JBJS | Instructional Course Lectures, The American Academy of Orthopaedic Surgeons - Femoral Bone Loss in Patients Managed with Revision Hip Replacement: Results of Circumferential Allograft Replacement*†

The Journal of Bone & Joint Surgery, Volume 81, Issue 3

Instructional Course Lecture | March 01, 1999

Instructional Course Lectures, The American Academy of Orthopaedic Surgeons - Femoral Bone Loss in Patients Managed with Revision Hip Replacement: Results of Circumferential Allograft Replacement*†

FARES S. HADDAD, B.SC., F.R.C.S.(ORTH)‡; DONALD S. GARBUZ, M.D., F.R.C.S.(C)‡; BASSAM A. MASRI, M.D., F.R.C.S.(C)‡; CLIVE P. DUNCAN, M.D., F.R.C.S.(C)‡, VANCOUVER, BRITISH COLUMBIA; CAROL R. HUTCHISON, B.SC., M.D., M.ED., F.R.C.S.(C)§; ALLAN E. GROSS, M.D., F.R.C.S.(C)§, TORONTO, ONTARIO, CANADA

View Disclosures and Other Information

An Instructional Course Lecture, American Academy of Orthopaedic Surgeons

The Journal of Bone & Joint Surgery. 1999; 81:420-36

5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case

text A A A

Introduction

Surgeons who perform reconstructive procedures about the hip are faced increasingly with complex cases in which severe loss of bone makes conventional revision techniques difficult or impossible^{18,47,48,51,90,91}. In such situations, one alternative is the use of a composite consisting of a proximal femoral allograft and a prosthesis to restore mechanical integrity to the proximal part of the femur. This technique has attracted increased attention because of its potential for reconstituting bone stock in young patients⁹⁶ and because the allograft-cement-implant composite is a strong reconstruction that does not require sacrifice of host tissue and theoretically improves the bone stock for future reconstructions. Allografts also have been used in revision hip arthroplasty to replace or reinforce the calcar^24,47,48,60 and to treat or prevent periprosthetic fractures^16,17,40,62. The present study concentrates on the indications, biological mechanisms, techniques, and results associated with use of circumferential proximal femoral allografts in the revision of a failed total hip replacement.

*Printed with permission of the American Academy of Orthopaedic Surgeons. This article will appear in Instructional Course Lectures, Volume 49, American Academy of Orthopaedic Surgeons, Rosemont, Illinois, March 2000.

†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. One of the authors (F. S. H.) received funds in support of this study. The funding sources were The Charnley Trust, The British Orthopaedic Association Wishbone Trust, and the Norman Capener Travelling Fellowship.

‡Division of Reconstructive Orthopaedics (F. S. H., D. S. G., and B. A. M.) and Department of Orthopaedics (C. P. D.), University of British Columbia, Laurel Pavilion, Room 3114, 910 West 10th Avenue, Vancouver, British Columbia V5Z 4E3, Canada. E-mail address for Dr. Haddad: fareshaddad@compuserve.com.

§Department of Orthopaedic Surgery, Mount Sinai Hospital, 600 University Avenue, Suite 476A, Toronto, Ontario M5G 1X5, Canada.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG1-A:: Figs. 1-A, 1-B, and 1-C: Schematic illustrations of proximal femoral reconstruction with use of a structural allograft. (Reprinted, with permission, from: Gross, A. E.: Revision arthroplasty of the hip using allograft bone. In Allografts in Orthopaedic Practice, p. 160. Edited by A. A. Czitrom and A. E. Gross. Baltimore, Williams and Wilkins, 1992.) Fig. 1-A: The allograft (left) is prepared distally with a step-cut. The prosthesis is fixed with cement in the allograft. The host bone (right) is split and retracted to form a vascularized bed for the allograft implant.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG1-B:: Fig. 1-B: The allograft-implant composite is inserted into the host femur with no cement distally or at the host-allograft junction. The residual proximal host bone is wrapped around the allograft-host junction to promote union.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG1-C:: Fig. 1-C: The greater trochanter and the host femur are wired to the allograft.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG2-A:: Figs. 2-A and 2-B: Anteroposterior radiographs made before and after revision with use of a structural femoral allograft reinforced with strut allografts at the graft-host junction. This procedure was performed according to the protocol that we described, with insertion of cement into the allograft proximally but not into the host bone distally. Fig. 2-A: Preoperative radiograph.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG2-B:: Fig. 2-B: Radiograph made 6.5 years after the operation.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG3-A:: Figs. 3-A, 3-B, and 3-C: Radiographs made before and after revision with a structural femoral allograft. Fig. 3-A: Preoperative anteroposterior radiograph.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG3-B:: Fig. 3-B Anteroposterior and lateral radiographs made six years postoperatively, showing wiring at the site of the step-cut osteotomy and union with florid callus formation. The greater trochanter had been resorbed but had not migrated. Some residual host bone, used as vascularized autogenous graft, can be seen proximally.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG3-C:: Fig. 3-C Anteroposterior and lateral radiographs made six years postoperatively, showing wiring at the site of the step-cut osteotomy and union with florid callus formation. The greater trochanter had been resorbed but had not migrated. Some residual host bone, used as vascularized autogenous graft, can be seen proximally.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG4:: Fig. 4 Radiograph made nine years postoperatively, showing trochanteric union and union between the allograft and the residual proximal host femur.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG5:: Fig. 5 Anteroposterior radiograph made four years postoperatively, showing severe resorption, down to the allograft-cement junction, in the absence of infection. Little more than the cement mantle remains proximally. The patient had no pain and was delighted with the reconstruction.

View Larger

TABLE I COMPLICATIONS AFTER REVISION OF A FEMORAL COMPONENT WITH USE OF A STRUCTURAL ALLOGRAFT

*The values are given as the number of allografts, with the percentage in parentheses.

Study	No. of Allografts	Duration of Follow-up (mos.)	Complication*
Nonunion	Dislocation	Infection
Gross and Hutchison⁵¹ (1998)	200	24	7 (4)	11 (6)	6 (3)
Masri et al.⁸⁹ (1995)	39	61	4 (10)	2 (5)	2 (5)
Chandler and Penenberg¹⁶ (1989)	30	22	2 (7)	5 (17)	1 (3)
Roberson¹¹¹ (1992)	24	48	2 (8)	1 (4)	2 (8)
Head et al.⁶⁰(1993)	22	28	2 (9)	5 (23)	0
Zmolek and Dorr¹²⁴ (1993)	11	48	2	6	1

Add to My JBJS

TABLE I COMPLICATIONS AFTER REVISION OF A FEMORAL COMPONENT WITH USE OF A STRUCTURAL ALLOGRAFT

*The values are given as the number of allografts, with the percentage in parentheses.

Study	No. of Allografts	Duration of Follow-up (mos.)	Complication*
Nonunion	Dislocation	Infection
Gross and Hutchison⁵¹ (1998)	200	24	7 (4)	11 (6)	6 (3)
Masri et al.⁸⁹ (1995)	39	61	4 (10)	2 (5)	2 (5)
Chandler and Penenberg¹⁶ (1989)	30	22	2 (7)	5 (17)	1 (3)
Roberson¹¹¹ (1992)	24	48	2 (8)	1 (4)	2 (8)
Head et al.⁶⁰(1993)	22	28	2 (9)	5 (23)	0
Zmolek and Dorr¹²⁴ (1993)	11	48	2	6	1

Indications

History

In 1908, Lexer⁷⁶ reported on what we believe was the first series of patients managed with allograft transplantation; twenty-three whole joints and eleven hemijoints were used. Lexer subsequently reported the results in 1925⁷⁷. After the excellent work of Bonfiglio et al.⁶, Chase and Herndon¹⁹, and Curtiss et al.²⁰ on the immunogenicity of frozen allografts, interest in the use of bulk allografts was rekindled in the late 1960s and early 1970s. This work centered on the reconstruction of defects after excision of a tumor^{101,102,105,106} and led to larger series, such as that reported on by Mankin et al.⁸⁵. With the success of such limb-sparing procedures, use of structural allografts soon attracted the interest of surgeons who performed hip arthroplasty in patients who had extensive bone loss^{7,58,59,81,100}. This work has been carried forward in a number of specialized units and represents one solution to the increasing problem of bone loss after hip arthroplasty^{1,16-18,46-51,60,61,65,89}.

Classification Systems

In order to better understand and to apply the various reconstructive options that are available, a comprehensive and easily applied classification system is of paramount importance⁹⁰. Such a system is also essential for the uniform reporting of the results of these various reconstructive options. Several systems have been proposed^{22,25-28,53,68,82,104}, but none has received universal acceptance. The emphasis of the different classification systems varies according to the objective that each was designed to achieve.

Three classification systems that could be used when use of a structural femoral allograft is being considered for a patient who has severe bone loss are those of Chandler and Penenberg¹⁶, Gross et al.⁴⁷, and the American Academy of Orthopaedic Surgeons Committee on the Hip²².

System of Chandler and Penenberg¹⁶: These investigators described a comprehensive system for the classification of femoral bone loss that was based on the outcome of revision hip arthroplasty for forty-three hips in thirty-seven patients who had been followed clinically for nineteen months. This classification system comprises six main categories, each of which may exist in isolation or in conjunction with another category. The six basic categories are calcar deficiency, trochanteric deficiency, cortical thinning, cortical perforation, femoral fracture, and circumferential deficiency.

System of Gross et al.⁴⁷: One of us (A. E. G.) and colleagues described a simple classification system that emphasizes the determination of the necessity for allograft bone in a reconstruction. Bone loss is classified into two main categories: intraluminal and cortical. The bone loss is considered intraluminal when the femur has a strong cortical shell but a widened canal that is devoid of cancellous bone. The cortical skeleton that remains usually is strong enough to support an implant. The bone loss is considered cortical when the femur has some loss of cortical bone. Cortical defects may be noncircumferential (and need only strut grafts) or circumferential (and need a segmental graft). The latter category is further subdivided into calcar defects and defects of the proximal part of the femur in which there is more than five centimeters of circumferential bone loss; the implication is that segmental defects that are less than five centimeters in length can be treated with a calcar-replacing implant, whereas those that are more than five centimeters in length must be treated with an allograft or a tumor prosthesis.

