Radial Head Replacement in Capitellocondylar Total Elbow Arthroplasty 2- to 6-year Follow-up Evaluation in Rheumatoid Arthritis
Elly Trepman,
M D , I. M i c h a e l Vella, M D , a n d F r e d e r i c k C. E w a l d , M D
Abstract: We reviewed six capitellocondylar metal-to-plastic total dbow replacement prostheses with radial head components, implanted in patients with rheumatoid arthritis. At an average of 4.7 - 1.5 years, relief of pain, improvement of function, and a functional range of motion were preserved. Five of tile six elbows (83%) were clinically rated good or excellent. Radiolucent lines Were seen at tile bone-cement interface in 50% of the humeral component stems and in all ulnar component boats within 2 years after surgery; none had progressed at subseqnent examination. Most were --< I mm wide, and none were associated with clinical deterioration. Only one of tile humeral components was radiographically loose. No radiolucent lines ware seen along the stems of the ulnar or radial components. There were no postoperative dislocations when tile radial component was used, presumably because the prosthetic radial head provided increased constraint. None of these elbows Itave required revision. Radial head replacement in capite]locondylar arthroplasty had been discontinued because radiolucent lines were observed at early review. However, the absence of clinical failure, dislocation, or progression of radiolucency at long-term follow-up examination favor radial head replacement in primary unconstrainer total elbow arthroplasty. Key words: radial head, elbow, capitellocondylar, unconstrained, arthroplasty, rheumatoid.
The major complication of-uniaxial hinge total elbow replacement has been loosening of the components, leading to pain and reoperation (3, 8, 13, 14,'i 6, 21, 28). The large forces generated across the elbow joint during elbow motion (1, 2), and tile
changes in carrying angle and forearm axial rotation during elbow flexion and extension (20), are belie~,ed to cause high torques at the b o n e - c e m e n t interface of uniaxial hinge prostheses, and thereby to cause loosening (12, 13). The unconstrained humeroulnar total elbow protheses, including the capitellocondylar model, allow tile dissipation of these torques by the soft tissues, resulting in a low incidence of clinical loosening (7, 11, 12, 23, 24, 31). However, these models are inherently less stable
From the Department of Orlhopaedic Surgery, ttarvard Medical School' Brigham and Women's ltospital' Boston, ~fassachusetts.
Reprint requests: Frederick C. E~x*,ald,MD, 75 Francis Street, Boston, MA O2115.
67
68 The Journal of Arthroplasty Vol. 6 No. 1 March 1991
than the hinged prostheses and have been complicated by a significant incidence of dislocation (7, 11, 12, 17, 24, 27, 31). A theoretical advantage of radial head replacement in unconstrained total elbow arthroplasty is the potential improvement of prosthetic joint stability (9). Furthermore, radial head replacement has been recommended in order to reduce the medial collateral ligament tension and humeroulnar joint force, which increase after radial head excision, and which may contribute to component loosening (1, 32). Radial head components have been designed for several total elbow prostheses (3, 4, 23), but the frequency of use or long-term clinical effect of the radial component has not been reported. The early capitellocondylar humeroulnar prosthesis also included an optional radial head com-
ponent. This component consisted of a titanium alloy tray and intramedullary stem with a high molecular weight polyethylene surface which articulated with the metal surface of the humeral component (Fig. 1). Preliminary clinical results using this prosthesis with radial component in eight patients included a high incidence of asymptomatic, early radiolucent lines at the bone-cement interface ( 11 ), which were not observed without the radial component (12). Therefore, even though no clinical loosening had occurred, the use of the radial component was discontinued. We analyzed this series of patients after a longer postoperative period to determine the clinical significance of these radiolucent lines, the effect of radiocapitellar replacement in total elbow arthroplasty, and the appropriateness of further clinical use of the radial component.
B Fig. 1. (A) Radiocapitellocondylar total elbow prosthesis, ttumeral component articulating surface is in valgus alignment relative to the fixation stem; ulnar component consists of metal boat and fixation stem with polyethylene articulating liner; radial component consists of metal tray and stem with polyethylene head. (B) Prosthesis (early prototype) implanted in a bone model.
