Osteoarthritis

Senior Capstone Design Project

Submitted to:

Dr. Miguel Bagajewicz University of Oklahoma

“The Hyaluronan Solution to a Devastating Problem”

By: Chris Clark And Kim Fink

Spring 2006

1

Executive Summary

The primary goal of this project was to create a treatment for osteoarthritis that is superior to those currently found on the market. There are nearly 23 million of people in the US alone that suffer from this degenerative joint disease, and this number is continually increasing. Although most people with osteoarthritis are over the age of 65, anyone is susceptible and can develop the disease for a number of reasons including impact injury, obesity, and genetic defects. The actual mechanism for the development of osteoarthritis is not fully understood but one thought is that the viscoelastic properties of synovial fluid, the fluid that separates the articular surfaces of the knee joint, are reduced. Since the actual cause of the disease is unknown, current treatments are not able to stop or reverse the degeneration. One of the newest treatments being used is viscosupplementation in which derivatives of hyaluronic acid, the primary component of synovial fluid, is injected into the knee to restore the viscoelastic properties. These current treatments must be administered once per week for 3-5 weeks and only last for approximately 6 months. This leads to accumulating costs in addition to the inconvenience to patients causing an increase in demand for more successful treatments. Therefore, our task was to create a novel solution to provide longer-lasting effects over other products. The treatment being proposed is more stable – meaning less likely to degrade – than current treatments due to a novel crosslinker being introduced. A hyaluronic acid derivative will be made through “bottom-up” synthesis and modified with 2-vinyl. Then ammonium peroxydifulfate will be introduced as a crosslinker to create a hydrogel. This treatment will use hyaluronan as a basis for the structure due to its chemical properties and biocompatibility. Through the properties of hyaluronan and the crosslinking associated with the structure a viscosity of around 16 Pa·s will be reached under moderate shear rates. This value of viscosity has been found to be adequate through the lubrication theory. The structure will have a molecular weight around 3 million Daltons which will consist of approximately 6700 monomers and 1000 crosslinks. It has been found through economic analysis of the competition that in order for our product to be competitive our product would have to cost around $2,400 per total treatment. Through the application of microeconomics equations it has been found that the demand for our product at this price should range from 350,000 people in the beginning all the way up to 625,000 people. Using this expected demand and looking at the possibility that the expected demand could be wrong by a deviation of 200,000 people, risk was assessed to the Net Present Value. It was found that the mean expected Net Present Value was around $240 million after 10 years. We also aim to receive approval from the FDA and modeled this process by looking at different possible scenarios. By analyzing the risk of failure it was found that it would be most profitable and less risky to have 10 workers and 85 experiments to go through each module of the FDA process.

2

Table of Contents

I. Introduction……………………………………………………………… ….3

II. Anatomy of the Human Knee………………………………………………3

III.1 The Joint………… ……………………………………………………3

III.2 Articular Cartilage………………………………………………....... 4

III.3 Synovial Fluid……………………………………………………… ...6

III. Osteoarthritis……………………………………………………………….8

IV. Traditional Treatments…………………………………………………….9

V. New Treatments…………………………………………………………...11

VI. Our Solution…………………………………………………………….....12

VII. Lubrication Theory………………………………………………………..23

VIII. FDA Approval Process……………………………………………………27

IX. FDA Risk Analysis………………………………………………………...46

X. Demand…………………………………………………………………….48

XI. Cost Analysis………………………………………………………………51

XII. Conclusion…………………………………………………………………57

References………………………………………………………………………....59

3

I. Introduction

An estimated 47.8 million Americans currently suffer from arthritis, an often debilitating

joint disease, costing the United States’ economy nearly $6.1 billion per year.44 With the

average life span ever-increasing, Americans will likely work longer, even well into their 60s

and 70, leading to two-thirds of the working population being afflicted by arthritis.44

Osteoarthritis, the most common form of arthritis accounting for over 50% of the patients, is a

degenerative joint disease that can cause severe pain and hinder people from everyday activities

such as working, gardening, and playing with grandchildren. This paper provides the

background information necessary to understand this disease, discusses the problems with

current treatments, and offers a novel solution of intra-acticular injections of a derivative of

hyaluronic acid intended to increase efficacy of treatments, lower lifetime costs, and return social

activity to those suffering with osteoarthritis.

II. Anatomy of the Human Knee

Although extremely important to the study of osteoarthritis, the structure of the knee is

introduced here only to provide enough background information necessary to understand and

recognize any terms or ideas discussed later in the paper. We will begin with the joint itself in

order to build a complete picture of the main components affected by the disease.

II.1 The Joint

The knee joint is a type of synovial joint – one in which the ends of two bones are

free moving, joined only by connecting ligaments and a fluid-filled cavity called the

synovial space.4 Figure II.a shows the bones – the femur, tibia, fibula, and patella – the

major ligaments and tendons – the anterior cruciate ligament, collateral ligament, and

patella tendon – and the meniscal and articular cartilage.5

4

II.2 Articular Cartilage

II.2 Articular Cartilage

Although very thin, approximately only 2-3 mm thick for an average healthy

knee, the main function of articular cartilage is to transfer loads from one surface to

another preventing direct contact between opposing bones.19 Its flexibility allows for

pressure to be distributed over a larger area instead of at a concentrated point, while its

elasticity allows for the return of its original shape upon removal of the load. These two

properties are due to the highly organized extracellular matrix (ECM) that makes up the

articular cartilage. The ECM has a biphasic characteristic as it is comprised of a porous

solid phase of collagen, proteoglycans, and chondrocytes, and a liquid phase of mostly

water. This two-phase characteristic of cartilage is vital because when at a relaxed state,

the extracellular matrix is filled with water, but once a load is applied, the water is

flushed out of the porous ECM, leaving a stiff solid structure to support the load. Then

when pressure is removed, the water rushes back into the ECM returning the cartilage to

its original shape.

Collagen, a triple helical protein, makes up about 20% of the wet weight of

articular cartilage and is responsible for the tensile and shear strength of the extracellular

matrix.

Figure I I.a – The major bones, ligaments, and cartilage ofthe hu man knee extended (left) and bent (right). 5

5

Proteoglycans, accounting for 10-15% wet weight, are “bottlebrush”

macromolecules composed of a glycosaminoglycan backbone and noncovalently attached

core proteins with covalently attached smaller glycosaminoglycans.12

Glycosaminoglycans (GAGs) are unbranched polysaccharides found throughout the

body, and will be discussed in more detail later. The proteoglycans give compressive

strength to articular cartilage and provide the pores into which water flows and becomes

trapped.11

Condrocytes contribute only 5% wet weight but are essential to its maintenance

because they are the cells responsible for secreting the necessary components for

regeneration.11 However, without a vascular or nervous system in the cartilage these

cells do not receive the nutrients and signals necessary to activate proliferation in times of

needed repair. Therefore, once damaged it is very difficult for cartilage to be restored,

hence the need for this study.

As previously mentioned, the extracellular matrix of articular cartilage is highly

structured as seen in Figure II.b. The zones of the cartilage are shown with a short

description of each.12

(i) Synovial Cavity – Very thin space between articular surfaces containing synovial fluid (to be discussed later)

(ii) Lamina Splendens (Articular surface) – highly fibrous, numerous collagen fibrils parallel to surface to resist tension

(iii) Tangential Zone – Flatter chondrocytes, thicker fibrils parallel to surface

(iv) Transitional Zone – Thickest zone, spherical condrocytes, collagen fibers perpendicular to surface to hold zones together

(v) Tidemark – Change in cartilage stiffness from transitional to calcified (vi) Subchondral Bone – The surface layer of bone covered with articular cartilage

Figure II.b – Zones of normal articular cartilage; shows the joints of the knee separated by the synovial cavity (i) 12

6

II.3 Synovial Fluid

Although articular cartilage has been the focus of many other studies of

osteoarthritis, the focus of this paper will be on the synovial fluid which also plays a vital

role in understanding and treating the disease. The synovial fluid is found in the synovial

space as was seen in Figure II.b and is contained inside the synovial membrane as shown

in Figure II.c. This semi-permeable membrane covers the non-cartilaginous surfaces

inside the joint cavity reaching around the entire ends of the femur and tibia. Although

more viscous than water, synovial fluid provides low-friction lubrication for the gliding

articular surfaces to help reduce wear of the cartilage.

Components

The main components of synovial fluid are water, proteins, and a particularly

important glycosaminoglycan, hyaluronic acid. The body’s natural hyaluronic acid (HA)

is a linear polysaccharide consisting of the two monomers D-glucuronic acid and N-

acetyl-D-glucosamine with a molecular weight of 105-107 Daltons and length of 500-

25,000 disaccharide units and appears in healthy synovial fluid at a concentration of 2-4

Figure II.c – Normal joint showing synovial fluid and synovial

membrane

7

mg/ml.18 One of the most important characteristics of HA is its polyanionic nature and

multiple reactive sites. The carboxylate group (-COOH) on each D-glucuronic acid

monomer carries a negative charge causing repulsion within the molecule itself while

attracting water molecules through hydrogen bonding. The backbone of hyaluronic acid

is stiff due to these intermolecular and solvent interactions. The axial hydrogen atoms

arrange in a non-polar, relatively hydrophobic face, while the equatorial side chains form

a polar hydrophilic face creating a coil-like conformation.18 The three reactive sites are

shown in Figure II.d and will be the possible sites for modification.

Figure II.d – HA molecule (2 disaccharide un its shown) indicating active sites 49

Because of the hydrogen bonding holding the HA molecules together and the

repulsion between the negative charges on the D-glucuronic acid monomer, HA forms a

mesh-like viscoelastic network that reconfigures upon loading and unloading.18 The

viscosity of hyaluronic acid in solution ranges based on molecular weight and shear rate.

For example, the relative viscosity – the viscosity of the pure substance divided by the

viscosity of the solvent – for HA with molecular weights 140, 500, and 2000 kDa is 1.11,

1.17, and 1.53, respectively, when the solvent is water at normal body temperature.39

Properties

The synovial space in the human knee joint is very narrow with an average width

of approximately 25-50 µm, but this can be as small as 0.8-1.5 µm in the high-load-

8

bearing areas.13,14 The volume of synovial fluid varies with age and health of the joint

but averages approximately 1 ml with a relative density close to unity. Healthy human

synovial fluid exhibits non-Newtonian shear-thinning behavior resulting in a decreased

viscosity at higher shear rates.19 Because of this, reporting the viscosity of synovial fluid

as a single value is impossible as it can range from 1-10,000 kg/m·s.39 As with articular

cartilage, the synovial fluid provides low-friction lubrication with a coefficient of friction

of 0.02.39

III. Osteoarthritis

Osteoarthritis is one of the most common of the several musculoskeletal disorders, but

unfortunately for millions of Americans, it is still not fully understood. For years it has been

referred to as simply a “wear-and-tear” disease, but researchers know that it is much more

complex than this term implies. Although osteoarthritis is most prevalent in people over the age

of 65, anyone is susceptible, even children. Several risk factors have been identified such as age,

obesity, injury, prolonged elevated activity, increased bone mineral content, endocrine and

metabolic disorders, and even genetic defects.15 Although the name “osteoarthritis” implies

inflammation, swelling is minor compared to the degeneration of the articular cartilage that leads

to bone-on-bone contact, hardening of underlying bone, and formation of osteophytes, more

commonly known as bone spurs, which are bony protrusions that grow near the end of the joint

as the body attempts to heal itself.

