Summary
Total knee arthroplasty in the rheumatoid patient pre- sents unique challenges, including the systemic nature of the patient’s disease, the presence of significant soft-tis- sue deformities and osteopenic bone, and an increased risk of complications such as wound healing and persis- tent contractures. In order to maximize the probability of a successful outcome, the surgeon must optimize the pa- tient’s pre-operative medical status, pay meticulous at- tention to soft-tissue balancing and contracture release in the operating room, and closely monitor the patient’s postoperative course.Adherence to these principles opti- mizes the results of total knee arthroplasty in the rheuma- toid patient, making this a very rewarding procedure for both patient and surgeon.
Introduction
The knee joint is affected in approximately 90% of pa- tients with chronic rheumatoid arthritis [1]. Total knee arthroplasty (TKA) provides the rheumatoid patient with substantial alleviation of pain and deformity.However,the rheumatoid knee presents several challenges to the sur- geon in the operating room as well as in the pre- and post- operative stages. Rheumatoid arthritis is a systemic dis- ease that affects multiple organ systems, and rheumatoid patients frequently take several immunosuppressive med- ications that must be addressed in the perioperative peri- od.The surgeon encounters several important issues when planning knee arthroplasty, including the timing of knee surgery relative to other arthritic joints and the choice of anesthesia. At the time of surgery, the rheumatoid knee is characterized by osteopenic bone, valgus deformity with a frequently incompetent medial collateral ligament, and soft-tissue contractures. The level of constraint of the prosthesis is an important decision.Extreme care must be given to soft-tissue balancing. Postoperatively, the rheumatoid patient may be affected by wound-healing problems, infection, and loss of full extension. This chap- ter discusses the issues of importance to the surgeon per- forming knee arthroplasty in the rheumatoid patient.
Preoperative Considerations and Planning
The systemic involvement in rheumatoid patients is an important issue in the preoperative period. Several organ systems are affected by the disease as well as by the im- munosuppressive medications commonly used to treat it.Approximately 10% of rheumatoid patients undergoing total knee arthroplasty are taking maintenance corticos- teroids [1].
Systemic Manifestations of Rheumatoid Arthritis
Rheumatoid patients should routinely undergo a com- plete medical evaluation prior to knee arthroplasty. This evaluation should also include a complete blood count, urinalysis, urine culture, electrolytes, and an electrocar- diogram [2]. The evaluation should include examining the patient for remote sites of potential infection such as the oral cavity. The skin over the knee in rheumatoid pa- tients may be thin and atrophic secondary to chronic steroid therapy or as a manifestation of the disease process. Rheumatoid arthritis is considered to be a cata- bolic, wasting disease. Therefore, many rheumatoid pa- tients may be malnourished even if they are not clinical- ly underweight [3].
Management of Corticosteroids
It has been common practice to administer stress-dose steroids at the time of surgery to patients who take chron- ic maintenance steroids. The purpose of this is to prevent adrenal insufficiency, particularly in patients who take relatively high doses of steroids (more than 20-30 mg hy- drocortisone daily) [2].It is our practice to administer 100 mg hydrocortisone iv before surgery and then Q8H for three doses after surgery. However, the need for routine exogenous steroid administration has been questioned in one study. Friedman et al. [4] reported on 28 patients tak- ing chronic steroids who underwent 35 major orthopedic
31 Understanding the Rheumatoid Knee
K. K. Anbari, J. P. Garino
surgeries. Patients were given only their usual doses of oral steroids without supplementation. No patients had evidence of adrenal insufficiency on physical exam or lab- oratory criteria.
Management of Other Immunosuppressive Medications
There exists conflicting evidence regarding the decision to discontinue non-steroid rheumatoid medications at the time of surgery. Grennan et al. [5] described a prospective randomized study of 388 rheumatoid pa- tients undergoing orthopedic surgery. The study found that certain remittive agents such as penicillamine, hy- droxychloroquine,and cyclosporine were associated with statistically significant increased risk of wound problems.
