Arthrodesis of the Knee

Conway, Janet D. MD; Mont, Michael A. MD; Bezwada, Hari P. MD

Journal of Bone & Joint Surgery - American Volume:
Current Concepts Review
Abstract

The most common indication for arthrodesis of the knee is an infection at the site of a total knee arthroplasty.

Deficiencies in bone stock and poor bone apposition adversely affect the success of a knee arthrodesis.

Arthrodesis of the knee can provide a stable, painless extremity for high-functioning patients who are able to walk.

Patient function after arthrodesis of the knee is superior to that after above-the-knee amputation.

Conversion of a solid knee fusion to a total knee arthroplasty has a substantial complication rate.

Author Information

1 Rubin Institute for Advanced Orthopaedics, 2401 West Belvedere Avenue, Baltimore, MD 21215. E-mail address for J.D. Conway: jconway@lifebridgehealth.org. E-mail address for M.A. Mont: mmont@lifebridgehealth.org. E-mail address for H.P. Bezwada: hbezwada@yahoo.com

Article Outline

Arthrodesis of the knee has been performed since the early 1900s to treat the pain and instability associated with advanced osteoarthritis, posttraumatic arthritis, Charcot arthropathy, infectious arthritis, poliomyelitis, and reconstruction following tumor resection1-10. Currently, the indications for knee arthrodesis have been narrowed for multiple reasons. Poliomyelitis has been virtually eliminated by vaccination. Improved antimicrobial therapy for patients with tuberculosis and syphilis has almost eliminated knee involvement by these diseases. In addition, the success of modern knee arthroplasty has substantially decreased the number of patients who formerly would have been candidates for a knee arthrodesis. Presently, the most common indication for a knee arthrodesis is pain and instability in an unreconstructable knee following an infection at the site of a knee arthroplasty (Figs. 1-A and 1-B)11-17. In addition, knees with substantial metaphyseal bone loss, inadequate ligamentous restraints, multiple failed revisions, inadequate soft-tissue coverage with loss of the extensor mechanism, and infection with virulent organisms should be considered for arthrodesis. This review will focus on these modern indications for knee arthrodesis; the operative techniques, results, and complications of the procedure; and alternatives to knee arthrodesis. The controversies surrounding conversion from a solid fusion to a total knee replacement will also be reviewed.

Back to Top | Article Outline

Indications for Knee Arthrodesis

Knee arthrodesis is indicated for pain and instability of the knee in a young, vigorously active patient who has had a reconstruction following tumor resection, has post-poliomyelitis syndrome, or has Charcot gonarthrosis. Historically, arthrodesis was indicated for young patients with posttraumatic arthritis. Total knee arthroplasty formerly was complicated by early failures and high rates of revision16, but recent reports have demonstrated success rates ranging from 86% to 96% (in series ranging from thirty to fifty-two patients, with durations of follow-up of approximately five to ten years) in patients less than fifty years of age18-20. Thus, the indications for total knee arthroplasty have expanded, and this procedure is now considered to be more appropriate for younger patients (less than fifty years of age).

Current indications for arthrodesis include tumors that require an extensive resection in which the joint surface cannot be preserved and there is an adequate margin of soft tissue between the resected tissue and the posterior neurovascular structures. Tumors that do not meet these criteria are best managed with amputation. Patients with a proximal tibial or distal femoral tumor are candidates for an arthrodesis21. Enneking and Shirley21 performed knee arthrodesis to treat recurrent tumors about the knee in twenty patients with a mean age of twenty-five years (range, ten to fifty-two years). Fifteen of sixteen patients resumed their previous occupation. Currently, many patients with a tumor about the knee have reconstruction with a modular tumor prosthesis22.

Knee arthrodesis is also indicated for patients with a failed total knee replacement that is unsalvageable, most commonly secondary to infection. It can be difficult to determine unsalvageable situations in which a knee arthrodesis would provide a better and more reliable outcome than would multiple revisions of a total knee arthroplasty. Each case must be carefully evaluated to determine which treatment plan will lead to the best outcome. A knee arthrodesis can be considered when reimplantation following infection at the site of a primary total knee arthroplasty has a high chance of failure. Husted and Toftgaard Jensen23 reported on twenty-four patients with an infection at the site of a primary total knee arthroplasty. Seventeen patients underwent a two-stage revision arthroplasty, and the other seven underwent an arthrodesis. Fifteen of the seventeen patients treated with the two-stage revision arthroplasty were considered to have successful eradication of the infection; however, eight of these patients continued to have pain on a daily basis. The remaining two patients had a failure of the two-stage reconstruction; one required an arthrodesis and the other, an above-the-knee amputation. This rate of failed total knee revisions is similar to that in other studies, in which the prevalence of failure, requiring a knee arthrodesis, of two-stage exchange arthroplasties performed to treat infection following a primary total knee replacement ranged from 6% to 13%24,25.

The rate of failure following rerevision of a revision total knee replacement that was complicated by infection is considerably higher11. In a study by Hanssen et al.11 of twenty-four patients who had an infection at the site of a revision total knee arthroplasty, twelve patients eventually had a knee arthrodesis. Ten of the twelve had a successful arthrodesis with a stable, pain-free knee.

Knutson et al.26,27 reported the results of twenty arthrodeses in patients with a failed total knee arthroplasty and advocated a revision arthroplasty when a primary total knee replacement fails as a result of a correctable technical error, such as component malposition or improper ligament balancing. They advocated an arthrodesis when there is infection associated with soft-tissue problems and virulent and/or resistant organisms27.

Massive bone loss may substantially reduce the success rate of arthrodesis26,28-32. Wilde and Stearns28 performed twelve knee arthrodeses and noted a decreased fusion rate with greater degrees of bone loss. They suggested that failure of a two-stage implantation may be an indication for arthrodesis before multiple failures lead to more bone loss. In a series of nine patients in whom an arthrodesis had been performed with external fixation, Behr et al.29 found a decrease in the fusion rate when the failed arthroplasty had involved use of a constrained prosthesis because of greater bone resection after removal of such implants. All four knees in which the arthrodesis was performed following removal of a semiconstrained prosthesis healed compared with four of five knees treated following removal of a fully constrained implant. Hak et al.33 reported the results of thirty-six knee arthrodeses and found a decrease in the fusion rate with an increase in the number of previous procedures.

It may be difficult for the surgeon and patient to decide on a knee arthrodesis because preserving motion with a revision total knee arthroplasty is an attractive option. Although the surgeon may consider an arthrodesis to be a poor outcome, a limb with a fusion is more efficient and functional than is one with an above-the-knee amputation34. Arthrodesis following a knee arthroplasty that was complicated by infection is associated with a low risk of reinfection16,26,35,36 and provides a stable, pain-free limb. In the authors' opinion, the objective Knee Society score following a successful knee arthrodesis is approximately 70 points compared with a score of <50 points after an amputation or a poor revision.

Although we are not aware of any prospective, randomized studies comparing knee arthrodesis with above-the-knee amputation, a review of studies in which one technique or the other was used suggests that an arthrodesis is preferable to an amputation in any population. When compared with an above-the-knee amputation, a knee arthrodesis offers a stable, durable, and dependable limb without pain. Isiklar et al.37 reported on nine above-the-knee amputations performed in eight patients because of complications after knee arthroplasty. The factors that led to the amputations were multiple revisions with chronic infection, severe bone loss, and intractable pain. At the time of the final follow-up, only two of the eight patients were able to walk. Pring et al38 reported that, of twenty-three patients who had an above-the-knee amputation following a knee arthroplasty, only seven (30%) were able to walk. In addition, the energy expenditure with walking is greater after amputation than it is after arthrodesis. Patients with a knee arthrodesis used 0.16 mL/kg/min of oxygen compared with 0.20 mL/kg/min used by patients with an above-the-knee amputation34,39.

