Infection following total elbow arthroplasty is a relatively rare complication, which was previously reported to have occurred in twenty-five (3.3%) of 757 cases at our institution1,2, but the rate is higher than that reported after hip or knee arthroplasty3-7. This difficult complication has been treated in the past with simple débridement, resection arthroplasty, arthrodesis, or removal followed by reimplantation of the total elbow components. Although resection arthroplasty may be a reasonable option for definitive treatment of an infection in elderly patients who make relatively low demands on the elbow, the procedure can have unsatisfactory results. The loss of the medial and lateral humeral condyles can result in gross instability, and marked shortening of the limb decreases muscle strength, ultimately resulting in a flail elbow. Arthrodesis of the elbow is another option, but it subjects the patient to the morbidity of another surgical procedure and it can be disabling as a result of loss of functional motion. Another solution is prosthetic reimplantation; however, there is a paucity of literature regarding the outcome of reimplantation of total elbow prostheses after resection arthroplasty. The purpose of this study was to review our experience with reimplantation following resection arthroplasty to treat an infection at the site of a total elbow arthroplasty and to report the outcomes, the risk factors for an unsatisfactory result, and the rate of failure.
Materials and Methods
A review of our institution's computerized total joint registry revealed that, between 1976 and 2003, thirty patients had undergone reimplantation of a total elbow prosthesis after the performance of a resection arthroplasty to treat an infection at the site of a total elbow arthroplasty. The medical records of these patients were retrospectively reviewed. One patient had died less than two years after reimplantation and was excluded from the study. Thus, we report on twenty-nine elbows in twenty-nine patients. This study was approved by our institution's internal review board, and the patients provided informed consent allowing their medical information to be used in this study.
Postoperatively, all patients either were clinically examined at our institution (nine patients) or completed a questionnaire that had been sent by mail (sixteen patients) or was administered by telephone (four patients). No patients were lost to follow-up. The range of motion of the elbow and function as measured with the Mayo Elbow Performance Score (MEPS)8 were documented. The outcome of the reimplantation was characterized as a success or a failure with regard to recurrence of the infection and the functional outcome. Postoperative complications were also documented.
Our operative technique has been previously described9. The first stage, resection of the components and débridement, is done through a midline incision, and when possible the posteromedial portion of the triceps is reflected in a subperiosteal fashion in continuity with the forearm fascia. The prosthetic components and any remaining cement are removed in a meticulous fashion with small curets, a motorized burr, and a motorized router. Recently, this step has been facilitated by removing the posterior cortex of the distal part of the humerus to create a window. Care is taken to preserve the integrity of the humerus so that the tip of a reimplanted component will be two diameters proximal to the osteotomized cortical window (Fig. 1). Every effort is made to preserve the integrity of the medial and lateral condyles because they provide structural stability for a resection arthroplasty in which the ulna should articulate within the space between them. The condyles also serve as the origins of the common flexor and extensor muscles of the forearm, and disruption of these origins can compromise their function.
Removal of a well-fixed ulnar component is more difficult. A pencil-tipped burr is used to remove cement from the proximal aspect of the ulnar component in the region of the plasma coating. This process then continues distally as needed for removal of the device. Only occasionally have we had to perform an extended osteotomy along the medial-proximal aspect of the ulna.
When staged reimplantation is anticipated, antibiotic-impregnated cement (2 g of tobramycin and 2 g of vancomycin per 40 g of cement) is mixed and placed as needed as a spacer down the shafts and between the humerus and ulna, and the incision is closed. The patient is treated with a six-week course of intravenous antibiotics, based on organism sensitivity. Prior to reimplantation of a total elbow prosthesis and completion of the six-week course of antibiotics, we routinely check that a complete blood-cell count with differential, the erythrocyte sedimentation rate, and the C-reactive protein level are within normal limits. An aspiration of the elbow with a cell count and culture is also performed to confirm the absence of active infection.
