- Email: [email protected]
Revision surgery for stiff total knee arthroplasty
Revision Surgery for Stiff Total Knee Arthroplasty
Danny
W . N i c h o l l s , D O , M a j o r , U S A F , M C FS*, a n d Lawrence
D. DOLT, M D t
Abstract: Thirteen stiff total knee arthroplasties were studied in 12 patients who did not have a stiff knee prior to the index arthroplasty. The reasons for replacement were osteoarthritis in seven knees, post-traumatic arthritis in two, and inflammatory collagen disease in four. Four knees were revised for fixed flexion deformity and nine for poor arc of motion. Tile cause of stiffness in eight knees was malpositioned components. Following postoperative revision for stiffness, three knees received excellent ratings, two received good ratings, four received fair ratings, and four received poor ratings. Satisfaction was achieved in 11 of 12 patients because of pain relieL Improved motion was achieved in patients with primary osteoarthritis and patients with malposition of components. Poor motion occurred in those with patella baja and inflammatory collagen disease. A V-Y patella turndown helped prevent patella ligament complications. Key words: stiff total knee arthroplasty, revision total knee anhroplasty, total knee arthrop]asty, knee, stiff knee.
Relief o f pain and satisfactory function, including a satisfactory range of motion for activities of daily living, are the goals of total knee arthroplasty, whether it is used as a primary or revision procedure (4). Joint stability and mobility with proper component a n d limb alignment are necessary to achieve these goals. Loss of flexion and flexion contracture are early a n d disturbing complications of knee arthroplasty. P e r m a n e n t flexion or extension contractures (stiff total knee) m a y result in pain and/or disability t h a t cause the patient to request revision. A stiff total k n e e can result from technical difficulties (5), aggressive fibroblastic scar formation, or the patient's p o o r cooperation due to postoperative reaction to h i s / h e r pain. Regardless of the cause, stiffness following t o t a l knee arthroplasty can result in a pain*From the Dallas/Fart Worth ~fedical Center, Grand Prairie, Texas. tFrom the Kerlan-Jabe Orthopaedic Clinic, hrglewood, California. Reprint requests: Lawrence D. Doff, MD, Kerlan-Jobe Orthopaedic Clinic, 501 East Hardy, Suite 300, Inglewood, CA 90301.
ful knee with significant loss of function. We have reoperated u p o n stiff total knees in an attempt to relieve pain and regain motion. This report evaluates the results of tile revision of stiff total knees.
Materials and Methods Twelve patients (13 knees) w h o u n d e r w e n t revision total knee arthroplasty for a flexion contracture greater than 20 ~ or flexion arc less t h a n 45 ~ were included in this study (Tables 1, 2). All patients had both pain and disability. There were six m e n and six w o m e n (one with bilateral total k n e e arthroplasties) whose m e a n age was 54 years (range, 2 8 - 7 3 years). The patients had had an average of two (range, 1 4) previous surgical procedures prior to revision surgery. All patients had had at least 90 ~ of flexion and
$74
The Journal of Arthroplasty Vol. 5 Supplement October 1990 Table l. Flexion Contracture Group ROM
Knee #
Age/Sex
Diagnosis
1 2 3 4
55/M 72/M 61/M 56/F
PTA OAP OAP OAP
Cause of Stiffness anterior tibial tilt anterior tibial tilt quadriceps dysfunction oblique femoral component posterior tibial tilt
Preoperative
}tSS Score
Postoperative
250-85 * 300-60 ~ 45*-90 ~ 300-60 ~
00-80 ~ 10"-100 ~ 50-95 * 15~ ~
Preoperative
Postoperative
32 38 30 23
82 61 82 67
PTA, posttraumatic arthritis; OAP, osteoarthritis primary; ROM, range of motion; HSS ltospital for Special Surgery; M, male; F, female.