The degeneration progresses slowly through four stages. During the stationary phase, the

joint cavity narrows and small osteophytes begin to form. This narrowing is likely due to the

reduced viscoelasticity of the synovial fluid because of reduced hyaluronic acid concentration

and increased water content.21 Studies have shown that the HA present in osteoarthritic synovial

fluid has decreased molecular weight and increased crosslinking density, which allows the HA to

9

escape through the semi-permeable synovial membrane and also leaves the molecule more

susceptible to degradation. Increased pressure on the rough cartilage surface causes the normally

highly structured extracellular matrix to be disrupted, and the disease progresses. Next the

synovial space is obliterated, and high friction bone-on-bone contact occurs, leading then to the

formation of cysts in the subchondral bone. Finally large osteophytes form throughout the joint

causing pressure on surrounding nerves and severe pain.16 Figure III.a shows a comparison

between a normal healthy knee and a knee with fully developed osteoarthritis.

IV. Traditional Treatments

When first diagnosed, doctors will generally suggest simple treatments to help relieve the

early symptoms of osteoarthritis such as dieting, exercise, reduced activity, and support

devices.17 Dieting and weight loss can help reduce the amount of pressure supported by the

joints and possibly help slow the progression of OA. Exercise is essential to keep the joints

flexible as well as to maintain strong surrounding muscles; although, also important is decreasing

the frequency of strenuous activities will obviously help relieve pain by removing extra stress

Figure III.a – Normal vs. Osteoarthritic Knee

10

and allowing time for swelling to decrease. Furthermore, if a patient is having pain during

normal daily activities they are encouraged to use a walker, cane, brace to help redistribute some

of the load.17

Most people who are seeking medical treatment for osteoarthritis likely have more pain

than that which the above physical treatments are capable of reducing and therefore opt to take

oral analgesics, or pain relievers. Common over the counter pain relievers such as aspirin and

acetaminophens (Tylenol®), taken four times per day at 1000 mg/dose can be equally effective at

reducing inflammation as prescription pain medication.17 Although there are few adverse side

effects of acetaminophens, taking several doses a day for an extended period of time can be quite

costly as well as inconvenient. Different from acetaminophens but often prescribed are

nonsteroidal anti-inflammatory drugs (NSAIDs) which provide slightly better relief for moderate

pain, but have been shown to cause side effects on the kidney and stomach.17 Aspirin is grouped

into this category. A newer class of NSAIDs is the cylcooxygenase (COX-2) inhibitors, such as

the marketed Celebrex® and Vioxx®, which have fewer adverse effects but are more expensive

and have not been shown to be any safer than acetaminophens, as they can cause serious

interactions with other medications leading to gastrointestinal bleeding.17 Most important to

note, however, about oral analgesics is that while they can relieve moderate pain, they cannot

slow the degeneration.

As an alternative to oral pain relievers, patients are often suggested to receive an intra-

articular steroidal injection which can be administered three to four times per year. These have

been proven to significantly reduce pain for up to four weeks, but not necessarily are they able to

restore function. This is due to the power of the active ingredient cortisone to reduce severe

inflammation, but it is thought to possibly quicken the degeneration of the articular cartilage.

Therefore steroid injections are not recommended to be used repeatedly.21

11

With even increasing severity of pain, osteoarthritic patients have the option of

arthroscopic lavage and debridement in which a surgeon inserts a small camera into the knee

cavity, locates the damaged cartilage, washes the area with saline, and collects the loose debris.46

By smoothing out the articular surfaces, the treatment usually relieves pain for nearly 3 years.

However, this procedure, like the steroidal injections, greatly reduces pain, but because the

articular surfaces are already deformed, the range of motion is not fully restored and the

deterioration continues.

After the cartilage has completely deteriorated and the previous treatments have been

exhausted, the patient is forced into a total knee replacement. This invasive surgery involves the

removal of the osteoarthritic joint from of the knee and replacement with a mechanical

prosthesis. Although the parts do loosen over time, this procedure is by far the longest lasting

solution with a lifetime of 15-25 years. However, the costly surgery poses serious health risks

including infection and rejection by the body while requiring extensive physical therapy.

Although these traditional treatments have been widely used and studied, with the

exception of a total knee replacement, they must be used repeatedly, result in accumulating costs,

can cause adverse side effects, and still do not return full functionality to the patient.

V. New Treatments

Recently a new type of treatment for the relief of osteoarthritis has been in use in which

hyaluronic acid is injected into the joint cavity to restore the viscoelasticity to the synovial fluid;

this is known as viscosupplementation. These injections must be administered once per week for

3-5 weeks, depending on the particular product, and repeated after 6 months. The two main

available products are hyaluronan and Hylan G-F 20. Hyaluronan is intended to mimic the

body’s natural linear hyaluronic acid with an average molecular weight of 2x106 Da. Some

subgroups of clinical trials treated with hyaluronan showed positive results, while many others

12

showed no benefit over placebo.21 Hylan G-F 20 consists of crosslinked hyaluronic acid that is

intended to increase the average molecular weight of the HA in the synovial fluid in attempt to

restore its viscoelasticity. Clinical trials with hylan indicated significant superiority of the

injections over placebo as well as NSAIDs.21 Between the two products, crosslinked hylans

seemed to be more efficacious.

Because the exact mechanism of the role of hyaluronic acid in the synovial fluid is not

completely understood, the exact mechanism of how the HA injections provide relief are not

fully known either. However, it has been mentioned frequently that the problem with the current

HA injections is that the solution is not being retained inside the synovial space. This can either

be due transport across the synovial membrane, enzymatic degradation, or simply missing the

synovial cavity when the injection is administered.

Although more effective and less invasive than any other traditional treatment, it would

be beneficial to patients to reduce the frequency of injections by increasing retention time and

slowing degradation rates while maintaining a competitive cost. Our solution to this problem

will now be presented.

VI. Solution (HYAL-VYNE)

We propose the creation of a hyaluronan derivative that is chemically cross-linked by

ammonium peroxydisulfate. The new solution will be superior due to this enhanced cross-

linking. The enhanced cross-linking that is associated with the new solution will increase the

stability of molecule protecting it from degradation. This increased stability will require fewer

injections because of the longer life of the new solution in the body.

The typical chemical structure of HA found in the body has a repeating disaccharide unit

consisting of D-glucuronic acid and N-acetyl-D-glucosamine which can be better seen in Figure

VI.a33.

13

The chemical name is β-D-glucuronyl-(1→3)-N-acetyl-D-glucosamine

(GlcAβ(1→3)GlcNAc, N-acetylhyalobiuronate), which is linked in a β(1→4) fashion. The

molecular weight for this structure is typically around 2 million Dalton. One major advantage of

HA in the joints is its ability to bind large quantities of water, giving rise to its ability to lubricate

the joints and absorb shock.33 This will be explained in more detail later on.

The reason that HA was used in the creation of our new solution was because of HA’s

ability as stated earlier to bind to water. When HA is crosslinked it is able to form a three-

dimensional network of flexible chains. For an example of this crosslinking, look at Figure VI.b.

Figure VI.a – Disaccharide unit

Figure VI.b – Crosslinking Hydrogel Formation

14

Since HA has the ability to absorb water, when it forms this three-dimensional network of

flexible chains it turns the crosslinked material into a hydrogel. One important property of a

hydrogel is its ability to have the swelling controlled. By controlling the swelling, the degree of

hydration and consistency are able to reach properties similar to those of highly hydrated body

tissues.43 It has been found that solutions containing no hyaluronate do not lubricate a soft

tissue system, such as the space between joints, nearly as well as solutions containing

hyaluronate.39 HA in the synovial joints has been found to act as a lubricant and shock absorber,

storing energy between opposing cartilages.30

The new solution was created through a series of steps. The monomer created and used

in our reaction was an oxazoline-type monomer derived from N-acetylhyalobiuronate also

known as HA or hyaluronan. The method that was used to create the monomer that was used in

synthesis of our solution is described by Ochiai, Ohmae, Mori, and Kobayashi.35 Our HA

derivative monomer was created using a total of 8 steps as shown in Figure VII.c. This figure

represents the changes that are occurring to the structure between each step. The steps are

described in more detail after the figure. The figure also shows the final structure of the

substrate monomer that is created from this bottom-up synthesis. This monomer will be the basis

for our cross-linked HA derivative.

15

Substrate Synthesis Procedure

Here are the structures and steps used to create our substrate monomer in relation to the

figure. Each number in parentheses represents the chemical being described or used from the

figure, while each Roman numeral in parentheses represents the actual additives and step in the

reaction forming the substrate monomer:

Substrate Monomer was formed in the following manner:

(i) A solution of 3,4,6-tri-O-acetyl-2-azido-2-deoxy-α-D-glucopyranosyl

trichloroacetimidate (1) and benzyl alcohol in dry toluene in with presence of activated

Figure V II.c – Bottom -up synthesis of hyaluronic acid

16

MS4A is added to a solution of BF3·OEt2 in dry CH2Cl2 at -40 oC. This is then poured

into aqueous NaHCO3, and extracted with CHCl3. The organic layer is washed with

aqueous NaCl and then dried over MgSO4, filtered, and concentrated. The dried residue

is then purified by silica gel column chromatography (10:1 to 4:1 n-hexanes-EtOAc) to

give Benzyl 3,4,6-Tri-O-acetyl-2-azido-2-deoxy-β-D-glucopyranoside (2).

(ii) Compound 2 is then dissolved in 10:1 (v/v) MeOH-CH2Cl2, and then NaOMe in

MeOH is added dropwise. The compound was then neutralized by adding Dowex 50W-

X4 (H+ form). The resulting solution was Benzyl 2-Azido-2-deoxy-β-D-

glucopyranoside (3).

(iii) Compound 3 is then dissolved in MeCN followed by the addition of 2,2-

dimethoxypropane, Drierite, and camphor-10-sulfonic acid. This mixture is then

concentrated and put through silica gel column chromatography (10:1 n-hexanes-EtOAc)

to give Benzyl 2-Azido-2-deoxy-4,6-O-isopropylidene-β-D-glucopyranoside (4).

(iv) Compound 4 is used with the addition of donor 353 in dry CH2Cl2 with activated

MS4A. To this solution BF3•OEt2 in dry CH2Cl2 was added. This mixture is stirred for

30 minutes under an argon atmosphere. This is used to activate the reaction. The

reaction is terminated by the addition of Et3N. The mixture is then filtered through Celite

and poured into aqueous NaHCO3, and extracted with CHCl3. The organic layer is

washed with saturated aqueous NaCl and dried over MgSO4, filtered, and concentrated

resulting in the formation of solution Benzyl (Methyl 2,3,4-Tri-O-acetyl-β-D-

glucopyranosyluronate)-(1→3)-2-azido-2-deoxy-4,6-O-isopropylidene-β-D-

glucopyranoside (5).

