The past 10 years have seen the introduction of newer techniques in femoral component revision. There now are several strategies of femoral reconstruction. When the bone loss in the femur affects the metaphysis but the diaphysis is well-preserved (the most common situation) distal fixation with a fully porous-coated long-stem is a popular form of reconstruction in North America especially if the canal is relatively narrow and the femoral cortices adequate. In instances where the bone loss is cavitary, affects metaphysis and diaphysis and the femoral canal diameter is very large, the method of impaction cancellous allograft with a cemented prosthesis is very attractive.
This technique first described by Sloof et al, 24 and was popularized by Gie et al. 8 Although the presence of an uninterrupted cortical envelope is desirable, segmental cortical defects do not contraindicate this technique as long as a satisfactory bony tube can be reconstructed. Gie and colleagues 8 reported excellent clinical results with this technique and a low incidence of complications, although subsidence of the stem within the cement mantle was common. During the past 8 years, there has been a plethora of reports from Europe 4,7,9,16,19,21,25 and North America 6,11,12,17 of femoral revisions using this technique. Although the overall experience seems favorable, problems reported included dislocation and early subsidence and fractures of the femur that occur intraoperative and postoperative. 10,11,17,21 All these problems may be related to the technique that has a steep learning curve.
In 1999, we reported our early experience in the first consecutive group of patients in whom this technique was used. Excellent clinical results were observed after 3 years of followup. Radiographic results were equally good with remodeling of the graft observed in more than 1/2 of the patients. The worst problem was the high incidence of postoperative femoral fractures (five, 8.8%) none of which was associated with prosthetic loosening; all fractures were treated successfully with osteosynthesis with plate and strut graft.
We are reporting on the same group of patients after a longer followup.
MATERIALS AND METHODS
Between 1993 and 1997, impaction grafting with particulate fresh-frozen cancellous allograft was done in 57 femoral component revisions by the three authors. During this period, these three surgeons did 404 femoral component revisions. This group of patients represents 14% of our revision practice in the period of the study. This study was approved by our Institutional Review Board and all patients gave informed consent to be included on the study. This technique was used selectively for severe cavitary or combined cavitary and segmental proximal femoral bone loss (AAOS types II or III as reported by D’Antonio et al 3) or Paprosky Types III or IV femoral bone loss as reported by Aribindi and Paprosky. 1 There were 24 men and 30 women (three had bilateral procedures). The right hip was operated on 33 times and the left hip was operated on 24 times. The mean age of the patients was 62.7 years (range, 36–79 years). All patients had revision surgery using compressed particulate fresh-frozen cancellous allograft and a cemented collarless, polished tapered stainless steel (CPT®, Zimmer, Warsaw, IN). The acetabular component was revised in the same procedure in 42 hips (12 with Harris-Galante® (Zimmer), 13 Trilogy® (Zimmer), 11 Omnifit PSL® (Howmedica-Osteonics, Allendale, NJ), four with Ganz (Sulzer Medical, Baar, Switzerland) support rings and in two a Burch-Schneider (Sulzer Medical) cage was used; the liner insert only was replaced in 11 hips and in three hips the socket was not revised.
Preoperatively, 46 hips caused moderate or severe pain, seven were only mildly painful, and in four there was no pain but surgery was deemed advisable because of the severe bone loss.
The technique used was the same in all cases and it evolved as follows. After hip exposure was accomplished most commonly using a posterolateral approach, the previous component and the cement, fibrous membrane and debris were removed. Every effort was made to ensure that the endosteal surface of the femur was cleaned perfectly. At this point, reconstruction of existing segmental defects was done to convert the femur into a tube; this was accomplished using some type of screen (vicryl or CoCr) or allograft struts. The femoral canal then was occluded approximately 3 cm below the most distal cavitary deficiency or below the distal tip of the implant to be used, whichever was more distal. A plastic cement restrictor or a cement plug were used commonly.
Cancellous bone graft from femoral heads or distal femurs was prepared using rongeurs or a mechanical bone mill, or both. Small fragments, 4 to 6 mm in size, were used. Then using a centering guide to ensure a uniform bone mantle, progressively-sized cylindrical packers starting at the bottom of the canal were used to compress the morsellized cancellous allograft increasing the diameter of the packers until the canal is 2/3 full. At this point, tamps that have the same shape but are sized larger than the real prosthesis were used to shape the proximal femoral endosteal cavity, creating in this manner a new medullary canal. Impaction must be vigorous so that rotational stability of the tamp is achieved. Careful attention to tamp rotation is essential to avoid component malposition. After a trial reduction with the tamp in place the cement is vacuum mixed, injected in retrograde fashion and pressurized, and the stem is cemented.