Radial Headin Total ElbowArthroplasty 9 Trepmanet al. 69 Materials and Methods Eight radiocapitellocondylar total elbow prostheses were implanted in eight patients with rheumatoid arthritis in our institution. Six patients (six elbows) were available for follow-up evaluation. The other two patients were lost to follow-up evaluation soon postoperatively and were excluded from the analysis that follows. The average time of clinical follow-up observation of the six patients was 4.7 _ 1.5 years (range 2 6 74 mon), and radiographic follow-up evaluation was performed at a mean 4.9 _ 1.3 years (range 3 3 - 7 3 mon). The average at operation was 57 _ 10 years (range 4 2 - 7 0 y). The study population consisted of four w o m e n and two men. The functional severity of rheumatoid disease (29) was class 2 in four patients w h o were diagnosed within 3 - 1 3 years prior to surgery, and class 3 in two w o m e n who had had rheumatoid arthritis for 40 years. The anatomic severity of elbow involvement (5, 29) was stage 3 in all five elbows for which preoperative radiographs were available. In all patients, the indication for operation was intractable pain and limitation of function. A scoring system was used to rate each elbow for pain, function, motion, and deformity, as previously described (10, "12). All six patients were rated poor preoperatively, with an average elbow score of 26 _ 10 (Table 1). A posterolateral surgical approach was used in all cases, as previously described (11). Ulnar nerve transposition was not performed in any of these patients. A 5 ~ (valgus alignment) humeral c o m p o n e n t was used in all six cases; a regular ulnar c o m p o n e n t
Table 1. Radiocapitellocondylar Total Elbow Replacement: Clinical Results* Preoperative Range of motion, degrees Flexion Extension Pronation Supination PFC#
130 -28 60 40 28
Elbow score Pain (max 50) Function (max 30) Motion (max 10) PFC# (max 5) Carrying angle (max 5) Total (max tO0)
8 7 6 l 5 26
• l0 -4- 6 • 30 ~ 20 • 6 -. •
8 4 3 1 1 10
was used in five elbows, and a short-stemmed ulnar component in one (11). Radial osteotomy was performed proximal to the annular ligament. The radial component consisted of a standard titanium alloy stem (28 m m long) and tray, with a high molecular weight polyethylene head; the radial head thickness used iavailable range 8 - 1 5 mm) was determined intraoperatively: 10 m m in four cases, 12 m m in one, and 13 m m in one. Concurrent fixation of all three components with methylmethacrylate cement was performed in five elbows (order of insertion usually either humeral or ulnar c o m p o n e n t first, and radial component last); in one case, the humeral and ulnar components were implanted initially, followed by the radial c o m p o n e n t after the humeral and ulnar cement had cured. The elbow was held reduced in full extension while the cement hardened, except in one patient (patient 1), in w h o m the elbow was held at 90 ~ of flexion because of bone deficiency (see Results). Average tourniquet time from the beginning of the procedure to completion of c o m p o n e n t fixation, w h e n the tourniquet was deflated, was 103 _+ 20 minutes. After hemostasis was secured, the w o u n d was closed in layers. Gentle active assisted range of motion was started postoperatively within several days, as permitted by the level of comfort. Postoperative radiographs were analyzed to evaluate component position, cement technique, and radiolucent lines. A scheme of radiographic zones was devised, analogous to those used for hip (30) and
H3 H2 H4
h2_ f 3 _
HI
Follow-up Findings'[133 -20 60 40 20 50 24 8 1.7 5 89
• • • • • • • • •
8 tO 20 30 1"0 0 8 3 0.8 l 7
* Average • standard deviation. "t"Average clinical follow-up evaluation at 57 • 20 months. PFC# = Permanent flexion contracture.
u'
'U' I Lateral
Anteroposterior
Fig. 2. Schemeof zonesdevisedfor analysis of radiographic results of radiocapitellocondylar total elbow replacement.
70
The Journal of Arthroplasty Vol. 6 No. 1 March 1991
knee (26, 34) arthroplasty, to standardize and simplify the analysis and reporting of data (Fig. 2). The presence and thickness of any radiolucent lines, either at the b o n e - c e m e n t interface or the c e m e n t prosthesis interface, was noted for each zone and recorded on a diagram similar to Figure 2. The valgus-varus humeral shaft-stem alignment was determined from the anteroposterior radiograph, and the flexion-extension ulnar stem-shaft alignment was measured from the lateral radiograph; the radial stem-shaft alignment could not be determined in a standardized fashion because of variation in forearm rotation with different radiographs. The carrying angle was estimated, w h e n possible, from the anteroposterior radiograph, and any miscellaneous findings were noted.
