Pierre Mansat
Epidemiology
A meta-analysis of 22 publications reviewing 838 total elbow arthroplasties (TEAs) was published in 1996. With an average follow-up of 5 years, the complication rate was 43% with a revision rate of 18%. Main complications were the following: aseptic loosening (radiographic 17.2% and clinical 6.4%), infection (8.1%), ulnar nerve involvement (10.4%), instability (7% to 19%), and periprosthetic fracture (3.2%). The French Orthopedic and Traumatology Society reviewing 370 TEAs found a complication rate of 27% with a revision rate of 17%. In this chapter, several specific features of the presentation are discussed and surgical options available to deal with failed TEAs are presented.
Assessment
Failed elbow arthroplasty can be caused by sepsis, device failure, instability, periprosthetic fracture, and loosening. Assessment to exclude the possibility of sepsis is the most important consideration prior to any revision procedure, but especially in those with early unexplained or unanticipated failure. Analysis of the sedimentation rate and C-reactive protein are regularly performed along with aspiration of the joint if there is any question of sepsis. Based on the plain radiograph to analyze bone quality, the appropriate preoperative plan is formulated. Surgery is not performed if radiolucent lines are not painful. In such cases, the patient is followed on a regular basis. Stiffness, scarring, and contracture of the soft tissue and prior evidence of sepsis and status of the ulnar nerve must also be noted.
Presentation—Treatment Options
Device Failure
Wear of the polyethylene in TEA has been the most common mode of material failure. Although often contributing to component loosening owing to polyethylene debris, mechanical symptoms may develop because of metal-on-metal articulation or dislocation of the component. Fracture of the metal components has been reported but is now less common owing to improvements in implant design and materials. Isolated bushing exchange is a successful procedure if there is no osteolysis compromising component fixation.
Instability
Instability is typically seen with unlinked resurfacing types of TEA. Although usually occurring in the early postoperative period as a frank dislocation, this also can present more insidiously as weakness, giving way, clunking, or other mechanical symptoms. Instability can be caused by ligament insufficiency, malpositioning of the components, or uneven wear of the polyethylene. Examination under fluoroscopy may show the cause of instability and allow the diagnosis to be made. Splinting for a few weeks may help restore stability, but if the elbow remains unstable after a period of splinting, a surgical procedure may be indicated. Attempts to salvage an unlinked TEA that is unstable can be unpredictable. One can attempt at least one soft tissue procedure before undertaking removal of a well-cemented unlinked prosthesis. Attention should be paid to maintaining or restoring an adequate lateral collateral ligament complex by firmly reattaching it to the lateral epicondyle with use of sutures through drill holes in bone or by reconstructing the ligament with a tendon graft. Elbows with malpositioned components do not respond to conservative treatment or ligament reconstruction. Revision of the component positioning or conversion to a semiconstrained implant is usually required.
Periprosthetic Fracture
Fractures in proximity to TEA can occur at the time of implant insertion, as a consequence of neglected component loosening with bone loss, or as a result of a traumatic
P.414
event. Mayo has proposed to classify these fractures as type I—metaphyseal; type II—stem involvement; type III—proximal or distal to the stem tip (Fig. 61-1).
|
Figure 61-1 Mayo Classification of periprosthetic elbow fractures. Type I: metaphyseal; type II: involves stem; type III: beyond stem. |
Humeral periprosthetic fractures that occur in the periarticular segment (type I) usually do not require surgical treatment if a linked prosthesis has been used. In unlinked prostheses, fixation of the fragment is necessary to preserve implant stability. Fractures that involve the olecranon, however, should be fixed in all patients since this will restore triceps function. Proximal or distal to the stem (type III), cerclage wire or plate and screw fixation may be required to stabilize the fracture (Fig. 61-2). Fractures around the stem of the prosthesis (type II) almost always require revision surgery because the implant is often loose. A longer revision stem should be used to bypass the fracture, with cortical strut allograft around the fracture.
|
Figure 61-2 Periprosthetic type III fracture at the stem of the humeral component; the implant is well fixed (A). Osteosynthesis with plate and screw and cerclage wire (B, C). |
Implant Loosening
Loosening of the implant is the most frequent cause of long-term implant failure. A loose implant can be associated with bone resorption, cortical thinning, and ballooning of the humerus or ulna. Revision options are predicated on the quality of bone and presence of a periprosthetic fracture. A salvage procedure such as arthrodesis is rarely indicated. Arthrodesis requires a sufficient amount of bone present. Resection arthroplasty is indicated in the presence of a septic prosthesis. A stable resection arthroplasty can provide a relatively comfortable joint. Not all patients after resection arthroplasty will desire a reimplantation procedure. A revision procedure with reimplantation can be performed when infection has been eradicated.
