Biomaterials 23 (2002) 41674176

Bone repair in radii and tibias of rabbits with phosphorylated

chitosan reinforced calcium phosphate cements
Xiaohong Wanga, Jianbiao Maa,b,*, Yinong Wanga, Binglin Hea
a

The State Key Laboratory of Functional Polymer Materials for Adsorption and Separation, Institute of Polymer Chemistry,
Nankai University, Tianjin 300071, Peoples Republic of China
b
Research Center of Biomaterials and Biomedical Devices, Tianjin Polytechnic University, Tianjin 300160, Peoples Republic of China
Received 25 June 2001; accepted 18 April 2002

Abstract
Biocompatibility of two calcium phosphate cements (CPCs), reinforced with phosphorylated chitosan (P-chitosan), was
investigated in rabbits in present study. The two CPCs are monocalcium phosphate monohydrate (MPCM) with calcium oxide
(CaO) in 1 m phosphate buffer (i.e. MCPM/CaO/1 m phosphate buffer cement, CPC-I) and dicalcium phosphate dihydrate (DCPD)
with calcium hydroxide [Ca(OH)2] in 1 m Na2HPO4 solution (i.e. DCPD/Ca(OH)2/1 m Na2HPO4 cement, CPC-II). Different
amount of P-chitosan was added to the liquid phase before the power phase was mixed with the liquid phase. The MCPM/CaO/1 m
phosphate buffer/P-chitosan cements (P-CPC-I) with neutral pH were lled into the holey defects of rabbit tibias. While the DCPD/
Ca(OH)2/1 m Na2HPO4/P-chitosan cements (P-CPC-II) shaped as prehardened cylinders were implanted into rabbit radial defects.
After operation, the two serial groups and CPC-II controls were observed for 1, 4, 12 and 22 weeks, respectively. Histological and
histomorphological studies proved that P-chitosan containing cements are biocompatible, bioabsorbable and osteoinductive. The
biodegradation rate has a negative relationship with the P-chitosan content. Progressive substitution took place at the interface of
implants and host bones. No adverse effects were found in tissues around the bone defects. Thus, they could be used as bone
substitutes in clinic. r 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Skeletal repair; Phosphorylated chitosans; Calcium phosphate cements; Bone substitutes; Osseoinduction

1. Introduction
Large bone repair is always a tough problem of
surgeons in clinics. The natural biological materials for
bone defect repair, including autografts and allografts,
each has its own shortcomings such as donor site
morbidity and donor shortage for autografts [2],
immunologic response and endemic risk for allografts
[3]. Although numerous synthetic bone substitutes using
metals, ceramics, and polymers have been developed to
promote bone regeneration for several decades, there
exists a lack of condence in their biological performances, particularly in view of long term, in vivo safety
and efcacy [4]. For example, metals are difcult to
*Corresponding author. State Key Laboratory of Functional
Polymer Materials for Adsorption and Separation, Institute of
Polymer Chemistry, Nankai University, Tianjin 300071, Peoples
Republic of China. Tel.: +86-22-350-1386; fax: +86-22-350-2749.
E-mail address: jbma@nankai.edu.cn (J. Ma).

shape and have a problem of corrosion which leads to

inammatory responses of surrounding tissue [5,6].
Polymers are usually encapsulated and do not integrate
with tissue, resulting in implant movement, inammatory responses and implant extrusion [5,7]. Ceramics are
fragile and difcult to sculpt to skeletal defects [5,8,9].
Recently, much attention has been paid to calcium
phosphate cements (CPCs), for their highly osteoconductive, readily osseointegrated, easily shaped (injected)
and gradually degradable properties [10,11]. However,
when compared with natural bone, the lack of toughness
and binding ability are still their undesired properties.
So some researchers added water-soluble polymers such
as sodium alginate [12], hydroxypropyl methylcellulose
and carboxymethylcellulose [13] into tetracalcium phosphate (TTCP) containing cements to improve the
handling properties. But these polymers were hardly
biodegradable in vivo and their usage in bone substitutes had been greatly limited. In our previous study,
water-soluble phosphorylate chitosans (P-chitosans),