(v) Compound 5 is dissolved in 80% aqueous AcOH and diluted with CHCl3. This

mixture is washed with aqueous NaHCO3, and aqueous NaCl. The organic layer is dried

17

over MgSO4, filtered through Celite, and concentrated. The residue is then purified by

silica gel column chromatography (1:1 to 1:4 n-hexanes-EtOAc) to give Benzyl (Methyl

2,3,4-Tri-O-acetyl-β-D-glucopyranosyluronate)-(1→3)-2-azido-2-deoxy-β-D-

glucopyranoside (6) 78%.

(vi) Compound 6 is put in MeOH containing palladium(II) hydroxide on activated

carbon. The reaction takes place in a hydrogen atmosphere giving a mixture with

compound 6.

(vii) The mixture is filtered and rinsed with MeOH. Then Et3N and acryloyl chloride

were added to the mixture. After the addition of pyridine the mixture is concentrated to

give (Methyl 2,3,4-Tri-O-acetyl-β-D-glucopyranosyluronate)-(1→3)-1,4,6-tri-O-

acetyl-2-acrylamido-2-deoxy-D-glucopyranose (Crude Product).

(viii) The crude product is treated with Ac2O and pyridine. Then MeOH is added to

stop the reaction. The residue is purified by silica gel column chromatography resulting

in the formation of (Methyl 2,3,4-Tri-O-acetyl-β-D-glucopyranosyluronate)-(1→3)-

1,4,6-tri-O-acetyl-2-acrylamido-2-deoxy-D-glucopyranose (7) (54%).

(ix) Compound 7 is added to dry 1,2-dichloroethane and then treated with trimethylsilyl

triflate (TMSOTf) in dry 1,2-dichloroethane using an argon atmosphere to cause the

reaction. To stop the reaction Et3N is added. This residue is purified through silica gel

column chromatography to give 2-Vinyl-4,5-dihydro-[4,6-di-O-acetyl-1,2-dideoxy-3-O-

(methyl 2,3,4-tri-O-acetyl-β-D-glucopyranosyluronate)-α-D-glucopyranoso][2,1-d]-

1,3-oxazole (8) (78%).

(x) Compound 8 is treated with NaOMe in MeOH and stirred until dry. This in turn

yields 2-Vinyl-4,5-dihydro-[1,2-dideoxy-3-O-(sodium β-D-glucopyranosyluronate)-

α-D-glucopyranoso][2,1-d]-1,3-oxazole ( crude substrate monomer).

18

(xi) The remaining methyl ester from step (x) is hydrolyzed in a carbonate buffer to form

2-Vinyl-4,5-dihydro-[1,2-dideoxy-3-O-(sodium β-D-glucopyranosyluronate)-

α-D-glucopyranoso][2,1-d]-1,3-oxazole (substrate monomer).

Polymerization

The idea of “transition-state analogue substrate” from Ochiai, Ohmae, Mori, and

Kobayashi was used in the enzymatic polymerization of the substrate monomer. It was found

from previous studies that ovine testicular HAase (OTH) can be used for successful enzymatic

polymerization of similar monomers. This enzymatic polymerization leads to a HA derivative

having an approximate Mn (number average molecular weight) of 9100 (containing about 44

saccharides) and approximate Mw (weight average molecular weight) of 19,500 after purification

which is much larger than the typical HA polymer with a Mn of 6000 and Mw of around 14,000.35

The time required for this reaction to completely consume the monomer is approximately 48 h.

The enzymatic polymerization effect can be seen in Figure VII.c below.

Enzymatic Polymerization Procedure

A solution of substrate monomer is put in a carbonate buffer and incubated with OTH.

After approximately 48 h, the reaction mixture is heated at around 90 °C for 5 min to inactivate

the enzyme. The resulting structure is the hyaluronic derivative.

The new hyaluronic derivative that will be used in the formation of our solution was

created with an addition of a new functional group which is represented by R in Figure VII.c.

Figure VII.c – Enzymatic polymerization of HA

19

The new functional group replaced the methyl group that is found in synthetic HA. The

synthesis for this hyaluronic derivative came from Ochiai, Ohmae, Mori, and Kobyashi.35 This

new functional group that was introduced was a 2-vinyl. The formation of this 2-vinyl came

from steps [(vi)-(viii)] that were shown earlier. The reason we chose to use the hyaluronic

derivative containing 2-vinyl was because of the 2-vinyl’s ability to be used to form crosslinks

with several different chemicals. The use of the 2-vinyl also gave the new hyaluronic derivative

the ability to react with ammonium peroxydisulfate, which forms a strong crosslink. This

crosslinking can be seen in Figure VII.d. The new crosslinking introduced in our material is a

chemical crosslink formed through the reaction of ammonium peroxydisulfate and the vinyl

group of our polymer. The reason for using ammonium peroxydisulfate as the crosslinking

reagent is because ammonium peroxydisulfate contains disulfide bonds which have been found

to have the ability to stabilize a structure against denaturation.25 Another reason for the use of

ammonium peroxydisulfate is because it has been shown to be non-toxic or biocompatible in the

human body which means that it is more likely to make our final product be safe in the human

body and easier to get through the FDA approval process.25 One more reason for using

peroxydisulfate for our crosslinker is because it has been found to be a very efficient system,

which gives rapid formation of a gel.38 As stated earlier this gel formation is important because

of the properties associated with a gel. The gel provides a structure that is less subseptible to

chemical denaturation and has the ability to mimic tissue in the human body.43 From looking at

previous experiments done by other scientists on peroxydisulfate the time required for the

crosslinking to occur should be around 400 seconds using a concentration of .05 mol/L for

ammonium peroxydisulfate. This time will vary with concentration increasing with decreasing

concentrations of ammonium peroxydisulfate.36 The reason for using an ammonium

peroxydisulfate concentration of .05 mol/L as a basis for the time required for crosslinking is

20

because after calculating the required concentration of peroxydisulfate in relation to the

concentration of the new HA derivative to get the desired percent crosslinking of the final

product, it was found that a concentration of approximately .055 mol/L of ammonium

peroxydisulfate is required for a concentration of .28 mol/L of the polymerized HA derivative in

water. The percent of crosslinking associated with the new material will be 15% with this

concentration of ammonium peroxydisulfate assuming that the concentration of the HA polymer

is kept constant. This percentage of crosslinking is associated with the number of crosslinks.

For our material we will have 3 crosslinks for every 20 substrate monomers or 24 crosslinks for

our molecule. The reason for the low percent of crosslinking is because if crosslinking becomes

too high the structure will become stiff and inflexible.

Crosslinking Procedure

The HA derivative with the addition of ammonium peroxydisulfate under oxidizing conditions

such as an air interface leads to a spontaneous chemical crosslinking reaction resulting in the

formation of our new HA derivative final product.

This crosslinking introduced is done through a process know as free radical

polymerization which is also more commonly known as a method of chemical crosslinking. The

importance of this chemical crosslink is that it decreases the molecular freedom of the material.

Figure VII.d – Free radical crosslinking of hyaluronic acid units

21

This means that the chemical has less freedom to react with other substances that may be present

in the surrounding environment such as the knee’s synovial fluid. The reduction of molecular

freedom helps to stabilize the material by reducing the chance of degradation through bond

breaking from occurring. The crosslinking also helps to shield the active sites from enzymatic

degradation. Some of the changes that can accompany a chemical crosslink include a decrease in

solubility. Also a chemically crosslinked polymer in the presence of a solvent tends to swell as

the solvent penetrates the network that was shown earlier. This in turn creates a hydrogel

because of the solvent for this new material being water.

The viscosity of the substrate monomer should be fairly close to the current viscosity of

synthetic HA monomer due to the fact that the addition of a 2-vinyl is a relatively small change

in the base structure of hyaluronan. The significant increase in viscosity will be attributed to the

crosslinking and polymerization of hyaluronan. This is the reason that polymerization and

crosslinking are so important for the new material. From previous studies it has been found that

crosslinking a material to form a gel can lead to an increase in the zero shear viscosity of the

base material by about 80 times the original value.29 For the normal HA polymer the zero shear

viscosity has been found to be 0.2 Pa·s @ 37oC.29 By relating the idea of the zero shear

viscosity being approximately 80 times the original polymer’s viscosity after crosslinking, it

resulted in the new HA derived polymer’s zero shear viscosity after sterilization being around 16

Pa·s @ 37oC which is comparable to our competitor’s treatment viscosities. This is important

because it has been found through studies that the viscosity of current treatments appear to be

adequate. This viscosity has also been tested using the lubrication theory to find whether or not

this value would be adequate. The lubrication theory is explained later on. The main problem

with current treatments has been the treatment life. Current treatments have problems with

degradation which through the use of ammonium peroxydisulfate will be fixed. Ammonium

22

peroxydisulfate has been found to have the ability to protect a structure against denaturation and

degradation.25 This benefits the stability of the hydrogel that is formed through crosslinking.

Basic steps for final material on a lab basis

The construction of the final material will be performed in a step manner as follows:

1. Basic construction of the substrate monomer as described earlier.

2. Polymerization of the monomer using OTH to create the HA derivative.

3. The polymer will be diluted in a pure water solvent for remaining steps and final

injection.

4. Crosslinking the HA derivative using ammonium peroxydisulfate with a molar

concentration relationship of 20:100 (disulfate / HA monomer). The concentration of

our polymerized HA derivative should be around 2.8*10-4 mol/ml and our concentration

of ammonium peroxydisulfate should be around 5.5*10-5 mol/ml.

5. Repeat crosslinking step 3, 150 times. This will provide a large molecule connected

through a crosslinking network. This will also result in a structure with an approximate

weight of 3 million Daltons which is comparable in size to current treatments.

6. The final step in the creation of this new material is sterilization which consists of the

removal of remaining unreacted peroxydisulfate along with any by products before in

vivo application.

The final product our newly described solution will be an intra-articular injection also

known as viscosupplementation for patients with osteoarthritis. The reason that our treatment

will be done in this manner is because increases the likely hood of being approved by the FDA.

It is more likely to be approved because there are already intra-articular injections for

osteoarthritis treatment approved and on the market.31 Another reason for using this method of

23

injection is that the location that the gel is placed in the body is easier to control through intra-

articular injection.

Advantages

The reason that this new crosslinking method is important is because of the disulfates’

ability to stabilize a structure against denaturation and because it gives our solution a viscosity

that is desired for its use.25 This stabilization from denaturation will help keep the new HA

derivative from degrading. This will in turn lead to a longer life span for the HA that is being

injected and in turn lead to fewer treatments being required for the patients.

VII. Lubrication Theory

To assess the effect HYAL-VYNE will have on the lubrication of the knee joint,

lubrication theory was studied to predict the minimum film thickness. The minimum film

thickness in any system is the distance between two surfaces separated by a lubricating substance

when a compressive force is applied. This will be calculated for the two articular surfaces of the

knee with our new hydrogel as the lubricant to ensure that the increased viscosity will not create

too much separation and not restrict the patient’s movement.