System of the American Academy of Orthopaedic Surgeons Committee on the Hip: D'Antonio et al.²² presented the collaborative efforts of the American Academy of Orthopaedic Surgeons Committee on the Hip in the classification of femoral deficiencies in patients being managed with total hip arthroplasty. The two basic categories of bone loss are segmental and cavitary. In addition, the Committee introduced categories for femoral ectasia, malalignment, stenosis, and discontinuity in order to provide a comprehensive system of classification for the full range of bone abnormalities found during hip arthroplasty. This classification system also encompasses the level and grade of bone loss. There are three femoral levels of bone loss. Level I is the area proximal to the inferior portion of the lesser trochanter; level II, the area between the inferior border of the lesser trochanter and ten centimeters distal to it; and level III, the area distal to level II. There are also three grades of bone loss. Grade I is assigned when there is minimum bone loss, host-prosthesis contact is maintained, and no bone-grafting is needed. Grade II is given when there is some loss of host-prosthesis contact but the host bone is still able to support the prosthesis and only morseled bone graft is needed. Grade III indicates that loss of host-prosthesis contact necessitates use of a structural bone graft (such as a proximal femoral allograft) for the reconstruction.

Treatment Options

Regardless of the classification system that is used, it is clear that femoral bone loss is a problem that continues to challenge orthopaedic surgeons. Moreover, this problem is likely to increase in the future as there is an association between the degree of bone loss and the number of previous total hip operations. A number of treatment options are available. These include excision arthroplasty, which is unacceptable when there is severe bone loss^45,57; arthrodesis, which is difficult to achieve when there is severe bone loss⁷³; disregarding and bypassing the proximal part of the femur with diaphyseal fixation of a component either with or without cement^26-28,104; reconstruction with onlay or inlay cortical allograft struts fixed without cement proximally or distally or fixed with cement (for example, repair of noncircumferential structural defects with use of strut grafts, which has had uniformly good results^49,62); impaction allografting for intraluminal and contained femoral defects⁴¹; use of a calcar-replacing prosthesis or a long-neck prosthesis⁸² for short circumferential defects that are less than five centimeters in length; and segmental replacement of the badly damaged proximal part of the femur with use of either metal or bone.

If segmental replacement is chosen, the options that are available to substitute for the missing bone include use of a massive proximal femoral replacement such as a tumor prosthesis, a more complex design of prosthesis that involves modularity or custom manufacture, or a composite consisting of a segmental allograft and a contained component^1,3,114. The disadvantages of use of a metal component alone include late fracture (due to the fatigue life of the metal), loosening (because the stem is fixed only distally), and instability (because of poor reattachment of the abductors).

Advantages of Structural Allografts

Use of an allograft-implant composite is attractive for a number of reasons. The allograft has an inherent capacity to incorporate and become an integral part of the femur, and the risk of instability should be lower because of the theoretically superior potential for soft-tissue attachment. The allograft provides a dry virgin lattice for good cementing, but it does not need to be fixed to the distal part of the host femur with cement; hence, it does not compromise subsequent revisions. Moreover, union with the remaining part of the femur decompresses the site of the distal fixation and may reduce the risk of loosening of the implant. Theoretically, the allograft also allows a more normal gradation of forces from the implant to the host bone and may be used, in some patients, with an implant of conventional design.

The precise indications for use of a structural allograft vary from center to center, but essentially they are a circumferential defect that is more than five centimeters in length; grade-III bone loss, according to the system of the American Academy of Orthopaedic Surgeons²²; some combined defects; and some cases of femoral discontinuity. In simpler clinical terms, these indications represent a failed tumor prosthesis, a failed calcar replacement with subsidence, a femur with little potential for bone ingrowth in its distal aspect (a femoral diaphysis that allows less than five centimeters of satisfactory bone growth into the distal part of the stem), or a femur that is too deficient or too patulous for bone-packing.

Disadvantages of Structural Allografts

Segmental replacement with an allograft-implant composite is technically demanding and expensive. It is a risky procedure that needs careful and constant appraisal. Loosening of an allograft-implant composite has not yet been proved to be less troublesome than loosening of a proximal femoral replacement.

The major disadvantages of structural allografts are the potential for the transmission of disease and the period of time necessary for healing at the graft-host interface. Moreover, the available allografts may not meet the requirements of the surgeon^38,70, and there is great variability in the quality of grafts, which is difficult to assess⁵.

The biological drawbacks of allograft bone, which are discussed later, may lead to nonunion and fracture or resorption of the graft.

In addition to transmission of the human immunodeficiency virus, which has attracted the greatest attention, the possible transmission of hepatitis and other viral agents should also be considered^98,99. Adherence to the standards of the American Association of Tissue Banks reduces the risk of transmission of the human immunodeficiency virus to less than one in one million^9,10. Irradiation of the graft further reduces this risk³¹, but radiation may weaken the allograft bone⁵⁴.

The late effects of antigenicity and immune rejection remain unclear¹⁰⁹. Our experience and that of other centers include some cases of resorption, but the etiology of this finding has yet to be fully clarified^65,89,111.

Biological, Biomechanical, and Immunological Characteristics of Allografts

Structural allografts have become a major part of the armamentarium of surgeons who perform reconstructive procedures. These allografts are particularly attractive because of their potential for reconstituting bone stock. Their use, however, requires specialized knowledge and technical skills as well as a thorough understanding of the biological, immunological, and biomechanical characteristics of allografts. A detailed description of the biological characteristics of allograft bone has been presented previously³⁹, but some important features will be highlighted in the present report.

An understanding of allograft incorporation and the various mechanisms that can alter this biological process is critical to the successful application of allografts in revision joint arthroplasty. The healing and incorporation of allografts is very similar to the healing of autogenous grafts except for the absence of donor cells that can contribute to healing⁴³. The healing (union) of cortical allografts therefore is generally slower than that of equivalent autogenous grafts. When a massive cortical graft is used in revision arthroplasty, the host bone has to unite to the allograft for a successful clinical result but the allograft does not have to, and will not, incorporate or remodel completely^29,30,44. Despite the absence of biological incorporation, an allograft construct can be a clinical success when it unites and is supported by internal fixation. Complete incorporation may be a success from a biological point of view, but it may not be necessary or desirable for clinical function. In fact, it is well recognized that revascularization increases the porosity and decreases the strength of an allograft^14,15 and that this increases the risk of fatigue fractures in the allograft^4,29.

Structural allografts have limited biological activity. The first step in the healing of a structural allograft is the inflammatory response. This response brings in the pluripotential cells required for new-bone formation. The union and incorporation processes are initiated by osteoclasts, which resorb the haversian systems of the allograft until osteoblasts appear and fill this area. The new-bone formation is derived from the endosteal lining of the host and from the surrounding soft tissues. This is a slow process that can take as much as eight times as long as the healing of an autogenous bone graft in rats^14,15. In most cases, union then occurs at the allograft-host junction. Rigid fixation appears to promote union^85,105,106. The progression to union is highly dependent on the host. Revascularization, creeping substitution, and remodeling occur to a very limited degree in processed allografts^15,116. Many factors have been implicated in healing, including the immune response to the allograft, the mechanical response, and graft-host stability^2,85. Despite these multiple confounding factors, experiments that have controlled for them have shown that the host will unite to structural allograft bone but that cortical allografts lack the ability to remodel and rely on internal fixation for clinical function^12,115,118.

Retrieval studies have provided great insights into the biological behavior of frozen allografts in humans. Enneking and Mindell studied sixteen massive retrieved allografts³⁰. They noted that union occurred slowly at the graft-host junction by the formation of external callus from the host. In addition, they reported that remodeling took place at the superficial ends of the graft and involved less than 20 percent of the graft. Incorporation of an allograft may be undesirable because incorporation of cortical bone involves vascularization of that bone, which leads to the problems of weakening, fracture, and resorption. An important clinical finding in the study by Enneking and Mindell was that soft tissues became firmly attached to the graft by a seam of new bone.

Frozen structural allografts primarily possess the osteoconductive, and not the osteoinductive, property of bone grafts. They are biologically inert and therefore function as implants. This has important biological and biomechanical consequences. In terms of structural function, acellular dead allograft supported by internal fixation is biomechanically as strong as living bone because it is the mineralized matrix of bone that serves this structural function. Therefore, an allograft-prosthesis composite, such as would be used in the reconstruction of a massive proximal femoral defect, can be considered a clinical success when union has occurred at the graft-host junction. Achieving union can be difficult because the allograft is only osteoconductive. The rate of union may be improved by placing autogenous bone graft at the graft-host junction. The autogenous bone, which can be obtained from the iliac crest or from the residual host femur, can be wrapped around the junction as has been suggested by one of us (A. E. G.) and colleagues^50,51. This autogenous bone graft possesses cells capable of osteogenesis and proteins capable of osteoinduction.

Once union has been achieved at the graft-host junction, allograft composites function well clinically despite their inability to remodel. This inability to remodel has important clinical implications in terms of methods of fixation of the allograft to the host bone and is responsible for the fatigue fractures seen in large weight-bearing allografts. Holes in allografts and the use of plates have been shown to increase the risk of fracture in allografts^121-123. Intramedullary fixation decreases the risk of fracture. When allografts are used in femoral reconstruction, the prosthesis serves as the intramedullary fixation device¹¹⁰. The prosthesis should completely span the allograft because any stress-riser eventually may lead to fracture in this biologically inert implant.

The incorporation of an allograft is therefore a very complex and limited process that is dependent on many factors. Clinical success ultimately depends on the graft providing mechanical support for the skeleton. Large cortical grafts in adults do not become completely incorporated (that is, replaced by new bone derived from the host). Of the many variables that can modulate the incorporation process, the immune response is increasingly recognized as the fundamental factor that leads to the success or failure of an allograft.

Fresh allograft bone evokes an immune response in the host. This immune response can result in resorption of the graft or in a marked delay in the incorporation of the graft. For this reason, techniques have been developed to decrease the immunogenicity of transplanted allograft bone. The most common clinical techniques that are used today are deep-freezing and freeze-drying. These methods allow for long-term preservation and storage of allograft bone.

Freeze-dried bone is weaker in bending and torsion than fresh-frozen bone^69,107,108, but it has the advantage of being less antigenic³². Makley found no difference in the time to osseous union or to incorporation between freeze-dried and fresh-frozen grafts⁸³. Originally, these techniques also were thought to eliminate the immune response of the host. Because these processed allografts had no living cells, it was thought that they would not elicit a strong immune response when transplanted^107,108.