0142-9612/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 1 4 2 - 9 6 1 2 ( 0 2 ) 0 0 1 5 3 - 9

4168

X. Wang et al. / Biomaterials 23 (2002) 41674176

Fig. 1. Radiographs of the prehardened samples (P-CPC-II) after implantation: (a) 1 week (P-chitosan: 0.12 g/ml); (b) 4 weeks (P-chitosan: 0.12 g/
ml); (c) 12 weeks (P-chitosan: 0.12 g/ml); (d) 12 weeks (P-chitosan: 0.07 g/ml); (e) 12 weeks (P-chitosan: 0.02 g/ml); (f) 12 weeks (P-chitosan: 0 g/ml);
(g) 22 weeks (P-chitosan: 0.12 g/ml).

which could be quickly biodegraded by lyzosome in

vivo, was prepared and admixed with two easily
prepared CPCs [14,15]. The two CPCs are monocalcium
phosphate monohydrate (MPCM) with calcium oxide
(CaO) in 1 m phosphate buffer (i.e. MCPM/CaO/1 m

phosphate buffer cement, CPC-I) and dicalcium phosphate dihydrate (DCPD) with calcium hydroxide
[Ca(OH)2] in 1 m Na2HPO4 solution (i.e. DCPD/
Ca(OH)2/1 m Na2HPO4 cement, CPC-II). With
certain amount of P-chitosan, the resulted P-chitosan

X. Wang et al. / Biomaterials 23 (2002) 41674176

4169

Fig. 2. Tissue responses to a P-CPC-II specimen in radius (P-chitosan:

0.12 g/ml) 1 week after implantation (decalcied, magnication
40,
MT stain).

Fig. 4. Bone formation induced by P-CPC samples 12 weeks after

operation: (a) verticle section of a CPC-II sample (P-chitosan: 0.02 g/
ml, decalcied, magnication
40, MT stain); (b) bone formation in
a tibia (P-chitosan: 0.05 g/ml, decalcied, magnication
40, HE
stain).
Fig. 3. Tissueimplant interface of a P-CPC-II (P-chitosan: 0.12 g/ml)
sample in radius 4 weeks after implantation (decalcied, magnication

40, MT stain).

containing CPCs (P-CPCs) have high mechanical

strength and slightly prolonged setting time. In vitro,
experiments have shown that P-chitosan can increase the
dissolubility of the start materials of cements and bind
calcium phosphate strongly afterwards [16].
Since slight difference in calcium phosphate cement
compositions may lead to a considerable change of
clinical result, further animal tests of the CPC and
P-CPCs in radial and tibial defects of rabbits were
performed. Their biocompatible, bioabsorbable and
osteoinductive properties were evaluated.

2. Materials and methods

2.1. Materials
DCPD, MCPM, CaO, Ca(OH)2, phosphorus pentoxide and methane sulphonic acid (analysis grade) were

purchased from Tianjin Chemical Co. Ltd., China.

Chitin was obtained from Tianjin Zhengtiancheng
Company, China.
2.2. Preparation of the P-chitosan reinforced cements
(P-CPCS)
P-CPCs were fabricated using two similar procedures
as mentioned by Boudeville [14] and Takagi [15]. Briey,
cement powders were prepared by mixing a calcium
phosphate (DCPD or MCPM) with a basic calcium
compound [Ca(OH)2 or CaO] in a stoichiometic ratio
according to Eq. (1) or Eq. (2) in an agate mortar. The
liquid phase was 1 m Na2HPO4 solution or 1 m
phosphate buffer (pH 7.4) of NaH2PO4 and Na2HPO4.
A cement powder was mixed with a liquid phase using
the powder-to-liquid (P/L) ratio (g/ml) of 0.96 or 2.29
separately. Different amount of P-chitosan (Mw
1:7
104 ; degree of deacetylation, DD: 63%; degree
of substitution, DS: 0.15) were added to the liquid
phase before the power phase was mixed with the liquid