However, patients who continued to take methotrexate had no increase in wound complications and experienced fewer rheumatoid flares than those who did not receive methotrexate.In contrast,Bridges et al.[6] reported on 38 rheumatoid patients undergoing elective orthopedic surgery. There were four infections in 19 procedures per- formed on patients who took methotrexate around the time of surgery, compared with no complications in 34 procedures for patients who discontinued methotrexate 4 weeks before surgery. Other sources recommend dis- continuing methotrexate and other similar agents 1-2 weeks before surgery and restarting them 1-2 weeks after surgery [2]. It remains our practice to discontinue these medications for 2 weeks before surgery and to restart them 1 week thereafter.
There is scant information in the literature about the treatment of the newer anticytokine agents such as etan- ercept in the perioperative period.One case report [7] de- scribes disseminated joint infections and fatal septic shock in a rheumatoid patient on etanercept who had a history of bilateral hip and knee prostheses. The authors caution that etanercept may mask the signs of acute in- fection and inflammation,and that a patient on this agent should be monitored closely for early symptoms of infec- tion and be treated aggressively.
Timing of Knee Arthroplasty Relative to Other Orthopedic Surgery
The rheumatoid patient presenting with end-stage knee arthritis may also suffer from joint pain and deformity in the spine, upper extremities, hips, and feet.
Cervical Spine .It is important for the arthroplasty surgeon to evaluate the rheumatoid patient for evidence of cervi- cal spine disease. It is estimated that 88% of rheumatoid patients have some degree of cervical spine involvement
[8]. The most common manifestations of this are at- lantoaxial subluxation, basilar invagination, and subaxi- al subluxation.The surgeon should obtain a thorough his- tory, looking for worrisome signs such as neck pain ex- tending to the head, upper or lower extremity weakness, urinary or bowel incontinence,dysphagia,and loss of fine motor coordination.The physical examination should in- clude neurological motor and sensory testing of the up- per and lower extremities.Any positive finding should be further investigated with cervical spine radiographs in- cluding AP, flexion and extension lateral, and odontoid views. Because of the devastating complications that may arise from spinal instability, stabilization of the cervical spine takes first priority in the rheumatoid patient. It is important to note that despite the progressive nature of cervical rheumatoid disease, only 15% of rheumatoid pa- tients require cervical spine surgery [9]. Our practice has been to obtain cervical spine radiographic series if there are any concerns in the history or physical, and subse- quently to refer the patient for orthopedic spine evalua- tion if the radiographs demonstrate instability. Finally, cervical spine involvement complicates the options for anesthesia as discussed below.
Upper Extremity. It is usually advisable for the rheumatoid patient to undergo lower extremity joint reconstruction before shoulder or elbow reconstruction to avoid stress- ing upper extremity prosthetic joints with the use of crutches or a walker. Furthermore, addressing lower ex- tremity joints promptly may spare the debilitated patient from becoming wheelchair bound and preserve mobility.
Occasionally, a rheumatoid patient may have such severe limitation of upper extremity or hand function that use of assistive devices after knee arthroplasty may be im- possible. In these cases, we turn our attention to the up- per extremity first.
Hip Versus Knee Reconstruction. When both the hip and
the knee need surgical reconstruction, it is usually
preferable to begin with the hip [3, 8]. Rehabilitation of a
hip arthroplasty may proceed with a diseased ipsilateral
knee more effectively than rehabilitation of a knee with
a diseased ipsilateral hip. However, there are important
exceptions. First, a rheumatoid knee with severe valgus
deformity would jeopardize the stability of an ipsilater-
al hip because of the resulting hip internal rotation and
adduction. In this circumstance, it may be safer to pro-
ceed with knee arthroplasty first. Second, a patient with
severe bilateral knee flexion contractures would have dif-
ficulty standing erect until both knees are reconstruct-
ed. Such a patient would be a good candidate for bilater-
al knee arthroplasty if medically appropriate. We have
found that in selected cases, simultaneous ipsilateral hip
and knee arthroplasty could be given consideration as
well.
rheumatoid patient because of the systemic nature of the disease and the multiple medications used to treat it. An anesthesiologist with experience in this situation is cru- cial to a good outcome.If intubation is necessary,fiberop- tic management of the airway may be advisable to avoid excessive manipulation of the cervical spine [10].The tho- racic, lumbar, and sacral portions of the spine are usual- ly spared in rheumatoid patients, making regional spinal anesthesia ideal in these patients. Regional anesthesia also provides the advantages of blunting the neurohor- monal response to surgery and offering effective postop- erative analgesia.
