Albert first described ankle arthrodesis in 1878. It has remained the standard in surgical treatment of end stage arthrosis of the ankle primarily because of the dissatisfaction with the results of ankle arthroplasty. While it is most commonly performed for posttraumatic arthrosis, it is also an effective and reliable treatment for patients with inflammatory arthritis, uncorrectable ankle instability, failed arthroplasty, and osteoarthrosis. 1–11 Over the years, many different techniques have been advocated in the literature, demonstrating that no one technique has proven optimal.
The two primary options for fixation are internal fixation and external fixation. External fixation techniques are fraught with problems, including pin tract infections, high rates of malunion, delayed union, and nonunion, as well as patient dissatisfaction. 1,10,12–15 For this reason we believe that external fixation techniques should only be considered in situations in which the patient has an end stage septic arthrosis or severe osteopenia.
In contrast, internal compression techniques utilizing cross-screw fixation maintain position more effectively, and have much greater patient acceptance. When quality of bone and screw purchase is good, screws are equal to or superior to external fixation for all forms of loading tested. 10,15 In addition, a shorter period of immobilization is required. Thus, the patient can begin weight bearing and range of motion of adjacent joints sooner, leading to an earlier return of function. 2,16
We use cannulated screws because the technique offers numerous advantages. The guide pins hold the ankle in position, can be easily repositioned if necessary, and can be placed percutaneously in both antegrade and retrograde fashion. 17
Two or more screws can be placed. Early in our series we used two crossed screws. As we gained more experience, we chose three screws. Three screws provide significantly greater compression and better resistance to torque. 7 The most important screw is the posterior to anterior screw. 10 When placed first, this screw allows compression, posterior translation, and correction of equinus. Finally, we place the screws in a crossed fashion, because it has been suggested that the crossed screw technique improves the likelihood of obtaining a successful fusion. 18
In addition to rigid internal fixation, the use of bone graft is indicated and improves the rate of union. 3,5,6,9,10 Although some surgeons use an anterior trough, we do not feel that this is necessary. We currently use allograft at all index procedures unless structural graft is required. This decreases the surgical time and eliminates the risks associated with iliac crest bone grafting.
As with the surgical technique, numerous rehabilitation protocols have been developed, ranging from no weight bearing for 6 weeks to full weight bearing from postoperative day 2. 1,3,4,8,10,19 The two senior authors use similar protocols. Weight bearing is progressed as radiographic evidence of union allows.
With internal compression arthrodesis high fusion rates, acceptable clinical function, high patient satisfaction, and low complication rates have been achieved. 1–9
The purpose of this paper is to describe the results of anatomic compression arthrodesis. The technique is easily reproducible, provides a high rate of fusion, and should be considered the standard surgical treatment of end-stage arthrosis of the ankle that has failed conservative management.
For patients presenting with ankle pain, a thorough history and physical examination must be performed to determine the etiology of the pain. The history should provide details as to the quality and location of the pain, the inciting and alleviating factors, the association of instability, the presence of rest or night pain, and any associated systemic factors. In patients with arthritis that affects numerous joints, the patient must be questioned about the presence of symptoms at the knee. All forms of previous treatment must be known.
The examination must include an analysis of the patient's gait and an assessment of the alignment of the entire lower extremity with weight bearing. The presence or absence of swelling, erythema, and warmth is noted. The range of motion at the ankle, subtalar, and transverse tarsal joints is measured. The presence of pain with range of motion is noted. If the pain seems to be coming from the ankle after evaluating all surrounding structures and insuring that the pain is not associated with a nerve mediated source, imaging studies are obtained.
Weight bearing radiographs are performed to fully appreciate the severity of arthritic involvement. A mortise and lateral view of the ankle are obtained (Fig. 1). In addition, an anteroposterior and lateral radiograph of the foot are obtained to evaluate the surrounding joints.
If imaging confirms the presence of arthrosis at the ankle, but there is still a question as to the exact source of the pain, diagnostic injections are a reliable method for isolating the source of the pain. The ankle can be injected with a combination of a short acting and a long acting anesthetic agent in the office. If necessary, at a different visit the subtalar joint is injected with fluoroscopic guidance to insure correct placement of the needle. The patient is then asked to walk around the office for a short period of time. If the injection relieves the pain then the source is confirmed and treatment is planned.
