Keywords
locking plates - fracture - radius ulna
Introduction
Fractures of the radius and ulna are common in dogs, particularly in miniature and
toy breeds,[1]
[2]
[3]
[4] and many fixation methods have been described. Conservative treatment using rigid
bandages in toy breed dogs commonly results in complications including malunion or
nonunion in up to 80% of fractures.[5]
[6] External skeletal fixation has been successfully used to manage radius fractures
in small breed dogs, but specific postoperative management and frequent follow-up
examinations require owner compliance.[7]
[8]
[9] Bone plating is a popular method for fixation of radius and ulna fractures[10]
[11]
[12]
[13]; however, complications can occur in up to 54% of dogs that weigh less than 6 kg.[14]
[15] These complications include delayed union, nonunion, re-fracture after implant removal
and osteopenia due to stress protection.[16]
[17] Nonunion is an interruption of the fracture healing process, which necessitates
surgical intervention to allow normal healing. Nonunion fractures are characterized
by formation of fibrous or cartilaginous tissue between fragments. This is a serious
complication in small animal orthopaedics, particularly in the treatment of radius
and ulna fractures in toy breed dogs. Nonunion occurs in the radius and ulna in 60%
of cases, in the tibia in 25% and in the femur in 15%.[18]
[19]
[20] Several factors contribute to nonunion, including inappropriate surgical treatment,
instability of fracture sites, infection and poor blood supply. Therapeutic planning
for septic nonunion fractures must consider several factors, such as local blood supply,
mechanical stability and regenerative ability of the involved tissues. The surgical
treatment of nonunion fractures remains a therapeutic challenge for orthopaedic surgeons,
especially in the presence of infection, bone loss or both.[17]
One of the primary objectives in treating septic nonunion fractures is adequate debridement
to reduce bacterial load and to remove necrotic tissue and sequestra. The debrided
fracture must be re-stabilized with appropriate internal or external fixation, providing
adequate inter-fragmentary compression if bone apposition is possible.[18]
[19]
[20] Enhancement of fracture healing and bone defect filling are important steps that
can be accomplished with a variety of materials including autologous or allogeneic
cancellous bone grafts, demineralized bone matrix, artificial bone materials such
as β-tricalcium phosphate and growth factors such as bone morphogenetic proteins (BMPs).[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
Fractures and nonunion of the proximal radius occur infrequently but are particularly
challenging because of the anatomic constraints to surgical approach and limited bone
stock for implant fixation.[4] When the proximal fragment of the radius is too small to achieve adequate fixation,
treatment options include use of an interlocking nail or a bone plate on the ulna
only, or the use of an external skeletal fixator with fixation pins applied to the
olecranon and to the distal radius fragment.[4]
[31]
[32] The purpose of this report was to describe the use of a locking plate fixed to the
proximal ulna and distal radius for treatment of nonunion of proximal radius and ulna
fractures in a toy breed dog.
Case Report
Clinical History
A 14-month-old, 4.5 kg, Maltese cross-breed dog was referred because of nonunion of
a fracture of the proximal radius and ulna of the left limb. The dog had sustained
a mildly comminuted fracture of the left proximal radius and ulna at 12 months of
age after being bitten by a dog ([Fig. 1A]). The fracture had been initially treated with an external skeletal fixator. ([Fig. 1B]). Three weeks later, surgical revision at the same veterinary practice was done
to remove the external skeletal fixator and place intramedullary pins in the radius
and ulna ([Fig. 1C]). Those implants were subsequently removed 1 month later because of implant failure,
and the limb was then splinted ([Fig. 1D]).
Fig. 1 Radiographs obtained at the primary veterinary care clinic. (A) Preoperative images show a mildly comminuted fracture of the left radius and ulna.
(B) Postoperative images after application of an external skeletal fixator. (C) Postoperative images after application of intramedullary pins. (D) Radiographs at the time of failure of intramedullary pinning showing bending of
the radial pin and invasion of the elbow and carpal joints.
The dog was admitted to our clinic 2 months after the initial injury with non-weight-bearing
lameness of the left forelimb. Mediolateral and craniocaudal radiographic views revealed
nonunion of the proximal radius and ulna fractures with periosteal reaction on the
distal fragment of the radius and a suspected bone sequestrum. The proximal fragment
of the radius was 4 × 5 mm ([Fig. 2A]).
