Follow Ups | Post Followup | Back to Discussion Board | VegSource

From: TSS ()
Subject: TSE Update On: Leukocyte Reduction of Blood and Blood Components (Transcript) Posted: 8/26/2005, Meeting Date: 7/20/2005
Date: August 27, 2005 at 7:14 pm PST

DEPARTMENT OF HEALTH AND HUMAN SERVICES

FOOD AND DRUG ADMINISTRATION

CENTER FOR BIOLOGICS EVALUATION AND RESEARCH

UPDATE ON: LEUKOCYTE REDUCTION OF

BLOOD AND BLOOD COMPONENTS

PUBLIC WORKSHOP

Wednesday, July 20, 2005

8:00 a.m.

National Institutes of Health

Lister Hill Center Auditorium

Building 38A

8600 Rockville Pike

Rockville, Maryland 20852

snip...

We have the final session still to come.

Dr. Luisa Gregori is here. Dr. Gregori works with

Dr. Rohwer at the University of Maryland on some

very exciting prion technology, and she is going to

talk to us about New Technologies in Filtration,

Prion Reduction from Blood by Filtration.

Dr. Gregori.

Before I forget, please, you have

evaluation forms in your folder. We would

appreciate it if you could complete those. It

helps us to make better workshops in the future, so

please complete your evaluation.

New Technologies in Filtration

Prion Reduction from Blood by Filtration

DR. GREGORI: Thank you. Actually, Dr.

Rohwer sends his apologies, he could not be here

today. He really wanted to, but he had an

emergency, family emergency, so I am here to

327

present the data from our laboratory.

The presentation is divided in update on

TSE blood infectivity. We will talk about

leukoreduction, leukoreduction and PRDT technology

in the context of control of TSE pathogens. Then,

if I have time, I would like to spend a couple of

minutes just on talking about diagnostic still in

the context of TSE removal.

There is a large body of evidence at this

point that there is TSE infectivity in blood. This

evidence comes from experimental animals, from

natural TSE infections in diverse strains of the

TSE agents, and all this information are consistent

with, and predictive of, transfusion-transmission

of TSE in humans.

Unfortunately, we have seen already two

cases. This is a summary slide of the UK TMER

study. We obtained the data from Dr. Robert Will

in the UK as a personal communication, but he

allowed us to show this data.

What you see here, those are all the

recipients of blood from donors who later on

328

developed variant CJD. There are less than 50, I

think there are 49.

Those are the years since transfusion.

The blue dot corresponds to individuals who died,

and the red dot are the individuals who are still

alive.

This is the first case of variant CJD

transmission that is reported in the literature.

This was an incubation period of 7 1/2 years. This

is the second case, the heterozygote individual who

died without showing signs of variant CJD, but

later on PrP-res was found in the spleen of this

individual.

So, what this picture says, one way of

looking at this data is let's say that the

incubation time of variant CJD in blood is five

years. All these individuals here, they cannot be

counted in this calculation because they died

before five years, so they could be incubating the

disease, but they died too early.

If we just count the individuals who died

and after they had enough exposure to the

329

infectivity, longer than five years, then, there

are already 2 cases out of 7. That is a very high

percentage of transmission.

If we include also the living individuals,

that is still a very high transmission, 10 percent,

this is what we see. This is not what we see with

animals.

If we take 2 years as the incubation time,

that is still a very high transmission rate that we

find in variant CJD individuals. So, this could

mean that either the titer of the variant CJD blood

is higher than we have anticipated based on the

animal models that we use, the mouse or rodent

models, or the virulence of the variant CJD strain

is higher again than the mouse or hamster model.

The other new piece of information that

came out last year is this study by Hilton and

Coworkers in which they looked at the presence of

PrP-res, the marker for TSE infectivity in appendix

and tonsil. They started with a very high number

of samples. Assuming 100 percent ascertainment,

this study concluded that there are about 3,800

330

cases of incubating variant CJD in United Kingdom

at this time.

This is very different from the number

that we actually have now. There is about 150

cases that we know of, so there is something there.

Perhaps these individuals are incubating, but they

are symptomatic, so that means there is some silent

potential transmission that we have to take into

account that perhaps is going on.

Also, estimating that 10 percent of

individuals here donating blood, then, we have

currently 380 variant CJD infected blood donors in

the UK.

So, this is not to make the picture too

grim, I think in these days it needs to be looked

at for what the numbers say, and we have to try to

understand what they actually mean.

In our laboratory, we work with the

hamster model of blood-borne TSE infection. We

have done a lot of studies for several years. I

will show you some of the studies, some of the

results, but before I get to that, the results, I

331

just want to spend a couple minutes describing the

various ways in which we titer hamster tissue.

If the tissue is brain, for example, brain

has a lot of infectivity, so then we use the

endpoint dilution titration method, which is a

conventional method in which the inoculum is

diluted in a 10-fold serodilution that you see

here. At each dilution, a cohort of animals are

inoculated. This is a little syringe, IC

inoculation.

Then, after a year, we just look. The

gray animals are the ones that died, the yellow,

the ones that are still alive, and then we use the

Spearman-Karber method to calculate the titer.

That is a very conventional method to do. We can

do this with brain because brain has a very high

titer of infectivity.

Another way to look at it is also using

the incubation time in the dose response curve.

There is an inverse correlation between the titer

of infectivity inoculated into the animal and the

time that the animal takes to develop the disease.

332

The higher the titer, the shorter is the incubation

time.