Routine postoperative care includes ambulation with partial weightbearing for 8 weeks with progression to full weightbearing thereafter as tolerated.
Our patients all were treated with unwashed fresh-frozen cancellous chips; the source was femoral heads or distal femurs. In addition, strut allografts were used for shaft reinforcement in 40 hips and they were fixed with wires or cables.
Radiographic evaluations were done by two of us (DJB and MEC). Comparison of the immediate postoperative radiographs with those obtained at 1 year and at the time of maximum followup were done. Quality of the cement mantle and compressed graft incorporation, stem position, and subsidence were assessed and recorded. Interobserver variations were minimal, but no statiscal test documenting measurement reliability can be provided.
All patients were followed up for 3 to 9.3 years (mean, 6.3 years) and no patient was lost to followup. Clinically the results were very satisfactory as far as pain relief is concerned. Thirty-six patients reported no pain and 15 patients reported slight pain (five of these were reporting pain around cables on the lateral thigh and two had preexisting trochanteric nonunions that remained ununited). Two patients had moderate pain. Three patients died 3, 5.3, and 6 years postoperatively, respectively of causes unrelated to the hip. All three patients had well-functioning hips at the time of their deaths. One patient had a deep infection develop and his prosthesis was removed at another hospital 18 months after surgery.
Radiographic evaluation showed all stems but one to be in neutral, less than 5° varus or less than 5° valgus. The one stem in varus greater than 5° subsided, became infected and failed within 18 months of surgery. Assessment of cement mantle was made on AP and lateral radiographs and proved to be difficult because the cement and the impacted cancellous graft at times were indistinguishable. It is our impression that the cement mantle was complete in most instances. Most of our patients’ radiographs were studied for quality of the cement mantle by an independent observer and this was reported elsewhere. 15
Long-term radiographic followup is available for 48 of the 53 surviving prostheses (Fig 1). No patient has shown evidence of radiographic loosening. Subsidence of the stem within the cement mantle was observed commonly. In 44 hips it measured 0 to 3 mm and two hips it measured 4 to 6 mm. There was one instance of massive subsidence but this was the deep infection mentioned previously. In 42 radiographs there was clear evidence of partial or complete cancellous bone remodeling.
There were, however, some complications. In addition to the aforementioned infection, there were four cases (7%) of prosthetic instability. One was acute in the postoperative period, related to malposition of the socket and required early socket liner revision. The second was related to severe abductor insufficiency and required revision to a constrained liner. The third occurred in a patient whose limb was shortened and who experiences episodes of subluxation without frank dislocation undoubtedly related to poor myofascial tension; this patient has not required revision surgery. The fourth one occurred early postoperatively and has not recurred after 6 years.
There was one partial peroneal palsy that recovered spontaneously and four intraoperative femoral fractures. This occurred during the process of cancellous graft impaction and the patients were treated with cerclage cables or wires and strut allograft as necessary. They did not have an influence on the postoperative recovery or the ultimate result. The most disturbing complications by far were the postoperative femoral shaft fractures. This problem occurred six times (10.5%). Two fractures occurred early (within 6 months of the surgery) and likely were related to known small defects on the wall of the femur created at the time of the revision; both were treated with internal fixation with a plate and strut graft augmentation placed at 90° to the plate, and both patients achieved uneventful healing. One occurred 2 years postoperatively at the level of a cable in a patient who was treated elsewhere with internal fixation with a plate; this failed and a second internal fixation with a plate and strut graft augmentation was done at our institution with complete healing of the fracture. Three occurred late (after 4 years); of these one occurred at 5 years at the level of a cable (Fig 2), and two patients initially were treated elsewhere with internal fixation with a plate, one with a satisfactory outcome and the other requiring a second internal fixation procedure (plus strut graft augmentation) at our institution to achieve union. All fractures occurred near the prosthetic tip. Five occurred in women and one occured in a man. There were no particular specific characteristics of the patients with fractures. No fracture was associated with prosthetic loosening. All united with internal fixation with a metal plate (augmented by a strut cortical graft at 90° to the plate in five).
In our revision hip replacement practice, the technique of impaction cancellous grafting plus cement was reserved for femurs with significant cavitary deficiencies and other techniques were used for most revisions. As mentioned during the period of this study, this technique encompassed 14% of our practice.
The clinical results reported here are similar to other results. 4,6–8,19,24 Excellent pain relief is observed commonly. This may be related to the very good mechanical prosthetic stability achieved with this method. 23 Our radiographic results also are very satisfactory with little subsidence seen after long-term followup; this is in contradistinction to the observations of others 10,11,17,21 and it may be related to how meticulously the technique is done. Vigorous packing of the cancellous allograft is, we think, essential; this would ensure rotational stability of the prosthesis, and place the graft in a restful environment that would facilitate its incorporation and minimize prosthetic subsidence. This was confirmed in laboratory studies done on adult sows by Hostner et al 9 who suggested that implant stability after impaction grafting is an effect of surgical technique. However, vigorous packing carries the risk of intraoperative fracture of the femur and this has been reported before 11,12,17,19; we had this complication four times and it may be related to the learning curve because all occurred early in our experience. A compromise must be kept between firm graft compression and fracture avoidance.