at most a 10 ~ flexion contracture prior to the index arthroplasty. Four knees were revised for fixed flexion deformity and nine knees for poor arc of motion. No revisions were done for infected arthroplasties. The diagnosis necessitating the original arthroplasty was primary osteoarthritis in seven knees, posttraumatic arthritis in two knees, and inflammatory connective tissue disease in four knees. All primary arthroplasties had been semiconstrained designs. Seven arthroplasties were cemented, five were noncemented, and one was hybrid with a cemented tibial component. Revision arthroplastics were of the semiconstrained (7 knees), constrained condylar (5 knees), and rotating hinge (l knee) designs. Preoperative radiographs demonstrated five knees with anterior tilt of the tibial c o m p o n e n t range from 5~ to 14 ~ Medial tibial tilts of 9 ~ and 12 ~ were seen in two knees. Posterior translocation of the tibial component with femoral c o m p o n e n t obliquity was present in one knee. A loose tibial c o m p o n e n t without malalignment was another radiographic finding. Revision arthroplasties were followed from 2 to 7 postoperative years and all patients were evaluated using the Hospital for Special Surgery (HSS) standard knee-rating scale. The surgical treatment was individualized to address the probable source of pain and/or loss of mo-
tion. Previous incisions were used that were medial in 12 and lateral parapatellar in one (2). To obtain optimal exposure in eight knees, a Co,rise-Adams patella t u r n d o w n was used (2, 7). To allow flexion without avulsion fracture of the femoral medial epicondyle, a subperiosteal release of the superficial medial collateral ligament should be done. At the time of revision surgery, correction of the fixed deformity in these patients was achieved by removal of all capsular scar tissue from the iliotibial band to the medial collateral ligament. Excision of the popliteus tendon, posterior cruciate ligament, and posterior capsule was done as needed to release soft tissue contractures. Resection of additional distal femoral bone was done as necessary to allow complete and full extension of the knee. Soft tissue release and removal of bony blocks often resulted in inequality of flexion and extension spaces, necessitating the use of a Total Condylar III implant. All components were cemented in correct alignment at time of revision to include 30-9 ~ valgus with the tibial component in 00-8 ~ posterior tilt. Intraoperative range of motion was full for all knees at the completion of the reconstruction. If the patella did not allow full flexion, a V-Y t u r n d o w n was done. W h e n a V-Y is done, the patella tendon is reattached at that level that results in the patella being "rock solid" in the trochlear groove at 90 ~ flexion. No manual motion of the patella should be present with the knee
Table 2. P o o r A r c o f M o t i o n G r o u p ROM Knee #
Age/Sex
5 6* 7* 8 9 10 11 12 13
52/M 30/1: 28/F 29/F 72/t:: 60/M 73/M 60/1:
Diagnosis PTA RA RA CTD RA OAP OAP OAP OAP
Cause of Stiffness medial tibial tilt joint fibrosis joint fibrosis joint fibrosis loose tibial component medial tibial tilt anterior tibial tilt anterior tibial tilt anterior tibial tilt
Preoperative 10~ ~ 10~ ~ 10~ ~ 0*.-20 ~ 15~ ~ 15*-30* 15~ ~ 15~ * 10~ ~
ttSS Score
Postoperative 50-25 ~ 10~-30 ~ 10~ ~ 0~ ~ 00-110 ~ 0*-90 ~ 10"-95" 10~ * 5~ ~
Preoperative
Postoperative
57 35 35 52 57 35 40 31 38
66 35 35 52 90 98 74 65 91
PTA, posttraumatic arthritis; OAP, osteoarlhritis primary; M, male; F, female; CTD, connective tissue disease; ROM, range of motion; HSS, Hospital for Special Surgery. * K n e e # 6 and knee # 7 are the right and left knees of the same patient.
Revision Surgery for Stiff TKA 9 Nicholls and Dorr
held at 90 ~ flexion. Continuous passive motion (CPM) was begun in the recovery room and was continued for the duration of hospitalization. Active range of motion (ROM) was begun on the third postoperative day. If a V-Y t u r n d o w n was used, the CPM was limited to 60 ~ for 10 days, then active motion was started. No brace was used for these patients. For all patients, full weight bearing was immediately encouraged with promotion of a heel-to e gait. In some patients a 89 heel lift was used on the opposite shoe to effectively lengthen the opposite leg. This leglengt h discrepancy forced extension of the affected leg. Manipulation, if necessary, was done within 1 6 weeks following surgery. Criteria for manipulation was active flexion of less than 70 ~ in the hospital, loss .of motion, and/or failure to progress after discharge. A cast followed by a long leg brace was used for one patient with a poor quadriceps and preoperative flexion contracture. With time and physical therapy muscle function improved and the brace was discontinued. With V-Y turndown, an extension lag of 15 ~ is often present at discharge and usually overcome by 3 months. Final range of motion (ROM) for all patients cannot be determined for ! year. Some patients gradually improved and three patients, two with patella baja and one with intraoperative patella tendon rupture, initially achieved 70 ~ flexion, then gradually returned to 250-30 ~ flexion in spite of physical therapy.