This gives a correlation that is very

linear to some degree, and this can be predictive

of the titer of the infectivity. However, when we

go down to very low titer, high dilutions, then you

see that the infectivity, the incubation time

doesn't really correlate with infectivity. This is

very flat line. This is the dose response that

disappears limiting dilutions.

This here, each dot corresponds to an

animal that was inoculated with blood, and you can

see that starting from 150 days to 450 days, these

all are animals spread out almost randomly.

So, if we need to titer blood, we cannot

use the dose response curve, we cannot do the

serodilution, we have to use a different method,

and the method that we use, we call it dilution

titration, a method that was developed in our

laboratory and basically, with this method, we take

a 5 mL aliquot of the test material that we want to

titer. This 5 mL are inoculated, 50 microliters

333

each, into 100 animals.

At the end of the study, about a year and

a half later, we just count the number of animals

that are positive. Their number correspond

approximately to the number of infectious doses,

and then in this case we divide by the volume, and

it gives a rough estimate of what the titer is.

We can get a very precise estimate if this

titer is then corrected for the Poisson

distribution that takes into account the

probability that one animal gets more than one dose

of infectivity.

So, this type of titration is dictated

basically by the Poisson distribution because the

titer is so low that you can anticipate that 50

microliters either has one unit of infectivity or

no unit of infectivity.

So, all the infectivity studies done with

hamster blood or blood components are done this

way. We have done many of these studies over the

years, so this is a composite of some of those.

The red corresponds to the pool and you see that

334

they are all clustered around about 10 infectious

dose per mL although when we looked at individual

animals, we found the most variation. Those are

the blue symbols.

So, usually, we found around 10 infectious

dose per mL. We also looked at titer in blood of

animals incubating the disease at different times

during the incubation. We found that there is

infectivity early on before the clinical signs of

the disease.

We find it here after 80 days, 100 days,

and 120 days where the clinical manifestation of

disease occurs. This is some sort of type of linear

correlation. One might think that this is very low

level of infectivity.

Up here it is about less than 2 infectious

doses per mL, but if you consider that a unit of

blood has 450 mL, then, you can calculate this,

already something like 800 infectious dose per

unit, so it becomes a significant amount of

infectivity.

In terms of control of TSE pathogens, what

335

we have used at this point is sourcing and

deferral, but this is more a moving target as we

learn more about TSE infections, how they spread

and how to control them, it is a good option that

we have, but cannot be the only option that we

apply.

Screening, of course, it would be very

useful if we had one, but the screening at this

point is technically problematic for blood.

Inactivation, it is incompatible with blood

products, so we are left at the end with one

option, which is removal.

Removal is relatively low risk and is

technically possible, so we focused on removal, and

we think that this the best option that we have at

the moment.

Removal in a certain way, leukoreduction

is a form of removal of TSE infectivity. You are

expert here on leukoreduction. We only looked at

leukoreduction in terms of TSE removal.

Leukoreduction was implemented in Great Britain

several years ago.

336

The idea was, the rationale was the

infectivity is concentrating in buffy coat, PrPsc,

which is the marker for infectivity, and

infectivity itself was demonstrated in lymphoid

tissue and some TSE infections, and also there was

some involvement of B lymphocytes.

So, they were the scientific basis for

leukoreduction. Since then, other countries are

also implementing universal leukoreduction. Three

years ago we did the study with Health Canada. We

worked with Tony DuLeve [ph]. They had just

implemented the universal leukoreduction in the

country, and he came to us because he just wanted

to know whether leukoreduction actually removed TSE

infectivity in blood or not.

This is the study we did for Health

Canada, together with Health Canada. We prepared a

human size unit of hamster blood, that is about 140

hamsters, the blood from 140 hamsters was collected

in one bag. We filter it.

This is the Pall filter that is used

currently. I believe it is currently used in the

337

Canadian blood centers, and we collected and

leukoreduced the whole blood. Then, we titer the

blood pre- and post-leukoreduction.

We had to confirm and verify that the

leukofilter performed according to specifications

and that hamster blood behaved similar to human

blood, and so on. I am not going to go through

that because that work has been published already

last year, so I just go to the bottom line.

This one is the distribution of animals.

This is the day post-inoculation. That is the

number of animals, and this is just to emphasize

what I said earlier, there is no dose response

here. These animals came down from 150 days to 550

days, pretty randomly.

We calculated the titer

pre-leukofiltration, post-leukofiltration, and what

we find is that the leukofilter removed 42 percent,

that is, 58 percent of infectivity went through the

filter.

In different words, if infectivity in the

unit before leukofiltration was 4,500 ID, after

338

leukofiltration it was 2,600 infectious doses.

The way we concluded the study,

leukofiltration is necessary for TSE removal

because it targets a specific blood cell type that

will have to be removed anyway in terms of TSE

infectivity, but obviously, it is not sufficient to

reduce the risk of transmission by TSE by blood

transfusion.

So, we propose to look at alternative

methods together with leukofiltration. One method

that we feel very strong about, as I said, is the

removal. The advantage of removal is that it

removes also infectivity that cannot be detected by

diagnostics. Even if we have a diagnostic test,

there is still going to be a limit of detection for

that test, so a removal of infectivity.

If we have a device that removes

infectivity, will it remove also for dose unit that

escaped diagnostic. Also, for removal, we don't

need to differentiate, discriminate between the

abnormal form or the normal form of PrP, so that is

another advantage.

339

So, we think this is more comprehensive

and perhaps less costly, but I am not sure about

that.

In terms of removal, we start a

collaboration with a company called PRDT, and

actually, to be perfectly clear, Bob Rohwer is one

of the founders, Dave Hammond, the American Red

Cross, Ruben Carbonel at the University of North

Carolina. This company is a joint venture of

American Red Cross and Prometic [ph] Corporation,

and more recently Maco Pharma enter in partnership

with PRDT for the manufacture and marketing of the

final product.