At the second stage (reimplantation of a new prosthesis), we first obtain at least three tissue samples for frozen-section evaluation to confirm the absence of acute inflammation. If an antibiotic spacer was used, it is removed. If there is no histologic evidence of acute inflammation, we proceed with reimplantation with use of the same incision as was used in the first stage. The new prosthesis is implanted with use of antibiotic impregnated cement (1 g of tobramycin per 40 g of cement). If a humeral or ulnar osteotomy was required in the first stage, the site is repaired with strut allografts and cerclage wires. No additional antibiotics are given after the immediate perioperative period of the reimplantation. The patients are reminded that there will be lifetime postoperative activity restrictions: no lifting of >10 lb (>4.5 kg) as a single event and no repetitive lifting of >2 lb (>0.9 kg).
The MEPS was used to assess function and patient satisfaction preoperatively and at the time of the latest follow-up. The MEPS is the sum of the following parameters: pain (0 to 45 points), motion (5 to 20 points), stability (0 to 15 points), and function (0 to 25 points). An excellent score is ≥90 points, a good score is from 75 to 89 points, a fair score is from 60 to 74 points, and a poor score is <60 points8.
Survival free of revision surgery was estimated as a function of time since the index procedure with the use of the Kaplan-Meier method. A nonparametric Wilcoxon rank sum test was performed to assess associations between preoperative and postoperative MEPS and nominal variables. A p value of <0.05 was considered to be significant.
The average age at the time of reimplantation of the total elbow prosthesis was 61.6 years (range, twenty-eight to seventy-eight years). Twenty-three of the patients were women, and six were men. The dominant extremity was involved in twenty-six cases. The underlying diagnosis at the time of the index arthroplasty was rheumatoid arthritis in nineteen patients, posttraumatic arthritis in five, osteoarthritis in three, and acute fracture in two (Table I). A Coonrad-Morrey total elbow prosthesis (Zimmer, Warsaw, Indiana) was used at the time of reimplantation in twenty patients; a Coonrad III prosthesis (Zimmer), in eight; and a Pritchard-Walker prosthesis (DePuy, Warsaw, Indiana), in one.
Eleven of the twenty-nine patients had had at least one procedure performed on the elbow before the index arthroplasty. One patient each had had open reduction and internal fixation, open reduction and internal fixation with subsequent removal of hardware, implantation of a silicone radial head and subsequent removal, irrigation and débridement because of infection when the patient was a child, interposition arthroplasty, and placement of an external fixator to treat a closed fracture-dislocation. Five patients had had a prior total elbow arthroplasty, so the index operation was a revision total elbow arthroplasty. Two of those revisions were done because of loosening; two, because of instability; and one, because of a periprosthetic ulnar fracture.
The presenting symptom was pain and swelling in eighteen elbows and wound dehiscence with a draining sinus in eleven. The average number of débridements prior to the resection arthroplasty was 1.9 (range, zero to three) (Table II). The mean time interval was 123 weeks (range, seven to 365 weeks) from the primary total elbow arthroplasty to the resection arthroplasty, 72.5 weeks (range, eight to 707 weeks) from the resection arthroplasty to the reimplantation, and 7.4 years (range, 2.2 to 28.6 years) from the reimplantation to the last follow-up.
All patients were treated with a course of antibiotic therapy after the resection arthroplasty. Twenty-five patients received intravenous antibiotics, for a mean duration of forty-two days. Three of those patients received intravenous antibiotics for forty-two days followed by a long-term course of oral antibiotics for lifelong suppression. Three patients who had had negative intraoperative cultures and one who had had cultures that were positive for Staphylococcus epidermidis were treated with oral antibiotics alone for forty-two days after the standard period of twenty-four hours of antibiotic prophylaxis.
The mean total MEPS before the reimplantation was 35.5 points (range, 15 to 60 points): 26.7 points (range, 0 to 45 points) for pain, 5 points (0 to 20 points) for motion, 1.7 points (range, 0 to 15 points) for stability, and 2.2 points (range, 0 to 15 points) for function (Table III). The mean total MEPS after the reimplantation was 66.3 points (range, 20 to 100 points): 26.9 points (range, 0 to 45 points) for pain, 16 points (range, 5 to 20 points) for motion, 8.8 points (range, 5 to 10 points) for stability, and 14.7 points (range, 0 to 25 points) for function. There was a significant improvement in the total MEPS after the reimplantation (p < 0.001).