Results Overall, of 13 knees, three were rated as excellent after revision, two as good, four as fair, and four as poor with three of these unchanged. Forty percent were therefore excellent or good according to the Table 3. Pain Scores Knee #
Preoperative
Postoperative
Gain
1 2 3 4 5 6* 7* 8 9 10 11 12 13
15 10 10 0 10 10 10 5 10 10 10 5 10
30 20 30 25 30 10 I0 30 30 30 25 30 30
15 10 20 25 20 N/C N/C 25 20 20 15 25 20
N/C, no change. * Knee # 6 an d knee # 7 are the right and left knees of the s a me patient.
S75
HSS knee score. The average prerevision knee score was 39 (range, 2 3 - 5 7 ) and all knees were considered failures. The average postoperative knee score was 69 (range, 3 5 - 9 8 ) . Because of pain relief, patient satisfaction improved in 11 of 12 patients even with poor improvements in range of motion. Three patients improved their pain score by 1 0 - 1 5 points and eight by 2 0 - 2 5 points. Poor results were caused by extensor mechanism problems or connective tissue disease that caused fibrous ankylosis (Tables 3, 4). Patella tendon rupture can be prevented by use of the V-Y patella turndown. Stiffness from connective tissue disease will ahvays be a risk.
Deformity Fixed flexion deformities in four k n e e s before revision averaged 32 ~ (range, 25~176 Fixed flexion deformity averaged 7 ~ (range, 5~ ~) after revision. In the remaining nine knees arc of motion averaged 18 ~ (range, 10~ ~) before revision. Improvement in arc of motion to 44 ~ (range 2 0 ~ ~ occurred after revision with three knees remaining unchanged (Tables 1, 2).
Disease Of two knees with posttraumatic arthritis, an increase in total arc of motion of 20 ~ in one knee and loss of 10 ~ in the second was obtained. However, both had no pain after revision. Of four knees with inflammatory connective tissue disorder, two knees (one patient) remained unchanged and continued with mild pain on walking. One knee lost 5 ~ arc of motion and another gained 90 ~ with both patients flee of pain. Of seven knees with primary osteoarthritis, six gained an average of 53 ~ (range, 5~ ~) While one knee lost 5 ~. Pain scores improved after operation in all seven patients. Four knees were asymptomatic, two had mild pain with walking, and one that had mild pain at rest and with walking was diagnosed as having spinal stenosis.
Pain Relief Of the 11 knees improving in pain levels, eight had an increased total arc of motion. Eight knees were noted to have c o m p o n e n t malalignment on preoperative radiographic evaluation and five had total painrelief, while three had pain with walking.
$76
The Journal of Arthroplasty Vol. 5 Supplement October 1990
Table 4. Poor Results Knee #
Postoperative R O M
Postoperative Pain Score
2
10~
~
4 5 6* 7* 8
150-40 `' 50-25 ~ 10~ ~ N/C 10"-30 ~ N/C 0~ ~
25 30 10 N/C 10 N/C 30
10~
30
~
20
C a u s e of Poor Result Limited a m b u l a t i o n secondary to cervical m y e l o p a t h y Elevated joint line-patella baja Intraoperative patellar t e n d o n rupture Severe connective tissue fibrosis Severe connective tissue fibrosis Severe connective tissue fibrosis (Stickler Syndrome) Elevated joint line-patella baja
ROM, range of motion; N/C, no change. * Knee # 6 a n d knee # 7 are the left and right knees of the s a m e patient.
Fixation Seven knees had cemented components prior to revision. Two were posttraumatic arthritic joints, one had inflammatory connective tissue disease, and four had primary osteoarthritis. Five knees had improvements in total arc of motion and two lost motion after revision. Five knees had uncemented implants prior to revision. Three had inflammatory connective tissue disease and two had primary osteoanhritis. In those with connective tissue disease, an improvement of 90 ~ arc of motion occurred in one, and no improvement occt~rred in arc of motion or pain in the other two knees (one patient). One osteoanhritic patient lost 5~ but improved to having only mild pain with walking, and one patient who had a loose uncemented tibial component gained 75 ~ total arc of motion with complete relief of pain.
