After receiving institutional review board approval, we retrospectively reviewed the medical records of all patients treated with the Coonrad-Morrey total elbow arthroplasty at the Mayo Clinic between 1982 and 2003. Seven hundred and fifty-eight primary total elbow arthroplasties had been performed, and sixty-one had been done in fifty-five patients who were forty years of age or less. After the exclusion of total elbow arthroplasties performed after tumor resection or in patients with a hemophilic condition, fifty-five implants in forty-nine patients (six of whom had a bilateral arthroplasty) constituted our study group.
The arthroplasty was performed primarily as a salvage procedure in thirteen patients (two with a bilateral arthroplasty) with rheumatoid arthritis, in seventeen patients (four with a bilateral arthroplasty) with juvenile rheumatoid arthritis, and in nineteen patients with a posttraumatic condition. All of the cases of rheumatoid arthritis and juvenile rheumatoid arthritis were stage 3 or 4 according to the Mayo Clinic classification system8. The patients with a posttraumatic condition all exhibited severe osteoarthritis, nonunion, or gross instability that was thought to be not amenable to any other salvage reconstruction or had been treated, before the joint replacement, with several (range, two to six) operative procedures that had failed.
The Coonrad-Morrey prosthesis (Zimmer, Warsaw, Indiana) used in all patients was a titanium linked semiconstrained implant with cement fixation of both components8. The implant incorporates an anterior flange and a "loose hinge" with approximately 7° to 8° of varus-valgus laxity and 7° to 8° of rotational freedom10. During the study period, three surface preparations were employed for the ulna. The humeral and ulnar components were available in several sizes and lengths to better accommodate anatomic variations.
The results of the preoperative and postoperative clinical evaluations were rated with use of the Mayo Elbow Performance Score (MEPS)8. The maximum score with this system is 100 points: 45 points for no pain, 20 points for >100° of motion, 10 points for joint stability, and 25 points for the capacity to perform selected daily activities. The results are categorized as excellent (90 to 100 points), good (75 to 89 points), fair (60 to 74 points), or poor (<60 points). For this study, excellent and good results were considered to be satisfactory and fair and poor results were considered to be unsatisfactory. Final follow-up radiographs, or the reports on radiographic evaluations that the patients had locally, were examined for radiolucency around the implants.
Sixteen patients with sixteen involved elbows (29%; three with rheumatoid arthritis, seven with juvenile rheumatoid arthritis, and six with posttraumatic arthritis) returned to our institution for the clinical and radiographic evaluations. All of the other patients were assessed on the basis of evaluations performed after the longest follow-up time by their local physician using the MEPS questionnaire and a table to guide measurement of the range of motion. The reports of radiographic evaluations performed locally were sent to us for review.
Statistical Assessment
The numerical differences between the results for the patients with an inflammatory condition and those with a traumatic condition were assessed with a chi-square statistical analysis. Differences with less than a 5% likelihood of occurring by chance (p < 0.05) were considered significant. Comparisons among the diagnostic categories were conducted with use of a two-sample t test, assessing unequal variances.
Source of Funding
No extramural funds were received for this study. Department infrastructure was used for the preparation of this manuscript.
Overall Clinical Results
The mean patient age in this series was thirty-two years (range, twenty-four to forty years) at the time of total elbow arthroplasty, and the mean duration of follow-up was ninety-one months (range, twenty-four to 242 months). Thirteen elbows (24%) had the last clinical and radiographic assessments within five years after the surgery, twenty-eight elbows (51%) were evaluated between five to ten years postoperatively, and fourteen elbows (25%) had the last follow-up more than ten years after the arthroplasty.
Two major diagnostic groups were represented in this study. Thirty-six replacements (65%) were performed for an inflammatory condition and nineteen replacements (35%) were done for posttraumatic arthritis.
The mean preoperative MEPS (and standard deviation) was 37 ± 15.8 points, and the mean latest postoperative MEPS was 91 ± 13.2 points. According to the MEPS, the result at the latest follow-up evaluation was excellent in thirty-six elbows, good in fifteen, poor in three, and fair in one (Table I). We also stratified the patients by diagnosis and assessed the MEPS scores before and after the surgery. In all three subgroups, the difference between the MEPS before the surgery and that after the surgery was significant (p < 0.004 for all three groups).
