Bifocal Femoral Deformity Correction and Lengthening Using a Circular Fixator Construct in a Dog

Introduction

Correction of femoral deformities requiring lengthening is problematic in dogs because anatomic constraints make application of traditional circular constructs difficult proximal to the stifle.¹ There are two reports describing femoral lengthening in dogs.^1,2 In those reports, the angular and rotational components of the deformity were corrected acutely at surgery with subsequent distraction performed through the same osteotomy/ostectomy site. All dogs had implant complications and inconsistent regenerate bone formation in the distraction gap.^1,2

In human patients with traumatically induced deformities, subsequent lengthening through the deformity correction site often yields poor regenerate bone formation.³ Poor regenerate bone formation has been ascribed to alterations in the osseous circulation and the periosseous soft-tissue envelope resultant from the trauma that produced the deformity.³ Distraction performed at a second osteotomy made remote to the location of deformity correction has been used to perform lengthening through a region of the bone that has normal morphology.^4–6

This report describes bifocal deformity correction and subsequent distraction osteogenesis performed using a circular fixator construct in a dog with a malunion fracture of the distal femur. Distal metaphyseal wedge ostectomies were performed to acutely correct the angular and rotational components of the deformity, with subsequent distraction osteogenesis performed through a second proximal diaphyseal osteotomy.

Case Report

A 7 mo old male rottweiler weighing 36 kg was evaluated 5 mo after sustaining a Salter-Harris type II fracture of the right distal femur. The dog had an obvious weight-bearing right hind limb lameness. The right thigh was visibly shortened, and the right hind limb was externally rotated at the level of the stifle. Moderate crepitus could be elicited on flexion and extension of the right stifle. The right patella was positioned laterally throughout the range of stifle motion; however, the patella could not be luxated. A positive Ortolani sign was elicited in the right coxofemoral joint.

Radiographs and computed tomography (CT) of both hind limbs were obtained (Figures 1A–D). There was a malunion fracture involving the distal physis of the right femur with smooth periosteal new bone partially bridging the fracture. An irregular, obliquely oriented radiolucent defect was present laterally within the distal metaphysis. Measurements of femoral alignment were calculated from frontal and sagittal radiographs (Table 1).^7–11 Femoral frontal plane angulation and torsion were measured from the CT images (Table 2).¹² There was 40° of distal femoral valgus, moderate procurvatum, and the femoral condyles were estimated to be externally rotated by approximately 45°. The right femur was 35% shorter than the left femur. Mild incongruity of the right coxofemoral joint was also noted.

TABLE 1 Measurements Obtained from the Preoperative Radiographs and Radiographs Obtained 14 Mo Post Surgery

Procurvatum was defined as mCDFA < 79°, and recurvatum was defined as mCDFA > 87°. aLPFA, anatomic lateral proximal femoral angle; aLDFA, anatomic lateral distal femoral angle; mLPFA, mechanical lateral proximal femoral angle; mLDFA, mechanical lateral distal femoral angle; aCPFA, anatomic caudal proximal femoral angle; aCDFA, anatomic caudal distal femoral angle; mCDFA, mechanical caudal distal femoral angle.

View Fullscreen

TABLE 2 Measurements Obtained from the Preoperative CT Images and CT Images Obtained 14 Mo Post Surgery

*

Positive value represents abaxial angulation, negative value represents axial angulation.

†

Positive value represents axial angulation, negative value represents abaxial angulation.

§

Positive value represents cranial deviation of the femoral neck.

**

Positive value represents cranial deviation of the medial femoral condyle.

††

fTA was calculated by subtracting tTCAA from the tFHNA.

CT, computed tomography; fTA, femoral torsion; tFHNA, torsion femoral head and neck axis; tTCAA, torsion transcondylar axis angle; vPFLAA, varus proximal femoral long axis angle; vTCAA, varus transcondylar axis angle.

