HYDROXYAPATITE IN REVISION OF TOTAL HIP REPLACEMENTS WITH MASSIVE ACETABULAR DEFECTS 4- TO 10-YEAR CLINICAL RESULTS H. OONISHI, Y. IWAKI, N. KIN, S. KUSHITANI, N. MURATA, S. WAKITANI, K. IMOTO From Osaka-Minami National Hospital, Osaka, Japan Hydroxyapatite (HA) granules of 100 to 300 m, 0.9 to 1.2 mm and 3.0 to 5.0 mm were mixed in a ratio of 10:45:45 and packed into massive bone deficiencies in revision operations for total hip arthroplasty. We did not use additional graft or cup support for deficiencies of the lateral and medial wall. The procedure was carried out in 40 hips between 1986 and 1992. The radiographic spaces seen at the interface between HA and bone immediately after surgery disappeared within three months. Some spaces appeared between HA granules near the bone in the lateral part of two joints, and three sockets migrated in patients with severe segmental and cavitary deficiencies. Direct bonding of HA to bone was observed radiologically without morphological changes, except in the three joints with migration. All patients could walk without pain but the three with definite loosening needed crutches. J Bone Joint Surg [Br] 1997;79-B:87-92. Received 3 May 1996; Accepted after revision 9 July 1996 Excellent results have been reported from Europe and the USA with the use of frozen allografts during revision arthroplasty when there has been massive bone deficiency. 1-3 Allografts have also been used in Japan, 4 although material other than femoral heads and pieces of femoral condyle or tibial plateau from patients undergoing arthroplasty are still very difficult to obtain. Increasing interest in bioactive ceramics, particularly in hydroxyapatite (HA) over the past ten years, has resulted in a considerable increase in their clinical application. 5-7 We have made clinical use of sintered HA granules. This material is not resorbable 8,9 and binds to the bone physicochemically. 10 Porous and irregularly-shaped HA granules of several sizes were packed as densely as possible and, before the socket was fixed, the entire surface of the exposed HA granules was covered and stabilised with bone cement. We obtained good results when performing autogenous grafting of defects in the femur in three patients in 1984 by placing fine HA granules (300 to 500 m) between the bone cement and the bone graft. 7 Since 1985 we have been filling massive bone defects at the time of revision with small HA granules 6,7 and we now describe our most recent findings. PATIENTS AND METHODS Between 1986 and 1992 we have carried out this procedure on 40 hips. The bone loss was assessed according to the H. Oonishi, MD, Orthopaedic Surgeon and Vice-Director Y. Iwaki, MD, Orthopaedic Surgeon N. Kin, MD, Orthopaedic Surgeon S. Kushitani, MD, Orthopaedic Surgeon N. Murata, MD, Orthopaedic Surgeon S. Wakitani, MD, Orthopaedic Surgeon K. Imoto, MD, Orthopaedic Surgeon Department of Orthopaedic Surgery, Artificial Joint Section and Biomaterial Research Laboratory, Osaka-Minami National Hospital, 677-2, Kido-Cho, Kawachinagano-Shi, Osaka 586, Japan. Correspondence should be sent to Dr H. Oonishi. 1997 British Editorial Society of Bone and Joint Surgery 0301-620X/97/11290 $2.00 Table I. Classification of the deficiencies of the acetabulum according to the AAOS criteria (D'Antonio et al 1989) Type Deficiency I II III IV V Segmental A Peripheral 1 Superior 2 Anterior 3 Posterior B Central (medial wall absent) Cavitary A Peripheral 1 Superior 2 Anterior 3 Posterior B Central (medial wall intact) Combined Pelvic discontinuity Arthrodesis VOL. 79-B, NO. 1, JANUARY 1997 87
88 H. OONISHI, Y. IWAKI, N. KIN, ET AL Table II. The type of acetabular deficiency in 40 hips, by number and percentage Type I A1 B II A1.2.3 18 (45) I A1.2.3 II A1.2.3 B 13 (33) I B II A 2 (5) I A1 II A1 B 7 (18) AAOS classification 11 of acetabular deficiencies (Table I). Whole peripheral segmental and cavitary deficiencies with the medial wall intact were found in 13 joints (33%) and generalised peripheral cavitary deficiencies with the medial wall absent in 18 (45%) (Table II). There were two men and 38 women and their age at operation ranged from 35 to 81 years. The follow-up was for nine to ten years in one, for eight to nine in three, for seven to eight in five, for six to seven in nine, for five to six in seven, and for four to five in 15. The original THR was performed because of osteoarthritis in 33 patients, rheumatoid arthritis in five, avascular necrosis of the head of the femur in one, and systemic lupus erythymatosus in one. In 36 joints this was the first revision, in three the second and in one the third. None of the patients required a further revision. Radiographs were used to study the interface between bone and HA, the interface between bone cement and HA, the changes in the volume and shape of HA granules, the absorption of HA, and movement of the component. Reconstructive procedures. The soft tissue and necrotic tissue adhering to the acetabulum are completely removed but only necrotic or inflammatory tissue is excised from the defect. HA granules of sizes 100 to 300 m (G-1), 