n Feature Article
Rotating-hinge Total Knee for Revision Total Knee Arthroplasty Alessandro Bistolfi, MD; Giuseppe Massazza, MD; Federica Rosso, MD; Maurizio Crova, MD
abstract Full article available online at ORTHOSuperSite.com. Search: 20120222-34 Rotating-hinge knee implants are used for revision total knee arthroplasty in patients with severe ligament instability and bone loss. This study evaluated the outcomes of a series of rotating-hinge knees. Thirty-one NexGen Rotating Hinge Knees (Zimmer, Warsaw, Indiana) were implanted in 29 patients (2 bilateral), with an average age of 72.8 years. Indications for surgery were aseptic loosening (n523), septic loosenings (n54), tibiofemoral instability (n53), and wear (n=1). The Hospital for Special Surgery Knee Score and the Knee Society Roentgenographic Evaluation System were used. Statistical and cumulative survival rate analyses were performed. Average follow-up was 60.3 months (range, 32-100 months). The Hospital for Special Surgery Knee Score results indicated statistically significant improvement; the total score increased from 65.5 preoperatively to 88.4 postoperatively. Average range of motion increased from 90.9° preoperatively to 114.4° postoperatively. Radiographs showed no periprosthetic bone fractures or implant ruptures. Radiolucent lines were found in 20 of 26 patients and were progressive in 2 (both revised). Complications occurred in 10 patients.
Figure: Kaplan-Meier survival estimate curve with revision of the implant as the endpoint. Four (78.6%) cases (SD, 611.5%) at 5-year follow-up. Time is reported in months.
The rigidity of the hinge may be associated with a risk of aseptic loosening due to the increased stress transfer to the bone from the prosthesis through the locked hinge. Rotating-hinge knee implants provided acceptable mid-term outcomes for revision knee surgery with ligamentous instability. They are not at higher risk for early loosening unless short tibial stems are used. The high percentage of failures is more related to the complex surgery and to the status of the patients than to the hinged mechanism.
Drs Bistolfi, Massazza, and Crova are from the Department of Orthopedics, Traumatology and Health Medicine, CTO/M, Adelaide Hospital, and Drs Massazza, Rosso, and Crova are from the University of the Studies of Turin, Turin, Italy. Drs Bistolfi, Massazza, Rosso, and Crova have no relevant financial relationships to disclose. Correspondence should be addressed to: Alessandro Bistolfi, MD, Department of Orthopedics, Traumatology and Health Medicine, CTO/M, Adelaide Hospital, Via Zuretti 29, 10126 Turin, Italy (
[email protected]). doi: 10.3928/01477447-20120222-34
MARCH 2012 | Volume 35 • Number 3
e325
n Feature Article
R
otating-hinge knee implants have been available since the 1970s as an improvement of the fixed-hinge models. The flexion–extension movement in these implants is combined with rotation of the femur on the tibial component, or with rotation of the tibial polyethylene liner on the metal tibial tray, thus allowing a more physiologic range of motion (ROM) and reducing the stress transfer to the bone–prosthesis interface compared with the fixed-hinge models.1,2 Hinged implants are currently more commonly used for revision total knee arthroplasty (TKA) but can also be used for primary TKA.3 Moderate to severe knee instability, ligament deficiency (eg, absence of 1 or both collateral ligaments), severe bone loss, or the presence of varus, valgus, or flexion deformities, are common indications for this type of implant. The NexGen Rotating Hinge Knee (Zimmer, Warsaw, Indiana) was introduced in 2002 for use with the patellar components, augments, and stem extensions of the NexGen Complete Knee System (Zimmer). Optimal results with hinged implants in terms of recovery of ROM and intrinsic stability have been described in the literature for complex primary implantation and revision surgery.1,4-6 Nevertheless, some authors reported that the high rigidity of the hinge is associated with a greater risk of aseptic loosening because the system may increase stress transfer to the bone from the prosthesis.2,7,8 Therefore, constrained condylar knee implants are used more frequently than rotating-hinge knee implants for revision TKA, even in patients with bone loss and ligament instability. Because a constrained condylar knee is typically less constrained and rigid compared with a rotating-hinge knee, it is possible that the stress transfer would be reduced with a constrained condylar knee. The purpose of this study was to evaluate the mid-term clinical and radiographic outcomes of a series of NexGen Rotating Hinge Knees (Zimmer), which were implanted consecutively for revision surgery in patients with ligament instability.
