Ankle arthrodesis remains a very effective and reliable treatment method for ankle arthritis. In patients with minimal deformity, poor skin quality, or at risk for wound complications, the ability to perform an arthroscopic ankle arthrodesis allows for a minimally invasive method to achieve a successful fusion and relieve the pain. The use of a mini-arthrotomy is an excellent approach for minimal to moderate deformity, allowing for bone grafting of defects without the need to perform a large anterior approach or fibulectomy. In cases of severe deformity or bone loss, the use of an anterior plate is superior
Key Wordsankle fusion, arthrodesis, arthroscopic, anterior plating, tension band
The optimal technique for ankle arthrodesis yields a predictable outcome, gives a high rate of fusion, and can be used in most patients. The repair should be secured with internal fixation to achieve both clinical and biomechanical stability. The approach that we have found to meet these criteria is either performed arthroscopically, mini-arthrotomy, or with an anterior approach and use of a precontoured anatomic plate. Common to all of these techniques is preservation of the fibula and the overall anatomy of the mortise. We feel that although a lateral approach may be used, this offers two disadvantages. One is that there is definitive injury to the peroneal artery, which may compromise vascularity of the lateral tibial plafond. Second, more relevant for younger patients, is that with the ability to perform ankle replacement following an arthrodesis, it is critical to preserve their anatomy so that this option is available to them in the future.
The mini-arthrotomy technique was originally described for use in ankles without much deformity and without bone defects or the presence of avascular segments of the talus and distal tibia. We have found that arthroscopic ankle arthrodesis is superior for these patients, as it significantly minimizes the soft tissue injury resulting in minimal postoperative soft tissue swelling and rapid arthrodesis. We have found that the mini-arthrotomy technique is feasible even in the presence of significant deformity. This approach requires little periosteal stripping and is associated with rapid bone healing and an increased likelihood of a solid fusion as a result of maintaining the periarticular blood supply. We have not found that flat cuts on the ankle joint offer any advantage; indeed, such cuts require more extensive exposure and periosteal stripping. Performing a fibular osteotomy along with the arthrodesis adds an unnecessary sequence to the procedure, requiring more periosteal stripping as well as creating a larger area for the fusion with increased potential for failure with nonunion. Even less advisable in this clinical setting is the removal of one or both malleoli, which significantly decreases vascularity as well as the surface area for arthrodesis.
For patients in whom there is dysplasia of the talar body or bony loss within the distal tibia, an anterior approach with fixation achieved with a precontoured anatomic plate is the most ideal choice. The exposure allows for deformity correction, bone grafting of any defect, interposition bone block grafting if required, and fixation is achieved from the intact distal tibial shaft and the talar neck and head. In a patient with poor bone quality, bony defect, or significant flattening of the talus, we prefer an anterior approach.
The desired position for ankle arthrodesis is with the foot in neutral with respect to the leg. There is never any reason to fuse the foot in equinus relative to the leg; this positioning will markedly increase the potential for complications, including subsequent arthritis of the peritalar joints as well as foot and knee pain. Much importance has been ascribed in the literature to the position of the foot and the need to translate the foot posteriorly under the tibia, which is not possible with either arthroscopic or mini-arthrotomy arthrodesis. The range of motion of the foot after arthrodesis depends on the preexisting deformity, and in particular the motion of the remaining joints. The true sagittal motion of the tibiotalar joint accounts for approximately 70% of the sagittal plane motion. The sagittal motion after arthrodesis is in plantar flexion through the transverse tarsal joints ( Fig. 31.1 ). This motion may have a deleterious effect on both the talonavicular and the subtalar joints, as confirmed by an increase in arthritis in these joints in some patients ( Figs. 31.2 and 31.3 ).
Incision and Joint Exposure, Ankle Arthroscopy/Mini-Arthrotomy Technique ( )
Two incisions are used for the surgical approach to the ankle. The first is a medial incision made between the notch of the malleolus and the anterior tibial tendon, over a 2.5 cm length. On the lateral aspect of the ankle, the interspace between the peroneus tertius tendon and the fibula is identified, lateral to the lateral cutaneous branch of the superficial peroneal nerve, which is retracted medially with the peroneus tertius tendon. The same incisions are used for ankle arthroscopy, with the main difference that the size of the incision is limited to 1 cm. On the medial aspect of the ankle, the incision is deepened through subcutaneous tissue through the capsule, which is incised down to the periosteum, and a retractor is inserted. Stripping the anterior distal periosteum off the tibia is helpful; then the articular debris, osteophytes, and loose bodies should be removed. Both gutters should be debrided extensively, with a focus on the lateral gutter between the talus and the fibula. We pay a lot of attention to obtaining an arthrodesis between the medial and lateral aspects of the talus and the medial malleolus and the fibula. Either a rongeur or a smooth nontoothed laminar spreader is inserted in the medial incision, followed by insertion of a toothed laminar spreader laterally ( Fig. 31.4 ).
The goal of joint debridement is to remove the joint surface down to bleeding bone. With use of the laminar spreader alternating between the medial and the lateral aspects of the joint, the articular surface of the talus and tibia is completely denuded. We use copious irrigation to ensure that good articular apposition is obtained and that good bleeding cancellous bone is present. One of the most significant changes in our preparation of the joint over recent years has been the creation of broad bleeding bone surfaces with use of a 5-mm osteotome, as well as thoroughly perforating the joint surfaces with a 2-mm drill bit. Before fixation, we check the alignment of the ankle fluoroscopically to ensure good bone-to-bone apposition. Large bone gaps are to be avoided; ideally, the prepared bone surfaces should give the appearance that the joint is already fused. Given the recent literature on the use of bone graft in ankle and hindfoot fusions, we do routinely use allograft combined with concentrated bone marrow aspirate (BMA) in our fusions, with the exception of arthroscopic ankle arthrodesis. Although concentrated BMA and allograft is not required, some type of graft with osteoinductive properties is advocated. Arthroscopic ankle arthrodesis has demonstrated such a predictable rate of fusion that graft is not only difficult to place within the fusion site, it is also not necessary, as the slurry that is created from the debridement serves as excellent autograft.
