BIOMATERIAL FOR SOFT TISSUE REPLACEMENTS
David N. Ku and Jin Wu Fan
Georgia Institute of Technology, Atlanta, GA 30332-0405, USA
Keywords: Biomaterial, mechanics, elasticity, strength, wear, biocompatibility, medical devices.
Abstract: Typical biomaterials are stiff, difficult to manufacture, and not initially developed for medical implants. A
new biomaterial is proposed that is
similar to human soft tissue. The biomaterial provides mechanical
properties similar to soft tissue in its mechanical and physical properties. Characterization is performed for
modulus of elasticity, ultimate strength and wear resistance. The material further exhibits excellent
biocompatibility with little toxicity and low inflammation. The material can be molded into a variety of
anatomic shapes for use as a cartilage replacement, heart valve, and reconstructive implant for trauma
victims. The biomaterial may be suitable for several biodevices of the future aimed at soft-tissue
replacements.
1 BACKGROUND
Most of the existing biomaterial technology is
limited to materials such as silicones, Teflon®,
polyethylene, metal and polyurethanes that do not
exhibit the mechanical and physical properties of
natural tissue. These materials are stiff, difficult to
manufacture, and not initially developed for medical
implants. Artificial tissue substitutes have not been
found to withstand the rigors of repetitive motion
and associated forces of normal life. Cadaver tissue
is limited in supply and due to the risk of infection is
coming under increased scrutiny by FDA and is not
accepted in Europe.
As an example, one of the most successful
med
ical implants is the artificial knee replacement
for the treatment of arthritis. Arthritis and joint pain
as a result of injury are major medical problems
facing the US and patients worldwide. Worldwide,
approximately 190 million people suffer from
osteoarthritis. This condition affects both men and
women, primarily over 40 years of age. The spread
of arthritis is also fueled by the rise of sports
injuries. Activities enjoyed by many can translate
into injury or joint damage that may set up a process
of deterioration that can have devastating effects
decades later. The number of patients that have
arthritis is staggering and growth is expected with
baby boomers entering the prime “arthritis years”
with prolonged life expectancies. Growth of the
world’s elderly is expected to increase three times
faster than that of the overall population.
Current standard treatmen
t is to surgically
implant a total knee replacement (TKR) that is made
of metals such as cobalt chromium or titanium.
These devices are highly rigid, providing no shock
absorption. Further, the metal integrates poorly with
bone and the HMWPE caps often create micro-
particulate debris with strong inflammation. The
invasive surgery not indicated for those under age 60
and usually reserved for end-stage patients. A
revision procedure is technically demanding, and
amputation may be required.
An alternative biodevice may be a soft tissue
replacem
ent. Arthritis stems from damage to
cartilage, the soft tissue between the bones.
Damaged cartilage leads to grinding of bone on bone
and eventual pain and limited joint function.
Biodevices that replace the soft tissue would restore
diarthroidal joint function much better and protect
further damage by a more natural stress distribution.
A similar problem exists for heart valves.
Prosth
etic heart valves are made from metal and
pyrolitic carbon which do not function like native
heart valves. The use of hard materials creates high
shear zones for hemolysis and platelet activation.
Tissue valves are subject to calcification, again a
problem of hard tissues not acting like natural soft
tissue. An alternative would be to design a biodevice
with soft tissue flexibility and endothelial cell
23
N. Ku D. and Wu Fan J. (2008).
BIOMATERIAL FOR SOFT TISSUE REPLACEMENTS.
In Proceedings of the First International Conference on Biomedical Electronics and Devices, pages 23-29
Copyright
c
SciTePress
covering to provide a wide-open central flow and
low-thrombogenic surface.
Yet another example for the need for soft tissue
replacements is a replacement part for reconstructive
surgery after traumatic accidents or cancer resection.
For many parts of the body, a replacement shape
needs to have smooth contours as well as soft tissue
compliance to yield a natural shape. The base
biomaterial should not have chemical composition
that is non-organic, such as silicone, which can
induce a hyper-immunogenic response.