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

An anatomic, three-dimensional model was created to assist in planning the surgical correction. Acute correction of the deformity was planned by performing a cuneiform closing wedge (a 40° wedge with the apex positioned laterally in the frontal plane to correct the valgus, and a 45° wedge in the sagittal plane with the apex positioned caudally to correct the procurvatum) ostectomy through the center of rotation of angulation.^{1,7–9,13,14} Axial rotation would also be corrected acutely with subsequent distraction osteogenesis performed through a separate, proximal, diaphyseal osteotomy to lengthen the femur.

The dog was anesthetized and the right hind limb was aseptically prepared for surgery. The dog was positioned in left lateral recumbency, and a lateral approach to the right femur and stifle was made.¹⁵ A 40° distal femoral wedge ostectomy was made in the frontal plane, with the apex of the wedge positioned laterally to correct the valgus deformity. A second 45° wedge of bone was excised from the proximal end of the distal femoral segment in the sagittal plane, adjacent to the initial ostectomy, with the apex of the wedge positioned caudally to correct excessive procurvatum. The ostectomized surfaces of the proximal and distal femur were opposed, and external rotation of the condylar segment was corrected subjectively by visually aligning the trochlear groove while palpating the position of the greater trochanter. The femur was stabilized using a 3.2 mm Steinmann pin placed through the intercondylar notch of the distal femur in normograde fashion. Two convergent interfragmentary 1.6 mm Kirschner wires were also placed for adjunctive stability. Autogenous bone graft harvested from the greater tubercle of the right humerus was placed in the recesses where the ostectomy surfaces were not in direct apposition. A subperiosteal transverse osteotomy was initiated in the proximal diaphysis of the right femur, leaving the medial cortex intact. All osteotomies were performed with a sagittal bone saw^a and intraoperative fluoroscopy was used to assess the position of the osteotomies, anatomic alignment, and implant placement.

A circular construct was applied to the femur^b (Figure 2). An incomplete 118 mm diameter stretch ring was placed perpendicular to the longitudinal axis of the femur, circumscribing the femoral condyles, with the open section of the ring oriented caudally. A 1.6 mm Kirschner wire and two divergent 1.6 mm olive wires were used to secure the stretch ring to the distal femoral segment. A 1.6 mm Kirschner wire was placed through the distal femoral diaphysis and secured to the stretch ring. All fixation wires were tensioned to 60 kg. Two linear motors, oriented parallel with the longitudinal axis of the femur, were mounted craniolaterally and caudolaterally on the stretch ring. Those motors were secured proximally to an 84 mm diameter one-third ring arch positioned laterally subjacent to the greater trochanter and perpendicular to the longitudinal axis of the femur. Three half pins were placed in the proximal femur from the ring arch. A short threaded rod, secured with paired nylon nuts, was mounted on the craniomedial portion of the stretch ring using a two-hole plate. The proximal portion of this rod was articulated with the ring arch using double connecting clamps and a carbon fiber rod (Figure 2). After the construct was in place, a small osteotome was used to complete the proximal diaphyseal transverse osteotomy through the medial cortex. Closure was routine.

View Figure

• Download as Powerpoint
• Download Figure

Rotational and angular alignment was deemed to be acceptable on postoperative radiographs (Figures 3A, B). The dog received cefazolin^c (22 mg/kg IV perioperatively then q 6 hr thereafter), postoperative hydromorphone^d (0.1 mg/kg IV q 4 hr), and carprofen^e (2 mg/kg subcutaneously q 24 hr). On the second day postoperatively, oral carprofen (2.2 mg/kg q 24 hr) and oral tramadol^f (2.2 mg/kg q 8 hr) was initiated and continued for 10 days. Cephalexin^g (22 mg/kg per os q 8 hr) was also prescribed for 14 days.

