A technique for enhancing union of allograft to host bone

S. M. Kumta, P. C. Leung, J. F. Grifth, D. J. Roebuck, L. T. C. Chow, C. K Li

From the Prince of Wales Hospital, Hong Kong

he aim of limb-salvage surgery in malignant bone tumours in children is to restore function and eradicate local disease with as little morbidity as possible. Allografts are associated with a high rate of complications, particularly malunion at the allograft-host junction. We describe a simple technique which enhances union of allograft to host bone taking advantage of the discrepancy in size between the adult allograft and the childs bone. This involves lifting a ap of periosteum before resection from the host bone, which is then telescoped into the allograft medullary canal, which may require internal burring or splitting, for a distance of 1.5 to 2 cm and covering the bone junction with the periosteal ap. This is more stable than conventional end-to-end opposition. For each centimetre of telescoping the surface area available for bony union is increased more than three times. The periosteal ap also augments union. Additional surface xation with a plate and screws is not necessary. We have used this technique in nine children, in eight of whom there was complete union at a mean of 16 weeks. Delayed union, associated with generalised limb osteoporosis, occurred in one. Early mobilisation, with weight-bearing by three weeks, was possible. There was only one fracture of the allograft.

for massive segmental skeletal reconstruction for several 1,2 decades. Their use in children, however, has not been universally accepted because of mechanical and other com3,4 plications. Diaphyseal intercalary resection with allograft reconstruction is increasingly used as a means of preserving joint function. Some authors have reported a low rate of 5 complications, but others have found a rate from 16% to 40%, particularly because of nonunion at the host-allograft 6-11 junction. Paediatric allografts are rarely used for weight-bearing reconstruction since the unfused physeal plate is a potential source of mechanical failure. Bone allografts are usually taken from skeletally mature donors and are often appreciably larger than the recipient childs bone. The discrepancy in size after segmental diaphyseal resection hinders the use of surface xation devices such as plates, which may require bending to ensure good surface contact. We were able to use this size discrepancy to our advantage in a group of children who had segmental diaphyseal resection and allograft reconstruction. The larger diameter allograft was telescoped over the host bone and the junction covered with a periosteal ap. The aim was to improve both the initial stability and the potential for early allografthost bony union compared with conventional end-to-end opposition. Stability was reinforced by intramedullary splintage.

J Bone Joint Surg [Br] 1998;80-B:994-8. Received 30 March 1998; Accepted after revision 13 May 1998

Patients and Methods

We performed segmental diaphyseal resection and telescoped allograft reconstruction in nine patients (four boys, five girls) aged from 7 to 12 years who had osteogenic sarcoma of the femur or the tibia (Table I). After staging of the tumour which included radiography, MRI, bone scintigraphy, thoracic CT and biopsy, all patients received preoperative chemotherapy followed by resection of the growth to an extent based on the preoperative MRI findings. We reconstructed the skeletal defect using adult allograft bone obtained from the Musculoskeletal Tissue Bank of the Chinese University of Hong Kong. Operative technique. Before removal of the tumour, the periosteum 2 to 3 cm distal to the level of bone resection was elevated as a ap (Fig. 1a). After excision, the intercalated defect was measured using the resected specimen as
THE JOURNAL OF BONE AND JOINT SURGERY

The standard treatment of malignant bone tumours in children is by wide local excision followed by one of several reconstructive procedures. Metal prostheses have been used
S. M. Kumta, MS Orth, Assistant Professor of Orthopaedics and Traumatology P. C. Leung, FRCS Orth, Professor of Orthopaedics and Traumatology J. F. Grifth, FRCR, Associate Professor of Radiology D. J. Roebuck, FRCR, Associate Professor of Radiology L. T. C. Chow, FRCP, Senior Medical Ofcer in Pathology C. K Li, MRCP, Consultant in Paediatric Oncology Prince of Wales Hospital, Chinese University of Hong Kong, New Territories, Shatin, Hong Kong. Correspondence should be sent to Dr S. M. Kumta. 1998 British Editorial Society of Bone and Joint Surgery 0301-620X/98/68982 $2.00
994

A TECHNIQUE FOR ENHANCING UNION OF ALLOGRAFT TO HOST BONE

995

Table I. Details of the nine children and the type of resection-reconstruction, follow-up and eventual outcome Case 1 2 3 4 5 6 7 8 9 Gender F M M M F F M F F Age (yr) 10 11 9 9 11 7 12 10 10 Site Femur Femur Femur Femur Femur Femur Tibia Tibia Tibia Stage II-B II-B II-B II-B II-B II-B II-B II-B II-B Type of reconstruction Intercalary Osteoarticular Osteoarticular Intercalary Intercalary Intercalary Allo-arthrodesis Intercalary Allo-arthrodesis Time to union (wk) 15 17 14 16 15 13 15 14 13 Complication Fracture None Local recurrence None None None None None None Follow-up (mth) 30 20 25 11 24 34 22 21 28 Function Excellent Fair Fair Excellent Fair Excellent Excellent Excellent Excellent Status No evidence of disease Died from disease Died from disease Died from complications No evidence of disease No evidence of disease Died from disease No evidence of disease No evidence of disease

Fig. 1 Diagrams showing a) the elevated periosteal ap of native bone and b) the telescoping of host bone into the allograft and covering of the junction by periosteum.

