentations on whatever happened to the HCV lookback and bring closure to that item as well. So all of those of the Committee who are in favor of putting off the smallpox discussion until a few weeks?

[Show of hands.]

DR. BRECHER: That motion carries.

CAPTAIN McMURTRY: I want to do one more housekeeping item just for the record. He's not here now, but I want the record to show that Larry Allen did come in yesterday. I had him as absent.

DR. BRECHER: This morning we're going to move away from viruses, which we basically concentrated on yesterday. I had said we would start on time, if not a few minutes earlier, and given the weather outside, I thought this would be an appropriate way to start here.

I stole this from Sunny Dzek, who presented it at an NHLBI Transfusion Medicine Hemostasis Clinical Network Steering Committee meeting last week, and I thought it was quite appropriate.

I think there's only one other person in the room who was there.

[Laughter.]

DR. BRECHER: We have to watch this closely here.

[Laughter.]

DR. BRECHER: If nothing else, it breaks the ice. Now, for those of you who weren't paying attention, watch his foot.

[Laughter.]

DR. BRECHER: Now, if that wasn't bad enough, this is called the evil penguin returns.

[Laughter.]

DR. BRECHER: So those two Sunny presented at the meeting, but I was really puzzled by all of this, and so I did a little searching on the web that night, and this apparently is the original footage from which the other two were digitally altered.

[Laughter.]

DR. BRECHER: So that second penguin was never really there. So seeing is not necessarily believing, and sometimes appearances are deceiving.

So let's talk about bacteria, and I fudged a lot bit. That's actually a Chagas organism flowing through the bloodstream there, but we're going to talk about bacteria, and I'm going to talk about red-cell contamination and then platelet contamination. I'm going to go zipping through the red-cell contamination because that's not really where the problem is.

But these are the kind of headlines we, as blood bankers, really don't like seeing, but it's a very important problem. You know, how safe is our blood supply? And for those of you who can't see how important this is, it even overshadowed the OJ Simpson story up here. So this is really important, and it began by talking about this man, Rollin Tobin, who was the public safety director for Southfield, Michigan, who was undergoing a total joint replacement, had donated three autologous units, but needed two allogeneic units, red cells, in the operating room.

To make a long story short, he became septic and died about 24 hours after his surgery from overwhelming Yersinia enterocolitica sepsis. It was a complicated story, and I'm not going to go too much into it, but this particular case went to a jury trial. Many times these cases lead to a lawsuit, but you never hear about them or you never can talk about them because they tend to be settled out of court. This particular one went to court, and the jury awarded the family $5.6 million in a wrongful death suit. So these can be very expensive lawsuits.

Now, it's not just in the U.S. that this happens. We think of the U.S. as being a very litigious country, but I took this one off the BBC website from about almost a year ago, where another red-cell contamination, where the family was awarded 300,000 pounds for brain damage that resulted from a patient becoming septic. So these cases can be quite expensive for a hospital if they do occur.

Now, with red-cell contamination, the two major organisms tend to be Yersinia enterocolitica and serratia liquifaciens. The reason is that these two enterobacteriaceae can grow quite well in the cold, in the refrigerator. They have endotoxin, and when you get a unit that is contaminated, generally the bottom falls out clinically--hypotension.

And then for Yersinia, this is a summary of 20 cases that were put together by the CDC, a couple major points. Sixty-percent of people who received Yersinia-contaminated unit died. So roughly half the people will die, and they tend to die within 24 hours.

If you have DIC, seven patients had DIC, including Rollin Tobin, six of these seven died. So if you're going to DIC, it's a very bad prognosis.

The incidence for red-cell contamination for Yersinia, there's some regional variability. The highest numbers have been reported from New Zealand, where they had an incidence of 1 in 65,000, and again roughly half of the cases resulted in a fatality. No one knows why there was such a high incidence in New Zealand back when this was reported. The numbers seems to have decreased in more recent years.

In the United States and Canada, it's estimated that the chance of dying from a red-cell contaminated unit is less than one in a million and may be on the order of one in nine million. Interestingly, I've been talking to Matt Ardvino at the CDC. There has not been a fatality from Yersinia in over a year. We're not quite sure why that is, but it may have, we've just been speculating, it may have something to do with the leukoreduction of red cells in this country that you were actually filtering out the Yersinia pre-source leukoreduction.

Several years ago we did experiments on red cell-contaminated units, and we found that over time the units turned darker, and you could actually see the dark color extending down the tubing until you get to the--it's a little hard to see in the slide, I'm afraid--but I'll show it to you better.

Here you see the dark blood coming from the bag, and then you get into the segmented tubings where they're not dark, and this is a sterile tubing up here. The reason they become dark is that as the organisms grow in the red cells, the oxygen drops to zero. So we have two units that have Yersinia growing in them here, as opposed to sterile units, where the PO2 in the units started around 40 sort of venous blood and very slowly creep up to the PO2 of room air.

We talk about these plastic bags breathing, but they breathe very slowly. In addition, there is some hemolysis of these units, and cell-free hemoglobin is darker than cellular hemoglobin. I know there have been some scattered reports suggesting there may be some methemoglobin formation inducing this as well.

Now, here's a particularly striking example. See the dark color of this unit compared to the segmented tubing here. Of course, when the PO2 drops, the hemoglobin becomes completely desaturated, and so it is darker.

This is actually a unit that was hung in a hospital. I got this picture from the CDC, and this is actually serratia liquifaciens. It gives you a very dramatic color change. We've grown this in our lab, and the units look just like this. In this case, with serratia, homolysis precedes a drop in the PO2. So the hemolyzed unit is bad news.

Now, the other interesting thing is, CDC, about a year ago, published seven case studies of serratia liquifaciens. Two of those cases discussed that once they figured out there was a contaminated red cell unit, they said whatever happened to the recipient of the platelet that was made from that whole blood donation? And so they looked back.

In one case, the patient had died of overwhelming sepsis, and of course the organism that grew was serratia liquifaciens from that patient, but they had not put together that it had come from the platelet, and this is a very common story. The other one was sick and had a bacteremia with serratia, but fortunately did survive.

So this is a very common thread with platelet reactions is that they tend not to be recognized, and it's sort of, through serendipity that you figure it out that it happened.

Now, some of this material was presented yesterday. I'll go through it, but this one approach to looking at the math. There are about four million platelet bags transfused in the U.S. every year; roughly, one million apheresis platelets, single-donor apheresis bags, and three million random donor or whole blood-derived platelet concentrates.

The contamination rate, based on a number of studies, principally using aerobic culture techniques, suggest that the rate is about one in a thousand to one in two thousand bags are bacterially contaminated. That means that in the U.S. we are transfusing 2,000 to 4,000 bacterially contaminated bags.

Now, the data in the literature suggest that maybe a quarter to a sixth will result in clinical sepsis of varying severity, but that works out to roughly 333 to 1,000 cases per year, of which perhaps a fifth to a third result in fatalities. So 67 to 333 deaths per year or a fatality rate of 1 in 60,000 units transfused to 1 in 120,000 transfused.

Now, are these numbers real? Data from hospitals that have closely looked for these reactions, such as Johns Hopkins, which has had a long interest in bacterial contamination platelets, Paul Ness published a paper in the past year where he reported that the risk of dying from a pool of random platelets 1 in 17,000 in his institution, out of a pool with six. So you'd have to multiply it times six to get the unit number. But for an apheresis pack, it was 1 in 61,000. So we're right in this range.

And, of course, he only knows about the cases that were reported back to him. So he may have missed a few cases. So I think that this estimate is a reasonably good estimate for what's happening in this country.

Similar numbers have been reported from the University Hospital of Cleveland, another institution that has really concentrated on bacterial contamination of blood products. So there's a heightened recognition of these reactions in those two institutions.

Now, I think this is a third or maybe the fourth time we've seen a figure like this. We've done a great job, both industry and in the blood centers, in reducing the risk of HIV, HBV, and HCV over the years per unit. Actually, this was in the handout of the article that was in the Lancet last week.

But for the risk of bacterial contamination per bag, we have not done as good a job. The risk has remained approximately 1 in 1,000 to 1 in 2,000 for years. Now, the one thing we have done is there has been a push in this country to switch to apheresis platelets, and I think roughly two-thirds of doses being handed out in this country now are apheresis platelets, but the risk per bag has not really changed.

Now, there are a variety of organisms that contaminate platelets, unlike what you see with red cells. The difference here is that platelets have to be stored at room temperature, and so they are a good growth media. If you were to just look at the organisms that grow from platelet bags, about two-thirds of them tend to be gram-positive organisms, like Staph epi and Bacillus cereus, so skin saprophytes.

However, if you look at what actually kills people in the U.S., and this is data from the FDA, roughly, 23 years, 51 fatalities reported to the FDA, it tends to be the gram negatives that kill people more than the gram positives, like Klebsiella, Proteus mirabilis, E.coli, Enterobacter, Pseudomonas Salmonella and Serratia. Most of these are enterobacteriaceae, except for Pseudomonas here. So it's the gram negatives or the organisms that we need to worry the most about.

Unfortunately, the patients who tend to receive platelets often are immunosuppressed. More than half of all platelets that are transfused in this country tend to go to heme oncology patients or bone marrow transplant patients, and they are not in a good position to fight off bacteria.

The gram positives, probably many people who are immunocompetent might be able to handle gram negatives. I think if you get a unit that's heavily contaminated with gram negatives, it doesn't matter whether you're immunosuppressive or not, you are probably going to die.

A variety of strategies have been suggested to address the problem of bacteria contamination. Growth inhibition, the question has been raised about what if we just put some antibiotics in every bag of platelets? It probably would solve the problem, but it would present new problems, and there's been a great reluctance to, one, trade a relatively rare reaction, fatality from bacterially contaminated platelet, for idiosyncratic drug reaction.

And, two, we would be spreading a little bit antibiotics all over the hospital, and we would be selecting for antibiotic-resistant bacteria, and so this has not really been considered a real possibility.

Temperature, a lot of research has gone into trying to refrigerate platelets, but to date this has not worked. Bacterial avoidance, let's try to keep the bacteria out of the bag. I'll come back and talk about that in just a bit. Bacterial detection we'll talk about, and I'll just briefly mention elimination, but we're going to have a whole talk on pathogen reduction later this morning so I'll stay away from that.

What about just trying to get a safer bag of platelets, less chance that there will be bacteria in the bag. Bacteria comes from two sources. It comes from the skin or it comes from a unit who has, it's come from a donor who is an asymptomatic bacteremia. Most of the gram negatives are from donors of asymptomatic bacteremias.

Recognizing this, the group at Johns Hopkins made a conscious decision back in 1986 to try to move toward an all apheresis blood supply. So, in 1986, roughly, 52 percent of their platelets were apheresis platelets, and 48 percent were pools of random donor platelets, so a six-pack.

By 1998, they had gotten to the point where 99.4 percent of all platelet doses being transfused at Johns Hopkins were apheresis platelets. During this time, their reaction rate, where people who actually had clinical signs and symptoms from these transfusions related to bacteria went from roughly 1 in 5,000 transfusions to 1 in 15,000 transfusions.

The difference in risk between these two products was between five to sixfold, as you would expect from a six-pack, compared to one bag. And so there were many institutions that have used this strategy. In fact, to be honest, when a patient needs a platelet transfusion, and they receive an informed consent for transfusion, how informed is it really?

Do you really say, well, we can give you a six-pack of random donor platelets for X number of dollars that's going to cost the hospital, it's cheaper, but it has six times the risk of bacteria or we can give you a single-donor apheresis platelet that's going to cost the hospital a few dollars more, but has one-sixth the risk of bacteria. That question never gets asked.