With the advent of modern immunological assay techniques, it has become apparent that fresh and processed allograft bone can evoke an immune response^33-37. In one study by Friedlaender et al., both deep-frozen and freeze-dried allograft bone elicited a humoral and cell-mediated immune response in rabbits³². However, both methods of processing decreased this response in comparison with the response to fresh allogenic bone. Goldberg and Stevenson suggested that the immune response to allograft bone diminishes osteoinduction by destroying or displacing progenitor cells⁴³. Although most investigators would agree that processed allograft bone elicits a host immune response^32-37,74, the effect on incorporation of the graft and any subsequent resorption is unclear¹¹². It has been suggested that this immune reaction may play a role in the unpredictable outcome associated with allografts in revision arthroplasty².

The major histocompatibility complex class-I and class-II antigens on specialized antigen-presenting cells are responsible for the immune response to allograft bone^21,64,116. Bos et al. studied the effect of histocompatibility matching on the incorporation of frozen bone transplants in rats⁸. Their study confirmed that frozen bone elicits a strong immune response in rats when there is a major histocompatibility difference between the donor and the recipient and that the immune response is almost undetectable when the donor and the recipient are closely related. Despite these marked differences in immune response, no effect was seen on the biological process of graft incorporation. Two other studies addressed the effect of freezing and histocompatibility matching on the incorporation of cortical bone grafts in weight-bearing sites in rats^115,118. Incorporation was followed sequentially over time. Those investigators concluded that histocompatibility matching and freezing were the most important determinants of incorporation of the graft. Frozen grafts with a major mismatch with the recipient, the most commonly used grafts in clinical practice, were unpredictable in terms of incorporation, with a trend toward less revascularization and more failure. Other studies of animals have confirmed the role of major histocompatibility complex matching in the incorporation of bone graft¹¹⁷. In this setting, mechanical rigidity appears to play an additional independent role^19,74,94.

Studies with use of immunosuppressive agents have provided additional evidence that the immunological mechanisms play a critical role in the incorporation of bone graft. In a study in which cyclosporine was used for immunosuppression in rats with a major histocompatibility mismatch, the incorporation of allografts was found to be similar to that of autogenous grafts⁷⁵. Burchardt et al. reported similar results with use of azathioprine immunosuppression in dogs^11,13. At the present time, immunosuppression has no role to play in revision hip arthroplasty because of the serious side effects of these medications.

The effect of human leukocyte antigen-matching on the immune response to allograft bone has been examined in a few studies⁹⁵. A recent multicenter study showed sensitization to human leukocyte antigen after the transplantation of frozen allograft bone¹¹⁹. Whether this immune response had any biological or clinical effects was not ascertained. Friedlaender also documented sensitization to human leukocyte antigen in nine of forty-three patients who had received a freeze-dried allograft; eight of those patients had no evidence of any untoward sequelae on follow-up³⁴. Muscolo et al. studied the effect of human leukocyte antigen matching on the radiographic incorporation of allografts⁹⁷. The data showed a higher radiographic score for patients who matched with the donor for at least one class-I human leukocyte antigen compared with that for patients who were totally mismatched. However, this difference was not found to be significant. Thus, immune recognition plays a role in the incorporation of a bone graft, although the exact nature of the interplay between the immune system and the physiology of bone has yet to be determined. The ideal situation would be to eliminate antigenicity and maintain biological activity.

Technique

Careful planning is necessary for a proximal femoral replacement with a structural allograft. This planning includes detailed templating, procurement of a suitable allograft, determination of the availability of an appropriate implant and familiarity with its use, and ensuring that a wide range of equipment that may be needed during the procedure is readily available. The intraoperative techniques include a versatile approach, precise measurement and preparation of the allograft, suitable fixation of the component to the allograft, and then fixation of this composite to the remaining part of the femur. The experience at a number of centers suggests that, ideally, proximal femoral replacement with a structural allograft should include cement for fixation of the implant to the allograft but not for fixation of the graft to the host, a stable graft-host junction either through press-fit fixation (with use of modular components) of the composite to the remaining part of the femur or through step-cut or oblique preparation of the allograft-host junction, and preservation of the bivalved proximal part of the host femur as a vascularized autogenous graft wrapped around the allograft and the graft-host junction.

Approach

The operative approach to the hip is either through a trochanteric slide^16,17,92 or through a transtrochanteric approach. The trochanteric slide is probably a better approach as it decreases the risk of trochanteric escape (proximal migration of the trochanter as a result of nonunion at the site of the osteotomy). A transverse osteotomy may be indicated when more than six centimeters of lengthening is needed; in such cases, the tension in the sciatic nerve must be carefully observed both during and at the end of the procedure, and a wake-up test may be indicated. If it is, the anesthesia is lightened such that function of the sciatic nerve can be assessed before closure. An extended trochanteric osteotomy also may be a suitable option in certain circumstances. The femur can then be split longitudinally to the distal extent of its incompetent part. Extensive stripping of the femur is avoided as this bone will subsequently act as a vascularized autogenous graft.

The acetabulum should be reconstructed first, with care taken not to lever on the weakened femur or to manipulate forcibly as even gentle movement may lead to a fracture when there is severe bone loss. The length of the allograft that is necessary is assessed in vivo by placing the femoral implant into the host bone and reducing it into the trial cup. The selected length depends on stability and limb-length discrepancy.

Selection and Preparation of the Graft

Fresh allografts are rejected by the host immune system. The initial response to fresh allografts is inflammation followed by a marked delay in incorporation of the graft or by complete resorption of the graft²¹. Therefore, the allografts in clinical use are processed. The most common methods of processing are freezing and freeze-drying. These techniques allow long-term preservation of the graft. Bone frozen at -70 degrees Celsius has a shelf life of five years²¹. These techniques have been shown to decrease or to eliminate the immunogenicity of bone allografts but at the price of decreasing their biological activity by removing all living cells. Irradiation of the allograft reduces the risk of viral transmission³¹ and theoretically may further decrease its immunogenicity.

At the time of reconstruction, the segmental allograft should be brought into the operating room only after infection has been ruled out as the cause of prosthetic failure and bone loss. The allograft is unwrapped, specimens are obtained from it for culture, and it is then placed in a warm antibiotic-containing solution.

Interface of the Allograft and the Prosthesis

There is good evidence, both theoretical and clinical, that the prosthesis should be fixed to the allograft with cement. Bone ingrowth cannot occur within the allograft, and a prosthesis inserted without cement therefore is likely either to subside or to split the calcar of the allograft^16,17,124. Moreover, it is difficult to shape the graft to the prosthesis without weakening it. The allograft bone provides a good surface for cement interlock and has not been, in our experience, a source of membrane formation. Prophylactic antibiotics should be added to the bone cement¹⁰³. The results to date have shown a stable bone-cement interface. The host canal is nearly always larger than the allograft⁶⁶; it is important to resist the temptation to overream the allograft in order to obtain a press-fit of the femoral component into the host. A long, thin femoral component or a modular femoral component should be used so that the allograft does not have to be excessively reamed.

Junction of the Graft and the Host Bone

In order to have a successful outcome with a large structural allograft, union at the allograft-host junction is critical and is mainly dependent on the host. It is therefore important to obtain stability at the allograft-host junction and to use the residual part of the host femur as an autogenous graft at the junction. The optimum technique for fixation of the allograft to the host bone remains controversial. A rigid junction favors union of the allograft and the host bone. One of the two most accepted techniques is a step-cut or an oblique cut, either of which is reinforced with cerclage wires, at the allograft-host junction to obtain rotational stability^{46,47,50-52,87}. The other is the use of a modular long-stem component that gains fixation to host bone by a distal press-fit¹⁷. Other solutions that have been attempted include the use of double plates^16,17,85,113, allograft struts¹⁶, and cementing of both the prosthesis into the allograft and the allograft into the host^60,89. Screws placed through a plate may interfere with the stem within the allograft and lead to stress-risers that increase the risk of fracture of the allograft¹²¹. Struts may weaken before union is achieved¹⁷.

The implications of rigid distal fixation are still unknown. If adequate stability of the junction is achieved, a tight distal press-fit may not be necessary⁴⁹. Although biomechanical tests on cadavera have shown good stability with use of cement in the distal part of the host femur^71,72,86, such cementing may increase the risk of nonunion and may not achieve good fixation as the distal part of the femur is often eburnated by the previous prosthesis. In some cases, however, when the junction is beyond the isthmus of the host femur, use of cement or interlocking screws may be necessary in order to provide primary distal stability²⁴.

When cement is not used in the distal aspect of the host bone, the allograft has the potential to load-share once union has been achieved. Moreover, the distal aspect of the femur can still support another reconstruction if one becomes necessary. At the time of reconstruction, meticulous attention to keeping the graft-host interface free of cement, the use of a step-cut or oblique cut, adjunctive fixation such as strut allografts and cerclage wires, the preservation of the maximum possible vascularity in the host femur, and autogenous bone-grafting all help to ensure union. After gaining initial fixation, any residual proximal host femoral bone should be wrapped around the allograft-host junction to function as autogenous bone graft to promote union.

If the remaining host trochanter is substantial, osseous reattachment can be attempted either with wires or with use of a cable-grip system. This technique provides better soft-tissue support and may limit the considerable risk of dislocation, but late trochanteric fracture is common. Reattachment of the abductors, either to the fascia lata or to the allograft, is worthwhile even when the trochanter is deficient.

Postoperative Care

The perioperative management includes appropriate prophylactic antibiotic therapy as these are time-consuming, lengthy, complex reconstructions in patients who may already be predisposed to infection. The intraoperative cultures of specimens from both the host and the allograft must be checked.

The patients are not allowed unprotected weight-bearing until there is good radiographic evidence of allograft-host union; this usually takes between three and six months. For some time during this period, external stabilization of the hip may be necessary if the joint is unstable and the abductors should be protected (either by avoiding active abduction of the hip for six to twelve weeks or with the use of braces) in order to assist union of the greater trochanter and stabilization of the soft tissue¹²⁴. If the risk of postoperative instability is high, consideration should be given to a socket that captures the femoral head^42,78.

Our Preferred Technique

In our practice, the graft is usually stored in the hospital bone bank at -70 degrees Celsius after irradiation with 2.5 megarad (25,000 gray). At the time of reconstruction, the segmental allograft is brought into the operating room only after infection has been ruled out. If there is any doubt about the potential for preoperative infection, a staged procedure is performed. The allograft is unwrapped, specimens are obtained from it for culture, and it is then placed in a warm solution of Betadine (povidone-iodine). Preparation of the allograft is carried out on a separate back table.

