Extensive arthrodesis of the hindfoot and ankle has been a mainstay in the management of complex deformities of the foot and ankle for nearly a century. The triple arthrodesis consists of fusing the subtalar, talonavicular, and calcaneocuboid joints, whereas the pantalar arthrodesis includes the triple joint complex with the addition of the tibiotalar (ankle) joint. Both procedures were initially utilized to treat equinovarus deformities and the sequelae of paralytic disease1,2 with the goal of maintaining ambulation with minimal or no bracing in neuromuscular patients. Over the years, these 2 powerful categories of arthrodesis have become more common in the management of all types of hindfoot/ankle deformities including arthritis caused by trauma,3,4 inflammatory arthropathies,5–7 rigid pes planovalgus or cavus feet,8,9 revision and failed surgery, as well as recent popularity in Charcot neuroarthropathy.10,11
Both triple and pantalar arthrodesis are very valuable tools for the correction of severe foot deformities because of their ability to realign the hindfoot, resolve pain, and provide stability. Both procedures have demonstrated themselves to be reliable over the years and are well tolerated. For triple arthrodesis, long-term satisfaction rates have been reported as high as 95% after 44 years.2 However, despite the overall good patient satisfaction, the procedure itself can be technically demanding and the final rigid osseous reconstruction can be unforgiving of errors. For equinovarus deformities, triple arthrodesis has shown recurrence of deformity in 23% of patients and progression to ankle arthrodesis in 20%,12 which, in part, has increased the use of the pantalar or tibiotalocalcaneal arthrodesis. Even in procedures that are performed well and have a good clinical outcome, the adjacent joints have been demonstrated to accelerate degenerative changes secondary to changes in peak pressures.13–15 Despite these degenerative changes in adjacent joints, the procedure is still well tolerated in older adults.16 In attempts to limit some of the issues associated with these fusions, both have been modified over the years to address different concerns. Many surgeons have moved toward “double” arthrodesis to only include the talonavicular and subtalar joints leaving the calcaneocuboid joint alone. Double arthrodesis may be somewhat faster as far as operative time and allows for variability in incision placement with the so-called medial approach in cases where lateral skin may be vulnerable to postoperative skin complications. In cases of pes planovalgus deformities, foregoing the calcaneocuboid joint also allows the lateral column to lengthen without need for bone graft. For many surgeons, pantalar arthrodesis has been similarly modified as well. At times a tibiotalocalcaneal arthrodesis will achieve similar correction but give some midfoot accommodation through the talonavicular joint, or the talonavicular joint may be included, but not the calcaneocuboid joint for similar reasons as mentioned previously. Modifications of these workhorse arthrodesis techniques may achieve the same product as far as deformity correction and stability but may also allow for more mobility and higher functional status.3
The triple and pantalar arthrodesis are generally thought of as salvage procedures for severe and rigid deformities because of their role in accelerating degenerative changes of the adjacent joints. Other complications frequently include malunion, nonunion, wound dehiscence, and superficial or deep infections. The most common complication of both procedures is reportedly nonunion. For triple arthrodesis, overall nonunion rates range from 0% to 30%.2,16–22 The most common joint of nonunion associated with a triple arthrodesis has been controversial, with some reporting the talonavicular and others reporting the calcaneocuboid joint as the more common site. The talonavicular joint nonunion rate has been reported as high as 20% to 22%.18,22 Reported outcomes for the pantalar arthrodesis are much more limited in the literature, with only a few small case series with irregularly reported outcomes. Among the limited literature, the most common site of reported nonunion is the ankle, which occurs as often as 7% to 10% of the time.10,23
Complications of triple and pantalar arthrodesis are common and occasionally require revision surgery, particularly in cases of painful nonunion or malunion. These complications may be increased in patients with risk factors such as smoking and diabetes mellitus.24 The nonunion rates for revision arthrodesis procedures are generally thought to be much higher compared with primary arthrodesis.25 However, revision of failed triple arthrodesis has been shown to obtain as high as 100% union rates with significant improvements in function and satisfaction.26,27 Although we often consider some cases of nonunion or malunion to be asymptomatic, a prospective computed tomography (CT) study of foot and ankle fusions suggested that the concept of an asymptomatic nonunion may not be valid.28 Unless routine postoperative CT scanning is employed, though, it is likely that we will continue to undertreat nonunion, even though our chances of improving outcomes remain high.
Triple and pantalar arthrodesis are considered definitive procedures and often thought of as a last resort prior to amputation for a number of pedal deformities. The indications for a triple or pantalar arthrodesis cover a wide range of pathologies that cause a painful and/or malaligned hindfoot/ankle including inflammatory arthropathies, end-stage posterior tibial tendon dysfunction, progressive or static neuromuscular-mediated foot deformities, and Charcot neuroarthropathy. The indications for a revisional triple or pantalar arthrodesis are much more limited and are generally considered for nonunion and malunion. Progressive deformities treated with triple arthrodesis may require revision with pantalar arthrodesis, even when the primary procedure healed well. Revision may also be necessary in cases of new Charcot neuroarthropathy deformities or in cases of infection or septic nonunion.