View Figure

• Download as Powerpoint
• Download Figure

The dog was intermittently bearing weight on the right hind limb 3 days after surgery when distraction was initiated at a rate of 1 mm/day, fractionated into four daily increments. Radiographs were obtained q 7–10 days to monitor the distraction process. Pin-skin and wire-skin interfaces were cleaned q 24 hr using 0.05% chlorhexidine solution^h, followed by application of a topical triple antibiotic ointmentⁱ.¹⁶ The dog was consistently placing moderate weight on the right hind limb by 10 days post surgery, and the lameness steadily improved until there was a sudden increase in lameness 21 days after surgery. Palpation of the right stifle revealed a decreased range of motion, pain, and crepitus. Radiographs revealed early regenerate bone formation at the margins of the distraction gap. It was also noted that the intramedullary pin had migrated distally into the stifle. The pin was removed through a lateral stifle arthrotomy and replaced with a 3.9 mm Steinman pin^b.

Distraction was performed for 53 days, although there were intermittent lapses in performing distraction due to pin migration, wire tract drainage, and lapses in owner compliance. Although the distal wire tracts had mild to moderate serosanguinous discharge nearly the entire time the fixator was in place, drainage had become purulent and flexion of the stifle had decreased by 56 days after surgery. Radiographs revealed that the distal ostectomy had obtained union. The fixation wire placed through the distal femoral diaphysis as well as one of the interfragmentary Kirschner wires, which had migrated, were removed (Figures 4A, B). Approximately 30 mm of femoral lengthening had been achieved. Distraction was terminated, and the linear motors were replaced with threaded rods to statically secure the construct. Consolidation of the regenerate bone progressed sufficiently, and the fixator was removed 86 days following surgery.

View Figure

• Download as Powerpoint
• Download Figure

The dog was re-evaluated 14 mo post surgery. The owner reported that the dog had a mild lameness that did not curtail the dog’s normal daily activities. The dog stood with the right hip, stifle, and hock held in a slightly more extended posture than the contralateral limb. When the dog was placed in lateral recumbency there was still a visible length discrepancy between femurs. The dog had a subtle, asymeteric hind limb gait in which the stride in the right hind limb was shortened, and the dog circumducted the left hind limb during the swing phase of the stride. A positive Ortolani sign could still be elicited from the right coxofemoral joint. The circumference of the right thigh (45.3 cm) was smaller than the left (51.5 cm), and range of motion of the right stifle was slightly decreased (extension, 155°; flexion, 70°) compared to the left stifle (extension, 158°; flexion, 55°). Radiographs and CT images revealed that the conformation of the right femur was substantially improved compared to the preoperative images. The proximal osteotomy and distal ostectomy sites had remodeled and were ill-defined with smooth, contiguous cortical margins (Figures 5A–D). Mild subluxation of the right femoral head persisted, femorotibial and patellofemoral alignment was excellent, and mild right stifle osteoarthritis was evident. The right femur was still 16% shorter than the left femur, and the right tibia was 5% longer than the left tibia.

View Figure

• Download as Powerpoint
• Download Figure

Discussion

The authors’ decision to lengthen this dog’s femur through an osteotomy performed remote to the site of acute angular and rotational deformity correction was based on their previous experience with three dogs that underwent femoral deformity correction involving distraction osteogenesis.¹ Distraction osteogenesis was performed at the osteotomy/ostectomy used to correct the angular and/or rotational deformity and yielded poor regenerate bone formation in those three dogs. All three dogs required revision surgery, which included bone grafting of the distraction gap.¹ The bifocal approach used in the dog reported herein allowed the authors to distract the femur in a region that had normal morphology and osseous circulation. Identification, elevation, and preservation of the periosteum were not feasible when performing the distal ostectomies due to the abundant callus formation and adhesions enveloping the malunion; however, the authors were able to effectively elevate and preserve the periosteum when the osteotomy was performed in the proximal diaphysis.