There have been some data from Europe that have advocated the diversion of the initial couple mLs of collection. It's thought that this is the most contaminated from the skin and that there may be a skin plug that comes off the needle. There have been some large studies from Europe that support this.

This study from France, they looked at, roughly, 3,400 collections, where they looked at the first 15 mLs of blood, where 76 units were contaminated, and then they looked at the second 15 mLs, where only 21 were contaminated. So it wouldn't take care of the entire problem, but it would cut down on some of the skin contaminants coming into the bag.

Similarly, a larger study from the Netherlands, 18,000 collections, .35 percent were contaminated. They diverted the first 10 mLs, and the contamination rate dropped to .21 percent. There is some move in this country to go in this direction. It has been discussed at the BPAC. It was not thought that the data was strong enough to warrant that this be a requirement, but people are moving in this direction. It's a small step or, as my colleague, Ros Yom Tovian, likes to say, we shouldn't be diverted by diversion. It will do something, but it's not a very good solution.

What about skin disinfection? This is a study from Canada, from Indie Goldman. It's sort of a busy slide. Basically, they did touch preps on the antecubital fossa after the skin had been prepared by a variety of skin disinfectants, and they looked at how many colonies were present afterwards.

The message here is that when we cleanse the skin, we do not make the skin sterile. All we are really doing is reducing the bacterial load on the skin. There's a suggestion in this paper and in another study from London that the use of tincture of iodine is superior to the povidone iodine, which tends to be the standard of care in this country.

Interestingly, another thing that came out of this study was they looked at green soap and isopropyl alcohol. Green soap was commonly used in this country to prepare the antecubital fossa of donors who were allergic to iodine. In one-third of the cases in this study, there were more bacteria after the green soap than before the green soap. All they did was stir things up, and so there has been a move, particularly from the AABB, to say green soap is not an acceptable alternative.

There is another skin prep, chlorhexidine, which does a reasonably nice job, and actually the new version of the AABB standards that will be coming out this year will specifically address this, and green soap was dropped from the AABB technical manual in the last addition.

When you're talking about platelet contamination, timing is a major issue. Data from Mo Blajchman from Canada, they looked at 16,000 random platelets on the day of collection, which they called Day One. A lot of people would have called that Day Zero, but we'll let him get away with it, where they found four positives, a contamination rate of .02 percent.

They came back two days later. There were still 10,000 of these bags in their inventory. Now, they had a culture-positive rate of 7. So the rate went up to .07 percent, so it more than tripled. The message here is that if you want to find the bacteria, a sample from the bag, right after you've collected it, is not going to detect most of the cases. You have to allow some time for the bacteria to proliferate so that a random small sample from the bag will have bacteria in it that you can actually identify.

These are growth curves for several of the most common, contaminating organisms, and clinically significant organisms from my lab at UNC, where we spiked in at low concentrations on Day Zero, usually 10 to 50 CFUs per mL, colony forming units bugs per mL. What you can see from these drawings, and this is Bacillus, Pseudomonas, Klebsiella and Serratia, is that usually by Day One or Two, you have significant amounts of bacteria-- these are log curves--and that generally you reach a plateau by Day Three or Four.

With Bacillus, often you're on plateau growth by one day, 24 hours, and probably a Day Three- or Four-day-old-platelet and is probably no more dangerous than a Day Six or Seven platelet, in terms of bacteria. If they're going to grow, these things tend to grow early.

Now, there is one organism that I don't think you can say that about, and that's Staph epi. That tends to be a slow grower. There's some data to suggest that the initial concentration has an effect on the lag phase and that Staph epi may take quite a while to grow.

This is just to remind me to emphasize that you can't test the whole unit. I guess I have to change the slide again because we're coming to the end of January 2003. You can see it's yellow, it's a platelet.

Detection techniques. A variety of both high-tech and low-tech approaches have been described in the literature. I've been fortunate, many of these technologies have played through my laboratory over the years, and I'm just going to concentrate on a couple of the technologies that have actually made it out into the world, that have actually been used.

One is bacterial staining. It doesn't matter whether you use a gram stain or a Wright stain. Some people prefer a Wright stain because the labs have these automated Wright stainers in the hematology lab. You put a little slide on a little conveyor belt, and it goes and out comes a perfectly stained slide at the other end, and we don't care whether it's a gram negative or a gram positive. We just want to see that there's bacteria there.

The bottom line is that with a gram or a Wright stain, you pick up at around 10 to the 6, maybe 10 to the 7 CFUs per mL, so it's not particularly sensitive. You could do acridine orange, where you can make the bacteria, the DNA, glow back at you, but it requires a fluorescent microscope, and it only gets you about one log better, so not all that great, but it has been used sporadically, particularly at the University Hospital of Cleveland, where they were able to interdict several bacterially contaminated--heavily contaminated--units.

Some people have advocated using dipsticks, urine dipsticks, looking at the drop in pH, where it would turn orange, or the drop in glucose. As the bacteria grow, they consume the glucose in the bags, and this is a slide from one of my papers on transfusion. Here, we've got Klebs pneumoniae and Staph aureus, and you can see the dipstick is blue, blue here from glucose, or if you can't see it, use your imagination, and the pH is orange.

Now, both of these units were Day Three after contamination, contaminated on Day Zero, and the organisms were at 10 to the 7 CFUs per mL. However, we would have missed this one, Serratia, where the glucose was not dropped sufficiently to turn it blue, and the pH was not orange because it was only at 10 to the 3rd CFUs per mL. So sort of across-the-board dipsticks pick up about 10 to the 7 CFUs per mL, about comparable to a gram stain. Now, they're somewhat easier to use, and they're relative cheap, pennies a dipstick.

There's a recent report from M.D. Anderson, where they screened 3,000 random platelets, and they found two that were contaminated with bacillus cereus using this dipstick technology, and they were able to interdict those units.

Now, to do that, they also found 28 units that did not pass the dipstick. You may say, well, it's specificity isn't very good, but you could come back and say, well, if the pH was really 6.5 or there was no glucose left in the bag, maybe those platelets weren't any good any way.

Most of the interest has concentrated in recent months on the two bacterial culture methods that have recently been approved by the FDA.

One is the bioMerieux BacT/ALERT microbial detection system. This is an automated liquid detection system, where you put your sample and question into one of these bottles, load them onto these machines, and over time, as the bacteria grow, they generate CO2, and the CO2 causes a color change in this colorimetric sensor at the bottom of the bottle.

There's a very similar system out by Becton Dickinson called the BacTec system, but it is not approved for platelets in this country, where the color change goes from green to yellow, and roughly every 10 minutes every bottle is scanned. There's a little light that reflects off here, and it goes to a sensor. There's a computer hooked to the system where it looks at both the absolute color, but also what is the rate of change of the color so that it picks up these units before the human eye would pick them up.

In my lab, we've looked at 14 or so different organisms in platelets with these machines using all of the types of bottles that they have. For the vast majority of them, we would have picked, at 10 CFUs per mL or lower--some of our units were at 1 CFU per mL--we would have picked up most of these bacteria, roughly, at 12 to 14 hours. In some cases, it would have taken 24 hours, with the notable exception of propionibacteria acnes, which is an anaerobic organism, which actually takes days, but the clinical significance of propionibacteria acnes is not clear and probably has little clinical significance.

The other system that was recently approved, and both of these systems, I should note, are approved for in-run quality control. They are not release control, so you cannot make a claim of sterility if you use these two techniques.

There's a system from Pall, their Bacterial Detection System or BDS, and what happens here is you push over about 6 mLs of your platelet solution in question, and it goes through a filter. It wouldn't be a Pall product unless it had a filter in it. I know the Pall guys are here.

The filter filters out white cells and platelets, but lets bacteria pass. And sort of across the board, about 50 percent of bacteria cross this filter, and so then you have 2 mLs that make it into this little side pouch, and there's a little SPS tablet in here which inhibits the inhibitors of bacterial growth, and so it allows bacteria to grow better in this little bag.

You clip off this little bag, you put it in a 35-degree incubator for 24 to 30 hours, and then you measure the PO2 of the headspace gas of this bag to see if it's below a cutoff. As the bacteria grow, they will consume oxygen, similar to the diagrams I showed you for Yersinia and red cells before.

Now, I'm not going to say much about pathogen activation--I'll leave that for Steve Wagner--other than to say Cerus, which is the one that has the most studies out there, is very good at inactivating in five logs a variety of bacteria, although it may have some trouble with spore-forming organisms. If it forms a spore, the chemicals may not be able to get into the spore.

Now, here's that crystal ball somebody was talking about yesterday. What's the future? The future tends to lie in the past, and referring back to our IOM report which we went over yesterday, and I think Lola mentioned this, Recommendation 6, this is my favorite one in the report: The perfect should not be the enemy of the good. Implementation of partial solutions that have little risk of causing harm should be encouraged.

I think that's where we are with bacteria. We don't a perfect solution right now, but we have a lot of partial solutions that would take us most of the way.

The FDA has sponsored three workshops that dealt with bacteria over the last seven years. This is a summary, summary comments made by Ed Snyder in the '99 meeting. I have Ed's permission to use this.

He concluded that the imperative is to act, so you don't have to explain yourself on Nightline and regulation is necessary to achieve the goal. Nothing says I care like a page of 483s--483s are the citation forms left from the FDA. So if you get dinged, you do something about it.

When all else fails, do something. Give us a mandate, and we'll do the rest. I think that's sort of where we're at. Unless blood banks are told to do something by some higher organism, be it FDA, AABB, CAP or some organism, the hospital administrators are not going to let us do this.

Following the meeting last August on pathogen reduction, five of the speakers and moderators from the meeting got together, and we wrote this letter which was published as an open letter to the blood banking community. In this letter, we basically said:

"It's our feeling that pathogen reduction won't be here in the short term. Nevertheless, bacteria contamination of platelets is a major problem in blood banking, and we feel that the implementation of detection strategies should be implemented now."

This got a lot of coverage. It was put up on a lot of electronic sites going out in the blood bank newspapers. Interesting results. In fact, there was one comment posted on the California Blood Bank Society page that accused us of moral blackmail, and I think it's only moral blackmail if we're moral.

[Laughter.]

DR. BRECHER: Now, AABB has been talking about this, and they did publish for comment their proposed 20-second AABB standard, which was the new standard 5.1.5.1, that the Blood Bank and Transfusion Service shall have a method to test for bacteria contamination of all platelet components.

I have to say that I'm aware that the wording has been changed. The final wording is not approved yet, and so it's not clear exactly what will happen with the AABB, but this may be the mandate for bacterial testing of platelets that the country needs.

The other thing is we need to recognize that in '82 the platelets were extended from three to five days; in '83, from five to seven days, because of good data function and survival of these platelets. But in '86, because of reports of bacterially contaminated Day Six and Day Seven platelets, it was rolled back to five days. There is now a lot of interest in going back to seven-day platelets. If anything, the platelets are better now than they were back in 1986. The plastics are better, the white cells are out of the bag, the platelets survive better.

There have been a variety of reports from Europe that have coupled culture to the extension of the platelets in Europe from Denmark, the Netherlands, Yugoslavia, United Kingdom. In fact, several countries, such as Norway and Sweden, are routinely culturing all of their platelets.

Several institutions in other countries, including HemaQuebec, but in this country, Dartmouth, the University of North Carolina have been culturing their platelets.