4170

X. Wang et al. / Biomaterials 23 (2002) 41674176

phase.
3CaH2 PO4 2  H2 OMCPM 7CaO
-Ca10 PO4 6 OH2 HA 8H2 O;

6CaHPO4  2H2 ODCPD 4CaOH2

-Ca10 PO4 6 OH2 HA 18H2 O:

For the prehardened cement system (P-CPC-II), formed

from DCPD/Ca(OH)2/1 m Na2HPO4/P-chitosan (Pchitosan concentration in the liquid phases was 0.02,
0.07 or 0.12 g/ml), the pastes were loaded into
3 mm
9 mm plastic mounds and stored in a 371C, 80
RH box for 24 h. The mechanical strength of zero,
low (0.02 g/ml), middle (0.07 g/ml) and high (0.12 g/ml)

P-chitosan content cylinders after setting 24 h was 2.63,

6.81, 8.78 and 9.05 Mpa, respectively. Ca/P ratios of
these samples were between 1.50 and 1.52. The settled
cylinders were sterilized under 60Co for 8 h before use.
The CPC pastes (P-CPC-I), composed of MCPM/
CaO/1 m phosphate buffer/P-chitosan (P-chitosan: 0.05
or 0.08 g/ml), were used to ll the tibial defects of rabbits
directly.
2.3. Operation of the animal tests
24 adult white rabbits (2.03.0 kg) were assigned to
one of four groups: the 1-, 4-, 12- and 22-week groups.
Each group consisted of six rabbits, and each rabbit was

Fig. 5. (a) BSE image of a P-CPC-II (P-chitosan: 0.02 g/ml) 12 weeks after operation; (b) EDX analysis of the grey region in Fig. 5a; (c) EDX
analysis of the trabeculae in Fig. 5a.

X. Wang et al. / Biomaterials 23 (2002) 41674176

4171

Fig. 5 (continued).

used for one of the cements (PCPC-I, CPC-II and

PCPC-II). The rabbits were shaved and disinfected in
areas over right radii or tibias after general anaesthesia
with a mixture of 15% ketamin hydrochloride and
xylazinehydrochloride (1 ml/kg). The radii or tibias were
exposed through skin incisions. Bone defects (9 mm in
length for radii; 3 mm in depth and diameter for tibias)
were surgically created by saw and drill. Twelve rabbits
were used for implanting the prehardened samples
(P-CPC-II). Six rabbits were used for lling the tibial
defects with P-chitosan containing cements (P-CPC-I) in
paste form. The other six rabbits were used for
implanting prehardened CPC-II cylinders in radii as
controls. The wounds were closed with silk threads.
2.4. Microscopic observation and histological study
The rabbits were sacriced with an air injection,
respectively, at 1, 4, 12 and 22 weeks postoperation.
X-ray radiographs of the rabbit radii and tibias with the
implanted materials were taken on Philips 1025 X-ray
Radiograph System (Japan). The materials were harvested with the surrounding tissue and xed in 10%
neutral buffered formalin. Half of the samples were
dehydrated in graded ethanol and embedded in polymethylmethacrylate (PMMA). Thin sections about
20 mm thickness were sawn by a diamond saw (QP301, China) and stained with Giemsa solution for
microscopic observation (Olympus X-70 microscope,
Japan). The other half were decalcied with 8% nitric

acid, dehydrated, embedded in parafn. These specimens were then sectioned (Leitz, Western German) to
7 mm thick and stained with Masson trichroism (MT)
or Haematoxylin-eosin (HE). Some remaining blocks
were analyzed with back-scattered scanning electron
microscopy (BSE) and energy dispersive X-ray spectrometry (EDX) (Philips, XL-30 W/TMP, Japan).

