Elbow resection due to infection with a good result: (a) anteroposterior (AP) and lateral view of resection arthroplasty; (b) flexion and (c) extension range of motion at final follow-up
11.2.4.3 Debridement and Implant Retention
Debridement and implant retention is the treatment of choice for patients with well-fixed components and an acute infection (less than 3 months) not caused by S. epidermidis. Yamaguchi et al. [7] reported the results of this approach in 14 patients in whom they unlinked both components, replaced the polyethylene and performed a throughout debridement acutely. Local antibiotics, including tobramycin, were placed in the joint and also intravenous antibiotic was administered. Patients underwent several debridements until the joint was considered free of infection and the cultures were negative. This treatment approach was only effective in 50% of cases. None of the patients infected with S. epidermidis healed.
11.2.4.4 One-Stage Revision
One-stage revision knee or hip arthroplasty has been successful in some scenarios. However, there is little information available on this approach for treating infections after total elbow replacement.
11.2.4.5 Two-Stage Revision
Two-stage revision arthroplasty is considered the gold standard in the treatment of periprosthetic elbow infections. After the humeral and ulnar components are extracted, the whole cement mantle is removed. Usually, a cement spacer loaded with antibiotics is placed in the joint. After a minimum of 6 weeks of intravenous antibiotic treatment adapted to the isolated microorganism, patients are considered for re-implantation. Before re-implantation, all blood markers must be within normal limits and joint aspiration or open biopsy must be negative. The reported success rate of two-two-stage re-implantation is between 70% and 90% [13, 14].
In most cases requiring several surgical procedures to address an infection, patients will present with significant bone loss. Surgeons dealing with this problem must be prepared to restore bone stock with bone struts, impaction grafting techniques or allograft-prosthetic composites depending on the specific situation.
11.2.5 Management of Bone Loss
Failed elbow replacement commonly presents with significant bone loss. Bone reconstruction strategies include augmentation with bone struts; impaction grafting for expanded, contained, cavitary defects and the use of allograft-prosthetic composites for large segmental defects.
Distal humerus bone loss is defined as grade I when the subchondral architecture is intact, grade II when the medial and lateral supracondylar columns are preserved, grade III when either the medial or the lateral columns were absent and grade IV when the entire distal humerus to or proximal to the level of the olecranon fossa is absent [15].
Revision humeral stems could accommodate up to 8 cm of distal humeral bone loss by using longer implants with an extended anterior flange. Larger defects may require some degree of humeral shortening. According to the location and degree of bone loss, several surgical strategies can be implemented.
11.2.5.1 Impaction Grafting
Impaction grafting is a reliable technique for treating osteolysis in patients undergoing revision total elbow arthroplasty when there are expanded, contained, cavitary defects. The technique is very similar to the one described for reconstructing the proximal femur in revision hip arthroplasty.
Bone stock is restored by filling the medullary canal with impacted morselized allograft. If any segmental cortical defect of cortical thinning is present, it must be previously augmented with allograft bone struts. Once the canal is cleaned and the cancellous allograft introduced in the medullary canal, it is sequentially impacted with a trial component. The defect surrounding the stem must be tightly packed in order to add stability to the final construct. Once stabilized, the trial component is removed, gentamicin loaded low-viscosity cement is injected and the definitive component implanted. Morrey et al. popularized the use of two tubes when impacting the bone graft, the outer tube could be the standard femoral cementation tube and, the inner one, the thinner cement injector tube used in elbow replacement. The allograft is pressed around the outer tube, and both tubes are removed while cement is injected.
Loebenberg et al. [16] published their results with elbow impaction grafting in 12 patients. After a minimum 2 years follow-up, eight patients showed radiographic restoration of bone quality without signs of loosening, and four patients required new revision surgery: for loosening in two patients, and infection and fracture in one patient each.
More recently, Rhee et al. [17] analyzed 16 cases of revision total elbow arthroplasty with impaction grafting in aseptic loosening. At the latest follow-up, 15 of the 16 patients showed significant improvement. Only two patients required further surgery: one periprosthetic humeral fracture and one superficial infection that resolved with debridement.