Conservative measures are always pursued initially no matter how severe the arthrosis. These include injection therapy, nonsteroidal antiinflammatories, acetaminophen, elastic ankle supports, shoe wear modifications, and activity modifications. If these measures fail then the patient is placed into a short leg walking cast or a removable walking boot. If the patient obtains relief from a 4-week period of immobilization, the option of bracing with an ankle foot orthoses or the option of surgery can be considered.
Surgical options include ankle arthrodesis and total ankle arthroplasty. Although the option of motion preserving, total ankle arthroplasty appears to be an attractive alternative to arthrodesis, with newer techniques and implants currently being used, only short-term follow up is available. In addition, many questions pertaining to ankle arthroplasty must be addressed in well-controlled trials before it can be recommended over arthrodesis, the current standard of care for end stage arthrosis of the ankle. If the patient decides to undergo arthrodesis, after a thorough discussion of the risks and benefits of each procedure, then preoperative planning is performed.
The radiographs are reviewed and the alignment is evaluated. The quality of the bone is assessed. If bone loss is a problem then the amount of bone graft necessary for a successful arthrodesis is determined. If a structural bone graft is necessary then autograft iliac crest is used. For small defects and to simply fill voids encountered at the time of surgery, crushed cancellous allograft with or without additional demineralized bone matrix is used. For patients with malalignment, plans are made to correct the alignment through a combination of bone grafting and limited bone resection.
The patient is placed supine on a table with a radiolucent extension, antibiotics are given, and a thigh tourniquet is used. The anterior approach to the ankle is used in all cases unless previous incisions would lead to narrow skin bridges (Fig. 2). In such settings the previous incisions are used. The anterior approach uses the interval between the tibialis anterior tendon and the extensor hallucis longus tendon. This approach provides an optimal view of the ankle, and facilitates preparation of symmetrical opposing surfaces for correction of malalignment. 20 The retinaculum is incised over the extensor hallucis longus. The neurovascular bundle is retracted laterally. The capsule is incised and preserved.
Any remaining cartilage is removed from the tibiotalar joint and both the medial and lateral gutters. This is performed with a sharp elevator (Fig. 3). The sclerotic subchondral bone is breached with a combination of the sharp elevator, small osteotomes, and a 2.0 mm drill bit. No other bone resection is performed. The goal is to expose good bleeding cancellous bone on both sides of the joint while maintaining the patient's anatomy.
The ankle is then placed into the optimal position of fusion, neutral sagital alignment, 5 degrees of valgus, rotation comparable to the opposite extremity, usually 5 to 10 degrees of external rotation, and posterior translation of the talus. 17 The posterior translation shortens the lever arm and is believed to improve gait dynamics. 17
Once adequate alignment is achieved the guide pins for the screws are placed as follows: the first pin is cut exactly 1cm short to provide two sharp ends for a retrograde-antegrade drilling technique. The first pin is placed from the posterior malleolus into the neck of the talus. This is performed by maximally plantar-flexing the foot. The center of the tibial plafond is identified. The pin is placed into the ankle joint and drilled in a retrograde manner just lateral to the Achilles' tendon (Fig. 4). The pin is grasped posteriorly and pulled flush with the joint surface. The second pin is drilled from the medial tibial metaphysis to the joint surface so that it will enter the talar body. The third pin is drilled from the fibula to the joint surface so that it will enter the talar body posterior to the other two guide pins. In revision cases in which the fibula has already been resected, the third pin may be placed from the lateral tibial metaphysis into the posterior talar body. Visual inspection of the guide pins is performed to insure adequate positioning. Crushed cancellous allograft bone is placed into the joint, both gutters, and the region immediately anterior to the joint (Fig. 5). The ankle is reduced and the pins are advanced into the talus.