Soft tissue replacements should start with a
biomaterial that has compliance ranges similar to
human soft tissue, be strong and wear resistant,
manufactured to personal shapes, and have long-
term biocompatibility. Cellular in-growth or
preloading of cells can then be performed on this
established scaffold. These features are
demonstrated in a new biomaterial described in this
paper.
2 METHODS
2.1 Biomaterials
Soft tissue-like devices can be made from polymers
such as poly vinyl alcohol as thermoset materials.
As an example, a PVA cryogel can be made
according to the full descriptions in US Patent U.S.
Patent Numbers 5,981,826 and 6,231,605. The
cryogels are made in a two stage process. In the first
stage a mixture of poly (vinyl alcohol) and water is
placed in a mold, and repeatedly frozen and thawed,
in cycles, until a suitable cryogel is obtained.
Poly(vinyl alcohol) having an average
molecular weight of from about 85,000 to 186,000,
degree of polymerization from 2700 to 3500, and
saponified in excess of 99% is preferred for creating
soft tissue-like mechanical properties. High
molecular weight poly (vinyl alcohol) in crystal
form is available from the Aldrich Chemical
Company. The PVA is then solubilized in aqueous
solvent. Isotonic saline (0.9% by weight NaCl,
99.1% water) or an isotonic buffered saline may be
substituted for water to prevent osmotic imbalances
between the material and surrounding tissues if the
cryogel is to be used as a soft tissue replacement.
Once prepared, the mixture can be poured into
pre-sterilized molds. The shape and size of the mold
may be selected to obtain a cryogel of any desired
size and shape. Vascular grafts, for example, can be
produced by pouring the poly (vinyl alcohol)/water
mixture into an annular mold. The size and
dimensions of the mold can be selected based upon
the location for the graft in the body, which can be
matched to physiological conditions using normal
tables incorporating limb girth, activity level, and
history of ischemia.
The new biomaterial, commercially available as
Salubria® from SaluMedica, LLC, Atlanta, GA is
similar to human tissue in its mechanical and
physical properties. The base organic polymer is
known to be highly biocompatible and hydrophilic
(water loving). The hydrogel composition contains
water in similar proportions to human tissue. Unlike
previous hydrogels, Salubria is wear resistant and
strong, withstanding millions of loading cycles; yet
it is compliant enough to match normal biological
tissue. The material can be molded into exact
anatomic configurations and sterilized without
significant deterioration.
2.2 Mechanical Characterization
2.2.1 Tensile Testing
Tensile test specimens were cut from sheets of
Salubria. They were tested in accordance with
ASTM 412 (die size D) in tension to failure using an
Instron Model 5543 electro-mechanical load frame
pulling at a rate of 20 inches per minute.
2.2.2 Stress-Strain Constitutive Relationship
The stress is a function of the load and the cross-
sectional area. However, the cross-sectional area
was difficult to measure. But the stretch ratio relates
the final and initial area due to the assumption of
incompressibility. That means the final area equals
the initial area divided by the stretch ratio.
Therefore, the ultimate stress calculation is a
function of the load at the breaking point of the
sample, the stretch ratio and the initial cross-
sectional area. The initial cross-sectional area is the
product of the initial width of the sample, w
o
, and the
initial thickness, t
o
.
Stretch Ratio:
oCC /
=
λ
(1)
Initial Cross-Section Area: (2)
0
* twA
oo
=
Final Stress:
o
ult
ult
A
F
λ
σ
*
=
(3)
In order to get an estimation of the pressure in
an intact tube the following simplified assumptions
were used. It was assumed that a tubular specimen
will burst when the circumferential wall stress is
BIODEVICES 2008 - International Conference on Biomedical Electronics and Devices
24
equal to the ultimate stress
ult
σ
. However, when an
artery is under physiologic load conditions it is in a
state of plane strain and undergoes finite two
dimensional stretches.