Acute deformity correction with subsequent distraction of a separate, second osteotomy located in a region of the bone with normal morphology has been described for treatment of pseudoarthrosis, osteomyelitis, and hip reconstructions in human patients.^4,5,17–19 This bifocal approach has been used in human patients to ensure more consistent regenerate bone formation by performing distraction osteogenesis in a region of the bone remote to the deformity.^6,20 Regenerate bone was visible at the margins of the distraction gap on radiographs obtained 18 days after distraction was initiated in the dog described in this report, and a confluent column of regenerate bone with a viable fibrous interzone was present when distraction was terminated 56 days after surgery. Consolidation was sufficient to allow removal of the fixator 30 days after distraction was terminated.

The construct used allowed for efficient distraction of the femur. Traditional circular fixator constructs, which use only small diameter wires as fixation elements that traverse the diameter of the ring, cannot be applied to the proximal femur.^21,22 The authors had concerns that the construct might result in eccentric distraction and induce angulation because the proximal femoral segment was only stabilized by laterally placed half pins; however, the distraction was linear. The intramedullary pin was placed primarily to align and stabilize the realigned distal femur, but may have helped maintain axial alignment during the subsequent distraction. The diameter of the pin was small relative to the diameter of the medullary canal, and the authors credit the stability afforded by the fixator for the linearity of the distraction. The original linear motors were easily exchanged with longer linear motors to allow for continued lengthening during the distraction process.

The authors had used a hybrid construct articulated with an interlocking nail in previous attempts to lengthen the femur.¹ There was considerable morbidity associated with the nail protruding proximally from the femur, and based on the results in the dog in the current report, the nail is not necessary. In the previous three dogs, the authors also used full pins rather than wires as fixation elements on the distal ring.¹ The dog in the current report had mild wire tract drainage, which became purulent 8 wk following surgery, influencing the authors’ decision to terminate distraction. The wire tract complications were initially minor compared with those in the dogs in which full pins were used to stabilize the distal femoral segment.¹ Fixation wires efficiently erode through the periosseous soft tissues and are potentially less problematic than larger diameter fixation pins.²³

Bifocal treatment affords the latitude to compress the site of angular and rotational correction while distraction is performed at a remote location.^18,20,24,25 Compression of the ostectomy site was achieved by placing a fixation wire through the distal femoral diaphysis, which was subsequently attached to the stretch ring and tensioned. Interfragmentary compression has been shown to facilitate bone healing at sites that are prone to delayed/nonunion, and radiographic union of the ostectomy was obtained by 8 wk following surgery.^20,25,26 An additional advantage of utilizing a bifocal approach is that distraction osteogenesis has been shown to increase appendicular blood flow distal to the distraction site, which enhances bone healing throughout the limb.^3,27,28

The dog in the current report had a mild lameness when evaluated 11 mo following fixator removal, which we ascribed to persistent hind limb asymmetry. The affected femur was 16% shorter than the left femur, with 5% compensatory overgrowth of the ipsilateral tibia.^29–31 A previous study reported that shortening the length of one femur by 20% did not produce significant gait abnormalities. Dogs can accommodate length discrepancy between femurs by altering the joint angles in both the ipsilateral and contralateral hind limb, similar to what was observed in this dog at the time of final evaluation.³² The dog in the current report also circumducted the left hind limb during the swing phase of the stride which seemed to be compensatory for the residual hind limb length discrepancy. There was also persistent laxity of the right coxofemoral joint, which may have contributed to the dog’s lameness. Coxofemoral joint laxity has been reported in other dogs with femoral deformities with length discrepancies and may be a consequence of altered weight-bearing on the deformed limb.¹

Conclusion

The bifocal approach used in the dog in this report appears to be an effective technique for managing femoral deformities that require lengthening. Performing distraction osteogenesis at an osteotomy made remote to the site of angular and rotational correction yielded a confluent column of regenerate bone. Although the authors still experienced complications, including implant migration and wire tract drainage, postoperative morbidity was considerably less than in the authors’ previous experience with femoral lengthening in dogs with angular and/or rotational deformities.¹ There is still an obvious need for further refinement of methods for lengthening the femur following angular and/or rotational correction in dogs.