1 s2.0 S089184220200054X Main PDF
1 s2.0 S089184220200054X Main PDF
1 s2.0 S089184220200054X Main PDF
20 (2003) 65 – 96
* Corresponding author.
E-mail address: jfleming@rocketmail.com (J. Fleming).
0891-8422/03/$ – see front matter D 2003, Elsevier Science (USA). All rights reserved.
PII: S 0 8 9 1 - 8 4 2 2 ( 0 2 ) 0 0 0 5 4 - X
66 S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96
a modular device that can accommodate the complex limb deformities while
performing multiple tasks. This article summarizes the authors’ experience with
compression arthrodesis of the rearfoot and ankle using external fixation.
the talar body results in substantial loss of bone substance, greatly compromising
screw placement and thread purchase in an attempt to create rigid fixation in the
hindfoot. This distorted architecture can be easily stabilized by multiple small
diameter transosseous wires for ankle and subtalar fusions. Similarly, Charcot
neuroarthropathy presents significant challenges to the foot and ankle surgeon
from two perspectives. The severe midfoot and rearfoot collapse and osseous
destruction in combination with regional osteoporosis lends itself to fine wire
fixation. Second, cast immobilization and single-limb weightbearing during the
postoperative period allows weight transfer to the contralateral limb, possibly
inciting a similar breakdown on a healthy foot. Stabilization and compression of a
failed infected arthrodesis site or pseudoarthrosis can result in satisfactory fusion
and salvage of a functional extremity [19]. External fixation is used at the authors’
institution primarily for compression arthrodesis of hindfoot joints, especially in
patients with complex deformity and significant structural abnormalities.
Rigid fixation constructs are possible in the distal lower extremity, though with
certain drawbacks. Safe corridors for pin and wire placement in this territory are
narrow. Transosseous wires have the potential to violate hazardous or unsafe zones
and impale myotendinous or neurovascular structures. Accidental joint penetration
with half-pins or wires can occur, potentially leading to arthrosis or joint sepsis. As
with any form of rigid fixation, stress shielding of bone is inevitable. When external
fixation is used for passive mechanical compression across an arthrodesis site, bone
demineralization is inevitable and restoration of cortical integrity through gradual
weightbearing will be required [20]. Radiographic visualization of the surgical site
is often difficult because of interposed hardware, but many external systems have
now incorporated carbon-fiber rings to eliminate this problem.
Immunocompromise, reflex sympathetic dystrophy, and venous stasis disease
are also relative contraindications [21,22]. There is a steep learning curve to this
method of fixation, which becomes magnified in consideration of the exceptional
anatomy of the foot and ankle and specific biomechanics involved in fixation of
this region.
arm. Single-ring fixation blocks are used at the talus and foot level. Alterations in
this basic frame setup will depend on the specific goals of the operation and the
fixation mode desired between each block.
Two half-rings are selected that are 2 to 3 cm larger than the maximal limb
girth. Approximately two fingerbreadths distance from the ring to the skin surface
is required posteriorly and one fingerbreadth anteriorly. Additional space is
needed posteriorly in the leg to accommodate for the increased soft tissue edema
postoperatively and requires the frame to be slightly offset in this direction.
Eccentrically positioning the bone within the frame has also been shown to
increase axial and torsional frame stiffness [13,16]. Inferiorly, the distance from
the weightbearing surface of the plantar foot to the foot plate should be
approximately two fingerbreadths to allow for postoperative ambulation. The
forefoot half ring should provide enough clearance for unobstructed dorsiflexion
of the digits. Too little space between the fixator and the soft tissue may result in
impingement and ulceration, whereas too much space will decrease the stiffness
and compromise the stability of the fixator.
Rings are available in sizes from 80 to 240 mm, and most adults require 150 to
160 mm rings for the tibia. These are assembled into two coplanar rings that are
placed in the distal tibial segment approximately 8 to 10 cm apart. Longitudinal
connecting elements (threaded or telescoping rods) are spaced to provide access
to sites on the frame where half-pins and wires will be affixed the frame. This is
generally from 2 o’clock to 5 o’clock position laterally and from 7 o’clock to
11 o’clock position medially for a right limb. Four connecting elements will
S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96 69
connect the rings, and three will connect the distal ring to the foot plate below.
The two connected rings comprise the supramalleolar tibial ring block. These
rings are attached to a frontal plane foot plate, which may be ‘‘closed’’ with the
addition of a transverse plane half-ring distally to resist frame deformation with
wire tensioning. The half-ring is also affixed to the inferior tibial ring through a
post-and-rod assembly. As the tibial ring block is important in stabilizing the
proximal tibial lever arm, so is the foot plate important in stabilizing the distal
lever arm (foot) in an attempt to create sagittal plane stiffness and neutralize
bending moments at the ankle level.
The tibial rings will be secured with at least two transosseous smooth or olive
wires each and possibly additional half-pins, which will enter the medial face of
the tibia. Efforts should be made to make space in this region of the frame
assembly for attachment of half-pins. Likewise, the foot plate will be secured
with two oblique crossing smooth or olive wires through the posterior calcaneus.
Midfoot and metatarsal wires are accommodated by distal holes on the foot plate
(Fig. 1).
Wire placement
To prevent damage to neurovascular elements and myotendinous structures
with the insertion of wires and pins, a three-dimensional understanding of cross-
sectional anatomy is essential (Figs. 2, 3). There are many publications that
detail this regional anatomy [23,24]. Although a 90° orientation or delta
configuration between the wires creates the most stable construct, this is not
always possible because of limb anatomy. Fleming et al [13] demonstrated a
decrease in the bending stiffness of the frame by a factor of two by decreasing
the angulation between two wires from 90 to 45°. Counterposed or ‘‘dueling’’
olive wires are used to fixate the osseous segment in space, creating a sandwich
or vice effect when the wire orientation is less than 90°. Additionally, 5-mm or
6-mm half-pins can be placed in the medial face of the tibia to increase
resistance to anterior-posterior bending forces acting on the tibia [21]. Olive
wires should be used wherever possible because they have been shown to
increase bending, torsional, and axial stiffness relative to smooth wires [13].
