Adequate planning—prepare for all possible scenarios.
After refusion of a nonunion of a syndesmosis, a fibula shaft osteotomy proximal to the fusion is recommended.
Be prepared to fuse the subtalar joint at the time of revision.
Be prepared to fuse the ankle if it is not possible to revise.
It is almost impossible to revise the tibial component without removing the talus as well.
Insidious onset of swelling, with or without pain, is an indication of loosening or wear.
If the subtalar joint is violated a stemmed component is required. These components are custom ordered; adequate preparation is required.
A high-quality computed tomography scan is necessary to plan the custom implant.
This chapter describes my management of the compromised or frankly failed total ankle replacement, developed over a 15-year period. My experience with this problem is almost exclusively limited to issues surrounding the Agility Total Ankle (DePuy Orthopaedics, Warsaw, IN). For all intents and purposes, the solutions I have employed can be used when attempting to salvage other prostheses as well.
HISTORY/INTRODUCTION/SCOPE OF THE PROBLEM
Patients presenting with a failed total ankle arthroplasty usually complain of severe pain and functional limitation. The pain generally does not occur with rest but is pronounced with ambulation and limits ambulation to prearthroplasty levels. If the prosthesis has settled unevenly, varus or valgus of the hindfoot will be apparent clinically, with the patient complaining of uneven shoe wear and a loss of balance. There are several distinct types of failure, and each may require a different solution. These types of failure include (1) impingement, (2) malposition of components, and (3) loosening of components.
As mentioned, patients will present with pain and discomfort, and most will enter the office using an ambulatory aid. Swelling about the ankle is obvious. Gait is antalgic, and dorsiflexion is limited. Patients will admit to restarting or increasing their dose of nonsteroidal or narcotic medication. When standing, limited weight-bearing is noted, and a varus, valgus, equinus, or a combination deformity is evident, most easily seen from behind ( Fig. 20-1 ).
Radiology and Imaging
To discern loosening or collapse, standing radiographs, as well as fluoroscopically controlled anteroposterior (AP), lateral, and mortise views, should be obtained. In most cases seen with the Agility Total Ankle (Depuy Orthopaedics, Warsaw, IN), the predominant problem is loosening of the talar component. This can be best seen on the AP and presents as apparent overgrowth of the sides of the prosthesis ( Fig. 20-2 ). However, when looking at the component on the lateral view, settling will be evident. Syndesmotic nonunion is seen best on the mortise view. If any question exists about prosthetic loosening, live fluoroscopy should be used. Because of the amount of metal in the prosthesis, neither computed tomography scans nor magnetic resonance images are possible. Unless a workup for infection is required, no further testing is needed.
If infection is suspected, a full workup is required. This would include a white blood cell count with differential, erythrocyte sedimentation rate, C-reactive protein, a triple-phase technetium bone scan, and an indium scan (definitive treatment of the infected total ankle is found in Chapter 18 ).
Nonunion of Malleoli and Symptomatic Implants
Malleolar nonunions are a frequent cause of pain in patients with total ankle replacements ( Fig. 20-3 A, B ). Although intraoperative fractures should be fixed when they occur, often nonunions will develop as the patient begins weight-bearing. These must be treated relative to both clinical symptoms and mechanical considerations. If the prosthesis is stable and the patient can tolerate the associated pain, nothing need be done. If the nonunion will result in component failure, then an attempt at a correction of the nonunion must be performed. On the other hand, if the prosthesis is secure, the syndesmosis and malleoli are healed and intact, but screws and/or plates are prominent, then these implants should be removed, typically in an outpatient setting.
Most of the problems associated with impingement are a result of earlier versions of the talar prosthesis. This component did not cover the entire surface of the talus and as a result is somewhat unsupported. That being said, the talar component can become loose for a variety of reasons, including improper insertion and/or talar preparation, improper size selection, or failure of bony in-growth. Once the component becomes loose, it begins to move in its slot causing an inflammatory reaction and subsequent erosion of the talar surface. The combination of bone debris and inflammation-induced fibrous tissue results in a further deterioration of the joint. At some point, the prosthesis settles and the cut surfaces of the talus (the surfaces not covered by the prosthesis) now begin to make contact with the malleolar sides of the tibial component. This is defined as impingement ( Fig. 20-4 ).