Contraindications to knee arthrodesis include a contralateral knee amputation, contralateral knee or hip arthrodesis, and degenerative changes in the ipsilateral hip and ankle16. Compensatory mechanisms for walking after a successful knee arthrodesis include increased pelvic tilt, increased ipsilateral hip abduction, and increased ipsilateral ankle dorsiflexion39. Ipsilateral hip and ankle arthritis preclude effective compensatory efforts during walking. Patients with degenerative arthritis of the spine are also poor candidates for knee arthrodesis because the compensatory pelvic tilt required during walking increases forces across the lumbar spine16. In addition, knee arthrodesis is contraindicated for patients with a contralateral amputation because of the increase in energy required for walking. The energy required to walk with a knee fusion alone is 25% to 30% higher than that required for normal walking39, and the energy required to walk with an above-the-knee amputation is 25% greater than that required with a knee fusion. Thus, the energy required with a combination of amputation and contralateral fusion would make walking extremely difficult and most likely dangerous to cardiovascular function in many patients. Other relative contraindications to knee arthrodesis include an inability to walk and medical comorbidities that preclude an extensive surgical procedure.

Back to Top | Article Outline

Preoperative Considerations

Many patients who are considered for a knee arthrodesis have underlying medical comorbidities that should be addressed in order to optimize the patient's condition. Host factors such as smoking, diabetes, and obesity increase the risk of wound problems, nonunion, and infection and should be managed preoperatively with a smoking cessation program, optimization of blood sugar control, and a weight-loss program. Ideally, patients who have had a tumor resection should have completed the chemotherapy and radiation treatment to minimize wound-healing problems.

The operative duration and blood loss associated with knee arthrodeses are often considerable13,40. Donley et al.40 reported on a series of twenty knee arthrodeses performed with a long intramedullary nail for a variety of indications. The mean estimated blood loss was 1574 mL (range, 520 to 3200 mL), and a mean of 3.9 units (range, zero to seven units) of packed red blood cells was transfused. The mean operative time was 4.1 hours (range, 2.5 to 8.3 hours). Arroyo et al.13 reported on a series of knee arthrodeses performed with use of a short nail that couples at the knee (Neff nail; Zimmer, Warsaw, Indiana). The mean estimated blood loss was 2199 mL (range, 325 to 5300 mL) for patients who had the arthrodesis after tumor resection and 865 mL (range, 325 to 1300 mL) for those who had it after a total knee arthroplasty complicated by infection.

All previous incisions must be critically inspected and taken into consideration to prevent wound necrosis. Many patients with an infection at the site of a knee arthroplasty have had previous wound problems. Rand and Bryan41 evaluated the outcomes of failed knee arthrodeses and found that eight of twenty-five patients had soft-tissue complications, including hematoma, drainage, and flap necrosis. If wound difficulties are anticipated, soft-tissue coverage with the help of soft-tissue expanders as well as specific flap techniques may be required41,42. We recommend preoperative vascular evaluation in the presence of diminished or absent peripheral pulses. In our experience, intraoperative Doppler assessment of the pedal pulses has been helpful in selected patients. Furthermore, during acute bone compression, vascular insufficiency can develop as a result of the extensive scarring that occurs in the posterior aspect of the knee following a total knee arthroplasty21.

All patients have shortening of the affected extremity following knee arthrodesis. The limb-length discrepancy has ranged from 2.5 to 6.4 cm29,43-47 and can be estimated from a preoperative teleoroentgenogram (a long, 51-in [129.5-cm] radiograph that shows the limbs from the hips to the ankles). This radiograph is also helpful for evaluating malalignment and for sizing of intramedullary devices. If the limb-length discrepancy is large, a concomitant lengthening procedure can be performed with use of another osteotomy combined with an external fixator48-51 (Figs. 2-A and 2-B). Hessmann et al.50 performed knee arthrodeses with simultaneous lengthening by 5 cm with use of an Ilizarov external fixator. We perform a concomitant lengthening when the predicted discrepancy is >5 cm because shoe-lifts become cumbersome with discrepancies of this magnitude. In our experience with simultaneous distraction osteogenesis and knee arthrodesis, the consolidation of the regenerated bone has not increased the time in the fixator51.

Preoperative planning is essential before reconstruction of the large bone and soft-tissue defects that follow tumor resection. Reconstructive techniques that preserve length may include the use of vascularized fibular grafts, allografts, and distal femoral and proximal tibial sliding grafts21,52,53.

Many authors have advocated the eradication of infection prior to arthrodesis because it results in a higher fusion rate and more options for fixation54-60. A single-stage arthrodesis can be performed when a surgeon thinks that an adequate débridement has been done and the arthrodesis will be performed with an external fixator. However, other techniques for arthrodesis (those involving plates, nails, and so on) are probably best done with a two-stage approach. The initial operative treatment consists of surgical débridement, removal of components, and insertion of an antibiotic-impregnated cement spacer. The use of an antibiotic-impregnated spacer appears to lead to less bone loss at the time of knee reconstruction28. We prefer to use a combination of 3.6 g of tobramycin and 1 g of vancomycin per pack of Palacos bone cement (Biomet, Warsaw, Indiana). Several authors have reported good results with the careful use of intramedullary nails after staged eradication of infection28,56. Waldman et al.56 reported on a series of twenty-one patients in whom a short intramedullary nail was used to perform an arthrodesis following failure of a knee arthroplasty due to infection. Sixteen knees had insertion of a cement spacer and no infection prior to the arthrodesis. Twenty of the twenty-one patients had a successful arthrodesis without recurrence of infection.

In order to avoid contamination of the intramedullary canal, Wilde and Stearns28 advocated preoperative aspiration of the knee as well as normalization of laboratory values such as the erythrocyte sedimentation rate and C-reactive protein level prior to arthrodesis with an intramedullary nail. Despite these precautions, five of their twelve patients had a positive culture at the time of the arthrodesis and an infection developed after union in two of those patients, who had to undergo nail removal in order to eradicate the infection. Fortunately, to our knowledge, there have been no reports of osteomyelitis of the tibia or femur after the use of long intramedullary nails for knee arthrodesis following infection61. Lidgren and Törholm62 found that local eradication of infection in the femur and subsequent diaphyseal reaming were not associated with postoperative infection in a study of eighteen patients. In addition, several authors demonstrated that one-stage arthrodesis with use of an intramedullary nail following failure of a total knee arthroplasty due to infection was safe and effective and not associated with recurrent infection in the absence of gross purulence and gram-positive organisms47,59,63. Puranen et al.47 reported a solidly healed fusion, clinically and radiographically, without recurrent infection after eight of eight knee arthrodeses that had been performed with a long intramedullary nail following infection at the site of a knee arthroplasty. They recommended that, in patients with a failed arthroplasty due to infection, the intramedullary nail be used in a staged fashion—i.e., following an eight to ten-week interval with a cement spacer and normalization of serologic markers. We believe that patients should complete a six-week course of antibiotic therapy followed by a period of two to four weeks without antibiotic treatment. Surgeons may also consider evaluating frozen sections of tissue from the knee joint to assess the number of polymorphonuclear leukocytes per high-powered field64,65. We recommend that patients with virulent or polymicrobial infection and extensive bone loss be treated with circular or biplanar external fixation.

Back to Top | Article Outline

General Intraoperative Considerations

We abide by several general operative principles for a knee arthrodesis regardless of the indication or technique. For patients treated following a tumor resection or infection, two separate intraoperative instrument sets are required—one “dirty” one and one “clean” one. The “dirty” set is used for the resection of the tumor or area of infection, and the “clean” set is used for the reconstruction. This potentially minimizes contamination of the unaffected tissues with residual bacteria or tumor cells. All gowns, gloves, and drapes should be changed between the stages of the procedure, also to minimize contamination. The entire lower limb should be prepared, from the anterior superior iliac spine to the toes, to determine rotational alignment. Sterile tourniquets can be a useful adjunct to limit blood loss. Both femoral and tibial bone ends must be adequately dissected free of the posterior capsule to allow for good bone apposition without soft-tissue interposition. In addition, acute shortening can result in kinking of arteries and loss of pulses with vascular embarrassment.

