UNITED STATES OF AMERICA

P‑R‑O‑C‑E‑E‑D‑I‑N‑G‑S

(1:35 p.m.)

CALL TO ORDER

EXECUTIVE SECRETARY KRAUSE: We are ready

to begin the afternoon portion of the General and

Plastic Surgery Devices meeting. Before I turn the

meeting back over to Dr. Chang, I would just like to

remind everybody here that you are requested to sign

in on the attendance sheets, which are outside the

door, if you didn't do so this morning.

Also, there is information out at that

table that is right outside the door, such as an

agenda or a roster of the panel. There is also

information regarding today's meeting. There is also

information out there that you can obtain on methods

of finding out how to find out about future meetings

and things like that.

Now that we all seem settled, I am going

to turn the meeting back over to Dr. Chang. Wait. We

do have a public testimony session before you start.

So could you just clear that table until the public

testifiers are done? Thank you. I appreciate that.

Okay. Now let's start the afternoon and

go back to Dr. Chang. Thank you.

ACTING CHAIRPERSON CHANG: Good afternoon.

We will now proceed with the next open public hearing

session of this meeting. All persons addressing the

panel are asked to speak clearly into the microphone

as the transcriptionist is dependent on this means of

providing an accurate record of this meeting.

As mentioned this morning, both the Food

and Drug Administration and public believe in a

transparent process for information gathering and

decision‑making to ensure such transparency at the

open public hearing of the advisory committee meeting.

FDA believes that it is important to

understand the context of an individual's

presentation. For this reason, the FDA encourages

you, our public hearing speaker, at the beginning of

your written or oral statement to advise the committee

of any financial relationship that you may have with

the sponsor; its product; and, if known, its direct

competitors. This may include sponsor's payment of

your travel, lodging, or other expenses in connection

with your attendance at this meeting.

Likewise, FDA encourages you at the

beginning of your statement to advise the committee if

you do not have such financial relationships. If you

choose not to address this issue, it will not preclude

you from speaking.

We have one individual who has notified

the FDA of their intent to testify during this open

public session. Is Dr. Diana Zuckerman or her

representative available to speak this afternoon?

OPEN PUBLIC COMMENTS

MS. SANTORO: Good afternoon. My name is

Elizabeth Santoro. I'm a health policy fellow at the

National Center for Policy Research for Women and

Families.

I will be reading the testimony of our

president, Dr. Diana Zuckerman, who, regretfully,

could not be here today. Our center is a think tank

that translates scientific research findings into

meaningful information for the public. We use that

research information to advocate for policies that

benefit the health and safety of children, women, and

families. We have several concerns about Hylaform.

As I stated at this morning's session, it

is difficult for us to testify before the data are

presented. We are basing this testimony on the

information that was made available on the FDA Web

site yesterday. Our concern about Hylaform is the

lack of data for African Americans and Asian

Americans. Only three of the patients are African

American, and only five are Asian Americans.

Our concerns were addressed with the

National Medical Association and to the FDA

commissioner a few months ago. Research clearly shows

that African Americans are more likely to produce

keloids and can respond differently to procedures

involving the skin. In addition, African Americans

are more likely to develop autoimmune diseases than

white women. The company has not studied a reasonable

number of African Americans or Asian Americans to

approve the product for those populations.

It is not appropriate to require studies

of minority populations on a post‑market basis since

the FDA does not have the authority to enforce such

requirements. The company should be required to do

the studies before the product is approved.

It is also inappropriate to label the

product "For whites only." This would be acceptable

if a product were found to be safe for whites but

unsafe for other racial or ethnic groups, but such a

label is not an appropriate way around a sponsor's

failure to conduct research on people of color.

And, of course, if the product were

approved, it would probably be used off label for

people of color. And that could potentially be

dangerous. Research is needed. It won't take long to

do it. And it should be done.

We are also concerned about the sample

size. The sample starts with only 133 people, and

only 123 are still in the study after 12 weeks. Since

this is a cosmetic procedure that is likely to be used

by hundreds of thousands, perhaps millions of people,

the product should be tested on a larger sample to

determine if there are rare adverse reactions that are

serious enough to consider before approval.

The study only lasted for 12 weeks, which

is a major concern. According to the FDA slides,

safety data, immunological responses were at four

weeks. This obviously is too short a time to prove

whether this product is safe.

Since this product doesn't last long,

women would need to undergo the procedure multiple

times. It is clear from published reports that women

who have a good outcome the first or second time they

use this product may have serious adverse reactions

after the third or later procedure. This needs to be

studied before approval since it is clear that their

product will be used more than once or twice.

The sponsor also excluded women who had

procedures in the previous six months. This is not

how the product will be used in the real world. And,

again, it raises safety concerns.

A major shortcoming of this research is

the high adverse reaction rate. I don't consider this

necessarily a problem of the product but, rather, of

the study design.

It is not helpful for the sponsor to

evaluate adverse reactions in a way that almost all

the women using this product or Zyplast all have

adverse reactions.

The additions of serious or severe adverse

reactions is helpful, but it seems likely that there

is a continuum of problems between what is listed as

an adverse reaction, which almost everybody

experiences, and what is listed as a serious adverse

reaction, which almost nobody experiences.

The standard needs to be set in a way that

is more meaningful. For example, it should measure

adverse reactions lasting more than a day or two.

That would enable the FDA to determine how safe this

product is compared to other products.

I don't think the FDA should be approving

a cosmetic product where 88 percent of the patients

have adverse reactions. Either the FDA should reject

this product or require the company to provide a more

meaningful measure of adverse reactions.

In conclusion, I have even more concerns

about this product than about Restylane. According to

the company's own data, this product is not

necessarily better than the comparison product Zyplast

and apparently may not last as long. For that reason,

I believe rushing this product to market without

gathering the additional data listed above is

unwarranted.

Thank you.

ACTING CHAIRPERSON CHANG: Thank you.

Is there anyone from the general public

who wishes to make a public comment limited to five

minutes?

(No response.)

ACTING CHAIRPERSON CHANG: Hearing none,

we would like to at this time invite Genzyme

Corporation to present their PMA information.

APPLICANT PRESENTATION, GENZYME CORPORATION,

HYLAFORM

INTRODUCTION

MS. STEWART: Good afternoon, Dr. Chang,

members of the panel and FDA. My name is Susan

Stewart. I'm a vice president of regulatory affairs

at Genzyme Corporation. We are happy to be here today

to present Hylaform to you. And I would like to start

with a brief introduction and overview.

Dr. Burns will begin our presentations

with a description of our device and a presentation on

the preclinical studies in support of the safety of

Hylaform.

Dr. Polisson will then summarize our

rationale, clinical history, and discuss and describe

the study that we are later going to present results

on with Dr. Holmdahl that shows we are comparable to

Zyplast, the standard of care.

We also have with us a number of

colleagues and invited experts that are available to

assist us in answering questions as they arise.

They're listed here on these two slides. I

specifically would like to point out that we have with

us today principal investigator for the study Dr.

Ellen Gendler as well as Dr. Gary Monheit and Dr.

Leslie Baumann, also investigators in our study.

Now I would like to give some background

on Hylaform. Hylaform is a clear, colorless hylan B

gel. Hylan B is composed of cross‑linked HA, which is

a naturally occurring polysaccharide found in all

human tissue, including skin, and is found to be

identical across all species.

Hylan B used in Hylaform is also used in

our product Synvisc, which is a visco supplement used

to treat osteoarthritis. We sold approximately nine

million doses of Synvisc around the world.

