Ankle arthrodesis remains an important procedure in the salvage of the painfully arthritic ankle joint, enabling the surgeon to create a painless, stable, and plantigrade foot. Surgical alternatives for the treatment of ankle joint arthritis, such as total ankle replacement, corrective supramalleolar osteotomy, and distraction joint arthroplasty, have improved over the past 10 years. These surgical alternatives have focused on changing the distribution of loads on the ankle joint, improving the biomechanics of the lower extremity, or replacing the joint. These techniques have gained increasing popularity as a result of long-term consequences and complications of ankle arthrodesis; in particular, the markedly increased loading of the adjacent hindfoot joints. For many surgeons, however, ankle arthrodesis remains the treatment of choice. 1
The operative techniques for reliably obtaining an arthrodesis have steadily improved over the past several decades and now include more than 20 different techniques. The basic surgical principles for a successful ankle arthrodesis were identified as broad, congruent cancellous bleeding surfaces that are stabilized with rigid internal fixation under compression after correct positioning of the foot in the coronal, transverse, and sagittal planes. 4
Despite the advances over time in instrumentation and operative approach, problems with ankle arthrodesis persisted. Problems with delayed union, particularly nonunion, 7,10 led to the evolution of techniques involving minimal exposure, such as arthroscopically assisted ankle arthrodesis and the miniarthrotomy procedure. These techniques provided minimal biologic disruption with a subsequently excellent and rapid rate of fusion with a reduced recovery period. 6,8,9,11,12,14,15 Initially, these techniques were not suitable to address complex deformities and were limited to more basic arthrodeses. The arthroscopic method was demanding, had a steep learning curve, and was far more time-consuming than most other procedures. 5,12,14
Encouraged by the advantages of arthroscopically assisted ankle arthrodesis, the senior author slightly enlarged the two anterior arthroscopic portals to obtain full open visualization of the joint. This led to the gradual evolution of the miniarthrotomy ankle arthrodesis technique, with the benefits of decreased soft tissue injury, markedly decreased periosteal stripping, and a rapid healing time comparable to that of arthroscopic procedures. 8,9,15 In addition, the procedure was relatively easy to learn and to perform.
We hypothesized that the miniarthrotomy arthrodesis technique allowed improved healing because it avoided the disruption of small blood vessels found in other ankle arthrodesis procedures. We conducted a laboratory study to identify the impact of exposure of the ankle on vascular disruption using digital subtraction angiography. 9 Eight pairs of below-knee specimens were obtained from eight cadavers. In each leg the popliteal artery was cannulated and a radiopaque mixture was infused. An ankle arthrodesis was then performed on one leg of each pair via an open arthrotomy and partial fibulectomy and on the other leg via a miniarthrotomy. Preoperative and postoperative radiographs were scrutinized for vascular damage. In two of the eight ankles in the open arthrodesis group, there was substantial vascular damage and disruption of the peroneal artery. In addition, there was substantial loss of the periarticular vessels in all eight of the open arthrodesis group. 9
Early clinical results of the miniarthrotomy arthrodesis technique were comparable to results with open and arthroscopic procedures. Early successes and the technical demands of the arthroscopic procedure led the senior author to fully adopt the miniarthrotomy technique for ankle arthrodesis.
A variety of surgical alternatives to ankle arthrodesis that address the arthritic ankle joint include supramalleolar osteotomy, total ankle replacement, distraction joint arthroplasty, and anterior cheilectomy. 1 The choice of surgical technique depends on a variety of factors including the patient's age, level of activity, and general medical status. The position of the foot, the degree of deformity, the neurovascular status of the limb, bone quality, overall limb alignment, and the presence of degeneration in the joints of the foot are also important factors. The severely destroyed joint with obliteration of the articular surface can be addressed only by arthrodesis or total ankle replacement and not by the other procedures. Total ankle replacement has specific indications. The ideal candidate for total ankle replacement is the person with good bone quality, minimal deformity, and reasonable weight who places low demand on the lower limb. 2
When we initiated the miniarthrotomy in 1992, there were very few limitations to the technique. In our early series, 9,15 the indications for this procedure were 1) severe arthritis with marked joint obliteration; 2) considerable ankle pain interfering with daily activities and walking ability; 3) failure of conservative treatment including nonsteroidal anti-inflammatory drugs (NSAIDs), intra-articular steroid injections, physical therapy, or use of and ankle-foot orthoses; 4) absence of mechanical malalignment above the ankle joint; 5) a minimally deformed ankle joint with varus or valgus less than 10°; 6) less than 25% posterior or anterior subluxation; 7) avascular necrosis (AVN) of the talus involving less than 25% of its articular surface; 8) articular surface cavitations smaller than 1 × 2 cm; and 9) absence of severe neuroarthropathy. These indications were similar to those described for an arthroscopically assisted in situ fusion.