e326
Materials and Methods Between September 2002 and May 2008, thirty-one NexGen Rotating Hinge Knees (Zimmer) were implanted in a consecutive series of revision total knee arthroplasties in patients with ligament instability. The rotating-hinge knees were implanted in 29 patients (2 bilateral). Twenty-four (82.8%) patients were women, and 5 (17.2%) were men, with an average age of 72.8 years (95% confidence interval [CI], 69.7-75.8; range, 43-81 years; standard deviation [SD]68.6 years). The indications for revision were aseptic loosening in 23 (74.2%) cases, septic loosening in 4 (12.9%), tibiofemoral instability in 3 (9.7%), and polyethylene tibial liner wear in 1 (3.2%).
Surgical Technique The NexGen Rotating Hinge Knee (Zimmer) provides varus/valgus and anteroposterior constraint, while allowing flexion, extension, and rotation. The hinge is secured to the femoral component, and a hinge postextension connects to the hinge while being inserted through the articular surface and into a hole in the tibial component. The polyethylene articular surface is free to rotate on the proximal tibial surface up to 25° in internal or external rotation (50° total), allowing the polyethylene to follow the movement of the femoral condyles. The tibial and femoral components are compatible with stem extensions and augments to address bone loss. All surgeries were performed under general anesthesia and were conducted with an ischemic limb using a pneumatic tourniquet (240-280 mmHg) at the thigh for a maximum of 120 minutes. If the surgery lasted .120 minutes, the tourniquet was released, and an accurate hemostasis was performed. According to the situation and the time, the tourniquet was reinflated for cementation or the surgery was finished without the tourniquet. The previous cutaneous approach was followed in all cases, extending it proximally and distally as necessary. A medial parapatellar capsulotomy was used. Osteotomy
of the tibial tubercle was performed in 1 case, and a Coonse-Adams detachment of the quadriceps tendon was performed in 6 cases. Combination of the Coonse-Adams detachment and osteotomy of the tibial tubercle was necessary in 1 case because of poor patellar tendon quality. A technique of hybrid fixation, metaphyseal cementation and uncemented press-fit intramedullary tibial and femoral stems, was chosen. In 12 (38.7%) of 31 cases, a 75-mm tibial stem was used, whereas longer tibial stems were used in 19 (61.3%) cases. The procedures were 1-stage revisions, except 3 of the 4 cases of septic loosening, in which a 2-stage revision technique was used. In these cases, a fixed antibiotic bone cement spacer was used, and the final prosthesis was implanted after approximately 3 months. The selective patellar treatment was revision of the patella if mobilized, broken, or delaminated, or in the presence of dislocation and maltracking. Similarly, in cases where the primary patella was not resurfaced, selective resurfacing was performed in the presence of dislocation and maltracking. Therefore, patellar components were implanted in 20 (64.5%) of 31 cases (3 revisions of the previous patella and 17 resurfacings of the native patella). The patella was thermally denervated in all cases. Bone loss was quantified intraoperatively according to the classification proposed by Engh9,10 (Table 1). Table 2 specifies the method used to fill the bony defects: cement in 27 cases; femoral wedges in 1 case (case #2), tantalum tibial spacer wedges in 2 cases (cases #13, 17), and autologous bone in 1 case (case #3). The rehabilitative protocol was different for each case. In general, a cautious approach was preferred to an aggressive one to reduce the risk of complications. A knee brace was used after osteotomy of the tibial tubercle and the Coonse-Adams detachment, usually fixed in extension for the first month and then progressively articulating. Weight bearing was allowed according to the nature of the bone defect but was generally partially protected by 2 crutches for 1 month and progressive in
ORTHOPEDICS | ORTHOSuperSite.com
Rotating Total Hinge Knee | Bistolfi et al
the second month. All patients received an antithrombotic prophylaxis with lowmolecular-weight heparin and an antibiotic prophylaxis with 1 g of vancomycin 1 hour preoperatively and at intervals of 12 and 24 hours postoperatively. Septic cases were treated with specific antibiotic therapy according to the antibiogram as well as the clinical findings and laboratory examinations. Clinical Evaluation The Hospital for Special Surgery Knee Score (HSS-KS) was used for clinical evaluation. The scores were defined as excellent (85-100 points), good (7084 points), fair (60-69 points), and poor (,60 points).11 The patients were clinically evaluated preoperatively and postoperatively at 3 and 6 months and annually thereafter. Clinical evaluations and medications prior to 3 months were not counted for scoring purposes. Radiographic Evaluation The Knee Society Roentgenographic Evaluation System (KS-RES) was used for radiographic evaluation.12 All patients were radiographically and clinically evaluated at the same intervals (preoperatively, 3 and 6 months postoperatively, and annually thereafter), with views including anteroposterior, lateral, and patellar skyline. Radiographs were studied to assess the position of the implants, the presence of periprosthetic fracture, signs of loosening (progressive radiolucent lines), and the presence of osteolysis. Radiolucent lines were defined as lines with a bone–implant (cement) distance .2 mm and were subdivided into nonprogressive (not typically indicative of migration) and progressive (usually associated with high probability of implant loosening). The analysis was conducted on digitized radiographs available in our hospital database, or on films provided by the patients for older radiographs (up to 2004) and for radiographs taken in other facilities.