Complications No postoperative infections were encountered and no further revisions have been performed. Four knees underwent manipulation under anesthesia 2 8 weeks following revision. In one knee a patella tendon avulsion occurred during operation. This was treated by longitudinally splitting the tibial tubercle and placing the tendon in the bony bed with fixation by use of a staple.
Discussion Stiff total knees in this study were caused by incorrectly positioned or loose components in nine knees, quadriceps dysfunction in one knee, and a
tendency towards knee-joint ankylosis in two patients (3 knees) with inflammatory connective tissue disease. All patients in this study had both pain and disability. We would recommend this surgery primarily for pain. If only stiffness is a complaint, surgery should be considered in only those patients with a good chance for improvement. Improvement in arc of motion was best achieved in those patients with primary osteoanhritis as the underlying disease and in patients with incorrect component position. Failure to significantly improve motion of total knees was found in 6 of 7 patients with inflammatory connective tissue disease and posttraumatic arthritis. Intraoperative patellar tendon avulsion occurred in one post.traumatic arthritic revision and the resultant prolonged postoperative immobilization contributed to a poor result. Those knees requiring extensive soft tissue release or having substantial bone loss resulting in large flexion-extension gaps required Total Condylar III prostheses to produce stability. A larg.e tibial insert resulted in elevation of the joint line, iatrogenic patella baja, and poor improvement in motion. Because of this experience we recommend that bone grafts or augmented femoral components should be used to allow maintenance of a lower joint line (5; 8). All uncemented knees were revised to cemented components to minimize postoperative pain and permit rapid full weight bearing. Those knees with a tendency to undergo fibrosing anklyosis after arthroplasty should be cemented to allow full weight bearing using a heel-toe gait, thus promoting range of motion. In patients who have demonstrated aggressive fibrous tissue formation, components should be placed so that mild ligamentous laxity is present. A judicious postoperative physical therapy program is invaluable to all total knee anhroplasty patients. Those with low pain thresholds, inflammatory connective tissue disease, or poor muscle function
Revision Surgery for Stiff TKA 9 Nicholls and Dorr
can benefit from supervised physical therapy (3) rather than self administered exercises. Although statistical trends cannot be derived from this data, good results in revision of a stiff total knee can be expected w h e n recognized mechanical problems are corrected (6). W h e n no connective tissue disease is present, the joint line is maintained so that patella baja is avoided, and adequate quadriceps function is present. A V-Y patella t u r n d o w n should be used w h e n needed to avoid extensor m e c h a n i s m rupture. This approach is used either w h e n the patella cannot be laterally dislocated and everted with the midline exposure or if the patella dislocates during flexion even with a lateral release. We prefer patella t u r n d o w n to tibial tubercle elevation because of o u t favorable experience and because we desire to avoid possible h e m a t o m a formation from b o n e bleeding at the level of the distal incision. The results of revision for stiff total knees is not as good as those reported for primary arthroplasty of stiff knees (1). Forty percent of stiff total knees had an excellent or good rating c o m p a r e d to 81% of index operations on stiff knees. In both studies, pa-
$77
tients with osteoarthritis had a good chance for satisfactory results.