There is no consensus regarding the optimal alignment of a knee fusion in the sagittal or coronal plane. In order to prevent further limb-shortening, the majority of knee arthrodeses are placed in 0° of extension. Some authors have advocated 5° to 10° of flexion1,61. Fusion in 10° to 15° of flexion allows a better sitting position and improves gait; however, it also increases limb-length discrepancy. In a study of the gait of forty-one patients who had had a knee arthrodesis, Siller and Hadjipavlou61 described two types of gait when the knee was fused in 0° of extension: gait with an excessive pelvic rise and gait en fauchant (“the gait while mowing fields”). The gait with an excessive pelvic rise causes the patient to “bounce up and down” in an exaggerated fashion. With the gait en fauchant, the patient compensates for the stiff knee by abducting the limb at the hip and swinging the limb out to the side. Both compensatory gait patterns are less efficient, resulting in a 25% increase in energy expenditure when compared with that of normal walking34. Siller and Hadjipavlou found gait to be more efficient with the fusion in 15° of flexion because of the change in the push-off direction of the gastrocnemius muscle. At 0° of knee flexion, the gastrocnemius has a vertical thrust, but there is more of a horizontal or propulsive thrust at 15° of flexion. However, 15° of flexion is associated with approximately 2 cm of shortening, and many surgeons are reluctant to increase the limb-length discrepancy in this way. Conway et al.39 studied the optimal position of an artificially fused knee by placing a long leg cylinder cast on the limbs of ten healthy volunteers with the knee fixed in 20° or 0° of flexion. Four of the ten volunteers selected a faster walking speed with the knee flexed 20° than they did with it in 0° of flexion.

Five to seven degrees of valgus has been proposed as the ideal normal frontal alignment47. This is easier to obtain with an external fixator than it is with a fixed intramedullary nail. Use of a straight nail often places the knee in 2° to 5° of varus, which increases the varus moment on the hip. This theoretically increases the risk of arthritis in the ipsilateral hip, according to Puranen et al.47. However, ipsilateral hip disease did not develop in any of their thirty-three patients treated with knee arthrodesis with an intramedullary nail and followed for a minimum of two years.

Maximum bone contact is extremely important to optimize the outcome of knee arthrodesis31,66,67. Charnley68 believed that the success of 99% (169) of the 171 arthrodeses in his patients was due to two perfectly coapted surfaces of cancellous bone with intact circulation. In a series in which seven of eleven knee arthrodeses were successful, Hagemann et al.67 determined that the most important factor for a successful fusion was good bone contact. Woods et al.66 also advocated good bone contact, after all of the sixteen arthrodeses performed following a failed knee arthroplasty in their series were successful. They stated that “no compression device is secure enough to overcome poor carpentry.”

Bone contact diminishes with bone loss, especially following removal of a hinged knee prosthesis and after tumor resection26,31,39,65. Rand et al.31 reported on twenty-eight patients treated with a knee arthrodesis following a failed total knee arthroplasty. Their series included seven patients with a hinged prosthesis, and four of the seven subsequently had a nonunion. Brodersen et al.43 reported that, of nine arthrodeses performed after removal of a failed hinged prosthesis, only five resulted in a successful knee fusion. The authors thought that these nonunions were secondary to the large loss of metaphyseal bone stock.

Back to Top | Article Outline

Techniques for Arthrodesis

Intramedullary Nailing

The intramedullary nail that has been used most commonly for knee arthrodesis, as reported in the literature, is the Küntscher nail (Biomet, Warsaw, Indiana)26,40,55,63, which is no longer available; however, similar long nails are currently in use. When intramedullary nailing is performed, the smaller-diameter bone of the tibia is often reamed first in order to prevent overreaming of the femur. Subsequently, the larger-diameter femoral canal is reamed 1 mL larger than the tibial canal. If the femur or tibia is overreamed, the nail will lack a secure fit. Usually, a nail diameter that is 0.5 mm smaller than the diameter of the last reamer is chosen. The nail length should span from the greater trochanter to 2 cm proximal to the ankle. Cancellous bone fragments from the reaming can be placed around the knee, and the nail is anchored at the greater trochanter with either an 18-gauge wire or a locking screw. The locking screw prevents proximal migration of the nail, which can lead to gluteal pain40 (Figs. 3-A and 3-B).

The long nail has the advantage of providing excellent stability across the long flexion-extension lever arm of the knee. Incavo et al.69 used a custom long intramedullary nail with a proximal locking screw to obtain an average of 3.1° of valgus (range, 1° to 5° of valgus) as well as a 5° flexion angulation of the nail. They reported a 100% fusion rate in twenty-two knees that had been treated after a failed knee arthroplasty, although four patients needed additional procedures to obtain union.

Another option is the use of a short nail such as the Neff nail (Zimmer), the Wichita nail (Stryker Howmedica Osteonics, Allendale, New Jersey), or the Huckstep nail (Downs Surgical, Sheffield, United Kingdom). Short nails have a coupling device at the knee that can accommodate different-sized femoral and tibial components. The nail is inserted 5 to 6 cm beyond the isthmus of the femur and tibia, and each segment is fixed with locking screws13,14. This nail is an excellent option for patients with an ipsilateral hip replacement, for whom a long nail would be unsuitable. One disadvantage is that the coupling device is large, so that the tibial canal must be reamed to a diameter of >20 mm, and a segment must be removed from the anterior aspect of the femur and tibia to allow access to the coupling mechanism. This exposure is less troublesome when the arthrodesis is performed after the failure of a total knee arthroplasty. It is also difficult to remove the nail once it has been inserted. This may require removal of anterior cortical panels from both the femur and the tibia. Christie et al.70 reported a 100% rate of both clinical and radiographic fusion after using the Wichita nail to treat fifty-three knees with a failed total knee arthroplasty.

Back to Top | Article Outline
External Fixation

External fixation was popularized as a method of arthrodesis by Charnley71, who used two transfemoral pins, two transtibial pins, and a Thomas splint to maintain 60 to 80 lb (27.2 to 36.3 kg) of compression at the arthrodesis site. His indications for the procedure were advanced osteoarthritis and tuberculous arthritis. His patients had rapid healing, which he attributed to compression and excellent bone contact. The main advantage of external fixators is their ability to exert compression at the arthrodesis site. This approach also allows better wound management. Since Charnley's report, many types of fixators have been described. Any pins or wires should avoid neurovascular structures and should be placed in healthy bone with bicortical purchase. The duration of fixation ranges from four weeks71 to several (nine to eleven) months26,72,73. We use hydroxyapatite-coated pins that promote bone ingrowth, which may decrease the risk of pin track infection and loosening74.

Knutson et al.75 examined the stability of external fixators used for arthrodeses performed after failures of knee arthroplasties due to infection and found that a modified Hoffmann-Vidal frame (Jaquet Orthopedie S.A., Geneva, Switzerland) and the Ace-Fischer apparatus (Ace Orthopedic, Los Angeles, California) had the greatest stiffness. Both devices consist of sagittal pins connected to an anterior fixator. As a biplanar configuration increases stability, it provides better knee fusion rates than does single-plane fixation67,72. A single-plane external fixator such as the Charnley device has minimum stability in the flexion-extension plane. In a series of thirty-six knee arthrodeses reviewed by Hak et al.33, ten of nineteen performed with a uniplanar external fixator compared with eleven of seventeen performed with a biplanar external fixator were followed by healing.

There have also been several studies of circular external fixators such as the Ilizarov device, with the reported fusion rates ranging from 68% to 100%44-46,76,77. Advantages of this type of fixation are the ability to use all-wire fixation, which provides the greatest stability in patients with poor bone quality, and the fact that it allows immediate weight-bearing. However, knee arthrodesis with the Ilizarov device can be technically challenging and may not be tolerated well by obese patients. Another disadvantage is that the proximal full ring at the distal part of the femur needs to be very large to accommodate the abundant soft tissues and does not allow the patient to walk with his or her lower limbs together. We use the Orthofix Limb Reconstruction System (Orthofix, Verona, Italy) as a biplanar fixator. Sixty-five and eighty-centimeter rails are utilized to allow fixation from the lesser trochanter to the distal tibial metaphysis in both the frontal and the sagittal plane for maximum stability across the long lever arm of the knee (Figs. 4-A and 4-B). This construct also allows a concomitant lengthening procedure if the limb-length discrepancy is particularly large (>5 cm). In three cases in which concomitant limb lengthening was performed at our center, the time for healing of the knee fusion was longer than that for consolidation of the regenerated bone51.