Finally, we heard from Jim AuBuchon yesterday, and Jim is probably the king of cost-effectiveness studies in this country. He's published a lot of major cost-effectiveness studies, and it tends to be that every time he says something is not cost-effective, p24 testing, NAT or whatever, he says it's not cost-effective, we go ahead and do it anyway; is that a fair statement?

He hasn't published a paper, but he had an abstract back in '99, and basically he said, hey, this is a cost-effective strategy. If we culture and get an extra two days of outdates on our platelets, it will easily pay for all of the culturing in this country, and so this could be a win-win for everybody--safer blood at less cost, as unbelievable as that may sound.

Also, there have been data from other countries that also support this stuff. They actually save money by culturing and extending an extra two days on the shelf life.

Finally, I want to end with this quote, which is actually how we ended this Lancet paper about looking to the future that was published last week and was distributed to the committee members yesterday.

"Actions are right to the degree that they tend to promote the greatest good for the greatest number. By John Stuart Mill."

Now I can take some questions.

Celso?

DR. BIANCO: Mark, thanks for a very nice review. I have two questions for you. The first one, if we went the other way around and looked at what contribution, that is, probably a lot of things will be implemented in the next year or so. It's the better skin prep, more attention to the skin prep, the diversion bag, and the culture. How much each one of those procedures will contribute to reduce the risk of bacterial contamination--ballpark? I know it's difficult.

DR. BRECHER: I think the skin prep and the diversion have the potential to cut down on the gram-positive contamination, but I don't think it's going to be more than half of the gram positives.

What worries me more in that two-thirds of the debts are from gram negatives, and neither diversion or the skin prep are going to impact on the gram-negative organisms. So the only way we're going to get to where the real fatalities are is with a detection system or with an inactivation system.

DR. BIANCO: The second issue is surprisingly both systems that we have now for bacterial detection were approved for quality control, not for unit release. That, in the short term, certainly helps in the sense of making those systems available on a wide scale, but in the long term, and I'm giving my personal opinion, I think that the approach was detrimental because it doesn't allow us to really say that the unit has a lower bacterial load because the culture was negative or something like that.

It interferes with our extension of the seven-day platelet and also discourages the manufacturers from pursuing full clinical trials and licensure of their tests in a format that would be really compatible with life.

I wonder if you could give us your opinion about that and what kind of quality control we could do, in a limited number of units, that could reflect the overall process and give us more certainty or more--

DR. BRECHER: The problem, as I see it, what we would require for release control would be a study of the magnitude of the NAT testing implementation. It would be an IND that would have to go across the country because the incidence is so low, and so it's almost impossible to prove true clinical efficacy on a large scale that I think that the FDA might want. I don't want to speak for the FDA.

Jay, do you want to say something about that?

DR. EPSTEIN: Yes. The problem, in a nutshell, is that to validate the actual benefit of the up-front test, you need a follow-up culture at either the time of issue or outdate on the very unit that you cultured either on Day One or Day Two in order to validate the sensitivity of that procedure, and the companies have not stepped forward to do that study, nor has it been done otherwise in an investigator-sponsored study.

So we're left not knowing what the up-front culture really does. We know it sometimes detects contamination, and if it's done "in real time," which means while you sell the unit in inventory, of course, there's an obvious benefit pulling that unit, but we really just don't now. Was that 50 percent of the units, 75 percent of the units?

So, in order to be able to make a statement that the up-front culture, done in a certain way, has a certain predictive value for a culture-negative product at issue, you have to do a more comprehensive study. Now, there are two barriers; one is the added cost of the follow-up culture, and the other is the need for large numbers.

The problem of large numbers is, in fact, easily remedied if any large part of the system does move to a routine quality control culture. You're going to be doing it. The question then is how do you fund the endpoint culture to validate the up-front culture.

I agree that it makes perfect sense to link this to extension of dating. The bug bear there is that we need, just as the platelet bags, oxygenation, et cetera, have improved, we need the validation data that we still have a quality platelet by today's standards. In other words, what might have been accepted in '86 for a five-day platelet may no longer be acceptable today, and we want to know that the seven-day platelet is, in fact, acceptable.

What we do know is that the platelets are deteriorating progressively on storage, and so it's a bit of a question of what's the belt line. But these issues have to be put together. Well, there are one of two ways to get the correct study funded, either to have antecedent data that you can extend data and therefore funded by extension of dating or to have supplemental funding so that you can study them at the same time.

I mean, neither of those approaches is infeasible, but the parties that need to step forward have not done so. So the Agency, you see, was left with a very difficult problem. We had culture systems that were validated to sometimes detect bacteria. Their actual clinical value was never measured, and the conditions of use were therefore not validated in any predictive way. So the most they could say is that, you know, if you do this, you'll sometimes detect bacteria.

Well, if you approve them for quality control, you eliminate the issue of whether it's being done real time. A quality control culture to monitor the frequency of contamination to ensure that you have a system operating under control with the expectation that the contamination rate is of the order of 1,000 to 2,000, as is the current state-of-the-art, is feasible whether you're doing an online culture or an offline culture. You could do quality control simply by sampling at issue or outday. Indeed, you could do it solely at outday, and the time taken to do to the culture and report the culture does not have to be commensurate with the shelf life of the product.

So the standard for approving it for quality control was a lot lower than a logical standard to approve it as a release test; in other words, a test with a certain predictive value of a nonculturable unit at issue.

We brought this whole matter to the Advisory Committee. There will be a summary later today of the discussion at BPAC, but the Advisory Committee agreed with the FDA that what we had described as the proper design of a study to validate a claim for a release test is scientifically appropriate.

Now, of course, we have an open mind, if people have a better idea. But the problem, as I see it, is that there has not been the will, you know, expressed through a funding mechanism, to simply do the appropriate study, but it's nothing arcane. It just requires a follow-up culture.

DR. BRECHER: Let me just comment on that. There have been two centers that have been running pilot studies like that, but they're small. Dartmouth and UNC have done this on a small scale. Dartmouth has done about 4,000 apheresis platelets, and UNC has done 2,500, roughly, to date.

At least we can say that the contamination rate that we see on the early cultures are similar to what's been reported in the literature. At least at UNC we have not seen any that we didn't pick up on a Day Two culture, and we've, actually, in 2,500, we stopped three Staph epi units from being transfused.

That said, when you do the statistics, and I've had several statisticians look at this and none of them seem to agree, but the message seems to be that you need about 120,000 platelets to validate this, to show that it really is statistical.

If you're only going to do the outdated platelets, Day Seven, and roughly 10 percent of platelets outdate in this country, we're talking about basically enrolling about a million platelets into the study and only picking up the Day Sevens.

Now, there may be some give, and maybe we can do issue platelets, instead of just Day Seven, but it is a large study, and I actually have had some discussions with the Red Cross, and they are interested in, at least initially, they said that they would be interested in reculturing their outdated platelets, and so maybe within a couple of years we can generate this data.

Is that safe enough?

DR. EPSTEIN: Well, I'm not going to comment on the study designs and the data submitted. I can't do that publicly.

DR. BRECHER: Yes, well, this data is not submitted.

DR. EPSTEIN: All I can say is that what the Agency has seen to date is not sufficient and that there is a problem of numbers, but as I say, if the system moves toward 100-percent quality control testing, you have the infrastructure to very rapidly do a large study if there's a way to fund the follow-up culture.

DR. BRECHER: Celso?

DR. BIANCO: My problem, and I think that it's very much the purview of this committee, instead of BPAC or the FDA, is that the fact that those companies were licensed or had their systems approved for quality control, those companies have no encouragement to go in and invest into the next step. So my question to Mark and to the Committee is how can we encourage them? Because I think that our goal is to have 100-percent bacterial detection.

DR. BRECHER: One strategy we thought was going to be a good one was to take us to the new transfusion medicine, hemostasis clinical network, but when we ran the numbers the cost just exceeded the money that was available in that network, and so that project was put on hold. I don't know where the money is going to come from, unless the blood centers just agree to do it because, in the long run, it will benefit them.

DR. DAVEY: Mark, I think we can all agree that moving to single-donor platelets is a good idea from whole blood-derived platelets, but that's going to be hard to do, at least in the near term.

What do you think about the proposal that's been floated to pool whole blood-derived platelets at the blood center and then test the pool, similar to what's been done in Europe, to some extent.

And perhaps a question for Jay, what would be needed to approve such a technology?

DR. BRECHER: I think logically it makes a lot of sense. Europe has been doing this for decades. My guess is that it would require an IND from the FDA, but I think it's feasible. I think we should move ahead. Someone should do a big study like that.

Keith?

DR. HOOTS: In terms of the sensitivity, I mean, compared to culture, some of these techniques, are they in Day One or Day Two, how many bacteria per mL are they capable of detecting, did you say?

DR. BRECHER: Well, there are papers out on the bioMerieux system that suggests that they can detect down to one or even less CFUs per mL. So let's just call it at least one CFU per mL. The abstract presentations on the Pall BDS suggest that they pick up down to 10 to 100 CFUs per mL.

DR. HOOTS: Just in terms of at the far end, since you made a great point about incremental things until you can solve the whole problem, in terms of practice, it's been my experience that a lot of times, and probably it applies most to immunosuppressed patients, BMTs, and leukemias and that sort of thing, that when blood products are ordered and are hanging or getting ready to be hung, that they will just come to the floor, and they'll get hung.

If the patient happens to be on antibiotics, say, because they have leukemia, and they've had neutropenia, they'll delay the antibiotics to get the blood hung, particularly if they only have limited access. I just wonder if just something as simple as saying, you know, make sure that if they're on antibiotics, they get their antibiotics, and then hang the blood or the platelets, particularly the platelets, if they happen to have a small contaminated unit, as opposed to something that you might see causing a fatal transfusion bacterial contamination. It might buy them some time.

DR. BRECHER: It might, but, clearly, there have been fatalities in people who were receiving around-the-clock antibiotics, and I think it's going to depend on the organism, the sensitivity of that antibiotic, and if you're getting a big flush of endotoxin, it doesn't matter what antibiotic you're on.

John?

DR. PENNER: I think this is the body that should be recommending a program of study and funding for it. If it's apparent at this point that we're running into a question of having something done, and if it needs to be accomplished, we need to get behind it with some form of resolution, and I don't see why this can't be done, even today to provide at least some prodding for proceeding in that direction.

Incidently, the unit of blood to protect Michigan came from Wisconsin, for the Detroit case, and it was a dairy farmer who was a carrier for Yersinia.

DR. BRECHER: I tend to agree with that. I think this is the right body to make a recommendation. And one of the things, when we went back through the grid, that I think we've been successful is identifying areas that needed to be funded and that have subsequently been funded.

Now, I'm in a difficult position, being Chair of this Committee and having a real interest in this bacteria, so I'm going to try to step back and just let the Committee members decide what they want to do with this particular issue.

I think we need to go on to the next speaker. We've run over a little bit, but now we're going to run to parasite contamination, David Leiby.

In case you were interested, that little cartoon that I had running at the beginning of the talk was actually a Chagas organism in blood, so I fudged a little bit.

[Pause.]

DR. LEIBY: We're going to go ahead and move forward.

I've been asked to come here and talk to you about parasitic contamination. In fact, if one wants to think about the parasites that are possibly transfused by the blood supply, this is pretty much the short or long list, however you want to look at it.

There's a group of parasites that are highlighted here in white, and we aren't going to talk about those today because I'll deem those as being less important threats to the blood supply. The only one I'd qualify is perhaps Leishmaniasis. If you recall, about 10 years ago, we had some concerns about Leishmania after Operation Desert Storm. Seeing that we are now back in the same part of the world, this may once again rear its head.

