Abstract
We reviewed the results an average of fifty months (range, twenty-four to 120 months) after the use of thirty-five allografts in thirty patients during primary or revision total knee replacement. Twenty-nine femoral-head allografts, five distal femoral allografts, and one proximal tibial allograft were used in conjunction with a long-stemmed implant to reconstruct large osseous defects. The patients were evaluated clinically, radiographically, and subjectively (with use of a questionnaire). Twenty-six (87 per cent) of the thirty patients had a good or excellent clinical result, and no revisions were necessary. As none of the patients had collapse of the graft, subsidence of the implant, or revision, we believe that the outcome of treatment with a femoral-head allograft, particularly in association with a component inserted with cement, is excellent.Four non-porous-coated components were placed without cement on structural allografts. Radiographically, three of those components subsided, but none of the three needed revision and two were associated with a good clinical result. Our current practice is to cement components in all arthroplasties involving grafting.Our findings suggest that the use of a stemmed component reduces the stress on the allograft, host bone, and fixation interface. In addition, such a component contributes to the longevity of a total knee replacement associated with a bone graft. Additional studies with long-term follow-up are necessary to confirm this outcome.
Large osseous defects occasionally are present at the planned site of a primary or revision total knee arthroplasty. Recently, promising short and intermediate-term clinical results have been reported after total knee arthroplasty with bone-grafting7,9,22,23. In those studies, in which the longest average duration of follow-up was less than five years, none of the bulk grafts failed as a result of collapse7,9,22,23. The findings in those reports contrast with the poor clinical results of the use of major structural allografts to repair acetabular bone defects and to provide support for total hip implants10,14,16.
Since 1985, we have used large structural allografts to reconstruct damaged segments of bone, in conjunction with insertion of a long-stemmed component to reduce the load on the graft. Femoral-head allografts were used whenever there was sufficient metaphyseal bone in the femur or the tibia to stabilize the graft and to provide support for the component. The purpose of the present study was to determine the intermediate-term clinical and radiographic results of primary or revision total knee arthroplasties performed with use of this technique.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
†Anderson Orthopaedic Reasearch Institute, 2445 Army Navy Drive, Arlington, Virginia 22206.
‡Orthopaedic Associates of Augusta, 820 St. Sebastian Way, Suite 8A, Augusta. Georgia 30901.
Study Group
Between January 1985 and December 1992, thirty-seven patients had a primary or revision total knee arthroplasty in which a bulk allograft was used to reconstruct a major defect of bone in the tibia or the femur, or both. Data on all of the involved knees, including the type of prosthesis, the method of fixation, and information regarding the bone-grafting, were entered into a computer database at the time of the operation.
Three of the thirty-seven patients died less than two years postoperatively. None of those patients had had any additional procedures on the involved knee. Four other patients were unable to return for follow-up. The remaining thirty patients had been followed clinically and radiographically for at least two years (average, fifty months; range, twenty-four to 120 months). We report the results for those patients.
Twenty-six patients had the graft reconstruction during a revision total knee arthroplasty, and five of those patients had it during a repeat revision operation. The other four patients, who had a major structural osseous deficiency, had the reconstruction during a primary total knee arthroplasty. An average of 1.2 procedures (range, zero to two procedures) had been performed on each knee before reconstruction with the allograft.
Eighteen patients were women and twelve were men. The average age at the time of the index procedure was seventy years (range, forty-three to eighty-six years). The primary diagnosis was osteoarthrosis in twenty-six patients and rheumatoid arthritis in four.
Thirty-five structural allografts were used in the thirty total knee arthroplasties. Twenty-nine femoral-head allografts, five distal femoral allografts, and one proximal tibial allograft were employed (Table I). Three patients needed a graft in both the femur and the tibia at the time of the arthroplasty. In addition, two patients each needed two femoral-head allografts to reconstruct a large metaphyseal defect that involved the medial and lateral femoral condyles.