A transtrochanteric approach is used because of the need for an extensive exposure. A sliding trochanteric osteotomy is performed because it is associated with a lower risk of trochanteric escape than is the more traditional transverse osteotomy⁶⁵. The proximal part of the femur is exposed by reflecting the vastus lateralis off the intermuscular septum anteriorly, with care taken to avoid extensive stripping of the remaining part of the femur as it subsequently acts as a vascularized autogenous graft. A transverse cut is made distally at the level of good-quality bone, but it is completed through only half of the circumference so that a step-cut can be fashioned later. The femur then is split longitudinally down to the transverse cut and pried open, exposing the implant and the cement. Copious irrigation is used throughout the procedure.

The acetabulum is reconstructed first. The length of allograft that is needed is determined in vivo by placing the femoral implant into the host bone and reducing it into the trial cup. This length is compared with the previously templated measurements. The selected length ultimately depends on stability and limb-length discrepancy.

A thin long-stem femoral component, which is narrow proximally and designed to be used with an allograft, is employed so that the allograft does not have to be excessively reamed. The distal part of the host femur is then reamed gently over a guide-wire to assess the size of the canal for the implant rather than to enlarge the canal. When the reamers are of a size with which definite reaming is taking place, that diameter of implant is selected. The allograft is then reamed and broached until a good fit for the implant is achieved. It is important not to overream the allograft in an attempt to achieve a press-fit of the femoral component into the host. The host canal is nearly always larger than the allograft, and thus it is usually impossible to achieve a press-fit into the host without overreaming the allograft, which weakens it. We therefore fix the implant to the allograft with cement and use a step-cut or an oblique cut of approximately two by two centimeters at the junction of the host and allograft to stabilize the whole construct. If there is a large discrepancy between the diameters of the host femur and the allograft, a step-cut may be technically difficult and stabilization may have to be achieved by other means. Cortical struts with circumferential wires and autogenous bone-grafting may be used or, in some cases, the allograft may be telescoped inside the host for a distance of one or two centimeters (intussusception) (Figs. 1-A, 1-B, and 1-C).

Wires are passed through and around the allograft for trochanteric reattachment. The prosthesis is fixed to the allograft with cement on a back table, but it is not fixed to the host femur with cement. Great care is taken to keep cement off of the interface that will be in contact with the host bone distally. After initial fixation is gained with the step-cut and cerclage wires, the remaining part of the host femur should be wrapped around the allograft-host junction to function as autogenous graft to promote union and to improve rotational control. Cortical strut grafts are used to reinforce any stress-risers or windows and to stabilize the junction further. These are wired into position, and autogenous bone-grafting is used, if possible, to encourage union.

The perioperative management includes intravenous administration of antibiotics for three days and then oral administration of antibiotics for five days, although the duration of antibiotic prophylaxis is currently under review. A cephalosporin is usually prescribed, and gentamicin is added if the patient needs urinary catheterization. Weight-bearing is not allowed until there is good radiographic evidence of union at the allograft-host junction. We have had satisfactory results with this technique (Figs. 2-A, 2-B, 3-A, 3-B, 3-C and 4).

Clinical Studies

Concerns

The use of structural allografts in revision hip arthroplasty remains controversial. Although early clinical results have been successful, the long-term results have been inconsistent. The interpretation of reports of clinical studies of the use of structural femoral allografts remains difficult. There is still no universally accepted common language to define the femoral deficiencies being addressed, and there is a wide variation in the indications for allograft reconstruction. The types of allograft used, their preoperative processing, and their operative preparation can differ markedly between centers. Moreover, a wide range of prostheses has been employed, with no uniformity in the method of fixation either within the allograft or to the host. The criteria for follow-up evaluation also differ between studies, and the interpretation of complications such as trochanteric escape or resorption may be very dependent on the observer. In the assessment of the outcomes of reconstruction of the proximal part of the femur with a circumferential allograft, a number of confounding factors render the interpretation and comparison of data from different centers difficult. These factors include the degree of immunological matching between the graft and the host; the extent of muscle and soft-tissue coverage remaining; the amount of host bone available to act as autogenous graft; the vascularity of the host bed, which may be influenced by factors such as smoking and diabetes; and the amount of dissection and soft-tissue stripping that is necessary in each reconstruction. Overall, the literature reflects the fact that these patients represent the most complex cases of reconstruction of the hip and that a variety of techniques continue to be used while more long-term data and a clearer understanding of the role of structural allografts in femoral reconstruction are awaited.

To our knowledge, all clinical studies to date have involved series of patients from centers in which the personnel have had expertise in the use of allograft reconstruction. Most of these series have included a small number of patients and relatively short-term follow-up.

The results for twenty-nine patients (thirty hips) who had been followed for a mean of twenty-two months were reported by Chandler and Penenberg¹⁶. Their technique included use of a long-stem modular prosthesis fixed with cement to the proximal femoral allograft and press-fit into the distal part of the host femur. The mean hip score improved from 35 to 78 points. There were two nonunions, five dislocations, and one deep infection as well as one hip that had resorption of the graft. Three hips had trochanteric escape. Four hips (13 percent) needed an additional operation.

Head et al. reported the results for twenty-two of their first twenty-five patients who had been managed with a proximal femoral allograft at the time of revision hip arthroplasty and who had been followed for a mean of twenty-eight months⁶⁰. This heterogeneous group of reconstructions included ten in which the graft had been fixed with cement into both the proximal and the distal part of the femur, three in which it had been fixed with cement only in the distal aspect of the femur, and nine in which it was fixed without cement. Sixteen patients (73 percent) had a good or excellent result, two had a fair result, and four had a poor result. Two hips had a nonunion with a stable implant, and one had a nonunion with partial resorption of the allograft. Complications included dislocations in five patients, but there were no infections.

The results for twenty-one patients (twenty-four hips) who had had reconstruction with a structural femoral allograft after a failed total hip replacement and had been followed for a mean of four years were described by Roberson¹¹¹. According to the Harris hip rating⁵⁶, twelve patients had a good or excellent result, six had a fair result, and only two had a poor result. The result was not given for one patient. The complications included two nonunions, six trochanteric nonunions, two deep infections, and one dislocation. Five hips had resorption of the graft.

Mankin et al. reported that sixty-four (70 percent) of ninety-one hips had a good or excellent result a minimum of two years after resection of a tumor and reconstruction with an allograft-implant composite⁸⁵. Failures usually occurred within two years. When early major complications were avoided, the results remained stable at eight, nine, or ten years. The longevity of successful reconstruction with an allograft-prosthesis composite after resection of a tumor has also been noted in other reports⁹⁶.

The results at a mean of four years after eleven reconstructions with a proximal femoral allograft-implant composite were reported by Zmolek and Dorr¹²⁴. Nine hips had radiographic union at a mean of thirteen months, and two had failure because of nonunion. Six of the eleven procedures were complicated by at least one episode of postoperative dislocation.

Martin and Sutherland reported on four hips that had a total of six major complications because the structural allograft had acted as an unsupported weight-bearing device⁸⁸. These complications included three nonunions, two fractures of the allograft, and one deep infection. In all of these hips, however, a short-stem component had been fixed with cement into the allograft, which had been secured to the host with two plates. Their findings reinforce the need for stable, durable fixation between the allograft and the host and the importance of avoiding use of fixation screws in the allograft. These goals are usually achieved with insertion of a long-stem prosthesis into the host femur.

In a recent review of the experience at Rush Medical College in Chicago, Illinois, Rosenberg and Jacobs reported on sixty-four patients who had had a total of sixty-six structural allografts¹¹³. Six patients were excluded from the study as they had died within two years, leaving fifty-nine hips that had been followed for a mean duration of 8.5 years. The procedures included a variety of techniques; a short-stem component was used in twenty reconstructions, a long-stem component was used in thirty-five, and intussusception was used in four. There were six nonunions, and three patients later had a fracture at the host-allograft junction. Ten hips (17 percent) needed a reoperation. The authors confirmed several previous observations that cement at the graft-host interface delays healing, reconstructions with a short-stem component fail more frequently than those with a long-stem construct, and the creation of screw-holes within the allograft leads to stress-risers and substantial weakening of the allograft.

The experience of the Division of Reconstructive Orthopaedics at the University of British Columbia includes sixty reconstructions to date: fifty-eight for a failed hip replacement and two for the treatment of a neoplasm in the proximal part of the femur⁸⁹. Forty-three patients were followed for more than two years, and forty-one of them were available for follow-up (one patient was lost to follow-up after returning to Europe, and one died); the mean duration of follow-up was four years (range, two to seven years). The mean Harris hip score⁵⁶ was 83 points (range, 5 to 95 points). The complications included a late fracture of the allograft in one patient (2 percent), a deep infection necessitating removal of the allograft in two patients (5 percent), nonunion of the host-allograft junction in four patients (10 percent), severe resorption of the allograft in ten patients (24 percent), and nonunion of the greater trochanter in eleven patients (27 percent). A major revision of the allograft was done in five patients (12 percent) because of a fracture, nonunion, infection, or resorption. Another five patients (12 percent) needed a reoperation for bone-grafting or treatment of a dislocation, leading to a total rate of reoperation of 24 percent.

One of us (A. E. G.) and colleagues reviewed the results for 168 hips that had been treated with a proximal femoral allograft-prosthesis composite and were followed for a mean duration of 4.8 years⁴⁹. One hundred and thirty of them were followed for more than two years. The mean Harris hip score⁵⁶ increased from 30 to 66 points, and only 10 percent (seventeen hips) needed another operation. Five hips had a nonunion, eight had a dislocation, and three had a deep infection. Six hips also had resorption of the graft. These results were incorporated into a later series by one of us (A. E. G.) and Hutchison, who reported on 200 reconstructions, performed no later than January 1996, with use of a circumferential allograft that was more than five centimeters long⁵¹. In that study, twenty-five hips (13 percent) needed a revision procedure. Two of the revision procedures were unsuccessful and were followed by an excision arthroplasty. The indications for revision were instability (eleven hips), infection (six hips), nonunion (seven hips), and loosening (one hip). All nonunions were treated successfully with use of a plate and autogenous bone-grafting, leaving the allograft-implant composite in situ. Radiographs showed union as early as six weeks after the procedure, but typically it occurred between three and six months. Resorption of the graft was identified in six hips. The resorption occurred on the periosteal surface of the graft, was never full-thickness, and was less than one centimeter long in all but one hip, in which it was four centimeters long. No revisions were performed because of resorption of the graft. With success defined as an increase in the clinical score of 20 points, a stable implant, and no additional operations related to the allograft, one of us (A. E. G.) and colleagues reported a rate of success of 85 percent (111 of 130 patients) at a mean duration of follow up of 4.8 years (minimum, two years)⁴⁹. Trochanteric escape of more than one centimeter was seen in thirty-three (25 percent) of 130 patients, although it should be noted that a transverse osteotomy had been performed in these hips. In a study with a longer minimum duration of follow-up of five years (mean, nine years), one of us (A. E. G.) and Hutchison also reported a rate of success of 85 percent (fifty-five of sixty-five hips)⁵¹.