The position of the ankle and the pins is checked before placing the compression screws (Figs. 6a and 6b). Depending on the size of the patient's bones, titanium cannulated 6.5 mm or 8.0 mm partially threaded, cancellous lag screws (Ace-DePuy, Warsaw, IA) are used (Fig. 7). Care is taken to insure that the screws do not penetrate the subtalar or talonavicular joints.
To further augment the stability at the fusion site a 3.5 mm, one-third tubular, anterior plate (Ace-DePuy, Warsaw, IA) may be applied (Figs 8a and 8b). This technique is used for patients at increased risk for nonunion, 1,10,11,19,24,28,29 and in cases in which slight motion exists with attempted dorsiflexion despite cannulated lag screw placement. Depending on the size of the patient, a 3- or 4-hole plate is used. The plate is contoured to the anterior surface of the distal tibia and talus. One cancellous 3.5mm screw is placed into the talus and one is placed into the tibia. We use the plate in approximately 70% of cases. Although we have not compared the use of the plate and screws to the use of screws alone in a randomized, prospective fashion, it is our feeling that the plate leads to increased stability and therefore a higher likelihood of obtaining a successful arthrodesis. In addition, a one third tubular plate is low profile and has not created problems with wound healing.
The patient is hospitalized for 1 day to receive antibiotics and to begin physical therapy. The patient is discharged in a non-weight bearing, short leg cast or splint. Sutures are removed in the office at 2 weeks. The cast is replaced and the patient is maintained at strict nonweight bearing for 4 more weeks. At 6 weeks after the surgery the patient is evaluated, the cast is changed to a walking cast or boot, and partial weight bearing is permitted. The patient is again evaluated between 10 and 12 weeks postoperatively. Weight bearing is advanced based on clinical and radiographic evidence of union. Clinical union is present when the patient is able to bear full weight without any type of immobilization and without pain. Radiographic union is considered to be present when trabecularization is noted over 50% of the joint on both the anterior-posterior and lateral radiographs. Often this determination is very difficult. If there is any question as to whether union has occurred, the ankle is stressed under fluoroscopy to check for motion at the fusion site. Delayed union is present if the patient has not demonstrated fusion by 6 months. 6 This can often be detected by noting lucency around the screws in a patient with persistent pain. Nonunion is present if fusion has not occurred by 12 months. However, patients usually undergo revision arthrodesis if union had not occurred by 6 months, and if gross motion is demonstrated on fluoroscopic examination.
From January 1986 until June 1998, 137 ankle arthrodeses were performed in 133 patients. One hundred and twenty-six of the 137 ankles (92%) had adequate radiographs to assess for fusion. One hundred and fourteen of the 126 ankles (90%) fused after the index surgery. Average time to fusion was 16 weeks. After revision surgery for the nonunions, 119 of the 126 ankles (94%) fused. Seven ankles did not fuse. Of these seven, one patient died 3 months after revision surgery, two patients with delayed unions were lost to follow up, and four patients have refused further surgery.
To better assess our rates of union we stratified our findings according to indication for arthrodesis (Table 2). Those patients with post-traumatic arthrosis, osteoarthrosis, instability, history of nonunion after prior arthrodesis, and miscellaneous arthropathy all had final union rates of 99 to 100%.
In the 11 patients with rheumatoid arthritis the initial fusion rate was only 73%, and the final fusion rate was 82%. This group often presents the surgeon with numerous challenges. Obtaining a successful fusion is difficult, secondary to multiple joint involvement, prior hindfoot fusion, osteopenia, fragile soft tissues, and medications such as corticosteroids, methotrexate, and anti-inflammatories. 21–23 Cracchiolo et al. presented similar fusion rates for their patients with rheumatoid arthritis. 21
Patients with paralysis or neuropathy also presented us with a significant challenge. Only two of six patients (33%) went on to successful fusion.
To analyze outcome we performed the American Orthopaedic Foot and Ankle Society (AOFAS) ankle and hindfoot score for patients who were examined. A phone survey based on the AOFAS ankle and hindfoot score was performed for patients who could not return to the office. One hundred and five ankles in 102 patients were followed for the minimum of 24 months required for functional assessment. Seventeen patients (18 ankles) in this group are deceased. Average follow up was 58 months (24–185 months). Overall, the functional outcome was evaluated for 89 ankle arthrodeses, including 2 arthrodeses in deceased patients who were evaluated before their death. These results are stratified according to indication (Table 3).
