J Oral Maxillofac Surg

48:1305-1309 1990

Poly(L-Lactide) Implants in Repair of

Defects of the Orbital Floor:
An Animal Study
FRED R. ROZEMA, DDS,* RUDOLF R.M. BOS, DDS, PHD,t
ALBERT J. PENNINGS, MS, PHD,~ AND HENK W.B. JANSEN, MS, PI-IDS

Because of the life-long presence of alloplastic, nonresorbable orbital

floor implants and the complications of their use mentioned in literature,
the use of a resorbable material appears to be preferable in the repair of
orbital floor defects. A high-molecular-weight, as-polymerized poly(L-
lactide) (PLLA) was used for repair of orbital floor defects of the blowout
type in goats. An artificial defect was created in the bony floor of both
orbits. Reconstruction of the orbital floor was then carried out using a
concave PLLA implant of 0.4-mm thickness. At 3, 6, 12, 19, 26, 52, and 78
weeks postoperatively, one goat was killed. Microscopic examination
showed full encapsulation of the implant by connective tissue after 3
weeks. After 6 weeks, resorption and remodeling of the bone at the points
of support of the implant could be detected. A differentiation between the
sinus and orbital sides of the connective tissue capsule was observed. The
orbital side showed a significantly more dense capsule than the antral
side, which had a loose appearance. At 19 weeks, a bony plate was pro-
gressively being formed, and at 78 weeks, new bone had fully covered the
plate on the antral and orbital side. No inflammation or rejection of the
PLLA implant was seen.

Many authors have advocated surgery for trau-
matically induced defects of the orbital floor.‘-6 The
main purpose of such surgery is to prevent the pa-
tient from having residual diplopia and enophthal-
mus. In the past, a variety of materials have been
used to reconstruct defects of the orbital floor or to
support the orbital contents. These include not only
autografts, but also xenografts and allografts. The
disadvantages of autografts are the additional oper-
ation required to procure these grafts and the risk of
Received from the University of Groningen, The Netherlands.
* Department of Oral and Maxillofacial Surgery.
t Department of Oral and Maxillofacial Surgery.
$ Department of Polymer Chemistry.
§ Department of Histology and Cell Biology.
Address correspondence and reprint requests to Dr Rozema:
Department of Oral and Maxillofacial Surgery, University Hos-
pital Groningen. Oostersingel59.9713 EZ Groningen, The Neth-
erlands.
0 1990 American Association of Oral and Maxillofacial
geons
0278-2391/90/4812-0010$3.00/O
rejection. One objection to the use of an allograft
such as lyophilized human dura is the insufficient
stiffness of the wet material, which can cause dif-
ficulties in bridging larger defects of the orbital
floor.7~8
Alloplastic materials are popular today because
of availability without an additional operation and
their ease of use. The alloplastic implants com-
monly used are of the nonresorbable type. How-
ever, materials such as Silastic (Dow Corning, Mid-
land, MI), Teflon (Dow Chemical Co, Wilmington,
DE), methylmethacrylate, and polyethylene have
been known to cause complications.9-” These in-
clude incidental cases of extrusion or migration of
the implant, residual diplopia, infection of the for-
eign-body implant, dental abscess, or infection of
the maxillary sinus.
Considering the above mentioned complications
and the life-long presence of these implants, the use
Sur- of resorbable alloplastic material appears to be pref-
erable. The resorbable implants should permit suf-
ficient temporary support of the orbital tissues, es-

1305
1306
pecially during the period of formation of scar tissue
or new bone, but disappear after healing is com-
plete.
Resorbable alloplastic implants such as polylactic
acid,” GeltXm,‘3 polyglactin,14 and polydioxanone
(PDS) (Ethicon, Somerville, NJ) have already been
used. However, these materials have certain disad-
vantages, such as rapid resorption, the bulkiness of
the implant (up to 2-mm thickness), and inferior me-
chanical properties. Poly(L-lactide) (PLLA), a new
material with improved mechanical properties, has
been developed by Leenslag et al. l5 This PLLA has
been successfully tested preclinically and clinically
in the form of boneplates and screws for internal
fixation of mandibular and zygomatic fractures.‘6-20
These favorable results have led us to undertake the
present animal study to determine the suitability of
high-molecular-weight as-polymerized PLLA im-
plants for repair of orbital floor fractures of the
blowout type.