After reimplantation, five elbows (17%) were rated as excellent, ten (34%) were rated as good, three (10%) were rated as fair, and eleven (38%) were rated as poor. Eight of the eleven elbows that were rated as poor were considered to be treatment failures because the infection recurred and required a repeat resection arthroplasty, at a mean of twenty-two months (range, one to seventy-two months) postoperatively.
With the numbers studied, no significant associations were identified between the postoperative MEPS and age, hand dominance, underlying diagnosis, duration of symptoms, or organism grown on culture at the time of the surgery. In addition, the outcome with regard to reinfection or the need for a reoperation was not associated with the above parameters.
The most common infecting organism grown on culture of specimens taken at the time of the resection arthroplasty was Staphylococcus epidermidis, which was present in thirteen (45%) of the twenty-nine elbows, followed by methicillin-sensitive Staphylococcus aureus (seven) and Klebsiella pneumonia (two). Cultures of the specimens from three elbows were negative. In those cases, infection was diagnosed clinically on the basis of wound dehiscence or a draining wound with an exposed prosthesis. Group-A Streptococcus, aerobic diphtheroids, Propionibacter acnes, and methicillin-resistant Staphylococcus aureus accounted for the infections in the remaining four patients.
There were three complications, in addition to the eight recurrences of infection. Two patients had a periprosthetic ulnar fracture, at one year and four years postoperatively. Both were treated with ulnar component strut grafting. One of these fractures healed, and the patient had a MEPS of 80 points at five years postoperatively. In the other patient with an ulnar fracture, a deep infection subsequently developed, requiring resection arthroplasty. The patient had a poor functional result, with a MEPS of 25 points at two years postoperatively. The third patient had a periprosthetic humeral fracture with an associated radial nerve palsy one month postoperatively; this eventually required a free vascularized fibular graft to secure healing of the fracture and sural nerve grafting to restore radial nerve function. The patient had partial neurological recovery and a MEPS of 85 points postoperatively.
The three-year estimate for survival free of a reoperation was 77.0% (95% confidence interval, 62.3% to 91.7%), and the eight-year estimate of survival free of a reoperation was 48.0% (95% confidence interval, 27.4 to 68.6%) (Fig. 2).
It has been our experience that resection arthroplasty can be a quite acceptable salvage procedure for treatment of an infection at the site of a total elbow arthroplasty. However, for it to be successful, the distal part of the humerus must serve as a fulcrum; thus, both the medial and the lateral condyles must be present. These are often lost either before or at the time of implant removal for treatment of an infection at the site of a total elbow arthroplasty. In our experience, resection arthroplasty has tended to be an unacceptable option for most patients if the condyles are not present.
Overall, eight (28%) of the twenty-nine patients were considered to have had a treatment failure because of recurrent infection. Four of the eight were infected with Staphylococcus epidermidis, which was the most common organism causing the infections around the total elbow prostheses at the time of resection arthroplasty; it was identified in thirteen (45%) of the twenty-nine elbows. Five of these thirteen elbows were regarded as treatment failures. Vigilance with gentle soft-tissue handling, meticulous débridement at the time of the resection arthroplasty to include any remnants of cement left within the humeral and ulnar medullary canals, and consultation with infectious disease specialists should be performed in order to optimize the chance of eradicating this difficult organism. It would perhaps not be unreasonable to suggest that patients who are elderly or medically unfit for multiple surgical procedures should consider accepting the resection arthroplasty as definitive treatment, to avoid the morbidity of revision procedures. Instability may be addressed in these cases with the use of a hinged elbow brace. An adjunctive treatment option for patients with a Staphylococcus epidermidis infection, after surgery and completion of a course of intravenous antibiotics, includes long-term oral antibiotic suppression.