This is the indication that we are seeking

for Hylaform. I would like to point out it is also

the indication for which we have been granted approval

in about 30 countries around the world. I would like

to read it for you now, "Hylaform is injected into the

mid to deep dermis for correction of soft tissue

contour deficiencies, such as wrinkles or acne scars."

Outside the U.S., Hylaform is sold as a

product family. They include Hylaform, Hylaform Plus,

and Hylaform Fineline. Hylan B is the main component

to these three products. The only difference in these

dermal fillers is that they have been processed to

optimize delivery through a range of needle sizes.

These were produced in reaction 2 position preference

and surgical practice.

I would like to be clear that although the

subject of the PMA is Hylaform, we do refer to

Hylaform Plus later in the presentation. The

commercial history of Hylaform begins with our CE mark

in 1995, which allowed us to begin launching product

in all the European Union countries in 1996. Since

then, Hylaform has been used to treat a variety of

wrinkles, including the nasolabial fold, those around

the lip and periorbital areas.

No pre‑implant skin test has been required

when using this product. And we report a worldwide

overall post‑market adverse event rate of about 0.1

percent.

This time line illustrates our current

status in the United States. Please note that the

sponsor of this PMA is Genzyme and that our worldwide

marketing partner is Inamed Corporation.

Initial clinical studies in the U.S. were

conducted under an IDE sponsored by Biomatrix,

Incorporated of Ridgefield, New Jersey. Safety and

efficacy of Hylaform were evaluated in a 216‑patient

open label single arm study. A PMA was later

submitted, which was then withdrawn by Biomatrix due

to issues concerning the clinical design. Please note

that this study did form the basis of our approvals

outside the United States.

In 2000, Genzyme Corporation acquired

Biomatrix. And by 2000, we had begun new studies with

Hylaform using a protocol that we believe addressed

the issues raised in the original PMA.

Our PMA was then submitted and has been

reviewed as a modular PMA. We have been submitting

sections over the past year, with our final clinical

module submitted in August of 2003.

Now I would like to ask Dr. Burns to begin

the preclinical presentation. Thank you.

DR. BURNS: Thank you and good afternoon.

PRECLINICAL REVIEW

DR. BURNS: I am Jim Burns. I am senior

vice president for biomaterials research at Genzyme.

I am going to provide some brief background on

Hylaform and a brief description of our preclinical

studies that preceded the clinical trials that you

will be hearing about.

As you have heard, Hylaform is a

hyaluronan‑based gel that is injected into the dermal

tissue to provide space‑occupying augmentation of the

extracellular matrix, thereby providing a temporary

correction of skin contour deficiencies, such as

folds, wrinkles, as well as acne scars.

By the way, during this presentation and

subsequent presentations, you will hear hyaluronate,

hyaluronic acid, hyaluronan. And for simplicity's

sake, we will very often just call it HA.

HA is particularly well‑suited for this

application of dermal augmentation because it is

already an important component of extracellular

matrix. And within the extracellular matrix, it is in

a proteoglycan complex called aggrecan.

You can see on this slide the HA molecule,

which is a slender ribbon here, which is associated

with protein as well as other glycocyamine glycans,

chondroitin sulfate, and keratin sulfate. If you

looked at an electron micrograph of this complex, you

would see a ribbon of HA that is decorated with

comb‑like structures shown here.

This structure is very important for wound

repair. It is important for allowing cells to migrate

into a dermal wound. In this hylan‑rich milieu, it

also helps regulate the fluid retention. And it keeps

the tissue hydrated.

When it is not associated with this

proteoglycan complex, relatively dilute solutions, it

can form fairly slippery or lubricous solutions, such

as occur in synovial fluid for joint protection. In

high concentrations, it will form a gel‑like structure

that is useful for mechanical support to occupy space

as in a vitreous humor of the eye.

Not surprisingly, it's found its way into

a number of products that are available in the U.S. as

well as throughout the world. Here are some examples

of those.

HA is a primary component of a product for

adhesion prevention and gynecologic as well as

abdominal surgery. It's also in a class of products

shown here for prevention or dealing with the pain of

osteoarthritis.

In this case, the viscoelastic solutions

are injected intra‑articularly in the knee. They also

have found their way into viscoelastic solutions that

are used in cataract and intraocular lens surgery.

Actually, an important point of this slide

I want to point out is that HA is a very interesting

molecule in the sense that it is biogenetically

conserved, that structurally and chemically, it is

identical, whether it comes from the protective cell

coat of a bacterium from an avian source or from our

own synovial fluid. The differences among HA

preparations come from purification processes from the

source, from low molecular weight or low‑level

contaminants or molecular weight of the HA.

So let me briefly describe some of the

design characteristics that we wanted to have in a

dermal filler. First, we want to balance durability

with safety. Any biomaterial that you implant in the

body is going to elicit a reaction. And we wanted to

ensure that we could provide temporary augmentation

without generating an untoward response.

So we feel we have done that with this

particular formulation. So our material provides

temporary correction and biorebsorbs with minimal

tissue reaction. We feel that it possesses acceptable

viscoelastic properties so that it can withstand some

of the forces that would be present at the site of

injection that will remain at the injection site for

an appropriate period of time and also be easily

injectable.

Natural HA will not meet these criteria

because it will resorb and clear from the site fairly

quickly. So we get to this product by starting with

avian tissue and treating that tissue in situ to

increase the molecular weight of the HA through a

formaldehyde treatment.

This is still water‑soluble HA of high

molecular weight of about four to six million. We

then cross‑link that further to form a gel. So we

have hylan A, which comes from the avian tissue

through out modification process, and then hylan B gel

is the component that goes into Hylaform.

This is just going to illustrate the

cross‑linking with hyalanosulphone. So we see two

parallel chains of hyaluronic acid on the left in

yellow. We see the basic monomeric unit of HA, which

is glucuronic acid or neutral pH sodium glucouron and

acetyl glycocyamine. Under basic conditions and only

under basic conditions divinyl sulphone will react

with free hydroxyl groups on the HA chains to form a

gel.

We have conducted a number of preclinical

studies to characterize the safety of this material.

These studies were consistent with the ISO 10993

standards as well as the FDA guidance document for an

implant that will reside for greater than 30 days.

We conducted 20 GLP and 7 non‑GLP

preclinical studies, which included cytotoxicity,

irritation, intercutaneous as well as subcutaneous,

acute systemic toxicity, subchronic as well as chronic

toxicity going out to one year, a number of

genotoxicity studies, sensitization and immunogenicity

implantation in the rabbit muscle implant test,

pyrogenicity, hemocompatibility, reproduction, as well

as clearance.

We found in these studies that Hylan B was

safe and nontoxic at amounts that were equivalent to

or in great excess to that which was studied in the

human clinical trial that you will be hearing about.

So, in summary, we had developed an

HA‑based formulation that we feel is consistent with

our design inputs for a temporary dermal filler. Our

preclinical studies show that Hylan B did not elicit

a significant biological response, indicating that it

made sense to go ahead and evaluate this material in

human clinical trials.

I would now like to ask Dr. Dick Polisson

to describe the design of our trial.

DR. POLISSON: Thank you, Jim.

CLINICAL STUDY DESIGN

DR. POLISSON: Ladies and gentlemen, my

name is Richard Polisson. I am senior vice president

of clinical research at Genzyme Corporation. My

charge in the next 14 minutes or so is to go over

three major areas: first, to provide a very brief

background regarding the biology of aging skin and how

dermal fillers are used in this condition; second, to

refresh for you a prior clinical experience with

Hylaform, which exists from two open label, single arm

studies, one in the U.S., one in Sweden; and, finally,

to spend most of my time going over three distinct

clinical research activities involving the clinical

endpoint development, which I will discuss in the

context of what we were calling our control material

characterization study.