Radiographic evidence of graft remodeling was common in our experience and it has been observed by others. 8,17 That the graft is remodeled has been confirmed by scintigraphic assessment 16 and by histologic studies in laboratory goats 22 and in biopsy specimens from human retrievals. 13,14,18,20 However, even in biopsy samples obtained from patients years after surgery, it is obvious that graft incorporation is incomplete and fragments of dead bone are observed surrounded by thick fibrous tissue, which is said to form a composite structure strong enough to carry load. 13 Long-term evidence of graft incorporation and bone stock accretion currently is not available, but the information collected to date suggest that this technique may be able to restore femoral bone stock in revision arthroplasty.
Complication occurrence with femoral impaction grafting is high. Nerve lesions, trochanteric-related problems, dislocations, or infections occurred with similar frequency as in any other type of revision. Postoperative instability occurred four times in our group of patients and although these incidence is not excessive, it is worth emphasizing that when impacting the allograft to prepare the neomedullary canal, exquisite attention must be given to the version of the preparing tamp. In one of our cases, the impacted graft had to be removed and the procedure had to be repeated, when it was realized that the femoral component would be retroverted. Another cause of postoperative instability in one of our patients was the increased offset of the prosthesis we used that would not allow us to regain leg length thereby reducing myofascial tension. Preoperative templating would help choose the prosthetic design that would restore offset and leg length. Different designs now are available to use with the technique of impaction grafting.
Intraoperative femoral fractures, a complication of any revision hip surgery, probably are related to this specific technique as mentioned previously (the need to obtain excellent compression of the graft) and their incidence can be reduced as expertise is gained. Postoperative fractures of the femur are the most commonly cited complication of impaction grafting. 2,5,6,8,11,17,19,21,25
Early postoperative fractures are most commonly the consequence of an open section defect either previously present or created at surgery and not corrected; this was present in at least two of our three early postoperative fractures. However, in our experience, the most disturbing fracture is the one that occurs late, 4 or 5 years after surgery in patients who were doing very well until then; these represent 1/2 of our postoperative fractures and they occurred with minimal or no trauma. It could be speculated that they are related to bone remodeling and/or the increased mechanical stresses at the junction of a stiff with a more flexible zone in the femur because they all seem to occur at a level just distal to the tip of the prosthesis. Another factor could be the localized cortical resorption induced by cables and in two of the three, the late fracture occurred at the level of a cable. We do not know what possible influence the frequent use in our series of allograft cortical bone plates could have in the etiology of these fractures. Although in our series none of the fractures was associated with a loose prosthesis and although all healed satisfactorily with internal fixation with metal and a cortical allograft bone plate, the incidence of this complication remains unacceptably high. It is possible that the use of long stem prostheses now available may help reduce it.
We have used the technique of impaction grafting only in association with collarless, polished and tapered stems as suggested by Gie et al. 8 Linder 13 used precoated or rough surfaced stems and the short-term results are similar. The rational of the technique is based on the ductility of the cement and the ability of a polished tapered prosthesis to sink in the cement mantle and in this manner apply circumferential loads to the bone graft that would encourage its remodeling and incorporation. It remains to be seen whether the impacted bone will incorporate in the different mechanical environment produced by a precoated or rough-surfaced prosthesis that by definition would not be prone to subside.
We continue to use the technique of impaction grafting selectively for femoral metaphyseal or diaphyseal defects primarily cavitary in nature. We think that the presence of a medial femoral cortex proximally is essential, although small segmental cortical deficiencies can be accepted if a satisfactory envelope can be reconstructed. The results presented here justified, in our opinion, this attitude. Careful attention to the technical details and the use of long stemmed prostheses may help reduce the incidence of complications, particularly that of late postoperative femoral fractures.
1. Aribindi R, Paprosky WG: Bypass fixation. Orthop Clin North Am 29:319–329, 1998.
2. Berry DJ, Trousdale RT, Cabanela ME: Selective use of impaction grafting in revision hip arthroplasty for severe cavitary femoral bone loss. J Bone Joint Surg 83B(Suppl III):314, 2001.
3. D’Antonio J, McCarthy JC, Barger WL, et al: Classification of femoral abnormalities in total hip arthroplasty. Clin Orthop 296:133–139, 1993.