For full disclosure, the study that I am

showing you that we did with PRDT was fully funded

by PRDT.

One of the things that interested us about

this company was that they were using combinatorial

peptide libraries to find ligands for specific

targets, so that technology was very appealing to

us, because we thought that we might be able to use

that to capture PrP-res or PrPsc, the target

340

protein for infectivity.

The peptide library can be actually this

many combination. We only looked at the

subpopulation on this combination, and they went

through different screening.

I am not going to go through many details

on this screening, but you can see that at each

step, there was a significant reduction. Now we

are down here to one ligand, and this ligand has

been tested now for removal of endogenous

infectivity.

The primary and secondary screening was

done in vitro, looking at PrP proteins with Western

blot. The tertiary or final screen is done with

infectivity. I showed you just to give an idea of

what the secondary screening was.

The 200 ligands that were found from the

primary screening were immobilized on resins and

then we tested those resins. We challenged the

resins with brain homogenate spiked into red cells.

Then, we looked at what was captured on the resin,

and the darker the signal, the better the resin was

341

because we must have captured a lot of PrP.

You can see here we tested with humans,

with hamster, scrapie, and with different forms of

mouse, mouse-derived TSE. From this study, a

handful of ligands turned out to be very

interesting, so we moved those ligands to what we

call the tertiary screening, and that involved

infectivity, brain spiked infectivity.

At this level, what we wanted to know is

to verify that those ligands performed well in

vitro, removing PrP-res. They actually also

removed the infectivity. The study that we did, we

started with one unit of human leukoreduced red

cells that were spiked with hamster scrapie brain

homogenate.

Each ligand was challenged in a series of

5. We collected the effluent from each step and we

used the incubation time as the measurement. We

don't usually use incubation time, we think it's

not very accurate, but we thought that for this

study, since we were looking at very dramatic

reductions on the order of 2, 3, 4 logs of

342

reduction, so we thought that it is a suitable

method, and the number of animals that they needed

for incubation time compared to endpoint dilution

titration is much less.

We also used an empiric endpoint using the

animal weight loss that we have used for the first

time, and it worked very well. For the incubation

time, we had to make a dose response curve, and I

showed you the results with that.

This is very briefly schematic to clarify

what you are going to see later. This is the

challenge, red cells in a homogenate passthrough,

ligand chromatographic column format. There were

actually five of them in line, but I showed you

only the results of the last one.

What we did, we looked at incubation time

of the challenge solution and incubation time of

the effluent, and we compared the two and looked at

the level of removal.

This is the dose response curve that I was

talking about earlier. We need to have a curve, so

that we can compare and determine the level of

343

removal from the incubation time only. The

challenge was 10

-3 dilution compared to brain. It

was a 0.1 percent scrapie brain homogenate.

This was serially diluted 10 times. We

inoculated a cohort of animals. We determined the

average incubation time, which is this blue symbol,

and this gave this curve that you see here.

The next step was to titer the effluent.

Actually, the challenge was 89 days

post-inoculation. Then, we started looking at the

effluent. This was our negative control. So, the

negative control showed that the animals inoculated

with this effluent from the resin 4 did not have a

decrease in the incubation time indicating that the

infectivity that was in the challenge was still

present in the flowthrough, in the effluent.

But then we found other resins that

performed differently. This one had 99 days

post-inoculation, 123, 140. To make a long story

short, the one that we focused the most were on

these 3 resins here, 8 or 3 or 1. They showed the

most reduction in infectivity titer because they

344

had the longest incubation time.

So, if we now look at this incubation time

here and report on the axis here, it looks like the

incubation time corresponds to the brain homogenate

at 10 -7, so we started from 10-3,

now we get 10-7, we

got 4 logs of removal with those 3 ligands.

So, the conclusions for this infectivity

study is that ligands showed around 4 logs of

removal of TSE infectivity in red cells. The

negative control did not remove infectivity, so it

was not a mechanical or some other artifact going

on during the chromatography.

The infectivity in the challenge was

200,000-fold higher than in 1 unit of infected

blood. This was necessary because we had to use

brain, so that is what we had to do, so obviously,

it was overloaded.

Also, what we found maybe wasn't--I forgot

to mention here--all the animals died. That means

not all infectivity was removed. So, there is some

infectivity that is still going through, and when

we calculated, it was 1 part per 10,000 unit of

345

infectivity. So, the filter removed 4 logs, but

the leftover infectivity that went through the

filters is on the order of 1 part per 10,000.

What is the implication for this for an

endogenous infectivity, we really don't know,

because it depends on how blood infectivity is

distributed. If the blood infectivity is

distributed in the same way as in brain, this 1 out

of 10,000 units, it doesn't correspond to a lot.

Actually, it will leave 0.5 infectivity per unit in

the blood after the device.

On the other hand, if blood is enriched in

that type of infectivity that did not get trapped

by the chromatography columns, then, it is more

problematic because we will have to then look if

this actually is an effective device.

So, the only way to distinguish these two

very different scenarios is to just do the

experiment. We had to look at endogenous

infectivity in blood as the proof of principle and

also to validate the relevance of our studies.

The endogenous infectivity study is still

346

ongoing. I just showed you what we have at this

point, but first I want to just take you, step by

step, how we get to do what we actually did.

The first thing that we had to do was to

choose the test material. We went through, there

are ideal test material and then there are

realistic test material. The ideal test material

that we really wanted to use obviously doesn't

exist. It would be 1 unit of variant CJD infected

blood from human patient.