Complications

Although a high rate of satisfaction has been reported in most studies of revision hip arthroplasties performed with use of an allograft, a great deal of emphasis has been placed on the postoperative morbidity. Use of allografts in patients with cancer has been associated with high rates of infection, fracture, and nonunion^{69,79,80,85,96}. Likewise, the most commonly reported complications associated with use of an allograft in femoral reconstruction after a failed total hip arthroplasty are nonunion, dislocation, and infection (Table I).

Infection

The prevalence of infection associated with structural allografts has been reported to be as high as 13 percent (twelve of ninety-one) in a series of patients with cancer who may have been immunosuppressed^80,85,120, but it has also been as low as 4 percent (six of 168)^51,65. Factors that have been implicated include the duration of the procedures, the extensive dissection of soft tissue, contamination of the graft, and the large loss of blood in these patients^48,120,122. Although the etiology of infection in these patients is undoubtedly multifactorial, contamination of the graft remains a problem²³.

The emphasis must be on prevention, with careful screening of patients both preoperatively and intraoperatively to exclude a preexisting infection. The allograft must be obtained from a recognized bone bank where standard preparation protocols are followed, and specimens from the allograft always should be cultured before insertion of the graft. Prophylactic antibiotics should be used both systemically and in the bone cement within the allograft¹⁰³. The procedures should be performed in specialized centers where the personnel have the greatest experience with the use of allografts and where the operative time and blood loss may be minimized. Intraoperative irrigation with Betadine (povidone-iodine) or antibiotics, or both, should be considered. When infection does develop, a standard two-stage protocol with the reinsertion of an additional allograft can be used^49,65.

Instability

The rate of dislocation is higher if only the femoral component is revised⁴⁹. We do not, however, necessarily recommend revision of a well fixed cup. A number of investigators have recommended postoperative bracing with use of a hip orthosis^82,124. A large proportion of the instability after proximal femoral reconstruction with use of a structural allograft may be related to the lack of soft-tissue stabilization, particularly when there is trochanteric migration.

Nonunion

Incorporation does not occur in the absence of union^14,15, although incorporation is not a prerequisite for clinical success. In a review of the results for 264 patients who had been managed with a humeral or femoral allograft after resection of a tumor, Tan and Mankin found that increased perioperative blood loss and blood-transfusion requirements had a significant effect on the time to osseous union of the graft and the host (p < 0.001)¹²⁰. The rate of nonunion is variable and has been as high as 23 percent (five of twenty-two)⁸⁸ and as low as 4 percent (seven of 168)⁴⁹. A number of these nonunions are related to interposition of cement at the graft-host junction¹⁶ and possibly to devascularization of the host femur when double plates are used. Insertion of cement in the distal part of the host femur may also distract the graft-host junction and prevent or delay union^55,63. Autogenous bone-grafting and use of a plate appears to be a reliable treatment for nonunion^51,85.

Trochanteric nonunion is common but leads to problems only when the trochanter escapes. It is noteworthy that trochanteric nonunion is not associated with health-status measures or functional outcome⁶⁵. The high rate of trochanteric nonunion may be related to the fact that this junction is under distraction rather than compression. Moreover, the blood supply to the junction of the trochanter and the graft comes from only the greater trochanter. The resulting instability presents a difficult management scenario, which is best avoided with use of a trochanteric slide rather than a transverse osteotomy. As a result, the distal pull of the vastus lateralis is maintained by good operative stabilization of the trochanter and by judicious postoperative precautions with use of active mobilization.

Fracture

The rate of fracture appears to be higher when reconstruction is performed after resection of a tumor^4,85. These fractures frequently occur between two and three years after the reconstruction^4,88, which reinforces the role of weakening of the graft after incorporation or fatigue of the graft after nonunion.

An understanding of the inert nature of allografts and the fact that any stress-riser, such as a screw-hole, does not remodel will help a surgeon to avoid the devastating effects caused by a fracture of the allograft^88,121. The technique that we have described emphasizes preservation of the intact allograft and avoidance of excessive reaming, which weakens the construct. Fractures nevertheless still may occur if the host bone is weak⁹³.

Resorption

A high rate of resorption of the graft was not recognized in the initial reports on allografts^{7,16,17,47,48,58-60,81,100}, but this is a potential problem that warrants careful surveillance of the graft^65,89,111 (Fig. 5). There is no universally accepted system for measuring the resorption of proximal femoral allografts.

In the classification system of the University of British Columbia, the allograft was divided into five zones similar to those used in the system of Gruen et al. for total hip arthroplasty⁵². Zones 1 and 4 were excluded because of the absence of an allograft trochanter (zone 1) and because of the allograft-host junction (zone 4). In each zone, resorption of allograft bone was categorized as involving less than one-third of the cortex, between one-third and two-thirds of the cortex, more than two thirds of the cortex, or complete resorption of bone down to the cement mantle. The severity of bone resorption was arbitrarily graded as mild (resorption of less than one-third of the cortex in one zone), moderate (resorption of less than one-third of the cortex in two zones or between one-third and two-thirds of the cortex in one zone), or severe (resorption of less than one-third of the cortex in more than two zones, between one-third and two-thirds of the cortex in two zones or more, or resorption of more than two-thirds of the cortex or complete resorption in any zone). According to the classification system of Mount Sinai Hospital in Toronto, mild resorption was defined as partial-thickness resorption of less than one centimeter in width and length; moderate, as partial-thickness resorption of more than one centimeter in width and length; and severe, as full-thickness resorption of any length.

Only a small part of the allograft, where it comes into contact with the host femur, is replaced by host bone. Resorption, therefore, eventually may be a problem. Resorption on the endosteal surface poses less of a problem than that on the external surface because the cement inhibits access by host granulation tissue and the dead allograft is not a source of viable cells. On the external surface, however, there is access to host tissue and vascularization that could lead to resorption and ultimately to a late fracture. With use of strong cortical allograft bone, this process should be very slow and the reconstruction with the graft-cement-implant composite should last for a number of years.

In our experience at the University of British Columbia, resorption of the graft was evident in 34 percent of forty-one hips and it was graded as severe in 24 percent⁸⁹. A few hips had almost complete resorption of the graft, which led to a revision procedure in one of them. It is unclear whether resorption is related to an immune phenomenon, to the local soft-tissue vascularity, to revascularization, to the way in which the allograft is processed or prepared intraoperatively, or to the use of bone cement. It is also possible that this is the natural history of reconstruction with a segmental allograft. Hutchison et al. found resorption at the site of cerclage wires and concluded that this was related to a local vascular phenomenon⁶⁵.

Resorption of the graft is one of the most important reasons for expressing caution about the use of proximal femoral grafts in hip reconstruction. Its natural history is unknown, and it will be important to determine whether resorption stabilizes or progresses over time. Continued careful long-term follow-up is necessary to determine the clinical implications of resorption and to modify our technique in order to avoid it.

Overview

The technique of segmental replacement with an allograft-implant composite requires careful and constant appraisal. Allograft bone is an attractive alternative to autogenous bone. It is available in large quantities and in various shapes that can be tailored to any defect that is encountered, and use of allograft avoids donor-site morbidity. These characteristics make it a very useful alternative in revision arthroplasty, during which bone-stock deficiencies are often encountered. Moreover, this technique does not require sacrifice of host tissue and could facilitate subsequent interventions. The clinical results are good, with a substantial improvement in clinical scores and a decreasing rate of complications. The results reported after difficult reconstructions in hips that have had multiple revisions have been encouraging, and the procedure is certainly one useful alternative in these challenging situations^18,47,48,51.

The exact role of replacement of the proximal part of the femur with a segmental allograft in revision hip arthroplasty is controversial. These allografts have been used successfully in the management of very disabled patients who have had many previous interventions and for whom the primary goals of such a complex reconstruction are the relief of pain and the restoration of function. The limited capacity of a structural allograft to act as a truly biological segmental replacement must be appreciated. We are not aware of any conclusive evidence regarding the long-term results of reconstruction with use of a proximal femoral allograft. At present, such allografts have a definite role in the reconstruction of femora with a severe bone deficiency. In hips that have less severe bone loss, the role of segmental allografts is unknown and awaits additional study.

A number of technical issues with regard to the preparation and fixation of an allograft have been resolved over the past decade. The importance of rigid fixation and good soft-tissue reconstruction has become clear. The indications for the procedure also have become clearer as the risks associated with it now are appreciated and more data are available on the alternatives for reconstruction of the proximal part of a femur that has loss of bone. We have seen a substantial rate of morbidity and a high rate of reoperation after reconstruction with use of an allograft, but, with better preparation of the allograft and with better operative techniques, the results have been improving. Careful observation and cautious optimism are necessary.

Although there is a wealth of knowledge on the subject in the literature, there are still many unknowns. The role of the immune system in bone transplantation has been known for a long time, but only recently has it become apparent that the bone-remodeling system and the immune system interact to affect the clinical success of bone transplantation. Neither of these two systems, nor their interaction, is completely understood. Future research in the field of bone transplantation will be aimed at achieving a better understanding of these systems individually and, more importantly, of how they interact in humans.

As the number of failing primary and revision arthroplasties increases, allograft bone will continue to play an important role in revision hip arthroplasty. There is clearly a need, however, for the continued careful surveillance of reconstructions with proximal femoral allografts, which must be performed in a specialized and well equipped setting. A better understanding of the biological characteristics of allografts is needed to improve the results. In addition, continued clinical research is needed to define the exact roles for the different techniques of reconstruction with an allograft. In the interim, structural femoral allografts can be used with success if the biological and immunological roles of the allograft and the important role of the host environment in bone transplantation are respected.