Defects were classified with use of the system described by one of us (G. A. E.) and Parks3,4. Type I indicates that the metaphyseal bone is intact and the defect usually can be reconstructed with particles of autologous bone graft or with cement and a primary component. Type-II defects have damaged metaphyseal bone that can be replaced with augments or wedges and a revision-type implant. Type-III defects have deficient metaphyseal bone that necessitates reconstruction with a bulk allograft and a long-stemmed revision implant or a custom component. All patients in the present study had type-III defects. The defects were predicted from preoperative radiographs and confirmed intraoperatively. There were twenty tibial defects and thirteen femoral defects in the thirty patients.
Revision arthroplasty was performed for osteolysis, loosening or failure of a component, instability, or femoral fracture (Table II). Stemmed components were used in all patients. For the primary arthroplasties, the length of the stem was based on the availability of the implant system and the size of the tibial tray. Canal-filling stems were not used in the primary arthroplasties. Sixteen of the tibial stems were partially cemented, and four were inserted without cement. Four femoral stems were inserted with cement and nine, without cement.
In general, custom implants with a press-fit canal-filling stem were used in the patients who were seen earliest in the study; four patients were managed with such a stem. In subsequent revisions, implants with a stem of intermediate length that did not fill the canal were inserted with cement. Implants that had a longer stem and good stability from intramedullary fit were inserted with cement only in the metaphyseal segment.
Operative Technique
The technique that was used for the insertion of the femoral-head allografts was similar for both the femoral and the tibial defects. When the metaphyseal segment was extensively damaged, a metaphyseal-diaphyseal graft was fashioned and was intercalated between the host bone and the long-stemmed component. Allografts were packaged and stored at -70 degrees Celsius and were transported in dry ice. Grafts that were not inserted were maintained in a frozen state and safely returned to storage for later use; thus, unnecessary costs were avoided. The twenty-seven defects of bone that were to be repaired with femoral-head allograft first were debrided of all granulomatous and membranous tissue. Acetabular reamers were used to prepare the host bed (Fig. 1-A). In the knees that had a primary arthroplasty, the densely sclerotic bed of host bone was prepared with an acetabular reamer until there was viable cancellous bone at the base of the defect. In knees that had a revision arthroplasty, reaming continued until a hemispherical shape was formed to provide optimum containment of the graft and stability without destroying the attachment of the ligaments to the condyles or to the tibial plateau. A vise system (Allogrip; DePuy, Warsaw, Indiana) for holding bone was used to stabilize the femoral-head allograft on a separate table during preparation of the graft. Femoral-head shapers (Allogrip; DePuy) were used to remove all cartilage and subchondral bone from the graft (Fig. 1-B). The femoral-head allograft was reamed to the cancellous level, and its diameter was one to two millimeters larger than the size of the host bed. The graft then was impacted into the recipient site to achieve an interference fit, and it was temporarily stabilized with use of Kirschner wires (Fig. 1-C). The graft was trimmed to accept the respective cutting guides for preparation of the tibia or the femur (Fig. 1-D). Bone cement provided final fixation between the graft and the implant and, in some knees, between the stem and the host bone. The Kirschner wires were removed from all but one patient, in whom the wires were cut flush with the bone and left in situ.
When a proximal tibial or distal femoral allograft was used as a whole segment, it was attached as a composite with the implant. When present, the host femoral epicondyles were osteotomized to preserve a sleeve of tissue in continuity with the posterior aspect of the capsule for varus-valgus stability. After fashioning the whole-segment allograft to accept the selected component, the composite was positioned with the long stem of the component within the intramedullary canal to provide stability for the construct. When a femoral graft was used, stability of the knee was established in flexion and the length of the graft was adjusted by removing bone from its diaphyseal end to permit full extension of the knee without hyperextension. A step-cut was performed at the host-graft junction to provide rotational stability.
Thirty-one grafts were fixed to the component with bone cement and were stabilized to host bone with a stemmed component that bypassed the host-graft junction. Although bone cement was placed around the stem to include any modular junction, no attempt was made to plug the intramedullary canal or to pressurize the cement, except in one patient. We did not supplement fixation with a plate. A cortical strut allograft was used in one patient for ancillary support at the host-graft junction in order to enhance rotational stability. Autologous bone graft was not used in any of these knees.