I will try to explain to you our

non‑inferiority approach in the context of describing

our pivotal trial design and then finally end up

describing our repeat treatment study, which was

designed primarily to look at the safety of Hylaform

after repeat treatment.

Okay. So this is the condition under

study, if you will. It's not really a disease per se.

It's really a process of life. Most of us already

have some of this and if not now, then very near in

the future.

If you can't tell, this is old skin. I

will just point out to you the fact that there are

amorphous collections of eosinophilic material in the

dermis that has been termed by dermatopathologists

"elastosis." I would like to focus our discussion

here on the dermis because I think that is very

important for this particular discussion.

So the dermis has cells, of course. But,

in addition, it has extracellular matrix molecules.

Three I think deserve mention. One is collagen, which

by virtue of its triple helical structure provides a

structural scaffold to the dermis.

Elastin by virtue of its elasticity

provides resiliency to the dermis such that the skin

would snap back into place after being retracted.

And, finally, glycocyaminic glycans,

particularly hyaluronic acid, which has a unique

physicochemical state and relationship with water,

provides an osmotic force that gives the dermis its

turgidity. The bad news is that with aging, all of

these molecules become deficient to one degree or

another.

So that's the biology. The clinical

problem is this, and that is wrinkles. There are many

different types of wrinkles. I have been learning

about all of these types of wrinkles. We have

marionette lines. We have the forehead wrinkles,

glabellar lines. And the nasolabial fold is the one

that we have chosen to focus on for the purposes of

our clinical trial.

We have done this primarily because we

feel it's the most challenging in terms of filler. It

exists in a very high‑motion area and because of its

sheer length and depth requires a significant volume

of dermal filler in order to correct.

I think the idea is really quite simple in

that in the dermis, which has deficient extracellular

matrix, we are really injecting into that space a

material that primarily is manufactured extracellular

matrix molecules in the form of collagen or hyaluronic

acid in an attempt to fill out the wrinkle from below

to give you a more pleasing visual contour.

And then this in the bottom is sort of an

idealized great treatment response, patient before and

after. For me as a clinical trialist, however, these

two panels are very important. The question is, how

do we quantify the degree of correction? I will get

into that when I describe our clinical endpoint.

Now to go over some of the prior studies

that we have. Again, these are single arm, open label

trials. One of them was performed in Sweden. It

involved 63 patients from 8 centers, followed for up

to 24 weeks. Multiple defects, not just the

nasolabial fold but other defects in the face, were

corrected, but only one touch‑up was allowed, which

doesn't really reflect practice at this point in time.

The endpoint was assessed by the

investigator, a 100‑millimeter visual analog scale,

and expressed as percent improvement from baseline.

The top‑level results are that 49 percent

of the defects had a greater than 50 percent

improvement at 12 weeks, 39 percent of the defects had

greater than a 50 percent improvement out to 24 weeks,

and there were no concerning safety trends in this

study.

The U.S. study was a bit different in a

number of areas. It was larger, certainly, involving

216 patients, 6 centers. These individuals were

followed up for up to 12 months, primarily for safety.

Multiple defects were treated, again

similar to the Swedish study. But in this particular

case, greater than one touch‑up was allowed, which

reflects more of the standard practice that we see

nowadays.

Again, the endpoint was assessed in

similar fashion. Top‑line results, greater than 60

percent of the defects that were treated in this study

had greater than or equal to a 33 percent improvement

out to 18 weeks. And, again, as with the Swedish

study, there were no concerning safety trends.

This is the list of the different types of

defects that were treated in those two studies. I

believe, as you can see, there were many different

types of wrinkles, many locations, including the

nasolabial fold but also including the lips, the

forehead. And, in particular, scars are treated.

These are for the most part acneiform or post‑viral

scars.

Our clinical strategy really drives our

clinical trial design. This is very important. We

wanted to develop a non‑bovine, non‑collagen source

dermal filler that was clear, not opaque, and it did

not require a pre‑implantation skin test. We wanted

this dermal filler to be comparable to, not superior

to, but comparable to, the current standard of care,

which is Zyplast.

We recognized that in order to do that, in

order to get this approved, that we have to perform a

well‑designed, well‑controlled clinical trial that

meets regulatory standards.

Let me now talk a little bit about our

clinical endpoint development. Lemperle published his

scale in 2001. That group designed the scale

specifically to look at efficacy of dermal fillers.

They evaluated this in three domains: the visual

assessment, live assessment; a photographic

assessment; and a profilometric assessment, in which

one takes a negative silicon impression of the wrinkle

and actually gets a depth. That then provides the

gold standard to which the other two domains are

compared and validated.

Now, this is a great scale, six‑point

scale. One I think would agree that that going from

here to there goes from almost no nasolabial fold

wrinkle to a very severe one. The problem is that

there are other visual signals in this scale that are

distracting to the individual who might be scaling

these wrinkles. Those include different lighting;

different pigments; and, indeed, different wrinkles on

other parts of the face.

So for our particular trial, we were very

keyed on the nasolabial fold, as I mentioned before.

So what we decided to do was to take photographic

images of nasolabial folds of varying degrees of

severity and digitize them and morph them onto a

standardized face. Again, the whole idea here is to

focus the individual who is scoring this on the scale.

Now, this is an animation showing how this

was done. We are aging this person's face 40 years in

5 seconds. And so what we see here is a very, very

severe nasolabial fold at the end with redundant

tissue. Although you may not remember the first one,

it was a perfectly normal face with no nasolabial

fold.

Now, the second part of this whole process

that we think is incredibly important is that we

establish an independent panel review. We recognize

that when investigators evaluate wrinkles in the live

situation, that there is a potential for variability

in measurement and a potential for unblinding because

oftentimes there is a team of individuals where one

would be the injector and the other would be the

evaluator.

In order to get over that, we elect to

establish this independent panel review and to have

the photographs of the patients evaluated outside of

the site at a central location.

So the IPR consisted of two groups of

three board‑certified dermatologists. They were

trained to the scale. They were tested and retested

to establish high intra and inter‑rater correlation.

At the time of the patient visit,

standardized photographic stereotactic methodology and

standardized letting was used to take the picture.

These pictures were then compared to the photographs

independently by each IPR member, who reviewed them in

random order. We have no idea whether this was the

first or the last treatment. They were blinded to

subject, treatment, to site, and to time.

This was the data collection form that

they used, sort of annotating the previous comments

about the systematic nature of how one collects the

data in this particular situation. This data

collection form then goes on to data management and

then is analyzed.

Okay. Now, what do we do with that?

Well, we wanted to evaluate this in our control

material characterization study. We wanted to get

experience with this scale and experience with the

process of the IPR review.

We also needed to evaluate intra and

inter‑rater correlations using the Genzyme scale.

Remember, this is almost exactly like the Lemperle

scale, even to the degree that the written descriptors

of the wrinkles were the same; and, finally, to

confirm the patient to patient variation so we could

convince ourselves that the sample size that we

calculated for the pivotal trial was accurate.

This was an open label trial involving

Zyplast only, 32 patients. They had to have a grade

three or four on our six‑point scale nasolabial fold.

They were touched up at two weeks if less than a

one‑point improvement. And then a final assessment

was made at 12 weeks.

What we found was that this new way of

evaluating wrinkles, this photographic review process,

was logistically feasible, that the standard deviation

of 1.28 confirmed the appropriateness of the sample

size of 108 in each treatment group, and that the IPR

group showed good inter and intra‑rater correlations.