But I will talk today about these three or four organisms, and I'll briefly just call them by their disease names. I'll start with malaria, then move on to Chagas, talk a little bit about ehrlichiosis, in particular, human granulocytic ehrlichiosis, and then lastly I'll finish up with babesiosis.

First of all, I'll talk about malaria, and there's actually four agents listed here that are actually the etiologic agents of malaria, human malaria; those being Plasmodium falciparum, P. vivax, P. malariae, and P. Ovale, and vivax and falciparum are by far the ones of greatest concern.

As you can see, they're intracellular pathogens of red cells, so they have convenient vehicles for being transmitted by blood. They are mosquito-borne, generally by Anophelene mosquitoes, and primarily they're limited to the tropics throughout the world, and they cause what are generally characterized as flu-like symptoms, but then they have some periodicity, meaning that these symptoms reoccur every two, three or four days, depending upon the organism.

As I say here, it varies by the infecting species, and that has to do with the periods at which these parasites, and they are synchronized, and they break out of the red cells. At that time, humans have reactions to the parasites in many of the products which they release.

Malaria certainly causes morbidity and mortality throughout the world and still is one of the number one killers of children.

What about transfusion transmission, particularly here in the United States? Well, our present prevention strategy relies solely on travel history. People are asked questions about where they've been, and if they've been to what is considered a malaria-endemic part of the world. They are deferred from blood donation for a certain period of time.

There are no screening tests available at this time. It's not something I think that's actively considered, but I'll maybe suggest that we should.

In the U.S., there are approximately one to two cases, transfusion-transmitted cases, of malaria per year. And generally these fall into two categories; the first one being an asymptomatic carrier, and that's going to be a common theme throughout all of these parasitic infections I'll talk about, is that the asymptomatic carriers are the ones we need to be most concerned about.

The others are semi-immune carriers or those in a semi-immune state; people who were infected long periods ago and appear to be, by all intents and purposes, clear of the infection or partially immune, but can reacquire the infection or, in some cases, they have relapses of infections they've never lost, and some of these infections have been measured out 40 years after their initial appearance.

Now, when we talk about malaria and how it gets into the United States, generally, we think about international travel--individuals from here going international and coming back with malaria--or, in the case of people who have lived their lives in malaria-endemic countries coming here and living in the United States and bringing the infection with them.

As I mentioned before, our prevention strategy has largely depended upon travel history. However, in many cases, we are probably unnecessarily deferring blood donors because they got off the cruise ship in Cozumel or somewhere else for a brief shopping excursion, were never exposed to the parasite, were never there in the evening or the morning, when the parasite actually feeds, the mosquito actually feeds, I should say, and so we're actually deferring a lot of donors.

So, if anyone wants to consider a test for malaria, its greatest benefit may be, in fact, increasing the donor pool, as opposed to preventing transmission. As you see, as I said before, there's only been two transmissions.

The last one I'll talk about is something that we need to consider, what about endemic foci in the United States? These were a series of headlines that came out of the Washington Post just last year.

As you're aware, there is a malaria outbreak not too far from here, just up the Potomac, and what it turned out to be was that there were two teenage boys in Loudon County who were infected with malaria. That was Plasmodium vivax, I believe. They did not live near one another, at least they were far enough apart that the source was not the same, and as they went farther, they found that there was actually some infected mosquitoes on some of the islands in the Potomac.

As I understand it, the thinking is that there are some actually sod farms on the islands in the Potomac, in which there were some immigrants from parts of Latin America who work there, and perhaps they were the source of the malaria infection, but whatever it was, they got into the local mosquitoes. So down the road we need to consider whether or not malaria may once again become endemic in parts of the United States.

Let me shift gears to something a little different, something that probably poses a much greater concern to the blood supply, and that's Trypanosoma cruzi.

Here, you can see it's a very small protozoan parasite. This is actually the extracellular stage. You can see it's about the size of a red blood cell. It's not in a ring. It's just curled up there. It actually has a tail that goes there, you can see. Most importantly, it causes a chronic, asymptomatic, and perhaps most importantly, untreatable infection.

It's endemic to portions of Mexico, Central America and South America, and transmission primarily of our concern is by four routes, and I'll go over each one of these, briefly--vectoral, by an insect; congenital, from an infected mother to a child in her womb; via organ transplant; and, lastly, by blood transfusion.

Primarily, as I said, in a natural state, it's transmitted by the bug, and this is any one of a number of reduviid bugs that contains the parasite. Most interesting, unlike some of the mosquito-born agents that are transmitted by the front end of the bug, this one is transmitted by the back end of the bug.

So, actually, when the bug feeds and does take a blood meal, it defecates in the process and the parasites are in the feces of the bug. If that feces is rubbed into the bite wound or into other mucosal surfaces, like the eye of this young girl, the parasite can enter the human host. That's what is commonly called a chagoma. It doesn't always occur, but it's a swelling at the site of the infection.

Here, again, we see the parasite in the blood, and eventually probably the most important area where the parasite ends up is in cardiac tissue. It causes cardiomyopathy and other complications in the heart, which will many times not be obvious for 20 or 30 years, but down the road can lead to serious complications and death.

Why is this parasite, if it's endemic to Latin America, of concern here? Well, quite simply, it's due to immigration, changing demographs. These are some statistics out of the 2000 Census that show there's about 12 million immigrants from Latin America. These are legal immigrants. Certainly, there are many more residents than that in this country, many who are also blood donors.

So we have a large population moving into the U.S. This is also 2000 Census data showing the great rise between 1990 and 2000 in the levels of Hispanic population. And if you read the headlines in the last couple days, I think, as of July, the Hispanic community is now the largest minority group in this country.

What about congenital transmission? We speak so much about the immigration population, but one thing I want to stress is we don't need to worry about that first generation of immigrants of Chagas. We also need to be concerned about the second generation and the third generation because it does seem to pass down, in some cases, through the families.

And this is out of a study we did in Texas. We identified a donor in Waco, Texas, down here in the lower left, who was a 17-year-old boy who had Chagas disease or antibodies of Chagas. He was certainly infected.

What was interesting as we talked to this boy and his family, we found out that the older brother, who was also born in Texas, was also medicated for arrhythmias, and arrhythmias are one of the common characteristics of Chagas disease.

The mother, also born in Texas--see, all of these were born in Texas. Please take note of that. They weren't born in Latin America--was also medicated for arrhythmias, and her grandmother, who gave us most of this information, also was born in Texas a couple of generations back, with a history of heart ailments. She lost her brother, who died at 55 of an enlarged heart. Once again, another common characteristic of individual Chagas.

And what we think is happening here is this was all traced back to the great-grandmother, who actually immigrated from Monterrey, Mexico, who also died at a very early age of an enlarged heart.

Now, we tried to get these individuals in to test the whole family, to show that this is really an actual occurrence, and initially they agreed, and they became a little skittish in the end. I am not sure why. But we really suspect that this is probably passed down the maternal line, and there is literature out there that will support this contention. So this isn't just pie-in-the-sky.

Back last fall there was a great deal of concern about West Nile Virus, and I'm sure it hasn't gone away, and one of the things that really set this all off was the stories of West Nile being transmitted by organ donors. Well, before that, earlier in 2002, there was actually a report of a case of Chagas disease, which was transmitted by organ transplantation as well.

What happened was there was a single cadaver donor who actually the organs were split up among three recipients; one received a kidney and pancreas, another one received a liver, and actually the third one received the other kidney, and all three individuals, all three recipients came down with Chagas disease. In fact, one of them died of acute Chagasic myocarditis.

This is actually a blood smear from one of the recipients showing four trypanosomes in one field. That is quite unique to see that many parasites in a single field.

What was interesting anecdotal information I was told, when this donor recipient, when they pulled the organs, they also pulled the heart. When they looked at the heart, they found the heart to be riddled with I would say an extent of pathology that deemed it not worthy of being transplanted, and so it was just curious enough that obviously it was probably the damage from the parasite that made the heart not useful.

What about transfusion cases? There has been seven cases in the U.S. and Canada since 1987. These are seven cases that have been recognized, and that's an important point I'll talk about later. The most recent one, some you may not know about, occurred last year in Rhode Island.

One thing or two things I want you to notice in this that transfusion cases are not limited to people living in Texas border towns, Miami or in Los Angeles. Now, there are cases in California, Houston and Miami. You'll also notice one in New York City, two in Manitoba of all places, and one in Rhode Island. So they do appear in Northern climes, suggesting that there are individuals there who are infected and donate the blood, but you'll also notice that the donors were also immigrants from Latin America--Mexican, there's two from Bolivia, two from Paraguay, and lastly a Chilean donor.

So why so few transfusion cases, the question I get all of the time, and I think the answer is really rather simple. The reported cases that we see are, in fact, the sentinels or, if someone wants to say, the tip of the iceberg.

These individuals have all been immunosuppresed. They generate fulminant disease in which the parasites are very obvious. In some cases, they were detected in urine, other places very easily in blood smears. So these are the real obvious cases we see. More than likely, most of the cases are missed. These are the immunocompetent recipients, ones who may be diagnosed. Just to make this clear, most cases, in fact, are not even recognized.

A few years ago we did a study that was subsequently published in circulation, which we looked at over 11,000 cardiac surgery patients. We were actually looking at them from the standpoint of lookback because we were hoping to demonstrate transmission because cardiac surgery patients receive multiple blood transfusions.

What we found when we tested this repository that is held by Johns Hopkins, was that six out of the 11,000 or .05 percent were positive, they had Chagas disease. What was interesting was when we looked at the preoperative/postoperative samples, everyone had Chagas prior to the operation, so no one got it through blood transfusion.

It was also interesting that when you do the numbers, and you look at the demographics of this 11,000 in this repository, 3 percent of the Hispanic population in this repository actually were positive for the parasite, had Chagas disease; in fact, these being cardiac surgery patients, some of them with cardiomyopathies and other arrhythmias and other associated problems that might be suggestive of Chagas disease.

Along with the fact that they were immigrants from Latin American countries, one might think the medical community would actually consider testing them for Chagas. Well, no, not a single one of these had any medical history of Chagas or any tests for Chagas disease. So it just points out that this is not something which physicians in this country, by and large, recognize, probably receive little training in medical school-- physicians here could back me up on that one--and something that, unless it's very obvious as something, it's probably missed.

This is some of our seroprevalence data, studies we did in Los Angeles and Miami. It was published just last year, and it points out that in Los Angeles we looked at slightly over 1.1-million donations and a smaller number in Miami. This is a study in which we asked a simple risk question: Were you born in Mexico, Central America or South America or have you spent more than six months?

As you can see, you can pretty significant populations will answer, yes, to your question. In L.A. it's 7 percent and in Miami it's 14 percent. Some have suggested just using that question of the deferred donors outright, and certainly with blood shortages, I don't think Miami would want to give up 14 percent of their donors.

If you follow these numbers down to the bottom, we tested them by EIA and then confirmed them by RIPA in my laboratory. The seropositivity rates were about 1 in 7,500 overall donors in L.A. and about 1 in 9,000 donors in Miami.

If one takes that L.A. data and begins to break it down, it becomes much more interesting. This is how it looks if you look at the data from 1968 through 1998. It's the years, and then this is the percent of donors who are positive or who have antibodies, but don't worry about this scale, just look at these numbers up here.

From 1996 through 1998, the rate of positive donors in L.A. incrementally increased from 1 in 9,900 to 1 in 7,200, to 1 in 5,400, and that is a significant increase. So what's going on? Well, we broke this data down a little bit farther in another direction. We look at differences by donation type. This is the same data set broken down by allogeneic, apheresis and directed donors.