In addition, we performed chart review and patient interview (when possible) to obtain information pertaining to satisfaction with the surgery, pain level before and after surgery, functional level before and after surgery, the need for braces, orthoses, or shoe wear modifications, and the need for a walking aid. Information pertaining to satisfaction, pain level, and functional level was available for 93 patients. Eighty-two of 93 patients (88%) were satisfied. Eighty-two of 93 patients (88%) had less pain. Seventy-nine of 93 patients (85%) had increased function (Table 4). Information pertaining to the need for braces, orthoses, or shoe wear modifications was available for 48 patients. Sixteen patients (33%) used such a device. Finally, information pertaining to the need for a walking aid was available for 94 patients. Nineteen patients (20%) used walking aids on a regular basis. Seven patients (7%) used a walking aid on an as needed basis.
When evaluating radiographs for progression of arthrosis we noted that evidence of progression before 24 months was rare. Thus, only patients with postoperative radiographs with a minimum of 24 months follow up were evaluated for progression of arthrosis at the subtalar and transverse tarsal joints based on the classification system of Morrey and Wiedeman. 14 Although we were able to perform functional evaluations on 89 ankles, only 41 patients with the required minimum 24-month follow up returned for radiographs. The mean radiographic follow up for these patients was 57 months. Thirty-three of 41 patients, 80%, demonstrated progression of arthrosis at at least one joint (Table 5). Additional fusions were performed in only 2 of these 41 patients because of their symptoms. One patient with osteoarthrosis underwent triple arthrodesis, and one patient with osteoarthrosis had a subtalar fusion.
Two patients had additional fusions for reasons unrelated to progressive arthrosis. One patient with post-traumatic arthrosis had a subtalar fusion after developing a pes planus deformity, and 1 patient with Charcot-Marie-Tooth disease had a subtalar fusion after developing instability at the subtalar joint.
Of the original 133 patients 29 (22%) developed one or more complications. There were a total of 41 complications in these 29 patients. Twelve patients developed a delayed union or nonunion. Six ankles (4%) developed a superficial infection. Eight ankles (6%) were fused in a sub-optimal position. Five of 70 patients (7%) had a complication related to the iliac crest donor site. One patient developed reflex sympathetic dystrophy, and 9 patients developed various minor complications. In addition, 16 patients (12%) complained of painful hardware or had suboptimal position of the hardware (Table 6).
POSSIBLE CONCERNS, FUTURE OF THE TECHNIQUE
Ankle arthrodesis is a reliable procedure for the treatment of arthrosis of the ankle and is the current treatment of choice. 7,10 Anatomic compression arthrodesis limits the amount of overall shortening and prevents impingement of the malleoli. In addition, it provides increased stability, a greater bony surface for fusion leading to higher fusion rates, and a more functional and cosmetic outcome. 3,10
In our series 90% of the ankles achieved union after the index procedure, and 94% of the ankles achieved union after revision surgery. Excluding patients with rheumatoid arthritis and neurologic conditions, the final union rates were between 99%–100%. These results compare favorably with the results of other studies utilizing compression arthrodesis with cannulated screws, including rates of 100% reported by Paremain et al. and Maurer et al. 2,4,10,13,14,23–26
The average time to union in our series of patients was 16 weeks. This falls within the range of times encountered in the literature, from 6 weeks as reported by Paremain et al., utilizing a mini-arthrotomy technique, to 16 weeks as noted by Holt et al. 2,3,5,7–10 Although the mini-arthrotomy techniques and arthroscopic techniques may result in more rapid fusion, they may only be used for ankles with minimal deformity.
Despite the low AOFAS ankle and hindfoot scores encountered in our series and in series reported by others, 3,6,27 the majority of our patients were satisfied and reported less pain and increased function.