So now on to our pivotal trial. What did

that look like? The primary objective, very simply,

was to evaluate the safety and efficacy, but the key

thing here is we wanted to do so in a non‑inferiority

type of approach. And we're comparing ourselves to

the 20‑year gold standard of Zyplast; and,

secondarily, wanted to measure other clinical

effectiveness measures, including the investigator

live assessment at the site face to face with the

patient as well as patient and physician global

assessments.

A couple of comments about the two fillers

under study. Many of you may know this, but I'll just

review it for you. Zyplast is cross‑linked bovine

collagen and has 0.3 percent lidocaine in it.

Hylaform is cross‑linked avian hyaluronic acid. It

does not contain lidocaine. Zyplast is opaque.

Hylaform is clear. Zyplast has a higher concentration

of 35 milligrams per ml; whereas, Hylaform is 5.5

milligrams per ml. Skin testing is required for

Zyplast, and skin testing is not for Hylaform.

Our methodology is as follows. It was a

double‑blinded, randomized multi‑center study against

an active comparator. And patients were followed up

and observed for 12 weeks. Both nasolabial folds were

corrected with the same material with one touch‑up

allowed at two weeks but only if less than one‑point

improvement was noted by the investigator at the site.

The key point here, patients were

blindfolded at the site, blindfolded before the

injector and the investigator entered the room and

before any material was visible to the patient. They

remained blinded to the treatment arm throughout the

study.

Finally, patients in order to really

evaluate whether or not blinding was appropriate and

good were queried about treatment allocation at the

conclusion of the study. Dr. Holmdahl will review

that data for you.

Major inclusion/exclusion criteria:

healthy men and women between 30 and 55 years of age,

‑‑ ethnicity was not an exclusion in this trial ‑‑ a

negative double skin test to collagen before

randomization, two visible nasolabial folds with a

live score of three or four, no exposure to

potentially confounding cosmetic therapy or procedures

for six months, ‑‑ the reason for this was to evaluate

confounding measurement of the endpoint ‑‑ and no

history of scar‑related diseases, delayed wound

healing, or keloid formation.

This is a schematic of the trial. Again,

patients were skin tested here about six weeks before

randomization. Two weeks later if they were clean,

they were randomized to Hylaform or Zyplast, followed

at three days for safety and efficacy.

A key point here is that two weeks after

the first implantation, they were touched up if the

degree of improvement was less than one in the

investigator's eye. And then again, they were

followed up for three days, two weeks. If they were

not touched up, they went on to 4, 8, and 12 weeks for

both safety and efficacy.

I will parenthetically state that the

proportion of patients touched up was actually quite

small in this study. It is a very important point.

Now I will try to explain non‑inferiority.

Clinically interpreted, it really implies that if one

is non‑inferior, the test device in this case is

comparable to the control. If you remember, that is

compatible with our development strategy in which we

wanted to develop a dermal filler that was comparable

to the effect that has been seen in Zyplast but had

different other characteristics that we think were

important.

Now, when one does non‑inferiority, one

has to define a margin or window of difference between

the treatment and control arm that is small enough or

smaller than a clinically meaningful treatment

benefit.

Remember that we used a six‑point scale

here. A one‑point difference was considered

clinically meaningful. A one‑point difference was

used as the decision note to touch up or not touch in

the pivotal study. And so 50 percent of that, or .5,

was considered to be the window that was most

appropriate.

Now, by convention, when one does a

non‑inferiority study, ‑‑ the reason I am going into

this in great detail is that most of us have done

trials where we are looking at superiority or drug A

is greater or better than drug B. But in a

non‑inferiority type of approach, one really does it

by virtue of confidence intervals, which are

constructed around the difference between the

treatment and the control mean.

In this case, if the 97.5 percent lower

confidence interval falls within the margin that I

have just described, the .5, then one is considered

non‑inferior.

I think Dr. Holmdahl will show you that in

our trial, we actually met this particular endpoint.

On the other hand, if one falls outside of that range,

if the treatment differences are significantly

different, then one is not considered non‑inferior.

Finally, the repeat treatment phase, very

simply, to evaluate safety of repeat treatment with

Hylan B products and to measure rates of adverse

events associated with re‑treatment and, very

importantly, to measure Hylan B anti‑IgG antibodies

and, even more importantly, to determine whether or

not there is an association of those particularly

laboratory tests with adverse events.

The schematic at that trial, again,

remember, Hylaform‑treated patients only. So they

have already been in the trial 12 weeks. If they

agreed to go into this study, they were re‑randomized

to receive Hylaform or Hylaform Plus in either the

right or the left nasolabial fold. They were followed

three days, two weeks, and four weeks for safety.

So I conclude. It is our opinion that

this study design just described meets the principles

of a well‑controlled clinical investigation as

outlined in 21 CFR 860 in that we identified and

recruited and randomized appropriate subjects, who

will receive this product in practice if it is

approved. We used an active control, which is the

gold standard. And we did so in a double‑blinded,

randomized fashion.

Our photographic endpoint provides a

systematic method of observation that we feel

minimizes bias and maximizes the blind at the site.

And a non‑inferiority statistical approach matches our

clinical development strategy.

Thank you for your attention. I am now

going to turn the podium over to Dr. Lena Holmdahl,

who will discuss for you the results of safety and

efficacy.

Thank you.

CLINICAL STUDY RESULTS

DR. HOLMDAHL: Good afternoon. My name is

Lena Holmdahl. I am a senior director in clinical

research at Genzyme. I am also professor of surgery

with about 20 years of clinical experience. I will

present the results to you from the trial that has

been introduced to you by Dr. Polisson.

This is an overview of what I will

present: first, that the two populations were

comparable at baseline; that the use of Hylaform

produces a cosmetic correction that is comparable to

the active comparator, Zyplast; and that the primary

efficacy endpoint was thereby met; that Hylaform does

so with the use of less material, leading to less

exposure of patients to implanted material; that

treatment and repeat treatment are safe, at least

after four weeks after treatment; that the overall

patients; tolerability to Hylaform is excellent; and

that this leads to an overall favorable risk‑benefit

ratio with Hylaform over standard of care.

This slide summarizes important

information about this study population. Three

hundred, thirty‑nine patients were consented, of which

23 percent were screen failures, most of which because

they didn't meet the wrinkle criteria.

Eleven tested positive for bovine collagen

skin tests. So they were not randomized. Please keep

this in mind when you compare the safety results that

patients with positive skin tests to bovine collagen

were already sorted out and did not participate in the

study.

In total, 261 patients were randomized,

133 to the Hylaform arm and 128 to the Zyplast arm.

And, as you can see in the table, there were three

discontinuations in both groups. In the Hylaform

treatment group, one subject requested to withdraw and

two were lost to follow‑up. And then in the Zyplast

group, there were two discontinuations due to adverse

events that were unrelated to the use of the product

and one patient that requested to withdraw.

This slide summarizes important baseline

characteristics that potentially could affect outcome

for the two populations that are compared. All of

these factors were the same in the two groups, which

shows that randomization worked and that the two

groups were comparable at baseline.

Ethnic background was not an exclusion

criteria in the study. This graph shows the ethnic

background of subjects seeking cosmetic improvement,

as published on the internet by the American Society

of Plastic Surgeons and American Society of Aesthetic

Plastic Surgery pertaining to 2002. The ethnic

distribution of patients in the trial is shown in

yellow.

The population in the trial is similar to

what has been published. We, therefore, believe that

the trial population accurately reflects the target

population.

I want to draw your attention to that 80

percent of the patients were Caucasian. So that means

that 20 percent were subjects with skin of color.

Wrinkles were scored at baseline using the

scoring scale that has been described to you by Dr.