At our institution, the orthopaedic infectious disease service is consulted at the time of an initial diagnosis of a periprosthetic infection. Prior to reimplantation of a total elbow prosthesis and completion of a six-week course of antibiotics after resection arthroplasty, we routinely check the complete blood-cell count with differential, the erythrocyte sedimentation rate, and the C-reactive protein level, and aspiration of the elbow with a cell count and culture is performed to confirm the absence of active infection. Intraoperative tissue samples are sent for frozen-section analysis to confirm the absence of acute inflammation prior to cementing a new prosthesis.
Of note, four patients in this series were treated with oral antibiotics alone, more than fifteen years before the time of this review. Three of these patients had negative intraoperative cultures, and the fourth patient was infected with Staphylococcus epidermidis. One of the three patients who had negative cultures had a good result (a MEPS of 80 points), and the remaining two patients had a poor result (a MEPS of 55 points). The patient who was infected with Staphylococcus epidermidis had a poor result, with recurrence of the infection. Thus, we do not recommend treatment of an infection at the site of a total elbow arthroplasty with surgery and oral antibiotics alone.
Limitations of this study include its retrospective nature and the relatively small number of patients. In addition, there were variations in treatment algorithms that correspond to the long period of time over which the patients were treated at this institution. However, to our knowledge, this is the largest series to date in which reimplantation of a total elbow prosthesis following resection arthroplasty for infection was analyzed.
This study suggests that reimplantation of a total elbow prosthesis after a prior resection arthroplasty for the treatment of infection can be a reasonable option, with marked improvement in function in patients who are dissatisfied with the function of the resection arthroplasty (Figs. 3-A through 3-D). As with other joints, function of the elbow is compromised after reimplantation, even when the infection is treated successfully.
Disclosure: The authors did not receive any outside funding or grants in support of their research for or preparation of this work. One or more of the authors, or a member of his or her immediate family, received, in any one year, payments or other benefits in excess of $10,000 or a commitment or agreement to provide such benefits from a commercial entity (Zimmer). No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, division, center, clinical practice, or other charitable or nonprofit organization with which the authors, or a member of their immediate families, are affiliated or associated.
Investigation performed at the Mayo Clinic, Rochester, Minnesota
1. Yamaguchi K, Adams RA, Morrey BF. Infection after total elbow arthroplasty. J Bone Joint Surg Am. 1998;80:481-91.
2. Yamaguchi K, Adams RA, Morrey BF. Semiconstrained total elbow arthroplasty in the context of treated previous infection. J Shoulder Elbow Surg. 1999;8:461-5.
3. Callaghan JJ, Albright JC, Goetz DD, Olejniczak JP, Johnston RC. Charnley total hip arthroplasty with cement. Minimum twenty-five-year follow-up. J Bone Joint Surg Am. 2000;82:487-97.
4. Morrey BF, Bryan RS. Infection after total elbow arthroplasty. J Bone Joint Surg Am. 1983;65:330-8.
5. Haleem AA, Berry DJ, Hanssen AD. Mid-term to long-term followup of two-stage reimplantation for infected total knee arthroplasty. Clin Orthop Relat Res. 2004;428:35-9.
6. Schoifet SD, Morrey BF. Treatment of infection after total knee arthroplasty by débridement with retention of the components. J Bone Joint Surg Am. 1990;72:1383-90.
7. Berry DJ, Harmsen WS, Cabanela ME, Morrey BF. Twenty-five-year survivorship of two thousand consecutive primary Charnley total hip replacements: factors affecting survivorship of acetabular and femoral components. J Bone Joint Surg Am. 2002;84:171-7.
8. Morrey BF, An K-N. Functional evaluation of the elbow. In: Morrey BF, editor. The elbow and its disorders. 3rd ed. Philadelphia: WB Saunders; 2000. p 74-83.
Copyright 2008 by The Journal of Bone and Joint Surgery, Incorporated
9. Bryan RS, Morrey BF. Extensive posterior exposure of the elbow. A triceps-sparing approach. Clin Orthop Relat Res. 1982;166:188-92.