In four knees, a custom component was press-fit without cement and a single femoral-head allograft was implanted. The graft was secured by the stem of the component and by the interference fit between the host bone and the graft (Table II).
Evaluation of the Patients
All patients were evaluated clinically and radiographically. The average duration of follow-up was fifty months (range, twenty-four to 120 months). Preoperatively and at every annual follow-up visit, scores were determined for each patient, according to the system of The Hospital for Special Surgery13 or that of The Knee Society12, and were recorded in a computer database.
Postoperative annual radiographs were analyzed by one of us (P. J. H.), who was not involved in the operation and was blinded to the clinical results. Radiographic analysis included inspection for the presence of radiolucent lines, osteolysis, subsidence of the component, time to incorporation of the graft (when possible), and any change in the position of the implant relative to the long axis of the bone. Incorporation of the graft was defined as the presence of trabeculae bridging the host-graft junction and the obliteration of the host-graft junction on two separate radiographic views. Incorporation of the femoral grafts was usually difficult to assess radiographically because the femoral component obscured the host-graft junction.
In 1995, all patients completed a questionnaire concerning their satisfaction, adapted from the Arthritis Impact Measurement Scales17. The responses to questions regarding pain in the involved knee and over-all function were incorporated into the results in the present study.
Clinical Data
The average range of motion of the knee for the thirty patients was 90 degrees (range, 27 to 121 degrees) preoperatively and 103 degrees (range, 71 to 128 degrees) postoperatively. The average preoperative score, according to the system of The Hospital for Special Surgery, for all thirty patients was 54 points (range, 29 to 82 points). Postoperatively, the average score was 74 points (range, 58 to 87 points) for four knees rated according to the system of The Hospital for Special Surgery and 86 points (range, 44 to 99 points) for twenty-six knees rated according to the system of The Knee Society. At the most recent clinical evaluation, twenty-six patients had a good or excellent score, according to the system of The Hospital for Special Surgery or that of The Knee Society.
The average preoperative score, according to the system of The Hospital for Special Surgery, for the four patients who had a primary total knee arthroplasty was 48 points (range, 37 to 57 points), and the average postoperative score, according to the system of The Knee Society, was 93 points (range, 85 to 97 points). These patients had an excellent clinical result (Figs. 2-A, 2B, and 2C). Of the twenty-six patients who had a revision arthroplasty, fifteen had an excellent result; seven, a good result; two, a fair result; and two, a poor result (Figs. 3-A, 3B, and 3C). The two patients who had a fair result had a moderate score (only 20 of 50 possible points) for pain. The two patients who received a poor clinical rating had pain, and they each had an extensor lag that had not been present preoperatively.
Three of the four patients who had been excluded from the study because of inadequate follow-up were contacted by telephone, and they reported no problems with the involved knee. Thirty-six months postoperatively, one of them sent radiographs that showed the femoral-head allograft in the medial tibial condyle to be intact and incorporated. Clinical and radiographic information was not available for the other excluded patients.
Radiographic Evaluation
Incorporation of the graft was demonstrated in twenty of the thirty patients, at an average of seven months (range, two to twenty-four months) postoperatively. The graft or grafts in the remaining ten patients were obscured radiographically by the femoral component. The serial radiographs showed that the position of the component had not changed by more than 1 degree in any knee. None of the patients had radiographic evidence of resorption of the graft.
No progressive radiolucent lines were seen around any component supported by a graft. The largest radiolucent line was two millimeters wide and was noted in four of six zones around a tibial component, but the lucency had not progressed during the three years before the most recent follow-up examination.
Three components that were seated on a structural allograft subsided during the period of the study; two were press-fit femoral components and one was a press-fit tibial component. During a ten-year period, one femoral component had subsided nine millimeters and the other, five millimeters. The tibial component had subsided seven millimeters over nine years. Two of the components had stabilized and had showed no additional signs of subsidence after two years. None of the components inserted with cement and with a structural allograft had evidence of subsidence on serial radiographs. No osteolytic lesions were seen around any component that was supported by a graft.
