On review of our hospital records, we identified a consecutive series of fifteen patients (seventeen elbows) with rheumatoid arthritis in whom an acute distal humeral fracture had been initially treated, within ten days after the injury, either with open reduction and internal fixation or with primary total elbow arthroplasty between January 1, 1982, and June 30, 2002. Of the two patients with concomitant bilateral fractures, one was treated with bilateral open reduction and internal fixation and the other was treated with open reduction and internal fixation on one side and with total elbow arthroplasty on the other side. Records were available for all seventeen elbows. Seven patients (seven elbows) had died by the time that this study was initiated; one of these patients (one elbow) had had open reduction and internal fixation and six patients (six elbows) had had total elbow arthroplasty. Six of the patients who had died (six elbows) had been followed clinically and radiographically for more than two years and therefore were included in the study. One patient (one elbow) died of causes unrelated to the elbow arthroplasty at seven months after the surgery and was not included in the study. Fourteen patients with sixteen involved elbows were included in the final analysis. The study was approved by the local institutional review board but did not require patient consent.
Demographic Data on the Patients
The eight men and six women had a median age of sixty-two years (range, forty-seven to eighty-three years) at the time of the fracture. Two men had a bilateral fracture. There were six right and ten left elbows. All sixteen fractures were the result of a fall on the upper extremity without elbow dislocation. There were no open fractures and no neurovascular injuries. Two patients had an associated injury: one had an olecranon fracture and the other, a radial head fracture. No patient had had previous surgery on the involved elbow.
Thirteen patients had adult rheumatoid arthritis, and one patient had juvenile rheumatoid arthritis. The duration of the disease averaged twenty-eight years (range, twenty to fifty-five years).
Surgical Treatment
The treatment of the acute distal humeral fractures consisted of either open reduction and internal fixation (Group 1) or primary total elbow arthroplasty (Group 2). The type of operative treatment was planned before the surgery on the basis of an evaluation of the preoperative radiographs. The potentially poor-quality, osteoporotic bone with poor holding power for screws or Kirschner wires as well as the altered joint morphology in patients with rheumatoid arthritis were considered in the preoperative decision-making process. The definitive decision regarding the operation was made intraoperatively by the operating surgeon. We strove to perform open reduction and internal fixation; however, when there was concern about the ability to achieve a stable open reduction and internal fixation, a total elbow arthroplasty was done instead. Because of the long study period (twenty years), a total of eight surgeons were involved in the performance of these procedures.
Group 1 (Open Reduction and Internal Fixation)
In Group 1, six distal humeral fractures were treated with open reduction and internal fixation with use of plates, screws, and pins after an average delay of three days (range, zero to five days) after the injury (Figs. 1-A through 1-D). The median age of the patients at the time of the fracture was fifty-nine years (range, fifty-six to sixty-nine years). All elbows were approached posteriorly. Four fractures had a single intra-articular extension that could be reduced and fixed anatomically by mobilizing the triceps medially and laterally, without the need for an olecranon osteotomy. In one elbow (Case 1; see Appendix), a concomitant olecranon fracture was used to facilitate an osteotomy-like posterior approach. Following application of plates and supplemental Kirschner wires, the olecranon fracture was fixed with a single screw. A second patient (Case 4) had fracture fixation through an olecranon osteotomy. Bone graft substitutes were not used in this series. Details are depicted in a table in the Appendix. The ulnar nerve was mobilized in all patients and was transposed anteriorly in the subcutaneous tissue.
Group 2 (Primary Total Elbow Arthroplasty)
Ten acute distal humeral fractures were treated with a primary total elbow arthroplasty. After an average delay of five days (range, one to ten days), a semiconstrained Coonrad-Morrey total elbow prosthesis (Zimmer, Warsaw, Indiana) was implanted (Figs. 2-A through 2-D). Both components were cemented. The median age of the patients at the time of the fracture was sixty-six years (range, forty-seven to eighty-three years). A Bryan-Morrey approach was used in five cases; a triceps-sparing approach, in four; and a triceps-splitting approach, in one4. In the patient with the radial head fracture (Case 11; see Appendix), the radial head was excised as part of the total elbow arthroplasty procedure.