Consideration should be given to increasing the number of wires in the obese
patient or in the presence of osteopenic bone. Half-pins may be used if
additional stability is needed because of the narrow safe corridors present and
dense cortical architecture of the anterior tibia. Half-pins do not traverse a
muscle compartment and are therefore less painful than a transosseous wire,
which may facilitate early weightbearing. Two half-pins are required to simulate
the stiffness of a single tensioned wire [25].
If olive wires are not used, half-pins will provide resistance to sliding of the
osseous segment on smooth tensioned wires.
70 S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96
Fig. 1. Basic Ilizarov frame construct for ankle and hindfoot reconstruction. Two transverse coplanar
rings stabilize the distal tibia, and a ‘‘closed’’ foot plate is used to stabilize the hindfoot and midfoot
regions. Counterposed or ‘‘dueling’’ olive wires and half-pins stabilize bony segments. An additional
ring may be added at the level of the talus for an isolated tibiotalar fusion.
Wire tensioning
The functionality of the smooth wire external fixator depends on tensioned
transosseous wires. Placing the wires under tension provides the stability
necessary to achieve axial compression and the mechanical resistance to
eliminate distractive forces. The appropriate wire tensions for the foot and leg
are listed in Box 2.
There are several techniques for tensioning smooth wires, including the Russian
slotted bolt head torque and turnbuckle techniques, the post tilt technique, and the
Italian dynamometric wire tensioner technique, each with specific advantages and
disadvantages. The Russian slotted bolt head torque consists of rotating the bolt
head and nut simultaneously against a wire fixed to the contralateral side of the
frame (Fig. 4B). This technique may add 10 to 20 kg of tension and can be done in
S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96 71
Fig. 2. Cross-sectional anatomy for transosseous wire/pin placement in the leg. (A) Mid-leg. Cross-
section just distal to the midpoint between the knee and ankle joints. The posterior neurovascular (NV)
bundle lies at the geometric center of the leg and the anterior NV bundle lies on the anterior surface of
the interosseous membrane. The subcutaneous border provides 140° safe corridor for wire/pin
insertion. Transverse and medial face wires combined with a 5-mm or 6-mm half-pin provide stability
at this level. (B) Distal leg. This section is approximately 10 to 12 cm proximal to the level of the
ankle joint. The anterior NV bundle begins to move anteriorly between the extensor hallucis longus
(EHL) and tibialis anterior (TA) muscle bellies; the posterior NV bundle runs posterior to the
interosseous membrane in the deep posterior compartment. The safe corridor decrease at this level to
120°. A similar wire/pin configuration is seen. Half-pins may be readily used at A and B levels
because of the high percentage of cortical bone present in the tibia. (C) Ankle. Cross-section just
proximal to the level of the ankle joint. If the fibula is intact, it may be stabilized by a lateral oblique
wire directed from posterolateral to anteromedial. Remaining wires are directed from medial to lateral.
The major NV bundle is located in the posteromedial quadrant.
the office. The turnbuckle used in the original Russian system is affixed to the wire
and tensioned against the frame. The post tilt technique involves tightening a post
tilted toward the limb with its affixed smooth wire straight, thus straightening the
post and increasing tension on the wire. Manual wire tensioning with the torqued
slotted bolt or turnbuckle methods should be avoided in the foot and leg because it
leads to inconsistent wire tensions and precludes the use of these modalities in the
72 S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96
Fig. 3. Cross-sectional anatomy for transosseous wire/pin placement in the foot. (D) Hindfoot. Axial
section through the proximal hindfoot demonstrating location of anterior and posterior neurovascular
structures. The anterior structures traverse the ankle joint between the extensor hallucis longus (EHL)
and tibialis anterior (TA) tendons. The posterior bundle enters the foot just posterior and inferior to the
medial malleolus in the third compartment of the tarsal tunnel and courses distally into the plantar
vault. (E) Hindfoot. Calcaneus is stabilized by two counterposed crossing olive wires at approximately
90° to each other. These wires stabilize the hindfoot to the foot plate. Although half-pins may be used
in the same orientation, the cancellous architecture of the calcaneus may be better suited for fine wire
fixation. (F, G) Midfoot/forefoot. Sections through the metatarsal bases and metatarsal shafts
respectively. Just distal to the Lisfranc articulation, dorsally directed oblique crossed wires may be
used to stabilize this region. Small diameter threaded half-pins may be placed in the first metatarsal
base. The distal forefoot is similarly stabilized with a combination of dorsally directed crossed wires
and pins with care taken to avoid the dorsal NV structures.
S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96 73
Half-rings/drop wires: 50 kg
Pediatric patients (1.5 mm): 100 –110 kg
Adults patients (1.8 mm): 120 –130 kg
Foot: 70 –80 kg
Increase wire size, wire tension, and number of wires with
increasing body weight.
Fig. 4. Wire tensioning techniques. (A) The Italian dynamometric wire tensioner. The wire tensioner
device is calibrated between 50 and 130 kg and can be dialed to deliver the appropriate amount of
tension. This device consistently achieves the greatest wire tensions, although caution should be used
when tensioning against the softer cancellous bones of the foot. ‘‘Dueling’’ olive wires should be
tensioned simultaneously in the foot and leg to prevent shifting of the osseous segments. (B) Russian
tensioning technique. Torqued slotted bolt head, 10 mm hexagonal slotted bolt attached to frame,
friction fit to smooth wire. Nut affixed to bolt from opposite side of ring. Opposing wrenches turn the
secured nut-bolt-wire assembly away from the bone segment after the opposite wire end is fastened to
the frame. This torque movement of the bolt head creates wire tension.