Inflammatory Arthritis
Juvenile Rheumatoid Arthritis
Three men and fourteen women had juvenile rheumatoid arthritis (see Appendix); four patients had bilateral implants. The mean age at the time of surgery was thirty-one years (range, twenty-four to forty years), and the mean time between the diagnosis of juvenile rheumatoid arthritis and the elbow replacement was 19.2 years (range, eleven to twenty-five years). Two patients had had a previous arthroscopic synovectomy of the elbow.
The average duration of follow-up was ninety-one months (range, twenty-four to 242 months). On the basis of the MEPS, fourteen (67%) of the twenty-one elbows had an excellent result; six, a good result; and one, a fair result.
Rheumatoid Arthritis
One man and twelve women had rheumatoid arthritis (see Appendix); two patients had bilateral implants. The mean age of this group at the time of the surgery was thirty-five years (range, twenty-seven to forty years), and the mean time between the diagnosis of rheumatoid arthritis and the elbow replacement was 12.7 years (range, seven to twenty years). Three patients had had prior surgery: arthroscopic débridement and synovectomy had been performed in two and an interpositional arthroplasty, in one.
According to the MEPS, twelve of the fifteen elbows had an excellent result and three had a good result at an average of eighty-eight months (range, thirty to 185 months) postoperatively.
Posttraumatic Arthritis
Seven men and twelve women had posttraumatic arthritis (see Appendix). The mean age at the time of the elbow replacement was thirty-three years (range, twenty-two to forty years). The mean time between the elbow injury and the replacement was 6.6 years (range, one to thirty-three years), and the average number of previous operations was three (range, one to six). The mean duration of follow-up was 102 months (range, twenty-four to 242 months). According to the MEPS, ten of the nineteen elbows had an excellent result, six had a good result, two had a fair result, and one had a poor result.
Pain Relief and Motion
Relief of pain was similar among the three diagnostic categories. Motion was more restricted before and after the replacement in the group with posttraumatic arthritis.
Daily Function
Before the surgery, the ability to perform activities of daily living (such as combing the hair, feeding oneself, personal hygiene, and putting on pants and shoes) was severely limited by pain and a restricted range of motion in all three groups. The mean MEPS for function (maximum, 25 points) before the surgery was 1.4 points in the juvenile rheumatoid arthritis group, 6 points in the rheumatoid arthritis group, and 7 points in the posttraumatic arthritis group. After elbow replacement, the mean scores were 21, 24, and 17 points, respectively. Thirty-one (56%) of the fifty-five elbows in the three groups combined were considered by the patient to cause no difficulties with any of these activities of daily living at the time of the last follow-up.
Radiographic Results
Of the fifty-five implants, forty-nine (89%) had no evidence of progressive radiolucency around either the humeral or the ulnar component at the time of the final assessment. One elbow in the rheumatoid arthritis group had a 1-mm-thick radiolucent line around <50% of the ulnar component at ten years after the surgery; it was clinically asymptomatic at the time of the most recent assessment. Three elbows, all in the posttraumatic arthritis group, had a complete radiolucent line around the ulnar component, with minimal symptoms. One elbow in the posttraumatic arthritis group had a complete radiolucent line around the humeral component and was asymptomatic. One elbow in the rheumatoid arthritis group had complete loosening of the implant seen radiographically but was not painful and was being observed at the time of this report.
Complications and Reoperations
There were no intraoperative complications. Three patients in the posttraumatic arthritis group and two in the rheumatoid arthritis group reported postoperative ulnar nerve sensory deficiency without motor deficiency. The sensory resolved in all of these patients. Two patients had a superficial wound infection, which was treated effectively with antibiotics.
There were fourteen delayed complications (25%): two superficial infections, two deep infections, four cases of aseptic loosening of a precoated ulnar component, three cases of triceps insufficiency, and three implants with bushing wear. These delayed complications developed in seven of the nineteen patients with posttraumatic arthritis and seven (23%) of the thirty patients with an inflammatory condition. With the small numbers of patients, we could not identify a significant difference between the two groups.
Twelve elbows, seven in the posttraumatic arthritis group and five in the two inflammatory arthritis groups, required a revision procedure because of a complication (see Appendix). Four operations were done because of loosening; three, because of triceps weakness; three, because of wear; and two, because of deep infection. The complication rate for the patients with posttraumatic arthritis was significantly greater than that in the groups with inflammatory arthritis (p = 0.02). The mean time after the index arthroplasty for the twelve reoperations was fifty-nine months (range, nine to 120 months).