Fig. 2 (A) Radiographic evaluation performed at the time of presentation showing nonunion of
radioulnar fractures, the presence of a sequestrum (red arrow) and periosteal reaction
on the distal radial fragment. (B) Image taken after removal of the splinted bandage showing skin lesions on the caudal
and lateral aspects of the antebrachium.
Clinical evaluation of the leg after splint removal revealed skin lesions, severe
instability of the fracture site and generalized muscle wasting of the limb ([Fig. 2B]). The dog was treated with a broad-spectrum antibiotic medication (amoxicillin + clavulanic
acid, Synulox: Pfizer, Rome, Italy, 20 mg/kg orally three times a day [TID]) without
performing a culture and sensitivity test, the skin lesions were medicated daily and
the limb was protected with a modified Robert-Jones bandage. Ten days after referral,
the skin lesions had completely healed, and radiographic evaluation showed remodelling
of the periosteal reaction. The results of a complete blood cell count, serum biochemical
profile and urinalysis were unremarkable. Surgical revision was then performed.
Revision Surgery
The dog was premedicated with fentanyl (Fentanest: Actavis Italy S.p.A, Nerviano,
Italy, 4 µg/kg, intramuscularly [IM]), morphine (Morfina Cloridrato: Molteni, Italy,
0.15 mg/kg, IM) and acepromazine (Prequilan, Fatro SpA, Ozzano Emilia [BO], Italy,
0.02 mg/kg IM). Anaesthesia was induced with propofol (2–4 mg/kg intravenously [IV])
and maintained with a mixture of isoflurane and oxygen (IsoFlo: Aesica Queenborough
Limited, Kent, United Kingdom) after endotracheal intubation. A constant rate infusion
of fentanyl (Fentanest: Actavis Italy S.p.A, Nerviano, Italy, 10 mcg/kg/h) provided
analgesia, and cefazolin (Cefazolina Dorom: Teva Pharma Italia, Milano, Italy, 20 mg/kg,
IV) was administered 1 hour before surgery and repeated 120 minutes later. The anaesthetized
dog was positioned in dorsal recumbency, and the limb was aseptically prepared in
a hanging position.
A standard craniomedial approach at the level of the fracture site was used to remove
the sequestrum and to debride the fracture fragments ([Fig. 3A]).[33] The proximal radius fragment was too small and friable to be used for fixation.
The fracture site was swabbed for bacterial culture and susceptibility testing. A
second incision extending from the proximal olecranon to the ulnar styloid was then
made on the lateral side. After caudal retraction of the flexor carpi ulnaris muscle
and cranial retraction of the ulnaris lateralis muscle, the fracture was manually
realigned, but no attempt was made to reduce the radial head. An 11-hole titanium
advanced locking plate (ALPS 6.5, Kyon, Zurich, Switzerland) was applied in bridging
fashion to the proximal ulnar fragment with three 2.4-mm locking screws in the three
most proximal plate holes. A fourth screw was inserted in the mid-shaft of the distal
ulnar fragment. Three 2.4-mm locking screws were inserted in the distal most holes
of the plate through the distal ulna to engage the distal radial fragment ([Fig. 3B]). A recombinant bone morphogenetic protein‐2 (rhBMP‐2) graft was prepared according
to the manufacturer's instructions by applying 1.4 mL of solution to an absorbable
collagen sponge (TruScient; Zoetis Inc., Madison, New Jersey, United States), which
was left to soak for 15 minutes before application. Two-thirds of the prepared sponge
were inserted into the radius and ulna fracture sites ([Fig. 3C]). The tissue layers were routinely closed, and a modified Robert Jones bandage was
applied to the limb.
Fig. 3 Intraoperative images. (A) Removal of the sequestrum via a craniomedial approach. (B) Application of the twisted bone plate on the lateral surface of the ulna. (C) Application of the recombinant bone morphogenetic protein‐2 graft delivered in an
absorbable collagen sponge (yellow arrow). Note the tip of two screws protruding from
the medial cortex of the distal radius, confirming proper orientation of the screws
through the distal ulna (green arrows).
Postoperative radiographic views showed appropriate implant positioning and satisfactory
alignment of the radius and ulnar fracture segments, with the exception of a mild
recurvatum ([Fig. 4A]). A 3-mm gap was evident between the proximal and distal fracture fragments. The
overall length of the radius from the radial head to the radiocarpal joint was 6.45 cm,
compared with 7.5 cm in the contralateral limb ([Fig. 4B] and [C]).
