Graft Incorporation After Acetabular and Femoral Impaction Grafting With Washed Irradiated Allograft and Autologous Marrow

Patients and Methods 

Between December 1991 and March 2004, 103 consecutive patients underwent acetabular and or femoral revisions with impaction grafting by a single surgeon (GCB). All prosthetic and fibrous material was removed to expose bleeding bone. Femoral heads irradiated with 0.025 or 0.05 MGy from the local bone bank were ground to chips of 5-mm thickness using a Howix (Orthosonics, Devon, UK) bone mill. The bone chips were serially washed in normal saline at 37°C until they were clear of fat.

Immediately after this, a 5-mm incision was made over the ipsilateral iliac crest. Using a Manatech marrow needle (Manatech, Burton-upon-Trent, UK) marrow needle, bone marrow was aspirated in volumes of between 5 and 10 mm from the ipsilateral iliac crest from multiple sites, rotating the needle while withdrawing marrow. This was then added to the washed bone to make up 40% of the combined allograft-autologous marrow mixture.

Before preparation of the bone allograft-autologous marrow composite, the acetabular and femoral defects were contained with mesh. Immediately after preparation of the bone allograft-autologous marrow composite, it was impacted into the acetabulum by reverse reaming and into the femur by CPT (Zimmer, Swindon, UK) impaction grafting system. The acetabulum was then replaced predominantly with an uncemented hemispherical cup and the femoral component with the CPT (Zimmer) collarless polished tapered stem cemented with Palacos R with gentamicin (Schering-Plough, Kenilworth, NJ).

Radiographs were available for 85 patients, including 51 acetabular and 59 femoral revisions. Preoperative bone loss was classified using the Paprosky [6] and Endo-Klinik classification systems [7]. Preoperative bone loss was also identified in each Johnstone zones 1 to 6 for acetabular revisions and Gruen zones 1 to 14 for femoral revisions [8]. Immediate postoperative films were then reviewed to confirm defect filling. Graft incorporation was determined by the appearance of progressive cortication and trabeculation in grafted bone. Cortication was defined and noted when the endosteal erosions regained normal cortical structure and thickness. Trabeculation was defined and noted when the grafted bone changed into a pattern of trabeculae running obliquely from the endosteal surface into the cement along the normal direction of strain (Fig. 1, Fig. 2, Fig. 3, Fig. 4). Both authors reviewed consecutive postoperative films together. Only when both authors agreed on the presence of trabeculation and or cortication in each grafted zone was this recorded. The time this occurred postoperatively was noted. Postoperative complications were identified and confirmed from medical records. Femoral subsidence was determined by the method described by Eldridge et al [9].

View Large Image Download to PowerPoint

Fig. 1. A, Radiograph of a preoperative femoral stem showing cortical erosion and bone lysis. B, Radiograph of the femoral stem shown in A at 3 months postrevision showing cortication and trabeculation.

View Large Image Download to PowerPoint

Fig. 2. A, Radiograph showing preoperative acetabular component showing cortical erosion and large cement block. B, Six months postimpaction grafting showing cortication and trabeculation in place of previous cement block.

View Large Image Download to PowerPoint

Fig. 3. A, Preoperative femoral stem extensive cortical erosion and bone lysis. B, Posroperative femoral stem at 6 months showing cortication and trabeculation.

View Large Image Download to PowerPoint

Fig. 4. A, Preoperative femoral stem showing extensive cortical erosion and bone lysis. B, Postoperative femoral stem at 6 months showing significant cortical thickening.

Results 

Forty-two men and 43 women, with a mean age of 69 years and a mean follow-up of 44 months (range, 6-132 months), underwent revision hip arthroplasty with this modified technique of impaction grafting. Thirty-five underwent isolated femoral revision and 27 underwent acetabular revisions. (One patient had both hips grafted.) Most of the acetabular defects were Paprosky grade III and femoral defects Endo-Klinik II and III (Table 1, Table 2). Most of the femoral revisions were to cemented CPT or long CPT stems (Zimmer). Most of the acetabular revisions before 1998 were to Harris-Galante cups (Zimmer) and after 1998 to ABG cups (Stryker Howmedica Osteonics, Newbury, UK).

Table 1. Preoperative Classification of Acetabular Defects by Paprosky Grade

	Paprosky grade	Patients (%)
	I	23
	IIa	12
	IIb	6
	IIc	4
	IIIa	10
	IIIb	45

Table 2. Preoperative Classification of Femoral Defects by Endo-Klinik Classification

	Endo-Klinik grade	Patients (%)
	I	15
	II	41
	III	44
	IV	0

Graft Incorporation 

Ninety-six percent and 90% of acetabular and femoral revisions, respectively, showed graft incorporation in all zones by 12 months. Between 80% and 90% of these changes were first noted on the first review radiograph, after either 3 or 6 months. Four percent and 10% of acetabular and femoral grafts, respectively, failed to demonstrate graft incorporation in most of the zones (Table 3).