The stretch ratios are:
r
R
θ
λ
= (4)
z
l
L
λ
=
(5)
Rewritten to solve for the pressure, P:
2
z
T
P
R
θ
θ
σ
λ
=
(6)
This equation can calculate the pressure if we
know that data of the initial dimensions, the stress
and the stretch ratios. From the equation (6) we can
estimate the corresponding pressures.
2.2.3 Unconfined Compression
Cylindrical unconfined compression samples were
cast in a custom mold. Samples were tested in
unconfined compression on an Instron Model 5543
electro-mechanical load frame and on a DMTA IV
dynamic mechanical analyzer. Rates of 1% strain
per second and 20 inches per minute were tested.
2.2.4 Ultimate Strength
Ring specimens were pulled in tension until they
failed. Ring specimens of Salubria were
preconditioned twenty five cycles. Then the
specimens were distracted at 0.1mm/s, 1m/s,
10mm/s, and 100mm/s using a MTS 858Mini Bionix
Test System. Comparisons are made to normal
coronary arteries using identical protocols. The load
at failure was recorded as the ultimate load, and the
ultimate stress was calculated. Failure of the ring
specimen was defined as a complete tear of the ring
through the entire wall. The stress was derived
based on the assumption of incompressibility and
was defined as the ratio of load and cross-sectional
area. The stretch ratio was defined as the ratio of the
final and initial circumference. The final stress at
failure represented the ultimate strength for the
tension tests. To determine the final stress, an
equation was derived based on the assumption of
incompressibility [3] which means that the initial
volume V
o
and final volume V are equal. In the
present experiments the stretch ratio is defined as the
ratio of the final and initial circumference, Equation
(1). The ultimate stress,
ult
σ
defined by the load at
the breaking point of the sample divided by the final
cross-sectional area, was calculated using Equation
(3).
2.2.5 Fatigue Resistance
Ring specimens were cycled at different cycles, and
then pulled in tension until failure. The frequency of
the cyclic tests was set at 2 Hz because this value is
close to physiologic frequency of heart beats (~1.2
Hz) and strain rates effects testing showed that there
are no significant difference to do cyclic test under
1HZ to 5HZ. The purpose of the cyclic tests was to
experimentally determine how the fatigue affects the
ultimate strengths of porcine common carotid
arteries.
2.2.6 Cyclic Compression
A cyclic compression study was performed to assess
the response of Salubria biomaterial cylinders to
repetitive compressive loading at physiologic stress
of approximately 1.3 MPa. The specimen is loaded
for 1 million loading cycles at a rate of
approximately 1.5 Hz. Dimensional integrity was
measured using an optical comparator and
mechanical modulus of elasticity was determined at
20% strain.
2.2.7 Wear Testing
An accelerated wear tester was built to test the wear
rate of Salubria biomaterial. Polished stainless steel
rollers with a diameter of 1 5/8 inches (chosen to
approximate the average radius of the femoral
condyles) are rotated so that they slide and roll
across the test sample, creating a peak shear load of
approximately 90N (0.2 MPa). Separate testing has
shown the coefficient of friction for polished
stainless steel against Salubria biomaterial to be
equivalent to that of porcine femoral condyle with
the cartilage surface abraded away to subchondral
bone or roughly 4 times that of porcine femoral
condyle with intact cartilage surface. The rotating
cylinders exert a normal load of approximately 200N
on the sample. The wear tester is operated at a rate
that subjects the sample to 1000 wear cycles per
hour where one cycle is defined as one roller to
sample contact. Data has been collected with the
sample lubricated and hydrated with water (a worst
case scenario since synovial fluid should provide
some surface lubrication). Wear rate is measured by
BIOMATERIAL FOR SOFT TISSUE REPLACEMENTS
25
weight loss of the sample over a number of cycles.
Salubria biomaterial is tested for >10,000,000 cycles
against polished stainless steel rollers.
2.3 Biocompatibility
Biomaterials for use in humans must pass a full
complement of biocompatibility tests as specified by
ISO and the USFDA. The material was tested for
ability to produce cytotoxicity, intracutaneous
irritation, sensitization by Kligman maximization,
Ames mutagenicity, chromosomal aberration, and
chronic toxicity.