MARCH 2012 | Volume 35 • Number 3
Table 1
Distribution and Incidence of Bone Loss by Engh Classification10 Classification
Femur, No. (%)
Tibia, No. (%)
1
11 (35.5)
12 (38.7%)
2A
7 (22.6)
8 (25.8)
2B
6 (19.4)
4 (12.9)
3
7 (22.6)
7 (22.6)
Table 2
Bone Losses and the Treatment to Fill Them Identification Number
Femoral Losses
Tibial Losses
Type of Filling
1
F1
T1
Cement
2
F3
T3
Femoral Wedges
3
F3
T3
Autologous Bone
4
F2a
T2a
Cement
5
F1
T1
Cement
6
F2b
T2b
Cement
7
F3
T3
Cement
8
F1
T1
Cement
9
F1
T1
Cement
10
F2a
T2a
Cement
11
F2b
T3
Cement
12
F1
T1
Cement
13
F3
T2b
Tantalum tibial spacer wedges
14
F2a
T2a
Cement
15
F2a
T2a
Cement
16
F1
T1
Cement
17
F2a
T2b
Tantalum tibial spacer wedges
18
F1
T1
Cement
19
F3
T3
Cement
20
F2b
T2a
Cement
21
F3
T3
Cement
22
F1
T1
Cement
23
F2a
T2a
Cement
24
F2b
T2b
Cement
25
F2b
T2a
Cement
26
F1
T1
Cement
27
F2a
T2a
Cement
28
F3
T3
Cement
29
F2b
T1
Cement
30
F1
T1
Cement
31
F1
T1
Cement
e327
n Feature Article
Results
Table 3
Preoperative and Last Available Postoperative Averages for Hospital for Special Surgery Knee Score Subcategories Mean6Standard Deviation (95% Confidence Interval) Subcategory Pain score
Preoperative
Last Follow-up
P
14.566.9 (11.8-17.1)
24.468.6 (21.1-27.7)
.000002
Function score
9.564.9 (7.6-11.3)
15.565.4 (13.4-17.6)
.0002
Muscle strength
8.860.9 (8.4-9.2)
9.460.9 (8.9-9.7)
.0215
Flexion deformity
9.960.4 (9.8-10.1)
9.960.4 (9.7-10.1)
.9101
961.4 (8.5-9.5)
9.960.4 (9.7-10.1)
.0021
90.9°615.1° (85.1°-96.7°)
114.4°615.8° (108.3°-120.5°)
.00001
65.5610.1 (61.7-69.4)
88.4619.7 (81.1-95.7)
.00006
Instability Range of motion Total score
1 Figure 1: Kaplan-Meier survival estimate curve taking revision of the implant as the endpoint. Four (78.6%) cases (SD611.5%) at 5-year follow-up. Time is reported in months.