Table 3. Percentage of Patients Showing Trabeculation and or Cortication in Most or All Zones by Time

		Acetabulum	Femur
	Percentage showing graft incorporation in all grafted zones on first review radiograph (3 or 6 mo)	90% (46/51)	80% (47/59)
	Percentage showing graft incorporation in all grafted zones by 12 mo	96% (49/51)	90% (53/59)
	Percentage failing to show graft incorporation in all grafted zones	4% (2/51)	10% (6/59)

Discussion 

In this series, graft incorporation was determined by the presence of radiographic trabeculation and or cortication in grafted bone. A comparison of the radiographic and histologic changes in 14 patients undergoing femoral impaction grafting demonstrated that both these radiographic changes correspond to histologically incorporated live graft [10]. Two patients in this series required rerevision for late infection and instability several years after initial revision. Both grafts showed full radiographic incorporation before rerevision, and neither required further impaction grafting, suggesting restoration of the original bone stock and clinically incorporated graft.

The incorporation rate of irradiated bone allograft is significantly lower than that of nonirradiated bone allograft, approximately 40% compared with approximately 90% (P < .01) (Table 1). The advantage of irradiating bone allograft is sterile graft free from potentially infective microorganisms 11, 12. Concerns have been raised regarding the sterility of nonirradiated bone allografts [13], and there are reported cases of HIV transmission from fresh frozen bone allograft [14]. The combined literature on irradiated graft reports infection rates of lower than 1% compared with a mean of higher than 4% in nonirradiated graft (Table 1, Table 4). However, it appears that any potential advantage of increased sterility of irradiated bone allograft is bought at the price of impaired graft incorporation. Irradiating allograft damages the little remaining osteoinductive capacity of allograft 4, 5. The addition of autologous marrow in this series overcame this problem by combining a sterile allograft with a rich source of autologous osteoinductive cells and growth factors. Autologous marrow is easily obtainable in all patients. The incorporation rates in this series are better than irradiated bone alone 1, 2, 3 and are as good as any in the literature (Table 1), with incorporation occurring rapidly after impaction grafting.

Table 4. Graft Incorporation and Infection Rates by Author

	Author	Grafted area	Graft used	No. of grafts	Incorporation rate	Postoperative infection (%)
	Hassaballa et al [2]	Femur	Irradiated allograft	79	38% incorporation, no trabecular remodeling	0
	Robinson et al [1]	Femur	Irradiated unwashed allograft	57	39% graft incorporation	0
	Mehendale et al [3]	Acetabular	Irradiated allograft	58	49% graft incorporation, 6% demonstrated remodeling	NA
	Halliday et al [15]	Femur	Unwashed nonirradiated allograft	226	87% cortication, 34% trabecular remodeling	4.8
	Slooff et al [16]	Acetabular	Frozen nonirradiated allograft	88	88% demonstrated incorporation	2.4
	Gie et al [7]	Femur	Unwashed nonirradiated allograft	56	89%⁎ demonstrated incorporation	5.3
	Elting et al [17]	Femur	Unwashed nonirradiated allograft	56	88% demonstrated incorporation and remodeling	3.6
	Tokgözoğlu et al [18]	Femur	Freeze-dried nonirradiated allograft	33	100% incorporation and remodeling	3.1
	Schreurs et al [19]	Acetabular	Fresh frozen nonirradiated allograft	62	80% radiographically stable	6.9
	Winter et al [20]	Acetabular	Frozen nonirradiated allograft	38	100% incorporation	2.6
	This series	Femur	Washed, irradiated	59	90% incorporation	3.5
		Acetabulum	Allograft with autologous marrow	51	96% incorporation

NA indicates not applicable.

⁎ Not including patients excluded because of postoperative fracture.

The addition of autologous marrow to bone allograft or xenograft as a method for stimulating osteoinduction has been established since the 1960s [21]. Burwell [21] investigated histologic changes and incorporation rates in bone autograft compared with bone allograft with autologous red marrow in rats and found similar results in both types of graft. In this series, the addition of autologous marrow provides a rich source of stromal cells and osteoinductive factors to a bone scaffold [22]. Osteogenesis occurs because of an interaction between these autologous stromal cells and the donor graft 22, 23.

The use of synthetic substitutes in revision arthroplasty has been reported 24, 25, 26. The use of synthetic substitutes may become more relevant as the demand for donor allograft increases. However, there are limited data available regarding the efficacy of synthetic substitutes in revision arthroplasty with most of the current literature describing animal studies [27].

The use of impaction grafting in acetabular and femoral revision arthroplasty involves 2 different biologic environments—one a cemented polished tapered stem and the other an uncemented hemispherical component. Mechanically, both environments sustain compression and shear forces. Biologically, acetabular revision usually involves grafting on cancellous bone. Femoral revision usually involves grafting onto an eroded cortex with increased osteoclastic and osteoblastic activity [28]. Despite these differences, similar incorporation rates occur in both types of revision, both in this and other published series (Table 1).

The rapid incorporation rates of higher than 90% reported in this and other series are comparable to those of fracture union and repair. This suggests that impaction grafting is comparable to fracture fixation where there is a race between fracture union and fixation failure. If dead bone does not incorporate, it inevitably collapses. We suggest that if irradiated bone is used, osteoinductive agents should be added. Autologous marrow is the most easily available and least expensive of these.