References 1. Aglietti P, Windsor RE, Buzzi R, Insall JN. Arthroplasty for the stiff or ankylosed knee. J Arthroplasty 4:1, 1989 2. Coonse KK, Adams JD: A new operative approach to the knee joint. Surg Gynecol Obstet 77:344, 1934 3. Dorr LD: Optimizing results of total joint arthroplasty. A.A.O.S. Instructional Course Lectures. Vol. XXXIV, p. 410, 1985. 4. Dorr LD: Technical considerations in total knee arthroplasty. Clin Orthop 205:5, 1986 5~ ttungerford DS, Kenna RV: Revising failed total knees/ technical considerations in total knee arthroplasty. Techniques in Orthopaedics, 1984 6. Jacobs MS, Hungerford DS: Revision total knee arthroplasty of aseptic failure. Clin O~hop 226:78, 1988 7. Insall JN: A midline approach to the knee. p. 1584. JBJS. 53A: , 1971 8. Scott RD: Revision total knee arthroplasty. Clin Orthop 226:65, 1988
Danny
W . N i c h o l l s , D O , M a j o r , U S A F , M C FS*, a n d Lawrence
D. DOLT, M D t
Abstract: Thirteen stiff total knee arthroplasties were studied in 12 patients who did not have a stiff knee prior to the index arthroplasty. The reasons for replacement were osteoarthritis in seven knees, post-traumatic arthritis in two, and inflammatory collagen disease in four. Four knees were revised for fixed flexion deformity and nine for poor arc of motion. Tile cause of stiffness in eight knees was malpositioned components. Following postoperative revision for stiffness, three knees received excellent ratings, two received good ratings, four received fair ratings, and four received poor ratings. Satisfaction was achieved in 11 of 12 patients because of pain relieL Improved motion was achieved in patients with primary osteoarthritis and patients with malposition of components. Poor motion occurred in those with patella baja and inflammatory collagen disease. A V-Y patella turndown helped prevent patella ligament complications. Key words: stiff total knee arthroplasty, revision total knee anhroplasty, total knee arthrop]asty, knee, stiff knee.
Relief o f pain and satisfactory function, including a satisfactory range of motion for activities of daily living, are the goals of total knee arthroplasty, whether it is used as a primary or revision procedure (4). Joint stability and mobility with proper component a n d limb alignment are necessary to achieve these goals. Loss of flexion and flexion contracture are early a n d disturbing complications of knee arthroplasty. P e r m a n e n t flexion or extension contractures (stiff total knee) m a y result in pain and/or disability t h a t cause the patient to request revision. A stiff total k n e e can result from technical difficulties (5), aggressive fibroblastic scar formation, or the patient's p o o r cooperation due to postoperative reaction to h i s / h e r pain. Regardless of the cause, stiffness following t o t a l knee arthroplasty can result in a pain*From the Dallas/Fart Worth ~fedical Center, Grand Prairie, Texas. tFrom the Kerlan-Jabe Orthopaedic Clinic, hrglewood, California. Reprint requests: Lawrence D. Doff, MD, Kerlan-Jobe Orthopaedic Clinic, 501 East Hardy, Suite 300, Inglewood, CA 90301.
ful knee with significant loss of function. We have reoperated u p o n stiff total knees in an attempt to relieve pain and regain motion. This report evaluates the results of tile revision of stiff total knees.
Materials and Methods Twelve patients (13 knees) w h o u n d e r w e n t revision total knee arthroplasty for a flexion contracture greater than 20 ~ or flexion arc less t h a n 45 ~ were included in this study (Tables 1, 2). All patients had both pain and disability. There were six m e n and six w o m e n (one with bilateral total k n e e arthroplasties) whose m e a n age was 54 years (range, 2 8 - 7 3 years). The patients had had an average of two (range, 1 4) previous surgical procedures prior to revision surgery. All patients had had at least 90 ~ of flexion and
$74
The Journal of Arthroplasty Vol. 5 Supplement October 1990 Table l. Flexion Contracture Group ROM
Knee #
Age/Sex
Diagnosis
1 2 3 4
55/M 72/M 61/M 56/F
PTA OAP OAP OAP
Cause of Stiffness anterior tibial tilt anterior tibial tilt quadriceps dysfunction oblique femoral component posterior tibial tilt
Preoperative
}tSS Score
Postoperative
250-85 * 300-60 ~ 45*-90 ~ 300-60 ~
00-80 ~ 10"-100 ~ 50-95 * 15~ ~
Preoperative
Postoperative
32 38 30 23
82 61 82 67
PTA, posttraumatic arthritis; OAP, osteoarthritis primary; ROM, range of motion; HSS ltospital for Special Surgery; M, male; F, female.