References

Allan, D. G.; Lavoie, G. J.; McDonald, S.; Oakeshott, R.; and Gross, A. E.: Proximal femoral allografts in revision hip arthroplasty. J. Bone and Joint Surg.,73-B(2): 235-240, 1991.73-B(2)235 1991

Aro, H. T., and Aho, A. J.: Clinical use of bone allografts. Ann. Med.,25: 403-412, 1993.25403 1993 [PubMed]

Bargar, W. L.: Shape the implant to the patient. A rationale for the use of custom-fit cementless total hip implants. Clin. Orthop.,249: 73-78, 1989.24973 1989 [PubMed]

Berrey, B. H., Jr.; Lord, C. F.; Gebhardt, M. C.; and Mankin, H. J.: Fractures of allografts. Frequency, treatment, and end-results. J. Bone and Joint Surg.,72-A: 825-833, July 1990.72-A825 1990

Bloebaum, R. D.; Lauritzen, R. S.; Skedros, J. G.; Smith, E. F.; Thomas, K. A.; Bennett, J. T.; and Hofmann, A. A.: Roentgenographic procedure for selecting proximal femur allograft for use in revision arthroplasty. J. Arthroplasty,8: 347-360, 1993.8347 1993 [PubMed]

Bonfiglio, M.; Jeter, W. S.; and Smith, C. L.: The immune concept: its relation to bone transplantation. Ann. New York Acad. Sci.,59: 417-432, 1955.59417 1955

Borja, F. J., and Mnaymneh, W.: Bone allografts in salvage of difficult hip arthroplasties. Clin. Orthop.,197: 123-130, 1985.197123 1985 [PubMed]

Bos, G. D.; Goldberg, V. M.; Zika, J. M.; Heiple, K. G.; and Powell, A. E.: Immune responses of rats to frozen bone allografts. J. Bone and Joint Surg.,65-A: 239-246, Feb. 1983.65-A239 1983

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]

Buck, B. E.; Resnick, L.; Shah, S. M.; and Malinin, T. I.: Human immunodeficiency virus cultured from bone. Implications for transplantation. Clin. Orthop.,251: 249-253, 1990.251249 1990 [PubMed]

Burchardt, H.; Glowczewskie, F. P.; and Enneking, W. F.: Allogeneic segmental fibular transplants in azathioprine-immunosuppressed dogs. J. Bone and Joint Surg.,59-A: 881-894, Oct. 1977.59-A881 1977

Burchardt, H.; Jones, H.; Glowczewskie, F.; Runder, C.; and Enneking, W. F.: Freeze-dried allogeneic segmental cortical-bone grafts in dogs. J. Bone and Joint Surg.,60-A: 1082-1090, Dec. 1978.60-A1082 1978

Burchardt, H.; Glowczewskie, F. P.; and Enneking, W. F.: Short-term immunosuppression with fresh segmental fibular allografts in dogs. J. Bone and Joint Surg.,63-A: 411-415, March 1981.63-A411 1981

Burchardt, H.: The biology of bone graft repair. Clin. Orthop.,174: 28-42, 1983.17428 1983 [PubMed]

Burchardt, H.: Biology of bone transplantation. Orthop. Clin. North America,18: 187-196, 1987.18187 1987

Chandler, H. P., and Penenberg, B. L.: Bone Stock Deficiency in Total Hip Replacement: Classification and Management. Thorofare, New Jersey, Slack, 1989.

Chandler, H.; Clark, J.; Murphy, S.; McCarthy, J.; Penenberg, B.; Danylchuk, K.; and Roehr, B.: Reconstruction of major segmental loss of the proximal femur in revision total hip arthroplasty. Clin. Orthop.,298: 67-74, 1994.29867 1994 [PubMed]

Chandler, H. P.: Reconstruction of major segmental loss of the proximal femur in revision total hip replacement. Orthopedics,20: 801-803, 1997.20801 1997 [PubMed]

Chase, S. W., and Herndon, C. H.: The fate of autogenous and homogenous bone grafts. A historical review. J. Bone and Joint Surg.,37-A: 809-841, July 1955.37-A809 1955

Curtiss, P. H., Jr.; Powell, A. E.; and Herndon, C. H.: Immunological factors in homogenous-bone transplantation. III. The inability of homogenous rabbit bone to induce circulating antibodies in rabbits. J. Bone and Joint Surg.,41-A: 1482-1488, Dec. 1959.41-A1482 1959

Czitrom, A. A.: Biology of bone grafting and principles of bone banking. In The Pediatric Spine: Principles and Practice, edited by S. L. Weinstein. Vol. 2, pp. 1285-1298. New York, Raven Press, 1994.

D'Antonio, J.; McCarthy, J. C.; Bargar, W. L.; Borden, L. S.; Cappelo, W. N.; Collis, D. K.; Steinberg, M. E.; and Wedge, J. H.: Classification of femoral abnormalities in total hip arthroplasty. Clin. Orthop.,296: 133-139, 1993.296133 1993 [PubMed]

Deijkers, R. L. M.; Bloem, R. M.; Petit, P. L. C.; Brand, R.; Vehmeyer, S. B. W.; and Veen, M. R.: Contamination of bone allografts. Analysis of incidence and predisposing factors. J. Bone and Joint Surg.,79-B(1): 161-166, 1997.79-B(1)161 1997

Emerson, R. H., Jr.; Malinin, T. I.; Cuellar, A. D.; Head, W. C.; and Peters, P. C.: Cortical strut allografts in the reconstruction of the femur in revision total hip arthroplasty. A basic science and clinical study. Clin. Orthop.,285: 35-44, 1992.28535 1992 [PubMed]

Engelbrecht, E., and Heinert, K.: Klassification und Behandlungsrichtlinien von Kflochensubstanzverlustem bei Revisions operationen am Huftgelenk-mittelfristige Ergebnisse. Primare und Revisionsalloarthroplastik Hrsg-Endo-Klinik, Hamburg, pp. 189-201. Berlin, Springer, 1987.

Engh, C. A.; Glassman, A. H.; Griffin, W. L.; and Mayer, J. G.: Results of cementless revision for failed cemented total hip arthroplasty. Clin. Orthop.,235: 91-110, 1988.23591 1988 [PubMed]

Engh, C. A., and Glassman, A. H.: Cementless revision of failed total hip replacement. Orthop. Rev.,19 (Supplement 9): 23-28, 1990.19 (Supplement 9)23 1990

Engh, C. A., and Glassman, A. H.: Cementless revision of failed total hip replacement: an update. In Instructional Course Lectures, American Academy of Orthopaedic Surgeons. Vol. 40, pp. 189-197. Park Ridge, Illinois, American Academy of Orthopaedic Surgeons, 1991.

Enneking, W. F.; Burchardt, H.; Puhl, J. J.; and Piotrowski, G.: Physical and biological aspects of repair in dog cortical-bone transplants. J. Bone and Joint Surg.,57-A: 237-252, March 1975.57-A237 1975

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

Fideler, B. M.; Vangsness, C. T., Jr.; Moore, T.; Li, Z.; and Rasheed, S.: Effects of gamma irradiation on the human immunodeficiency virus. A study in frozen human bone-patellar ligament-bone grafts obtained from infected cadavera. J. Bone and Joint Surg.,76-A: 1032-1035, July 1994.76-A1032 1994

Friedlaender, G. E.; Strong, D. M.; and Sell, K. W.: Studies on the antigenicity of bone. I. Freeze-dried and deep-frozen bone allografts in rabbits. J. Bone and Joint Surg.,58-A: 854-858, Sept. 1976.58-A854 1976

Friedlaender, G. E.: Editorial comment. Clin. Orthop.,174: 2-4, 1983.1742 1983

Friedlaender, G. E.: Immune responses to osteochondral allografts. Current knowledge and future directions. Clin. Orthop.,174: 58-68, 1983.17458 1983 [PubMed]

Friedlaender, G. E.; Mankin, H. J.; and Sell, K. W.: Osteochondral Allografts. Biology, Banking, and Clinical Applications. Boston, Little, Brown, 1983.

Friedlaender, G. E.: Editorial. Bone allografts: the biological consequences of immunological events. J. Bone and Joint Surg.,73-A: 1119-1122, Sept. 1991.73-A1119 1991

Friedlaender, G. E., and Goldberg, V. M.: Bone and Cartilage Allografts: Biology and Clinical Applications. Park Ridge, Illinois, American Academy of Orthopaedic Surgeons, 1991.

Galea, G.; Kopman, D.; and Graham, B. J. M.: Supply and demand of bone allograft for revision hip surgery in Scotland. J. Bone and Joint Surg.,80-B(4): 595-599, 1998.80-B(4)595 1998

Garbuz, D. S.; Masri, B. A.; and Czitrom, A. A.: Biology of allografting. Orthop. Clin. North America,29: 199-204, 1998.29199 1998

Garbuz, D. S.; Masri, B. A.; and Duncan, C. P.: Periprosthetic fractures of the femur: Principles of prevention and management. In Instructional Course Lectures, American Academy of Orthopaedic Surgeons. Vol. 47, pp. 237-246. Rosemont, Illinois, American Academy of Orthopaedic Surgeons, 1998.

Gie, G. A.; Linder, L.; Ling, R. S. M.; Simon, J.-P.; Slooff, T. J. J. H.; and Timperley, A. J.: Impacted cancellous allografts and cement for revision total hip arthroplasty. J. Bone and Joint Surg.,75-B(1): 14-21, 1993.75-B(1)14 1993

Goetz, D. D.; Capello, W. N.; Callaghan, J. J.; Brown, T. D.; and Johnston, R. C.: Salvage of a recurrently dislocating total hip prosthesis with use of a constrained acetabular component. A retrospective analysis of fifty-six cases. J. Bone and Joint Surg.,80-A: 502-509, April 1998.80-A502 1998

Goldberg, V. M., and Stevenson, S.: Biology of bone and cartilage allografts. In Allografts in Orthopaedic Practice, pp. 1-13. Edited by A. A. Czitrom and A. E. Gross. Baltimore, Williams and Wilkins, 1992.

Gouin, F.; Passuti, N.; Verriele, V.; Delecrin, J.; and Bainvel, J. V.: Histological features of large bone allografts. J. Bone and Joint Surg.,78-B(1): 38-41, 1996.78-B(1)38 1996

Grauer, J. D.; Amstutz, H. C.; O'Carroll, P. F.; and Dorey, F. J.: Resection arthroplasty of the hip. J. Bone and Joint Surg.,71-A: 669-678, June 1989.71-A669 1989

Gross, A. E.: Revision arthroplasty of the hip using allograft bone. In Allografts in Orthopaedic Practice, pp. 147-173. Edited by A. A. Czitrom and A. E. Gross. Baltimore, Williams and Wilkins, 1992.

Gross, A. E.; Allan, D. G.; Lavoie, G. J.; and Oakeshott, R. D.: Revision arthroplasty of the proximal femur using allograft bone. Orthop. Clin. North America,24: 705-715, 1993.24705 1993

Gross, A. E.; Allen, G.; and Lavoie, G.: Revision arthroplasty using allograft bone. In Instructional Course Lectures, American Academy of Orthopaedic Surgeons. Vol. 42, pp. 363-380. Rosemont, Illinois, American Academy of Orthopaedic Surgeons, 1993.