Results Clinical Results The preoperative range of motion was not significantly changed at follow-up examination and was within the limits of function for most activities of daily living for all patients (19). The complete reduction of pain achieved by surgery was preserved, with all patients having a perfect pain score of 50. Functional use of the elbow was also dramatically improved, evidenced by the marked increase in function score. The average total elbow score was 89 __
7. Of the six elbows, three were rated excellent, tivo good, and one fair; the patient whose elbow was rated as fair had a function score of 10 (total score 77) because activities of daily living were restricted by poor hand, not elbow, function. These findings are summarized in Table 1.
Radiographic Evaluation Three of the six elbows had b o n e - c e m e n t radiolucent lines along the humeral c o m p o n e n t stem (Table 2). In one patient (patient 1), this radiolucent line was greater than 1 m m wide (Fig. 3). This patient had had severe rheumatoid destruction of the troehlea and olecranon notch preoperatively and therefore was treated with medial bone grafting and a 24-rrim cancellous screw left 5 m m prominent to buttress the humeral c o m p o n e n t at the trochlea (Fig. 3). The humeral radiolucent line and screw failure were evident 15 months postoperatively and remained unchanged at 33 months with the exception of further displacement of the screw fracture (Fig. D and E). The radiolucent line at the ulnar boat was present immediately postoperatively (Fig. 3C). Although radiographically loose, this elbow was free of pain (pain score 50), had excellent function (function score 30), and was rated clinically excellent (total score 92). The other two humeral stem b o n e - c e m e n t radi-
Table 2. Radiocapitellocondylar Tolal Elbow Replacement: Bone-Cement Radiolucent Lines* Patient
Time postop, mo Lateral radiograph zones ti1 It2 H3 H4 U1 U2 U3
1
2
3
4
5
6
33
45
65
73
69
67
Elly Trepman,
M D , I. M i c h a e l Vella, M D , a n d F r e d e r i c k C. E w a l d , M D
Abstract: We reviewed six capitellocondylar metal-to-plastic total dbow replacement prostheses with radial head components, implanted in patients with rheumatoid arthritis. At an average of 4.7 - 1.5 years, relief of pain, improvement of function, and a functional range of motion were preserved. Five of tile six elbows (83%) were clinically rated good or excellent. Radiolucent lines Were seen at tile bone-cement interface in 50% of the humeral component stems and in all ulnar component boats within 2 years after surgery; none had progressed at subseqnent examination. Most were --< I mm wide, and none were associated with clinical deterioration. Only one of tile humeral components was radiographically loose. No radiolucent lines ware seen along the stems of the ulnar or radial components. There were no postoperative dislocations when tile radial component was used, presumably because the prosthetic radial head provided increased constraint. None of these elbows Itave required revision. Radial head replacement in capite]locondylar arthroplasty had been discontinued because radiolucent lines were observed at early review. However, the absence of clinical failure, dislocation, or progression of radiolucency at long-term follow-up examination favor radial head replacement in primary unconstrainer total elbow arthroplasty. Key words: radial head, elbow, capitellocondylar, unconstrained, arthroplasty, rheumatoid.
The major complication of-uniaxial hinge total elbow replacement has been loosening of the components, leading to pain and reoperation (3, 8, 13, 14,'i 6, 21, 28). The large forces generated across the elbow joint during elbow motion (1, 2), and tile
changes in carrying angle and forearm axial rotation during elbow flexion and extension (20), are belie~,ed to cause high torques at the b o n e - c e m e n t interface of uniaxial hinge prostheses, and thereby to cause loosening (12, 13). The unconstrained humeroulnar total elbow protheses, including the capitellocondylar model, allow tile dissipation of these torques by the soft tissues, resulting in a low incidence of clinical loosening (7, 11, 12, 23, 24, 31). However, these models are inherently less stable
From the Department of Orlhopaedic Surgery, ttarvard Medical School' Brigham and Women's ltospital' Boston, ~fassachusetts.