74 S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96
through the myofascial compartment; continued drilling may engage the neuro-
vascular bundle, ‘‘wrapping’’ it around the wire. Sterile pin loosening and loss of
stability resulting from thermal bone necrosis can be prevented by intermittent
start-stop drilling at low speeds using the sharper bayonet tips at all times. A
moistened sponge or irrigation may also be helpful during wire insertion to
reduce temperatures at the wire/pin interface. Any wire that is a significant
distance from the ring and requires multiple stacked washers or a component
greater than a single hole post should be replaced as wire stiffness decreases with
increasing distance from the ring. ‘‘Drop’’ wires should be avoided as a third
point of fixation because they cannot be appropriately tensioned without frame
deformation. Reduced wire tension leads to instability and motion at the wire/
bone interface, producing osteolysis and higher infection rates secondary to soft
tissue irritation [24]. Any skin tension around the wires should be released to
prevent soft tissue irritation and discomfort.
Tibiotalar arthrodesis
Literature review/indications
Despite recent developments in the treatment of ankle arthritis, such as total
ankle arthroplasty, ankle arthrodesis remains the gold standard for end-stage
ankle arthrosis. Continuing advancements in osteoarthritits, such as arthroscopy,
viscosupplementation, arthrodiastasis, and realignment osteotomies, have
improved clinical outcomes and deferred the time to fusion.
Hammerschlag achieved 100% union in a series of 10 ankle fusions using
anterior arthrotomy and a small wire fixator with a single supramalleolar and talar
ring [27]. Two wires were placed in the talus and all patients were allowed to
weightbear postoperatively without forefoot stabilization from the fixator. Ham-
merschlag’s indications included osteopenia from chronic inflammatory disease,
septic arthritis, neuromuscular disease, and severe deformity or posttraumatic
deformities. Hawkins et al also described good success in 17 patients with the
Ilizarov technique for ankle fusion in patients with complex distal tibial
pathology or failed arthrodesis [28]. They reported an 80% union rate with a
22-month follow-up. Approximately 50% of their patients underwent concom-
itant leg lengthening for segmental bone defects, which likely accounts for their
high rate of complication and increased time in the fixator. Laughlin and Calhoun
also reported successful ankle arthrodesis in 16 of 20 patients with small wire
fixators [29]. Fusion was combined with concomitant tibial lengthening for
segmental bone loss in 12 patients. Kenzora et al [8], Kitaoka et al [30], and
Rothhacker et al [9] all reported similar fusion rates employing biplanar pin
fixators for revisional arthrodeses and fusion secondary to posttraumatic con-
ditions. Each study used comparable frame constructs and pin placement with
multiple transverse tibial pins and two transverse talar pins. Kenzora et al found
substantially dissimilar failure rates for patients who experienced high-energy or
S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96 75
low-energy injuries. The energy level of the injury was defined by the degree of
bony comminution, the presence of joint dislocation, and the condition of the soft
tissue envelope at the time of injury. High-energyy and low-energy groups
achieved 69% and 100% primary union respectively. They also reported a 43%
prevalence of pin tract infections, which may be secondary to the large diameter
pins characteristic of these devices.
Perioperative considerations
The patient is positioned in the lateral decubitis position with the affected
extremity up. If a radiolucent operating table is available, the contralateral
extremity may be flexed at the hip and knee and safely secured to the border
of the table to facilitate intraoperative fluoroscopy. A pneumatic thigh tourniquet
is used to facilitate visualization during dissection and joint resection. The patient
should not be fully paralyzed after intubation so that wires that penetrate or
contact nerve structures during insertion will produce a muscular reaction and
possible nerve damage may be avoided.
A transfibular approach is the most commonly employed approach at the
authors’ institution when external fixation is chosen. The fibula may also be left
intact. Regardless of the fixation method used or the incisional technique chosen,
ankle arthrodesis must strictly adhere to Glissan’s [31] four requirements to
achieve solid bony union: (1) complete removal of all cartilage, fibrous tissue,
and any other material that may prevent contact of raw bone surfaces; (2) accurate
and close fitting of the fusion surfaces; (3) optimal position of the ankle joint; and
(4) maintenance of the bone apposition in an undisturbed fashion until the fusion
is complete. As with any joint arthrodesis, the position of fusion will ultimately
dictate the outcome and long-term functional capabilities, and largely relies on
minimizing stress transfer to adjacent joints. This is most critical in the ankle joint
because of its implications on the ascending skeletal mechanics of the knee and
hip articulations in addition to the subtalar and midfoot joints. The optimum
position of fusion of the ankle is neutral flexion, slight hindfoot valgus (0– 5°),
and approximately 5 to 10° of external rotation or rotation, which mimics a
nonpathologic contralateral limb [32]. Posterior displacement of the talus beneath
the tibia creates a shorter lever arm and decreased stresses at the knee level. This
is of primary importance in Charcot reconstructions as the shorter lever arm
minimizes bending forces across the midfoot, which may potentially reactivate a
neurotrophic breakdown.