Several types of lubrication exist, so the first step is to determine which applies to the

synovial joint. The main types that could possibly apply to our system are hydrodynamic,

boundary, and elastohydrodynamic. Hydrodynamic lubrication was first used to explain the low

friction of human joints.47 It is based on the pressure forces formed between two surfaces in the

wedge configuration that occurs with cylindrical bearings. The idea is that the moving surface

will carry with it a fluid layer at the same velocity, and that this velocity will diffuse through the

rest of the fluid. Therefore the fluid entering the gap between the wedge and the moving surface

must partially move backward causing a pressure that supports the load. However, this model

has been found to work best for high speeds and low loads while the joint is characterized by low

24

speeds and high loads. Although it is now widely accepted that hydrodynamic lubrication does

not apply for biological joints, it was the only system available for nearly 30 years.

Another attempt at classifying the lubrication of the knee joint was boundary

lubrication.47 This type is different from hydrodynamic lubrication as it depends mostly on the

quality of the moving surfaces and the lubricating fluid. However, the viscosity of the lubricant

and geometry of the system are unimportant. Boundary lubrication can explain the low

coefficient of friction encountered in the synovial joint as motion begins, but it does not take into

consideration non-Newtonian behavior that exists in natural synovial fluid as well as the new

hydrogel.

To better characterize systems that include a solid surface covered with an elastic

material, elastohydrodynamic lubrication was developed. This type applies to systems of solid

surfaces of low geometric conformity that undergo elastic deformation and have a minimum film

thickness on the order of 1 µm.48 Within elastohydrodynamic lubrication are 4 regimes that can

be used to simplify the solution of the complex equations: (1) rigid-isoviscous, (2) rigid-

piezoviscous, (3) elastic-isoviscous, and (4) elastic-piezoviscous. Because the pressures incurred

in the knee joint are not high enough to cause a non-uniform viscosity in the synovial space, our

system falls into the elastic-isoviscous regime. The geometry of the knee joint can be estimated

as two convex bodies that would touch at a point if allowed. For point contact in the elastic-

isoviscous regime, the minimum film thickness can be determined by first calculating the film

thickness parameter, gHmin,

[ ]κ31.067.0min 85.0170.8 −−= egg EH

where gE is the elasticity parameter, and κ is the ellipticity, both defined below.

2/1U

WgE = (VII.1)

25

636.0

0339.1

=

x

y

R

Rκ (VII.2)

In the above equations, W is the load parameter, U is the velocity parameter, and Rx and Ry are

the radii of the ellipse in the x and y directions, respectively. We have estimated the ends of the

femur and tibia to be spheres with 20 mm radius. The load and velocity parameters are defined as

'2ER

wW

x

= (VII.3)

'

0

ER

VU

x

µ= (VII.4)

where w is the compressive load, E’ is the effective elasticity modulus, µ0 is the viscosity of the

lubricant, and V is the velocity parameter. E’ and V are defined as

−+

−=

2

22

1

21 11

2

1'

EEE

νν (VII.5)

( )22 ν+= uV (VII.6)

where E1 and E2 are Young’s modulus for each surface, v1 and v2 are the Poisson ratios for each

surface, and u is the speed of the moving surface. Based on typical values for the human knee

and articular cartilage, the parameters used in these equations are as follows:

Poisson’s ratio: v1 = v2 = 0.45

Young’s modulus: E1 = E2 = 12 MPa

Speed: u = 0.03 m/s

Radius: Rx = 0.002 m

Load: w = 3.26 MPa

The only parameter that was actually based on our new hydrogel was the viscosity. We used 16

Pa·s for these calculations as it was found from similar studies to exist under conditions such as

26

walking. For this reason, the compressive load was chosen to be 3.26 MPa which is typical of an

adult male during walking. With all the parameters defined, the minimum film parameter is

( )

( ) ( )

( )( )( )( )

( ) ( )[ ] 50339.131.07min

72.1

7

6

7

62

6

22

22

1050.685.011086.77.8

1086.70245.0

1023.1

0245.0m 002.0Pa 100665.0

2034.0sPa 16

1023.1m100665.0m 002.0

Pa 1026.3

2034.045.003.0

MPa 0665.012

45.01

12

45.01

2

1'

×=−×=

×=×=

=×

⋅=

×=×

×=

=+=

=

−+−=

−eg

g

U

W

V

E

H

E

The film thickness parameter is related to the actual film thickness by

x

H

RHh

HU

Wg

⋅=

=

(VII.7, 8)

where H is another film thickness parameter, and h is the film thickness. Therefore, the

minimum film thickness for our new gel in the synovial joint is

( )

( )( ) mh

H

µ6.2106.2002.00013.0

0013.01023.1

0245.01050.6

6

7

5

=×==

=

××=

−

If these same calculations are done for natural synovial fluid using a viscosity of 10 Pa·s,

which is typical for walking, a minimum film thickness of 1.9 µm is obtained. This indicates

27

that the increased viscosity of HYAL-VYNE will not affect the minimum film thickness

significantly and the range of motion should be retained.

VIII. FDA Approval Process

As was mentioned in the objective of this project, one of the requirements, and possibly

the most crucial to the survival of the project, is the approval of the Food and Drug

Administration (FDA) for the new product HYAL-VYNE. This process takes extensive amounts

of time and money, with risk of complete failure. The average cost of the entire FDA approval

process, from pre-clinical testing to actual review by the FDA to post-approval costs for a new

drug is $800-900 million as reported in 2003 taking anywhere between 4 and 12 years23. Before

beginning, our new gel of crosslinked hyaluronic acid derivative must be classified as either a

drug or a device. According to the Food, Drug, and Cosmetic Act, a drug is an article

“…intended for use in the diagnosis, cure, mitigation, treatment, or prevention of disease in man or other animals; and…articles (other than food) intended to affect the structure or any function of the body of man or other animals…,”22

while a device is defined as

“…an instrument, apparatus, implement, machine, contrivance, implant, in vitro reagent, or other similar or related article, including any component, part, or accessory, which is… intended for use in the diagnosis of disease or other conditions, or in the cure, mitigation, treatment, or prevention of disease …and which does not achieve its primary intended purposes through chemical action within or on the body of man or other animals and which is not dependent upon being metabolized for the achievement of its primary intended purposes.”22

Based on these definitions, as well as on the classification of the similar products Synvisc and

Hyalgan49, the intra-articular hydrogel falls under the definition of a medical device because of

its intent to treat and slow the progression of a disease (osteoarthritis), and because it does not

achieve this primary purpose through metabolic processing.

The branch of the FDA that deals with medical devices is the Center for Devices and

Radiological Health (CDRH), which places devices into three categories based on the control

28

required to ensure safety and effectiveness of the device. A description of each classification as

provided by the CDRH follows.

Class I – General Controls:

This class is subject to the least stringent control as these devices pose minimal

risk to consumers and are most often simpler in design than those in Class II and

III. They are subject to the general controls which include establishment

registration, medical device listing, and labeling. A premarket notification must

be filed before marketing. Examples in this class include elastic bandages, latex

gloves, and hand-held surgical tools.

Class II – Special Controls:

This class is also subject to the general controls described for Class I devices but

these regulations alone may not be sufficient to ensure safety and effectiveness

although methods to do so are already in existence. These devices may also be

subject to special controls such as special labeling, mandatory performance

standards, and postmarket surveillance. Examples in this class include powered

wheelchairs and surgical drapes.

Class III – Premarket Approval:

This class is subject to the most stringent control as the general and special

controls alone cannot provide sufficient information to ensure the safety and

effectiveness of these devices. These devices often have the purpose of

supporting or sustaining human life and/or preventing impairment of health while

posing a risk of injury or illness. All Class III devices require a premarket

approval. Some examples in this class are replacement heart valves, silicon breast

implants, and implanted cerebella stimulators.

29

Because OsteoGel does pose serious risks such as localized pain, infection, and rejection by the

body, it will fall under the Class III medical devices and will require premarket approval.

For FDA approval of Class III devices, the premarket approval (PMA) is used to review

the extensive non-clinical and clinical data to assure that the device is safe and effective and will

ultimately determine approval for marketing. The FDA is officially given 180 days to review a

submitted PMA, but this often takes longer. If the PMA is denied, the applicant – the person,

company, or institution seeking approval – has 30 days to petition the decision in order to be

reconsidered. Because approval is based on the content provided in the PMA, it is vital to

provide adequate technical data in addition to thorough background information and device

purpose and description. Although not the only deciding factor, the technical section must

provide the data necessary for the FDA to make their decision. This is normally divided into two

sections: non-clinical laboratory studies and clinical investigations. The non-clinical section

should contain information on microbiology, toxicology, immunology, biocompatibility, stress,

wear, and other laboratory or animal tests, while the clinical investigations section should

include study protocols, safety and effectiveness data, adverse reactions, complications

encountered, device failures, patient information and complaints, data tables, and statistical

analyses50

There are two types of PMA that can be submitted, traditional and modular. With a

traditional PMA all necessary data are first collected then presented all at once to be reviewed for

approval. This type of PMA is most often used for devices that have already undergone clinical

testing and/or have previously been approved by a medical regulation organization of another

country. With a modular PMA, the necessary contents are broken down into specific

components, or modules, where the research group completes one subset of experiments, submits

the results, and waits for approval while concurrently working on the next module. This type of

30

submission is best for products that are still in the early stages of clinical trials. While the

modular PMA may take more total review time than the traditional PMA, we will follow the

modular type for this project since we are in the still in the pre-clinical stage. Also to avoid the

risk of spending extensive amounts of time and money to complete the entire testing process

before submitting for review, we see it best to ensure approval after smaller portions are

complete so that in the case of FDA denial, only that particular module will need to be revised

and repeated.

The modular premarket approval will be divided into three discrete sections. The FDA

does not provide an actual form to follow for each module, but instead requires a PMA shell to

first be submitted. The PMA shell provides an outline of the testing that will be performed in

each module and gives a time line of approximately when each module will be completed. A

team from the FDA will review the PMA shell and provide assistance to develop a customized

plan to fit the particular device and situation. We have designed each module in attempt to best

manage the time and scope involved for each one; this is summarized in our PMA shell shown in

Table VIII.a

Table VIII.a - PMA Shell

HYAL-VYNE

Module Number Contents Time to Complete

1

Nonclinical Laboratory Studies: • Physical and Chemical Property Tests • Degradation Tests • Toxicity Tests

3 Years

2

Nonclinical Laboratory Studies: • Animal Testing • Sterilization and Packaging • Injection Procedure

3 Years

3

Clinical Studies: • Human Patient Testing • Physician Instructions • Patient Instructions

5 Years

31

Because of the numerous possible scenarios that could occur during these experiments,

we have created a model based on the work of Kim, Shreve, and Yankovich for a relatively

similar proposal.50 Before beginning the actual laboratory testing for FDA approval, several

choices must be made. Most important are the number of PhDs and lab technicians that will be

employed to conduct the experiments as well as the number of experiments they will perform.