at most a 10 ~ flexion contracture prior to the index arthroplasty. Four knees were revised for fixed flexion deformity and nine knees for poor arc of motion. No revisions were done for infected arthroplasties. The diagnosis necessitating the original arthroplasty was primary osteoarthritis in seven knees, posttraumatic arthritis in two knees, and inflammatory connective tissue disease in four knees. All primary arthroplasties had been semiconstrained designs. Seven arthroplasties were cemented, five were noncemented, and one was hybrid with a cemented tibial component. Revision arthroplastics were of the semiconstrained (7 knees), constrained condylar (5 knees), and rotating hinge (l knee) designs. Preoperative radiographs demonstrated five knees with anterior tilt of the tibial c o m p o n e n t range from 5~ to 14 ~ Medial tibial tilts of 9 ~ and 12 ~ were seen in two knees. Posterior translocation of the tibial component with femoral c o m p o n e n t obliquity was present in one knee. A loose tibial c o m p o n e n t without malalignment was another radiographic finding. Revision arthroplasties were followed from 2 to 7 postoperative years and all patients were evaluated using the Hospital for Special Surgery (HSS) standard knee-rating scale. The surgical treatment was individualized to address the probable source of pain and/or loss of mo-
tion. Previous incisions were used that were medial in 12 and lateral parapatellar in one (2). To obtain optimal exposure in eight knees, a Co,rise-Adams patella t u r n d o w n was used (2, 7). To allow flexion without avulsion fracture of the femoral medial epicondyle, a subperiosteal release of the superficial medial collateral ligament should be done. At the time of revision surgery, correction of the fixed deformity in these patients was achieved by removal of all capsular scar tissue from the iliotibial band to the medial collateral ligament. Excision of the popliteus tendon, posterior cruciate ligament, and posterior capsule was done as needed to release soft tissue contractures. Resection of additional distal femoral bone was done as necessary to allow complete and full extension of the knee. Soft tissue release and removal of bony blocks often resulted in inequality of flexion and extension spaces, necessitating the use of a Total Condylar III implant. All components were cemented in correct alignment at time of revision to include 30-9 ~ valgus with the tibial component in 00-8 ~ posterior tilt. Intraoperative range of motion was full for all knees at the completion of the reconstruction. If the patella did not allow full flexion, a V-Y t u r n d o w n was done. W h e n a V-Y is done, the patella tendon is reattached at that level that results in the patella being "rock solid" in the trochlear groove at 90 ~ flexion. No manual motion of the patella should be present with the knee
Table 2. P o o r A r c o f M o t i o n G r o u p ROM Knee #
Age/Sex
5 6* 7* 8 9 10 11 12 13
52/M 30/1: 28/F 29/F 72/t:: 60/M 73/M 60/1:
Diagnosis PTA RA RA CTD RA OAP OAP OAP OAP
Cause of Stiffness medial tibial tilt joint fibrosis joint fibrosis joint fibrosis loose tibial component medial tibial tilt anterior tibial tilt anterior tibial tilt anterior tibial tilt
Preoperative 10~ ~ 10~ ~ 10~ ~ 0*.-20 ~ 15~ ~ 15*-30* 15~ ~ 15~ * 10~ ~
ttSS Score
Postoperative 50-25 ~ 10~-30 ~ 10~ ~ 0~ ~ 00-110 ~ 0*-90 ~ 10"-95" 10~ * 5~ ~
Preoperative
Postoperative
57 35 35 52 57 35 40 31 38
66 35 35 52 90 98 74 65 91
PTA, posttraumatic arthritis; OAP, osteoarlhritis primary; M, male; F, female; CTD, connective tissue disease; ROM, range of motion; HSS, Hospital for Special Surgery. * K n e e # 6 and knee # 7 are the right and left knees of the same patient.
Revision Surgery for Stiff TKA 9 Nicholls and Dorr
held at 90 ~ flexion. Continuous passive motion (CPM) was begun in the recovery room and was continued for the duration of hospitalization. Active range of motion (ROM) was begun on the third postoperative day. If a V-Y t u r n d o w n was used, the CPM was limited to 60 ~ for 10 days, then active motion was started. No brace was used for these patients. For all patients, full weight bearing was immediately encouraged with promotion of a heel-to e gait. In some patients a 89 heel lift was used on the opposite shoe to effectively lengthen the opposite leg. This leglengt h discrepancy forced extension of the affected leg. Manipulation, if necessary, was done within 1 6 weeks following surgery. Criteria for manipulation was active flexion of less than 70 ~ in the hospital, loss .of motion, and/or failure to progress after discharge. A cast followed by a long leg brace was used for one patient with a poor quadriceps and preoperative flexion contracture. With time and physical therapy muscle function improved and the brace was discontinued. With V-Y turndown, an extension lag of 15 ~ is often present at discharge and usually overcome by 3 months. Final range of motion (ROM) for all patients cannot be determined for ! year. Some patients gradually improved and three patients, two with patella baja and one with intraoperative patella tendon rupture, initially achieved 70 ~ flexion, then gradually returned to 250-30 ~ flexion in spite of physical therapy.