Such material doesn't exist and even if it

existed, we don't really know how to measure

infectivity in that blood, so it would have been a

problem anyway. The second best choice will have

been variant CJD infected with blood from a

phylogenetic human mouse.

The problem with that is that the human

mouse, the humanized mouse doesn't seem to be

working very well with variant CJD, so that also

didn't work very well. Sheep blood, of course, we

could have studied with sheep blood. The advantage

is that we can produce 1 unit of sheep blood with

347

no problem.

The disadvantage would be where are we

going to titer it. Well, we can titer it in sheep,

but in the same host, that is possible to do, but

it will take more than five years before we know

the results, so that is a little too long.

We could have done it in the mouse,

transgenic sheep mouse. There are several

laboratories including our laboratory that has the

transgenic sheep mouse, but it has not been

characterized enough to know if there is enough

sensitivity to do these type of studies.

So, at the end, we ended up with a rodent

blood model. This is what we are very familiar.

We can inoculate the blood into the same host. It

takes a year and a half. This is a long time to

wait, but it is still better than five years, so we

settled with the hamster.

The second choice we had to make was the

challenge. We had three options: whole blood, red

cells, or plasma. What are we going to challenge

these ligands with?

348

We ended up, we decided to use whole

blood. That is because it represented the worst

case and contains all the infectivity that are

present in blood. Also, we had already experience

with leukoreduction. We already knew that there

would be enough infectivity in the leukoreduced

blood to be able to run this experiment.

We also looked at the interference of

protein in plasma. So, at the end, the model was

hamster-infected whole blood. The titer was done

with the limiting dilution method, and I just want

to clarify and point out here an important point,

that the demonstrable removal function on the

volume of the sample assay.

We usually test 5 mL. This is in 100

animals. This gives a limited detection of 0.2

infectious dose per mL. We could have started with

1 mL. There would be much less animals, but it

would give us much higher limits of detection.

So, we prefer to use, we are convinced

that this is the best option, the model that we

have, 100 animals is a good balance between a study

349

that would be too huge if we use more animals, and

a study that would be too small and won't give us a

very clear answer.

So, the endogenous infectivity was

basically done this way. We started with a PRDT

leader ligand. It was a scaledown. We are going

to do another experiment with the full 4 units of

blood, but this was a pre-prototype, so we just

want to have an idea whether these resins actually

remove endogenous infectivity.

Leukoreduce whole blood in the challenge.

It was the challenge, and we did limited dilution

titration on the challenge and on the effluent from

the PRDT devices. The study is ongoing. It is 87

percent completed, and it is going to be completed

in January 2006.

What we have now. I notice it is not a

month old, but not much happened in the past month.

This is the not leukoreduced blood. This is the

titer that we found, that we have extrapolated to

100 percent completion.

This is the leukoreduced whole blood and

350

this is the titer. The final flowthrough, we found

no animal that came down with the disease as of 330

days, and so we have reduced the level of

infectivity to the limit of detection on the assay,

which corresponds to about 1 log on infectivity

reduction.

I will come back to the leukoreduction

results in a moment, at the end of the

presentation, and this is what the filter looks

like. It has been prepared by Maco Pharma.

So, the summary at this point is that PRDT

has a lead resin defined. This ligand appears to

have a high affinity for prion protein, and it

works with brain of rodent and human with different

forms of TSE strains.

It works in vitro with red cells, whole

blood, and plasma. In the removal of infectivity,

we show 99.99 percent removal using brain that

infectivity corresponds to 4 logs, and we also

showed about 90 percent of removal, which is 1 log

of the infectivity. These ligands appear to have

no impact on blood.

351

I said I thought that this audience might

be interested in what we find here. I mentioned

the Health Canada study with the leukoreduction

earlier on. Those studies concluded that 42

percent of infectivity was removed by the

leukofilter.

In the study we did now with PRDT, we find

that the same leukofilter, I mean the same type of

leukofilter removed 71 percent of infectivity, not

42. So, we are trying to understand why there is

this difference.

Of course, it could be the filters behave

different because there were two different filters,

but that would be too obvious. I think there might

be something else going on, and we tried to address

this point by looking at some other data that we

have in our laboratory, and I am going to just

briefly mention to you.

When we spin blood to produce the three

fractions, and we titer plasma, buffy coat, and red

cells, we find that most infectivity is in buffy

coat, 45 percent of the total infectivity in whole

352

blood is in buffy coat.

That matched very well with the

infectivity that was removed by the leukofilter.

The leukofilter removes white cells. It was about

42 percent. This is most in white cells is 45

percent, so we thought that we got everything

clear.

So, there would be two pools of

infectivity in blood. One is in plasma, one is in

white cells. The leukofilter removes infectivity

in white cells. We just have to go after the

infectivity in plasma. We thought that the two

pools of infectivity could be separated in separate

compartments.

Later on we did another experiment in

which we took buffy coat from blood, separated by

centrifugation and washed the buffy coat

extensively. We just wash away with PBS, nothing

special. Then, we titered the washed buffy coat.

What we found is that after washing, most

of the infectivity was gone. As much as 80 percent

of infectivity was gone from buffy coat. So, our

353

conclusion at that point was that was, well, maybe

infectivity is not very tightly bound to white

cells, and it is easy to wash it off.

So, if that is the case, then, that could

explain the difference in leukoreduction. After

all, the white cells are trapped in the

leukofilter, there may be some factors that we

haven't identified yet, that removes more or less

infectivity from the white cells that are trapped

in the filter.