Materials and Methods
PREPARATION OF PLLA IMPLANTS
The shape and dimensions of the implants were
based on a study of the anatomical dimensions in
the orbital region of both dried goat and human
skulls. The implants, which were machined from a
block of as-polymerized PLLA,” were concave,
0.4 mm thick, and 30 mm in diameter. Perforations
of 2 mm diameter were made to allow tissue in-
growth (Fig 1).
ANIMAL STUDY
A total of 15 goats (European breed, 60 to 80 kg)
were used. A study on dried skull collections
showed that the goat has an orbital floor with a
shape that closely resembles that of humans.
After intravenous injection of diazepam (1 mg/kg
body weight), glycopyrronium bromide (2 mg), and
ketamine hydrochloride 10% (10 mg/kg body
weight), intubation was performed and anesthesia
was maintained using a nitrous oxide-oxy-
gen-fluothane mixture supplemented with sufficient
FIGURE 1. Photograph of the
PLLA implant. The implant has a
glassy appearance.
POLY(L-LACTIDE) ORBITAL FLOOR IMPLANTS
ketamine hydrochloride for relaxation (1,500
mg/500 mL saline, 20 to 30 drops/min).
Both eye sockets were operated on using an in-
ferior orbital rim approach. The orbital floor was
exposed by incising and elevating the periosteum
and an artificial defect, approximately 15 mm in di-
ameter, was created in the bone and sinus epithe-
lium with the aid of a blunt instrument and a
rongeur. Care was taken not to damage the infraor-
bital rim. The orbital contents were allowed to pro-
lapse in the maxillary sinus by perforating the sup-
porting tissues using scissors. A concave implant
of PLLA was adapted to the floor of the orbit, tak-
ing care that the entire defect was covered. Excess
areas of the implant were trimmed with scissors.
Just before the implant was inserted, the orbital
contents that had prolapsed were repositioned. The
implant was fixed to the infraorbital rim with one
Dexon (Davis & Geck; Cyanamid, Seoul, Korea)
resorbable suture. The wound was closed in layers
with Dexon sutures. The operation on both eyes
was identical.
EVALUATION OF HEALING
The goats were examined clinically, without mea-
suring diplopia and enophthalmus. They were ob-
served daily during the 1st week, and regularly
thereafter. Photographs were taken at 1, 2 and 3
weeks postoperatively.
At 3, 6, 12, 19, 26, 52, and 78 weeks postopera-
tively, one goat was killed using pentobarbital 200
mg/mL intravenously. After the orbital area was ex-
cised, it was quick-frozen and sawed in half sagit-
tally through the lens and the artificially created
defect. Photographs were taken of each half to rec-
ord the position of the implant.
After fixation in 4% formaldehyde, slices of the
orbital region (5 mm) were dehydrated in graded
series of ethanol and acetone and embedded in
polyester resin. Sections were then made of the re-
gion comprising the orbit, periorbit, maxillary si-
nus, sinus roof, and the fractured area in a plane
through the lens and the artificially made defect.
After staining of the surface of the block with tolu-
idine blue and basic-fuchsin, 30+m sections were
cut with a Leitz 1600 diamond saw (Ernst Leitz,
GNBH, Wetzlar, FRG) for light-microscopic eval-
uation.
RemIts
MACROSCOPIC FINDINGS
Macroscopic examination of the orbital halves
showed that all intra- and periorbital structures
were clearly visible, including a cross-section of the
implant. The implant was still in place. It was cov-
ROZEMA ET AL
ered by a thin layer of fibrous tissue on both the
orbital and the maxillary sinus side. From week 12,
the orbital tissue covering the implant became
somewhat thicker. The maxillary sinus side resem-
bled the epithelial layer seen in a normal sinus. At
week 26, 52, and 78, this situation appeared to be
stable. No pathologic reactions could be seen.
MICROSCOPICFINDINGS
The 3-week sample showed an implant fully cov-
ering the defect, totally encapsulated by loose fi-
brous tissue. No inflammation was seen. At 6
weeks, resorption and remodeling of bone at the
points of support of the implant could be detected.
The fibrous tissue capsule on the sinus side was
very thin. On the orbital side, the capsule was
slightly thicker. There was no sign of inflammation.