Fig. 4 Immediate postoperative radiographs showing proper implant positioning and alignment
(A). Fourteen percent length discrepancy is evident between the operated limb (B) and the contralateral limb (C).
The dog was hospitalized for 1 week after surgery. Amoxicillin with clavulanic acid
(Synulox: Pfizer, Rome, Italy, 20 mg/ kg orally TID) was administered for 5 days postoperatively
and was then discontinued because bacteriological culture results were negative. Meloxicam
(Metacam: Boehringer Ingelheim Vetmedica GmbH, Ingelheim, Germany, 0.1 mg/kg orally)
was administered once daily for 1 week and then on alternate days for another 2 weeks.
The bandage was removed 2 days after surgery to assess wound healing, which proceeded
uneventfully, and then replaced. The dog was discharged 7 days after surgery with
instructions to restrict activity to in-house confinement and short leash walks for
4 weeks, after which time the amount of exercise was gradually increased. Wound healing
was uneventful and the skin sutures and bandage were removed 12 days postoperatively.
Re-evaluation 4 weeks after surgery revealed persistent moderate weight-bearing lameness
of the left forelimb. The range of motion of the elbow was normal, but the carpal
range of motion was moderately reduced. Mediolateral and craniocaudal radiographic
views ([Fig. 5A]) showed progression of bone healing with interfragmentary callus formation that
was starting to bridge the fracture gap, but the fracture lines were still visible.
Fig. 5 Radiographs taken at postoperative re-evaluation. (A) Four weeks after surgery, bridging of the fracture gap is evident but fracture lines
are still visible. (B) Eight weeks after surgery, complete fracture healing is evident. Decreased bone
density at the level of the radial and ulnar diaphysis attributable to stress protection
is seen and implant dynamization was undertaken (C). Twenty weeks after surgery, improved bone density is evident (D). Radioulnar synostosis is visible after implant removal (E). Thirty weeks after surgery, fracture callus appears remodelled and degenerative
changes are evident at the radiocarpal joint.
Re-evaluation 8 weeks postoperatively revealed mild intermittent weight-bearing lameness.
The same radiographic views showed complete healing of the fracture ([Fig. 4B]). The radiopacity of the radial shaft, particularly the distal third, appeared decreased,
which was thought to be attributable to stress shielding. For this reason, dynamization
of the implant by removal of two screws in the distal radial fragment and another
screw in the distal ulnar fragment was done through separate stab incisions ([Fig. 4C]).
Re-evaluation 20 weeks after surgery revealed mild intermittent weight-bearing lameness
with moderate palmigrade stance. Radiographic evaluation showed remodelling of the
fracture callus, increased radiopacity of the radius and ulna and signs of degenerative
changes at the level of the radiocarpal joint ([Fig. 4D]). The dog underwent implant removal, and postoperative radiographs showed radioulnar
synostosis at the level of the proximal radius ([Fig. 4E]).
A final re-evaluation was done 36 weeks after surgery. The dog had mild intermittent
lameness and a moderate palmigrade stance. The elbow range of motion was normal and
manipulation elicited no pain, but the range of motion of the radiocarpal joint was
decreased during flexion and increased during extension. Radiographs showed progression
of the degenerative changes at the radiocarpal joint ([Fig. 4F]).
Discussion
A serious complication of fracture treatment is nonunion, which has a higher rate
in radioulnar fractures in toy dog breeds because of mechanical and biological factors.[34] In our case, initial osteosynthesis using external fixation was inadequate and intramedullary
pinning in a subsequent revision surgery led to development of a nonviable necrotic
nonunion. Guidelines for the treatment of this type of nonunion include removal of
necrotic bone fragments, debridement of fracture edges until bleeding from the periosteum
and endosteum is observed, grafting with cancellous bone, bone substitutes or growth
factors and providing stable external or internal fixation.[18]
[19]
[20] Regeneration of new bone, particularly large amounts, ensuring the shape of the
newly formed bone is adequate and overcoming soft tissue problems associated with
nonunion are challenging. Amputation may be the choice in many cases because of financial
burden and a guarded prognosis with respect to achieving normal limb function.
Transverse fractures or nonunion of the radial neck can be stabilized with small T-plates
when the radial head fragment provides sufficient bone for the placement of two screws.