Polisson. This graph illustrates that the two groups

were comparable at baseline. It also shows that the

investigators scored at baseline the wrinkles within

the range specified in the protocol for inclusion into

the study.

As degree of cosmetic correction is by

virtue of its nature challenging to quantify and at

the same time is one of the very few endpoints of a

trial that easily lends itself to illustration, I

thought it could be helpful to show some pictures of

patients in the study before and after treatment. As

a side note, the patients have consented to having

their pictures used.

On the left‑hand side is shown Hylaform

patients before treatment and at two weeks after the

last treatment. This area here is a nasolabial fold.

Here you can see the results after treatment.

On the right‑hand side is shown a Zyplast

patient at the same time points. As you can see,

there was an improvement in both cases. In fact,

two‑thirds of the patients in both treatment groups

assessed themselves as better at the conclusion of the

study.

The FDA has specifically asked us to

present examples of various cosmetic outcomes. Here

is shown a good outcome of treatment. In the middle,

we see an average outcome of treatment. And to the

right we can see is an example of a poor outcome of

the treatment.

The primary efficacy endpoint was met.

This diagram again illustrates the results of the

primary efficacy endpoint, showing that the lower

bound of the one‑sided confidence interval was within

the specified non‑inferiority margin. The reason that

I am showing this diagram is to avoid confusion about

the difference, or the delta.

The results from the model were based on

the adjusted score, taking into consideration baseline

factors, including treatment group, study center,

baseline median IPR score, and nasolabial fold. That

difference, or delta, was ‑.2. What I will show in

the following graphs is the raw score.

This difference, or delta, is ‑.1. This

graph illustrates the primary efficacy endpoint, IPR

score, at 12 weeks after the last treatment. The bars

illustrate the mean values of wrinkles of the error

bars, one standard deviation.

The majority of wrinkles did not have any

touch‑ups. As shown in the graph to the left, the IPR

score was the same in the two groups. This group

consisted of 111 of the 133 Hylaform‑treated patients

and 119 of the 128 Zyplast‑treated patients.

The middle bar shows the IPR scores in the

subset of wrinkles that had a touch‑up. That was 22

patients, or 42 wrinkles, in the Hylaform group and 9

patients, or 14 wrinkles, in the Zyplast group.

The bars to the right show combined

results. As illustrated in the graph, the numerical

difference that exists at 12 weeks is being driven by

the subset of wrinkles that had an initial correction

that was less than one score and, thus, qualified for

a touch‑up.

The difference, or delta, between the two

groups is ‑.1. That difference is located in the

range between IPR scores two and three.

The cosmetic appearance of wrinkles grades

two and three is shown in this picture. The numerical

difference at 12 weeks, thus, translates into a tenth

of the difference between these 2 wrinkles, something

that is not discernible by the eye.

To be clinically meaningful, any

difference needed to be at least one point, 50 percent

of which was used as the non‑inferiority limit.

According to these predefined criteria, Hylaform is

statistically and clinically comparable to Zyplast.

As dermal fillers operate through a

mechanism that is depending on occupational space to

be efficacious, the volume administered needs to be

taken into consideration.

The purpose of this slide is to illustrate

and emphasize that Hylaform exhibited a comparable

efficacy to Zyplast with a smaller volume of material

used.

This graph is showing the volume

administered to achieve the results. From this

comparison, it is evident that Hylaform can achieve a

similar result as the comparator with less volume.

This is particularly apparent in the majority of

wrinkles, that they are not qualified for a touch‑up.

In this subset, the volume of Zyplast needed to

achieve the same correction was 1.4 times greater.

This graph illustrates the live assessment

of wrinkles at 12 weeks after the last treatment.

That was the secondary endpoint in the study. This

assessment was done by the investigator in a

face‑to‑face meeting enabling a three‑dimensional

perception of the contour defect, as opposed to the

IPR scores that were assessed from standardized

photographs. The bars illustrate the mean values and

the error bars one standard deviation.

Again, the difference, or delta, between

the two groups is ‑.1. And that difference is located

in the range between live scores two and three. The

same grading template was used for the live scores as

for the IPR scores. And the cosmetic appearance or

wrinkles live scores two and three is shown in this

picture.

The numerical difference at 12 weeks,

thus, translates into a tenth of the difference

between these 2 wrinkles, a difference that is not

visible.

The patients were blinded to their

treatment assignment, as described here earlier by Dr.

Polisson. They were, actually, as a part of the study

asked if they knew their treatment assignment.

As shown in this graph, some of them

thought that they knew the treatment assignment. Some

were right; some were wrong. But, most importantly,

the majority did not know. A proper blinding of the

patients makes patient assessment of outcome valid.

Given the IPR scores I just presented to

you, one might ask if either of the two dermal fillers

had any effect at the conclusion of the study with a

strict treatment regimen that was employed. The

answer to that question is yes, they both had an

effect.

To a certain degree, cosmetic outcome is

in the eye of the beholder. And, therefore, patient

assessment of patient satisfaction must be considered

as this drives the need for additional treatments.

Global assessment is the qualitative

composite outcome measure consisting of positive and

negative effects produced by the use of the product.

This was a secondary endpoint in the study. Global

assessment was done by both the investigators and the

patients.

The bars illustrate the mean values of

patients' assessment at 12 weeks and the error bars

one standard deviation. At the last visit, the

difference in patients' global assessment was ‑.1 on

a 5‑point scale. Both the treatment and the control

group were on average about one, which means that they

assessed the outcome as better than before treatment.

None of the Hylaform patients considered themselves

worse or much worse.

This is important in patient satisfaction

or dissatisfaction with cosmetic correction, governs

the timing of repeat treatment. This is what for a

cosmetic product one might call a clinical benefit.

This notion of improvement from baseline

was confirmed by the investigators. The investigators

were also asked to qualitatively assess outcome. The

bars illustrate the mean values of investigators'

global assessment at 12 weeks and the error bars one

standard deviation.

The difference was, again, ‑.1 on a

5‑point scale. Both the treatment and the control

group were on average about one, which means that the

investigator assessed the outcome as better than

before treatment.

The next section of the presentation will

summarize results pertaining to the safety of the

product. Based on the mechanism of action of

Hylaform, our expectation was that the patients would

come back for repeat treatment and would, therefore,

have an interim safety assessment in the repeat

treatment phase.

At one percent, adverse events were

captured both during the initial phase and after four

weeks of the repeat treatment phase. That was a time

point that was considered by us and by the FDA to be

sufficient to evaluate safety of repeat treatment.

And the data that we have from the initial phase

support that. Lastly, immunogenicity results will be

presented.

This diagram shows the definition of

adverse events as stated in the protocol.

Treatment‑emergent adverse events could be either

procedure‑related or not procedure‑related.

Adverse events with an onset of up to

three days could be deemed by the investigator to be

procedure‑related. Regardless of onset, adverse

events could be deemed as not procedure‑related. In

this case, the investigator was asked for a

relationship with the device and anesthetic procedure

or if the adverse event was unrelated to the

treatment.

This slide shows the treatment‑emergent

adverse events during the initial phase. The overall

incidence of adverse events was the same in the two

groups. The incidence of procedure‑related adverse

events was also similar. The incidence of

device‑related adverse events was two percent for

Hylaform‑treated and seven percent for Zyplast‑treated

patients.

There was one serious adverse event that

was in the Hylaform group. That adverse event was

seen by the investigator to be unrelated to the

product.

There were three patients experiencing

severe adverse events in the Hylaform group and seven

in the Zyplast group. Finally, there was no

discontinuation because of adverse events in the

Hylaform group, but two patients chose to discontinue

because of adverse events in the Zyplast group.