Subjective Evaluation
Twenty-eight of the thirty patients completed the subjective evaluation form. In response to the question addressing over-all function and satisfaction with the outcome of the total knee arthroplasty, eleven patients answered very well; ten, well; and seven, fair. None of the patients rated the result as poor or very poor. When asked to describe pain in the knee, eleven patients answered none; five, very mild; ten, mild; and two, severe.
Complications
Seven patients had a total of eight postoperative complications, and only one of the complications necessitated a reoperation. Deep venous thrombosis developed postoperatively in three patients, and phlebitis also developed in one of them. Staphylococcus haemolyticus grew on culture of specimens obtained at the time of arthroplasty from two patients. Both patients were managed with antibiotics for six weeks, and the components were retained. One of these two patients needed early manipulation under anesthesia within one month after the operation. The most recent clinical result for that patient was described as fair because of persistent moderate pain. The other patient had an excellent clinical and radiographic result at the most recent follow-up examination.
One patient had an episode of atrial fibrillation postoperatively; it resolved without sequelae. Finally, the only complication that necessitated a reoperation was in a patient in whom the pin that held the polyethylene insert in place dislodged two months postoperatively. The pin was replaced without changing the tibial insert. This patient had no other problems, and the clinical result, according to the system of The Knee Society, was excellent (a score of 98 points) at the most recent follow-up visit.
Contemporary components used for revision arthroplasty are designed with modular augments to replace deficient bone encountered at revision. Defects of bone, particularly in knees that have had multiple revisions, may exceed the size limitations available for these augments. Large metaphyseal defects of bone can be filled with polymethylmethacrylate, with or without the use of cancellous-bone or cortical reinforcement screws, allograft bone, or a linked prosthesis such as a rotating hinge or custom implant that replaces the missing segment of bone. We are aware of no outcome studies with long-term results that can be used for determining the best method for managing large defects of bone at revision knee arthroplasty.
Metaphyseal bone is a particularly attractive material for the repair of bone defects because of its cancellous structure, inherent strength, and osteoconductive properties. Allograft bone is available in a variety of sizes and shapes. Femoral heads obtained from osteoarthrotic hips at the time of total hip arthroplasty are a readily available source of bone graft. Pelker and Friedlaender showed that bone obtained from patients who are in their seventies maintains 70 to 85 per cent of its strength. They also maintained that the size and geometry of the graft are more important than the age or gender of the donor6,19.
Living donors of allografts can be retested for communicable diseases, so that the risk of transmitting a viral disease is effectively eliminated2. Large-segment allografts, such as distal femoral and proximal tibial grafts, are obtained from cadavera in a sterile environment and are sterilized with a low-dose gamma irradiation that does not adversely weaken the bone.
During a revision knee arthroplasty, allograft bone can be morselized or prepared in large chunks and packed into defects of bone whenever an intact peripheral rim of bone is present. However, bone graft prepared in such a manner provides minimum support for axial loads produced during weight-bearing activities. Fixation of the component is tenuous if the stability of the implant depends on support provided by a large volume of morselized graft. This fact may account for failures of fixation and the dislocations that have been reported after total knee arthroplasty with use of a morselized allograft24. A structural graft is a better option when the host bone is compromised. Intact bone graft retains anisotropic properties and provides both support and load transmission to adjacent host bone. Stability of the implant is less dependent on the prosthetic stem when a structural graft, which provides mechanical support, is used.
A structural femoral-head allograft was used in twenty-seven knees in the present study. Our method for preparing the graft with use of male and female-type reamers is distinct from other grafting techniques that have been reported7,23. The subchondral plate of a femoral head or the distal portion of a femur is removed easily with a so-called cheese-grater type of reamer. The reaming exposes a trabecular structure that rapidly unites with host bone by ingrowth of woven bone and that provides an ideal surface for interdigitation of cement between the graft and the implant. Our technique for preparing the graft and for achieving an interference fit of the graft, a long-stemmed component, and bone cement at the graft-prosthesis interface proved to be a reliable method for stabilization of the implant and the graft. Other authors have advocated invaginated grafts, fixation with screws, and supplemental fixation with a plate for rotational stability9,18,21,23. We do not favor either supplemental fixation with a plate or complete cementing of the stem of the component when a bulk structural allograft is used. These techniques reduce compressive loads at the graft-host junction and may result in non-union of the graft15.