In the 12-week sample, the orbital side showed a
significantly more dense capsule than the antral
side, which had a loose appearance. Also, ingrowth
was observed in the perforations of the implant.
Normal sinus epithelium containing goblet cells
covered the maxillary sinus side. Capillaries were
observed in the loose connective tissue. The num-
ber of layers of connective tissue was estimated at
30, with a total thickness of approximately 150 pm.
Bone resorption was noted at the points of support
of the implant. New bone was being formed, a thin
layer of which just covered the outline of the im-
plant.
At 19 weeks, a progressive bony plate was found,
which partially covered the implant from the border
toward the center. This bony plate had been formed
in apposition to the outer side of the fibrous con-
nective tissue capsule. It was more prominent on
the antral side than on the orbital side. In the per-
forations, new bone formation was also observed,
arising from the antral side.
At week 26, the antral side showed a mature con-
nective tissue capsule. The whole implant, includ-
ing the perforations, was covered with a layer of
dense connective tissue. A thin bony plate was seen
in apposition to this fibrous complex. In the perfo-
rations bone formation could be detected. The de-
fect was healed and the maxillary sinus side was
covered with pseudostratified epithelium containing
goblet cells. Capillaries were observed in the sub-
epithelial connective tissue (Fig 2). On the orbital
side, a thin layer of dense connective tissue was
observed. The adjacent orbital fat was normal in
appearance. No inflammatory or foreign-body reac-
tions were seen.
The 78-week specimen showed a mature connec-
tive tissue capsule. New bone had now fully cov-
ered the PLLA implant on the antral and orbital
1307
FIGURE 2. Photomicrograph showing the implant (P), fully en-
capsulated by connective tissue. A thin, newly formed bony
plate is in apposition to this fibrous complex. On the bottom side,
a layer of the maxillary sinus mucosa containing goblet cells can
be recognized (arrows) (26 weeks; hematoxylin-eosin stain, orig-
inal magnification x 16).
sides. The perforations in the implant also were
filled with new bone. Both the orbital aspect and
maxillary sinus side had a normal appearance. The
orbital fat was also normal. The implant was still in
place, with no visible changes (Fig 3). No intlam-
mation or foreign-body reactions were seen (Fig 4).
Discussion
After 3 weeks, the orbital floor implant was to-
tally encapsulated by connective tissue. Experi-
ments on polymer implants in general, and also in
the orbital floor region as reported in literature,
showed similar encapsulation patterns. 12~14,18*21*22
As is generally accepted, the implant acts as a for-
eign body, causing irritation to the surrounding tis-
sue. It is well known that any tissue damage causes
a proliferative and reparative connective tissue re-
sponse. In the first period of implantation, no sub-
stantial resorption of the implant occurred. There-
fore, no disturbance in capsule formation took
place. The formation of fibrous tissue eventually
turns into encapsulation of the implant.
Davilla et al23 and Matlaga et al” showed that the
degree of encapsulation is directly dependent on the
size, geometrical shape, and certain physical as-
pects of the implant. This makes the PLLA implant
suitable for application in the orbital floor area. Our
experiment has demonstrated that the tissue reac-
tion to the PLLA implant is very mild, causing only
a thin capsule of connective tissue.
The degree of soft-tissue response can be in&t-
1308
enced by the surface texture of an polymer implant.
Taylor and Gibbon? showed that surface texturing
in polytetrafluoroethylene (PTFE) implants in-
creased the number of giant cells and macrophages.
The smooth surface of the PLLA implant did not
attract giant cells and macrophages.
The implant is fixed with only one suture to the
infraorbital rim. During postoperative observation,
and in the histologic findings, no signs of migration
were found. Brown et alz6 observed an increased
host-tissue reaction when an implant was mobile for
a long time. The histologic examination did not
show signs of increased tissue response.
There was a striking difference between the tis-
sue reaction on the orbital side, showing a dense
appearance, and the loose connective tissue on the
FIGURE 4. High-power photograph of the interface of the
PLLA implant. A thin layer of fibrous connective tissue (CT)
containing spindle-shaped fibroblasts and collagen fibers is seen
between the bone (B) and the implant (P). No signs of intlamma-
tion are visible (hematoxylin-eosin stain, original magnification
x 160).
POLY(L-LACTIDE) ORBITAL FLOOR IMPLANTS
FIGURE 3. Photomicro-
graph showing a PLLA im-
plant in situ 78 weeks after im-
plantation. On the left side,
the implant is completely em-
bedded in newly formed bone.
