Abstract
Animal fracture models have been extensively applied to preclinical
research as a platform to identify and characterize normal and abnormal
physiological processes and to develop specific maneuvers that alter the
biology and biomechanics being examined. The choice of animal model employed
in a study bears a direct relationship to the specific intervention being
analyzed. The animal models employed should be described clearly,
control-group data should be established, and reproducibility should be
defined from experiment to experiment and from institution to institution so
that quantitative and qualitative outcomes can be reliably compared and
contrasted to other related studies.
Animal fracture models have been extensively applied to preclinical
research as a platform to identify and characterize normal and abnormal
physiological processes and to develop specific maneuvers that alter the
biology and biomechanics being examined. By facilitating the translation of in
vitro discovery into in vivo testing, animal fracture models allow researchers
to establish strategies to further understand the physiology of bone-healing
and to improve the rate, speed, and quality of fracture-healing.
The choice of animal model employed in a study bears a direct relationship
to the specific intervention being analyzed. Thus, to allow accurate and
reliable interpretation of preclinical animal studies, it is important that
animal fracture models are well characterized. The animal models employed
should be described clearly, control group data should be established, and
reproducibility should be defined from experiment to experiment and from
institution to institution so that quantitative and qualitative outcomes can
be reliably compared and contrasted to other related studies.
The typical progression of animal experimentation, depending on the nature
of the study, is an evolution from small animals (mice, rats, rabbits) to
higher phylogenetic species (dogs, sheep, pigs, goats) in an effort to mimic
the human conditions as closely as possible. This path typically culminates in
nonhuman primates and ultimately clinical studies.
The aim of this review is to discuss the different types of animal models
used in fracture-healing research in an attempt to characterize them in terms
of their applicability for particular methods of analysis and provide guidance
for selection of animal models. To do so, we have reviewed all of the articles
over the past ten years in a number of prominent orthopaedic and
musculoskeletal journals (Bone, Clinical Orthopaedics and Related
Research, The Journal of Bone and Joint Surgery, the Journal of Bone
and Mineral Research, the Journal of Orthopaedic Research, and
the Journal of Orthopaedic Trauma) and have identified 291 studies
that utilized animal fracture models. This review describes and discusses the
animal model data employed in these 291 studies and in a number of other
seminal articles that are directly referenced in the article. Reference
details for all of these articles can be found in a supplementary reference
list and corresponding table in the Appendix that guides the reader toward
various types of animal fracture studies.
Factors Affecting Fracture-Healing
The typical healing pattern in bone is a path toward either osteoblastic or
chondroblastic lineage with a resultant osseous repair. The osteoblastic
lineage leads to direct new bone formation (analogous to intramembranous
ossification in the developing animal), and the chondroblastic pathway forms
bone via endochondral ossification. Histological studies in human
fracture-healing have shown that ossification is typically manifest before
cartilage is present and that both processes coexist in normal bone
repair1. A similar
healing pattern exists in higher animal species (e.g., canine) along with the
involvement of periosteal and medullary
calluses2. However,
in the example of bone-healing in rodents (e.g., rats, mice), although the two
processes of bone formation are demonstrable, endochondral bone formation
predominates3-5.
Fracture repair is affected by the level of injury, the specific site, the
fracture gap, and by any necessary fixation. Local biology, including the
condition of the vasculature available at the site, and specific patient
factors, such as age, gender, drugs, and diseases, all affect fracture repair.
Additional integral factors in successful fracture repair include viability of
local tissue, vasculature, and access to stem cells. Rhinelander et al.
demonstrated through vascular studies that there is a small vascular pattern
in uninjured bone but that this vascular pattern rapidly increased in size at
the time of
fracture2.
Boström et al. demonstrated that the newly generated vasculature produced
large amounts of bone morphogenetic protein (BMP) with invading pericyte-like
cells6. Brighton et
al. demonstrated that the oxygen tension remained low well into the healing
stages despite
vasculature7-9.
While the vascular process provides a source of nutrients and stem cells in
the form of pericytes, the blood flow itself has only a secondary role.