Technique of Revision Arthroplasty
Implant Selection
The most important consideration is whether adequate fixation can be obtained with another stemmed implant given the amount and quality of the intact cement mantle. If bone
P.415
stock remains, a nonconstrained implant can be used. However, most of the time a semiconstrained implant is preferred. Long-stem devices should be available. For humeral revisions, a 15- or 20-cm-long humeral stem is often needed. Long-stemmed ulnar components should also be available.
Surgical Technique
Preparation for iliac crest bone should be routine. A sterile tourniquet is often used. The lateral decubitus position is preferred by some surgeons to allow wide exposure of the humerus and radial nerve if necessary. The use of a posterior midline elbow incision is preferred, but use of a previous skin incision is recommended. The technical features of all revision options must address the preservation of the triceps, identifying and protecting all neurovascular structures, and protection of the cortical bone. If the distal humeral columns are deficient, a triceps-preserving approach should be considered. If the olecranon process is fractured, a trans-olecranon approach is used for revision of the elbow arthroplasty and repaired with tension band wire technique at the end of the procedure. Proximally, the radial nerve is always identified, at least by palpation. The ulna is extensively exposed in a subcutaneous fashion as distally as necessary to have adequate exposure and to avoid violation of the ulnar cortex. Extensive synovectomy is always necessary; tissue samples are sent for pathology and culture.
Extraction of the components is straightforward in cases of aseptic loosening but more problematic if they are firmly fixed. If the implant design is tapered, the component sometimes can be extracted by grasping the articulating surface and tapping on the prosthesis in a retrograde direction. Removing as much cement as possible allows the device to be removed more easily. Fracture or further bone loss from aggressive attempts at cement removal should be avoided. A cortical window around the stem or at the tip of the prosthesis should be considered to avoid intraoperative fracture caused by more aggressive attempts to remove the components. It should be fixed with cerclage wire at the end of the procedure before cement injection. Once the prosthesis is removed, cement that is firmly adherent should be left in place unless it interferes with placement of the new components (Fig. 61-3). Powered cement removal instruments should be used with caution. Ultrasonic cement removal devices can be particularly useful.
|
Figure 61-3 Revision of a loose constrained implant (A). A long-stem semiconstrained implant has been used to bypass the area of loosening (B). |
The implant chosen for revision should have stems of adequate length to bypass cortical defects or fractures and have sufficient constraint to provide adequate joint stability (Fig. 61-4). Modern cement technique is used with cement restrictors and pressurized gun. Antibiotic cement should be routinely used. Caution should be exercised in patients with cortical perforations or fractures to ensure that cement does not damage adjacent neurovascular structures.
Closure is done with steps to ensure triceps function and the arm placed in extension with a splint to avoid tension or pressure on the wound. The arm is elevated for 48 hours. Gentle active motion can then be initiated. If the bone is fractured or if a bone graft has been used, protection in a cast brace or splint for several weeks may be undertaken. No formal therapy is prescribed, but activities of daily living are encouraged.
Specific Revision Options
Osseous Enhancement Options
If the process is associated with bone resorption, osteolysis, or a periprosthetic fracture, an impaction or strut grafting augmentation procedure is indicated.
Impaction Grafting
Impaction grafting is designed for two specific purposes, to restore bone stock and to enhance the bone/cement interface. At least 2 to 3 cm stem depth into intact bone in addition to the augmentation fixators is needed. The
P.416
P.417
P.418
medullary canal is plugged with a silastic device. A double-tube apparatus is assembled. The nozzle of tubing used for femoral cementation is cut to the length that corresponds to the extent of the lytic process. The elbow cement injector tube is then inserted within the femoral tube, extending distally into normal host bone. Cancellous bone graft or graft substitute is tightly packed around the outer tube. The cement is then mixed in the canister of the smaller elbow injector system and inserted on the nozzle in situ. It is injected through the nozzle while withdrawing to the level of the outer tube. At this point both inner and outer tubes are simultaneously withdrawn while injecting cement into the void created by the larger tube. The implant is carefully inserted to the desired length (Fig. 61-5). This technique may be used for both the humerus and ulna, as an adjunct in the re-establishment of bone stock (Fig. 61-6).