More comminuted radial head and neck fractures, proximal fragments that are too small
to achieve stable fixation or fractures that offer little opportunity for load sharing
with internal fixation will likely require additional stabilization. This includes
the placement of a circular external skeletal fixator or a hybrid-circular external
skeletal fixator with wires that run from a proximal ring to engage the radial head.[4] In our case, because of the iatrogenic damage resulting from previous surgeries,
the radial head did not provide sufficient bone for implant placement and therefore
an alternative technique was needed. Ideally the proximal ulna would be fixed to the
distal radius allowing the proximal radius to ‘float’ while still attached to the
ulna by soft tissue structures (annular ligament, cranial crura of the collateral
ligaments and joint capsule) to maintain its position. This can be achieved by using
an external skeletal fixator that engages the proximal olecranon and extends to the
radius and ulna distal to the fracture or by simply stabilizing the ulna alone with
a bone plate or interlocking nail.[4]
[31]
[32]
In toy breed dogs, the shape of the distal ulna precludes stable fixation using a
bone plate on the ulna alone, and the small diameter of the medullary canal of the
distal ulna does not allow insertion of an interlocking nail.[35] Therefore, we felt that proximal fixation of the olecranon and distal fixation of
the radius was the best treatment option. The use of an external skeletal fixator
was not chosen because of reluctance of the owners to provide postoperative home care
and the aggressive nature of the patient.
Locking bone plates, also called fixed angle implants, have been developed over the
past two decades in an attempt to overcome limitations of conventional plates and
screws. Plates with a fixed-angle locking system do not require bone-plate contact
to achieve stable fixation, which aids in maintaining desired fragment positioning
and eliminates the need for plate contouring. This type of implant aligns with the
philosophy of biological osteosynthesis, which highlights functional alignment, relative
fracture stability and promotion of an optimal biological environment to promote fracture
healing. The biomechanics of locking plates and screws is similar to that of external
skeletal fixators and their transfixation pins; locking screws act as transverse supporting
components subjected to cantilever bending. The angular stability of the construct
converts shear stress created during axial loading or bending to compressive stress
at the screw–bone interface.[36]
Several locking plate systems are commercially available and currently used at our
clinic. The ALPS system was chosen because of its material properties and ease of
contouring, which can be accomplished in three planes. The plates are made of grade
4 titanium (Cp Ti, ASTM F‐ 67, ISO 5832‐2) and the screws are made of titanium alloy
(Ti‐6A1‐4V, ASTM F1472, ISO 5832‐3).[37]
An experimental study on rabbits showed increased resistance to localized infection
when titanium implants were used, and we therefore decided that this material would
be of benefit in our patient with necrotic nonunion.[38] Even when results of culture and sensitivity testing are negative, nonunion attributable
to the presence of bacteria is possible and likely underdiagnosed. A study that compared
molecular diagnostics and traditional culture methods in 24 human patients with nonunion
found bacteria via molecular diagnostics in 30 samples but positive culture results
in only eight.[39]
The plate was positioned on the lateral surface of the proximal and distal ulnar fragments
and contoured. The lateral surface of the ulna is straight and thus only minor contouring
was required to restore alignment in the frontal plane, while a more pronounced ‘on
plane’ contouring could have been made to avoid the mild iatrogenic recurvatum that
resulted from the application of the straight plate. Locking screws need to be inserted
perpendicular to the plate, which means the direction of the screw cannot be adjusted.
Because the distal ulna is positioned slightly caudal with respect to the distal radius,
the plate had to be twisted to direct the screws through the ulna and into the distal
radius in a caudolateral to craniomedial direction.[35] The degree of twisting was determined visually after insertion of the proximal screws.