0.9 to 1.2 mm (G-4) and 3.0 to 5.0 mm (G-6) (Bone Ceram P; Sumitomo Cement Co) are mixed in a ratio of 10:45:45. Physiological saline is added to increase the mixing density to facilitate the adhesion of the granules. The acetabulum is then filled with the mixture. A space is left for the cup and a hemispherical compressor, 3 mm larger than the outer diameter of the socket, is then placed in this space and knocked into the acetabulum with a plastic hammer in the optimum position (Figs 1a and 1b). When there was a large defect in the medial wall HA granules were packed in as required, escape being inhibited by the soft tissues. A layer of cement 2 to 3 mm thick is then placed over the entire surface of the acetabulum and compressed with a mould 1 mm larger than the outer diameter of the socket. Time is allowed for polymerisation to take place. Further granules of HA are then placed in any remaining defects of the acetabulum and fixed, if necessary, by another layer of cement (Fig. 1c). In 35 cases, the superior peripheral deficiency was covered with bone cement after filling with HA granules. In one patient a bone graft from the ilium was used Since 1993, we have used an allograft from the tibial plateau to cover large deficiencies at the superior periphery of the acetabulum. 12 When the hardened cement surface has dried the cup is Fig. 1 Stages in the revision of a THR using hydroxyapatite granules to fill massive bone deficiencies in the acetabulum. THE JOURNAL OF BONE AND JOINT SURGERY
HYDROXYAPATITE IN REVISION OF HIP REPLACEMENTS WITH MASSIVE ACETABULAR DEFECTS 89 Figure 2a Immediately after operation in a 65-year-old woman. Figure 2b Four months later. The spaces initially seen between the HA granules and the bone have disappeared. Fig. 2a Fig. 2b Figure 3a Before revision in a 67-year-old woman. Figure 3b Ten years later. This was the first case in which HA granules had been used to fill massive defects in the acetabulum. Fig. 3a Fig. 3b fixed with viscous cement which contains a high proportion of monomer (Fig. 1d). This facilitates insertion of the cup and allows the hardened and the viscous cement to polymerise and bond solidly. Postoperative management. Partial weight-bearing was allowed at 10 to 12 weeks after operation and full weightbearing two to three weeks later. In exceptional cases careful, supervised partial weight-bearing was allowed three to six weeks after operation. RESULTS Radiological evaluation. When the HA granules are firmly packed like a stone wall, a stable filling is attained. Although some spaces were observed at the interface between the HA granules and bone immediately after operation they gradually disappeared in the next three months (Fig. 2), probably because new bone entered the space between the granules encircling the cavity and became bound to them. Sclerotic bone which had surrounded the loose cup was seen to change to the appearance of cancellous bone over a period of one to three years after revision. We saw no radiolucent lines or spaces at the interface between the HA granules and the bone base (Fig. 2). Morphological changes or a decrease in volume were not seen (Fig. 3) except in some cases with very specific complications. Of the 38 patients with a superior peripheral deficiency in which exposed granules had been covered with bone cement, 34 showed good radiological results, but in two spaces had appeared between the granules and nearby bone in the lateral part of the acetabulum, and in the other two the socket had migrated (Figs 4 and 5). In the only patient in whom a superior peripheral deficiency had been covered with autograft from the ilium, the VOL. 79-B, NO. 1, JANUARY 1997
90 H. OONISHI, Y. IWAKI, N. KIN, ET AL Fig. 4 Radiograph eight years after revision in a 65-year-old woman. Spaces can be seen between the granules and bone and individual granules can be distinguished. graft had absorbed within three years of operation. The socket had been placed medial to the graft which did not transmit loading. In one patient in whom an excess of granules had been placed medially the cup migrated but without any detrimental effect. Even in patients with massive bone loss, the granules were very stable after operation. Complications. In two patients, spaces were observed in between granules and nearby bone (Fig. 4), but there were no clinical problems. We consider that rather wide spaces had been left between the bone base and the packed granules, and hence the bond between the bone and the granules was unstable from the onset, immediately after surgery. Continuous micromotion probably occurs at these gaps before sufficient bone growth can provide bonding. If the superior peripheral deficiency had been covered with an allograft plate, such as a tibial plateau, the granules would probably have filled the spaces more satisfactorily. In three patients in whom the central part of the medial wall was absent the overall deficiency was too great to allow stable filling by the granules (Fig. 5) and the prostheses migrated. At latest review all three patients could walk without pain, but were