Intraoperative Options
The surgeon faces several important choices when decid- ing on the optimal prosthesis and surgical technique for a rheumatoid knee. This section reviews the literature and gives the authors’ perspective.
Cemented Versus Cementless. The evidence in the litera- ture has been inconsistent regarding whether cementless fixation has a role in the treatment of the rheumatoid knee. There are several studies with a relatively small number of patients that show adequate results for ce- mentless knees in the intermediate term, but with some concerns, nonetheless. Schroder et al. reported on 41 ce- mentless knees in rheumatoid patients with an average follow-up of 54 months [11] and described one tibial revi- sion and five tibias with radiolucencies. While these au- thors quoted a favorable success rate of 97% (40/41) after 4-5 years, these results were not long term.Another study of 103 cementless knees implanted in patients with rheumatoid arthritis or osteoarthritis described two tib- ial revisions for aseptic loosening,two tibias with asymp- tomatic loosening, and four tibias with lucent lines at a follow-up period of 3 years [12]. Rosenqvist et al. [13]
showed that more than half of 34 porous-coated anatom- ical (PCA) knee arthroplasties demonstrated radi- ographic evidence of displacement, and all knees had a radiolucent zone at an average of 17 months. Analysis of 3054 rheumatoid knees from the Swedish Knee Arthro- plasty Register for which data on cementation were avail- able showed statistically significant reduction in loosen- ing as well as revision rate in the cemented knees [14].Bo- goch and Moran commented on the issue of cementless knees for rheumatoid patients [15].They stated that in the absence of convincing evidence favoring cementless knees and because of the good long-term results with ce- mented knees,cemented knee arthroplasty was preferred in this patient population. Our practice is to perform ce- mented knee arthroplasty in this population because of
weight-bearing.
Patellar Resurfacing. We favor routine resurfacing of the patella in all rheumatoid patients, based on extensive ev- idence in the literature. Remaining articular cartilage on the undersurface of the patella may provide an antigenic stimulus in the rheumatoid knee that allows the synovial inflammation to persist, leading to anterior knee pain [16]. In a large study involving more than 27 000 total knee replacements from the Swedish Knee Arthroplasty Register [17], there was a higher proportion of satisfied patients in the patellar resurfacing group, both for rheumatoid patients and overall.An interesting study ex- amined in a prospective fashion 35 rheumatoid patients who underwent simultaneous bilateral TKA with resur- facing of one randomly chosen patella while leaving the contralateral one unresurfaced [18, 19]. At 2-year mini- mum follow-up, Hospital for Special Surgery knee scores and range of motion were similar in the two sets of knees [18]; this result was corroborated at 6-year minimum fol- low-up [19]. However, the authors evaluated parameters specific to the patellofemoral joint at 6-year follow-up that they did not at 2-year follow-up, namely, tenderness over the patellofemoral joint and pain with using stairs.
They reported that patellofemoral tenderness and pain with stairs were found in eight and nine unresurfaced knees,respectively,out of 26 knees remaining in the study.
None of the resurfaced knees had these symptoms. They therefore recommended routine patellar resurfacing based on these results. Finally, even when another group of authors [20] resurfaced the patella selectively (i.e.,only when the surgeon found loss of cartilage under the patel- la, exposed bone, gross surface irregularity, or abnormal tracking), they found a higher proportion of anterior knee pain in the unresurfaced group versus the resur- faced group at a mean of 6.5 years.These authors similarly recommended routine resurfacing of the patella.
Posterior Cruciate-Retaining Versus -Substituting Designs . While both cruciate-retaining and cruciate-substituting designs have excellent results in the osteoarthritic knee, outcomes in rheumatoid knees have been less consistent.
The concern with cruciate-retaining prostheses is that the inflammatory process in rheumatoid arthritis compromis- es the PCL, exposing the patient to the risk of late rupture of the PCL and posterior instability.To be sure,cruciate-re- taining prostheses have strong proponents in the literature.