So, that is one possibility that we are at

this point considering. Also, I just want to point

out that we did the leukofiltration reduction only

twice, so it is not that we have a large number of

data in our laboratory that we can interpret, did

it two times, and we got rather different results.

Finally, just to mention how we see

removal. We see removal as a form of

concentration. Once infectivity is removed, we

have a device, we have a filter or something where

the infectivity is basically concentrated, and this

concentration may be useful for diagnostic

354

development.

So, there is a lot of interest in

diagnostic on the TSE field at the time. We think

that this interest should also be placed on looking

at ways of removing infectivity and use that

removal step as a concentration step.

We all know we need to concentrate PrP-rs

from blood if we want to have a diagnostic. So, we

are proposing that plan, too.

Thank you.

[Applause.]

DR. WILLIAMS: I think we have a little

time for questions.

PARTICIPANT: I am just curious. In your

washing experiments, do you retain the platelets

with the white cells, or do you wash the platelets

away with the plasma?

DR. GREGORI: In the buffy coat wash? The

platelets are mostly in the buffy coat, and they

stay in the buffy coat, so it's in the buffy coat,

yes, they are retained.

PARTICIPANT: Thank you.

355

DR. GREGORI: But I mean in terms of

infectivity, we have done a study a couple of years

ago, looking at platelets purified from hamster

blood, and we showed that there is no infectivity

in platelets, so that was one of the reasons why I

wasn't really looking at the platelets.

I know in the study also that I didn't

mention is during the wash, white cells were not

lysed. We were very careful not to do that, so we

think that the loss of infectivity is not because

we lysed white cells and infectivity was inside the

white cells and that is how it got lost. We think

the infectivity, the majority of it is localized on

the surface of the white cells.

PARTICIPANT: Would you expect that the

kinetics of removal would be the same in high-titer

material than low-titer material? You did most of

your experiments with high-titer material, so that

you could actually detect it at the end, but with

low-titer material, would the time for removal be

much longer, because the collision between the

prion protein and what is absorbing it would be

356

less frequent and limited by diffusion?

If so, and if that is the case, are the

conditions that you are using the filtration and

the time it takes for filtration going to be an

accurate estimate for what might occur actually in

blood before someone becomes symptomatic?

DR. GREGORI: The studies, in both cases,

the spiked study and endogenous study were done as

chromatography, so I think the diffusion might be

initially for we are doing it benchwise, but in

terms of, as chromatography, and the flow rate we

used, we think we gave enough time, contact time,

for the prion protein to bind the ligand that was

immobilized on the resin. We don't really know.

The other question, I am not sure I

remember what the second question was, but the

brain infectivity might be very, very different

from blood infectivity. Actually, it is different.

So, the extrapolation from the brain results, one

has to be very, very careful of how to interpret

the brain results experiment.

I think that is very important data

357

because it shows that the ligand in that

conformation has the capacity of binding that much

infectivity, 4 logs of infectivity, it is huge, but

in terms of how that translates in terms of removal

of infectivity in blood, it could be a very

different story.

At this point, it looks the infectivity in

blood is being removed by the ligand, but we still

have six months, and six months is a long time, and

we can get even one animal coming down and the

whole experiment is not useful, I mean it is not

working then.

So, we hope that we are going to see the

same situation six months from now. We have to

wait. I am not sure I am answering your second

question because I don't remember your second

question anymore.

PARTICIPANT: Well, basically, what I am

thinking of is that I guess the objective is to

remove prion protein in individuals who show no

symptoms, and so you would expect that the titers

would be moderate or low.

358

I guess the concern I have is you have a

process that is being driven by affinity, and that

is dependent on the number of collisions of the

prion protein with its ligand to take it out. If

the ligand is very rare, that is not going to occur

very quickly, for example, at a rate of a

leukoreduction filter, so you may not get down to a

level that would not necessarily transmit BSE or

infectivity.

DR. GREGORI: I guess then we will have to

look at different flow rate for this type of

filtration. Also, the final device, the way Maco

Pharma--it is not going to be like a

chromatographic column, it is going to be a large

surface area. I guess it would be not the best

solution.

I don't really know how they are going to

solve that issue. I understand what you are saying

and I think with flow rate, maybe it is just a

matter of sending it slower, but we haven't done

that study yet.

DR. EPSTEIN: Is there any feasibility to

359

make a concentrate of the hamster blood to use as a

higher titer inoculum for a clearance experiment?

Can you use the affinity ligand to generate a

concentrate, or can you use centrifugal method to

generate a concentrate?

You might need hundreds of hamsters, but

if you were able to spike hamster blood, not with

pooling units from infected hamsters, but with

concentrates from a larger number of hamsters, you

might be able to raise the titer inoculum at least

1, maybe 2 logs, because then you would have a more

convincing experiment.

DR. GREGORI: That would be nice, but I

don't know a way to concentrate blood without

incurring so many other problems. One way of

concentrating infectivity in blood is by preparing

buffy coat. Buffy coat has a tenth of a volume of

blood and has half of the infectivity of whole

blood, but the buffy coat, because the device that

we are using will be placed after the leukofilter,

so obviously, we can't make a buffy coat. So, I

don't know if there is a way to concentrate blood.

360

What I was mentioning at the end is that

we could use the concentrated infectivity in the

PrP-res presumably in the device and use it as a

diagnostic concentration step. Whether that could

be used in some other purposes actually, I never

thought about it. Maybe it is possible.

DR. BIANCO: Can you elute your material

from the resin?