2.3.1 Rabbit Osteochondral Defect Model
In addition to the standard biocompatibility testing,
the ability of Salubria biomaterial to withstand load
or cause local inflammatory responses in a widely
used rabbit osteochondral defect model was assessed
(e.g., Hanff et al., 1990). A cylindrical plug (3.3
mm diameter, 3 mm depth) was implanted in the
right knee of each of sixteen New Zealand white
rabbits. An unfilled drill hole was made in the left
knee of each animal to serve as an operative control.
After three months of implantation in the
patellofemoral groove, eight rabbits were sacrificed
for histologic analysis of the implant site and
surrounding synovium. The remaining eight animals
were sacrificed and the implant assessed for any
change in physical properties. In addition, the
distant organ sites that are known targets for PVA
injected intravenously were assessed histologically
for any sign of toxicity due to implantation. Tissues
assessed included liver, spleen, kidney, and lymph
node.
2.3.2 Particulate Inflammation
Salubria biomaterial was tested for particulate
toxicity or inflammatory reaction in the joint. The
study design was based on a study conducted by Oka
et al. (2000) on a similar PVA-based biomaterial in
comparison to UHMWPE. Particulate sizes for the
study varied from approximately 1 micron to 1000
microns. The total volume of particulate injected
over the 2 divided doses was designed to represent
full-thickness wear of 10 x 10 mm diameter cartilage
implants.
2.3.3 Ovine Knee Inflammation Model
In vivo testing was performed using a meniscus
shaped device made of Salubria and implanted into
the sheep knee joint. Devices were removed at 2
week, 3 week, 2 month, 4 month and 1 yr. intervals.
Animals were fully load-bearing on the day of
operation and after. Full range of motion with no
disability was observed. Gross examination of
surrounding tissues and histology of target end
organs (liver, kidney, spleen, and lymph nodes) were
evaluated for acute or chronic toxicity.
3 RESULTS
3.1 Tensile Testing
Tensile Stress vs. Strain of Salubria
Sample 480-G-1
0
1
2
3
4
5
6
7
0 50 100 150 200 250 300 350
Strain (%)
Stress (MPa)
Figure 1: Representative Tension Curve.
Plots of stress versus strain in tension (figure 1)
show a non-linear response. Due to the non-linearity
of the loading curve, tangent modulus values at a
defined percent strain are used to characterize the
material stiffness. Tangent modulus ranges from
1.2-1.6 MPa. Ultimate tensile strength is 8-10 MPa.
The stress-strain curve exhibits a non-linear elastic
behavior similar to natural soft tissue.
3.2 Unconfined Compression
Figure 2 is a representative curve of stress versus
strain in compression. Compression loading curves
show a non-linearity suggesting that Salubria is a
viscoelastic material similar to cartilage. Tangent
modulus values in compression range from 0.1 to 7
MPa. Plastic compressive failure occurs at or above
65% strain.
3.3 Creep and Creep Recovery
Creep and creep recovery experiments were
performed to assess the performance of Salubria
biomaterial under long-term static loading at
physiologic loads of up to 480 N-force. High loads
were applied for 24 hours creating deformation of
BIODEVICES 2008 - International Conference on Biomedical Electronics and Devices
26
3.8 Animal Testing
50% of the initial height. Initial loading
demonstrates a biphasic visco-elastic behavior.
After 24 hours of recovery in saline, sample height
had returned to within 5% of the original height.
The compressive modulus of the material before the
test and after creep recovery was unchanged.
After three months of implantation in the
patellofemoral groove, eight rabbits were sacrificed
for histologic analysis of the implant site and
surrounding synovium. Tissues assessed included
liver, spleen, kidney, and lymph node. The Salubria
biomaterial was well-tolerated with subchondral
bone formation surrounding the implant, no fibrous
tissue layer or inflammatory response, no implant
failures or evidence of wear debris formation, no
osteolysis, and no toxic effects on the implant site or
distant organ tissues.