2 Figure 2: Kaplan-Meier survival estimate curve taking all causes of failure as endpoint. Seven (70.1%) cases (SD612.1%) at 5-year follow-up. Time is reported in months.
e328
Data Collection and Statistical Analysis In 1998, we developed computer software based on the HSS-KS and the KS-RES to collect data regarding TKA and revision TKA. All data were collected and stored in the database, starting with the first preoperative evaluation and continuing at every clinical and radiographic evaluation. Therefore, this study was a prospective analysis. The program allows for statistical analysis and the comparison of different groups. The Kaplan-Meier method was used to estimate the cumulative survival rate, using failure with re-revision of the implant as the endpoint. Clinical and radiographic data were analyzed by arithmetic mean and compared using 95% CIs and SD. Statistical significance was assessed by calculating the P value using Student’s t test, with a threshold of P,.05.
The study ended in January 2011, with an average follow-up of 60.3 months (95% CI, 53.7-66.7 months; range, 32100 months, SD616.8 months). One patient died during follow-up, but this case evaluation was still included in the study. Three (9.6%) patients were lost at followup. Therefore, 28 cases (26 patients, 2 bilateral) were included in the analysis. Clinical Results The HSS-KS results showed statistically significant improvements from the preoperative to the postoperative evaluations. The total score increased from an average score of 65.5 points (fair) preoperatively to 88.4 points (excellent) at last available followup. Average pain changed from 14.5 points (mild to moderate) preoperatively to 24.4 points (none to mild) postoperatively, and average ROM increased from 90.9° preoperatively to 114.4° postoperatively. Table 3 reports the preoperative and last available postoperative averages for principal subcategories of the HSS-KS. Radiographic Analysis Of 28 cases, 26 (92.9%) underwent radiographic evaluation. Radiographs did not show signs of periprosthetic bone fractures or implant fracture. The mean tibiofemoral angle was 6.2° valgus (SD62.7°). Radiolucent lines were found in 20 of 26 cases. Of these, 18 presented as nonprogressive radiolucent lines, which were distributed around the femoral component in 11 cases and around the tibial component in 13, with 6 cases presenting femoral and tibial radiolucent lines. Progressive radiolucent lines were found in 2 cases around the tibial component, with involvement of zones 5 to 7 of the anteroposterior view and zones 1 to 3 of the lateral view. These cases were revised for failure and substitution of the tibial component with a longer stem. Complications and Survival Estimate Postoperative complications occurred in 10 (35.7%) of 28 cases. One case had
ORTHOPEDICS | ORTHOSuperSite.com
Rotating Total Hinge Knee | Bistolfi et al
a permanent deficit of the external popliteal sciatic nerve, which partially compromised functionality. Another case had an intra-articular hematoma, which was drained along with articular wash and poly exchange. One patient sustained a wound dehiscence, requiring surgical revision of the skin for complete healing. No cases of deep venous thrombosis occurred. Seven additional cases (25%) with complications were considered failures: aseptic loosening of the tibial component (n52), septic loosening after 2-stage revision for previous infection (n52), and 3 additional cases were considered clinical failures but were not revised. Of the 2 septic loosenings, 1 was re-revised, and the other was treated with biological arthrodesis using an external circular fixator in a chronic alcoholic patient. Thus, failure resulting in revision of an implant occurred in 4 (14.3%) cases. Three additional cases were considered clinical failures, although the implants are still in situ: 2 traumatic lesions of the extensor apparatus and 1 tibial fracture. The cumulative survival estimates using endpoints of (1) revision of the implant and (2) all causes of failure are reported in Figures 1 and 2, respectively.