Incisional approach/exposure
The modified transfibular approach, originally described by Adams [33], is
performed through two incisions. The lateral incision begins over the lateral
aspect of the fibula at the distal third of the leg. It courses distally over the fibula
to the level of the sinus tarsi. Dissection is carried down to the level of the
periosteum overlying the fibula. The cutaneous portion of the superficial peroneal
nerve exits the crural fascia at variable levels and must be anticipated in the
76 S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96
Joint preparation
Before joint resection, all loose osteochondral fragments, periarticular ossicles,
and exuberant synovitis should be excised to enhance visualization and allow for
a greater appreciation of the joint anatomy and borders. Joint surfaces may be
prepared by either planal resection or curettage and decortication. The authors
prefer to reserve planal resection for severe multiplanar deformities that require
significant repositioning and realignment. Curettage and decortication is per-
formed using large curettes to remove the cartilage from the opposing surfaces of
the tibial plafond and talar dome, including the medial talar facet and its
corresponding articulation on the tibia. Curettage and decortication allows for
minor angular adjustments while maintaining the ‘‘anatomic’’ concave-convex
contours. This construct affords greater rotational stability and allows for
continual intraoperative repositioning and adjustment. This technique also max-
imizes limb length in patients who may already have preexisting limb-length
discrepancies secondary to posttraumatic conditions. For greater visualization of
the joint, a lamina spreader may be sequentially placed in the anterior and
posterior joint margins to facilitate visualization preparation of the fusion sites. If
a half-pin fixator is chosen as the primary form of fixation, then the fixator may
be applied before the dissection and used for joint distraction purposes before
final reduction and compression of the arthrodesis.
S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96 77
Frame construct
The subtalar joint may be preserved in those patients who possess supple joint
motion without coexisting pathology. Joint preservation allows for greater
compensatory motion postfusion and can be accomplished by the addition of a
transverse plane ring at the level of the talus (Fig. 5). The talus is transfixed to the
ring by counterposed olive wires, effectively suspending it relative to the
calcaneus. This permits isolated compression at the tibiotalar joint level.
Compression occurs between the tibial ring block and the talar ring/foot plate
with preservation of the subtalar complex (Fig. 6). Alternatively, two wires may
be posted from the foot plate transfixing the talar body to create a ‘‘virtual ring’’
[22]. This may be another effective method of preventing compression and
arthrofibrosis at the level of the subtalar joint (Fig. 7).
Following dissection and joint resection, the arthrodesis site is then reduced,
positioned, and temporarily fixated with a 5/6400 Steinmann pin in a retrograde
fashion through the heel. The pin may be inserted through the heel pad to the
lateral side of the calcaneus and ‘‘walked’’ medially on the plantar cortex to
dislodge the plantar lateral neurovascular bundle and avoid injuring this structure
[36]. The alignment of the fusion is critical and is continually reassessed. The
relationship of the tibial tubercle to the second digit on the contralateral limb
should be noted during the preoperative evaluation and symmetrical rotation
should be achieved intraoperatively. Overall position and apposition of the
arthrodesis should be confirmed clinically and with an image intensifier. Any
78 S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96
Fig. 5. Tibiotalar arthrodesis. A 55-year-old man with posttraumatic ankle deformity requiring
subsequent ankle fusion. (A, B) Anteroposterior and lateral radiographs depicting varus mal-
alignment and arthrosis of the ankle joint. (C) Postoperative lateral radiograph demonstrating
Ilizarov technique for an isolated ankle arthrodesis. The talus is suspended and the subtalar joint is
well preserved by the addition of a transverse plane talar ring. Compression is achieved between the
talar ring and the tibial ring block. (D) Ilizarov frame for ankle arthrodesis.
defects may be filled with the autogenous bone to maximize contact of the fusion
surfaces. The tourniquet is released and the wounds are closed in layers over
drains. If the fibula is to be used as an onlay graft, it should be debulked and
applied with the decorticated surfaces against the talus and tibia. This may be
secured to the tibia with cannulated 4.0-mm screws. Fixation to the talus and tibia
will block compression following application of the frame.
S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96 79
Fig. 6. Compression arthrodesis constructs. (A) Ankle arthrodesis. Fixation blocks consist of a tibial
segment (proximal two rings) and a rearfoot block (distal two rings). The talus and calcaneus are
compressed proximally as a unit to avoid subtalar joint compression. Separate threaded rod assemblies
join the central two rings to effect compression at the ankle joint. Note the foot plate above the
weightbearing plane of the foot. (B) Pantalar arthrodesis. The tibial fixation block and the foot plate
neutralize forces across the leg and foot lever arms, respectively, while compression occurs
simultaneously between these segments across the ankle and subtalar joints. (C) Subtalar arthrodesis.
This is a variant of the frame assembly shown in (A). The tibial ring block is used and again is
attached to the foot plate; however, compression is achieved through an arched wire technique across
the subtalar joint posted off of the foot plate. This compresses the subtalar joint while maintaining
position of the ankle joint. The posted wire replaces the talar ring in the ankle fusion assembly.
Frame application
The preassembled frame should be placed over the limb and any final
adjustments made. The first two wires are the most critical to align the long
axes of the foot and leg with the fixator. The mounting sequence of the
transfixion wires is outlined as follows:
Fig. 7. Tibiotalar arthrodesis with ‘‘virtual talar ring.’’ A 63-year-old woman with complaint of pain
and disability for 15 years following an untreated ankle fracture. (A,B) Preoperative anteroposterior
(AP) and radiographs depicting significant arthritic change and varus deformity of the ankle joint.
There is adaptation and widening of the talar dome and tibial plafond with complete obliteration of the
joint space. (C) Postoperative AP radiograph showing reduction of varus alignment and good bony
apposition of the arthrodesis site. Note the fibular onlay graft secured to the tibia with screw fixation.
(D) Clinical appearance of the ankle and external fixator on the second postoperative day. Again,
offset sagittal plane half-pins have been placed into the medial face of the tibia achieving three points
of fixation per ring.
S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96 81
Fig. 8. Tibiotalocalcaneal fusion. A 36-year-old man presents with distal tibial pathology resulting
from a high-energy pilon fracture secondary to a fall from a 30-foot height during a suicide attempt.
Secondary involvement of the subtalar joint required combined tibiotalar and talocalaneal fusions.