For an arthroscopic approach, attention is placed on debridement of the anterior osteophytes and synovitis to allow for visualization ( ). Noninvasive distraction is then used, and a 4-mm burr is introduced to debride the remaining articular cartilage. The joint should be addressed from anterior to posterior and medial to lateral, as this will help with visualization and ensure the entire joint is denuded of articular cartilage. Upon completion of the cartilage removal, the subchondral bone should be removed with the burr taking care not to create large divots within the tibia or talus as this will limit the amount of bone apposition. Following removal of the subchondral bone, one should be presented with two large surfaces of bleeding cancellous bone. To increase the surface area of bleeding cancellous bone, we do prefer to create small divots as would be seen on a golf ball on both the tibia and talus. The final maneuver is to create a bone slurry; this is accomplished by turning off the suction and using the burr to debride a thin layer of bone, which will result in a slurry that will act as autograft. The instrumentation is then removed and fixation performed in the same fashion as for a mini-arthrotomy.
Crossing the plane of the screws is important. No screw construct should be created in which all of the screws are perpendicular to the axis of the joint. Parallel screw fixation should be avoided because this technique is ineffective in controlling torsion and rotational forces. The fixation is performed with two or three screws placed over cannulated guide pins. The first guide pin is inserted from the medial malleolus to extend obliquely down toward the anterior aspect of the sinus tarsi. Why insertion in the medial malleolus first? We find that this is a more predictable placement that affords good compression of the joint. This screw is introduced from the medial aspect of the medial malleolus and extends down inferolaterally into the lateral body of the talus, directly above the lateral process of the talus. The screw must not enter the subtalar joint, and one rarely uses a screw longer than 40 mm here. The foot can then be held in position while the second guide pin is introduced from the posterior leg adjacent to the Achilles tendon between the tendon and the sural nerve. The posterior axial screw is introduced next, from the posterior aspect of the tibia into the anteromedial neck of the talus; a 6.5–7.5 mm partially threaded or differentially pitched fully threaded screw, usually approximately 65 mm in length, is used. The only disadvantage of using this posterior to anterior screw is the marked difficulty in removing this screw when either a nonunion develops or when one is converting the ankle arthrodesis to a total ankle replacement. Although this “home run” screw was thought to generate the most compression, this has not proven to be the case. For this reason, where possible, we will use the first screw as described, and then a second screw either parallel with this one from the posterior medial malleolus distally into the head of the talus or from the sinus tarsi into the posteromedial ankle.
The insertion point for the third screw is variable and depends on the anterolateral bone shape. If a ridge is observed on the distal lateral tibia, then the third screw is inserted from the anterolateral tibia immediately adjacent to the fibula into the distal medial talus. If no shoulder is present on the distal tibia, then one or two smaller screws can be inserted from the fibula into the talus, but these screws are not our preference ( Fig. 31.5 ).
Alternatively, the placement of two anterior tibial screws directed into the talar body has been described using the Achilles tendon as a “tension band.” The concept is to dorsiflex the ankle, which will place tension on the Achilles tendon, subsequently compressing the ankle joint and ensuring neutral dorsiflexion. Two wires for cannulated screws are placed from anterior medial and anterolateral into the talar body, “locking” this compression in place. Although this technique has been described only with these two screws, we prefer a third screw as discussed earlier to improve torsional rigidity ( Fig. 31.6 ). There is no proven superior method of fixation; the principle of creating absolute rigidity is all that must be adhered to, which ensures that multiple screws in different planes that do not intersect at the site of fusion are placed ( Fig. 31.7 ).
In complex cases and when the bone quality is poor, the arthrodesis can be extended up into the syndesmosis, with insertion of a few extra screws from the fibula into the tibia ( Figs. 31.8 and 31.9 ). Sometimes the second posterior-to-anterior screw does not go in smoothly and hits the first medial screw. In such instances, a more lateral starting point is preferable, with the guide pin inserted through the posterolateral tibia. The three screws should ideally go through the tibia into the talus, rather than through the fibula. If, however, placement through the fibula is necessary, it should be drilled through to allow the screw to compress the fibula against the tibia. We like to use compression here if at all possible. Clearly, the first medially inserted screw gives the most compression, with less afforded with each subsequent screw, despite the use of partially threaded screws. If the bone quality is very poor, we forego compression; after making sure that the joint surfaces are very well debrided, we then use fully threaded screws ( Fig. 31.10 ). One must plan the approach to arthrodesis in patients with poor skin, with deformity, and with a lot of previous hardware from prior ankle trauma. There is nothing inappropriate with the use of fully threaded screws of any size in any location. If the joint has been well debrided and all of the cartilage removed, compression may not be absolutely necessary. This is particularly the case where poor purchase has been obtained with the first or second partially threaded “compression” screws, but little purchase is obtained, movement of the joint is present, and it is pointless to continue with further attempts at compression screw fixation. Indeed, we have used four fully threaded 4.5-mm screws in place of two or three 6.5-mm screws quite successfully in these cases.