There is no news here. This table shows

the procedure‑related adverse events. They are in

agreement with what has been reported with the use of

dermal fillers. That's what one would expect.

The overall incidence between the two

treatment groups was similar. The most common adverse

events were erythema, bruising, swelling, pain, and

pruritus.

As a side note, the adverse event profile

was the same in subjects with skin of color. And we

have no occurrence of scarring or keloid formation or

pigmentation disorders in patients with skin of color

in the Hylaform group.

These are the device‑related adverse

events in the two treatment groups. Two Hylaform

patients experienced device‑related adverse events.

They were erythema, induration, and pruritus. Nine

Zyplast patients experienced deice‑related adverse

events listed here: erythema, bruising, pain, nodule,

necrosis, and stomatitis.

Duration of adverse events is another

factor that needs to be considered in the evaluation

of the overall safety profile. Using a conservative

approach, the adverse events that could be

attributable to the treatment are the procedure and

the device‑related adverse events. Most of them had

a short duration and had resolved within a week in

both treatment groups, as shown in this table. None

of the adverse events in the Hylaform group had an

onset after one month. This is, of course, evaluating

safety after two weeks after repeat treatment was

reasonable.

Blood sampling for antibodies to Hylan B

was done in all randomized patients prior to

treatment, 2 weeks after last treatment, and at 12

weeks and measured with an ELISA test. Most

importantly, we looked specifically if titer levels

were associated with adverse events. And they were

not.

One of the 261 patients tested developed

a significant increase in Hylan B IgG. That patient

had been treated with a single injection of Hylaform.

The increase in titer was detected at the four‑week

visit. Keep this in mind when assessing repeat

treatment results.

The adverse events that that specific

patient had were injection site bruising and headache,

none of them consistent with an immune‑mediated

response.

In this context, it might be appropriate

to mention that the incidence of bovine collagen skin

tests was 3.2 percent of the screened population and

that screening for potential hyperreactivity to

Hylaform was not done.

Finally, a brief summary of the interim

safety data from the repeat treatment phase. These

results pertain to data collected up to four weeks

after the repeat treatment. Ninety‑six of the 133

patients who were treated with Hylaform opted to enter

into the repeat treatment phase. Overall, the types

of adverse events that did develop were the same as in

the initial phase.

One patient developed an abscess at the

treatment site that needed intervention and, thus, was

classified as serious. It did resolve without

sequelae and was of moderate intensity. There were no

severe adverse events in the repeat treatment phase

and there were no discontinuations because of adverse

events.

The patients who enrolled in the repeat

treatment phase were again tested for Hylan B

antibodies in blood samples. All of these patients

had previously been exposed to Hylaform, some of them

twice. None of the 92 or the 96 patients from which

antibody results are available developed a significant

increase in Hylan B IgG during the first four weeks of

repeat treatment.

So, to conclude, we believe that these

results demonstrate that Hylaform exhibits the same

degree of cosmetic correction, that Hylaform is

well‑tolerated without the skin test, that Hylaform

has a similar safety profile after repeat treatment.

And we believe that this translates into a favorable

overall risk‑benefit ratio.

Thank you.

CONCLUSION

MS. STEWART: I would just like to present

some of the key points that we have provided today:

the fact that Hylaform has been available worldwide

and in Europe beginning in 1996, we have shown in

extensive preclinical testing that Hylaform is safe,

we have demonstrated in a robust clinical trial that

it is comparable to Zyplast in both safety and

efficacy, and we truly believe it is safe to use in

repeat use and that our results support its use in

correction of soft tissue contour deficiencies.

Thank you. Now we are available to answer

any questions the panel may have.

ACTING CHAIRPERSON CHANG: Questions from

the panel? Dr. Blumenstein, do you have any specific

questions regarding the study design?

MEMBER BLUMENSTEIN: I might have missed

it, but did you show just changed scores, just changed

scores?

DR. HOLMDAHL: No. What I showed was the

scores at 12 weeks. I didn't show any changes in

scores.

MEMBER BLUMENSTEIN: What I'm trying to do

is get the magnitude of change that happened in the

Zyplast group, for example.

DR. LARHOLT: I'm Kay Larholt from Genzyme

Corporation, the biostatistics group.

We had established that the efficacy

parameter that we were measuring was the week 12

score. and, therefore, that's what we looked at. We

did not look at changes from baseline.

MEMBER BLUMENSTEIN: Do you have baseline

scores?

DR. LARHOLT: We have baseline scores that

we have shown. Those are the IPR median scores that

were at baseline, the live score as well.

Lena had showed the IPR scores at

baseline. She had shown that they were 3.5 in the

Hylaform group and 3.6. The mean was 3.5 in the

Hylaform group and 3.6 in the Zyplast group, which

shows the means at baseline for using the live score.

This was what was used for enrollment into the study.

MEMBER BLUMENSTEIN: I may have questions

later.

ACTING CHAIRPERSON CHANG: Dr. Newburger?

MEMBER NEWBURGER: I have a question. In

the population of people who were treated a second

time, where you say that the incidence of adverse

events is similar between the two groups, in the

people who were treated de novo in your study the

first time, very few people are reported to have

nodules. I think it was at a zero or one.

In the repeat study, individuals who had

Hylaform the second time, I count 12 people who had a

3 or more days induration. Now, I understand there's

a difference in that there is a diary the second time

around and there wasn't. However, since the protocol,

I think, didn't it show investigator observation the

first time at three days? Why is there such a

disparity? Because 12 out of 96 is a tremendous

proportion.

DR. HOLMDAHL: I would like to call on Dr.

Kingma to answer that question.

DR. KINGMA: Hi. My name is Wytske Kingma.

I am heading up the safety department at Genzyme.

With regard to the nodules we find in the

study, it's important to note that in the initial

phase, the nodules as such were assessed by the

physician. We did not have any in the Hylaform group,

and we have three in the Zyplast group.

In the repeat treatment phase, as you

pointed out, the patients used a diary. It was

actually anything that were lumps or bumps that were

coded. In order to collect adverse events, you need

to code them in the dictionary. So they code to

nodules.

As we are, of course, concerned about

whether or not that nodule is a true nodule in the

immunological sense, we actually looked at those. The

majority of the mean onset for all those 22 patients

who had reported lumps and bumps was 1.2 days, with

the majority of them starting right at the time of

injection and actually some commenting that they felt

that the material under their skin and noting that as

a lump or bump.

In addition, the mean duration of the

entire group was only 12 days. There were no nodules

that actually had the delayed onset, which is what you

would otherwise expect with regards to a possible

immunological reaction.

Does that answer your question?

MEMBER NEWBURGER: Yes, sort of.

ACTING CHAIRPERSON CHANG: Dr. Halsey?

MEMBER HALSEY: I have a question about

the exclusion criteria. Reading the documents, it

seemed that patients were excluded from the study if

they had allergies to avian proteins as well as bovine

proteins. Is that correct?

How was that determined? Was that

patients' reported history of allergy or was there

some test, either a skin test or a blood test of

atopic status?

DR. POLISSON: Those exclusions were by

history.

MEMBER HALSEY: Reported history from the

patient, not a physician's evaluation?

DR. POLISSON: I would like actually to

get some clarification from that. May I ask Laura

Fleming to come forward and specifically address that?

MS. FLEMING: Hi. I'm Laura Fleming from

Genzyme clinical research.

The history is avian source protein

histories. The patients were asked if they were

allergic to eggs. It was patient‑reported. Patients

were only excluded for bovine if they had positive

skin tests.

MEMBER HALSEY: You did not exclude

patients with other allergies, seasonal allergies,

perennial rhinitis, or anything like that?