Assessment
Clinical Assessment
At the time of the follow-up examination, the patients were asked to subjectively rate their satisfaction with the result of the procedure as very satisfied, satisfied, disappointed, or not satisfied. They were also examined with use of the Mayo Elbow Performance Score (MEPS)5. The maximum MEPS is 100 points, with a maximum of 45 points for pain, 20 points for motion, 10 points for stability, and 25 points for function. Pain, as reported by the patients, was graded as none (45 points), mild (30 points), moderate (15 points), or severe (0 points). Motion was assessed as the degrees of flexion, extension, pronation, and supination, and the flexion-extension and pronation-supination arcs were defined. Furthermore, the flexion-extension arc of motion was graded, according to the MEPS, as >100° (20 points), 50° to 100° (15 points), or <50° (0 points). Elbow stability was graded as stable (10 points), moderately unstable (5 points), or grossly unstable (0 points). Finally, five daily activities were assessed, including the ability to comb one's hair (5 points), feed oneself (5 points), attend to personal hygiene (5 points), don a shirt (5 points), and don shoes (5 points). A MEPS of >90 points is classified as excellent; 75 to 89 points, as good; 60 to 74 points, as fair; and <60 points, as poor.
Radiographic Assessment
Preoperatively, the fracture was assessed radiographically with use of the AO classification system for distal humeral fractures6. The classification includes three major groups: extra-articular fractures (type A); partial articular fractures, in which a portion of the joint remains in continuity with the diaphysis (type B); and complete articular fractures (type C). Further subdivisions were not used.
The radiographic presentation of the rheumatoid arthritis in the elbow was stratified according to the Mayo classification7,8. In stage I, there are no radiographic changes other than osteoporosis, and the joint surface is normal. In stage II, there is narrowing of the joint space but the subchondral architecture of the joint is maintained. In stage III, there has been alteration in the architecture of the joint with partial loss of the joint space (IIIA without deformity and IIIB with deformity). A complete loss of the joint space with gross joint destruction is considered to be stage IV. Because stages I and II and stages III and IV can be difficult to distinguish from one another, especially in the presence of a fracture, the radiographic classification was simplified, with the presentation graded as mild (stage I or II) or severe (stage III or IV).
Postoperative radiographs were assessed for fracture-healing after the open reduction and internal fixation procedures and for loosening of the prosthetic components after the total elbow arthroplasties. The latest available radiographs were compared with the immediate postoperative radiographs to distinguish signs of loosening from those of irregular cementing. Radiographic signs of loosening (radiolucencies) were looked for around the ulnar and humeral components. The length of a radiolucent line along the component was measured in 5-mm increments (0 to 5 mm, 5 to 10 mm, and so on). Radiographic signs of partial loosening were differentiated from those of complete loosening. The ulnar or humeral component was graded as loose when the radiolucent line was present around the entire component.
Comparison of Groups
Group 1 and Group 2 were compared to determine possible differences in age, fracture type, extent of rheumatoid involvement of the elbow joint, and the clinical results (MEPS, pain, and range of motion) at the time of final follow-up.
Statistical Analysis
The Mann-Whitney U test for unpaired groups was used to compare numeric data between Group 1 and Group 2, and the Fisher exact test was used to assess differences in frequencies within groups. When there was less than a 5% likelihood that a difference had occurred by chance (p < 0.05), that difference was considered to be significant.
Demographic Data
Age
With the numbers studied, we could not identify a difference between groups with respect to age (median, fifty-nine years in Group 1 and sixty-six years in Group 2) (Table I). The age range in Group 1 (fifty-six to sixty-nine years) was narrower than that in Group 2 (forty-seven to eighty-three years). Three of the ten patients were older than seventy years of age in Group 2.
Fracture Type
The initial fracture type, according to the AO classification, was A in one elbow, B in nine elbows, and C in six elbows. In Group 1, two elbows had a type-B fracture and four had a type-C fracture. In Group 2, one patient had a type-A fracture, seven patients had a type-B fracture, and two patients had a type-C fracture. There was no predominant fracture type in Group 1 or Group 2.