Reprint requests: Frederick C. E~x*,ald,MD, 75 Francis Street, Boston, MA O2115.
67
68 The Journal of Arthroplasty Vol. 6 No. 1 March 1991
than the hinged prostheses and have been complicated by a significant incidence of dislocation (7, 11, 12, 17, 24, 27, 31). A theoretical advantage of radial head replacement in unconstrained total elbow arthroplasty is the potential improvement of prosthetic joint stability (9). Furthermore, radial head replacement has been recommended in order to reduce the medial collateral ligament tension and humeroulnar joint force, which increase after radial head excision, and which may contribute to component loosening (1, 32). Radial head components have been designed for several total elbow prostheses (3, 4, 23), but the frequency of use or long-term clinical effect of the radial component has not been reported. The early capitellocondylar humeroulnar prosthesis also included an optional radial head com-
ponent. This component consisted of a titanium alloy tray and intramedullary stem with a high molecular weight polyethylene surface which articulated with the metal surface of the humeral component (Fig. 1). Preliminary clinical results using this prosthesis with radial component in eight patients included a high incidence of asymptomatic, early radiolucent lines at the bone-cement interface ( 11 ), which were not observed without the radial component (12). Therefore, even though no clinical loosening had occurred, the use of the radial component was discontinued. We analyzed this series of patients after a longer postoperative period to determine the clinical significance of these radiolucent lines, the effect of radiocapitellar replacement in total elbow arthroplasty, and the appropriateness of further clinical use of the radial component.
B Fig. 1. (A) Radiocapitellocondylar total elbow prosthesis, ttumeral component articulating surface is in valgus alignment relative to the fixation stem; ulnar component consists of metal boat and fixation stem with polyethylene articulating liner; radial component consists of metal tray and stem with polyethylene head. (B) Prosthesis (early prototype) implanted in a bone model.
Radial Headin Total ElbowArthroplasty 9 Trepmanet al. 69 Materials and Methods Eight radiocapitellocondylar total elbow prostheses were implanted in eight patients with rheumatoid arthritis in our institution. Six patients (six elbows) were available for follow-up evaluation. The other two patients were lost to follow-up evaluation soon postoperatively and were excluded from the analysis that follows. The average time of clinical follow-up observation of the six patients was 4.7 _ 1.5 years (range 2 6 74 mon), and radiographic follow-up evaluation was performed at a mean 4.9 _ 1.3 years (range 3 3 - 7 3 mon). The average at operation was 57 _ 10 years (range 4 2 - 7 0 y). The study population consisted of four w o m e n and two men. The functional severity of rheumatoid disease (29) was class 2 in four patients w h o were diagnosed within 3 - 1 3 years prior to surgery, and class 3 in two w o m e n who had had rheumatoid arthritis for 40 years. The anatomic severity of elbow involvement (5, 29) was stage 3 in all five elbows for which preoperative radiographs were available. In all patients, the indication for operation was intractable pain and limitation of function. A scoring system was used to rate each elbow for pain, function, motion, and deformity, as previously described (10, "12). All six patients were rated poor preoperatively, with an average elbow score of 26 _ 10 (Table 1). A posterolateral surgical approach was used in all cases, as previously described (11). Ulnar nerve transposition was not performed in any of these patients. A 5 ~ (valgus alignment) humeral c o m p o n e n t was used in all six cases; a regular ulnar c o m p o n e n t
Table 1. Radiocapitellocondylar Total Elbow Replacement: Clinical Results* Preoperative Range of motion, degrees Flexion Extension Pronation Supination PFC#
130 -28 60 40 28
Elbow score Pain (max 50) Function (max 30) Motion (max 10) PFC# (max 5) Carrying angle (max 5) Total (max tO0)
8 7 6 l 5 26
• l0 -4- 6 • 30 ~ 20 • 6 -. •
8 4 3 1 1 10