In a thorough and seminal review, McKibbin assessed fracture-gap bridging
speed, rigidity, tolerance of movement, and the importance of soft tissue at
various healing
stages10. The
primary cortical periosteal response is very rapid, tolerates movement
successfully, and is unrelated to the soft-tissue response. The external
bridging callus is slightly slower, highly dependent on the soft-tissue
environment, and dramatically slowed by immobilization. Thus, it favors
motion. The intramedullary bridging callus is relatively slow, tolerates
motion minimally, and prefers a stable environment. The primary osteon is
exquisitely slow, does not bridge gaps, and will not tolerate motion,
preferring complete immobilization.
Each animal model is best suited to answer a particular question. No
predictable correction exists for size, weight, or surface area across
different species. A much larger BMP-2 dose was necessary in monkeys than
would be suggested by merely correcting established canine dosages for surface
area and weight11.
Due to the species-specific and unpredictable nature of dosage, dose-finding
studies may need to be performed directly in humans after animal research. In
attempts to emulate human conditions, research commonly progresses
incrementally from smaller to larger animals.
Bone-healing rate and speed varies among
animals3-5,12,13.
In the rat, normal fracture-healing takes four to five weeks and is
accomplished almost exclusively by the external periosteal callus, with
minimal medullary callus
involvement14.
Rabbits, dogs, and nonhuman primates heal with greater involvement of the
medullary callus, and thus the healing takes longer.
Age affects fracture repair rate and biology. Meyer and Meyer extensively
studied age-associated genetic changes as they pertain to fracture-healing and
found that the delayed fracture-healing in old rats is associated with reduced
messenger ribonucleic acid expression of genes forming the mitochondrial
energy pathways15.
Many other specific pathways are affected by age. This is an area of intense
investigation.
In the initial design of any animal fracture model, appropriate power
studies are imperative. With animal studies, hesitancy to calculate the
appropriate number is common. Consequently, inadequate sample size leads to
trends rather than true observations. In adjuvant testing, which can be
directed at modes of administration, single doses, multiple dosing, delayed
dosing, local versus systemic administration, and duration, the carrier may be
as critical as the material being tested.
Rodents and Smaller Animals
Small rodent models, such as mice and rats, allow manipulation of molecular
biology, incorporation of a greater number of animals, and potentially quicker
healing rates. Furthermore, mice permit specific and conditional knockout
studies. Athymic mice and rats can accept human cells, factors, and tissues.
Inbred mice and rats are preferred in non-time-constrained tissue maneuvers.
Mouse and rat calvarial defects have been utilized extensively for
bone-healing and are well suited for specialized observation (e.g., bone
windows, continuous monitoring, and tissue
sampling)16-24.
Rodent critical-size defect models, currently a popular theme, allow
experimentation with growth factors, local drugs, and small peptides. The
principal limitation with this model is the resultant membranous repair, which
correlates poorly with the endochondral healing process. In the metaphysis and
diaphysis, critical-size defects will not heal. They are often unloaded and
weakly correlate to fracture repair. They can be evaluated with use of
radiography, microquantitative computed tomography, histology, and various
biomechanical tests. They have been used to test biologics (growth factors,
small peptides), synthetics (allografts, ceramics), and osteogenic factors.
One critical-size defect model, which has been used in studies such as the
evaluation of BMP-2, involves a 5-mm defect in the rat femur stabilized by two
pins proximally and distally through a polyethylene plate that will not
obscure
radiographs25.
Rodents can be manipulated and various diseases can be engineered to create
challenges to bone-healing. Examples include chemically induced oophorectomy
(which produces osteoporosis within three months), chemically induced
diabetes, nicotine exposure, smoking, chemotherapy,
cyclooxygenase-2-inhibiting medication, knockout animals, and
osteomyelitis.
Mouse
Murine models were used in 15% of the reviewed studies. Mice allow for the
most sophisticated genetic manipulation, such as specific and conditional
"knockout" studies. Nude mice are ideal when an immunologic
response to implanted graft material is a concern. Their small size, however,
makes surgical procedures more challenging and renders assessment of fixation
mode and implant assessment often impractical. Although many of the techniques
in the rat have been converted to the mouse, it is generally more feasible to
perform such analyses in the rat.