The disadvantages of all external fixation devices are the possibility of pin track infection, fusion rates that are lower than those following intramedullary fixation, and difficulty in assessing the adequacy of healing and thus the timing of fixator removal. Cunningham et al.78 performed a mechanical assessment of the bending stiffness of the knee fusion following arthrodesis with a unilateral external fixator in ten patients. They advocated a bending stiffness of 10 N-m/degree of movement in the construct prior to removal of the fixator. However, measuring bending stiffness is not always feasible. As an alternative, the surgeon can usually examine the patient in the office with the fixator disconnected. Under fluoroscopy, the surgeon evaluates pain and motion at the arthrodesis site with manual stress. This technique has been successful in most cases; however, it may not adequately reproduce the forces of full weight-bearing.

Back to Top | Article Outline
Plates

There have been few reports on the use of plates for knee arthrodesis79. Lucas and Murray80 described knee arthrodesis with use of one medial and one anterior plate in eighteen patients. Seventeen of the eighteen fusions healed, but there was difficulty with wound closure over the anterior plate. No wound problems occurred, but five patients required plate removal because of painful, prominent hardware. Nichols et al.81 reported on a dual-plate technique in which conforming, broad, staggered twelve-hole AO dynamic compression plates are placed on the medial and lateral sides of the knee (Fig. 5). All eleven patients in their series were treated postoperatively with a cast and had healing at a mean of 5.6 months (range, three to ten months). Munzinger et al.82 reported on the use of one large conforming lateral plate in a series of patients with a failed knee arthroplasty due to infection, and the fusion rate was 80% (twenty-seven of thirty-four knees) at six months.

Back to Top | Article Outline
Combination of Nails and Plates or Nails and External Fixation

We are aware of two studies on combined techniques for difficult indications such as a failed previous arthrodesis or Charcot arthropathy. Fahmy et al.9 combined Charnley's compression technique with an intramedullary Küntscher nail in thirteen patients who had either a failed previous arthrodesis or Charcot arthropathy, and all knees healed at a mean of six months (range, two to eleven months). Stiehl and Hanel54 reported on a series of eight patients treated with a long intramedullary rod and a medial compression plate. Their indications for the procedure included a failed arthrodesis after a knee arthroplasty, chronic osteomyelitis, and Charcot arthropathy. Union occurred within three to eight months in all patients.

Back to Top | Article Outline
Vascularized Fibular Graft

Vascularized fibular grafts have been used in conjunction with intramedullary nails for arthrodesis in patients with massive bone loss following tumor resection or total knee arthroplasty complicated by infection (Figs. 6-A and 6-B)52,55,83. The grafts can be used as a supplement or to bridge large defects, as reported by Usui et al.52. There are two types of fibular grafts: a free vascularized graft and a pedicled vascularized graft. Pedicled vascularized fibular grafts are rotated 180° on the peroneal artery. Usui et al. reported union in sixteen of seventeen patients who had been treated with a fibular graft for a large segmental defect (average, 14 cm). Rasmussen et al.83 used these grafts to treat segmental bone loss of >6 cm associated with tumor resection or infection after knee arthroplasty. Twelve of thirteen patients had a successful arthrodesis.

Back to Top | Article Outline
Supplemental Techniques

Other techniques for supplementing the arthrodesis include iliac crest bone-grafting and the use of pulsed electromagnetic fields. Several authors have recommended iliac crest bone-grafting when no signs of healing are present at four months after the index procedure26,29,31,82. Bigliani et al.84 reported the results of twenty arthrodeses supplemented with pulsed electromagnetic fields in twenty patients who had demonstrated delayed healing or no healing. Fusion was obtained in seventeen patients, at a mean of 5.8 months (range, three to twelve months) after the initiation of therapy.

Back to Top | Article Outline

Results According to Diagnosis

Early studies by Charnley1,71 and others6,85,86 demonstrated a fusion rate of 95% to 98% in patients with osteoarthritis, rheumatoid arthritis, or posttraumatic arthritis. However, the rates of fusion in patients with Charcot arthropathy or with infection following total knee arthroplasty usually have been lower (range, 33% to 100%16,26,33,36,43,45,56,67,87). Rates of successful fusion following tumor resection have approached 100%17,21,52,53,88. The following sections describe outcomes according to diagnosis.

Back to Top | Article Outline
Tumor

Knee arthrodesis after tumor resection is most commonly performed with either a short or a long intramedullary rod, and the fusion rates have been excellent, ranging from 82% to 100%13,17,21,89. Enneking and Shirley21 demonstrated that, in patients with a malignant tumor, limb preservation with a knee arthrodesis has the same results, with respect to tumor control, as does an amputation. They found a 100% fusion rate in twenty patients treated with a nail and a cortical autograft to bridge the osseous defect at the knee. Weiner et al.53 reported that thirty-two (82%) of thirty-nine patients had a healed fusion at a mean of 4.5 years (range, one to eleven years) following an arthrodesis performed with an intramedullary nail through an allograft. All patients were generally satisfied with the function of the lower extremity, with a mean functional score89 of 23 points (maximum score, 30 points). Other authors have reported high fusion rates after tumor resection13,17,52,88,90.

Back to Top | Article Outline
Failed Total Knee Arthroplasty

The fusion rates in studies of biplane or circular external fixation for arthrodesis following failed knee arthroplasty have ranged from 66% to 93% and are slightly lower than the rates in studies of intramedullary nailing, which have ranged from 67% to 100%26,28,45,66. In a study of arthrodeses performed with an Ace-Fischer biplanar external fixator (Ace Orthopedic, Hawthorne, California), Rand et al.31 reported an initial fusion in twenty of twenty-eight patients. Donley et al.40 reported an 85% fusion rate (seventeen of twenty patients) following use of an intramedullary nail. Damron and McBeath16 reviewed the recent literature comparing knee arthrodesis techniques in patients with a failed knee arthroplasty. Five reports, on a total of fifty-six patients, were included in the review. Intramedullary nailing led to successful fusion in thirty-four (94%) of thirty-six patients, and external fixation with use of a variety of biplanar fixators achieved a successful fusion in eight of thirteen patients. Despite the small number of patients in each study, Damron and McBeath concluded that intramedullary nailing leads to a higher fusion rate in patients with a failed knee arthroplasty.

Fusion rates following hinged knee arthroplasties generally have been lower than such rates following nonconstrained arthroplasties (range, 55% to 75% and 50% to 95%, respectively)26,30,43,67,84,91.

Arthrodeses following infections are more likely to be followed by delayed union or nonunion26,36. In a review of the results of ninety-one arthrodeses performed in patients with a failed total knee arthroplasty, Knutson et al.57 reported that fusion was achieved in thirty-one (62%) of fifty knees in which the infection was initially controlled compared with four of twenty-one knees in which active infection had been present at the time of the arthrodesis. Rothacker and Cabanela30 reported a mean time to clinical union of 20.8 weeks (range, seven to sixty weeks) after arthrodeses in patients with an infection at the site of a total knee arthroplasty compared with 11.3 weeks (range, six to sixteen weeks) in those with an aseptic knee. In a study of twenty-seven patients, Woods et al.66 found that it took a mean of 2.2 months longer for the fusion to heal in patients who had had the arthrodesis at the site of an infection.

Back to Top | Article Outline
Osteoarthritis

Charnley68 reported a 99% fusion rate (169 of 171) in patients with osteoarthritis and tuberculous arthritis. That is the largest reported series of arthrodeses for osteoarthritis because the majority of patients with osteoarthritis now are treated with total knee arthroplasty. Many recent studies have included only a few primary arthrodeses in patients with posttraumatic arthritis or osteoarthritis40,50,92. The time to fusion was exceptionally short in Charnley's series, with the majority of patients returning to work, without a brace, after twelve weeks. The indications for knee arthrodesis in patients with osteoarthritis have decreased dramatically because of the expanded indications for total knee arthroplasty in the younger population18.