We found in the rates here that allogeneic was about 1 in 7,200; apheresis, 1 in 93,000; directed donors, 1 in 2,400. What we've found, as we looked through all of this and we started breaking it down and seeing who answered yes to the questions, we found a large number in the directed donors, 10.2 percent, and only 7.5 percent in the allogeneic. What it all really comes down to is the number of at-risk donors in your population who are donating.

So what was really happening during this time period, as we found out when we talked to the people in L.A., is because of changing donor demographics in L.A., they changed the recruitment efforts and started targeting Hispanic populations in Los Angeles. So as we began to change our demographics in this country, as I showed you the census data earlier, as the demographics change in this country, we are certainly going to be recruiting more Hispanic donors, and as we do that, we're likely to see increases in that number of positive individuals.

The other question I get, and we've done lookbacks at the Red Cross, is why haven't we demonstrated transmission by lookback? Actually, we're 0 for 19. That's where you always keep striking out, some might say. Well, I'm going to tell you perhaps some reasons why we see that and maybe why this shouldn't be something that we hang onto and be of that great concern.

So what, we're 0 for 19, but we know transmission occurs in Latin America, somewhere on the order of between 13 and 49 percent of positive individuals are thought to transmit the infection. We see transfusion cases here in North America as well, so we know it actually happens.

In conjunction with the CDC, we actually looked at some of our seropositive donors. Some have said, well, they have antibodies. They're just not infected. Well, the general thinking is, once you're infected with this parasite, you are infected for life. In fact, when we tested our seropositive donors, we could demonstrate that 33 out of 52 were 63 percent were actually parasitemic. They had parasites circulating in their blood. So that means 63 percent of the time we were actually transfusing blood with parasites in it.

What is important, and I think is relevant, is that we found this parasitemia is intermittent. It wasn't there every time we tested, even on individual donors. We could test them one time, we could not demonstrate parasitemia. The next time we would test them it was there.

What was also interesting and relevant to our lookbacks is, if you break down this 19 by the products, the recipients received, we found that 11 were red cells, 3 were FFP, 2 cryo and then 3 platelet units, and we kind of focused in on these platelet units because we thought, perhaps, platelet units are actually the one that causes the greatest amount of problem, as far as transmitting Ti cruzi.

In fact, if you look at excluding the last case in Rhode Island, at least five of the six reported transfusion cases in the United States have involved platelet units. Perhaps the reason why we see this is that platelet recipients tend to be more likely to be immunocompromised than those receiving red cell units, but our thinking also suggests that maybe Ti cruzi may separate with the platelets during whole-blood centrification. So we've actually done some survival studies in the lab, seeing how long the parasite survives, and where it ends up when it separates out.

What we've seen in whole blood, the parasite survives quite well up to about three weeks. In platelets, it seems to survive for at least four days. Considering the shelf life at present is five days, that means almost the entire shelf life of the platelet unit, the parasites survive and are capable probably of transmitting infection.

Much to our surprise, the red cell units also do quite well, and the parasites again surviving up to three weeks. So there may be some discrepancy here, but it may then get back to who is immunocompromised or where we actually can detect the infection. Lastly, in plasma, we didn't see any parasite surviving.

So what about nationwide risk? Let me walk you through this slide. If we consider in this country there's 13.2 million donations per year, now, if each donor in this country donates, on average, 1.6 times, if we divide 13.2 by 1.6, we get 8.25 million donors per year in this country.

Now, based on some surveys we did nationally, we think probably about 2.5 percent of present donors are at risk. They have risk factors. They are born in endemic countries. So if we multiply 2.5 percent times 8.25 million donors, we have approximately 206,000 at-risk donors in the U.S.

Based on some of our studies in the laboratory and in other locations, we think about one out of every 625 of those will actually turn out to be seropositive donors. They will confirm as being infected. So that leaves us with 330 seropositive donors.

Once again, if each one of these donors donates 1.6 times, we likely have 528 seropositive donations per year in the U.S. Now, if each of those donations, on average, is broken down into 1.17 components, we have at least 618 potentially infectious components transfused or produced in this country each year, not necessarily transfused.

Keep in mind, this is all based on at-risk donors. That doesn't include congenitally acquired infections. So, in many ways, this can be considered as a conservative estimate.

Let me shift gears to the last phase which I'll talk about. I grouped these together because these are all organisms transmitted by ticks. If you look at the infections, this is the big three so to speak, that are all transmitted by the same tick, the deer tick, Ixodes scapularis. Those are Lyme disease, human granulocytic ehrlichiosis and babesiosis.

First of all, I'll qualify this by saying I'm not going to talk about Lyme disease. There has, to my knowledge, never been a transfusion case involving Lyme disease, much maybe to our surprise. There are certainly explanations that the spirochete does not survive well under blood bank conditions. The period of spirochetemia in the donors may be very short, but the bottom line is we have not seen a transfusion case, so I'm going to focus on these two, in which there has been transfusion cases.

Now, this life cycle, by and large, involves a group of animals. There are deer involved, not because the deer are actually reservoir hosts, but they are actually good places where these ticks, the adult ticks live, and eventually lay their eggs and so forth. They're actually good transport hosts, too, because they travel quite far so they can take the ticks quite a distance.

The real culprit is this little guy, the white-footed mouse, who is actually the reservoir host for babesiosis and Lyme and so forth, perhaps bartonellosis. We'll come back maybe in a couple of years about that one, but we're finding new organisms all the time.

The first one I'm going to talk about is the agent of human granulocytic ehrlichiosis, now called anaplasma phagocytophilum. Its name changes quite frequently it seems these days. Not too long ago, it was always just known as the agent--I always found kind of funny--the agent of human granulocytic ehrlichiosis.

Then, for a while, everyone called it ehrlichios species, and then it was renamed ehrlichia phagocytophilum. Then, most recently in some paper by Steve Dumler's group in Hopkins, it was renamed anaplasma phagocytophila, with an "A" on the end. Actually, I published a paper. It came out in the December issue of Transfusion that had anaplasma phagocytophila in the title, and then I saw another paper somewhere and it had the "u-m" on the end. So I sent an e-mail to Steve and said, "What the heck is going on here?"

And he said, "Well, the Bacteria Systematic people don't think it should have an "A" on the end because that's plural. So now the correct terminology, the latest is a "u-m" on the end.

The important thing is it's actually rickettsia. It's newly emergent and appeared in 1994. There are a series of agents that have just come out in the last 10 years that are of some concern. It actually lives, as the name suggests, inside granulocytes, and actually the parasite is this little circular thing living in what's called a morula.

It is, in fact, a tick-born zoonosis; again, the same tick, I. scapularis. On the West Coast, the thinking is maybe it's I. pacificus, Ixodes pacificus.

As with most of these agents, the symptoms are generally mild and flu-like. However, there can be severe symptoms--renal failure, gastrointestinal bleeding, secondary infection, and 5-percent fatality rate. If you can diagnose incorrectly, treatment is not too bad--Doxycycline.

What about the seroprevalence of this organism? Well, we did a couple of studies, one which we just published, and was using samples we collected in 1996 in Wisconsin and Connecticut. In Wisconsin we found about 5 out of 1,000 or .5 percent of donors had antibodies to this parasite. In Connecticut the levels were surprisingly 3.5 percent. It's not so surprising any more, because we've used samples collected in 2001, on which we found almost the same rate, again, 4.1 percent of donors have antibodies to this parasite.

Well, what about transfusion transmission? Well, there has been a case involving this parasite. It was actually reported in Minnesota, involving a red-cell unit. It was confirmed by symptoms, serology, 1:512 antibody titer as well as PCR. The donor involved had a history of lyme disease, extensive tick bites, and a very high serology, a very high antibody titer for this parasite. Certainly the history of lyme disease suggests exposure to the ticks, and we know that these ticks can carry two or three of these organisms. In fact, these ticks can transmit two or three of these at the same time.

The agent also survives quite well in blood, at least 18 days in laboratory experiments, and based on this transfusion case, we know it can survive at least 30 days in blood banks.

So given some of the seroprevalence figures I gave you, why don't we see more cases? Well, first of all, it's most likely misdiagnosed. Flu-like symptoms could be lots of things. Probably subclinical, and some of the thinking, and talking to some of my colleagues at the CDC, is that it likely has a very short bacteremic base, so donors who are infected have a very narrow window in which they can transmit the infection through blood transfusion.

And one last possibility is the fact that much of the blood in this country now is leukoreduced. Considering that the parasite is inside granulocytes, may be in fact pulling out most of the granulocytes that contain the parasite.

And last but not least is Babesia, certainly an up and coming agent, as you'll see. This is the agent of babesiosis, once again like malaria, and it's a very close cousin of malaria, actually in the same group, lives, resides within red blood cells. Again, carried by the same tick, as I showed you before, causes flu, malaria-like symptoms. Infections are generally asymptomatic or self limiting. Most of us can handle Babesia quite well. For those of who can't there are antibiotics that can be used to treat the infection. However, it can be severe and/or fatal, primarily in people who are elderly, immunocompromised and asplenic. And I think that last three, elderly, immunocompromised and asplenic, all tend to be people who get a lot of blood donations.

When one looks at the geographical distribution of this parasite, the transmission or the distribution is actually expanding. B. microti is largely in two foci, one the upper midwest, Minnesota and Wisconsin, and one in the northeast, New York and New Jersey and Rhode Island and Connecticut as well as Massachusetts.

And within the last 10 years there's been this group of what are called Babesia-like organisms, and they're designated by the state they are found, but we commonly call them WA-1, MO-1, et cetera, for California and Washington and for Missouri.

What's interesting about these, there's been at least two transfusion cases of this new parasite, WA-1 already, and they're endemic ranges appear to be expanding. It's one of those instances I think, when we start looking for these organisms we find them. What's also appeared within the last year is the description of B. divergens, Babesia divergens, which actually is a cattle parasite which causes most of the Babesia, human babesiosis cases in Europe, and actually causes a much more severe disease. And it was found in Kentucky. There's some thought that this MO-1 parasite in Missouri may in fact be B. divergens, so this is certainly an area or a consideration for these organisms that's rapidly growing.

Well, what about transfusion cases with Babesia? I up this figure all the time, and I think it's reasonable. It's probably even much greater than this, but I think there's at least 50 known cases of transfusion-transmitted babesiosis at this point from 1979. Most of the cases at this point aren't published because I think most individuals feel everything that's said about it has been said about it, and so these cases just aren't getting into literature. But those of who know, and the CDC, the Red Cross, up through New York and Connecticut health departments, we certainly are aware of these cases.

The recipients are anywhere from neonates to 79 years of age, and as I said, most of them are immunocompromised. Platelets and red cells have been involved. Of course platelets can be contaminated by red cells, so if red cells are infected they can be transmitted through platelets, up to 35 days, so basically almost for the entire shelf life of the red cell unit. About 2 to 8 weeks is the incubation period in the recipient.

And lastly, these infections can be identified by a variety of techniques, including serology, PCR and/or animal inoculations. What that refers to is hamsters are actually exquisitely sensitive to infection of Babesia, so if you inoculate a hamster with human blood and then do a little smear on the blood at weekly intervals, you can actually pick up the infection.