Schai et al. [21] reported on 81 rheumatoid knee arthro- plasties using cruciate-retaining components followed for 10-13 years. The authors reported 13-year survivorship of 97%. Furthermore, they did not observe any late instabili- ty from PCL attenuation or rupture. Two knees were re- vised for reasons unrelated to the cruciate-retaining design
31
(failed metal-backed patella and severe synovitis necessi- tating synovectomy).Archibeck et al.[22] described 72 cru- ciate-retaining TKAs in 51 rheumatoid patients for a mean of 10.5 years.Nine of 72 knees required revision operations, but six of these revisions were performed for failure of a metal-backed patellar component. The rate of 10-year sur- vival was 93% with the end point being femoral or tibial re- vision.Two of the 72 knees had posterior instability.There- fore, the authors of both of these reports preferred cruci- ate-retaining TKA, even in the rheumatoid patient. The findings of these reports were contradicted by Laskin and O’Flynn [23],who reviewed the outcomes of 98 rheumatoid knees treated with cruciate-retaining prostheses at mini- mum 6-year follow-up. Fifty percent of 98 knees demon- strated posterior instability of 10 mm or more, compared with 1% of 80 rheumatoid knees treated with a posterior- stabilized prosthesis. Eleven patients in the cruciate-re- taining group required revision for instability,and the PCL was not identifiable in any of them during the revision surgery.These authors advocated posterior-stabilized TKA in rheumatoid knees. Basic science research demonstrates that the PCL in rheumatoid knees shows evidence of de- generation and collagen breakdown using light and elec- tron microscopy, respectively, and that the PCL is biome- chanically less elastic and less resistant to rupture than in normal knees [24-26]. We favor substitution of the PCL in all rheumatoid patients because of the consistently excel- lent results obtainable with this design and because of in- sufficient evidence that cruciate-retaining TKA provides significant advantages to the patient.
Surgical Technique
While some rheumatoid patients have ligamentous laxi- ty with no fixed deformity, rheumatoid knees often pre-
sent with severe soft-tissue contractures [1]. Additional- ly, rheumatoid bone is usually weakened secondary to disuse osteopenia and steroid intake. This section de- scribes the surgical technique for rheumatoid TKA with special attention to the difficult areas of exposure, angu- lar deformity, weak bone, and flexion contracture (
⊡ Fig.31-1
).
The Typical Rheumatoid Knee. The surgeon must be mind- ful of the compromised bone quality in the rheumatoid knee. The bone may be directly involved in the inflam- matory process by infiltration of the rheumatoid granu- lation tissue into subchondral bone, and prostaglandin release by nearby synovial tissue can lead to bone re- sorption and osteopenia in the rheumatoid knee [1]. The stiffness of proximal tibial cancellous bone in rheumatoid patients is approximately 675 N/mm, compared with 1287 N/mm for normal bone and 1116 N/mm for osteoarthrit- ic bone [15]. Therefore, excessive retraction or manipula- tion of the extremity during surgery can lead to inadver- tent fractures. Furthermore, rheumatoid knees more of- ten than not have an angular deformity. In a study of 99 rheumatoid knees [27], 38% were in valgus, 31% in varus, and 30% had no angular deformity. Forty-four percent of all knees studied had a flexion contracture over 10°.
Exposure. We use the standard anterior midline skin inci- sion followed by a medial parapatellar arthrotomy. A complete synovectomy is recommended to minimize the likelihood of a recurrent inflammatory synovitis after TKA.The fatty tissue between the anterior femur and syn- ovium should be preserved to avoid adhesions.The patel- la is everted, and the rest of the knee exposure is per- formed in the usual manner.
If exposure of the knee proves to be difficult,and prox- imal and distal extension of the arthrotomy does not pro-
⊡ Fig. 31-1a-d.Total knee arthroplasty in a 46-year-old rheumatoid patient. a, b AP and lateral radiographs of the knee preoperatively. c, d AP and lateral radiographs of the knee 3.5 years postoperatively
a b c d
times necessary. This allows easier eversion of the patella.
The quadriceps snip is repaired securely at the end of surgery and it does not typically result in postoperative quadriceps weakness [27]. A complete patellofemoral turndown is very rarely required; this involves extending the capsular incision laterally and distally, starting at the proximal end of the medial parapatellar arthrotomy. This technique provides unparalleled exposure but it com- monly results in quadriceps weakness and an extensor lag.