After further experience with the miniarthrotomy technique, we recommend that ankle joints with moderate deformity be approached with this technique by resecting minimal wedges of bone using flexible chisels and appropriate size osteotomes. The few contraindications to performing this procedure include severe deformities amenable to correction only by angular and rotational realignment and malalignment above or below the ankle joint. We have used the miniarthrotomy with AVN of either the distal tibia or the talus and where large bone defects of the distal tibia are present. The limited skin incisions make this procedure appealing even in the presence of dermatologic, diabetic, autoimmune, and vascular conditions, and previous surgical incisions that may be contraindications to more extensive surgical procedures. Currently, we use the miniarthrotomy procedure almost exclusively. More extensive procedures are used to correct massive bone loss requiring structural bone graft.
Preoperative evaluation and planning includes careful clinical assessment and various diagnostic studies. When evaluating a patient with a painful ankle joint, the clinician should assess the ankle as well as the adjacent foot joints for range of motion, overall alignment, and pain and tenderness on palpation. A manual motor test of the major muscular groups about the foot and ankle is performed because any weakness may affect the position of the arthrodesis. The degree of ankle joint deformity in all planes (with the patient sitting, standing, and walking) is carefully evaluated, and the adjacent joints are also evaluated because the ankle arthrodesis places increased stresses on them. Any preexisting stiffness or arthritis in these joints will likely deteriorate over time and must therefore be evaluated separately with diagnostic lidocaine blocks. This evaluation may be particularly helpful in planning for a more extensive ankle and hindfoot arthrodesis. Reduced sensation or vascularity, skin changes, and the presence of venous stasis should be evaluated further before surgery is performed.
Preoperatively, full length, weightbearing, bilateral radiographs are obtained to determine overall limb alignment, the extent of arthritis, deformity, bone defects in the distal tibia or talus, presence and degree of AVN in either the talar body or the distal tibia, and the presence of arthritis in the adjacent foot joints. A computed tomogram is done to evaluate the extent of arthritis of the subtalar joint and any cavitation or defects in the distal tibia or talus. Whereas these do not present a contraindication to the miniarthrotomy procedure, large defects must be filled with autograft or allograft. If AVN is noted, an MRI should be obtained to further determine the extent of the disease. If AVN is extensive, the miniarthrotomy procedure is not contraindicated, but healing may be compromised and internal or external bone stimulation must be planned for.
Finally, the most important step is to determine the proper position of the arthrodesis site. Although the optimal position for an ankle arthrodesis has been debated, there is a consensus that the ankle should be in neutral position in the sagittal plane, minimal valgus (up to 5°), and external rotation symmetric with the contralateral uninvolved side (usually about 5° to 10°). 3,4 Malunion of the arthrodesis in the sagittal plane is a problem because a plantarflexed or equinus arthrodesis results in a dynamic genu recurvatum, a vaulting type of gait pattern, and symptomatic transverse tarsal arthritis. A dorsal malunion produces increased stresses on the calcaneus during the heel phase of gait, which may lead to heel pain and diminished push-off strength. The calcaneus malunion position is very difficult to treat, and even with appropriate heel support and cushioning further surgery may be necessary. A valgus malunion leads to a pronated flatfoot and increased stresses along the medial side of the knee and hindfoot joints, while a varus malposition leads to increased stresses along the lateral aspect of the foot, especially under the fifth metatarsal. Increased internal or external rotation of the fused ankle results in abnormal rotation of the lower extremity during gait with painful foot and ipsilateral hip and knee joints.
Because most ankle fusions do not require manipulation for correction, positioning of the ankle joint is not difficult, as there is little preoperative deformity. In arthroscopic and miniarthrotomy procedures, studies considered whether there was a need to translate the talus posteriorly with respect to the tibia. 3,10 Theoretically, this translation reduces the anterior lever arm of the foot on the fusion site and improves its ground clearance during gait. While this is anatomically correct, we have never found that it is necessary to perform such a translation, and we have not noted any awkwardness or difficulty with gait as a result of the in situ position of the ankle.
