THE CLASSIC: An Artificial Ankle Joint

Abstract

St. Elmo Newton, III was born in Memphis, Tennessee in 1934. His undergraduate education was obtained from Vanderbilt University in Nashville. He then earned his Doctor of Medicine degree in 1958 at the University of Tennessee School of Medicine in Memphis. After a year of surgical internship at New York Hospital, he moved to Rochester, Minnesota where he spent 4 years as an orthopaedic resident at the Mayo Clinic. A fellowship in hand surgery with Dr. Joseph Boyes in Los Angeles completed his postgraduate training. In 1965 Dr. Newton joined the Seattle Orthopaedic and Fracture Clinic, with which he is still associated. Dr. Newton has been a very active member of the Western Orthopaedic Association and has served in many capacities, including a term as president in 1992–1993. He has held an appointment as a clinical professor of orthopaedic surgery at the University of Washington since 1981. Dr. Newton’s interest in total ankle joint replacement began in the early 1970s. He presented a preliminary report on total ankle arthoplasty before the Seattle Surgical Society in January 1974. The following paper summarizes his experience with this procedure. Leonard F. Peltier, MD, PhDFigure: Saint Elmo Newton, IIIAnkle fusion has been the time-honored treatment for painful debilitating conditions of the ankle due to osteoarthritis and rheumatoid arthritis. However, sacrificing motion to relieve pain in the ankle produces additional stress on the knee and midtarsal joints. If there is pre-existing disease in the knee or midtarsal joints, the pain is frequently made worse by ankle fusion. If there is no pre-existing disease, then painful degenerative changes frequently occur in the midtarsal joints subsequently. Ankle fusion is not a uniformly successful operation and has a high complication rate. Charnley1 reports a 20% pseudoarthrosis rate and feels that the ankle does not possess the same natural potential for union as the knee. Johnson4 reports a 60% complication rate on 140 fusions, both tibiotalar and pantalar, with 15 pseudoarthroses and 6 amputations. In an effort to relieve pain, yet retain motion in the arthritic ankle joint, a prosthesis was devised and has been inserted successfully since 1973 in those patients who otherwise would require ankle fusion. Results of the first 50 consecutive patients with a minimum one year follow-up were presented at the American Academy of Orthopaedic Surgeons 44th Annual Meeting, and will be detailed in a subsequent publication.5 Inasmuch as the use of methylmethacrylate in the ankle was not yet approved by the FDA, a special IND number was obtained for the use of methylmethacrylate and all patients were graded on an Ankle Analysis form which has been described previously by Evanski and Waugh.2 The ankle joint is not a hinge, but has multiple parameters of motion, confirmed both by anatomic study3 and cineradiography. These motions include gliding, rocking, and rotation on simple flexion and extension, and slight inversion and eversion on stress. There is no constant center of motion. It was my feeling that any attempt to restrict ankle motion in any way would cause loosening of one or both units of the ankle prosthesis. Therefore, the ankle device was designed to allow freedom of motion in every plane. The high density polyethylene tibial unit is only one cm thick and requires minimal bone resection. The Vitallium® talar unit rests on the dome of the talus with its stem in the longitudinal groove. The dome of the talus is not resected. The obvious advantage of this design is ease of fusion, should it later become necessary, inasmuch as only minimal amounts of bone already have been removed. The tibial unit is a section of a cylinder, while the talar unit is a section of a sphere with a slightly smaller radius. This configuration allows freedom of motion in all planes. A further advantage is that imperfect placement does not compromise end result or motion obtained. The articular surface between the malleoli and talus is separated by distraction of the joint during fitting and insertion of the prosthesis as shown in these clinical X-rays. MATERIALS AND METHODS Technique of Insertion The ankle is exposed through an anterior incision approximately 20 cm in length through the interval between the extensor hallucis longus and anterior tibial tendon. The nerve and artery are retracted laterally with the extensor hallucis longus. Periosteum and capsule are elevated from the distal tibia and talus. The center of the talus is identified and marked with a Methylene Blue marking stick. The grooved sizing template is placed against the front of the tibia with the inferior groove in line with the marked center of the talus. The superior fin of the template is parallel to, and in line with, the tibia. The defect therefore created will be parallel to the floor when the patient stands. The tibia is marked around the template with Methylene Blue and a defect is created in the distal tibia, 2.54 cm in width and 1 cm in depth, using an air drill. It is important to preserve the posterior cortex of the tibia, to avoid malposition of the tibial