This is an important decision because with a PhD salary of $66,000/year and a lab technician

salary of $26,000/year, this will account for a large portion of project expenses. The number of

each type of experiment to be performed is important because while the more experiments that

are run the more consistent the results will be, at the same time the greater the total cost and time

spent will be. Therefore, three different combinations of employees will be considered – 1 PhD

and 10 lab techs, 1 PhD and 5 lab techs, 1 PhD and 2 lab techs – along with three selections for

the number of experiments – 100, 85, 50 – that will be performed to prove consistency of results.

Figure VIII.a shows a decision tree for the number of employees, number of experiments, and

the amount of time required based on these decisions.

32

Stage One Variables

Determine Number of workers

1 PhD10 Lab Technicians

$415,000

1 PhD5 Lab Technicians

$245,000

1 PhD2 Lab Technicians

$143,000

Determine Number of Experiments

Determine Number of Experiments

100 Experiments$170,000100 Days

85 Experiments$150,00090 Days

50 Experiments$100,00065 Days

Determine Number of Experiments

100 Experiments$170,000105 Days

85 Experiments$150,00095 Days

50 Experiments$100,00070 Days

100 Experiments$170,000150 Days

85 Experiments$150,000140 Days

50 Experiments$100,000120 Days

Begin Pre-FDA Experimentation

The specific tasks included in each module along with probabilities assigned for success

and for failure due to various scenarios are presented in the following sections so that later a risk

analysis for the entire FDA approval process can be performed. The probabilities that will be

seen in each module are assigned based on anticipated difficulty of the particular experiment,

predictability of results, and most importantly on the number of experiments performed.

Probabilities are given for each set of number of experiments – 100, 85, 50 – with the likelihood

of success decreasing with decreasing number of experiments; the fewer experiments that are

performed, the larger the deviation of results, and therefore the greater the possibility of rejection

Figure VIII.a – First stage decision: Number of emp loyees and experiments

33

by the FDA. It is important to note that the number of employees does not affect the success or

failure rates. This decision only determines the amount of time and money required for each set

of experiments.

Module 1

Module 1 will consist of extensive laboratory tests to determine material properties such

as compressive and tensile strength, viscosity, elasticity, coefficient of friction, and molecular

weight of HYAL-VYNE as well as degradation rates and toxicity to cells. These experiments

are summarized below.

• Polymer Synthesis

The synthesis procedure for the novel HA derivative has already been described

in detail (see section VI) and therefore will not be repeated here. However, after the

enzymatic polymerization of the substrate monomer, it is important to know the actual

degree of modification of the N-acetyl reactive site with the newly introduced 2-vinyl

functional group. This will be quantified using H-NMR spectroscopy with the Varian

Inova-500 spectrophotometer. Because the hydrogens of the 2-vinyl group will resonate

at a different frequency than those of the methyl group of a non-substituted HA

monomer, two different peaks will be produced on the H-NMR spectrum. To determine

the degree of modification, the relative integrations of the 2-vinyl and methyl protons will

be analyzed.

It is unlikely the FDA would deny the proposal based on the polymer synthesis,

but these experiments will provide data to prove consistency in the product produced

from the devised steps. These experiments are also helpful in ruling out the possibility of

an improperly modified polymer or contaminants in the case of failure in other steps.

34

The H-NMR experiments will take a minimal amount of time since only a very

small sample is required for analysis, which can be performed at the same time as other

experiments.

• Degree of Crosslinking

Although we have previously predicted a theoretical degree of crosslinking of the

new HA hydrogel, the actual degree of crosslinking will be determined by a method

similar to that described by Miyamoto et al.51 1 g of the gel will be dissolved in 1.0 ml of

1.0 M sodium hydroxide and shaken on an orbital shaker for at least 1 hour. Next 1.1 ml

of 1.0 M HCl will be added to the solution, followed by extraction of the HA sample with

ethyl acetate. Then solvent will be evaporated. The sample will then be analyzed using

high performance liquid chromatography (HPLC) equipped with an ODS column and a

UV detector from Toshoh Corp.

• Force Profiles

A Surface Forces Apparatus (SFA) Mark III will be used with a geometry similar

to that of the human knee to create a force profile using our new HA gel as the lubricant.

The SFA Mark III consists of two crossed cylinder plates separated by a gap of distance

D in which the gel is placed. A normal force is applied to the upper plate while the lower

plate is connected to a spring, and the induced deflection is measured to determine the

force between the two plates for various gap distances. This information will be used to

assure that the gel will be able to hold loads typical of the knee joint, which can reach a

maximum of 473 psi during walking, or up to 50% higher in extreme motions such as

running or stair climbing.47 Because the gap distances in the synovial joint can vary from

50 to <0.8 µm, D will be varied from 100 µm to 0.01 µm to develop a force profile over

the entire range of distances experienced in the knee.

35

• Viscosity Measurements

Because we are attempting to provide improved viscoelastic properties over

existing synovial fluid, viscosity measurements will also be taken using the SFA Mark III

in order to ensure that the viscosity of HYAL-VYNE is in fact equal to or greater than

that of natural synovial fluid, which has a viscosity of 300 cP or more depending on

shear. This is done by applying a sinusoidal normal motion of specified amplitude and

frequency to the upper plate of the apparatus. Normal walking cycles have a frequency

of approximately 1 Hz, so our gel will be tested for a frequency range of 0.01 Hz (to

simulate rest) to 10 Hz (significantly faster than normal running). The viscosity is then

calculated for these conditions by

2/12

221

12)(

−

=A

A

vR

KDD o

πη

where η(D) is the viscosity at gap distance D, K is the spring constant, R is the radius of

the curved plate of the apparatus, v is the frequency, Ao is the applied amplitude, and A is

the amplitude of the vibrating surface. To best mimic the human knee, the radius R of

the plate will be 25 mm.

• Hydrogel Degradation

To analyze the rate of degradation of the new crosslinked HA hydrogel, the gel

will be subjected to enzymatic degradation experiments as suggested by Leach et al.52 To

test the enzymatic degradation, the hydrogel will be incubated at 37°C in solutions of

various concentrations of the enzyme hyaluronidase (2.6 µg/ml, 5 µg/ml, 50 µg/ml, and

500 µg/ml in citrate buffer) to create a profile of percent mass degraded relative to

concentration of enzyme. Hyaluronidase (HAase) is the enzyme secreted naturally in the

synovial fluid and appears at an average concentration of 2.6 µg/ml and has an active life

36

of approximately 5 hours. First, the gel samples will be placed in 24-well plates and

soaked in citrate buffer overnight to reach their equilibrium swelling. The next day, the

gels will be removed, dried of excess buffer with filter paper, and weighed to determine

their initial mass. Then 0.5 ml of bovine testicular hyaluronidase will be added to the

samples. The well plates will be gently agitated on an orbital shaker at 37°C. After

every hour since HAase was added, the gel will be removed from the well, excess

enzyme solution blotted away, the gel reweighed, and fresh HAase solution added. This

will be done for 48 hours for each enzyme concentration in triplicate to obtain consistent

results. Three control samples of the gel in citrate buffer only and three samples of

synthetic HA (unmodified, no crosslinks) in the each of the concentrations of HAase will

also be included for comparison.

Because of the functional group modification and more importantly the crosslinks

that are introduced to hyaluronic acid, we anticipate a significantly reduced degradation

rate; however, if the FDA yields this as a reason for denial of the PMA, we will reassess

the chosen modification and crosslinking densities. The degree of modification with 2-

vinyl can be altered by changing the concentrations and reaction times from steps vi, vii,

and viii in the “bottom-up” synthesis. The degree of crosslinking can be changed by

changing the concentration of ammonium peroxydisulfate and/or the number of times the

crosslinking step is repeated. For this reason, there is a relatively high probability that

failure will occur due to rapid degradation times; however, because this problem can be

readdressed easily, the probability of this being a reason for failure in later modules is

significantly lessened.

• Cytotoxicity Tests

37

To test the toxicity at the cellular level, we will use the elution method which is

widely used for medical device testing.53 L929 mouse fibroblast cells will be grown in

cell culture medium to produce a monolayer of cells attached to the bottom of culture

flasks. The new hydrogel will be placed in cell culture medium and incubated at 37°C

for 24 hours. The culture medium from the gel samples containing any extractable

fragments will then replace the culture medium of the cells. Each day after exposure, the

cells will be monitored under a microscope to check for cell death (dead cells will detach

from the flask bottom) and changes in configuration such as cell size and appearance for

1 week.

If these tests indicate significant toxicity due to the HA gel, we will first run the

experiment again to try to rule out the possibility of random contamination. If the

problem persists, we will then test for impurities such as reagents remaining from the

synthesis steps by using H-NMR spectroscopy as discussed earlier.

These tests will take approximately 5 months to complete since the cells will need

to be cultured for 7 days prior to and 3 days after exposure and only 5 flasks will be

tested at one time so that in the case of contamination, only this small number will need

to be retested.

Figure VIII.b shows the possible pathways that could occur after Module 1 is submitted

to the FDA. The probability of each path of the module is given for each number of

experiments; the probability for success is higher when more experiments are performed.

However, because a greater number of experiments also corresponds to higher costs and more

time, there must be an optimum that will produce the minimum net present cost. Microsoft

Excel was used to generate values for this for each possible pathway through the PMA modules

and will be discussed after the modules are presented.

38

Module 1 Testing$500,0001 year

Module 2

Module 1 testing$500,0003 years

Abandon Project

Failure due to polymer synthesis

4%, 5%, 7%

Failure due to mechanical properties

9%, 12%, 15%

Approval Granted

60%, 45%, 30%

Approval Granted

90%, 85%, 75%

Failure due to polymer synthesis

2%, 3%, 5%

Failure due to purification problems

9%, 11%, 15%

Change polymer synthesis procedure$15,0001 months

Change crosslinking procedure$60,0003 month

Change purification procedure$12,0001 months

Reapply for Module 1

Failure due to crosslinking reaction

3%, 6%, 9%

Failure due to degradation

5%, 6%, 11%

Abandon Project Abandon Project

Failure due to degradation

12%,15%, 21%

Failure due to poor cell survival

6%, 8%, 12%

Add additional crosslinks$40,0002 months

Abandon Project

Module 2

Module 2

Module 2 (see Figure VIII.c on the next page) will consist of animal testing where the

new crosslinked hyaluronic acid derivative will administered to guinea pigs, rabbits, pigs, and

dogs. The tests will be conducted in order of increasing biological complexity starting with the

smallest animals, and once satisfactory results are obtained, the next larger animal will be used,

and so on, concluding with dogs, which should have the greatest similarity to humans. The

Figure VIII.b – Module 1: Laboratory Testing

39

experiments to be run for this section are outlined below followed by the pathways for Module 2

with probabilities assigned for anticipated failures due to biocompatibility, retention time,

degradation, and injection procedure.