DR. GREGORI: Yes, the method that we use

at this point is a denaturation method, so that

will work as a diagnostic, but it will not work for

anything else. We didn't really evaluate that, but

one could imagine a milder way of eluting out and

trying to spike into blood. I never thought about

it.

DR. WILLIAMS: Thank you very much, Luisa,

very nice work.

[Applause.]

DR. WILLIAMS: Our final presentation will

be by Dr. Jerry Ortolano with Pall Corporation, and

he is going to describe the Pall Leukotrap Affinity

Filtration System.

361

Pall Leukotrap Affinity Filtration System

DR. ORTOLANO: The best part of going last

is I get to say anything I want.

I do want to make a comment about the cost

of universal leukocyte reduction because it really

prompted our interest in prion removal. One of the

things that we became very well aware of is the

fact that ULR has been estimated to cost the

American healthcare system some $600 million for

full implementation.

It also turns out that about 75 percent of

the blood product is already paid for right now in

one way or another with adoption of universal or

nearly universal leukocyte reduction, leaving about

$150 million to be spent.

If you divide that $150 million by 6,000

hospitals, we are talking about $25,000 per

hospital, which in the grand scheme of things seems

to be a pretty small price to pay.

That notwithstanding, we are constantly

looking towards ways of improving leukocyte

reduction and adding attributes to it to kind of

362

improve or increase the value of leukoreduction,

and our pursuits with respect to prion removal are

really spirited by that thought.

So, I would like to share now with you

some results of our prion removal capability, which

was targeted for a leukocyte reduction filter, but

as you will see when we talk about this technology

right now, we have broken down the project into two

phases.

The first phase was really for a more

immediate use in Europe where the pressing need for

prion removal was higher, and for that, we have

really characterized this filter for us with an

already leukoreduced blood product.

The second phase of the research, I am not

going to speak about today, is really encompassing

leukocyte reduction with prion removal. Suffice to

say that we have made some sufficient progress on

that, as well, and we expect that to be coming up

pretty shortly.

Dr. Gregori has really provided a

wonderful background for my presentation, which

363

means I can spend a lot less time and you can get

out a little bit earlier than originally planned.

The Pall Leukotrap Affinity Prion

Reduction Filter really targets all prions, both

cell and non-cell associated. Even in a

leukoreduced blood product, there are still some

cells remaining, and with this technology, since it

was based on integrating prion removal with

leukoreduction in its early concept, still has

leukoreduction capability, so we also effect

leukoreduction on top of the removal of

non-leukocyte associated prions.

Surface modification technology does not

impact red cell stability. We did some survival

studies, 42-day storage studies, and demonstrated

that to be true, and the filtration is a commonly

used process, as you all well know, so integrating

the two is ideal for use in the American blood

centers.

Dr. Gregori has provided you with some

information about the kinds of testing that can be

done, and I would just like to summarize for you

364

the advantages of each. The Western blot is the

most economical approach for screening. Basically,

what that involves is taking hamsters that have

developed disease, taking the brains of those

hamsters and preparing a 10 percent homogenate, and

then spiking an aliquot of the 10 percent

homogenate in the blood product of choice.

In this case, we talk largely about

previously leukoreduced human blood. You can then

filter that blood product and then try to measure a

pre-filtration aliquot by Western blot compared to

a post-filtration aliquot to get some idea of the

magnitude of prion removal.

I am happy to admit that spiking prion

brain homogenate into blood was probably not going

to be the same as blood that has become infected by

a more natural route, but nonetheless, there is

some value in doing the Western blot with respect

to screening.

The exogenous bioassay really is an

extension of kind of the Western blot format. What

it involves is taking very homogenate, putting it

365

into blood, filtering the blood, and then taking an

aliquot of pre-filtration product, as well as the

post-filtration product, and then serially diluting

those, and then injecting those intracerebrally

into hamsters, so that we can get some idea of the

proportion of animals that die over a fixed period

of time to see at what dilution animals do die in

both the pre-filtered sample compared to the

post-filtered sample.

This will give us some idea of the log

reduction overall effected or accomplished by

filtration.

The endogenous infectivity study, which

Dr. Gregori explained so well, is also really a

very valuable tool in that now the nature of the

prion is much more closely aligned with what we

might expect to see in an asymptomatic or

symptomatic blood donor. Obviously, blood donors

who are symptomatic wouldn't be donors, but you are

getting very close to the very low levels of prion

that you would expect to see in those blood

products.

366

If it isn't obvious to you already, please

do appreciate that the concentration of pathogenic

prion present in a brain homogenate is about 10

9

infectious unit per milliliter, and you can

contrast that with what Dr. Gregori has shown in a

publication about 10 infectious units in blood when

endogenous infectivity has elaborated.

Here is an example of the Western blot

studies that we have performed. Again, this is

scrapie-infected hamster brain homogenate spiked

into human blood, filtered, and then determined log

removal using the Western blot.

Keep in mind now that, as Steve has

pointed out, this is a very high concentration that

we spite into the blood. There are some

limitations with respect to understanding these

data or interpreting these data, not the least of

which is the fact that we may very well be

saturating the leukocyte reduction capability by

virtue of having such a high concentration of prion

in that blood.

That is a requirement, however, because

367

the limit of sensitivity of the Western blot is

actually not very low at all, it is not very

sensitive at all, and without sophisticated or I

should say the more technologically advanced

approach used recently, effective 2002, the limit

of resolution capability was about 2 logs, and now

we can get it to be a little bit better than 3

logs>

What you see in the panel on the left is

leukocyte filtration. There is a pre-filtration

sample Western blot and a post-filtration sample

Western blot, and no surprises here, 42 percent

reduction in infectivity, as Dr. Gregori has

published, barely shows any change in the Western

blot, both pre- and post-filtration with standard

leukocyte reduction technology.