A composite or femoral-head graft placed in the tibia was seen easily on anteroposterior and lateral radiographs. Therefore, we readily could assess trabeculae bridging the graft and incorporation of the graft on the two radiographic views. However, the host-graft junction was not visible on the anteroposterior radiographs made after the ten reconstructions of a defect in the distal aspect of the femur with a femoral-head allograft. Without clear anteroposterior and lateral radiographs of the graft, we could not confirm that it had been incorporated. In order to assess these ten femoral-head grafts radiographically, we measured the angle between the femoral component and the femoral axis and we assessed the alignment of the limb on the anteroposterior radiograph. We also examined lateral radiographs for evidence of radiolucent lines, osteolysis, and resorption of the graft.
Stemmed components have not been used routinely in studies of the repair of bone defects with allograft in revision total knee replacements21,25. A variety of prosthetic components, including short-stemmed components and hinged implants, were used in one study and may have accounted for fracture of the graft in two and infection in three of the twenty knees22.
Use of a long-stemmed component may be essential for the long-term success of an allograft in revision total knee arthroplasty. It seems unlikely that revascularization is a cause of collapse of the graft, as allografts become only slightly incorporated over a long period of time5,8,20. Although acetabular allografts used in conjunction with revision total hip arthroplasty were reported to have failed because of trabecular fracture, the grafts had not revascularized10,11. Collapse of the graft in such instances probably is due to trabecular fatigue and to the inability of the graft to repair and remodel. A stem reduces the stress on the graft and on the component-host interface1. Therefore, long-stemmed implants may extend the fatigue life of a graft by reducing stresses. We believe that the good clinical results demonstrated in our study, compared with the poor clinical results reported in studies of load-bearing acetabular allografts10,16, support this theory.
In conclusion, use of a structural allograft, particularly a femoral-head graft, in conjunction with a stemmed component inserted with cement provided an excellent option for the treatment of a large skeletal defect during both primary and revision total knee arthroplasty. The allografts in our patients had not deteriorated after an average of fifty months. Long-term follow-up studies are necessary to confirm the durability of such allografts used in conjunction with stemmed components in revision arthroplasty.
Brooks, P. J.; Walker, P. S.; and Scott, R. D.: Tibial component fixation in deficient tibial bone stock. Clin. Orthop.,184: 302-308, 1984.184302
1984
[PubMed]
Buck, B. E.; Malinin, T. I.; and Brown, M. D.: Bone transplantation and human immunodeficiency virus. An estimate of risk of acquired immunodeficiency syndrome (AIDS). Clin. Orthop.,240: 129-136, 1989.240129
1989
[PubMed]
Engh, G. A.: Bone defect classification. In Revision Total Knee Arthroplasty, pp. 63-120. Edited by G. A. Engh and C. H. Rorabeck. Baltimore, Williams and Wilkins, 1997.
