Experimental Animal Models for Tissue-Engineered Bone Regeneration - Engineered Bone: Tissue Engineering Intelligence Unit - page 97

Figure 1. Distal femoral filling defects in rabbit. a) filled with coral alone; b) filled with a mixture of fibrin
sealant and coral.
Rabbit ulna is one of the most currently used segmental long bone resection models. 41,42
The surgical procedure can be performed bilaterally, 43 is easy to perform and internal fixation
is not theorically required as ulnar stability is afforded by radio-ulnar synostosis. The ulna is a
weight bearing bone which allows implant loading and morbidity is low. Several limitations
must nevertheless be pointed out: (i) spontaneous bone healing occurs in 30% of animals after
10 mm long resections and some authors have observed bone healing with 20 mm long ostec-
tomies making this model more a model for delayed bone healing than a CSD, 44 (ii) bone
curvature and synostosis makes implant positioning and stabilisation challenging for massive
materials, (iii) individual variations in bone curvatures impair reproducibility of bone sam-
pling for histologic analysis and biomechanical assays (Fig. 2). Radial resections have been
performed in rabbits and carry the same disadvantages. 45 Tibial resections stabilised by un-
locked intramedullary nailing have been associated with rotational instability and high mor-
bidity which preclude their use. 46
A highly reproducible CSD model has been developed in the cat tibia. 9 Femoral, 42,47,48
tibial, 49 radial 40 or ulnar 50 resections have been performed in dogs and were either stabilised by
bone plating 47,48 or external fixation. 40,49 Some of these models have shown spontaneous bone
healing and are not thus CSD. 40 25 mm long ulnar resections have been performed in dogs and
carry the same advantages and disadvantages as in the rabbit. Radial synostosis is nevertheless
more localised than in the rabbit which simplifies implant positioning and stabilisation.
Femoral resections stabilised by bone plates and tibial resections either stabilised by bone
plates 51 or an intramedullary interlocking nail 52 have been performed in sheep. The recently
described metatarsal resection 53 carries many advantages: (i) it is a true CSD as fibrous non-
union occurs for 25 mm long bone defects after 4 months observation, (ii) it is performed in a
straight bone allowing reproducible histomorphometric and biomechanical assays, (iii) it has a
low morbidity (Fig. 3).
Post-operative handling of pigs is often difficult making orthopaedic surgery challenging. A
radial defect has been described in the miniature pig with minimal morbidity. 54 Femoral de-
fects stabilised either by bone plate or external fixation have been proposed by Sendowski in
Pitmann Moore minipigs and Large White pigs. Morbidity with these models has been high:
external fixators have failed in both sized animals and bone plates have been satisfactory in
minipigs only. 55
Femoral, tibial, ulnar and humeral resections have been performed in NHP. 56-58 Although
developed in a species close to man as far as bone healing and biomechanical loading are con-
cerned, the experimental designs suffer from major drawbacks: (i) lack of negative controls 56-58