3. Results
3.1. Macroscopic and radiograph observation
After the materials were implanted in the rabbit
defects, all the wounds healed gradually and the rabbits
were active with no postsurgery complications. No
osteolysis, hyperplasia or other negative tissue responses
were found in the CPC and P-CPC containing samples
through the study of 22 weeks. Radiographic evidence in
the healing areas (Fig. 1) demonstrated that the cements
both in prehardened and paste forms could combine
with the around tissue tightly during the whole process.
After implanted for 1, 4, 12 and 22 weeks, X-ray
radiographs showed that the P-chitosan containing
samples and the controls permitted stablization of the
defected bones and maintenance of correct position. The
new bone formation and the implant biodegradation in
the bone defects could be clearly reected by the
gradually weakened umbral on the radiographs. During

4172

X. Wang et al. / Biomaterials 23 (2002) 41674176

the whole process, bone repair in Controls was not as

perfect as those in P-CPC samples (Fig. 1).
3.2. Histological observation
3.2.1. After 1 week
According to the observation, P-CPCs (both in
prehardened and paste forms) existed as a block buried
in the radii and tibias. Very little haematoma appeared
around the specimens. Bone formation from the cement
could not be identied (Fig. 2). Tissues around the

implants were normal. In contrast, CPC-II was

surrounded by a great deal of haematoma and a thin
granulation tissue that contained many macrophages
and foreign body giant cells. Moderate inammatory
was observed.
3.2.2. After 4 weeks
From the sections retrieved at the end of 4 weeks,
newly formed woven bone clearly appeared in all the
P-CPCs containing rabbits. High-quality bone bonding
between the implants and host bone were observed

Fig. 6. (a) BSE image of a P-CPC-II sample (P-chitosan: 0.12 g/ml), 12 weeks after operation; (b) EDX analysis of the grey layer in Fig. 6a; (c) EDX
analysis of the trabeculae in Fig. 6a.

X. Wang et al. / Biomaterials 23 (2002) 41674176

4173

Fig. 6 (continued).

(Fig. 3). The voids between terminals of bone defects

and implants, even the marrow cavities nearby, were
lled with newly formed trabeculae with active osteoblasts. Direct contact between trabeculae and materials
occurred. There was no evident random dissolution at
the edges of the implants. No macrophages were found
around the implants. In control, the CPC-II cylinder
was changed into osteoid tissue on its peripheral
portion. Particles of the CPC-I were phagocytosed by
the surrounding macrophages. It was expected that at
this early termination, biological activity was new bone
integration rather than remodelling.
3.2.3. After 12 weeks
Tissue responses to P-CPC-I paste in the tibias and
prehardened P-CPC-II in the radii were similar. A part
of the P-CPC-II implants remained at the end of 12
weeks. But the sizes of the remaining blocks are quite
different. The more P-chitosan contained in a P-CPC,
the more material was left. In the low P-chitosan content
(P-chitosan: 0.02 g/ml) sample, nearly one-fourth of the
implant could be detected on the X-ray radiograph
(Fig. 1e) as well as on the histological section. There was
clear evidence of remodelling around the implant
surface (Fig. 4a). In all regions of inspection, excellent
bone bonding between the host bone and implant was
displayed. On the outside of the remained material there
was a thin and grey body uid inltrated layer. EDX
analysis shows that Ca/P (At) ratio in the grey layer was