Engh, G. A., and Parks, N. L.: The use of a bone defect classification system in revision total knee arthroplasty. Orthop. Trans.,18: 1138-1139, 1994-1995.181138
1994-1995
Enneking, W. F., and Mindell, E. R.: Observations on massive retrieved human allografts. J. Bone and Joint Surg.,73-A: 1123-1142, Sept. 1991.73-A1123
1991
Friedlaender, G. E.: Current concepts review. Bone grafts. The basic science rationale for clinical applications. J. Bone and Joint Surg.,69-A: 786-790, June 1987.69-A786
1987
Ghazavi, M. T.; Stockley, I.; Yee, G.; Davis, A.; and Gross, A. E.: Reconstruction of massive bone defects with allograft in revision total knee arthroplasty. J. Bone and Joint Surg.,79-A: 17-25, Jan. 1997.79-A17
1997
Goldberg, V. M., and Stevenson, S.: Natural history of autografts and allografts. Clin. Orthop.,225: 7-16, 1987.2257
1987
[PubMed]
Harris, A. I.; Poddar, S.; Gitelis, S.; Sheinkop, M. B.; and Rosenberg, A. G.: Arthroplasty with a composite of an allograft and a prosthesis for knees with severe deficiency of bone. J. Bone and Joint Surg.,77-A: 373-386, March 1995.77-A373
1995
Hooten, J. P., Jr.; Engh, C. A., Jr.; and Engh, C. A.: Failure of structural acetabular allografts in cementless revision hip arthroplasty. J. Bone and Joint Surg.,76-B(3): 419-422, 1994.76-B(3)419
1994
Hooten, J. P., Jr.; Engh, C. A.; Heekin, R. D.; and Vinh, T. N.: Structural bulk allografts in acetabular reconstruction. Analysis of two grafts retrieved at post-mortem. J. Bone and Joint Surg.,78-B(2): 270-275, 1996.78-B(2)270
1996
Insall, J. N.; Dorr, L. D.; Scott, R. D.; and Scott, W. N.: Rationale of The Knee Society Clinical Rating System. Clin. Orthop.,248: 13-14, 1989.24813
1989
[PubMed]
Insall, J. N.; Ranawat, C. S.; Aglietti, P.; and Shine, J.: A comparison of four models of total knee-replacement prostheses. J. Bone and Joint Surg.,58-A: 754-765, Sept. 1976.58-A754
1976
Jasty, M., and Harris, W. H.: Salvage total hip reconstruction in patients with major acetabular bone deficiency using structural femoral head allografts. J. Bone and Joint Surg.,72-B(1): 63-67, 1990.72-B(1)63
1990
Kraay, M. J.; Goldberg, V. M.; Figgie, M. P.; and Figgie, H. E., III: Distal femoral replacement with allograft/prosthetic reconstruction for treatment of supracondylar fractures in patients with total knee arthroplasty. J. Arthroplasty,7: 7-16, 1992.77
1992
[PubMed]
Kwong, L. M.; Jasty, M.; and Harris, W. H.: High failure rate of bulk femoral head allografts in total hip acetabular reconstructions at 10 years. J. Arthroplasty,8: 341-346, 1993.8341
1993
[PubMed]
Meenan, R. F.; Gertman, P. M.; and Mason, J. H.: Measuring health status in arthritis. The Arthritis Impact Measurement Scales. Arthrit. and Rheumat.,23: 146-152, 1980.23146
1980
Mnaymneh, W.; Emerson, R. H.; Borja, F.; Head, W. C.; and Malinin, T. I.: Massive allografts in salvage revisions of failed total knee arthroplasties. Clin. Orthop.,260: 144-153, 1990260144
[PubMed]
Pelker, R. R., and Friedlaender, G. E.: Biomechanical considerations in osteochondral grafts. In Bone and Cartilage Allografts: Biology and Clinical Applications, pp. 155-162. Edited by G. E. Friedlaender and V. M. Goldberg. Park Ridge, Illinois, The American Academy of Orthopaedic Surgeons, 1991.
Prolo, D. J., and Rodrigo, J. J.: Contemporary bone graft physiology and surgery. Clin. Orthop.,200: 322-341, 1985.200322
1985
[PubMed]
Stevenson, S., and Horowitz, M.: Current concepts review. The response to bone allografts. J. Bone and Joint Surg.,74-A: 939-950, July 1992.74-A939
1992
Stockley, I.; McAuley, J. P.; and Gross, A. E.: Allograft reconstruction in total knee arthroplasty. J. Bone and Joint Surg.,74-B(3): 393-397, 1992.74-B(3)393
1992
Tsahakis, P. J.; Beaver, W. B.; and Brick, G. W.: Technique and results of allograft reconstruction in revision total knee arthroplasty. Clin. Orthop.,303: 86-94, 1994.30386
1994
[PubMed]
Whiteside, L. A.: Cementless revision total knee arthroplasty. Clin. Orthop.,286: 160-167, 1993.286160
1993
[PubMed]
Wilde, A. H.; Schickendantz, M. S.; Stulberg, B. N.; and Go, R. T.: The incorporation of tibial allografts in total knee arthroplasty. J. Bone and Joint Surg.,72-A: 815-824, July 1990.72-A815
1990