(A,B) Anteroposterior (AP) and lateral radiographs. Status post-ORIF with valgus and procurvatum
deformity and severe end-stage ankle arthrosis. (C,D) Postoperative anteroposterior and lateral
radiographs demonstrating ankle and subtalar arthrodesis with Ilizarov technique. An EBI spinal
stimulator has been used to enhance fusion. High-energy injuries increase the potential for nonunion
because of the periosteal degloving and bony devascularization associated with these fractures. Note
the wire deformation at the foot plate level and the associated ‘‘pin drag.’’ Two offset 6-mm half-pins
have been added to increase frame stiffness.
82 S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96
6. Talar ring
(a) Opposing olive wires in anterior and posterior talar body
7. Foot plate
(a) Opposing olive forefoot wire from medial to lateral
(b) Forefoot, midfoot wires are important to eliminate the lever arm acting
against the ankle, which may create sagittal plane ‘‘rocking’’
8. Talar ring (virtual ring, not necessary with talar ring application)
(a) Two wires in talus
(b) Ring or drop wire off posts
9. Additional half-pins or wires as needed
Fig. 9. Tibiotalocalcaneal fusion. Severe fixed equinovarus deformity resulting from delayed
recognition of a compartment syndrome secondary to a gunshot wound. A pantalar fusion was
required to attain a rectus foot and ankle. (A,B) Fixed equinovarus deformity. Lateral radiograph
demonstrating severe cavus deformity with fixed ankle plantarflexion. (C) Anteroposterior (AP)
radiograph demonstrating moderate arthritic change along the lateral column and proximal midfoot
with healed stress fractures from altered distribution of weightbearing forces. Long-standing adaptive
changes will require repositional midfoot arthrodesis to achieve a plantigrade foot. (D) Oblique
postoperative ankle radiograph of a pan talar arthrodesis. A 7.3-mm cannulated screw provides
supplemental fixation of the subtalar complex. The ankle fusion is well reduced and compressed with
a fibular onlay graft acting as a lateral buttress. (E) Postoperative AP clinical photograph
demonstrating weightbearing in Encore fixator with Dynasplint device.mbf
S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96 83
Fig. 9 (continued).
84 S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96
Half-pin fixator
Half-pin uniplanar/biplanar devices have gained popularity because of their
relative ease of application and simplicity compared with circular wire fixators
(Fig. 10). EBI1, Parsippany, NJ and Orthofix1, McKinney, TX have developed
Fig. 10. Oblique and posterior photographs (A,B) of the EBI Dynafix1 unilateral half-pin fixator
technique for ankle arthrodesis. Pins are placed in the posterior talar body and talar neck employing the
ankle T-clamp.
S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96 85
The diameter of the pins should not exceed one third of the
diameter of the bone.
The fixator may be applied before the surgical dissection and used to distract
the ankle joint to aid in joint preparation. If this is anticipated, the compression/
distraction module should be maintained at the midway point, which allows
2.5 cm of compression/distraction.
Tibiocalcaneal arthrodesis
Tibiocalcaneal arthrodesis is often necessary when there is destruction of the
talus resulting from neuroarthropathy, avascular necrosis, infection, or severe
deformity (Figs. 11, 12). Talectomy or loss of talar body height in combination
with preparation of the fusion sites on the tibia and calcaneus often leads to a 4-cm
loss of limb length [21]. As a result, many surgeons prefer to combine these fusion
sites with a proximal tibial lengthening and simultaneous compression at the ankle
level using the Ilizarov technique [21,28]. Exposure is accomplished through a
lateral hockey stick-type incision over the fibula proximally and extending to the
level of the cuboid distally. If the talar body is to be extracted, a transverse incision
centered approximately 3 to 5 cm above the tip of the lateral malleolus may be
employed to ensure adequate closure without wound ‘‘puckering’’ from resultant
limb shortening [21]. If the fibula exists, it is excised and morcelized in a bone mill
to augment any defects at the fusion site. The remaining portions of the talar body
are then removed, and the area should be debrided of all fibroligamentous tissues
and loose bodies to allow flush contact between the tibia and calcaneus. If the talar
head and neck are viable, they may be left intact to avoid harvesting additional
graft. Planal resection is now performed on the tibial plateau and dorsal calcaneal
surfaces. Resection may be aided by the insertion of two orthogonal guide wires in
the distal tibia perpendicular to the long axis of the tibia. Additional wires may be
inserted into the calcaneus parallel to the sole of the foot. Saw cuts may now be
made parallel to these wires to ensure congruous fusion surfaces [21]. The lateral
surface of the medial malleolus may also be decorticated to allow fusion, or the
entire malleolus may be resected if it hinders reduction of the primary fusion site.
The anterior surface of the distal tibia and talar neck may be prepared in a similar
fashion. In prior sites of infection, a ‘‘slurry’’ of fresh-frozen irradiated cancellous
chips, autologous platelet gel, and vancomycin/tobramycin powder may be used
to optimize surgical results. Iliac crest bone grafts may be harvested as needed but
should be avoided if possible to decrease postoperative morbidity. Provisional
fixation is achieved in a likewise fashion to the tibiotalar arthrodesis, and the
wound is closed in layers over a drain.