The use of a polyaxial locking plate would have made orientation of the distal screws
easier, but this system was not available at our clinic at that time.[36]
[40] Another option would have been to contour the plate so that its distal aspect fits
the cranial surface of the radius. However, we used the lateral surface of the ulna
for plate placement because application was easier and it eliminated the need for
more aggressive dissection of tissues and more surgical time. The plate in this configuration
acts as a uniplanar unilateral linear external skeletal fixator, which is the least
stiff configuration of an external skeletal fixator. The locking plate, however, is
relatively closer to the bone than the connecting bar of an external skeletal fixator,
providing increased mechanical advantage for the construct.[36]
Different grafting techniques are effective in treatment of nonunion, and we used
rhBMP‐2 delivered in a collagen sponge. The BMPs initiate a cascade of developmental
events, in which pluripotent mesenchymal cells are induced to differentiate into osteoprogenitor
cells, osteoblasts and ultimately into osteocytes thus resulting in new bone formation.[41]
[42]
[43]
[44] The BMPs were effective in healing of experimentally-induced osteotomy and for repair
of mandibular defects and nonunion in dogs.[27]
[28]
[29] Successful use of rhBMP-2 depends on the quantity and concentration of the product,
time of application and incorporation of a delivery vehicle. A single product approved
for augmentation of fracture healing in dogs was available in the European Union (TruScient;
Zoetis Inc., Madison, New Jersey, United States), which is rhBMP‐2 (0.2 mg/mL) delivered
in an absorbable collagen sponge. A paucity of data on the dose of rhBMP-2 used in
clinical cases renders comparison of results difficult. When the dose of rhBMP-2 is
too high, hypertrophic bone and soft tissue inflammation may occur.[45] We found that using two-thirds of an absorbable collagen sponge soaked in 1.4 mL
of solution provided rapid filling of the bone defect with formation of abundant bone
callus at the 4-week re-evaluation. Immediate postoperative soft tissue swelling adjacent
to the graft is a potential complication of rhBMP‐2 because of an associated increase
in angiogenesis. Leakage of high protein transudate from a large number of new blood
vessels is thought to be the source of rhBMP‐2-induced postoperative swelling when
the product is used at pharmacological doses. Excessive soft tissue swelling is associated
with an increased risk of incisional dehiscence.[45] In our case, there were no signs of soft tissue swelling at the time of bandage
change 2 days after surgery and wound healing proceeded uneventfully.
We elected staged removal of the implants because of suspected osteopenia of the diaphyseal
bone secondary to stress protection seen on radiographs at the 8-week re-evaluation.
Stress-protection osteopenia occurs when bone undergoes cortical atrophy because of
subphysiological loading of the bone associated with implants or external devices.[46] In toy breeds, stress-protection osteopenia is a concern and thus the standard of
care at our clinic includes staged removal of implants to avoid this complication.
In this case, the positioning of a screw in the proximal aspect of distal ulnar fragment,
shortening the working length of the plate, could have contributed in increasing the
stiffness of the construct. Dynamization or sequential removal of the screws and plate
allows a gradual increase in strain and a controlled increase in load to stimulate
remodelling in the affected bone.[47]
[48] Dynamization was effective in promoting an increase in bone density and eliminating
the radiographic signs of osteopenia in our case, but required two additional operations
with all of the associated costs and risks.
Compared with the contralateral limb, severe bone loss resulted in a total length
reduction of 14% in the left radius of our patient. To the authors' knowledge, data
on the amount of discrepancy between the length of the forelimbs that is acceptable
without causing significant lameness are not available. Dogs can tolerate up to 20%
reduction in femoral length by increasing the standing angle of the stifle and hock.[49] It can therefore be presumed that a certain amount of discrepancy in forelimb length
can be overcome by increasing the standing angle of the shoulder and elbow. The mild
intermittent lameness and moderate palmigrade stance seen at the final re-evaluation
were likely caused by degenerative changes involving the radiocarpal joint and by
the shortening of the antebrachium that prevented the right tension and subsequent
antigravity activity of the flexor carpi ulnaris muscle on the accessory carpal bone
in the stance phase. Mild recurvatum of the radius could also play a role in carpal
hyperextension. The degenerative changes of the radiocarpal joint may have been attributable
to previous intramedullary pinning from the distal articular surface of the radius
at the primary care clinic. Repair of radial diaphyseal fractures using intramedullary
pins is not recommended because this technique has been associated with a high rate
of postoperative complications.[6]
Other possible cause of lameness includes changes to radial head caused by previous
surgeries and radioulnar synostosis at the site of fracture healing. Synostosis is
a common complication of radioulnar fracture repair in dogs and cats.[7] Even if the effects of this complication are unknown, in dogs the radius and ulna
have an average of 15 degrees pronation and 17 degrees supination, respectively, during
flexion and extension while walking and trotting on a treadmill and thus radioulnar
synostosis abolishes physiological joint kinematics, possibly predisposing to lameness.[50]
In conclusion, internal fixation using a locking plate with proximal screws in the
olecranon and distal screws in the radius can be considered an effective treatment
method in dogs with proximal radial nonunion and a small proximal radial fragment.