MS. FLEMING: No, we did not.

MEMBER HALSEY: Not even if it was severe?

MS. FLEMING: That's correct.

MEMBER HALSEY: I would also like to ask

a little bit of detail about the IgG ELISA that you

used. How was that standardized and quantified? Are

these titer units, this 50, or is that nanograms per

ml of IgG or what?

MS. RICHARDS: I'm Sue Richards, Genzyme

immunology department.

The immunologic assay that was used was an

indirect ELISA. So basically Hylaform was coded into

the microtiter plate well. It was then blocked using

nonfat dry milk. Patient sera was then added using a

twofold serial dilution series. And then any bound

antibody was detected using an HRP anti‑human IgG

conjugate. Certain colorometric reaction was then

determined.

Now, the titer was assessed using the

negative control, which was a normal human serum

sample that was tetanus‑positive because we also used

tetanus as a positive control for coding in the assay.

In that particular sample from the

validation, it was determined that an ODE value

greater than .5 was what was considered as negative

reactivity of normal human serum because we did

discover that we had a human serum effect in the

assay.

So the titer values are the reciprocal of

the highest dilution that gave that particular cut

point value.

MEMBER HALSEY: Then you used a fourfold

rule. So it would have to be four times the titer of

that to have been concluded what you concluded?

MS. RICHARDS: No. Since we do have

background reactivity of normal human serum in this

particular assay, the fourfold rule was based on each

patient's own baseline. So we had a true baseline

value of that particular individual. So a sera sample

was drawn before treatment and then a titer determined

on that if there was background reactivity. And then

a fourfold or greater increase above that patient's

own baseline was what was determined as showing

positivity in the assay.

MEMBER HALSEY: Just to address one

further technical detail, the IgG ELISA, did you show

inhibition with soluble forms of hyaluronic acid to

show that what we're really measuring is antibody to

HA, rather than something else?

MS. RICHARDS: What we had looked at

subsequently, we did not do inhibition analysis. What

we had looked upon subsequently in the positive

patient was we had looked at alternative avian sources

and alternative bacterial sources of HA and had

ascertained that the reactivity was specific to

Hylaform B, the cross‑linked avian material.

MEMBER HALSEY: So these patients didn't

have antibody that you could measure to other avian

proteins?

MS. RICHARDS: Correct.

MEMBER HALSEY: Did you measure IgE to any

of these?

MS. RICHARDS: The initial conjugate that

was used in the screening assay, although the vendor

did specify it was anti‑IgG, we did do

cross‑reactivity analysis and ascertained that it is

cross‑reactive to IgM and also IgE.

So the positivity at the beginning would

show cross‑reactivity to IgE if it was present. So if

we did see an increase, we would have seen it also for

IgE.

ACTING CHAIRPERSON CHANG: Dr. Bartoo?

MEMBER BARTOO: You've presented data

outside the U.S. in 30‑plus countries with over

500,000 units distributed. My first question is, what

is the ethnic diversity within those treatments

outside the U.S. that you have presented here?

The second question I have is that I

understand that in the U.S. pivotal study, your

representation of ethnic groups is in proportion to

what typically is treated. However, you still have

very small numbers of certain groups, especially

African Americans. There's only three subjects in

that group.

So I would like to hear your justification

for why you feel that is enough in terms of safety

data in this group.

DR. POLISSON: With respect to the two

clinical trials that I reviewed and presented to you,

most of those patients were Caucasian, but I can

actually get those numbers for you and present to you

later in the panel. I don't have them on the tip of

my tongue.

Now, with respect to other worldwide use,

I, quite frankly, don't think we actually track that

data. But, again, if that's a mistaken or misspoken

comment, then we'll correct that later.

ACTING CHAIRPERSON CHANG: And the second

question was, do you as a sponsor have a feel that the

study is adequate, even though there are three

patients who are of African American descent?

DR. HOLMDAHL: I would like to call on one

of our investigators, Dr. Monheit, who has experience

with patients with skin of color from his practice.

DR. MONHEIT: Hi. I'm Dr. Gary Monheit.

I'm a dermatologist from Birmingham, Alabama

associated with the Department of Dermatology in the

University of Alabama Medical Center.

I have a 20‑plus‑year history of using

injectables in a diverse population in Birmingham and

through a portion of Alabama. First, I'll address the

fact that in our study group, we have the three

African American patients, whom I took care of and

treated. Their efficacy was the same and as good as

all of the other patients treated, and there were no

adverse events in those three particular patients.

In addition to that, I've got many years

of experience injecting patients of color, including

African Americans, with Zyderm, Zyplast injectable

fillers. I have found that there is very little in

the way of adverse events among the populations I've

treated with Zyplast and Zyderm of African American

descent as I have found among Caucasians.

The things that we have always worried

about have been keloid scars, the possibilities of

pigmentary dischromias, hyper and hypopigmentation.

I found that we get probably the same degree of

pigmentation problems as we get in people who have

prolonged erythema, who then may get pigmentation.

That's probably the derivation of it. It's an

inflammatory response that can lead to

post‑inflammatory hyperpigmentation.

When we have seen this with Zyderm and

Zyplast, it is treated in the same way we have treated

hyperpigmentation of other causes. And that's with

bleaching agents and skin creams and exfoliating

agents.

I have never seen a keloid in my 20 years

of experience treating African American patients. I

think part of the reason why is we're treating mainly

the mid‑face area and keloids are very rare in all

diverse populations of people.

Thank you.

ACTING CHAIRPERSON CHANG: Dr. Newburger?

MEMBER NEWBURGER: Obviously you have

given a great deal of thought and care in the design

of the study and have considered immunologic

compatibility. I am wondering since in the

literature, the reports of reactions, adverse

reactions, to Hylaform have been what appear to be of

the delayed hypersensitivity type, have you looked at

all that t‑cell response to the material?

DR. DeLUSTRO: My name is Dr. Frank

DeLustro. I'm with Global Biodevice Development. I

am here as a consultant for Inamed. I have no

financial involvement with either of those companies.

I would like to address that question a

little bit. Perhaps you could put on slide H010. The

company did look at the data and sorted out reactions

which could potentially be ascribed or immunological

in nature.

What we did was to take a look at any

incidences of erythema, redness, induration, swelling,

pruritus in any combination which occurred in patients

in the absence of bruising at those sites and which

occurred at a duration of greater than 14 days.

MEMBER NEWBURGER: Excuse me. Is it

occurred at greater than 14 days or ‑‑

DR. DeLUSTRO: Excuse me. No, no. Had a

duration of grater than 14 days. We also did a cut at

greater than seven days with the same results.

What I will show you is a cut at greater

than 14 days since our experiences with other

injectables demonstrates that those reactions

typically last four weeks at a time. And we're real

happy when they go away.

You can see from this table that, in fact,

what we saw in the initial phase of the study was that

there were only two patients in the Hylaform group

that fit any of this criteria. They were typical

erythema and in a second patient some pruritus and

induration, I suppose, to nine subjects that

demonstrated this type of erythema or pruritus in the

Zyplast group.

These reactions, as you have heard

probably too extensively by now, are not truly

indicative of immunologically mediated events but are

events of reaction within the dermis to the material.

You can see that with removal of

hypersensitive patients from the Zyplast group with

the double skin testing procedure which resulted in

3.2 percent of the population being eliminated prior

to treatment ‑‑ and that is consistent with the

literature of allergic reaction rates to injectable

collagen.

With that elimination, one still sees

approximately a 7 percent rate of these nonspecific

indicators, as opposed to the 1.5 percent seen with

Hylaform. So it's clearly far less.