When correcting a moderate deformity, the position of the arthrodesis becomes critical in the attempt to produce a plantigrade foot. At our institution, the positioning is based upon the knee-foot-ground axis as determined intraoperatively. Any other pathology of the lower extremity above the arthrodesis site is taken into account. For example, when there is a quadriceps weakness, the ankle should preferably be positioned in 5° to 10° of plantarflexion. Because the plantarflexed fused ankle leads to genu recurvatum, the knee joint is stabilized with the deficient extensor mechanism. Finally, the alignment of the foot distal to the fusion site must be carefully assessed and taken into account to create a plantigrade foot. A forefoot deformity dictates a compromised position of the ankle joint arthrodesis or other procedures, such as osteotomies. For example, in the presence of fixed midfoot or forefoot equinus deformity, it is preferable to position the ankle in slight dorsiflexion to accommodate the deformity.
The ankle arthrodesis using the miniarthrotomy technique is performed as an outpatient procedure with the patient under regional ankle or popliteal block and intravenous sedation. 13 It is rare to use general anesthesia, although that is decided by the patient and the anesthesiologist. It is important to drape the leg well above the knee and to use the patella as a reference point for the final positioning of the ankle. Two 1.5 cm incisions are made, one anteromedial and one anterolateral, in approximately the same positions as the arthroscopic portals (Fig. 1).
The first incision is made just medial to the anterior tibial tendon, avoiding inadvertent injury of the saphenous vein and nerve. The second incision is performed immediately lateral to the peroneus tertius tendon, with more careful subcutaneous dissection to avoid injury to the dorsal cutaneous branch of the superficial peroneal nerve. The ankle retinaculum is identified and incised along the same line with the skin incision. The tendons are retracted to identify and open the joint capsule.
The key to this procedure is adequate visualization, which is facilitated by internal joint distraction with a lamina spreader and exposure of the anterior joint with a Weitlander retractor (Fig. 2). A rongeur is used to remove the hypertrophied synovial tissue from the anterior aspect of the joint, further improving the working space and visualization. Any visible cartilage of the anterior ankle joint is resected with curettes of various sizes and shapes and a set of small rongeurs (Fig. 3). Although it is not necessary, we remove most osteophytes from the anterior tibia with flexible chisels to further improve visualization.
A small lamina spreader is inserted into either the medial or the lateral joint space, and further debridement is performed. A rongeur can be used on one side of the joint to open the space when necessary and the lamina spreader is inserted on the opposite side of the joint. This process is alternated between medial and lateral incisions. It is important to position the lamina spreader correctly to avoid tilting the talus from the neutral position. With the joint distracted, a variety of instruments (rongeurs, curettes, and chisels) are used to debride any remnants of cartilage, synovial tissue, loose bodies, and sclerotic subchondral bone. Originally we used a high-speed burr, but this seemed to burn bone and was frequently followed by anterior joint synovitis during healing, probably as a result of the slurry created by the burr. We now predominantly use a flexible chisel until bleeding is uniform. The joint must be irrigated frequently to visualize the cancellous bone surfaces and confirm uniform bleeding.
Small bone wedges may be resected to obtain the ideal joint position, particularly when moderate deformity is present. After joint debridement, the ankle can often be moved and manipulated for correction (particularly in the coronal plane) without resorting to bone resection or wedges. We routinely use cancellous allograft chips to fill in any defects. Any dense sclerotic subchondral bone can be drilled with a 2.5 mm drill bit to enhance revascularization. The use of a drill bit is preferable to use of a Kirschner wire, which may compress the bone and lead to burning and sclerosis.
Any remaining cartilage on the lateral articular surfaces of the talus and the articular surfaces of the malleoli is then meticulously removed. It is not known to what extent the medial or lateral malleolus contributes to the fusion process, but the possible effectiveness of this contribution can be facilitated by debridement and filling the void in the gutter with cancellous graft. The miniarthrotomy technique does not allow perfect visualization and debridement of the most posterior portion of the ankle joint. In our experience, the opening and the lack of arthrodesis posteriorly does not adversely affect the success of fusion. This result seems reasonable, considering the effectiveness of ankle arthrodesis in the setting of AVN where a slotted graft is slid down the anterior aspect of the ankle only.