Gross, A. E.; Hutchison, C. R.; Alexeeff, M.; Mahomed, N.; Leitch, K.; and Morsi, E.: Proximal femoral allografts for reconstruction of bone stock in revision arthroplasty of the hip. Clin. Orthop.,319: 151-158, 1995.319151 1995 [PubMed]

Gross, A. E.; Allan, D. G.; Leitch, K. K.; and Hutchison, C. R.: Proximal femoral allografts for reconstruction of bone stock in revision arthroplasty of the hip. In Instructional Course Lectures, American Academy of Orthopaedic Surgeons. Vol. 45, pp. 143-147. Rosemont, Illinois, American Academy of Orthopaedic Surgeons, 1996.

Gross, A. E., and Hutchison, C. R.: Proximal femoral allografts for reconstruction of bone stock in revision arthroplasty of the hip. Orthop. Clin. North America,29: 313-317, 1998.29313 1998

Gruen, T. A.; McNeice, G. M.; and Amstutz, H. C.: "Modes of failure" of cemented stem-type femoral components. A radiographic analysis of loosening. Clin. Orthop.,141: 17-27, 1979.14117 1979 [PubMed]

Gustilo, R. B., and Pasternak, H. S.: Revision total hip arthroplasty with titanium ingrowth prosthesis and bone grafting for failed cemented femoral component loosening. Clin. Orthop.,235: 111-119, 1988.235111 1988 [PubMed]

Hamer, A. J.; Strachan, J. R.; Black, M. M.; Ibbotson, C. J.; Stockley, I.; and Elson, R. A.: Biochemical properties of cortical allograft bone using a new method of bone strength measurement. A comparison of fresh, fresh-frozen and irradiated bone. J. Bone and Joint Surg.,78-B(3): 363-368, 1996.78-B(3)363 1996

Hanson, P. D.; Warner, C.; Kofroth, R.; Osmond, C.; Bogdanske, J. J.; Kalscheur, V. L.; Frassica, F. J.; and Markel, M. D.: Effect of intramedullary polymethylmethacrylate and autogenous cancellous bone on healing of frozen segmental allografts. J. Orthop. Res.,16: 285-292, 1998.16285 1998 [PubMed]

Harris, W. H.: Traumatic arthritis of the hip after dislocation and acetabular fractures: treatment by mold arthroplasty. An end-result study using a new method of result evaluation. J. Bone and Joint Surg.,51-A: 737-755, June 1969.51-A737 1969

Harris, W. H., and White, R. E., Jr.: Resection arthroplasty for nonseptic failure of total hip arthroplasty. Clin. Orthop.,171: 62-67, 1982.17162 1982 [PubMed]

Head, W. C.; Malinin, T. I.; and Berklacich, F.: Freeze-dried proximal femur allografts in revision total hip arthroplasty. A preliminary report. Clin. Orthop.,215: 109-121, 1987.215109 1987 [PubMed]

Head, W. C.; Berklacich, F. M.; Malinin, T. I.; and Emerson, R. H., Jr.: Proximal femoral allografts in revision total hip arthroplasty. Clin. Orthop.,225: 22-36, 1987.22522 1987 [PubMed]

Head, W. C.; Wagner, R. A.; Emerson, R. H., Jr.; and Malinin, T. I.: Restoration of femoral bone stock in revision total hip arthroplasty. Orthop. Clin. North America,24: 697-703, 1993.24697 1993

Head, W. C.; Wagner, R. A.; Emerson, R. H., Jr.; and Malinin, T. I.: Revision total hip arthroplasty in the deficient femur with a proximal load-bearing prosthesis. Clin. Orthop.,298: 119-126, 1994.298119 1994 [PubMed]

Head, W. C.; Malinin, T. I.; Mallory, T. H.; and Emerson, R. H., Jr.: Onlay cortical allografting for the femur. Orthop. Clin. North America,29: 307-312, 1998.29307 1998

Hodge, W. A.: Cemented proximal femoral allograft in revision total hip replacement. Orthop. Internat.,2: 259-266, 1994.2259 1994

Horowitz, M. C., and Friedlaender, G. E.: Induction of specific T-cell responsiveness to allogeneic bone. J. Bone and Joint Surg.,73-A: 1157-1168, Sept. 1991.73-A1157 1991

Hutchison, C. R.; Mahomed, N.; Agnidis, Z.; Leitch, K.; Alexeeff, M.; and Gross, A. E.: Proximal femoral allografts in revision hip arthroplasty—minimum five-year follow-up. Orthop. Trans.,20: 94, 1996.2094 1996

Incavo, S. J., and Ames, S. E.: Allograft-host mismatch in revision total hip replacement. Orthop. Rev.,23: 832-836, 1994.23832 1994 [PubMed]

Jofe, M. H.; Gebhardt, M. C.; Tomford, W. W.; and Mankin, H. J.: Reconstruction for defects of the proximal part of the femur using allograft arthroplasty. J. Bone and Joint Surg.,70-A: 507-516, April 1988.70-A507 1988

Johnston, R. C.; Fitzgerald, R. H., Jr.; Harris, W. H.; Poss, R.; Müller, M. E.; and Sledge, C. B.: Clinical and radiographic evaluation of total hip replacement. A standard system of terminology for reporting results. J. Bone and Joint Surg.,72-A: 161-168, Feb. 1990.72-A161 1990

Kang, J. S., and Kim, N. H.: The biomechanical properties of deep freezing and freeze drying bones and their biomechanical changes after in-vivo allograft. Yonsei Med. J.,36: 332-335, 1995.36332 1995 [PubMed]

Khan, M. T.; Stockley, I.; and Ibbotson, C.: Allograft bone transplantation: a Sheffield experience. Ann. Roy. Coll. Surgeons England,80: 150-153, 1998.80150 1998

Kohles, S. S.; Markel, M. D.; Rock, M. G.; Chao, E. Y.; and Vanderby, R., Jr.: Mechanical evaluation of six types of reconstruction following 25, 50, and 75% resection of the proximal femur. J. Orthop. Res.,12: 834-843, 1994.12834 1994 [PubMed]

Kohles, S. S.; Markel, M. D.; Rock, M. G.; Chao, E. Y.; and Vanderby, R., Jr.: Fixation of femoral allograft/prosthesis composites after 25%, 50% and 75% resection. Med. Eng. and Phys.,18: 115-121, 1996.18115 1996

Kostuik, J., and Alexander, D.: Arthrodesis for failed arthroplasty of the hip. Clin. Orthop.,188: 173-182, 1984.188173 1984 [PubMed]

Langer, F.; Czitrom, A.; Pritzker, K. P.; and Gross, A. E.: The immunogenicity of fresh and frozen allogeneic bone. J. Bone and Joint Surg.,57-A: 216-220, March 1975.57-A216 1975

Lee, W. P.; Pan, Y. C.; Kesmarky, S.; Randolph, M. A.; Fiala, T. S.; Amarante, M. T.; Weiland, A. J.; and Yaremchuk, M. J.: Experimental orthotopic transplantation of vascularized skeletal allografts: functional assessment and long-term survival. Plast. and Reconstr. Surg.,95: 336-353, 1995.95336 1995

Lexer, E.: Substitution of whole or half joints from freshly amputated extremities by free plastic operations. Surg., Gynec. and Obstet.,6: 601-607, 1908.6601 1908

Lexer, E.: Joint transplantations and arthroplasty. Surg., Gynec. and Obstet.,40: 782-809, 1925.40782 1925

Lombardi, A. V., Jr.; Mallory, T. H.; Kraus, T. J.; and Vaughn, B. K.: Preliminary report on the S-ROM constraining acetabular insert: a retrospective clinical experience. Orthopedics,14: 297-303, 1991.14297 1991 [PubMed]

Lord, C. F.; Gebhardt, M. C.; Tomford, W. W.; and Mankin, H. J.: Infection in bone allografts. Incidence, nature, and treatment. J. Bone and Joint Surg.,70-A: 369-376, March 1988.70-A369 1988

Loty, B.; Tomeno, B.; Evrard, J.; and Postel, M.: Infection in massive bone allografts sterilised by radiation. Internat. Orthop.,18: 164-171, 1994.18164 1994

McGann, W.; Mankin, H. J.; and Harris, W. H.: Massive allografting for severe failed total hip replacement. J. Bone and Joint Surg.,68-A: 4-12, Jan. 1986.68-A4 1986

McLaughlin, J. R., and Harris, W. H.: Revision of the femoral component of a total hip arthroplasty with the calcar-replacement femoral component. Results after a mean of 10.8 years postoperatively. J. Bone and Joint Surg.,78-A: 331-339, March 1996.78-A331 1996

Makley, J. T.: The use of allografts to reconstruct intercalary defects of long bones. Clin. Orthop.,197: 58-75, 1985.19758 1985 [PubMed]

Mallory, T. H.: Preparation of the proximal femur in cementless total hip revision. Clin. Orthop.,235: 47-60, 1988.23547 1988 [PubMed]

Mankin, H. J.; Doppelt, S.; and Tomford, W.: Clinical experience with allograft implantation. The first ten years. Clin. Orthop.,174: 69-86, 1983.17469 1983 [PubMed]

Markel, M. D.; Gottsauner-Wolf, F.; Rock, M. G.; Frassica, F. J.; and Chao, E. Y.: Mechanical characteristics of proximal femoral reconstruction after 50% resection. J. Orthop. Res.,11: 339-349, 1993.11339 1993 [PubMed]

Markel, M. D.; Wood, S. A.; Bogdanske, J. J.; Rapoff, A. J.; Kalscheur, V. L.; Bouvy, B. M.; Rock, M. G.; Chao, E. Y.; and Vanderby, R., Jr.: Comparison of allograft/endoprosthetic composites with a step-cut or transverse osteotomy configuration. J. Orthop. Res.,13: 639-641, 1995.13639 1995 [PubMed]

Martin, W. R., and Sutherland, C. J.: Complications of proximal femoral allografts in revision total hip arthroplasty. Clin. Orthop.,295: 161-167, 1993.295161 1993 [PubMed]

Masri, B. A.; Spangehl, M. J.; Duncan, C. P.; Beauchamp, C. P.; and Myerthal, S. L.: Proximal femoral allografts in revision total hip arthroplasty: a critical review. J. Bone and Joint Surg.,77-B (Supplement III): 306-307, 1995.77-B (Supplement III)306 1995

Masri, B. A., and Duncan, C. P.: Classification of bone loss in total hip arthroplasty. In Instructional Course Lectures, American Academy of Orthopaedic Surgeons. Vol. 45, pp. 199-208. Rosemont, Illinois, American Academy of Orthopaedic Surgeons, 1996.

