See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/327564879

In vivo evaluation of resorbable supercritical CO 2 -treated collagen

membranes for class III furcation-guided tissue regeneration: IN VIVO
EVALUATION OF RESORBABLE scCO 2 -TREATED...

Article  in  Journal of Biomedical Materials Research Part B Applied Biomaterials · September 2018

DOI: 10.1002/jbm.b.34225

CITATIONS READS

0 63

9 authors, including:

Nick Tovar Lukasz Witek

New York University College of Dentistry New York University College of Dentistry
127 PUBLICATIONS   1,034 CITATIONS    104 PUBLICATIONS   488 CITATIONS   

SEE PROFILE SEE PROFILE

Rodrigo Neiva Heloisa Fonseca Marão

University of Florida Universidade Santo Amaro (UNISA)
72 PUBLICATIONS   1,541 CITATIONS    21 PUBLICATIONS   101 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Color determination in ceramic restorations View project

on efficacy of thread design on enhancing osseointegration View project

All content following this page was uploaded by Heloisa Fonseca Marão on 30 April 2019.

The user has requested enhancement of the downloaded file.

In vivo evaluation of resorbable supercritical CO2-treated collagen
membranes for class III furcation-guided tissue regeneration

Nick Tovar,1 Lukasz Witek ,1 Rodrigo Neiva,2 Heloisa F. Marão,1 Luiz F. Gil,3 Pablo Atria ,6
Ryo Jimbo ,4 Eduardo A. Caceres,7 Paulo G. Coelho1,5
1
Department of Biomaterials and Biomimetics, New York University College of Dentistry, New York, New York
2
Department of Periodontology, University of Florida in Gainesville, Gainesville, Florida
3
Department of Morphological Sciences, Universidade Federal de Santa Catarina, Florianopolis, Brazil
4
Department of Applied Prosthodontics, Nagasaki University Graduate School of Biomedical Sciences, Nagasaki, Japan
5
Hansjörg Wyss Department of Plastic Surgery, New York University Langone Medical Center, New York, New York
6
Biomaterials Department, Universidad de los Andes, Santiago, Chile
7
Deparment of Biomaterials, Universidad Andes Bello, Faculty of Dentistry Vina del Mar, Chile

Received 2 August 2017; revised 26 July 2018; accepted 2 August 2018

Published online 00 Month 2018 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jbm.b.34225

Abstract: The study evaluated the effects of a Supercritical CO2
(scCO2) on a commercially available decellularized/delipidized
naturally derived porcine pericardium collagen membrane,
Vitala®. The Vitala® and scCO2 treated experimental membranes
were evaluated for guided tissue regeneration (GTR) of periodon-
tal tissue in class III furcation defects utilizing a dog model. Physi-
cal material characterization was performed by scanning electron
microscopy (SEM), thermogravimetric analysis (TGA), and differ-
ential scanning calorimetry (DSC). The in vivo portion of the study
was allocated to three-time points (6, 12, and 24-weeks) using
standardized class III furcation defects created in the upper sec-
ond and third premolars. The experimental defects (n = 5) were
covered with either a collagen membrane (positive control),
scCO2-treated collagen membrane (experimental) or no mem-
brane (negative control). Following sacrifice, histologic serial sec-
tions were performed from cervical to apical for morphologic/
morphometric evaluation. Morphometric evaluation was carried
out by ranking the presence of collagen membrane, amount of
bone formation within the defect site and inflammatory cell infil-
trate content. SEM showed the experimental scCO2-treated mem-
brane to have a similar gross fibrous appearance and chemical
structure in comparison to the Vitala® Collagen membrane. A sig-
nificant increase in membrane thickness was noted in the scCO2-
treated membranes (366  54 μm) vs non-treated membranes
(265  75 μm). TGA and DSC spectra indicated no significant
qualitative differences between the two membranes. For the
in vivo results, both membranes indicated significantly greater
amounts of newly formed bone (scCO2: 2.85  1.1; Vitala®: 2.80 
1.0) within the covered defects relative to uncovered controls (0.8
 0.27) at 24 weeks. Both membrane types gradually degraded as
time elapsed in vivo from 6 to 12 weeks, and presented nearly
complete resorption at 24 weeks. The inflammatory infiltrate at
regions in proximity with the membranes was commensurate
with healthy tissue levels from 6 weeks in vivo on, and periodon-
tal ligament regeneration onset was detected at 12 weeks in vivo.
The effect of the supplementary scCO2 treatment step on the col-
lagen membrane was demonstrated to be biocompatible, allow-
ing for the infiltration of cells and degradation over time. The
treated membranes presented similar performance in GTR to
non-treated samples in Class III furcation lesions. Defects treated
without membranes failed to achieve regeneration of the native
periodontium. © 2018 Wiley Periodicals, Inc. J Biomed Mater Res B Part
B: Appl Biomater, 00B: 000–000, 2018.
Key Words: collagen membrane, supercritical CO2, guided tis-
sue regeneration, furcation defect, dog

How to cite this article: Tovar N, Witek L, Neiva R, Marão HF, Gil LF, Atria P, Jimbo R, Caceres EA, Coelho PG. 2018. In vivo evalu-
ation of resorbable supercritical CO2-treated collagen membranes for class III furcation-guided tissue regeneration. J Biomed
Mater Res B Part B. 2018:00:00:1–9.