After debridement is complete, the joint is copiously irrigated and the extent of the joint preparation checked. The ankle is then positioned with guide pins from a cannulated, self-tapping screw system. We use three cannulated, 6.5-mm partially threaded cancellous titanium screws. The guide wires are placed in the following pattern: the first pin and screw are inserted from the posterolateral aspect of the tibia in an anteromedial direction into the talar head. The guide pin is inserted immediately adjacent to the Achilles tendon, approximately 3 cm proximal to the ankle joint (Fig. 4). The second pin and screw are inserted from the anteromedial aspect of the tibia directly above the medial malleolus distally and anteriorly toward the sinus tarsi (Fig. 5). The third guide pin is inserted from the lateral aspect of the joint anterior to the fibula and directed toward the medial talar neck. Occasionally, there is not enough space to insert the pin if there is no flare of the distal lateral tibia. In this case, the screw is inserted through the fibula into the talus. The positions of the guide wires and the screws are then checked under fluoroscopy. The screw from the medial malleolus should be carefully checked because of its proximity to the subtalar joint. The screw inserted from the posterolateral tibia is critical because it obtains the best purchase in the talus and is in the plane of the most direct line of compression across the joint. Because the screws are not introduced parallel to each other, eccentric loading of the arthrodesis site may occur as the first one is inserted. This can be avoided by alternately tightening each screw until compression is obtained. Once the screws have been inserted, the joint must be manipulated and the screws tightened or repositioned if any motion is present. Further bone graft can be inserted at this time.
After closure of the retinaculum, the skin is routinely closed without a drain. A bulky, cotton dressing is applied with a medial to lateral coaptation-type splint and posterior mold of plaster. The patient is given oral narcotics, NSAIDs, and oral antibiotics. If the operation was performed under regional ankle block, the patient is discharged the same day.
The authors have performed this procedure since 1992 and published the early results of the first 32 patients with the miniarthrotomy technique. 9 Of these 13 men and 19 women with an average age of 52.5 years (range, 29–76 ± 13.5 years), 27 had posttraumatic arthritis, one had rheumatoid arthritis, and four had degenerative arthritis. For three of the patients, the miniarthrotomy procedure was a revision of a prior arthrodesis attempt. The average follow-up for these patients was 23 months (range, 7–37 months). In 30 patients, a radiographic fusion was achieved at an average time of 8 weeks (range, 6–22 ± 4 weeks). Radiographic fusion was judged as the presence of good bone apposition without hardware loosening and osseous trabeculae across the ankle arthrodesis site. Postoperative radiographs demonstrated a fusion of the anterior three-quarters of the joint because of inadequate visualization and debridement of the most posterior portion of the ankle joint. We have found that the lack of bony fusion across the posterior portion of the ankle does not appear to be a clinical problem (Fig. 6).
Fusion is very difficult to determine radiographically. We use clinical and radiographic criteria to determine the onset of fusion, including the absence of pain with weightbearing and ambulation and, most importantly, cessation of warmth and swelling at the joint site.
In the original series of patients, 9 one patient had delayed union and two had nonunions. The patient with the delayed union was noncompliant with instructions and walked extensively immediately after surgery. Fusion finally occurred in this patient at 22 weeks without further intervention. The two patients with nonunion were both heavy smokers, and successful revision surgery using the miniarthrotomy technique was performed at 34 and 40 weeks after the original surgery. More recent results with this technique have been very similar to those previously reported. We have performed this procedure on 114 patients to date, with successful fusion in 105 patients at an average time to fusion of 10 weeks.
We have encountered only minimal complications using the miniarthrotomy technique, with complications related mostly to the burring technique initially used. Seven of 32 patients in our early series 9 experienced transient swelling and warmth limited to the anterior aspect of the ankle at 4 to 10 weeks after surgery. There was no concomitant pain and no signs of delayed healing or hardware loosening. All seven of the patients were treated conservatively with NSAIDs, and three required a local steroid injection. Inflammation resolved completely in all seven patients within 4 weeks. We attributed this complication to the leakage of bone slurry, generated by the pneumatic high-speed burr that was used in our early series to debride and roughen the subchondral bone. This problem has not occurred since we discontinued use of the high-speed burr.