That little tiny footnote in the bottom is

also important. What that states down there is that

in the second group, which received not only the

primary treatment but went on into the repeat phase

and were retreated, of that 96 patients, none of these

symptomologies were seen. No conditions that fit this

inflammatory categorization were seen at all.

Could I have the next slide? I think this

data specifically on the Hylaform study in conjunction

with what we know about Hylaform from a worldwide

study demonstrates that, in fact, it does not have

significant incidence of immunologically mediated

hypersensitivity.

As you can see in the first point, the

published studies of hypersensitivity to hyaluronic

acid dermal fillers is approximately .42 percent, but,

in fact, if you look specifically at Hylaform within

those numbers, it actually is about .2 percent. That

is clearly a logarithmic level lower than what is seen

with the injectable collagens in the dermal area. In

addition, the adverse event rate seen with Hylaform on

a worldwide basis is approximately 0.1 percent, a

very, very low level.

So I think in terms of your question, I

think the incidence, potential incidence, of

immunological reactions to Hylaform is exceedingly

unlikely based on what we have seen.

MEMBER NEWBURGER: I'm sorry. That wasn't

my question. My question was, have you looked at

t‑cell recognition?

DR. DeLUSTRO: T‑cell recognition, as we

know from bovine collagen, would occur with the same

symptomology. So if you look at the capture of

clinical symptoms, one sees, in fact, the same

clinical picture of reaction.

When you look at a delayed type

hypersensitivity reaction in a bovine

collagen‑mediated reaction, one sees the same

symptomologies occurring. And that we have studied

for quite some time.

So by looking at the clinical

symptomology, one captures the potential reactions

that are going to occur without looking at the

specific etiology, whether it's t‑cell or

antibody‑mediated.

ACTING CHAIRPERSON CHANG: Dr. Miller?

MEMBER MILLER: In your study, you

excluded people who skin‑tested positive against

collagen and also I guess avian proteins. That was

done by skin test or by history only that they exclude

them?

DR. DeLUSTRO: No. That was by history.

And that history of allergy or sensitivity to avian

proteins or bovine proteins is consistent with the

existing package labeling for Zyplastic control

material.

MEMBER MILLER: So what do you think would

happen if you took a group of patients who were

positive in their reaction to bovine collagen and you

gave them your product?

DR. DeLUSTRO: I think our worldwide

experience actually speaks to that. In Europe,

outside the United States, where Hylaform has been in

use since '96, the first population to be treated with

Hylaform, in fact, was the built‑up population of

subjects who had bovine collagen sensitivity. We know

from the literature and experience with Hylaform in

Europe that, in fact, those patients were treated very

successfully with the material.

MEMBER MILLER: Do you still recommend or

are you recommending for the product that if it is

approved, that it not be administered to people who

have a history of reactivity to avian proteins?

DR. DeLUSTRO: We would suggest that as a

precautionary statement in the package insert.

MEMBER MILLER: Just one other question.

Is there some avian protein present in the Hylaform

preparation, some minimal amount?

DR. BURNS: There's about 0.01 percent

avian protein per ml in Hylaform.

MEMBER HALSEY: My understanding is that

there was .45 percent of the mass in the product.

DR. BURNS: Absolutely not.

MEMBER HALSEY: That's what was here.

Okay.

DR. BURNS: 0.01 percent protein.

ACTING CHAIRPERSON CHANG: Dr.

Blumenstein? Dr. Doull next.

MEMBER BLUMENSTEIN: Was Zyplast effective

in this study?

DR. LARHOLT: We chose Zyplast as a

comparator because of this data.

ACTING CHAIRPERSON CHANG: Please speak

into the microphone. Identify yourself.

DR. LARHOLT: I'm Kay Larhout from

biostatistics from Genzyme.

We chose Zyplast as the comparator because

it is the gold standard. It has been used 20 years.

We show that at the end of the study, as Dr. Holmdahl

has shown, the life scores were on average one point

better than they had been at baseline, that the

investigator and patient‑level assessments were both

positive at 12 weeks. And those were significantly

different from zero. So we thought that yes, Zyplast

was effective in this study.

I will let Dr. Holmdahl talk a little bit

about the differences in this study compared to how

Zyplast is used in practice.

MEMBER BLUMENSTEIN: All I'm asking is,

was there a statistical test done to show that Zyplast

was effective in this study?

DR. LARHOLT: No, there was not.

MEMBER BLUMENSTEIN: Well, then can I

conclude from that that it wasn't or that you didn't

do the statistical test? That's what I'm trying to

get to because if you demonstrate non‑inferiority to

something that is not effective, then you haven't done

anything.

DR. LARHOLT: We did not test

statistically whether Zyplast was effective. However,

we feel that the result we have shown shows that

Zyplast still was effective at 12 weeks in the study

and, therefore, shown comparability to it shows that

we are non‑inferior to an active comparator.

MEMBER BLUMENSTEIN: I am convinced that

you have shown non‑inferiority to Zyplast, but I am

concerned about whether Zyplast was effective.

DR. LARHOLT: I think that maybe Dr.

Holmdahl can talk about the actual use of Zyplast in

the study that may be different from how it's used and

how it's been shown in different studies.

DR. HOLMDAHL: Well, if you recall the

graphs that I showed, both the live scores show that

both Zyplast and Hylaform had an average of about one

point improvement at 12 weeks and also the global

assessment, both investigator and patients' global

assessment, were about average of whether it shows or

it translates into that they rated themselves as

better.

So I think from a clinical perspective,

there is no doubt in the study that there was an

effect at 12 weeks. I would also like to call on Dr.

Monheit, who was one of the investigators, to give his

view of this.

MEMBER BLUMENSTEIN: Well, the problem is

I don't have the standard deviation. I have no idea

of the variability for the patients in this study. So

I don't know whether Zyplast was truly effective in

this study.

ACTING CHAIRPERSON CHANG: Do you have

data to show about the efficacy, sir, of Zyplast to

answer that question?

DR. MONHEIT: No. I'm not talking about

data. I'm talking about clinical experience.

MEMBER BLUMENSTEIN: That doesn't matter.

ACTING CHAIRPERSON CHANG: Thank you.

MEMBER BLUMENSTEIN: My next question is,

in your statement of the primary analysis that you

were going to do, you said that you were going to have

a combination of a two‑sided test of superiority and

a one‑sided test of non‑inferiority that was used for

the primary efficacy endpoint. The test of

non‑inferiority was considered to be the primary

analysis. So you presented that. And I believe it,

for a change.

(Laughter.)

MEMBER BLUMENSTEIN: And then you said if

the non‑inferiority test was demonstrated, the test of

superiority was to be performed as a secondary

analysis.

So you demonstrated non‑inferiority.

Where is your test of superiority?

DR. LARHOLT: We did not present the test

for superiority because it was obvious from the

results that said that the Hylaform because of the

negative difference, the ‑.2, would not show

superiority to Zyplast.

MEMBER BLUMENSTEIN: I just wanted to say

it. You're not superior, right?

DR. LARHOLT: No, no.

MEMBER BLUMENSTEIN: Okay. Thank you.

DR. LARHOLT: We're not, but that was not

the primary endpoint. The primary endpoint was

non‑inferiority.

ACTING CHAIRPERSON CHANG: Dr. Doull has

a question and then Dr. Halsey.

MEMBER DOULL: I may have missed it in

your material, but can you tell me how much divinyl

sulphone and formaldehyde are in your product?

DR. BURNS: There's approximately 2.3

parts per million total formaldehyde in the product.

About one part per million is actually free. And for

divinyl sulphone, it's about two parts per million