Rat
Of the reviewed studies, 38% used a rat as their animal model, making it
the most commonly used animal (Fig.
1). Athymic rats allow for implant studies. In comparison to mice,
this model is typically more predictable, given its larger anatomy and
practicality when performing both surgery and analyses such as biomechanical
testing. Although biomechanical testing is possible in the mouse, it can be
difficult and the standard deviation of the results can be quite large. On the
other hand, while current gene-modification and knockout tools for the mouse
have been well established and are extremely versatile and
flexible26, they
involve the use of special pluripotent embryonic stem cells that have not been
identified as of yet in the rat. Although there have been other methods
described that aim to achieve similar
results27, knockout
rats, to the best of our knowledge, are not readily available or in widespread
use to date.
Rabbit
The lapine model was used in 19% of the reviewed articles. The New Zealand
White rabbit is the most commonly used rabbit model in fracture-healing
studies. Rabbits have been used in many biomechanical and cartilage studies,
given their larger joints compared with rodents. Most lapine fracture studies
utilized an open fracture model.
Higher-Order Animals
Higher-order mammals and nonhuman primates mimic the human more effectively
than rodents do, especially in terms of fixation methods and biomechanics. The
relatively larger size facilitates surgery and biomechanical analysis
(Fig. 2). They have answered
mechanical, osteoconductive, and osteoinductive questions. They are used to
confirm human dose determination, but not as a screening tool. Biomechanical
testing was far more common among larger animals than in smaller animals.
Dog
Canine models were used in 9% of the reviewed studies. Depending on the
breed of dog used, they are easy to handle. Their use can vary worldwide
depending on the societal acceptance of their role in research. Open fractures
were generated in the majority of canine fracture studies. Depending on the
breed of dog used, dogs are easy to handle. Of note, their use is often
limited by lack of societal acceptance of their place in research.
Sheep and Goat
Ovine and caprine models were used in 11% and 4% of the studies,
respectively. In the case of the ovine model, all were used only in open
fracture studies. A disadvantage may be higher cost compared with smaller
animal models. Given that both of these animals are ruminants, there are
special considerations to bear in mind when handling and with regard to
housing, given their size. Special attention should also be paid to anesthesia
and positioning in ventral recumbency. Regurgitation of ruminal contents while
under general anesthesia (regardless of positioning) should be avoided.
Withholding all solid food for a minimum of twenty-four hours is necessary in
all small-ruminant anesthesia.
Closed Fracture Compared with Open Fracture
(Fig. 3)
Closed injury has less iatrogenic alteration and comminution. Bonnarens and
Einhorn developed a guillotine-induced closed model that makes endochondral
ossification quite reproducible and that has been widely used in the rat femur
and subsequently downsized to the
mouse28. This model
has been rigorously defined and characterized broadly in terms of histology,
biomechanics, radiology, and molecular patterns. In contrast, open fracture
with an osteotomy causes local tissue damage with periosteal stimulation. It
forms a bone within a bone, is not consistently reproducible, has a much lower
rate of success in union, and provides a greater window of opportunity for
augmentation of fracture-healing.
Fixation (Fig.
4)
The mode of fracture fixation affects the repair process. Intramedullary
fixation includes a pin (rigid or flexible), nail, wire, or plate. The
fixation device can be metal or polyethylene. A fixator can cause distraction
or variable stability. Fixation is not necessary in radial or ulnar defects.
These have been extensively used and each one has a different mechanical
environment, providing either an excellent or a deleterious environment for
repair.
Gap
A gap delays healing and is desired for testing adjuvants. In the absence
of a gap, with adequate contact and apposition, the fracture heals too rapidly
to assess the influence of growth factors.
Gene Therapy
Gene therapy has recently entered the arena of fracture-healing, and
studies involving in vivo versus ex vivo modes of transfection, duration, and
gene markers have evolved as testing modalities. These methods have often been
used in defect models, but they require a carrier for presentation.