• Biocompatibility and Biodegradation Tests

To evaluate the biocompatibility and biodegradation of the crosslinked HA

derivative in animal models, the gel will be injected subcutaneously into guinea pigs

similar to the method described by Miyamoto et al.51 A group of 40 6-week old guinea

pigs (20 male, 20 female; Elm Hill Breeding Labs, Chelmsford, MA) will be given

sodium pentobarbital for general anesthesia, then a large area of dorsal skin will be

shaved. The injections will be administered in 0.2 ml samples through a 25-gauge

needle. The guinea pigs will be divided into two groups of 20 (equal gender in each

group). Group 1 will receive 3 injections at separate locations with samples of the new

gel of crosslink densities 15%, 20%, and 25%. Group 2 will serve as a sort of control

group each receiving samples of noncrosslinked HA derivative, unmodified HA, saline,

and competitor product Hylan G-F 20. Each day the injection locations will be observed,

looking for signs of bioincompatibility such as erythema (redness of the skin due to

inflammation), edema (swelling due to fluid accumulation), and vesiculation (blistering),

as well as signs of degradation seen as a reduction in initial protrusion of skin from the

injection material. At the end of days 3, 7, 14, 21, and 28, 4 animals from each group (2

males, 2 females) will be sacrificed and samples of their injection sites removed. These

specimens will be placed in 10% neutral buffered formalin, paraffin embedded, and

sectioned. Part of these sections will be stained with hematoxylin and eosin while the

others will be reviewed histologically. The hematoxylin and eosin stains allow for

microscopic examination to score the degree of inflammation based on the number of

40

cells accumulated in the skin sample. Histological review of the sections will show the

dispersion of the gel into surrounding tissue, indicating the degree of degradation during

the specific time period.

If the FDA determines that HYAL-VYNE is not biocompatible based on the

results obtained in Module 2, the project will be abandoned due to the high probability

that if the gel is non-biocompatible in animals, it will also be so in humans. On the other

hand, if the FDA declines approval due to rapid degradation, we will reconsider using a

higher crosslinking density to introduce more sulfate groups. However, there is not much

room left for change in this area because as more crosslinks are added to the molecule,

the more stiff the gel becomes, possibly becoming too stiff and losing its functionality.

Therefore, we will also consider changing the crosslinker used, ammonium

peroxydisulfate, to another sulfate-containing crosslinking agent in attempt to keep the

sulfate’s stability against degradation.

Since the longest the guinea pigs will be used in this test is 28 days, it is estimated

that these biocompatibility and biodegradation tests will take 30 days to complete with an

approximate cost of $6,000.

• Immunogenicity Test

To determine whether the injected gel causes an immunogenic response, i.e. the

production of antibodies to fight the foreign substance, the method of Miyamoto et al will

be used.52 A tissue sample from the guinea pigs from the biocompatibility tests that were

sacrificed after 28 days will be tested for the presence of antibodies by passive cutaneous

anaphylaxis (PCA) analysis. This will detect any biologically active antibodies,

especially immunoglobulin G1 and immunoglobulin E, by adding a dye solution to the

test serum. If a blue spot is observed, the test is positive for antibodies. Each of the

41

injection areas on the 8 animals from Day 28 will be tested along with 1 male and 1

female that had not be subjected to any injections.

If consistent positive results are obtained from the PCA analysis, the project will

be abandoned due to the high probability that if the gel produces an immunogenic

response in animals, it will do so also in humans.

This portion of the Module 2 testing is estimated to take 1 day at a cost of $400.

• Effectiveness

The effectiveness of the new hydrogel will be tested on rabbits, pigs, and dogs.

The rabbits will first be administered the injection by a veterinary surgeon. After each

animal receives the injection, they will be monitored for signs of change in mobility

relative to their actions prior to treatment. Because we want to know how long the

effectiveness of the treatment lasts, we will monitor some of the animals for as long as 6

months. If it appears that the animals are still in pain and do not increase their activity,

the FDA may deny request for approval. In this case we may have to reevaluate the

crosslinking density because the gel likely is not being retained in the synovial cavity. If

there appears to be no improvement after injection, another possible problem could lie

within the injection procedure. The surgeon may have not gotten the hydrogel into the

proper space, therefore allowing the gel to disperse throughout the body. In this case we

will need to examine other possible methods of administration.

This part of Module 2 will take the most amount of time, possibly up to 3 years in

order to get sufficient data on the effectiveness and treatment life of the new injection.

42

Module 2 Testing$500,0003 years

Module 3

Module 2 testing$500,0002 years

Failure due to lack of retention

12%, 15%, 19%

Failure due to biocompatibility issues

4%, 7%, 8%

Approval Granted

65%, 50%, 38%

Approval Granted

70%, 63%, 45%

Failure due to injection procedure

7%, 9%, 11%

Reevaluate molecular size

$40,0002 month

Additional purification, sterilization$15,0001 months

Test new injection procedures$50,0002 months

Reapply for Module 2

Failure due to in vivotoxicity

5%, 8%, 9%

Failure due to degradation

7%, 11%, 15%

Abandon Project

Alter polymer formulation or M.W.

to find optimal$30,0002 months

Abandon Project

Failure due to biocompatibility

5%, 7%, 12%

Failure due to in vivotoxicity

5%, 7%, 12%

Failure due to degradation

10%, 12%, 16%

Abandon Project Abandon Project

Failure due to injection procedure

10%, 11%, 15%

Purchase advanced equipment$200,0001 month

Module 3

Reapply for Module 2

Module 2 testing$500,0001 year

Approval Granted

90%, 80%, 60%

Failure due to surgical procedure

10%, 20%, 40%

Module 3 Abandon Project

Figure VIII.c - Module 2: Animal Testing

43

Module 3

The final stage before approval is granted consists of extensive human clinical trials. In

this stage people suffering from osteoarthritis will be administered the treatment or placebo to

determine the efficacy and safety of the treatment. Again, each scenario has been assigned a

probability so that a complete analysis of the module pathway and the costs associated with each

path can be generated (Figure VIII.d).

• Safety

HYAL-VYNE will be injected into the synovial cavity of the knee joint of 100

human patients. On days 1, 7, 21, and 28 (where day 0 = day of injection) the patients

will be examined to check for signs of infection, excessive swelling and tenderness, and

fever. Swelling and pain near the injection site is normal for similar products and will

likely occur with HYAL-VYNE as well, but this should clear up within about a week and

can be treated with anti-inflammatory drugs. If inflammation persists after the second

week in several patients, the FDA may find this as reason to deny approval, and the

project will have to be abandoned. Fever is often in indication of infection, so patients

experiencing fever will be treated with antibiotics. If the fever subsides, there likely was

an infection present. Again, if this occurs in several patients, the FDA will likely find

this as reason to deny approval, and the project will have to be abandoned. Patients will

also be asked if any other problems have been noticed since administration such as

abdominal pain, nausea, diarrhea, headaches, body aches, fatigue, or dizziness. Since

hyaluronic acid is created naturally in the body and because it should not be metabolized,

we do not anticipate these side effects to arise but obviously need to note them if they do.

• Effectiveness

44

Before receiving the injection, each patient will be asked to rate their current pain

and ability to perform certain activities such as walking, stair-climbing, bending,

standing, etc. Once the patients have been administered the new HA injection, they will

be asked to return on days 1, 7, 21, and 28 to examine the effectiveness of the gel at

relieving their symptoms of osteoarthritis and returning their mobility. They will rate

their pain on a scale of 1-10, ten being the most painful. The patients will also be asked

to rate their range of motion, that is, how much more or less they are able to extend and

flex the knee as well and their ability to apply weight during normal activities such as

those listed above. These will also be rated on a scale of 1-10, ten being not able to move

at all.

If several patients do not experience significant pain relief and return of mobility,

the FDA will possibly deny approval. If this is the case, it is likely due to lack of

retention after injection and we will have to reassess our decisions for crosslinking

density and molecular weight and then reapply for Module 3. It is fairly unlikely that this

will happen because of the numerous tests already completed in Module 1 and 2.

45

Module 3 Testing$4,000,0005 years

Medical Market

Module 3 testing$4,000,0004 years

Abandon Project

Failure due to low retention time

15%, 20%, 22%

Failure due to rapid degradation

20%, 25%, 33%

Approval Granted

55%, 40%, 30%

Approval Granted

70%, 60%, 35%

Failure due to damaged surrounding tissue

5%, 9%, 15%

Failure due to inflammation of

surrounding tissue

10%, 15%, 15%

Change molecular size, crosslink

density$300,0006 months

Increase functional group substitution, crosslink density

$500,0008 months

Treat with anti-inflammatory medication$300,0006 months

Reapply for Module 3

Failure due to rapid degradation

13%, 16%, 25%

Increase functional group substitution,

crosslink density$50,0001 month

Failure due to lack of benefit over current

treatments

10%, 18%, 28%

Change polymer constiuent

concentrations$750,0006 months

Improve degradative properties of polymer

$500,0001 year

Reapply Module 3 and Testing$4,000,0002 years

Failure

10%, 20%, 30%

Abandon Project

Failure due to low retention time

12%, 15%, 25%

Change molecular size, crosslink

density $500,0001 year

Reapply for Module 3

Module 3 testing$4,000,0003 years

Approval Granted

80%, 70%, 55%

Failure due to poor efficacy

10%, 12%, 17%

Approval Granted

90%, 80%, 70%

Medical Market

Medical Market

Medical Market

Figure VIII.d - Module 3: Clinical Trials

46

IX. FDA Risk Analysis

Before the FDA process can be completed first stage decisions must be made such as

number of employees and number of experiments. These variables were important because they

were needed to determine the time and money that would be required for the FDA process to be

completed. To decide the best combination of workers and experiments, a risk analysis was

done. The risk was based upon the possibility of failure or success of the project with respect to

the net present value (NPV) of the project. There were 9 possible scenarios associated with this

risk analysis. The possible scenarios can be seen in Table IX.a with the number in each center

box being associated with the number of days required for completion of the given number of

experiments with the combination of workers. The days required were estimated by looking at

the time it takes for each experiment to run and the time that can be cut down by each worker.

100 Experiments 85 Experiments 50 Experiments

10 workers 100 90 65

5 workers 105 95 70

2 workers 150 140 120

The overall probabilities of success or failure were determined by looking at each module

and the probabilities associated with each step in the module occurring. In our case there were

more than 1800 pathways in which success or failure could occur. When the numbers of

experiments and workers is considered there are more than 9000 possible ways that success can

possibly be reached. To narrow down the best choice for the number of workers and

experiments risk analysis was performed. Figure IX.a shows risk associated with 5 lab

technicians. From looking at the figure it can be seen that 85 experiments with 5 techs is the

riskiest combination in comparison with 50 and 100 experiments. The scenario consisting of 85

Table IX.a Time required in days for various combinations of experiments and workers

47

experiments and 5 techs led to a probability of around 0.37 for not making money. The 50 and

100 experiment scenarios were shown to have about the same probability as each other having

probabilities of around 0.26. It can also be seen from the figure that 85 experiments and 100

experiments had the chance to be the most profitable giving net present values of around $295

million

5 Lab Technicians

0

0.1

0.2

0.3

0.4

0.5

0.6

-25 -5 15 35 55 75 95 115 135 155 175 195 215 235 255 275 295

NPV (million$)

5 techs 100 exp

5 techs 85 exp

5 techs 50 exp

The optimal combination that was found for our circumstances was found using 10 lab

techs and 85 experiments. The risk associated with 10 lab technicians can be seen in Figure

IX.b. From the graph it can be seen that the risk of losing money for 100, 85, and 50

experiments is very close with a probability of 0.26. With our optimal choice it was found that

the NPV could reach as much as $295 million. The reason that this combination of 10 lab techs

and 85 experiments was optimal was because it gave the least amount of risk for the time

Figure IX.a Risk Analysis on 5 lab technician s

48

required to complete the FDA process. By having more lab techs the time was able to be

decreased for the FDA process which helps to get our product out on the market more quickly.