Although improved surgical techniques such as anatomic compression arthrodesis are highly successful, a number of issues must be kept in mind. First, the surgeon must be aware of and plan appropriately for certain problems that make arthrodesis more difficult. These include local factors such as massive loss of bone or poor quality bone, avascular necrosis, severe deformity, a history of combined plafond and talus fractures, open injuries, or septic arthrosis. In addition, systemic factors such as major medical or psychiatric disorders, smoking, alcohol use, and illegal drug use place patients at a higher risk for nonunion. 1,10,11,19,24,28,29
Second, authors such as Frey et al. have reported high complication rates. 28 In contrast, our study and numerous other studies of internal compression arthrodesis have demonstrated more acceptable rates of complications. 1–4,9
Finally, the greatest dissatisfaction and concern regarding the use of ankle fusion is the risk of developing progressive arthrosis of the subtalar and midfoot joints. A high percentage of the patients in our series and in series reported by other authors have developed progressive arthrosis. 3,30 The majority of the patients in theses series with short to intermediate follow up were asymptomatic. However, with 22 year follow up, Coester et al. reported that 23 patients who underwent isolated ankle arthrodesis for post-traumatic arthrosis, with good early relief of pain, developed premature deterioration of other joints in the foot with the development of pain and dysfunction. 31
Although we feel that anatomic internal compression arthrodesis with cannulated screws is still the standard surgical treatment of end-stage arthrosis, these concerns need to be addressed. However, the only other option available at this time, total ankle arthroplasty, has shown disappointing results in the past, especially in young patients with post-traumatic arthrosis, and newer designs have limited follow up. 27,30,32–37 Thus, we feel that while motion-sparing alternatives to arthrodesis must continue to be pursued, the results of these procedures must equal or surpass the high satisfaction rate of patients with an ankle arthrodesis before arthroplasty can be recommended.
1. Chen YJ, Huang TJ, Shih HN, et al. Ankle arthrodesis
with cross-screw fixation good results in 36/40 cases followed 3–7 years. Acta Orthop Scand 1996; 67:473–478.
2. Dennis DA, Clayton ML, Wong DA, et al. Internal fixation
of the ankle
. Clin Orthop 1990; 253:212–220.
3. Mann RA, Rongstad KM. Arthrodesis
of the ankle
: a critical analysis. Foot Ankle
Int 1998; 19:3–9.
4. Mann RA, Van Manen JW, Wapner K, et al. Ankle fusion
. Clin Orthop 1991; 268:49–55.
5. Maurer RC, Cimino WR, Cox CV, et al. Transarticular cross-screw fixation a technique of ankle arthrodesis
. Clin Orthop 1991; 268:56–64.
6. Monroe MT, Beals TC, Manoli A. Clinical outcome of arthrodesis
of the ankle
using rigid internal fixation
with cancellous screws. Foot Ankle
Int 1999; 20:227–231.
7. Paremain GD, Miller SD, Myerson MS. Ankle arthrodesis
: results after the miniarthrotomy technique. Foot Ankle
Int 1996; 17:247–252.
8. Partio EK, Hirvensalo E, Partio E, et al. Talocrural arthrodesis
with absorbable screws 12 cases followed for 1 year. Acta Orthop Scand 1992; 63:170–172.
9. Stranks GJ, Cecil T, Jeffery ITA. Anterior ankle arthrodesis
with cross-screw fixation a dowel graft method used in 20 cases. J Bone Joint Surg [Br], 1994; 76-B:943–946.
10. Holt ES, Hansen ST, Mayo KA, et al. Ankle arthrodesis
using internal screw fixation. Clin Orthop 1991; 268:21–28.
11. Perlman MH, Thordarson DB. Ankle fusion
in a high risk population: an assessment of nonunion risk factors. Foot Ankle
Int 1999; 20:491–496.
12. Hagen RJ. Ankle arthrodesis
problems and pitfalls. Clin Orthop 1986; 202:152–162.
13. Lynch AF, Bourne RB, Rorabeck CH. The long-term results of ankle arthrodesis
. J Bone Joint Surg [Br] 1988; 70-B:113–116.
14. Morrey BF, Wiedeman GP. Complications and long-term results of ankle arthrodesis
following trauma. J Bone and Joint Surg 1980; 62-A:777–784.