Preclinical fracture-healing models play a very important role in
understanding the physiology of bone-healing and provide an efficient and
effective means to assess various factors and their effects on the normal
healing process. It is clear that there is a pattern of animal selection
corresponding to the question being posed. The mouse is increasingly being
used in studies involving genetic manipulations. The rat model is both
effective and robust. Rabbits and the larger animals, such as dogs and sheep,
allow for more feasible biomechanical studies. With the progression to higher
phylogenetic species, the primate offers the last step before a clinical
trial.
In the current era of advanced preclinical research into fracture-healing,
there is a large array of sophisticated animal models available. By choosing
an established model that has been well characterized in the literature, one
can focus on the specific biological question being studied. Alternatively,
one can attempt to develop a new, more specific animal model. However, this
will require extensive characterization and testing prior to application of
the model to the question at hand. Whether the model is established or novel,
the ability of the animal model selected to reliably and accurately represent
the biological process being studied is paramount to the integrity of the
study.
A table showing the various categories of animal fracture studies and the
supplemental reading list are available with the electronic versions of this
article, on our web site at
(go to
the article citation and click on "Supplementary Material") and on
our quarterly CD-ROM (call our subscription department, at 781-449-9780, to
order the CD-ROM). ?
Sevitt S. Bone repair and
fracture healing in man. Edinburgh: Churchill Livingstone;
1981. p 35-51,64-107.35
1981
Rhinelander FW, Phillips RS, Steel WM,
Beer JC. Microangiography in bone healing. II. Displaced closed fractures.
J Bone Joint Surg Am.
1968;50:
643-62.50643
1968
[PubMed]
Bostrom MP, Lane JM, Berberian WS,
Missri AA, Tomin E, Weiland A, Doty SB, Glaser D, Rosen VM. Immunolocalization
and expression of bone morphogenetic proteins 2 and 4 in fracture healing.
J Orthop Res. 1995;13:
357-67.13357
1995
[PubMed][CrossRef]
Chakkalakal DA, Strates BS, Mashoof AA,
Garvin KL, Novak JR, Fritz ED, Mollner TJ, McGuire MH. Repair of segmental
bone defects in the rat: an experimental model of human fracture healing.
Bone. 1999;25:
321-32.25321
1999
[PubMed][CrossRef]
Kawaguchi H, Kurokawa T, Hanada K,
Hiyama Y, Tamura M, Ogata E, Matsumoto T. Stimulation of fracture repair by
recombinant human basic fibroblast growth factor in normal and
streptozotocin-diabetic rats. Endocrinology.
1994;135:
774-81.135774
1994
[PubMed][CrossRef]
Boström K, Watson KE, Horn S,
Wortham C, Herman IM, Demer LL. Bone morphogenetic protein expression in human
atherosclerotic lesions. J Clin Invest.
1993;91:
1800-9.911800
1993
[PubMed][CrossRef]
Brighton CT, Krebs AG. Oxygen tension of
nonunion of fractured femurs in the rabbit. Surg Gynecol
Obstet. 1972;135:
379-85.135379
1972
Brighton CT, Heppenstall RB. Oxygen
tension in zones of the epiphyseal plate, the metaphysis and diaphysis. An in
vitro and in vivo study in rats and rabbits. J Bone Joint Surg
Am. 1971;53:
719-28.53719
1971
Brighton CT, Krebs AG. Oxygen tension of
healing fractures in the rabbit. J Bone Joint Surg Am.
1972;54:
323-32.54323
1972
[PubMed]
McKibbin B. The biology of fracture
healing in long bones. J Bone Joint Surg Br.
1978;60:
150-62.60150
1978
[PubMed]
Martin GJ Jr, Boden SD, Marone MA,
Marone MA, Moskovitz PA. Posterolateral intertransverse process spinal
arthrodesis with rhBMP-2 in a nonhuman primate: important lessons learned
regarding dose, carrier, and safety. J Spinal Disord.
1999;12:
179-86.12179
1999
[PubMed]
Nakamura T, Hara Y, Tagawa M, Tamura M,
Yuge T, Fukuda H, Nigi H. Recombinant human basic fibroblast growth factor
accelerates fracture-healing by enhancing callus remodeling in experimental
dog tibial fracture. J Bone Miner Res.