10 Lab Technicians

0

0.1

0.2

0.3

0.4

0.5

0.6

-25 -5 15 35 55 75 95 115 135 155 175 195 215 235 255 275 295

NPV (million $)

10 techs 100 exp

10 techs 85 exp

10 techs 50 exp

X. Demand

The demand for our solution was determined through a series of steps. The first step in

determining the demand was a happiness function. The happiness function for the new

treatment was based off of the current treatment with a happiness value of 40%. It was

determined by assigning values to three different variables. The three variables were treatment

life, cost, and pain of injections. Treatment life was the most important of the three variables

given a weight of .75, while the other two variables were given values of .125. These variables

were all related to different parameters such as amount of crosslinking, size of injections, amount

polymerized, and length of polymer chains. The purpose of the happiness function was to

provide a relationship between current treatments and our treatment. The relationship between

these two values can be seen in X. 1 which was used to find a value called β.

Figure IX.b Risk analysis on 10 lab technicians

49

happinesstreatmentnew

happinesstreatmentold

__

__=β (X.1)

It was assumed that β would change from year to year because of competition improving

treatments to compete with our treatment. It was also assumed that after 5 years this new

treatment will be replaced by a treatment that is better. Another variable that was important that

was used with β was α. The value α was important because it was based on how much someone

was going to know about our treatment in relationship to current treatments. For example if both

treatments were known about equally then the value would be 1, but if our treatment is know half

as much as current treatments then the value is 0.5. The value of α was assumed to increase each

year because of advertisement and information about our treatment spreading through word of

mouth. Table X.i below shows values of α and β that were used over the 5 year period.

Year α β

0 0 0.645

1 0.15 0.715

2 0.4 0.785

3 0.89 0.855

4 0.99 0.925

5 1 0.995

These values were applied to equations X.2 and X.3. Important values that had to be

determined before these equations could be used were our treatment price (p1), our competition’s

treatment price (p2), the total market demand for our product (D), and currently what is being

spent on a yearly basis for treatments (Y). From research it was found that our current

Table X.a. Values of α and β over the life of the treatment

50

competition’s treatment cost about $1250 per year and that around 7 million people per year visit

physicians for OA treatments. It was also found that approximately $1.5 billion a year is spent

on viscosupplementation treatments. With this information it was determined that for our

treatment to be competitive it should be priced at about $2400 per treatment because of the

longer life of our treatment. Using the gathered information we were able to apply equations X.

2 and X. 3 to find our treatment’s demand (d1) and our competition’s demand (d2).

Ydpdp ≤+ 2211 (X. 2)

β

α

αβ2

12211 d

ddpdp = (X. 3)

Combining X.2 and X.3 led to the formation of X.4 and X.5. These equations were used to find

d1 and d2 which had to fit within the boundaries of X.6.

α

βα

112

1

21

21 dd

p

p

p

Y

p

pd

−

= (X. 4)

12

1

2

2 dp

p

p

Yd −= (X. 5)

Ddd ≤+ 21 (X. 6)

Using X.4 a graph was able to be created that related demand. This graphical relationship was

used to find the demand. The demand was determined to be the place of intersection between the

two lines that were graphed where the maximum demand occurred. For example in year 1 you

can see from graph 1 that the demand is around 370,000 people. This same idea was applied to

years 2, 3, 4, and 5. From using the same principles it was found that in year 2 the demand

would be approximately 620,000 people and in years 3, 4, and 5 it would be a maximum at

around 625,000 people.

51

demand in the market

0

100,000

200,000

300,000

400,000

500,000

600,000

700,000

0 100,000 200,000 300,000 400,000 500,000 600,000

d1

d1

From calculating the demand for our treatment (d1) we were able to solve for the competitions

demand (d2). These demands were checked in X.6 to make sure that they were not greater than

the overall demand.

XI. Cost Analysis

Although the primary goal of this research is to develop a novel, more effective method

of relieving the symptoms of osteoarthritis, attention must be given to the economics involved.

To analyze this project from a business point of view, the total capital investment, total product

cost, annual cash flow, net present worth, return on investment, and pay-out time will be

discussed in this section.

Total Capital Investment

The total capital investment (TCI) is the money required to start a business

operation before revenue is available and is equal to the sum of the fixed capital

investment (FCI) and the working capital (WC).8

Graph X.a - Demand in the market at year 1

52

WCFCITCI += (XI.1)

The fixed capital investment is that necessary to obtain, develop, and equip the facilities

to be used for production, while the working capital is that needed to sustain operation.8

To obtain a preliminary estimate, the method of percentage of delivered-equipment costs

as in Peters and Timmerhaus will be used with an anticipated accuracy of ± 20-30

percent, which is an acceptable approximation at this early stage of the proposal.8

Fixed Capital Investment

The fixed capital investment can be divided into direct and indirect costs, and

these can be estimated as a percentage of the total FCI as given by Peters and

Timmerhaus (241).

To obtain a better assessment for the cost of our process, the direct and indirect

costs were estimated as a percentage of the delivered equipment costs after a list of

purchased equipment was generated as shown in TableXI.i. The list includes major

laboratory equipment and accessories needed for pre-clinical experiments, as well as

many of those required for clinical testing. There of course are items that will be added

as the need arises, hence the inclusion of miscellaneous cost. The prices shown were

found from online product catalogues of Fisher-Scientific and VWR.8,10 The percentage

estimates presented in Peters and Timmerhaus (251) are for solid, solid-fluid, and fluid

processing plants, so because these are not suitable representations for biochemical

processes, an economic study done in the biotechnology industry was used to obtain

better ratios. TableXI.ii on the following page shows these ratios and the calculated

costs.

53

Table XI.a – Estimated equipment costs 8,10

Equipment Quantity Unit Cost Total Cost

Autoclave 1 $ 4,330.00 $ 4,330.00

Torque rheometer 1 $ 40,000.00 $ 40,000.00

Centrifuge 1 $ 2,995.00 $ 2,995.00

Computers 4 $ 1,200.00 $ 4,800.00

Culture Flasks 100 $ 1.30 $ 130.00

Eppendorf pipets 4 $ 345.00 $ 1,380.00

Falcon tubes 500 $ 0.45 $ 225.00

Glassware, stirrers, etc Multi $ 750.00

Hemacytometer 1 $ 215.00 $ 215.00

Incubator 2 $ 6,245.00 $ 12,490.00

Laminar hood 1 $ 7,421.00 $ 7,421.00

Liquid N2 1 $ 2,610.00 $ 2,610.00

Microscope 1 $ 750.00 $ 750.00

Orbital shaker 2 $ 550.00 $ 1,100.00

Osmometer 1 $ 8,418.00 $ 8,418.00

Pasteur pipets 1000 $ 0.08 $ 80.00

Printer/Fax/Scanner 1 $ 300.00 $ 300.00

Reaction Vessel 1 $ 2,000.00 $ 2,000.00

Refrigerator/Freezer 1 $ 1,417.00 $ 1,417.00

Scale 1 $ 3,957.00 $ 3,957.00

Spectrophotometer 1 $ 6,700.00 $ 6,700.00

Sterile pipets (various sizes) 1600 $ 0.34 $ 544.00

Stirrer/Hot Plate 2 $ 420.00 $ 840.00

Syringes 100 $ 0.27 $ 27.00

Vortex 1 $ 248.00 $ 248.00

Water bath 1 $ 2,245.00 $ 2,245.00

Water purification 1 $ 4,000.00 $ 4,000.00

Total Equipment Cost $ 109,972.00

54

Table XI.b - Estimated TCI based on delivered-equip ment cost

Components % of Delivered-Equipment Cost

Cost

Direct Costs

Purchased equipment 100 $ 109,972.00

Purchased-equipment installation 41 $ 45,088.52

Instrumentation (installed) 40 $ 43,988.80

Piping (installed) 35 $ 38,790.20

Electrical (installed) 10 $ 10,797.20

Buildings (including services) 40 $ 43,188.80

Yard improvements 15 $ 16,195.80

Service facilities (installed) 40 $ 43,188.80

Total direct costs 280 $ 353,190.12

Indirect Costs

Engineering and supervision 82 $ 90,737.04

Construction expense 116 $ 127,847.52

Contractor's fee 26 $ 28,672.72

Contingency 53 $ 58,025.16

Total indirect costs 277 $ 304,282.44 Fixed-capital investment $ 657,472.56 Working capital (15% of TCI) $ 116,730.45 Total Capital Investment $ 773,203.01

Total Product Cost

The total product cost (TPC) includes all costs associated with processing and

selling the product, recovering the capital investment, and continuing research and

development.8 These can be subdivided into manufacturing costs and general expenses.

Table XI.iii below shows estimates for these costs with a total product cost of

$244,600,000. The injections to be produced per year were determined from the demand

model as described earlier in the demand section.

55

Raw Materials Factor Quantity/year Cost per rate unit Calculated Cost

Injection materials 350000 $688.29 injection $240,976,535.83 *See Table IX.i for list Total $240,976,535.83 Operating Labor $287/hr 7884 hr/year $2,262,708.00 Operating Supervision 0.15 of operating labor $339,406.20 Utilities $14,000.00 Maintenance and Repairs 0.07 of FCI $46,453.08 Operating Supplies 0.4 of maintenance and repairs $18,267.96 Laboratory Charges 0.15 of operating labor $339,406.20

Variable Product Costs $243,964,777.27

Depreciation Straight line depreciation (not included here)

Taxes (property) 0.02 of FCI $13,129.45 Insurance 0.01 of FCI $6,564.73 Rent 0 Already have plant Fixed Charges $19,194.18 Plant Overhead 0.1 of total product costs $77,820.30 Administrative 0.2 of maintenance, operating labor, operating supervision $529,113.46 Distribution and Marketing 0.1 of total product costs $77,820.30 Research and Development 0.05 of total product costs $38,910.15

General Expenses $723,664.21 Total Product Cost $244,622,635.65

Based upon the estimate that the demand model gave for the number of patients

expected per a year risk was assessed on the NPV based on whether a person would or

would not get the injection. The risk that was used was based upon the idea that this

value could be larger or smaller depending upon how the market reacted. The reaction

was determined to be ± 200,000 people per year receiving the injection. This would lead

to a best and worst case scenario and give an idea of how this would affect the business in

marketing the solution. After risk analysis was done a positive mean net present worth

was observed with a good possibility of still making a profit. This sensitivity analysis

can be seen in Figure XI.a, having a probability as shown by the histogram in Figure XI.b

shown below.