At 2 weeks after surgery, the postoperative bulky cotton dressing is changed to a below-the-knee cast, and no bearing of weight is permitted for 6 weeks or until early radiographic and clinical signs of healing are noted. Once walking commences, a below-the-knee cast is applied with a posteriorly placed rubber heel until healing of the fusion is radiographically and clinically confirmed. The role of the heel is to provide an axial compressive force across the ankle joint with weightbearing, rather than any torque by pressure on the forefoot. We do not use a removable walker boot for recovery after ankle arthrodesis because the commercially available boots all have a rocker-type sole, which may increase the force across the anterior ankle. After healing is confirmed, patients are gradually weaned to a walker-type boot and then to regular shoes. Most patients find that a rocker bottom shoe assists with a more normal gait.
POSSIBLE CONCERNS AND FUTURE OF THE TECHNIQUE
The miniarthrotomy is our procedure of choice for ankle arthrodesis. Preservation of the fibula is essential regardless of the technique of ankle arthrodesis to preserve the blood supply and, more importantly, to maintain the anatomy of the ankle for future conversion to a total ankle replacement if necessary. In our initial series, 9 most of the patients had deformities that were not complex in terms of overall ankle alignment and bone quality. We now use this procedure for correction of both simple and complex deformities associated with almost all types of ankle arthritis, and the results have been well maintained (Fig. 7). It has become clear that the miniarthrotomy procedure may be used for AVN of the talus, particularly partial AVN. Ankle deformity is not a contraindication to this procedure, although we would not recommend starting out with this procedure on a deformed ankle because of the technical familiarity required.
The success rate of the miniarthrotomy has been high with the advantages of decreased soft tissue injury, markedly decreased periosteal stripping, and rapid healing time. It is relatively easy to learn and perform and should continue to prove a valuable alternative for ankle arthrodesis.
1. Acevedo JI, Myerson MS. Reconstructive alternatives for ankle
arthritis. Foot Ankle
Clin 1999; 4:409–30.
2. Alvine FG. Total ankle
arthroplasty. In: Myerson MS, ed. Foot and Ankle
Disorders, Philadelphia: WB Saunders Co., 2000:1085–11102.
3. Buck P, Morrey BF, Chao EY. The optimum position of arthrodesis
of the ankle
. A gait study of the knee and ankle
. J Bone Joint Surg 1987; 69A:1052–62.
4. Coughlin MJ, Mann RA. Arthrodesis
of the foot and ankle
. In: Myerson MS, ed. Surgery of the Foot and Ankle
, Ed. 7. St.Louis: Mosby, 1999:651–99.
5. Crosby LA, Yee TC, Formanek TS, et al. Complications following arthroscopic ankle arthrodesis
. Foot Ankle
Int 1996; 17:340–2.
6. Dent CM, Patil M, Fairclough JA. Arthroscopic ankle arthrodesis
. J Bone Joint Surg 1993; 75B:830–2.
7. Frey C, Halikus NM, Vu-Rose T, et al. A review of ankle arthrodesis
: predisposing factors to nonunion. Foot Ankle
Int 1994; 15:581–4.
8. Miller SD, Myerson MS. Tibiotalar arthrodesis
. Foot Ankle
Clin 1996; 1:151–62.
9. Miller SD, Paremain GP, Myerson MS. The miniarthrotomy
technique of ankle arthrodesis
: a cadaver study of operative vascular compromise and early clinical results. Orthopedics 1996; 19:425–30.
10. Morrey BF, Wiedeman Jr. GP, Complications and long-term results of ankle
arthrodeses following trauma. J Bone Joint Surg 1980; 62A:777–84.
11. Myerson MS, Allon SM. Arthroscopic ankle arthrodesis
. Contemp Orthop 1989; 19:21–7.
12. Myerson MS, Quill G. Ankle arthrodesis
. A comparison of an arthroscopic and an open method of treatment. Clin Orthop 1991; 268:84–95.
13. Myerson MS, Ruland CM, Allon SM. Regional anesthesia for foot and ankle
surgery. Foot Ankle
14. Ogilvie-Harris DJ, Lieberman I, Fitsialos D. Arthroscopically assisted arthrodesis
for osteoarthrotic ankles. J Bone Joint Surg 1993; 75A:1167–74.
15. Paremain GD, Miller SD, Myerson MS. Ankle arthrodesis
: results after the miniarthrotomy
technique. Foot Ankle
Int 1996; 17:247–52.