Back to Top | Article Outline
Charcot Arthropathy

The numbers of arthrodeses for Charcot arthropathy are decreasing because there are fewer patients with this disease9,93. Also, many patients with Charcot arthropathy are choosing to preserve motion with a total knee arthroplasty, knowing that, if it fails, they can then have a knee arthrodesis. Early reports in the literature attest to the difficulty, with high failure rates, of fusing knees with Charcot arthropathy5-7,85,94-96. In 1957, Holt97 reported successful fusion in five of five knees with Charcot arthropathy that had been treated with an intramedullary rod combined with the Charnley clamp. Drennan et al.87 reported successful fusion in nine of nine knees treated with an intramedullary nail.

Back to Top | Article Outline

Complications

Complication rates following knee arthrodesis are high, ranging from 20% to 84%26,30,45,78. Complications include recurrent tumor or infection, peroneal nerve palsy, nonunion, and thrombophlebitis. Peroneal nerve palsy is relatively common26,54,55,71. Arroyo et al.13 reported three peroneal nerve palsies and Enneking and Shirley21 reported four following proximal tibial tumor resection. Conway51 reported two peroneal nerve palsies—one acute and one late (more than three months after the arthrodesis)—and they were both treated with peroneal nerve decompression, which completely restored function. Mont et al.98 and Nogueira et al.99 recommended early nerve decompression as it leads to a quicker and more reliable recovery.

Nonunion following knee arthrodesis is a very difficult problem to manage. Behr et al.29 reported one symptomatic nonunion in their series of nine patients who had had an arthrodesis because of a failed total knee arthroplasty. The non-union was treated with repeat external fixation and bone-grafting, and the result was a painless pseudarthrosis. Brodersen et al.43 reported that five of forty-five arthrodeses were followed by a symptomatic nonunion, which was attributed to severe bone loss and recurrent infection. Rand and Bryan41 reported on twenty-five knees with a failed arthrodesis. The knees had a mean of 1.6 attempts at arthrodesis (range, one to four attempts), and the most common indication for revision arthrodesis was pain and instability. The majority of the failures were due to bone loss. The first repeat attempt at arthrodesis with an external fixator succeeded in five of fourteen knees, and the second attempt succeeded in four of eight. Clearly, a failed knee arthrodesis remains a difficult problem.

Complications following arthrodeses with an external fixator include pin track infection30,76,100 and fracture through a pin site30,100. Several authors have reported one, two, or three cases of fracture through a pin site30,31,33,57. The majority of the reported pin track infections were treated with oral antibiotics; however, in a few cases, the pins became loose and had to be removed and replaced26,43-45,77.

Complications specific to rod insertion include perforation of the tibial shaft, disengagement of the coupling device, stress fracture at the site of the arthrodesis, pain at the end of the nail, proximal prominence of the nail with gluteal pain, and breakage of the nail13,17,55,63,101. Knutson et al.27 reported two cases of proximal nail migration that required reinsertion and proximal fixation to prevent additional migration. Arroyo et al.13 reported three stress fractures at the arthrodesis site after intramedullary nail insertion that were treated with a cast, resulting in complete pain relief. Goldberg et al.101 reported one perforation of the tibial shaft in a series of seventeen arthrodeses performed with a Huckstep nail. The perforation was corrected intraoperatively by repositioning of the nail. Goldberg et al. also reported nail breakage through a nonunion; this was treated with revision nailing, which was followed by union.

Back to Top | Article Outline

Function Following Knee Arthrodesis

Several authors have evaluated patient function following a knee arthrodesis. Harris et al.102 compared patients with a knee arthrodesis with those with a constrained total knee arthroplasty and with those with an above-the-knee amputation. They found that the patients with an arthrodesis performed the most physically demanding activities and had superior stability (less symptomatic giving-way). Benson et al.103 compared nine patients who had had a knee arthrodesis with nine patients who had had a primary total knee arthroplasty. The Short Form-36 scores for pain, health, vitality, and social and emotional health were similar in the two groups. The arthroplasty group scored higher for physical functioning, but the arthrodesis group had a slightly higher mean global score (74 compared with 73 points). The Arthritis Impact Measurement Score was better after the arthroplasties because of increased mobility (0.97 compared with 2.5 points) and physical activity (4 compared with 6.3 points). Patients with an arthrodesis had a better mean score on the pain scale (3.3 compared with 3.9 points).

Rud and Jensen104 evaluated the activities of twenty-three patients following knee arthrodesis and found that eighteen had returned to work. Siller and Hadjipavlou61 indicated that patients with a fused knee had difficulty with stairs, rugs, and ladders, and those who had performed strenuous work preoperatively rarely resumed that type of work after the arthrodesis. Enneking and Shirley21 reported that fifteen of twenty patients resumed their previous occupation and sixteen patients walked without a cane or walker. Some of those patients were highly active, with pursuits that included golfing, water-skiing, sailing, hunting, and racquetball. In a study of twenty-eight knee arthrodeses, Rand et al.31 reported that nine patients with a successful arthrodesis could walk more than six blocks and seven could walk one to three blocks. David et al.44 reported that all of their thirteen patients with a successful knee arthrodesis, which had been performed with the Ilizarov technique, needed a cane to walk. In another series79, which included nineteen Ilizarov knee arthrodeses, nine of thirteen patients with a successful arthrodesis used a cane or walker and only two patients did not require any assistive device. In a study of twenty-one patients, Arroyo et al.13 reported that twelve of eighteen with a successful arthrodesis walked without an assistive device, five used a cane, and one used a walker. Fourteen patients could walk the same distance or farther than they could preoperatively, four had no walking limitations, and four could not walk as far as they could preoperatively.

With regard to overall function, patients can expect a stable, painless extremity with difficulty climbing stairs and in sitting in movie theaters and in airplanes. It is difficult to quantify the degree of social problems associated with a knee arthrodesis. In a report by Kim et al.105 on thirty patients in whom a knee arthrodesis had been converted to a total knee arthroplasty, seventeen patients had attempted suicide preoperatively because they were unhappy with the conditions of the lower extremity. Counseling prior to the arthrodesis and appropriate patient selection are essential to prevent unrealistic postoperative expectations as well as to minimize postoperative depression.

Back to Top | Article Outline

Alternatives to Arthrodesis

Alternatives to knee arthrodesis include resection arthroplasty, artificial arthrodesis, and above-the-knee amputation. Falahee et al.106 reported that, in their series of twenty-eight resection arthroplasties performed because of an infection following a total knee arthroplasty, six knees had an unacceptable result and underwent knee arthrodesis. Fifteen patients were able to walk independently with a cane or walker, and ten of them walked with a knee-ankle-foot orthosis. The mean arc of knee motion was 40° (range, 20° to 90°). The advantage of a resection arthroplasty is a better sitting position, but the major disadvantage is pain and instability with walking. According to Thornhill et al.107, patients who are able to walk but have poor bone stock or poor soft tissue are not good candidates for resection arthroplasty because the limb will not be stable enough for weight-bearing.

Scott et al.108 reported the results of resection arthroplasty with a cement spacer at the knee in thirteen patients with an infection following a total knee arthroplasty who were medically compromised and not deemed to be candidates for other procedures. Three patients died within six weeks after the procedure, but three patients survived for five years. Only one of the surviving patients was satisfied with the result and was able to walk with a cane. Two survivors scored below the twenty-fifth percentile on the Short Form-36 questionnaire. The authors recommended resection arthroplasty only for severely medically compromised patients who are not candidates for procedures that provide better function.

Another option is an artificial arthrodesis, as described in a case report by Voss109, who performed it in a patient with a severe infection at the site of a total knee arthroplasty. This technique consists of insertion of a long intramedullary rod into a large cement spacer at the knee. The cement spacer fills the space between the femur and tibia and allows immediate full weight-bearing. This technique was first described by Campanacci and Costa110 for patients who had undergone a large tumor resection about the knee, and it has been used by other tumor surgeons as well110. This technique can provide a stable limb for weight-bearing.