If we look at the prototype design for a

prion removal filtration technology, again, this

was done when we didn't have as sophisticated a

Western blot assay as we have now, we see that

removal is at least 2 log, if not greater. That is

attested to by virtue of the fact that as you look

368

at the Western blot on the right, in the middle

panel, the post-filtration sample doesn't show any

image at all on the gel.

In terms of the final filtration design,

here we have just completed a series of studies

totaling 48 samples, which looked at Western blot

data and calculate removal to be about 2.9 log

overall.

In our bioassay in hamsters, again, just

to refresh your memory, we are spiking now the

brain homogenate into blood and then taking a

pre-filtered aliquot, serially diluting it along

with a post-filtration aliquot, and serially

diluting that, and injecting all of those serial

dilutions into naive hampsters, and then looking

for disease.

You can see here that with respect to the

concentrate with the dilution required to start to

impact on the prevention of manifestation of

disease, we are at about 10

-9. If you compare that

with post-filtration, it is about 10

-5, so the total

infectious prions removed here in this example is

369

about 3.7 log. Again, it is not infectivity, it's

an exogenous model.

Here, in our infectivity studies, again,

with this particular prototypical filter, we have

scrapie-infected hamster brain homogenate injected

into hamsters, 100 hamsters were used in this case.

Then, the blood was pooled after manifestation of

disease was apparent. Blood was pooled, and then

an aliquot of the pre-filtration sample was

injected into a series of animals, as well as

post-filtration aliquots. These are also

intracerebral injections into animals.

The Western blots, 3 of the animals that

we see in a pre-filtration sample is shown below in

the panel on the right. In total, we had 6 out of

43 hamsters that were infected, and that contrasted

with zero out of 38 in the post-filtration sample,

giving a p-value of 0.0384, which does establish a

level of statistical significance.

So, in summary, with respect to the

various types of assays, our prototype and the

Western blot gave us greater than 2 log removal, or

370

equal to or greater than 2 log removal. The final

design gave us 2.9 log reduction.

With respect to the exogenous bioassay, we

had a 3.7 log reduction with an N of 1. The final

design, the study is ongoing and is planned to end

in December of this year.

With respect to our endogenous infectivity

study, the prototype again showed significant

reduction in infectivity, but the final design data

will not be available until the spring of 06.

Some additional studies you may find of

interest involve taking hamster brain homogenate,

injecting them intracerebrally into normal

hamsters, and then taking the blood from those

hamsters, pool them, separate out the components,

and filter the packed red cells, and then subject

that to filtration.

As you might expect, these animals have a

very low level of pathogenic prion in the

circulation. We were at least able to detect a

little bit of it in the panel on the left, so

before filtration, you do see some signs there of

371

pathogenic prion on the Western blot, and PrP-res

post-filtration is clearly not evident on the

Western blot, but if you take that filter and

actually backflush it with a small volume, you can

recover and concentrate pathogenic prion, so much

like Dr. Gregori's model, our system appears to be

working pretty much the same way.

It is possible to use this filtration

technology as a way to concentrate sample, so that

we can utilize existing assays for pathogenic

prion, and actually increase the overall

sensitivity of the system.

In terms of safety studies, we actually

CE-marked this in Europe in May of this year, and

we had to subject our filter to a wide variety of

tests. All those tests are standard tests for

filtration products and all have passed.

In summary, the prototypical filters show

a 3.7 log in the bioassay, and the Western blot

data showed equal to or greater than 2 log. The

final filter showed a Western blot of 2.9 log, so

we expect that the bioassay data should be actually

372

higher than 3.7 log.

The quality of the blood cells is

unaffected by prion reduction filtration. The

safety study showed no cause for concern, and the

24-hour single and double isotope red cell survival

data is unaffected by filtration.

On top of that, the residual white blood

cells were further reduced by filtration to levels

less than 1 x 10

5, such that 98 percent of the time

we are able to ascribe that this occurs at 95

percent confidence, which is significantly lower

than the current standard for leukoreduced blood.

I would be happy to entertain any

questions you might have at this time, and I think

I gave that in record time.

[Applause.]

PARTICIPANT: I am just curious. Could

you tell us something about what the mechanism of

removal of the prions from the blood is?

DR. ORTOLANO: How much of that broaches

outside the area of being proprietary, I don't

know, but I will tell you this much. It is well

373

known in the literature that lots of things can tie

to prions. We tested early on a wide variety of

agents that are known to be able to latch on to

prions.

Some of that information was useful and

some of it led us in a certain direction, but the

direction ultimately wound up in that we have now,

not a ligand, not an affinity ligand, not a

protein, but rather a physical/chemical separations

technology.

So, if you look at the physical properties

of prion, that is basically where we honed our

work, so I will let you think about what that

means.

PARTICIPANT: So, is the mechanism

therefore somewhat nonspecific?

DR. ORTOLANO: Yes. It removes pathogenic

prion and it removes PrPsc. We have checked it for

a wide variety of proteins and found that with the

exception of factor IXA, it doesn't appear to

remove very many other proteins. As a matter of

fact, IXA might be a good surrogate QC for removal

374

of pathogenic prion.

PARTICIPANT: In the short time you have

on the market in Europe, what is the feedback you

have gotten?

DR. ORTOLANO: Well, Europe does things a

little bit differently than we do. We have the FDA

to keep us honest, and they go through a different

process. They go through a process of internal

validation. So, you can expect to put a product in

Europe and not expect a year to go by before you

actually sell anything. They undergo study, and

the Prion Working Group, which is a group of four

nations, the scientists from four countries in

Europe are actually convening now and reviewing our

data, much like the FDA would, and they are

prepared to initiate their own trials probably the

end of this year.