Discussion Revision TKA has become more frequent in recent years, and the incidence is destined to grow as the rate of primary TKA increases, with causes of TKA failure including aseptic loosening, infection, misalignments, instability, and mechanical failure. These situations are often associated with severe ligament damage and, therefore, may require hinged implants. Comparing the different hinged implants may be difficult, and concerns exist about the ideal hinge level. Initially, a high rate of failure was described with rotatinghinge knee implants, whereas some good results have been described more recently.1,5,13-16 Nevertheless, several authors report discouraging results,2,17,18 and many think that the hinge transfers high stresses
MARCH 2012 | Volume 35 • Number 3
to the implant–bone interface and causes loosening. This prospective study analyzed the mid-term results of an unselected series of NexGen Rotating Hinge Knee (Zimmer) implants used in revision TKA in patients with severe ligament instability. Analysis was based on clinical and radiographic evaluations, with the duration of followup and number of cases comparable withsimilar studies. However, our sample size was relatively small for the purpose of estimating survivorship and, therefore, resulted in a wide 95% CI for the KaplanMeier estimate. Statistically significant improvement of the overall HSS-KS score and its individual scoring components was demonstrated in all patients, including the 3 cases that were clinical failures, along with achievement of good clinical results, thus confirming the validity of the revision procedure despite cost and complications. Although the complication rate was high, it was similar to rates reported in the literature for this type of revision surgery.4,16,19 Deficit of the external popliteal sciatic nerve is not a frequent complication in knee arthroplasty but can occur in revision surgery. The authors preferred a wide cementation because the cement is easy to handle, can fill almost any kind of defect, is stable when polymerized, and gives excellent metaepiphyseal fixation. Taking revision of the implant as the endpoint, the survival rate was approximately 79% at 5 years (95% CI, 63.7%93.4%). According to the existing literature, this is a good result for this type of procedure. In addition, the 4 implant revisions can be commented on. One failure after aseptic revision was related to the use of a 75-mm tibial stem, which we do not recommend. This implant requires a more distal and stronger press-fit fixation compared with the constrained condylar knee model, where a shorter stem might be enough. The use of longer stems could reduce the risk of loosening. Persistence of infection occurred in 2 of the 4 septic revi-
sions. However, 1 of these patients was an elderly man who was a chronic alcoholic with diabetes mellitus and poor compliance and was, therefore, at high risk. The other 1 had no additional risk factors. This patient was treated with arthrodesis, with good results until his death. When possible, we adopted the 2-stage technique in septic cases, and the results were similar to those reported in the literature.20,21 Three more cases of failure occurred without implant revision. Two patients had complete traumatic lesions of the extensor apparatus, with necessity to use an aid during walking. In these cases, revision surgery was not possible. One patient was a 32-year-old woman with rheumatoid arthritis, severe sufferance of the skin, and low functional request, who recovered an acceptable level of ability with nonoperative treatment. The other was a 78-year-old man with several comorbidities who was contraindicated for further surgeries by the anesthesiologist. The third clinical failure was a 76-yearold woman with a periprosthetic tibial fracture associated with osteotomy for previous tibial stem removal and wound problems. Due to impaired general health conditions, we avoided revision surgery and instructed her to walk with a knee brace. She developed a nonunion below the tip of the tibial stem. A longer stem might have prevented the tibial fracture and should have been used. The rotating-hinge knee implant in this study demonstrated clinical and radiographic results comparable with those reported in the literature. Nevertheless, the complication and clinical failure rates were high. The use of short stems must be avoided when using this prosthesis. However, the strong hinge is not the main cause of failure for the implants, whereas other factors are more relevant. A more careful evaluation of the patient, considering all general medical conditions and local anatomy, is necessary. Revision surgery is a higher risk than primary surgery, and patients must be presented with de-
e329
n Feature Article
tailed and specific risk vs benefit information to provide informed consent. Patients with severely compromised health conditions and inadequate compliance should not be selected for revision surgery. In some cases, arthrodesis should be considered to reduce the rate of re-revision or failure and to achieve a higher rate of well-functioning and long-lasting revision TKA. According to our results, we do not implicate the hinge of the NexGen Rotating Hinge Knee System (Zimmer) as the main cause of failure this mid-term follow-up. The prostheses implanted in ideal conditions of bone anatomy and general patient health are still doing well. Rotating-hinge knee models provide better stability in cases of ligamentous deficiency and offer more confidence for the restoration of a physiologic joint line than constrained condylar knee models, particularly when constrained condylar knee models require significant tightening of the capsular structure to achieve adequate tension and ligament balancing. However, controlled studies are currently lacking and should be recommended to compare rotatinghinge and constrained condylar knee implants.
Conclusion Rotating-hinge knee implants provide acceptable clinical and radiological results at mid-term follow-up. They are not at higher risk for early loosening due to stress transfer through the locked hinge unless short tibial stems are used. According to our results, rotating-hinge knee implants are still the implant of
e330
choice for revision knee surgery in patients with ligamentous instability. The elevated percentage of failures is more related to the complex surgery and to the statuses of the patients than to the hinged mechanism.