1.75 (Figs. 5a and b), while in the trabeculae regions was

1.44 (Figs. 5a and c). The degradation rate of the
implant with high P-chitosan content (P-chitosan:
0.12 g/ml) was much slower than that with low
P-chitosan content. About three-fourths of the implant
with high P-chitosan content had not been degraded.
Trabeculae formed after the implant was gobbled up
(inltrated) by body uid. On the side of the cement,
trabeculae had encapsulated the implant body. Ca/P
(At) ratio in the body uid inltrated grey layer was 1.66
(Figs. 6a and b), in the black trabecula regions was
1.01(Figs. 6a and c). Sulphur (S) was found in the black
trabecula regions in signicant amount. Interestingly, in
the P-CPC-I paste sample, cement already disappeared
and the trabeculae formed only where once was lled
with the cement (Fig. 4b). In the two controls, about
one-fth of the implants were left (Figs. 1f and 7a).
Neither connective tissue nor body uid inltrated grey
layer could be detected. CPC-II was directly changed
into trabeculae after it was inltrated by body uid
(Fig. 7a). Ca/P (At) ratio in the centre part of the cement
body was 1.28 (Figs. 7a and b) and in the around
trabeculae region was 0.46 (Figs. 7a and c). Sulphur (S),
Silicium (Si) and Kalium (K) were found in the
trabecular regions (Figs. 47).
3.2.4. After 22 weeks
The implants with low and middle P-chitosan content
disappeared and bone remodelling almost nished at the

4174

X. Wang et al. / Biomaterials 23 (2002) 41674176

end of 22 weeks. Half of the implant with high

P-chitosan content was left (Fig. 1g). The newly formed
trabeculae in the high P-chitosan containing sample
were much denser than those in the low P-chitosan
containing sample (Figs. 8 and 4a). While in the control,
CPC-II disappeared and bone remodelling nished. The
edge of the repaired radius was blurred (Fig. 1f).

4. Discussion
As indicated by Ripamonti and Duneas [17], an ideal
biomaterial for bone tissue engineering should be
non-immunogenic, biodegradable, highly effective in

osteoinduction with relatively low doses of inducing

signals, ready for rapid vascular and mesenchymal cell
invasion, carvable, and amenable to contouring for
optimal adaptation to the various shapes of bone
defects, providing mechanical support when needed.
The P-CPCs fulll all of these requirements. Histological examination has revealed that bone repair in the
P-CPCs in a different manner with CPC, in which
microphages appear around implants. Much less haematoma appeared in the P-CPCs specimens than those
in the CPC at the end of 1 week which indicats that
P-chitosan has the property of hemostasis. The osseoinductive properties of the P-CPCs can be clearly reected
by Fig. 4b on which new trabeculae formed only where

Fig. 7. (a) BSE image of a CPC-II 12 weeks after operation; (b) EDX analysis of the grey region in Fig. 7a; (c) EDX analysis of the trabeculae in
Fig. 7a.

X. Wang et al. / Biomaterials 23 (2002) 41674176

4175

Fig. 7 (continued).

Fig. 8. Tissueimplant interface of a P-CPC- II specimen (P-chitosan:

0.12 g/ml) 22 weeks after implantation(decalcied, magnication
40,
MT stain).

once was lled with P-CPC-I. The results that a

complete osseointegration of bone and cements, new
bone progressively forms on the surface of implants and
incorporates with the host bone show that P-CPCs have
excellent biocampatibility. Under BSE, a seam of body
uid inltrated grey layer on the front of mineral
trabeculae. EDX analysis shows that Ca/P ratios in the
grey layers are very different from the mineral regions.
The change of the Ca/P ratios and the appearance of S
and other elements in the mineral regions mean that the
implants are changing into physiological calcium

phosphate tissues. Furthermore, the biodegradation rate

of the implants depends on the P-chitosan content. The
more P-chitosan contained, the slower degradation
rate will be. Most of the cements in low and middle
P-chitosan samples can be bioabsorbed in 16 weeks.
Trabeculae form gradually as the cements disappear,
and bone remodelling can nish nally. Thus, for the
rapid repair of no loading bone, the low P-chitosan
samples should be chosen rather than the high
P-chitosan samples. High P-chitosan samples may have
special use in high-loading bone-defect repair.
Histopathological examinations have proven that
P-chitosan containing samples are safety in vivo. There
was no bre along with the repair processes. Since bre
is the main reason of inammation and total joint
loosening [18] in calcium phosphate cement orthotics,
the prehardened P-CPCs will be especially useful in large
skeletal defect repairs.
In general, P-CPCs are biocompatible, osteoinductive
and bioresorbable. They can be nally replaced by new
bone in a creeping substitute manner. P-CPCs are
promising candidates for bone repair.