Rheumatoid Joint Involvement
At the time of presentation, rheumatoid involvement of the elbow joint was classified as stage II in four elbows, stage III in six (two IIIA and four IIIB), and stage IV in five. The initial radiograph was insufficient to allow staging of one elbow. Overall, four elbows were graded as having mild rheumatoid involvement and eleven, as having severe involvement. There was a tendency for the preoperative arthritic involvement to be less severe in Group 1 than in Group 2. There were no stage-II elbows in Group 2, and half of the patients in that group had stage-IV disease. When the rheumatoid joint involvement was categorized simply as mild or severe, four of the six elbows in Group 1 had mild involvement, but all nine of the elbows in Group 2 that could be staged had severe involvement. This difference was significant (p = 0.0110).
Clinical Results
Group 1
The duration of follow-up for the six elbows averaged forty-nine months (range, twenty-four to seventy-two months). Four patients with five involved elbows were very satisfied with the result, and one patient was disappointed. The average MEPS was 93 points. The patients reported no pain in five elbows and mild pain in one elbow. Flexion averaged 130°; loss of extension, 23°; pronation, 54°; and supination, 67°. Four elbows had an arc of elbow motion of =100°, and two elbows had an arc of between 50° and 100°. All six elbows were stable. Five of the six elbows could be used to perform all five requested daily activities. One patient (Case 4) could perform only two activities (he could feed himself and attend to personal hygiene). Overall, five of the six elbows had objectively and subjectively satisfactory outcomes.
Group 2
The mean duration of follow-up of the ten elbows was sixty-six months (range, twenty-four to 132 months). Nine patients were very satisfied and one patient was satisfied with the result. The MEPS averaged 96 points, with six elbows having the maximum of 100 points. No patient experienced pain. Flexion averaged 133°; loss of extension, 26°; pronation, 54°; and supination, 57°. Eight patients had an arc of elbow flexion-extension of =100°, and two had an arc of <100° (90° and 80°, respectively). All elbows were stable. Seven of the ten elbows could be used to perform all of the five daily activities, one could be used to perform four, one could be used to perform three, and one could be used to perform only two.
Comparison Between Groups
With the numbers studied, we could not identify a difference regarding the duration of follow-up, MEPS, pain, or range of motion between Groups 1 and 2 (Table I).
Radiographic Results
Group 1
After open reduction and internal fixation, all six fractures healed uneventfully. There were no nonunions, and no clear progression of arthritic joint involvement was observed in any of the patients at the time of the latest follow-up.
Group 2
Postoperatively, two total elbow arthroplasties in two patients were revised; in one patient the humeral component was changed as part of the treatment for infection, and the second patient had a revision of a loose ulnar component. Details of these complications are provided below. At the time of final follow-up, at an average of 5.5 years, none of the current prostheses were considered to be loose as seen radiographically. Minor signs of loosening or partial loosening (radiolucencies) were seen in three elbows.
Complications and Reoperations
Group 1
There was one major complication and one minor complication, and three reoperations, in two patients. In one patient treated with plates, screws, and Kirschner wires, elbow stiffness developed with a painful Kirschner wire in the lateral epicondyle. The wire was removed ten months after the open reduction and internal fixation. Then, after twelve months, complete hardware removal was combined with an arthrolysis to treat persistent elbow stiffness. At the sixty-month follow-up examination, the MEPS was 100 points; flexion was 120°, extension was 15°, pronation was 30°, and supination was 85°. The second patient had a painful fixation screw at the site of an olecranon osteotomy; the screw was removed six months postoperatively. At the time of final follow-up at twenty-four months, this patient had a MEPS of 70 points with mild pain. None of the patients treated with open reduction and internal fixation underwent a subsequent replacement arthroplasty.
Group 2
Two patients had two major complications and a total of three reoperations. In one patient, symptoms suggestive of an acute infection developed three weeks after implantation. The humeral component was removed, and the elbow was débrided. No organisms were isolated. Another humeral component was implanted two weeks later. The following course was uneventful, with a MEPS of 80 points at the twenty-four-month follow-up examination and until the patient died sixty months after the revision. The second patient had loosening of the ulnar component, which was revised thirty-three months after implantation. After the revision, the clinical course was uneventful, with a MEPS of 100 points at the time of the latest follow-up (sixty months after the fracture; twenty-four months after the revision).