INTRODUCTION important element in the maintenance of the body.1,2 Periodon-

Bone is a complex porous composite with unique properties
allowing for adaptive remodeling of its microstructure due to
external mechanical stresses and repair of damage due to
injury. Bone plays a crucial role both in terms of functional
support and metabolism, so rapid repair and replacement is an
tal bone defects may result from trauma, surgery, or diseases,3
and typically result in damage to the complex periodontal
attachment system and surrounding soft tissues as well.
In order to improve the efficacy and predictability of
periodontal regeneration procedures, the use of resorbable

Correspondence to: L. Witek; e-mail: lukasz.witek@nyu.edu

Contract grant sponsor: Osteogenics Biomedical, Inc.

© 2018 WILEY PERIODICALS, INC. 1

or non-resorbable barrier membranes has been introduced.
Used with or without a particulate grafting material or
growth factors, the primary role of the barrier membrane is
to provide physical isolation of the defect, preventing the
early invasion of the defect by faster growing gingival epi-
thelial and connective tissues and allowing regeneration of
the defect by bone, cementum, and periodontal ligament tis-
sue. Thus, with Guided Tissue Regeneration (GTR) the regen-
eration of a functional periodontal attachment system may
be achieved, rather than a simple repair of the defect by scar
tissue, limited bone formation, and soft tissue recession.4–6
Several resorbable or non-resorbable biocompatible mate-
rials have been previously used as membranes in GTR. Although
previous clinical and histologic studies have demonstrated that
similar results can be achieved using either material,7,8 previous
studies shown the advantages of collagen-based resorbable
membranes, implying their superiority to their non-resorbable
counterparts.9 The most commonly reported clinical advantage
of resorbable membranes is that a secondary procedure for
retrieval of the GTR membrane is not necessary.10,11
In order to preserve the biological and biomechanical
properties of collagen-based GTR membranes, numerous tech-
niques have been studied and reported for preparing and ster-
ilizing devices for use in clinical applications.12,13 However, it
is well accepted that all of these techniques may potentially
alter the structure and characteristics of biomaterials through
degradation and or crosslinking of the naturally derived colla-
gen matrix, thus potentially affecting the tissue response to
the membrane. Traditional methods for decellularizing and
processing of biological tissues include the use of mechanical
disruption, ultrasonic treatment, osmotic treatment, acids,
bases, organic solvents, enzymes, and surfactants. Typically,
the membranes are sterilized by electron beam, gamma irradi-
ation or ethylene oxide (ETO) treatment. Irradiation of colla-
gen has the potential for weakening the material structurally,
while ETO sterilization introduces the potential complication
of ETO residuals that must be monitored and controlled.
An alternative decellularization and sterilization method
is the use of supercritical CO2 (scCO2), which has previously
been used as a processing method for allogenic and xenoge-
neic tissues and most recently has been employed for sterili-
zation as well. An advantage of scCO2 for sterilization is that
it is done using a cold process, thus lowering the potential to
alter the collagen material structurally.12–19 As a processing
method for tissues these Supercritical fluids (SCFs), specifi-
cally, scCO2 which is performed using high-pressure carbon
dioxide (CO2) has an exceptional ability to penetrate and dif-
fuse into biological tissues such as dermis or bone.12,17,18
The value of scCO2 sterilization has been established for var-
ious collagen scaffolds, including amnion membranes,13
decellularized lungs,14 and decellularized heart valves.16
Additionally, scCO2 as an extraction media is reported to be
an effective mechanism.20 The primary advantage of SCF
technology is that it occurs in moderate conditions, therefore
decellularization of biologic tissues may be done with
reduced treatment time and minimal or reduced exposure to
harsh chemicals. This technology offers great promise in the
processing of xenogenic and allogenic tissues on its own or
2 TOVAR ET AL.
as an adjunctive processing step to ensure natural tissue
derived decellularization and hence its clinical safety.
Although scCO2 theoretically has the capability to achieve
both decellularization and sterilization using a single step
process,12–19 the effect of such processing in the physico-
chemical properties and in vivo performance of collagen-
based membranes has not been addressed to date. The pur-
pose of this study was to evaluate the effects of scCO2 treat-
ment on a collagen membrane derived from porcine
pericardium as a potential adjunctive processing step.
The study design called for the physicochemical character-
ization and in vivo evaluation of porcine pericardium derived
collagen membrane devices (Positive control—commercially
available Vitala® Collagen Membrane vs. Experimental—
Vitala® subsequently treated with scCO2). Membranes were
analyzed for bone forming capabilities, membrane degrada-
tion rate, and biocompatibility (assessed through inflamma-
tory infiltrate presence) for guided tissue regeneration (GTR)
following a standardized class III furcation defect at the maxil-
lary premolar regions of beagle dogs.
MATERIALS AND METHODS
A commercially available FDA cleared porcine pericardium
collagen-derived membrane, Vitala® (V) Collagen
(Osteogenics Biomedical, Inc., Lubbock, TX), was chosen for
this study as a positive control. The experimental membrane
(TC) was the commercial porcine pericardium collagen mem-
brane subjected to scCO2 treatment. The scCO2 treatment
method was applied with a previously published protocol
utilizing a Nova2200™ vessel (NovaSterilis Inc., Lansing,
NY).16,21 The supercritical vessel was sprayed with water
(H2O) to maintain humidity during the process along with
peracetic acid and 3% hydrogen peroxide (H2O2). The 2-h
cycle operated at ~35 C and ~1430 psi.16,22 Both the control
and experimentally treated membranes were proved by the
manufacturer, Osteogenics Biomedical, Inc., in sterile condi-
tion after exposure to low dose e-beam irradiation (1 keV–
10 MeV).23
Scanning electron microscopy (SEM) (Zeiss EV-50, Ober-
kochen, Germany) was used to analyze the surface topogra-
phy under an accelerating voltage of 15 kV. Membrane
thickness was measured using calipers for both the TC and V
groups of 10 membranes (n = 5/type) at five varying loca-
tions across the length of the membrane. Both TC and V
membranes were cut to smaller sections, each weighing
~1.2 mg, and subjected to a 24-h period in a desiccator. To
ensure they were free of moisture from the environment the
vacuum chamber was used per the manufacturer’s protocol.
Subsequently, the membranes (n = 9) were subjected to
thermogravimetric analysis (TGA) and differential scanning
calorimetry (DSC) (SDT Q600, TA Instruments-Waters, LLC,
New Castle, DE), in an Nitrogen (N) atmosphere, ramped to
1000 C at 20 C/min. All analysis was completed using the
Universal Analysis 2000 software.
The in vivo study consisted of skeletally mature female
beagle dogs, ~1.5 years of age, with an average weight ~16
kg, and in good health. The study was performed under
IN VIVO EVALUATION OF RESORBABLE scCO2-TREATED COLLAGEN MEMBRANES