15. Thordarson DB, Markolf K, Cracchiolo A. Stability of an ankle arthrodesis
fixed by cancellous-bone screws compared with that fixed by an external fixator. A biomechanical study. J Bone and Joint Surg 1992; 74-A:1050–1055.
16. Ross SDK, Matta J. Internal compression arthrodesis
of the ankle
. Clin Orthop 1985; 199:54–60.
17. Walling AK, Padrta BJ. Ankle arthrodesis
and cannulated screw fixation. Mechanics, operative techniques, clinical implications. Springer Verlag: New York. 1996, pp. 260–267.
18. Friedman RL, Glisson RR, Nunley JA. A biomechanical comparative analysis of two techniques for tibiotalar arthrodesis
. Foot Ankle
Int 1994; 15:301–305.
19. Scranton Jr. PE, An overview of ankle arthrodesis
. Clin Orthop 1991; 268:96–101.
20. Mears DC, Gordon RG, Kann SE, et al. Ankle arthrodesis
with an anterior tension plate. Clin Orthop 1991; 268:70–77.
21. Cracchiolo III, A Cimino WR, Lian G. Arthrodesis
of the ankle
in patients who have rheumatoid arthritis
. J Bone Joint Surg 1992; 74-A:903–909.
22. Felix NA, Kitaoka HB. Ankle arthrodesis
in patients with rheumatoid arthritis
. Clin Orthop 1998; 349:58–64.
23. Smith EJ, Wood PLR. Ankle arthrodesis
in the rheumatoid patient. Foot Ankle
Int 1990; 10:252–256.
24. Cobb TK, Gabrielsen TA, Campbell DC, et al. Cigarette smoking and nonunion after ankle arthrodesis
. Foot Ankle
Int 1994; 15:64–67.
25. Kish G, Eberhart R, King T, et al. Ankle arthrodesis
place ment of cannulated screws. Foot Ankle
Int 1993; 14:223–224.
26. Ogilvie-Harris DJ, Fitsialos D, Hedman TP. Arthrodesis
of the ankle
: a comparison of two versus three screw fixation in a crossed configuration. Clin Orthop 1994; 304:195–199.
27. Newton SE. Total ankle
arthroplasty clinical study of fifty cases. J Bone and Joint Surg 1982; 64-A:104–111.
28. Frey C, Halikus NM, Vu-Rose T, et al. A review of ankle arthrodesis
: predisposing factors to nonunion. Foot Ankle
Int 1994; 15:581–584.
29. Stuart MJ, Morrey BF. Arthrodesis
of the diabetic neuropathic ankle
joint. Clin Orthop 1990; 253:209–211.
30. Takakura Y, Tanaka Y, Sugimoto K, et al. Long term results of arthrodesis
for osteoarthritis of the ankle
. Clin Orthop 1999; 361:178–185.
31. Coester LM, Saltzman CL, Leupold J, et al. Long-term results following ankle arthrodesis
for post-traumatic arthritis
. J Bone Joint Surg 2001; 83-A:219–228.
32. Bolton-Maggs BG, Sudlow RA, Freeman MAR. Total ankle
arthroplasty a long-term review of the London hospital experience. J Bone Joint Surg 1985; 67-B:785–790.
33. Kitaoka HB, Patzer GL. Clinical results of the Mayo total ankle
arthroplasty. J Bone Joint Surg 1996; 78-A:1658–1664.
34. McGuire MR, Kyle RF, Gustilo RB. Comparative analysis of ankle
arthroplasty versus ankle arthrodesis
. Clin Orthop 1988; 226:174–181.
35. Newton SE. An artificial ankle
joint. Clin Orthop 1979; 142:141–145.
36. Stauffer RN, Segal NM. Total ankle
arthroplasty: four years' experience. Clin Orthop 1981; 160:217–221.
37. Unger AS, Inglis AE, Mow CS, et al. Total ankle
arthroplasty in rheumatoid arthritis
: a long-term follow-up study. Foot Ankle
Int 1988; 8:173–179.
Keywords:© 2002 Lippincott Williams & Wilkins, Inc.
ankle; arthrodesis; fusion; internal fixation; arthritis