If concomitant leg lengthening is performed, a retrograde nail may be inserted
through the arthrodesis site to the proximal tibial metaphysis and locked distally
into the tibia and calcaneus. A proximal tibial corticotomy is then performed and
the Ilizarov device is applied spanning the corticotomy and arthrodesis sites for the
distraction phase. When the desired length is achieved, the nail may be locked
proximally and the fixator removed for the consolidation phase, thus reducing the
time in the fixator [38]. Proximal lengthening without associated intramedullary
S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96 87
Fig. 11. Tibiocalcaneal arthrodesis. Tibiocalcaneal fusion Ilizarov technique with concomitant medial
and lateral column stabilizations in a 48-year-old noninsulin diabetic who had deformity and chronic
ulcerations 1 year before presentation. (A) Preoperative anteroposterior (AP) weightbearing
photograph showing Charcot breakdown with secondary varus deformity of the hindfoot and
increased load bearing along the lateral forefoot. (B) Charcot neuroarthropathy involving the subtalar
and ankle joints with dissolution of the talar body. (C) Lateral radiograph of tibiocalcalneal fusion and
‘‘nailing’’ of the medial and lateral columns with large cannulated screws. Iliac crest grafts were
needed to supplement fusion at the tibiocalcaneal and tibionavicular regions. (D) Lateral photograph
of Encore fixator for tibiocalcaneal fusion.
88 S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96
Fig. 12. Tibiocalcaneal arthrodesis using a Taylor spatial frame. A 45-year-old woman with a history
of spina bifida and the secondary development neuropathic osteoarthopathy of the ankle and hindfoot
requiring tibiotalocalcaneal arthrodesis. (A) Preoperative weigthbearing photograph. Note the severe
rotary deformity of the foot and leg. Patient underwent a symes amputation on the contralateral limb
from previous complications related to her medical condition. (B,C) Anteroposterior and lateral
weightbearing radiograph demonstrating complete collapse with tibiofibular diastasis and ankle
valgus. There is also significant subtalar and midtarsal joint involvement with the development of a
‘‘rockerbottom’’ deformity. (D) Axial CT through the posterior ankle and hindfoot articulations. End-
stage arthrosis at the ankle and subtalar levels. (E) Postoperative photograph of tibiotalocalcaneal
arthrodesis using a Taylor Spatial Frame. Restoration of the rotary alignment and frontal plane
hindfoot position.
S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96 89
Fig. 12 (continued).
nailing will significantly increase the treatment period in the fixator as the
consolidation phase is generally twice the time required for the lengthening alone.
Triple arthrodesis
Conditions such as rheumatoid arthritis or posttraumatic deformitites such as
calcaneal fractures require triple arthrodesis to relieve pain and create stability in
the hindfoot. This procedure is performed through a two-incisional approach. The
medial incision between the tibialis posterior and tibialis anterior tendons
provides access to the talonavicular joint; the lateral incision is placed superior
to the peroneal tendons overlying the floor of the sinus tarsi. Planal saw resection
or minimal resection techniques may be employed with care taken to respect the
blood supply to the talus, especially the vessels penetrating the dorsal neck
90 S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96
region. Following removal of the articular cartilage, temporary fixation with 5/6400
Steinmann pins is employed. Several relationships must be observed, including the
position of the lateral talar process and congruency of the posterior subtalar joint
facet, talar coverage, the position of the calcaneus to leg, and forefoot to rearfoot
relationship. The ideal alignment of the arthrodesis is slight hindfoot valgus
(3 –5°), talonavicular congruity, and a valgus attitude of the forefoot to rearfoot.
Failure to accomplish the latter will result in compensatory pronation through the
ankle joint postoperatively with resulting deltoid failure and valgus deformity of
the mortise. The Steinmann pins are then bent and locked into the adjacent cortices,
and the wounds closed in layers over drains.
The preconstructed frame, consisting of one or two tibial rings and a foot
plate, is stabilized to the tibia and calcaneus using the standard mounting
sequence. Under fluoscopic guidance, a 1.8-mm smooth wire is placed through
the talar neck/body and arched or bent backward to create a perpendicular
relationship between the posterior facet and the vector of wire pull [39]. Either
ends of the wire are then stabilized to the foot plate and tensioned. This ‘‘arched
wire’’ compression technique is illustrated in Fig. 13. A similar method is
employed for arthrodesis of the midtarsal region (Fig. 14). Additional wires are
added as needed.
In ankle and tibiocalcaneal applications, compression is achieved in a ring-to-
ring fashion or decreasing the distance between foot and leg rings. Triple
arthrodesis is unique in that it relies on wire-to-ring compression (‘‘arched
wire’’) and may be supplemented with internal fixation (see Fig. 6). This method
has a high learning curve and is not recommended for those who have minimal
experience with ring fixators.
Distraction arthrodiastasis
Joint distraction is the most recent joint-sparing advancement in the treatment
of osteoarthritis (OA) and arthrofibrosis in the lower extremity. Van Valburg et al
[40,41] demonstrated that Ilizarov joint distraction delays and may obviate the
Fig. 13. ‘‘Arch’’ wire compresssion technique. (A) A transosseous wire placed through the talar body
and deflected or ‘‘arched’’ perpendicular to subtalar joint. The wire ends are then affixed to foot plate.
(B) Talar wire is then placed under tension which encourages the wire to assume a linear configuration
and compress the subtalar joint. This technique may also be used for midfoot fusion.
S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96 91
Fig. 14. Triple arthrodesis. (A) Lateral radiograph of failed adult flatfoot reconstruction with an
arthroeresis device. Significant arthritic changes have occurred within the subtalar and talonavicular
articulations. (B,C) Postoperative radiographs of a triple arthrodesis demonstrating internal splintage
of the talonavicular and calcaneo-cuboid joints with arch wire compression across these segments
encouraging fusion. This technique is also employed at the level of the subtalar joint.
need for arthrodesis in patients with severe ankle OA. In a preliminary 2-year
follow-up of 11 patients, they reported symptomatic relief and stimulation of
articular cartilage repair as evidenced by an increased joint space. Further
investigation of the effects of this technique on canine cartilage showed
normalization of chondrocyte function and improved changes in cartilage
metabolism [42]. At least superficially, this joint preservation technique seems
to have some promise and may play a crucial role in the treatment of OA,
especially in the younger active patient.