Total elbow arthroplasty yields good results in patients with various disorders of the elbow joint1,9-11. Semiconstrained total elbow arthroplasty also performs well in elderly patients with a severely comminuted distal humeral fracture1-3,12-16 and in patients with rheumatoid arthritis7,17-19. To date, we are aware of only one article in the English-language literature that exclusively addresses the treatment of fractures around the elbow (the distal part of the humerus and the olecranon) in patients with rheumatoid arthritis by total elbow arthroplasty partially combined with open reduction and internal fixation3. We are not aware of any study comparing the outcomes of open reduction and internal fixation with those of total elbow arthroplasty for the treatment of acute distal humeral fractures in patients with rheumatoid arthritis. The goal of this study was to better understand the best treatment strategy on the basis of demographic factors, the extent of the injury, and the stage of rheumatoid joint involvement.
In this study of two different fracture-treatment modalities, primary total elbow arthroplasty as well as open reduction and internal fixation, there was an excellent or good result in fifteen of sixteen cases. There are three considerations that may shape one's judgment regarding the best type of intervention in this setting: patient age, fracture type, and extent of rheumatoid involvement of the elbow joint.
Age
An attempt at open reduction and internal fixation may be considered preferable to a primary total elbow arthroplasty in relatively young patients. In the literature, surgical treatment of acute distal humeral fractures with a primary total elbow arthroplasty1,2,12-14 seems to have had a better outcome than open reduction and internal fixation in elderly patients without rheumatoid arthritis1,20,21. The median age of our patients (sixty-two years) was clearly younger than that described in the literature for patients without rheumatoid arthritis2,12-14. Following this current experience, it is our preference to perform open reduction and internal fixation in younger patients. However, this study does not provide information about patients less than forty-seven years old.
Fracture Type
In our review, the fracture type did not appear to be a major factor influencing the choice of fixation or joint replacement. There was no predominant fracture type in Group 1 or Group 2. Thus, at least in this series, the fracture type appeared to be less important than the degree of involvement by the rheumatoid process.
Rheumatoid Involvement
All of the patients in this study had a long history (more than twenty years) of rheumatoid arthritis. On radiographic evaluation, all of the elbows were graded as at least stage II. Four of the six elbows treated with open reduction and internal fixation had mild rheumatoid joint involvement (stage II), but all of the elbows treated with total elbow arthroplasty had severe (stage-III or IV) rheumatoid involvement. Thus, this review suggests that patients with radiographic stage-II arthritis of the elbow can be treated successfully with open reduction and internal fixation. Our data also suggest that primary total elbow arthroplasty is the preferred operative treatment for stage-III or IV elbow joint involvement. Total elbow arthroplasty may be favored over open reduction and internal fixation in the presence of severe arthritis because of difficulty in achieving rigid osteosynthesis acutely and increasing stiffness postoperatively secondary to distorted joint anatomy. Therefore, the use of total elbow arthroplasty for acute distal humeral fractures associated with severe rheumatoid joint involvement, even in a younger patient, seems to be justified2,12-14.
The rate and types of complications were acceptable in both groups. Two of the six elbows treated with open reduction and internal fixation had a complication. Interestingly, the olecranon was involved in both cases (one concomitant fracture and one osteotomy). In the other four elbows, the fracture could be reduced and fixed anatomically without the need for an olecranon osteotomy. We can conclude that olecranon osteotomy should be avoided when possible in these patients.
This study had several limitations. First, it is a retrospective review of a heterogeneous patient population. Second, it lacks statistical power. The number of patients in both groups is too small to allow meaningful statistical comparison. Third, for the duration of the study, the reported follow-up is rather short. This may be due to the fact that six of the fourteen patients had died when the study was initiated. Fourth, there may have been a surgeon selection bias in favor of total elbow arthroplasty. Fifth, there were eight surgeons, which may have led to inconsistencies regarding treatment. Finally, comorbidities influencing the decision regarding either the treatment or the results were incompletely documented and therefore we could make no meaningful statement regarding the influence of comorbidities on our therapeutic decisions. Nevertheless, this study provides some previously unknown information that should assist in the management of these injuries.