|
Figure 61-4 Ulna fenestration has been made to remove a fractured well-fixed ulnar component (A). Cerclage wire has been used to fix the cortical window before cementing a new ulnar component (B). |
|
Figure 61-5 A double tube is used and allograft bone is packed around the outer tube (A); The cement is injected through the smaller tube, which is slowly withdrawn to the level of the outer tube; then both tubes are withdrawn simultaneously while injecting cement into the void created by the larger tube (B). |
|
Figure 61-6 Marked osteolysis (A) effectively treated with impaction grafting at 6 years (B). |
|
Figure 61-7 Patient with severe bone loss and fracture after failed Coonrad/Morrey implant (A). At 2 years, the struts have incorporated with a successful clinical and radiographic outcome (B). |
Strut Graft
This technique is especially useful for types II and III periprosthetic fractures and for distal humeral or proximal ulnar bone loss. The most effective application to the humerus is that of an anterior strut that transverses the osteolysis or fracture and captures the flange anteriorly. A posterior strut is used to enhance stability and to prevent the wire cutting through the host bone. The use of an extended flange, anterior strut graft, and 2 cm of shortening allows management of distal humeral deficiencies of ≤7 cm (Fig. 61-7).
Allograft-Prosthetic Composite (APC) Reconstruction
When bone loss is significant, either a custom implant, a component reconstruction with strut graft, or an allograft-prosthesis composite (APC) are options. In this last option, the prosthesis is fitted and cemented to the allograft. When that has hardened, the prosthesis is cemented in the host. The difficulty of obtaining interface union between the host and allograft bones has limited the use of this strategy.
Conclusion
Revision elbow arthroplasty is a challenging procedure. An evolution in prosthetic components and varied surgical options has allowed the surgeon to deal with different presentations of TEA failure. Although the complication rate is higher than in primary TEA, the frequency of problems continues to decrease as experience with revision surgery increases.
Suggested Readings
Augereau B, Mansat P. Total Elbow Arthroplasty [in French]. Rev Chir Orthop. 2005;91(suppl 5):2S31–2S96.
Gschwend N, Simmen BR, Matejovsky Z. Late complications in elbow arthroplasty. J Shoulder Elbow Surg. 1996;5:86–96.
Kamineni S, Morrey BF. Proximal ulnar reconstruction with strut allograft in revision total elbow arthroplasty. J Bone Joint Surg Am. 2004;86:1223–1229.
King GJ, Adams RA, Morrey BF. Total elbow arthroplasty: revision with use of a non-custom semiconstrained prosthesis. J Bone Joint Surg Am. 1997;79:394–400.
Lee BP, Adams RA, Morrey BF. Polyethylene wear after total elbow arthroplasty. J Bone Joint Surg Am. 2005;87:1080–1087.
Loebenberg MI, Morrey BF, O'Driscoll SW. Impaction grafting in revision total elbow arthroplasty. J Bone Joint Surg Am. 2005;87:99–106.
Mansat P, Adams RA, Morrey BF. Allograft-prosthesis composite for revision of catastrophic failure of total elbow arthroplasty. J Bone Joint Surg Am. 2004;86:724–735.
Morrey BF. Complications of elbow replacement surgery. In: Morrey BF, ed. The Elbow and Its Disorders. 3rd ed. WB Saunders; 2000: 667–677.
O'Driscoll SW, Morrey BF. Periprosthetic fractures about the elbow. Orthop Clin North Am. 1999;30:319–325.
Redfern DRM, Dunkley AB, Trail IA, et al. Revision total elbow replacement using the Souter-Strathclyde prosthesis. J Bone Joint Surg Br. 2001;83:635–639.
Ring D, Kocher M, Koris M, et al. Revision of unstable capitellocondylar (unlinked) total elbow replacement. J Bone Joint Surg Am. 2005;87:1075–1079.
Sanchez-Sotelo J, O'Driscoll SW, Morrey BF. Periprosthetic humeral fracture after total elbow arthroplasty treatment with implant revision and strut allograft augmentation. J Bone Joint Surg Am. 2002;8:1642–1650.