How much is known about seroprevalence of Babesia in this country? And the answer is, well, not that much. There's been a series of studies that's shown the rates anywhere from .3 to as high as 9.5 percent from a variety of individuals. However, very few of these studies have actually been done on blood donors. There was an early study in Cape Cod by Marc Papovsky, showed about 3.7 percent of donors are positive. One in 2000 by Jean Linden, 4.3 percent. And then we published some recently in Wisconsin and Connecticut showing slightly lower numbers. The important thing though I think is that there is a fair amount of Babesia out there in the general population as well as in blood donors.

One of the questions that came to mind rather quickly was why not just defer donors based on tick bites? Well, we sent out some postcards to a variety of donors in quite distinct, geographically distinct areas, and asked them a simple question, if they had been bitten by a tick in the last six years. Surprisingly, out of 6,000 postcards, we got 2,400 back. What we found was that 4 percent of the donors actually reported a tick bite within the last year, and what was particularly interesting was the difference in some of these regions, blood collection regions.

In Tulsa, Oklahoma, 9 percent of the donors reported they had been bitten by a tick in the last six months. Similarly in Atlanta, down near the CDC, 8.4 percent. Once again, deferring donors based on this criteria alone would certainly be unacceptable.

What we did notice was that donors were also very good at distinguishing large and small ticks. And from this standpoint we were trying to differentiate large ticks being dog ticks versus the smaller deer ticks. And the donors seemed to quite good at that as well. And as you can see, a large number of the smaller ticks.

One thing I should point out is that when you talk about tick-borne transmitted diseases, most individuals--and this is true of lyme disease, babesiosis--most infected individuals do not recall a tick bite, as I'll show you. In fact, when we looked at the seroprevalence of tick bits, we thought maybe the next level, if we can't defer them based on a tick bite, maybe we could test donors who report a tick bite. And this was published just a couple of months ago. And if we looked at individuals with tick bites and those who are controls from Connecticut, we found that the percent positive was virtually the same, .3 versus .4 percent. So actually asking people about tick bites was of little use. In fact, as we talked to some of these donors in more detail, we think that maybe asking donors about a tick bite may be actually negatively select them, because those who report tick bites are the ones who are looking for the ticks. Those are the ones, who after they come in from outside, check themselves out. Those are the ones who probably use DEET when they go outside. So it's the ones who don't look for ticks that may be the ones we ought to be worried about.

We've done some studies in Connecticut too, and some of these are ongoing. This was an early study in 1999 comparing endemic and nonendemic regions. There again, these are arbitrary lines. Certainly there are ticks that are infected all through the central part of Connecticut. I mean they do cross that line. If you look at the endemic versus nonendemic areas, we had roughly--not roughly--we had the same number of donors, 1,745, and the percent infected was, in the endemic area, 1.4 percent; in nonendemic was .3 percent. Gave us an overall rate in Connecticut in 1999 of almost 1 percent of the blood donors.

Well, that was the antibodies. What about whether or not they had the parasite? We called back 19 of those seropositive donors from 1999 and we did PCR on them, nested PCR, and we found that 10 out of 19 or 53 percent of them had the parasite in their blood system. Once again, not unlike Chagas disease, a fairly large number of donors were in fact parasitemic, capable of transmitting infection.

That led us to an interesting study which is still ongoing, and it's actually a cooperative agreement the Red Cross has with the CDC to look at the issues of serology and parasitemia, how all these things relate with blood donors. And as I said, it's a 3-year study. It's actually kind of going into its fourth summer. And actually we're enrolling seropositive donors, donors who have antibodies to B. microti. And every 30, 60 days we're testing them by serology, blood smear, PCR, hamster inoculation, and also we're asking them a brief questionnaire to find out if they've been exposed to ticks in the interim, see if they've become reinfected or what maybe is going on. And as I said, we're looking for the relationship, if any, between serology and parasitemia.

This is the compilation of data from the three summers which we did this, once again about the same number each year, from about 2,100 up to about 2,600 donors. Seroprevalence rates, very, very little, from .8 this year, as high as 1.4 last year. Again, we think it's probably about 1 percent each year in Connecticut. When one looks at the PCR rates there was some differences, as in '99 we had 53 percent, in 2000 it was 56 percent. Then it dropped to 8, and back up to 14 this year. The hamster rates are also variable, a little bit less, not quite as sensitive.

But what was interesting is the question why does the rate change so much from a couple years being in the 50s to lower rates? Well, if we look at these--first of all, all the donors smear negative, so we can't really detect them by blood smear, it's not sensitive enough. Several of the donors were actually repeatedly or intermittently PCR position and I'll show you some individual data.

And then lastly, the differences in PCR positivity. These infections of Babesia in general are certainly affected by climactic and ecologic factors. The ticks themselves have two-year life cycles. So what happens, one winter may have a very important effect on what happens to the ticks the following year and their ability to transmit infection. Certainly this being a very cold winter, ticks don't do well in cold winters unless it's very snowy, because then the snow protects them. So it will be interesting to see what happens after this winter being cold for a change. Climactic and ecologic factors effects the other hosts involved, the deer population as well as the rodent population. So all these things are tied into why we might see differences year to year.

The other factor is, for old donors, we're actually deferring all IFA positive donors, so we may be in fact pulling out of the donor pool those donors who are actually most susceptible and most likely to be parasitemic. Perhaps we'll learn more about that.

A couple more slide, then I'll be done. And this one is just a few slides about some of our donors in the study. This is typical of what we've seen. This is subject 1426, first identified in July of 2000. We followed him every couple of months after that. Initially they had a titer of 1:512. When they came back, were entered in the study, they were at 1:256, and they were parasitemic both by PCR and hamster. Within the next couple of draws the parasitemic went away. As you can see, the antibody titer dropped below baseline, and they were released from the study. This is what we would typically expect to see, someone was infected, parasitemic, clears the infection and is fine.

Then we have donors like this, and this is not a uncommon occurrence, donor 2348, first identified in August 2000, very high IFA. Again, parasitemic. Was treated actually for babesiosis, received a 10-day treatment I believe of clindamycin and quinine. No longer parasitemic and has not been since then. But for that time--and we're now I think somewhere into December of last year, so we're a good 24 months along--the antibody titer of this individual has not dropped. So we're not really sure what this means, but it's not out of the question that this individual can in fact be a chronic carrier, someone who was infected with Babesia and does not clear the infection.

And these are the kind of individuals that concern us as far as blood donation. What do you do if someone who still has an antibody titer, but yet you can't measure parasitemia?

Then the other category donors we see, starting now to see on some basis, is like this one, 1078. Once again a lower antibody titer, was both parasitemic by both methods, cleared the parasitemia. And as we got down to this level, by ELISA as well as IFA, was at baseline. In fact our criteria for removing these donors from our study was after three months of being negative on all tests, they are dropped from the study. So we thought when this donor came back in May, would have another low titer and be dropped. Well, suddenly the levels jumped back up and have remained that way since then, although they're right around the cutoff, which suggests this person may have been re-exposed to the agent. It's not surprising because most of these individuals live on large properties, which they have deer on their properties and probably have a chance for re-exposure, even though we couldn't measure it be parasitemia.

Perhaps the most interesting data I'll show you is from our lookback investigations. These involved individuals in Red Cross who had previous donations from IFA positive and/or parasitemic donors, and we went back as far as 12 months. Recipients were then tested by IFA and PCR, and this is ongoing, but we've had 44 donors who donated or 118 donations. And if one looks at the data--and let's just go right to the bottom; this is the most important part--number of products transfused is 204. The number of recipients we've tested out of these is 28, and the number of recipients that are positive by antibody and PCR is 7. 25 percent of the recipients of the blood were infected by this parasite. So one out of four. That's, I would have to say, a pretty high transfusion rate, and certainly increases the concern for this organism.

So in summary then, I'll say that parasitic agents pose an ongoing and increasing risk to blood safety. Most importantly these infected donors are not often or almost always asymptomatic. They appear to be quite healthy. Even those ones who have infections with things like Babesia and Chagas disease, outwardly most times don't even know they're infected.

The implications for recipients certainly vary. As I said, in most cases recipients go unrecognized. In some cases, like Babesia, we can give them antibiotics, but in some cases like Chagas disease, it's an untreatable infection. Might as well forget it now about question strategies because they by and large lack sensitivity and specificity.

A problem now for all of these agents is that licensed tests are unavailable, so if we make the decision today--not we--if you would make the decision that we need to test the blood supply, at this point there are no licensed tests available for any of these agents. That also brings up always the question, what potential role might pathogen and activation have? If we could just implement, if we had effective methods of pathogen activation, perhaps all these agents could be eliminated as well, but then again that's the promise of pathogen activation.

So what about donor management strategies? I'll just give you four possibilities here. First one for malaria, we are already doing questioning, and it seems to do quite well for the most part with only one or two transfusion cases per year. But I raise the issues of screening, not from the standpoint so much as preventing infections, you know, perhaps it would prevent those other two before increasing the number of donors in the donor pool.

Chagas disease seems to be that we are moving towards universal screening, and I think Jay can correct me if I'm wrong, but at the BPAC meeting in September the FDA expressed an interest in having manufacturers submit tests for blood screening for t. cruzi, and they suggested that if such a test was submitted and approved, that we would perhaps move towards screening the blood supply for Chagas.

HGE is just something that we need to monitor more. There's been very little information. However, perhaps leukoreduction is already doing the job, but some studies of that nature actually would be beneficial.

Now, babesiosis is perhaps a little more complicated, but there are some possible solutions. Certainly because we see this early parasitemic phase, this is the one we might consider NAT screening. I would not suggest NAT screening for Chagas disease because these are individuals who are infected as children and they have very strong antibody titers, so they are easily detected. And babesiosis certainly has window periods where NAT screening may be involved. Certainly blood screening situations are something for consideration. However, given its regional nature and the fact that most individuals can handle infection quite well, the route to go here may in fact by the CMV model, in which we provide tested units for individuals who are at risk.

Now, since this is a panel that's trying to prioritize the issues, I thought for your help, your sake, I'd prioritize the parasites. This is my own list. So if I would prioritize these as standing, I would put Chagas and babesiosis as 1A and 1B. Always the problem here is while there's lots of cases of transfusion babesiosis, but which would you rather get, babesiosis or Chagas? I think it would opt for the treatable one. But either way, I think these are both ones that are worthy of consideration.

Certainly granulocytic ehrlichiosis is right up there as No. 3 and I would now put malaria as 4, but if we start establishing endemic populations of the parasite in this country, maybe that's something else we should consider, and certainly all the other agents.

Thanks you.

DR. BRECHER: Thank you, Dave. That was a very nice review, very comprehensive.

We have time for a couple questions and comments. Jean?

DR. LINDEN: Thank you very much. That was a truly excellent summary. I just have one very quick question on the Chagas. You mentioned finding I believe that 63 percent of the donors are parasitemic, which I understood was at some point, not on a single test.

DR. LEIBY: Correct.

DR. LINDEN: So how many tests did you do and over what time period?

DR. LEIBY: The maximum number of tests we did was three, and so most of them were--we saw various patterns. Some people were positive on more than one test. Some it was every other test, and so it really varied. And what you have to keep in mind is that not only are they intermittently parasitemic, the levels in the blood bank may be low, and all the problems with PCR is that you're taking a small sample. So if the organism's not in the small sample you take, you may in fact be missing it. So it gets back to what Mark said, while we can't the whole unit, we're testing a small portion

DR. LINDEN: And those three tests were over what time period?

DR. LEIBY: Over anywhere from about 3 to 6 or 9 months, depending on the cooperation of the donor, which was sometimes quite difficult.

DR. BRECHER: Celso, want to ask if it's automated?

[Laughter.]