Absolute contraindications for triple and pantalar arthrodesis include the presence of an active infection and arterial insufficiency. Relative contraindications might include acute Charcot neuroarthropathy, uncontrolled comorbidities, active smoking status, presence of ulceration, and lack of attempted conservative therapy. For revision arthrodesis, the contraindications would mirror those of the primary procedures. Many of these contraindications may contribute to the need for a revision procedure and should be considered carefully when evaluating a patient for a revision procedure.
The preoperative planning for reconstructive hindfoot and ankle procedures should always include a detailed history, a thorough clinical examination, and review of all pertinent radiographic imaging. When obtaining the patient history, the surgeon should pay particular attention to progression of pain or deformity and to patient expectations and desires. Although the triple and pantalar arthrodesis are great procedures for significant deformities, it is important to note that they do not restore normal biomechanics.29 These procedures provide reliable pain relief but should not be construed as complete relief, as they often still have some level of discomfort.2 When evaluating radiographs, it is important to consider which joints are involved and carefully evaluate for the global bone density and/or the presence of any intraosseous lesions or cysts.
For revision triple and pantalar arthrodesis, much of the preoperative evaluation is the same as it is for the primary procedures with a few additional considerations. A detailed history, a thorough examination, and review of all pertinent imaging are necessary. The patient’s medical history, laboratory findings, and imaging studies should be evaluated very carefully to optimize their chance at a successful outcome. Also inquiring about possible signs of infection, including wound healing or extended antibiotic use, would be pertinent. All modifiable risk factors should be managed prior to any attempt at a revision arthrodesis.
Prior incision placement and the location and type of hardware utilized for the primary procedure also need consideration when planning revision procedures. Fixation type, size, amount, and orientation must be well understood. This is often overlooked, but inadequate fixation is a very modifiable factor. Understanding fixation, its placement, and techniques to aid in stability is important for primary outcomes and even more important in revision procedures. It is also important to note that removal of old hardware can be complicated by broken hardware, stripped hardware, and bone overgrowth from nonunion. Intraoperative fluoroscopy can be utilized to help locate fixation, and having the proper equipment for removal may save frustration. This is often overlooked but important, as removing old hardware is only the beginning. The surgeon still has to revise the surgery and place new hardware at the time of the revision index surgery. CT scanning should be considered preoperatively to evaluate the extent of arthrodesis from prior procedures.
There are a number of risk factors for nonunion that have been suggested over the years. Smoking greater than 2 packs of cigarettes per day, lack of stability across an arthrodesis site, and poor local vascularity are commonly perceived as risks for nonunion among foot and ankle surgeons.30 However, a multitude of risk factors have been demonstrated, including endocrine disorders and vitamin D deficiency,31 poorly controlled diabetes mellitus and peripheral neuropathy,32,33 tobacco use,34 nutritional deficiencies,35 and certain pharmaceutical regimens.36,37
For pantalar arthrodesis, the incision approach is similar, except the lateral or dorsal incisions are extended proximally. The lateral incision can extend proximal along the fibula, exposing the ankle joint by resecting the distal 5 to 7 cm portion of the fibula. With this approach, care must be taken to avoid the perforating peroneal artery with fibular resection. The resected fibula can be morselized for autograft and placed into the arthrodesis sites, or it can be split vertically and replaced laterally to provide additional stability. Exposure of the ankle joint through the dorsal incision is obtained by extending the incision proximally through the extensor retinaculum and continuing in the tibialis anterior and extensor hallucis longus tendon interval, taking care to protect the neurovascular bundle lateral to the extensor hallucis longus tendon. This approach may still require osteotomy of the fibula, as the length of the fibula may prevent adequate compression across the ankle and subtalar joints.
The medial incision approach to a triple arthrodesis extends an incision along the medial foot at the interval between the posterior tibial tendon and the tibialis anterior tendon. This incision allows for direct visualization of the talonavicular joint as well as the medial side of the subtalar joint and sustentaculum tali. The calcaneocuboid joint can also be accessed here, directly lateral to the talonavicular joint. This approach avoids putting any lateral skin and structures at risk38; however, a portion of the deltoid ligament can easily be sacrificed through the medial exposure of the subtalar joint. Despite this potential deltoid ligamentous injury, it seems that, when comparing a medial double arthrodesis with a traditional 2-incision triple arthrodesis, the risk of ankle valgus still remains as much as 3.64 times higher in the triple arthrodesis group compared with the medial double arthrodesis group.39
In the cases of revision procedures, the initial incisions have already been made, and the surgeon must decide if using the same incisions twice or starting with a clean interval would be a benefit for the second procedure. Removal of existing hardware and any previous wound complications or infections will certainly factor into the decision-making process. Once the incisions have been made and previous hardware has been removed, the joints must be mobilized, inspected, and prepared for the revision arthrodesis. In cases of revision, aggressive curettage and violation of the subchondral plate is absolutely necessary. Use of autogenous bone graft or allograft should also be considered for revision arthrodesis. Choosing between autograft and allograft is not always straightforward. Generally, autograft is considered the gold standard; however, in cases of nonunion with suspected poor host biology it might make more feasible to supplement the arthrodesis sites with allograft materials, which may actually help provide some of the missing growth factors and cells necessary to generate a solid fusion.