approval of the committee for animal experimentation at
Ecole Nationale Veterinaire D’Alfort, Maisons-Alfort, France.
The dogs were acquired 2 weeks prior to the first surgical
procedure to allow for acclimation, and were treated sepa-
rately in five batches (randomly allocated) for 6, 12, and
24 weeks (n = 5/time point). The number of subjects per time
in vivo was based on a previous study24 of the same animal
and bone defect model, which considered a minimum bone
formation amount size effect of 30% at a statistical power of
β = 0.80 and α = 0.05 between control and membrane sites
for each experimental time (6, 12, and 24 weeks).
All surgical procedures were performed under general
anesthesia and strict sterile guidelines. The pre-anesthetic pro-
cedure comprised of an intramuscular (IM) administration of
atropine sulfate (0.044 mg/kg) and xylazine hydrochloride
(8 mg/kg). General anesthesia was then obtained following an
IM injection of ketamine chlorate (15 mg/kg). This study
included three surgical sites in each dog: one untreated control
(C), one experimental site using V membrane, and one experi-
mental site using TC membrane. Both membrane types, which
were cut to cover defect size, are presented in Figure 1(A).
The pre-surgical aspect of the beagle dog posterior max-
illa is presented in Figure 1(B). Following an intrasulcular
incision, mucoperiosteal flaps were made and the bone tis-
sue was exposed [Figure 1(C)]. The standardized defect was
FIGURE 1. Surgical procedure sequence. (A) Collagen membrane preparation, (B) pre-surgical aspect of the beagle dog posterior maxilla, (C) total
flap reflection, (D) creation of standardized defects at the furcation region, (E) collagen membrane placement, (F ) suture.
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B | MONTH 2018 VOL 000B, ISSUE 0
ORIGINAL RESEARCH REPORT
created at the furcation region of the second and third upper
premolars presented 4 mm in height from the bone crest
level by 2.5 mm of horizontal width, and were extended
toward the lingual wall to comprise a class III furcation
lesion [Figure 1(D)]. The defects were created by means of a
2-mm diameter cylindrical bur. After the creation of the
defects, the two collagen membranes (V—lot number
IPARD062413; TC—lot number IPARD080513-B) were
placed over two of the three defects [Figure 1(E)], and the
remaining defect was left uncovered (negative control
group). The sites of the membrane placement [Figure 1(F )]
and control groups were randomized between dogs in order
to balance the study design and minimize any variations and
bias due to site location. The flaps were coronally reposi-
tioned with resorbable sutures (4–0 Vicryl) [Figure 1(F )].
After surgery, the animals were subjected to a soft diet
for a period of 7 days. Postsurgical medication included anti-
biotics (penicillin, 20,000 UI/Kg) and analgesics (ketopro-
phen, 1 mL/5 kg) for a period of 48-h postoperatively. The
euthanasia was performed by anesthesia overdose after
scheduled times in vivo.
Following euthanasia, the specimens were fixed in 10%
buffered formalin and reduced to blocks for histological pro-
cessing. The blocks were then washed in running water for
24 h and steadily dehydrated in a series of alcohol solutions
3
ranging from 70 to 100% ethanol. Following dehydration, the
samples were embedded in a methacrylate-based resin
(Technovit 9100, Heraeus Kulzer GmbH, Wehrheim, Germany)
according to the manufacturer’s instructions. The blocks were
then cut into slices (approximate thickness, 200 μm) in the
buccolingual direction from the cervical to apical direction
within the defect with a precision diamond saw (Isomet 2000,
Buehler Ltd., Lake Bluff, IL), and glued to acrylic plates with
an acrylate-based cement. After a 24 h setting time, the sec-
tions were reduced to a final thickness of ~80 μm by means
of a series of SiC abrasive papers (600, 800, 1200, and 2400
grit) (Buehler Ltd., Lake Bluff, IL, USA) using a grinding/pol-
ishing machine (Metaserv 3000, Buehler Ltd., Lake Bluff, IL)