Arthrodiastasis may be accomplished through the application of two supra-
malleolar tibial rings and a foot plate. Telescoping rods are used as longitudinal
92 S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96
Complications
Complications resulting from external fixators, primarily fine wire ring
fixators, may be divided into three categories: major, minor, and permanent
[22]. Major complications are those that require operative treatment to resolve
and can compromise the final outcome if not addressed appropriately (ie,
nonunion at the docking site). Dahl and Valezquez [43,44] noted that major
complications decrease with increasing surgeon experience. Minor complications
are the most common and may be resolved nonoperatively with little difficulty
(ie, simple pin tract infection). Those complications that cannot be alleviated and
often preclude the original goals of treatment are termed permanent. These
complications may arise intraoperatively, postoperatively, or following removal
of the frame.
Intraoperative complications are most often related to placement of smooth
wires through myofascial compartments with direct injury to neurovascular
structures. Complications of this nature are directly proportional to the surgeon’s
understanding of cross-sectional anatomy and technique of wire insertion. Wire
penetration should begin on the side of the limb away from the SAR (neuro-
vascular bundle), and the wire should be tapped through the myofascial
compartment with a mallet to avoid ‘‘wrapping’’ of the neurovascular structures.
Vascular lesions occur from direct injury to arterial structures, though these
lesions rarely lead to problems because of the small diameter of the wires [22]. In
the presence of an acute lesion intraoperatively, the wire is removed and direct
pressure is applied to the vessel. Vascular disruption may also arise from a tibial
corticotomy or fibular osteotomy/resection. Compartment syndrome can occur
postoperatively from vascular penetration, and the affected compartments should
be monitored directly with compartment pressure measurements. Passive stretch
of the myotendinous structures is generally unreliable in the presence of wires
S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96 93
Table 1
Dahl’s classification for pin tract complications
Grade Appearance Treatment
0 Clear NaCl and Bactroban
1 Slightly red NaCl and Bactroban
2 Red/tender yellow drainage possible PO abx with TID pin care
3 Red/painful/purulent drainage Definite PO abx
4 Radiolucency in combination with purulence Removal of pin, possible IV abx
5 Sequestrum Removal of pins/debridement of
pin tract with intravenous abx
PO, oral; abx, antibiotics; TID, three times daily.
Data from Sontich JK. Essentials of Ilizarov. Comprehensive Ilizarov solutions. Atlanta, Georgia,
April 29, 2002.
94 S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96
technique or wire tensioner if possible. Pin tract infections may range from clear
serous drainage to osteomyelitis requiring pin removal and surgical debridement.
Dahl devised a classification and treatment protocol for pin tract infections
(Table 1).
Additional complications such as joint contractures or joint luxation, wire or
component failure, and difficulties related to limb lengthening (eg, delayed or
premature consolidation, poor regenerate formation, and angular deviations) may
be encountered during treatment.
Frame removal
The appropriate time for frame removal depends on when consolidation of
bone or the arthrodesis site occurred radiographically and clinically. The fixator
may be removed in piecemeal fashion over several weeks, beginning with wires
and threaded rods, to allow gradual axial micromotion on the stress-shielded
bones. Delayed bony union may be accelerated by dynamization of the frame
through a sequential ‘‘destabilization’’ of the rings and wires. The frame may also
be removed as a unit followed by a period of strict nonweightbearing in a short
leg cast. Pin and wire sites require local wound care and observation until skin
closure occurs. Frame removal is typically performed in the operating room or
patient holding area under mild sedation.
References
[1] Malgaigne JG. Considerations cliniques sur les fractures de la rotule et leur traitement par les
griffes. J Connaissance Md Pratiques 1853;16(9).
[2] Parkhill C. A new apparatus for the fixation of bone after resection and in fractures with
tendency to displacement. Trans Am Surg Assoc 1897;15:251.
[3] Lambotte A. Le traitment des fractures. Paris: Masson; 1907.
[4] Anderson R. An automatic method for treatment of fractures of the tibia and fibula. Surg
Gynecol Obstet 1934;58:639.
[5] Hoffmann R. Rotules a os pour la reduction dirigee, non saglante, des fractures (osteotaxis). Helv
Med Acta 1938;6:844.
[6] Charnley J. Compression arthrodesis of the ankle and shoulder. J Bone Joint Surg Br
1951;33:180 – 91.
[7] Charnley J. Compression arthrodesis of the knee. J Bone Joint Surg Br 1948;30:478.
[8] Kenzora JE, Simmons SC, Burgess AR, et al. External fixation arthrodesis of the ankle joint
following trauma. Foot Ankle Int 1986;7(1):49 – 61.
[9] Rothacker GW, Cabanela ME. External fixation for arthrodesis of the foot and ankle. Clin
Orthop 1983;180:101 – 8.
[10] Williams Jr JE, Marcinko DE, Lazerson A, et al. The Calandruccio triangular compression
device: a schematic introduction. J Am Podiatr Assoc 1983;10:536.
[11] Ilizarov GA. New principles of osteosynthesis by means of crossing pins and rings. Kurgan,
Russia: Internal Publication Book; 1954.
[12] Gasser B, Bowman B, Wyder D, et al. Stiffness characteristics of the circular Ilizarov external
fixator device as opposed to conventional external fixators. J Biomech Eng 1990;112:15.
[13] Fleming B, Paley D, Kristiansen T, et al. A biomechanical analysis of the Ilizarov external
fixator. Clin Orthop 1989;241:95.
S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96 95
[14] Kummer FJ. Biomechanics of the Ilizarov external fixator. Bull Hosp Jt Dis Orthop Inst
1989;49:140.
[15] Paley D, Fleming B, Catagni M, et al. Mechanical evaluation of external fixators used in limb
lengthening. Clin Orthop 1990;250:50.