Masri, B. A.; Masterson, E. L.; and Duncan, C. P.: The classification and radiographic evaluation of bone loss in revision hip arthroplasty. Orthop. Clin. North America,29: 219-227, 1998.29219 1998

Masri, B. A.; Campbell, D. G.; Garbuz, D. S.; and Duncan, C. P.: Seven specialized exposures for revision hip and knee replacement. Orthop. Clin. North America,29: 229-240, 1998.29229 1998

Moran, M. C.: Postoperative fracture of the host femur in allograft-prosthetic reconstruction of the hip. J. Arthroplasty,10: 550-553, 1995.10550 1995 [PubMed]

Muscolo, D. L.; Kawai, S.; and Ray, R. D.: In vitro studies of transplantation antigens present on bone cells in the cat. J. Bone and Joint Surg.,59-B(2): 342-348, 1977.59-B(2)342 1977

Muscolo, D. L.; Caletti, E.; Schajowicz, F.; Araujo, E. S.; and Makino, A.: Tissue-typing in human massive allografts of frozen bone. J. Bone and Joint Surg.,69-A: 583-595, April 1987.69-A583 1987

Muscolo, D. L.; Petracchi, L. J.; Ayerza, M. A.; and Calabrese, M. E.: Massive femoral allografts followed for 22 to 36 years. Report of six cases. J. Bone and Joint Surg.,74-B(6): 887-892, 1992.74-B(6)887 1992

Muscolo, D. L.; Ayerza, M. A.; Calabrese, M. E.; Redal, M. A.; and Araujo, E. S.: Human leukocyte antigen matching, radiographic score, and histologic findings in massive frozen bone allografts. Clin. Orthop.,326: 115-126, 1996.326115 1996 [PubMed]

Nemzek, J. A.; Arnoczky, S. P.; and Swenson, C. L.: Retroviral transmission by the transplantation of connective-tissue allografts. An experimental study. J. Bone and Joint Surg.,76-A: 1036-1041, July 1994.76-A1036 1994

Nemzek, J. A.; Arnoczky, S. P.; and Swenson, C. L.: Retroviral transmission in bone allotransplantation. The effects of tissue processing. Clin. Orthop.,324: 275-282, 1996.324275 1996 [PubMed]

100

Oakeshott, R. D.; Morgan, D. A. F.; Zukor, D. J.; Rudan, J. F.; Brooks, P. J.; and Gross, A. E.: Revision total hip arthroplasty with osseous allograft reconstruction. A clinical and roentgenographic analysis. Clin. Orthop.,225: 37-61, 1987.22537 1987 [PubMed]

101

Ottolenghi, C. E.: Massive osteoarticular bone grafts. Transplant of the whole femur. J. Bone and Joint Surg.,48-B(4): 646-659, 1966.48-B(4)646 1966

102

Ottolenghi, C. E.: Massive osteo and osteo-articular bone grafts. Technic and results of 62 cases. Clin. Orthop.,87: 156-164, 1972.87156 1972 [PubMed]

103

Ozaki, T.; Hillmann, A.; Bettin, D.; Wuisman, P.; and Winkelmann, W.: Intramedullary, antibiotic-loaded cemented, massive allografts for skeletal reconstruction. 26 cases compared with 19 uncemented allografts. Acta Orthop. Scandinavica,68: 387-391, 1997.68387 1997

104

Paprosky, W. G.; Lawrence, J.; and Cameron, H.: Femoral defect classification: clinical application. Orthop. Rev.,19 (Supplement 9): 9-15, 1990.19 (Supplement 9)9 1990

105

Parrish, F. F.: Treatment of bone tumors by total excision and replacement with massive autologous and homologous grafts. J. Bone and Joint Surg.,48-A: 968-990, July 1966.48-A968 1966

106

Parrish, F. F.: Allograft replacement of all or part of the end of a long bone following excision of a tumor. Report of twenty-one cases. J. Bone and Joint Surg.,55-A: 1-22, Jan. 1973.55-A1 1973

107

Pelker, R. R.; Friedlaender, G. E.; and Markham, T. C.: Biomechanical properties of bone allografts. Clin. Orthop.,174: 54-57, 1983.17454 1983 [PubMed]

108

Pelker, R. R., and Friedlaender, G. E.: Biomechanical aspects of bone autografts and allografts. Orthop. Clin. North America,18: 235-239, 1987.18235 1987

109

Poitout, D. G.; Lempidakis, M.; Loncle, X.; and Bernat, M.: Reconstructions massives du cotyle et du fémur proximal. Chirurgie,120: 254-263, 1994-1995.120254 1994-1995

110

Ries, M. D.; Gomez, M. A.; Eckhoff, D. G.; Lewis, D. A.; Brodie, M. R.; and Wiedel, J. D.: An in vitro study of proximal femoral allograft strains in revision hip arthroplasty. Med. Eng. and Phys.,16: 292-296, 1994.16292 1994

111

Roberson, J. R.: Proximal femoral bone loss after total hip arthroplasty. Orthop. Clin. North America,23: 291-302, 1992.23291 1992

112

Rodrigo, J. J.; Heiden, E.; Hegyes, M.; and Sharkey, N. A.: Immune response inhibition by irrigating subchondral bone with cytotoxic agents. Clin. Orthop.,326: 96-106, 1996.32696 1996 [PubMed]

113

Rosenberg, A., and Jacobs, J.: Circumferential proximal femoral allografting. Read at the Harvard Hip Course, Boston, Massachusetts, Sept. 1998.

114

Sim, F. H., and Chao, E. Y. S.: Hip salvage by proximal femoral replacement. J. Bone and Joint Surg.,63-A: 1228-1239, Oct. 1981.63-A1228 1981

115

Stevenson, S.; Li, X. Q.; and Martin, B.: The fate of cancellous and cortical bone after transplantation of fresh and frozen tissue-antigen-matched and mismatched osteochondral allografts in dogs. J. Bone and Joint Surg.,73-A: 1143-1156, Sept. 1991.73-A1143 1991

116

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

117

Stevenson, S.; Shaffer, J. W.; and Goldberg, V. M.: The humoral response to vascular and nonvascular allografts of bone. Clin. Orthop.,326: 86-95, 1996.32686 1996 [PubMed]

118

Stevenson, S.; Li, X. Q.; Davy, D. T.; Klein, L.; and Goldberg, V. M.: Critical biological determinants of incorporation of non-vascularized cortical bone grafts. Quantification of a complex process and structure. J. Bone and Joint Surg.,79-A: 1-16, Jan. 1997.79-A1 1997

119

Strong, D. M.; Friedlaender, G. E.; Tomford, W. M.; Springfield, D. S.; Shives, T. C., II; Burchardt, H.; Enneking, W. F.; and Mankin, H. J.: Immunologic responses in human recipients of osseous and osteochondral allografts. Clin. Orthop.,326: 107-114, 1996.326107 1996 [PubMed]

120

Tan, M. H., and Mankin, H. J.: Blood transfusion and bone allografts. Effect on infection and outcome. Clin. Orthop.,340: 207-214, 1997.340207 1997 [PubMed]

121

Thompson, R. C., Jr.; Pickvance, E. A.; and Garry, D.: Fractures in large-segment allografts. J. Bone and Joint Surg.,75-A: 1663-1673, Nov. 1993.75-A1663 1993

122

Tomford, W. W.; Thongphasuk, J.; Mankin, H. J.; and Ferraro, M. J.: Frozen musculoskeletal allografts. A study of the clinical incidence and causes of infection associated with their use. J. Bone and Joint Surg.,72-A: 1137-1143, Sept. 1990.72-A1137 1990

123

Vander Griend, R. A.: The effect of internal fixation on the healing of large allografts. J. Bone and Joint Surg.,76-A: 657-663, May 1994.76-A657 1994

124

Zmolek, J. C., and Dorr, L. D.: Revision total hip arthroplasty. The use of solid allograft. J. Arthroplasty,8: 361-370, 1993.8361 1993 [PubMed]

Topics

View Large | Download Slide(.ppt)
Add to My JBJS

FIG1-C:: Fig. 1-C: The greater trochanter and the host femur are wired to the allograft.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG2-B:: Fig. 2-B: Radiograph made 6.5 years after the operation.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG3-A:: Figs. 3-A, 3-B, and 3-C: Radiographs made before and after revision with a structural femoral allograft. Fig. 3-A: Preoperative anteroposterior radiograph.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG4:: Fig. 4 Radiograph made nine years postoperatively, showing trochanteric union and union between the allograft and the residual proximal host femur.

View Large | Download Slide(.ppt)
Add to My JBJS

View Larger

TABLE I COMPLICATIONS AFTER REVISION OF A FEMORAL COMPONENT WITH USE OF A STRUCTURAL ALLOGRAFT

*The values are given as the number of allografts, with the percentage in parentheses.

Study	No. of Allografts	Duration of Follow-up (mos.)	Complication*
Nonunion	Dislocation	Infection
Gross and Hutchison⁵¹ (1998)	200	24	7 (4)	11 (6)	6 (3)
Masri et al.⁸⁹ (1995)	39	61	4 (10)	2 (5)	2 (5)
Chandler and Penenberg¹⁶ (1989)	30	22	2 (7)	5 (17)	1 (3)
Roberson¹¹¹ (1992)	24	48	2 (8)	1 (4)	2 (8)
Head et al.⁶⁰(1993)	22	28	2 (9)	5 (23)	0
Zmolek and Dorr¹²⁴ (1993)	11	48	2	6	1

Add to My JBJS

TABLE I COMPLICATIONS AFTER REVISION OF A FEMORAL COMPONENT WITH USE OF A STRUCTURAL ALLOGRAFT

*The values are given as the number of allografts, with the percentage in parentheses.

Study	No. of Allografts	Duration of Follow-up (mos.)	Complication*
Nonunion	Dislocation	Infection
Gross and Hutchison⁵¹ (1998)	200	24	7 (4)	11 (6)	6 (3)
Masri et al.⁸⁹ (1995)	39	61	4 (10)	2 (5)	2 (5)
Chandler and Penenberg¹⁶ (1989)	30	22	2 (7)	5 (17)	1 (3)
Roberson¹¹¹ (1992)	24	48	2 (8)	1 (4)	2 (8)
Head et al.⁶⁰(1993)	22	28	2 (9)	5 (23)	0
Zmolek and Dorr¹²⁴ (1993)	11	48	2	6	1