On the right side, one of the
perforations is visible. Note
the tearing artifact in the max-
illary sinus mucosa due to cut-
ting (hematoxylin-eosin stain,
original magnification X 1).
antral side. Sevastjanova21 described the biome-
chanical features that may influence collagen com-
position and appearance. Eye movement and grav-
ity may be responsible for a more dense composi-
tion of the connective tissue covering the orbital
side of the PLLA implant. The antral side has no
load-bearing function, resulting in a loose connec-
tive tissue.
After insertion, the implant was covered by peri-
osteum. At 6 weeks, probably due to eye movement
and loading of the implant, resorption and remod-
eling of bone at the points of support of the implant
could be detected. Cutright and Hunsuck’* ob-
served similar recontouring and osteoblastic activ-
ity using their fast-resorbing implant of compres-
sion-molded polylactic acid (PLA).
At week 12, the PLLA implant showed new bone
formation in apposition to the fibrous tissue cap-
sule. The perforations were also filled with new
bone. Both processes started on the antral side.
Cutright and Hunsuck12 reported in a monkey ex-
periment that PLA implants do not inhibit new bone
formation at the site of the artificially made defect.
In a study on osseous repair in discontinuity de-
fects in dog mandibles treated with the copolymer
of PLA:PGA (polyglycolic acid), Hollinger2’
showed that these fractures healed more rapidly
than untreated defects. A linear increase of bony
reparative elements was suggested, but this phe-
nomena might be explained by the combination of
the copolymer with a phospholipid. In our study it
was difficult to determine the osteogenic or osteo-
conductive potential of the PLLA. For this pur-
pose, a control side is needed to compare defects
treated with PLLA implants with those left un-
treated.
This study showed that in goats, the specially de-
signed PLLA implant, of 0.4-mm thickness, gave
sufficient support to the orbital tissue during heal-
ing. Histologic analysis showed encapsulation and
new bone formation completely covering the im-
ROZEMA ET AL
plant. The PLLA was very well tolerated and did
not give rise to any clinically detectable inflamma-
tory or foreign-body reaction.
At the end of the longest period of observation
(78 weeks), the PLLA implants were not yet com-
pletely disintegrated. However, in vitro (phosphate-
buffered aqueous saline solution) and in vivo deg-
radation of PLLA samples performed in rats,
sheep, and dogs showed a decrease of molecular
weight (a,) and tensile strength (ub) up to 95% after
12 weeks. Loss of mass was observed from 26
weeks onward. 18,28Based on the animal studies, it
can be estimated that the implant will be fully re-
sorbed in about 3.5 years. It would be preferable to
shorten this total resorption time, as the long pres-
ence of the implant may result in some of the same
complications seen with nonresorbable implants.
As clinical examination showed no complications
in the healing process, the method of treatment ap-
pears to be well tolerated. The implant support was
sufficient for the eyeball to assume a normal posi-
tion. Considering the disadvantages of today’s im-
plant materials, the high-molecular-weight, as-
polymerized PLLA appears to have promise in the
management of orbital blowout fracture.
Acknowledgment
The authors wish to thank Professor Geert Boering (head of
the Department of Oral and Maxillofacial Surgery) for his ideas
and stimulating help, and for critically reviewing the manuscript.
The technical assistance of Adams B. Verweij (Department of
Polymer Chemistry, University of Groningen) is greatly appre-
ciated.
References
I. Converse JM. Smith B, Obear MF, et al: Orbital blowout
fractures: A IO-year survey. Plast Reconstr Surg 39:20,
1967
2. Crikelair GF, Rein JM, Potter GD, et al: A critical look at the
“blow-out” fracture. Plast Reconstr Surg 49:374, 1972
3. van Herk W, Hovinga J: Choice of treatment of orbital floor
fractures as part of facial fractures. J Oral Surg 31600,
1973
4. Hawes JM, Dortzbach RK: Surgery on orbital floor frac-
tures. Ophthalmology 9Ck1066, 1983
5. Bleeker GM, van Ommen B: Early treatment of orbital frac-
tures. Onhthalmoloaica 13840. 1959
6. de Man K: Fractures-of the orbital floor: Indications for
exploration and for the use of a floor implant. J Maxillofac
Surg 12:47. 1984
7. de Man K: Orbitabodemfracturen. Thesis, Utrecht, Ant-
werpen, Bohn, Scheltema & Holkema, 1982, p 61
8. Webster K: Orbital floor repair with lyophilized porcine der-
mis. Oral Surg 65161, 1988
1309
9. Mauriello JA, Flanagan JC, Peyster RG: An unusual late
complication of orbital floor fracture repair. Ophthalmol-
ogy 91:102, 1984
10. Sewall SR, Pemoud FG, Pemoud MJ: Late reaction to sili-
cone following reconstruction of an orbital floor fracture.