Medial Release. Some degree of medial dissection is nec- essary to allow enough exposure to perform the surgery in any knee. The extent of medial release is dependent on the pre-existing varus deformity, and as mentioned above, varus alignment is not uncommon in rheumatoid knees. The medial release begins with raising a periosteal flap on the proximal medial tibia and extending distally and posteriorly as necessary. This involves raising the deep portion of the medial collateral ligament, elevation of the semimembranosus insertion, and removal of me- dial osteophytes. It is sometimes necessary to elevate a medial soft-tissue flap more distally on the tibia to achieve adequate medial release. Despite these releases, the knee may, rarely, remain asymmetric with lateral lax- ity and medial tightness.In this case,we use a constrained condylar knee prosthesis.
Lateral Release. Valgus deformity is comparatively more common in the rheumatoid population and can present a challenge to achieving soft-tissue balance. The lateral structures, including the iliotibial band, lateral collateral ligament (LCL), popliteus tendon, and lateral joint cap- sule are contracted in a valgus knee with relative laxity of the medial structures. After femoral and tibial lateral os- teophytes have been resected, the tightest structure is usually the iliotibial band, and it should be released from its tibial insertion. The posterolateral capsule is subse- quently detached from the femur with a periosteal eleva- tor. If more release is necessary (such as when the valgus deformity is greater than 15°), the popliteus tendon, fol- lowed by the LCL,may be released from their femoral ori- gin. Rarely, the femoral origin of the lateral head of the gastrocnemius may require release by sharp dissection with the knee in flexion [27]. In the course of this dissec- tion, the inferolateral geniculate branch should be cau- terized to prevent excessive bleeding. The purpose of these releases is to match the relative laxity of the medial side of the knee and achieve a rectangular gap in flexion and extension. This may not always be possible in severe- ly deformed knees and may necessitate the use of a con- strained prosthesis.As described below,this becomes par- ticularly important in combined valgus and flexion con- tracture deformities.
cept the prosthesis. We use the measured resection tech- nique, removing an amount of bone equal to the dimen- sions of the prosthesis. The distal femur is cut in 6° of val- gus, based on an intramedullary guide rod. The anterior and posterior cuts are made in a few degrees of external rotation, as determined by Whiteside’s line and the transepicondylar axis. The tibia is cut with an ex- tramedullary guide, removing 8-10 mm of the proximal tibia.
Flexion Contracture. The flexion and extension gaps can now be assessed. The bone cuts provide access to the pos- terior aspect of the knee, and posterior osteophytes are removed with a curved osteotome.This occasionally is all that is required to correct a mild flexion contracture. If more correction is necessary, a periosteal elevator is used to release the capsule from the posterior femur [28] (
⊡ Fig.31-2
).
The attachments of the cruciate ligaments are re- leased completely from the intercondylar notch. To achieve even more correction, the origin of the gastroc- nemius muscle is dissected from the posterior distal fe- mur.If necessary,the release can be brought medially and laterally around the femur to the posterior aspects of the insertions of the collateral ligaments without sacrificing the integrity of the ligaments. At this point, the next step in achieving flexion-extension balance is resecting more distal femoral bone. This should be used as a last resort after the posterior structures are fully released, since re- secting more distal femur elevates the joint line and can
31
⊡ Fig. 31-2.Posterior capsular release for flexion contracture after os- teophyte removal. (Drawn by Lisa Khoury, MD; University of Pennsylvania Health System)
result in lax collateral ligaments, particularly in mid flex- ion.If these releases achieve good,stable extension but re- sult in looseness in flexion, the surgeon may consider up- sizing the femoral component and using augments pos- teriorly. This would add stability to the knee in flexion.
Doing so would be appropriate as long as the femoral component remains well sized in the mediolateral di- mension.
If the knee can achieve adequate extension only by sacrificing stability in flexion, then we have a low thresh- old to using a constrained device in this case. Of course, constraint has the disadvantage of transferring shear and rotational stresses to the bone-cement interface, which can adversely affect implant longevity. Therefore, we would resort to it only when the flexion laxity is severe and would lead to a functional disability, and after a full posterior release has been performed. We also favor the use of stems with a constrained device.
With loss or incompetence of the medial collateral ligament, the stresses normally attenuated by this struc- ture can lead to premature loosening. Stem use has been shown to be advantageous when higher levels of con- straint are necessary [29] (
⊡ Fig. 31-3).