Comparison of Outcomes Between Inflammatory and Posttraumatic Arthritis Groups
There was a significant difference in the mean MEPS score between the inflammatory arthritis (93 points) and posttraumatic arthritis (84 points) groups (p = 0.03). There was also a significantly greater rate of radiographic evidence of loosening in the posttraumatic arthritis group (p < 0.02). Finally, there was a greater rate of reoperations in the posttraumatic arthritis group (37% compared with 11%) (p < 0.05).
Between 1982 to 2003 at our institution, 8% of all total elbow arthroplasties were performed in patients who were forty years of age or less. The forty-nine patients included in our study were carefully selected for the arthroplasty, and the arthroplasties represented salvage interventions in two major diagnostic groups: posttraumatic and inflammatory conditions. The salvage character of the operation is reflected by the fact that all of the patients with a traumatic diagnosis had had two or more prior operations, and all of those with inflammatory arthritis had advanced disease8. The surgical management of these patients is very challenging with few reliable options11-18.
Synovectomy with débridement and radial head resection is recommended for elbows with stage-1 or 2 rheumatoid involvement11,12. However, elbow replacement is considered to be the most reasonable option for elbows with stage-2, 3, or 4 rheumatoid involvement, even in a younger patient with considerable loss of bone stock. The stability of the linked implant does not depend on soft-tissue integrity, and the anterior flange of the implant provides stable fixation of the humeral component even when there is loss of the distal humeral bone. Alternative reconstructive procedures such as arthrodesis are problematic in young patients since fusion is difficult to obtain when the bone stock is deficient19. More importantly, fusion has a high complication rate and is functionally disabling20-22.
The use of a distal humeral allograft has been reported as an option for treatment of posttraumatic conditions in patients who are considered too young for total elbow arthroplasty19. However, several major complications have been described, including nonunion, infection, and graft resorption23,24. Furthermore, total elbow arthroplasties have been reported to have less favorable results when they are performed following failure of a reconstruction with a distal humeral allograft25. Although we did perform allograft reconstruction during this time period in selected circumstances, we believe that the results of this procedure are too unpredictable to be widely used in the young population22.
Interposition arthroplasty is a viable option for young patients with elbow arthritis26-29. However, this technique is not considered suitable in the presence of extensive bone loss30 or in patients with gross instability or a fixed deformity20. Over the time of this study, we performed 133 interposition arthroplasties and traumatic arthritis was the indication for ninety-four (71%) of the 133 procedures.
During the study period, our postoperative recommendations included avoiding lifting of >1 kg repetitively or 4 kg as a single event. On the basis of discussions with our patients, it appears that the vast majority modified their daily living activities to follow the postoperative protocol.
Most of the overuse complications occurred in the posttraumatic arthritis group. Although the overall complication rate in our series (25%) was high, it is comparable with that reported in other studies involving older populations of patients and is lower than that reported following some of the alternative procedures29-34.
We were surprised to learn that a group of older patients treated at our institution for posttraumatic arthritis with the same prosthetic implant had similar results30. According to the MEPS, 83% of the older patients30 and 84% of the patients with posttraumatic arthritis in the current study had a satisfactory outcome (a score exceeding 75 points). The outcomes in our patients with rheumatoid arthritis were similar to those reported for older patients with rheumatoid arthritis as well6,30-34.
In our experience, the most important factors that have negatively influenced the results in this younger age group have been involvement of the dominant side, activity level, and a traumatic diagnosis.
This study had several weaknesses. First, it is retrospective and has the inherent weaknesses of such an analysis. Second, the activity levels of the patients were, at best, estimates. Finally, and most importantly, the majority of the patients did not return to us for a final assessment. Thirty-nine (71%) of the fifty-five evaluations were carried out by local physicians, which introduces an additional assessment variable.
In conclusion, our experience indicates that elbow replacement in patients who are forty years of age or less can provide good pain relief and increase the elbow range of motion with an acceptable survival rate of the implant. However, the revision rate is high and we continue to prefer non-replacement options such as synovectomy and interposition arthroplasty when possible.