prosthesis during insertion. Several anchoring holes are made in the tibia for better methylmethacrylate fixation. When the grooved sizing template can be place in the defect while the foot is distracted, the talus is marked by running a Methylene Blue applicator stick into the inferior groove of the template. The reciprocating saw is used to make a groove in the talus 1.25 cm deep and 0.5 cm in width to accomodate the stem of the talar unit. The dome of the talus is not resected and the posterior talus is preserved to prevent malposition during insertion. A trial fit is mandatory. The talar unit is inserted first. As viewed from the front, the talar unit should be in the center of the talus and the tibial unit parallel to the plane of the floor when the patient stands. In the imaginary lateral view, both units should be centered over the dome of the talus. The fit should be snug enough so that the tibial unit goes into the defect only while the foot is firmly distracted. This separates the medial and lateral talar surface from the malleoli. The talar unit is cemented first, removing excess cement from the joint. It is important to insert the tibial unit temporarily while the talar cement hardens. Then the tibial component is removed, and any excess cement is removed. A 2.5 cm strip of cellophane cut from the packaging of the methylmethacrylate is inserted over the talar unit to protect it from cement. A small portion of cement is placed into the tibial defect, filling the anchor holes and coating all sides of the raw cancellous defect as well as all but the articular surface of the tibial prosthesis. The foot is distracted firmly and the tibial component forced into place. While the tibial cement is still in a soft state, the ankle should be plantarflexed and dorsiflexed several times to compress any excess cement which might be in the back of the joint, since a bolus of hardened cement in the back of the joint would block flexion. After routine wound closure, a Robert Jones dressing with plaster splints is used 3 days. Mobilization, physiotherapy, and weight-bearing as tolerated are started after the third day. Sutures are generally left in place for 2 weeks or longer to prevent wound dehiscence. Crutches are used for 6 weeks. RESULTS Of the first 50 cases of total ankle arthroplasty, there were 34 ankles with degenerative joint disease, 10 ankles with rheumatoid arthritis, 3 ankles with avascular necrosis of the talus, and 3 ankles with pseudarthrosis of prior ankle fusion.6 There were 10 failures in the group of 34 degenerative joints. Four prostheses were loose at the bone cement junction. One became infected secondarily, and 5 remained painful without apparent reason. Treatment of these 10 ankles included 5 fusions, 2 revisions, 2 below knee amputations, and one received no treatment. The remaining 24 of the 34 patients are extremely happy with their ankles. They are virtually symptom free, have no restriction of activities, and require no external support. All 3 of the ankles with pseudarthrosis of prior ankle fusion failed. Two of the 3 ankles with avascular necrosis of the talus failed. There were 10 ankles with rheumatoid arthritis. Of the 6 patients on long-term Prednisone, 5 failed. Of the 4 patients not on Prednisone, 3 are completely satisfied with their ankles, and one has foot pain due to involvement of the midtarsal joints. Predictably good results can be obtained in ankles with good stability, reasonably normal anatomy, and rheumatoids who are not on long-term steroid therapy. Contraindications to ankle replacement are any recent infection, valgus or varus deformity of the talus of 20° or greater, any significant ligamentous instability of the ankle, severe distortion of the normal anatomy of the ankle, avascular necrosis of the talus, pseudarthrosis of prior ankle fusion, and rheumatoid arthritis in a patient who is on long-term steroid therapy. SUMMARY Ankle fusion has been the time-honored treatment for painful debilitating conditions of the ankle due to osteoarthritis and rheumatoid arthritis. It is not a uniformly successful operation and has a high complication rate. If there is pre-existing disease in the knee or midtarsal joints, the pain is frequently made worse by ankle fusion. If there is no pre-existing disease, then painful degenerative changes frequently occur in the midtarsal joints subsequently. An artificial ankle joint has been used successfully since 1973 in patients who otherwise would require ankle fusion. The design of the prosthesis allows unrestricted motion in all planes. It also allows imperfect placement of the device without compromising results. The dome of the talus is not resected during insertion of the prosthesis. Only 1 cm of distal tibia is resected, therefore allowing fusion without difficulty should it later become necessary. The superior position of the polyethylene device decreases wear from particulate matter which might settle by gravity. Study of 50 total ankle arthroplasty cases shows that predictably good results can be obtained in ankles with good stability, reasonably normal anatomy, and rheumatoids who are not on long-term steroid therapy.