References 1. Barrack RL. Evolution of the rotating hinge for complex total knee arthroplasty. Clin Orthop Relat Res. 2001; 392:292-299. 2. Pour AE, Parvizi J, Slenker N, Purtill JJ, Sharkey PF. Rotating hinged total knee replacement: use with caution. J Bone Joint Surg Am. 2007; 89(8):1735-1741. 3. Sculco TP. The role of constraint in total knee arthroplasty [published online ahead of print April 17, 2006]. J Arthroplasty. 2006; 21(4 suppl 1):54-56. 4. Deehan DJ, Murray J, Birdsall PD, Holland JP, Pinder IM. The role of the rotating hinge prosthesis in the salvage arthroplasty setting [published online ahead of print March 4, 2008]. J Arthroplasty. 2008; 23(5):683-688. 5. Joshi N, Navarro-Quilis A. Is there a place for rotating-hinge arthroplasty in knee revision surgery for aseptic loosening [published online ahead of print April 14, 2011]? J Arthroplasty. 2008; 23(8):1204-1211. 6. Barrack RL, Lyons Johnson JC. The use hinge component in tal knee arthroplasty. 15(7):858-866.
TR, Ingraham RQ, of a modular rotating salvage revision toJ Arthroplasty. 2000;
7. Guenoun B, Latargez L, Freslon M, Defossez G, Salas N, Gayet LE. Complications following rotating hinge Endo-Modell (Link) knee arthroplasty [published online ahead of print October 17, 2009]. Orthop Traumatol Surg Res. 2009; 95(7):529-536. 8. Rand JA, Chao EY, Stauffer RN. Kinematic rotating-hinge total knee arthroplasty. J Bone Joint Surg Am. 1987; 69(4):489-497. 9. Engh GA, Ammeen DJ. Bone loss with revision total knee arthroplasty: defect classification and alternatives for reconstruction. Instr Course Lect. 1999; 48:167-175.
10. Engh GA, Parks NL. The management of bone defects in revision total knee arthroplasty. Instr Course Lect. 1997; 46:227-236. 11. Insall JN, Dorr LD, Scott RD, Scott WN. Rationale of the Knee Society clinical rating system. Clin Orthop Relat Res. 1989; (248):13-14. 12. Ewald FC. The Knee Society total knee arthroplasty roentgenographic evaluation and scoring system. Clin Orthop Relat Res. 1989; 248:9-12. 13. Walker PS, Manktelow AR. Comparison between a constrained condylar and a rotating hinge in revision knee surgery. Knee. 2001; 8(4):269-279. 14. Lombardi Jr AV, Mallory TH, Eberle RW. Constrained knee arthroplasty. In: Scott WN, ed. The Knee. St Louis, MO: CV Mosby; 1994:1305. 15. Westrich GH, Mollano AV, Sculco TP, Buly RL, Laskin RS, Windsor R. Rotating hinge total knee arthroplasty in severely affected knees. Clin Orthop Relat Res. 2000; (379):195-208. 16. Neumann DR, Hofstaedter T, Dorn U. Follow-up of a modular rotating hinge knee system in salvage revision total knee arthroplasty. J Arthroplasty. [published online ahead of print October 12, 2011]. 17. Hui FC, Fitzgerald RH Jr. Hinged total knee arthroplasty. J Bone Joint Sur Am. 1980; 62(4):513-519. 18. Wang CJ, Wang HE. Early catastrophic failure of rotating hinge total knee prosthesis. J Arthroplasty. 2000; 15:387-391. 19. Gudnason A, Milbrink J, Hailer NP. Implant survival and outcome after rotating-hinge total knee revision arthroplasty: a minimum 6-year follow-up [published online ahead of print June 9, 2011]. Arch Orthop Trauma Surg. 2011; 131(11):1601-1607. 20. Haddad FS, Masri BA, Campbell D, McGraw RW, Beauchamp CP, Duncan CP. The PROSTALAC functional spacer in two-stage revision for infected knee replacements. Prosthesis of antibiotic-loaded acrylic cement. J Bone Joint Surg Br. 2000; 82(6):807-812. 21. Cuckler JM. The infected total knee: management options. J Arthroplasty. 2005; 20(4 suppl 2):33-36.
ORTHOPEDICS | ORTHOSuperSite.com