Results Overall, of 13 knees, three were rated as excellent after revision, two as good, four as fair, and four as poor with three of these unchanged. Forty percent were therefore excellent or good according to the Table 3. Pain Scores Knee #
Preoperative
Postoperative
Gain
1 2 3 4 5 6* 7* 8 9 10 11 12 13
15 10 10 0 10 10 10 5 10 10 10 5 10
30 20 30 25 30 10 I0 30 30 30 25 30 30
15 10 20 25 20 N/C N/C 25 20 20 15 25 20
N/C, no change. * Knee # 6 an d knee # 7 are the right and left knees of the s a me patient.
S75
HSS knee score. The average prerevision knee score was 39 (range, 2 3 - 5 7 ) and all knees were considered failures. The average postoperative knee score was 69 (range, 3 5 - 9 8 ) . Because of pain relief, patient satisfaction improved in 11 of 12 patients even with poor improvements in range of motion. Three patients improved their pain score by 1 0 - 1 5 points and eight by 2 0 - 2 5 points. Poor results were caused by extensor mechanism problems or connective tissue disease that caused fibrous ankylosis (Tables 3, 4). Patella tendon rupture can be prevented by use of the V-Y patella turndown. Stiffness from connective tissue disease will ahvays be a risk.
Deformity Fixed flexion deformities in four k n e e s before revision averaged 32 ~ (range, 25~176 Fixed flexion deformity averaged 7 ~ (range, 5~ ~) after revision. In the remaining nine knees arc of motion averaged 18 ~ (range, 10~ ~) before revision. Improvement in arc of motion to 44 ~ (range 2 0 ~ ~ occurred after revision with three knees remaining unchanged (Tables 1, 2).
Disease Of two knees with posttraumatic arthritis, an increase in total arc of motion of 20 ~ in one knee and loss of 10 ~ in the second was obtained. However, both had no pain after revision. Of four knees with inflammatory connective tissue disorder, two knees (one patient) remained unchanged and continued with mild pain on walking. One knee lost 5 ~ arc of motion and another gained 90 ~ with both patients flee of pain. Of seven knees with primary osteoarthritis, six gained an average of 53 ~ (range, 5~ ~) While one knee lost 5 ~. Pain scores improved after operation in all seven patients. Four knees were asymptomatic, two had mild pain with walking, and one that had mild pain at rest and with walking was diagnosed as having spinal stenosis.
Pain Relief Of the 11 knees improving in pain levels, eight had an increased total arc of motion. Eight knees were noted to have c o m p o n e n t malalignment on preoperative radiographic evaluation and five had total painrelief, while three had pain with walking.
$76
The Journal of Arthroplasty Vol. 5 Supplement October 1990
Table 4. Poor Results Knee #
Postoperative R O M
Postoperative Pain Score
2
10~
~
4 5 6* 7* 8
150-40 `' 50-25 ~ 10~ ~ N/C 10"-30 ~ N/C 0~ ~
25 30 10 N/C 10 N/C 30
10~
30
~
20
C a u s e of Poor Result Limited a m b u l a t i o n secondary to cervical m y e l o p a t h y Elevated joint line-patella baja Intraoperative patellar t e n d o n rupture Severe connective tissue fibrosis Severe connective tissue fibrosis Severe connective tissue fibrosis (Stickler Syndrome) Elevated joint line-patella baja
ROM, range of motion; N/C, no change. * Knee # 6 a n d knee # 7 are the left and right knees of the s a m e patient.
Fixation Seven knees had cemented components prior to revision. Two were posttraumatic arthritic joints, one had inflammatory connective tissue disease, and four had primary osteoarthritis. Five knees had improvements in total arc of motion and two lost motion after revision. Five knees had uncemented implants prior to revision. Three had inflammatory connective tissue disease and two had primary osteoanhritis. In those with connective tissue disease, an improvement of 90 ~ arc of motion occurred in one, and no improvement occt~rred in arc of motion or pain in the other two knees (one patient). One osteoanhritic patient lost 5~ but improved to having only mild pain with walking, and one patient who had a loose uncemented tibial component gained 75 ~ total arc of motion with complete relief of pain.