Finally, above-the-knee amputation is an option, although the results following total knee arthroplasty are poor. Pring et al.38 reported that only seven of twenty-three patients who had undergone an above-the-knee amputation were able to walk. Isiklar et al.37 reported on nine above-the-knee amputations in eight patients; only two of the eight patients were able to walk with a walker, and one was able to walk with a prosthetic device. The authors recommended avoiding this procedure in patients who are able to walk. However, there may be situations in which a patient is best served by an above-the-knee-amputation, such as when the extremity is dysvascular. Surgeons should be aware that these patients are unlikely to ever walk.

Back to Top | Article Outline

Conversion of Arthrodesis to Total Knee Arthroplasty

Conversion of a knee arthrodesis to a total knee arthroplasty has been addressed in several series105,111-119. Kreder et al.113 reported on eighteen conversions, which were associated with multiple complications. Two patients had an infection, and three required removal of components within the first four years after the conversion. Two of the patients who had the components removed had a repeat arthrodesis of the knee. Cameron and Hu118 reported that eight of seventeen patients in whom a formal knee arthrodesis had been converted to a total knee arthroplasty had a good or excellent result according to The Hospital for Special Surgery scoring system. Postoperative complications occurred in nine patients and included myositis ossificans, loosening of the tibial component, ligamentous laxity, rupture of the quadriceps tendon, wound breakdown, and repeat arthrodesis. Kim et al.105 reported conversion of sixteen knee arthrodeses after spontaneous ankylosis and conversion of fourteen after a formal knee arthrodesis. The mean active range of flexion was 75° (range, 70° to 95°), and the mean extension lag was 9° (range, 0° to 20°). The complication rate was substantial (53%; sixteen of thirty patients). Nineteen patients had skin-edge necrosis, two patients had infection, and one patient had a quadriceps tendon rupture.

Henkel et al.117 reported the results of total knee arthroplasty after a formal knee arthrodesis in seven patients, five of whom were satisfied with the outcome and two of whom had a repeat arthrodesis. Six of the seven patients underwent a reoperation because of complications. Three patients had an open arthrolysis, and two had a gastrocnemius flap procedure because of wound necrosis. Five patients had increased pain with walking postoperatively. Even though a mobile knee can improve the walking ability of these patients, the rate of complications and repeat arthrodesis is substantial. The procedure is fraught with complications and guarded outcomes; as a result, it should not be undertaken lightly by either the patient or the surgeon. This procedure is typically patient-driven and should be selected on an individual basis. In our experience, it has been successful in selected cases when it has been combined with a Judet quadricepsplasty to mobilize the extensor mechanism119. This technique is performed in phases until adequate flexion is obtained. In phase 1, the suprapatellar pouch is mobilized and intra-articular adhesions are released. In phase 2, the vastus lateralis is divided from the linea aspera and the vastus intermedius is dissected extraperiosteally off the lateral femoral condyle. In phase 3, the rectus femoris is released from the anterior inferior iliac spine120.

Back to Top | Article Outline

Overview

Knee arthrodesis is indicated most commonly for patients with an infection at the site of a total knee arthroplasty that is not amenable to revision arthroplasty. The success of knee arthrodesis, regardless of the indication, is directly related to the patient's bone stock, physical demands, and expectations. Knee arthrodesis is technically challenging, and surgeons performing this procedure need to be adept at managing the multiple complications that may occur. A patient with a successful knee arthrodesis may be able to walk effectively, particularly in comparison with a patient with an above-the-knee amputation. Conversion of a knee fusion to a total knee arthroplasty is associated with a high complication rate and should be performed only in very selected cases.

The authors did not receive grants or outside funding in support of their research or preparation of this manuscript. They did not receive payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the authors are affiliated or associated.

Investigation performed at Rubin Institute for Advanced Orthopaedics, Baltimore, Maryland