It is important to guard against losing extension post- operatively in these knees. This may be accomplished by casting the knee in extension and by diligent physical therapy to maintain extension [1].
Implantation and Closure. When the surgeon is satisfied with the soft-tissue balancing and trial components, the knee is irrigated and the components are cemented. The
arthrotomy is repaired.We do not typically use a drain.A meticulous closure is done while minimizing trauma to the subcutaneous tissues.
Postoperative Rehabilitation
The reconstructed knee is examined at the end of the procedure to determine if there is any residual flexion contracture. If the knee easily comes to full extension, a soft dressing is applied and a continuous passive motion device is used in the recovery unit. But if the last few de- grees of extension can be achieved only with passive pressure applied anteriorly by the surgeon, the knee should be placed in a cylinder cast. We also utilize a cylinder cast when the soft tissue overlying the pros- thesis is thin and tenuous; this enhances the potential for uneventful wound healing. The cast can be removed at 3-7 days postoperatively, and then physical therapy for gentle range of motion is started. The physical therapist should pay particular attention to maintaining full ex- tension, particularly in knees that were contracted chronically prior to surgery. The surgeon must be vigi- lant regarding signs or symptoms of infection. There is a significantly increased risk of deep prosthetic infec- tion in rheumatoid patients compared with osteo- arthritic patients [30]. Prolonged drainage from a sero- ma or hematoma should be treated aggressively with débridement to prevent bacterial seeding of the pros- thetic joint.
Results
Overall, the results of knee arthroplasty in the rheuma- toid population have been excellent.Robertsson et al.[14]
desribed the results of 4143 primary tricompartmental knee replacements from the Swedish Knee Arthroplasty Register. They reported a cumulative revision rate of 5%
at 10 years.A survey of patient satisfaction from the same group of patients [17] demonstrated higher satisfaction rates among rheumatoid patients than among patients with osteoarthritis. These findings may be secondary to the lower demand placed on the prosthesis by sedentary rheumatoid patients.Therefore,despite the multiple chal- lenges inherent in knee arthroplasty in the rheumatoid patient, excellent pain relief and durable restoration of function can be expected from the procedure.
References
1. Chmell MJ, Scott RD (1999) Total knee arthroplasty in patients with rheumatoid arthritis. Clin Orthop 366:54-60
2. MacKenzie CR, Sharrock NE (1998) Perioperative medical considerations in patients with rheumatoid arthritis. Rheum Dis Clin North Am 24:1-17
⊡ Fig. 31-3.Bilateral total knee arthroplasty using constrained prosthe- ses and stems in a 68-year-old rheumatoid patient. Instability with stan- dard components was found to be excessive, and constrained femoral components with stems were used
4. Friedman RJ, Schiff CF, Bromberg JS (1995) Use of supplemental steroids in patients having orthopaedic operations. J Bone Joint Surg [Am]
77:1801-6
5. Grennan DM, Gray J, Loudon J, Fear S (2001) Methotrexate and early post- operative complications in patients with rheumatoid arthritis undergo- ing elective orthopaedic surgery. Ann Rheum Dis 60:214-217 6. Bridges SL, Lopez-Mendez A, Han KH, Alarcon GS (1991) Should
methotrexate be discontinued before elective orthopedic surgery in pa- tients with rheumatoid arthritis? J Rheum 18:984-988
7. Baghai M, Osmond DR, Wolk D, Wold LE, Haidukewych GJ, Matteson E (2001) Fatal sepsis in a patient with rheumatoid arthritis treated with etanercept. Mayo Clin Proc 76:653-656
8. Dunbar RP, Alexiades MM (1998): Decision making in rheumatoid arthri- tis. Rheum Dis Clin North Am 24:35-54
9. Pellicci PM, Ranawat CS, Tsairis P, et al (1981) The natural history of rheumatoid arthritis of the cervical spine. J Bone Joint Surg [Am] 63:342- 350