under water irrigation.25 The sections were then stained with
Stevenel’s Blue and Van Gieson Picro Fuchsin and observed
by optical microscopy at 50–200× magnification (Leica
DM2500M, Leica Microsystems GmbH, Wetzlar, Germany) for
histomorphologic evaluation. Five sections were generated in
the buccolingual direction from the cervical to apical direction
within the defect. The section representing the central region
of the defect length from cervical to apical was evaluated. The
amount of collagen membrane presence, bone formation
within the defect, and presence of inflammatory infiltrate was
ranked as follows26:
• Bone growth within the defect site between mesial and
distal roots: The morphology of the surgical sites was
assessed histologically to assure any old bone was not
considered as new bone. A 0–5 scale was used, where a
0 indicated that no new bone filled the region, while a
5 indicated that new bone filled the entire region.
• Collagen Membrane Presence: Visual observation of
the membrane across the cross-section of the mesial/
distal roots. A 5 constituted entire coverage of the sur-
gical site with visible presence of the membrane, while
a 0 denotes fibrous structure at the surgical site with
no visible presence of the membrane.
• Inflammation: A 0–5 scale was used to assess the
degree of infiltration of inflammatory cells at the sur-
gical site and membrane/host tissue interface. A 0 indi-
cated no inflammation, while 5 indicated a significant
amount of inflammation.
FIGURE 2. SEM cross-sectional view of (A) TC and (B) V membrane.
4 TOVAR ET AL.
For each time in vivo, parameter rankings were submit-
ted to statistical analyses and valuated using Friedman anal-
ysis comparison between groups. Significance levels for all
groups are presented where these were detected. The statis-
tical unit considered was the number of subjects for each
time in vivo. Statistical significant differences were consid-
ered at α = 0.05.
RESULTS
The SEM images (Figure 2) and analysis showed a fibrous
extracellular matrix structure for both TC and V membranes.
Caliper measurements of the TC and V were significantly dif-
ferent (p = 0.002) at 366  54 μm and 265  75 μm, respec-
tively. The TGA spectra [Figure 3(A)] did not indicate any
significant differences in weight loss % (p = 0.184), onset
point (p = 0.06) and derivative weight loss peak temperature
(p = 0.935). Although the DSC spectra [Figure 3(B)] yielded
nearly identical curves, the Vitala® Collagen membrane, dif-
fered slightly with an additional endothermic peak at ~400 C.
The surgical procedures and follow-up demonstrated no
complications due to procedural conditions or post-operative
infection. Histomorphologic evaluation of the slides at
6 weeks for the control group exhibited apical migration of
the gingival epithelium and fibrous connective tissue, result-
ing in minimal bone formation within the defect area [-
Figure 4(A)]. On the other hand, both TC [Figure 4(B)] and V
[Figure 4(C)] presented substantial bone formation. At
6 weeks in vivo, the collagen membranes were in direct con-
tact with both roots and were lined by newly formed bone
within the defect region and by a layer of randomly oriented
connective tissue (covered by dense connective tissue and
gingival epithelium) on its gingival aspect [Figure 3(B,C)].
The randomly oriented connective tissue in contact with the
membranes presented blood vessels in close proximity with
both membranes. Inflammatory infiltrate was evenly spread
through the tissue within the control defect, and it was
restricted to regions in proximity with the membrane for
both membrane groups at mild levels. For both membrane
groups, cellular cementum was often observed along
denuded roots in tandem with parallel bone formation, indi-
cating initial periodontal ligament healing [Figure 4(B,C)]. No
full buccal/lingual plate regeneration was observed for any
of the experimental groups.
IN VIVO EVALUATION OF RESORBABLE scCO2-TREATED COLLAGEN MEMBRANES
 ORIGINAL RESEARCH REPORT