3.4 Wear Testing
Results for one formulation of Salubria biomaterial
tested for >2,000,000 cycles against polished
stainless steel rollers demonstrate minimal wear.
3.5 Cyclic Compression
At time of explantation, the samples were
essentially unchanged (see Fig. 3). Based on
histologic examination in comparison to the
operative control, there was no evidence of
inflammatory reaction in either the surrounding
cartilage/bone (see Fig. 4) or in the synovium. In
fact, a layer of normal hyaline cartilage partially
covered the implant surface. The cartilage surface
of the patella also showed no changes in the area
articulating against the Salubria implant. There was
no sign of toxicity on histologic examination of the
distant organ sites.
Repetitive compressive loading at physiologic stress of
approximately 1.3 MPa was imposed. After 1 million
loading cycles at a rate of approximately 1.5 Hz, there
was minimal change (<5%) in sample dimensions and
no change in modulus of elasticity at 20% strain.
3.6 Ultimate Strength
Sixty-four specimens were pulled at four different
stain rates. Ultimate stress increased as a weak
function of increasing strain rates. The ultimate
stress at 100 mm/s was 4.54 MPa, greater than the
3.26 MPa at 0.1 mm/s. The differences between
0.1mm/s and 100 mm/s was highly significant with
p<0.001. The differences between 0.1 mm/s and 10
mm/s gave p=0.013; and 1 mm/s to 100 mm/s was
p=0.018. The difference between 1 mm/s and 10
mm/s was not statistical significance. Strain rates
between 1 and 100 mm/s correspond to a cyclic
frequency of 0.5 Hz to 5 Hz for fatigue testing.
Compressive Stress vs. Strain of Salubria
Sample 482-G-1
0
1
2
3
4
5
6
7
0 20406080
Strain (%)
Stress (MPa)
Figure 3: The left-hand knee shows a Salubria implant in
the patellofemoral joint of a rabbit knee after 3 months
implantation. The right-hand knee is an operative control.
On excision for mechanical testing, the sample
edges firmly adhere to the surrounding bone. The
indentation force (i.e., the force required to cause a
certain amount of sample deformation) of the
implant is unchanged from a non-implanted control.
Comparison material characterization testing showed
that the implanted samples were not different from
non-implanted controls.
Figure 2: Representative Compression Curve.
3.7 Biocompatibility
The following table outlines the results of standard
biocompatibility testing performed on Salubria
biomaterial, in accordance with ISO 10993-1 and
FDA Blue Book Memorandum #G95-1.
BIOMATERIAL FOR SOFT TISSUE REPLACEMENTS
27
Table 1: Biocompatibility Testing.
ISO 10993-1 Recommended Testing
Requirement
Test Performed Test Results
ISO MEM Elution L929 cells,
GLP.
Pass
Cytotoxicity, ISO 10993-5
Direct Contact Neurotoxicity
Pass
Kligman Maximization Method
Pass
Sensitization and Irritation,
ISO 10993-10
Primary Vaginal Test: Repeat
Exposure
Pass
Sub-acute and Sub-Chronic Toxicity,
ISO 10993-6
Sub-acute and sub-chronic
toxicity
Pass
Ames Mutagenicity: Dimethyl
Sulfoxide Extract, 0.9% Sodium
Chloride Extract
Pass
Genotoxicity, ISO 10993-3
Chromosomal Abberation
Pass
Subacute or site Specific
Implantation with chronic
Toxicity
Pass
Biocompatibility study in
Rabbits following Intra-articular
injections.
Pass
Implantation,
ISO 10993-6 and 10993-11
Rabbit Trochlear Osteochondral
Defect
Pass
(b)
(a)
Figure 4: (a) This digital scan of a paraffin tissue block containing a Salubria implant demonstrates that the implant remains
in place over 3 months of implantation in the rabbit patellofemoral groove. (b) Hematoxylin and eosin stain of a section
from the tissue block in (a) showing the implant site – the implant has been removed during the staining process. There is
no evidence of inflammatory reaction; the surrounding cartilage and bone are normal in histologic appearance.