11.2.5.2 Allograft Bone Struts
Bone strut allografts are generally used in cortical defects. Depending on the size, location and morphology of the defect, struts could be used to cover a discrete cortical deficiency, bypass a fracture or to augment a thinned cortex during impaction grafting. Allograft struts have also been used to augment a deficient olecranon to provide a good triceps attachment site [18].
The source of the allograft should be selected according to the size of the patient, the characteristics of the bony defect and the bone affected. A large structural defect in a small person could be treated with a fibular allograft while the same defect in a larger person might require a femoral shaft allograft.
Sanchez-Sotelo et al. [3] demonstrated satisfactory results in the treatment of periprosthetic humeral fractures around a loose humeral component using strut allograft augmentation in revision arthroplasty. Clinical and radiographic results were satisfactory and fracture union was achieved in 10 of 11 patients. However, the complication rate was elevated with four patients reporting one complication and two patients reporting two complications. Complications included olecranon fracture, permanent ulnar nerve injury, periprosthetic humeral fracture and a case of triceps insufficiency.
Kaminemi and Morrey [18] reported their experience treating aseptic failure of total elbow arthroplasty associated with proximal ulnar bone deficiency with allograft bone struts. In 21 patients, the mean Mayo Elbow Performance Score (MEPS) improved from 34 pre-operatively to 79 points at the time of latest follow-up (2–11 years). Eight patients (38%) suffered a complication.
Foruria et al. [19] reported on 21 patients with periprosthetic ulnar fractures associated with loosening that were treated with revision of the ulnar stem and strut allografts in 12 cases (eight of them with associated impaction grafting), three impaction grafting alone and five with allograft ulnar prosthesis composite. In two elbows, fracture fixation was achieved with a revision longer stem only. All patients had fracture healing and postoperative MEPS of 82 points. Complications included four infections, one ulnar component loosening and one case of transient dysfunction of median and radial nerves.
11.2.5.3 Allograft-Prosthetic Composites (APC)
Allograft composites consist of a hybrid structure composed of a segmental bone allograft with a cemented prosthetic stem inside. The distal part of the stem is left free to be cemented into the native bone. This technique is mostly used in severe bone loss, typically leading to olecranon insufficiency or distal humerus resorption.
Allograft selection is made according to the affected bone. Humeral allografts are commonly used in humeral defects and ulnar or fibular allografts in ulnar defects. The length of the graft can be estimated by sizing the length of the contralateral side. Composite-host bone union is typically made step-cut to increase the contact area and promote healing.
Fixation is achieved by cementing the implant into the prosthetic composite and also in the host bone. The implant should bypass at least two cortical diameters’ length of the prosthetic component into the host bone. Cancellous bone graft can be packed around the junction. Stability may be increased by plating or a cerclage fixation. In proximal ulnar reconstruction, drill holes could be used to reattach the triceps.
Renfree et al. [20] reported their results in 14 allograft-prosthetic composites with a mean follow-up of 6.5 years. They demonstrated 79% radiographic healing but only four patients achieved a functional range of motion and five patients failed and required another operation.
Mansat et al. [21] reported the results of revision elbow arthroplasty with an APC in 13 elbows. Eighty-five percent of cases achieved union. The MEPS scale was excellent for four elbows, good for three, fair for one and poor for five. Range of motion was limited to an average of 28° of extension to 125° of flexion. The revision rate was 38% and the complication rate was high, infection being the most common one.
Type-I (intussusception): the cemented implant and the allograft were inserted into the host bone canal. It is indicated in contained defects with intact cortical bone.
Type-II (strut-like coaptation): similar to the original technique, the implant and the allograft composite are fixed to the host bone in a step-cut osteotomy. Two cortical widths are left nude to be fixed to the native bone. It is used when there is a major cortical defect at the implant insertion site.
Type-III (side-to-side): a side-to-side contact between the cortices of the allograft composite and the native bone is formed.