TableXI.c – Total product cost estimation

56

Selling Price

In order to be competitive with other HA injection companies with a cost of

$1300 per year, we estimated our cost to the patient to be $2,400 per treatment. This is

more expensive than the leading hylan injection Synvisc®, which has a wholesale cost

(not including physician expenses) of $660 for the five-week injection series and must be

administered twice per year. The reason for our price being higher was because of the

Distribution for NPV (millions)/M19

Val

ues

in 1

0 ̂-

3

0.000

0.500

1.000

1.500

2.000

2.500

3.000

Mean=222.61

-200 0 200 400 600 800-200 0 200 400 600 800

5% 90% 5% -17.9703 442.2469

Mean=222.61

Distribution for NPV (millions)/M19

0.000

0.200

0.400

0.600

0.800

1.000

Mean=222.61

-200 0 200 400 600 800-200 0 200 400 600 800

5% 90% 5% -17.9703 442.2469

Mean=222.61

Figure XI.a – Distribution of net present worth (without depreciation)

Figure XI.b – Probability of net present worth

57

longer life of our treatment. Our treatment is expected to be superior and last about twice

as long hence the reason for our price being about twice the amount of our competitor’s.

As more patients use our product and more positive results are documented, the demand

for our product will increase making for a very profitable business.

XII. Conclusion

Osteoarthritis creates such a large consumer market in the United States which is

expected to grow as the average age of the population continues to increase, it is without

question that viscosupplementation serves as a potential solution to return patients suffering from

the disease to their normal pain-free active lives. However, because this industry is still in its

early stages, there are several issues with current available treatments that need to be addressed

before these intra-articular injections can truly be effective. Some patients do not feel the relief

provided is worth the pain involved in the injections, which in most cases must be repeated every

6 months. The main problem with current products is their lack of retention time in the synovial

cavity due to either degradation, transport across the synovial membrane, or simply missing the

synovial space during injection. We propose that a superior solution is to use hyaluronic acid

modified with 2-vinyl which will then be crosslinked to form a hydrogel. This crosslinking will

result in a more viscous, more elastic gel that will provide the load support and lubrication

necessary to keep bone-on-bone contact from occurring. We chose to use ammonium

peroxydisulfate because of the proven ability of sulfate groups to buffer against degradation and

prevent denaturation.

We suggest in the future that further research be done on the scale up and

commercialization of HYAL-VYNE to bring the product to the market. This will need to

include a complete business plan with a full design of the production process and investigation of

the market and plant locations. Also of concern is the possibility that ammonium

58

peroxydisulfate will significantly reduce the pH when added to the modified HA. This may be

avoided by using a buffer solution but because it will be injected into the body, it must be non-

toxic and must not affect the integrity of the gel. Since HYAL-VYNE is a novel idea, one final

issue is the patenting of the product to protect the technology. This will greatly increase total

product costs and will need to be included in the cost analysis.

59

References

1. “Report Measures Impact of Knee Conditions.” Knee1.com. 11 November 2003. http://www.knee1.com/news/mainstory.cfm/220/1

2. “Osteoarthritis Fact Sheet.” Arthritis Foundation. 2005.

http://www.arthritis.org/conditions/fact_sheets/OA_Fact_Sheet.asp

4. “Dissectors Answers – Joints of the Upper and Lower Limbs.” University of Michigan Medical School. 2000. http://images.google.com/imgres?imgurl=http://anatomy.med.umich.edu/images/knee_joint_post.jpg&imgrefurl=http://anatomy.med.umich.edu/musculoskeletal_system/joints_ans.html&h=363&w=400&sz=22&tbnid=Ngvb4pYw0_GrKM:&tbnh=108&tbnw=120&hl=en&start=162&prev=/images%3Fq%3Dhuman%2Bknee%26start%3D160%26svnum%3D10%26hl%3Den%26lr%3D%26sa%3DN

5. “Knee Joint – Anatomy and Function.” The Center for Orthopaedics and Sports

Medicine. Marietta, GA. 3 April 2003. http://www.arthroscopy.com/sp05001.htm

6. “Chemical and Other Safety Information.” The Physical and Theoretical Chemistry Laboratory, University of Oxford. 7 March 2006. http://ptcl.chem.ox.ac.uk/MSDS/#MSDS

7. “Product Search – FisherScientific.” 2006. <https://www1.fishersci.com/>

8. Peters, Max S., Klaus D. Timmerhaus, Ronald E. West. Plant Design and Economics for

Chemical Engineers, 5th ed. New York: McGraw Hill, 2003.

9. “End of Small Volume High Value Myth in Biotechnology – Process Design for a Mega-plant Producing γ-Interferon for Mega Profit.”

10. “2006 – 2008 General Catalogue.” VWR.

http://www.vwrsp.com/catalog/index.cgi?parent_id=9000002

11. Mironowicz, Magdalena and Sarah Saw. “Electronic Texbook – Joints: Part 1.” WorldOrtho, Inc. 1997. http://www.worldortho.com/database/etext/joints.html

12. “Articular Cartilage Zones in Normal Articular Cartilage.” Histology Atlas. School of

Physical Therapy, Slippery Rock University. http://www.sru.edu/depts/pt/histo/artic_cartilage.htm

13. Knox, P., J.R. Levick and J.N McDonald. Quarterly Journal of Experimental

Physiology. 73.1, 33-45.

14. Tadmor, Rafael; Nianhuan Chen, and Jacob N. Israelachvili. Journal of Biomedical Materials Research. 61.4, 514-523.

60

15. Hinton, Ralph et al. “Osteoarthritis: Diagnosis and Therapeutic Considerations.” American Family Physician. 1 March 2002.

16. “Osteoarthritis Overview.” eMedicine Consumer Health. <

http://www.emedicinehealth.com/articles/5494-1.asp >

17. “Osteoarthritis: medical information about the symptoms, diagnosis and treatment of degenerative joint disease.” MedicineNet.com 16 April 2004.

18. Hascall, Vincent and Torvard Laurent. “Hyaluronan: Structure and Properties.”

Glycoforum/Science of Hyaluronan-1.

19. Dinnar, U. “Lubrication Theory in Synovial Joints.” CRC Critical Reviews in Bioengineering 1975.

20. Griffith, Robert W., MD. “Knee Osteoarthritis: Is Arthroscopy Any Help?” Health and

Age. 24 January 2003.

21. Wen, Dennis Y., MD. “Hyaluronic Acid Injections for knee Osteoarthritis.” American Family Physician. The American Academy of Family Physicians. 1 August 2000.

22. “Federal Food, Drug, and Cosmetic Act – Chapter II.” US Food and Drug

Administration.

23. “Total Cost to Develop New Prescription Drug, Including Cost of Post-Approval Research, is $897 million.” Tufts Center for the Study of Drug Development. 13 May 2003.

24. “Device Advice – Premarket Approval.” US Food and Drug Administration Center for

Devices and Radiological Health. 1 November 2002.

25. Jones, D.B. and Middelberg, A. P.J. Langmuir 2002, 18, 10357-10362.

26. Kluger, R. and Alagic, A. Bioorganic Chemistry 2004, 32, 451-472.

27. Langer, Robert S.; Elisseeff, Jennifer, H.; Anseth, Kristi; Sims, Derek; “Semi- interpenetrating or interpenetrating polymer networks for drug delivery and tissue engineering” May 3, 1997.

28. Binette, J.S.; Garon, M.; Savard, P.; McKee, M. D.; Buschmann, M. D.; Journal of

Biomechanical Engineering 2004, 126, 475-484.

29. Huin-Amargier C.; Marchal, P.; Payan, E.; Netter, P.; Dellacherie, E.; Journal of biomedical materials research 2006, 76A, 416-424.

30. Uebelhart, Daniel MD.; Williams, James M. PhD.; Current Opinion in Rheumatology

1999, 11(5), 427.

61

31. Soltes, L. and Mendichi, R. Biomedical Chromatography 2003, 17, 376-384.

32. Burdick, J. A.; Khademhosseini, A.; Langer, R.; Langmuir 2004, 20(13), 5153-5156.

33. Ponnuraj, K. and Jedrzejas, M. J. Journal of Molecular Biology 2000, 299, 885-895.

34. Waddell, D. MD; Rein, A. MS; Panarites, C. PhD; Coleman, M. MD; Weiss, C. MD;

The American Journal of Managed Care 2001, 7(10), 981-991.

35. Ochiai, H.; Ohmae, M.; Mori, T.; Kobayashi, S.; Biomacromolecules 2005, 6, 1068-1084.

36. Han, T. Journal Zhejiang University SCI 2004, 5(8), 928-931.

37. Chang, W.; Chang, Y.; Chen, Y.; Sung, H.; Artificial Cells, Blood Substitutes, and

Biotechnology 2004, 32(2), 243-262.

38. Hennink, W.E.; Nostrum, C.F.; Advanced Drug Delivery Reviews 2002, 54, 13-36.

39. Lipowitz, A. J.; “Synovial Fluid”, Ch. 86, 2006, 1-19.

40. Anggiansah, C. L.; Scott, D.; Poli, A.; Coleman, P. J.; Badrick, E.; Mason, R. M.; Levick, J. R.; The Journal of Physiology 2003, 550.2, 631-640.

41. Milas et al. Biopolymers 2000, 59.4, 191-204. 42. Kirker, Kelly R. and Glenn D. Prestwich. Journal of Polymer Science: Part B: Polymer

Physics 2004. 42.23, 4344-4356. 43. Benedetto, Biasco, Pisinano, and Cingolani Nanotechnology 2005, 16, 165-170. 44. “Osteoarthritis Fact Sheet.” Arthritis Foundation. 2005. 45. Prestwich, Glenn D. “Biomaterials for Chemically-Modified Hyaluronan.” GlycoForum

29 March 2001. 46. “Arthroscopic Lavage.” The Knee Guru Dictionary. www.kneeguru.co.uk 47. Dinnar, U. CRC Critical Reviews in Bioengineering 1975, 2.2, 159-177. 48. Szeri, Andras Z. Fluid Film Lubrication Theory & Design. Cambridge University Press:

Cambridge, UK. 1998. p. 267-310. 49. “Is the Product a Medical Device?” Food and Drug Administration, 30 June 1998.

62

50. Shreve, Mark et al. NKOB. University of Oklahoma, ChE 4273. Spring 2005.

51. Miyamoto, Kenji et al. Journal of Biomedical Materials Research: Part A. 70.4, 550-559.

52. Leach, Jennie Baier et al. Biotechnology and Bioengineering. 82.5, 578-589.

53. Wallin, Richard F. and Edward F. Arscott. Medical Device and Diagnostic Industry

Magazine. April 1998.