PARTICIPANT: What is your blood loss per

unit?

DR. ORTOLANO: It is about 40 mL, and if

you add that on top of the blood loss you get with

a standard leukoreduction product, it is excessive

375

in the sense that it is just not something you are

accustomed to.

If we were to integrate it together with

leukofiltration as we plan to, particularly if it

needs to be released in the United States, you

know, if it's an issue in the United States and it

becomes important enough, we can pump up the

developmental process I am sure, you know, making

it a higher priority, and get that technology

integrated into one single filter, so your loss

would be no more than what you would see now with a

standard leukofiltration.

PARTICIPANT: You showed a Western blot in

which you found protease-resistant,

protease-resistant prion protein in serum of

hamsters, if I understood correctly.

DR. ORTOLANO: Yes.

PARTICIPANT: Nobody else has been able to

do that.

DR. ORTOLANO: I know, and I don't know

what to make of that either. We have been

questioned by everyone, Neil Cashman [ph] and

376

others, and I mean the data are what they are.

Could it be a contaminant? I don't know honestly,

but it's what we got.

You know, a lot of these studies, we have

an N of 1, pooling hundreds of hamsters together to

give you sufficient sample size is difficult, so

that is why we look at the Western blot as an

alternative.

PARTICIPANT: Conventional Western blot or

is it some sort of--

DR. ORTOLANO: This was the--I am blanking

on the paper--it's the enhanced phosphotungstic

acid precipitation

PARTICIPANT: Jerry, were these stained

with Coomassie or Silver stain?

DR. ORTOLANO: The readings were done by

densitometric readings on Western blot.

DR. BIANCO: What I heard from Dr. Gregori

and you is that we are in the range of 4 logs. Dr.

Jay Epstein just left the room. I wanted him to

hear the question.

DR. ORTOLANO: He will read about it.

377

DR. BIANCO: Is that enough? Can we get

rid of the deferral of donors, geographic deferrals

that gives us only 1 log?

DR. ORTOLANO: You know donor deferral is

not going to work, you know that. But we asked the

experts and we did convene a panel of experts in

this field, and they seem to think 3 to 4 log will

be good. Whether or not it is going to work, I

don't think they have a crystal ball, I am not

sure. But that is what we are targeting.

DR. GREGORI: The 3 to 4 logs of

infectivity that you said the filter removes, that

is not endogenous infectivity.

DR. ORTOLANO: I understand.

DR. GREGORI: That is brain infectivity.

DR. ORTOLANO: Yes, and we did an

endogenous infectivity study, and we showed that 6

animals out of 43 came down with disease, and none

out of 38 came down with disease. The question is

how much was that. We can't answer that with that

assay.

But does it attenuate infectivity? Yes,

378

it appears to.

DR. GREGORI: Yes. So, in your

calculations, maybe I missed it, what did you say

the calculated log of removal from endogenous

infected hamster red cells?

DR. ORTOLANO: I didn't say.

DR. GREGORI: Oh, you didn't say.

DR. ORTOLANO: No, because we haven't

calculated it.

DR. GREGORI: You could calculate it based

on the data you have.

DR. ORTOLANO: Yes. That was a

prototypical experiment, an experiment of the

prototype filter. The next series of experiments

is being done in 600 animals, 200 per group per

filter.

PARTICIPANT: What level of log reduction

do you think ultimately you will need to prevent

infection, human infection?

DR. ORTOLANO: What level of log reduction

do we need to prevent? As I mentioned before, I am

not the expert on this, but when we polled the

379

experts, they seemed to think that 3 to 4 log is

acceptable. I hear Dr. Gregori saying how do you

know. I don't know.

DR. GREGORI: The experts, I think they

will agree that whatever you can do, the best you

can do, they will work around that, because

basically, the infectivity in brain, you can huge

amount of infectivity removal, 4 logs, even more if

you work hard to prepare a filter that will

specifically do that, but who wants to remove

infectivity from brain?

From blood, you cannot demonstrate 4 logs

of removal, it is just impossible, because what you

can demonstrate is if you inoculate 5 mL of blood,

and none of the animals come down with the scrapie,

then, you can say that there are zero or less than

1 infectious dose in 5 mL, that is all you can say.

You cannot say there is no infection in that unit

because you didn't measure the whole unit.

If you want to titer the whole unit, you

need thousands and thousands of animals.

DR. ORTOLANO: It's impractical.

380

DR. GREGORI: So, the maximum you can

actually show is I think it's a log and a half, 2

logs at the most. That is if there is no

infectivity in the flowthrough of your device.

So, the agencies in Europe, and I am sure

here, too, they are perfectly aware of this

limitation. That is on what we can do, and they

accept it.

DR. ORTOLANO: Thank you very much.

DR. WILLIAMS: Thank you, Dr. Ortolano.

[Applause.]

DR. WILLIAMS: I think before we get off

on a discussion of geographic deferrals, we will

call it a day. Thanks. You have been a great

audience, great speakers, and I want to thank again

Rhonda Dawson, Susan Zula [ph], Marty Edwards, and

our unnamed audiovisual support person who did a

great job.

It has been a good workshop. Thank you.

[Whereupon, at 5:37 p.m., the workshop

concluded.]

- - -

http://www.fda.gov/cber/minutes/leuko072005t.pdf

TSS

• PDF]Freas, William TSS SUBMISSION

Follow Ups:

Post a Followup