Fig. 3 Fig. 2 Diagrams showing a) the cleavage of the allograft to accommodate host bone and b) the securing of the allograft-host junction with cerclage wires. Lateral radiograph of an allograft-tibial bone junction and intramedullary splintage with bular autograft. Bony union is visible on both the endosteal and periosteal margins of the allograft.

a guide. An appropriate length of allograft was prepared with allowance for telescoping. In seven patients, the diaphysis of the host bone was telescoped into the allograft for 1.5 to 2 cm (Fig. 1b). In three of these patients internal burring of the allograft and in two splitting of the allograft down three sides in a cloverleaf manner was needed to allow telescoping over the host bone (Fig. 2a). We used a
VOL. 80-B, NO. 6, NOVEMBER 1998

cerclage wire to secure the split allograft (Fig. 2b). In all nine patients the periosteal ap was positioned over the site of union of the host and allograft bone. Internal splintage was provided by an intramedullary nail in seven patients and a bular autograft in two (Fig. 3). At the metaphyseal end, the width of the allograft was trimmed to match that of the host in both osteoarticular and intercalary resections.

996

S. M. KUMTA,

P. C. LEUNG,

J. F. GRIFFITH,

D. J. ROEBUCK,

L. T. C. CHOW,

C. K LI

Fig. 4 Anteroposterior radiographs of an allograft-femoral bone junction. Progressive bony union is seen at a) six weeks, b) four months and c) 14 months after reconstruction.

Metaphyseal stabilisation was secured by wire loops and Kirschner wires. The stability of the reconstruction was tested clinically before closure of the wound. The patients were allowed to mobilise the leg with the aid of a hinged brace in the rst week after operation within the limits of comfort. Partial weight-bearing was allowed by the third postoperative week. All patients were able to bear weight fully in the eighth postoperative week.

tion with osteoporosis of the adjacent host bone (Fig. 5). There were no allograft failures. Fracture of the allograft occurred in one patient at 20 months, and was treated by a vascularised bular graft (Table I). One year after resection, clinical function was assessed using the grading system developed by the Musculoskeletal 12 Tumour Society. Function was excellent in six and fair in three patients (Table I).

Results
The postoperative clinical progress of the patients was closely monitored and regular radiographs were taken to assess the state of union at the host-allograft junction. The median follow-up was 25.5 months (mean 25). All patients received postoperative chemotherapy for 16 weeks which did not have an adverse effect on callus formation or bony union. Early periosteal callus was noticeable at six weeks in eight of the nine patients. This progressively enveloped the host-allograft junction and corresponded well to the extent of coverage achieved by the periosteal ap (Fig. 4). Bony union progressed simultaneously at the endosteal surface. There was progressive obliteration of the allograft endosteal and host-bone periosteal interface. Union of the allograft-host junction was deemed complete when this was no longer visible radiologically. The mean time to union was 15 weeks (Table I). Protective external braces were discarded when union was solid and joint movement satisfactory. No patient had nonunion. In one patient the periosteal callus was poor in associa-

Discussion
The stability of the host-allograft bone junction is of great importance for achieving rapid bony union. The reported rates of failure for host-allograft bony union suggest that conventional methods of osteosynthesis which are so successful in the treatment of fractures may not be entirely suitable for allograft stabilisation. A multicentre study 13 involving 113 cases reported a 57% incidence of delayed union after intercalary allograft reconstruction or arthrodesis. In the same study, fractures were seen in 16% of allografts in which a plate was used, and in 12.5% of those with an intramedullary nail. Other studies have shown that often many years after implantation, an allograft will remain as an inert piece of bone with very little or no remodelling. It is not surprising that plates and screws inserted in a tissue that is avascular and has no remodelling capacity may result in fatigue failure of the allograft through stress risers beyond the plate or through the screw holes. Multivariate analysis of frac14 tures of large segment allograft showed signicant correTHE JOURNAL OF BONE AND JOINT SURGERY

A TECHNIQUE FOR ENHANCING UNION OF ALLOGRAFT TO HOST BONE

997

Fig. 5 Lateral radiograph of an allograft-femoral bone junction. Poor progression of bony union is seen at a) two months, b) four months, c) ten months and d) 15 months after reconstruction.

