INDICATIONS

We reserve resection arthroplasty for septic total elbow replacements, especially when reimplantation cannot be done.

TECHNIQUE

To remove a well-fixed implant, a trapezoid-shaped cortical window is removed from the posterior aspect of the distal humerus (Fig. 67-1). Care is taken to preserve the epicondyles (Fig. 67-2). The olecranon is positioned between the two columns. We use a No. 3 PDS suture to help stabilize this relationship.

FIGURE 67-1 Resecting a well-fixed total elbow arthroplasty can be accomplished by removing a trapezoid-shaped segment of the posterior humerus. This allows access to the medullary canal. The length of this should allow at least two diameters of reinsertion of an implant should that be necessary. Possibly more importantly, this should carefully be done in order to maintain the medial and lateral condyles.

FIGURE 67-2 Patient with resected arthroplasty after an infected total elbow (A and B). Notice that the medial and lateral columns have been maintained (A). This patient has a Mayo Elbow Performance Score of 80 and a functional elbow.

AFTERCARE

A cast is applied for 6 weeks. Protected motion is begun using a Mayo Elbow Brace (Donjoy International, Vista, CA) and continued for an additional 6 to 8 weeks (Fig. 67-3).

FIGURE 67-3 The Mayo Elbow Brace is helpful in immobilizing and stabilizing the extremity in the early phases and then allowing flexion and extension in the later phases.

RESULTS

We have just completed an assessment of 59 resected elbows after failed total elbow arthroplasty (TEA) for sepsis at a mean time of 12 years after surgery. Although the study is still under way, preliminary analysis demonstrates a clear distinction between those with “contained” articulation (Fig. 67-4) having medial and lateral condyles doing markedly better than those with more extensive resection preventing any opportunity for a fulcrum effect (Fig. 67-5). Custom orthotic braces are occasionally prescribed, but infrequently worn on a regular basis (Fig. 67-6).

FIGURE 67-4 Two years after resection arthroplasty, notice the maintenance of the medial and lateral columns.

FIGURE 67-5 A, In some instances, a complete resection of the joint may be indicated. This is usually in the case of sepsis. However, the result is dysfunctional instability. B, The use of the extremity is markedly compromised due to a windmill type of global instability.

FIGURE 67-6 A custom-fit prosthesis is sometimes employed after resection arthroplasty. Compliance is low, however.

SEGMENTAL BONE LOSS

The unstable or “flail” elbow with segmental bone loss, that is, complete loss of articulation and ligament attachments, is difficult to manage. Treatment with a primary replacement is discussed in Chapters 59 and 66. This problem can be a sequel of trauma, infection, failed elbow arthroplasty, or tumor resection. Bone may be lost at the original injury following a severe open fracture, or it may be removed subsequent to infection, particularly if the fragment is avascular. In treating recalcitrant union of supracondylar fractures, surgeons have in the past (and disastrously) succumbed to the temptation to remove the condylar fragment of the humerus and treat the problem by inserting an endoprosthesis.

If the problem is limited to just the articular region of the humerus, a hemireplacement may be effective (Fig. 67-7). If, however, all stability and distal contour of the humerus is lost, today, a linked replacement is performed for such patients (see Chapter 59).

FIGURE 67-7 The instances in which there is loss of articular surface and bone but maintenance of collateral origins allows a hemireplacement of the joint and is an attractive option that has recently emerged.

On the other hand, in the situation of a flail elbow, particularly in a young patient, arthrodesis may be considered (see Chapter 70). However, failure of fusion, even under optimal conditions, has been high in some series^2,14,16 and successful arthrodesis is unlikely in patients with large bone stock deficiencies. Reconstruction procedures, such as those involving reattachment of muscles more proximally on the humerus or interposing a tongue of triceps muscle into the gap, are unlikely to be successful. Muscle transfer procedures can improve flexion in cases of flail elbow associated with paralysis, but in these cases, the problem is of a different nature. A successful outcome will occur only if a stable fulcrum is provided and muscles are restored to their normal functioning length.