Once the preparatory work has been done, the foot and/or ankle should be reduced and temporarily fixated. Alignment and fixation should be performed from proximal to distal. The ankle should be in neutral or slight plantarflexion (0°–5°), the hindfoot should be neutral or slightly valgus (0°–5°), and the forefoot should be slightly abducted (10°–15°) so the second ray (toe and metatarsal) aligns with the anterior crest of the tibia. Fine-tuning of the overall alignment of the foot and ankle may require additional procedures including tendo-Achilles lengthening, calcaneal osteotomies, or plantarflexion osteotomies of the medial cuneiform.
Hardware placement and selection is often dependent on the prior fixation utilized both because the surgeon must access old hardware to remove it and because prior areas of hardware fixation may no longer support additional hardware. For instance, failure of multiple large-diameter screw fixations through the posterior calcaneus might make obtaining any solid fixation through that surface area difficult and necessitate alternate means of fixation. Frequently, when revising aseptic nonunion, malunion, or Charcot neuroarthropathy, the result will be an increase in the amount and the size of the hardware to increase construct rigidity. The hardware may also extend beyond previous surgical sites (ie, conversion of triple to pantalar arthrodesis) to provide more stable fixation. However, in cases of infection, the revision may be a removal of internal hardware and consist primarily of antibiotic spacers and external fixation.
A 55-year-old female patient with a history of hypertension and chronic left foot pain had presented to an outside physician with eventual triple arthrodesis to address the symptomatic acquired pes planovalgus deformity. After several months, she continued to have pain to the left foot and an external bone stimulator was initiated. She ultimately presented for a second opinion 16 months after her let foot triple arthrodesis. She was still complaining of pain and swelling that progressed throughout the day. She was utilizing a walking boot almost daily with a postoperative CT scan revealing the nonunion.
Original left foot postoperative radiographs were available after her index surgery (Figure 13.1). The anterior-posterior and lateral views show an overall well-aligned left foot triple arthrodesis; however, the talonavicular screws were noted plantarly with little fixation given to a majority of the talonavicular joint. Likewise, the subtalar screw was noted lateral within the talus body with also limited fixation to a large portion of the talonavicular joint. These screws would generally be considered adequate, but with nonunion and failure, other options may be considered. Radiographs obtained just over 1 year after the initial procedure (Figure 13.2) showed no overt hardware failure but did show the classic nonunion findings of radiolucency at all 3 joints. The CT scan (Figure 13.3) confirmed nonunion of all 3 joints with lucency of the joints and bone sclerosis.
The decision was made to revise all 3 joints. It was felt that inadequate fixation largely contributed to her nonunion, so the type and amount of fixation as well as the addition of autogenous bone graft from the distal tibia was utilized. The revision triple arthrodesis was performed with all 3 joints being exposed by utilizing the original surgical incisions and by removing all of the retained hardware. The joints were prepared with curettes and a rotary burr. Subchondral drilling was performed to fenestrate the sclerotic subchondral bone of the nonunion sites, and bone graft was harvested from the distal tibia through an anterior approach. The incision was 4 to 5 cm proximal to the ankle joint between the tibialis anterior and extensor hallucis longus tendons. The tendons were retracted and the bone window was made to gain access to the metaphyseal bone. This example utilized a square “window” (Figure 13.4), although more recently circular bone harvesting equipment is available. The round window may be mechanically advantageous and the instrumentation makes harvest more efficient (Figure 13.5).
The left foot position was verified and hardware was placed. Typically, the subtalar joint is reduced and fixated first. In this revision case, 2 screws were utilized for additional fixation compared with the single screw construct of the original surgery (Figure 13.6). The construct with increased number of screws, locking plates, and bone graft went on to heal without any complications (Figure 13.7). The patient was kept non-weight-bearing for 6 weeks with slow progression to a walking boot and shoes over the following months. Since the patient had an external bone stimulator prior to surgery, it was utilized postoperatively as well.
On physical examination her BMI was 45 and she had well-healed surgical scars of both feet consistent with triple arthrodesis as well as a long medial column incision on the right foot. She presented with a forefoot varus on the right side, which was unreducible owing to the triple arthrodesis. The arthrodesis seemed to have good clinical union but the talonavicular joint and medial column were in varus position. Her tenderness was along the entire lateral column as she could not evert to offload the lateral aspect of the right foot.
On her presenting radiographs (Figure 13.8) a well-healed arthrodesis was noted with an elevated medial column. Her prior first ray surgery was noted as well. On further discussion and medical record review, the patient already had an attempt at a bone graft Cotton osteotomy noted by the well-healed medial cuneiform. From a revision standpoint, we discussed with the patient a revision medial cuneiform osteotomy with bone or metal wedge but clinically the deformity appeared to be more related to the hindfoot in nature and with revision surgery, to attempt to be more aggressive with a corrective osteotomy of the fusion site.