Host bone telescoped into allograft

End-to-end alignment

Fig. 6 Diagrams showing how telescoping increases the surface area available for bony union compared with conventional end-to-end oppositional alignment. 2 Area of Contact in A = dx + 1/2 2 d 1/2 Area of Contact in B = d where d = diameter of host bone and x = length of telescoped portion. The difference is 3.142 times the diameter for every centimetre of telescoping ( = 3.142).

lation with the use of screws and plates. Immune-related resorption of the margin of the allograft may further jeopardise the stability of the host-allograft junction. Histological studies have shown that bone formation at the host-allograft junction is initiated by enveloping callus 12,15 from the host periosteum. Surface plates interfere mechanically with this process, which may explain the unsatisfactory results associated with their use. Telescoping the host bone into the allograft results in a more secure
VOL. 80-B, NO. 6, NOVEMBER 1998

method of reconstruction than conventional end-to-end opposition. The contact area of the bone surface between host and allograft is increased more than threefold for every centimetre of telescoped length (Fig. 6). This encourages better union over a longer segment than the conventional spot-weld type of annealing seen with end-to-end opposition. Additional surface metal is rarely required. This method of reconstruction is sufciently stable to allow early activity at one week and partial weight-bearing at three weeks after operation. Telescoping opposing bone ends, in a similar fashion to that described, has been used previously to good effect in 16 the treatment of difcult fractures. The contact surface area of the host and allograft may also be increased by using a Z-shaped interlocking to oppose both bone ends. This arrangement does not, however, confer any additional stability which must be achieved by metal implants. The aim of limb-salvage surgery in children with a malignant growth is to restore function and eradicate local disease with as little morbidity as possible. Allografts are a useful means of retaining limb function despite a high rate of complications. This simple technique of telescoping the host bone into the allograft has allowed us to avoid some of the problems of host-allograft union with the minimal use of metal implants.
No benets in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.

998

S. M. KUMTA,

P. C. LEUNG,

J. F. GRIFFITH,

D. J. ROEBUCK,

L. T. C. CHOW,

C. K LI

References
1. Kotz R, Ritschl P, Trachtenbrodt J. A modular femur tibia reconstruction system. Orthopaedics 1986;9:1639-52. 2. Sim FH, Chao EYS. Segmental prosthetic replacement of the hip and knee. In: Chao EY-S, Ivins JC, eds. Tumor prostheses for bone and joint reconstruction: the design and application. New York: Thieme Stratton, 1983:247-66. 3. Gebhardt MC, Flugstad DI, Springeld DS, Mankin HJ. The use of bone allografts for limb salvage in high-grade extremity osteosarcoma. Clin Orthop 1991;270:181-96. 4. Schindler OS, Cannon SR, Briggs TW, Blunn GW. Stanmore custom-made extendible distal femoral replacements: clinical experience in children with primary malignant bone tumours. J Bone Joint Surg [Br] 1997;79-B:927-37. 5. Mankin HJ, Doppelt S, Tomford W. Clinical experience with allograft implantation: the rst ten years. Clin Orthop 1983;174: 69-86. 6. Clohisy DR, Mankin HJ. Osteoarticular allografts for reconstruction after resection of a musculoskeletal tumor in the proximal end of the tibia. J Bone Joint Surg [Am] 1994;76-A:549-54. 7. Hernigou P, Delepine G, Goutallier D, Julieron A. Massive allografts sterilised by irradiation: clinical results. J Bone Joint Surg [Br] 1993;75-B:904-13.

8. Loty B, Courpied JP, Tomeno B, et al. Bone allografts sterilised by irradiation: biological properties, procurement and results of 150 massive allografts. Int Orthop 1990;14:237-42. 9. Makley JT. The use of allografts to reconstruct intercalary defects of long bones. Clin Orthop 1985;197:58-75. 10. Mnaymneh W, Malinin T. Massive allografts in surgery of bone tumors. Orthop Clin North Am 1989;20:455-67. 11. Rock MG. Intercalary allograft reconstruction. In: Citzrom AA, Winkler H, eds. Orthopaedic allograft surgery New York: Springer, 1996;1:181-9. 12. Enneking WF, Mindell ER. Observations on massive retrieved human allografts. J Bone Joint Surg [Am] 1991;73-A:1123-42. 13. Donati D, Capanna R, Campanacci D, et al. The use of massive bone allografts for intercalary reconstruction and arthrodeses after tumor resection: a multicentric European study. Chir Organi Mov 1993;78:81-94. 14. Thompson RC, Pickvance EA, Garry D. Fractures in large-segment allografts. J Bone Joint Surg [Am] 1993;75-A:1663-73. 15. Aho AJ, Ekfors T, Dean PB, et al. Incorporation and clinical results of large allografts of the extremities and pelvis. Clin Orthop 1994;307:200-13. 16. Dimon JH, Hughston JC. Unstable intertrochanteric fractures of the hip. J Bone Joint Surg [Am] 1967;49-A:440-50.

THE JOURNAL OF BONE AND JOINT SURGERY