LONG-TERM CONSIDERATIONS

The fate of implanted allograft material depends on many factors, some of which are not fully understood. Large segmental allografts must unite with host bone (by a process of osteoconduction) if they are to succeed.⁷ This union occurs frequently, provided that the fixation to the host is technically satisfactory (Fig. 67-9). The standard AO plating technique is usually effective. Considerable loss of bone is observed to occur with time in that portion of the graft remote from the donor host junction. The exact nature of this neurotrophic type response is unknown at this time. One of the advantages of the method is that this gradual loss of allograft over time appears to allow some functional adaptation of muscles so that joint stability sufficient to permit continuing good function remains.³ Subluxation may also occur. Even if further reconstruction is required on some future occasion, it is facilitated, not prejudiced, by the augmentation of the patient’s bone stock that is achieved by the allograft. In fact, if we perform their procedure, it is considered a staged bone augmentation procedure that facilitates elbow replacement should this be necessary.

FIGURE 67-9 Radiograph of a 20-year-old woman with flail elbow following infected open fracture (A). The elbow soft tissues were augmented with a free vascularized skin flap 6 months before elbow reconstruction (here, seen immediately before allograft implantation surgery) (B). Radiograph immediately following the procedure. In this case, a fresh-frozen graft was used (C). Eight months postoperatively, union has occurred (D). The patient is doing office work. The arm is not braced and the joint is stable (no soft tissue reconstruction or ligamentous support of any kind). Three years postoperatively (E). Five years postoperatively (F and G). Note progressive bone loss. The elbow is braced, but the fulcrum is maintained and the arm is functional. No dislocation is present, but there is some posterior subluxation in extension.

BICOMPATIBILITY

The immune response is not universally deleterious to the outcome of these transplants because satisfactory and full incorporation of the allograft into the host’s skeleton does occur. Graft incorporation depends on the mechanical environment and the type of graft as well as immune factors that are complicated and poorly understood.^7,9 The immunogenicity of allografts is reduced by freezing or freeze drying, which renders them more suitable for clinical use. However, even preserved allografts provoke recognizable cellular and humoral antibody responses. Donor-specific anti—human leukocyte antigen (HLA) antibodies and cellular antibodies have been identified in frozen and freeze-dried grafts.⁵ The major reactivity appears to be cellular because viable bone marrow cells and, to a lesser extent, bone cells are highly immunogenic and induce both a cell-mediated and a humoral response. Antigens in the extracellular matrix (collagen and the noncollagenous proteins) seem to be of much less significance in the clinically observed allograft response.⁹ At this time, it has not been possible to identify the specific criteria that define an immune rejection of these grafts, as opposed to failure caused by some other process. Any infection inevitably spells failure for this procedure. As more becomes known of the nature of the rejection phenomenon, it may be possible to favorably manipulate the response, perhaps by tissue typing and immunosuppression of the host.

Some anxiety with disease transmission continues to exist concerning the use of these large bone grafts. In spite of improved screening techniques, one cannot offer these patients absolute assurance that one is not transferring disease from donor to host. Unlike small grafts, such as femoral head grafts, these larger upper limb allografts are inevitably from dead donors. Obviously, this means that the opportunity for donor interrogation is not available. Also, there is no way to identify donors who may convert and become seropositive at a later date.

RESULTS

Urbaniak¹⁵ found total joint transplantation to be successful in six of eight cases, with a relatively short follow-up period of up to 6 years. Subsequent surveillance has tempered this enthusiasm due to predictable deterioriation of the articulation.³

Others have reported somewhat similar results. A French report described patient satisfaction in five of seven patients with TEA replacements.¹ In five of six patients, progressive joint resorption occurred. In our limited experience, ultimate joint deterioration universally occurs, but can be readily managed by a joint replacement (Fig. 67-10). Otherwise, there is little information in the peer-reviewed literature regarding this treatment option.

FIGURE 67-10 A, As shown, a 44-year-old patient with an elbow replacement and early loosening with osseous resorption four years after insertion. The patient was found to have a nickel allergy (A and B). A resection of the joint and allograft elbow replacement was carried out (C and D). Sixteen years later, the allograft has undergone a neurotrophic change. The patient’s elbow was unstable and painful (E and F). An elbow replacement was carried out. The allograft provided a structural basis for the total elbow arthroplasty (G and H).