FIGURE 3. (A) TGA results for weight loss (%) and (B) DSC as a function of temperature for the TC and V membranes.

When compared with 6 weeks in vivo, V and TC samples
at 12 weeks roughly presented the same morphology, while
the control group [Figure 5(A)] presented apical migration
of the gingival epithelium surrounding the defect area filled
with fibrous connective tissue and minimal bone formation
within the defect area. At 12 weeks, no gains in the amount
of bone for the TC [Figure 5(B)] and V [Figure 5(C)] defects
was observed. There was also no substantial collagen mem-
brane degradation occurring via soft tissue remodeling
throughout the thickness of both membranes. Relative to
6 weeks, a decrease in the amount of inflammatory infiltrate
was observed for all groups at soft tissue regions in proxim-
ity to both membrane types, where inflammatory cells were
seldom observed and levels were commensurate with
healthy tissue. At this time in vivo, in regions where no direct
contact between membrane and denuded roots occurred,
cellular cementum was observed along the roots in tandem
with parallel bone formation. Cellular cementum and bone
were bridged by Sharpey fibers indicating full periodontal
regeneration at these regions (Figure 6). No full buccal/lin-
gual plate regeneration was observed for any of the experi-
mental groups.
The 24-week samples were remarkably similar to the
12-week samples, with exceptions, which included further
apical migration of the gingival epithelium for control sam-
ples [Figure 7(A)] and further membrane degradation
through its thickness for both TC and V [Figure 7(B,C),
respectively] groups. At this time in vivo, substantial soft tis-
sue migration through the membranes resulted in their
nearly complete resorption. Additionally, the healing

FIGURE 4. At 6 weeks, the (A) control defect area depicted dense connective tissue (CT) along with gingival epithelium migration (E) within the
defect and reduced amounts of bone formation. (B) TC and (C) V groups presented substantial new bone (NB) formation through the central region
of the defect. At this time in vivo, the collagen membranes (thickness bound between white arrows) were observed bounding the defect area and
were lined by bone in the inner aspect of the defect and by a layer of randomly oriented connective tissue (thickness bound by red arrows) present-
ing blood vessels and a mild inflammatory infiltrate in the outer defect aspect.

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B | MONTH 2018 VOL 000B, ISSUE 0 5

FIGURE 5. At 12 weeks, the (A) control defect area depicted dense CT within the defect and reduced amounts of bone formation. Gingival epithe-
lium (E) apical migration was also observed between roots for control groups. (B) TC and (C) V groups maintained the NB amounts formed at
6 weeks in vivo at the central region of the defect. At this time in vivo, substantial collagen membrane (marked by white arrows) degradation was
observed through tissue remodeling. Relative to 6 weeks, a substantial decrease in the amount of inflammatory infiltrate was observed for all
groups at soft tissue regions in proximity and away from both membrane types, where inflammatory cells were seldom observed and levels were
comparable to healthy tissue. A layer of randomly oriented connective tissue (thickness bound by red arrows) presenting blood vessels in the outer
defect aspect was still present.

connective tissue that was in direct contact with the mem-
branes at 6 and 12 weeks was fully remodeled toward a
dense connective tissue configuration. No full buccal/lingual
plate regeneration was observed for any of the experimental
groups. For all groups, inflammatory cell infiltrate evaluation
seldom showed inflammatory cells.
The data for collagen membrane presence, bone amount,
and inflammation are presented in Figure 8. At 6weeks, sig-
nificant bone formation occurred for the membrane-treated
sites, ~64%  17 (TC) and ~65%  13 (V) along with both
collagen membrane presence and inflammation [Figure 8
(A)]. The amount of bone observed for the membrane groups
at 12 weeks [Figure 8(B)] and 24 weeks [Figure 8(C)] was
similar to those observed at 6 weeks (Figure 8). However, a
substantial decrease in collagen membrane presence was
noted as time in vivo elapsed from 6 to 12 to 24 weeks
in vivo. Within each time in vivo the only parameter evalu-
ated, which presented statistically significant differences was
bone amount for both membrane groups relative to controls
at all times in vivo.

FIGURE 6. At 12 weeks, initial PDL regeneration was observed at areas where bone growth occurred parallel to the denuded roots for both mem-
brane groups. This Vitala membrane figure is a representative of all groups, which showed a certain level of cellular cementum (CC) and bone were
bridged by Sharpey’s fibers depicting regenerated PDL (RPDL) along the length of the reforming periodontal ligament space. Where no RPDL was
observed, fibrous connective tissue parallel to the bone and denuded root was observed. The white arrows depict the membrane presence.

6 TOVAR ET AL. IN VIVO EVALUATION OF RESORBABLE scCO2-TREATED COLLAGEN MEMBRANES

 ORIGINAL RESEARCH REPORT

FIGURE 7. The 24 week samples were remarkably similar to the 12 week samples. Exceptions included (A) further apical migration of the gingival
epithelium (E) occupying the defect along with CT for control samples, and almost complet membrane degradation for both (B) TC and (C) V
groups. At this time in vivo, substantial soft tissue migration through the membranes resulted in their nearly complete resorption. For all groups,
inflammatory cell infiltrate evaluation seldom showed inflammatory cells, and levels were comparable to healthy tissue.