10. Matti MV, Sharrock NE (1998) Anesthesia on the rheumatoid patient.
Rheum Dis Clin North Am 24:19-33
11. Schroder HM, Aaen K, Hansen EB, Nielsen PT, Rechnagel K (1996) Ce- mentless total knee arthroplasty in rheumatoid arthritis. A report on 51 AGC knees followed for 54 months. J Arthroplasty 11:18-23
12. Nielsen PT, Hansen EB, Rechnagel K (1992) Cementless total knee arthro- plasty in unselected cases of osteoarthritis and rheumatoid arthritis. A 3- year follow-up study of 103 cases. J Arthroplasty 7:137-143
13. Rosenqvist R, Bylander B, Knutson K, Rydholm U, Rooser B, Egund N, Lid- gren L (1986) Loosening of the porous coating of bicompartmental pros- theses in patients with rheumatoid arthritis. J Bone Joint Surg [Am]
68:538-542
14. Robertsson O, Knutson K, Lewold S, Goodman S, Lidgren L (1997) Knee arthroplasty in rheumatoid arthritis; a report from the Swedish Knee Arthroplasty Register on 4381 primary operations 1985-1995. Acta Or- thop Scand 68:545-553
15. Bogoch ER and Moran EL (1999) Bone abnormalities in the surgical treat- ment of patients with rheumatoid arthritis. Clin Orthop 366:8-21 16. Burnett RS and Bourne RB (2003) Indications for patellar resurfacing in to-
tal knee arthroplasty. J Bone Joint Surg [Am] 85:728-745
17. Robertsson O, Dunbar M, Pehrsson T, Knutson K, Lidgren L (2000) Patient satisfaction after knee arthroplasty: a report on 27,372 knees operated on between 1981 and 1995 in Sweden. Acta Orthop Scand 71:262-267
19. Kajino A, Yoshino S, Kameyama S, Kohda M, Nagashima S (1997) Com- parison of the results of bilateral total knee arthroplasty with and with- out patellar replacement for rheumatoid arthritis. A follow-up note. J Bone Joint Surg [Am] 79:570-574
20. Boyd AD Jr, Ewald FC, Thomas WH, Poss R, Sledge CB (1993) Long-term complications after total knee arthroplasty with or without resurfacing of the patella. J Bone Joint Surg [Am] 75:674-681
21. Schai PA, Scott RA, Thornhill TS (1999) Total knee arthroplasty with pos- terior cruciate retention in patients with rheumatoid arthritis. Clin Orthop 367:96-106
22. Archibeck MJ, Berger RA, Barden RM, Jacobs JJ, Sheinkop MB, Rosenberg AG, Galante JO (2001) Posterior cruciate ligament-retaining total knee arthroplasty in patients with rheumatoid arthritis. J Bone Joint Surg [Am]
83:1231-1236
23. Laskin RS and O’Flynn HM (1997) Total knee replacement with posterior cruciate ligament retention in rheumatoid arthritis: problems and com- plications. Clin Orthop 345:24-28
24. Alexiades M, Scuderi G, Vigorita V, et al (1983) A histological study of the posterior cruciate ligament in the arthritic knee. Am J Knee Surg 64A:1328-1333
25. Neurath MF (1993) Detection of Luse bodies, spiraled collagen, dysplas- tic collagen, and intracellular collagen in rheumatoid connective tissues:
an electron microscopic study. Ann Rheum Dis 52:278-284
26. Hagena FW, Hoffman GO, Mittlemeier T, et al (1999) The cruciate liga- ments in knee replacement. Int Orthop 249:9-12
27. Aglietti P, Buzzi R (2002) Correction of combined deformity. In: Scuderi GR, Tria A (eds) Surgical techniques in total knee arthroplasty, 1st edn.
Springer-Verlag, Berlin Heidelberg New York
28. Lotke PA, Simon RG (2002) Flexion contracture in total knee arthroplasty.
In: Scuderi GR, Tria A (eds) Surgical techniques in total knee arthroplasty, 1st edn. Springer-Verlag, Berlin Heidelberg New York
29. Garino JP, Lotke PA (1999) Fixation techniques in revision total knee surgery: stem designs, rationale, and fixation. In: Lotke PA, Garino JP (eds) Revision total knee arthroplasty, 1st edn. Lippincott-Raven, Philadelphia 30. Sculco TP (1998) The knee joint in rheumatoid arthritis. Rheum Dis Clin
North Am 24:143-156