Complications No postoperative infections were encountered and no further revisions have been performed. Four knees underwent manipulation under anesthesia 2 8 weeks following revision. In one knee a patella tendon avulsion occurred during operation. This was treated by longitudinally splitting the tibial tubercle and placing the tendon in the bony bed with fixation by use of a staple.
Discussion Stiff total knees in this study were caused by incorrectly positioned or loose components in nine knees, quadriceps dysfunction in one knee, and a
tendency towards knee-joint ankylosis in two patients (3 knees) with inflammatory connective tissue disease. All patients in this study had both pain and disability. We would recommend this surgery primarily for pain. If only stiffness is a complaint, surgery should be considered in only those patients with a good chance for improvement. Improvement in arc of motion was best achieved in those patients with primary osteoanhritis as the underlying disease and in patients with incorrect component position. Failure to significantly improve motion of total knees was found in 6 of 7 patients with inflammatory connective tissue disease and posttraumatic arthritis. Intraoperative patellar tendon avulsion occurred in one post.traumatic arthritic revision and the resultant prolonged postoperative immobilization contributed to a poor result. Those knees requiring extensive soft tissue release or having substantial bone loss resulting in large flexion-extension gaps required Total Condylar III prostheses to produce stability. A larg.e tibial insert resulted in elevation of the joint line, iatrogenic patella baja, and poor improvement in motion. Because of this experience we recommend that bone grafts or augmented femoral components should be used to allow maintenance of a lower joint line (5; 8). All uncemented knees were revised to cemented components to minimize postoperative pain and permit rapid full weight bearing. Those knees with a tendency to undergo fibrosing anklyosis after arthroplasty should be cemented to allow full weight bearing using a heel-toe gait, thus promoting range of motion. In patients who have demonstrated aggressive fibrous tissue formation, components should be placed so that mild ligamentous laxity is present. A judicious postoperative physical therapy program is invaluable to all total knee anhroplasty patients. Those with low pain thresholds, inflammatory connective tissue disease, or poor muscle function
Revision Surgery for Stiff TKA 9 Nicholls and Dorr
can benefit from supervised physical therapy (3) rather than self administered exercises. Although statistical trends cannot be derived from this data, good results in revision of a stiff total knee can be expected w h e n recognized mechanical problems are corrected (6). W h e n no connective tissue disease is present, the joint line is maintained so that patella baja is avoided, and adequate quadriceps function is present. A V-Y patella t u r n d o w n should be used w h e n needed to avoid extensor m e c h a n i s m rupture. This approach is used either w h e n the patella cannot be laterally dislocated and everted with the midline exposure or if the patella dislocates during flexion even with a lateral release. We prefer patella t u r n d o w n to tibial tubercle elevation because of o u t favorable experience and because we desire to avoid possible h e m a t o m a formation from b o n e bleeding at the level of the distal incision. The results of revision for stiff total knees is not as good as those reported for primary arthroplasty of stiff knees (1). Forty percent of stiff total knees had an excellent or good rating c o m p a r e d to 81% of index operations on stiff knees. In both studies, pa-
$77
tients with osteoarthritis had a good chance for satisfactory results.
References 1. Aglietti P, Windsor RE, Buzzi R, Insall JN. Arthroplasty for the stiff or ankylosed knee. J Arthroplasty 4:1, 1989 2. Coonse KK, Adams JD: A new operative approach to the knee joint. Surg Gynecol Obstet 77:344, 1934 3. Dorr LD: Optimizing results of total joint arthroplasty. A.A.O.S. Instructional Course Lectures. Vol. XXXIV, p. 410, 1985. 4. Dorr LD: Technical considerations in total knee arthroplasty. Clin Orthop 205:5, 1986 5~ ttungerford DS, Kenna RV: Revising failed total knees/ technical considerations in total knee arthroplasty. Techniques in Orthopaedics, 1984 6. Jacobs MS, Hungerford DS: Revision total knee arthroplasty of aseptic failure. Clin O~hop 226:78, 1988 7. Insall JN: A midline approach to the knee. p. 1584. JBJS. 53A: , 1971 8. Scott RD: Revision total knee arthroplasty. Clin Orthop 226:65, 1988