DR. BIANCO: Two issues, David, and I have to concur it was an excellent summary. In your last point with babesiosis and the lookback, I don't think that you were saying that those cases are transfusion transmitted. Your past experience with Chagas showed that a lot of those people with Chagas were positive. All these patients may be coming from endemic areas, may have been infected before.

DR. LEIBY: Well, I'll address that. If you look at the rate there as 25 percent, our transmission rate, and if the rate in Connecticut is only 1 percent, it's very unlikely that all those individuals would get it through--

DR. BIANCO: Well, I think we need more than that. The other point, and I think a point that you raised very appropriately about priorities and that is the them of this Committee, I think that we have to put those--and I want your opinion--all these agents under a bigger picture and context. Certainly you put the priorities for the parasites, but you didn't include them into the bigger issue of priorities. Do we do first bacterial contamination or Chagas, or bacterial detection or Chagas?

The second issue, I think that we have, and you have, particularly with all these years that you have dedicated to that, looked at the issue of transmission of parasites by blood in the classical setting of donor prevalence and lookback. But I--and I want Dr. Chamberland--I'd like to see more the other side. That is, to have 7 cases of Chagas in 15 years, but you have 50 of babesiosis in 50 years or 30 years, that are being followed, one of HGE. All of them are less than what we have in malaria every year. And there is something that when you look at Chagas and you see 618 positive individuals and you don't hear about it, you have 7 cases in 15 years, there is something where the clinical implication of those transmission, even if they are occurring, does not appear to be significant so that was picked up by the clinicians. Yet, people are not informed by Chagas. People don't look for it. But in the higher contest of priorities I wish we could have a more national epidemiological look at the impact of those diseases in the country with the only one for which I know there is serious follow up is lyme disease. But so that we could see the other side of the picture, what is the prevalence in the population, what is the impact. And then go back and see how many of those were recipients of transfusion and how much of that transmission could be attributed to transfusion.

And I think that is very important experience, particularly in South America, where over the years with better protection methods for instance for Chagas, transfusion became the most important means of transmitting Chagas in the '80s because people were so effective in combatting the vector and all the other things, and so transfusion became really the focus.

I'm open to your comments.

DR. LEIBY: I think that is the challenge the Committee faces, is actually prioritizing these agents, and I'm not going to try to pit bacteria against parasites or those kind of things. That's your job, not mine I guess.

I would say though that if you look at Chagas, obviously Babesia if there's a lot of transfusion cases, and there again, I'm sure there's many more than that we're not seeing. I think Jean will agree with that. In Babesia I would say there's probably 5 to 10 cases per year at this point that we know about. Chagas, I go back to the point that, yeah, there's only been 7 cases, but you know if the rate in Los Angeles was up to 1 in 5,400 donors, and if we know that 63 percent of those donors are parasitemic, then you have to ask the question: are you willing to transfuse that blood that has parasites in it? And then I think that becomes the bottom line.

DR. BIANCO: What we have to understand is the disconnect, David, and you as the expert--and we're talking about experts since yesterday here--has to help us understand the disconnect. That is if you have 63 percent of the people that are positive on PCR, those parasites are in the unit of blood. If you go to any more experiments, at least in mouse models of Chagas, you can transmit with one or two bugs.

DR. LEIBY: Sure.

DR. BIANCO: What's happened?

DR. LEIBY: I think they are being transmitted, and I think the cases are there. I think the people are just not being recognized. That was the whole point of the early slide, the cases that we do see are the obvious fulminant cases. And I think the other ones are there, and I think those donors, those recipients, 20, 30, 40 years down the road are going to have the cardiomyopathies and the other problems.

DR. BIANCO: Are there studies of frequent or particular patients like thalassemias and others in areas like LA or New York, to see the frequency or the prevalence of those markers in those individuals that are receiving red cells every couple weeks?

DR. LEIBY: And the answer is no, of course. You know that as well as I do. But I still it gets back to--I mean you're looking for concrete studies, and I'm not sure at this point, generating more studies for something we already know about. I mean you know the South American experience. You just cited it yourself quite correctly, that South Americans have made great strides, and that in fact, yes, transfusion is the greatest or the primary way Chagas is transmitted in Latin America these days because they've, in many cases, eliminated vectorial transmission. And the same lines, all of Latin America tests for T. cruzi, even Mexico now, and we don't. And yet we see this recurring cases, and I think we cannot bury our heads in the sands because we can't demonstrate these by lookback when we know in fact it occurs.

DR. BRECHER: I think since we're running behind, we're going to have to stop now. Why don't we take a 10-minute break and we'll come back.

[Brief recess.]

DR. BRECHER: Everyone take their seats, please. We're now going to move on to pathogen reduction. Steve, are you ready? It's all yours, Steve.

DR. WAGNER: Good morning. I've been asked to talk a bit about pathogen reduction in cellular blood components, and I'll try to give an overview and a bit of perspective on this subject.

There's a number of rationale for pathogen reduction or inactivation. Everyone recognizes that there is some residual infectivity in blood, even in potentially tested products because of window periods and whatnot. There was a great concern, obviously, for plasma products that are pooled, and I'd like to remind everyone that platelets are pooled as well, platelets derived from whole blood. And so that also is a rationale because that increases probability that an infected unit could be combined with others.

Pathogen inactivation might constitute an additional layer of safety beyond all the donor deferral mechanisms that we have in place now with respect to questions or infectious disease tests.

Pathogen inactivation has been suggested for agents that we are familiar with but for which we have no test, and I think David Leiby and Mark Brecher have gone over some of those agents this morning.

In addition, there are agents that we are well aware of which may mutate and might not be detected in a mutated form, and so it's been suggested that pathogen inactivation might be useful for these variant agents.

And, finally, there are those who believe that pathogen inactivation might be very effective against new agents, and I think this is somewhat controversial because one might--one would have to show that a particular method would be active against all agents to be sure--not even be sure, to hope that it might effective against new agents. But the argument has been put forth.

Then, finally, there's a great public and political expectation to have a zero risk blood supply, difficult though it might be.

There are a number of challenges, though, to solving a pathogen reduction problem. First of all is that pathogens, as you saw with David's lecture occur in all different cell compartments. They can be extracellular. They can be intracellular. They can be pro-viral forms and they can also be virus-associated, for example, where a virus is attached to a white cell membrane.

One of the other difficult things to deal with in pathogen reduction is that the different pathogens are all different and they have different susceptibilities to a particular agent. And one good example is hepatitis A, which is a non-enveloped picornavirus which has a very closely knit or packed virus capsid structure, so the proteins are so tightly packed that very few molecules are small enough to go through the pores of the virus. This makes the virus quite insensitive to most known disinfectants and agents that might be used for pathogen reduction.

Then, finally, there are some agents that can be present in very high quantities in blood, attain very high titers. And most of the methods for pathogen reduction probably are neither robust enough to be able to detect in a test system, the many number of logs that might need to be reduced--an example might be parvovirus B19--or it may just be very difficult using any one method to be able to inactivate all the infectious particles that might be present to prevent transmission.

There's a number of approaches to inactivation. Today I'll talk about those that have been used for cellular blood components. I'll talk about psoralens and, for example, S-59. I'll talk a bit about riboflavin. And for red cells I'll talk about what's called a FRALE compound, and I'll explain that a bit later, which is termed S303. And then another molecule called INACTINE. I'm not going to talk about plasma today.

S-59 is a psoralen. It's a heterocyclic aromatic structure made of three rings that line in a plane. It readily intercalates into nucleic acids because of its planar structure and also because it's an amphiophile. It's got a portion of the molecule that can form a positive charge that can potentially interact with a negative charge on the phosphate backbone of nucleic acids.

As I said before, psoralens intercalate between the bases of double-stranded regions of DNA, and also there are double-stranded regions of RNA as well. And upon the absorption of ultraviolet A light, psoralens can make mono- and di-adducts with pyrimidine bases in nucleic acid. And the presence of these adducts prevents the subsequent nucleic acid replication. And the logic here is if you have something that goes against nucleic acid and you're working on a way to inactivate pathogens in platelets, well, the only nucleic acids in platelets is in mitochondrial DNA, which is not thought to be necessary for storage or for transfusion. And so it's possible to distinguish the pathogens from the platelets by using this target.

The investigators who are looking into this method prepare apheresis platelets, and it's suspended in a platelet additive solution. S-59 is added to the platelets, and the mixture is then transferred into a UVA-permeable plastic container. They then shine that UVA-permeable plastic container with light of the appropriate wavelength in the UVA region, and when that is finished--and it takes just a couple minutes, for the light exposure, at least--they transfer the platelets to another container, which contains an absorbing resin which removes a large amount of the free S-59 that's in solution.

I should mention, though, that it would not be expected to remove any S-59 which is bound to the cells in any way. And so that would be transfused with the unit.

The platelets then that contain much lower levels of reduced S-59 that are free, that are not bound, are transferred to a storage container.

S-59 is quite robust in inactivating enveloped extracellular viruses and intracellular viruses. In fact, there are some non-enveloped viruses that it's effective against, and the companies involved in this have begun to show some effectiveness for inactivation of some parasites.

As far as I know, since it is a nucleic acid agent, it will probably not be effective against prions. It is effective against both gram positive and gram negative organisms. It's probably not effective against bacterial spores; however, one might argue that spores, given time to incubate in a blood unit, would probably germinate and with time would grow up. And so that would speak to the fact that you might want to do the inactivation process not immediately after the blood was collected, but a little later to be able to let any spores that might be present to germinate.

In addition, S-59 is probably not effective against endotoxins since that's, again, not nucleic acid-based. An endotoxin, as you know, is a great risk for sepsis in transfusion-associated sepsis. So you could actually kill the organisms with the inactivation mechanism, yet transfuse enough endotoxin to kill an individual. And what that speaks to is that you probably can't wait one or two days to do your inactivation process to give the possibility for the bacteria to completely grow up, because then you could accumulate endotoxin. And so there has to be a timing of the process in order to allow spores to germinate first, but not allow endotoxin to accumulate.

Another pathogen reduction process that's being investigated by a different company is based on riboflavin. As you know, riboflavin is a vitamin. Again, the common theme is this aromatic tricyclic, a planar structure that intercalates into nucleic acids. It's got a sugar for a tail. And there is some literature, certainly not as much as the psoralens, that riboflavin binds to DNA by intercalation. And in the presence of light riboflavin induces guanine oxidation, single-strand breaks in the formation of covalent adducts between riboflavin and nucleic acid.

And, similarly to what I just spoke about, riboflavin seems to be able to kill extracellular enveloped viruses, and there's some evidence for killing of intracellular viruses. Again, the same viruses that might be difficult to inactivate by some of the psoralens would probably be difficult to inactivate with riboflavin, those viruses that have very tightly packed capsid structures.

Riboflavin is probably not effective against prions, and the same thing goes for bacteria. Some bacteria have been demonstrated to be inactivated by riboflavin, but, again, one might not expect it to be active against spores or against endotoxin.

In riboflavin right now--let me back up. Riboflavin is being used predominantly for inactivation in platelets, although some work has been done in red cells.

This is a compound that's being developed. The company terms is a FRALE compound. It has a structural similarity to quinacrine mustard for use in red cells for inactivation. Quinacrine mustard is an alkylating agent. It's got a nitrogen mustard moiety coupled to an acridine ring so it intercalates into nucleic acids by the bases of the acridine ring and then forms covalent cross-links with the nitrogen mustard.

Quinacrine mustard is a known closterogenic(?) agent, and so it obviously would be difficult to use that in a blood supply. And what the investigators did was put an ester linkage in the middle of the octyl chain, and we'll ese what that does.