[16] Podolsky A, Chao E. Biomechanical performance of Ilizarov external fixators. Trans Orthop Res
1990;15:416.
[17] Kershaw C, Cunningham J, Kenwright J. Tibial external fixation, weight bearing and fracture
movement. Clin Orthop 1993;293:28 – 36.
[18] Kenwright J, Goodship A. Controlled mechanical stimulation in the treatment of tibial fractures.
Clin Orthop 1989;241:36 – 47.
[19] Cierny III G, Cook WG, Mader JT. Ankle arthrodesis in the presence of ongoing sepsis:
indications, methods, results. Orthop Clin North Am 1989;20:709.
[20] Fischer DA. Skeletal stabilization with a multiplane external fixation device: design rationale and
preliminary clinical experience. Clin Orthop 1983;180:50 – 62.
[21] Paley D, Herzenberg JE. Applications of external to foot and ankle reconstruction. In: Myerson
MS, editor. Foot and ankle disorders. Philadelphia: WB Saunders; 1999. p. 1135 – 88.
[22] EBI1. Essentials of Ilizarov and comprehensive Ilizarov solutions. Parsippany, NJ 2002.
[23] Behrens F. General theory and principles of external fixation. Clin Orthop 1989;241:15 – 23.
[24] Catagni MA. Atlas for the insertion of transosseous wires and half-pins: Ilizarov method. In:
Maiocchi AB, editor. Memphis (TN): Smith and Nephew; 2002.
[25] Calhoun JH, Li F, Bauford WL, et al. Rigidity of half-pins for the Ilizarov external fixator. Bull
Hosp Jt Dis 1992;52(1):21 – 6.
[26] EBI. Product brochure. Parsippany, NJ: EBI, 2002.
[27] Hammershlag WA. Ankle arthrodesis using a ring external fixator. Tech Orthop 1996;
11(3):263 – 8.
[28] Hawkins BJ, Langerman RJ, Anger DM, et al. The Ilizarov technique in ankle fusion. Clin
Orthop 1994;303:217 – 24.
[29] Laughlin RT, Calhoun JH. Ring fixators for reconstruction of traumatic disorders of the foot and
ankle. Orthop Clin North Am 1995;26:287 – 94.
[30] Kitaoka HB, Anderson PJ, Morrey BF. Revision of ankle arthrodesis with external fixation for
non-union. J Bone Joint Surg Am 1992;74:1191 – 200.
[31] Glissan DJ. The indications for inducing fusion at the ankle joint by operation with description of
two successful techniques. Aust N Z J Surg 1949;19:64 – 71.
[32] Buck P, Morrey BF, Chao EY. The optimum position of the arthrodesis of the ankle: a gait study
of the knee and ankle. J Bone Joint Surg Am 1987;69:1052 – 62.
[33] Adams JC. Arthrodesis of the ankle joint: experiences with the transfibular approach. J Bone
Joint Surg Am 1948;30:506.
[34] Peterson L, Goldie I, Lindell D. The arterial supply of the talus. Acta Orthop Scand
1974;45:260 – 70.
[35] Paley D. Ankle malalignment. In: Kelikian AS, editor. Operative treatment of the foot and ankle.
Stamford (CT): Appleton and Lange; 1999. p. 547 – 86.
[36] Pochantko DJ, Smith JW, Philips RA, et al. Anatomic structures at risk: combined subtalar and
ankle arthrodesis with a retrograde intramedullary rod. Foot Ankle Int 1995;16:542.
[37] Nepola JV. DFS1 Standard Fixator, DFS1 Ankle Clamp, DFS1 T-Clamp: surgical technique.
EBI Fracture Management. Product brochure. EBI1, Parsippany, NJ, 2002.
[38] Herzenberg JE, Paley D. Tibial lengthening over nails (LON). Tech Orthop 1997;12(4):250 – 9.
[39] LaBianco GJ, Rush SM, Vito GR. External fixation. In: Banks AS, Downey MS, Martin DE,
Miller SJ, editors. McGlamry’s comprehensive textbook of foot and ankle surgery. 3rd edition.
Philadelphia: Lippincott Williams and Wilkins; 2001. p. 107 – 38.
[40] Van Valburg AA, van Roermund PM, Lammens J, et al. Joint distraction in the treatment of
osteoarthritis: a two-year follow-up of the ankle. Osteoarthritis Cartilage 1999;7:474 – 9.
[41] Van Valburg AA, van Roermund PM, Lammens J, et al. Can Ilizarov joint distraction delay the
need for an arthrodesis of the ankle? A preliminary report. J Bone Joint Surg Am 1995;77:720 – 5.
96 S. Kalish et al / Clin Podiatr Med Surg 20 (2003) 65–96
[42] Van Valburg AA, van Roermund PM, Marijnissen AC, et al. Joint distraction in the treatment of
osteoarthritis (II): effects on cartilage in a canine model. Osteoarthritis Cartilage 2000;8:1 – 8.
[43] Dahl MT, Gulli B, Berg T. Complications of limb lengthening, a learning curve. Clin Orthop
1994;301: 10 – 8.
[44] Valezquez RJ, et al. Complications of use of the Ilizarov technique in the correction of limb
deformities in children. J Bone Joint Surg Am 1993;75:1148 – 56.
[45] Polak WG, Pawlowski S, Skora J, et al. Vascular complications after the treatment with Ilizarov
external fixators. Vasa 2001;30(2):138 – 40.
[46] Slomka R. Complications of ring fixators in the foot and ankle. Clin Orthop 2001;391:115 – 22.
[47] Makrov M, Birch JG, Delgado MR, et al. Effects of external fixation and limb lengthening on
peripheral nerve function. Clin Orthop 1996;329:310 – 6.