1998;13:
942-9.13942
1998
[PubMed][CrossRef]
Shirley D, Marsh D, Jordan G, McQuaid S,
Li G. Systemic recruitment of osteoblastic cells in fracture healing. J
Orthop Res. 2005;23:
1013-21.231013
2005
[CrossRef]
Wray JB, Lynch CJ. The vascular response
to fracture of the tibia in the rat. J Bone Joint Surg Am.
1959;41:
1143-8.411143
1959
[PubMed]
Meyer MH, Meyer RA Jr. Genes with
greater up-regulation in the fracture callus of older rats with delayed
healing. J Orthop Res.
2007;25:
488-94.25488
2007
[PubMed][CrossRef]
Aalami OO, Nacamuli RP, Lenton KA, Cowan
CM, Fang TD, Fong KD, Shi YY, Song HM, Sahar DE, Longaker MT. Applications of
a mouse model of calvarial healing: differences in regenerative abilities of
juveniles and adults. Plast Reconstr Surg.
2004;114:
713-20.114713
2004
[PubMed][CrossRef]
Bosch C, Melsen B, Vargervik K.
Importance of the critical-size bone defect in testing bone-regenerating
materials. J Craniofac Surg.
1998;9:
310-6.9310
1998
[PubMed][CrossRef]
Zhang X, Kuroda S, Carpenter D,
Nishimura I, Soo C, Moats R, Iida K, Wisner E, Hu FY, Miao S, Beanes S, Dang
C, Vastardis H, Longaker M, Tanizawa K, Kanayama N, Saito N, Ting K.
Craniosynostosis in transgenic mice overexpressing Nell-1. J Clin
Invest. 2002;110:
861-70.110861
2002
Schmitz JP, Hollinger JO. The critical
size defect as an experimental model for craniomandibulofacial nonunions.
Clin Orthop Relat Res.
1986;205:
299-308.205299
1986
[PubMed]
Schmitz JP, Schwartz Z, Hollinger JO,
Boyan BD. Characterization of rat calvarial nonunion defects. Acta Anat
(Basel). 1990;138:
185-92.138185
1990
[CrossRef]
Perlyn CA, Morriss-Kay G, Darvann T,
Tenenbaum M, Ornitz DM. A model for the pharmacological treatment of Crouzon
syndrome. Neurosurgery.
2006;59:
210-5.59210
2006
[PubMed][CrossRef]
Kanou M, Ueno T, Kagawa T, Fujii T,
Sakata Y, Ishida N, Fukunaga J, Sugahara T. Osteogenic potential of primed
periosteum graft in the rat calvarial model. Ann Plast Surg.
2005;54:
71-8.5471
2005
[PubMed][CrossRef]
Hyun SJ, Han DK, Choi SH, Chai JK, Cho
KS, Kim CK, Kim CS. Effect of recombinant human bone morphogenetic protein-2,
-4, and -7 on bone formation in rat calvarial defects. J
Periodontol. 2005;76:
1667-74.761667
2005
[CrossRef]
Develioglu H, Saraydin SU,
Bolayir G, Dupoirieux L. Assessment of the effect of a biphasic ceramic on
bone response in a rat calvarial defect model. J Biomed Mater Res
A. 2006;77:
627-31.77627
2006
Lane JM, Yasko AW, Tomin E, Cole BJ,
Waller S, Browne M, Turek T, Gross J. Bone marrow and recombinant human bone
morphogenetic protein-2 in osseous repair. Clin Orthop Relat
Res. 1999;361:
216-27.361216
1999
[CrossRef]
Cox RD, Brown SD. Rodent models of
genetic disease. Curr Opin Genet Dev.
2003;13:
278-83.13278
2003
[PubMed][CrossRef]
Smits BM, Mudde JB, van de Belt J,
Verheul M, Olivier J, Homberg J, Guryev V, Cools AR, Ellenbroek BA, Plasterk
RH, Cuppen E. Generation of gene knockouts and mutant models in the laboratory
rat by ENU-driven target-selected mutagenesis. Pharmacogenet
Genomics. 2006;16:
159-69.16159
2006
Bonnarens F, Einhorn TA. Production of a
standard closed fracture in laboratory animal bone. J Orthop
Res. 1984;2:
97-101.297
1984
[CrossRef]