DISCUSSION biological tissues in moderate temperature conditions with-

The periodontium is a complex anatomical structure com-
posed of bone, cementum and soft tissues such as periodon-
tal ligament and gingiva. Loss of any of these tissues due to
trauma or disease can result in compromised function, dis-
comfort and ultimately tooth loss. Once these tissues are
lost, spontaneous regeneration of the periodontium is lim-
ited while tissue repair becomes predominant.27 Typically,
healing following conventional surgical approaches results in
limited formation of new bone and cementum, as well as a
disorganized or non-functional periodontal ligament fibers
that run parallel to the root surface.28 Numerous approaches
based on periodontal tissue regeneration, such as tissue-
engineering29,30 and GTR24,31–33 have been used with some
degree of success and have lead to regeneration of periodon-
tal tissues.
Unlike in previous studies, where scCO2 was used as a
method of terminal sterilization of xenogeneic dermal
matrix, the focus of this study aimed to determine the physi-
cochemical properties and regenerative capacity of periodon-
tal tissues and the biologic response of scCO2 as an
adjunctive treatment to a conventional commercially used
method for decellularizing porcine pericardium derived col-
lagen GTR membrane, with no additional osteogenic or
osteoinductive biomaterials, in a Class III furcation lesion.
The rationale for the use of scCO2 technology was to evalu-
ate its potential use as an adjunctive in the processing of
out the potential for toxic residuals and with minimal
changes to the collagen structure.13 In theory, this method
could result in improvements in biological tissue proces-
sing/cleaning/safety with little to no disruption of the native
collagen structure (e.g., cross-linking or denaturation).
From a physicochemical perspective, SEM analysis
revealed a similar gross fibrous structural appearance
between membranes, while thickness measurements exhib-
ited an increased membrane thickness for the experimental
group, which was likely due to the temperature and pressure
necessary during scCO2 treatment.34 Such conditions
increase the diffusivity of CO2, making penetration into the
membrane scaffold easier without affecting its structure
(denaturing of the collagen or excessive crosslinking) and
potentially accounts for the structural expansion observed.
Although the thermal analysis via DSC characterization
showed no significant differences, aside from a small inflec-
tion difference which was observed for the control group,
Vitala®, the results are remarkably similar to work by Weh-
meyer et al.,13 which showed no degradation or loss of the
ECM proteins, while inactivating bioburden.
The degradation kinetics of a collagen membrane
becomes an important aspect in eliminating infiltration of
the fast-growing gingival tissue, thereby allowing the healing
of the periodontium through Sharpey fibers bridging cemen-
tum and bone as observed for both membrane experimental

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B | MONTH 2018 VOL 000B, ISSUE 0 7


FIGURE 8. Collagen membrane presence (N = 30 measurements (5/time
point/membrane)), bone amount (N = 30 measurements (5/time point/
membrane)), and inflammation (N = 30 measurements (5/time point/
membrane)) (mean  95%CI) at (A) 6 weeks, (B) 12 weeks and
(C) 24 weeks. Undoubtedly the bone amount for all samples was signifi-
cantly less for all control samples, independent of the time point.
groups at 12 weeks. The present investigation of the treated
and non-treated membrane samples showed low degrees of
resorption at 6 weeks (~63% on average), with an abun-
dance of vascularized bone formation in the furcation site.
Such conditions within the defects covered by membranes
from both groups promoted periodontal regeneration over
time as membrane mass slowly decreased. Control samples
8 TOVAR ET AL.
did not fare well and rapid-proliferating epithelial and con-
nective gingival tissue occupied the furcation site affecting
the osteogenic healing potential. Such a scenario did not
improve for the control group at 12 and 24 weeks.
At 12 weeks further bone remodeling is observed with
signs of complete periodontium: PDL, alveolar bone, and
cementum. The regenerated PDL was vascularized and
anchored by Sharpey’s fibers, showing features of native tis-
sue. Approximately 25% of both membranes remained at
12 weeks. At 24 weeks, samples showed a minimal amount
of membrane for both groups, while the amount of bone in
the site did not change. Overall, both membranes presented
significantly higher degrees of bone formation relative to the
control group and no detrimental effect of scCO2 treatment
was detected through histology.
Despite differences in fabrication methods, both mem-
branes’ degradation occurred to the same degree over time
despite the increased thickness observed for the scCO2-
treated group. The absorption rate of the membrane was
previously affected by its porosity, which is affected by the
purity of the source and manufacturing technique. Porosity
allows for the infiltration of cells, such as mononuclear cells,
which degraded the collagen membranes of both types with
minimal inflammatory infiltrate despite morphologic and
thickness differences, suggesting a low antigenic character.35
Our morphologic and ranked data showed that as the colla-
gen membranes were reabsorbed/remodeled the inflamma-
tion decreased showing the membrane biocompatibility
allowing for regeneration in the class III furcation site.
Pontoriero et al.36 reported that the regeneration of class
III furcation defects was always incomplete through a GTR
procedure, which is in direct agreement with our results.
Although significantly higher amounts of bone were
observed between membrane sites and control, full peri-
odontal tissue regeneration was never complete for any
group due to the surgical approach taken where the mem-
branes were placed in direct contact with the denuded roots
in an attempt to cover the class III furcation defects (i.e., no
full buccal plate regeneration occurred through the whole
defect extension, especially at the interproximal root buccal
areas). However, in areas where bone healing around
denuded roots did occur due to the lack of membrane con-
tact with denuded roots, initial periodontal ligament regen-
eration was observed through the bridging of highly cellular
cementum to regenerated bone.
In another study concerning periodontal ligament regener-
ation implementing a membrane barrier,37 the healing envi-
ronment within the volume bounded by the membranes
resulted in multiple cell differentiation to form bone (osteo-
blasts) and cementum (cementoblasts) at 12 weeks. These
observations further confirm that space maintenance by the
barriers enabled localized cellular events to occur, allowing for
the natural healing without disruption. No morphologic differ-
ences were observed for regenerated periodontal ligament
between 12 and 24 weeks, whereas the randomly oriented
connective tissue that was in proximity with the membrane
toward soft tissue was fully remodeled to dense connective tis-
sue typically observed along with gingival epithelium.
IN VIVO EVALUATION OF RESORBABLE scCO2-TREATED COLLAGEN MEMBRANES