J Oral Maxillofac Surg 44:821, 1986
11. Mauriello JA, Fiore PM, Ketch M: Late complication of
orbital floor fracture repair with implant. Ophthalmology
941248, 1987
12. Cutright DE, Hunsuck EE: The repair of fractures of the
orbital floor using biodegradable polylactic acid. Oral
Surg 33~28, 1972
13. Burres SA, Cohn AM, Mathog RH: Repair of orbital blow-
out fractures with Marlex mesh and gelftlm. Laryngo-
scope 91:1881, 1981
14. Holtje WJ: Wiederstellung von Orbitabodendefekten mit
Polyglactin. Eine tierexperimentelle Studie. Fortschritte
der Kiefer und Gesichts- Chirurgie Bd 28. Stuttgart, New
York, Thieme, 1983, p 65
15. Leenslag JW, Pennings AJ: Synthesis and morphology of
high-molecular weight poly-(L-lactide). Makromol Chem
188:1809, 1987
16. Bos RRM, Boering G, Rozema FR, et al: Resorbable
Poly(L-lactide) plates and screws for the fixation of Zy-
gomatic fractures. J Oral Maxillofac Surg 45751, 1987
17. Rozema FR, Bos RRM, Boering G, et al: Experimental frac-
tures of the mandibular body of sheep and dogs. A new
technique. Br J Oral Maxillofac Surg 27: 163, 1989
18. Bos RRM, Rozema FR, Boering G, et al: Degradation on
and tissue reaction to biodegradable poly(L-lactide) for
use as internal fixation of fractures. A study on rats. Bi-
omaterials (accepted for publication)
19. Bos RRM, Rozema FR, Boering G, et al: Bone plate and
screws of bioabsorbable poly(L-lactide). An animal pilot
study. Br J Oral Maxillofac Surg 27:467, 1989
20. Bos RRM, Rozema FR, Boering G, et al: Bio-absorbable
plates and screws for the fixation of mandibular fractures.
A study in 6 dogs. Int J Oral Maxillofac Surg 18365, 1989
21. Sevastjanova NA, Mansurova LA, Dombrovska LE, et al:
Biochemical characterization of connective tissue reac-
tion to synthetic polymer implants. Biomaterials 8:242,
1987
22. Gerlach KL: Biologisch abbaubare Polymere in der Mund-,
Kiefer-, Gesichtschirurgie. Tierexperimentelle Untersu-
chungen. Mtinchen, Wien, Hanser, 1988, p 77
23. Davilla JC. Lautsch EV, Palmer TE: Some physical factors
affecting the acceptance of synthetic materials as tissue
implants. Ann NY Acad Sci 146:138, 1968
24. Matlaga BF, Yasenchak LP, Salthouse TN: Tissue response
to implanted polymers: The significance of sample shape.
J Biomed Mater Res 10:391, 1976
25. Taylor SR, Gibbons DF: Effect of surface texture on the soft
tissue response to polymer implants. J Biomed Mater Res
17:205, 1983
26. Brown JB, Fryer MP, Ohlwiler DA: Study and use of syn-
thetic materials such as silicone and Teflon, as subcuta-
neous prostheses. Plast Reconstr Surg 26:264, 1960
27. Hollinger JO, Schmitz JP: Restoration of bone discontinui-
ties in dogs using a biodegradable implant. J Oral Maxil-
lofac Surg 45:954, 1987
28. Bos RRM, Rozema FR, Boering G, et al: In vivo and in vitro
degradation of poly(L-lactide) used for fracture fixation,
in Putter C de, Lange GL, Groot K de, et al (eds): Implant
Materials in Biofunction. Advances in Biomaterials. ~018.
Amsterdam, Elsevier, 1988, pp 245-250