Salubria biomaterial was tested for particulate
toxicity or inflammatory reaction in the joint.
Salubria particulates were biologically well-
tolerated. The biomaterial particulate was deposited
on the superficial synovium with minimal
inflammatory reaction. There was no evidence of
migration from the joint space or toxicity in the knee
or at distant organ sites. There was no evidence of
third body wear or osteolysis.
For the goat study, the native articular cartilage
surfaces were protected in the test group compared
to extensive damage from the control meniscectomy
group. No local inflammation was noted on MRI or
histology. No distant organ inflammation was seen
BIODEVICES 2008 - International Conference on Biomedical Electronics and Devices
28
in these large animals, confirming the observations
in the rabbits.
4 DISCUSSION
The biomaterial described here exhibits the requisite
characteristics for soft tissue replacements. For
knee cartilage, the material has non-linear
viscoelastic properties similar to native tissue. The
strength and fatigue properties exceed the
requirements for a fully loaded knee articular joint
(Stammen, 2001). For heart valves, the material
must be moldable to complex anatomies and exhibit
low thrombogenicity. For reconstructive anatomic
parts, the biomaterial should be easily molded to
custom shapes and have low inflammation potential.
The biomaterial presented here exhibits these
properties and opens the potential for soft tissue
replacements that more closely match the anatomic
and physiologic requirements.
Although ring specimens and dumbbell shape
specimens are both one-dimensional tests, ring
specimens were used because they provide a good
gripping connection. Ring samples can relieve the
experimental error comes from the inappropriate
clamping dumbbell specimens which can cause the
specimens to slide or break in the neighborhood of
the clamp. There may be damage from preparing
uniaxial dumbbell shaped strips. Dumbbell strips
would also be difficult to obtain because of the small
diameter of the tubular samples.
The biocompatibility testing for Salubria
reflects previous carcinogenicity testing on other
PVA-based biomaterials. PVA hydrogels in the
literature are non-carcinogenic with rates of
tumorigenicity similar to the well-accepted medical-
grade materials, silicone and polyethylene.
Nakamura (2000) reports on a 2-year carcinogenicity
study conducted on a PVA-based biomaterial
subcutaneously implanted in rats. Pre-clinical
investigation of other PVA-based hydrogels and
Salubria biomaterial demonstrates that these
materials are biocompatible in the joint space (Oka
et al., 1990). The rabbit is the most commonly
published cartilage repair model with study lengths
varying from 3 months to 1 year, with little
difference in results at 3 months from those at 1
year. These studies indicate that 3 months is
sufficient to assess biocompatibility and early
treatment failure in the rabbit model. These results
are further confirmed by clinical results on
SaluCartilage
TM
.
Based on this study, Salubria soft tissue
biomaterial has been shown to be biocompatible
with long-term implantation. There is no evidence
of inflammatory response or local or distant toxicity.
Furthermore, the biomaterial has stable, durable
physical properties over the period of implantation
in joints and would be suitable for use as structure
deceives such as a cartilage replacement. The
biomaterial presented here opens the potential for
soft tissue replacements that more closely match the
anatomic and physiologic requirements for human
biodevices.
REFERENCES
Hanff, G., Sollerman, C., Abrahamsson, S. O., and G.
Lundborg. 1990. Repair of osteochondral defects in
the rabbit knee with Gore-TEXTM (expanded
polytetrafluoroethylene). Scandinavian Journal of
Plastic and Reconstructive Hand Surgery. 24: 217-
223.
Oka, M., et al., Development of an artificial articular
cartilage. Clin Mater, 1990. 6(4): p. 361-81.
Stammen, J.A., Williams, S., Ku, D. N., and R. E.
Guldberg. 2001. Mechanical properties of a novel
PVA hydrogel in shear and unconfined compression.
Biomaterials 22(8): 799-806.
BIOMATERIAL FOR SOFT TISSUE REPLACEMENTS
29