1. . Arthrodesis of the knee. Clin Orthop. 1960;18: 37-42.
2. . The treatment of tuberculosis of the joints of the lower extremities by operative fusion. J Bone Joint Surg. 1930;12: 749-54.
3. . Intramedullary pinning for arthrodesis of the knee joint. J Bone Joint Surg Am. 1948;30: 728-34.
4. . Knee fusion by the use of a three-flanged nail. J Bone Joint Surg. 1946;28: 550-4.
5. . The Charcot knee (abstract). J Bone Joint Surg Br. 1958;40: 148.
6. , Bland WG. Compression in arthrodesis. A comparative study of methods of fusion of the knee in ninety-three cases. J Bone Joint Surg Am. 1958;40: 585-606.
7. , Marchetti PG. Arthrodesis of the knee. Review of 81 cases. Bull Hosp Spec Surg. 1961;4: 81-97.
8. . Fusion in Charcot's disease of the knee: new technique for arthrodesis. Ann Surg. 1938;108: 124-6.
9. , Barnes KL, Noble J. A technique for difficult arthrodesis of the knee. J Bone Joint Surg Br. 1984;66: 367-70.
10. . Substitution of whole or half joints from freshly amputated extremities by free plastic operations. Surg Gynecol Obstet. 1908;6: 601-7.
11. , Trousdale RT, Osmon DR. Patient outcome with reinfection following reimplantation for the infected total knee arthroplasty. Clin Orthop. 1995;321: 55-67.
12. , Mont MA, Orlando JC, Delanois RE, Hungerford DS. Operative management and outcome of complex wounds following total knee arthroplasty. Plast Reconstr Surg. 1999;104: 1688-97.
13. , Garvin KL, Neff JR. Arthrodesis of the knee with a modular titanium intramedullary nail. J Bone Joint Surg Am. 1997;79: 26-35.
14. , Pellicci PM, Windsor RE, Conrad EU, Insall JN, Salvati EA. Contralateral total hip arthroplasty or ipsilateral total knee arthroplasty in patients who have a long-standing fusion of the hip. J Bone Joint Surg Am. 1989;71: 1355-62.
15. , Hulard O, Zalta R, Bardot J, Magalon G. Management of wounds of exposed or infected knee prostheses. Scand J Plast Reconstr Surg Hand Surg. 2001;35: 71-7.
16. , McBeath AA. Arthrodesis following failed total knee arthroplasty: comprehensive review and meta-analysis of recent literature. Orthopedics. 1995;18: 361-8.
17. , Makley JT, Locke M, Gentili A, Heiner J. Resection arthrodesis of the knee for tumor: large intercalary allograft and long intramedullary nail technique. Semin Arthroplasty. 1994;5: 76-84.
18. , Hershman S, Mont M, Lotke PA. Total knee arthroplasty in patients 40 years and younger with osteoarthritis. Clin Orthop. 2000; 380: 85-90.
19. , Chan KC, Mills DM. 5- to 18-year follow-up study of cemented total knee arthroplasty for patients 55 years old or younger. J Arthroplasty. 1997;12: 49-54.
20. , Lee CW, Sheldon M, Lennon WC, Hungerford DS. Total knee arthroplasty in patients </= 50 years old. J Arthroplasty. 2002;17: 538-43.
21. , Shirley PD. Resection-arthrodesis for malignant and potentially malignant lesions about the knee using an intramedullary rod and local bone grafts. J Bone Joint Surg Am. 1977;59: 223-36.
22. , Rechl H, Burgkart R, Messmer C, Schelter R, Hipp E, Gradinger R. Limb salvage with tumor endoprostheses for malignant tumors of the knee. Clin Orthop. 2002;405: 207-15.
23. , Toftgaard Jensen T. Clinical outcome after treatment of infected primary total knee arthroplasty. Acta Orthop Belg. 2002;68: 500-7.
24. , Butler RA, Andrews P, Rorabeck CH, Engh G. Managing the infected knee: as good as it gets. Orthopedics. 2000;23: 991-2.
25. , Barden RM, Rosenberg AG. Results of different surgical procedures on total knee arthroplasty infections. J Arthroplasty. 1996;11: 931-8.
26. , Hovelius L, Lindstrand A, Lidgren L. Arthrodesis after failed knee arthroplasty. A nationwide multicenter investigation of 91 cases. Clin Orthop. 1984;191: 202-11.
27. , Lindstrand A, Lidgren L. Arthrodesis for failed knee arthroplasty. A report of 20 cases. J Bone Joint Surg Br. 1985;67: 47-52.
28. , Stearns KL. Intramedullary fixation for arthrodesis of the knee after infected total knee arthroplasty. Clin Orthop. 1989;248: 87-92.
29. , Chmell SJ, Schwartz CM. Knee arthrodesis for failed total knee arthroplasty. Arch Surg. 1985;120: 350-4.
30. Jr, Cabanela ME. External fixation for arthrodesis of the knee and ankle. Clin Orthop. 1983;180: 101-8.
31. , Bryan RS, Chao EY. Failed total knee arthroplasty treated by arthrodesis of the knee using the Ace-Fischer apparatus. J Bone Joint Surg Am. 1987;69: 39-45.
32. , Denham RA. Arthrodesis of the knee after failed knee replacement. J Bone Joint Surg Br. 1984;66: 362-6.
33. , Lieberman JR, Finerman GA. Single plane and biplane external fixators for knee arthrodesis. Clin Orthop. 1995;316: 134-44.
34. , Perry J, Antonelli D, Hislop H. Energy cost of walking of amputees: the influence of level of amputation. J Bone Joint Surg Am. 1976; 58: 42-6.
35. , Morrey BF. Persistent infection after successful arthrodesis for infected total knee arthroplasty. A report of two cases. J Arthroplasty. 1990;5: 277-9.
36. , Knutson K, Lidgren L. Treatment of infected knee arthroplasty. Clin Orthop. 1989;245: 173-8.
37. , Landon GC, Tullos HS. Amputation after failed total knee arthroplasty. Clin Orthop. 1994;299: 173-8.
38. , Marks L, Angel JC. Mobility after amputation for failed knee replacement. J Bone Joint Surg Br. 1988;70: 770-1.
39. , Starr R, Bhave A, Tennis S. A biomechanical and functional comparison of the artificially fused knee at zero and twenty degrees of knee flexion as determined by gait analysis and oxygen consumption. Presented as a poster exhibit at the Annual Meeting of the Limb Lengthening and Reconstruction Society; 2003 July 25-27; Boston, MA.
40. , Matthews LS, Kaufer H. Arthrodesis of the knee with an intramedullary nail. J Bone Joint Surg Am. 1991;73: 907-13.
41. , Bryan RS. The outcome of failed knee arthrodesis following total knee arthroplasty. Clin Orthop. 1986;205: 86-92.
42. , Cushner FD, Craig-Scott S, Scott WN. Long-term results of total knee arthroplasty after the use of soft tissue expanders. Clin Orthop. 2000;380: 133-9.
43. , Fitzgerald RH Jr, Peterson LF, Coventry MB, Bryan RS. Arthrodesis of the knee following failed total knee arthroplasty. J Bone Joint Surg Am. 1979;61: 181-5.
44. , Shtarker H, Horesh Z, Tsur A, Soudry M. Arthrodesis with the Ilizarov device after failed knee arthroplasty. Orthopedics. 2001;24: 33-6.
45. , van Roermund PM. Arthrodesis of the knee after an infected arthroplasty using the Ilizarov method. J Bone Joint Surg Br. 2001; 83: 50-4.
46. , Pullen C, Deromedis B, Catagni MA. Knee arthrodesis after infected total knee arthroplasty using the Ilizarov method. Clin Orthop. 2001; 389: 143-9.
47. , Kortelainen P, Jalovaara P. Arthrodesis of the knee with intramedullary nail fixation. J Bone Joint Surg Am. 1990;72: 433-42.
48. , Pullen C, Guerreschi F, Catagni MA. Knee arthrodesis and limb lengthening in the treatment of infected total knee arthroplasty: case report. J Trauma. 2002;52: 359-63.
49. . Knee arthrodesis and simultaneous limb lengthening using the Ilizarov method. Read at the Annual Meeting of the Limb Lengthening and Reconstruction Society; 2002 June 10-11; Irving, TX.
50. , Gotzen L, Baumgaertel F. Knee arthrodesis with a unilateral external fixator. Acta Chir Belg. 1996;96: 123-7.
51. . Arthrodesis of the knee using biplanar external fixation. Read at the Annual Meeting of the Limb Lengthening and Reconstruction Society; 2003 July 25-27; Boston, MA.
52. , Ishii S, Naito T, Wada T, Nagoya A, Takahashi T, Tsuchida Y. Arthrodesis of knee joint by vascularized fibular graft. Microsurgery. 1996; 17: 2-8.
53. , Scarborough M, Vander Griend RA. Resection arthrodesis of the knee with an intercalary allograft. J Bone Joint Surg Am. 1996;78: 185-92.
54. , Hanel DP. Knee arthrodesis using combined intramedullary rod and plate fixation. Clin Orthop. 1993;294: 238-41.
55. , Rand JA. Intramedullary arthrodesis of the knee after failed total knee arthroplasty. J Bone Joint Surg Am. 1994;76: 870-7.
56. , Mont MA, Payman KR, Freiberg AA, Windsor RE, Sculco TP, Hungerford DS. Infected total knee arthroplasty treated with arthrodesis using a modular nail. Clin Orthop. 1999;367: 230-7.
57. , Hovelius L, Lindstrand A, Lidgren L. Arthrodesis after failed knee arthroplasty. A nationwide multicenter investigation of 91 cases. Clin Orthop. 1984;191: 202-11.
58. , Froehlich J. Knee fusion with intramedullary rods for failed total knee arthroplasty. Clin Orthop. 1985;197: 209-16.
59. , Shen WJ, Yang CY. Arthrodesis with a short Huckstep nail as a salvage procedure for failed total knee arthroplasty. J Bone Joint Surg Am. 1998;80: 380-8.
60. , Torholm C. Arthrodesis after infected knee arthroplasty using long arthrodesis nail. A report of five cases. Am J Knee Surg. 1995; 8: 110-3.