was used in five elbows, and a short-stemmed ulnar component in one (11). Radial osteotomy was performed proximal to the annular ligament. The radial component consisted of a standard titanium alloy stem (28 m m long) and tray, with a high molecular weight polyethylene head; the radial head thickness used iavailable range 8 - 1 5 mm) was determined intraoperatively: 10 m m in four cases, 12 m m in one, and 13 m m in one. Concurrent fixation of all three components with methylmethacrylate cement was performed in five elbows (order of insertion usually either humeral or ulnar c o m p o n e n t first, and radial component last); in one case, the humeral and ulnar components were implanted initially, followed by the radial c o m p o n e n t after the humeral and ulnar cement had cured. The elbow was held reduced in full extension while the cement hardened, except in one patient (patient 1), in w h o m the elbow was held at 90 ~ of flexion because of bone deficiency (see Results). Average tourniquet time from the beginning of the procedure to completion of c o m p o n e n t fixation, w h e n the tourniquet was deflated, was 103 _+ 20 minutes. After hemostasis was secured, the w o u n d was closed in layers. Gentle active assisted range of motion was started postoperatively within several days, as permitted by the level of comfort. Postoperative radiographs were analyzed to evaluate component position, cement technique, and radiolucent lines. A scheme of radiographic zones was devised, analogous to those used for hip (30) and
H3 H2 H4
h2_ f 3 _
HI
Follow-up Findings'[133 -20 60 40 20 50 24 8 1.7 5 89
• • • • • • • • •
8 tO 20 30 1"0 0 8 3 0.8 l 7
* Average • standard deviation. "t"Average clinical follow-up evaluation at 57 • 20 months. PFC# = Permanent flexion contracture.
u'
'U' I Lateral
Anteroposterior
Fig. 2. Schemeof zonesdevisedfor analysis of radiographic results of radiocapitellocondylar total elbow replacement.
70
The Journal of Arthroplasty Vol. 6 No. 1 March 1991
knee (26, 34) arthroplasty, to standardize and simplify the analysis and reporting of data (Fig. 2). The presence and thickness of any radiolucent lines, either at the b o n e - c e m e n t interface or the c e m e n t prosthesis interface, was noted for each zone and recorded on a diagram similar to Figure 2. The valgus-varus humeral shaft-stem alignment was determined from the anteroposterior radiograph, and the flexion-extension ulnar stem-shaft alignment was measured from the lateral radiograph; the radial stem-shaft alignment could not be determined in a standardized fashion because of variation in forearm rotation with different radiographs. The carrying angle was estimated, w h e n possible, from the anteroposterior radiograph, and any miscellaneous findings were noted.
Results Clinical Results The preoperative range of motion was not significantly changed at follow-up examination and was within the limits of function for most activities of daily living for all patients (19). The complete reduction of pain achieved by surgery was preserved, with all patients having a perfect pain score of 50. Functional use of the elbow was also dramatically improved, evidenced by the marked increase in function score. The average total elbow score was 89 __
7. Of the six elbows, three were rated excellent, tivo good, and one fair; the patient whose elbow was rated as fair had a function score of 10 (total score 77) because activities of daily living were restricted by poor hand, not elbow, function. These findings are summarized in Table 1.
Radiographic Evaluation Three of the six elbows had b o n e - c e m e n t radiolucent lines along the humeral c o m p o n e n t stem (Table 2). In one patient (patient 1), this radiolucent line was greater than 1 m m wide (Fig. 3). This patient had had severe rheumatoid destruction of the troehlea and olecranon notch preoperatively and therefore was treated with medial bone grafting and a 24-rrim cancellous screw left 5 m m prominent to buttress the humeral c o m p o n e n t at the trochlea (Fig. 3). The humeral radiolucent line and screw failure were evident 15 months postoperatively and remained unchanged at 33 months with the exception of further displacement of the screw fracture (Fig. D and E). The radiolucent line at the ulnar boat was present immediately postoperatively (Fig. 3C). Although radiographically loose, this elbow was free of pain (pain score 50), had excellent function (function score 30), and was rated clinically excellent (total score 92). The other two humeral stem b o n e - c e m e n t radi-
Table 2. Radiocapitellocondylar Tolal Elbow Replacement: Bone-Cement Radiolucent Lines* Patient
Time postop, mo Lateral radiograph zones ti1 It2 H3 H4 U1 U2 U3
1
2
3
4
5
6
33
45
65
73
69
67