CONCLUSION
In conclusion, membranes which underwent the supplemen-
tal treatment with scCO2, demonstrated to be biocompatible
and simultaneously presented similar GTR performance in
class III furcation lesions, in comparison to their counter-
parts, which were processed following the conventional
methodology. Similarly, implementation of scCO2 sterilization
instead of irradiation could be used as a safe adjunctive
treatment to improve the quality of the collagen products for
GTR purposes.
ACKNOWLEDGMENTS
This study was funded by Osteogenics Biomedical, Inc.
(Lubbock, TX).
REFERENCES
1. Service RF. Tissue engineers build new bone. Science 2000;
289(5484):1498–1500.
2. Hench LL, Wilson J. An Introduction to Bioceramics. Advanced
Series in Ceramics. Singapore: World Scientific Publishing; 1993
p. 386 https://doi:10.1142/9789814317351_0001.
3. Chaput C, Selmani A, Rivard C. Artificial scaffolding materials for
tissue extracellular matrix repair. Curr Opin Orthop 1996;7:62–68.
4. Polimeni G, Xiropaidis AV, Wikesjo UM. Biology and principles of
periodontal wound healing/regeneration. Periodontol 2006;41:
30–47.
5. Needleman IG, Worthington HV, Giedrys-Leeper E, Tucker RJ.
Guided tissue regeneration for periodontal infra-bony defects.
Cochrane Database Syst Rev 2006;2:CD001724.
6. Siciliano VI, Andreuccetti G, Siciliano AI, Blasi A, Sculean A,
Salvi GE. Clinical outcomes following treatment of non-contained
Intrabony defects with enamel matrix derivative (EMD) or guided
tissue regeneration (GTR). A 12-month randomized controlled clini-
cal trial. J Periodontol 2011: https://doi10.1902/jop.2010.100144.
7. de Andrade PF, de Souza SL, de Oliveira Macedo G, Novaes AB, Jr.,
de Moraes Grisi MF, Taba M, Jr., Palioto DB. Acellular dermal
matrix as a membrane for guided tissue regeneration in the treat-
ment of class II furcation lesions: a histometric and clinical study in
dogs. J Periodontol 2007; 78(7):1288–99.
8. Eickholz P, Pretzl B, Holle R, Kim TS. Long-term results of guided
tissue regeneration therapy with non-resorbable and bioabsorbable
barriers. III. Class II furcations after 10 years. J Periodontol 2006;
77(1):88–94.
9. Behring J, Junker R, Walboomers XF, Chessnut B, Jansen JA.
Toward guided tissue and bone regeneration: morphology, attach-
ment, proliferation, and migration of cells cultured on collagen bar-
rier membranes. A systematic review. Odontology 2008;96(1):1–11.
10. Keles GC, Cetinkaya BO, Baris S, Albayrak D, Simsek SB. Compari-
son of platelet pellet with or without guided tissue regeneration in
the treatment of class II furcation defects in dogs. Clin Oral Investig
2009;13(4):393–400.
11. Camargo PM, Lekovic V, Weinlaender M, Vasilic N, Madzarevic M,
Kenney EB. Platelet-rich plasma and bovine porous bone mineral
combined with guided tissue regeneration in the treatment of
intrabony defects in humans. J Periodontal Res 2002;37(4):300–306.
12. Qiu QQ, Leamy P, Brittingham J, Pomerleau J, Kabaria N,
Connor J. Inactivation of bacterial spores and viruses in biological
material using supercritical carbon dioxide with sterilant. J Biomed
Mater Res B Appl Biomater 2009;91(2):572–578.
13. Wehmeyer JL, Natesan S, Christy RJ. Development of a sterile
amniotic membrane tissue graft using supercritical carbon dioxide.
Tissue Eng Part C Methods 2015;21(7):649–659.
14. Balestrini JL, Liu A, Gard AL, Huie J, Blatt KM, Schwan J, Zhao L,
Broekelmann TJ, Mecham RP, Wilcox EC and others. Sterilization
of lung matrices by supercritical carbon dioxide. Tissue Eng Part C
Methods 2016;22(3):260–269.
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B | MONTH 2018 VOL 000B, ISSUE 0
View publication stats
ORIGINAL RESEARCH REPORT
15. Fages J, Marty A, Delga C, Condoret JS, Combes D, Frayssinet P.
Use of supercritical CO2 for bone delipidation. Biomaterials 1994;
15(9):650–656.
16. Hennessy RS, Jana S, Tefft BJ, Helder MR, Young MD,
Hennessy RR, Stoyles NJ, Lerman A. Supercritical carbon dioxide-