61. , Hadjipavlou A. Knee arthrodesis: long-term results. Can J Surg. 1976;19: 217-9.
62. , Törholm C. Intramedullary reaming in chronic diaphyseal osteomyelitis: a preliminary report. Clin Orthop. 1980;151: 215-21.
63. , Stewart HD, Newton G. Curved Kuntscher nail arthrodesis after failure of knee replacement. J Bone Joint Surg Br. 1989;71: 588-90.
64. , Bogner E, Desai P, Lonner JH, Adler E, Zuckerman JD, Di Cesare PE. Analysis of frozen sections of intraoperative specimens obtained at the time of reoperation after hip or knee resection arthroplasty for the treatment of infection. J Bone Joint Surg Am. 1999;81: 684-9.
65. , Waldman BJ, Mont MA, Hungerford DS. Infections associated with dental procedures in total hip arthroplasty. J Bone Joint Surg Br. 1999; 81: 56-9.
66. , Lionberger DR, Tullos HS. Failed total knee arthroplasty. Revision and arthrodesis for infection and noninfectious complications. Clin Orthop. 1983;173: 184-90.
67. , Woods GW, Tullos HS. Arthrodesis in failed total knee replacement. J Bone Joint Surg Am. 1978;60: 790-4.
68. . Arthrodesis of the knee. Clin Orthop. 1960;18: 37-42.
69. , Lilly JW, Bartlett CS, Churchill DL. Arthrodesis of the knee: experience with intramedullary nailing. J Arthroplasty. 2000;15: 871-6.
70. , DeBoer DK, McQueen DA, Cooke FW, Hahn DL. Salvage procedures for failed total knee arthroplasty. J Bone Joint Surg Am. 2003; 85 (Suppl 1): S58-62.
71. , Lowe HG. A study of the end-results of compression arthrodesis of the knee. J Bone Joint Surg Br. 1958;40: 633-5.
72. Jr, Hansen NM Jr. The biplane frame: modified compression arthrodesis of the knee. Clin Orthop. 1981;160: 163-7.
73. . Knee arthrodesis following prosthesis removal. Use of the Wagner apparatus. J Bone Joint Surg Br. 1983;65: 29-31.
74. , Heikkila J, Magyar G, Toksvig-Larsen S, Giannini S. Fixation strength and pin tract infection of hydroxyapatite-coated tapered pins. Clin Orthop. 2001;388: 209-17.
75. , Bodelind B, Lidgren L. Stability of external fixators used for knee arthrodesis after failed knee arthroplasty. Clin Orthop. 1984;186: 90-5.
76. , Verebelyi DM. One-stage debridement and knee fusion for infected total knee arthroplasty using the hybrid frame. J Arthroplasty. 2002;17: 129-34.
77. , Fitch RD, Hoffmann ED, Hardaker WT, Vail TP, Scully SP. Knee arthrodesis with circular external fixation. Clin Orthop. 2001;382: 168-78.
78. , Richardson JB, Soriano RM, Kenwright J. A mechanical assessment of applied compression and healing in knee arthrodesis. Clin Orthop. 1989;242: 256-64.
79. , Mallin BA, Matthews AC. Knee arthrodesis with a tension-band plate. J Bone Joint Surg Am. 1988;70: 285-8.
80. , Murray WR. Arthrodesis of the knee by double-plating. J Bone Joint Surg Am. 1961;43: 795-808.
81. , Landon GC, Tullos HS. Arthrodesis with dual plates after failed total knee arthroplasty. J Bone Joint Surg Am. 1991;73: 1020-4.
82. , Knessl J, Gschwend N. [Arthrodesis following knee arthroplasty]. Orthopade. 1987;16: 301-9. German.
83. , Bishop AT, Wood MB. Arthrodesis of the knee with a vascularized fibular rotatory graft. J Bone Joint Surg Am. 1995;77: 751-9.
84. , Rosenwasser MP, Caulo N, Schink MM, Bassett CA. The use of pulsing electromagnetic fields to achieve arthrodesis of the knee following failed total knee arthroplasty. A preliminary report. J Bone Joint Surg Am. 1983;65: 480-5.
85. , Parkes JC 2nd, Stinchfield FE. Arthrodesis of the knee. A followup study. J Bone Joint Surg Am. 1967;49: 1065-78.
86. Jr. Knee joint tuberculosis. Two hundred twenty-two patients treated by operative fusion. Surg Gynecol Obstet. 1939;68: 1029-37.
87. , Fahey JJ, Maylahn DJ. Important factors in achieving arthrodesis of the Charcot knee. J Bone Joint Surg Am. 1971;53: 1180-93.
88. , Nerubay J, Caspi I, Horoszowski H. Resection arthrodesis of the knee in the treatment of tumours—a long-term follow-up. Int Orthop. 1997; 21: 101-3.
89. . Modified system for functional evaluation of surgical management of musculoskeletal tumors. In: Enneking WF, editor. Bristol-Meyers/Zimmer orthopaedic symposium. Limb salvage in musculoskeletal oncology. New York: Churchill Livingstone; 1987. p 626-39.
90. , Bishop AT, Wood MB. Arthrodesis of the knee with a vascularized fibular rotatory graft. J Bone Joint Surg Am. 1995;77: 751-9.
91. . Arthrodesis in failed knee replacement in eight rheumatoid patients. Ann Chir Gynaecol. 1979;68: 57-62.
92. , Heidt RS Jr, Miller EH, Welch MC. Arthroscopic-assisted arthrodesis of the knee. Arthroscopy. 1991;7: 237-40.
93. , Shen WJ, Yang CY. Arthrodesis with a short Huckstep nail as a salvage procedure for failed total knee arthroplasty. J Bone Joint Surg Am. 1998;80: 380-8.
94. . Experiences with intramedullary fixation in knee fusion. Am J Surg. 1952;83: 291-6.
95. . The value of intramedullary nail for knee fusion particularly for the Charcot joint. Am J Surg. 1954;87: 63-5.
96. Jr, Urist MR. Arthrodesis of the knee with intramedullary nail fixation. Clin Orthop. 1960;18: 43-53.
97. . Fusion of the Charcot knee. South Med J. 1957;50: 1215-20.
98. , Dellon AL, Chen F, Hungerford MW, Krackow KA, Hungerford DS. The operative treatment of peroneal nerve palsy. J Bone Joint Surg Am. 1996;78: 863-9.
99. , Paley D, Bhave A, Herbert A, Nocente C, Herzenberg JE. Nerve lesions associated with limb-lengthening. J Bone Joint Surg Am. 2003;85: 1502-10.
100. , Bryan RS, Chao EY. Failed total knee arthroplasty treated by arthrodesis of the knee using the Ace-Fischer apparatus. J Bone Joint Surg Am. 1987;69: 39-45.
101. , Drummond RP, Bruce WJ, Viglione W, Lennon WP. Huckstep nail arthrodesis of the knee: a salvage for infected total knee replacement. Aust NZ J Surg. 1989;59: 147-50.
102. , Leff AR, Gitelis S, Simon MA. Function after amputation, arthrodesis, or arthroplasty for tumors about the knee. J Bone Joint Surg Am. 1990;72: 1477-85.
103. , Resine ST, Lewis CG. Functional outcome of arthrodesis for failed total knee arthroplasty. Orthopedics. 1998;21: 875-9.
104. , Jensen UH. Function after arthrodesis of the knee. Acta Orthop Scand. 1985;56: 337-9.
105. , Kim JS, Cho SH. Total knee arthroplasty after spontaneous osseous ankylosis and takedown of formal knee fusion. J Arthroplasty. 2000;15: 453-60.
106. , Matthews LS, Kaufer H. Resection arthroplasty as a salvage procedure for a knee with infection after a total arthroplasty. J Bone Joint Surg Am. 1987;69: 1013-21.
107. , Dalziel RW, Sledge CB. Alternatives to arthrodesis for the failed total knee arthroplasty. Clin Orthop. 1982;170: 131-40.
108. , Hennessey MS, Parkinson RW, Molloy AP. Long-term outcome of the `Beefburger' procedure in patients unsuitable for two-stage revision following infected total knee replacement. Knee. 2001;8: 281-6.
109. . A new technique of limb salvage after infected revision total knee arthroplasty: artificial fusion. J Arthroplasty. 2001;16: 524-8.
110. , Costa P. Total resection of distal femur or proximal tibia for bone tumours. Autogenous bone grafts and arthrodesis in twenty-six cases. J Bone Joint Surg Br. 1979;61: 455-63.
111. , Biagini R, Ruggieri P, Bettelli G, Casadei R, Campanacci M. Temporary resection-arthrodesis of the knee using an intramedullary rod and bone cement. Int Orthop. 1989;13: 253-8.
112. , Jackson DW. Considerations in total knee arthroplasty following previous knee fusion. Clin Orthop. 1988;227: 223-8.
113. , Williams JI, Jaglal S, Axcell T, Stephen D. A population study in the Province of Ontario of the complications after conversion of hip or knee arthrodesis to total joint replacement. Can J Surg. 1999;42: 433-9.
114. , Freeman MA, Albrektsson BE. Total prosthetic replacement of ankylosed knees. J Arthroplasty. 1987;2: 179-83.
115. . Role of total knee replacement in failed knee fusions. Can J Surg. 1987;30: 25-7.
116. Jr, Lotke PA, Pagnano MW, Hanssen AD. Total knee arthroplasty in a previously ankylosed or arthrodesed knee. Clin Orthop. 1996; 331: 234-7.
117. , Boldt JG, Drobny TK, Munzinger UK. Total knee arthroplasty after formal knee fusion using unconstrained and semiconstrained components: a report of 7 cases. J Arthroplasty. 2001;16: 768-76.
118. , Hu C. Results of total knee arthroplasty following takedown of formal knee fusion. J Arthroplasty. 1996;11: 732-7.
119. , Mont MA. Results of total knee arthroplasties after fusion take-downs using Judet quadricepsplasty procedure. Unpublished data.
120. . The Judet quadricepsplasty for management of severe posttraumatic extension contracture of the knee. A report of a bilateral case and review of the literature. Clin Orthop. 1990;256: 169-73.
Copyright 2004 by The Journal of Bone and Joint Surgery, Incorporated