Five years ago, UC Davis veterinary oral surgeons started using recombinant human bone morphogenetic protein-2 (rhBMP-2) as a bone stimulator to repair fractures and large defects in mandibles. To date, they have successfully regrown nearly three dozen jawbones.
Now that technology has made its way to orthopaedic surgery, and data from those cases are being utilised to facilitate the application of this substance in the repair of nonunion long-bone fractures in dogs.
“Nonunion after long-bone fracture repair in dogs represents a potentially devastating complication,” comments Dr Amy Kapatkin, chief of the Orthopedic Surgery Service at the UC Davis Veterinary Medical Teaching Hospital. “We are excited about this new treatment and are optimistic that our use of orthopaedics can have the same long-term positive results our oral surgeons have seen with jawbones.”
Though the rate of nonunion bone fractures is unknown in veterinary medicine, it can be as high as 10% in reports in human literature.
What is rhBMP-2?
Bone morphogenetic proteins are endogenous, signalling peptides that stimulate mesenchymal stem cells to differentiate into osteocytes, creating new bone formation.
Two recombinant human bone morphogenetic proteins (rhBMP-2 and rhBMP-7) are FDA-approved for use in specific human medical conditions in the United States, including maxillofacial, spinal fusions, open tibial fractures and compassionate use in long-bone unions.
Application in dogs
Nine dogs with 11 fractures were enrolled into the study at UC Davis. All patients had had at least one previous surgery, while some had had up to five previous surgeries, with continued nonunion of their fracture. These cases were considered at high risk of failure with the traditional technique of rigid fixation and autogenous or allogenic bone graft, and amputation may otherwise have been recommended.
The dogs underwent general anaesthesia, with removal of previous implants and debridement of fracture ends. The bones were then secured with bridging implants. The defect was measured and compression-resistant matrix infused with rhBMP-2 was placed, ensuring bone contact.
Nine of 11 nonunion limbs returned to full function postoperatively; the other two patients had acceptable function. One patient, followed for 24 months, had decreased carpal range of motion and mild muscle atrophy with limb shortening resulting in a mechanical lameness.
The other dog had occasional proprioceptive deficits at 10-week follow-up. Radiographs demonstrated progressive integration of bone at the site of rhBMP-2 placement, with time to healing varying from seven to 20 weeks (median: 10 weeks).
Repairing Charlie
Charlie, a one-year-old Jack Russell terrier, presented to orthopaedic surgeons at UC Davis following surgery at a different veterinary hospital. That surgery came after a traumatic event (most likely hit by a car) in which he suffered a right open (unknown grade), comminuted tibial fracture and a craniodorsal left hip luxation.
Charlie had a femoral head osteotomy performed on his left hip and a bilateral uniplanar external fixator was placed on his right tibia and fibula fractures; 1cc of Osteoallograft was placed during the procedure.
On radiographic follow-up, the tibia progressed to an avascular nonunion and the fibula showed signs of a hypertrophic nonunion, so he was brought to UC Davis for revision.
Charlie’s general examination was all within normal limits. He was ambulating on all limbs, but was lame in the right pelvic limb. He had a bilateral, uniplanar external fixator (with five pins – three full and two half) placed on his right tibia.
There was significant muscle atrophy of the right pelvic limb. The mid diaphysis of the right tibia felt thin and the fracture had movement. There was some pain on extension of left stifle and hip. Radiographs at UC Davis of his right tibia and fibula showed a suspect atrophic nonunion of the tibia with punctate radiolucencies that suggested a lowgrade bone infection.
Without histopathologic and ultrasonographic analysis, atrophic nonviable nonunions cannot be definitively differentiated from oligotrophic viable nonunions. The fibular fracture was consistent with hypertrophic nonunion secondary to increased motion. There was also poor alignment of the right femur and right tibiotarsal joint.
The external fixator was removed from Charlie’s right tibia and movement at the fracture site was confirmed. Dr Kapatkin performed an ostectomy (approximately 1cm distal and 2cm proximal) of the avascular tibia. A 2mm, 13-hole locking plate was placed bridging the entire tibia. A total fracture gap of approximately 3.5cm was created with this procedure.
This could open the door for further applications that may allow regrowth of bones and repair of fractures
Three locking screws proximal and two distal were placed. A 3.5cm compression-resistant matrix (CRM) was soaked with rhBMP-2 and placed in the 3.2cm gap, ensuring contact with the bone ends. Vancomycin osteoset antibiotic beads were then placed all along the edges of the locking plate and within the screw holes.
The procedure yielded no complications; post-operative radiographs showed adequate plate and screw positioning, and joint alignment. The rhBMP-2 and Vancomycin beads were also visible on the radiographs.
Charlie’s owners were warned to strictly limit his activity for at least eight weeks to allow for the bones to properly heal and to prevent the plate from breaking. He was to be carried outside for elimination purposes.
They were given instructions to keep Charlie’s incisions clean and dry and warned that it was not unusual for the incision to be mildly swollen and red with a small amount of red-tinged discharge for the first few days.
Additionally, the rhBMP-2 commonly causes a flaretype reaction and the entire leg may appear red and inflamed for about a week. This is expected with the use of BMP; excessive swelling and inflammation have led to the discontinuation of its use in cervical spinal fusion in humans due to the risk of upper airway obstruction.
At his four-week recheck examination, Charlie’s right pelvic limb was weight-bearing and not painful on palpation. He did, however, display a partial hock drop and conscious proprioceptive deficits.
Radiographs of his right tibia and fibula showed complete resorption of the CRM and the antibiotic beads. There was incorporation of the rhBMP-2 at both sites with significant boney callus, but no remodelling had occurred yet – expected at only four weeks.
At 10 weeks, radiographs of the right tibia and fibula showed complete healing and a start to remodelling of the bone. Charlie was green-lit to resume normal activity, being able to walk and run with no adverse effects.
Though his conscious proprioceptive deficits were present at that visit, the owner reports they have since resolved.
Implications of this study
This study combined fracture fixation with regenerative technique to provide successful outcomes in cases with otherwise guarded to poor prognoses. This could open the door for further applications that may allow regrowth of bones and repair of fractures not previously considered for surgical repair.