4. DeRoeck, Drabu KJ: Impaction bone grafting using freeze-dried allograft in revision hip arthroplasty. J Arthroplasty 16:201–206, 2001.
5. Duncan LP, Masterson EL, Masri BA: Impaction allografting with cement for the management of femoral bone loss. Orthop Clin North Am 29:297–305, 1998.
6. Elting JJ, Mikhail WE, Zicat BA, et al: Preliminary report of impaction grafting for exchange femoral arthroplasty. Clin Orthop 319: 159–167, 1995.
7. Flugsrud GB, Ovre S, Grogaard B, Nordsletten L: Cemented femoral impaction bone grafting for severe osteolysis in revision hip arthroplasty: Good results at 4-year followup of 10 patients. Arch Orthop Trauma Surg 120:386–389, 2000.
8. Gie GA, Linder L, Ling RS, et al: Impacted cancellous allografts and cement for revision total hip arthroplasty. J Bone Joint Surg 75B:14–21, 1993.
9. Hostner J, Hultmark P, Karrholm J, Malchau H, Tveit M: Impaction technique and graft treatment in revision of the femoral component: Laboratory studies and clinical validation. J Arthroplasty 16:76–82, 2001.
10. Karrholm J, Hultmark P, Carlsson L, Malchau H: Subsidence of a nonpolished stem in revisions of the hip using impaction allograft: Evaluation with radiostereometry and dual-energy x-ray absorptiometry. J Bone Joint Surg 81B:135–142, 1999.
11. Knight JL, Helming C: Collarless polished tapered impaction grafting of the femur during revision total hip arthroplasty: Pitfalls of the surgical technique and followup in 31 cases. J Arthroplasty 15:159–165, 2000.
12. Leopold SS, Berger RA, Rosenberg AG, et al: Impaction allografting with cement for revision of a femoral component: A minimum four-year followup study with use of a pre-coated femoral stem. J Bone Joint Surg 81A:1080–1092, 1999.
13. Linder L: Cancellous impaction grafting in the human femur: Histological and radiographic observations in six autopsy femurs and eight biopsies. Acta Orthop Scand 71:543–552, 2000.
14. Ling RSN, Timperley AJ, Linder L: Histology of cancellous impaction grafting in the femur. J Bone Joint Surg 75B:693–696, 1993.
15. Masterson EL, Masri BA, Duncan CP, et al: The cement mantle in femoral impaction allografting: A comparison of three systems from four centers. J Bone Joint Surg 79B:908–913, 1997
16. Mazhar Tokgozoglu A, Aydin M, Atilla B, Caner B: Scintigraphic evaluation of impaction grafting for total hip arthroplasty revision. Arch Orthop Trauma Surg 120:416–419, 2000.
17. Meding JB, Ritter MA, Keating EM, Faris RM: Impaction bone-grafting before insertion of a femoral stem with cement in revision total hip arthroplasty: A minimum two-year follow-up study. J Bone Joint Surg 79A:1834–1841, 1997.
18. Mikhail WE, Weidenhielm LR, Wretenberg P, Mikhail N, Bauer TW: Femoral bone regeneration subsequent to impaction grafting during hip revision: Histologic analysis of a human biopsy specimen. J Arthroplasty 14:849–853, 1999.
19. Mikhail WE, Wretenberg PF, Weidenhielm LR, Mikhail MN: Complex cemented revision using polished stem and morselized allograft: Minimum 5 years’ follow-up. Arch Orthop Trauma Surg 119:288–291, 1999.
20. Nelissen RG, Bauer TW, Weidenhielm LR, LeGolvan DP, Mikhail WE: Revision hip arthroplasty with the use of cement and impaction grafting: Histological analysis of four cases. J Bone Joint Surg 77A:412–422, 1995.
21. Pekkarinen J, Alho A, Lepisto J, et al: Impaction bone grafting in revision hip surgery: A high incidence of complications. J Bone Joint Surg 82B:103–107, 2000.
22. Schreurs BW, Buma P, Huiskes R, Slagter SL, Slooff TJ: Morsellized allografts for fixation of the hip prosthesis femoral component: A mechanical and histological study in the goat. Acta Orthop Scand 65:267–275, 1994.
23. Schreurs BW, Huiskes S, Slooff TJJH: The initial stability of cemented and noncemented femoral stems fixated with a bone grafting technique. Clin Mater 16:105, 1994.
24. Slooff TJ, Schimmel JW, Buma P: Cemented fixation with bone grafts. Orthop Clin North Am 24:667–677, 1993.
© 2003 Lippincott Williams & Wilkins, Inc.
25. Van Biezen FC, ten Have BL, Verhaar JA: Impaction bone-grafting of severely defective femora in revision total hip surgery: 21 hips followed for 41–85 months. Acta Orthop Scand 71:135–142, 2000.