based sterilization of decellularized heart valves. JACC Basic Transl
Sci 2017;2(1):71–84.
17. Setlow B, Korza G, Blatt KM, Fey JP, Setlow P. Mechanism of Bacil-
lus subtilis spore inactivation by and resistance to supercritical CO2
plus peracetic acid. J Appl Microbiol 2016;120(1):57–69.
18. White A, Burns D, Christensen TW. Effective terminal sterilization
using supercritical carbon dioxide. J Biotechnol 2006;123(4):
504–515.
19. Sawada K, Terada D, Yamaoka T, Kitamura S, Fujisato T. Cell
removal with supercritical carbon dioxide for acellular artificial tis-
sue. J Chem Technol Biotechnol 2008;83(6):943–949.
20. Lin HM, Yang Z, Chen LF. Inactivation of Saccharomyces cerevisiae
by supercritical and subcritical carbon dioxide. Biotechnol Prog
1992;8(5):458–461.
21. Joslyn A, Walden T, AlloSource, assignee. Supercritical carbon
dioxide tissue processing methods patent 14/794, 354.2016.
22. Guler S, Aslan B, Hosseinian P, Aydin HM. Supercritical carbon
dioxide-assisted Decellularization of aorta and cornea. Tissue Eng
Part C Methods 2017;23(9):540–547.
23. Armentano I, Dottori M, Fortunati E, Mattioli S, Kenny JM. Biode-
gradable polymer matrix nanocomposites for tissue engineering: a
review. Polym Degrad Stab 2010;95(11):2126–2146.
24. Coelho PG, Giro G, Kim W, Granato R, Marin C, Bonfante EA,
Bonfante S, Lilin T, Suzuki M. Evaluation of collagen-based mem-
branes for guided bone regeneration, by three-dimensional com-
puterized microtomography. Oral Surg Oral Med Oral Pathol Oral
Radiol 2012;114(4):437–443.
25. Donath K, Breuner G. A method for the study of undecalcified
bones and teeth with attached soft tissues. The Sage-Schliff (saw-
ing and grinding) technique. J Oral Pathol 1982;11(4):318–326.
26. Tovar N, Bourke S, Jaffe M, Murthy NS, Kohn J, Gatt C, Dunn MG.
A comparison of degradable synthetic polymer fibers for anterior
cruciate ligament reconstruction. J Biomed Mater Res A 2010;93(2):
738–747.
27. Bartold PM, McCulloch CA, Narayanan AS, Pitaru S. Tissue engi-
neering: a new paradigm for periodontal regeneration based on
molecular and cell biology. Periodontol 2000 2000;24:253–269.
28. Liu Y, Zheng Y, Ding G, Fang D, Zhang C, Bartold PM, Gronthos S,
Shi S, Wang S. Periodontal ligament stem cell-mediated treatment
for periodontitis in miniature swine. Stem Cells 2008;26(4):1065–1073.
29. Yang Y, Rossi FM, Putnins EE. Periodontal regeneration using engi-
neered bone marrow mesenchymal stromal cells. Biomaterials
2010;31(33):8574–8582.
30. Tobita M, Uysal CA, Guo X, Hyakusoku H, Mizuno H. Periodontal
tissue regeneration by combined implantation of adipose tissue-
derived stem cells and platelet-rich plasma in a canine model.
Cytotherapy 2013;15(12):1517–1526.
31. Araujo MG, Berglundh T, Lindhe J. GTR treatment of degree III fur-
cation defects with 2 different resorbable barriers. An experimental
study in dogs. J Clin Periodontol 1998;25(3):253–259.
32. Araujo MG, Berglundh T, Albrekstsson T, Lindhe J. Bone formation
in furcation defects. An experimental study in the dog. J Clin Peri-
odontol 1999;26(10):643–652.
33. Caffesse RG, Dominguez LE, Nasjleti CE, Castelli WA, Morrison EC,
Smith BA. Furcation defects in dogs treated by guided tissue regen-
eration (GTR). J Periodontol 1990;61(1):45–50.
34. Chiara G. Porous Polymeric Bioresorabable Scaffolds for Tissue
Engineering. Berlin, Heidelberg: Spring-Verlag; 2011.
35. Taga RCT, Taga EM. Avaliação histológica, radiográfica e morfomé-
trica da reparação de defeito ósseo perene em crânio de cobaia tra-
tado com mistura de Osseobond e Biohapatita e membrana
absorvível de cortical óssea bovina. J Bras Endo/Perio 2000;1:78–87.
36. Pontoriero R, Nyman S, Ericsson I, Lindhe J. Guided tissue regener-
ation in surgically-produced furcation defects. An experimental
study in the beagle dog. J Clin Periodontol 1992;19(3):159–163.
37. Consolaro A. The four mechanisms of dental resorption initiation.
Dental Press J Orthod 2013;18(3):7–9.
9