So these FRALE compounds stand for frangible anchor linker effector compounds, and the anchor, an acridine moiety of FRALE compounds, is responsible for intercalation between the bases of double-stranded regions of DNA and RNA, and the nitrogen mustard moiety, or the effector of the FRALEs, make an adduct with nucleic acid bases. And the di-adducts form a cross-link between the nucleic acid strands that prevent subsequent nucleic acid replication of the pathogens. And like platelets, red cells contain no nucleic acid, so there's a way to distinguish the pathogens from the red cells.

The ester moiety in the FRALEs is what's termed the frangible linker, and with time, it hydrolyzes forming a negatively charged acridine compound that should not further interact with nucleic acids because nucleic acids are negatively charged and like charges repel. And, in addition, the investigators have used a removal device to reduce the concentration of remaining compound even after the hydrolyzation.

S-303 has a spectrum which is similar to the other compounds I spoke about. It inactivates enveloped extracellular viruses and intracellular viruses, some non-enveloped viruses. Because its major target is nucleic acid, it shouldn't be effective against prions. It can inactivate a wide range of bacteria but, again, is not effective against spores and probably not effective against endotoxin.

Even though all these agents I speak about are fairly specific for nucleic acids, there are side reactions that occur, and for FRALE compounds, one of the consequences of using the FRALE compound is that it can react with surface proteins on the red cell membrane, and this is demonstrated by this FRALE compound called PIC-1. And if you use an antibody against acridine which would pick up the molecule and do flow cytometry, you can see it's binding to the red cell membrane compared to a control which contains no FRALE compound. And this is quite striking, the difference.

However, in the presence of glutathione, as you can see here, for example, at the 2 millimolar level, that binding can be reduced quite a bit towards but not to baseline values. So this is a situation again where there might be some binding of the compound to the red cells that might be transfused.

So if exogenously added glutathione reacts with the FRALE compound, and that acts with an extracellular quencher, and then--and if FRALE compounds can permeate cells and inactivate intracellular viruses, can FRALE compounds permeate red cells and deplete their intracellular glutathione pool? And that's a concern because there are some drugs that people take that interact with red cells and make red cells susceptible to lysis when glutathione levels are low. These are oxidant drugs.

And there are some patients, for example, those who are glutathione-deficient patients or reduced glutathione levels, who might be more susceptible to an agent that might react with glutathione.

INACTINE is a different type of molecule that's being studied for use in red cells. It's a molecule that has--that is different than the others. It doesn't have this tricyclic aromatic ring structure. It does have a three-ring structure which is joined by an alkyl group that has a repeating positive charge. And the tail, the cationic tail, confers DNA binding to nucleic acids, presumably through interaction with the phosphate backbone of nucleic acid. It's said to stabilize the molecule, and one of the things that distinguishes this molecule from the others that I talked about is that its molecular weight is smaller. And so it's able to--its range of organisms that it can inactivate is a little bit broader, although still agents like hepatitis A might be challenging for this agent.

INACTINEs refer to compounds that have aziridine moiety, which is this tricyclic ring here, followed by alkyl groups. And you see the amino groups are the positively charged groups that might interact with the phosphate backbone. This one is called PEN 102, and the investigators are using another one for clinical trials called PEN 110.

The molecule primarily acts at the N7 region of guanine and forms an alkylation. I should mention also that at least we know that ethyleneimine is an agent that causes cancer in animals.

The N7 adduct can stop replication of the pathogen, but it also--DNA repair enzymes can recognize this adduct and remove the base, which can lead to strand breakage, and so there's another--there's more than one mechanism of damaging the nucleic acid.

As I mentioned, INACTINE blocks nucleic acid replication, and this is just evidence of this from the company that's developing it. This is just a DNA sequencing reaction here, and you can see that if you allow the reaction to go to completion, you get very long high-molecular-weight products. But in the presence of INACTINE, you get lower-molecular-weight products. And if you notice that the stops here at C residues, which on the template would correspond to guanine bases, indicating that the stop is at guanine.

The inactivation process for INACTINE involves a typically collected red cell unit to which INACTINE is added or delivered, and this is incubated for a period of time. I believe it's at room temperature. And then there's an extensive removal step which involves automated washing many times, and this is an issue to what to do with the wash. Does it go down the sink? How is that processed? I'm not sure. But in the end, you have a pathogen-inactivated RBC unit.

And susceptible pathogens, again, include extracellular enveloped viruses and intracellular viruses. The range of non-enveloped viruses is probably broader than a number of the other methods because of the small size of the molecule. But, again, I think that hepatitis A would be a challenge for this molecule. The company has done some work with some parasites, and it looks like there may be some inactivation of some parasites.

There's been a claim by those who work with it that the washing step removes prion protein, but I would remind you that prion proteins tend to be a bit sticky and people have found them associated with platelets and some white cells, and I'm not sure that all the studies have been done yet to show that those are removed with this process.

In terms of bacteria, there seems to be evidence that it inactivates bacteria, and, again, I'm not sure about spores. I'm not sure if those studies have been done to show whether or not spores are sensitive or not.

With respect to endotoxin, I don't know. Perhaps someone needs to measure endotoxin levels before and after the washing process to see what occurs in that case.

So there's a number of challenges for pathogen reduction. Processing may lead to unwanted reduction in cellular yields. Some of these have been documented in clinical trials. Although the agents may be specific for nucleic acid, as I said before, side reactions may occur. Some of the most notable ones are reactions, covalent adduct formation to lipids. In the case of psoralens, there's a good body of literature, scientific literature on that.

There is some literature on covalent reaction to proteins. You saw some data I presented on one of the FRALE compounds, and even ethyleneimine can form covalent bonds with some peptides.

To a lesser extent, you see protein alkylations with riboflavin and psoralen, but if you look hard enough, there are some literature to suggest that these might occur as well.

In addition, the compounds that are photochemicals, which are psoralens and riboflavins, they can generate reactive oxygen species, and so instead of going to adduct route, they can interact and product single oxygen or hydroxyl radicals or super-oxide. And these are species that are small that can diffuse, and the diffusion of these molecules can go to other places in cells. Often membrane is a site of damage for these types of oxygen radicals. And that could be responsible for damage to membranes, whether it be platelet membranes or red cell membranes.

And so side reactions may be responsible for the loss of survival or function of the blood component, and these losses of survival or function are modest for some of these agents in clinical trials but, nevertheless, has been observed.

The side reactions may be responsible for unwanted, low-frequency adverse events, because if you have covalent reactions to proteins, that might act as a haptin to which an antibody response could be generated. And so of concern, but not yet observed, are potential immunological reactions, including anaphylaxis. The other concern I had mentioned before, that there might be an increased sensitivity of blood cells to other pharmaceuticals, for example, oxidative drugs, if intracellular glutathione is depleted, and also some patients, for example, glucose 6 phosphate dehydrogenase deficiency, people with that that already have low glutathione, that might be a concern.

With some agents, an unexpected accidental exposure of the staff who are making some of these alkylating agents and other things, to the manufacturing, transportation, or blood center staff, that could lead to an increased genotoxic risk. And so you have to kind of weigh what the current safety of the blood supply is with what you might expect if these things might be implemented.

These low-frequency risks, which have not yet been measured, cannot be estimated from the results of Phase I to III clinical trials, and even with the system that has been studied the most, which is probably the psoralen, the current experience with any one of the pathogen reduction systems probably hasn't involved more than 400 or so, maybe 500 patients, who have been transfused with roughly three, four thousand units. And so if these events are rarer than that, then more experience would be necessary to measure a low-frequency adverse event.

So my own view is that evaluation of these methods requires measurement of these events, these low-frequency adverse events. So without implementation and long-term study, it may be difficult to predict the risk to blood bank workers or to recipients by accidental exposure or by residual drug. And this accidental exposure is real. Sometimes blood bags break. That may happen in the centrifuge occasionally. Even though the manufacturers wish it weren't so, some blood bags just have defects--pinholes or bad seals.

Without implementation and surveillance, it may be difficult to assess the risk of allergic or hypersensitivity or anaphylactic reactions in susceptible recipients caused by alkylations to proteins or by drug metabolites. And then, finally--and this is the bottom line, I think--without implementation and long-term surveillance, it might be impossible to determine if the risk of fatal outcomes from implementing an inactivation process is greater than the current risk of fatalities from infectious disease transmission. And certainly you don't want to do more harm than good.

So what's the risk of fatality from transfusion-transmitted infectious diseases? And you guys have heard about this the last two days, so I probably don't need to talk about it very much. For HIV, it's probably less than 1 in 2 million units. For bacteria in platelets--and I'm citing the CDC study here. I know Mark mentioned a number that was more frequent. But fatality in platelets is at least 1 in 450,000 units, according to the CDC BaCON study. And in red cells, it's much less, about 1 in 7.7 million units.

Certainly for bacteria, it's possible that this number is quite a bit lower because of unreported events that you might not look for. But the bottom line is, no matter how you look at it, these are very small numbers. And so I think it would be that a pathogen reduction system has to be quite safe to pose less a risk than the reported risk of fatalities from transfusion-associated bacterial sepsis. And so that's an important point. Should bacterial screening be put into play, I think it would have to be even safer. The bar would be raised.

So, in conclusion, all methods target nucleic acids. The methods can reduce the infectious titer of extracellular and intracellular enveloped viruses. Some non-enveloped viruses, bacterial spores, endotoxins, and prions will probably not be susceptible to inactivation. Implementation and surveillance may be required to assess low-frequency risks. These low-frequency risks and their assessment is essential for establishing that fatalities from pathogen reduction, from a process of pathogen reduction are less than the current fatalities from infectious disease transmission. And, again, these potential side reactions of these agents against molecules that are not nucleic acids, they may be important to understanding some recipient reactions as well as to explain any loss of cellular function recovery or survival.

Thanks.

DR. BRECHER: Thank you, Steve.

Questions? Comments?

[No response.]

DR. BRECHER: Amazing. Oh, John?

DR. PENNER: The genetic toxicity with the very small amount of mustard-like agents, it's a little difficult for me to think that this is much of an event that can occur with the quantities we're talking about compared to what we use in our chemotherapeutic approaches and the rather low incidence of mutations or carcinogenesis. Is that in your perspective?

DR. WAGNER: Well, it's kind of hard to measure in cancer patients who might reoccur whether their treatment with chemotherapeutic agents increases the risk of cancer. Some people believe so, but other people believe that reoccurrence occurs and metastasis in different cancers coming up. So I think it's a controversial question.

DR. PENNER: In addition, the pharmacists who handle most of our chemotherapeutic agents are certainly under the gun in exposure, and incidence has really not been revealed in this group, who, of course, take rather careful management of these products.

DR. WAGNER: Yes, I think that implementation of these sorts of agents might require that level of expertise and care to make sure that people aren't harmed. You have to have people that are knowledgeable and trained in order to reduce risk.

DR. GILCHER: You didn't remark about the possible advantage in emerging agents and potentially killing an emerging agent, and then potentially eliminating the need for additional testing. A good example would be West Nile.

DR. WAGNER: I didn't mention emerging agents. I didn't mention that that would enable you to not implement a new test. I think that's possible. And I think what's evolving is that we're devising this testing strategy for blood that right now is involving adding one test after another to blood. But it doesn't seem alternatively that it's not an impossibility that in the future there may be tests that can deal with a whole number of agents on things like chips and other things that we can only imagine.

And so you have these two competing processes that are trying to get at the same question, and right now screening is in place, which determines a risk factor of the blood supply. And so that's the hurdle that the inactivation--the peopl