Acknowledgements
The project is supported by the funds of the National
Natural Science Foundation of China and Ministry of
Education.

4176

X. Wang et al. / Biomaterials 23 (2002) 41674176

References
[2] Damien JC, Parson JR. Bone graft and bone graft substitutes: a
review of current technology and applications. J Appl Biomater
1991;2:187208.
[3] Binderman I, Fin N. Bone substitutes-organic, inorganic, and
polymeric: cell material interactions. In: Yamamuro I, Hench L,
Wilson J, editors. CPC Handbook of Bioactive Ceramics. Boca
Raton, FL: CPC Press, 1990. p. 4551.
[4] Langstaff S, Sayer M, Smith TJN, Pugh SM. Resorbable
bioceramics based on stabilized calcium phosphates. Part II:
evalution of biological response. Biomaterials 2001;22:13550.
[5] Costantino PD, Friedman CD, Lane A. Synthetic biomaterials in
facial plastic and reconstructive surgery. Facial Plast Surg
1993;9:115.
[6] Kobayashi S, Hara H, Okudera H, Takemae T, Sugita K.
Usefulness of ceramic implants in neurosurgery. Neurosurgery
1987;21:7515.
[7] Wolfe SA. Correction of a lower eyelid deformity caused by
multiple extrusion of alloplastic orbital oor implants. Plast
Reconstr Surg 1981;68:42933.
[8] Munting E, Mirtchi AA, Lemaitre J. Bone repair of defects lled
with phosphocalcic hydraulic cement: an in vivo study. J Mater
Sci Med 1993;4:33744.
[9] Kent JN, Quinn Jh, Zide MF, Guerra LR, Boyne PJ. Alveolar
ridge augmentation using non-resorbable hydroxylapatite with or
without autogenous cancellous bone. J Oral Maxillofac Surg
1987;41:62942.

[10] Yoshimine Y, Sumi M, Isobe R, Anan H, Maeda K. In vitro

interaction between tetracalcium phosphate-based cement and
calvarial osteogenic cells. Biomaterials 1996;23:22413.
[11] Patrick F, Laurent G, Philippe C, Jacques F, Nicole R. Shortterm implantation effects of a DCPC-based calcium phosphate
cement. Biomaterials 1998;19:9717.
[12] Ishikawa K, Miyamoto Y, Kon M, Nagayama M, Asaoka K.
Non-decay type fast-setting calcium phosphate cement: composite
with sodium alginate. Biomaterials 1995;16:52732.
[13] Cherng A, Takagi S, Chow LC. Effects of hydroxypropyl
methylcellulose and other gelling agents on the handling properties of calcium phosphate cement. J Biomed Mater Res
1997;35:2737.
[14] Boudeville P, Serraj S, Leloup JM, Margerit J, Pauvert B, Terol
A. Physical properties and self-setting mechanism of calcium
phosphate cements from calcium bisdihydrogenophosphate
monohydrate and calcium oxide. J Mater Sci: Mater Med
1999;10:99109.
[15] Takagi S, Chow LC, Ishikawa K. Formation of hydroxyapatite in
new calcium phosphate cements. Biomaterials 1998;19:15939.
[16] Wang X, Ma J, Wang Y, He B. Structural characterization of
phosphorylated chitosan and their applications as effective
additives of calcium phosphate cements. Biomaterials
2001;22:224755.
[17] Ripamonti U, Duneas N. Tissue engineering of bone by osteoinductive biomaterials. MRS Bull 1996;21:369.
[18] Thomas WB, Schils J. The pathology of total joint arthoplasty II.
Mechanisms of implant failure. Skeletal Radiol 1999;28:48397.