obliged to use crutches. Clinical results. Hip pain was alleviated in all the patients. One patient sustained a fracture of the femur at the time of the revision with subsequent persistence of pain in the thigh and greater trochanter (Fig. 6a). Walking ability was markedly improved except in patients with either tetraplegia associated with moderate sensory and motor paralysis due to myelopathy, with Parkinson s disease, or with contracture of both knees and ankles due to rheumatoid arthritis and in the one patient in whom the socket became loose after the third revision (Fig. 6b). The range of movement also improved except for the patient with the loose cup and the one with rheumatoid contractures (Fig. 6c). DISCUSSION The advantages of using HA granules for the management of massive bone deficiency are that immunoreaction can be completely ignored, morphological changes and a decrease in volume do not occur if the granules are of adequate size and are packed densely and firmly, absorption of the HA is Fig. 5a Fig. 5b Fig. 5c Figure 5a Before operation in a 78-year-old woman. Figure 5b Immediate postoperative radiograph. Figure 5c Fourteen months after operation the socket has migrated. THE JOURNAL OF BONE AND JOINT SURGERY
HYDROXYAPATITE IN REVISION OF HIP REPLACEMENTS WITH MASSIVE ACETABULAR DEFECTS 91 Fig. 6a Fig. 6b Fig. 6c Clinical evaluation using the criteria of Merle d Aubigné and Postel (1954) for pain (a), walking (b) and range of movement (c). Fig. 7 A superior peripheral segmental defect is covered from the outside with an allograft from the tibial plateau. The graft bone should protrude from the rim of the socket. minimal and it becomes bound to bone physicochemically. Osteolysis due to the presence of particles of polyethylene at the interfaces between cement, HA and bone does not appear to occur. The granules are packed tightly together and bonding occurs between them; this new bone creates a very stable structure which gives firm anchorage to the socket. The early radiological disappearance of the spaces between the granules and the bone and the replacement of sclerotic by cancellous bone emphasise the incorporation of the granules into bone (see Fig. 2). In many patients the HA layer was 3 to 4 cm thick, and we found at further operation in one patient after two years that the granules had formed a homogeneous mass which was difficult to cut with a chisel. This tends to confirm our results for experiments in goats; we found that the same size range of granules used to fill a 2 cm cavity in the shaft of the tibia achieved complete bone ingrowth by one year. 13 We also found that loosening and similar complications could be reduced when major superior peripheral segmental defects were stabilised by thick allografts. This allowed firm packing of the granules and provided loading of the grafted site (Fig. 7). Similarly, the reinforcement of large deficiencies of the medial wall with additional allograft will VOL. 79-B, NO. 1, JANUARY 1997
92 H. OONISHI, Y. IWAKI, N. KIN, ET AL improve the stability of the grafted area. The authors choose not to respond to the request for a conflict of interest statement. REFERENCES 1. Borja FJ, Mnaymneh W. Bone allografts in the salvage of difficult hip arthroplasties. Clin Orthop 1985;197:125-31. 2. Gross AE, Lavoie MV, McDermott P, Marks P. The use of allograft bone in revision of total hip arthroplasty. Clin Orthop 1985;197:115-22. 3. McGann W, Mankin HJ, Harris WH. Massive allografting for severe failed total hip replacement. J Bone Joint Surg [Am] 1986;68-A:4-12. 4. Itoman M, Sunabe S. Revision total hip replacement supplemented with allogenic bone grafting. J Joint Surg (Japan) 1988;7:83-93. 5. Oonishi H. Revision of THR for massive bone defects. J Joint Surg (Japan) 1988;7:49-60. 6. Oonishi H, Yamamoto M, Ishimaru H, et al. The effect of hydroxyapatite coating on bone growth into porous titanium alloy implants J Bone Joint Surg [Br] 1989;71-B:213-6. 7. Oonishi H. Orthopaedic applications of hydroxyapatite. Biomaterials. 1991;12:171-8. 8. Denissen HW, de Groot K, Makkes PCh, van den Hooff K, Klopper PJ. Tissue response to dense apatite implants in rats. J Biomed Mat Res 1980;5:713-21. 9. Hoogendoorn HA, Renooij W, Akkermans LMA, Visser W, Wittebol P. Long-term study of large ceramic implants (porous hydroxyapatite) in dog femora. Clin Orthop 1984;187:281-8. 10. Jarcho M, Kay JF, Gumaer KI, Doremus RH, Drobeck HP. Tissue, cellular and subcellular events at a bone-ceramic hydroxyapatite interface. J Bioeng 1977;1:79-92. 11. D'Antonio JA, Capello WN, Borden LS, et al. Classification and management of acetabular abnormalities in total hip arthroplasty. Clin Orthop 1989;243:126-37. 12. Oonishi H. Long term clinical results after revision total hip arthroplasty by using HA. J Joint Surg (Japan) 1995;14:51-64. 13. Oonishi H, Kushitani S, Murata N, et al. Long term bone growth behaviour into the spaces of HAp granules packed into massive bone defect cavity. In: Ducheyne P, Christiansen, eds. Bioceramics. Vol. 6. Butterworth-Heinemann Ltd, 1993:157-61. THE JOURNAL OF BONE AND JOINT SURGERY